diff --git a/README.md b/README.md index cc31743..f1a01c3 100644 --- a/README.md +++ b/README.md @@ -1,8 +1,8 @@ -

Agent-R1: Training Powerful LLM Agents with
End-to-End Reinforcement Learning

+

Agent-R1: Training Powerful LLM Agents with
End-to-End Reinforcement Learning

Paper Arxiv - Documentation + Documentation Ask DeepWiki.com GitHub Repo stars GitHub forks @@ -10,125 +10,251 @@

Agent-R1 Logo

-## News +**Agent-R1** is a unified, modular framework for **Agentic Reinforcement Learning**. It trains multi-step LLM agents through a step-native RL loop, where the model observes an environment, generates an action, receives tool or environment feedback, and continues until the task is solved or terminated. + +Unlike single-turn RL pipelines that treat interaction as one growing prompt-response sequence, Agent-R1 models every turn as a **step-level MDP transition**. This makes tool use, environment state, context management, reward assignment, and policy optimization explicit parts of the same training substrate. -- [2026.03.23] **Agent-R1 v0.1.0 marks the first official version of the project.** It introduces a fully refactored codebase, the **Step-level MDP** foundation, and new **Layered Abstractions**. The previous version has been archived to the `legacy` branch. +## News -- [2026.03.04] **We've launched [Claw-R1](https://agentr1.github.io/Claw-R1/)**, a more advanced framework designed to empower General Agents (OpenClaw etc.) with Agentic RL through a Middleware design. Check it out at [AgentR1/Claw-R1](https://github.com/AgentR1/Claw-R1). +- [2026.05.26] **Broader algorithm and benchmark support.** Agent-R1 now supports StepPO, RLOO, REINFORCE++ Baseline, and GiGPO in addition to GRPO, PPO, and REINFORCE++; beyond GSM8K-tool, it also includes HotpotQA, Paper Search, ALFWorld, and WebShop recipes. +- [2026.03.23] **Agent-R1 v0.1.0 is the first official release of the refactored architecture.** It introduces the **Step-level MDP** foundation and new **Layered Abstractions**. The previous implementation is archived on the `legacy` branch. +- [2026.03.04] **[Claw-R1](https://agentr1.github.io/Claw-R1/) is released.** It extends Agentic RL to general agents such as OpenClaw through a middleware-style design. See [AgentR1/Claw-R1](https://github.com/AgentR1/Claw-R1).
Earlier Updates +- [2026.01.10] **PaperScout** is released: an autonomous academic paper search agent trained with Agent-R1 and Proximal Sequence Policy Optimization. Read the paper [here](https://arxiv.org/abs/2601.10029). +- [2025.11.18] The Agent-R1 technical report is released on [arXiv](https://arxiv.org/abs/2511.14460). +- [2025.05.06] Tool environments are redesigned to support more flexible agent-tool interaction patterns. +- [2025.05.06] GRPO and REINFORCE++ training crashes caused by NaN values are fixed. See [issue #30](https://github.com/0russwest0/Agent-R1/issues/30). +- [2025.04.01] Basic inference scripts and an interactive chat interface are added. +- [2025.03.18] Multi-modal support is added for vision-language model agents. +- [2025.03.18] `verl` is moved to a git submodule and Agent-R1 extensions are separated from upstream code. +- [2025.03.16] Process rewards are supported for per-tool-call feedback. +
-- [2026.01.10] **New Application Released**: We are excited to introduce **PaperScout**, an autonomous agent for academic paper search trained using Agent-R1. It introduces a novel *Proximal Sequence Policy Optimization (PSPO)* method. Read the paper [here](https://arxiv.org/abs/2601.10029). +## Why Agent-R1 -- [2025.11.18] **Technical Report**: We have released the technical report on arXiv. Read the paper [here](https://arxiv.org/abs/2511.14460). +Modern LLM infrastructure already has strong serving systems such as vLLM and SGLang, and strong distributed training systems such as DeepSpeed, FSDP, and Megatron-LM. Agentic RL needs to reconnect these two sides into a **rollout -> reward -> replay -> update** loop where the model interacts with tools and environments over multiple turns. -- [2025.05.06] **Tool Environment Redesign**: Completely redesigned and abstracted tool environments to support more flexible and diverse agent-tool interactions patterns. +Agent-R1 is built around three design goals: -- [2025.05.06] **Critical Bug Fixes**: Fixed GRPO and Reinforce++ training crash issues that were causing NaN values during training. See [issue #30](https://github.com/0russwest0/Agent-R1/issues/30) for details. +- **Step-level trajectory representation**: each transition stores observation, action, environment feedback, reward, termination state, and next observation while preserving action boundaries and avoiding fragile `Token -> Text -> Token` reconstruction. +- **Flexible context management**: the environment decides what the model sees next, so history can be appended, truncated, summarized, rewritten, or augmented. +- **Algorithm-system decoupling**: task workflows, environments, rollout, rewards, advantage estimators, and policy objectives can evolve independently. -- [2025.05.06] **New Tutorials**: Added comprehensive tutorials for creating custom tools and tool environments, including the first open-source runnable implementation of ReTool. +

Agent-R1 Framework

+## Core Idea: Step-level MDP -- [2025.04.01] Added basic **inference scripts** and a simple interactive chat interface. You can now easily deploy and interact with your trained models. See [inference guide](docs/inference/inference.md) for details. +In multi-turn agent training, the model is not just continuing a token sequence. Each model output can invoke tools, change the environment state, receive external feedback, and shape the next observation. Agent-R1 therefore treats the **agent step** as the basic interaction unit: a step records what the model saw, what action it produced, what feedback and reward the environment returned, and what observation should be exposed next. This step-level trajectory representation keeps rollout, replay, context construction, and credit assignment aligned with real agent decisions, while still allowing token-level policy losses inside each generated action. -- [2025.03.18] Added comprehensive **multi-modal support**! Agent-R1 now seamlessly integrates with vision-language models (VLMs), enabling agents to process and reason with both text and visual inputs in rich multi-modal environments. +

Step-level MDP

-- [2025.03.18] Refactored our codebase to improve maintainability! We've converted verl from a static folder to a **git submodule** and separated our custom code extensions. This makes it easier to update `verl` and understand the project structure. +## Architecture - > **Important:** After pulling this update, you'll need to reinitialize your environment. Run `git submodule update --init --recursive` and reinstall verl locally from this directory. +Agent-R1 uses layered abstractions so new tasks can reuse the same trainer without rewriting the full RL stack. -- [2025.03.16] Added support for **process rewards**! You can now assign rewards for each tool call based on its effectiveness. To balance process rewards with outcome rewards, we implemented reward normalization inspired by [PRIME](https://github.com/PRIME-RL/PRIME). +| Layer | Responsibility | When to Use | +|---|---|---| +| `AgentFlowBase` | Full control over prompt construction, model calls, and step assembly. | Custom workflows or experimental agent logic. | +| `AgentEnvLoop` | The main multi-step loop connecting model generation with environment `reset()` / `step()`. | Most agentic RL tasks. | +| `AgentEnv` | Task environment interface returning observations, rewards, termination, and metadata. | When your task has state transitions. | +| `ToolEnv` | Built-in environment for parsing tool calls, executing tools, and feeding observations back. | Tool-augmented tasks such as GSM8K-tool. | +| `BaseTool` | Standard interface for registering executable tools. | Adding calculators, search tools, APIs, or task-specific checkers. | - +The main loop is: -## Overview +1. Load a sample containing `prompt`, `agent_name`, `reward_model`, and optional `env_kwargs`. +2. Create the configured `AgentFlow` and environment. +3. Generate an action from the current observation. +4. Parse the action, execute tools or update the environment, and return feedback. +5. Record the step and continue until `done=True` or `max_steps` is reached. +6. Convert the structured trace into rewards, advantages, masks, and policy updates. -**Agent-R1** is an open-source framework for training powerful language **agents** with **end-to-end reinforcement learning**. It is designed for **multi-step agent tasks**, where the model interacts with environments and tools across multiple rounds instead of producing a single final answer. +## Getting Started -The core idea behind Agent-R1 is **Step-level MDP**: each interaction step is treated as a proper RL transition, with an environment-defined state, an LLM action, and the next observation produced by the environment. This replaces the usual "append everything into one ever-growing token sequence" view with a more principled and more flexible training abstraction. +Agent-R1 follows the environment setup of [verl](https://verl.readthedocs.io/en/latest/start/install.html). Use a recent source installation of `verl` that includes AgentFlow, async rollout, reward-loop, and `verl.trainer.config` package-data APIs. -With Agent-R1, you can build custom agent workflows, define interactive environments and tools, and train multi-step agents in a unified RL pipeline. +Agent-R1 itself does not need a separate package install. Clone the repository, prepare a compatible `verl` environment, and run scripts from the repository root. -> **Also check out [Awesome-Agent-RL](https://github.com/0russwest0/Awesome-Agent-RL)**: Our curated collection of papers and resources on unlocking the potential of Agents through Reinforcement Learning. +### Stage 1: Sanity Check with GSM8K-tool -

Agent-R1 Framework

+Start with the smallest complete tool-use loop before moving to larger agent tasks: -## Why Agent-R1 v0.1.0 +```bash +python3 examples/data_preprocess/gsm8k_tool.py --local_save_dir ~/data/gsm8k_tool +bash examples/run_qwen3-4b_gsm8k_tool.sh +``` -Agent-R1 v0.1.0 is the first official release of the new architecture. It is built to address two common failure modes in RL training for LLM agents: +This checks the model path, dataset path, rollout engine, `AgentEnvLoop`, `ToolEnv`, tool execution, rewards, and trainer wiring. -- **Retokenization drift in text-based pipelines**: if rollout data is collected as text and later tokenized again for training, the `Token -> Text -> Token` conversion is not reversible. -- **Rigid token-only trajectory construction**: if the whole interaction is represented as a single growing token list, context handling becomes hard-wired to simple append-only logic. +You can append Hydra overrides directly: -Agent-R1 addresses these issues with a **step-level trajectory representation**: +```bash +bash examples/run_qwen3-4b_gsm8k_tool.sh \ + actor_rollout_ref.model.path=/path/to/Qwen3-4B \ + trainer.n_gpus_per_node=4 +``` -- each step stores its own prompt and response -- the environment, not raw token concatenation, controls the next observation -- context can be **truncated**, **summarized**, **rewritten**, or **augmented** between steps -- standard RL loops such as `obs -> action -> step -> next_obs` map naturally onto agent training +### Stage 2: Run Training + +```bash +bash examples/run_qwen3-4b_gsm8k_tool.sh +bash examples/run_hotpotqa_grpo.sh +bash examples/run_papersearch_grpo.sh +bash examples/run_alfworld_grpo.sh +bash examples/run_webshop_grpo.sh +``` -This makes Agent-R1 a better fit for real multi-step agent tasks with tool use, environment feedback, and flexible context management. +These scripts are a convenient way to verify that Agent-R1 can reuse the same training stack across different tasks, environments, tools, rewards, and data recipes. -## Version Guide +## Datasets and Scripts -- The default [`main`](https://github.com/AgentR1/Agent-R1/tree/main) branch contains the new **v0.1.0** architecture based on **Step-level MDP** and **Layered Abstractions**. -- The previous implementation is preserved in the [`legacy`](https://github.com/AgentR1/Agent-R1/tree/legacy) branch for reference. -- The current version uses the same runtime environment as `verl` and requires **`verl==0.7.0`**. +Agent-R1 is not tied to one benchmark. Current recipes cover math reasoning, multi-hop retrieval QA, academic paper search, text embodied environments, and web shopping. +| Dataset / Environment | Task Type | Training Script | Data Preparation | +|---|---|---|---| +| GSM8K Tool | Tool-augmented math reasoning | `examples/run_qwen3-4b_gsm8k_tool.sh` | `examples/data_preprocess/gsm8k_tool.py` | +| HotpotQA | Multi-hop retrieval QA | `examples/run_hotpotqa_grpo.sh` | `recipe/hotpotqa/prepare_hotpotqa_agent_r1.py` | +| Paper Search | Academic search agent | `examples/run_papersearch_grpo.sh` | `recipe/paper_search/prepare_paper_search_agent_r1.py` | +| ALFWorld | Text embodied household interaction | `examples/run_alfworld_grpo.sh` | `recipe/alfworld/prepare_alfworld_agent_r1.py` | +| WebShop | Simulated online shopping | `examples/run_webshop_grpo.sh` | `recipe/webshop/prepare_webshop_agent_r1.py` | +Data preparation examples: -## Getting Started +```bash +# GSM8K Tool +python3 examples/data_preprocess/gsm8k_tool.py --local_save_dir ~/data/gsm8k_tool -Agent-R1 uses the same environment setup as [verl](https://verl.readthedocs.io/en/latest/start/install.html), and the current version requires `verl==0.7.0`. You only need to clone this repository; there is no separate Agent-R1 installation step. +# HotpotQA + retrieval corpus +python3 recipe/hotpotqa/prepare_hotpotqa_agent_r1.py \ + --output_dir data/corpus/hotpotqa \ + --corpus_output_path data/corpus/hotpotqa_corpus/hpqa_corpus.jsonl + +# Paper Search from bundled AutoScholarQuery JSONL files +python3 recipe/paper_search/prepare_paper_search_agent_r1.py \ + --input_dir recipe/paper_search/inference/datasets/AutoScholarQuery \ + --output_dir data/pasa + +# ALFWorld from local ALFWorld raw data +python3 recipe/alfworld/prepare_alfworld_agent_r1.py \ + --input_dir alfworld_data/json_2.1.1 \ + --output_dir data/alfworld + +# WebShop small/full data +python3 recipe/webshop/prepare_webshop_agent_r1.py \ + --dataset_mode small \ + --input_dir webshop_data \ + --output_dir data/webshop +``` -The recommended path is: +Paper Search raw JSONL files are included under `recipe/paper_search/inference/datasets/`. Larger generated artifacts such as parquet files, retrieval indexes, environment caches, and copied game or product data should stay local under `data/`. -1. Read the [Getting Started](https://agentr1.github.io/Agent-R1/getting-started/) page for the minimal setup flow. -2. Use [`examples/data_preprocess/gsm8k.py`](examples/data_preprocess/gsm8k.py) and [`examples/run_qwen2.5-3b.sh`](examples/run_qwen2.5-3b.sh) as a sanity check that the environment is wired correctly. -3. Move to the [Agent Task Tutorial](https://agentr1.github.io/Agent-R1/tutorials/agent-task/) for the main Agent-R1 workflow based on multi-step interaction and tool use. +## Training Data Contract -### Stage 1: Sanity Check the Base Training Stack +Agent-R1 uses parquet files compatible with the `verl` trainer. Agent tasks normally include: -Prepare a minimal GSM8K dataset and run the single-step script: +| Field | Required | Meaning | +|---|---:|---| +| `data_source` | Yes | Dataset or benchmark name. | +| `prompt` | Yes | Chat messages passed to the tokenizer and rollout engine. | +| `ability` | Recommended | Task category used for logging and reward routing. | +| `reward_model` | Yes | Rule/model reward metadata, usually including `ground_truth`. | +| `extra_info` | Recommended | Split, index, raw question, raw answer, or task-specific metadata. | +| `agent_name` | Agent tasks | Agent flow name, usually `agent_env_loop`. | +| `env_kwargs` | Tool/env tasks | JSON config consumed by `AgentEnvLoop._create_env`. | -```bash -python3 examples/data_preprocess/gsm8k.py --local_save_dir ~/data/gsm8k -bash examples/run_qwen2.5-3b.sh +Minimal `env_kwargs` for GSM8K-tool: + +```json +{ + "env_type": "tool", + "tools": ["calc_gsm8k_reward"], + "tool_format": "hermes", + "tools_kwargs": {"ground_truth": ""} +} ``` -This stage is only a **setup check**. It helps confirm that your environment, model path, dataset path, and training stack are wired correctly. +## Supported Algorithms -### Stage 2: Run the Main Agent-R1 Workflow +The trainer routes `algorithm.adv_estimator` to multiple estimators: -Prepare the tool-augmented dataset and launch the multi-step agent training script: +| Method | Configuration | Granularity | Critic Required | +|---|---|---|---:| +| StepPO | `adv_estimator=gae` + `loss_mode=gspo` | Step-level advantage + sequence-level policy loss | Yes | +| GRPO | `algorithm.adv_estimator=grpo` | Trajectory / group-relative | No | +| PPO / GAE | `algorithm.adv_estimator=gae` | Step-level actor-critic | Yes | +| RLOO | `algorithm.adv_estimator=rloo` | Trajectory outcome | No | +| REINFORCE++ | `algorithm.adv_estimator=reinforce_plus_plus` | Token return | No | +| REINFORCE++ Baseline | `algorithm.adv_estimator=reinforce_plus_plus_baseline` | Prompt / trajectory baseline | No | +| GiGPO | `algorithm.adv_estimator=gigpo` | Trajectory + step group | No | -```bash -python3 examples/data_preprocess/gsm8k_tool.py --local_save_dir ~/data/gsm8k_tool -bash examples/run_qwen3-4b_gsm8k_tool.sh +`actor_rollout_ref.actor.policy_loss.loss_mode` controls the policy objective separately from advantage estimation. This separation makes it easier to compare credit-assignment strategies under the same environment and rollout setup. + +## Experimental Snapshot + +The Agent-R1 report evaluates Qwen3-4B across representative agent scenarios. The table below summarizes the main results; see [Experiments](docs/experiments.md) for the experimental setting, task coverage, optimizer comparison, and context-management analysis. + +| Method | GSM8K Acc. (%) | HotpotQA Acc. (%) | ALFWorld SR Seen (%) | ALFWorld SR Unseen (%) | WebShop Score (%) | WebShop SR (%) | +|---|---:|---:|---:|---:|---:|---:| +| ReAct | 53.1 | 25.8 | 7.14 | 2.98 | 51.58 | 23.8 | +| GRPO | **83.3** | **59.4** | **81.29** | **74.58** | 65.83 | 44.2 | +| PPO | 78.1 | 56.7 | 76.42 | 72.38 | **70.18** | **46.0** | +| REINFORCE++ | 78.9 | 52.8 | 73.84 | 69.57 | 63.41 | 41.8 | +| RLOO | 81.6 | 55.2 | 79.08 | 73.46 | 68.02 | 45.1 | + +## Building a New Agent Task + +For a new task, keep the trainer intact and implement the task-specific layers: + +```text +recipe// + base.yaml + prepare__agent_r1.py + _agent_flow.py + reward_fn.py + prompts.py + utils.py + env/ # optional environment service or wrappers ``` -This is the main Agent-R1 path, where `AgentEnvLoop` drives multi-step rollout and `ToolEnv` handles tool calls and environment feedback. +Typical migration checklist: -Core concepts: +- **Data**: emit parquet rows with `prompt`, `reward_model`, `agent_name`, and `env_kwargs`. +- **Environment / tools**: define how state updates, tool observations, rewards, and termination work. +- **Agent flow**: connect model actions to the environment loop and expose step records. +- **Training script**: set paths, rollout steps, batch sizes, estimator, and policy loss through Hydra overrides. -- [Step-level MDP](https://agentr1.github.io/Agent-R1/core-concepts/step-level-mdp/) -- [Layered Abstractions](https://agentr1.github.io/Agent-R1/core-concepts/layered-abstractions/) +See: -## Awesome Projects Using Agent-R1 +- [Step-level MDP](docs/core-concepts/step-level-mdp.md) +- [Layered Abstractions](docs/core-concepts/layered-abstractions.md) +- [Agent Task Tutorial](docs/tutorials/agent-task.md) +- [Datasets and Algorithms](docs/tutorials/datasets-and-algorithms.md) + +## Documentation -Here are some representative projects built on top of Agent-R1: +- Project homepage: [https://agentr1.github.io/agent-r1](https://agentr1.github.io/agent-r1) +- Documentation: [https://agentr1.github.io/agent-r1/docs/](https://agentr1.github.io/agent-r1/docs/) + +## Version Guide + +- `main` contains the current v0.1.0 architecture based on Step-level MDP and layered abstractions. +- `legacy` preserves the previous implementation for reference. +- Use a recent source checkout of `verl` that includes the AgentFlow / async rollout stack required by this repository. + +## Awesome Projects Using Agent-R1 -- **[TableMind](https://arxiv.org/abs/2509.06278)**: An autonomous programmatic agent for tool-augmented table reasoning. TableMind is built upon the Agent-R1 framework and leverages its end-to-end reinforcement learning pipeline to train a specialized agent for structured table understanding. -- **[PaperScout](https://arxiv.org/abs/2601.10029)**: An autonomous agent for academic paper search built with Agent-R1. It introduces Proximal Sequence Policy Optimization (PSPO), a process-aware method for aligning token-level optimization with sequence-level agent interactions. -- **[Cast-R1](https://arxiv.org/abs/2602.13802)**: A learned agentic framework that reformulates time series forecasting as a sequential decision-making problem. Built upon Agent-R1, it features a memory-based state management mechanism and a tool-augmented workflow, trained via a two-stage strategy combining supervised fine-tuning with multi-turn reinforcement learning to autonomously gather evidence, reason, and iteratively refine forecasts. +- **[TableMind](https://arxiv.org/abs/2509.06278)**: an autonomous programmatic agent for tool-augmented table reasoning. +- **[PaperScout](https://arxiv.org/abs/2601.10029)**: an autonomous academic paper search agent trained with Agent-R1 and Proximal Sequence Policy Optimization. +- **[Cast-R1](https://arxiv.org/abs/2602.13802)**: an agentic framework that reformulates time-series forecasting as sequential decision making. +- **[StepPO](https://arxiv.org/pdf/2604.18401)**: Step-Aligned Policy Optimization for Agentic Reinforcement Learning, a step-level Agentic RL method that treats the agent step as the action unit and aligns credit assignment with multi-turn agent decisions. - ## Acknowledgements -This work is conducted at the **State Key Laboratory of Cognitive Intelligence, USTC**. We gratefully acknowledge the inspiring ideas and early insights from [DeepSeek-R1](https://github.com/deepseek-ai/DeepSeek-R1), [veRL](https://github.com/volcengine/verl), and [RAGEN](https://github.com/ZihanWang314/ragen), which have significantly influenced the development of Agent-R1. We also sincerely thank [**Prof. Qi Liu**](http://staff.ustc.edu.cn/~qiliuql/) and [**Prof. Mingyue Cheng**](https://mingyue-cheng.github.io/) for their guidance and support. +This work is conducted at the **State Key Laboratory of Cognitive Intelligence, USTC**. We gratefully acknowledge the ideas and infrastructure from [DeepSeek-R1](https://github.com/deepseek-ai/DeepSeek-R1), [veRL](https://github.com/volcengine/verl), and [RAGEN](https://github.com/ZihanWang314/ragen). We also thank [Prof. Qi Liu](http://staff.ustc.edu.cn/~qiliuql/) and [Prof. Mingyue Cheng](https://mingyue-cheng.github.io/) for their guidance and support. ## Citation diff --git a/agent_r1/trainer/main_agent_ppo.py b/agent_r1/trainer/main_agent_ppo.py index 6cb4a56..6cc9f58 100644 --- a/agent_r1/trainer/main_agent_ppo.py +++ b/agent_r1/trainer/main_agent_ppo.py @@ -22,10 +22,10 @@ import ray from omegaconf import OmegaConf -from agent_r1.trainer.ppo.ray_trainer import RayAgentTrainer +from agent_r1.trainer.ppo.ray_trainer import RayAgentTrainer, need_critic_agent_ppo from verl.trainer.constants_ppo import get_ppo_ray_runtime_env from verl.trainer.ppo.reward import load_reward_manager -from verl.trainer.ppo.utils import need_critic, need_reference_policy +from verl.trainer.ppo.utils import need_reference_policy from verl.utils.config import validate_config from verl.utils.device import auto_set_device, is_cuda_available @@ -286,7 +286,7 @@ def run(self, config): validate_config( config=config, use_reference_policy=need_reference_policy(self.role_worker_mapping), - use_critic=need_critic(config), + use_critic=need_critic_agent_ppo(config), ) # Download the checkpoint from HDFS to the local machine. diff --git a/agent_r1/trainer/ppo/core_algos.py b/agent_r1/trainer/ppo/core_algos.py index 8f74861..36c8b8f 100644 --- a/agent_r1/trainer/ppo/core_algos.py +++ b/agent_r1/trainer/ppo/core_algos.py @@ -19,6 +19,7 @@ """ from collections import defaultdict +from difflib import SequenceMatcher from typing import Any, Optional import numpy as np @@ -27,6 +28,27 @@ import verl.utils.torch_functional as verl_F +def _to_hashable(value): + """Convert common observation objects to hashable keys for GiGPO grouping.""" + if isinstance(value, (int, float, str, bool)): + return value + if isinstance(value, (np.integer, np.floating)): + return value.item() + if isinstance(value, np.ndarray): + return tuple(value.flatten()) + if isinstance(value, (list, tuple)): + return tuple(_to_hashable(item) for item in value) + if isinstance(value, dict): + return tuple(sorted((key, _to_hashable(val)) for key, val in value.items())) + raise TypeError(f"Unsupported observation type for GiGPO grouping: {type(value)}") + + +def _are_similar(a: str, b: str, threshold: float) -> bool: + if not isinstance(a, str) or not isinstance(b, str): + raise ValueError("Similarity-based GiGPO only supports text observations.") + return SequenceMatcher(None, a, b).ratio() >= threshold + + def compute_gae_advantage_return( token_level_rewards: torch.Tensor, values: torch.Tensor, @@ -318,6 +340,302 @@ def compute_grpo_outcome_advantage( return scores, scores +def compute_reinforce_plus_plus_outcome_advantage( + token_level_rewards: torch.Tensor, + response_mask: torch.Tensor, + gamma: float, +) -> tuple[torch.Tensor, torch.Tensor]: + """Compute REINFORCE++ token-level discounted returns and whitened advantages.""" + with torch.no_grad(): + returns = torch.zeros_like(token_level_rewards) + running_return = 0 + + for t in reversed(range(token_level_rewards.shape[1])): + running_return = token_level_rewards[:, t] + gamma * running_return + returns[:, t] = running_return + running_return = running_return * response_mask[:, t] + + advantages = verl_F.masked_whiten(returns, response_mask) + advantages = advantages * response_mask + + return advantages, returns + + +def _trajectory_total_scores( + token_level_rewards: torch.Tensor, + response_mask: torch.Tensor, + index: np.ndarray, + trajectory_uids: np.ndarray, +) -> tuple[torch.Tensor, dict[object, torch.Tensor], dict[object, object]]: + step_scores = (token_level_rewards * response_mask).sum(dim=-1) + traj2total_score: dict[object, torch.Tensor] = {} + traj2index: dict[object, object] = {} + + for i in range(step_scores.shape[0]): + traj_uid = trajectory_uids[i] + if traj_uid in traj2total_score: + traj2total_score[traj_uid] = traj2total_score[traj_uid] + step_scores[i] + else: + traj2total_score[traj_uid] = step_scores[i] + traj2index[traj_uid] = index[i] + + return step_scores, traj2total_score, traj2index + + +def compute_reinforce_plus_plus_baseline_outcome_advantage( + token_level_rewards: torch.Tensor, + response_mask: torch.Tensor, + index: np.ndarray, + trajectory_uids: np.ndarray, +) -> tuple[torch.Tensor, torch.Tensor]: + """Compute RF++-baseline advantages using a per-prompt trajectory mean baseline.""" + with torch.no_grad(): + step_scores, traj2total_score, traj2index = _trajectory_total_scores( + token_level_rewards=token_level_rewards, + response_mask=response_mask, + index=index, + trajectory_uids=trajectory_uids, + ) + + id2score = defaultdict(list) + for traj_uid, total_score in traj2total_score.items(): + id2score[traj2index[traj_uid]].append(total_score) + + id2mean: dict[object, torch.Tensor] = {} + for idx, scores in id2score.items(): + id2mean[idx] = torch.mean(torch.stack(scores)) if len(scores) > 1 else step_scores.new_tensor(0.0) + + traj2adv = { + traj_uid: total_score - id2mean[traj2index[traj_uid]] + for traj_uid, total_score in traj2total_score.items() + } + + scores = step_scores.clone() + for i in range(step_scores.shape[0]): + scores[i] = traj2adv[trajectory_uids[i]] + + scores = scores.unsqueeze(-1).tile([1, response_mask.shape[-1]]) * response_mask + scores = verl_F.masked_whiten(scores, response_mask) * response_mask + + return scores, scores + + +def compute_rloo_outcome_advantage( + token_level_rewards: torch.Tensor, + response_mask: torch.Tensor, + index: np.ndarray, + trajectory_uids: np.ndarray, +) -> tuple[torch.Tensor, torch.Tensor]: + """Compute RLOO advantages over trajectory outcomes.""" + with torch.no_grad(): + step_scores, traj2total_score, traj2index = _trajectory_total_scores( + token_level_rewards=token_level_rewards, + response_mask=response_mask, + index=index, + trajectory_uids=trajectory_uids, + ) + + id2score = defaultdict(list) + for traj_uid, total_score in traj2total_score.items(): + id2score[traj2index[traj_uid]].append(total_score) + + id2mean: dict[object, torch.Tensor] = {} + for idx, scores in id2score.items(): + id2mean[idx] = torch.mean(torch.stack(scores)) if len(scores) > 1 else step_scores.new_tensor(0.0) + + traj2adv: dict[object, torch.Tensor] = {} + for traj_uid, total_score in traj2total_score.items(): + prompt_idx = traj2index[traj_uid] + response_num = len(id2score[prompt_idx]) + if response_num > 1: + traj2adv[traj_uid] = ( + total_score * response_num / (response_num - 1) + - id2mean[prompt_idx] * response_num / (response_num - 1) + ) + else: + traj2adv[traj_uid] = total_score + + scores = step_scores.clone() + for i in range(step_scores.shape[0]): + scores[i] = traj2adv[trajectory_uids[i]] + + scores = scores.unsqueeze(-1) * response_mask + + return scores, scores + + +def compute_step_discounted_returns( + token_level_rewards: torch.Tensor, + response_mask: torch.Tensor, + trajectory_uids: np.ndarray, + step_indices: np.ndarray, + gamma: float, +) -> torch.Tensor: + """Compute per-step discounted returns from Agent-R1 multi-step trajectory rows.""" + device = token_level_rewards.device + step_rewards = (token_level_rewards * response_mask).sum(dim=-1) + + with torch.no_grad(): + unique_traj_np, traj_inv_np = np.unique(trajectory_uids, return_inverse=True) + num_traj = len(unique_traj_np) + traj_inv = torch.as_tensor(traj_inv_np, dtype=torch.long, device=device) + step_ids = torch.as_tensor(step_indices, dtype=torch.long, device=device) + max_step = int(step_ids.max().item()) + 1 if step_rewards.numel() > 0 else 0 + + rewards_map = torch.zeros((num_traj, max_step), dtype=step_rewards.dtype, device=device) + rewards_map[traj_inv, step_ids] = step_rewards + + returns_map = torch.zeros_like(rewards_map) + running_return = torch.zeros((num_traj,), dtype=step_rewards.dtype, device=device) + for t in reversed(range(max_step)): + running_return = rewards_map[:, t] + gamma * running_return + returns_map[:, t] = running_return + + return returns_map[traj_inv, step_ids] + + +def _build_step_groups( + anchor_obs: np.ndarray, + index: np.ndarray, + enable_similarity: bool, + similarity_thresh: float, +) -> np.ndarray: + if enable_similarity and not 0.0 < similarity_thresh < 1.0: + raise ValueError("GiGPO similarity_thresh must be in (0, 1) when similarity grouping is enabled.") + + step_group_uids = np.empty(len(anchor_obs), dtype=object) + for prompt_idx in np.unique(index): + locs = np.where(index == prompt_idx)[0] + + if not enable_similarity: + clusters = defaultdict(list) + for loc in locs: + clusters[_to_hashable(anchor_obs[loc])].append(loc) + for cluster_id, cluster_locs in enumerate(clusters.values()): + group_uid = (prompt_idx, cluster_id) + for loc in cluster_locs: + step_group_uids[loc] = group_uid + continue + + clusters: list[dict[str, object]] = [] + for loc in locs: + obs = anchor_obs[loc] + placed = False + for cluster in clusters: + if _are_similar(obs, cluster["rep"], similarity_thresh): + cluster["locs"].append(loc) + placed = True + break + if not placed: + clusters.append({"rep": obs, "locs": [loc]}) + + for cluster_id, cluster in enumerate(clusters): + group_uid = (prompt_idx, cluster_id) + for loc in cluster["locs"]: + step_group_uids[loc] = group_uid + + if np.any(step_group_uids == None): # noqa: E711 + missing = np.where(step_group_uids == None)[0] # noqa: E711 + raise ValueError(f"Failed to assign GiGPO step groups for rows: {missing}") + return step_group_uids + + +def _normalize_group_scores( + scores: torch.Tensor, + group_uids: np.ndarray, + epsilon: float, + remove_std: bool, + single_mean_zero: bool = False, +) -> torch.Tensor: + id2score = defaultdict(list) + id2mean: dict[object, torch.Tensor] = {} + id2std: dict[object, torch.Tensor] = {} + + for i in range(scores.shape[0]): + id2score[group_uids[i]].append(scores[i]) + + for group_uid, group_scores in id2score.items(): + stacked = torch.stack(group_scores) + if single_mean_zero and len(group_scores) == 1: + id2mean[group_uid] = scores.new_tensor(0.0) + else: + id2mean[group_uid] = torch.mean(stacked) + id2std[group_uid] = torch.std(stacked) if len(group_scores) > 1 else scores.new_tensor(1.0) + + normalized = scores.clone() + for i in range(scores.shape[0]): + group_uid = group_uids[i] + if remove_std: + normalized[i] = scores[i] - id2mean[group_uid] + else: + normalized[i] = (scores[i] - id2mean[group_uid]) / (id2std[group_uid] + epsilon) + return normalized + + +def compute_gigpo_outcome_advantage( + token_level_rewards: torch.Tensor, + step_rewards: torch.Tensor, + response_mask: torch.Tensor, + anchor_obs: np.ndarray, + index: np.ndarray, + trajectory_uids: np.ndarray, + epsilon: float = 1e-6, + step_advantage_w: float = 1.0, + mode: str = "mean_std_norm", + enable_similarity: bool = False, + similarity_thresh: float = 0.95, +) -> tuple[torch.Tensor, torch.Tensor]: + """Compute GiGPO advantages for Agent-R1 agent trajectories.""" + if mode == "mean_std_norm": + remove_std = False + elif mode == "mean_norm": + remove_std = True + else: + raise ValueError(f"Unknown GiGPO mode: {mode}") + + with torch.no_grad(): + step_scores, traj2total_score, traj2index = _trajectory_total_scores( + token_level_rewards=token_level_rewards, + response_mask=response_mask, + index=index, + trajectory_uids=trajectory_uids, + ) + + traj_uids = np.array(list(traj2total_score.keys()), dtype=object) + traj_scores = torch.stack([traj2total_score[traj_uid] for traj_uid in traj_uids]) + traj_groups = np.array([traj2index[traj_uid] for traj_uid in traj_uids], dtype=object) + traj_adv = _normalize_group_scores( + scores=traj_scores, + group_uids=traj_groups, + epsilon=epsilon, + remove_std=remove_std, + single_mean_zero=True, + ) + traj2adv = {traj_uid: traj_adv[i] for i, traj_uid in enumerate(traj_uids)} + + episode_advantages = step_scores.clone() + for i in range(step_scores.shape[0]): + episode_advantages[i] = traj2adv[trajectory_uids[i]] + + step_group_uids = _build_step_groups( + anchor_obs=anchor_obs, + index=index, + enable_similarity=enable_similarity, + similarity_thresh=similarity_thresh, + ) + step_advantages = _normalize_group_scores( + scores=step_rewards, + group_uids=step_group_uids, + epsilon=epsilon, + remove_std=remove_std, + ) + + advantages = episode_advantages + step_advantage_w * step_advantages + advantages = advantages.unsqueeze(-1) * response_mask + + return advantages, advantages + + def agg_loss( loss_mat: torch.Tensor, loss_mask: torch.Tensor, @@ -481,6 +799,54 @@ def compute_policy_loss_reinforce( return pg_loss, pg_metrics +def compute_policy_loss_gspo( + old_log_prob: torch.Tensor, + log_prob: torch.Tensor, + advantages: torch.Tensor, + response_mask: torch.Tensor, + loss_agg_mode: str = "seq-mean-token-mean", + config: Optional[Any] = None, + rollout_is_weights: torch.Tensor | None = None, +) -> tuple[torch.Tensor, dict[str, Any]]: + """Compute GSPO policy loss with sequence-level importance ratios.""" + assert config is not None + + clip_ratio = config.clip_ratio + clip_ratio_low = config.clip_ratio_low if config.clip_ratio_low is not None else clip_ratio + clip_ratio_high = config.clip_ratio_high if config.clip_ratio_high is not None else clip_ratio + + negative_approx_kl = log_prob - old_log_prob + seq_lengths = torch.sum(response_mask, dim=-1).clamp(min=1) + negative_approx_kl_seq = torch.sum(negative_approx_kl * response_mask, dim=-1) / seq_lengths + + log_seq_importance_ratio = log_prob - log_prob.detach() + negative_approx_kl_seq.detach().unsqueeze(-1) + log_seq_importance_ratio = torch.clamp(log_seq_importance_ratio, max=10.0) + seq_importance_ratio = torch.exp(log_seq_importance_ratio) + + pg_losses1 = -advantages * seq_importance_ratio + pg_losses2 = -advantages * torch.clamp(seq_importance_ratio, 1 - clip_ratio_low, 1 + clip_ratio_high) + pg_losses = torch.maximum(pg_losses1, pg_losses2) + + if rollout_is_weights is not None: + pg_losses = pg_losses * rollout_is_weights + + pg_loss = agg_loss( + loss_mat=pg_losses, + loss_mask=response_mask, + loss_agg_mode=loss_agg_mode, + **getattr(config, "global_batch_info", {}), + ) + + pg_clipfrac = verl_F.masked_mean(torch.gt(pg_losses2, pg_losses1).float(), response_mask) + ppo_kl = verl_F.masked_mean(-negative_approx_kl, response_mask) + pg_metrics = { + "actor/pg_clipfrac": pg_clipfrac.detach().item(), + "actor/ppo_kl": ppo_kl.detach().item(), + "actor/pg_clipfrac_lower": torch.tensor(0.0, device=pg_loss.device).detach().item(), + } + return pg_loss, pg_metrics + + def compute_policy_loss_bypass_mode( old_log_prob: torch.Tensor, log_prob: torch.Tensor, @@ -558,6 +924,7 @@ def compute_policy_loss_bypass_mode( def get_policy_loss_fn(name: str): local_policy_loss_fns = { "vanilla": compute_policy_loss_vanilla, + "gspo": compute_policy_loss_gspo, "reinforce": compute_policy_loss_reinforce, "bypass_mode": compute_policy_loss_bypass_mode, } diff --git a/agent_r1/trainer/ppo/ray_trainer.py b/agent_r1/trainer/ppo/ray_trainer.py index f1dc480..b415d8d 100644 --- a/agent_r1/trainer/ppo/ray_trainer.py +++ b/agent_r1/trainer/ppo/ray_trainer.py @@ -82,6 +82,32 @@ def get_valid_data(data: DataProto) -> tuple[DataProto, torch.Tensor]: return valid_data, valid_mask +def _agent_adv_estimator_key(adv_estimator: AdvantageEstimator | str) -> str: + """Normalize Hydra / enum values to the algorithm.adv_estimator string.""" + if isinstance(adv_estimator, AdvantageEstimator): + return adv_estimator.value + return str(adv_estimator) + + +def need_critic_agent_ppo(config) -> bool: + """Whether RayAgentTrainer must load a critic/value net.""" + from verl.trainer.ppo.utils import need_critic as verl_need_critic + + adv_key = _agent_adv_estimator_key(config.algorithm.adv_estimator) + if adv_key in ("gae", "token_gae"): + return True + return verl_need_critic(config) + + +def _critic_vf_loss_response_mask(response_mask: torch.Tensor, adv_key: str) -> torch.Tensor: + """Return the mask used for critic value loss under each advantage estimator.""" + if adv_key == "token_gae": + return response_mask.clone() + value_mask = torch.zeros_like(response_mask) + value_mask[:, 0] = 1 + return value_mask + + def assign_global_mini_batch_ids(batch: DataProto, mini_batch_size: int, dp_size: int) -> None: """Assign global PPO mini-batch ids while preserving the existing DP dispatch layout.""" if dp_size <= 0: @@ -199,11 +225,15 @@ def build_trajectory_dump_entries( def compute_advantage( data: DataProto, - adv_estimator: AdvantageEstimator, + adv_estimator: AdvantageEstimator | str, gamma: float = 1.0, lam: float = 1.0, num_repeat: int = 1, norm_adv_by_std_in_grpo: bool = True, + gigpo_step_advantage_w: float = 1.0, + gigpo_mode: str = "mean_std_norm", + gigpo_enable_similarity: bool = False, + gigpo_similarity_thresh: float = 0.95, config: Optional[AlgoConfig] = None, ) -> DataProto: # TODO: 重写所有 core_algos 中的 advantage 函数，适配新型的 agent flow 数据结构 @@ -235,8 +265,9 @@ def compute_advantage( valid_data, valid_mask = get_valid_data(data) - # prepare response group - if adv_estimator == AdvantageEstimator.GAE: + adv_key = _agent_adv_estimator_key(adv_estimator) + + if adv_key == "gae": # Compute advantages and returns using Generalized Advantage Estimation (GAE) from agent_r1.trainer.ppo.core_algos import compute_gae_advantage_return @@ -251,7 +282,21 @@ def compute_advantage( ) advantages[valid_mask] = valid_advantages returns[valid_mask] = valid_returns - elif adv_estimator == AdvantageEstimator.GRPO: + elif adv_key == "token_gae": + from agent_r1.trainer.ppo.core_algos import compute_token_gae_advantage_return + + valid_advantages, valid_returns = compute_token_gae_advantage_return( + token_level_rewards=valid_data.batch["token_level_rewards"], + values=valid_data.batch["values"], + response_mask=valid_data.batch["response_mask"], + trajectory_uids=valid_data.non_tensor_batch["trajectory_uids"], + step_indices=valid_data.non_tensor_batch["step_indices"], + gamma=gamma, + lam=lam, + ) + advantages[valid_mask] = valid_advantages + returns[valid_mask] = valid_returns + elif adv_key == "grpo": # Call compute_grpo_outcome_advantage with parameters matching its definition from agent_r1.trainer.ppo.core_algos import compute_grpo_outcome_advantage @@ -264,6 +309,73 @@ def compute_advantage( ) advantages[valid_mask] = valid_advantages returns[valid_mask] = valid_returns + elif adv_key == "reinforce_plus_plus": + from agent_r1.trainer.ppo.core_algos import compute_reinforce_plus_plus_outcome_advantage + + valid_advantages, valid_returns = compute_reinforce_plus_plus_outcome_advantage( + token_level_rewards=valid_data.batch["token_level_rewards"], + response_mask=valid_data.batch["response_mask"], + gamma=gamma, + ) + advantages[valid_mask] = valid_advantages + returns[valid_mask] = valid_returns + elif adv_key == "reinforce_plus_plus_baseline": + from agent_r1.trainer.ppo.core_algos import compute_reinforce_plus_plus_baseline_outcome_advantage + + valid_advantages, valid_returns = compute_reinforce_plus_plus_baseline_outcome_advantage( + token_level_rewards=valid_data.batch["token_level_rewards"], + response_mask=valid_data.batch["response_mask"], + index=valid_data.non_tensor_batch["uid"], + trajectory_uids=valid_data.non_tensor_batch["trajectory_uids"], + ) + advantages[valid_mask] = valid_advantages + returns[valid_mask] = valid_returns + elif adv_key == "rloo": + from agent_r1.trainer.ppo.core_algos import compute_rloo_outcome_advantage + + valid_advantages, valid_returns = compute_rloo_outcome_advantage( + token_level_rewards=valid_data.batch["token_level_rewards"], + response_mask=valid_data.batch["response_mask"], + index=valid_data.non_tensor_batch["uid"], + trajectory_uids=valid_data.non_tensor_batch["trajectory_uids"], + ) + advantages[valid_mask] = valid_advantages + returns[valid_mask] = valid_returns + elif adv_key == "gigpo": + from agent_r1.trainer.ppo.core_algos import compute_gigpo_outcome_advantage, compute_step_discounted_returns + + if "anchor_obs" not in valid_data.non_tensor_batch: + raise KeyError( + "algorithm.adv_estimator='gigpo' requires non_tensor_batch['anchor_obs']. " + "Set step.extra_fields['anchor_obs'] in the agent flow before using GiGPO." + ) + step_rewards = compute_step_discounted_returns( + token_level_rewards=valid_data.batch["token_level_rewards"], + response_mask=valid_data.batch["response_mask"], + trajectory_uids=valid_data.non_tensor_batch["trajectory_uids"], + step_indices=valid_data.non_tensor_batch["step_indices"], + gamma=gamma, + ) + valid_advantages, valid_returns = compute_gigpo_outcome_advantage( + token_level_rewards=valid_data.batch["token_level_rewards"], + step_rewards=step_rewards, + response_mask=valid_data.batch["response_mask"], + anchor_obs=valid_data.non_tensor_batch["anchor_obs"], + index=valid_data.non_tensor_batch["uid"], + trajectory_uids=valid_data.non_tensor_batch["trajectory_uids"], + step_advantage_w=gigpo_step_advantage_w, + mode=gigpo_mode, + enable_similarity=gigpo_enable_similarity, + similarity_thresh=gigpo_similarity_thresh, + ) + advantages[valid_mask] = valid_advantages + returns[valid_mask] = valid_returns + else: + raise ValueError( + f"Unsupported algorithm.adv_estimator={adv_estimator!r}. Supported: " + "'gae', 'token_gae', 'grpo', 'reinforce_plus_plus', " + "'reinforce_plus_plus_baseline', 'rloo', 'gigpo'." + ) data.batch["advantages"] = advantages data.batch["returns"] = returns @@ -282,6 +394,16 @@ def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) self.use_reward_loop = True + adv_key = _agent_adv_estimator_key(self.config.algorithm.adv_estimator) + if adv_key in ("gae", "token_gae"): + if self.config.critic.enable is False: + raise ValueError( + f"algorithm.adv_estimator={adv_key!r} requires a value network, but critic.enable=False." + ) + if Role.Critic not in self.role_worker_mapping: + raise ValueError(f"algorithm.adv_estimator={adv_key!r} requires Role.Critic in role_worker_mapping.") + self.use_critic = True + def _update_actor(self, batch: DataProto) -> DataProto: rollout_config = self.config.actor_rollout_ref.rollout batch.meta_info["multi_turn"] = rollout_config.multi_turn.enable @@ -1005,6 +1127,16 @@ def fit(self): lam=self.config.algorithm.lam, num_repeat=self.config.actor_rollout_ref.rollout.n, norm_adv_by_std_in_grpo=norm_adv_by_std_in_grpo, + gigpo_step_advantage_w=self.config.algorithm.get("gigpo", {}).get( + "step_advantage_w", 1.0 + ), + gigpo_mode=self.config.algorithm.get("gigpo", {}).get("mode", "mean_std_norm"), + gigpo_enable_similarity=self.config.algorithm.get("gigpo", {}).get( + "enable_similarity", False + ), + gigpo_similarity_thresh=self.config.algorithm.get("gigpo", {}).get( + "similarity_thresh", 0.95 + ), config=self.config.algorithm, ) @@ -1013,11 +1145,8 @@ def fit(self): with marked_timer("update_critic", timing_raw, color="pink"): # Temporarily replace response_mask for critic response_mask = batch.batch["response_mask"] - # For "sequence = action", the critic value used by GAE is at action start. - # In `dp_critic.py`, returned `values` are sliced as `values[:, -resp_len-1:-1]`, - # so index 0 corresponds to the prompt-last position (before response[0]). - value_mask = torch.zeros_like(response_mask) - value_mask[:, 0] = 1 + adv_key = _agent_adv_estimator_key(self.config.algorithm.adv_estimator) + value_mask = _critic_vf_loss_response_mask(response_mask, adv_key) sample_mask = batch.batch.get("sample_mask", None) if sample_mask is not None: value_mask[~sample_mask.to(dtype=torch.bool)] = 0 diff --git a/data/README.md b/data/README.md new file mode 100644 index 0000000..87f15b1 --- /dev/null +++ b/data/README.md @@ -0,0 +1,23 @@ +# Data + +This directory is reserved for generated or downloaded local datasets. Large generated artifacts such as parquet files, retrieval indexes, model outputs, logs, and environment caches should stay local and are not meant to be committed. + +Agent-R1 includes lightweight raw/example data where it is practical: + +| Dataset | Included in Repository | Generated Output | +|---|---|---| +| GSM8K / GSM8K Tool | No raw copy; downloaded by `datasets` in `examples/data_preprocess/`. | `~/data/gsm8k*` by default. | +| Paper Search | Yes. JSONL files are included under `recipe/paper_search/inference/datasets/`. | `data/pasa/{train,test}.parquet`. | +| HotpotQA | No raw copy; downloaded from HuggingFace by the preparation script. | `data/corpus/hotpotqa/*.parquet` and optional retrieval corpus/index. | +| ALFWorld | No raw copy; prepare from local ALFWorld raw data. | `data/alfworld/*.parquet` plus copied game files. | +| WebShop | No raw copy; prepare from local WebShop product data or prebuilt goals. | `data/webshop*/{train,test}.parquet` plus environment artifacts. | + +Paper Search conversion example: + +```bash +python3 recipe/paper_search/prepare_paper_search_agent_r1.py \ + --input_dir recipe/paper_search/inference/datasets/AutoScholarQuery \ + --output_dir data/pasa +``` + +See `docs/tutorials/datasets-and-algorithms.md` for the full matrix. diff --git a/docs/experiments.md b/docs/experiments.md new file mode 100644 index 0000000..b78dc77 --- /dev/null +++ b/docs/experiments.md @@ -0,0 +1,58 @@ +# Experiments + +This page summarizes the experimental analysis for Agent-R1. The experiments ask two questions: + +1. Whether the same Agent-R1 framework transfers across different agent tasks. +2. Whether the context-management interface affects learning quality under a fixed training setup. + +## Experimental Setting + +We instantiate Agent-R1 with Qwen3-4B on GSM8K, HotpotQA, ALFWorld, and WebShop. These tasks cover arithmetic reasoning with tool interaction, retrieval-based multi-hop question answering, embodied household interaction, and simulated online shopping. + +For controlled comparisons, GSM8K is used as the main isolation setting. The environment, tool-based interaction format, rollout configuration, and reward definition are fixed, so differences can be attributed more directly to the optimizer or the context-management rule. The reward combines answer accuracy with a format component. + +## Main Results Across Scenarios + +The table below reports one representative metric for each task. Agent-R1 supports multiple RL methods under the same multi-turn interaction framework. + +| Method | GSM8K Acc. (%) | HotpotQA Acc. (%) | ALFWorld SR Seen (%) | ALFWorld SR Unseen (%) | WebShop Score (%) | WebShop SR (%) | +|---|---:|---:|---:|---:|---:|---:| +| ReAct | 53.1 | 25.8 | 7.14 | 2.98 | 51.58 | 23.8 | +| GRPO | **83.3** | **59.4** | **81.29** | **74.58** | 65.83 | 44.2 | +| PPO | 78.1 | 56.7 | 76.42 | 72.38 | **70.18** | **46.0** | +| REINFORCE++ | 78.9 | 52.8 | 73.84 | 69.57 | 63.41 | 41.8 | +| RLOO | 81.6 | 55.2 | 79.08 | 73.46 | 68.02 | 45.1 | + +All four RL methods outperform the training-free ReAct baseline across these settings. The best optimizer varies by task: GRPO leads on arithmetic reasoning, retrieval QA, and embodied interaction, while PPO is strongest on WebShop. This suggests that Agent-R1 is broad enough to support heterogeneous agent environments while preserving meaningful algorithm-specific behavior. + +## Learning Across Tasks + +Representative training curves on GSM8K, HotpotQA, and ALFWorld show clear upward trends under the same framework. The learning dynamics differ across tasks: GSM8K improves quickly and stabilizes early, HotpotQA shows slower and more fluctuating gains, and ALFWorld improves in a more stage-wise pattern with late jumps. + +This is useful for interpreting Agent-R1 as a framework rather than a single benchmark recipe. The same rollout and training abstraction can transfer across tasks, but each environment still exposes its own optimization dynamics. + +

Agent-R1 training curves across GSM8K, HotpotQA, and ALFWorld

+ +## Optimizer Comparison on GSM8K + +We compare PPO, GRPO, REINFORCE++, and RLOO under the same GSM8K environment, prompts, tool format, rollout configuration, and reward definition. The curves report reward, accuracy, and response length. + +Two patterns are notable. First, GRPO and RLOO reach the strongest late-stage accuracy, while PPO is more volatile. Second, REINFORCE++ behaves differently: it can reach relatively high accuracy while receiving a lower reward, which is consistent with shorter later-stage responses. Since the reward includes both answer accuracy and a format component, this indicates that high task accuracy does not necessarily mean the policy maximizes the full training signal. + +The takeaway is that Agent-R1 does not wash out optimizer-specific behavior. It makes that behavior observable under a common interaction setup. + +

GSM8K optimizer comparison under Agent-R1

+ +## Context-Management Strategies + +To test whether flexible context construction matters in practice, we compare three GSM8K context strategies under the same GRPO setup: + +- **Append-only context**: keeps growing the interaction history. +- **Sliding-window context**: keeps the original question plus the most recent tool output and model analysis. +- **LLM-summarized context**: compresses the evolving interaction history with an LLM summary. + +Sliding-window context performs best, append-only context is weaker, and LLM-summarized context underperforms in this small-model setting. This supports the main Agent-R1 design claim: context management is not just a presentation detail. Once the framework exposes context construction explicitly, different memory rules can be studied under the same rollout and optimizer. + +The summary result should not be read as a general rejection of summary-based memory. It shows that the quality of the transformation itself becomes part of the training problem. In this setting, preserving the most relevant recent evidence gives a cleaner learning signal than either unbounded history growth or noisy compression. + +

GSM8K context-management comparison under Agent-R1

diff --git a/docs/getting-started/installation-guide.md b/docs/getting-started/installation-guide.md index baf6ab6..0e416e7 100644 --- a/docs/getting-started/installation-guide.md +++ b/docs/getting-started/installation-guide.md @@ -4,7 +4,7 @@ Agent-R1 uses the same environment setup as `verl`. ## Base Environment -Follow the official [`verl` installation guide](https://verl.readthedocs.io/en/latest/start/install.html), but make sure the environment ends up with `verl==0.7.0`. +Follow the official [`verl` installation guide](https://verl.readthedocs.io/en/latest/start/install.html), but use a recent source checkout of `verl` rather than the old release used by earlier Agent-R1 versions. Agent-R1 relies on the newer AgentFlow / async rollout / reward-loop APIs and on `verl.trainer.config` being available as package data. If you want a broader overview of the base training workflow, the [`verl` quickstart](https://verl.readthedocs.io/en/latest/start/quickstart.html) is also useful. @@ -12,7 +12,7 @@ If you want a broader overview of the base training workflow, the [`verl` quicks Once the `verl` environment is working, Agent-R1 should run in the same environment. In practice, that means you can: -- prepare a Python environment with `verl==0.7.0` +- prepare a Python environment with a compatible recent source installation of `verl` - clone this repository - run Agent-R1 commands directly from the repository root diff --git a/docs/getting-started/quick-start.md b/docs/getting-started/quick-start.md index d205158..6d54f3a 100644 --- a/docs/getting-started/quick-start.md +++ b/docs/getting-started/quick-start.md @@ -1,38 +1,38 @@ # Quick Start -This quick start is a **sanity check**, not the main Agent-R1 workflow. Its purpose is to verify that your environment, dataset path, model path, and training stack are wired correctly. +This quick start is a compact **sanity check** for the current Agent-R1 workflow. Its purpose is to verify that your environment, dataset path, Qwen3 model path, rollout engine, and tool loop are wired correctly. -## 1. Prepare a Minimal Dataset +## 1. Prepare a Tool Dataset -Use the GSM8K preprocessing script: +Use the GSM8K tool preprocessing script: ```bash -python3 examples/data_preprocess/gsm8k.py --local_save_dir ~/data/gsm8k +python3 examples/data_preprocess/gsm8k_tool.py --local_save_dir ~/data/gsm8k_tool ``` This produces: -- `~/data/gsm8k/train.parquet` -- `~/data/gsm8k/test.parquet` +- `~/data/gsm8k_tool/train.parquet` +- `~/data/gsm8k_tool/test.parquet` ## 2. Run the Sanity Check Script -Use the provided single-step script: +Use the provided Qwen3 multi-step tool-use script: ```bash -bash examples/run_qwen2.5-3b.sh +bash examples/run_qwen3-4b_gsm8k_tool.sh ``` If needed, adjust the following values before running: - `CUDA_VISIBLE_DEVICES` - `actor_rollout_ref.model.path` -- dataset paths under `~/data/gsm8k` +- dataset paths under `~/data/gsm8k_tool` -The script entrypoint is [`examples/run_qwen2.5-3b.sh`](https://github.com/AgentR1/Agent-R1/blob/main/examples/run_qwen2.5-3b.sh), which launches `python3 -m agent_r1.trainer.main_agent_ppo`. +The script entrypoint is [`examples/run_qwen3-4b_gsm8k_tool.sh`](https://github.com/AgentR1/Agent-R1/blob/main/examples/run_qwen3-4b_gsm8k_tool.sh), which launches `python3 -m agent_r1.trainer.main_agent_ppo`. ## 3. What to Do Next - Read [`Step-level MDP`](../core-concepts/step-level-mdp.md) to understand the main training abstraction. - Read [`Layered Abstractions`](../core-concepts/layered-abstractions.md) to see how `AgentFlowBase`, `AgentEnvLoop`, and `ToolEnv` fit together. -- Continue to the [`Agent Task Tutorial`](../tutorials/agent-task.md) for the main Agent-R1 workflow based on multi-step interaction. +- Continue to the [`Datasets and Algorithms`](../tutorials/datasets-and-algorithms.md) guide for StepPO, HotpotQA, Paper Search, ALFWorld, WebShop, and baseline scripts. diff --git a/docs/index.md b/docs/index.md index a4aa9d6..fa01e4e 100644 --- a/docs/index.md +++ b/docs/index.md @@ -31,6 +31,7 @@ Agent-R1 is an open-source framework for training powerful language agents with - Start with [`Getting Started`](getting-started/index.md) if you want the minimal path: use the same environment as `verl`, run a sanity check, and confirm the repository is ready. - Read [`Step-level MDP`](core-concepts/step-level-mdp.md) and [`Layered Abstractions`](core-concepts/layered-abstractions.md) if you want to understand the framework design before touching code. - Follow [`Agent Task Tutorial`](tutorials/agent-task.md) if you want to see the main Agent-R1 workflow: multi-step interaction through `AgentEnvLoop` and `ToolEnv`. +- Use [`Datasets and Algorithms`](tutorials/datasets-and-algorithms.md) as the script reference for data preprocessing, runnable examples, and supported RL baselines. ## Scope of This Documentation diff --git a/docs/tutorials/datasets-and-algorithms.md b/docs/tutorials/datasets-and-algorithms.md new file mode 100644 index 0000000..302e911 --- /dev/null +++ b/docs/tutorials/datasets-and-algorithms.md @@ -0,0 +1,154 @@ +# Datasets, Data Processing, and Algorithms + +Agent-R1 is designed as a framework rather than a single benchmark implementation. A task recipe only needs to provide the same training data contract, an optional environment/tool configuration, and an agent flow. Once those pieces are present, the same trainer can run multiple RL algorithms by changing Hydra overrides. + +This page documents the datasets, preparation scripts, task recipes, and algorithm entry points included in this repository. StepPO is included as a first-class Agent-R1 recipe, while the remaining entries are generic baselines that can run on the same dataset and agent-flow contracts. + +## Current Runnable Scripts + +| Dataset | Main StepPO Script | Baseline Scripts | Data Preparation | +|---|---|---|---| +| GSM8K Tool | `examples/run_qwen3-4b_gsm8k_tool_steppo.sh` | GRPO, PPO, token GAE, RLOO, REINFORCE++ scripts under `examples/` | `examples/data_preprocess/gsm8k_tool.py` | +| HotpotQA | `examples/run_hotpotqa_steppo.sh` | `examples/hotpotqa/run_{grpo,gspo,token_adv,rloo,gigpo,reinforce_plus_plus}.sh` | `recipe/hotpotqa/prepare_hotpotqa_agent_r1.py` | +| Paper Search | `examples/run_papersearch_steppo.sh` | `examples/papersearch/run_{grpo,gspo,token_adv,rloo,gigpo,reinforce_plus_plus}.sh` | Built-in JSONL under `recipe/paper_search/inference/datasets/` + `recipe/paper_search/prepare_paper_search_agent_r1.py` | +| ALFWorld | `examples/run_alfworld_steppo.sh` | `examples/alfworld/run_{grpo,gspo,token_adv,rloo,gigpo,reinforce_plus_plus}.sh` | `recipe/alfworld/prepare_alfworld_agent_r1.py` | +| WebShop | `examples/run_webshop_steppo.sh` | `examples/webshop/run_{grpo,gspo,token_adv,rloo,gigpo,reinforce_plus_plus}.sh` | `recipe/webshop/prepare_webshop_agent_r1.py` | + +The repository includes the Paper Search raw JSONL files from the StepPO recipe: + +- `recipe/paper_search/inference/datasets/AutoScholarQuery/{train,dev,test,test_lt_5}.jsonl` +- `recipe/paper_search/inference/datasets/RealScholarQuery/test.jsonl` +- `recipe/paper_search/inference/datasets/test_case.jsonl` + +HotpotQA, ALFWorld, and WebShop are larger benchmark/environment datasets. They are prepared through the included scripts but are not committed as generated parquet, retrieval indexes, environment caches, or copied game/product artifacts. + +## Data Preparation + +```bash +# GSM8K Tool +python3 examples/data_preprocess/gsm8k_tool.py --local_save_dir ~/data/gsm8k_tool + +# HotpotQA + retrieval corpus +python3 recipe/hotpotqa/prepare_hotpotqa_agent_r1.py \ + --output_dir data/corpus/hotpotqa \ + --corpus_output_path data/corpus/hotpotqa_corpus/hpqa_corpus.jsonl + +# Optional HotpotQA FAISS assets +python3 recipe/hotpotqa/process_hotpotqa.py + +# Paper Search from bundled AutoScholarQuery JSONL files +python3 recipe/paper_search/prepare_paper_search_agent_r1.py \ + --input_dir recipe/paper_search/inference/datasets/AutoScholarQuery \ + --output_dir data/pasa + +# ALFWorld from local ALFWorld raw data +python3 recipe/alfworld/prepare_alfworld_agent_r1.py \ + --input_dir alfworld_data/json_2.1.1 \ + --output_dir data/alfworld + +# WebShop small/full data +python3 recipe/webshop/prepare_webshop_agent_r1.py \ + --dataset_mode small \ + --input_dir webshop_data \ + --output_dir data/webshop +``` + +Then choose any supported algorithm script. The StepPO entry points are: + +```bash +bash examples/run_qwen3-4b_gsm8k_tool.sh +bash examples/run_qwen3-4b_gsm8k_tool_steppo.sh +bash examples/run_hotpotqa_steppo.sh +bash examples/run_papersearch_steppo.sh +bash examples/run_alfworld_steppo.sh +bash examples/run_webshop_steppo.sh +``` + +All scripts accept additional Hydra overrides through `"$@"`, for example: + +```bash +bash examples/run_hotpotqa_steppo.sh \ + actor_rollout_ref.model.path=/path/to/model \ + trainer.n_gpus_per_node=4 +``` + +## Training Data Contract + +Agent-R1 uses parquet files compatible with the veRL trainer. For agent tasks, each row should include: + +| Field | Required | Meaning | +|---|---:|---| +| `data_source` | Yes | Dataset or benchmark name. | +| `prompt` | Yes | Chat messages passed to the tokenizer and rollout engine. | +| `ability` | Recommended | Task category used for logging and reward routing. | +| `reward_model` | Yes | Rule/model reward metadata, usually including `ground_truth`. | +| `extra_info` | Recommended | Split, index, raw question, raw answer, or task-specific metadata. | +| `agent_name` | Agent tasks | Agent flow name. `agent_env_loop` is the default runnable flow. | +| `env_kwargs` | Tool/env tasks | JSON string consumed by `AgentEnvLoop._create_env`. | + +`examples/data_preprocess/gsm8k_tool.py` is the smallest complete reference. It stores the environment configuration in `env_kwargs`: + +```json +{ + "env_type": "tool", + "tools": ["calc_gsm8k_reward"], + "tool_format": "hermes", + "tools_kwargs": {"ground_truth": ""} +} +``` + +This is the key extension point for new datasets. A HotpotQA recipe can point to a retrieval tool, a WebShop recipe can point to a web-shopping environment server, and an ALFWorld recipe can point to a text-world environment. The trainer does not need a dataset-specific rewrite as long as the recipe emits the same fields and the configured agent flow knows how to step the environment. + +## Dataset Recipe Pattern + +Agent-R1-derived recipes generally use the following layout: + +```text +recipe// + base.yaml # task-specific Hydra defaults + prepare__agent_r1.py # raw data -> parquet + _agent_flow.py # maps prompts/actions/observations into Agent-R1 steps + reward_fn.py # task reward + prompts.py # prompt templates + utils.py # task helpers + env/ # optional environment service or wrappers +``` + +This repository now includes the following concrete recipes: + +| Dataset / Environment | Data Processing | Environment / Flow | Notes | +|---|---|---|---| +| HotpotQA | `recipe/hotpotqa/prepare_hotpotqa_agent_r1.py`, `process_hotpotqa.py`, `build_retrieval_corpus.py`, `verify_dataset.py` | `HotpotQAAgentFlow` | Multi-hop QA with local retrieval. | +| Paper Search | `recipe/paper_search/prepare_paper_search_agent_r1.py`, bundled AutoScholarQuery/RealScholarQuery JSONL files, inference utilities | `PaperSearchAgentFlow` | Academic search and citation expansion. | +| ALFWorld | `recipe/alfworld/prepare_alfworld_agent_r1.py` | `AlfworldAgentFlow`, ALFWorld wrapper | Text-world embodied household tasks. | +| WebShop | `recipe/webshop/prepare_webshop_agent_r1.py`, index/artifact builders, environment server | `WebShopAgentFlow` | Web shopping navigation and scoring. | + +These recipes demonstrate that the Agent-R1 abstraction is not tied to GSM8K. They add task-specific data preparation and environment logic, while reusing the same rollout, reward-loop, and trainer interfaces. + +## Supported Algorithms + +Agent-R1 supports StepPO as a composed recipe: + +```bash +algorithm.adv_estimator=gae +actor_rollout_ref.actor.policy_loss.loss_mode=gspo +actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean +``` + +The first setting computes credit assignment over the agent step timeline. The GSPO policy loss then uses sequence-level importance ratios for the complete generated action at each step. + +The Agent-R1 trainer also routes `algorithm.adv_estimator` to the following estimators: + +| `algorithm.adv_estimator` | Granularity | Critic Required | Typical Use | +|---|---|---:|---| +| `gae` | Step-level | Yes | PPO-style actor-critic over agent steps. | +| `token_gae` | Token-level | Yes | Token-level actor-critic baseline for multi-step rollouts. | +| `grpo` | Trajectory outcome | No | Group-relative outcome optimization. | +| `rloo` | Trajectory outcome | No | Leave-one-out baseline over multiple rollouts. | +| `reinforce_plus_plus` | Token return | No | REINFORCE++ baseline with KL in reward. | +| `reinforce_plus_plus_baseline` | Trajectory outcome | No | REINFORCE++ with prompt-level mean baseline. | +| `gigpo` | Trajectory + step group | No | Requires `anchor_obs` in the agent flow for step grouping. | + +The policy objective is controlled separately by `actor_rollout_ref.actor.policy_loss.loss_mode`. Agent-R1 keeps this axis separate from advantage estimation so that a task recipe can compare different credit-assignment strategies without changing the environment. + +For convenience, the `examples/run_*_steppo.sh` scripts set the StepPO combination directly. Other algorithms can still be selected by overriding `algorithm.adv_estimator` and, when needed, `actor_rollout_ref.actor.policy_loss.loss_mode`. diff --git a/docs/tutorials/index.md b/docs/tutorials/index.md index d3ea23d..2f62af8 100644 --- a/docs/tutorials/index.md +++ b/docs/tutorials/index.md @@ -5,6 +5,7 @@ This section contains task-oriented walkthroughs built on the real examples in t ## In This Section - [`Agent Task Tutorial`](agent-task.md): the main Agent-R1 workflow based on multi-step interaction through `AgentEnvLoop` and `ToolEnv`. +- [`Datasets and Algorithms`](datasets-and-algorithms.md): scripts, data-processing contracts, compatible dataset recipes, and supported RL baselines. ## Scope diff --git a/examples/alfworld/run_gigpo.sh b/examples/alfworld/run_gigpo.sh new file mode 100755 index 0000000..22d132b --- /dev/null +++ b/examples/alfworld/run_gigpo.sh @@ -0,0 +1,21 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-alfworld_gigpo}" +export ALFWORLD_VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$ROOT_DIR/outputs/alfworld_validation/gigpo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GIGPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_alfworld_grpo.sh" \ + algorithm.adv_estimator=gigpo \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GIGPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + ++algorithm.gigpo.step_advantage_w="${AGENT_R1_GIGPO_STEP_ADVANTAGE_W:-1.0}" \ + ++algorithm.gigpo.mode="${AGENT_R1_GIGPO_MODE:-mean_std_norm}" \ + ++algorithm.gigpo.enable_similarity="${AGENT_R1_GIGPO_ENABLE_SIMILARITY:-False}" \ + ++algorithm.gigpo.similarity_thresh="${AGENT_R1_GIGPO_SIMILARITY_THRESH:-0.95}" \ + "$@" diff --git a/examples/alfworld/run_grpo.sh b/examples/alfworld/run_grpo.sh new file mode 100755 index 0000000..d9b782b --- /dev/null +++ b/examples/alfworld/run_grpo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_alfworld_grpo.sh" "$@" diff --git a/examples/alfworld/run_gspo.sh b/examples/alfworld/run_gspo.sh new file mode 100755 index 0000000..4277257 --- /dev/null +++ b/examples/alfworld/run_gspo.sh @@ -0,0 +1,19 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-alfworld_gspo}" +export ALFWORLD_VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$ROOT_DIR/outputs/alfworld_validation/gspo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GSPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_alfworld_grpo.sh" \ + algorithm.adv_estimator=grpo \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GSPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + "$@" diff --git a/examples/alfworld/run_reinforce_plus_plus.sh b/examples/alfworld/run_reinforce_plus_plus.sh new file mode 100755 index 0000000..a615083 --- /dev/null +++ b/examples/alfworld/run_reinforce_plus_plus.sh @@ -0,0 +1,18 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-alfworld_reinforce_plus_plus}" +export ALFWORLD_VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$ROOT_DIR/outputs/alfworld_validation/reinforce_plus_plus}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_alfworld_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/alfworld/run_reinforce_plus_plus_baseline.sh b/examples/alfworld/run_reinforce_plus_plus_baseline.sh new file mode 100755 index 0000000..0e690ac --- /dev/null +++ b/examples/alfworld/run_reinforce_plus_plus_baseline.sh @@ -0,0 +1,18 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-alfworld_reinforce_plus_plus_baseline}" +export ALFWORLD_VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$ROOT_DIR/outputs/alfworld_validation/reinforce_plus_plus_baseline}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_alfworld_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus_baseline \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/alfworld/run_rloo.sh b/examples/alfworld/run_rloo.sh new file mode 100755 index 0000000..0b0b66d --- /dev/null +++ b/examples/alfworld/run_rloo.sh @@ -0,0 +1,17 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-alfworld_rloo}" +export ALFWORLD_VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$ROOT_DIR/outputs/alfworld_validation/rloo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_RLOO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_alfworld_grpo.sh" \ + algorithm.adv_estimator=rloo \ + actor_rollout_ref.actor.use_kl_loss=False \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_RLOO_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_RLOO_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/alfworld/run_step_adv.sh b/examples/alfworld/run_step_adv.sh new file mode 100755 index 0000000..4cd6703 --- /dev/null +++ b/examples/alfworld/run_step_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_alfworld_step_adv.sh" "$@" diff --git a/examples/alfworld/run_steppo.sh b/examples/alfworld/run_steppo.sh new file mode 100755 index 0000000..052d10c --- /dev/null +++ b/examples/alfworld/run_steppo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_alfworld_steppo.sh" "$@" diff --git a/examples/alfworld/run_token_adv.sh b/examples/alfworld/run_token_adv.sh new file mode 100755 index 0000000..c229567 --- /dev/null +++ b/examples/alfworld/run_token_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_alfworld_token_adv.sh" "$@" diff --git a/examples/hotpotqa/run_gigpo.sh b/examples/hotpotqa/run_gigpo.sh new file mode 100755 index 0000000..5f2e7c9 --- /dev/null +++ b/examples/hotpotqa/run_gigpo.sh @@ -0,0 +1,20 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-hotpotqa_gigpo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GIGPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_grpo.sh" \ + algorithm.adv_estimator=gigpo \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GIGPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + ++algorithm.gigpo.step_advantage_w="${AGENT_R1_GIGPO_STEP_ADVANTAGE_W:-1.0}" \ + ++algorithm.gigpo.mode="${AGENT_R1_GIGPO_MODE:-mean_std_norm}" \ + ++algorithm.gigpo.enable_similarity="${AGENT_R1_GIGPO_ENABLE_SIMILARITY:-False}" \ + ++algorithm.gigpo.similarity_thresh="${AGENT_R1_GIGPO_SIMILARITY_THRESH:-0.95}" \ + "$@" diff --git a/examples/hotpotqa/run_grpo.sh b/examples/hotpotqa/run_grpo.sh new file mode 100755 index 0000000..c521cc7 --- /dev/null +++ b/examples/hotpotqa/run_grpo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_grpo.sh" "$@" diff --git a/examples/hotpotqa/run_gspo.sh b/examples/hotpotqa/run_gspo.sh new file mode 100755 index 0000000..a2fe733 --- /dev/null +++ b/examples/hotpotqa/run_gspo.sh @@ -0,0 +1,18 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-hotpotqa_gspo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GSPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_grpo.sh" \ + algorithm.adv_estimator=grpo \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GSPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + "$@" diff --git a/examples/hotpotqa/run_reinforce_plus_plus.sh b/examples/hotpotqa/run_reinforce_plus_plus.sh new file mode 100755 index 0000000..7e3d0f8 --- /dev/null +++ b/examples/hotpotqa/run_reinforce_plus_plus.sh @@ -0,0 +1,17 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-hotpotqa_reinforce_plus_plus}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/hotpotqa/run_reinforce_plus_plus_baseline.sh b/examples/hotpotqa/run_reinforce_plus_plus_baseline.sh new file mode 100755 index 0000000..5702a87 --- /dev/null +++ b/examples/hotpotqa/run_reinforce_plus_plus_baseline.sh @@ -0,0 +1,17 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-hotpotqa_reinforce_plus_plus_baseline}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus_baseline \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/hotpotqa/run_rloo.sh b/examples/hotpotqa/run_rloo.sh new file mode 100755 index 0000000..d678afc --- /dev/null +++ b/examples/hotpotqa/run_rloo.sh @@ -0,0 +1,16 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-hotpotqa_rloo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_RLOO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_grpo.sh" \ + algorithm.adv_estimator=rloo \ + actor_rollout_ref.actor.use_kl_loss=False \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_RLOO_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_RLOO_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/hotpotqa/run_step_adv.sh b/examples/hotpotqa/run_step_adv.sh new file mode 100755 index 0000000..64155b7 --- /dev/null +++ b/examples/hotpotqa/run_step_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_step_adv.sh" "$@" diff --git a/examples/hotpotqa/run_steppo.sh b/examples/hotpotqa/run_steppo.sh new file mode 100755 index 0000000..4b1de23 --- /dev/null +++ b/examples/hotpotqa/run_steppo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_steppo.sh" "$@" diff --git a/examples/hotpotqa/run_token_adv.sh b/examples/hotpotqa/run_token_adv.sh new file mode 100755 index 0000000..df113f3 --- /dev/null +++ b/examples/hotpotqa/run_token_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_hotpotqa_token_adv.sh" "$@" diff --git a/examples/papersearch/run_gigpo.sh b/examples/papersearch/run_gigpo.sh new file mode 100755 index 0000000..246c7d4 --- /dev/null +++ b/examples/papersearch/run_gigpo.sh @@ -0,0 +1,20 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-papersearch_gigpo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GIGPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_papersearch_grpo.sh" \ + algorithm.adv_estimator=gigpo \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GIGPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + ++algorithm.gigpo.step_advantage_w="${AGENT_R1_GIGPO_STEP_ADVANTAGE_W:-1.0}" \ + ++algorithm.gigpo.mode="${AGENT_R1_GIGPO_MODE:-mean_std_norm}" \ + ++algorithm.gigpo.enable_similarity="${AGENT_R1_GIGPO_ENABLE_SIMILARITY:-False}" \ + ++algorithm.gigpo.similarity_thresh="${AGENT_R1_GIGPO_SIMILARITY_THRESH:-0.95}" \ + "$@" diff --git a/examples/papersearch/run_grpo.sh b/examples/papersearch/run_grpo.sh new file mode 100755 index 0000000..b436620 --- /dev/null +++ b/examples/papersearch/run_grpo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_papersearch_grpo.sh" "$@" diff --git a/examples/papersearch/run_gspo.sh b/examples/papersearch/run_gspo.sh new file mode 100755 index 0000000..a1b8746 --- /dev/null +++ b/examples/papersearch/run_gspo.sh @@ -0,0 +1,18 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-papersearch_gspo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GSPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_papersearch_grpo.sh" \ + algorithm.adv_estimator=grpo \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GSPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + "$@" diff --git a/examples/papersearch/run_reinforce_plus_plus.sh b/examples/papersearch/run_reinforce_plus_plus.sh new file mode 100755 index 0000000..14d59e0 --- /dev/null +++ b/examples/papersearch/run_reinforce_plus_plus.sh @@ -0,0 +1,17 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-papersearch_reinforce_plus_plus}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_papersearch_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/papersearch/run_reinforce_plus_plus_baseline.sh b/examples/papersearch/run_reinforce_plus_plus_baseline.sh new file mode 100755 index 0000000..5dd48fb --- /dev/null +++ b/examples/papersearch/run_reinforce_plus_plus_baseline.sh @@ -0,0 +1,17 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-papersearch_reinforce_plus_plus_baseline}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_papersearch_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus_baseline \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/papersearch/run_rloo.sh b/examples/papersearch/run_rloo.sh new file mode 100755 index 0000000..78eb7f1 --- /dev/null +++ b/examples/papersearch/run_rloo.sh @@ -0,0 +1,16 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-papersearch_rloo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_RLOO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_papersearch_grpo.sh" \ + algorithm.adv_estimator=rloo \ + actor_rollout_ref.actor.use_kl_loss=False \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_RLOO_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_RLOO_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/papersearch/run_step_adv.sh b/examples/papersearch/run_step_adv.sh new file mode 100755 index 0000000..92e6fec --- /dev/null +++ b/examples/papersearch/run_step_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_papersearch_step_adv.sh" "$@" diff --git a/examples/papersearch/run_steppo.sh b/examples/papersearch/run_steppo.sh new file mode 100755 index 0000000..2ce77f1 --- /dev/null +++ b/examples/papersearch/run_steppo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_papersearch_steppo.sh" "$@" diff --git a/examples/papersearch/run_token_adv.sh b/examples/papersearch/run_token_adv.sh new file mode 100755 index 0000000..d0c6002 --- /dev/null +++ b/examples/papersearch/run_token_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_papersearch_token_adv.sh" "$@" diff --git a/examples/run_alfworld_grpo.sh b/examples/run_alfworld_grpo.sh new file mode 100755 index 0000000..7627bf6 --- /dev/null +++ b/examples/run_alfworld_grpo.sh @@ -0,0 +1,108 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/alfworld}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES="${CUDA_VISIBLE_DEVICES:-0,1,2,3}" +export VLLM_USE_V1=1 +export HF_ENDPOINT="${HF_ENDPOINT:-https://hf-mirror.com}" +export CUDA_HOME="${CUDA_HOME:-/usr/local/cuda}" +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI="${MLFLOW_TRACKING_URI:-http://127.0.0.1:5000}" + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/alfworld/base.yaml" + +ALFWORLD_MODEL_PATH="${ALFWORLD_MODEL_PATH:-Qwen/Qwen3-4B-Instruct-2507}" +ALFWORLD_MAX_PROMPT_LEN="${ALFWORLD_MAX_PROMPT_LEN:-8192}" +ALFWORLD_MAX_RESPONSE_LEN="${ALFWORLD_MAX_RESPONSE_LEN:-4096}" +ALFWORLD_TRAIN_PATH="${ALFWORLD_TRAIN_PATH:-$PROJECT_DIR/data/alfworld/train.parquet}" +ALFWORLD_VAL_SEEN_PATH="${ALFWORLD_VAL_SEEN_PATH:-$PROJECT_DIR/data/alfworld/valid_seen.parquet}" +ALFWORLD_VAL_UNSEEN_PATH="${ALFWORLD_VAL_UNSEEN_PATH:-$PROJECT_DIR/data/alfworld/valid_unseen.parquet}" +export ALFWORLD_DATA_ROOT="${ALFWORLD_DATA_ROOT:-$PROJECT_DIR/data/alfworld}" +VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$PROJECT_DIR/outputs/alfworld_validation/grpo}" + +# GRPO: multiple independent rollouts per sampled task for group-relative advantages (verl rollout.n). +AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GRPO_ROLLOUT_N:-8}" +# Match token_gae script's 128 unique tasks per step: train_batch_size * rollout.n ~= 128 (fewer prompts, more rollouts). +ALFWORLD_GRPO_BASE_TRAIN_BATCH="${ALFWORLD_GRPO_BASE_TRAIN_BATCH:-128}" +ALFWORLD_GRPO_BASE_LOG_PROB_MICRO_BATCH="${ALFWORLD_GRPO_BASE_LOG_PROB_MICRO_BATCH:-32}" +ALFWORLD_TRAIN_BATCH_SIZE="$((ALFWORLD_GRPO_BASE_TRAIN_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +ALFWORLD_LOG_PROB_MICRO_BATCH="$((ALFWORLD_GRPO_BASE_LOG_PROB_MICRO_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +if [[ "$ALFWORLD_TRAIN_BATCH_SIZE" -lt 1 ]]; then + echo "❌ ALFWORLD_GRPO_BASE_TRAIN_BATCH ($ALFWORLD_GRPO_BASE_TRAIN_BATCH) must be >= AGENT_R1_GRPO_ROLLOUT_N ($AGENT_R1_GRPO_ROLLOUT_N)." >&2 + exit 1 +fi +if [[ "$ALFWORLD_LOG_PROB_MICRO_BATCH" -lt 1 ]]; then + ALFWORLD_LOG_PROB_MICRO_BATCH=1 +fi + +PROJECT_NAME="${PROJECT_NAME:-ALFWorld_Agent-R1}" +EXP_NAME="${EXP_NAME:-alfworld_grpo}" + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=grpo \ + algorithm.norm_adv_by_std_in_grpo="${AGENT_R1_NORM_ADV_BY_STD_IN_GRPO:-True}" \ + data.train_files="$ALFWORLD_TRAIN_PATH" \ + data.val_files="[\"$ALFWORLD_VAL_SEEN_PATH\",\"$ALFWORLD_VAL_UNSEEN_PATH\"]" \ + data.train_batch_size="$ALFWORLD_TRAIN_BATCH_SIZE" \ + data.max_prompt_length="$ALFWORLD_MAX_PROMPT_LEN" \ + data.max_response_length="$ALFWORLD_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$ALFWORLD_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size="$ALFWORLD_TRAIN_BATCH_SIZE" \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu="$ALFWORLD_LOG_PROB_MICRO_BATCH" \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.5 \ + actor_rollout_ref.rollout.n="$AGENT_R1_GRPO_ROLLOUT_N" \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu="$ALFWORLD_LOG_PROB_MICRO_BATCH" \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=alfworld_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/alfworld/reward_fn.py \ + custom_reward_function.name=compute_score \ + critic.enable=False \ + algorithm.use_kl_in_reward=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.validation_data_dir="$VAL_DUMP_DIR" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=50 \ + trainer.test_freq=5 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.total_epochs=10 "$@" diff --git a/examples/run_alfworld_step_adv.sh b/examples/run_alfworld_step_adv.sh new file mode 100755 index 0000000..df68d73 --- /dev/null +++ b/examples/run_alfworld_step_adv.sh @@ -0,0 +1,101 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/alfworld}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES="${CUDA_VISIBLE_DEVICES:-0,1,2,3}" +export VLLM_USE_V1=1 +export HF_ENDPOINT="${HF_ENDPOINT:-https://hf-mirror.com}" +export CUDA_HOME="${CUDA_HOME:-/usr/local/cuda}" +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI="${MLFLOW_TRACKING_URI:-http://127.0.0.1:5000}" + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/alfworld/base.yaml" + +ALFWORLD_MODEL_PATH="${ALFWORLD_MODEL_PATH:-Qwen/Qwen3-4B-Instruct-2507}" +ALFWORLD_MAX_PROMPT_LEN="${ALFWORLD_MAX_PROMPT_LEN:-8192}" +ALFWORLD_MAX_RESPONSE_LEN="${ALFWORLD_MAX_RESPONSE_LEN:-4096}" +ALFWORLD_TRAIN_PATH="${ALFWORLD_TRAIN_PATH:-$PROJECT_DIR/data/alfworld/train.parquet}" +ALFWORLD_VAL_SEEN_PATH="${ALFWORLD_VAL_SEEN_PATH:-$PROJECT_DIR/data/alfworld/valid_seen.parquet}" +ALFWORLD_VAL_UNSEEN_PATH="${ALFWORLD_VAL_UNSEEN_PATH:-$PROJECT_DIR/data/alfworld/valid_unseen.parquet}" +export ALFWORLD_DATA_ROOT="${ALFWORLD_DATA_ROOT:-$PROJECT_DIR/data/alfworld}" +VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$PROJECT_DIR/outputs/alfworld_validation/step_adv}" +ROLLOUT_DUMP_DIR="${ALFWORLD_ROLLOUT_DUMP_DIR:-$PROJECT_DIR/outputs/alfworld_rollout/step_adv}" + +PROJECT_NAME="${PROJECT_NAME:-ALFWorld_Agent-R1}" +EXP_NAME="${EXP_NAME:-alfworld_step_adv}" + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=gae \ + data.train_files="$ALFWORLD_TRAIN_PATH" \ + data.val_files="[\"$ALFWORLD_VAL_SEEN_PATH\",\"$ALFWORLD_VAL_UNSEEN_PATH\"]" \ + data.train_batch_size=128 \ + data.max_prompt_length="$ALFWORLD_MAX_PROMPT_LEN" \ + data.max_response_length="$ALFWORLD_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$ALFWORLD_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=128 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.5 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=alfworld_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/alfworld/reward_fn.py \ + custom_reward_function.name=compute_score \ + critic.model.path="$ALFWORLD_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=16 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.validation_data_dir="$VAL_DUMP_DIR" \ + trainer.rollout_data_dir="$ROLLOUT_DUMP_DIR" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=50 \ + trainer.test_freq=5 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=10 "$@" diff --git a/examples/run_alfworld_steppo.sh b/examples/run_alfworld_steppo.sh new file mode 100755 index 0000000..22bd039 --- /dev/null +++ b/examples/run_alfworld_steppo.sh @@ -0,0 +1,6 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_alfworld_step_adv.sh" "$@" diff --git a/examples/run_alfworld_token_adv.sh b/examples/run_alfworld_token_adv.sh new file mode 100755 index 0000000..6be2901 --- /dev/null +++ b/examples/run_alfworld_token_adv.sh @@ -0,0 +1,98 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/alfworld}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES="${CUDA_VISIBLE_DEVICES:-0,1,2,3}" +export VLLM_USE_V1=1 +export HF_ENDPOINT="${HF_ENDPOINT:-https://hf-mirror.com}" +export CUDA_HOME="${CUDA_HOME:-/usr/local/cuda}" +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI="${MLFLOW_TRACKING_URI:-http://127.0.0.1:5000}" + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/alfworld/base.yaml" + +ALFWORLD_MODEL_PATH="${ALFWORLD_MODEL_PATH:-Qwen/Qwen3-4B-Instruct-2507}" +ALFWORLD_MAX_PROMPT_LEN="${ALFWORLD_MAX_PROMPT_LEN:-8192}" +ALFWORLD_MAX_RESPONSE_LEN="${ALFWORLD_MAX_RESPONSE_LEN:-4096}" +ALFWORLD_TRAIN_PATH="${ALFWORLD_TRAIN_PATH:-$PROJECT_DIR/data/alfworld/train.parquet}" +ALFWORLD_VAL_SEEN_PATH="${ALFWORLD_VAL_SEEN_PATH:-$PROJECT_DIR/data/alfworld/valid_seen.parquet}" +ALFWORLD_VAL_UNSEEN_PATH="${ALFWORLD_VAL_UNSEEN_PATH:-$PROJECT_DIR/data/alfworld/valid_unseen.parquet}" +export ALFWORLD_DATA_ROOT="${ALFWORLD_DATA_ROOT:-$PROJECT_DIR/data/alfworld}" +VAL_DUMP_DIR="${ALFWORLD_VAL_DUMP_DIR:-$PROJECT_DIR/outputs/alfworld_validation/token_adv}" + +PROJECT_NAME="${PROJECT_NAME:-ALFWorld_Agent-R1}" +EXP_NAME="${EXP_NAME:-alfworld_token_adv}" + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=token_gae \ + data.train_files="$ALFWORLD_TRAIN_PATH" \ + data.val_files="[\"$ALFWORLD_VAL_SEEN_PATH\",\"$ALFWORLD_VAL_UNSEEN_PATH\"]" \ + data.train_batch_size=128 \ + data.max_prompt_length="$ALFWORLD_MAX_PROMPT_LEN" \ + data.max_response_length="$ALFWORLD_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$ALFWORLD_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=128 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.5 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=alfworld_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/alfworld/reward_fn.py \ + custom_reward_function.name=compute_score \ + critic.model.path="$ALFWORLD_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=16 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.validation_data_dir="$VAL_DUMP_DIR" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=50 \ + trainer.test_freq=5 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=10 "$@" diff --git a/examples/run_hotpotqa_grpo.sh b/examples/run_hotpotqa_grpo.sh new file mode 100755 index 0000000..16627a5 --- /dev/null +++ b/examples/run_hotpotqa_grpo.sh @@ -0,0 +1,125 @@ +set -x + +# 4 GPUs: training + vLLM use 0–3; each of the 4 agent workers runs BGE on its own GPU (cuda:0..3). +export CUDA_VISIBLE_DEVICES=${CUDA_VISIBLE_DEVICES:-0,1,2,3} +export HOTPOTQA_EMBEDDING_PER_WORKER_GPU=${HOTPOTQA_EMBEDDING_PER_WORKER_GPU:-1} +export VLLM_USE_V1=1 +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI=${MLFLOW_TRACKING_URI:-http://172.17.0.1:5000} + +# GRPO: multiple rollouts per task for group-relative advantages (verl rollout.n). +AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GRPO_ROLLOUT_N:-8}" +# Match run_hotpotqa_token_adv.sh (256 unique tasks): train_batch_size * rollout.n ~= 256. +HOTPOTQA_GRPO_BASE_TRAIN_BATCH="${HOTPOTQA_GRPO_BASE_TRAIN_BATCH:-256}" +HOTPOTQA_GRPO_BASE_LOG_PROB_MICRO_BATCH="${HOTPOTQA_GRPO_BASE_LOG_PROB_MICRO_BATCH:-8}" +HOTPOTQA_TRAIN_BATCH_SIZE="$((HOTPOTQA_GRPO_BASE_TRAIN_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +HOTPOTQA_LOG_PROB_MICRO_BATCH="$((HOTPOTQA_GRPO_BASE_LOG_PROB_MICRO_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +if [[ "$HOTPOTQA_TRAIN_BATCH_SIZE" -lt 1 ]]; then + echo "❌ HOTPOTQA_GRPO_BASE_TRAIN_BATCH ($HOTPOTQA_GRPO_BASE_TRAIN_BATCH) must be >= AGENT_R1_GRPO_ROLLOUT_N ($AGENT_R1_GRPO_ROLLOUT_N)." >&2 + exit 1 +fi +if [[ "$HOTPOTQA_LOG_PROB_MICRO_BATCH" -lt 1 ]]; then + HOTPOTQA_LOG_PROB_MICRO_BATCH=1 +fi + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/hotpotqa/base_faiss_cpu.yaml" + +export HOTPOTQA_DATA_ROOT="${HOTPOTQA_DATA_ROOT:-$PROJECT_DIR/data/corpus/hotpotqa}" + +HOTPOTQA_MODEL_PATH=${HOTPOTQA_MODEL_PATH:-Qwen/Qwen3-4B-Instruct-2507} + +# Length budget (vs. Agent-R1-legacy `run_ppo_hotpotqa.sh`; semantics differ): +# - Legacy: multi-turn tokens concatenated into one trajectory → data.max_prompt_length=8192, full response=8192, +# per-turn max_response_length_single_turn=1024. +# - This script (Agent-R1 AgentFlow): each step rebuilds the prompt + one generate per step; the user block adds a +# "Recent tool / format issues" section, so we use a larger prompt budget; per-step response matches legacy 1024 +# to reduce JSON truncation at max_tokens. +HOTPOTQA_MAX_PROMPT_LEN=${HOTPOTQA_MAX_PROMPT_LEN:-10240} +HOTPOTQA_MAX_RESPONSE_LEN=${HOTPOTQA_MAX_RESPONSE_LEN:-1024} + +TRAIN_PATH="$PROJECT_DIR/data/corpus/hotpotqa/train.parquet" +VAL_PATH="$PROJECT_DIR/data/corpus/hotpotqa/validation.parquet" +HOTPOTQA_WIKI2_VAL_PATH="${HOTPOTQA_WIKI2_VAL_PATH:-$PROJECT_DIR/data/corpus/hotpotqa/2wikimultihopqa_validation.parquet}" +HOTPOTQA_MUSIQUE_VAL_PATH="${HOTPOTQA_MUSIQUE_VAL_PATH:-$PROJECT_DIR/data/corpus/hotpotqa/musique_validation.parquet}" +HOTPOTQA_ENABLE_CROSS_EVAL="${HOTPOTQA_ENABLE_CROSS_EVAL:-auto}" + +build_val_files() { + local files=("$VAL_PATH") + local include_missing=false + if [[ "$HOTPOTQA_ENABLE_CROSS_EVAL" == "1" || "$HOTPOTQA_ENABLE_CROSS_EVAL" == "true" ]]; then + include_missing=true + fi + if [[ "$HOTPOTQA_ENABLE_CROSS_EVAL" != "0" && "$HOTPOTQA_ENABLE_CROSS_EVAL" != "false" ]]; then + if [[ -f "$HOTPOTQA_WIKI2_VAL_PATH" || "$include_missing" == true ]]; then + files+=("$HOTPOTQA_WIKI2_VAL_PATH") + fi + if [[ -f "$HOTPOTQA_MUSIQUE_VAL_PATH" || "$include_missing" == true ]]; then + files+=("$HOTPOTQA_MUSIQUE_VAL_PATH") + fi + fi + + local val_files="[" + local path + for path in "${files[@]}"; do + val_files+="\"$path\"," + done + val_files="${val_files%,}]" + echo "$val_files" +} + +VAL_FILES="$(build_val_files)" + +PROJECT_NAME='HotpotQA_Agent-R1' +EXP_NAME='hotpotqa_grpo' + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=grpo \ + algorithm.norm_adv_by_std_in_grpo="${AGENT_R1_NORM_ADV_BY_STD_IN_GRPO:-True}" \ + data.train_files="$TRAIN_PATH" \ + data.val_files="$VAL_FILES" \ + data.train_batch_size="$HOTPOTQA_TRAIN_BATCH_SIZE" \ + data.max_prompt_length="$HOTPOTQA_MAX_PROMPT_LEN" \ + data.max_response_length="$HOTPOTQA_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$HOTPOTQA_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.ppo_mini_batch_size="$HOTPOTQA_TRAIN_BATCH_SIZE" \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu="$HOTPOTQA_LOG_PROB_MICRO_BATCH" \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.n="$AGENT_R1_GRPO_ROLLOUT_N" \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=hotpotqa_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + critic.enable=False \ + algorithm.use_kl_in_reward=False \ + algorithm.gamma=0.99 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/hotpotqa/reward_fn.py \ + custom_reward_function.name=compute_score \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=10 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.total_epochs=5 "$@" diff --git a/examples/run_hotpotqa_step_adv.sh b/examples/run_hotpotqa_step_adv.sh new file mode 100755 index 0000000..a87c30b --- /dev/null +++ b/examples/run_hotpotqa_step_adv.sh @@ -0,0 +1,115 @@ +set -x + +# 4 GPUs: training + vLLM use 0–3; each of the 4 agent workers runs BGE on its own GPU (cuda:0..3). +export CUDA_VISIBLE_DEVICES=${CUDA_VISIBLE_DEVICES:-0,1,2,3} +export HOTPOTQA_EMBEDDING_PER_WORKER_GPU=${HOTPOTQA_EMBEDDING_PER_WORKER_GPU:-1} +export VLLM_USE_V1=1 +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI=${MLFLOW_TRACKING_URI:-http://172.17.0.1:5000} + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/hotpotqa/base_faiss_cpu.yaml" +export HOTPOTQA_DATA_ROOT="${HOTPOTQA_DATA_ROOT:-$PROJECT_DIR/data/corpus/hotpotqa}" + +HOTPOTQA_MODEL_PATH=${HOTPOTQA_MODEL_PATH:-Qwen/Qwen3-4B-Instruct-2507} + +# Length budget (vs. Agent-R1-legacy `run_ppo_hotpotqa.sh`; semantics differ): +# - Legacy: multi-turn tokens concatenated into one trajectory → data.max_prompt_length=8192, full response=8192, +# per-turn max_response_length_single_turn=1024. +# - This script (Agent-R1 AgentFlow): each step rebuilds the prompt + one generate per step; the user block adds a +# "Recent tool / format issues" section, so we use a larger prompt budget; per-step response matches legacy 1024 +# to reduce JSON truncation at max_tokens. +HOTPOTQA_MAX_PROMPT_LEN=${HOTPOTQA_MAX_PROMPT_LEN:-10240} +HOTPOTQA_MAX_RESPONSE_LEN=${HOTPOTQA_MAX_RESPONSE_LEN:-1024} + +TRAIN_PATH="$PROJECT_DIR/data/corpus/hotpotqa/train.parquet" +VAL_PATH="$PROJECT_DIR/data/corpus/hotpotqa/validation.parquet" +HOTPOTQA_WIKI2_VAL_PATH="${HOTPOTQA_WIKI2_VAL_PATH:-$PROJECT_DIR/data/corpus/hotpotqa/2wikimultihopqa_validation.parquet}" +HOTPOTQA_MUSIQUE_VAL_PATH="${HOTPOTQA_MUSIQUE_VAL_PATH:-$PROJECT_DIR/data/corpus/hotpotqa/musique_validation.parquet}" +HOTPOTQA_ENABLE_CROSS_EVAL="${HOTPOTQA_ENABLE_CROSS_EVAL:-auto}" + +build_val_files() { + local files=("$VAL_PATH") + local include_missing=false + if [[ "$HOTPOTQA_ENABLE_CROSS_EVAL" == "1" || "$HOTPOTQA_ENABLE_CROSS_EVAL" == "true" ]]; then + include_missing=true + fi + if [[ "$HOTPOTQA_ENABLE_CROSS_EVAL" != "0" && "$HOTPOTQA_ENABLE_CROSS_EVAL" != "false" ]]; then + if [[ -f "$HOTPOTQA_WIKI2_VAL_PATH" || "$include_missing" == true ]]; then + files+=("$HOTPOTQA_WIKI2_VAL_PATH") + fi + if [[ -f "$HOTPOTQA_MUSIQUE_VAL_PATH" || "$include_missing" == true ]]; then + files+=("$HOTPOTQA_MUSIQUE_VAL_PATH") + fi + fi + + local val_files="[" + local path + for path in "${files[@]}"; do + val_files+="\"$path\"," + done + val_files="${val_files%,}]" + echo "$val_files" +} + +VAL_FILES="$(build_val_files)" + +PROJECT_NAME='HotpotQA_Agent-R1' +EXP_NAME='hotpotqa_step_level_0.99_adv_mlflow_4gpu' + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=gae \ + data.train_files="$TRAIN_PATH" \ + data.val_files="$VAL_FILES" \ + data.train_batch_size=256 \ + data.max_prompt_length="$HOTPOTQA_MAX_PROMPT_LEN" \ + data.max_response_length="$HOTPOTQA_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$HOTPOTQA_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=256 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=8 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=hotpotqa_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + critic.model.path="$HOTPOTQA_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=4 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + algorithm.gamma=0.99 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/hotpotqa/reward_fn.py \ + custom_reward_function.name=compute_score \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=10 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=5 "$@" diff --git a/examples/run_hotpotqa_steppo.sh b/examples/run_hotpotqa_steppo.sh new file mode 100755 index 0000000..79f2659 --- /dev/null +++ b/examples/run_hotpotqa_steppo.sh @@ -0,0 +1,6 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_hotpotqa_step_adv.sh" "$@" diff --git a/examples/run_hotpotqa_token_adv.sh b/examples/run_hotpotqa_token_adv.sh new file mode 100755 index 0000000..b995e47 --- /dev/null +++ b/examples/run_hotpotqa_token_adv.sh @@ -0,0 +1,114 @@ +set -x + +# 4 GPUs: training + vLLM use 0–3; each of the 4 agent workers runs BGE on its own GPU (cuda:0..3). +export CUDA_VISIBLE_DEVICES=${CUDA_VISIBLE_DEVICES:-0,1,2,3} +export HOTPOTQA_EMBEDDING_PER_WORKER_GPU=${HOTPOTQA_EMBEDDING_PER_WORKER_GPU:-1} +export VLLM_USE_V1=1 +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI=${MLFLOW_TRACKING_URI:-http://172.17.0.1:5000} + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/hotpotqa/base_faiss_cpu.yaml" +export HOTPOTQA_DATA_ROOT="${HOTPOTQA_DATA_ROOT:-$PROJECT_DIR/data/corpus/hotpotqa}" + +HOTPOTQA_MODEL_PATH=${HOTPOTQA_MODEL_PATH:-Qwen/Qwen3-4B-Instruct-2507} + +# Length budget (vs. Agent-R1-legacy `run_ppo_hotpotqa.sh`; semantics differ): +# - Legacy: multi-turn tokens concatenated into one trajectory → data.max_prompt_length=8192, full response=8192, +# per-turn max_response_length_single_turn=1024. +# - This script (Agent-R1 AgentFlow): each step rebuilds the prompt + one generate per step; the user block adds a +# "Recent tool / format issues" section, so we use a larger prompt budget; per-step response matches legacy 1024 +# to reduce JSON truncation at max_tokens. +HOTPOTQA_MAX_PROMPT_LEN=${HOTPOTQA_MAX_PROMPT_LEN:-10240} +HOTPOTQA_MAX_RESPONSE_LEN=${HOTPOTQA_MAX_RESPONSE_LEN:-1024} + +TRAIN_PATH="$PROJECT_DIR/data/corpus/hotpotqa/train.parquet" +VAL_PATH="$PROJECT_DIR/data/corpus/hotpotqa/validation.parquet" +HOTPOTQA_WIKI2_VAL_PATH="${HOTPOTQA_WIKI2_VAL_PATH:-$PROJECT_DIR/data/corpus/hotpotqa/2wikimultihopqa_validation.parquet}" +HOTPOTQA_MUSIQUE_VAL_PATH="${HOTPOTQA_MUSIQUE_VAL_PATH:-$PROJECT_DIR/data/corpus/hotpotqa/musique_validation.parquet}" +HOTPOTQA_ENABLE_CROSS_EVAL="${HOTPOTQA_ENABLE_CROSS_EVAL:-auto}" + +build_val_files() { + local files=("$VAL_PATH") + local include_missing=false + if [[ "$HOTPOTQA_ENABLE_CROSS_EVAL" == "1" || "$HOTPOTQA_ENABLE_CROSS_EVAL" == "true" ]]; then + include_missing=true + fi + if [[ "$HOTPOTQA_ENABLE_CROSS_EVAL" != "0" && "$HOTPOTQA_ENABLE_CROSS_EVAL" != "false" ]]; then + if [[ -f "$HOTPOTQA_WIKI2_VAL_PATH" || "$include_missing" == true ]]; then + files+=("$HOTPOTQA_WIKI2_VAL_PATH") + fi + if [[ -f "$HOTPOTQA_MUSIQUE_VAL_PATH" || "$include_missing" == true ]]; then + files+=("$HOTPOTQA_MUSIQUE_VAL_PATH") + fi + fi + + local val_files="[" + local path + for path in "${files[@]}"; do + val_files+="\"$path\"," + done + val_files="${val_files%,}]" + echo "$val_files" +} + +VAL_FILES="$(build_val_files)" + +PROJECT_NAME='HotpotQA_Agent-R1' +EXP_NAME='hotpotqa_token_level_0.99_adv_mlflow_4gpu' + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=token_gae \ + data.train_files="$TRAIN_PATH" \ + data.val_files="$VAL_FILES" \ + data.train_batch_size=256 \ + data.max_prompt_length="$HOTPOTQA_MAX_PROMPT_LEN" \ + data.max_response_length="$HOTPOTQA_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$HOTPOTQA_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=256 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=8 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=hotpotqa_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + critic.model.path="$HOTPOTQA_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=4 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + algorithm.gamma=0.99 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/hotpotqa/reward_fn.py \ + custom_reward_function.name=compute_score \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=10 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=5 "$@" diff --git a/examples/run_papersearch_grpo.sh b/examples/run_papersearch_grpo.sh new file mode 100755 index 0000000..bb0e082 --- /dev/null +++ b/examples/run_papersearch_grpo.sh @@ -0,0 +1,104 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/papersearch}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES=${CUDA_VISIBLE_DEVICES:-0,1,2,3} +export VLLM_USE_V1=1 +export HF_ENDPOINT=${HF_ENDPOINT:-https://hf-mirror.com} +export CUDA_HOME=${CUDA_HOME:-/usr/local/cuda} +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI=${MLFLOW_TRACKING_URI:-http://127.0.0.1:8080} +export PAPER_SEARCH_BASE_URL=${PAPER_SEARCH_BASE_URL:-http://localhost:4000} +export PAPERSEARCH_SELECTOR_BASE_URL=${PAPERSEARCH_SELECTOR_BASE_URL:-http://localhost:8000} + +AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GRPO_ROLLOUT_N:-8}" +# Match run_papersearch_token_adv.sh (128 unique tasks): train_batch_size * rollout.n ~= 128. +PAPERSEARCH_GRPO_BASE_TRAIN_BATCH="${PAPERSEARCH_GRPO_BASE_TRAIN_BATCH:-128}" +PAPERSEARCH_GRPO_BASE_LOG_PROB_MICRO_BATCH="${PAPERSEARCH_GRPO_BASE_LOG_PROB_MICRO_BATCH:-32}" +PAPERSEARCH_TRAIN_BATCH_SIZE="$((PAPERSEARCH_GRPO_BASE_TRAIN_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +PAPERSEARCH_LOG_PROB_MICRO_BATCH="$((PAPERSEARCH_GRPO_BASE_LOG_PROB_MICRO_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +if [[ "$PAPERSEARCH_TRAIN_BATCH_SIZE" -lt 1 ]]; then + echo "❌ PAPERSEARCH_GRPO_BASE_TRAIN_BATCH ($PAPERSEARCH_GRPO_BASE_TRAIN_BATCH) must be >= AGENT_R1_GRPO_ROLLOUT_N ($AGENT_R1_GRPO_ROLLOUT_N)." >&2 + exit 1 +fi +if [[ "$PAPERSEARCH_LOG_PROB_MICRO_BATCH" -lt 1 ]]; then + PAPERSEARCH_LOG_PROB_MICRO_BATCH=1 +fi + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/paper_search/base.yaml" + +PAPERSEARCH_MODEL_PATH="Qwen/Qwen3-4B-Instruct-2507" +PAPERSEARCH_MAX_PROMPT_LEN=${PAPERSEARCH_MAX_PROMPT_LEN:-10240} +PAPERSEARCH_MAX_RESPONSE_LEN=${PAPERSEARCH_MAX_RESPONSE_LEN:-4096} +PAPERSEARCH_TRAIN_PATH="${PAPERSEARCH_TRAIN_PATH:-$PROJECT_DIR/data/pasa/train.parquet}" +PAPERSEARCH_VAL_PATH="${PAPERSEARCH_VAL_PATH:-$PROJECT_DIR/data/pasa/test.parquet}" +export PAPERSEARCH_SELECTOR_MODEL_NAME=${PAPERSEARCH_SELECTOR_MODEL_NAME:-selector-qwen-8b} + +PROJECT_NAME=${PROJECT_NAME:-FALCON} +EXP_NAME=${EXP_NAME:-papersearch_grpo} + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=grpo \ + algorithm.norm_adv_by_std_in_grpo="${AGENT_R1_NORM_ADV_BY_STD_IN_GRPO:-True}" \ + data.train_files="$PAPERSEARCH_TRAIN_PATH" \ + data.val_files="$PAPERSEARCH_VAL_PATH" \ + data.train_batch_size="$PAPERSEARCH_TRAIN_BATCH_SIZE" \ + data.max_prompt_length="$PAPERSEARCH_MAX_PROMPT_LEN" \ + data.max_response_length="$PAPERSEARCH_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$PAPERSEARCH_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size="$PAPERSEARCH_TRAIN_BATCH_SIZE" \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu="$PAPERSEARCH_LOG_PROB_MICRO_BATCH" \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.n="$AGENT_R1_GRPO_ROLLOUT_N" \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu="$PAPERSEARCH_LOG_PROB_MICRO_BATCH" \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=paper_search_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + critic.enable=False \ + algorithm.use_kl_in_reward=False \ + reward_model.enable=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=20 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.total_epochs=20 "$@" diff --git a/examples/run_papersearch_step_adv.sh b/examples/run_papersearch_step_adv.sh new file mode 100755 index 0000000..0cb3a7b --- /dev/null +++ b/examples/run_papersearch_step_adv.sh @@ -0,0 +1,96 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/papersearch}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES=${CUDA_VISIBLE_DEVICES:-0,3,4,7} +export VLLM_USE_V1=1 +export HF_ENDPOINT=${HF_ENDPOINT:-https://hf-mirror.com} +export CUDA_HOME=${CUDA_HOME:-/usr/local/cuda} +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI=${MLFLOW_TRACKING_URI:-http://127.0.0.1:5000} +export PAPER_SEARCH_BASE_URL=${PAPER_SEARCH_BASE_URL:-http://localhost:4000} +export PAPERSEARCH_SELECTOR_BASE_URL=${PAPERSEARCH_SELECTOR_BASE_URL:-http://localhost:8000} + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/paper_search/base.yaml" + +PAPERSEARCH_MODEL_PATH="Qwen/Qwen3-4B-Instruct-2507" +PAPERSEARCH_MAX_PROMPT_LEN=${PAPERSEARCH_MAX_PROMPT_LEN:-10240} +PAPERSEARCH_MAX_RESPONSE_LEN=${PAPERSEARCH_MAX_RESPONSE_LEN:-4096} +PAPERSEARCH_TRAIN_PATH="${PAPERSEARCH_TRAIN_PATH:-$PROJECT_DIR/data/pasa/train.parquet}" +PAPERSEARCH_VAL_PATH="${PAPERSEARCH_VAL_PATH:-$PROJECT_DIR/data/pasa/test.parquet}" +export PAPERSEARCH_SELECTOR_MODEL_NAME=${PAPERSEARCH_SELECTOR_MODEL_NAME:-selector-qwen-8b} + +PROJECT_NAME=${PROJECT_NAME:-Agent-R1} +EXP_NAME=${EXP_NAME:-papersearch_step_adv_mlflow_4gpu} + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=gae \ + data.train_files="$PAPERSEARCH_TRAIN_PATH" \ + data.val_files="$PAPERSEARCH_VAL_PATH" \ + data.train_batch_size=128 \ + data.max_prompt_length="$PAPERSEARCH_MAX_PROMPT_LEN" \ + data.max_response_length="$PAPERSEARCH_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$PAPERSEARCH_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=128 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.5 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=paper_search_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + critic.model.path="$PAPERSEARCH_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=16 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + reward_model.enable=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=20 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=20 "$@" diff --git a/examples/run_papersearch_steppo.sh b/examples/run_papersearch_steppo.sh new file mode 100755 index 0000000..a9e174d --- /dev/null +++ b/examples/run_papersearch_steppo.sh @@ -0,0 +1,6 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_papersearch_step_adv.sh" "$@" diff --git a/examples/run_papersearch_token_adv.sh b/examples/run_papersearch_token_adv.sh new file mode 100755 index 0000000..d3ff657 --- /dev/null +++ b/examples/run_papersearch_token_adv.sh @@ -0,0 +1,95 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/papersearch}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES=${CUDA_VISIBLE_DEVICES:-0,1,2,3} +export VLLM_USE_V1=1 +export HF_ENDPOINT=${HF_ENDPOINT:-https://hf-mirror.com} +export CUDA_HOME=${CUDA_HOME:-/usr/local/cuda} +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI=${MLFLOW_TRACKING_URI:-http://127.0.0.1:8080} +export PAPER_SEARCH_BASE_URL=${PAPER_SEARCH_BASE_URL:-http://localhost:4000} +export PAPERSEARCH_SELECTOR_BASE_URL=${PAPERSEARCH_SELECTOR_BASE_URL:-http://localhost:8000} + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/paper_search/base.yaml" + +PAPERSEARCH_MODEL_PATH="Qwen/Qwen3-4B-Instruct-2507" +PAPERSEARCH_MAX_PROMPT_LEN=${PAPERSEARCH_MAX_PROMPT_LEN:-10240} +PAPERSEARCH_MAX_RESPONSE_LEN=${PAPERSEARCH_MAX_RESPONSE_LEN:-4096} +PAPERSEARCH_TRAIN_PATH="${PAPERSEARCH_TRAIN_PATH:-$PROJECT_DIR/data/pasa/train.parquet}" +PAPERSEARCH_VAL_PATH="${PAPERSEARCH_VAL_PATH:-$PROJECT_DIR/data/pasa/test.parquet}" +export PAPERSEARCH_SELECTOR_MODEL_NAME=${PAPERSEARCH_SELECTOR_MODEL_NAME:-selector-qwen-8b} + +PROJECT_NAME=${PROJECT_NAME:-Agent-R1} +EXP_NAME=${EXP_NAME:-papersearch_token_adv_mlflow_4gpu} + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=token_gae \ + data.train_files="$PAPERSEARCH_TRAIN_PATH" \ + data.val_files="$PAPERSEARCH_VAL_PATH" \ + data.train_batch_size=128 \ + data.max_prompt_length="$PAPERSEARCH_MAX_PROMPT_LEN" \ + data.max_response_length="$PAPERSEARCH_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$PAPERSEARCH_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=128 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=paper_search_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + critic.model.path="$PAPERSEARCH_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=16 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + reward_model.enable=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=20 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=20 "$@" diff --git a/examples/run_qwen3-4b_gsm8k_tool_ppo.sh b/examples/run_qwen3-4b_gsm8k_tool_ppo.sh new file mode 100755 index 0000000..8948a92 --- /dev/null +++ b/examples/run_qwen3-4b_gsm8k_tool_ppo.sh @@ -0,0 +1,10 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_qwen3-4b_gsm8k_tool.sh" \ + algorithm.adv_estimator=gae \ + actor_rollout_ref.rollout.n="${AGENT_R1_PPO_ROLLOUT_N:-1}" \ + trainer.experiment_name="${AGENT_R1_EXP_NAME:-qwen3_4b_gsm8k_tool_ppo}" \ + "$@" diff --git a/examples/run_qwen3-4b_gsm8k_tool_reinforce_plus_plus.sh b/examples/run_qwen3-4b_gsm8k_tool_reinforce_plus_plus.sh new file mode 100755 index 0000000..0311d23 --- /dev/null +++ b/examples/run_qwen3-4b_gsm8k_tool_reinforce_plus_plus.sh @@ -0,0 +1,15 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_qwen3-4b_gsm8k_tool.sh" \ + algorithm.adv_estimator=reinforce_plus_plus \ + actor_rollout_ref.rollout.n="${AGENT_R1_REINFORCE_PLUS_PLUS_ROLLOUT_N:-5}" \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_COEF:-0.001}" \ + trainer.experiment_name="${AGENT_R1_EXP_NAME:-qwen3_4b_gsm8k_tool_reinforce_plus_plus}" \ + "$@" diff --git a/examples/run_qwen3-4b_gsm8k_tool_reinforce_plus_plus_baseline.sh b/examples/run_qwen3-4b_gsm8k_tool_reinforce_plus_plus_baseline.sh new file mode 100755 index 0000000..6385deb --- /dev/null +++ b/examples/run_qwen3-4b_gsm8k_tool_reinforce_plus_plus_baseline.sh @@ -0,0 +1,15 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_qwen3-4b_gsm8k_tool.sh" \ + algorithm.adv_estimator=reinforce_plus_plus_baseline \ + actor_rollout_ref.rollout.n="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_ROLLOUT_N:-5}" \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_COEF:-0.001}" \ + trainer.experiment_name="${AGENT_R1_EXP_NAME:-qwen3_4b_gsm8k_tool_reinforce_plus_plus_baseline}" \ + "$@" diff --git a/examples/run_qwen3-4b_gsm8k_tool_rloo.sh b/examples/run_qwen3-4b_gsm8k_tool_rloo.sh new file mode 100755 index 0000000..b9f546e --- /dev/null +++ b/examples/run_qwen3-4b_gsm8k_tool_rloo.sh @@ -0,0 +1,14 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_qwen3-4b_gsm8k_tool.sh" \ + algorithm.adv_estimator=rloo \ + actor_rollout_ref.rollout.n="${AGENT_R1_RLOO_ROLLOUT_N:-5}" \ + actor_rollout_ref.actor.use_kl_loss=False \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_RLOO_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_RLOO_KL_COEF:-0.001}" \ + trainer.experiment_name="${AGENT_R1_EXP_NAME:-qwen3_4b_gsm8k_tool_rloo}" \ + "$@" diff --git a/examples/run_qwen3-4b_gsm8k_tool_steppo.sh b/examples/run_qwen3-4b_gsm8k_tool_steppo.sh new file mode 100755 index 0000000..9d2e035 --- /dev/null +++ b/examples/run_qwen3-4b_gsm8k_tool_steppo.sh @@ -0,0 +1,14 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_qwen3-4b_gsm8k_tool.sh" \ + algorithm.adv_estimator=gae \ + algorithm.gamma="${AGENT_R1_STEPPO_GAMMA:-0.99}" \ + actor_rollout_ref.rollout.n="${AGENT_R1_STEPPO_ROLLOUT_N:-1}" \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + trainer.experiment_name="${AGENT_R1_EXP_NAME:-qwen3_4b_gsm8k_tool_steppo}" \ + "$@" diff --git a/examples/run_qwen3-4b_gsm8k_tool_token_gae.sh b/examples/run_qwen3-4b_gsm8k_tool_token_gae.sh new file mode 100755 index 0000000..ca419c7 --- /dev/null +++ b/examples/run_qwen3-4b_gsm8k_tool_token_gae.sh @@ -0,0 +1,10 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_qwen3-4b_gsm8k_tool.sh" \ + algorithm.adv_estimator=token_gae \ + actor_rollout_ref.rollout.n="${AGENT_R1_TOKEN_GAE_ROLLOUT_N:-1}" \ + trainer.experiment_name="${AGENT_R1_EXP_NAME:-qwen3_4b_gsm8k_tool_token_gae}" \ + "$@" diff --git a/examples/run_webshop_grpo.sh b/examples/run_webshop_grpo.sh new file mode 100755 index 0000000..bfc68cd --- /dev/null +++ b/examples/run_webshop_grpo.sh @@ -0,0 +1,107 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/webshop}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES="${CUDA_VISIBLE_DEVICES:-0,1,2,3}" +export VLLM_USE_V1=1 +export HF_ENDPOINT="${HF_ENDPOINT:-https://hf-mirror.com}" +export CUDA_HOME="${CUDA_HOME:-/usr/local/cuda}" +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI="${MLFLOW_TRACKING_URI:-http://127.0.0.1:5000}" +export WEBSHOP_ENV_BASE_URL="${WEBSHOP_ENV_BASE_URL:-http://127.0.0.1:4111}" + +AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GRPO_ROLLOUT_N:-8}" +# Match run_webshop_token_adv.sh (128 unique tasks): train_batch_size * rollout.n ~= 128. +WEBSHOP_GRPO_BASE_TRAIN_BATCH="${WEBSHOP_GRPO_BASE_TRAIN_BATCH:-128}" +WEBSHOP_GRPO_BASE_LOG_PROB_MICRO_BATCH="${WEBSHOP_GRPO_BASE_LOG_PROB_MICRO_BATCH:-32}" +WEBSHOP_TRAIN_BATCH_SIZE="$((WEBSHOP_GRPO_BASE_TRAIN_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +WEBSHOP_LOG_PROB_MICRO_BATCH="$((WEBSHOP_GRPO_BASE_LOG_PROB_MICRO_BATCH / AGENT_R1_GRPO_ROLLOUT_N))" +if [[ "$WEBSHOP_TRAIN_BATCH_SIZE" -lt 1 ]]; then + echo "❌ WEBSHOP_GRPO_BASE_TRAIN_BATCH ($WEBSHOP_GRPO_BASE_TRAIN_BATCH) must be >= AGENT_R1_GRPO_ROLLOUT_N ($AGENT_R1_GRPO_ROLLOUT_N)." >&2 + exit 1 +fi +if [[ "$WEBSHOP_LOG_PROB_MICRO_BATCH" -lt 1 ]]; then + WEBSHOP_LOG_PROB_MICRO_BATCH=1 +fi + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/webshop/base.yaml" + +WEBSHOP_MODEL_PATH="Qwen/Qwen3-4B-Instruct-2507" +WEBSHOP_MAX_PROMPT_LEN="${WEBSHOP_MAX_PROMPT_LEN:-8192}" +WEBSHOP_MAX_RESPONSE_LEN="${WEBSHOP_MAX_RESPONSE_LEN:-4096}" +WEBSHOP_DATA_ROOT="${WEBSHOP_DATA_ROOT:-$PROJECT_DIR/data/webshop_full}" +WEBSHOP_TRAIN_PATH="${WEBSHOP_TRAIN_PATH:-$WEBSHOP_DATA_ROOT/train.parquet}" +WEBSHOP_VAL_PATH="${WEBSHOP_VAL_PATH:-$WEBSHOP_DATA_ROOT/test.parquet}" +VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$PROJECT_DIR/outputs/webshop_validation/grpo}" + +PROJECT_NAME="${PROJECT_NAME:-WebShop_Agent-R1}" +EXP_NAME="${EXP_NAME:-webshop_grpo}" + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=grpo \ + algorithm.norm_adv_by_std_in_grpo="${AGENT_R1_NORM_ADV_BY_STD_IN_GRPO:-True}" \ + data.train_files="$WEBSHOP_TRAIN_PATH" \ + data.val_files="$WEBSHOP_VAL_PATH" \ + data.train_batch_size="$WEBSHOP_TRAIN_BATCH_SIZE" \ + data.max_prompt_length="$WEBSHOP_MAX_PROMPT_LEN" \ + data.max_response_length="$WEBSHOP_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$WEBSHOP_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size="$WEBSHOP_TRAIN_BATCH_SIZE" \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu="$WEBSHOP_LOG_PROB_MICRO_BATCH" \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.n="$AGENT_R1_GRPO_ROLLOUT_N" \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu="$WEBSHOP_LOG_PROB_MICRO_BATCH" \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=webshop_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/webshop/reward_fn.py \ + custom_reward_function.name=compute_score \ + critic.enable=False \ + algorithm.use_kl_in_reward=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.validation_data_dir="$VAL_DUMP_DIR" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=5 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.total_epochs=10 "$@" diff --git a/examples/run_webshop_step_adv.sh b/examples/run_webshop_step_adv.sh new file mode 100755 index 0000000..238bc25 --- /dev/null +++ b/examples/run_webshop_step_adv.sh @@ -0,0 +1,99 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/webshop}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES="${CUDA_VISIBLE_DEVICES:-0,1,2,3}" +export VLLM_USE_V1=1 +export HF_ENDPOINT="${HF_ENDPOINT:-https://hf-mirror.com}" +export CUDA_HOME="${CUDA_HOME:-/usr/local/cuda}" +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI="${MLFLOW_TRACKING_URI:-http://127.0.0.1:5000}" +export WEBSHOP_ENV_BASE_URL="${WEBSHOP_ENV_BASE_URL:-http://127.0.0.1:4111}" + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/webshop/base.yaml" + +WEBSHOP_MODEL_PATH="Qwen/Qwen3-4B-Instruct-2507" +WEBSHOP_MAX_PROMPT_LEN="${WEBSHOP_MAX_PROMPT_LEN:-8192}" +WEBSHOP_MAX_RESPONSE_LEN="${WEBSHOP_MAX_RESPONSE_LEN:-4096}" +WEBSHOP_DATA_ROOT="${WEBSHOP_DATA_ROOT:-$PROJECT_DIR/data/webshop_full}" +WEBSHOP_TRAIN_PATH="${WEBSHOP_TRAIN_PATH:-$WEBSHOP_DATA_ROOT/train.parquet}" +WEBSHOP_VAL_PATH="${WEBSHOP_VAL_PATH:-$WEBSHOP_DATA_ROOT/test.parquet}" +VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$PROJECT_DIR/outputs/webshop_validation/step_adv}" + +PROJECT_NAME="${PROJECT_NAME:-WebShop_Agent-R1}" +EXP_NAME="${EXP_NAME:-webshop_step_adv_mlflow_4gpu}" + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=gae \ + data.train_files="$WEBSHOP_TRAIN_PATH" \ + data.val_files="$WEBSHOP_VAL_PATH" \ + data.train_batch_size=128 \ + data.max_prompt_length="$WEBSHOP_MAX_PROMPT_LEN" \ + data.max_response_length="$WEBSHOP_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$WEBSHOP_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=128 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=webshop_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/webshop/reward_fn.py \ + custom_reward_function.name=compute_score \ + critic.model.path="$WEBSHOP_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=16 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.validation_data_dir="$VAL_DUMP_DIR" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=5 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=10 "$@" diff --git a/examples/run_webshop_steppo.sh b/examples/run_webshop_steppo.sh new file mode 100755 index 0000000..6099692 --- /dev/null +++ b/examples/run_webshop_steppo.sh @@ -0,0 +1,6 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" + +exec bash "$SCRIPT_DIR/run_webshop_step_adv.sh" "$@" diff --git a/examples/run_webshop_token_adv.sh b/examples/run_webshop_token_adv.sh new file mode 100755 index 0000000..b72c3d5 --- /dev/null +++ b/examples/run_webshop_token_adv.sh @@ -0,0 +1,98 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/webshop}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES="${CUDA_VISIBLE_DEVICES:-0,1,2,3}" +export VLLM_USE_V1=1 +export HF_ENDPOINT="${HF_ENDPOINT:-https://hf-mirror.com}" +export CUDA_HOME="${CUDA_HOME:-/usr/local/cuda}" +export HYDRA_FULL_ERROR=1 +export MLFLOW_TRACKING_URI="${MLFLOW_TRACKING_URI:-http://127.0.0.1:5000}" +export WEBSHOP_ENV_BASE_URL="${WEBSHOP_ENV_BASE_URL:-http://127.0.0.1:4111}" + +PROJECT_DIR="$(pwd)" +CONFIG_PATH="$PROJECT_DIR/recipe/webshop/base.yaml" + +WEBSHOP_MODEL_PATH="Qwen/Qwen3-4B-Instruct-2507" +WEBSHOP_MAX_PROMPT_LEN="${WEBSHOP_MAX_PROMPT_LEN:-8192}" +WEBSHOP_MAX_RESPONSE_LEN="${WEBSHOP_MAX_RESPONSE_LEN:-4096}" +WEBSHOP_DATA_ROOT="${WEBSHOP_DATA_ROOT:-$PROJECT_DIR/data/webshop_full}" +WEBSHOP_TRAIN_PATH="${WEBSHOP_TRAIN_PATH:-$WEBSHOP_DATA_ROOT/train.parquet}" +WEBSHOP_VAL_PATH="${WEBSHOP_VAL_PATH:-$WEBSHOP_DATA_ROOT/test.parquet}" +VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$PROJECT_DIR/outputs/webshop_validation/token_adv}" + +PROJECT_NAME="${PROJECT_NAME:-WebShop_Agent-R1}" +EXP_NAME="${EXP_NAME:-webshop_token_adv_mlflow_4gpu}" + +python3 -m agent_r1.trainer.main_agent_ppo \ + algorithm.adv_estimator=token_gae \ + data.train_files="$WEBSHOP_TRAIN_PATH" \ + data.val_files="$WEBSHOP_VAL_PATH" \ + data.train_batch_size=128 \ + data.max_prompt_length="$WEBSHOP_MAX_PROMPT_LEN" \ + data.max_response_length="$WEBSHOP_MAX_RESPONSE_LEN" \ + data.filter_overlong_prompts=True \ + data.truncation='error' \ + data.return_raw_chat=True \ + actor_rollout_ref.model.path="$WEBSHOP_MODEL_PATH" \ + actor_rollout_ref.actor.optim.lr=1e-6 \ + actor_rollout_ref.model.use_remove_padding=True \ + actor_rollout_ref.actor.ppo_mini_batch_size=128 \ + actor_rollout_ref.actor.ppo_micro_batch_size_per_gpu=4 \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef=0.001 \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + actor_rollout_ref.actor.clip_ratio_low=3e-4 \ + actor_rollout_ref.actor.clip_ratio_high=4e-4 \ + actor_rollout_ref.actor.clip_ratio_c=10.0 \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.entropy_coeff=0 \ + actor_rollout_ref.model.enable_gradient_checkpointing=True \ + actor_rollout_ref.actor.fsdp_config.param_offload=True \ + actor_rollout_ref.actor.fsdp_config.optimizer_offload=True \ + actor_rollout_ref.rollout.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.rollout.tensor_model_parallel_size=1 \ + actor_rollout_ref.rollout.name=vllm \ + actor_rollout_ref.rollout.gpu_memory_utilization=0.7 \ + actor_rollout_ref.rollout.agent.agent_flow_config_path="$CONFIG_PATH" \ + actor_rollout_ref.ref.log_prob_micro_batch_size_per_gpu=32 \ + actor_rollout_ref.ref.fsdp_config.param_offload=True \ + actor_rollout_ref.rollout.agent.num_workers=4 \ + actor_rollout_ref.rollout.agent.default_agent_flow=webshop_agent \ + actor_rollout_ref.rollout.trace.backend=mlflow \ + actor_rollout_ref.rollout.trace.token2text=True \ + actor_rollout_ref.rollout.trace.max_samples_per_step_per_worker=5 \ + reward_model.enable=False \ + custom_reward_function.path=recipe/webshop/reward_fn.py \ + custom_reward_function.name=compute_score \ + critic.model.path="$WEBSHOP_MODEL_PATH" \ + critic.optim.lr=1e-5 \ + critic.model.use_remove_padding=True \ + critic.model.enable_gradient_checkpointing=True \ + critic.ppo_micro_batch_size_per_gpu=16 \ + critic.model.fsdp_config.param_offload=True \ + critic.model.fsdp_config.optimizer_offload=True \ + algorithm.use_kl_in_reward=False \ + trainer.critic_warmup=0 \ + trainer.logger='["console","swanlab","mlflow"]' \ + trainer.project_name="$PROJECT_NAME" \ + trainer.experiment_name="$EXP_NAME" \ + trainer.validation_data_dir="$VAL_DUMP_DIR" \ + trainer.n_gpus_per_node=4 \ + trainer.nnodes=1 \ + trainer.val_before_train=True \ + trainer.save_freq=100 \ + trainer.test_freq=5 \ + trainer.max_actor_ckpt_to_keep=3 \ + trainer.max_critic_ckpt_to_keep=3 \ + trainer.total_epochs=10 "$@" diff --git a/examples/webshop/run_gigpo.sh b/examples/webshop/run_gigpo.sh new file mode 100755 index 0000000..2c58c67 --- /dev/null +++ b/examples/webshop/run_gigpo.sh @@ -0,0 +1,21 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-webshop_gigpo}" +export WEBSHOP_VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$ROOT_DIR/outputs/webshop_validation/gigpo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GIGPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_webshop_grpo.sh" \ + algorithm.adv_estimator=gigpo \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GIGPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + ++algorithm.gigpo.step_advantage_w="${AGENT_R1_GIGPO_STEP_ADVANTAGE_W:-1.0}" \ + ++algorithm.gigpo.mode="${AGENT_R1_GIGPO_MODE:-mean_std_norm}" \ + ++algorithm.gigpo.enable_similarity="${AGENT_R1_GIGPO_ENABLE_SIMILARITY:-False}" \ + ++algorithm.gigpo.similarity_thresh="${AGENT_R1_GIGPO_SIMILARITY_THRESH:-0.95}" \ + "$@" diff --git a/examples/webshop/run_grpo.sh b/examples/webshop/run_grpo.sh new file mode 100755 index 0000000..3f2ab64 --- /dev/null +++ b/examples/webshop/run_grpo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_webshop_grpo.sh" "$@" diff --git a/examples/webshop/run_gspo.sh b/examples/webshop/run_gspo.sh new file mode 100755 index 0000000..90417f1 --- /dev/null +++ b/examples/webshop/run_gspo.sh @@ -0,0 +1,19 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-webshop_gspo}" +export WEBSHOP_VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$ROOT_DIR/outputs/webshop_validation/gspo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_GSPO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_webshop_grpo.sh" \ + algorithm.adv_estimator=grpo \ + actor_rollout_ref.actor.policy_loss.loss_mode=gspo \ + actor_rollout_ref.actor.loss_agg_mode=seq-mean-token-mean \ + actor_rollout_ref.actor.use_kl_loss=True \ + actor_rollout_ref.actor.kl_loss_coef="${AGENT_R1_GSPO_KL_COEF:-0.001}" \ + actor_rollout_ref.actor.kl_loss_type=low_var_kl \ + algorithm.use_kl_in_reward=False \ + "$@" diff --git a/examples/webshop/run_reinforce_plus_plus.sh b/examples/webshop/run_reinforce_plus_plus.sh new file mode 100755 index 0000000..8b4d575 --- /dev/null +++ b/examples/webshop/run_reinforce_plus_plus.sh @@ -0,0 +1,18 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-webshop_reinforce_plus_plus}" +export WEBSHOP_VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$ROOT_DIR/outputs/webshop_validation/reinforce_plus_plus}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_webshop_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/webshop/run_reinforce_plus_plus_baseline.sh b/examples/webshop/run_reinforce_plus_plus_baseline.sh new file mode 100755 index 0000000..79f9b48 --- /dev/null +++ b/examples/webshop/run_reinforce_plus_plus_baseline.sh @@ -0,0 +1,18 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-webshop_reinforce_plus_plus_baseline}" +export WEBSHOP_VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$ROOT_DIR/outputs/webshop_validation/reinforce_plus_plus_baseline}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_webshop_grpo.sh" \ + algorithm.adv_estimator=reinforce_plus_plus_baseline \ + actor_rollout_ref.actor.use_kl_loss=False \ + actor_rollout_ref.actor.kl_loss_type=mse \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_REINFORCE_PLUS_PLUS_BASELINE_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/webshop/run_rloo.sh b/examples/webshop/run_rloo.sh new file mode 100755 index 0000000..dda3a57 --- /dev/null +++ b/examples/webshop/run_rloo.sh @@ -0,0 +1,17 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +export EXP_NAME="${EXP_NAME:-webshop_rloo}" +export WEBSHOP_VAL_DUMP_DIR="${WEBSHOP_VAL_DUMP_DIR:-$ROOT_DIR/outputs/webshop_validation/rloo}" +export AGENT_R1_GRPO_ROLLOUT_N="${AGENT_R1_RLOO_ROLLOUT_N:-${AGENT_R1_GRPO_ROLLOUT_N:-8}}" + +exec bash "$ROOT_DIR/examples/run_webshop_grpo.sh" \ + algorithm.adv_estimator=rloo \ + actor_rollout_ref.actor.use_kl_loss=False \ + algorithm.use_kl_in_reward=True \ + algorithm.kl_penalty="${AGENT_R1_RLOO_KL_PENALTY:-kl}" \ + algorithm.kl_ctrl.kl_coef="${AGENT_R1_RLOO_KL_COEF:-0.001}" \ + "$@" diff --git a/examples/webshop/run_step_adv.sh b/examples/webshop/run_step_adv.sh new file mode 100755 index 0000000..2a6d154 --- /dev/null +++ b/examples/webshop/run_step_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_webshop_step_adv.sh" "$@" diff --git a/examples/webshop/run_steppo.sh b/examples/webshop/run_steppo.sh new file mode 100755 index 0000000..6e34aa6 --- /dev/null +++ b/examples/webshop/run_steppo.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_webshop_steppo.sh" "$@" diff --git a/examples/webshop/run_token_adv.sh b/examples/webshop/run_token_adv.sh new file mode 100755 index 0000000..3c878e5 --- /dev/null +++ b/examples/webshop/run_token_adv.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" +ROOT_DIR="$(cd "$SCRIPT_DIR/../.." && pwd)" + +exec bash "$ROOT_DIR/examples/run_webshop_token_adv.sh" "$@" diff --git a/image/agent-r1-datasets.png b/image/agent-r1-datasets.png new file mode 100644 index 0000000..30436f3 Binary files /dev/null and b/image/agent-r1-datasets.png differ diff --git a/image/agent-r1-gsm8k-context.png b/image/agent-r1-gsm8k-context.png new file mode 100644 index 0000000..a9bc107 Binary files /dev/null and b/image/agent-r1-gsm8k-context.png differ diff --git a/image/agent-r1-gsm8k.png b/image/agent-r1-gsm8k.png new file mode 100644 index 0000000..ba2c0db Binary files /dev/null and b/image/agent-r1-gsm8k.png differ diff --git a/image/framework.png b/image/framework.png index e097ffd..7e35185 100644 Binary files a/image/framework.png and b/image/framework.png differ diff --git a/image/step-level-mdp.png b/image/step-level-mdp.png new file mode 100644 index 0000000..f1b41ec Binary files /dev/null and b/image/step-level-mdp.png differ diff --git a/mkdocs.yml b/mkdocs.yml index 56e0cf4..f0ff1cd 100644 --- a/mkdocs.yml +++ b/mkdocs.yml @@ -48,6 +48,7 @@ nav: - Tutorials: - tutorials/index.md - Agent Task Tutorial: tutorials/agent-task.md + - Datasets and Algorithms: tutorials/datasets-and-algorithms.md plugins: - search diff --git a/recipe/__init__.py b/recipe/__init__.py new file mode 100644 index 0000000..75aa618 --- /dev/null +++ b/recipe/__init__.py @@ -0,0 +1 @@ +"""Agent-R1 task recipes.""" diff --git a/recipe/alfworld/__init__.py b/recipe/alfworld/__init__.py new file mode 100644 index 0000000..fa92376 --- /dev/null +++ b/recipe/alfworld/__init__.py @@ -0,0 +1 @@ +"""ALFWorld recipe.""" diff --git a/recipe/alfworld/alfworld_agent_flow.py b/recipe/alfworld/alfworld_agent_flow.py new file mode 100644 index 0000000..9429a92 --- /dev/null +++ b/recipe/alfworld/alfworld_agent_flow.py @@ -0,0 +1,233 @@ +from __future__ import annotations + +import json +import logging +import os +import re +from typing import Any +from uuid import uuid4 + +from transformers import AutoProcessor, AutoTokenizer + +from agent_r1.agent_flow.agent_flow import AgentFlowBase, AgentFlowOutput, AgentFlowStep, register +from agent_r1.reward_loop.reward_loop import RewardLoopWorker +from recipe.alfworld.prompts import ALFWORLD_TOOL_SCHEMAS +from recipe.alfworld.utils import ( + INVALID_TOOL_CALL_ACTION, + AlfworldToolExecutor, + build_alfworld_messages, + build_invalid_tool_call_observation, + extract_task_text, +) +from verl.experimental.agent_loop.agent_loop import AsyncLLMServerManager, DictConfigWrap +from verl.experimental.agent_loop.tool_parser import FunctionCall, ToolParser +from verl.utils.profiler import simple_timer + +logger = logging.getLogger(__file__) +logger.setLevel(os.getenv("VERL_LOGGING_LEVEL", "WARN")) + +_TOOL_CALL_BLOCK = re.compile(r"(.*?)", re.DOTALL) + + +def _recover_tool_calls_from_text(text: str) -> list[FunctionCall]: + recovered: list[FunctionCall] = [] + for raw in _TOOL_CALL_BLOCK.findall(text): + try: + payload = json.loads(raw.strip()) + except Exception: + continue + if not isinstance(payload, dict): + continue + name = payload.get("name") + arguments = payload.get("arguments") + if not isinstance(name, str): + continue + if isinstance(arguments, str): + try: + arguments = json.loads(arguments) + except Exception: + continue + if not isinstance(arguments, dict): + continue + recovered.append( + FunctionCall( + name=name, + arguments=json.dumps(arguments, ensure_ascii=False), + ) + ) + return recovered + + +@register("alfworld_agent") +class AlfworldAgentFlow(AgentFlowBase): + def __init__( + self, + trainer_config: DictConfigWrap, + server_manager: AsyncLLMServerManager, + reward_loop_worker: RewardLoopWorker, + tokenizer: AutoTokenizer, + processor: AutoProcessor, + **kwargs, + ): + super().__init__(trainer_config, server_manager, reward_loop_worker, tokenizer, processor, **kwargs) + self.max_steps = kwargs.get("max_steps", 20) + self.max_parallel_calls = 1 + self.max_episode_steps = kwargs.get("max_episode_steps", 50) + + self.tool_parser = ToolParser.get_tool_parser( + self.config.actor_rollout_ref.rollout.multi_turn.format, + self.tokenizer, + ) + self.prompt_length = self.config.actor_rollout_ref.rollout.prompt_length + self.response_length = self.config.actor_rollout_ref.rollout.response_length + self.tool_schemas = ALFWORLD_TOOL_SCHEMAS + + self.executor = AlfworldToolExecutor(max_episode_steps=self.max_episode_steps) + self.current_observation: str = "" + self.current_admissible_commands: list[str] = [] + self.history_actions: list[str] = [] + self.steps: list[AgentFlowStep] = [] + + async def run(self, sampling_params: dict[str, Any], **kwargs) -> AgentFlowOutput: + extra_info = kwargs.get("extra_info") or {} + task_id = extra_info.get("task_id") + split = extra_info.get("split", "train") + task_type_raw = extra_info.get("task_type_raw") + task_family = extra_info.get("task_family") + game_relative_path = extra_info.get("game_relative_path") + if not game_relative_path: + raise ValueError("ALFWorld sample is missing extra_info.game_relative_path") + + self.current_observation, reset_info = self.executor.reset_with_info( + game_relative_path=game_relative_path, + task_id=task_id, + ) + admissible_commands = reset_info.get("admissible_commands") + self.current_admissible_commands = admissible_commands if isinstance(admissible_commands, list) else [] + task_text = extract_task_text(self.current_observation, extra_info.get("goal_text")) + self.history_actions = [] + self.steps = [] + + metrics: dict[str, Any] = {} + num_steps = 0 + done = False + final_success_flag: bool | None = None + dense_reward_sum = 0.0 + + def build_reward_extra_info(step_env_reward: float = 0.0) -> dict[str, Any]: + return { + "score": 0.0, + "step_env_reward": float(step_env_reward), + "dense_reward_sum": float(dense_reward_sum), + "success": bool(final_success_flag), + "num_steps": int(num_steps), + "task_id": str(task_id or ""), + "split": str(split or ""), + "task_type_raw": str(task_type_raw or ""), + "task_family": str(task_family or ""), + "is_action_valid": False, + } + + while num_steps < self.max_steps and not done: + num_steps += 1 + observation_before_action = self.current_observation + + messages = build_alfworld_messages( + task_text=task_text, + observation=observation_before_action, + history_actions=self.history_actions, + admissible_commands=self.current_admissible_commands, + ) + + prompt_ids = await self.apply_chat_template(messages, tools=self.tool_schemas) + + with simple_timer("generate_sequences", metrics): + output = await self.server_manager.generate( + request_id=uuid4().hex, + prompt_ids=prompt_ids, + sampling_params=sampling_params, + ) + + response_ids = output.token_ids[: self.response_length] + _, tool_calls = await self.tool_parser.extract_tool_calls(response_ids) + if not tool_calls: + response_text = self.tokenizer.decode(response_ids, skip_special_tokens=True) + tool_calls = _recover_tool_calls_from_text(response_text) + + env_reward = 0.0 + is_action_valid = False + invalid_reason: str | None = None + + if not tool_calls: + invalid_reason = "missing env_step tool call" + else: + tool_call = tool_calls[0] + if tool_call.name != "env_step": + invalid_reason = f"expected env_step tool, got {tool_call.name!r}" + else: + command = "" + try: + tool_args = json.loads(tool_call.arguments) + command = str(tool_args.get("command", "")).strip() + except Exception as e: + logger.warning("Failed to parse env_step arguments: %s", e) + invalid_reason = f"failed to parse env_step arguments: {e}" + + if not command and invalid_reason is None: + invalid_reason = "missing command argument" + + if command: + is_action_valid = command in self.current_admissible_commands + result = self.executor.step(command) + self.current_observation = result["observation"] + env_reward = float(result["reward"]) + done = bool(result["done"]) + info = result.get("info", {}) or {} + admissible_commands = info.get("admissible_commands") + self.current_admissible_commands = admissible_commands if isinstance(admissible_commands, list) else [] + self.history_actions = result.get("history_actions", self.history_actions) + if "success" in info: + final_success_flag = bool(info["success"]) + elif "won" in info: + final_success_flag = bool(info["won"]) + dense_reward_sum += env_reward + + if invalid_reason is not None: + self.current_observation = build_invalid_tool_call_observation( + self.current_observation, + invalid_reason, + ) + self.history_actions.append(f"{INVALID_TOOL_CALL_ACTION}: {invalid_reason}") + + step = AgentFlowStep( + prompt_ids=prompt_ids, + response_ids=response_ids, + response_logprobs=output.log_probs[: self.response_length] if output.log_probs else None, + reward_score=0.0, + extra_fields={ + "anchor_obs": observation_before_action, + "reward_extra_info": { + **build_reward_extra_info(env_reward), + "is_action_valid": bool(is_action_valid), + } + }, + ) + step = await self._postprocess(step, **kwargs) + self.steps.append(step) + + if done: + final_step = AgentFlowStep( + prompt_ids=prompt_ids, + response_ids=response_ids, + response_logprobs=output.log_probs[: self.response_length] if output.log_probs else None, + reward_score=None, + extra_fields={ + "anchor_obs": observation_before_action, + "reward_extra_info": build_reward_extra_info(), + }, + ) + final_step = await self._postprocess(final_step, **kwargs) + self.steps.append(final_step) + break + + return AgentFlowOutput(steps=self.steps, metrics=metrics) diff --git a/recipe/alfworld/base.yaml b/recipe/alfworld/base.yaml new file mode 100644 index 0000000..a6a6d22 --- /dev/null +++ b/recipe/alfworld/base.yaml @@ -0,0 +1,5 @@ +- name: alfworld_agent + _target_: recipe.alfworld.alfworld_agent_flow.AlfworldAgentFlow + max_steps: 20 + max_parallel_calls: 1 + max_episode_steps: 20 diff --git a/recipe/alfworld/env/alfworld_wrapper.py b/recipe/alfworld/env/alfworld_wrapper.py new file mode 100644 index 0000000..56a1cdc --- /dev/null +++ b/recipe/alfworld/env/alfworld_wrapper.py @@ -0,0 +1,199 @@ +from __future__ import annotations + +import os +from dataclasses import dataclass, field +from pathlib import Path +from typing import Any + + +def _default_data_root() -> str: + return os.getenv("ALFWORLD_DATA_ROOT", "data/alfworld") + + +@dataclass +class AlfworldTextworldEnv: + data_root: str = field(default_factory=_default_data_root) + max_episode_steps: int = 50 + _env: Any = field(init=False, default=None) + _current_game_path: Path | None = field(init=False, default=None) + + def _games_root(self) -> Path: + return Path(self.data_root).expanduser().resolve() / "games" + + def _resolve_game_path(self, game_relative_path: str) -> Path: + game_path = self._games_root() / game_relative_path + if not game_path.exists(): + raise FileNotFoundError( + f"ALFWorld game file not found: {game_path}. " + f"Expected runtime assets under {self._games_root()}." + ) + return game_path + + def _close_env(self) -> None: + if self._env is not None and hasattr(self._env, "close"): + self._env.close() + self._env = None + self._current_game_path = None + + def _build_env(self, game_path: Path): + try: + import textworld + import textworld.gym + except ImportError as e: + raise ImportError( + "TextWorld runtime is not installed. Please install ALFWorld/TextWorld dependencies first." + ) from e + + wrappers: list[Any] = [] + try: + from alfworld.agents.utils.misc import Demangler + + class AlfredDemangler(textworld.core.Wrapper): + def __init__(self, *args, shuffle: bool = False, **kwargs): + super().__init__(*args, **kwargs) + self.shuffle = shuffle + + def load(self, *args, **kwargs): + super().load(*args, **kwargs) + demangler = Demangler(game_infos=self._entity_infos, shuffle=self.shuffle) + for info in self._entity_infos.values(): + info.name = demangler.demangle_alfred_name(info.id) + + wrappers.append(AlfredDemangler(shuffle=False)) + except Exception: + # Friendly feedback still exists in game.tw-pddl, so demangling is optional. + pass + + class AlfredInfos(textworld.core.Wrapper): + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + self._gamefile = None + + def load(self, *args, **kwargs): + super().load(*args, **kwargs) + self._gamefile = args[0] + + def reset(self, *args, **kwargs): + state = super().reset(*args, **kwargs) + try: + state["extra.gamefile"] = self._gamefile + except Exception: + pass + return state + + wrappers.append(AlfredInfos) + + request_infos = textworld.EnvInfos( + won=True, + admissible_commands=True, + extras=["gamefile"], + ) + env_id = textworld.gym.register_games( + [str(game_path)], + request_infos, + batch_size=1, + asynchronous=False, + max_episode_steps=self.max_episode_steps, + wrappers=wrappers, + ) + return textworld.gym.make(env_id) + + def _ensure_env(self, game_path: Path) -> None: + if self._env is not None and self._current_game_path == game_path: + return + self._close_env() + self._env = self._build_env(game_path) + self._current_game_path = game_path + + @staticmethod + def _unwrap_batch_item(value: Any) -> Any: + if isinstance(value, (list, tuple)) and len(value) == 1: + return value[0] + return value + + @staticmethod + def _batch_info_to_single(info: dict[str, Any] | None) -> dict[str, Any]: + return { + key: AlfworldTextworldEnv._unwrap_batch_item(value) + for key, value in dict(info or {}).items() + } + + @staticmethod + def _won_to_success(value: Any) -> bool: + value = AlfworldTextworldEnv._unwrap_batch_item(value) + if hasattr(value, "item"): + try: + value = value.item() + except Exception: + pass + return bool(float(value)) + + def _normalize_reset_output(self, reset_output: Any) -> tuple[Any, dict[str, Any]]: + if isinstance(reset_output, tuple) and len(reset_output) == 2: + obs, info = reset_output + return obs, dict(info or {}) + return reset_output, {} + + def _normalize_step_output(self, step_output: Any) -> tuple[Any, float, bool, dict[str, Any]]: + if isinstance(step_output, tuple) and len(step_output) == 5: + obs, reward, terminated, truncated, info = step_output + reward = self._unwrap_batch_item(reward) + terminated = self._unwrap_batch_item(terminated) + truncated = self._unwrap_batch_item(truncated) + info = self._unwrap_batch_item(info) + return obs, float(reward), bool(terminated or truncated), dict(info or {}) + if isinstance(step_output, tuple) and len(step_output) == 4: + obs, reward, done, info = step_output + reward = self._unwrap_batch_item(reward) + done = self._unwrap_batch_item(done) + info = self._unwrap_batch_item(info) + return obs, float(reward), bool(done), dict(info or {}) + if isinstance(step_output, tuple) and len(step_output) == 3: + obs, reward, done = step_output + reward = self._unwrap_batch_item(reward) + done = self._unwrap_batch_item(done) + return obs, float(reward), bool(done), {} + raise RuntimeError(f"Unsupported step() output: {type(step_output)} / {step_output!r}") + + @staticmethod + def _state_to_observation(state: Any) -> str: + state = AlfworldTextworldEnv._unwrap_batch_item(state) + if isinstance(state, dict): + if "feedback" in state: + return str(state["feedback"]) + if "observation" in state: + return str(state["observation"]) + return str(state) + + @staticmethod + def _state_to_info(state: Any, base_info: dict[str, Any]) -> dict[str, Any]: + state = AlfworldTextworldEnv._unwrap_batch_item(state) + info = AlfworldTextworldEnv._batch_info_to_single(base_info) + if isinstance(state, dict): + for key in ("won", "admissible_commands", "extra.gamefile"): + if key in state: + info[key] = AlfworldTextworldEnv._unwrap_batch_item(state[key]) + info["success"] = AlfworldTextworldEnv._won_to_success(info.get("won", 0.0)) + return info + + def reset_with_info(self, game_relative_path: str, task_id: str | None = None) -> tuple[str, dict[str, Any]]: + del task_id # task_id is carried for logging / debugging; game_relative_path is the runtime selector. + game_path = self._resolve_game_path(game_relative_path) + self._ensure_env(game_path) + raw_state, base_info = self._normalize_reset_output(self._env.reset()) + raw_state = self._unwrap_batch_item(raw_state) + return self._state_to_observation(raw_state), self._state_to_info(raw_state, base_info) + + def reset(self, game_relative_path: str, task_id: str | None = None) -> str: + observation, _ = self.reset_with_info(game_relative_path=game_relative_path, task_id=task_id) + return observation + + def step(self, action_str: str) -> tuple[str, float, bool, dict[str, Any]]: + if self._env is None: + raise RuntimeError("Environment not initialized. Call reset() before step().") + + raw_state, reward, done, base_info = self._normalize_step_output(self._env.step([action_str])) + raw_state = self._unwrap_batch_item(raw_state) + info = self._state_to_info(raw_state, base_info) + observation = self._state_to_observation(raw_state) + return observation, reward, done, info diff --git a/recipe/alfworld/prepare_alfworld_agent_r1.py b/recipe/alfworld/prepare_alfworld_agent_r1.py new file mode 100644 index 0000000..7372e79 --- /dev/null +++ b/recipe/alfworld/prepare_alfworld_agent_r1.py @@ -0,0 +1,226 @@ +#!/usr/bin/env python3 +""" +Prepare ALFWorld TextWorld data for Agent-R1 / verl training. + +This script: +1. Reads raw ALFWorld data from a source directory. +2. Keeps only TextWorld-usable trials from train / valid_seen / valid_unseen. +3. Copies runtime assets into data/alfworld/games/. +4. Writes verl-compatible parquet files into data/alfworld/. +""" + +from __future__ import annotations + +import argparse +import json +import shutil +from collections import Counter +from pathlib import Path +from typing import Any + +import pandas as pd + +TASK_FAMILY_MAP = { + "pick_and_place_simple": "pick_and_place", + "look_at_obj_in_light": "look_at_obj_in_light", + "pick_clean_then_place_in_recep": "pick_clean_then_place", + "pick_heat_then_place_in_recep": "pick_heat_then_place", + "pick_cool_then_place_in_recep": "pick_cool_then_place", + "pick_two_obj_and_place": "pick_two_obj_and_place", +} + +SPLIT_TO_DATASOURCE = { + "train": "alfworld_train", + "valid_seen": "alfworld_valid_seen", + "valid_unseen": "alfworld_valid_unseen", +} + +SPLIT_TO_OUTPUT = { + "train": "train.parquet", + "valid_seen": "valid_seen.parquet", + "valid_unseen": "valid_unseen.parquet", +} + + +def _load_json(path: Path) -> dict[str, Any]: + with path.open("r", encoding="utf-8") as f: + return json.load(f) + + +def _extract_goal_text(game_data: dict[str, Any], traj_data: dict[str, Any]) -> str: + grammar_text = game_data.get("grammar") + if isinstance(grammar_text, str): + try: + grammar_obj = json.loads(grammar_text) + task_entries = grammar_obj.get("task") or [] + if task_entries and isinstance(task_entries[0], dict): + rhs = str(task_entries[0].get("rhs", "")).strip() + if rhs: + return rhs + except json.JSONDecodeError: + pass + + anns = ((traj_data.get("turk_annotations") or {}).get("anns") or []) + if anns and isinstance(anns[0], dict): + task_desc = str(anns[0].get("task_desc", "")).strip() + if task_desc: + return task_desc + + return str(traj_data.get("task_type", "")).strip() + + +def _copy_runtime_assets( + split: str, + trial_dir: Path, + output_games_root: Path, +) -> str: + task_dir = trial_dir.parent.name + trial_name = trial_dir.name + dest_dir = output_games_root / split / task_dir / trial_name + dest_dir.mkdir(parents=True, exist_ok=True) + + for filename in ("game.tw-pddl", "traj_data.json", "initial_state.pddl"): + src = trial_dir / filename + if src.exists(): + shutil.copy2(src, dest_dir / filename) + + game_path = dest_dir / "game.tw-pddl" + return str(game_path.relative_to(output_games_root)) + + +def _build_row( + *, + split: str, + row_index: int, + traj_data: dict[str, Any], + goal_text: str, + game_relative_path: str, + trial_dir: Path, +) -> dict[str, Any]: + task_id = str(traj_data.get("task_id") or trial_dir.name) + task_type_raw = str(traj_data.get("task_type", "")).strip() + task_family = TASK_FAMILY_MAP[task_type_raw] + + prompt = [{"role": "user", "content": goal_text}] + reward_model = { + "ground_truth": {"success": None}, + "style": "rule", + } + extra_info = { + "index": row_index, + "task_id": task_id, + "split": split, + "task_type_raw": task_type_raw, + "task_family": task_family, + "goal_text": goal_text, + "game_relative_path": game_relative_path, + "trial_dir": str(trial_dir), + } + + return { + "data_source": SPLIT_TO_DATASOURCE[split], + "prompt": prompt, + "reward_model": reward_model, + "extra_info": extra_info, + } + + +def _process_split( + input_root: Path, + output_games_root: Path, + split: str, +) -> tuple[list[dict[str, Any]], dict[str, Any]]: + split_dir = input_root / split + if not split_dir.exists(): + raise FileNotFoundError(f"Split directory not found: {split_dir}") + + rows: list[dict[str, Any]] = [] + family_counter: Counter[str] = Counter() + raw_type_counter: Counter[str] = Counter() + + row_index = 0 + for traj_path in sorted(split_dir.glob("*/trial_*/traj_data.json")): + trial_dir = traj_path.parent + game_path = trial_dir / "game.tw-pddl" + if not game_path.exists(): + continue + + traj_data = _load_json(traj_path) + task_type_raw = str(traj_data.get("task_type", "")).strip() + if task_type_raw not in TASK_FAMILY_MAP: + continue + + game_data = _load_json(game_path) + if not bool(game_data.get("solvable")): + continue + + goal_text = _extract_goal_text(game_data, traj_data) + game_relative_path = _copy_runtime_assets(split, trial_dir, output_games_root) + row = _build_row( + split=split, + row_index=row_index, + traj_data=traj_data, + goal_text=goal_text, + game_relative_path=game_relative_path, + trial_dir=trial_dir, + ) + rows.append(row) + row_index += 1 + + family_counter[row["extra_info"]["task_family"]] += 1 + raw_type_counter[task_type_raw] += 1 + + stats = { + "rows": len(rows), + "task_family_counts": dict(family_counter), + "task_type_raw_counts": dict(raw_type_counter), + } + return rows, stats + + +def main() -> None: + parser = argparse.ArgumentParser(description="Prepare ALFWorld TextWorld parquet data for Agent-R1.") + parser.add_argument( + "--input_dir", + type=str, + default="alfworld_data/json_2.1.1", + help="Raw ALFWorld data root.", + ) + parser.add_argument( + "--output_dir", + type=str, + default="data/alfworld", + help="Directory to write parquet files and runtime assets.", + ) + args = parser.parse_args() + + input_root = Path(args.input_dir).expanduser().resolve() + output_root = Path(args.output_dir).expanduser().resolve() + output_root.mkdir(parents=True, exist_ok=True) + output_games_root = output_root / "games" + output_games_root.mkdir(parents=True, exist_ok=True) + + overall_stats: dict[str, Any] = { + "input_root": str(input_root), + "output_root": str(output_root), + "splits": {}, + } + + for split in ("train", "valid_seen", "valid_unseen"): + rows, stats = _process_split(input_root, output_games_root, split) + df = pd.DataFrame(rows) + out_path = output_root / SPLIT_TO_OUTPUT[split] + df.to_parquet(out_path, index=False) + overall_stats["splits"][split] = stats + + print(f"[{split}] wrote {len(df)} rows -> {out_path}") + print(f"[{split}] task families: {stats['task_family_counts']}") + + stats_path = output_root / "stats.json" + with stats_path.open("w", encoding="utf-8") as f: + json.dump(overall_stats, f, ensure_ascii=False, indent=2) + print(f"Wrote stats -> {stats_path}") + + +if __name__ == "__main__": + main() diff --git a/recipe/alfworld/prompts.py b/recipe/alfworld/prompts.py new file mode 100644 index 0000000..75c95f6 --- /dev/null +++ b/recipe/alfworld/prompts.py @@ -0,0 +1,65 @@ +ALFWORLD_SYSTEM_PROMPT = ( + "You are acting in ALFWorld TextWorld. " + "Choose exactly one command from the provided admissible commands each turn. " + "You may reason briefly inside before acting. " + "Call the env_step tool with that exact command. " + "Do not output a final natural-language answer." +) + + +ALFWORLD_USER_PROMPT = """### Task +{task_text} + +### Current Observation +{observation} + +### History Actions +{history_actions} + +### Admissible Commands +{admissible_commands} + +### Instructions +- Think briefly about the current state inside `...` tags. +- Use exactly one command through the `env_step` tool. +- The command must exactly match one item from `Admissible Commands`. +- Follow ALFWorld TextWorld command style such as `go to dresser 1`, `take mug 1 from cabinet 3`, `use desklamp 1`. +- Use the official observation text as the source of truth. +- Do not output explanations or a final natural-language answer. + +### Output Format + +[Your brief reasoning about the current state and next command.] + + +{{"name": "env_step", "arguments": {{"command": "[one admissible command]"}}}} + +""" + + +EXEC_ACTION_TOOL_SCHEMA = { + "type": "function", + "function": { + "name": "env_step", + "description": ( + "Execute one ALFWorld TextWorld command and return the next official observation." + ), + "parameters": { + "type": "object", + "properties": { + "command": { + "type": "string", + "description": ( + "A single ALFWorld TextWorld command such as " + "`go to dresser 1`, `open cabinet 3`, `take mug 1 from cabinet 3`, `use desklamp 1`. " + "It must exactly match one currently admissible command." + ), + } + }, + "required": ["command"], + }, + }, +} + + +ALFWORLD_TOOL_SCHEMAS = [EXEC_ACTION_TOOL_SCHEMA] diff --git a/recipe/alfworld/requirements.txt b/recipe/alfworld/requirements.txt new file mode 100644 index 0000000..704da96 --- /dev/null +++ b/recipe/alfworld/requirements.txt @@ -0,0 +1,7 @@ +# Data prep +pandas +pyarrow + +# Runtime environment +alfworld +textworld diff --git a/recipe/alfworld/reward_fn.py b/recipe/alfworld/reward_fn.py new file mode 100644 index 0000000..c8768bf --- /dev/null +++ b/recipe/alfworld/reward_fn.py @@ -0,0 +1,97 @@ +from __future__ import annotations + +from typing import Any + +from verl.utils.reward_score import default_compute_score + + +def _native_value(value: Any) -> Any: + if value is None: + return None + if hasattr(value, "item"): + try: + value = value.item() + except Exception: + pass + if isinstance(value, (str, int, float, bool)): + return value + return str(value) + + +def _native_str(value: Any) -> str | None: + value = _native_value(value) + if value is None: + return "" + return str(value) + + +def _native_int(value: Any) -> int | None: + value = _native_value(value) + if value is None: + return 0 + try: + return int(value) + except (TypeError, ValueError): + return 0 + + +def _native_float(value: Any) -> float | None: + value = _native_value(value) + if value is None: + return 0.0 + try: + return float(value) + except (TypeError, ValueError): + return 0.0 + + +def _maybe_bool(value: Any) -> bool | None: + if value is None: + return None + value = _native_value(value) + if isinstance(value, bool): + return value + return bool(value) + + +def compute_score( + data_source: str, + solution_str: str, + ground_truth: Any, + extra_info: dict | None = None, + **kwargs, +) -> float | dict[str, Any]: + if not str(data_source).startswith("alfworld"): + return default_compute_score(data_source, solution_str, ground_truth, extra_info, **kwargs) + + extra_info = extra_info or {} + runtime_info = extra_info.get("reward_extra_info", {}) if isinstance(extra_info, dict) else {} + if not isinstance(runtime_info, dict): + runtime_info = {} + + success_flag = None + if isinstance(ground_truth, dict): + success_flag = _maybe_bool(ground_truth.get("success")) + else: + success_flag = _maybe_bool(ground_truth) + + if success_flag is None: + success_flag = _maybe_bool(runtime_info.get("success")) + if success_flag is None: + success_flag = _maybe_bool(extra_info.get("success")) if isinstance(extra_info, dict) else None + + score = 1.0 if success_flag else 0.0 + + result = { + "score": float(score), + "step_env_reward": _native_float(runtime_info.get("step_env_reward")) or 0.0, + "dense_reward_sum": _native_float(runtime_info.get("dense_reward_sum")), + "success": bool(success_flag), + "num_steps": _native_int(runtime_info.get("num_steps")), + "is_action_valid": bool(_maybe_bool(runtime_info.get("is_action_valid"))), + "task_id": _native_str(runtime_info.get("task_id", extra_info.get("task_id"))), + "split": _native_str(runtime_info.get("split", extra_info.get("split"))), + "task_type_raw": _native_str(runtime_info.get("task_type_raw", extra_info.get("task_type_raw"))), + "task_family": _native_str(runtime_info.get("task_family", extra_info.get("task_family"))), + } + return result diff --git a/recipe/alfworld/summarize_validation.py b/recipe/alfworld/summarize_validation.py new file mode 100644 index 0000000..1bda584 --- /dev/null +++ b/recipe/alfworld/summarize_validation.py @@ -0,0 +1,177 @@ +#!/usr/bin/env python3 +from __future__ import annotations + +import argparse +import json +from collections import defaultdict +from pathlib import Path +from statistics import mean +from typing import Any + +TASK_ORDER = [ + "pick_and_place", + "look_at_obj_in_light", + "pick_clean_then_place", + "pick_heat_then_place", + "pick_cool_then_place", + "pick_two_obj_and_place", +] + + +def _iter_jsonl_files(input_path: Path) -> list[Path]: + if input_path.is_file(): + return [input_path] + return sorted(input_path.glob("*.jsonl")) + + +def _load_entries(input_path: Path) -> list[dict[str, Any]]: + files = _iter_jsonl_files(input_path) + entries: list[dict[str, Any]] = [] + for file_path in files: + with file_path.open("r", encoding="utf-8") as f: + for line in f: + line = line.strip() + if not line: + continue + entries.append(json.loads(line)) + return entries + + +def _filter_entries(entries: list[dict[str, Any]], step: str) -> list[dict[str, Any]]: + if step != "latest": + target_step = int(step) + return [entry for entry in entries if int(entry.get("step", -1)) == target_step] + + if not entries: + return entries + max_step = max(int(entry.get("step", -1)) for entry in entries) + return [entry for entry in entries if int(entry.get("step", -1)) == max_step] + + +def _group_by_task(entries: list[dict[str, Any]]) -> list[dict[str, Any]]: + grouped: dict[tuple[str, str], dict[str, Any]] = {} + for entry in entries: + split = entry.get("split") + task_id = entry.get("task_id") + if not split or not task_id: + continue + key = (str(split), str(task_id)) + bucket = grouped.setdefault( + key, + { + "split": str(split), + "task_id": str(task_id), + "task_family": entry.get("task_family"), + "scores": [], + }, + ) + bucket["scores"].append(float(entry.get("score", 0.0))) + + grouped_rows = [] + for row in grouped.values(): + grouped_rows.append( + { + "split": row["split"], + "task_id": row["task_id"], + "task_family": row["task_family"], + "score": mean(row["scores"]) if row["scores"] else 0.0, + } + ) + return grouped_rows + + +def _compute_summary(rows: list[dict[str, Any]]) -> dict[str, Any]: + by_split: dict[str, list[dict[str, Any]]] = defaultdict(list) + for row in rows: + by_split[row["split"]].append(row) + + summary: dict[str, Any] = {} + for split_name, split_rows in sorted(by_split.items()): + per_task: dict[str, Any] = {} + for task_family in TASK_ORDER: + task_rows = [row for row in split_rows if row.get("task_family") == task_family] + if task_rows: + per_task[task_family] = { + "score": mean(float(row["score"]) for row in task_rows), + "num_tasks": len(task_rows), + } + + summary[split_name] = { + "overall": mean(float(row["score"]) for row in split_rows) if split_rows else 0.0, + "num_tasks": len(split_rows), + "tasks": per_task, + } + + valid_rows = [row for row in rows if row["split"] in {"valid_seen", "valid_unseen"}] + merged_tasks: dict[str, Any] = {} + for task_family in TASK_ORDER: + task_rows = [row for row in valid_rows if row.get("task_family") == task_family] + if task_rows: + merged_tasks[task_family] = { + "score": mean(float(row["score"]) for row in task_rows), + "num_tasks": len(task_rows), + } + + summary["all_valid"] = { + "overall": mean(float(row["score"]) for row in valid_rows) if valid_rows else 0.0, + "num_tasks": len(valid_rows), + "tasks": merged_tasks, + } + return summary + + +def _print_summary(summary: dict[str, Any]) -> None: + for split_name in ("valid_seen", "valid_unseen", "all_valid"): + if split_name not in summary: + continue + split_summary = summary[split_name] + print(f"\n[{split_name}] overall={split_summary['overall']:.4f} num_tasks={split_summary['num_tasks']}") + for task_family in TASK_ORDER: + task_stats = split_summary["tasks"].get(task_family) + if not task_stats: + continue + print( + f" {task_family:<28} score={task_stats['score']:.4f} " + f"num_tasks={task_stats['num_tasks']}" + ) + + +def main() -> None: + parser = argparse.ArgumentParser(description="Summarize ALFWorld validation dumps.") + parser.add_argument( + "--input", + type=str, + required=True, + help="Validation dump directory or a single jsonl file.", + ) + parser.add_argument( + "--step", + type=str, + default="latest", + help="Step to summarize. Use 'latest' or a concrete integer step.", + ) + parser.add_argument( + "--output_json", + type=str, + default="", + help="Optional path to write summary JSON.", + ) + args = parser.parse_args() + + input_path = Path(args.input).expanduser().resolve() + entries = _load_entries(input_path) + filtered = _filter_entries(entries, args.step) + grouped = _group_by_task(filtered) + summary = _compute_summary(grouped) + _print_summary(summary) + + if args.output_json: + output_path = Path(args.output_json).expanduser().resolve() + output_path.parent.mkdir(parents=True, exist_ok=True) + with output_path.open("w", encoding="utf-8") as f: + json.dump(summary, f, ensure_ascii=False, indent=2) + print(f"\nWrote summary -> {output_path}") + + +if __name__ == "__main__": + main() diff --git a/recipe/alfworld/utils.py b/recipe/alfworld/utils.py new file mode 100644 index 0000000..1bf8ddb --- /dev/null +++ b/recipe/alfworld/utils.py @@ -0,0 +1,96 @@ +from __future__ import annotations + +from dataclasses import dataclass, field +from typing import Any + +from recipe.alfworld.env.alfworld_wrapper import AlfworldTextworldEnv +from recipe.alfworld.prompts import ALFWORLD_SYSTEM_PROMPT, ALFWORLD_USER_PROMPT + + +INVALID_TOOL_CALL_ACTION = "" + + +@dataclass +class AlfworldToolExecutor: + max_episode_steps: int = 50 + _env: AlfworldTextworldEnv = field(init=False) + _history_actions: list[str] = field(default_factory=list) + + def __post_init__(self) -> None: + self._env = AlfworldTextworldEnv(max_episode_steps=self.max_episode_steps) + + def reset(self, game_relative_path: str, task_id: str | None = None) -> str: + self._history_actions.clear() + return self._env.reset(game_relative_path=game_relative_path, task_id=task_id) + + def reset_with_info(self, game_relative_path: str, task_id: str | None = None) -> tuple[str, dict[str, Any]]: + self._history_actions.clear() + return self._env.reset_with_info(game_relative_path=game_relative_path, task_id=task_id) + + def step(self, command: str) -> dict[str, Any]: + self._history_actions.append(command) + observation, reward, done, info = self._env.step(command) + return { + "observation": str(observation), + "reward": float(reward), + "done": bool(done), + "info": info, + "history_actions": list(self._history_actions), + } + + +def format_history_actions(actions: list[str]) -> str: + if not actions: + return "None" + return "\n".join(f"[Action {i + 1}] {action}" for i, action in enumerate(actions)) + + +def format_admissible_commands(commands: list[str] | None) -> str: + if not isinstance(commands, list) or not commands: + return "None" + return "\n".join(f"- {command}" for command in commands if command != "help") + + +def extract_task_text(observation: str, fallback: str | None = None) -> str: + marker = "Your task is to:" + if marker in observation: + task = observation.split(marker, 1)[1].strip() + task = task.split("\n", 1)[0].strip() + return f"{marker} {task}" + if fallback: + fallback = str(fallback).strip() + if fallback.lower().startswith(marker.lower()): + return fallback + return f"{marker} {fallback}" + return f"{marker} Unknown." + + +def build_alfworld_messages( + *, + task_text: str, + observation: str, + history_actions: list[str], + admissible_commands: list[str] | None, +) -> list[dict[str, str]]: + return [ + {"role": "system", "content": ALFWORLD_SYSTEM_PROMPT}, + { + "role": "user", + "content": ALFWORLD_USER_PROMPT.format( + task_text=task_text, + observation=observation, + history_actions=format_history_actions(history_actions), + admissible_commands=format_admissible_commands(admissible_commands), + ), + }, + ] + + +def build_invalid_tool_call_observation(previous_observation: str, reason: str) -> str: + return ( + "Invalid tool call. You must call the `env_step` tool exactly once with JSON arguments " + 'like {"command": ""}. ' + f"Reason: {reason}\n\n" + "The environment state did not change. Current Observation:\n" + f"{previous_observation}" + ) diff --git a/recipe/hotpotqa/__init__.py b/recipe/hotpotqa/__init__.py new file mode 100644 index 0000000..55a7591 --- /dev/null +++ b/recipe/hotpotqa/__init__.py @@ -0,0 +1 @@ +"""HotpotQA recipe.""" diff --git a/recipe/hotpotqa/base.yaml b/recipe/hotpotqa/base.yaml new file mode 100644 index 0000000..d7a3e60 --- /dev/null +++ b/recipe/hotpotqa/base.yaml @@ -0,0 +1,8 @@ +- name: hotpotqa_agent + _target_: recipe.hotpotqa.hotpotqa_agent_flow.HotpotQAAgentFlow + max_steps: 5 + max_parallel_calls: 4 + force_first_search: true + + embedding_model_name: ${oc.env:HOTPOTQA_EMBEDDING_MODEL,BAAI/bge-large-en-v1.5} + embedding_devices: null diff --git a/recipe/hotpotqa/base_faiss_cpu.yaml b/recipe/hotpotqa/base_faiss_cpu.yaml new file mode 100644 index 0000000..00c62e8 --- /dev/null +++ b/recipe/hotpotqa/base_faiss_cpu.yaml @@ -0,0 +1,15 @@ +- name: hotpotqa_agent + _target_: recipe.hotpotqa.hotpotqa_agent_flow.HotpotQAAgentFlow + max_steps: 5 + max_parallel_calls: 4 + force_first_search: true + + # BGE model id or local mirror (must match `process_hotpotqa.py --embedding_model` when building index.bin) + embedding_model_name: ${oc.env:HOTPOTQA_EMBEDDING_MODEL,BAAI/bge-large-en-v1.5} + # Retrieval corpus/index root, separate from train/validation parquet data root. + corpus_data_dir: ${oc.env:HOTPOTQA_CORPUS_DATA_ROOT,data/corpus/hotpotqa_corpus} + # BGE is separate from vLLM: no gpu_memory_utilization knob; sharing a GPU with vLLM(0.7) is best-effort—tune vLLM util downward or use CPU (`embedding_devices` / HOTPOTQA_EMBEDDING_*). + # null: use HOTPOTQA_EMBEDDING_DEVICE (default cpu), or per-worker cuda:i if HOTPOTQA_EMBEDDING_PER_WORKER_GPU=1 + embedding_devices: null + # Like Agent-R1-legacy: on parse/arg errors inject brief feedback into next user turn and try lenient JSON recovery + enable_tool_parse_feedback: true diff --git a/recipe/hotpotqa/build_retrieval_corpus.py b/recipe/hotpotqa/build_retrieval_corpus.py new file mode 100644 index 0000000..8c6fce5 --- /dev/null +++ b/recipe/hotpotqa/build_retrieval_corpus.py @@ -0,0 +1,251 @@ +#!/usr/bin/env python3 +"""Build hpqa_corpus.jsonl for 2WikiMultiHopQA and MuSiQue from downloaded raw data. + +Usage (conda env ``steppo``, repo root):: + + python recipe/hotpotqa/build_retrieval_corpus.py --dataset all + +Then build FAISS indexes (requires GPU for 2Wiki ~5.9M paragraphs):: + + export PYTHONPATH=$(pwd) + EMB=${HOTPOTQA_EMBEDDING_MODEL:-BAAI/bge-large-en-v1.5} + + python recipe/hotpotqa/process_hotpotqa.py \\ + --data_dir data/corpus/musique_corpus \\ + --corpus_path data/corpus/musique_corpus/hpqa_corpus.jsonl \\ + --embedding_model "$EMB" --devices cuda:0 --batch_size 1024 + + python recipe/hotpotqa/process_hotpotqa.py \\ + --data_dir data/corpus/2wikimultihopqa_corpus \\ + --corpus_path data/corpus/2wikimultihopqa_corpus/hpqa_corpus.jsonl \\ + --embedding_model "$EMB" --devices cuda:0 --batch_size 1024 +""" + +from __future__ import annotations + +import argparse +import json +from pathlib import Path +from typing import Any, Iterator + +_REPO_ROOT = Path(__file__).resolve().parents[2] +_DEFAULT_RAW_ROOT = _REPO_ROOT / "data" / "raw" +_DEFAULT_CORPUS_ROOT = _REPO_ROOT / "data" / "corpus" + + +def _write_corpus_entry(fout, title: str, text: str) -> None: + """Write one deduplicated corpus line if text is non-empty.""" + title = str(title).strip() + text = str(text).strip() + if not text: + return + fout.write(json.dumps({"title": title, "text": text}, ensure_ascii=False) + "\n") + + +def iter_2wiki_para_records(path: Path) -> Iterator[tuple[str, str]]: + """Yield (title, text) from 2Wiki ``para_with_hyperlink.jsonl``.""" + with path.open("r", encoding="utf-8") as fin: + for line_no, line in enumerate(fin, start=1): + line = line.strip() + if not line: + continue + rec = json.loads(line) + title = str(rec.get("title", "")).strip() + sentences = rec.get("sentences") or [] + if isinstance(sentences, list): + text = " ".join(str(s).strip() for s in sentences if str(s).strip()).strip() + else: + text = str(sentences).strip() + if not text: + continue + yield title, text + if line_no % 500_000 == 0: + print(f"[2wiki] scanned {line_no:,} lines...") + + +def build_2wiki_corpus( + input_path: Path, + output_path: Path, + *, + dedupe: bool = False, +) -> int: + """Convert 2Wiki paragraph jsonl to ``hpqa_corpus.jsonl``. + + Args: + input_path: Path to ``para_with_hyperlink.jsonl``. + output_path: Destination ``hpqa_corpus.jsonl``. + dedupe: If True, drop duplicate (title, text) pairs (uses more RAM). + + Returns: + Number of paragraphs written. + """ + if not input_path.is_file(): + raise FileNotFoundError(f"2Wiki corpus not found: {input_path}") + + output_path.parent.mkdir(parents=True, exist_ok=True) + seen: set[tuple[str, str]] | None = set() if dedupe else None + count = 0 + + print(f"[2wiki] reading {input_path}") + print(f"[2wiki] writing {output_path}") + with output_path.open("w", encoding="utf-8") as fout: + for title, text in iter_2wiki_para_records(input_path): + key = (title, text) + if seen is not None: + if key in seen: + continue + seen.add(key) + _write_corpus_entry(fout, title, text) + count += 1 + if count % 500_000 == 0: + print(f"[2wiki] wrote {count:,} paragraphs...") + + print(f"[2wiki] done: {count:,} paragraphs -> {output_path}") + return count + + +def iter_musique_paragraphs(data_dir: Path, *, include_full: bool) -> Iterator[tuple[str, str]]: + """Yield (title, text) from MuSiQue official jsonl files under ``data/``.""" + patterns = ["musique_ans_v1.0_*.jsonl"] + if include_full: + patterns.append("musique_full_v1.0_*.jsonl") + + paths: list[Path] = [] + for pattern in patterns: + paths.extend(sorted(data_dir.glob(pattern))) + + if not paths: + raise FileNotFoundError(f"No MuSiQue jsonl files under {data_dir}") + + for path in paths: + print(f"[musique] reading {path.name}") + with path.open("r", encoding="utf-8") as fin: + for line in fin: + line = line.strip() + if not line: + continue + ex: dict[str, Any] = json.loads(line) + for para in ex.get("paragraphs") or []: + if not isinstance(para, dict): + continue + title = str(para.get("title", "")).strip() + text = str(para.get("paragraph_text", "")).strip() + if text: + yield title, text + + +def build_musique_corpus( + data_dir: Path, + output_path: Path, + *, + include_full: bool = True, +) -> int: + """Extract deduplicated paragraphs from MuSiQue jsonl into ``hpqa_corpus.jsonl``. + + Args: + data_dir: Directory containing ``musique_*_v1.0_*.jsonl``. + output_path: Destination ``hpqa_corpus.jsonl``. + include_full: Include ``musique_full_*`` splits for a larger retrieval pool. + + Returns: + Number of unique paragraphs written. + """ + if not data_dir.is_dir(): + raise FileNotFoundError(f"MuSiQue data dir not found: {data_dir}") + + output_path.parent.mkdir(parents=True, exist_ok=True) + seen: set[tuple[str, str]] = set() + count = 0 + + print(f"[musique] writing {output_path}") + with output_path.open("w", encoding="utf-8") as fout: + for title, text in iter_musique_paragraphs(data_dir, include_full=include_full): + key = (title, text) + if key in seen: + continue + seen.add(key) + _write_corpus_entry(fout, title, text) + count += 1 + if count % 50_000 == 0: + print(f"[musique] wrote {count:,} unique paragraphs...") + + print(f"[musique] done: {count:,} paragraphs -> {output_path}") + return count + + +def main() -> None: + """CLI entry: build one or both retrieval corpora from raw downloads.""" + parser = argparse.ArgumentParser( + description="Build hpqa_corpus.jsonl for 2Wiki / MuSiQue retrieval indexes.", + ) + parser.add_argument( + "--dataset", + type=str, + default="all", + choices=("2wikimultihopqa", "musique", "all"), + help="Which corpus to build.", + ) + parser.add_argument( + "--raw_root", + type=str, + default=str(_DEFAULT_RAW_ROOT), + help="Root directory with raw/2wikimultihopqa and raw/musique.", + ) + parser.add_argument( + "--corpus_root", + type=str, + default=str(_DEFAULT_CORPUS_ROOT), + help="Output root under data/corpus/.", + ) + parser.add_argument( + "--wiki2_input", + type=str, + default="", + help="Override path to para_with_hyperlink.jsonl.", + ) + parser.add_argument( + "--musique_data_dir", + type=str, + default="", + help="Override MuSiQue data/ directory with jsonl files.", + ) + parser.add_argument( + "--wiki2_dedupe", + action="store_true", + help="Deduplicate 2Wiki paragraphs (not recommended for ~6M lines; high RAM).", + ) + parser.add_argument( + "--musique_skip_full", + action="store_true", + help="Only use musique_ans_* jsonl (smaller corpus).", + ) + args = parser.parse_args() + + raw_root = Path(args.raw_root).expanduser().resolve() + corpus_root = Path(args.corpus_root).expanduser().resolve() + + if args.dataset in ("2wikimultihopqa", "all"): + wiki_in = ( + Path(args.wiki2_input).expanduser().resolve() + if args.wiki2_input + else raw_root / "2wikimultihopqa" / "para_with_hyperlink.jsonl" + ) + wiki_out = corpus_root / "2wikimultihopqa_corpus" / "hpqa_corpus.jsonl" + build_2wiki_corpus(wiki_in, wiki_out, dedupe=args.wiki2_dedupe) + + if args.dataset in ("musique", "all"): + musique_dir = ( + Path(args.musique_data_dir).expanduser().resolve() + if args.musique_data_dir + else raw_root / "musique" / "data" + ) + musique_out = corpus_root / "musique_corpus" / "hpqa_corpus.jsonl" + build_musique_corpus( + musique_dir, + musique_out, + include_full=not args.musique_skip_full, + ) + + +if __name__ == "__main__": + main() diff --git a/recipe/hotpotqa/hotpotqa_agent_flow.py b/recipe/hotpotqa/hotpotqa_agent_flow.py new file mode 100644 index 0000000..f63b2cb --- /dev/null +++ b/recipe/hotpotqa/hotpotqa_agent_flow.py @@ -0,0 +1,424 @@ +""" +HotpotQA AgentFlow — multi-step search agent for multi-hop QA. + +Architecture follows recipe/paper_search style: +- Each step re-builds messages from current state (not multi-turn message accumulation) +- Passages and action history are maintained as structured state, rendered into prompt each step +- Prompt length is bounded: passages are truncated to fit within budget +- Tool / answer format: `tools=` in apply_chat_template **plus** user prompt aligned with + `recipe/paper_search/prompts.py` (``, ``, final ``); see + `recipe.hotpotqa.prompts.HOTPOTQA_USER_PROMPT`. +- Search tool backed by local FAISS + BGE (HotpotQASearchToolLegacy) +- Reward: tool steps get reward_score=0.0; final step gets reward_score=None (→ custom EM reward) +""" + +import ast +import json +import logging +import os +import re +from typing import Any +from uuid import uuid4 + +from transformers import AutoProcessor, AutoTokenizer + +from agent_r1.agent_flow.agent_flow import AgentFlowBase, AgentFlowOutput, AgentFlowStep, register +from agent_r1.reward_loop.reward_loop import RewardLoopWorker +from recipe.hotpotqa.prompts import ( + HOTPOTQA_SYSTEM_PROMPT, + HOTPOTQA_TOOL_SCHEMAS, + HOTPOTQA_USER_PROMPT, +) +from recipe.hotpotqa.utils import ( + DEFAULT_HOTPOTQA_EMBEDDING_MODEL, + HotpotQASearchToolLegacy, + parse_legacy_tool_result, + resolve_hotpotqa_embedding_devices, +) +from verl.experimental.agent_loop.agent_loop import AsyncLLMServerManager, DictConfigWrap +from verl.experimental.agent_loop.tool_parser import FunctionCall, ToolParser +from verl.utils.profiler import simple_timer + +logger = logging.getLogger(__file__) +logger.setLevel(os.getenv("VERL_LOGGING_LEVEL", "WARN")) + +_RETRIEVAL_TOOL_NAMES = frozenset({"search", "wiki_search"}) +_TOOL_CALL_BLOCK = re.compile(r"(.*?)", re.DOTALL) + + +def _json_or_python_dict(s: str) -> Any: + """Parse JSON; on failure try Python literal dict (common small-model mistake).""" + s = s.strip() + if not s: + return None + try: + return json.loads(s) + except Exception: + pass + if len(s) > 8192 or "__" in s: + return None + try: + return ast.literal_eval(s) + except Exception: + return None + + +def _normalize_tool_call_dict(obj: Any) -> dict[str, Any] | None: + """Return dict with str name and dict arguments, or None.""" + if not isinstance(obj, dict) or "name" not in obj: + return None + name = obj["name"] + args = obj.get("arguments") + if args is None: + return None + if isinstance(args, str): + inner = _json_or_python_dict(args) + if not isinstance(inner, dict): + return None + args = inner + if not isinstance(args, dict): + return None + return {"name": str(name), "arguments": args} + + +def _recover_tool_calls_from_text(text: str) -> list[FunctionCall]: + """ + Fallback when Hermes json.loads fails (e.g. single-quoted dicts, trailing commas). + Agent-R1-legacy NousToolEnv still fails JSON but feeds 'Error: JSONDecodeError' into the next + user turn; here we try to salvage calls and additionally add text feedback (see run loop). + """ + out: list[FunctionCall] = [] + for raw in _TOOL_CALL_BLOCK.findall(text): + obj = _json_or_python_dict(raw.strip()) + if obj is None: + continue + norm = _normalize_tool_call_dict(obj) + if norm is None: + continue + try: + out.append( + FunctionCall( + name=norm["name"], + arguments=json.dumps(norm["arguments"], ensure_ascii=False), + ) + ) + except Exception: + continue + return out + + +def _decode_tool_arguments(arguments: str) -> dict[str, Any] | None: + try: + obj: Any = json.loads(arguments) + except Exception: + return None + if isinstance(obj, str): + try: + obj = json.loads(obj) + except Exception: + return None + return obj if isinstance(obj, dict) else None + + +def _format_passage_list(passages: list[tuple[str, str]], max_chars: int = 0) -> str: + """Format accumulated passages for prompt. Each entry is (query, text).""" + if not passages: + return "None" + lines: list[str] = [] + total = 0 + for i, (query, text) in enumerate(passages, start=1): + snippet = text[:1200].replace("\n", " ") + line = f"[{i}] (query: {query}) {snippet}" + if max_chars > 0 and total + len(line) > max_chars: + lines.append(f"... ({len(passages) - i + 1} more passages truncated)") + break + lines.append(line) + total += len(line) + return "\n".join(lines) + + +def _format_history_actions(actions: list[str]) -> str: + if not actions: + return "None" + return "\n".join(f"[Search] {q}" for q in actions) + + +@register("hotpotqa_agent") +class HotpotQAAgentFlow(AgentFlowBase): + """ + Multi-step HotpotQA agent (paper_search style state management). + + Each step re-builds [system, user] from current state: + - user_query (fixed) + - passage_list (accumulated, truncated to fit) + - history_actions (list of past queries) + This avoids prompt length explosion from multi-turn message accumulation. + """ + + def __init__( + self, + trainer_config: DictConfigWrap, + server_manager: AsyncLLMServerManager, + reward_loop_worker: RewardLoopWorker, + tokenizer: AutoTokenizer, + processor: AutoProcessor, + **kwargs, + ): + super().__init__(trainer_config, server_manager, reward_loop_worker, tokenizer, processor, **kwargs) + + self.max_steps = int(kwargs.get("max_steps", 5)) + self.max_parallel_calls = int(kwargs.get("max_parallel_calls", 4)) + self.force_first_search = bool(kwargs.get("force_first_search", True)) + + self.tool_parser = ToolParser.get_tool_parser( + self.config.actor_rollout_ref.rollout.multi_turn.format, + self.tokenizer, + ) + self.prompt_length = self.config.actor_rollout_ref.rollout.prompt_length + self.response_length = self.config.actor_rollout_ref.rollout.response_length + self.tool_schemas = HOTPOTQA_TOOL_SCHEMAS + + corpus_data_dir = kwargs.get("corpus_data_dir") + embedding_model_name = kwargs.get("embedding_model_name", DEFAULT_HOTPOTQA_EMBEDDING_MODEL) + embedding_devices = resolve_hotpotqa_embedding_devices( + kwargs.get("embedding_devices"), + kwargs.get("agent_flow_worker_index"), + ) + self.search_tool = HotpotQASearchToolLegacy( + embedding_model_name=embedding_model_name, + embedding_devices=embedding_devices, + corpus_data_dir=corpus_data_dir, + ) + self.enable_tool_parse_feedback = bool(kwargs.get("enable_tool_parse_feedback", True)) + + def _build_messages( + self, + question: str, + passages: list[tuple[str, str]], + actions: list[str], + tool_feedback: str, + *, + max_passage_chars: int | None = None, + ) -> list[dict]: + """Build [system, user] messages from current state, with passage truncation for safety. + + Must stay in sync with `HOTPOTQA_USER_PROMPT.format(...)` keys: + user_query, history_actions, passage_list, tool_feedback. + """ + if max_passage_chars is None: + max_passage_chars = self.prompt_length * 3 + fb = tool_feedback.strip() if tool_feedback.strip() else "None" + user_content = HOTPOTQA_USER_PROMPT.format( + user_query=question, + passage_list=_format_passage_list(passages, max_chars=max_passage_chars), + history_actions=_format_history_actions(actions), + tool_feedback=fb, + ) + return [ + {"role": "system", "content": HOTPOTQA_SYSTEM_PROMPT}, + {"role": "user", "content": user_content}, + ] + + async def _prompt_ids_within_budget( + self, + question: str, + passages: list[tuple[str, str]], + history_actions: list[str], + fb_text: str, + *, + num_step: int, + ) -> list[int]: + """ + Tokenize prompt with tools= so the model always sees tool definitions (chat template prefix). + + PaperSearchAgentFlow does not truncate the tokenized prompt. Here we only shrink the + variable-length passage block; we must NOT use tail slicing on prompt_ids — that drops + the tool schema and breaks JSON adherence. + """ + max_chars = self.prompt_length * 3 + min_chars = 400 + last_ids: list[int] = [] + while True: + messages = self._build_messages( + question, passages, history_actions, fb_text, max_passage_chars=max_chars + ) + last_ids = await self.apply_chat_template(messages, tools=self.tool_schemas) + if len(last_ids) <= self.prompt_length: + return last_ids + if max_chars <= min_chars: + break + max_chars = max(min_chars, int(max_chars * 0.72)) + + logger.warning( + "[hotpotqa_agent][step=%d] prompt still too long (%d tokens, limit %d) after shrinking " + "passages; keeping PREFIX so tool definitions remain.", + num_step, + len(last_ids), + self.prompt_length, + ) + return last_ids[: self.prompt_length] + + def _make_anchor_obs( + self, + question: str, + passages: list[tuple[str, str]], + history_actions: list[str], + tool_feedback: str, + ) -> str: + fb = tool_feedback.strip() if tool_feedback.strip() else "None" + return HOTPOTQA_USER_PROMPT.format( + user_query=question, + passage_list=_format_passage_list(passages, max_chars=self.prompt_length * 3), + history_actions=_format_history_actions(history_actions), + tool_feedback=fb, + ) + + def _make_extra_fields(self, anchor_obs: str, history_actions: list[str], acc: float = 0.0) -> dict[str, Any]: + """Build extra_fields with consistent reward_extra_info keys across all steps.""" + return { + "anchor_obs": anchor_obs, + "reward_extra_info": { + "num_tool_steps": len(history_actions), + "acc": acc, + }, + } + + async def run(self, sampling_params: dict[str, Any], **kwargs) -> AgentFlowOutput: + raw_prompt = list(kwargs["raw_prompt"]) + question = raw_prompt[0]["content"] + + metrics: dict[str, Any] = {} + steps: list[AgentFlowStep] = [] + passages: list[tuple[str, str]] = [] + history_actions: list[str] = [] + + if self.force_first_search: + self._do_search(question, passages, history_actions) + + tool_feedback_lines: list[str] = [] + + num_steps = 0 + while num_steps < self.max_steps: + num_steps += 1 + + fb_text = "\n".join(tool_feedback_lines[-3:]) if tool_feedback_lines else "" + if not self.enable_tool_parse_feedback: + fb_text = "" + anchor_obs = self._make_anchor_obs(question, passages, history_actions, fb_text) + prompt_ids = await self._prompt_ids_within_budget( + question, passages, history_actions, fb_text, num_step=num_steps + ) + + with simple_timer("generate_sequences", metrics): + output = await self.server_manager.generate( + request_id=uuid4().hex, + prompt_ids=prompt_ids, + sampling_params=sampling_params, + ) + + response_ids = output.token_ids[: self.response_length] + _, tool_calls = await self.tool_parser.extract_tool_calls(response_ids) + response_text = self.tokenizer.decode(response_ids, skip_special_tokens=True) + + if not tool_calls: + recovered = _recover_tool_calls_from_text(response_text) + if recovered: + tool_calls = recovered + elif ( + self.enable_tool_parse_feedback + and "" in response_text + and "" in response_text + ): + tool_feedback_lines.append( + "Previous assistant message contained ... but the JSON " + "could not be parsed. Use strict JSON with double-quoted keys and string values, " + "comma between fields: name must be search and arguments must include query." + ) + + if not tool_calls: + step = AgentFlowStep( + prompt_ids=prompt_ids, + response_ids=response_ids, + response_logprobs=output.log_probs[: self.response_length] if output.log_probs else None, + reward_score=None, + extra_fields=self._make_extra_fields(anchor_obs, history_actions), + ) + step = await self._postprocess(step, **kwargs) + ri = step.extra_fields.get("reward_extra_info", {}) + step.extra_fields["reward_extra_info"] = { + "num_tool_steps": len(history_actions), + "acc": ri.get("acc", 0.0), + } + steps.append(step) + break + + tool_calls = tool_calls[: self.max_parallel_calls] + + queries: list[str] = [] + for tc in tool_calls: + if tc.name not in _RETRIEVAL_TOOL_NAMES: + continue + tool_args = _decode_tool_arguments(tc.arguments) + if not tool_args: + continue + query = tool_args.get("query") + if query: + queries.append(str(query)) + + if queries: + with simple_timer("tool_calls", metrics): + self._do_search_batch(queries, passages, history_actions) + elif self.enable_tool_parse_feedback and tool_calls: + tool_feedback_lines.append( + "Previous tool call was recognized but arguments were invalid. " + "For search, arguments must be a JSON object with a string field query." + ) + + step = AgentFlowStep( + prompt_ids=prompt_ids, + response_ids=response_ids, + response_logprobs=output.log_probs[: self.response_length] if output.log_probs else None, + reward_score=0.0, + extra_fields=self._make_extra_fields(anchor_obs, history_actions), + ) + step = await self._postprocess(step, **kwargs) + steps.append(step) + + return AgentFlowOutput(steps=steps, metrics=metrics) + + def _do_search(self, query: str, passages: list[tuple[str, str]], history_actions: list[str]) -> None: + """Execute a single search query and update state.""" + try: + results = self.search_tool.batch_execute([{"query": query}]) + self._ingest_results(query, results, passages) + history_actions.append(query) + except Exception as e: + logger.warning("[hotpotqa_agent] search failed for query=%r: %s", query, e) + history_actions.append(query) + + def _do_search_batch(self, queries: list[str], passages: list[tuple[str, str]], history_actions: list[str]) -> None: + """Execute multiple search queries and update state.""" + try: + results = self.search_tool.batch_execute([{"query": q} for q in queries]) + for query, item in zip(queries, results): + self._ingest_results(query, [item], passages) + history_actions.append(query) + except Exception as e: + logger.warning("[hotpotqa_agent] batch search failed: %s", e) + for q in queries: + history_actions.append(q) + + @staticmethod + def _ingest_results( + query: str, + results: list[dict[str, Any]], + passages: list[tuple[str, str]], + ) -> None: + """Parse search results and deduplicate into the passage list.""" + for item in results: + if not item.get("success", False): + continue + content = str(item.get("content", "")) + for p in parse_legacy_tool_result(content): + if not any(existing_text == p.text for _, existing_text in passages): + passages.append((query, p.text)) diff --git a/recipe/hotpotqa/prepare_hotpotqa_agent_r1.py b/recipe/hotpotqa/prepare_hotpotqa_agent_r1.py new file mode 100644 index 0000000..64651a2 --- /dev/null +++ b/recipe/hotpotqa/prepare_hotpotqa_agent_r1.py @@ -0,0 +1,344 @@ +#!/usr/bin/env python3 +# Copyright 2025 recipe contributors +# +# Download HotpotQA (distractor setting) and export: +# 1) train.parquet / validation.parquet for verl/agent_r1 RLHFDataset (prompt + reward_model + data_source) +# 2) hpqa_corpus.jsonl — deduplicated wiki paragraphs from all contexts (for FAISS indexing, see process_hotpotqa.py) +# +# Usage: +# pip install datasets pyarrow pandas +# python recipe/hotpotqa/prepare_hotpotqa_agent_r1.py \ +# --output_dir data/corpus/hotpotqa \ +# --corpus_output_path data/corpus/hotpotqa_corpus/hpqa_corpus.jsonl +# +# Cross-eval only (2WikiMultiHopQA + MuSiQue validation, no corpus rebuild): +# python recipe/hotpotqa/prepare_hotpotqa_agent_r1.py \ +# --skip_hotpotqa \ +# --include_cross_eval \ +# --skip_corpus + +from __future__ import annotations + +import argparse +import json +import os +from collections.abc import Sequence +from typing import Any + +import pandas as pd + +try: + from datasets import load_dataset +except ImportError as e: + raise SystemExit( + "Please install dependencies: pip install datasets pyarrow pandas" + ) from e + + +def _row_to_agent_r1( + ex: dict[str, Any], + split: str, + row_index: int, +) -> dict[str, Any]: + """Single HotpotQA example -> verl RLHFDataset row.""" + qid = ex.get("_id", f"{split}_{row_index}") + question = ex["question"].strip() + answer = ex["answer"] + if not isinstance(answer, str): + answer = str(answer) + + # Match paper_search-style rollout: first user message is the task text (agent reads raw_prompt[0]["content"]). + prompt = [{"role": "user", "content": question}] + + reward_model = { + "ground_truth": answer, + "style": "rule", + } + + extra_info: dict[str, Any] = { + "index": row_index, + "question_id": qid, + "split": split, + "type": ex.get("type"), + "level": ex.get("level"), + } + + return { + "data_source": "hotpotqa_distractor", + "prompt": prompt, + "reward_model": reward_model, + "extra_info": extra_info, + } + + +def _normalize_answers(answers: Any) -> list[str]: + """Normalize answer payloads from MultiHopQA-style datasets. + + Args: + answers: Raw answer value from the dataset. It can be a string, a + sequence of strings, or another scalar value. + + Returns: + A list of non-empty answer strings. + """ + if isinstance(answers, str): + answer = answers.strip() + return [answer] if answer else [] + if isinstance(answers, Sequence): + return [str(answer).strip() for answer in answers if str(answer).strip()] + if answers is None: + return [] + answer = str(answers).strip() + return [answer] if answer else [] + + +def _cross_eval_row_to_agent_r1( + ex: dict[str, Any], + data_source: str, + split: str, + row_index: int, +) -> dict[str, Any]: + """Convert one cross-eval QA example to the HotpotQA Agent-R1 row schema. + + Args: + ex: Raw example from a MultiHopQA-style dataset. + data_source: Metric/reward grouping name for this validation set. + split: Source split name. + row_index: Row index in the source split. + + Returns: + A dict containing ``data_source``, ``prompt``, ``reward_model``, and + ``extra_info`` columns for ``RLHFDataset``. + + Raises: + ValueError: If the example has no question/query field. + """ + question = (ex.get("query") or ex.get("question") or "").strip() + if not question: + raise ValueError(f"Missing question/query in {data_source} row {row_index}") + + qid = ex.get("query_id") or ex.get("id") or f"{data_source}_{split}_{row_index}" + answers = _normalize_answers(ex.get("answers", ex.get("golden_answers", ex.get("answer")))) + primary_answer = answers[0] if answers else "" + prompt = [{"role": "user", "content": question}] + reward_model = { + "ground_truth": primary_answer, + "style": "rule", + } + extra_info: dict[str, Any] = { + "index": row_index, + "question_id": qid, + "split": split, + "source_dataset": data_source, + "answers": answers, + } + + return { + "data_source": data_source, + "prompt": prompt, + "reward_model": reward_model, + "extra_info": extra_info, + } + + +def _write_cross_eval_parquets( + output_dir: str, + hf_name: str, + configs: list[str], + split: str, + max_samples: int, +) -> None: + """Write cross-eval validation parquets in HotpotQA Agent-R1 schema. + + Args: + output_dir: Directory where converted parquet files are written. + hf_name: HuggingFace dataset name. + configs: Dataset configs to load from ``hf_name``. + split: Split to load for each config. + max_samples: If positive, keep at most this many rows per config. + """ + for config_name in configs: + print(f"Loading {hf_name} / {config_name} / {split} from HuggingFace...") + dataset = load_dataset(hf_name, config_name, split=split) + n_rows = len(dataset) if max_samples <= 0 else min(len(dataset), max_samples) + rows = [ + _cross_eval_row_to_agent_r1(dataset[i], config_name, split, i) + for i in range(n_rows) + ] + out_path = os.path.join(output_dir, f"{config_name}_{split}.parquet") + pd.DataFrame(rows).to_parquet(out_path, index=False) + print(f"Wrote {n_rows} rows -> {out_path}") + + +def _iter_context_paragraphs(ex: dict[str, Any]): + """Yield (title, sentences_list) for one HotpotQA example. + + HuggingFace `hotpot_qa` uses context: {title: [...], sentences: [[...], ...]}. + Official JSON dumps use context: [[title, [sent, ...]], ...]. + """ + ctx = ex.get("context") or [] + if isinstance(ctx, dict) and "title" in ctx and "sentences" in ctx: + titles = ctx["title"] + sents_block = ctx["sentences"] + for title, sents in zip(titles, sents_block): + yield title, sents + else: + for item in ctx: + title, sents = item[0], item[1] + yield title, sents + + +def _contexts_to_corpus_entries(examples: list[dict[str, Any]]) -> list[dict[str, str]]: + """Flatten HotpotQA context paragraphs into {title, text} records.""" + seen: set[tuple[str, str]] = set() + out: list[dict[str, str]] = [] + for ex in examples: + for title, sents in _iter_context_paragraphs(ex): + text = " ".join(sents).strip() + title = str(title).strip() + key = (title, text) + if not text or key in seen: + continue + seen.add(key) + out.append({"title": title, "text": text}) + return out + + +def main() -> None: + parser = argparse.ArgumentParser(description="Prepare HotpotQA for Agent-R1 / verl RL training.") + parser.add_argument( + "--output_dir", + type=str, + default="data/corpus/hotpotqa", + help="Directory for train/validation parquet (created if missing).", + ) + parser.add_argument( + "--hf_name", + type=str, + default="hotpot_qa", + help="HuggingFace dataset name.", + ) + parser.add_argument( + "--hf_config", + type=str, + default="distractor", + help="HuggingFace config (distractor = 10 paragraphs per question).", + ) + parser.add_argument( + "--max_train", + type=int, + default=-1, + help="If >0, only keep first N training examples (debug).", + ) + parser.add_argument( + "--max_val", + type=int, + default=-1, + help="If >0, only keep first N validation examples (debug).", + ) + parser.add_argument( + "--skip_corpus", + action="store_true", + help="Do not write hpqa_corpus.jsonl.", + ) + parser.add_argument( + "--corpus_output_path", + type=str, + default="data/corpus/hotpotqa_corpus/hpqa_corpus.jsonl", + help="Output path for hpqa_corpus.jsonl used by search tool/index builder.", + ) + parser.add_argument( + "--skip_hotpotqa", + action="store_true", + help="Skip HotpotQA train/validation parquet generation.", + ) + parser.add_argument( + "--include_cross_eval", + action="store_true", + help="Also generate 2WikiMultiHopQA/MuSiQue validation parquets for training-time validation.", + ) + parser.add_argument( + "--cross_eval_hf_name", + type=str, + default="corag/multihopqa", + help="HuggingFace dataset containing cross-eval QA validation splits.", + ) + parser.add_argument( + "--cross_eval_configs", + type=str, + default="2wikimultihopqa,musique", + help="Comma-separated dataset configs to convert for cross-eval validation.", + ) + parser.add_argument( + "--cross_eval_split", + type=str, + default="validation", + help="Split name to convert for cross-eval datasets.", + ) + parser.add_argument( + "--max_cross_eval", + type=int, + default=-1, + help="If >0, only keep first N cross-eval examples per dataset (debug).", + ) + args = parser.parse_args() + + out_dir = os.path.abspath(os.path.expanduser(args.output_dir)) + os.makedirs(out_dir, exist_ok=True) + + ds = None + if not args.skip_hotpotqa: + print(f"Loading {args.hf_name} / {args.hf_config} from HuggingFace...") + ds = load_dataset(args.hf_name, args.hf_config) + + for split_name, parquet_name in [("train", "train.parquet"), ("validation", "validation.parquet")]: + if split_name not in ds: + print(f"Skip split {split_name} (not in dataset).") + continue + split = ds[split_name] + rows = [] + max_n = args.max_train if split_name == "train" else args.max_val + n = len(split) if max_n <= 0 else min(len(split), max_n) + for i in range(n): + rows.append(_row_to_agent_r1(split[i], split_name, i)) + df = pd.DataFrame(rows) + path = os.path.join(out_dir, parquet_name) + df.to_parquet(path, index=False) + print(f"Wrote {n} rows -> {path}") + + if args.include_cross_eval: + configs = [name.strip() for name in args.cross_eval_configs.split(",") if name.strip()] + _write_cross_eval_parquets( + output_dir=out_dir, + hf_name=args.cross_eval_hf_name, + configs=configs, + split=args.cross_eval_split, + max_samples=args.max_cross_eval, + ) + + if not args.skip_corpus and ds is not None: + corpus_path = os.path.abspath(os.path.expanduser(args.corpus_output_path)) + os.makedirs(os.path.dirname(corpus_path), exist_ok=True) + all_examples: list[dict[str, Any]] = [] + for split_name, max_n in ( + ("train", args.max_train), + ("validation", args.max_val), + ): + if split_name not in ds: + continue + sp = ds[split_name] + n = len(sp) if max_n <= 0 else min(len(sp), max_n) + for i in range(n): + all_examples.append(sp[i]) + + entries = _contexts_to_corpus_entries(all_examples) + with open(corpus_path, "w", encoding="utf-8") as f: + for rec in entries: + f.write(json.dumps(rec, ensure_ascii=False) + "\n") + print(f"Wrote {len(entries)} deduplicated paragraphs -> {corpus_path}") + elif not args.skip_corpus: + print("Skip corpus generation because --skip_hotpotqa was set.") + + +if __name__ == "__main__": + main() diff --git a/recipe/hotpotqa/process_hotpotqa.py b/recipe/hotpotqa/process_hotpotqa.py new file mode 100644 index 0000000..8cd4745 --- /dev/null +++ b/recipe/hotpotqa/process_hotpotqa.py @@ -0,0 +1,98 @@ +#!/usr/bin/env python3 +from __future__ import annotations + +import argparse +import json +import os +from pathlib import Path + +import faiss +from FlagEmbedding import FlagAutoModel +import numpy as np + +from recipe.hotpotqa.utils import DEFAULT_HOTPOTQA_EMBEDDING_MODEL + +_REPO_ROOT = Path(__file__).resolve().parents[2] +_DEFAULT_DATA_DIR = _REPO_ROOT / "data" / "corpus" / "hotpotqa_corpus" +_DEFAULT_CORPUS_PATH = _DEFAULT_DATA_DIR / "hpqa_corpus.jsonl" + + +def _load_corpus_texts(corpus_path: Path) -> list[str]: + corpus: list[str] = [] + with corpus_path.open("r", encoding="utf-8") as f: + for line in f: + line = line.strip() + if not line: + continue + rec = json.loads(line) + corpus.append(f'{rec.get("title", "")} {rec.get("text", "")}'.strip()) + return corpus + + +def main() -> None: + parser = argparse.ArgumentParser(description="Build HotpotQA FAISS index (legacy-compatible).") + parser.add_argument("--data_dir", type=str, default=str(_DEFAULT_DATA_DIR)) + parser.add_argument("--corpus_path", type=str, default=str(_DEFAULT_CORPUS_PATH)) + parser.add_argument("--embedding_model", type=str, default=DEFAULT_HOTPOTQA_EMBEDDING_MODEL) + parser.add_argument( + "--devices", + type=str, + default="", + help='Embedding devices. Examples: "cuda:0", "cuda:0,cuda:1", "cpu". Empty means library default.', + ) + parser.add_argument("--batch_size", type=int, default=1024) + parser.add_argument( + "--reuse_embeddings", + action="store_true", + help="Reuse existing hpqa_corpus.npy if present and skip encode_corpus.", + ) + parser.add_argument( + "--query_instruction", + type=str, + default="Represent this sentence for searching relevant passages: ", + ) + args = parser.parse_args() + + data_dir = Path(args.data_dir) + data_dir.mkdir(parents=True, exist_ok=True) + corpus_path = Path(args.corpus_path) if args.corpus_path else _DEFAULT_CORPUS_PATH + emb_path = data_dir / "hpqa_corpus.npy" + index_path = data_dir / "index.bin" + + if not corpus_path.exists(): + raise SystemExit(f"Corpus not found: {corpus_path}") + + os.makedirs(str(data_dir), exist_ok=True) + vectors: np.ndarray + + if args.reuse_embeddings and emb_path.exists(): + print(f"[hotpotqa] reuse existing embeddings: {emb_path}") + vectors = np.load(str(emb_path)).astype(np.float32) + else: + corpus = _load_corpus_texts(corpus_path) + model_kwargs = { + "query_instruction_for_retrieval": args.query_instruction, + } + if args.devices.strip(): + devices = [x.strip() for x in args.devices.split(",") if x.strip()] + model_kwargs["devices"] = devices[0] if len(devices) == 1 else devices + print( + f"[hotpotqa] encoding corpus, n={len(corpus)}, batch_size={args.batch_size}, devices={args.devices or 'default'}" + ) + model = FlagAutoModel.from_finetuned(args.embedding_model, **model_kwargs) + try: + vectors = model.encode_corpus(corpus, batch_size=int(args.batch_size)) + except TypeError: + vectors = model.encode_corpus(corpus) + vectors = np.asarray(vectors, dtype=np.float32) + np.save(str(emb_path), vectors) + print(f"[hotpotqa] saved embeddings to {emb_path}") + + dim = vectors.shape[-1] + index = faiss.index_factory(dim, "Flat", faiss.METRIC_INNER_PRODUCT) + index.add(vectors) + faiss.write_index(index, str(index_path)) + print(f"[hotpotqa] saved index to {index_path}") + +if __name__ == "__main__": + main() \ No newline at end of file diff --git a/recipe/hotpotqa/prompts.py b/recipe/hotpotqa/prompts.py new file mode 100644 index 0000000..3d1e2fa --- /dev/null +++ b/recipe/hotpotqa/prompts.py @@ -0,0 +1,65 @@ +""" +HotpotQA prompts — same layout as `recipe/paper_search/prompts.py` (system + user +sections, Instructions, Output Format with `` / `` placeholders). +HotpotQA-only: `### Retrieved Passages`, `### Recent tool / format issues`, and +`` when finishing from current evidence. +""" + +HOTPOTQA_SYSTEM_PROMPT = "You are a research agent. Your goal is to answer the User Query using Wikipedia search evidence." + +HOTPOTQA_USER_PROMPT = """### User Query +{user_query} + +### History Actions +{history_actions} + +### Retrieved Passages +{passage_list} + +### Recent tool / format issues +{tool_feedback} + +### Instructions +Analyze the **Retrieved Passages** and **History Actions** to determine the next set of actions. Enclose your analysis of the state and decision logic within `...` tags. +**You support parallel tool calling.** You should output multiple tool calls in a single step if several independent actions are valuable at the current state. +**Attend to the history actions and avoid repeating the same search queries.** +When you can answer the question from the current passages, put the short final answer inside `` tags (no explanation) instead of further tool calls. + +### Output Format + +[Your analysis of the current state and decision logic...] + + +[Tool call 1] + + +[Tool call 2] + +... +""" + +SEARCH_TOOL_SCHEMA = { + "type": "function", + "function": { + "name": "search", + "description": ( + "Search Wikipedia for passages relevant to the user question. " + "Use natural-language or keyword queries; must differ from prior history queries when possible." + ), + "parameters": { + "type": "object", + "properties": { + "query": { + "type": "string", + "description": ( + "A single search query (natural language or keywords). " + "Must differ from all history queries when seeking new evidence." + ), + } + }, + "required": ["query"], + }, + }, +} + +HOTPOTQA_TOOL_SCHEMAS = [SEARCH_TOOL_SCHEMA] diff --git a/recipe/hotpotqa/requirements.txt b/recipe/hotpotqa/requirements.txt new file mode 100644 index 0000000..685e9bb --- /dev/null +++ b/recipe/hotpotqa/requirements.txt @@ -0,0 +1,7 @@ +# HotpotQA FAISS search tool + BGE embeddings (recipe/hotpotqa/utils.py) +faiss-cpu>=1.7.4 +FlagEmbedding>=1.2.0 +# Data prep (recipe/hotpotqa/prepare_hotpotqa_agent_r1.py): datasets, pyarrow, pandas +datasets +pandas +pyarrow diff --git a/recipe/hotpotqa/reward_fn.py b/recipe/hotpotqa/reward_fn.py new file mode 100644 index 0000000..e581c5b --- /dev/null +++ b/recipe/hotpotqa/reward_fn.py @@ -0,0 +1,117 @@ +import re +import string +from typing import Any + +from verl.utils.reward_score import default_compute_score + +_LOCAL_EM_DATA_SOURCES = { + "hotpotqa_distractor", + "2wikimultihopqa", + "musique", + "searchR1_hotpotqa", + "searchR1_2wikimultihopqa", + "searchR1_musique", +} + + +def _normalize_answer(s: str) -> str: + def lower(text: str) -> str: + return text.lower() + + def remove_punc(text: str) -> str: + exclude = set(string.punctuation) + return "".join(ch for ch in text if ch not in exclude) + + def remove_articles(text: str) -> str: + return re.sub(r"\b(a|an|the)\b", " ", text) + + def white_space_fix(text: str) -> str: + return " ".join(text.split()) + + return white_space_fix(remove_articles(remove_punc(lower(s)))) + + +def _extract_answer_from_solution(solution_str: str) -> str: + """ + Prefer content inside .... If not present, fall back to full string. + """ + pattern = r"(.*?)" + matches = list(re.finditer(pattern, solution_str, flags=re.DOTALL | re.IGNORECASE)) + if not matches: + return solution_str.strip() + return matches[-1].group(1).strip() + + +def _iter_ground_truths(ground_truth: Any) -> list[str]: + """Convert a ground-truth payload into answer strings. + + Args: + ground_truth: A string answer, a list/tuple of aliases, or another + scalar value. + + Returns: + A list of non-empty answer strings. + """ + if ground_truth is None: + return [] + if isinstance(ground_truth, str): + gt_str = ground_truth.strip() + return [gt_str] if gt_str else [] + if isinstance(ground_truth, (list, tuple, set)): + return [str(item).strip() for item in ground_truth if str(item).strip()] + gt_str = str(ground_truth).strip() + return [gt_str] if gt_str else [] + + +def _candidate_ground_truths(ground_truth: Any, extra_info: dict | None) -> list[str]: + """Collect primary and alias answers for EM scoring. + + Args: + ground_truth: Primary answer from ``reward_model.ground_truth``. + extra_info: Optional row metadata that may contain ``answers`` aliases. + + Returns: + Deduplicated answer strings in scoring order. + """ + candidates = _iter_ground_truths(ground_truth) + if isinstance(extra_info, dict): + candidates.extend(_iter_ground_truths(extra_info.get("answers"))) + + seen: set[str] = set() + deduped: list[str] = [] + for answer in candidates: + if answer in seen: + continue + seen.add(answer) + deduped.append(answer) + return deduped + + +def compute_score( + data_source: str, + solution_str: str, + ground_truth: Any, + extra_info: dict | None = None, + **kwargs, +) -> float: + """ + Custom reward function for HotpotQA. + + - If data_source is a HotpotQA-style QA source: use simple exact match + (EM) between predicted answer and one or more ground-truth answers. + - Otherwise, fall back to verl's default_compute_score. + """ + if data_source not in _LOCAL_EM_DATA_SOURCES: + # Delegate to built-in reward logic for other datasets if any. + return default_compute_score(data_source, solution_str, ground_truth, extra_info, **kwargs) + + ground_truths = _candidate_ground_truths(ground_truth, extra_info) + if not ground_truths: + return 0.0 + + pred = _extract_answer_from_solution(solution_str or "") + norm_pred = _normalize_answer(pred) + norm_gts = {_normalize_answer(answer) for answer in ground_truths} + + return 1.0 if norm_pred in norm_gts else 0.0 + diff --git a/recipe/hotpotqa/test_search_tool.py b/recipe/hotpotqa/test_search_tool.py new file mode 100644 index 0000000..02fc324 --- /dev/null +++ b/recipe/hotpotqa/test_search_tool.py @@ -0,0 +1,82 @@ +#!/usr/bin/env python3 +""" +Smoke test for HotpotQASearchToolLegacy (local FAISS + BGE). + +Run from repo root (same Python env as training): + + python recipe/hotpotqa/test_search_tool.py + python recipe/hotpotqa/test_search_tool.py "Who founded Apple?" + +Optional env vars (same as recipe/hotpotqa/utils.py): + HOTPOTQA_CORPUS_DATA_ROOT defaults to /data/corpus/hotpotqa_corpus + HOTPOTQA_EMBEDDING_DEVICE defaults to cpu; for BGE encoding e.g. cuda:0 (when not using HOTPOTQA_EMBEDDING_PER_WORKER_GPU) + HOTPOTQA_EMBEDDING_PER_WORKER_GPU if 1, agent worker i uses cuda:i (colocate with training GPUs) +""" +from __future__ import annotations + +import argparse +import json +import sys +from pathlib import Path + +_REPO_ROOT = Path(__file__).resolve().parents[2] +if str(_REPO_ROOT) not in sys.path: + sys.path.insert(0, str(_REPO_ROOT)) + +from recipe.hotpotqa.utils import ( # noqa: E402 + DEFAULT_HOTPOTQA_EMBEDDING_MODEL, + HOTPOTQA_CORPUS_JSONL, + HOTPOTQA_CORPUS_DATA_ROOT, + HOTPOTQA_DATA_ROOT, + HOTPOTQA_INDEX_BIN, + HotpotQASearchToolLegacy, + default_hotpotqa_embedding_device, +) + + +def main() -> None: + parser = argparse.ArgumentParser(description="Test HotpotQA FAISS search tool.") + parser.add_argument( + "query", + nargs="?", + default="What is the capital of France?", + help="Search query (natural language).", + ) + parser.add_argument( + "--embedding_model", + type=str, + default=DEFAULT_HOTPOTQA_EMBEDDING_MODEL, + help="Must match the model used when building index.bin.", + ) + args = parser.parse_args() + + print(f"HOTPOTQA_CORPUS_DATA_ROOT = {HOTPOTQA_CORPUS_DATA_ROOT}") + print(f"HOTPOTQA_DATA_ROOT alias = {HOTPOTQA_DATA_ROOT}") + print(f"index.bin exists = {HOTPOTQA_INDEX_BIN.exists()} ({HOTPOTQA_INDEX_BIN})") + print(f"hpqa_corpus.jsonl = {HOTPOTQA_CORPUS_JSONL.exists()} ({HOTPOTQA_CORPUS_JSONL})") + print(f"query = {args.query!r}") + print(f"embedding_model = {args.embedding_model}") + print(f"HOTPOTQA_EMBEDDING_DEVICE (env) = {default_hotpotqa_embedding_device()}") + print("---") + + tool = HotpotQASearchToolLegacy(embedding_model_name=args.embedding_model) + print(f"HotpotQASearchToolLegacy.embedding_devices (after normalize) = {tool.embedding_devices}") + out = tool.execute({"query": args.query}) + print(f"success = {out.get('success')}") + content = out.get("content", "") + if out.get("success"): + try: + payload = json.loads(content) + results = payload.get("results", []) + print(f"n_results = {len(results)}") + for i, text in enumerate(results, start=1): + snippet = (text[:200] + "…") if len(str(text)) > 200 else text + print(f" [{i}] {snippet}") + except json.JSONDecodeError: + print("content (raw):", content) + else: + print("content (error):", content) + + +if __name__ == "__main__": + main() diff --git a/recipe/hotpotqa/utils.py b/recipe/hotpotqa/utils.py new file mode 100644 index 0000000..f548138 --- /dev/null +++ b/recipe/hotpotqa/utils.py @@ -0,0 +1,352 @@ +import logging +import os +from dataclasses import dataclass, field +import json +from pathlib import Path +from typing import Any, List, Optional +import threading + +import faiss +import torch +from FlagEmbedding import FlagAutoModel +import numpy as np + +# Retrieval corpus root: defaults to /data/corpus/hotpotqa_corpus +# (index.bin + hpqa_corpus.jsonl). Override with HOTPOTQA_CORPUS_DATA_ROOT. +_STEPPO_ROOT = Path(__file__).resolve().parents[2] +_DEFAULT_HOTPOTQA_CORPUS_DATA_ROOT = _STEPPO_ROOT / "data" / "corpus" / "hotpotqa_corpus" + + +def resolve_hotpotqa_corpus_data_root(corpus_data_dir: Optional[str] = None) -> Path: + """Resolve the directory that stores HotpotQA retrieval files. + + Args: + corpus_data_dir: Optional explicit directory from recipe YAML. + + Returns: + Absolute path containing ``index.bin`` and ``hpqa_corpus.jsonl``. + """ + raw = ( + corpus_data_dir + or os.environ.get("HOTPOTQA_CORPUS_DATA_ROOT") + or str(_DEFAULT_HOTPOTQA_CORPUS_DATA_ROOT) + ) + return Path(raw).expanduser().resolve() + + +HOTPOTQA_CORPUS_DATA_ROOT = resolve_hotpotqa_corpus_data_root() +# Backward-compatible alias for older smoke tests/imports. New code should use HOTPOTQA_CORPUS_DATA_ROOT. +HOTPOTQA_DATA_ROOT = HOTPOTQA_CORPUS_DATA_ROOT +HOTPOTQA_INDEX_BIN = HOTPOTQA_CORPUS_DATA_ROOT / "index.bin" +# Passage text for decoding search hits; must match hpqa_corpus.jsonl used when building the index. +HOTPOTQA_CORPUS_JSONL = HOTPOTQA_CORPUS_DATA_ROOT / "hpqa_corpus.jsonl" +# hpqa_corpus.npy is only produced by process_hotpotqa for embedding cache; not loaded at runtime. + +# Default BGE checkpoint (local dir or Hugging Face hub id). Override via YAML `embedding_model_name` or +# `HOTPOTQA_EMBEDDING_MODEL` for portability. +DEFAULT_HOTPOTQA_EMBEDDING_MODEL = ( + os.environ.get("HOTPOTQA_EMBEDDING_MODEL", "BAAI/bge-large-en-v1.5").strip() + or "BAAI/bge-large-en-v1.5" +) + +logger = logging.getLogger(__name__) + + +def default_hotpotqa_embedding_device() -> str: + """Read desired device string from env (e.g. cuda:N); see `normalize_embedding_device` for actual resolution.""" + return os.environ.get("HOTPOTQA_EMBEDDING_DEVICE", "cpu").strip() or "cpu" + + +def normalize_embedding_device(requested: str) -> str: + """ + Resolve configured device to a FlagEmbedding-compatible `devices` string. + + Ray AgentFlowWorker processes often see **no GPUs** (no CUDA device assigned or visibility + masked). Passing `cuda:*` then raises "no CUDA GPUs are available" — we fall back to `cpu` + with a warning. + """ + dev = (requested or "cpu").strip() or "cpu" + if dev.lower() == "cpu": + return "cpu" + dl = dev.lower() + if dl.startswith("cuda"): + if not torch.cuda.is_available(): + logger.warning( + "embedding device %r requested but torch.cuda.is_available() is False; " + "using cpu (typical for Ray agent workers without GPU allocation).", + dev, + ) + return "cpu" + if ":" in dev: + try: + idx = int(dev.split(":")[-1]) + except ValueError: + return dev + if idx < 0 or idx >= torch.cuda.device_count(): + logger.warning( + "embedding device %r invalid for torch.cuda.device_count()=%s; using cpu.", + dev, + torch.cuda.device_count(), + ) + return "cpu" + return dev + return dev + + +def resolve_hotpotqa_embedding_devices( + embedding_devices: Optional[str], + agent_flow_worker_index: Optional[int], +) -> Optional[str]: + """Pick BGE `devices` for HotpotQAAgentFlow before constructing `HotpotQASearchToolLegacy`. + + Args: + embedding_devices: Non-empty value from YAML ``embedding_devices``; if set, it wins. + agent_flow_worker_index: Index of the Ray ``AgentFlowWorker`` (0 .. N-1). + + Returns: + Device string for FlagEmbedding, or ``None`` to use ``HotpotQASearchToolLegacy`` defaults + (``HOTPOTQA_EMBEDDING_DEVICE`` env, then ``cpu``). + + If environment variable ``HOTPOTQA_EMBEDDING_PER_WORKER_GPU`` is truthy (``1``/``true``/``yes``) + and ``agent_flow_worker_index`` is not ``None``, uses ``cuda:{index}`` (after + :func:`normalize_embedding_device`), so each of 4 workers can colocate BGE on its training GPU. + """ + if embedding_devices is not None: + s = str(embedding_devices).strip() + if s and s.lower() != "null": + return s + flag = os.environ.get("HOTPOTQA_EMBEDDING_PER_WORKER_GPU", "").strip().lower() + if flag in ("1", "true", "yes", "on") and agent_flow_worker_index is not None: + raw = f"cuda:{int(agent_flow_worker_index)}" + return normalize_embedding_device(raw) + return None + + +@dataclass +class Passage: + pid: int + title: str + text: str + score: float = 0.0 + + +@dataclass +class PassagePool: + passages: List[Passage] = field(default_factory=list) + + def has_passage(self, pid: int) -> bool: + return any(p.pid == pid for p in self.passages) + + def add_passage(self, passage: Passage) -> None: + # Dedupe by passage text; pid is the index in-pool (avoids fixed 0~4 ids across searches dropping hits). + if any(p.text == passage.text for p in self.passages): + return + pid = len(self.passages) + self.passages.append( + Passage(pid=pid, title=passage.title, text=passage.text, score=passage.score) + ) + + @property + def passage_list(self) -> str: + if not self.passages: + return "None" + lines = [] + for i, p in enumerate(self.passages, start=1): + snippet = p.text[:512].replace("\n", " ") + lines.append(f"[{i}] (id={p.pid}) {p.title}: {snippet}") + return "\n".join(lines) + + +class HotpotQASearchToolLegacy: + """ + Local HotpotQA FAISS retrieval; behavior matches the production-validated implementation in + `Agent-R1-legacy/agent_r1/tool/tools/search_tool.py` (SearchTool). + + Extensions vs. upstream (success-path semantics unchanged): + - Data layout: `HOTPOTQA_CORPUS_DATA_ROOT` + `index.bin` / `hpqa_corpus.jsonl` + - Process-wide shared index/corpus/model to avoid reloading every trajectory + - `_format_results` bounds-checks ids (legacy direct indexing could raise IndexError) + - If `encode_queries` returns torch.Tensor, converts via `.cpu().numpy()` for FAISS CPU index I/O + + Upstream often uses `FlagAutoModel.from_finetuned(..., devices="cpu")`; CPU encode is recommended + for training. If `HOTPOTQA_EMBEDDING_DEVICE=cuda:*`, call from the same thread (see hotpotqa_agent_flow). + """ + + _shared_lock = threading.RLock() + _shared_key: Optional[str] = None + _shared_index: Optional[faiss.Index] = None + _shared_corpus: Optional[list[str]] = None + _shared_model: Optional[FlagAutoModel] = None + + def __init__( + self, + embedding_model_name: str = DEFAULT_HOTPOTQA_EMBEDDING_MODEL, + query_instruction: str = "Represent this sentence for searching relevant passages: ", + embedding_devices: Optional[str] = None, + corpus_data_dir: Optional[str] = None, + ) -> None: + self.data_dir = resolve_hotpotqa_corpus_data_root(corpus_data_dir) + self.index_path = self.data_dir / "index.bin" + self.corpus_path = self.data_dir / "hpqa_corpus.jsonl" + self.embedding_model_name = embedding_model_name + self.query_instruction = query_instruction + raw = (embedding_devices if embedding_devices is not None else default_hotpotqa_embedding_device()).strip() or "cpu" + self.embedding_devices = normalize_embedding_device(raw) + if self.embedding_devices != raw: + logger.info("HotpotQASearchToolLegacy: effective embedding_devices=%r (from requested %r)", self.embedding_devices, raw) + + self._index: Optional[faiss.Index] = None + self._corpus: list[str] = [] + self._model: Optional[FlagAutoModel] = None + self._ensure_loaded() + + def __enter__(self): + self._ensure_loaded() + return self + + def __exit__(self, exc_type, exc, tb): + self.close() + + def _ensure_loaded(self) -> None: + cache_key = f"{self.data_dir}|{self.embedding_devices}|{self.embedding_model_name}" + with self.__class__._shared_lock: + if ( + self.__class__._shared_key != cache_key + or self.__class__._shared_index is None + or self.__class__._shared_corpus is None + or self.__class__._shared_model is None + ): + if not self.index_path.exists(): + raise FileNotFoundError(f"FAISS index not found: {self.index_path}") + if not self.corpus_path.exists(): + raise FileNotFoundError(f"Corpus file not found: {self.corpus_path}") + + logger.info("HotpotQASearchToolLegacy: loading FAISS index from %s", self.index_path) + index = faiss.read_index(str(self.index_path)) + logger.info( + "HotpotQASearchToolLegacy: loading corpus jsonl from %s (may take several minutes)", + self.corpus_path, + ) + corpus: list[str] = [] + with self.corpus_path.open("r", encoding="utf-8") as f: + for line in f: + line = line.strip() + if not line: + continue + rec = json.loads(line) + title = str(rec.get("title", "")) + text = str(rec.get("text", "")) + corpus.append(f"{title} {text}".strip()) + + logger.info( + "HotpotQASearchToolLegacy: loading FlagEmbedding model=%s devices=%s", + self.embedding_model_name, + self.embedding_devices, + ) + model = FlagAutoModel.from_finetuned( + self.embedding_model_name, + query_instruction_for_retrieval=self.query_instruction, + devices=self.embedding_devices, + ) + self.__class__._shared_key = cache_key + self.__class__._shared_index = index + self.__class__._shared_corpus = corpus + self.__class__._shared_model = model + + if int(index.ntotal) != len(corpus): + logger.warning( + "FAISS index.ntotal (%s) != hpqa_corpus.jsonl rows (%s). " + "Ids from search may be out of range and passages will be empty; " + "rebuild index.bin with the same jsonl or fix the corpus file.", + int(index.ntotal), + len(corpus), + ) + + self._index = self.__class__._shared_index + self._corpus = self.__class__._shared_corpus or [] + self._model = self.__class__._shared_model + + def close(self) -> None: + # Keep shared model/index alive for whole training process. + # This matches legacy behavior where SearchTool is initialized once and reused. + self._index = self.__class__._shared_index + self._corpus = self.__class__._shared_corpus or [] + self._model = self.__class__._shared_model + + def execute(self, args: dict[str, Any]) -> dict[str, Any]: + try: + query = str(args["query"]) + embeddings = self._encode_queries([query]) + assert self._index is not None + _, ids = self._index.search(embeddings, 5) + result_str = self._format_results(ids[0]) + return {"content": result_str, "success": True} + except Exception as e: + return {"content": str(e), "success": False} + + def batch_execute(self, args_list: list[dict[str, Any]]) -> list[dict[str, Any]]: + """Same contract as Agent-R1-legacy `SearchTool.batch_execute`: one encode + search; on failure each row gets str(e).""" + if not args_list: + return [] + try: + queries = [str(x["query"]) for x in args_list] + embeddings = self._encode_queries(queries) + assert self._index is not None + _, ids = self._index.search(embeddings, 5) + results_str = [self._format_results(ids[i]) for i in range(len(ids))] + return [{"content": result_str, "success": True} for result_str in results_str] + except Exception as e: + logger.warning( + "HotpotQASearchToolLegacy.batch_execute failed (%s queries): %s", + len(args_list), + e, + exc_info=True, + ) + return [{"content": str(e), "success": False} for _ in args_list] + + def _encode_queries(self, queries: list[str]) -> np.ndarray: + self._ensure_loaded() + with self.__class__._shared_lock: + assert self._model is not None + out = self._model.encode_queries(queries) + # FAISS CPU Index::search expects host float32 ndarray; BGE on GPU may return torch.Tensor. + if torch.is_tensor(out): + out = out.detach().float().cpu().numpy() + arr = np.asarray(out, dtype=np.float32) + if not arr.flags.c_contiguous: + arr = np.ascontiguousarray(arr) + return arr + + def _format_results(self, results) -> str: + results_list: list[str] = [] + row_ids = [int(x) for x in np.asarray(results, dtype=np.int64).reshape(-1)] + for result in row_ids: + if result < 0 or result >= len(self._corpus): + continue + results_list.append(self._corpus[result]) + if ( + not results_list + and self._corpus + and row_ids + and max(row_ids) >= len(self._corpus) + ): + logger.warning( + "FAISS returned ids %s but corpus length is %s; dropping all hits.", + row_ids[:10], + len(self._corpus), + ) + return json.dumps({"results": results_list}, ensure_ascii=False) + + +def parse_legacy_tool_result(content: str) -> list[Passage]: + """Parse legacy `{"results":[...]}` tool content into Passage list.""" + passages: list[Passage] = [] + try: + payload = json.loads(content) + results = payload.get("results", []) + for idx, text in enumerate(results): + text_str = str(text) + passages.append(Passage(pid=idx, title="", text=text_str, score=0.0)) + except Exception: + return [] + return passages diff --git a/recipe/hotpotqa/verify_dataset.py b/recipe/hotpotqa/verify_dataset.py new file mode 100644 index 0000000..58f89b6 --- /dev/null +++ b/recipe/hotpotqa/verify_dataset.py @@ -0,0 +1,156 @@ +#!/usr/bin/env python3 +""" +Verify HotpotQA parquet files for Agent-R1 training readiness. + +Checks: +1. Required columns exist (data_source, prompt, reward_model, extra_info) +2. Every row has a non-empty ground_truth answer +3. reward_fn.compute_score works correctly on sample data +4. data_source matches what reward_fn expects + +Usage: + python recipe/hotpotqa/verify_dataset.py \ + --train data/corpus/hotpotqa/train.parquet \ + --val data/corpus/hotpotqa/validation.parquet +""" +from __future__ import annotations + +import argparse +import sys +from pathlib import Path + +import pandas as pd + + +def verify_parquet(path: str, label: str) -> bool: + p = Path(path) + if not p.exists(): + print(f"[FAIL] {label}: file not found: {p}") + return False + + df = pd.read_parquet(p) + print(f"\n{'='*60}") + print(f"[{label}] {p} ({len(df)} rows, columns={list(df.columns)})") + print(f"{'='*60}") + + ok = True + + required = {"data_source", "prompt", "reward_model"} + missing = required - set(df.columns) + if missing: + print(f" [FAIL] missing columns: {missing}") + ok = False + else: + print(f" [OK] required columns present") + + # data_source + ds_values = df["data_source"].unique().tolist() + print(f" data_source values: {ds_values}") + if "hotpotqa_distractor" not in ds_values: + print(f" [WARN] 'hotpotqa_distractor' not in data_source — reward_fn will delegate to default") + + # prompt structure + sample_prompt = df["prompt"].iloc[0] + if isinstance(sample_prompt, list) and len(sample_prompt) > 0: + first_msg = sample_prompt[0] + if isinstance(first_msg, dict) and "role" in first_msg and "content" in first_msg: + print(f" [OK] prompt format correct (list of message dicts)") + print(f" sample question: {first_msg['content'][:100]}...") + else: + print(f" [FAIL] prompt[0] not a message dict: {type(first_msg)}") + ok = False + else: + print(f" [FAIL] prompt not a list of messages: {type(sample_prompt)}") + ok = False + + # ground_truth in reward_model + no_answer = 0 + empty_answer = 0 + answer_lengths = [] + for i, rm in enumerate(df["reward_model"]): + if not isinstance(rm, dict): + print(f" [FAIL] row {i}: reward_model is not a dict: {type(rm)}") + ok = False + continue + gt = rm.get("ground_truth") + if gt is None: + no_answer += 1 + elif not str(gt).strip(): + empty_answer += 1 + else: + answer_lengths.append(len(str(gt))) + + if no_answer > 0: + print(f" [FAIL] {no_answer} rows have ground_truth=None") + ok = False + else: + print(f" [OK] all rows have ground_truth present") + + if empty_answer > 0: + print(f" [WARN] {empty_answer} rows have empty ground_truth (will get reward=0)") + else: + print(f" [OK] all ground_truth non-empty") + + if answer_lengths: + avg = sum(answer_lengths) / len(answer_lengths) + print(f" answer stats: min={min(answer_lengths)}, max={max(answer_lengths)}, avg={avg:.1f} chars") + + # Test reward_fn on a few samples + try: + from recipe.hotpotqa.reward_fn import compute_score, _normalize_answer + + sample_rm = df["reward_model"].iloc[0] + gt = sample_rm["ground_truth"] + # exact match + score_match = compute_score("hotpotqa_distractor", f"{gt}", gt) + # wrong answer + score_wrong = compute_score("hotpotqa_distractor", "WRONG_ANSWER_XYZ", gt) + # no answer tag + score_notag = compute_score("hotpotqa_distractor", gt, gt) + + print(f"\n reward_fn test (gt={gt!r}):") + print(f" exact match with tag: {score_match}") + print(f" wrong answer: {score_wrong}") + print(f" no tag (raw text): {score_notag}") + + if score_match != 1.0: + print(f" [FAIL] exact match should be 1.0 but got {score_match}") + ok = False + else: + print(f" [OK] reward_fn works correctly") + except Exception as e: + print(f" [WARN] could not import/test reward_fn: {e}") + + # Show first 3 samples + print(f"\n --- first 3 samples ---") + for i in range(min(3, len(df))): + q = df["prompt"].iloc[i][0]["content"][:80] + a = df["reward_model"].iloc[i]["ground_truth"] + print(f" [{i}] Q: {q}...") + print(f" A: {a}") + + return ok + + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument("--train", default="data/corpus/hotpotqa/train.parquet") + parser.add_argument("--val", default="data/corpus/hotpotqa/validation.parquet") + args = parser.parse_args() + + all_ok = True + for path, label in [(args.train, "TRAIN"), (args.val, "VAL")]: + if not verify_parquet(path, label): + all_ok = False + + print(f"\n{'='*60}") + if all_ok: + print("ALL CHECKS PASSED — dataset ready for training") + else: + print("SOME CHECKS FAILED — fix issues above before training") + print(f"{'='*60}") + sys.exit(0 if all_ok else 1) + + +if __name__ == "__main__": + main() diff --git a/recipe/paper_search/__init__.py b/recipe/paper_search/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/recipe/paper_search/base.yaml b/recipe/paper_search/base.yaml new file mode 100644 index 0000000..973b2a7 --- /dev/null +++ b/recipe/paper_search/base.yaml @@ -0,0 +1,7 @@ +- name: paper_search_agent + _target_: recipe.paper_search.paper_search_agent_flow.PaperSearchAgentFlow + max_steps: 5 + max_parallel_calls: 5 + search_top_k: 10 + citations_limit: 30 + references_limit: -1 diff --git a/recipe/paper_search/env/run_papersearch_selector_service.sh b/recipe/paper_search/env/run_papersearch_selector_service.sh new file mode 100755 index 0000000..34e2d2f --- /dev/null +++ b/recipe/paper_search/env/run_papersearch_selector_service.sh @@ -0,0 +1,41 @@ +#!/usr/bin/env bash +set -euo pipefail + +SCRIPT_NAME="$(basename "$0" .sh)" +LOG_ROOT="${LOG_ROOT:-$(pwd)/logs}" +LOG_DIR="${LOG_DIR:-$LOG_ROOT/papersearch}" +mkdir -p "$LOG_DIR" +TIMESTAMP="$(date -u +%Y%m%d_%H%M%S)" +LOG_FILE="${LOG_FILE:-$LOG_DIR/${SCRIPT_NAME}_${TIMESTAMP}.log}" + +exec > >(tee -a "$LOG_FILE") 2>&1 +echo "Logging to $LOG_FILE" +set -x + +export CUDA_VISIBLE_DEVICES=${CUDA_VISIBLE_DEVICES:-0} +export VLLM_USE_V1=1 +export HF_ENDPOINT=${HF_ENDPOINT:-https://hf-mirror.com} + +SELECTOR_MODEL_PATH=${PAPERSEARCH_SELECTOR_MODEL_PATH:-./checkpoints/selector_Qwen3_8B} +SELECTOR_MODEL_NAME=${PAPERSEARCH_SELECTOR_MODEL_NAME:-selector-qwen-8b} +SELECTOR_HOST=${PAPERSEARCH_SELECTOR_HOST:-0.0.0.0} +SELECTOR_PORT=${PAPERSEARCH_SELECTOR_PORT:-8000} +SELECTOR_DTYPE=${PAPERSEARCH_SELECTOR_DTYPE:-bfloat16} +SELECTOR_TENSOR_PARALLEL_SIZE=${PAPERSEARCH_SELECTOR_TENSOR_PARALLEL_SIZE:-1} +SELECTOR_GPU_MEMORY_UTILIZATION=${PAPERSEARCH_SELECTOR_GPU_MEMORY_UTILIZATION:-0.85} +SELECTOR_MAX_MODEL_LEN=${PAPERSEARCH_SELECTOR_MAX_MODEL_LEN:-32768} + +echo "Selector service target: http://localhost:${SELECTOR_PORT}" +echo "Set PAPERSEARCH_SELECTOR_BASE_URL=http://localhost:${SELECTOR_PORT} before training." + +vllm serve "$SELECTOR_MODEL_PATH" \ + --runner pooling \ + --host "$SELECTOR_HOST" \ + --port "$SELECTOR_PORT" \ + --served-model-name "$SELECTOR_MODEL_NAME" \ + --dtype "$SELECTOR_DTYPE" \ + --tensor-parallel-size "$SELECTOR_TENSOR_PARALLEL_SIZE" \ + --gpu-memory-utilization "$SELECTOR_GPU_MEMORY_UTILIZATION" \ + --max-model-len "$SELECTOR_MAX_MODEL_LEN" \ + --trust-remote-code \ + "$@" diff --git a/recipe/paper_search/http_retry.py b/recipe/paper_search/http_retry.py new file mode 100644 index 0000000..759787b --- /dev/null +++ b/recipe/paper_search/http_retry.py @@ -0,0 +1,211 @@ +import asyncio +import random +import time +from typing import Any, Iterable, Optional + +import httpx + +_DEFAULT_RETRY_STATUS_CODES = {429, 500, 502, 503, 504} + + +def _compute_backoff_seconds( + attempt: int, + *, + initial_backoff: float, + max_backoff: float, + jitter: float, +) -> float: + """ + Exponential backoff with jitter. + attempt=0 means first retry wait, attempt=1 means second retry wait, ... + """ + base = min(max_backoff, initial_backoff * (2**attempt)) + if jitter <= 0: + return base + lo = max(0.0, 1.0 - jitter) + hi = 1.0 + jitter + return base * random.uniform(lo, hi) + + +def _parse_retry_after_seconds(headers: httpx.Headers) -> Optional[float]: + # Retry-After can be seconds or an HTTP date; we only support seconds here. + value = headers.get("Retry-After") + if not value: + return None + try: + return float(value) + except ValueError: + return None + + +async def httpx_request_with_retry( + client: httpx.AsyncClient, + method: str, + url: str, + *, + semaphore: Optional[asyncio.Semaphore] = None, + max_retries: int = 3, + retry_status_codes: Iterable[int] = _DEFAULT_RETRY_STATUS_CODES, + retry_exceptions: tuple[type[BaseException], ...] = (httpx.RequestError, httpx.TimeoutException), + initial_backoff: float = 0.5, + max_backoff: float = 8.0, + jitter: float = 0.2, + **kwargs: Any, +) -> httpx.Response: + """ + Wrap httpx.AsyncClient.request with retries. + + Fix: Always release the connection on success by reading the full body and closing the response. + This prevents connection pool exhaustion caused by unclosed responses. + + Notes: + - Retries on network/timeout exceptions, and on retry_status_codes (e.g. 429/5xx). + - max_retries means "extra tries" (total attempts = max_retries + 1). + - If server returns Retry-After (seconds), it takes precedence. + - If caller passes stream=True, we will NOT auto-read/close; caller must manage the response. + """ + retry_status_set = set(retry_status_codes) + last_exc: Optional[BaseException] = None + stream = bool(kwargs.get("stream", False)) + + for attempt in range(max_retries + 1): + try: + if semaphore is not None: + await semaphore.acquire() + try: + resp = await client.request(method, url, **kwargs) + finally: + if semaphore is not None: + semaphore.release() + + if resp.status_code in retry_status_set and attempt < max_retries: + # Ensure connection can be reused before retrying. + try: + await resp.aread() + finally: + await resp.aclose() + + retry_after = _parse_retry_after_seconds(resp.headers) + delay = ( + retry_after + if retry_after is not None + else _compute_backoff_seconds( + attempt, + initial_backoff=initial_backoff, + max_backoff=max_backoff, + jitter=jitter, + ) + ) + await asyncio.sleep(delay) + continue + + if stream: + return resp + + # Success path: fully consume and close so the connection is released back to the pool. + try: + await resp.aread() + finally: + await resp.aclose() + return resp + except retry_exceptions as e: + last_exc = e + if attempt >= max_retries: + raise + delay = _compute_backoff_seconds( + attempt, + initial_backoff=initial_backoff, + max_backoff=max_backoff, + jitter=jitter, + ) + await asyncio.sleep(delay) + + assert last_exc is not None + raise last_exc + + +def requests_request_with_retry( + method: str, + url: str, + *, + max_retries: int = 3, + retry_status_codes: Iterable[int] = _DEFAULT_RETRY_STATUS_CODES, + retry_exceptions: tuple[type[BaseException], ...] = (), + initial_backoff: float = 0.5, + max_backoff: float = 8.0, + jitter: float = 0.2, + timeout: Optional[float] = 10.0, + **kwargs: Any, +): + """ + Wrap requests.request with retries. + + Fix: Always release the connection on success by reading the full body and closing the response. + If caller passes stream=True, we will NOT auto-read/close; caller must manage the response. + """ + import requests # lazy import + + retry_status_set = set(retry_status_codes) + if not retry_exceptions: + retry_exceptions = (requests.exceptions.RequestException,) + + last_exc: Optional[BaseException] = None + stream = bool(kwargs.get("stream", False)) + + for attempt in range(max_retries + 1): + try: + resp = requests.request(method, url, timeout=timeout, **kwargs) + if resp.status_code in retry_status_set and attempt < max_retries: + retry_after = None + ra = resp.headers.get("Retry-After") + if ra: + try: + retry_after = float(ra) + except ValueError: + retry_after = None + resp.close() + delay = ( + retry_after + if retry_after is not None + else _compute_backoff_seconds( + attempt, + initial_backoff=initial_backoff, + max_backoff=max_backoff, + jitter=jitter, + ) + ) + time.sleep(delay) + continue + + if not stream: + # Read and close so the underlying urllib3 connection returns to the pool. + _ = resp.content + resp.close() + return resp + except retry_exceptions as e: + last_exc = e + if attempt >= max_retries: + raise + delay = _compute_backoff_seconds( + attempt, + initial_backoff=initial_backoff, + max_backoff=max_backoff, + jitter=jitter, + ) + time.sleep(delay) + + assert last_exc is not None + raise last_exc + + +def requests_json_with_retry( + method: str, + url: str, + *, + max_retries: int = 3, + timeout: Optional[float] = 10.0, + **kwargs: Any, +) -> Any: + resp = requests_request_with_retry(method, url, max_retries=max_retries, timeout=timeout, **kwargs) + resp.raise_for_status() + return resp.json() diff --git a/recipe/paper_search/inference/.env.example b/recipe/paper_search/inference/.env.example new file mode 100644 index 0000000..7b21421 --- /dev/null +++ b/recipe/paper_search/inference/.env.example @@ -0,0 +1,32 @@ +# Copy to `.env` locally and fill in machine-specific values. + +# --- vLLM policy --- +PAPER_SEARCH_INFERENCE_MODEL_PATH=Qwen/Qwen3-4B-Instruct-2507 +VLLM_TENSOR_PARALLEL_SIZE=1 +VLLM_GPU_MEMORY_UTILIZATION=0.85 +VLLM_MAX_MODEL_LEN=10240 +PAPER_SEARCH_INFERENCE_MAX_NEW_TOKENS=4096 +PAPER_SEARCH_INFERENCE_TEMPERATURE=0.0 +PAPER_SEARCH_TOOL_PARSER=hermes + +# Optional tokenizer.apply_chat_template kwargs, e.g. {"enable_thinking": false} +# PAPER_SEARCH_APPLY_CHAT_TEMPLATE_KWARGS= + +# --- Paper retrieval --- +PAPER_SEARCH_BASE_URL=http://localhost:4000 +PAPER_AGENT_V2_SEARCH_SOURCE=serper +# PAPER_SEARCH_SERPER_API_KEYS= +# SERPER_API_KEY= +# SERPER_SEARCH_URL= +PAPER_AGENT_V2_PAPER_FROM= +PAPER_AGENT_V2_PAPER_TO= + +# --- Selector HTTP API --- +PAPERSEARCH_SELECTOR_BASE_URL=http://localhost:8000 +PAPERSEARCH_SELECTOR_MODEL_NAME=selector-qwen-8b + +# --- Batch IO --- +PAPER_AGENT_V2_DATASET=recipe/paper_search/inference/datasets/test_case.jsonl +PAPER_AGENT_V2_SAVE_DIR=outputs/paper_search +PAPER_AGENT_V2_RETRY_ROUNDS=1 +PAPER_SEARCH_INFERENCE_MAX_STEPS=5 diff --git a/recipe/paper_search/inference/__init__.py b/recipe/paper_search/inference/__init__.py new file mode 100644 index 0000000..e58e812 --- /dev/null +++ b/recipe/paper_search/inference/__init__.py @@ -0,0 +1 @@ +"""Offline batch inference for paper search agent (vLLM + optional Serper).""" diff --git a/recipe/paper_search/inference/datasets/AutoScholarQuery/dev.jsonl b/recipe/paper_search/inference/datasets/AutoScholarQuery/dev.jsonl new file mode 100644 index 0000000..60bc443 --- /dev/null +++ b/recipe/paper_search/inference/datasets/AutoScholarQuery/dev.jsonl @@ -0,0 +1,1000 @@ +{"question": "Could you list the works exploit knowledge from pre-trained vision-language models for text-guided queries in 3D scenes?", "answer": ["CLIP-Fields: Weakly Supervised Semantic Fields for Robotic Memory", "ConceptFusion: Open-set Multimodal 3D Mapping", "OpenScene: 3D Scene Understanding with Open Vocabularies"], "answer_arxiv_id": ["2210.05663", "2302.07241", "2211.15654"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_dev_0"} +{"question": "Could you provide me some works related to representer theorems in machine learning?", "answer": ["A representer theorem for deep kernel learning", "A representer theorem for deep neural networks"], "answer_arxiv_id": ["1709.10441", "1802.09210"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_1"} +{"question": "Which studies discuss the effects and mechanisms of weight decay regularization in machine learning?", "answer": ["Three Mechanisms of Weight Decay Regularization", "L2 Regularization versus Batch and Weight Normalization"], "answer_arxiv_id": ["1810.12281", "1706.05350v1"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_dev_2"} +{"question": "Could you mention some papers that propose neural algorithms for f-divergence regularized costs?", "answer": ["Stochastic Optimization for Large-scale Optimal Transport", "Large-Scale Optimal Transport and Mapping Estimation", "Score-based Generative Neural Networks for Large-Scale Optimal Transport"], "answer_arxiv_id": ["1605.08527", "1711.02283", "2110.03237"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_dev_3"} +{"question": "What works implemented temporal pixel-wise audio-visual interaction module in FCN-based methods for AVS?", "answer": ["Audio-Visual Segmentation"], "answer_arxiv_id": ["2207.05042"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_dev_4"} +{"question": "Which papers describe work aimed at increasing the faithfulness of Large Language Models (LLMs) by changing the prediction generation method?", "answer": ["Stay on topic with Classifier-Free Guidance", "Selection-Inference: Exploiting Large Language Models for Interpretable\n Logical Reasoning", "Question Decomposition Improves the Faithfulness of Model-Generated\n Reasoning", "Faithful Chain-of-Thought Reasoning", "Faithful Explanations of Black-box NLP Models Using LLM-generated\n Counterfactuals"], "answer_arxiv_id": ["2306.17806", "2205.09712", "2307.11768", "2301.13379", "2310.00603"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_5"} +{"question": "Which works focus on face verification models and discuss their performance in unconstrained environments?", "answer": ["ArcFace: Additive Angular Margin Loss for Deep Face Recognition", "CurricularFace: Adaptive Curriculum Learning Loss for Deep Face\n Recognition", "AdaFace: Quality Adaptive Margin for Face Recognition"], "answer_arxiv_id": ["1801.07698", "2004.00288", "2204.00964"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_dev_6"} +{"question": "Any works about hallucination assessment in GPT-4V?", "answer": ["Holistic Analysis of Hallucination in GPT-4V(ision): Bias and\n Interference Challenges"], "answer_arxiv_id": ["2311.03287"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_dev_7"} +{"question": "Which papers discussed alternative resolution schemes related to unit scaling?", "answer": ["High-Performance Large-Scale Image Recognition Without Normalization"], "answer_arxiv_id": ["2102.06171"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_dev_8"} +{"question": "What studies have shown the global convergence of Gradient Descent (GD) for simple linear networks and two-layer networks?", "answer": ["SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data", "The Implicit Bias of Gradient Descent on Separable Data", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks"], "answer_arxiv_id": ["1710.10174", "1710.10345", "1901.08584"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_dev_9"} +{"question": "What papers discuss the use of knowledge distillation for model compression?", "answer": ["Distilling Task-Specific Knowledge from BERT into Simple Neural Networks", "TinyBERT: Distilling BERT for Natural Language Understanding", "MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices"], "answer_arxiv_id": ["1903.12136", "1909.10351", "2004.02984"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_dev_10"} +{"question": "Could you tell me about the research that revealed different variants of pretrained transformer models?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_dev_11"} +{"question": "What research studies have been cited for applying the CLIP model for various downstream applications like image-based object detection, segmentation, and video applications?", "answer": ["Bridging the Gap between Object and Image-level Representations for\n Open-Vocabulary Detection", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Language-Grounded Indoor 3D Semantic Segmentation in the Wild", "Expanding Language-Image Pretrained Models for General Video Recognition", "ActionCLIP: A New Paradigm for Video Action Recognition"], "answer_arxiv_id": ["2207.03482", "2210.04150", "2204.07761", "2208.02816", "2109.08472"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_dev_12"} +{"question": "Which studies indicate a natural conflict between adversarial robustness and standard accuracy?", "answer": ["Robustness May Be at Odds with Accuracy", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1805.12152", "1901.08573"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_13"} +{"question": "Which paper introduced the concept of dataset distillation?", "answer": ["Dataset Distillation"], "answer_arxiv_id": ["1811.10959"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_14"} +{"question": "Which papers tried to accelerate Neural Differential Equations models using higher-order regularization terms?", "answer": ["Learning differential equations that are easy to solve"], "answer_arxiv_id": ["2007.04504"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_dev_15"} +{"question": "In what works are SMLD diffusion models considered?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_dev_16"} +{"question": "Which works use standard fine-tuning to attempt to localize edits in parameter updating methods?", "answer": ["Editing Factual Knowledge in Language Models", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2104.08164", "2110.11309"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_dev_17"} +{"question": "Could you provide me some studies performed on adversarial training?", "answer": ["CyCADA: Cycle-Consistent Adversarial Domain Adaptation", "Self-Ensembling with GAN-based Data Augmentation for Domain Adaptation\n in Semantic Segmentation", "Learning from Synthetic Data: Addressing Domain Shift for Semantic\n Segmentation", "Learning to Adapt Structured Output Space for Semantic Segmentation", "Both Style and Distortion Matter: Dual-Path Unsupervised Domain\n Adaptation for Panoramic Semantic Segmentation"], "answer_arxiv_id": ["1711.03213", "1909.00589", "1711.06969", "1802.10349", "2303.14360"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_dev_18"} +{"question": "Which papers worked on lower bound theory for ReLU Networks?", "answer": ["The Expressive Power of Neural Networks: A View from the Width", "Approximating Continuous Functions by ReLU Nets of Minimal Width"], "answer_arxiv_id": ["1709.02540", "1710.11278"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_dev_19"} +{"question": "Are there any studies which used conditional Gaussian distributions over feature spaces in the context of adversarial robustness?", "answer": ["Max-Mahalanobis Linear Discriminant Analysis Networks", "Shaping Deep Feature Space towards Gaussian Mixture for Visual\n Classification"], "answer_arxiv_id": ["1802.09308", "2011.09066"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_dev_20"} +{"question": "What research focused on the development of sequential action understanding datasets?", "answer": ["SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos"], "answer_arxiv_id": ["1804.04527"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_dev_21"} +{"question": "Which papers cover information-sharing strategies to mitigate heterogeneity in Federated Learning?", "answer": ["Federated Learning with Non-IID Data", "Communication-Efficient On-Device Machine Learning: Federated Distillation and Augmentation under Non-IID Private Data", "G-PATE: Scalable Differentially Private Data Generator via Private Aggregation of Teacher Discriminators"], "answer_arxiv_id": ["1806.00582", "1811.11479", "1906.09338v2"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_dev_22"} +{"question": "What were the methods proposed for learning the prompt from downstream data in continual input embedding space?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2101.00190"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_dev_23"} +{"question": "Could you provide me some works focused on 3D diffusion models based on implicit fields?", "answer": ["3D Neural Field Generation using Triplane Diffusion", "LION: Latent Point Diffusion Models for 3D Shape Generation", "Diffusion-SDF: Text-to-Shape via Voxelized Diffusion", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "Neural Wavelet-domain Diffusion for 3D Shape Generation, Inversion, and\n Manipulation", "3DQD: Generalized Deep 3D Shape Prior via Part-Discretized Diffusion\n Process", "HyperDiffusion: Generating Implicit Neural Fields with Weight-Space\n Diffusion"], "answer_arxiv_id": ["2211.16677", "2210.06978", "2212.03293", "2212.04493", "2302.00190", "2303.10406", "2303.17015"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_24"} +{"question": "Any research on building a prompting pipeline where the LLM reasons over the extracted KG subgraphs?", "answer": ["MindMap: Knowledge Graph Prompting Sparks Graph of Thoughts in Large\n Language Models"], "answer_arxiv_id": ["2308.09729"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_dev_25"} +{"question": "Who proposed improvements to the accuracy of optimization-based attacks using different image priors?", "answer": ["iDLG: Improved Deep Leakage from Gradients", "Inverting Gradients - How easy is it to break privacy in federated learning?", "See through Gradients: Image Batch Recovery via GradInversion"], "answer_arxiv_id": ["2001.02610", "2003.14053", "2104.07586"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_26"} +{"question": "Which studies are related to the empowerment of modern natural language processing systems by text embedders?", "answer": ["Dense Passage Retrieval for Open-Domain Question Answering"], "answer_arxiv_id": ["2004.04906"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_dev_27"} +{"question": "Could you provide me some works that lean on uncertainty or diversity criteria for their selection strategies?", "answer": ["LiDAL: Inter-frame Uncertainty Based Active Learning for 3D LiDAR Semantic Segmentation", "Dirichlet-based Uncertainty Calibration for Active Domain Adaptation", "Active Learning for Point Cloud Semantic Segmentation via Spatial-Structural Diversity Reasoning"], "answer_arxiv_id": ["2211.05997", "2302.13824", "2202.12588"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_dev_28"} +{"question": "What work proposed a model that estimates pixel-wise weights for pre-specified WB presets?", "answer": ["Auto White-Balance Correction for Mixed-Illuminant Scenes"], "answer_arxiv_id": ["2109.08750"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_29"} +{"question": "Which papers introduced the use of VAEs and 3D convolutional networks to generate voxelized molecules?", "answer": ["Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models"], "answer_arxiv_id": ["2010.08687"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_30"} +{"question": "Which research papers propose motion trajectory conditioned on scene image?", "answer": ["Long-term Human Motion Prediction with Scene Context"], "answer_arxiv_id": ["2007.03672"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_dev_31"} +{"question": "What is a noteworthy study that discusses Federated Learning with cyclic client participation?", "answer": ["On the Convergence of Federated Averaging with Cyclic Client Participation"], "answer_arxiv_id": ["2302.03109v1"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_32"} +{"question": "Can you provide references where rewards are shaped by training a reinforcement learning agent to learn and complete intermediate tasks guided by language?", "answer": ["Using Natural Language for Reward Shaping in Reinforcement Learning", "EAGER: Asking and Answering Questions for Automatic Reward Shaping in Language-guided RL", "ELLA: Exploration through Learned Language Abstraction"], "answer_arxiv_id": ["1903.02020", "2206.09674", "2103.05825"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_33"} +{"question": "Which work first introduced the idea of converting visual features into readable embeddings for LLMs?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models"], "answer_arxiv_id": ["2106.13884"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_34"} +{"question": "What researches used averaging and max pooling in feature aggregation on a multi-view rendering-based method?", "answer": ["Learning Local Shape Descriptors from Part Correspondences With\n Multi-view Convolutional Networks", "Self-Supervised Learning with Multi-View Rendering for 3D Point Cloud\n Analysis"], "answer_arxiv_id": ["1706.04496", "2210.15904"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_35"} +{"question": "Which studies proposed to increase the representation ability of quantization by replacing uniform quantization with non-uniform quantization?", "answer": ["Nonuniform-to-Uniform Quantization: Towards Accurate Quantization via Generalized Straight-Through Estimation", "Additive Powers-of-Two Quantization: An Efficient Non-uniform Discretization for Neural Networks", "LQ-Nets: Learned Quantization for Highly Accurate and Compact Deep Neural Networks"], "answer_arxiv_id": ["2111.14826", "1909.13144", "1807.10029v1"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_dev_36"} +{"question": "What research suggests that duplicate examples can hurt the performance in document retrieval?", "answer": ["Deduplicating Training Data Makes Language Models Better", "Scaling Laws and Interpretability of Learning from Repeated Data"], "answer_arxiv_id": ["2107.06499", "2205.10487"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_dev_37"} +{"question": "Could you list some works that tried to re-introduce hierarchical designs into transformer?", "answer": ["Multiscale Vision Transformers", "Video Swin Transformer"], "answer_arxiv_id": ["2104.11227", "2106.13230"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_38"} +{"question": "Can you name the studies that used Counterfactual examples as data augmentation in Natural Language Processing (NLP)?", "answer": ["Learning the Difference that Makes a Difference with\n Counterfactually-Augmented Data", "Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and\n Improving Models", "Evaluating Models' Local Decision Boundaries via Contrast Sets"], "answer_arxiv_id": ["1909.12434", "2101.00288", "2004.02709"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_dev_39"} +{"question": "What are some works that have improved the theoretical convergence rate of local stochastic gradient descent ascent algorithms in federated learning?", "answer": ["FedNest: Federated Bilevel, Minimax, and Compositional Optimization"], "answer_arxiv_id": ["2205.02215v3"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_dev_40"} +{"question": "What studies structured the learned RL algorithm as a black box using a neural network as a general purpose sequence model in meta-RL methods?", "answer": ["RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "Learning to reinforcement learn", "A Simple Neural Attentive Meta-Learner", "Generalization of Reinforcement Learners with Working and Episodic Memory", "Rapid Task-Solving in Novel Environments", "Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPs"], "answer_arxiv_id": ["1611.02779", "1611.05763", "1707.03141", "1910.13406", "2006.03662", "2110.05038"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_41"} +{"question": "What papers proposed text data augmentation techniques like synonym replacement, positional swaps and back translation?", "answer": ["Improving Neural Machine Translation Models with Monolingual Data"], "answer_arxiv_id": ["1511.06709"], "source_meta": {"published_time": "20220226"}, "qid": "AutoScholarQuery_dev_42"} +{"question": "What studies propose the usage of diffusion models for processing graph data?", "answer": ["Diffusion Models for Graphs Benefit From Discrete State Spaces", "Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations"], "answer_arxiv_id": ["2210.01549", "2202.02514"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_dev_43"} +{"question": "Can you show some applications of diffusion models in zero-shot classification and supervised segmentation?", "answer": ["Your Diffusion Model is Secretly a Zero-Shot Classifier", "SegDiff: Image Segmentation with Diffusion Probabilistic Models"], "answer_arxiv_id": ["2303.16203", "2112.00390"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_dev_44"} +{"question": "Any works on breaking the representational symmetry based on spatial coordinates or specific object types?", "answer": ["SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition"], "answer_arxiv_id": ["2001.02407"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_45"} +{"question": "What works talk about the LLM-Blender, which uses a pair-ranker model for optimal LLM output selection?", "answer": ["LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and\n Generative Fusion"], "answer_arxiv_id": ["2306.02561"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_dev_46"} +{"question": "What studies demonstrated the capability of LLMs to provide chain-of-thought explanations that elucidate their reasoning processes?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2201.11903", "2205.11916"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_47"} +{"question": "Which works focused on designing interventions for causal discovery?", "answer": ["Interventions, Where and How? Experimental Design for Causal Models at Scale"], "answer_arxiv_id": ["2203.02016"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_48"} +{"question": "What research focused on deploying the PW-learner and the RA-learner in the estimation of the CATE?", "answer": ["Nonparametric Estimation of Heterogeneous Treatment Effects: From Theory to Learning Algorithms"], "answer_arxiv_id": ["2101.10943v2"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_dev_49"} +{"question": "Which study approaches the problem of convergence rates of classic TD from the perspective of Ordinary Differential Equations (ODE) analysis?", "answer": ["Finite-Time Error Bounds For Linear Stochastic Approximation and TD Learning"], "answer_arxiv_id": ["1902.00923v3"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_50"} +{"question": "Which works focus on the application of Bayesian Optimization in the area of global non-convex optimization?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design", "Convergence rates of efficient global optimization algorithms", "Thompson Sampling for Contextual Bandits with Linear Payoffs", "On Lower Bounds for Standard and Robust Gaussian Process Bandit Optimization"], "answer_arxiv_id": ["0912.3995", "1101.3501v3", "1209.3352", "2008.08757"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_dev_51"} +{"question": "Which research contains over thousands of single-choice questions covering numerous different ability dimensions?", "answer": ["MMBench: Is Your Multi-modal Model an All-around Player?"], "answer_arxiv_id": ["2307.06281"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_52"} +{"question": "Which study used a weak supervision type, foreground mask, as a substitute for costly 3D CAD annotations?", "answer": ["Weakly supervised 3D Reconstruction with Adversarial Constraint"], "answer_arxiv_id": ["1705.10904"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_dev_53"} +{"question": "What are the studies manipulate the generation process of a pre-trained model to implicitly control the generated content?", "answer": ["Diffusion-based Image Translation using Disentangled Style and Content Representation", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image Translation", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Blended Diffusion for Text-driven Editing of Natural Images", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "Null-text Inversion for Editing Real Images using Guided Diffusion Models", "DiffEdit: Diffusion-based semantic image editing with mask guidance", "Leveraging Off-the-shelf Diffusion Model for Multi-attribute Fashion Image Manipulation", "Diffusion Models already have a Semantic Latent Space"], "answer_arxiv_id": ["2209.15264", "2211.12572", "2208.01626", "2111.14818", "2108.02938", "2211.09794", "2210.11427", "2210.05872", "2210.10960v2"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_dev_54"} +{"question": "What research works explore human bodies rendering through multi-view videos?", "answer": ["Neural Human Video Rendering by Learning Dynamic Textures and\n Rendering-to-Video Translation", "Neural Articulated Radiance Field", "Vid2Actor: Free-viewpoint Animatable Person Synthesis from Video in the\n Wild"], "answer_arxiv_id": ["2001.04947", "2104.03110", "2012.12884"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_dev_55"} +{"question": "What is the research that estimates a transformation between two point clouds by deforming a template shape?", "answer": ["3D-CODED : 3D Correspondences by Deep Deformation"], "answer_arxiv_id": ["1806.05228"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_56"} +{"question": "Which papers studied learning invariant representations for domain adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks"], "answer_arxiv_id": ["1505.07818"], "source_meta": {"published_time": "20201219"}, "qid": "AutoScholarQuery_dev_57"} +{"question": "Can you give me examples of papers that used optimization and translation models to recover atomic coordinates from generated voxel grids?", "answer": ["Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models"], "answer_arxiv_id": ["2010.08687"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_58"} +{"question": "Could you provide some works that tackle regression as an ordinal classification problem?", "answer": ["Deep Ordinal Regression Network for Monocular Depth Estimation", "Rank consistent ordinal regression for neural networks with application to age estimation", "Deep Neural Networks for Rank-Consistent Ordinal Regression Based On Conditional Probabilities"], "answer_arxiv_id": ["1806.02446", "1901.07884", "2111.08851"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_dev_59"} +{"question": "What research has been done on learning the relative position of texts in the latent embedding space?", "answer": ["Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", "SimCSE: Simple Contrastive Learning of Sentence Embeddings"], "answer_arxiv_id": ["1908.10084", "2104.08821"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_dev_60"} +{"question": "What applications for federated learning are discussed in the field of healthcare?", "answer": ["A Secure Federated Learning Framework for 5G Networks"], "answer_arxiv_id": ["2005.05752"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_dev_61"} +{"question": "Which papers used guidance to influence the sampling procedure within their diffusion models?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2105.05233", "2207.12598"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_dev_62"} +{"question": "What studies are about the emergence of sequence-structure co-design methods and their superiority over previous methods?", "answer": ["Iterative Refinement Graph Neural Network for Antibody Sequence-Structure Co-design", "Conditional Antibody Design as 3D Equivariant Graph Translation"], "answer_arxiv_id": ["2110.04624", "2208.06073"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_dev_63"} +{"question": "Which studies illustrate usage of semantic tokens for ASR or speech resynthesis?", "answer": ["Textless Speech-to-Speech Translation on Real Data", "CoBERT: Self-Supervised Speech Representation Learning Through Code\n Representation Learning"], "answer_arxiv_id": ["2112.08352", "2210.04062"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_dev_64"} +{"question": "What research papers joined the recent considerations of the regret minimization when the sender repeatedly interacts with receivers?", "answer": ["Multi-Receiver Online Bayesian Persuasion", "Learning to Persuade on the Fly: Robustness Against Ignorance", "Online Bayesian Recommendation with No Regret"], "answer_arxiv_id": ["2106.06480", "2102.10156", "2202.06135"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_dev_65"} +{"question": "Is there any recent work that used an implicit model to produce a non-parametric distribution over SO​(3) that can model objects with large symmetry groups?", "answer": ["Implicit-PDF: Non-Parametric Representation of Probability Distributions on the Rotation Manifold"], "answer_arxiv_id": ["2106.05965"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_66"} +{"question": "What are the studies that talk about using prompts and example-based definitions with regards to in-context learning (ICL)?", "answer": ["Fantastically Ordered Prompts and Where to Find Them: Overcoming\n Few-Shot Prompt Order Sensitivity", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2104.08786", "2201.11903"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_67"} +{"question": "Which studies have updated challenge test sets in a dynamic way similar to RealTime QA?", "answer": ["Dynabench: Rethinking Benchmarking in NLP", "DynaSent: A Dynamic Benchmark for Sentiment Analysis", "Dynaboard: An Evaluation-As-A-Service Platform for Holistic Next-Generation Benchmarking"], "answer_arxiv_id": ["2104.14337", "2012.15349", "2106.06052"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_dev_68"} +{"question": "Can you provide works that extended the zero-shot learning capability of CLIP to monocular depth estimation?", "answer": ["Can Language Understand Depth?", "Learning to Adapt CLIP for Few-Shot Monocular Depth Estimation"], "answer_arxiv_id": ["2207.01077", "2311.01034"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_69"} +{"question": "Which paper discusses indirect measurements of hypothesized theoretical entities known as constructs in social sciences?", "answer": ["Measurement and Fairness"], "answer_arxiv_id": ["1912.05511"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_dev_70"} +{"question": "What works have previously used continuation techniques when the objective function is differentiable?", "answer": ["Piecewise linear regularized solution paths", "The Lasso Problem and Uniqueness", "Complexity Analysis of the Lasso Regularization Path"], "answer_arxiv_id": ["0708.2197v1", "1206.0313", "1205.0079"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_dev_71"} +{"question": "Which work was first to use open-source models for QA systems with citation capability?", "answer": ["WebGLM: Towards An Efficient Web-Enhanced Question Answering System with\n Human Preferences"], "answer_arxiv_id": ["2306.07906"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_dev_72"} +{"question": "Could you provide me the work that scales the training dataset to billions in the field of Vision-Language Pre-training?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2102.05918"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_dev_73"} +{"question": "Which studies have focused on grounding elements in mobile UI based on instructions?", "answer": ["Mapping Natural Language Instructions to Mobile UI Action Sequences", "A Dataset for Interactive Vision-Language Navigation with Unknown\n Command Feasibility", "VUT: Versatile UI Transformer for Multi-Modal Multi-Task User Interface\n Modeling", "Spotlight: Mobile UI Understanding using Vision-Language Models with a\n Focus"], "answer_arxiv_id": ["2005.03776", "2202.02312", "2112.05692", "2209.14927"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_dev_74"} +{"question": "Any studies that introduced a special [LENGTH] token to the encoder for response length prediction?", "answer": ["Mask-Predict: Parallel Decoding of Conditional Masked Language Models", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1904.09324", "1810.04805"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_75"} +{"question": "Who used the spawning method to explore the connection between LMC and the Neural Tangent Kernel dynamics?", "answer": ["Deep learning versus kernel learning: an empirical study of loss landscape geometry and the time evolution of the Neural Tangent Kernel"], "answer_arxiv_id": ["2010.15110"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_dev_76"} +{"question": "Which papers researched on few-shot learning in drug discovery?", "answer": ["Low Data Drug Discovery with One-shot Learning", "Meta-Learning GNN Initializations for Low-Resource Molecular Property Prediction", "Few-Shot Graph Learning for Molecular Property Prediction", "Property-Aware Relation Networks for Few-Shot Molecular Property Prediction"], "answer_arxiv_id": ["1611.03199v1", "2003.05996", "2102.07916", "2107.07994"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_dev_77"} +{"question": "Which papers investigated the use of diffusion models in specific domains such as images?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Improved Denoising Diffusion Probabilistic Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2102.09672", "2205.11487", "2204.06125", "2112.10752"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_dev_78"} +{"question": "What are the studies that focus on quadratic reward functions?", "answer": ["Bandit Principal Component Analysis", "Bandit Phase Retrieval", "Stochastic Rank-1 Bandits", "Bilinear Bandits with Low-rank Structure", "Low-Rank Generalized Linear Bandit Problems", "Structured Stochastic Linear Bandits", "Low-rank Bandits with Latent Mixtures", "Stochastic Linear Bandits with Hidden Low Rank Structure"], "answer_arxiv_id": ["1902.03035", "2106.01660", "1608.03023v3", "1901.02470v2", "2006.02948", "1606.05693", "1609.01508", "1901.09490"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_dev_79"} +{"question": "What works developed contrastive learning for OOD detection?", "answer": ["Hybrid Discriminative-Generative Training via Contrastive Learning", "RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection", "Contrastive Training for Improved Out-of-Distribution Detection"], "answer_arxiv_id": ["2007.09070", "2204.02553v3", "2007.05566"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_80"} +{"question": "What work tried to estimate the global translation by marrying a supporting-foot-based method with an RNN-based root translation regression model?", "answer": ["TransPose: Real-time 3D Human Translation and Pose Estimation with Six\n Inertial Sensors"], "answer_arxiv_id": ["2105.04605"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_81"} +{"question": "Could you provide me some works related to policy-gradient methods?", "answer": ["Bayesian Model-Agnostic Meta-Learning", "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Multimodal Model-Agnostic Meta-Learning via Task-Aware Modulation", "Fast Context Adaptation via Meta-Learning"], "answer_arxiv_id": ["1806.03836", "1703.03400", "1910.13616", "1810.03642"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_82"} +{"question": "Which paper converts behavioral cloning into a conditional energy-based modeling problem?", "answer": ["Implicit Behavioral Cloning"], "answer_arxiv_id": ["2109.00137"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_dev_83"} +{"question": "Which works are related to Networks with discrete Key-Value bottlenecks?", "answer": ["Discrete Key-Value Bottleneck"], "answer_arxiv_id": ["2207.11240"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_84"} +{"question": "What research has been done on using variational inference or Latent Variable Modeling for radiance field uncertainty in NeRF?", "answer": ["Stochastic Neural Radiance Fields: Quantifying Uncertainty in Implicit\n 3D Representations", "Conditional-Flow NeRF: Accurate 3D Modelling with Reliable Uncertainty\n Quantification"], "answer_arxiv_id": ["2109.02123", "2203.10192"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_85"} +{"question": "In what works were approaches proposed to mitigate the computational burden of NeRF training?", "answer": ["Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "Improved Direct Voxel Grid Optimization for Radiance Fields\n Reconstruction", "Plenoxels: Radiance Fields without Neural Networks", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2111.11215", "2206.05085", "2112.05131", "2201.05989"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_86"} +{"question": "Can you list the references where they have classified the method of knowledge updating and model editing for language-literature models?", "answer": ["Editing Large Language Models: Problems, Methods, and Opportunities", "Knowledge Editing for Large Language Models: A Survey"], "answer_arxiv_id": ["2305.13172", "2310.16218"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_87"} +{"question": "Can you provide me some works where LLMs are utilized to perform summarization with the fixed aspects provided by humans?", "answer": ["News Summarization and Evaluation in the Era of GPT-3", "Exploring the Limits of ChatGPT for Query or Aspect-based Text\n Summarization"], "answer_arxiv_id": ["2209.12356", "2302.08081"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_88"} +{"question": "Can you name the papers that have discussed strategies to counteract the high variance issue of LR gradient?", "answer": ["High-Dimensional Continuous Control Using Generalized Advantage Estimation"], "answer_arxiv_id": ["1506.02438"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_89"} +{"question": "Are there any studies employing GPT-3 Codex in training-free neural-symbolic framework?", "answer": ["Binding Language Models in Symbolic Languages", "Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2210.02875", "2107.03374"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_dev_90"} +{"question": "What are some studies that have adapted the representation of Gaussian splatting for text-to-3D generation?", "answer": ["DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation", "Text-to-3D using Gaussian Splatting"], "answer_arxiv_id": ["2309.16653", "2309.16585"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_91"} +{"question": "What studies made advances in the photorealistic synthesis of images conditioned on text prompts?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "MaskGIT: Masked Generative Image Transformer", "Muse: Text-To-Image Generation via Masked Generative Transformers"], "answer_arxiv_id": ["2105.05233", "2205.11487", "2204.06125", "2112.10752", "2202.04200", "2301.00704"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_dev_92"} +{"question": "What papers involve prefix techniques used in LLMs?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally Across Scales and Tasks"], "answer_arxiv_id": ["2101.00190", "2110.07602"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_93"} +{"question": "What papers analyse stochastic optimization methods under the arbitrary sampling paradigm?", "answer": ["On Optimal Probabilities in Stochastic Coordinate Descent Methods", "Coordinate Descent with Arbitrary Sampling I: Algorithms and Complexity", "Accelerated Coordinate Descent with Arbitrary Sampling and Best Rates for Minibatches", "Stochastic Reformulations of Linear Systems: Algorithms and Convergence Theory", "Momentum and Stochastic Momentum for Stochastic Gradient, Newton, Proximal Point and Subspace Descent Methods", "Convergence Analysis of Inexact Randomized Iterative Methods", "A New Perspective on Randomized Gossip Algorithms", "Revisiting Randomized Gossip Algorithms: General Framework, Convergence Rates and Novel Block and Accelerated Protocols", "SGD: General Analysis and Improved Rates", "SGD for Structured Nonconvex Functions: Learning Rates, Minibatching and Interpolation", "SAGA with Arbitrary Sampling", "Nonconvex Variance Reduced Optimization with Arbitrary Sampling", "Unified Analysis of Stochastic Gradient Methods for Composite Convex and Smooth Optimization", "Stochastic Hamiltonian Gradient Methods for Smooth Games", "Stochastic Gradient Descent-Ascent and Consensus Optimization for Smooth Games: Convergence Analysis under Expected Co-coercivity", "Stochastic Extragradient: General Analysis and Improved Rates"], "answer_arxiv_id": ["1310.3438", "1412.8060", "1809.09354", "1706.01108", "1712.09677", "1903.07971", "1610.04714", "1905.08645", "1901.09401", "2006.10311v3", "1901.08669", "1809.04146v2", "2006.11573v1", "2007.04202", "2107.00052", "2111.08611v3"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_94"} +{"question": "What papers used Hermite expansion to study the Neural Tangent Kernel?", "answer": ["Toward Deeper Understanding of Neural Networks:The Power of Initialization and a Dual View on Expressivity", "Reverse Engineering the Neural Tangent Kernel", "Global Convergence of Deep Networks with One Wide Layer Followed by Pyramidal Topology", "Effect of Activation Functions on the Training of Overparametrized Neural Nets", "Fast Neural Kernel Embeddings for General Activations"], "answer_arxiv_id": ["1602.05897", "2106.03186", "2002.07867", "1908.05660", "2209.04121"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_dev_95"} +{"question": "What works defined the settings of the DDPM model based on the continuous limit of βt?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_dev_96"} +{"question": "Which papers established improvements over the no-regret framework when specific learning dynamics are in place?", "answer": ["Fast Convergence of Regularized Learning in Games", "Near-Optimal No-Regret Learning in General Games"], "answer_arxiv_id": ["1507.00407", "2108.06924"], "source_meta": {"published_time": "20220928"}, "qid": "AutoScholarQuery_dev_97"} +{"question": "What works researched the effect of adversarial perturbations on image-to-image tasks?", "answer": ["Deep Image Destruction: Vulnerability of Deep Image-to-Image Models against Adversarial Attacks", "Towards Adversarially Robust Deep Image Denoising", "Evaluating Robustness of Deep Image Super-Resolution Against Adversarial Attacks"], "answer_arxiv_id": ["2104.15022", "2201.04397", "1904.06097"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_98"} +{"question": "What works explore the vulnerability of models to inversion attacks causing the leakage of private information?", "answer": ["Membership Inference Attacks against Machine Learning Models", "Comprehensive Privacy Analysis of Deep Learning: Passive and Active\n White-box Inference Attacks against Centralized and Federated Learning"], "answer_arxiv_id": ["1610.05820", "1812.00910"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_dev_99"} +{"question": "What papers provided strategies to enable image generation conditioned on text and other modalities in guided diffusion models?", "answer": ["Classifier-Free Diffusion Guidance", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "Universal Guidance for Diffusion Models", "AudioToken: Adaptation of Text-Conditioned Diffusion Models for\n Audio-to-Image Generation", "The Power of Sound (TPoS): Audio Reactive Video Generation with Stable\n Diffusion", "Improved Denoising Diffusion Probabilistic Models", "Universal Guidance for Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2207.12598", "2110.02711", "2302.07121", "2305.13050", "2309.04509", "2102.09672", "2302.07121", "2112.10752"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_dev_100"} +{"question": "What are the studies that designed sample-efficient RL algorithms with general function approximations in static RL setting?", "answer": ["Model-based Reinforcement Learning and the Eluder Dimension", "Reinforcement Learning with General Value Function Approximation: Provably Efficient Approach via Bounded Eluder Dimension"], "answer_arxiv_id": ["1406.1853", "2005.10804"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_dev_101"} +{"question": "What work uses keyword-based retrieval (BM25) for semi-supervised learning?", "answer": ["NLP From Scratch Without Large-Scale Pretraining: A Simple and Efficient Framework"], "answer_arxiv_id": ["2111.04130"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_dev_102"} +{"question": "In what work do the researchers use a Python interpreter to make the prediction of an LLM more likely to be faithful?", "answer": ["Faithful Chain-of-Thought Reasoning"], "answer_arxiv_id": ["2301.13379"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_103"} +{"question": "What research papers have presented algorithms that can control the dynamic regret in non-stationary online learning?", "answer": ["Online Optimization : Competing with Dynamic Comparators", "Non-stationary Stochastic Optimization", "Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient", "Adaptive Online Learning in Dynamic Environments", "Dynamic Regret of Convex and Smooth Functions", "Improved Analysis for Dynamic Regret of Strongly Convex and Smooth Functions", "Non-stationary Online Learning with Memory and Non-stochastic Control", "Parameter-free Mirror Descent", "Second Order Path Variationals in Non-Stationary Online Learning"], "answer_arxiv_id": ["1501.06225", "1307.5449", "1605.04638", "1810.10815", "2007.03479", "2006.05876", "2102.03758", "2203.00444", "2205.01921v2"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_104"} +{"question": "What papers improved convergence rate for soft policies by analyzing NAC under Markovian sampling?", "answer": ["Improving Sample Complexity Bounds for (Natural) Actor-Critic Algorithms"], "answer_arxiv_id": ["2004.12956"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_105"} +{"question": "What research fine-tunes subset layers of cross-attention in the UNet for personalized visual content generation?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2212.04488"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_dev_106"} +{"question": "What studies have achieved strong results in text conditioned image generation through Diffusion generative models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Zero-Shot Text-to-Image Generation", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2112.10752", "2102.12092", "2112.10741", "2208.12242"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_dev_107"} +{"question": "Which references assert that even for graphs such as trees or paths, computing subgraph counts is NP-hard?", "answer": ["Everything you always wanted to know about the parameterized complexity of Subgraph Isomorphism (but were afraid to ask)"], "answer_arxiv_id": ["1307.2187"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_108"} +{"question": "Which paper introduced a hierarchical transformer structure to progressively shrink the spatiotemporal resolution of feature maps and increase channels in the case of action recognition?", "answer": ["Multiscale Vision Transformers"], "answer_arxiv_id": ["2104.11227"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_109"} +{"question": "Which papers introduced LLMs such as OPT, LLaMA, BLOOM and PaLM?", "answer": ["OPT: Open Pre-trained Transformer Language Models", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "PaLM: Scaling Language Modeling with Pathways", "PaLM 2 Technical Report"], "answer_arxiv_id": ["2205.01068", "2302.13971", "2307.09288", "2211.05100", "2204.02311", "2305.10403"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_110"} +{"question": "Which works used approaches like particle-based graph neural network dynamics predictors?", "answer": ["Interaction Networks for Learning about Objects, Relations and Physics", "Relational inductive biases, deep learning, and graph networks", "Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids", "Learning to Simulate Complex Physics with Graph Networks"], "answer_arxiv_id": ["1612.00222", "1806.01261", "1810.01566", "2002.09405"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_dev_111"} +{"question": "Could you list some works that improved performance using Auxiliary Learning methods?", "answer": ["Deep Auxiliary Learning for Visual Localization and Odometry", "VLocNet++: Deep Multitask Learning for Semantic Visual Localization and\n Odometry", "Deep Global-Relative Networks for End-to-End 6-DoF Visual Localization\n and Odometry"], "answer_arxiv_id": ["1803.03642", "1804.08366", "1812.07869"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_112"} +{"question": "Which works improved the efficiency of TRPO using an ensemble of environment models?", "answer": ["Model-Ensemble Trust-Region Policy Optimization"], "answer_arxiv_id": ["1802.10592"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_dev_113"} +{"question": "What works developed a unified framework for estimating expected information gain and optimizing designs with gradient-based methods in the field of Differentiable Bayesian Optimal Experimental Design?", "answer": ["Gradient-based stochastic optimization methods in Bayesian experimental design", "Variational Bayesian Optimal Experimental Design", "A Unified Stochastic Gradient Approach to Designing Bayesian-Optimal Experiments", "Bayesian Experimental Design for Implicit Models by Mutual Information Neural Estimation", "Gradient-based Bayesian Experimental Design for Implicit Models using Mutual Information Lower Bounds"], "answer_arxiv_id": ["1212.2228", "1903.05480v3", "1911.00294v2", "2002.08129", "2105.04379"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_dev_114"} +{"question": "Which studies have utilized dual pathways in Bird’s Eye View(BEV) detection?", "answer": ["Learning to Prompt for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language\n Modeling", "Conditional Prompt Learning for Vision-Language Models", "ActionCLIP: A New Paradigm for Video Action Recognition", "Robust fine-tuning of zero-shot models"], "answer_arxiv_id": ["2109.01134", "2111.03930", "2203.05557", "2109.08472", "2109.01903"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_dev_115"} +{"question": "Which studies showed that original optimization based attacks still work to some extent if very small learning rates are used?", "answer": ["Inverting Gradients - How easy is it to break privacy in federated learning?", "AGIC: Approximate Gradient Inversion Attack on Federated Learning"], "answer_arxiv_id": ["2003.14053", "2204.13784"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_116"} +{"question": "What research has been conducted on the convergence of PFL with respect to system heterogeneity?", "answer": ["Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization"], "answer_arxiv_id": ["2007.07481"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_117"} +{"question": "What research has been done on the usage of Transformer for low-level vision tasks?", "answer": ["Learning Texture Transformer Network for Image Super-Resolution", "Pre-Trained Image Processing Transformer", "SwinIR: Image Restoration Using Swin Transformer", "Restormer: Efficient Transformer for High-Resolution Image Restoration", "Uformer: A General U-Shaped Transformer for Image Restoration"], "answer_arxiv_id": ["2006.04139", "2012.00364", "2108.10257", "2111.09881", "2106.03106"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_118"} +{"question": "What research has been done to establish lower bounds for networks with monotone activation functions or its variants?", "answer": ["Deep, Skinny Neural Networks are not Universal Approximators"], "answer_arxiv_id": ["1810.00393"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_dev_119"} +{"question": "What studies started the research line in designing appropriate prompts for large language models?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?", "Multitask Prompted Training Enables Zero-Shot Task Generalization"], "answer_arxiv_id": ["2101.06804", "2110.08207"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_120"} +{"question": "What are the research papers that contribute to Diffusion Models?", "answer": ["Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Improved Denoising Diffusion Probabilistic Models", "HouseDiffusion: Vector Floorplan Generation via a Diffusion Model with Discrete and Continuous Denoising", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Image Super-Resolution via Iterative Refinement", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic Models", "Blended Latent Diffusion", "Label-Efficient Semantic Segmentation with Diffusion Models", "Diffusion-Based 3D Human Pose Estimation with Multi-Hypothesis Aggregation", "LEGO-Net: Learning Regular Rearrangements of Objects in Rooms"], "answer_arxiv_id": ["2006.11239", "2105.05233", "2102.09672", "2211.13287", "2112.10741", "2207.06635", "2204.06125", "2112.10752", "2104.07636", "2104.14951", "2206.02779", "2112.03126", "2303.11579", "2301.09629"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_dev_121"} +{"question": "What papers have implemented Generative Adversarial Imitation Learning (GAIL) for imitation learning from observations?", "answer": ["Generative Adversarial Imitation Learning", "Generative Adversarial Imitation from Observation", "Learning Agile Skills via Adversarial Imitation of Rough Partial Demonstrations"], "answer_arxiv_id": ["1606.03476", "1807.06158", "2206.11693"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_122"} +{"question": "What are some papers that used OpenWebText for pretraining of models?", "answer": ["RoBERTa: A Robustly Optimized BERT Pretraining Approach", "Megatron-LM: Training Multi-Billion Parameter Language Models Using\n Model Parallelism"], "answer_arxiv_id": ["1907.11692", "1909.08053"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_dev_123"} +{"question": "Which studies released LexFiles, an English legal corpus, and trained two new legal English PLMs using this corpus?", "answer": ["LeXFiles and LegalLAMA: Facilitating English Multinational Legal\n Language Model Development"], "answer_arxiv_id": ["2305.07507"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_dev_124"} +{"question": "Which works suggested synthesizing video sequences by extending the generated image tensors along a time dimension?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models", "StyleGAN-V: A Continuous Video Generator with the Price, Image Quality\n and Perks of StyleGAN2", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Video Probabilistic Diffusion Models in Projected Latent Space", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation", "Generating Long Videos of Dynamic Scenes"], "answer_arxiv_id": ["2210.02303", "2112.14683", "2304.08818", "2302.07685", "2211.11018", "2303.08320", "2206.03429"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_dev_125"} +{"question": "What references proposed the concept of a hypernetwork?", "answer": ["HyperNetworks"], "answer_arxiv_id": ["1609.09106v4"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_126"} +{"question": "What are the notable contributions in the field of text-infilling for generative data augmentation?", "answer": ["MELM: Data Augmentation with Masked Entity Language Modeling for\n Low-Resource NER", "GENIUS: Sketch-based Language Model Pre-training via Extreme and\n Selective Masking for Text Generation and Augmentation", "ACLM: A Selective-Denoising based Generative Data Augmentation Approach\n for Low-Resource Complex NER", "DALE: Generative Data Augmentation for Low-Resource Legal NLP", "BioAug: Conditional Generation based Data Augmentation for Low-Resource\n Biomedical NER"], "answer_arxiv_id": ["2108.13655", "2211.10330", "2306.00928", "2310.15799", "2305.10647"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_dev_127"} +{"question": "Which research papers observed linear mode connectivity (LMC) in models trained on MNIST starting from the same random initialization?", "answer": ["Uniform convergence may be unable to explain generalization in deep learning"], "answer_arxiv_id": ["1902.04742"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_dev_128"} +{"question": "Which papers discuss the exact methods for multi-objective combinatorial optimization (MOCO)?", "answer": ["Network Models for Multiobjective Discrete Optimization"], "answer_arxiv_id": ["1802.08637"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_dev_129"} +{"question": "Which works propose the use of learnable prompts at the CLIP text input for fine-tuning on few-shot examples?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_dev_130"} +{"question": "What research employ multi-modal models in the field of object detection?", "answer": ["Multi-Modal Fusion Transformer for End-to-End Autonomous Driving"], "answer_arxiv_id": ["2104.09224"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_131"} +{"question": "What research used a strategy that incorporates a prior probability derived from local color variance and further tracks photometric error throughout training with an adaptive quadtree structure?", "answer": ["Fast Learning Radiance Fields by Shooting Much Fewer Rays"], "answer_arxiv_id": ["2208.06821"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_dev_132"} +{"question": "Which paper proposed the idea of sharing all layers within a transformer model?", "answer": ["Universal Transformers"], "answer_arxiv_id": ["1807.03819"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_dev_133"} +{"question": "What works have emerged in the field of compression with INRs, and have been effective in compressing various data like images, climate data, videos and 3D scenes?", "answer": ["COIN: COmpression with Implicit Neural representations", "Implicit Neural Representations for Image Compression", "COIN++: Neural Compression Across Modalities", "NeRV: Neural Representations for Videos", "3D Scene Compression through Entropy Penalized Neural Representation Functions"], "answer_arxiv_id": ["2103.03123", "2112.04267v2", "2201.12904", "2110.13903", "2104.12456"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_dev_134"} +{"question": "What research prioritizes a sample if it significantly improves the probability of correctly predicting the true label?", "answer": ["What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation"], "answer_arxiv_id": ["2008.03703"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_dev_135"} +{"question": "What works discuss the challenge of hallucinations in Large Language Models?", "answer": ["Siren's Song in the AI Ocean: A Survey on Hallucination in Large\n Language Models", "HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large\n Language Models", "Explore Spurious Correlations at the Concept Level in Language Models\n for Text Classification", "Hallucination Detection: Robustly Discerning Reliable Answers in Large Language Models"], "answer_arxiv_id": ["2309.01219", "2305.11747", "2311.08648", "2407.04121v1"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_dev_136"} +{"question": "What works have studied and utilized the high-dimensional visual features in a diffusion model?", "answer": ["Your Diffusion Model is Secretly a Zero-Shot Classifier", "SegDiff: Image Segmentation with Diffusion Probabilistic Models", "Label-Efficient Semantic Segmentation with Diffusion Models", "Unsupervised Semantic Correspondence Using Stable Diffusion", "A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot\n Semantic Correspondence", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2303.16203", "2112.00390", "2112.03126", "2305.15581", "2305.15347", "2303.04803"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_dev_137"} +{"question": "Could you provide some works discussing the inference cost as a drawback of Seq2Seq models in GEC?", "answer": ["Instantaneous Grammatical Error Correction with Shallow Aggressive\n Decoding"], "answer_arxiv_id": ["2106.04970"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_dev_138"} +{"question": "What are some examples of research that explore improvement of communication efficiency in Federated Learning through methods based on gradient compression?", "answer": ["FetchSGD: Communication-Efficient Federated Learning with Sketching", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "FedPAQ: A Communication-Efficient Federated Learning Method with Periodic Averaging and Quantization", "Federated Learning with Compression: Unified Analysis and Sharp Guarantees", "AsySQN: Faster Vertical Federated Learning Algorithms with Better Computation Resource Utilization"], "answer_arxiv_id": ["2007.07682v2", "1610.02132", "1909.13014v4", "2007.01154v2", "2109.12519"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_dev_139"} +{"question": "Are there any studies that work on shuffling-based methods in Federated Learning?", "answer": ["Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond", "On the Convergence of Federated Averaging with Cyclic Client Participation"], "answer_arxiv_id": ["2110.10342", "2302.03109v1"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_140"} +{"question": "What research has been done on the effect of persona variables on hate speech detection?", "answer": ["Designing Toxic Content Classification for a Diversity of Perspectives", "Annotators with Attitudes: How Annotator Beliefs And Identities Bias\n Toxic Language Detection", "When Do Annotator Demographics Matter? Measuring the Influence of\n Annotator Demographics with the POPQUORN Dataset", "NLPositionality: Characterizing Design Biases of Datasets and Models", "How Crowd Worker Factors Influence Subjective Annotations: A Study of\n Tagging Misogynistic Hate Speech in Tweets"], "answer_arxiv_id": ["2106.04511", "2111.07997", "2306.06826", "2306.01943", "2309.01288"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_141"} +{"question": "Which studies used perturbation techniques similar to ours for measuring contamination in the test questions?", "answer": ["Extracting Training Data from Large Language Models", "Ethical Challenges in Data-Driven Dialogue Systems", "Investigating Data Contamination for Pre-training Language Models", "Understanding Unintended Memorization in Federated Learning"], "answer_arxiv_id": ["2012.07805", "1711.09050", "2401.06059", "2006.07490"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_142"} +{"question": "Which paper ensured diversity sampling by selecting core sets in LiDAR point clouds?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["1708.00489"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_dev_143"} +{"question": "What are some works that use sparse voxels for 3D scene understanding?", "answer": ["3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1711.10275", "1904.08755"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_dev_144"} +{"question": "Could you provide me some examples of research dealing with unauthorized data usage in diffusion models?", "answer": ["Conditional Generative Adversarial Nets", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["1411.1784", "2204.06125"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_dev_145"} +{"question": "Which works have explored visual representations supervised by language in the context of multi-modal studies?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Multimodal Contrastive Training for Visual Representation Learning", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "VirTex: Learning Visual Representations from Textual Annotations", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2104.12836", "2205.01917", "2006.06666", "2108.10904", "2204.14198"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_146"} +{"question": "Which paper evaluated only a single specialized meta-RL method?", "answer": ["VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning"], "answer_arxiv_id": ["1910.08348"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_147"} +{"question": "Any works about semantic segmentation applying multi-modal models?", "answer": ["SNE-RoadSeg: Incorporating Surface Normal Information into Semantic\n Segmentation for Accurate Freespace Detection"], "answer_arxiv_id": ["2008.11351"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_148"} +{"question": "Which papers demonstrated the performance of LLMs leveraging RLHF for alignment and generation?", "answer": ["Training language models to follow instructions with human feedback", "GPT-4 Technical Report", "Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback", "Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["2203.02155", "2303.08774", "2204.05862", "1909.08593"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_dev_149"} +{"question": "Could you provide me some works that enhance ray adjacency consistency for scene reconstruction?", "answer": ["InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Rendering"], "answer_arxiv_id": ["2112.15399"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_150"} +{"question": "Are there any works on the acceleration of the generation process of AudioLM?", "answer": ["SoundStorm: Efficient Parallel Audio Generation"], "answer_arxiv_id": ["2305.09636"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_151"} +{"question": "Are there any research papers that utilize LLMs to aid the training stage of CQR in conversational search?", "answer": ["Enhancing Conversational Search: Large Language Model-Aided Informative\n Query Rewriting"], "answer_arxiv_id": ["2310.09716"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_dev_152"} +{"question": "Have any research examined biasedness as a writing style?", "answer": ["Automatically Neutralizing Subjective Bias in Text"], "answer_arxiv_id": ["1911.09709"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_153"} +{"question": "Which work does the paper refer to when discussing a special case of AMPO?", "answer": ["On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2201.07443"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_154"} +{"question": "What works demonstrate linear convergence of PG for the softmax tabular policy without regularization?", "answer": ["Leveraging Non-uniformity in First-order Non-convex Optimization"], "answer_arxiv_id": ["2105.06072"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_155"} +{"question": "Which study proposed the Iterative Refinement Long Short-Term Memory approach in few-shot learning?", "answer": ["Low Data Drug Discovery with One-shot Learning"], "answer_arxiv_id": ["1611.03199v1"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_dev_156"} +{"question": "What works have attempted to accelerate the computation of Jacobian determinants in the ML objective by exploiting linear transformations with special structures?", "answer": ["MintNet: Building Invertible Neural Networks with Masked Convolutions", "Emerging Convolutions for Generative Normalizing Flows", "MaCow: Masked Convolutional Generative Flow", "Woodbury Transformations for Deep Generative Flows", "ButterflyFlow: Building Invertible Layers with Butterfly Matrices", "Improving Variational Auto-Encoders using Householder Flow"], "answer_arxiv_id": ["1907.07945", "1901.11137", "1902.04208", "2002.12229", "2209.13774", "1611.09630"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_157"} +{"question": "What studies exploit artificial techniques to increase the number of training tasks for generalization in RL, such as procedural generation, augmentations, or task interpolation?", "answer": ["Quantifying Generalization in Reinforcement Learning", "Leveraging Procedural Generation to Benchmark Reinforcement Learning", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Rotation, Translation, and Cropping for Zero-Shot Generalization", "Network Randomization: A Simple Technique for Generalization in Deep Reinforcement Learning", "Automatic Data Augmentation for Generalization in Reinforcement Learning", "Meta-Learning with Fewer Tasks through Task Interpolation"], "answer_arxiv_id": ["1812.02341", "1912.01588", "2004.13649", "2001.09908", "1910.05396", "2006.12862", "2106.02695"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_dev_158"} +{"question": "Could you provide me with the research that proposed an alternative dual form of UOT, which resembles the dual form of OT with a Lagrangian regularizer?", "answer": ["Robust Optimal Transport with Applications in Generative Modeling and Domain Adaptation"], "answer_arxiv_id": ["2010.05862"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_159"} +{"question": "Which research studies require training a reference model for data selection?", "answer": ["Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt"], "answer_arxiv_id": ["2206.07137"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_dev_160"} +{"question": "Could you tell me some research exploring the use of LLMs in recommendation tasks?", "answer": ["Large Language Models are Competitive Near Cold-start Recommenders for\n Language- and Item-based Preferences", "Personalized Prompt Learning for Explainable Recommendation"], "answer_arxiv_id": ["2307.14225", "2202.07371"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_dev_161"} +{"question": "What works are concerned with CXR-to-report generation in medical VLMs for chest radiographs?", "answer": ["Generating Radiology Reports via Memory-driven Transformer", "Cross-modal Memory Networks for Radiology Report Generation", "Cross-modal Prototype Driven Network for Radiology Report Generation", "Automatic Radiology Report Generation by Learning with Increasingly Hard\n Negatives", "Radiology Report Generation with a Learned Knowledge Base and\n Multi-modal Alignment"], "answer_arxiv_id": ["2010.16056", "2204.13258", "2207.04818", "2305.07176", "2112.15011"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_dev_162"} +{"question": "Which papers consider influence functions to estimate training data's influence on a test point?", "answer": ["Understanding Black-box Predictions via Influence Functions", "\"Influence Sketching\": Finding Influential Samples In Large-Scale Regressions", "Interpreting Black Box Predictions using Fisher Kernels", "If Influence Functions are the Answer, Then What is the Question?"], "answer_arxiv_id": ["1703.04730", "1611.05923", "1810.10118", "2209.05364"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_163"} +{"question": "Could you provide me works that discussed the upper bounds of ReLU networks?", "answer": ["The Expressive Power of Neural Networks: A View from the Width", "Approximating Continuous Functions by ReLU Nets of Minimal Width", "Universal Function Approximation by Deep Neural Nets with Bounded Width and ReLU Activations", "Minimum Width for Universal Approximation"], "answer_arxiv_id": ["1709.02540", "1710.11278", "1708.02691", "2006.08859"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_dev_164"} +{"question": "Can you provide references discussing similar sampling-based approaches for the graphlet kernel and frequent-subtree kernels?", "answer": ["Fast graph kernel with optical random features"], "answer_arxiv_id": ["2010.08270"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_165"} +{"question": "What studies use Graph Convolutional Networks for activity recognition?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks"], "answer_arxiv_id": ["1609.02907"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_166"} +{"question": "Can you provide me with the studies that focused on functional requirements in code generation?", "answer": ["Measuring Coding Challenge Competence With APPS", "Program Synthesis with Large Language Models", "Evaluating Large Language Models Trained on Code", "[2203.07814] Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2105.09938", "2108.07732", "2107.03374", "2203.07814"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_dev_167"} +{"question": "What studies further refined the without-replacement Policy Gradient (PG) estimator by using without-replacement samples as a free baseline?", "answer": ["Estimating Gradients for Discrete Random Variables by Sampling without Replacement"], "answer_arxiv_id": ["2002.06043"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_dev_168"} +{"question": "What works propose the hybridization of classical numerical methods with contemporary data-driven deep learning techniques?", "answer": ["A machine learning framework for data driven acceleration of computations of differential equations.", "Machine learning accelerated computational fluid dynamics", "A posteriori learning for quasi-geostrophic turbulence parametrization"], "answer_arxiv_id": ["1807.09519", "2102.01010", "2204.03911"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_dev_169"} +{"question": "Which studies describe autoregressive modeling of semantic tokens for generating speech continuations?", "answer": ["Generative Spoken Language Modeling from Raw Audio"], "answer_arxiv_id": ["2102.01192"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_170"} +{"question": "Which paper proposes the method of generating noise samples with the help of observed data in Conditional NCE?", "answer": ["Conditional Noise-Contrastive Estimation of Unnormalised Models"], "answer_arxiv_id": ["1806.03664v1"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_171"} +{"question": "What benchmark does the variant version of HumanEval-NFR in the study use as a reference?", "answer": ["Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2107.03374"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_dev_172"} +{"question": "What research proposes distilling features on various sophisticatedly-selected sub-regions of the feature map to solve imbalance issue?", "answer": ["Distilling Object Detectors with Fine-grained Feature Imitation", "Distilling Object Detectors with Task Adaptive Regularization", "General Instance Distillation for Object Detection", "Distilling Object Detectors via Decoupled Features", "Focal and Global Knowledge Distillation for Detectors"], "answer_arxiv_id": ["1906.03609", "2006.13108", "2103.02340", "2103.14475", "2111.11837"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_dev_173"} +{"question": "Are there any works that develop accurate probes for factuality detection in LLM without relying on annotated training data?", "answer": ["The Internal State of an LLM Knows When It's Lying", "Representation Engineering: A Top-Down Approach to AI Transparency", "Hallucination Detection: Robustly Discerning Reliable Answers in Large Language Models"], "answer_arxiv_id": ["2304.13734", "2310.01405", "2407.04121v1"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_dev_174"} +{"question": "Which studies have addressed the issue of capturing electron-electron interactions beyond a mean-field approximation?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_dev_175"} +{"question": "What work proposes to decompose the voxel tensor into feature planes and vectors?", "answer": ["TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2203.09517"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_dev_176"} +{"question": "What papers propose using model self-consistency for factuality detection in LLM?", "answer": ["Language Models (Mostly) Know What They Know", "The Internal State of an LLM Knows When It's Lying", "Representation Engineering: A Top-Down Approach to AI Transparency"], "answer_arxiv_id": ["2207.05221", "2304.13734", "2310.01405"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_dev_177"} +{"question": "Which papers provide interesting examples of recent work in online calibration in the adversarial sequence model?", "answer": ["Estimating Uncertainty Online Against an Adversary"], "answer_arxiv_id": ["1607.03594"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_dev_178"} +{"question": "Could you provide information about works which tried to enhance PLMs through fine-tuning on human or synthetic labels?", "answer": ["COMET: A Neural Framework for MT Evaluation", "Towards a Unified Multi-Dimensional Evaluator for Text Generation"], "answer_arxiv_id": ["2009.09025", "2210.07197"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_dev_179"} +{"question": "What studies focus on interpolation condition in overparameterized models?", "answer": ["SGD for Structured Nonconvex Functions: Learning Rates, Minibatching and Interpolation", "Fast and Faster Convergence of SGD for Over-Parameterized Models and an Accelerated Perceptron", "Implicit Regularization in Deep Matrix Factorization", "L4: Practical loss-based stepsize adaptation for deep learning", "SGD for Structured Nonconvex Functions: Learning Rates, Minibatching and Interpolation", "Stochastic Polyak Step-size for SGD: An Adaptive Learning Rate for Fast Convergence", "Painless Stochastic Gradient: Interpolation, Line-Search, and Convergence Rates", "On the Convergence of Stochastic Extragradient for Bilinear Games using Restarted Iteration Averaging"], "answer_arxiv_id": ["2006.10311v3", "1810.07288v3", "1905.13655", "1802.05074", "2006.10311v3", "2002.10542v3", "1905.09997", "2107.00464v4"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_180"} +{"question": "Which work focused on learning of entity embeddings for rule learning but had limitations?", "answer": ["Embedding Entities and Relations for Learning and Inference in Knowledge Bases"], "answer_arxiv_id": ["1412.6575"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_181"} +{"question": "What works proposed the generative alignment of multiple modalities into one joint embedding space?", "answer": ["GIT: A Generative Image-to-text Transformer for Vision and Language", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and\n Dataset", "VAST: A Vision-Audio-Subtitle-Text Omni-Modality Foundation Model and\n Dataset"], "answer_arxiv_id": ["2205.14100", "2301.12597", "2304.08345", "2305.18500"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_182"} +{"question": "Any works demonstrate that acoustic tokens can capture details of audio waveforms, ranging from multi-speaker speech to music and audio effects?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation", "MusicLM: Generating Music From Text", "AudioGen: Textually Guided Audio Generation"], "answer_arxiv_id": ["2209.03143", "2301.11325", "2209.15352"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_183"} +{"question": "What models were cited as significantly benefiting from the instructional tuning strategy?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_dev_184"} +{"question": "Which works called out self-supervised tasks for improving model training in the context of self-supervised learning?", "answer": ["Self-supervised Learning from a Multi-view Perspective", "Predicting What You Already Know Helps: Provable Self-Supervised Learning", "Contrastive learning, multi-view redundancy, and linear models"], "answer_arxiv_id": ["2006.05576", "2008.01064", "2008.10150"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_185"} +{"question": "Are there any papers that extend neural fields to inverse rendering, where geometry and reflectance are modeled as neural fields?", "answer": ["WildLight: In-the-wild Inverse Rendering with a Flashlight", "IRON: Inverse Rendering by Optimizing Neural SDFs and Materials from\n Photometric Images", "ShadowNeuS: Neural SDF Reconstruction by Shadow Ray Supervision", "NeRFactor: Neural Factorization of Shape and Reflectance Under an\n Unknown Illumination", "PhySG: Inverse Rendering with Spherical Gaussians for Physics-based\n Material Editing and Relighting", "Neural Reflectance Fields for Appearance Acquisition", "NeRD: Neural Reflectance Decomposition from Image Collections", "Extracting Triangular 3D Models, Materials, and Lighting From Images", "Shape, Light, and Material Decomposition from Images using Monte Carlo\n Rendering and Denoising"], "answer_arxiv_id": ["2303.14190", "2204.02232", "2211.14086", "2106.01970", "2104.00674", "2008.03824", "2012.03918", "2111.12503", "2206.03380"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_186"} +{"question": "What studies utilized model grafting for detecting skill neurons?", "answer": ["Task-Specific Skill Localization in Fine-tuned Language Models"], "answer_arxiv_id": ["2302.06600"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_dev_187"} +{"question": "Could you provide me some studies that have explored various augmentations on graphs, based on the data augmentation in image analysis?", "answer": ["Graph Contrastive Learning with Augmentations", "Graph Contrastive Learning Automated", "Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "Bringing Your Own View: Graph Contrastive Learning without Prefabricated Data Augmentations", "Let Invariant Rationale Discovery Inspire Graph Contrastive Learning"], "answer_arxiv_id": ["2010.13902", "2106.07594", "2106.05819", "2201.01702", "2206.07869"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_dev_188"} +{"question": "What works fine-tune a pretrained language model for aspect extraction and rely on a manual labeling of comparative data?", "answer": ["DILBERT: Customized Pre-Training for Domain Adaptation withCategory\n Shift, with an Application to Aspect Extraction"], "answer_arxiv_id": ["2109.00571"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_189"} +{"question": "Can you name some works about semantic occupancy prediction based on RGB data?", "answer": ["MonoScene: Monocular 3D Semantic Scene Completion", "TBP-Former: Learning Temporal Bird's-Eye-View Pyramid for Joint\n Perception and Prediction in Vision-Centric Autonomous Driving", "OCTraN: 3D Occupancy Convolutional Transformer Network in Unstructured\n Traffic Scenarios", "Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction", "Symphonize 3D Semantic Scene Completion with Contextual Instance Queries", "VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene\n Completion", "OccDepth: A Depth-Aware Method for 3D Semantic Scene Completion", "OVO: Open-Vocabulary Occupancy", "Scene as Occupancy", "PanoOcc: Unified Occupancy Representation for Camera-based 3D Panoptic\n Segmentation", "PET-NeuS: Positional Encoding Tri-Planes for Neural Surfaces", "SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Driving", "BEV-IO: Enhancing Bird's-Eye-View 3D Detection with Instance Occupancy", "OccFormer: Dual-path Transformer for Vision-based 3D Semantic Occupancy\n Prediction", "FB-OCC: 3D Occupancy Prediction based on Forward-Backward View\n Transformation", "S4C: Self-Supervised Semantic Scene Completion with Neural Fields"], "answer_arxiv_id": ["2112.00726", "2303.09998", "2307.10934", "2302.07817", "2306.15670", "2302.12251", "2302.13540", "2305.16133v2", "2306.02851", "2306.10013", "2305.05594", "2303.09551", "2305.16829", "2304.05316", "2307.01492", "2310.07522"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_dev_190"} +{"question": "Which works used fine-tuning in combination with weight interpolation to improve results on specific tasks?", "answer": ["Patching open-vocabulary modelsby interpolating weights"], "answer_arxiv_id": ["2208.05592"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_191"} +{"question": "What research work focused on enhancing generation quality in SLP using mixture density networks, Mixture-of-Experts, dictionary representations, and diffusion models?", "answer": ["Continuous 3D Multi-Channel Sign Language Production via Progressive\n Transformers and Mixture Density Networks", "Mixed SIGNals: Sign Language Production via a Mixture of Motion\n Primitives", "Signing at Scale: Learning to Co-Articulate Signs for Large-Scale\n Photo-Realistic Sign Language Production", "Neural Sign Actors: A diffusion model for 3D sign language production\n from text"], "answer_arxiv_id": ["2103.06982", "2107.11317", "2203.15354", "2312.02702"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_dev_192"} +{"question": "What studies propose the idea of labelled data points for efficient learning progress during online updates in the area of active learning?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Active Learning at the ImageNet Scale"], "answer_arxiv_id": ["1708.00489", "2111.12880"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_dev_193"} +{"question": "Could you provide me some studies about animating images using motion or 3D geometry priors?", "answer": ["Dense Optical Flow Prediction from a Static Image", "Visual Dynamics: Stochastic Future Generation via Layered Cross\n Convolutional Networks", "Photo Wake-Up: 3D Character Animation from a Single Photo", "Thin-Plate Spline Motion Model for Image Animation", "Dimensions of Motion: Monocular Prediction through Flow Subspaces", "Implicit Warping for Animation with Image Sets", "Conditional Image-to-Video Generation with Latent Flow Diffusion Models", "Animating Pictures with Eulerian Motion Fields", "Controllable Animation of Fluid Elements in Still Images", "Animating Landscape: Self-Supervised Learning of Decoupled Motion and\n Appearance for Single-Image Video Synthesis", "Novel View Synthesis with Diffusion Models", "Water Simulation and Rendering from a Still Photograph"], "answer_arxiv_id": ["1505.00295", "1807.09245", "1812.02246", "2203.14367", "2112.01502", "2210.01794", "2303.13744", "2011.15128", "2112.03051", "1910.07192", "2210.04628", "2210.02553"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_dev_194"} +{"question": "What works are about generating arguments for answering comparative questions?", "answer": ["Aspect-Controllable Opinion Summarization"], "answer_arxiv_id": ["2109.03171"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_195"} +{"question": "Could you provide some research papers that recourse to adversarial imitation learning to handle challenges in behavior cloning?", "answer": ["Symphony: Learning Realistic and Diverse Agents for Autonomous Driving Simulation", "Hierarchical Model-Based Imitation Learning for Planning in Autonomous Driving"], "answer_arxiv_id": ["2205.03195", "2210.09539"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_dev_196"} +{"question": "Which works decomposed the reconstructed volume into geometry, SVBRDF, and illumination?", "answer": ["NeRV: Neural Reflectance and Visibility Fields for Relighting and View Synthesis", "Neural-PIL: Neural Pre-Integrated Lighting for Reflectance Decomposition"], "answer_arxiv_id": ["2012.03927", "2110.14373"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_dev_197"} +{"question": "What research indicates that the trade-off between adversarial robustness and accuracy can be attributed to current adversarial training algorithms?", "answer": ["A Closer Look at Accuracy vs. Robustness", "Robustness and Accuracy Could Be Reconcilable by (Proper) Definition"], "answer_arxiv_id": ["2003.02460", "2202.10103"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_198"} +{"question": "What paper estimates the inner integral with a quadrature of N samples from a Q-network?", "answer": ["All-Action Policy Gradient Methods: A Numerical Integration Approach"], "answer_arxiv_id": ["1910.09093"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_dev_199"} +{"question": "What works fall under the category of post-training quantization (PTQ)?", "answer": ["Post-Training Quantization for Vision Transformer", "Up or Down? Adaptive Rounding for Post-Training Quantization", "Post training 4-bit quantization of convolutional networks for rapid-deployment", "Data-Free Quantization Through Weight Equalization and Bias Correction"], "answer_arxiv_id": ["2106.14156", "2004.10568", "1810.05723", "1906.04721"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_dev_200"} +{"question": "Could you tell me which works have optimized the mixing weights of the source and target text embeddings for disentangled image editing?", "answer": ["Uncovering the Disentanglement Capability in Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2212.08698"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_201"} +{"question": "Which works tried to reduce the dependency on densely collected data for scene reconstruction by utilizing local semantic relationships across multiple scenes?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "Stereo Radiance Fields (SRF): Learning View Synthesis for Sparse Views\n of Novel Scenes"], "answer_arxiv_id": ["2012.02190", "2104.06935"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_202"} +{"question": "What papers have been written on the uses of the Optimal Transport map in generative modeling?", "answer": ["Generative Modeling with Optimal Transport Maps"], "answer_arxiv_id": ["2110.02999"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_203"} +{"question": "Any works on the development of counter-measures to embedding inversion attacks?", "answer": ["When Federated Learning Meets Pre-trained Language Models'\n Parameter-Efficient Tuning Methods", "TextHide: Tackling Data Privacy in Language Understanding Tasks", "Differentially Private Representation for NLP: Formal Guarantee and An\n Empirical Study on Privacy and Fairness"], "answer_arxiv_id": ["2212.10025", "2010.06053", "2010.01285"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_dev_204"} +{"question": "Which studies used random features for node embeddings and node classification tasks?", "answer": ["Taming graph kernels with random features"], "answer_arxiv_id": ["2305.00156"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_205"} +{"question": "Could you provide me some research that consider global properties of the networks?", "answer": ["Size-Independent Sample Complexity of Neural Networks"], "answer_arxiv_id": ["1712.06541"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_dev_206"} +{"question": "Could you provide some references about global methods for offering sample-independent meaningful perturbations for each latent space?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "GANSpace: Discovering Interpretable GAN Controls", "Closed-Form Factorization of Latent Semantics in GANs"], "answer_arxiv_id": ["2103.17249", "2004.02546", "2007.06600v4"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_dev_207"} +{"question": "Which studies are there on pre-trained models for vision tasks?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Scaling Vision Transformers"], "answer_arxiv_id": ["2010.11929", "2106.04560"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_dev_208"} +{"question": "Which study introduces CLoM and CCLoM loss terms for notable cross-architecture improvements in distillation methods?", "answer": ["Can pre-trained models assist in dataset distillation?"], "answer_arxiv_id": ["2310.03295"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_209"} +{"question": "Do any subsequent studies existed that have improved upon the concept of implicit data augmentation for image classification tasks?", "answer": ["Imagine by Reasoning: A Reasoning-Based Implicit Semantic Data\n Augmentation for Long-Tailed Classification", "MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition", "Implicit Counterfactual Data Augmentation for Deep Neural Networks"], "answer_arxiv_id": ["2112.07928", "2103.12579", "2304.13431"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_dev_210"} +{"question": "Can you provide me the papers that applied learning invariant representations for multilingual machine translation?", "answer": ["Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation", "Massively Multilingual Neural Machine Translation", "On Learning Language-Invariant Representations for Universal Machine Translation"], "answer_arxiv_id": ["1611.04558", "1903.00089", "2008.04510"], "source_meta": {"published_time": "20201219"}, "qid": "AutoScholarQuery_dev_211"} +{"question": "What studies focused on training with the knowledge of some attacks in the context of adversarial detection?", "answer": ["On Detecting Adversarial Perturbations", "Detecting Adversarial Samples from Artifacts", "GraN: An Efficient Gradient-Norm Based Detector for Adversarial and Misclassified Examples", "Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning", "Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality"], "answer_arxiv_id": ["1702.04267", "1703.00410", "2004.09179", "1803.04765", "1801.02613"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_212"} +{"question": "Which paper initially demonstrated that contrastive learning framework can achieve a performance comparable to fully supervised baselines?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_dev_213"} +{"question": "Can you name a few recent models that have contributed significantly in the field of image captioning?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "CoCa: Contrastive Captioners are Image-Text Foundation Models"], "answer_arxiv_id": ["2201.12086", "2301.12597", "2204.14198", "2205.01917"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_214"} +{"question": "What works introduced temporal transformers for frame-level relationship encoding in action recognition?", "answer": ["Video Transformer Network", "An Image is Worth 16x16 Words, What is a Video Worth?"], "answer_arxiv_id": ["2102.00719", "2103.13915"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_215"} +{"question": "Could you provide a study that discusses how arbitrary input features could influence its reasoning process when generating the explanation, which could result in different reasoning processes for explanation and prediction, and hide the underlying drivers of the prediction?", "answer": ["Language Models Don't Always Say What They Think: Unfaithful\n Explanations in Chain-of-Thought Prompting"], "answer_arxiv_id": ["2305.04388"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_216"} +{"question": "What work uses the scene graph structure to generate reasoning questions on real-world images to test the compositional reasoning ability?", "answer": ["GQA: A New Dataset for Real-World Visual Reasoning and Compositional\n Question Answering"], "answer_arxiv_id": ["1902.09506"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_217"} +{"question": "What research introduced memory efficient schemes in the field of optimization-based meta-learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "On First-Order Meta-Learning Algorithms", "Meta-Learning with Implicit Gradients", "Large-Scale Meta-Learning with Continual Trajectory Shifting"], "answer_arxiv_id": ["1703.03400", "1803.02999", "1909.04630", "2102.07215"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_dev_218"} +{"question": "Any works about adapting pretrained vision-language model for fine-grained localization tasks?", "answer": ["Extract Free Dense Labels from CLIP", "CLIP Surgery for Better Explainability with Enhancement in\n Open-Vocabulary Tasks"], "answer_arxiv_id": ["2112.01071", "2304.05653"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_dev_219"} +{"question": "Could you provide me studies where it was hypothesized that attention can alleviate oversmoothing in attention-based GNNs?", "answer": ["Scattering GCN: Overcoming Oversmoothness in Graph Convolutional Networks"], "answer_arxiv_id": ["2003.08414"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_dev_220"} +{"question": "Which work first initiated 3D human geometry via a shape VAE network for 3D avatar generation?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars"], "answer_arxiv_id": ["2205.08535"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_dev_221"} +{"question": "Which papers have studied Meta-RL methods based on gradients?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-Reinforcement Learning of Structured Exploration Strategies"], "answer_arxiv_id": ["1703.03400", "1802.07245"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_dev_222"} +{"question": "What pioneer work uses deep learning models for generating realistic LiDAR point clouds?", "answer": ["LiDARsim: Realistic LiDAR Simulation by Leveraging the Real World"], "answer_arxiv_id": ["2006.09348"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_223"} +{"question": "Which works used Generative Adversarial Networks for text-to-image generation?", "answer": ["Generative Adversarial Text to Image Synthesis", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "CLIP2StyleGAN: Unsupervised Extraction of StyleGAN Edit Directions", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators"], "answer_arxiv_id": ["1605.05396", "1711.10485", "2112.05219", "2108.00946"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_dev_224"} +{"question": "Could you provide me with some works that converted preprocessed neuroimages to brain network datasets?", "answer": ["Explainable Classification of Brain Networks via Contrast Subgraphs", "TUDataset: A collection of benchmark datasets for learning with graphs"], "answer_arxiv_id": ["2006.05176", "2007.08663"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_dev_225"} +{"question": "Which works use dataset pruning by keeping hard samples with maximum entropy?", "answer": ["Selection via Proxy: Efficient Data Selection for Deep Learning"], "answer_arxiv_id": ["1906.11829"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_dev_226"} +{"question": "What paper estimated CMI by reformulating it as a minmax optimization problem?", "answer": ["C-MI-GAN : Estimation of Conditional Mutual Information Using MinMax Formulation"], "answer_arxiv_id": ["2005.08226"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_227"} +{"question": "Which papers suggest that group-wise quantization approaches can achieve higher accuracy compared to layer-wise or channel-wise methods?", "answer": ["Q-BERT: Hessian Based Ultra Low Precision Quantization of BERT", "GPTQ: Accurate Post-Training Quantization for Generative Pre-trained\n Transformers"], "answer_arxiv_id": ["1909.05840", "2210.17323"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_228"} +{"question": "Could you provide me examples of literature focusing on improving detection robustness to paraphrasing attacks?", "answer": ["On the Reliability of Watermarks for Large Language Models", "A Semantic Invariant Robust Watermark for Large Language Models"], "answer_arxiv_id": ["2306.04634", "2310.06356"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_229"} +{"question": "Could you provide me some works that applied joint-embedding strategy and generative approaches for self-supervised vision transformer?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "An Empirical Study of Training Self-Supervised Vision Transformers", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2104.14294", "2104.02057", "2111.06377"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_230"} +{"question": "What works demonstrated the usefulness of emergent relationships for learning fine-grained visual features?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "Relational Proxies: Emergent Relationships as Fine-Grained Discriminators"], "answer_arxiv_id": ["2104.14294", "2210.02149"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_dev_231"} +{"question": "Which papers proposed concept-based models for few-shot learning settings?", "answer": ["Concept Learners for Few-Shot Learning"], "answer_arxiv_id": ["2007.07375"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_dev_232"} +{"question": "Which papers provided the basis for the researcher's work on Transformer for program synthesis?", "answer": ["Unsupervised Translation of Programming Languages", "Program Synthesis with Large Language Models"], "answer_arxiv_id": ["2006.03511", "2108.07732"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_dev_233"} +{"question": "Could you provide me some works about the automatic debiased machine learner (Auto-DML) approach which can handle continuous treatments in the back-door adjustment?", "answer": ["Automatic Debiased Machine Learning of Causal and Structural Effects"], "answer_arxiv_id": ["1809.05224"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_dev_234"} +{"question": "Are there any studies that utilized differential equations to model continuous-time processes?", "answer": ["Neural Ordinary Differential Equations", "Learning Long-Term Dependencies in Irregularly-Sampled Time Series", "Neural Controlled Differential Equations for Irregular Time Series"], "answer_arxiv_id": ["1806.07366", "2006.04418", "2005.08926"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_dev_235"} +{"question": "What study proposed a method to specialize the model’s ability towards a target task with CoT prompting?", "answer": ["Specializing Smaller Language Models towards Multi-Step Reasoning"], "answer_arxiv_id": ["2301.12726"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_dev_236"} +{"question": "Can you name a study that employs selective search to crop out Regions of Interest for dense prediction?", "answer": ["Aligning Pretraining for Detection via Object-Level Contrastive Learning", "Unsupervised Object-Level Representation Learning from Scene Images"], "answer_arxiv_id": ["2106.02637", "2106.11952"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_237"} +{"question": "Which dataset synthetically generates commands for device-control tasks?", "answer": ["Mapping Natural Language Instructions to Mobile UI Action Sequences"], "answer_arxiv_id": ["2005.03776"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_238"} +{"question": "What papers are about the utilization of knowledge for reasoning purposes?", "answer": ["A Mechanistic Interpretation of Arithmetic Reasoning in Language Models\n using Causal Mediation Analysis", "Towards a Mechanistic Interpretation of Multi-Step Reasoning\n Capabilities of Language Models"], "answer_arxiv_id": ["2305.15054", "2310.14491"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_239"} +{"question": "Which research work proposes techniques to offload part of the model to the server using split learning?", "answer": ["FedAdapt: Adaptive Offloading for IoT Devices in Federated Learning", "Group Knowledge Transfer: Federated Learning of Large CNNs at the Edge", "FedFly: Towards Migration in Edge-based Distributed Federated Learning"], "answer_arxiv_id": ["2107.04271", "2007.14513", "2111.01516"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_dev_240"} +{"question": "What research introduced Meta-Prompting, a method that breaks down complex tasks into subtasks?", "answer": ["Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding"], "answer_arxiv_id": ["2401.12954"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_dev_241"} +{"question": "What work aims to approximate the homotopy only enough to generate the conformal prediction set but suffers from accuracy difficulties?", "answer": ["Computing Full Conformal Prediction Set with Approximate Homotopy"], "answer_arxiv_id": ["1909.09365"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_dev_242"} +{"question": "Which studies used techniques like spatial attention maps and an iterative refinement strategy?", "answer": ["Cascading Convolutional Color Constancy"], "answer_arxiv_id": ["1912.11180"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_243"} +{"question": "What paper first connected U-Nets and multi-resolution analysis?", "answer": ["A Multi-Resolution Framework for U-Nets with Applications to Hierarchical VAEs"], "answer_arxiv_id": ["2301.08187"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_244"} +{"question": "What early studies on mechanistic interpretability focused on how the model stores factual knowledge internally?", "answer": ["Locating and Editing Factual Associations in GPT", "Knowledge Neurons in Pretrained Transformers"], "answer_arxiv_id": ["2202.05262", "2104.08696"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_245"} +{"question": "What are some recent studies that leverage LLMs as Emotional Support Chat(ECS) systems through in-context learning?", "answer": ["Controllable Mixed-Initiative Dialogue Generation through Prompting", "Building Emotional Support Chatbots in the Era of LLMs"], "answer_arxiv_id": ["2305.04147", "2308.11584"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_dev_246"} +{"question": "What works introduce the feasibility of creating adversarial examples that can break LMMs?", "answer": ["Visual Adversarial Examples Jailbreak Aligned Large Language Models", "Are aligned neural networks adversarially aligned?", "Universal and Transferable Adversarial Attacks on Aligned Language\n Models", "Jailbroken: How Does LLM Safety Training Fail?", "Jailbreaking ChatGPT via Prompt Engineering: An Empirical Study"], "answer_arxiv_id": ["2306.13213", "2306.15447", "2307.15043", "2307.02483", "2305.13860v2"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_247"} +{"question": "Which work introduced Optima-TT, a promising algorithm for multidimensional optimization?", "answer": ["Optimization of Functions Given in the Tensor Train Format"], "answer_arxiv_id": ["2209.14808"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_dev_248"} +{"question": "Can you provide some papers that discuss how LLMs reproduce human-like text and also replicate biases present in the training data?", "answer": ["Persistent Anti-Muslim Bias in Large Language Models"], "answer_arxiv_id": ["2101.05783"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_249"} +{"question": "Which papers introduced semi-supervised methods that explore the entire eigenspectrum?", "answer": ["Adaptive Universal Generalized PageRank Graph Neural Network", "BernNet: Learning Arbitrary Graph Spectral Filters via Bernstein Approximation", "A Piece-wise Polynomial Filtering Approach for Graph Neural Networks"], "answer_arxiv_id": ["2006.07988", "2106.10994", "2112.03499"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_dev_250"} +{"question": "What studies use self-supervised and supervised methods to train UI understanding models?", "answer": ["ActionBert: Leveraging User Actions for Semantic Understanding of User Interfaces", "UIBert: Learning Generic Multimodal Representations for UI Understanding", "Lexi: Self-Supervised Learning of the UI Language", "Unblind Your Apps: Predicting Natural-Language Labels for Mobile GUI Components by Deep Learning", "Object Detection for Graphical User Interface: Old Fashioned or Deep Learning or a Combination?", "Screen Recognition: Creating Accessibility Metadata for Mobile Applications from Pixels"], "answer_arxiv_id": ["2012.12350", "2107.13731", "2301.10165", "2003.00380", "2008.05132", "2101.04893"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_251"} +{"question": "Can you give examples of studies on drug-sized molecules with respect to voxel and point cloud representations?", "answer": ["GEOM: Energy-annotated molecular conformations for property prediction and molecular generation"], "answer_arxiv_id": ["2006.05531"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_252"} +{"question": "Which studies discussed the interaction between privacy and model explainability?", "answer": ["When Differential Privacy Meets Interpretability: A Case Study", "On the Privacy Risks of Model Explanations"], "answer_arxiv_id": ["2106.13203", "1907.00164"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_dev_253"} +{"question": "Which work proposed using the attribution technique for detecting skill neurons?", "answer": ["Task-specific Compression for Multi-task Language Models using\n Attribution-based Pruning"], "answer_arxiv_id": ["2205.04157"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_dev_254"} +{"question": "What is the first work focusing on preconditioning of on-policy, linear, least-squares forms of TD?", "answer": ["Preconditioned Temporal Difference Learning"], "answer_arxiv_id": ["0704.1409v3"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_255"} +{"question": "Which works are about using GANs for scene text editing tasks?", "answer": ["STEFANN: Scene Text Editor using Font Adaptive Neural Network", "GenText: Unsupervised Artistic Text Generation via Decoupled Font and Texture Manipulation", "Look Closer to Supervise Better: One-Shot Font Generation via Component-Based Discriminator", "RewriteNet: Reliable Scene Text Editing with Implicit Decomposition of Text Contents and Styles", "De-rendering Stylized Texts", "SwapText: Image Based Texts Transfer in Scenes", "Spatial Fusion GAN for Image Synthesis"], "answer_arxiv_id": ["1903.01192", "2207.09649", "2205.00146", "2107.11041", "2110.01890", "2003.08152", "1812.05840"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_dev_256"} +{"question": "What studies extended their gradient-based recovery strategy from MLPs to Convolutional Neural Networks?", "answer": ["Exploring the Security Boundary of Data Reconstruction via Neuron Exclusivity Analysis", "R-GAP: Recursive Gradient Attack on Privacy", "When the Curious Abandon Honesty: Federated Learning Is Not Private⋄"], "answer_arxiv_id": ["2010.13356", "2010.07733", "2112.02918"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_257"} +{"question": "Can you provide works that have used Chain of Thought (CoT) in multi-step reasoning tasks?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Complexity-Based Prompting for Multi-Step Reasoning"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2210.00720"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_258"} +{"question": "Can you provide some studies that pointed out inherent ambiguity in questions from human users and proposed a benchmark that provides multiple answers to every question?", "answer": ["AmbigQA: Answering Ambiguous Open-domain Questions"], "answer_arxiv_id": ["2004.10645"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_dev_259"} +{"question": "Which work introduces error span detection and correction to address the GEC problem?", "answer": ["Improving the Efficiency of Grammatical Error Correction with Erroneous\n Span Detection and Correction"], "answer_arxiv_id": ["2010.03260"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_dev_260"} +{"question": "Any studies that attempted to construct 3D feature fields?", "answer": ["Decomposing NeRF for Editing via Feature Field Distillation", "Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D\n Image Representations", "Distilled Feature Fields Enable Few-Shot Language-Guided Manipulation", "In-Place Scene Labelling and Understanding with Implicit Scene\n Representation", "Panoptic Lifting for 3D Scene Understanding with Neural Fields"], "answer_arxiv_id": ["2205.15585", "2209.03494", "2308.07931", "2103.15875", "2212.09802"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_dev_261"} +{"question": "Which work introduced the concept of a diachronic word usage graph (DWUG)?", "answer": ["DWUG: A large Resource of Diachronic Word Usage Graphs in Four Languages"], "answer_arxiv_id": ["2104.08540"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_dev_262"} +{"question": "Could you provide me some studies about transformers based architectures for BEV feature generation?", "answer": ["PersFormer: 3D Lane Detection via Perspective Transformer and the\n OpenLane Benchmark", "FIERY: Future Instance Prediction in Bird's-Eye View from Surround\n Monocular Cameras", "BEVerse: Unified Perception and Prediction in Birds-Eye-View for\n Vision-Centric Autonomous Driving", "BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers"], "answer_arxiv_id": ["2203.11089", "2104.10490", "2205.09743", "2203.17270"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_dev_263"} +{"question": "What works integrate Knowledge Graph (KG) embedding methods using Graph Neural Networks (GNNs) with Language Learning Models (LLMs) during finetuning or pre-training stage?", "answer": ["QA-GNN: Reasoning with Language Models and Knowledge Graphs for Question\n Answering", "Deep Bidirectional Language-Knowledge Graph Pretraining"], "answer_arxiv_id": ["2104.06378", "2210.09338"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_dev_264"} +{"question": "Could you tell me about the studies that used volumetric entropy for scene geometry in NeRF?", "answer": ["Uncertainty Guided Policy for Active Robotic 3D Reconstruction using\n Neural Radiance Fields", "Active Implicit Object Reconstruction using Uncertainty-guided\n Next-Best-View Optimization"], "answer_arxiv_id": ["2209.08409", "2303.16739"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_265"} +{"question": "Which papers introduced neural construction methods for combinatorial optimization?", "answer": ["Pointer Networks", "Neural Combinatorial Optimization with Reinforcement Learning", "Reinforcement Learning for Solving the Vehicle Routing Problem"], "answer_arxiv_id": ["1506.03134", "1611.09940", "1802.04240"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_dev_266"} +{"question": "In what works LVLMs provide bounding boxes for objects while generating responses?", "answer": ["Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Kosmos-2: Grounding Multimodal Large Language Models to the World"], "answer_arxiv_id": ["2306.15195", "2306.14824"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_dev_267"} +{"question": "What papers consider the convergence of PFL with a partial client participation?", "answer": ["On the Convergence of FedAvg on Non-IID Data", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Achieving ​ Linear ​ Speedup ​ with ​​ Partial ​​ Worker ​​ Participation ​​ in ​​ Non-IID ​​ Federated ​​ Learning"], "answer_arxiv_id": ["1907.02189", "1910.06378", "2101.11203"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_268"} +{"question": "What works achieved state-of-the-art performance on generating unaligned complex objects using big generative models?", "answer": ["Zero-Shot Text-to-Image Generation", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Muse: Text-To-Image Generation via Masked Generative Transformers"], "answer_arxiv_id": ["2102.12092", "2206.10789", "2204.06125", "2205.11487", "2112.10752", "2211.01324", "2301.00704"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_dev_269"} +{"question": "Could you provide me some works that utilized a combination of strong data augmentation functions to improve robustness to common corruptions and random Gaussian noises?", "answer": ["AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "RandAugment: Practical automated data augmentation with a reduced search space"], "answer_arxiv_id": ["1912.02781", "1909.13719"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_dev_270"} +{"question": "Could you provide me some studies that have used unlearning to remove sensitive data from a trained model?", "answer": ["Making AI Forget You: Data Deletion in Machine Learning", "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep\n Networks", "Remember What You Want to Forget: Algorithms for Machine Unlearning", "Adaptive Machine Unlearning", "Deep Unlearning via Randomized Conditionally Independent Hessians"], "answer_arxiv_id": ["1907.05012", "1911.04933", "2103.03279", "2106.04378", "2204.07655"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_dev_271"} +{"question": "Could you provide me some papers that compute closed-form solutions to related or approximate versions of CMDP framework?", "answer": ["Constrained Policy Optimization", "Constrained Variational Policy Optimization for Safe Reinforcement Learning"], "answer_arxiv_id": ["1705.10528", "2201.11927"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_272"} +{"question": "Which work introduced the Factor-VAE that encourages disentanglement through a factorial distribution of features?", "answer": ["Disentangling by Factorising"], "answer_arxiv_id": ["1802.05983"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_273"} +{"question": "Could you provide me some works that discuss how LLM can explore and manipulate various attributes of texts?", "answer": ["Controllable Data Augmentation for Few-Shot Text Mining with\n Chain-of-Thought Attribute Manipulation"], "answer_arxiv_id": ["2307.07099"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_dev_274"} +{"question": "In what papers were the minimax regret rates first established for two-armed bandit and generalized to any number of arms?", "answer": ["Nonparametric Bandits with Covariates", "The multi-armed bandit problem with covariates"], "answer_arxiv_id": ["1003.1630", "1110.6084"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_dev_275"} +{"question": "What work uses gradients for data selection?", "answer": ["Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training", "Glister: Generalization based Data Subset Selection for Efficient and Robust Learning", "Optimizing Data Usage via Differentiable Rewards", "Deep Learning on a Data Diet: Finding Important Examples Early in Training", "Coresets for Data-efficient Training of Machine Learning Models"], "answer_arxiv_id": ["2103.00123", "2012.10630", "1911.10088", "2107.07075", "1906.01827"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_dev_276"} +{"question": "Which studies expanded tool use to general API function calling?", "answer": ["ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs", "ToolAlpaca: Generalized Tool Learning for Language Models with 3000\n Simulated Cases", "Gorilla: Large Language Model Connected with Massive APIs"], "answer_arxiv_id": ["2307.16789", "2306.05301", "2305.15334"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_dev_277"} +{"question": "Could you provide me some studies about the difficulties in finding reliable confidence thresholds in self-distillation?", "answer": ["Born Again Neural Networks", "Noisy Self-Knowledge Distillation for Text Summarization"], "answer_arxiv_id": ["1805.04770", "2009.07032"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_dev_278"} +{"question": "Which work first popularized the use of neural fields as a method of representing 3D scenes and objects?", "answer": ["Scene Representation Networks: Continuous 3D-Structure-Aware Neural\n Scene Representations"], "answer_arxiv_id": ["1906.01618"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_dev_279"} +{"question": "Could you provide me with some studies on soft prompts?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "What Makes Good In-Context Examples for GPT-3?"], "answer_arxiv_id": ["2104.08691", "2101.06804"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_280"} +{"question": "Could you provide me some works on top-down approaches for instance segmentation?", "answer": ["Learning Object Bounding Boxes for 3D Instance Segmentation on Point\n Clouds", "GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in\n Point Cloud", "3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans", "RevealNet: Seeing Behind Objects in RGB-D Scans"], "answer_arxiv_id": ["1906.01140", "1812.03320", "1812.07003", "1904.12012"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_dev_281"} +{"question": "Which works propose the use of a smaller threshold and a bias shift to better match the activation in ANN-SNN conversion?", "answer": ["RMP-SNN: Residual Membrane Potential Neuron for Enabling Deeper High-Accuracy and Low-Latency Spiking Neural Network", "Optimal Conversion of Conventional Artificial Neural Networks to Spiking Neural Networks"], "answer_arxiv_id": ["2003.01811", "2103.00476"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_dev_282"} +{"question": "Are there any papers where the tasks were modeled as sequence generation tasks within a unified paradigm?", "answer": ["Uni-Perceiver: Pre-training Unified Architecture for Generic Perception\n for Zero-shot and Few-shot Tasks", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "A Unified Sequence Interface for Vision Tasks", "UniTAB: Unifying Text and Box Outputs for Grounded Vision-Language\n Modeling"], "answer_arxiv_id": ["2112.01522", "2202.03052", "2206.08916", "2206.07669", "2111.12085"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_283"} +{"question": "Who developed a triangular attention to deduce relations from other relations?", "answer": ["Systematic Generalization with Edge Transformers"], "answer_arxiv_id": ["2112.00578"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_dev_284"} +{"question": "What studies create datasets using templates?", "answer": ["Large Language Model as Attributed Training Data Generator: A Tale of\n Diversity and Bias"], "answer_arxiv_id": ["2306.15895"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_dev_285"} +{"question": "Which studies discuss adaptation for better generalization in model's optimization and training?", "answer": ["Domain Generalization via Invariant Feature Representation", "Generalizing to Unseen Domains: A Survey on Domain Generalization"], "answer_arxiv_id": ["1301.2115", "2103.03097"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_286"} +{"question": "Could you provide me some works that introduced residual networks?", "answer": ["Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1512.03385"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_dev_287"} +{"question": "What research papers propose variants for regression learning that divide the regression range into small bins?", "answer": ["Deep Label Distribution Learning With Label Ambiguity"], "answer_arxiv_id": ["1611.01731"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_dev_288"} +{"question": "Which works used a pre-trained object detector to obtain region-of-interest (ROI) features from images?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "UNIMO: Towards Unified-Modal Understanding and Generation via\n Cross-Modal Contrastive Learning", "Product1M: Towards Weakly Supervised Instance-Level Product Retrieval\n via Cross-modal Pretraining"], "answer_arxiv_id": ["1909.11740", "2004.06165", "2012.15409", "2107.14572"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_dev_289"} +{"question": "Can you point out studies that applied hypernetworks to supervised learning?", "answer": ["HyperNetworks", "Principled Weight Initialization for Hypernetworks"], "answer_arxiv_id": ["1609.09106v4", "2312.08399"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_290"} +{"question": "Could you name any works that utilized variance bounds of the local objective gradients to capture data heterogeneity?", "answer": ["SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "A Unified Theory of Decentralized SGD with Changing Topology and Local Updates"], "answer_arxiv_id": ["1910.06378", "2003.10422"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_291"} +{"question": "Which studies focus on monocular building height estimation using deep neural networks?", "answer": ["THE Benchmark: Transferable Representation Learning for Monocular Height\n Estimation"], "answer_arxiv_id": ["2112.14985"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_dev_292"} +{"question": "Which research focused on understanding the situations where models may obtain correct answers through unfaithful or spurious reasoning shortcuts?", "answer": ["Language Models Are Greedy Reasoners: A Systematic Formal Analysis of Chain-of-Thought"], "answer_arxiv_id": ["2210.01240v4"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_dev_293"} +{"question": "What work proposed an efficient citation-based QA approach by fine-tuning much smaller LLMs?", "answer": ["WebGLM: Towards An Efficient Web-Enhanced Question Answering System with\n Human Preferences"], "answer_arxiv_id": ["2306.07906"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_dev_294"} +{"question": "Could you provide me some works in which fine-tuning was applied to learn new knowledge for model editing?", "answer": ["Modifying Memories in Transformer Models", "Editing Large Language Models: Problems, Methods, and Opportunities"], "answer_arxiv_id": ["2012.00363", "2305.13172"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_295"} +{"question": "Which works demonstrate the effectiveness of fine-tuning diffusion models for personalized image generation?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2106.09685", "2208.12242"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_dev_296"} +{"question": "Are there any studies that highlight the need of model simplification when working with causal models?", "answer": ["Causal Consistency of Structural Equation Models", "Multi-Level Cause-Effect Systems", "Jointly Learning Consistent Causal Abstractions Over Multiple Interventional Distributions", "Weakly supervised causal representation learning"], "answer_arxiv_id": ["1707.00819", "1512.07942v1", "2301.05893", "2203.16437"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_dev_297"} +{"question": "Which works are about the application of orthogonal polynomials in machine learning?", "answer": ["Do RNN and LSTM have Long Memory?", "Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling", "Multi-Dimensional Recurrent Neural Networks"], "answer_arxiv_id": ["2006.03860", "1412.3555", "0705.2011"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_dev_298"} +{"question": "Which works carried out the developments in contrastive learning using the InfoNCE loss with augmented data pairs?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Contrastive Multiview Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Residual Relaxation for Multi-view Representation Learning", "Big Self-Supervised Models are Strong Semi-Supervised Learners", "A Mutual Information Maximization Perspective of Language Representation Learning"], "answer_arxiv_id": ["1807.03748", "1906.05849", "2002.05709", "1911.05722", "2110.15348", "2006.10029", "1910.08350"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_dev_299"} +{"question": "Could you provide me some studies that develop BERT-based Transformers for code syntax?", "answer": ["Learning and Evaluating Contextual Embedding of Source Code", "CodeBERT: A Pre-Trained Model for Programming and Natural Languages", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "GraphCodeBERT: Pre-training Code Representations with Data Flow"], "answer_arxiv_id": ["2001.00059", "2002.08155", "1810.04805", "2009.08366"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_dev_300"} +{"question": "Can you name the papers that considered non-stationary RL under linear MDPs?", "answer": ["Nonstationary Reinforcement Learning with Linear Function Approximation", "Efficient Learning in Non-Stationary Linear Markov Decision Processes"], "answer_arxiv_id": ["2010.04244v3", "2010.12870"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_dev_301"} +{"question": "What papers suggested practical approaches to mitigate the oversmoothing problem in deep GNNs?", "answer": ["Simple and Deep Graph Convolutional Networks", "PairNorm: Tackling Oversmoothing in GNNs", "Representation Learning on Graphs with Jumping Knowledge Networks", "Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "DropEdge: Towards Deep Graph Convolutional Networks on Node Classification", "Bayesian Graph Neural Networks with Adaptive Connection Sampling", "Diffusion Improves Graph Learning"], "answer_arxiv_id": ["2007.02133", "1909.12223", "1806.03536v2", "1810.05997", "1907.10903", "2006.04064v3", "1911.05485"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_dev_302"} +{"question": "Does any work define multiple emergent tasks of interest to evaluate the performance of multi-modal LLMs?", "answer": ["MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities"], "answer_arxiv_id": ["2308.02490"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_303"} +{"question": "In what papers does the researcher address models’ resilience against various distribution shifts?", "answer": ["Wilds: A Benchmark of in-the-Wild Distribution Shifts"], "answer_arxiv_id": ["2012.07421"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_304"} +{"question": "Which studies are about establishing a functional mapping using eigenfunctions defined on surfaces?", "answer": ["Deep Geometric Functional Maps: Robust Feature Learning for Shape\n Correspondence"], "answer_arxiv_id": ["2003.14286"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_305"} +{"question": "Which studies introduced the semi-dual approaches that best approximate the OT map?", "answer": ["Generative Modeling with Optimal Transport Maps"], "answer_arxiv_id": ["2110.02999"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_306"} +{"question": "Which studies focus on augmenting the capability for MLLMs to follow visual instructions?", "answer": ["Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Cheap and Quick: Efficient Vision-Language Instruction Tuning for Large\n Language Models", "CogVLM: Visual Expert for Pretrained Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2304.08485", "2305.06500", "2305.15023", "2311.03079", "2304.14178"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_307"} +{"question": "Any works evaluating the potential of using TEES for machine learning in terms of low computing performance in comparison to GPUs?", "answer": ["3LegRace: Privacy-Preserving DNN Training over TEEs and GPUs"], "answer_arxiv_id": ["2110.01229"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_308"} +{"question": "In which paper the researchers introduced a distilling mechanism step-by-step, extracting LLM rationales as additional supervision for training small models within a multi-task framework?", "answer": ["Distilling Step-by-Step! Outperforming Larger Language Models with Less\n Training Data and Smaller Model Sizes"], "answer_arxiv_id": ["2305.02301"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_dev_309"} +{"question": "What are the works using supervised learning for AVS on AVSBench dataset?", "answer": ["Audio-Visual Segmentation", "Multimodal Variational Auto-encoder based Audio-Visual Segmentation", "AVSegFormer: Audio-Visual Segmentation with Transformer", "CATR: Combinatorial-Dependence Audio-Queried Transformer for\n Audio-Visual Video Segmentation", "Audio-aware Query-enhanced Transformer for Audio-Visual Segmentation", "Audio-Visual Segmentation by Exploring Cross-Modal Mutual Semantics", "Annotation-free Audio-Visual Segmentation"], "answer_arxiv_id": ["2207.05042", "2310.08303", "2307.01146", "2309.09709", "2307.13236", "2307.16620", "2305.11019"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_dev_310"} +{"question": "What papers introduced more precise retrieval by decomposing the problem into QDMR format?", "answer": ["Break It Down: A Question Understanding Benchmark"], "answer_arxiv_id": ["2001.11770"], "source_meta": {"published_time": "20240628"}, "qid": "AutoScholarQuery_dev_311"} +{"question": "Which papers have explored the Strong Growth Condition (SGC) assumption to control the rates at which the stochastic gradients decay comparing to the full gradient?", "answer": ["Fast Convergence of Stochastic Gradient Descent under a Strong Growth Condition", "Fast and Faster Convergence of SGD for Over-Parameterized Models and an Accelerated Perceptron"], "answer_arxiv_id": ["1308.6370", "1810.07288"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_dev_312"} +{"question": "What papers focus on methods treating factual knowledge as subject-relation-object tuples for transformers?", "answer": ["Locating and Editing Factual Associations in GPT", "Mass-Editing Memory in a Transformer"], "answer_arxiv_id": ["2202.05262", "2210.07229"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_dev_313"} +{"question": "Could you give me some studies that use an external camera or scene depth estimation for 3D pose estimation with volumetric heatmaps?", "answer": ["Estimating Egocentric 3D Human Pose in the Wild with External Weak\n Supervision", "Scene-aware Egocentric 3D Human Pose Estimation"], "answer_arxiv_id": ["2201.07929", "2212.11684"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_314"} +{"question": "What research has covered the topic of fine-tuning via bootstrapping models in Large Language Models?", "answer": ["STaR: Bootstrapping Reasoning With Reasoning", "Solving Quantitative Reasoning Problems with Language Models"], "answer_arxiv_id": ["2203.14465", "2206.14858"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_dev_315"} +{"question": "What studies discuss readability as a feature for article quality?", "answer": ["Automatic Quality Assessment of Wikipedia Articles -- A Systematic\n Literature Review"], "answer_arxiv_id": ["2310.02235"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_316"} +{"question": "Could you list some works discussing Countsketch-type operators used in tensor decomposition when A has Khatri-Rao product structure?", "answer": ["Fast and Guaranteed Tensor Decomposition via Sketching"], "answer_arxiv_id": ["1506.04448"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_dev_317"} +{"question": "Can you list any studies about the trade-off between model's standard accuracy and robustness?", "answer": ["Robustness May Be at Odds with Accuracy", "Towards Deep Learning Models Resistant to Adversarial Attacks", "AugMax: Adversarial Composition of Random Augmentations for Robust Training"], "answer_arxiv_id": ["1805.12152", "1706.06083", "2110.13771"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_318"} +{"question": "Can you cite some works about the use of chain-of-thought to generate the reasoning process by LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "The Unreliability of Explanations in Few-shot Prompting for Textual\n Reasoning", "Large Language Models are Zero-Shot Reasoners", "Program of Thoughts Prompting: Disentangling Computation from Reasoning\n for Numerical Reasoning Tasks"], "answer_arxiv_id": ["2201.11903", "2203.11171", "2205.03401", "2205.11916", "2211.12588"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_319"} +{"question": "Which works reassessed traditional benchmarks in light of the advent of Multimodal Large Language Models?", "answer": ["Microsoft COCO: Common Objects in Context", "OK-VQA: A Visual Question Answering Benchmark Requiring External\n Knowledge", "GQA: A New Dataset for Real-World Visual Reasoning and Compositional\n Question Answering"], "answer_arxiv_id": ["1405.0312", "1906.00067", "1902.09506"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_dev_320"} +{"question": "What work proposed utilizing a learned adversarial distribution as noise in NCE, and developed the Adversarial Contrastive Estimation method?", "answer": ["Adversarial Contrastive Estimation"], "answer_arxiv_id": ["1805.03642"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_321"} +{"question": "Which work indicates a method for editing pre-existing images based on instructions?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2211.09800"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_dev_322"} +{"question": "Which studies focus on target-agnostic zero-shot object navigation tasks?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Zero Experience Required: Plug & Play Modular Transfer Learning for Semantic Visual Navigation", "Zero-Shot Object Goal Visual Navigation"], "answer_arxiv_id": ["2103.00020", "2202.02440", "2206.07423"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_323"} +{"question": "What studies achieved the universal models by including dihedral angles or by introducing SO(3) equivariant models?", "answer": ["GemNet: Universal Directional Graph Neural Networks for Molecules", "Spherical Message Passing for 3D Molecular Graphs", "Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials"], "answer_arxiv_id": ["2106.08903", "2102.05013", "1802.08219", "2101.03164"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_dev_324"} +{"question": "Which research papers propose deadlock-specialized methods to solve object navigation tasks?", "answer": ["Learning to Learn How to Learn: Self-Adaptive Visual Navigation using Meta-Learning", "Learning Object Relation Graph and Tentative Policy for Visual Navigation"], "answer_arxiv_id": ["1812.00971", "2007.11018"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_325"} +{"question": "Are there any studies that use voxel method for segmenting 3D LiDAR point clouds?", "answer": ["3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "VV-NET: Voxel VAE Net with Group Convolutions for Point Cloud Segmentation", "Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR-based Perception"], "answer_arxiv_id": ["1711.10275", "1811.04337", "2109.05441"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_dev_326"} +{"question": "What work investigates how human users perceive fairness and transparency in recommender systems?", "answer": ["Fairness and Transparency in Recommendation: The Users' Perspective"], "answer_arxiv_id": ["2103.08786"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_dev_327"} +{"question": "Which study proposes a benchmark that covers both image and video modality to evaluate the performance of multi-modal LLMs?", "answer": ["SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension"], "answer_arxiv_id": ["2307.16125"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_328"} +{"question": "What works have been conducted about Electronic Health Records (EHRs) in the field of multi-modal databases?", "answer": ["DrugEHRQA: A Question Answering Dataset on Structured and Unstructured Electronic Health Records For Medicine Related Queries"], "answer_arxiv_id": ["2205.01290"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_dev_329"} +{"question": "Which works investigated the lower bounds of SGD-RR in the quadratic case?", "answer": ["How Good is SGD with Random Shuffling?", "Random Shuffling Beats SGD Only After Many Epochs on Ill-Conditioned Problems"], "answer_arxiv_id": ["1908.00045", "2106.06880"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_330"} +{"question": "Could you provide me some studies about inferring both light transport and density in portrait and face relighting?", "answer": ["NeLF: Neural Light-transport Field for Portrait View Synthesis and Relighting"], "answer_arxiv_id": ["2107.12351"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_dev_331"} +{"question": "Can you provide examples of research that implemented unsupervised OOD detection methods?", "answer": ["Deep Structured Energy Based Models for Anomaly Detection", "Do Deep Generative Models Know What They Don’t Know?", "WAIC, but Why? Generative Ensembles for Robust Anomaly Detection", "Implicit Generation and Modeling with Energy-Based Models", "Likelihood Ratios for Out-of-Distribution Detection", "Input Complexity and Out-of-distribution Detection with Likelihood-based Generative Models", "Your classifier is secretly an energy based model and you should treat it like one", "Energy-based Out-of-distribution Detection", "Density estimation using Real NVP", "Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery", "Generative Probabilistic Novelty Detection with Adversarial Autoencoders", "OCGAN: One-class Novelty Detection Using GANs with Constrained Latent Representations", "WAIC, but Why? Generative Ensembles for Robust Anomaly Detection", "Deep Anomaly Detection Using Geometric Transformations", "Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty", "Classification-Based Anomaly Detection for General Data", "CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances", "ViM: Out-Of-Distribution with Virtual-logit Matching"], "answer_arxiv_id": ["1605.07717", "1810.09136", "1810.01392", "1903.08689", "1906.02845", "1909.11480", "1912.03263", "2010.03759", "1605.08803", "1703.05921", "1807.02588", "1903.08550", "1810.01392", "1805.10917", "1906.12340", "2005.02359", "2007.08176", "2203.10807"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_332"} +{"question": "Can you provide me with some works that thoroughly used geometric insights on distances to surfaces in the context of classification?", "answer": ["Robustness of classifiers: from adversarial to random noise", "Are adversarial examples inevitable?"], "answer_arxiv_id": ["1608.08967", "1809.02104"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_333"} +{"question": "What study demonstrated that dividing complex problems into simpler sub-problems improved the performance of CoT prompting?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models"], "answer_arxiv_id": ["2205.10625"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_334"} +{"question": "Which paper proposed to constrain the weights in linear layers as lower triangular matrices to speed up training?", "answer": ["MintNet: Building Invertible Neural Networks with Masked Convolutions"], "answer_arxiv_id": ["1907.07945"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_335"} +{"question": "What are the cited examples of citation-based QA systems?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback", "Teaching language models to support answers with verified quotes", "WebGLM: Towards An Efficient Web-Enhanced Question Answering System with\n Human Preferences"], "answer_arxiv_id": ["2112.09332", "2203.11147", "2306.07906"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_dev_336"} +{"question": "Could you provide me some studies about HMT methods focusing on monocular images?", "answer": ["Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a\n Single Image", "End-to-end Recovery of Human Shape and Pose", "Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "Learning 3D Human Dynamics from Video", "VIBE: Video Inference for Human Body Pose and Shape Estimation", "PARE: Part Attention Regressor for 3D Human Body Estimation", "3D Human Pose Estimation via Intuitive Physics", "Neural MoCon: Neural Motion Control for Physically Plausible Human\n Motion Capture", "NIKI: Neural Inverse Kinematics with Invertible Neural Networks for 3D\n Human Pose and Shape Estimation"], "answer_arxiv_id": ["1607.08128", "1712.06584", "1904.05866", "1812.01601", "1912.05656", "2104.08527", "2303.18246", "2203.14065", "2305.08590"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_337"} +{"question": "Which research papers investigated techniques for demonstration retrieval as part of improving in-context learning?", "answer": ["Learning To Retrieve Prompts for In-Context Learning", "Active Example Selection for In-Context Learning"], "answer_arxiv_id": ["2112.08633", "2211.04486"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_dev_338"} +{"question": "What works used these loss functions to train a non-linear independent component estimation (NICE) model on high-dimensional tasks?", "answer": ["Sliced Score Matching: A Scalable Approach to Density and Score Estimation", "Efficient Learning of Generative Models via Finite-Difference Score Matching"], "answer_arxiv_id": ["1905.07088", "2007.03317"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_339"} +{"question": "Who designed extractive summarization approaches that adopt types of hierarchical architectures?", "answer": ["HIBERT: Document Level Pre-training of Hierarchical Bidirectional\n Transformers for Document Summarization", "Unsupervised Extractive Summarization by Pre-training Hierarchical\n Transformers", "HiStruct+: Improving Extractive Text Summarization with Hierarchical\n Structure Information", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding"], "answer_arxiv_id": ["1905.06566", "2010.08242", "2203.09629", "1810.04805"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_dev_340"} +{"question": "What are the papers about the Forgetting method that considers hard samples with the most forgetting events as important?", "answer": ["An Empirical Study of Example Forgetting during Deep Neural Network Learning"], "answer_arxiv_id": ["1812.05159"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_dev_341"} +{"question": "Which paper devised multiple tests to examine the faithfulness of CoT?", "answer": ["Measuring Faithfulness in Chain-of-Thought Reasoning"], "answer_arxiv_id": ["2307.13702"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_342"} +{"question": "Which are the initial research papers on generative data augmentation?", "answer": ["Data Augmentation Generative Adversarial Networks", "Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro"], "answer_arxiv_id": ["1711.04340", "1701.07717"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_dev_343"} +{"question": "Which researcher extended the classifier approach of CCMI by applying it directly to the estimation of CMI?", "answer": ["Neural Estimators for Conditional Mutual Information Using Nearest Neighbors Sampling"], "answer_arxiv_id": ["2006.07225"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_344"} +{"question": "What research introduced L2SP, a simple regularization method which minimizes the parameters between source and target models during fine-tuning?", "answer": ["Explicit Inductive Bias for Transfer Learning with Convolutional\n Networks"], "answer_arxiv_id": ["1802.01483"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_dev_345"} +{"question": "What research introduces OOD detection?", "answer": ["Generalized Out-of-Distribution Detection: A Survey"], "answer_arxiv_id": ["2110.11334"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_dev_346"} +{"question": "What research initially focused on content moderation in online social media platforms like Twitter and Reddit?", "answer": ["The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memes", "Ruddit: Norms of Offensiveness for English Reddit Comments", "SemEval-2019 Task 6: Identifying and Categorizing Offensive Language in\n Social Media (OffensEval)"], "answer_arxiv_id": ["2005.04790", "2106.05664", "1903.08983"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_dev_347"} +{"question": "Which works have been proposed for reducing the 3D annotation demand via weakly-supervised or semi-supervised learning in monocular 3D reconstruction?", "answer": ["Multiview-Consistent Semi-Supervised Learning for 3D Human Pose\n Estimation", "Robust Model-based Face Reconstruction through Weakly-Supervised Outlier\n Segmentation", "Semi-Supervised Adversarial Monocular Depth Estimation"], "answer_arxiv_id": ["1908.05293", "2106.09614", "1908.02126"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_dev_348"} +{"question": "What papers involve the use of an encoder-decoder network for height value regression?", "answer": ["Elevation Estimation-Driven Building 3D Reconstruction from Single-View\n Remote Sensing Imagery"], "answer_arxiv_id": ["2301.04581"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_dev_349"} +{"question": "In what work is SAM, a model that can segment any object of a given image based on visual prompts such as points and boxes, proposed?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_350"} +{"question": "What works apply reinforcement learning to rule learning by training agents to search for paths in the knowledge graph?", "answer": ["DeepPath: A Reinforcement Learning Method for Knowledge Graph Reasoning", "Variational Knowledge Graph Reasoning", "Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning", "Multi-Hop Knowledge Graph Reasoning with Reward Shaping", "M-Walk: Learning to Walk over Graphs using Monte Carlo Tree Search"], "answer_arxiv_id": ["1707.06690", "1803.06581", "1711.05851", "1808.10568", "1802.04394"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_351"} +{"question": "Could you give me examples of research on distributing matrix multiplication or more general, polynomial function computation?", "answer": ["Polynomial Codes: an Optimal Design for High-Dimensional Coded Matrix Multiplication", "Speeding Up Distributed Machine Learning Using Codes", "Universally Decodable Matrices for Distributed Matrix-Vector Multiplication", "Random Khatri-Rao-Product Codes for Numerically-Stable Distributed Matrix Multiplication", "Lagrange Coded Computing: Optimal Design for Resiliency, Security, and Privacy", "On the Optimal Recovery Threshold of Coded Matrix Multiplication", "Straggler Mitigation in Distributed Matrix Multiplication: Fundamental Limits and Optimal Coding", "Straggler-resistant distributed matrix computation via coding theory"], "answer_arxiv_id": ["1705.10464", "1512.02673", "1901.10674", "1907.05965", "1806.00939", "1801.10292v2", "1801.07487", "2002.03515"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_dev_352"} +{"question": "Which papers incorporated task-inference methods in their research?", "answer": ["Meta Reinforcement Learning As Task Inference", "VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning", "Learning Adaptive Exploration Strategies in Dynamic Environments Through Informed Policy Regularization", "Decoupling Exploration and Exploitation for Meta-Reinforcement Learning without Sacrifices", "Hypernetworks in Meta-Reinforcement Learning"], "answer_arxiv_id": ["1905.06424", "1910.08348", "2005.02934", "2008.02790", "2210.11348"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_353"} +{"question": "Which works are based on Inverse Dynamic Models (IDMs) in the field of imitation learning from observations?", "answer": ["Behavioral Cloning from Observation", "Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos", "Combining Self-Supervised Learning and Imitation for Vision-Based Rope Manipulation", "Zero-Shot Visual Imitation"], "answer_arxiv_id": ["1805.01954", "2206.11795", "1703.02018", "1804.08606"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_354"} +{"question": "What works have proposed node-to-graph Graph Contrastive Learning methods?", "answer": ["Deep Graph Infomax", "Contrastive Multi-View Representation Learning on Graphs"], "answer_arxiv_id": ["1809.10341", "2006.05582"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_dev_355"} +{"question": "What research showed how underlying generative factors can be recovered given certain conditions?", "answer": ["Independent mechanism analysis, a new concept?"], "answer_arxiv_id": ["2106.05200"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_dev_356"} +{"question": "What papers are about the extensions to NeRF for better image encoding?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images"], "answer_arxiv_id": ["2012.02190"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_dev_357"} +{"question": "Are there any works about low-rank models applied for matrix completion?", "answer": ["Distributed Matrix Completion and Robust Factorization"], "answer_arxiv_id": ["1107.0789"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_dev_358"} +{"question": "What works applied variational autoencoders to compressing various data modalities like video or point clouds?", "answer": ["DVC: An End-to-end Deep Video Compression Framework", "Density-preserving Deep Point Cloud Compression"], "answer_arxiv_id": ["1812.00101", "2204.12684"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_dev_359"} +{"question": "Which papers discuss the high cost of sample-based explanations that require repeated retraining?", "answer": ["Data Shapley: Equitable Valuation of Data for Machine Learning", "Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms", "Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning", "What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation"], "answer_arxiv_id": ["1904.02868v2", "1908.08619", "2110.14049", "2008.03703"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_360"} +{"question": "Are there any papers related to the application of Principal component analysis (PCA) to re-weighted samples?", "answer": ["Isotropic PCA and Affine-Invariant Clustering", "Fourier PCA and Robust Tensor Decomposition", "Structure from Local Optima: Learning Subspace Juntas via Higher Order PCA"], "answer_arxiv_id": ["0804.3575", "1306.5825", "1108.3329v3"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_dev_361"} +{"question": "What research papers mention the use of 3D attention-based models for instance segmentation?", "answer": ["Superpoint Transformer for 3D Scene Instance Segmentation"], "answer_arxiv_id": ["2211.15766"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_dev_362"} +{"question": "What research proposes the use of a meta-network to generate an image-conditioned prompt?", "answer": ["Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2203.05557"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_dev_363"} +{"question": "Which work extended the Sinkhorn method to address the problems in the field of UOT?", "answer": ["On Unbalanced Optimal Transport: An Analysis of Sinkhorn Algorithm"], "answer_arxiv_id": ["2002.03293"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_364"} +{"question": "Could you provide me some works that developed different variants of memory augmented networks?", "answer": ["Self-Attentive Associative Memory", "Relational recurrent neural networks", "Neural Turing Machines", "Neural Stored-program Memory"], "answer_arxiv_id": ["2002.03519", "1806.01822", "1410.5401", "1906.08862"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_dev_365"} +{"question": "What studies utilized deep learning methods in recent advances of anomaly detection (AD)?", "answer": ["A Unifying Review of Deep and Shallow Anomaly Detection"], "answer_arxiv_id": ["2009.11732"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_366"} +{"question": "Which papers provide studies about causative attacks on GNNs to reduce the accuracy of or intentionally change the outcome of node classification, link prediction, and graph classification?", "answer": ["Adversarial Attacks on Node Embeddings via Graph Poisoning", "Towards More Practical Adversarial Attacks on Graph Neural Networks", "Topology Attack and Defense for Graph Neural Networks: An Optimization Perspective", "Adversarial Attacks on Node Embeddings via Graph Poisoning", "Adversarial Attack on Graph Structured Data", "Graph Backdoor"], "answer_arxiv_id": ["1809.01093", "2006.05057", "1906.04214", "1809.01093", "1806.02371", "2006.11890"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_367"} +{"question": "Which studies propose the use of intermediate domains to mitigate domain shift?", "answer": ["CoSMix: Compositional Semantic Mix for Domain Adaptation in 3D LiDAR Segmentation", "DODA: Data-oriented Sim-to-Real Domain Adaptation for 3D Semantic Segmentation"], "answer_arxiv_id": ["2207.09778", "2204.01599"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_dev_368"} +{"question": "Are there any research contributions to the study of the noun-based Referring Expression Comprehension model?", "answer": ["Learning to Compose and Reason with Language Tree Structures for Visual Grounding", "Modeling Relationships in Referential Expressions with Compositional Modular Networks", "Learning to Assemble Neural Module Tree Networks for Visual Grounding", "Learning Two-Branch Neural Networks for Image-Text Matching Tasks", "Dynamic Graph Attention for Referring Expression Comprehension", "MAttNet: Modular Attention Network for Referring Expression Comprehension", "Grounding Referring Expressions in Images by Variational Context", "Parallel Attention: A Unified Framework for Visual Object Discovery through Dialogs and Queries", "Real-Time Referring Expression Comprehension by Single-Stage Grounding Network", "A Real-Time Cross-modality Correlation Filtering Method for Referring Expression Comprehension", "Zero-Shot Grounding of Objects from Natural Language Queries", "Improving One-stage Visual Grounding by Recursive Sub-query Construction", "A Fast and Accurate One-Stage Approach to Visual Grounding"], "answer_arxiv_id": ["1906.01784", "1611.09978", "1812.03299", "1704.03470", "1909.08164", "1801.08186", "1712.01892", "1711.06370", "1812.03426", "1909.07072", "1908.07129", "2008.01059", "1908.06354"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_dev_369"} +{"question": "What works are part of the 'BLIP-family' models that adopted both ITC and ITM as training objectives?", "answer": ["Vision-Language Pre-Training with Triple Contrastive Learning", "Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual\n Concepts", "GRIT-VLP: Grouped Mini-batch Sampling for Efficient Vision and Language\n Pre-training", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BUS:Efficient and Effective Vision-language Pre-training with Bottom-Up\n Patch Summarization", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Bootstrapping Vision-Language Learning with Decoupled Language\n Pre-training"], "answer_arxiv_id": ["2202.10401", "2111.08276", "2208.04060", "2201.12086", "2307.08504", "2301.12597", "2307.07063"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_dev_370"} +{"question": "What research papers applied attention mechanisms for single-task based deblurring?", "answer": ["Restormer: Efficient Transformer for High-Resolution Image Restoration"], "answer_arxiv_id": ["2111.09881"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_371"} +{"question": "Could you provide me some works about mitigating simplicity bias?", "answer": ["Evading the Simplicity Bias: Training a Diverse Set of Models Discovers Solutions with Superior OOD Generalization"], "answer_arxiv_id": ["2105.05612"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_dev_372"} +{"question": "Are there any works debating the correlation between high attention weights and high feature importance in attention-based methods?", "answer": ["Is Attention Interpretable?", "Attention is not Explanation", "Attention is not not Explanation"], "answer_arxiv_id": ["1906.03731", "1902.10186v3", "1908.04626v2"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_373"} +{"question": "Can you mention some papers that incorporate self-attention module for image restoration due to the limited ability of CNN to model long-range dependencies?", "answer": ["Learning Texture Transformer Network for Image Super-Resolution", "SwinIR: Image Restoration Using Swin Transformer", "Restormer: Efficient Transformer for High-Resolution Image Restoration", "Pre-Trained Image Processing Transformer"], "answer_arxiv_id": ["2006.04139", "2108.10257", "2111.09881", "2012.00364"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_374"} +{"question": "Could you list some studies in speech tokenization that focus on semantic tokens with high correlation with phonemes?", "answer": ["Textless Speech-to-Speech Translation on Real Data", "HuBERT: Self-Supervised Speech Representation Learning by Masked\n Prediction of Hidden Units", "W2v-BERT: Combining Contrastive Learning and Masked Language Modeling\n for Self-Supervised Speech Pre-Training"], "answer_arxiv_id": ["2112.08352", "2106.07447", "2108.06209"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_dev_375"} +{"question": "Are there any works that propose models to solve Task-oriented Object detection and segmentation tasks?", "answer": ["TOIST: Task Oriented Instance Segmentation Transformer with Noun-Pronoun Distillation"], "answer_arxiv_id": ["2210.10775"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_dev_376"} +{"question": "Which studies addressed pretraining legal language models for Italian, Romanian, and Spanish?", "answer": ["Spanish Legalese Language Model and Corpora"], "answer_arxiv_id": ["2110.12201"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_dev_377"} +{"question": "What work proposed a similar PAC-Bayes framework where the KL divergence is replaced by a general family of Integral Probability Metrics?", "answer": ["Integral Probability Metrics PAC-Bayes Bounds"], "answer_arxiv_id": ["2207.00614"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_378"} +{"question": "What works use traditional machine translation data like parallel data from the web for fine-tuning LLMs?", "answer": ["BigTranslate: Augmenting Large Language Models with Multilingual\n Translation Capability over 100 Languages", "Steering Large Language Models for Machine Translation with Finetuning\n and In-Context Learning", "Extrapolating Large Language Models to Non-English by Aligning Languages"], "answer_arxiv_id": ["2305.18098", "2310.13448", "2308.04948"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_dev_379"} +{"question": "Which work employs masked image modeling for in-context training in computer vision?", "answer": ["Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning"], "answer_arxiv_id": ["2212.02499"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_380"} +{"question": "What papers propose datasets that focus on testing the ability of visual question-answering (VQA) models to answer 'counterfactual' questions?", "answer": ["Counterfactual VQA: A Cause-Effect Look at Language Bias", "Don't Just Assume; Look and Answer: Overcoming Priors for Visual\n Question Answering"], "answer_arxiv_id": ["2006.04315v4", "1712.00377"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_381"} +{"question": "Can you provide the works about the large-scale web data containing billions of image-text pairs?", "answer": ["LAION-5B: An open large-scale dataset for training next generation image-text models", "Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts"], "answer_arxiv_id": ["2210.08402", "2102.08981"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_dev_382"} +{"question": "Which papers introduce generative models for LiDAR scene creation?", "answer": ["Deep Generative Modeling of LiDAR Data"], "answer_arxiv_id": ["1812.01180"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_383"} +{"question": "Could you name some concurrent tuning-free methods for personalized visual content generation which use an image encoder for accessibility?", "answer": ["FastComposer: Tuning-Free Multi-Subject Image Generation with Localized\n Attention", "Face0: Instantaneously Conditioning a Text-to-Image Model on a Face", "Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning"], "answer_arxiv_id": ["2305.10431", "2306.06638", "2307.11410"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_dev_384"} +{"question": "Which papers analyze the convergence of Minibatch RR and Local RR, a variant of Minibatch SGD and Local SGD in Federated Learning?", "answer": ["Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond"], "answer_arxiv_id": ["2110.10342"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_385"} +{"question": "What research papers used Bayesian approaches and evidential neural networks to quantify uncertainty in neural networks?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer\n Vision?", "Evidential Deep Learning to Quantify Classification Uncertainty", "Deep Evidential Regression"], "answer_arxiv_id": ["1703.04977", "1806.01768", "1910.02600"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_386"} +{"question": "Any papers that utilized the self-supervised training of transformer for vision tasks?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_387"} +{"question": "Which works focus on generating code directly from natural language descriptions?", "answer": ["CodeBERT: A Pre-Trained Model for Programming and Natural Languages", "Evaluating Large Language Models Trained on Code", "Language Models are Few-Shot Learners", "[2203.07814] Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2002.08155", "2107.03374", "2005.14165", "2203.07814"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_dev_388"} +{"question": "What studies improved GCL by introducing adaptive masking and dropping rate related to node centrality?", "answer": ["Graph Contrastive Learning with Adaptive Augmentation"], "answer_arxiv_id": ["2010.14945"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_dev_389"} +{"question": "Which works focus on 3D shape generation using variational autoencoders?", "answer": ["FoldingNet: Point Cloud Auto-encoder via Deep Grid Deformation", "Multiresolution Tree Networks for 3D Point Cloud Processing", "SetVAE: Learning Hierarchical Composition for Generative Modeling of Set-Structured Data"], "answer_arxiv_id": ["1712.07262", "1807.03520v2", "2103.15619"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_dev_390"} +{"question": "Which studies proposed methods to introduce hard orthogonality?", "answer": ["Efficient Riemannian Optimization on the Stiefel Manifold via the Cayley\n Transform", "Orthogonal Weight Normalization: Solution to Optimization over Multiple\n Dependent Stiefel Manifolds in Deep Neural Networks", "Orthogonalizing Convolutional Layers with the Cayley Transform", "Skew Orthogonal Convolutions", "Lipschitz regularity of deep neural networks: analysis and efficient\n estimation"], "answer_arxiv_id": ["2002.01113", "1709.06079", "2104.07167", "2105.11417", "1805.10965"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_dev_391"} +{"question": "Which work provided a lower bound theory for general activation functions?", "answer": ["Minimum Width for Universal Approximation"], "answer_arxiv_id": ["2006.08859"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_dev_392"} +{"question": "Which works contribute to the advancements in vector quantization for efficient coding of information?", "answer": ["Neural Discrete Representation Learning", "End-to-End Optimized Speech Coding with Deep Neural Networks", "Low Bit-Rate Speech Coding with VQ-VAE and a WaveNet Decoder", "Audio-Visual Speech Codecs: Rethinking Audio-Visual Speech Enhancement\n by Re-Synthesis", "SoundStream: An End-to-End Neural Audio Codec"], "answer_arxiv_id": ["1711.00937", "1710.09064", "1910.06464", "2203.17263", "2107.03312"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_dev_393"} +{"question": "Could you provide me with studies that used LLMs to generate queries from a document in information retrieval?", "answer": ["InPars: Data Augmentation for Information Retrieval using Large Language\n Models"], "answer_arxiv_id": ["2202.05144"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_dev_394"} +{"question": "Which work introduced an intermediate representation to enable the weakly-supervised training for hand pose estimation?", "answer": ["Hand Pose Estimation through Semi-Supervised and Weakly-Supervised\n Learning"], "answer_arxiv_id": ["1511.06728"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_dev_395"} +{"question": "Can you give examples of studies where new knowledge base was connected with the base model to implement a retrieve for needed new knowledge to a prompt or a question?", "answer": ["Fixing Model Bugs with Natural Language Patches", "Memory-Based Model Editing at Scale", "Large Language Models with Controllable Working Memory", "Memory-assisted prompt editing to improve GPT-3 after deployment"], "answer_arxiv_id": ["2211.03318", "2206.06520", "2211.05110", "2201.06009"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_396"} +{"question": "Which researches have studied sarcasm in NLP?", "answer": ["A Large Self-Annotated Corpus for Sarcasm"], "answer_arxiv_id": ["1704.05579"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_397"} +{"question": "What work suggests that direct pixel space parameterization is a key factor for the architecture transferability issue and proposes GLaD for enhancing generalization across any distillation method?", "answer": ["Generalizing Dataset Distillation via Deep Generative Prior"], "answer_arxiv_id": ["2305.01649"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_398"} +{"question": "What papers discuss research on post-hoc watermarking techniques involving synonym replacement or paraphrasing?", "answer": ["Watermarking Text Generated by Black-Box Language Models", "DeepTextMark: A Deep Learning-Driven Text Watermarking Approach for\n Identifying Large Language Model Generated Text"], "answer_arxiv_id": ["2305.08883v1", "2305.05773"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_399"} +{"question": "Could you provide me some works that studied spatial reasoning limitations in generative VLMs?", "answer": ["Benchmarking Spatial Relationships in Text-to-Image Generation", "DALL-Eval: Probing the Reasoning Skills and Social Biases of\n Text-to-Image Generation Models"], "answer_arxiv_id": ["2212.10015", "2202.04053"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_dev_400"} +{"question": "What studies applied TracIn for measuring training data importance?", "answer": ["Estimating Training Data Influence by Tracing Gradient Descent", "HYDRA: Hypergradient Data Relevance Analysis for Interpreting Deep Neural Networks"], "answer_arxiv_id": ["2002.08484", "2102.02515v5"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_401"} +{"question": "What research proposes a method for complex edits such as geometry deformation and texture swapping, filling, and painting in NeRF editing?", "answer": ["NeuMesh: Learning Disentangled Neural Mesh-based Implicit Field for\n Geometry and Texture Editing"], "answer_arxiv_id": ["2207.11911"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_dev_402"} +{"question": "What studies have focused on adapting foundation models, specifically through techniques such as prompt-tuning or fine-tuning with residual connections?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_dev_403"} +{"question": "Which work generated images containing specific text based on a large number of image-text pairs to improve the text generation capability of diffusion models?", "answer": ["Character-Aware Models Improve Visual Text Rendering"], "answer_arxiv_id": ["2212.10562"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_dev_404"} +{"question": "Which works have used ensemble predictions from clients’ models on an unlabeled dataset to guide the training of the server model?", "answer": ["Ensemble Distillation for Robust Model Fusion in Federated Learning", "FedBE: Making Bayesian Model Ensemble Applicable to Federated Learning", "FedMD: Heterogenous Federated Learning via Model Distillation"], "answer_arxiv_id": ["2006.07242v3", "2009.01974", "1910.03581"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_dev_405"} +{"question": "Can you provide me with research that focused on manipulating LLMs’ internal representations to guide them towards factuality during inference-time intervention?", "answer": ["Inference-Time Intervention: Eliciting Truthful Answers from a Language\n Model", "DoLa: Decoding by Contrasting Layers Improves Factuality in Large\n Language Models", "Contrastive Decoding: Open-ended Text Generation as Optimization", "Alleviating Hallucinations of Large Language Models through Induced\n Hallucinations"], "answer_arxiv_id": ["2306.03341", "2309.03883", "2210.15097", "2312.15710"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_dev_406"} +{"question": "Which studies employed meta-learning loops to reduce the fitting times during encoding in INR-based compression methods?", "answer": ["COIN++: Neural Compression Across Modalities", "Implicit Neural Representations for Image Compression", "Meta-Learning Sparse Compression Networks"], "answer_arxiv_id": ["2201.12904", "2112.04267v2", "2205.08957"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_dev_407"} +{"question": "What studies introduced various regularization techniques to maximize the utility of sparse input views for scene reconstruction?", "answer": ["Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis", "InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Rendering", "Learning Transferable Visual Models From Natural Language Supervision", "VGOS: Voxel Grid Optimization for View Synthesis from Sparse Inputs", "FreeNeRF: Improving Few-shot Neural Rendering with Free Frequency\n Regularization"], "answer_arxiv_id": ["2104.00677", "2112.15399", "2103.00020", "2304.13386", "2303.07418"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_408"} +{"question": "Could you provide me some works which decouples the continuous decision-making process into two steps?", "answer": ["Hierarchical Representations and Explicit Memory: Learning Effective Navigation Policies on 3D Scene Graphs using Graph Neural Networks"], "answer_arxiv_id": ["2108.01176"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_409"} +{"question": "Which research works propose different selection methods for scans, regions, points, or boxes to be labeled during training?", "answer": ["Bi3D: Bi-domain Active Learning for Cross-domain 3D Object Detection", "ReDAL: Region-based and Diversity-aware Active Learning for Point Cloud Semantic Segmentation", "LESS: Label-Efficient Semantic Segmentation for LiDAR Point Clouds", "Exploring Active 3D Object Detection from a Generalization Perspective"], "answer_arxiv_id": ["2303.05886", "2107.11769", "2210.08064", "2301.09249"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_dev_410"} +{"question": "What research papers are about physics-simulator-based methods in VR HMD settings?", "answer": ["QuestEnvSim: Environment-Aware Simulated Motion Tracking from Sparse\n Sensors", "QuestSim: Human Motion Tracking from Sparse Sensors with Simulated\n Avatars", "Isaac Gym: High Performance GPU-Based Physics Simulation For Robot\n Learning"], "answer_arxiv_id": ["2306.05666", "2209.09391", "2108.10470"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_411"} +{"question": "Which papers describe a knowledge graph as a type of a heterogeneous graph?", "answer": ["A Survey on Heterogeneous Graph Embedding: Methods, Techniques, Applications and Sources"], "answer_arxiv_id": ["2011.14867"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_dev_412"} +{"question": "Could you name few works that employed the residual approach to study the effect of syntactic and semantic properties on brain alignment?", "answer": ["Joint processing of linguistic properties in brains and language models"], "answer_arxiv_id": ["2212.08094"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_dev_413"} +{"question": "Which works established the SustainBench consisting of 15 public datasets covering sustainable development goals?", "answer": ["SustainBench: Benchmarks for Monitoring the Sustainable Development Goals with Machine Learning"], "answer_arxiv_id": ["2111.04724"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_dev_414"} +{"question": "Any works about the effectiveness of knowledge distillation in semi-supervised learning?", "answer": ["Weighted Distillation with Unlabeled Examples"], "answer_arxiv_id": ["2210.06711"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_dev_415"} +{"question": "Which papers extended IPS and SNIPS methods to implicit feedback data?", "answer": ["Unbiased Recommender Learning from Missing-Not-At-Random Implicit Feedback"], "answer_arxiv_id": ["1909.03601"], "source_meta": {"published_time": "20220510"}, "qid": "AutoScholarQuery_dev_416"} +{"question": "Which works are considered as the first to propose the task of Visual Question Answering (VQA)?", "answer": ["Exploring Models and Data for Image Question Answering", "VQA: Visual Question Answering"], "answer_arxiv_id": ["1505.02074", "1505.00468"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_417"} +{"question": "Could you provide me some studies about speaker identification in manga?", "answer": ["Manga109Dialog: A Large-scale Dialogue Dataset for Comics Speaker\n Detection"], "answer_arxiv_id": ["2306.17469"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_dev_418"} +{"question": "What studies propose hierarchical classifiers for CNN-based deep models?", "answer": ["B-CNN: Branch Convolutional Neural Network for Hierarchical\n Classification", "Visual Tree Convolutional Neural Network in Image Classification", "Network of Experts for Large-Scale Image Categorization"], "answer_arxiv_id": ["1709.09890", "1906.01536", "1604.06119"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_dev_419"} +{"question": "What works explored perturbations on different granularities in various aspects of NLP tasks?", "answer": ["Robust Multilingual Part-of-Speech Tagging via Adversarial Training", "FreeLB: Enhanced Adversarial Training for Natural Language Understanding", "InfoBERT: Improving Robustness of Language Models from An Information\n Theoretic Perspective", "Knowledge Graph Contrastive Learning Based on Relation-Symmetrical\n Structure"], "answer_arxiv_id": ["1711.04903", "1909.11764", "2010.02329", "2211.10738"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_dev_420"} +{"question": "Which papers have evaluated explanation benchmarks with correlation to system performance or human understanding of decisions?", "answer": ["Sanity Checks for Saliency Maps"], "answer_arxiv_id": ["1810.03292v3"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_dev_421"} +{"question": "What initiatives were mentioned as related to open science community initiatives in language modeling?", "answer": ["BigScience: A Case Study in the Social Construction of a Multilingual\n Large Language Model", "OLMo: Accelerating the Science of Language Models", "Dolma: an Open Corpus of Three Trillion Tokens for Language Model\n Pretraining Research"], "answer_arxiv_id": ["2212.04960", "2402.00838", "2402.00159"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_dev_422"} +{"question": "What papers utilize VAEs, normalizing flows, reinforcement learning, optimal transport and diffusion models for the task of predicting the 3D structure of molecules given a molecular graph?", "answer": ["Molecular Geometry Prediction using a Deep Generative Graph Neural Network", "A Generative Model for Molecular Distance Geometry", "Boltzmann Generators – Sampling Equilibrium States of Many-Body Systems with Deep Learning", "Symmetry-Aware Actor-Critic for 3D Molecular Design", "GeoMol: Torsional Geometric Generation of Molecular 3D Conformer Ensembles", "Learning Gradient Fields for Molecular Conformation Generation", "GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation", "Torsional Diffusion for Molecular Conformer Generation"], "answer_arxiv_id": ["1904.00314", "1909.11459", "1812.01729", "2011.12747", "2106.07802", "2105.03902", "2203.02923", "2206.01729"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_423"} +{"question": "Could you provide me some studies that used the theory of universal learning in their work?", "answer": ["A Theory of Universal Learning", "Fine-Grained Distribution-Dependent Learning Curves", "Multiclass Learnability Beyond the PAC Framework: Universal Rates and Partial Concept Classes"], "answer_arxiv_id": ["2011.04483", "2208.14615", "2210.02297"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_424"} +{"question": "What papers discuss about generating a reasoning process to enhance interpretability and extra supervision for answer generation?", "answer": ["Program Induction by Rationale Generation : Learning to Solve and\n Explain Algebraic Word Problems"], "answer_arxiv_id": ["1705.04146"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_425"} +{"question": "What research papers have performed theoretical studies on dataset shifts, specifically covariate shifts and label shifts?", "answer": ["On Causal and Anticausal Learning"], "answer_arxiv_id": ["1206.6471"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_426"} +{"question": "Which work proposed a meta-learning approach to align the linguistic spaces, enabling zero-shot and few-shot generalization?", "answer": ["Are Structural Concepts Universal in Transformer Language Models?\n Towards Interpretable Cross-Lingual Generalization"], "answer_arxiv_id": ["2310.12794"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_dev_427"} +{"question": "Could you list some works that generalise conformal prediction beyond the i.i.d. data setting?", "answer": ["Conformal Prediction Under Covariate Shift", "Distribution-free uncertainty quantification for classification under label shift", "Conformal Inference of Counterfactuals and Individual Treatment Effects", "Conformalized Survival Analysis"], "answer_arxiv_id": ["1904.06019", "2103.03323", "2006.06138", "2103.09763"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_dev_428"} +{"question": "Could you provide me some works that use prompt-based learning based on seq2seq models to solve ARA as a text-to-text generative task?", "answer": ["Prompt-based Learning for Text Readability Assessment"], "answer_arxiv_id": ["2302.13139"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_429"} +{"question": "Which research works address prompt engineering for LLM performance?", "answer": ["An Information-theoretic Approach to Prompt Engineering Without Ground Truth Labels", "Calibrate Before Use: Improving Few-Shot Performance of Language Models", "True Few-Shot Learning with Language Models"], "answer_arxiv_id": ["2203.11364", "2102.09690", "2105.11447"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_dev_430"} +{"question": "What references discuss the use of Differential Privacy in the context of machine learning to protect training data?", "answer": ["Deep Learning with Differential Privacy", "Differentially Private Empirical Risk Minimization", "Differentially Private Generative Adversarial Network"], "answer_arxiv_id": ["1607.00133", "0912.0071v5", "1802.06739"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_dev_431"} +{"question": "What papers explored downsizing frame resolution as a solution to GPU memory constraints?", "answer": ["Learning Salient Boundary Feature for Anchor-free Temporal Action\n Localization", "An Efficient Spatio-Temporal Pyramid Transformer for Action Detection"], "answer_arxiv_id": ["2103.13137", "2207.10448"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_432"} +{"question": "Can you tell me about the works that fine-tune language like models with privacy guarantees?", "answer": ["Large Language Models Can Be Strong Differentially Private Learners", "Differentially Private Fine-tuning of Language Models"], "answer_arxiv_id": ["2110.05679", "2110.06500"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_433"} +{"question": "What papers relate to training an additional router model to integrate multiple LLMs into one framework?", "answer": ["LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and\n Generative Fusion", "Routing to the Expert: Efficient Reward-guided Ensemble of Large\n Language Models"], "answer_arxiv_id": ["2306.02561", "2311.08692"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_dev_434"} +{"question": "What research used Householder orthogonal decomposition to achieve strict matrix orthogonality in neural networks?", "answer": ["What if Neural Networks had SVDs?"], "answer_arxiv_id": ["2009.13977"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_dev_435"} +{"question": "Could you provide me some works updating evaluation tasks?", "answer": ["Dynatask: A Framework for Creating Dynamic AI Benchmark Tasks"], "answer_arxiv_id": ["2204.01906"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_dev_436"} +{"question": "What studies provide a discussion on the importance of the cut distance and homomorphism counts in the graph learning context?", "answer": ["Lovász Meets Weisfeiler and Leman", "Optimal graphon estimation in cut distance", "word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings of Structured Data", "Graph Homomorphism Convolution"], "answer_arxiv_id": ["1802.08876", "1703.05101", "2003.12590", "2005.01214"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_437"} +{"question": "Could you provide me some studies about the utilization of the OT map in domain adaptation?", "answer": ["Optimal Transport for Domain Adaptation"], "answer_arxiv_id": ["1507.00504"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_438"} +{"question": "Could you provide me some references about neural audio codecs?", "answer": ["SoundStream: An End-to-End Neural Audio Codec", "High Fidelity Neural Audio Compression"], "answer_arxiv_id": ["2107.03312", "2210.13438"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_439"} +{"question": "What works are about discrete prompts in Large Language Models (LLMs)?", "answer": ["Language Models are Few-Shot Learners", "Making Pre-trained Language Models Better Few-shot Learners", "Efficient (Soft) Q-Learning for Text Generation with Limited Good Data", "What Makes Good In-Context Examples for GPT-3?", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts"], "answer_arxiv_id": ["2005.14165", "2012.15723", "2106.07704v4", "2101.06804", "2010.15980"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_440"} +{"question": "What works provide datasets specifically designed for collaborative perception?", "answer": ["V2X-Sim: Multi-Agent Collaborative Perception Dataset and Benchmark for\n Autonomous Driving", "Where2comm: Communication-Efficient Collaborative Perception via Spatial\n Confidence Maps", "DAIR-V2X: A Large-Scale Dataset for Vehicle-Infrastructure Cooperative\n 3D Object Detection", "V2X-Seq: A Large-Scale Sequential Dataset for Vehicle-Infrastructure\n Cooperative Perception and Forecasting", "V2V4Real: A Real-world Large-scale Dataset for Vehicle-to-Vehicle\n Cooperative Perception"], "answer_arxiv_id": ["2202.08449", "2209.12836", "2204.05575", "2305.05938", "2303.07601"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_dev_441"} +{"question": "What works utilized ResNet as a backbone network for visual object tracking?", "answer": ["Transformer Tracking", "Learning Spatio-Temporal Transformer for Visual Tracking"], "answer_arxiv_id": ["2103.15436", "2103.17154"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_dev_442"} +{"question": "Can you provide studies that detect visual objects and match ROI embeddings with textual embeddings?", "answer": ["UNIMO: Towards Unified-Modal Understanding and Generation via\n Cross-Modal Contrastive Learning"], "answer_arxiv_id": ["2012.15409"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_dev_443"} +{"question": "What works discuss extensions to hierarchical datasets for differential privacy marginals, such as geographic level and household composition?", "answer": ["The 2020 Census Disclosure Avoidance System TopDown Algorithm", "Differentially Private Hierarchical Count-of-Counts Histograms", "Private Synthetic Data with Hierarchical Structure"], "answer_arxiv_id": ["2204.08986", "1804.00370", "2206.05942"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_dev_444"} +{"question": "What studies propose augmenting techniques for improving the transferability of adversarial examples?", "answer": ["Identity Mappings in Deep Residual Networks", "Improving Transferability of Adversarial Examples with Input Diversity", "Evading Defenses to Transferable Adversarial Examples by\n Translation-Invariant Attacks", "Synthesizing Robust Adversarial Examples", "Nesterov Accelerated Gradient and Scale Invariance for Adversarial\n Attacks", "Admix: Enhancing the Transferability of Adversarial Attacks"], "answer_arxiv_id": ["1603.05027", "1803.06978", "1904.02884", "1707.07397", "1908.06281", "2102.00436"], "source_meta": {"published_time": "20220924"}, "qid": "AutoScholarQuery_dev_445"} +{"question": "Could you provide some studies about language-guided Diffusion Models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10741", "2204.06125"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_446"} +{"question": "What papers are about CLIP variants with focus on performance and efficiency improvements?", "answer": ["SLIP: Self-supervision meets Language-Image Pre-training", "MaskCLIP: Masked Self-Distillation Advances Contrastive Language-Image\n Pretraining", "Supervision Exists Everywhere: A Data Efficient Contrastive\n Language-Image Pre-training Paradigm", "Scaling Language-Image Pre-training via Masking", "Attentive Mask CLIP"], "answer_arxiv_id": ["2112.12750", "2208.12262", "2110.05208", "2212.00794", "2212.08653"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_447"} +{"question": "Can you name a few works that research personalized models in pFL using client-side local distillation?", "answer": ["Parameterized Knowledge Transfer for Personalized Federated Learning", "QuPeD: Quantized Personalization via Distillation with Applications to\n Federated Learning"], "answer_arxiv_id": ["2111.02862", "2107.13892"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_dev_448"} +{"question": "Which works adapt the preconditioning of on-policy, linear, least-squares forms of TD for nonlinear function approximation?", "answer": ["Zap Q-learning with Nonlinear Function Approximation", "TDprop: Does Jacobi Preconditioning Help Temporal Difference Learning?"], "answer_arxiv_id": ["1910.05405", "2007.02786"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_449"} +{"question": "Could you provide me some studies that popularized membership inference attacks (MIA) as a practical means to demonstrate leakage of private information in Machine Learning?", "answer": ["Membership Inference Attacks From First Principles", "Membership Inference Attacks Against Machine Learning Models", "Extracting Training Data from Large Language Models", "Auditing Data Provenance in Text-Generation Models"], "answer_arxiv_id": ["2112.03570", "1610.05820", "2012.07805", "1811.00513"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_450"} +{"question": "What works are about masking and scaled cosine error usage focused on feature reconstruction in generative self-supervised learning?", "answer": ["GraphMAE: Self-Supervised Masked Graph Autoencoders"], "answer_arxiv_id": ["2205.10803"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_dev_451"} +{"question": "Could you provide me some works that proposed approaches to increase the context length by making the attention mechanism more scalable?", "answer": ["Longformer: The Long-Document Transformer", "PRIMERA: Pyramid-based Masked Sentence Pre-training for Multi-document\n Summarization", "Poolingformer: Long Document Modeling with Pooling Attention", "Reformer: The Efficient Transformer", "HyperAttention: Long-context Attention in Near-Linear Time"], "answer_arxiv_id": ["2004.05150", "2110.08499", "2105.04371", "2001.04451", "2310.05869"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_dev_452"} +{"question": "Could you mention studies that adopt controlling character scripts for plot development?", "answer": ["Controlled Cue Generation for Play Scripts", "Controllable Multi-Character Psychology-Oriented Story Generation"], "answer_arxiv_id": ["2112.06953", "2010.05230"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_dev_453"} +{"question": "Which works demonstrate applying vision-language models to tasks such as visual question answering and object detection?", "answer": ["Unified Vision-Language Pre-Training for Image Captioning and VQA", "MERLOT: Multimodal Neural Script Knowledge Models", "TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models", "Pix2seq: A Language Modeling Framework for Object Detection"], "answer_arxiv_id": ["1909.11059", "2106.02636", "2109.10282", "2109.10852"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_dev_454"} +{"question": "What are some research papers that focus on unsupervised anomaly detection methods?", "answer": ["Adversarially Learned One-Class Classifier for Novelty Detection", "Memorizing Normality to Detect Anomaly: Memory-augmented Deep\n Autoencoder for Unsupervised Anomaly Detection", "Uninformed Students: Student-Teacher Anomaly Detection with\n Discriminative Latent Embeddings", "CutPaste: Self-Supervised Learning for Anomaly Detection and\n Localization", "Attribute Restoration Framework for Anomaly Detection", "CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via\n Conditional Normalizing Flows", "Anomaly Detection via Reverse Distillation from One-Class Embedding", "Towards Total Recall in Industrial Anomaly Detection", "Multiresolution Knowledge Distillation for Anomaly Detection"], "answer_arxiv_id": ["1802.09088", "1904.02639", "1911.02357", "2104.04015", "1911.10676", "2107.12571", "2201.10703", "2106.08265", "2011.11108"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_dev_455"} +{"question": "What studies have focused on generative self-supervised learning in graph representation learning?", "answer": ["GraphMAE: Self-Supervised Masked Graph Autoencoders", "GPT-GNN: Generative Pre-Training of Graph Neural Networks"], "answer_arxiv_id": ["2205.10803", "2006.15437"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_dev_456"} +{"question": "What works have proposed the use of Large Language Models (LLMs) as rewards through fine-tuning them on extensive user data in Reinforcement Learning from Human Feedback (RLHF)?", "answer": ["Training language models to follow instructions with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback"], "answer_arxiv_id": ["2203.02155", "2204.05862"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_457"} +{"question": "Which works have presented different views and understandings on the concept of memorization in LMs?", "answer": ["Preventing Verbatim Memorization in Language Models Gives a False Sense\n of Privacy", "Counterfactual Memorization in Neural Language Models", "Emergent and Predictable Memorization in Large Language Models"], "answer_arxiv_id": ["2210.17546", "2112.12938", "2304.11158"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_458"} +{"question": "What papers observed the failings of existing AD methods in detecting anomalies when distribution shifts occur?", "answer": ["AnoShift: A Distribution Shift Benchmark for Unsupervised Anomaly Detection"], "answer_arxiv_id": ["2206.15476"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_459"} +{"question": "What papers discuss the techniques that have been incorporated into previous Seq2Seq GEC models?", "answer": ["Synthetic Data Generation for Grammatical Error Correction with Tagged\n Corruption Models", "Data Weighted Training Strategies for Grammatical Error Correction", "Efficient Grammatical Error Correction Via Multi-Task Training and\n Optimized Training Schedule", "Improving Seq2Seq Grammatical Error Correction via Decoding Interventions"], "answer_arxiv_id": ["2105.13318", "2008.02976", "2311.11813", "2310.14534v1"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_dev_460"} +{"question": "Any works discuss post-hoc calibration methods dealing poorly with over-confident predictions in domain-shift scenarios?", "answer": ["Post-hoc Uncertainty Calibration for Domain Drift Scenarios"], "answer_arxiv_id": ["2012.10988"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_dev_461"} +{"question": "What paper considered generating adversarial perturbations for training with an auxiliary network?", "answer": ["Improving Robustness of Deep-Learning-Based Image Reconstruction"], "answer_arxiv_id": ["2002.11821"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_462"} +{"question": "Which research works apply slot attention to the domain of novel view synthesis?", "answer": ["Object Scene Representation Transformer"], "answer_arxiv_id": ["2206.06922"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_463"} +{"question": "Can you provide some references that introduced and developed the concept of Split Learning with regards to data protection?", "answer": ["Split learning for health: Distributed deep learning without sharing raw\n patient data", "Advancements of federated learning towards privacy preservation: from\n federated learning to split learning", "SplitFed: When Federated Learning Meets Split Learning"], "answer_arxiv_id": ["1812.00564", "2011.14818", "2004.12088"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_464"} +{"question": "Which paper reports that existing adaptations of foundation models for AD may generalize poorly to specific domains not covered in their massive training samples?", "answer": ["Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images"], "answer_arxiv_id": ["2205.11474"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_465"} +{"question": "Could you provide the papers that discussed the models capable of addressing real-scenario compositional reasoning?", "answer": ["Abstract Meaning Representation-Based Logic-Driven Data Augmentation for\n Logical Reasoning", "IDOL: Indicator-oriented Logic Pre-training for Logical Reasoning", "Exploring Self-supervised Logic-enhanced Training for Large Language\n Models", "MERIt: Meta-Path Guided Contrastive Learning for Logical Reasoning", "Discourse-Aware Graph Networks for Textual Logical Reasoning", "DAGN: Discourse-Aware Graph Network for Logical Reasoning"], "answer_arxiv_id": ["2305.12599", "2306.15273", "2305.13718", "2203.00357", "2207.01450", "2103.14349"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_dev_466"} +{"question": "Any works about GEM benchmark that assesses models on 40 language generation tasks?", "answer": ["The GEM Benchmark: Natural Language Generation, its Evaluation and\n Metrics", "Random walk models approximating symmetric space-fractional diffusion\n processes"], "answer_arxiv_id": ["2102.01672", "1210.6589"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_dev_467"} +{"question": "Can you provide some works about improving mathematical reasoning potentials by further training the generators with feedback from reward models?", "answer": ["Solving math word problems with process- and outcome-based feedback", "Let's Reinforce Step by Step"], "answer_arxiv_id": ["2211.14275", "2311.05821"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_468"} +{"question": "Which paper introduced the concept of implicit data augmentation in the context of image classification?", "answer": ["Implicit Semantic Data Augmentation for Deep Networks"], "answer_arxiv_id": ["1909.12220"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_dev_469"} +{"question": "What works explored video textures as a kind of texture in moving scenes?", "answer": ["Strumming to the Beat: Audio-Conditioned Contrastive Video Textures"], "answer_arxiv_id": ["2104.02687"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_dev_470"} +{"question": "Could you provide me with the work providing theoretical explanation about the credibility of generated pseudolabels in comparison to original labels?", "answer": ["Theoretical Analysis of Self-Training with Deep Networks on Unlabeled\n Data"], "answer_arxiv_id": ["2010.03622"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_dev_471"} +{"question": "Which papers talk about external tools called by a large language model to perform mathematical operations?", "answer": ["Solving Math Word Problems by Combining Language Models With Symbolic\n Solvers"], "answer_arxiv_id": ["2304.09102"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_dev_472"} +{"question": "What works have focused on the interactions between group fairness and differential privacy?", "answer": ["Differential Privacy Has Disparate Impact on Model Accuracy", "P"], "answer_arxiv_id": ["1905.12101", "0704.0320"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_dev_473"} +{"question": "What work is partially related to the proposed method in this paper and concerns the estimation of CT models with latent variables?", "answer": ["Learning Dynamical Systems from Partial Observations"], "answer_arxiv_id": ["1902.11136"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_dev_474"} +{"question": "Which work utilize OOD data for training-time regularization?", "answer": ["Deep Anomaly Detection with Outlier Exposure", "Unsupervised Out-of-Distribution Detection by Maximum Classifier Discrepancy", "VOS: Learning What You Don’t Know by Virtual Outlier Synthesis", "Semantically Coherent Out-of-Distribution Detection"], "answer_arxiv_id": ["1812.04606", "1908.04951", "2202.01197", "2108.11941"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_dev_475"} +{"question": "Can you name the work that applied self-supervised pre-trained features to detect instances in driving scenarios?", "answer": ["OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point\n Clouds"], "answer_arxiv_id": ["2210.04458"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_dev_476"} +{"question": "Which works explored generating photo-realistic sign language videos using GANs or diffusion models?", "answer": ["Signing at Scale: Learning to Co-Articulate Signs for Large-Scale\n Photo-Realistic Sign Language Production", "SignDiff: Learning Diffusion Models for American Sign Language\n Production", "Sign Language Production with Latent Motion Transformer"], "answer_arxiv_id": ["2203.15354", "2308.16082", "2312.12917"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_dev_477"} +{"question": "What works adapted random Fourier features to graphs and proposed a sampling-based variant of the global alignment graph kernel?", "answer": ["Scalable Global Alignment Graph Kernel Using Random Features: From Node Embedding to Graph Embedding"], "answer_arxiv_id": ["1911.11119"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_478"} +{"question": "Could you provide studies that propose two-stage training to improve the prediction performance for commonsense reasoning tasks?", "answer": ["Explain Yourself! Leveraging Language Models for Commonsense Reasoning"], "answer_arxiv_id": ["1906.02361"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_479"} +{"question": "Which work proposes CCMI and an estimator for the KL-divergence in the context of Conditional Mutual Information (CMI)?", "answer": ["CCMI : Classifier based Conditional Mutual Information Estimation"], "answer_arxiv_id": ["1906.01824"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_480"} +{"question": "Who applied a contrastive loss on a supervised setting in a multi-view rendering-based method?", "answer": ["Learning Local Shape Descriptors from Part Correspondences With\n Multi-view Convolutional Networks"], "answer_arxiv_id": ["1706.04496"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_481"} +{"question": "What works introduce task-specific gating networks in the sparse-MoE framework?", "answer": ["M$^3$ViT: Mixture-of-Experts Vision Transformer for Efficient Multi-task\n Learning with Model-Accelerator Co-design", "TaskExpert: Dynamically Assembling Multi-Task Representations with\n Memorial Mixture-of-Experts", "DSelect-k: Differentiable Selection in the Mixture of Experts with\n Applications to Multi-Task Learning"], "answer_arxiv_id": ["2210.14793", "2307.15324", "2106.03760"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_482"} +{"question": "What works utilize point methods for segmenting 3D LiDAR point clouds?", "answer": ["PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space", "PointSeg: Real-Time Semantic Segmentation Based on 3D LiDAR Point Cloud"], "answer_arxiv_id": ["1706.02413", "1807.06288"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_dev_483"} +{"question": "What are the studies that focus on improving the quality of annotations by including model adversaries into the annotation rounds?", "answer": ["Adversarial NLI: A New Benchmark for Natural Language Understanding"], "answer_arxiv_id": ["1910.14599"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_484"} +{"question": "Could you list the works using normalized flows for 3D shape generation?", "answer": ["PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "SoftFlow: Probabilistic Framework for Normalizing Flow on Manifolds", "Discrete Point Flow Networks for Efficient Point Cloud Generation"], "answer_arxiv_id": ["1906.12320", "2006.04604", "2007.10170"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_dev_485"} +{"question": "Can you suggest some researches that consider graph to be fully-connected when the underlying connectivity structure is unknown?", "answer": ["Attention Is All You Need", "VAIN: Attentional Multi-agent Predictive Modeling"], "answer_arxiv_id": ["1706.03762", "1706.06122"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_dev_486"} +{"question": "What works have proposed alternative diffusion processes closely related to Gaussian diffusion?", "answer": ["Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise", "Generative Modelling With Inverse Heat Dissipation", "Blurring Diffusion Models"], "answer_arxiv_id": ["2208.09392", "2206.13397v7", "2209.05557"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_dev_487"} +{"question": "Could you tell me if there are any studies that propose Neural Radiance Fields for novel-view synthesis?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_dev_488"} +{"question": "Could you provide some examples of datasets that involved scenarios with only a single API call?", "answer": ["ToolAlpaca: Generalized Tool Learning for Language Models with 3000\n Simulated Cases", "Gorilla: Large Language Model Connected with Massive APIs", "On the Tool Manipulation Capability of Open-source Large Language Models"], "answer_arxiv_id": ["2306.05301", "2305.15334", "2305.16504"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_dev_489"} +{"question": "Which works have incorporated equivariances into CNPs, but still suffer from the same scaling issues?", "answer": ["Convolutional Conditional Neural Processes", "Practical Conditional Neural Processes Via Tractable Dependent Predictions", "Group Equivariant Conditional Neural Processes", "Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data", "Equivariant Learning of Stochastic Fields: Gaussian Processes and Steerable Conditional Neural Processes"], "answer_arxiv_id": ["1910.13556", "2203.08775", "2102.08759", "2002.12880", "2011.12916"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_dev_490"} +{"question": "Which research is based on the task of RGBD-based 6D pose estimation in point cloud registration?", "answer": ["SuperPoint: Self-Supervised Interest Point Detection and Description", "Learning general and distinctive 3D local deep descriptors for point cloud registration", "Revisiting Fully Convolutional Geometric Features for Object 6D Pose\n Estimation", "OMNet: Learning Overlapping Mask for Partial-to-Partial Point Cloud\n Registration", "REGTR: End-to-end Point Cloud Correspondences with Transformers"], "answer_arxiv_id": ["1712.07629", "2105.10382v3", "2307.15514", "2103.00937", "2203.14517"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_dev_491"} +{"question": "Which works discussed adjusting quantization error in Post-training Quantization (PTQ) for Language Model (LLM)?", "answer": ["GPTQ: Accurate Post-Training Quantization for Generative Pre-trained\n Transformers", "QuIP: 2-Bit Quantization of Large Language Models With Guarantees"], "answer_arxiv_id": ["2210.17323", "2307.13304"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_492"} +{"question": "What research demonstrates the application of contrastive learning in text and image domains?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2002.05709", "2103.00020"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_dev_493"} +{"question": "Which works are about Seq2Seq models that have demonstrated high performance in Grammar Error Correction (GEC)?", "answer": ["BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "Approaching Neural Grammatical Error Correction as a Low-Resource\n Machine Translation Task", "A Neural Grammatical Error Correction System Built On Better\n Pre-training and Sequential Transfer Learning", "Improving Grammatical Error Correction via Pre-Training a Copy-Augmented\n Architecture with Unlabeled Data", "Stronger Baselines for Grammatical Error Correction Using Pretrained\n Encoder-Decoder Model"], "answer_arxiv_id": ["1910.13461", "1910.10683", "1804.05940", "1907.01256", "1903.00138", "2005.11849"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_dev_494"} +{"question": "Which studies showed that when the Polyak-Lojasiewicz condition is replaced by the weak PL condition, PG methods can also achieve linear convergence?", "answer": ["Sharp Analysis of Stochastic Optimization under Global Kurdyka-Łojasiewicz Inequality", "Stochastic Policy Gradient Methods: Improved Sample Complexity for Fisher-non-degenerate Policies"], "answer_arxiv_id": ["2210.01748", "2302.01734"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_495"} +{"question": "What studies focus on human-curated multilingual examples?", "answer": ["OpenAssistant Conversations -- Democratizing Large Language Model\n Alignment", "OpenAssistant Conversations -- Democratizing Large Language Model\n Alignment"], "answer_arxiv_id": ["2304.07327", "2304.07327"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_dev_496"} +{"question": "Which studies developed protein sequence design models using a BERT-style generative framework?", "answer": ["BERTology Meets Biology: Interpreting Attention in Protein Language Models"], "answer_arxiv_id": ["2006.15222"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_dev_497"} +{"question": "What research combines the property of Kronecker products with other techniques to produce accurate and updatable Kronecker sketching methods?", "answer": ["Optimal Sketching for Kronecker Product Regression and Low Rank Approximation", "Subquadratic Kronecker Regression with Applications to Tensor Decomposition"], "answer_arxiv_id": ["1909.13384v1", "2209.04876"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_dev_498"} +{"question": "Are there studies that allow for non-stationary environments but only explore regret to the best arm in hindsight?", "answer": ["Contextual Bandits with Similarity Information"], "answer_arxiv_id": ["0907.3986v5"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_dev_499"} +{"question": "Could you provide some studies where the idea of gating has been used in designing GNNs?", "answer": ["Residual Gated Graph ConvNets", "Gated Graph Sequence Neural Networks", "GaAN: Gated Attention Networks for Learning on Large and Spatiotemporal Graphs"], "answer_arxiv_id": ["1711.07553", "1511.05493", "1803.07294"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_dev_500"} +{"question": "Which works focus on improving efficiency of LLM inference using parallelism methods such as pipeline parallelism and tensor parallelism?", "answer": ["Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism", "GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism", "Colossal-AI: A Unified Deep Learning System For Large-Scale Parallel Training"], "answer_arxiv_id": ["1909.08053", "1811.06965", "2110.14883"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_501"} +{"question": "What papers presented the first framework of explaining GNN predictions?", "answer": ["GNNExplainer: Generating Explanations for Graph Neural Networks"], "answer_arxiv_id": ["1903.03894"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_dev_502"} +{"question": "Which research introduced CodeBLEU, which adds terms to measure Abstract Syntax Tree and data-flow similarity?", "answer": ["CodeBLEU: a Method for Automatic Evaluation of Code Synthesis"], "answer_arxiv_id": ["2009.10297"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_dev_503"} +{"question": "What paper has worked on improving the explainability of the information retrieval approach by inferring an adjacency matrix?", "answer": ["TransferNet: An Effective and Transparent Framework for Multi-hop Question Answering over Relation Graph"], "answer_arxiv_id": ["2104.07302"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_dev_504"} +{"question": "What studies focused on prompting methods to elicit the mathematical reasoning abilities of LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Cumulative Reasoning with Large Language Models", "Complexity-Based Prompting for Multi-Step Reasoning"], "answer_arxiv_id": ["2201.11903", "2308.04371", "2210.00720"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_505"} +{"question": "What research papers used a pooling layer and MLP classifier to predict the response length?", "answer": ["Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative Refinement", "FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow", "Iterative Refinement in the Continuous Space for Non-Autoregressive Neural Machine Translation", "Latent-Variable Non-Autoregressive Neural Machine Translation with Deterministic Inference Using a Delta Posterior"], "answer_arxiv_id": ["1802.06901", "1909.02480", "2009.07177", "1908.07181"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_506"} +{"question": "Could you give me examples of works that designed model aggregation schemes in the context of Federated Learning?", "answer": ["Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classification", "Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization", "Federated learning with matched averaging", "FedBN: Federated Learning on Non-IID Features via Local Batch Normalization"], "answer_arxiv_id": ["1909.06335", "2007.07481", "2002.06440", "2102.07623"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_dev_507"} +{"question": "What works used the ideas in neural ODEs and extended them to normalizing flows to efficiently model arbitrary probability distributions?", "answer": ["Normalizing Flows for Probabilistic Modeling and Inference", "FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models"], "answer_arxiv_id": ["1912.02762", "1810.01367"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_dev_508"} +{"question": "Which research found Gaussian noise addition beneficial for corruption robustness in classification?", "answer": ["A simple way to make neural networks robust against diverse image corruptions"], "answer_arxiv_id": ["2001.06057"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_509"} +{"question": "Which research works have contributed to the domain of TEE-based Private Learning for privacy-preserving machine learning?", "answer": ["secureTF: A Secure TensorFlow Framework", "MLCapsule: Guarded Offline Deployment of Machine Learning as a Service"], "answer_arxiv_id": ["2101.08204", "1808.00590v2"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_510"} +{"question": "What works are related to the applications of Latent Diffusion Models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_511"} +{"question": "Which work is similar to the proposed work in that they both index and shuffle the slot description in natural language?", "answer": ["Description-Driven Task-Oriented Dialog Modeling"], "answer_arxiv_id": ["2201.08904"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_512"} +{"question": "What work required actual calls to a real API to solve its problems, contrasting with other works that simulated API calls?", "answer": ["ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs", "On the Tool Manipulation Capability of Open-source Large Language Models"], "answer_arxiv_id": ["2307.16789", "2305.16504"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_dev_513"} +{"question": "Could you provide me some works which discuss about Non-Gaussian Component Analysis (NGCA)?", "answer": ["Polynomial Time and Sample Complexity for Non-Gaussian Component Analysis: Spectral Methods"], "answer_arxiv_id": ["1704.01041"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_dev_514"} +{"question": "What papers have focused on the subset of medical visual question answering that deals with image-based EHR QA?", "answer": ["Interpretable Medical Image Visual Question Answering via Multi-Modal Relationship Graph Learning"], "answer_arxiv_id": ["2302.09636"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_dev_515"} +{"question": "Could you provide me some studies that applies text prompting approach in multi-modal scenarios?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2109.01134", "2203.05557"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_dev_516"} +{"question": "What studies integrate detection results as supplementary information for Seq2Seq correction models?", "answer": ["Encoder-Decoder Models Can Benefit from Pre-trained Masked Language\n Models in Grammatical Error Correction"], "answer_arxiv_id": ["2005.00987"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_dev_517"} +{"question": "Is there any work that used zero-shot group equivariance in partially observable Markov decision processes?", "answer": ["Equivariant Networks for Zero-Shot Coordination"], "answer_arxiv_id": ["2210.12124"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_dev_518"} +{"question": "What research papers discuss the use of diffusion models for high-fidelity image synthesis?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Spot the fake lungs: Generating Synthetic Medical Images using Neural Diffusion Models", "Diffusion Probabilistic Models beat GANs on Medical Images", "Brain Imaging Generation with Latent Diffusion Models", "Diffusion Models for Medical Image Analysis: A Comprehensive Survey"], "answer_arxiv_id": ["2105.05233", "2211.00902v1", "2212.07501v1", "2209.07162", "2211.07804"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_dev_519"} +{"question": "Can you cite some early studies that investigated adversarial training in the computer vision domain?", "answer": ["Adversarial Machine Learning at Scale", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1611.01236", "1706.06083"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_dev_520"} +{"question": "Which papers introduced learning-augmented algorithms for weighted paging?", "answer": ["Learning-Augmented Weighted Paging"], "answer_arxiv_id": ["2011.09076"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_dev_521"} +{"question": "Which papers discuss the application of diffusion models in the field of image reconstruction from fMRI?", "answer": ["Reconstructing the Mind's Eye: fMRI-to-Image with Contrastive Learning\n and Diffusion Priors"], "answer_arxiv_id": ["2305.18274"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_522"} +{"question": "Which research papers used the RealNews dataset for pretraining?", "answer": ["Defending Against Neural Fake News", "Megatron-LM: Training Multi-Billion Parameter Language Models Using\n Model Parallelism", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1905.12616", "1909.08053", "1907.11692"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_dev_523"} +{"question": "What references propose LLM-Blender, a method to rank and fuse generations from different models?", "answer": ["LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and\n Generative Fusion"], "answer_arxiv_id": ["2306.02561"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_dev_524"} +{"question": "What are some works that have designed methods to learn the geocentric pose of buildings in off-nadir images for monocular height estimation?", "answer": ["Learning Geocentric Object Pose in Oblique Monocular Images"], "answer_arxiv_id": ["2007.00729"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_dev_525"} +{"question": "Could you name a few works that provide documentation guidelines for NLP and ML datasets, models, and systems?", "answer": ["Model Cards for Model Reporting", "Datasheets for Datasets", "Machine Learning Data Practices through a Data Curation Lens: An Evaluation Framework"], "answer_arxiv_id": ["1810.03993", "1803.09010", "2405.02703v1"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_dev_526"} +{"question": "What works propose weighting strategies in the fusion of multiple views considering the view quality?", "answer": ["Reconsidering Representation Alignment for Multi-view Clustering"], "answer_arxiv_id": ["2103.07738"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_dev_527"} +{"question": "Which works utilize spatio-temporal LSTM for action recognition?", "answer": ["Skeleton-Based Action Recognition Using Spatio-Temporal LSTM Network\n with Trust Gates"], "answer_arxiv_id": ["1706.08276"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_528"} +{"question": "Give me the examples of papers where surrogate gradient estimation of the firing function in SNNs has been studied.", "answer": ["Learning to Prune Deep Neural Networks via Layer-wise Optimal Brain Surgeon"], "answer_arxiv_id": ["1705.07565"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_dev_529"} +{"question": "What works focused on calibration in generative question answering?", "answer": ["How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answering"], "answer_arxiv_id": ["2012.00955"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_dev_530"} +{"question": "Are there papers that tried to use internal solver heuristics to control the learned dynamics of Neural Differential Equations?", "answer": ["Opening the Blackbox: Accelerating Neural Differential Equations by Regularizing Internal Solver Heuristics"], "answer_arxiv_id": ["2105.03918"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_dev_531"} +{"question": "Could you mention some papers that have discussed knowledge distillation in meta-learning?", "answer": ["Distilling the Knowledge in a Neural Network", "Few-Shot Learning with a Strong Teacher"], "answer_arxiv_id": ["1503.02531", "2107.00197"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_dev_532"} +{"question": "Did any research propose canonicalization-based methods to construct equivariant networks out of non-equivariant backbones?", "answer": ["Equivariance with Learned Canonicalization Functions"], "answer_arxiv_id": ["2211.06489"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_dev_533"} +{"question": "Which works have been particularly focused on introducing the concept of quantile temporal-difference learning?", "answer": ["Distributional Reinforcement Learning with Quantile Regression", "A Cramér Distance perspective on Quantile Regression based Distributional Reinforcement Learning", "An Analysis of Quantile Temporal-Difference Learning", "Implicit Quantile Networks for Distributional Reinforcement Learning", "Fully Parameterized Quantile Function for Distributional Reinforcement Learning"], "answer_arxiv_id": ["1710.10044", "2110.00535v2", "2301.04462", "1806.06923", "1911.02140"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_dev_534"} +{"question": "What works have studied the generalization from training data to test data both theoretically and practically?", "answer": ["Very Deep Convolutional Networks for Large-Scale Image Recognition"], "answer_arxiv_id": ["1409.1556"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_dev_535"} +{"question": "What papers require an L2-norm bound on the error of the linear approximation of Qt?", "answer": ["Linear Convergence for Natural Policy Gradient with Log-linear Policy Parametrization", "Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies"], "answer_arxiv_id": ["2209.15382", "2210.01400v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_536"} +{"question": "What works examined bandits and global optimization with neural function approximation?", "answer": ["Neural Contextual Bandits with UCB-based Exploration", "Neural Thompson Sampling", "Sample-Then-Optimize Batch Neural Thompson Sampling"], "answer_arxiv_id": ["1911.04462", "2010.00827", "2210.06850"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_dev_537"} +{"question": "What papers discuss the use of data augmentation or mixup to prevent robust overfittings?", "answer": ["Overfitting in adversarially robust deep learning", "ReRoGCRL: Representation-based Robustness in Goal-Conditioned\n Reinforcement Learning"], "answer_arxiv_id": ["2002.11569", "2312.07392"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_538"} +{"question": "Which studies focus on efficient FL by deploying model ensemble and sub-parameter sharing?", "answer": ["Think Locally, Act Globally: Federated Learning with Local and Global Representations"], "answer_arxiv_id": ["2001.01523"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_dev_539"} +{"question": "Could you provide me with studies that have demonstrated the importance of relationships between entities in deep learning?", "answer": ["Inductive Relation Prediction by Subgraph Reasoning", "Single-Stage Visual Relationship Learning using Conditional Queries", "A simple neural network module for relational reasoning", "Relation Networks for Object Detection", "Relational Knowledge Distillation", "Learning to Compare: Relation Network for Few-Shot Learning"], "answer_arxiv_id": ["1911.06962", "2306.05689", "1706.01427", "1711.11575", "1904.05068", "1711.06025"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_dev_540"} +{"question": "Are there any studies about ControlNet for image editing by providing reference images?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_dev_541"} +{"question": "Could you provide me some papers that added inductive bias into the neural network policy or learning algorithm?", "answer": ["Value Iteration Networks", "Neuro-algorithmic Policies enable Fast Combinatorial Generalization"], "answer_arxiv_id": ["1602.02867", "2102.07456"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_dev_542"} +{"question": "What work conducted further pre-training and instruction tuning on a speech dataset of semantic tokens?", "answer": ["SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal\n Conversational Abilities"], "answer_arxiv_id": ["2305.11000"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_543"} +{"question": "What paper proposed the utilization of the in-context learning method to revise the output of LLMs with demonstrations extracted from the corpus based on similarity?", "answer": ["Can We Edit Factual Knowledge by In-Context Learning?"], "answer_arxiv_id": ["2305.12740"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_544"} +{"question": "Could you provide research on enhancing in-context learning's capability in vision?", "answer": ["Exploring Effective Factors for Improving Visual In-Context Learning", "Towards In-context Scene Understanding"], "answer_arxiv_id": ["2304.04748", "2306.01667"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_545"} +{"question": "Which works have extended prototypical networks to few-shot anomaly detection?", "answer": ["One-Way Prototypical Networks", "Learning to Adapt to Domain Shifts with Few-shot Samples in Anomalous Sound Detection"], "answer_arxiv_id": ["1906.00820", "2204.01905"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_546"} +{"question": "What study observed that neural-network based deblurring is sensitive to adversarial perturbations despite being trained with Jittering?", "answer": ["On Adversarial Robustness of Deep Image Deblurring"], "answer_arxiv_id": ["2210.02502"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_547"} +{"question": "Could you provide me the papers which applied representations learned by large-scale models for semantic correspondence?", "answer": ["GAN-Supervised Dense Visual Alignment", "Emerging Properties in Self-Supervised Vision Transformers", "Deep ViT Features as Dense Visual Descriptors"], "answer_arxiv_id": ["2112.05143", "2104.14294", "2112.05814"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_548"} +{"question": "What research have made progress in predicting low-energy conformations given molecular graphs?", "answer": ["Learning Gradient Fields for Molecular Conformation Generation", "GeoMol: Torsional Geometric Generation of Molecular 3D Conformer Ensembles", "GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation", "Torsional Diffusion for Molecular Conformer Generation"], "answer_arxiv_id": ["2105.03902", "2106.07802", "2203.02923", "2206.01729"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_dev_549"} +{"question": "What papers explored vision-based UI Agents for web or mobile?", "answer": ["From Pixels to UI Actions: Learning to Follow Instructions via Graphical\n User Interfaces", "You Only Look at Screens: Multimodal Chain-of-Action Agents"], "answer_arxiv_id": ["2306.00245", "2309.11436"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_dev_550"} +{"question": "What papers studied methods that employed the use of motion in audio-visual learning?", "answer": ["Lip Reading Sentences in the Wild", "Looking to Listen at the Cocktail Party: A Speaker-Independent\n Audio-Visual Model for Speech Separation", "Audio-Visual Scene Analysis with Self-Supervised Multisensory Features", "The Sound of Motions", "Hear The Flow: Optical Flow-Based Self-Supervised Visual Sound Source\n Localization", "FlowGrad: Using Motion for Visual Sound Source Localization"], "answer_arxiv_id": ["1611.05358", "1804.03619", "1804.03641", "1904.05979", "2211.03019", "2211.08367"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_dev_551"} +{"question": "Which works used image generation models to create synthetic images for classification tasks?", "answer": ["Is synthetic data from generative models ready for image recognition?", "Synthetic Data from Diffusion Models Improves ImageNet Classification", "Leaving Reality to Imagination: Robust Classification via Generated Datasets"], "answer_arxiv_id": ["2210.07574", "2304.08466", "2302.02503"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_552"} +{"question": "Which papers propose the use of generative adversarial networks (GANs) and counterfactuals to augment training sets as dataset-level mitigation strategies against bias amplification?", "answer": ["Fair Attribute Classification through Latent Space De-biasing", "Contrastive Examples for Addressing the Tyranny of the Majority", "Learning the Difference that Makes a Difference with Counterfactually-Augmented Data", "Robustness to Spurious Correlations in Text Classification via Automatically Generated Counterfactuals"], "answer_arxiv_id": ["2012.01469", "2004.06524", "1909.12434", "2012.10040"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_dev_553"} +{"question": "Could you point to the literature that discusses the KL variation in relation to Proximal Policy Optimization?", "answer": ["Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1707.06347v2"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_554"} +{"question": "What studies have presented connections between RNNs and early versions of GNNs?", "answer": ["Graph Neural Networks: A Review of Methods and Applications"], "answer_arxiv_id": ["1812.08434"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_dev_555"} +{"question": "What paper proposed the integration of a generative adversarial network (GAN) framework for solving the primal formulation of unbalanced Monge OT?", "answer": ["Scalable Unbalanced Optimal Transport using Generative Adversarial Networks"], "answer_arxiv_id": ["1810.11447"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_556"} +{"question": "Which paper presented a modification to the decoder that allows SimulST using the wait-k policy and a fixed pre-decision module?", "answer": ["Streaming Simultaneous Speech Translation with Augmented Memory Transformer"], "answer_arxiv_id": ["2011.00033"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_dev_557"} +{"question": "Are there any papers about controllable Diffusion Models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_558"} +{"question": "Which paper discusses the usage of Logical Neural Networks in the Neuro-Symbolic RL framework?", "answer": ["Neuro-Symbolic Reinforcement Learning with First-Order Logic", "Logical Neural Networks"], "answer_arxiv_id": ["2110.10963", "2006.13155"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_dev_559"} +{"question": "Are there any papers that explain the reasoning behind the phenomenon of arithmetic operations, such as linear analogies, revealing semantic meaning?", "answer": ["Towards Understanding Linear Word Analogies"], "answer_arxiv_id": ["1810.04882"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_dev_560"} +{"question": "Which works indicated models trained with their methods yield less satisfactory results in comparison to the researcher's approach?", "answer": ["Sliced Score Matching: A Scalable Approach to Density and Score Estimation", "Efficient Learning of Generative Models via Finite-Difference Score Matching"], "answer_arxiv_id": ["1905.07088", "2007.03317"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_561"} +{"question": "Which works achieve cross-modal interaction by matching the visual tokens of fixed patches and textual tokens?", "answer": ["FILIP: Fine-grained Interactive Language-Image Pre-Training"], "answer_arxiv_id": ["2111.07783"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_dev_562"} +{"question": "What is the paper that launched SemEval-2020 Task 1 on Unsupervised Lexical Semantic Change Detection?", "answer": ["SemEval-2020 Task 1: Unsupervised Lexical Semantic Change Detection"], "answer_arxiv_id": ["2007.11464"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_dev_563"} +{"question": "Which papers discuss the evaluation of synthetic images in the medical domain?", "answer": ["How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Models", "Breast Mass Classification from Mammograms using Deep Convolutional Neural Networks", "GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium", "Assessing Generative Models via Precision and Recall"], "answer_arxiv_id": ["2102.08921", "1612.00542", "1706.08500", "1806.00035"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_dev_564"} +{"question": "Which work initiated the research agenda of replicable algorithm design?", "answer": ["Reproducibility in Learning"], "answer_arxiv_id": ["2201.08430"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_565"} +{"question": "What are some of the selected works that involve research on implicit modeling related to LazyGNN?", "answer": ["Neural Ordinary Differential Equations", "Implicit Deep Learning", "Deep Equilibrium Models", "Implicit Graph Neural Networks"], "answer_arxiv_id": ["1806.07366", "1908.06315", "1909.01377", "2009.06211"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_566"} +{"question": "Which studies focused on improving the transformer-based 2D-to-3D pose lifting method with the per joint temporal characteristics and frequency domain feature?", "answer": ["MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose\n Estimation in Video", "PoseFormerV2: Exploring Frequency Domain for Efficient and Robust 3D\n Human Pose Estimation"], "answer_arxiv_id": ["2203.00859", "2303.17472"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_567"} +{"question": "What work highlighted that using synthetic samples for augmented data can result in performance degradation?", "answer": ["How good is my GAN?"], "answer_arxiv_id": ["1807.09499"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_dev_568"} +{"question": "Which studies discuss diffusion-based text-to-image models in synthetic face generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2204.06125"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_dev_569"} +{"question": "Which works proposed IL+RL methods that are based on including prior data in the replay buffer for a value-based approach?", "answer": ["Overcoming Exploration in Reinforcement Learning with Demonstrations", "Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewards"], "answer_arxiv_id": ["1709.10089", "1707.08817"], "source_meta": {"published_time": "20220405"}, "qid": "AutoScholarQuery_dev_570"} +{"question": "Any papers around which indicate the application of Visual-Language Modeling in various scenarios?", "answer": ["Task Residual for Tuning Vision-Language Models", "GraphAdapter: Tuning Vision-Language Models With Dual Knowledge Graph", "Diffusion Model as Representation Learner", "Mutual-modality Adversarial Attack with Semantic Perturbation", "DeepCache: Accelerating Diffusion Models for Free"], "answer_arxiv_id": ["2211.10277", "2309.13625", "2308.10916", "2312.12768", "2312.00858"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_dev_571"} +{"question": "What works have uncovered inherent challenges in RLHF?", "answer": ["Open Problems and Fundamental Limitations of Reinforcement Learning from\n Human Feedback", "The History and Risks of Reinforcement Learning and Human Feedback"], "answer_arxiv_id": ["2307.15217", "2310.13595"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_dev_572"} +{"question": "Which papers propose to apply modern variance reduction techniques to efficiently solve the regression problem?", "answer": ["SARAH: A Novel Method for Machine Learning Problems Using Stochastic Recursive Gradient", "Momentum-Based Variance Reduction in Non-Convex SGD", "PAGE: A Simple and Optimal Probabilistic Gradient Estimator for Nonconvex Optimization", "An Improved Analysis of (Variance-Reduced) Policy Gradient and Natural Policy Gradient Methods"], "answer_arxiv_id": ["1703.00102", "1905.10018", "2008.10898", "2211.07937"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_573"} +{"question": "Which studies have developed supervised disentanglement metrics for latent spaces?", "answer": ["Disentangling by Factorising"], "answer_arxiv_id": ["1802.05983"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_dev_574"} +{"question": "Which works suggest that careful engineering of the provided prompts can influence LLMs behavior?", "answer": ["Prompt Programming for Large Language Models: Beyond the Few-Shot Paradigm", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts"], "answer_arxiv_id": ["2102.07350", "2010.15980"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_575"} +{"question": "Can you provide me the study of safe RL with linear function approximation?", "answer": ["Safe Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["2106.06239"], "source_meta": {"published_time": "20220628"}, "qid": "AutoScholarQuery_dev_576"} +{"question": "Which works address moment matching approaches to domain adaptation?", "answer": ["DACS: Domain Adaptation via Cross-domain Mixed Sampling", "Optimal Transport for Domain Adaptation"], "answer_arxiv_id": ["2007.08702", "1507.00504"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_577"} +{"question": "Which studies are about Large Language Models (LLMs)?", "answer": ["Training Compute-Optimal Large Language Models", "Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2203.15556", "2005.14165", "2204.02311"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_578"} +{"question": "What papers discussed the technique of randomized smoothing for obtaining robust classifiers?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing", "(Certified!!) Adversarial Robustness for Free!"], "answer_arxiv_id": ["1902.02918", "2206.10550"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_579"} +{"question": "Which paper established the theoretical framework for supervised adversarial training (sup-AT)?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_dev_580"} +{"question": "Can you list some works that used Image-text pre-training for Vision and Language tasks?", "answer": ["LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "UNITER: UNiversal Image-TExt Representation Learning", "VinVL: Revisiting Visual Representations in Vision-Language Models", "Unifying Vision-and-Language Tasks via Text Generation", "Scaling Up Vision-Language Pre-training for Image Captioning"], "answer_arxiv_id": ["1908.07490", "1909.11740", "2101.00529", "2102.02779", "2111.12233"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_dev_581"} +{"question": "What works predict 3D keypoints on the object from which the pose can be extracted?", "answer": ["PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimation"], "answer_arxiv_id": ["1911.04231"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_582"} +{"question": "Which papers focus on refining the latent dynamics model learning by proposing a joint learning scheme?", "answer": ["Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1912.01603", "2010.02193"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_dev_583"} +{"question": "Which research papers extend the distillation process to train NeRF for the 2D-to-3D task?", "answer": ["Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion\n Prior", "Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors"], "answer_arxiv_id": ["2303.14184", "2306.17843"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_dev_584"} +{"question": "Could you provide me some studies about medical anomaly detection?", "answer": ["Encoding Structure-Texture Relation with P-Net for Anomaly Detection in\n Retinal Images", "Viral Pneumonia Screening on Chest X-ray Images Using Confidence-Aware Anomaly Detection", "Proxy-bridged Image Reconstruction Network for Anomaly Detection in\n Medical Images", "BMAD: Benchmarks for Medical Anomaly Detection", "Dual-distribution discrepancy with self-supervised refinement for\n anomaly detection in medical images", "Multi-scale Cross-restoration Framework for Electrocardiogram Anomaly\n Detection"], "answer_arxiv_id": ["2008.03632", "2003.12338v4", "2110.01761", "2306.11876", "2210.04227", "2308.01639"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_dev_585"} +{"question": "Could you provide me some works about the successful use of Vision-language models in various downstream tasks?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "FLAVA: A Foundational Language And Vision Alignment Model", "Contrastive Learning of Medical Visual Representations from Paired\n Images and Text", "Florence: A New Foundation Model for Computer Vision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2103.00020", "2112.04482", "2010.00747", "2111.11432", "2102.05918", "2204.14198"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_dev_586"} +{"question": "What studies focused on bounding the GE specifically for deep iterative recovery algorithms?", "answer": ["Compressive Sensing and Neural Networks from a Statistical Learning Perspective", "Generalization Error Bounds for Iterative Recovery Algorithms Unfolded as Neural Networks"], "answer_arxiv_id": ["2010.15658", "2112.04364v3"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_dev_587"} +{"question": "What research work highlighted ALBEF's incorporation of the ITC loss and in-batch hard negative sampling strategy for ITM?", "answer": ["Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation"], "answer_arxiv_id": ["2107.07651"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_dev_588"} +{"question": "Which papers proposed a hyper neural network in meta-learning?", "answer": ["A Simple Neural Attentive Meta-Learner", "Conditional Neural Processes", "Meta Networks"], "answer_arxiv_id": ["1707.03141v3", "1807.01613", "1703.00837"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_dev_589"} +{"question": "Could you provide me research about data-driven methods generating 2D motions based on facial keypoints or movement frequencies?", "answer": ["Interactive Generative Adversarial Networks for Facial Expression\n Generation in Dyadic Interactions"], "answer_arxiv_id": ["1801.09092"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_dev_590"} +{"question": "What papers inspired the logical inductive bias of the study?", "answer": ["DeepProbLog: Neural Probabilistic Logic Programming", "Learning Explanatory Rules from Noisy Data"], "answer_arxiv_id": ["1805.10872", "1711.04574"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_dev_591"} +{"question": "Which research decomposes features for occupancy segmentation into a 3D space?", "answer": ["Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction"], "answer_arxiv_id": ["2302.07817"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_dev_592"} +{"question": "What research proposed building an offline memory bank or a backbone with reversible modules to address GPU memory constraints for TAD?", "answer": ["TALLFormer: Temporal Action Localization with a Long-memory Transformer", "Re^2TAL: Rewiring Pretrained Video Backbones for Reversible Temporal\n Action Localization"], "answer_arxiv_id": ["2204.01680", "2211.14053"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_593"} +{"question": "What work proposed E(3) Equivariant Diffusion Models (EDM) for molecular design?", "answer": ["Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2203.17003"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_594"} +{"question": "Which works have been proposed to solve image restoration problems using CNN?", "answer": ["Densely Residual Laplacian Super-Resolution", "Burst Image Restoration and Enhancement", "Learning Enriched Features for Real Image Restoration and Enhancement", "Multi-Stage Progressive Image Restoration", "Blueprint Separable Residual Network for Efficient Image Super-Resolution", "Attention in Attention Network for Image Super-Resolution"], "answer_arxiv_id": ["1906.12021", "2110.03680", "2003.06792", "2102.02808", "2205.05996", "2104.09497"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_595"} +{"question": "Can you name the works jointly consider bandit PCA and its rank-1 special cases?", "answer": ["Bandit Principal Component Analysis", "Bandit Phase Retrieval", "Stochastic Rank-1 Bandits"], "answer_arxiv_id": ["1902.03035", "2106.01660", "1608.03023v3"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_dev_596"} +{"question": "What studies have proposed GAN-based methods to learn OT plans?", "answer": ["Generative Adversarial Nets", "Large-Scale Optimal Transport via Adversarial Training with Cycle-Consistency", "On Scalable and Efficient Computation of Large Scale Optimal Transport", "GAN Estimation of Lipschitz Optimal Transport Maps"], "answer_arxiv_id": ["1406.2661", "2003.06635", "1905.00158", "2202.07965"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_dev_597"} +{"question": "Could you list some prior works that proposed special-purpose significance tests for different conditions in Machine Learning?", "answer": ["Replicability Analysis for Natural Language Processing: Testing Significance with Multiple Datasets"], "answer_arxiv_id": ["1709.09500"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_dev_598"} +{"question": "Could you provide me some studies about implementing Knowledge Distillation (KD) in various research fields?", "answer": ["Distilling the Knowledge in a Neural Network", "FitNets: Hints for Thin Deep Nets", "Cross-Image Relational Knowledge Distillation for Semantic Segmentation", "TinyBERT: Distilling BERT for Natural Language Understanding", "Compressing Visual-linguistic Model via Knowledge Distillation", "Multimodal Adaptive Distillation for Leveraging Unimodal Encoders for\n Vision-Language Tasks"], "answer_arxiv_id": ["1503.02531", "1412.6550", "2204.06986", "1909.10351", "2104.02096", "2204.10496"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_599"} +{"question": "Which works consider data augmentation as a viable option for improving NLI models?", "answer": ["Adversarially Regularising Neural NLI Models to Integrate Logical\n Background Knowledge", "Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and\n Improving Models", "Generating Data to Mitigate Spurious Correlations in Natural Language\n Inference Datasets"], "answer_arxiv_id": ["1808.08609", "2101.00288", "2203.12942"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_600"} +{"question": "What works apply score functions such as probability-based method, logit-based method, and feature-based method for out-of-distribution detection in computer vision?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Scaling Out-of-Distribution Detection for Real-World Settings", "On the Importance of Gradients for Detecting Distributional Shifts in the Wild", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "ReAct: Out-of-distribution Detection With Rectified Activations", "Out-of-Distribution Detection with Deep Nearest Neighbors", "Energy-based Out-of-distribution Detection", "Scaling Out-of-Distribution Detection for Real-World Settings", "ReAct: Out-of-distribution Detection With Rectified Activations", "A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "ViM: Out-Of-Distribution with Virtual-logit Matching"], "answer_arxiv_id": ["1610.02136", "1911.11132", "2110.00218", "1706.02690", "2111.12797", "2204.06507", "2010.03759", "1911.11132", "2111.12797", "1807.03888", "2203.10807"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_dev_601"} +{"question": "Could you provide me some works about association methods utilized for solving object navigation tasks?", "answer": ["Bayesian Relational Memory for Semantic Visual Navigation", "Visual Semantic Navigation using Scene Priors"], "answer_arxiv_id": ["1909.04306", "1810.06543"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_602"} +{"question": "Any studies about the Process Reward Model (PRM) and its comparison with the Outcome Reward Model (ORM)?", "answer": ["Let's Verify Step by Step"], "answer_arxiv_id": ["2305.20050"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_603"} +{"question": "Which papers describes the approach of using a small set of high-quality human-written translations or a set of translation instructions for fine-tuning LLMs in Machine Translation?", "answer": ["Eliciting the Translation Ability of Large Language Models via\n Multilingual Finetuning with Translation Instructions", "TIM: Teaching Large Language Models to Translate with Comparison"], "answer_arxiv_id": ["2305.15083", "2307.04408"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_dev_604"} +{"question": "What paper proposed a method that leverages a font-adaptive neural network and a color-preserving model for scene text editing?", "answer": ["STEFANN: Scene Text Editor using Font Adaptive Neural Network"], "answer_arxiv_id": ["1903.01192"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_dev_605"} +{"question": "Which works related to ccbo utilize an acquisition function in constrained BO methods?", "answer": ["Bayesian Optimization with Unknown Constraints", "Constrained Bayesian Optimization for Automatic Chemical Design", "Predictive Entropy Search for Efficient Global Optimization of Black-box Functions"], "answer_arxiv_id": ["1403.5607", "1709.05501", "1406.2541"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_606"} +{"question": "Which studies applied self-supervised contrastive learning methods?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance Discrimination", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Big Self-Supervised Models are Strong Semi-Supervised Learners"], "answer_arxiv_id": ["1805.01978", "1911.05722", "2002.05709", "2006.10029"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_607"} +{"question": "Which papers developed algorithms that can achieve a linear speedup for nonconvex-strongly-concave optimization problems in federated learning?", "answer": ["Federated Minimax Optimization: Improved Convergence Analyses and Algorithms", "Communication-Efficient Distributed Stochastic AUC Maximization with Deep Neural Networks"], "answer_arxiv_id": ["2203.04850", "2005.02426"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_dev_608"} +{"question": "What efforts have been made in the development of Tree-of-Thought prompting in the context of Large Language Models?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2305.10601"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_dev_609"} +{"question": "What work employs kernel ridge-regression with NTK to formulate dataset distillation?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_610"} +{"question": "What study refers to the annotation's issue of hypotheses alone being highly predictive of the label?", "answer": ["Annotation Artifacts in Natural Language Inference Data"], "answer_arxiv_id": ["1803.02324"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_611"} +{"question": "Which works discuss the area of test-time adaptation?", "answer": ["Improving robustness against common corruptions by covariate shift adaptation", "Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shift", "TTN: A Domain-Shift Aware Batch Normalization in Test-Time Adaptation", "Tent: Fully Test-Time Adaptation by Entropy Minimization", "Improving Test-Time Adaptation via Shift-agnostic Weight Regularization and Nearest Source Prototypes"], "answer_arxiv_id": ["2006.16971v2", "2006.10963", "2302.05155", "2006.10726", "2207.11707"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_612"} +{"question": "Could we name some research that utilized specifying noise for ensuring the coherent fluency between clips under different text commands?", "answer": ["Talking Head Generation with Probabilistic Audio-to-Visual Diffusion\n Priors"], "answer_arxiv_id": ["2212.04248"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_dev_613"} +{"question": "Could you give me some works that proposed methods of approximate inference or posterior sampling that could be used for a design framework?", "answer": ["Data Analysis with Bayesian Networks: A Bootstrap Approach", "Variational Causal Networks: Approximate Bayesian Inference over Causal Structures", "DiBS: Differentiable Bayesian Structure Learning", "BCD Nets: Scalable Variational Approaches for Bayesian Causal Discovery", "Bayesian Structure Learning with Generative Flow Networks", "Bayesian learning of Causal Structure and Mechanisms with GFlowNets and Variational Bayes"], "answer_arxiv_id": ["1301.6695v1", "2106.07635", "2105.11839", "2112.02761", "2202.13903", "2211.02763v3"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_dev_614"} +{"question": "What studies use “instance slots” in their models and solve their routing problem through mixture models with amortized variational inference?", "answer": ["Tagger: Deep Unsupervised Perceptual Grouping", "Multi-Object Representation Learning with Iterative Variational Inference"], "answer_arxiv_id": ["1606.06724", "1903.00450"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_615"} +{"question": "Any works targeting the enhancement of LLMs’ factuality?", "answer": ["Aligning Large Multimodal Models with Factually Augmented RLHF", "LIMA: Less Is More for Alignment", "Let's Verify Step by Step", "Check Your Facts and Try Again: Improving Large Language Models with\n External Knowledge and Automated Feedback", "Chain-of-Knowledge: Grounding Large Language Models via Dynamic\n Knowledge Adapting over Heterogeneous Sources", "When Not to Trust Language Models: Investigating Effectiveness of\n Parametric and Non-Parametric Memories", "A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of\n LLMs by Validating Low-Confidence Generation"], "answer_arxiv_id": ["2309.14525", "2305.11206", "2305.20050", "2302.12813", "2305.13269", "2212.10511", "2307.03987"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_dev_616"} +{"question": "Which work propose to learn a score-based Average Thresholded Confidence (ATC) by leveraging the softmax probability of a CNN classifiers?", "answer": ["Leveraging Unlabeled Data to Predict Out-of-Distribution Performance"], "answer_arxiv_id": ["2201.04234"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_dev_617"} +{"question": "Which studies showed the recovery of samples from a training dataset using the gradients generated during training?", "answer": ["Deep Leakage from Gradients"], "answer_arxiv_id": ["1906.08935"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_618"} +{"question": "What studies have applied parallel atrous convolutions in their method for deblurring?", "answer": ["Single Image Defocus Deblurring Using Kernel-Sharing Parallel Atrous\n Convolutions"], "answer_arxiv_id": ["2108.09108"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_619"} +{"question": "What papers are about white-box attacks, a type of adversarial attacks where the adversary has complete access to the model parameters?", "answer": ["Intriguing properties of neural networks", "Towards Evaluating the Robustness of Neural Networks", "Obfuscated Gradients Give a False Sense of Security: Circumventing\n Defenses to Adversarial Examples"], "answer_arxiv_id": ["1312.6199", "1608.04644", "1802.00420"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_620"} +{"question": "Could you name the studies that implemented recurrent neural networks (RNN) for improving real-time performance in full-body motion estimation?", "answer": ["Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time"], "answer_arxiv_id": ["1810.04703"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_621"} +{"question": "What papers have demonstrated the efficacy of label smoothing in both visual and language domains?", "answer": ["Rethinking the Inception Architecture for Computer Vision", "When Does Label Smoothing Help?", "Regularization via Structural Label Smoothing", "Adaptive Label Smoothing", "Adaptive Label Smoothing with Self-Knowledge in Natural Language\n Generation"], "answer_arxiv_id": ["1512.00567", "1906.02629", "2001.01900", "2009.06432", "2210.13459"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_dev_622"} +{"question": "In what studies were Error Imputation-based (EIB) unbiased learning method derived?", "answer": ["Collaborative Filtering and the Missing at Random Assumption"], "answer_arxiv_id": ["1206.5267"], "source_meta": {"published_time": "20220510"}, "qid": "AutoScholarQuery_dev_623"} +{"question": "What papers discuss the use of Wikipedia as a multilingual dataset for pretraining language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Megatron-LM: Training Multi-Billion Parameter Language Models Using\n Model Parallelism", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1810.04805", "1909.08053", "1910.10683", "2005.14165"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_dev_624"} +{"question": "What papers have studied the use of reinforcement learning or evolutionary algorithms in neural architecture search (NAS)?", "answer": ["Neural Architecture Search with Reinforcement Learning", "Designing Neural Network Architectures using Reinforcement Learning", "Practical Block-wise Neural Network Architecture Generation", "Large-Scale Evolution of Image Classifiers", "Hierarchical Representations for Efficient Architecture Search", "Efficient Multi-objective Neural Architecture Search via Lamarckian Evolution", "Regularized Evolution for Image Classifier Architecture Search"], "answer_arxiv_id": ["1611.01578", "1611.02167", "1708.05552", "1703.01041", "1711.00436", "1804.09081", "1802.01548"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_dev_625"} +{"question": "Which studies achieved success in self-supervised representation learning through contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "Debiased Contrastive Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2003.04297", "2007.00224", "2104.14294"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_dev_626"} +{"question": "What research efforts have been made to improve efficient INT8 quantisation?", "answer": ["ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers", "LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models"], "answer_arxiv_id": ["2206.01861", "2208.07339", "2211.10438"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_dev_627"} +{"question": "Which study first introduced the Dynamic routing in capsule networks?", "answer": ["Dynamic Routing Between Capsules"], "answer_arxiv_id": ["1710.09829"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_dev_628"} +{"question": "What papers discuss the influences of multiple training examples on a model's prediction?", "answer": ["On the Accuracy of Influence Functions for Measuring Group Effects"], "answer_arxiv_id": ["1905.13289"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_dev_629"} +{"question": "What works revealed that the expressiveness of MPNNs and k-GNNs is bounded by k-WL?", "answer": ["How Powerful are Graph Neural Networks?", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks"], "answer_arxiv_id": ["1810.00826", "1810.02244"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_630"} +{"question": "Which works enhanced diffusion based on an initial input graph in latent graph and topology inference?", "answer": ["Diffusion Improves Graph Learning", "On the Bottleneck of Graph Neural Networks and its Practical Implications"], "answer_arxiv_id": ["1911.05485", "2006.05205"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_dev_631"} +{"question": "What works proposed methods of improving the generation quality for SLP using adversarial training, mixture density networks, and dictionary representations?", "answer": ["Adversarial Training for Multi-Channel Sign Language Production", "Mixed SIGNals: Sign Language Production via a Mixture of Motion\n Primitives", "Signing at Scale: Learning to Co-Articulate Signs for Large-Scale\n Photo-Realistic Sign Language Production"], "answer_arxiv_id": ["2008.12405", "2107.11317", "2203.15354"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_632"} +{"question": "Which work drove the success of VLMs by training transformers on large scale image-text pairs data using contrastive learning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_dev_633"} +{"question": "Could you name some methods that require a neural network forward pass to get embeddings?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Beyond neural scaling laws: beating power law scaling via data pruning", "RETRIEVE: Coreset Selection for Efficient and Robust Semi-Supervised Learning"], "answer_arxiv_id": ["1708.00489", "2206.14486v6", "2106.07760v2"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_dev_634"} +{"question": "Which works proposed variations of non-local blocks for aggregating long-range context in semantic segmentation models?", "answer": ["Dual Attention Network for Scene Segmentation", "OCNet: Object Context for Semantic Segmentation", "CCNet: Criss-Cross Attention for Semantic Segmentation", "Non-local Neural Networks"], "answer_arxiv_id": ["1809.02983", "1809.00916", "1811.11721", "1711.07971"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_635"} +{"question": "In what work was mentioned the fine-tuning of GPT-3 (175B) for answering open-domain questions?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback"], "answer_arxiv_id": ["2112.09332"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_dev_636"} +{"question": "In which work were pretrained language models augmented with a mechanism to directly attend to a single context image?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_dev_637"} +{"question": "What works developed local methods for finding meaningful latent perturbations?", "answer": ["A Spectral Regularizer for Unsupervised Disentanglement", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated Images using Conditional Continuous Normalizing Flows", "Low-Rank Subspaces in GANs", "Do Not Escape From the Manifold: Discovering the Local Coordinates on the Latent Space of GANs"], "answer_arxiv_id": ["1812.01161v2", "2103.17249", "2008.02401", "2106.04488", "2106.06959"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_dev_638"} +{"question": "Which papers made contributions in designing hierarchical Transformer architectures for document classification?", "answer": ["Hierarchical Transformers for Long Document Classification", "Revisiting Transformer-based Models for Long Document Classification"], "answer_arxiv_id": ["1910.10781", "2204.06683"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_dev_639"} +{"question": "Could you provide me some studies that proposed strategies to mitigate biases in NLI models?", "answer": ["Unlearn Dataset Bias in Natural Language Inference by Fitting the\n Residual", "Don't Take the Easy Way Out: Ensemble Based Methods for Avoiding Known\n Dataset Biases", "End-to-End Bias Mitigation by Modelling Biases in Corpora"], "answer_arxiv_id": ["1908.10763", "1909.03683", "1909.06321"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_640"} +{"question": "Which papers focus on combining vision and language inputs in an embodied setting with the goal of direct action prediction?", "answer": ["Instruction-driven history-aware policies for robotic manipulations", "Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation", "CLIPort: What and Where Pathways for Robotic Manipulation", "Hierarchical Task Learning from Language Instructions with Unified Transformers and Self-Monitoring", "BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning", "Learning Language-Conditioned Robot Behavior from Offline Data and Crowd-Sourced Annotation", "Interactive Language: Talking to Robots in Real Time", "RT-1: Robotics Transformer for Real-World Control at Scale"], "answer_arxiv_id": ["2209.04899", "2209.05451", "2109.12098", "2106.03427", "2202.02005", "2109.01115", "2210.06407", "2212.06817"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_dev_641"} +{"question": "Which papers explored enhancing the accuracy of responses by concurrently generating reasoning processes while producing answers?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Navigate through Enigmatic Labyrinth A Survey of Chain of Thought\n Reasoning: Advances, Frontiers and Future"], "answer_arxiv_id": ["2201.11903", "2309.15402"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_642"} +{"question": "Which papers have conducted a sublinear convergence analysis of softmax tabular policies?", "answer": ["Adaptive Trust Region Policy Optimization: Global Convergence and Faster Rates for Regularized MDPs", "On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift", "Finite-Sample Analysis of Off-Policy Natural Actor-Critic Algorithm", "On the Convergence Rates of Policy Gradient Methods", "A Theory of Regularized Markov Decision Processes"], "answer_arxiv_id": ["1909.02769", "1908.00261", "2102.09318", "2201.07443", "1901.11275"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_643"} +{"question": "Which study proposed OpenSeg, a technique for fine-tuning a model using class-agnostic masks and image-text pair data?", "answer": ["Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2112.12143", "2102.05918"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_dev_644"} +{"question": "What are the works that handled volumetric radiative decomposition?", "answer": ["Deep Reflectance Volumes: Relightable Reconstructions from Multi-View Photometric Images", "DeRF: Decomposed Radiance Fields", "NeRFactor: Neural Factorization of Shape and Reflectance Under an Unknown Illumination", "Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance Fields"], "answer_arxiv_id": ["2007.09892", "2011.12490", "2106.01970", "2112.03907"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_dev_645"} +{"question": "Which studies discuss hallucinations in LLMs?", "answer": ["Generative Judge for Evaluating Alignment", "Siren's Song in the AI Ocean: A Survey on Hallucination in Large\n Language Models", "A Comprehensive Survey of Hallucination Mitigation Techniques in Large\n Language Models"], "answer_arxiv_id": ["2310.05470", "2309.01219", "2401.01313"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_dev_646"} +{"question": "What papers have been written on incorporating unsafe prompt detection into online ChatBot and LLM-integrated applications?", "answer": ["Augmented Language Models: a Survey", "Toolformer: Language Models Can Teach Themselves to Use Tools", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face"], "answer_arxiv_id": ["2302.07842", "2302.04761", "2303.17580"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_dev_647"} +{"question": "What studies have used prototypical networks and neural processes in the field of efficient meta-learning?", "answer": ["Prototypical Networks for Few-shot Learning", "Memory Efficient Meta-Learning with Large Images", "Neural Processes", "Conditional Neural Processes", "Meta-Learning surrogate models for sequential decision making"], "answer_arxiv_id": ["1703.05175", "2107.01105", "1807.01622", "1807.01613", "1903.11907"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_dev_648"} +{"question": "Which papers discuss about equivariant neural networks for voxel grids with respect to voxel and point cloud representations?", "answer": ["3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data", "An end-to-end SE(3)-equivariant segmentation network", "Sparse Steerable Convolutions: An Efficient Learning of SE(3)-Equivariant Features for Estimation and Tracking of Object Poses in 3D Space"], "answer_arxiv_id": ["1807.02547", "2303.00351", "2111.07383"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_649"} +{"question": "Which works are about the sampling-based uncertainty estimation methods?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Dynamic Bayesian Combination of Multiple Imperfect Classifiers", "BatchEnsemble: An alternative approach to Efficient Ensemble and Lifelong Learning", "Hyperparameter Ensembles for Robustness and Uncertainty Quantification", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Weight Uncertainty in Neural Networks", "A Simple Baseline for Bayesian Uncertainty in Deep Learning"], "answer_arxiv_id": ["1612.01474", "1206.1831", "2002.06715", "2006.13570", "1506.02142", "1505.05424", "1902.02476"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_dev_650"} +{"question": "Which works have explored white-box detection methods involving watermarks in LLM-generated texts?", "answer": ["The Science of Detecting LLM-Generated Texts", "Watermarking Text Generated by Black-Box Language Models"], "answer_arxiv_id": ["2303.07205", "2305.08883v1"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_651"} +{"question": "Could you give me examples of studies that made significant progress in multimodal response generation?", "answer": ["CM3: A Causal Masked Multimodal Model of the Internet", "Meta-Transformer: A Unified Framework for Multimodal Learning"], "answer_arxiv_id": ["2201.07520", "2307.10802"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_dev_652"} +{"question": "What papers have proposed methods to learn representations that are invariant to image distractors such as background colour?", "answer": ["Invariant Causal Prediction for Block MDPs", "Learning Invariant Representations for Reinforcement Learning without Reconstruction", "Domain Adversarial Reinforcement Learning", "Learning Markov State Abstractions for Deep Reinforcement Learning"], "answer_arxiv_id": ["2003.06016", "2006.10742", "2102.07097", "2106.04379"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_653"} +{"question": "Which paper provides a discussion on the many-body representation hypothesis in context of voxel and point cloud representations?", "answer": ["ATOM3D: Tasks On Molecules in Three Dimensions"], "answer_arxiv_id": ["2012.04035"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_654"} +{"question": "What is the first trial on instruction-following LMMs?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_dev_655"} +{"question": "What research work incorporated attention into the capsule routing via a non-iterative feed-forward operation?", "answer": ["Attention routing between capsules"], "answer_arxiv_id": ["1907.01750"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_dev_656"} +{"question": "Which works were mentioned in relation to the use of synthetic captions generated using BLIP and ranked using CLIP models?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2201.12086", "2103.00020"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_657"} +{"question": "Which research suggested that jittering can enhance worst-case robustness?", "answer": ["Solving Inverse Problems With Deep Neural Networks – Robustness Included?"], "answer_arxiv_id": ["2011.04268"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_658"} +{"question": "Which papers explored generating responses to queries using multi-modal knowledge sources?", "answer": ["WebQA: Multihop and Multimodal QA", "MuRAG: Multimodal Retrieval-Augmented Generator for Open Question Answering over Images and Text", "Conversational Question Answering on Heterogeneous Sources", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering", "MultiModalQA: complex question answering over text, tables and images", "Towards Multi-Modal DBMSs for Seamless Querying of Texts and Tables", "MultiHiertt: Numerical Reasoning over Multi Hierarchical Tabular and Textual Data"], "answer_arxiv_id": ["2109.00590", "2210.02928", "2204.11677", "2209.09513", "2104.06039", "2304.13559", "2206.01347"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_dev_659"} +{"question": "Could you tell me about some research papers that have used a 3D native pipeline for diffusion-based text-to-3D work?", "answer": ["Shap-E: Generating Conditional 3D Implicit Functions", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "ATT3D: Amortized Text-to-3D Object Synthesis", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "Single-Stage Diffusion NeRF: A Unified Approach to 3D Generation and\n Reconstruction", "3DGen: Triplane Latent Diffusion for Textured Mesh Generation", "AutoDecoding Latent 3D Diffusion Models"], "answer_arxiv_id": ["2305.02463", "2212.08751", "2306.07349", "2212.04493", "2304.06714", "2303.05371", "2307.05445"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_660"} +{"question": "What studies improve the approximation factor to 3/4, and then to 3/4+o​(1)?", "answer": ["An Improved Approximation Algorithm for Maximin Shares", "Simplification and Improvement of MMS Approximation"], "answer_arxiv_id": ["1903.00029v3", "2303.16788v2"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_dev_661"} +{"question": "What previous studies explored the trade-off between communication and straggler resiliency in Gradient Coding?", "answer": ["Communication-Computation Efficient Gradient Coding", "Communication-Efficient Gradient Coding for Straggler Mitigation in Distributed Learning"], "answer_arxiv_id": ["1802.03475", "2005.07184"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_dev_662"} +{"question": "What papers in the field of NLP have explored the topic of hate/offense or toxicity?", "answer": ["Automated Hate Speech Detection and the Problem of Offensive Language"], "answer_arxiv_id": ["1703.04009"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_663"} +{"question": "What are the early studies on Random Reshuffling (SGD-RR) that proposed upper bounds for strongly convex and twice-smooth objectives?", "answer": ["Why Random Reshuffling Beats Stochastic Gradient Descent", "Random Shuffling Beats SGD after Finite Epochs"], "answer_arxiv_id": ["1510.08560", "1806.10077"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_664"} +{"question": "What works proposed the approach of embedding images and text into a shared space?", "answer": ["Deep Fragment Embeddings for Bidirectional Image Sentence Mapping", "Explain Images with Multimodal Recurrent Neural Networks"], "answer_arxiv_id": ["1406.5679", "1410.1090"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_dev_665"} +{"question": "What work introduced rotating features to complex-valued activations by extending Convolutional Auto-Encoders?", "answer": ["Rotating Features for Object Discovery"], "answer_arxiv_id": ["2306.00600"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_666"} +{"question": "What papers explored spatio-temporal information in visual object tracking?", "answer": ["Unified Sequence-to-Sequence Learning for Single- and Multi-Modal Visual\n Object Tracking", "Compact Transformer Tracker with Correlative Masked Modeling", "MixFormer: End-to-End Tracking with Iterative Mixed Attention"], "answer_arxiv_id": ["2304.14394", "2301.10938", "2203.11082"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_dev_667"} +{"question": "What works introduced the prompts paradigm to Vision Transformer?", "answer": ["Visual Prompt Tuning", "Exploring Visual Prompts for Adapting Large-Scale Models"], "answer_arxiv_id": ["2203.12119", "2203.17274"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_dev_668"} +{"question": "Could you provide me some studies about the use of transformers in visual object tracking?", "answer": ["Attention Is All You Need", "Transformer Tracking", "Learning Spatio-Temporal Transformer for Visual Tracking", "SwinTrack: A Simple and Strong Baseline for Transformer Tracking", "MixFormer: End-to-End Tracking with Iterative Mixed Attention", "Joint Feature Learning and Relation Modeling for Tracking: A One-Stream\n Framework", "Correlation-Aware Deep Tracking"], "answer_arxiv_id": ["1706.03762", "2103.15436", "2103.17154", "2112.00995", "2203.11082", "2203.11991", "2203.01666"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_dev_669"} +{"question": "What studies exploit depth maps for view-morphing to augment sparse inputs?", "answer": ["VM-NeRF: Tackling Sparsity in NeRF with View Morphing"], "answer_arxiv_id": ["2210.04214"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_670"} +{"question": "Could you provide some of the initial explorations on LLMs, which involve prompting methods and model variants?", "answer": ["GPTScore: Evaluate as You Desire", "Large Language Models Are State-of-the-Art Evaluators of Translation\n Quality", "Is ChatGPT a Good NLG Evaluator? A Preliminary Study", "G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment"], "answer_arxiv_id": ["2302.04166", "2302.14520", "2303.04048", "2303.16634"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_dev_671"} +{"question": "What works propose optimizing a surrogate loss function to enhance stability in learning?", "answer": ["Trust Region Policy Optimization", "Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1502.05477", "1707.06347v2"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_672"} +{"question": "Are there papers that analyze the lower bounds of PFL?", "answer": ["Is Local SGD Better than Minibatch SGD?", "Minibatch vs Local SGD for Heterogeneous Distributed Learning", "Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond"], "answer_arxiv_id": ["2002.07839", "2006.04735", "2110.10342"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_673"} +{"question": "What studies used distillation techniques for 'gisting' to make shorter prompts?", "answer": ["Learning to Compress Prompts with Gist Tokens"], "answer_arxiv_id": ["2304.08467"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_dev_674"} +{"question": "Could you provide me some studies about probing-based methods for factuality detection in LLM ?", "answer": ["Understanding intermediate layers using linear classifier probes", "Language Models Represent Space and Time", "Language Models (Mostly) Know What They Know", "The Internal State of an LLM Knows When It's Lying", "Representation Engineering: A Top-Down Approach to AI Transparency", "Hallucination Detection: Robustly Discerning Reliable Answers in Large Language Models"], "answer_arxiv_id": ["1610.01644", "2310.02207", "2207.05221", "2304.13734", "2310.01405", "2407.04121v1"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_dev_675"} +{"question": "What are the references that discuss the estimation of Q functions and learned transition models under epistemic uncertainty?", "answer": ["Deep Exploration via Bootstrapped DQN", "Randomized Prior Functions for Deep Reinforcement Learning", "Conservative Safety Critics for Exploration", "Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models", "Model-Ensemble Trust-Region Policy Optimization", "When to Trust Your Model: Model-Based Policy Optimization", "A Game Theoretic Framework for Model Based Reinforcement Learning", "Constrained Policy Optimization via Bayesian World Models"], "answer_arxiv_id": ["1602.04621", "1806.03335v2", "2010.14497", "1805.12114", "1802.10592", "1906.08253", "2004.07804", "2201.09802"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_676"} +{"question": "Could you provide me some researches that develop learning-augmented algorithms for metrical task systems?", "answer": ["Online metric algorithms with untrusted predictions"], "answer_arxiv_id": ["2003.02144v3"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_dev_677"} +{"question": "Which studies highlighted the vulnerability of contrastive learning to adversarial attack in downstream classification tasks?", "answer": ["Contrastive Learning with Adversarial Examples", "Adversarial Self-Supervised Contrastive Learning"], "answer_arxiv_id": ["2010.12050", "2006.07589"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_dev_678"} +{"question": "Which research utilized the reverse KLD to improve the accuracy of language generation in the MINILLM?", "answer": ["MiniLLM: Knowledge Distillation of Large Language Models"], "answer_arxiv_id": ["2306.08543"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_679"} +{"question": "What works use MIAs to assess whether a given data point was used within the prompt prepended to the inputs of a trained LLM?", "answer": ["Membership Inference Attacks From First Principles", "Membership Inference Attacks Against Machine Learning Models"], "answer_arxiv_id": ["2112.03570", "1610.05820"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_680"} +{"question": "Could you provide studies that use diffusion models in relation to computer vision problems?", "answer": ["Diffusion Autoencoders: Toward a Meaningful and Decodable Representation", "Diffusion Based Representation Learning", "Label-Efficient Semantic Segmentation with Diffusion Models"], "answer_arxiv_id": ["2111.15640", "2105.14257", "2112.03126"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_dev_681"} +{"question": "Which papers propose methods to predict 6D pose of objects in an image to find applications in fields like robotics, autonomous vehicles, and microscopy?", "answer": ["Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects", "CryoAI: Amortized Inference of Poses for Ab Initio Reconstruction of 3D Molecular Volumes from Real Cryo-EM Images"], "answer_arxiv_id": ["1809.10790", "2203.08138"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_682"} +{"question": "What are the works that addressed the differences between individual annotators or the group-level attributes of annotators by adding individual layers?", "answer": ["Dealing with Disagreements: Looking Beyond the Majority Vote in\n Subjective Annotations", "Jury Learning: Integrating Dissenting Voices into Machine Learning\n Models", "When the Majority is Wrong: Modeling Annotator Disagreement for\n Subjective Tasks"], "answer_arxiv_id": ["2110.05719", "2202.02950", "2305.06626"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_683"} +{"question": "Who has recently explored an instruction-based text embedder?", "answer": ["One Embedder, Any Task: Instruction-Finetuned Text Embeddings"], "answer_arxiv_id": ["2212.09741"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_dev_684"} +{"question": "Which studies align the text with a paired image in the embedding space in visual-language learning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Iterative Prompt Learning for Unsupervised Backlit Image Enhancement"], "answer_arxiv_id": ["2103.00020", "2303.17569"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_685"} +{"question": "Are there any works that focus on image inpainting methods that don't require finetuning?", "answer": ["Blended Latent Diffusion", "Blended Diffusion for Text-driven Editing of Natural Images"], "answer_arxiv_id": ["2206.02779", "2111.14818"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_dev_686"} +{"question": "What works used model's weights to identify parts of the training dataset that influenced the model?", "answer": ["ORCA: Interpreting Prompted Language Models via Locating Supporting Data\n Evidence in the Ocean of Pretraining Data", "Studying Large Language Model Generalization with Influence Functions"], "answer_arxiv_id": ["2205.12600", "2308.03296v1"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_687"} +{"question": "What studies demonstrate that a simple image classifier trained on a specific CNN generator is able to generalize well to unseen architectures?", "answer": ["CNN-generated images are surprisingly easy to spot… for now"], "answer_arxiv_id": ["1912.11035"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_688"} +{"question": "What paper introduced a method for identifying close and robust counterfactuals which use interval neural networks?", "answer": ["Formalising the Robustness of Counterfactual Explanations for Neural Networks"], "answer_arxiv_id": ["2208.14878"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_dev_689"} +{"question": "Are there any studies focussed on managing false negatives in contrastive learning, particularly for the vision domain?", "answer": ["Probabilistic Embeddings for Cross-Modal Retrieval", "Polysemous Visual-Semantic Embedding for Cross-Modal Retrieval", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "GRIT-VLP: Grouped Mini-batch Sampling for Efficient Vision and Language\n Pre-training", "Boosting Contrastive Self-Supervised Learning with False Negative\n Cancellation", "Debiased Contrastive Learning", "Incremental False Negative Detection for Contrastive Learning", "Contrastive Learning with Hard Negative Samples"], "answer_arxiv_id": ["2101.05068", "1906.04402", "2107.07651", "2208.04060", "2011.11765", "2007.00224", "2106.03719", "2010.04592"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_dev_690"} +{"question": "Which study represents the embedding-based method in the method of jointly learning the logical rule form and the weights in a differentiable manner?", "answer": ["Embedding Entities and Relations for Learning and Inference in Knowledge Bases"], "answer_arxiv_id": ["1412.6575"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_691"} +{"question": "What studies focus on developing an estimator for model's classifier performance on unlabeled data from unknown distributions in the target domain?", "answer": ["Are Labels Always Necessary for Classifier Accuracy Evaluation?", "Leveraging Unlabeled Data to Predict Out-of-Distribution Performance", "Predicting Out-of-Distribution Error with the Projection Norm", "On the Strong Correlation Between Model Invariance and Generalization", "Predicting with Confidence on Unseen Distributions", "What Does Rotation Prediction Tell Us about Classifier Accuracy under Varying Testing Environments?"], "answer_arxiv_id": ["2007.02915", "2201.04234", "2202.05834", "2207.07065", "2107.03315", "2106.05961"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_dev_692"} +{"question": "Which studies discuss using modifications like larger/smaller learning rates and regularization-based methods for enhancing Fine-Tuning's performance?", "answer": ["SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization"], "answer_arxiv_id": ["1911.03437"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_dev_693"} +{"question": "What research studies proposed improved concentration coefficients than AMPO?", "answer": ["Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies"], "answer_arxiv_id": ["2210.01400v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_694"} +{"question": "Which studies developed physics-based LiDAR simulators?", "answer": ["CARLA: An Open Urban Driving Simulator"], "answer_arxiv_id": ["1711.03938"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_695"} +{"question": "Can you name some studies that focused on cross-style or zero-shot classification in NLP?", "answer": ["Zero-shot Text Classification With Generative Language Models", "Style is NOT a single variable: Case Studies for Cross-Style Language\n Understanding"], "answer_arxiv_id": ["1912.10165", "1911.03663"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_696"} +{"question": "Which works used models structured as an RNN in meta-RL methods?", "answer": ["RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "Learning to reinforcement learn", "Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPs"], "answer_arxiv_id": ["1611.02779", "1611.05763", "2110.05038"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_697"} +{"question": "Any studies tackling the challenge of Uncertainty Estimation in specific NLP tasks such as paraphrase detection and natural language inference?", "answer": ["Calibration of Pre-trained Transformers"], "answer_arxiv_id": ["2003.07892"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_698"} +{"question": "Which works explored the offline label shift problem in the domain adaptation literature and estimating mixture proportions of different classes in unlabeled data?", "answer": ["Detecting and Correcting for Label Shift with Black Box Predictors", "A Unified View of Label Shift Estimation", "Mixture Proportion Estimation and PU Learning: A Modern Approach", "Domain Adaptation under Open Set Label Shift", "Unsupervised Learning under Latent Label Shift"], "answer_arxiv_id": ["1802.03916", "2003.07554", "2111.00980", "2207.13048", "2207.13179"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_699"} +{"question": "What studies introduced improvements or advancements in stereo egocentric setups?", "answer": ["UnrealEgo: A New Dataset for Robust Egocentric 3D Human Motion Capture", "Ego3DPose: Capturing 3D Cues from Binocular Egocentric Views"], "answer_arxiv_id": ["2208.01633", "2309.11962"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_700"} +{"question": "Which paper introduces the MoTIF dataset with a large number of task demonstrations?", "answer": ["Mobile App Tasks with Iterative Feedback (MoTIF): Addressing Task Feasibility in Interactive Visual Environments"], "answer_arxiv_id": ["2104.08560"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_701"} +{"question": "Which papers propose creating synthetic data using copulas for answering marginal queries?", "answer": ["Differentially Private Release of High-Dimensional Datasets using the Gaussian Copula"], "answer_arxiv_id": ["1902.01499"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_dev_702"} +{"question": "Which studies evaluate the LM’s success at performing multi-hop inferences with the edited information?", "answer": ["MQuAKE: Assessing Knowledge Editing in Language Models via Multi-Hop Questions", "EasyEdit: An Easy-to-use Knowledge Editing Framework for Large Language Models", "Evaluating the Ripple Effects of Knowledge Editing in Language Models"], "answer_arxiv_id": ["2305.14795", "2308.07269v3", "2307.12976"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_dev_703"} +{"question": "Which research work provided both sublinear and linear convergence analysis of natural policy gradient (NPG) with softmax tabular policies or with log-linear policies?", "answer": ["On the Convergence Rates of Policy Gradient Methods", "Linear Convergence for Natural Policy Gradient with Log-linear Policy Parametrization", "Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies"], "answer_arxiv_id": ["2201.07443", "2209.15382", "2210.01400v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_704"} +{"question": "Which research work offered theoretical analyses on the issue of balancing the generator and discriminator in GAN training?", "answer": ["Generalization and Equilibrium in Generative Adversarial Nets (GANs)", "Approximability of Discriminators Implies Diversity in GANs"], "answer_arxiv_id": ["1703.00573", "1806.10586"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_dev_705"} +{"question": "Which studies decomposed the problem into a tree or constructed a reasoning graph instead?", "answer": ["Probabilistic Tree-of-thought Reasoning for Answering\n Knowledge-intensive Complex Questions", "Graph Elicitation for Guiding Multi-Step Reasoning in Large Language\n Models"], "answer_arxiv_id": ["2311.13982", "2311.09762"], "source_meta": {"published_time": "20240628"}, "qid": "AutoScholarQuery_dev_706"} +{"question": "Which studies discuss the threat of gradient inversion to Federated Learning (FL)?", "answer": ["Towards General Deep Leakage in Federated Learning", "Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix", "CAFE: Catastrophic Data Leakage in Vertical Federated Learning"], "answer_arxiv_id": ["2110.09074", "2106.06089", "2110.15122v4"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_707"} +{"question": "Any works that applied reinforcement learning (RL) and planning algorithms for code generation by formulating the code generation problem as a sequential decision-making problem?", "answer": ["Leveraging Grammar and Reinforcement Learning for Neural Program Synthesis"], "answer_arxiv_id": ["1805.04276"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_dev_708"} +{"question": "Can you indicate some studies addressing security and safety concerns in the deployment of multi-modal models in real-world applications such as autonomous driving?", "answer": ["SNE-RoadSeg: Incorporating Surface Normal Information into Semantic\n Segmentation for Accurate Freespace Detection", "DeepFusion: Lidar-Camera Deep Fusion for Multi-Modal 3D Object Detection", "MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving", "End-to-end Autonomous Driving with Semantic Depth Cloud Mapping and\n Multi-agent"], "answer_arxiv_id": ["2008.11351", "2203.08195", "1612.07695", "2204.05513"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_709"} +{"question": "Which papers have focused on highlighting decision words as a method for explaining predictions of neural NLP systems?", "answer": ["A causal framework for explaining the predictions of black-box\n sequence-to-sequence models", "Is Attention Interpretable?"], "answer_arxiv_id": ["1707.01943", "1906.03731"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_710"} +{"question": "Which work introduced the General Language Understanding Evaluation (GLUE) benchmark in NLP?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language\n Understanding"], "answer_arxiv_id": ["1804.07461"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_dev_711"} +{"question": "Which papers attempted to handle the problem of GAN compression through the use of pruning-based methods?", "answer": ["Co-Evolutionary Compression for Unpaired Image Translation", "Teachers Do More Than Teach: Compressing Image-to-Image Models"], "answer_arxiv_id": ["1907.10804", "2103.03467"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_dev_712"} +{"question": "What paper proposed the method ExpertPrompting to improve the reasoning capabilities of LLMs by generating expert-level responses?", "answer": ["ExpertPrompting: Instructing Large Language Models to be Distinguished\n Experts"], "answer_arxiv_id": ["2305.14688"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_dev_713"} +{"question": "Could you provide me the study that proposed the relative gradient method for optimizing flow-based models with arbitrary linear transformations?", "answer": ["Relative gradient optimization of the Jacobian term in unsupervised deep learning"], "answer_arxiv_id": ["2006.15090"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_714"} +{"question": "What research aimed to explain collaboration mechanism in a social psychology view?", "answer": ["Exploring Collaboration Mechanisms for LLM Agents: A Social Psychology\n View"], "answer_arxiv_id": ["2310.02124"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_715"} +{"question": "What works have been done on graph neural networks (GNNs) that are used for graph classification problems?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "Neural Message Passing for Quantum Chemistry", "Graph U-Nets"], "answer_arxiv_id": ["1609.02907", "1710.10903", "1704.01212", "1905.05178"], "source_meta": {"published_time": "20220226"}, "qid": "AutoScholarQuery_dev_716"} +{"question": "What are the examples of works on text-to-image diffusion models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "KNN-Diffusion: Image Generation via Large-Scale Retrieval"], "answer_arxiv_id": ["2112.10741", "2205.11487", "2204.06125", "2112.10752", "2204.02849"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_dev_717"} +{"question": "Which works are about extending instruction finetuned datasets outside of English through translation?", "answer": ["NusaX: Multilingual Parallel Sentiment Dataset for 10 Indonesian Local\n Languages"], "answer_arxiv_id": ["2205.15960"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_dev_718"} +{"question": "Can you provide some works that have incorporated human prior models like SMPL and imGHUM for text-driven 3D human generation?", "answer": ["imGHUM: Implicit Generative Models of 3D Human Shape and Articulated\n Pose", "EVA3D: Compositional 3D Human Generation from 2D Image Collections", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "AvatarCraft: Transforming Text into Neural Human Avatars with\n Parameterized Shape and Pose Control", "AvatarBooth: High-Quality and Customizable 3D Human Avatar Generation", "Text-Guided Generation and Editing of Compositional 3D Avatars", "TextDeformer: Geometry Manipulation using Text Guidance", "Zero-Shot Text-to-Parameter Translation for Game Character Auto-Creation", "Learning Hierarchical Cross-Modal Association for Co-Speech Gesture\n Generation"], "answer_arxiv_id": ["2108.10842", "2210.04888", "2304.00916", "2303.17606", "2306.09864", "2309.07125", "2304.13348", "2303.01311", "2203.13161"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_719"} +{"question": "Could you provide some studies discussing the high-level connection between stability, online learnability, and differential privacy?", "answer": ["Preserving Statistical Validity in Adaptive Data Analysis", "Algorithmic Stability for Adaptive Data Analysis", "A Limitation of the PAC-Bayes Framework"], "answer_arxiv_id": ["1411.2664", "1511.02513", "2006.13508"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_720"} +{"question": "Can you provide me with studies that explored variants of the Cross Entropy (CE) loss to improve discriminative power of learned feature representations of data?", "answer": ["FaceNet: A Unified Embedding for Face Recognition and Clustering", "CONTRASTIVE-CENTER LOSS FOR DEEP NEURAL NETWORKS"], "answer_arxiv_id": ["1503.03832", "1707.07391"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_dev_721"} +{"question": "What studies does AMPO recovers the best-known convergence rates in both the tabular and non-tabular setting?", "answer": ["A Theory of Regularized Markov Decision Processes", "On the Convergence Rates of Policy Gradient Methods", "Finite-time analysis of entropy-regularized neural natural actor-critic algorithm", "Linear Convergence for Natural Policy Gradient with Log-linear Policy Parametrization", "Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies"], "answer_arxiv_id": ["1901.11275", "2201.07443", "2206.00833", "2209.15382", "2210.01400v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_722"} +{"question": "What works contribute to the theoretical study of linear MDPs and linear convergence theory of AMPO?", "answer": ["Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1907.05388"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_723"} +{"question": "What papers propose code filtering strategies involving code execution with given or generated test cases?", "answer": ["Learning to Execute Actions or Ask Clarification Questions", "[2203.07814] Competition-Level Code Generation with AlphaCode", "CodeT: Code Generation with Generated Tests", "Enhancing Large Language Models in Coding Through Multi-Perspective\n Self-Consistency"], "answer_arxiv_id": ["2204.08373", "2203.07814", "2207.10397", "2309.17272"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_dev_724"} +{"question": "Which work propose Gaussian posterior approximation method (SWAG) based on the first two moments of SGD iterations explicitly for model calibration or uncertainty quantification?", "answer": ["A Simple Baseline for Bayesian Uncertainty in Deep Learning"], "answer_arxiv_id": ["1902.02476"], "source_meta": {"published_time": "20220924"}, "qid": "AutoScholarQuery_dev_725"} +{"question": "What recent works have focused on the intersection of FL and DG?", "answer": ["FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Continuous Frequency Space", "Federated Domain Generalization for Image Recognition via Cross-Client Style Transfer"], "answer_arxiv_id": ["2103.06030", "2210.00912"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_dev_726"} +{"question": "Which study first proposed A-MTRL?", "answer": ["Active Multi-Task Representation Learning"], "answer_arxiv_id": ["2202.00911"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_dev_727"} +{"question": "Which studies focused on dynamic regret for non-stationary tabular MDPs?", "answer": ["Reinforcement Learning for Non-Stationary Markov Decision Processes: The Blessing of (More) Optimism", "Dynamic Regret of Policy Optimization in Non-stationary Environments"], "answer_arxiv_id": ["2006.14389", "2007.00148"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_dev_728"} +{"question": "Which research shows that heuristic classification improves downstream few-shot performance for GLaM?", "answer": ["GLaM: Efficient Scaling of Language Models with Mixture-of-Experts"], "answer_arxiv_id": ["2112.06905"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_dev_729"} +{"question": "What are the studies that are based on anonymous temporal random walks for temporal graph learning?", "answer": ["Inductive Representation Learning in Temporal Networks via Causal Anonymous Walks"], "answer_arxiv_id": ["2101.05974"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_dev_730"} +{"question": "What study showcased improvement in results by jointly generating the 3D conformation and the connectivity graph of molecules?", "answer": ["MiDi: Mixed Graph and 3D Denoising Diffusion for Molecule Generation"], "answer_arxiv_id": ["2302.09048"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_731"} +{"question": "In which study did the experiments use results or hyperparameters from the original papers, while affording extra computation to tune the RNNs on each benchmark?", "answer": ["Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPs"], "answer_arxiv_id": ["2110.05038"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_732"} +{"question": "Can you name the works that introduced neural models in the development of code search models?", "answer": ["Multimodal Representation for Neural Code Search", "When Deep Learning Met Code Search", "CodeSearchNet Challenge: Evaluating the State of Semantic Code Search"], "answer_arxiv_id": ["2107.00992", "1905.03813", "1909.09436"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_dev_733"} +{"question": "Which research work focused on convergence analysis using variants of PMD methods for the linear MDP setting?", "answer": ["Cautiously Optimistic Policy Optimization and Exploration with Linear Function Approximation", "Actor-critic is implicitly biased towards high entropy optimal policies"], "answer_arxiv_id": ["2103.12923", "2110.11280"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_734"} +{"question": "Which studies use integer quantisation for accelerated 8-bit inference?", "answer": ["Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference"], "answer_arxiv_id": ["1712.05877"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_dev_735"} +{"question": "Which works have conducted studies on video action detection with datasets that only include single-person videos?", "answer": ["UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild"], "answer_arxiv_id": ["1212.0402"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_dev_736"} +{"question": "What are some recent variants that have improved on optimization time in novel-view synthesis?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2112.05131", "2103.14024", "2201.05989", "2308.04079"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_dev_737"} +{"question": "What papers study the Constrained MDP (CMDP) framework in the field of safe reinforcement learning?", "answer": ["Exploration-Exploitation in Constrained MDPs", "Safe Reinforcement Learning via Curriculum Induction", "Constrained Upper Confidence Reinforcement Learning", "A Sample-Efficient Algorithm for Episodic Finite-Horizon MDP with Constraints", "Learning Policies with Zero or Bounded Constraint Violation for Constrained MDPs", "Provably Efficient Model-Free Constrained RL with Linear Function Approximation"], "answer_arxiv_id": ["2003.02189", "2006.12136", "2001.09377", "2009.11348", "2106.02684", "2206.11889"], "source_meta": {"published_time": "20220628"}, "qid": "AutoScholarQuery_dev_738"} +{"question": "What work generates 3D models based on text prompts by optimizing the CLIP distance between the CLIP text embedding and NeRF renderings?", "answer": ["Zero-Shot Text-Guided Object Generation with Dream Fields"], "answer_arxiv_id": ["2112.01455"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_dev_739"} +{"question": "Which papers propose neural volumetric relighting approaches in face relighting?", "answer": ["Pixel Codec Avatars", "Self-supervised Learning of Detailed 3D Face Reconstruction", "Face Relighting with Geometrically Consistent Shadows"], "answer_arxiv_id": ["2104.04638", "1910.11791", "2203.16681"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_dev_740"} +{"question": "What studies are about unsupervised domain adaptation?", "answer": ["A review of domain adaptation without target labels"], "answer_arxiv_id": ["1901.05335"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_741"} +{"question": "What papers cover autoregressive models where bonds are added using separate algorithms after all atoms are generated?", "answer": ["Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules"], "answer_arxiv_id": ["1906.00957"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_dev_742"} +{"question": "Can you give me some works that utilized model mixtures for personalized Federated Learning?", "answer": ["Personalized Federated Learning with First Order Model Optimization", "Ditto: Fair and Robust Federated Learning Through Personalization"], "answer_arxiv_id": ["2012.08565", "2012.04221"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_dev_743"} +{"question": "What papers discussed approaches for personalized image synthesis?", "answer": ["Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "Unified Multi-Modal Latent Diffusion for Joint Subject and Text\n Conditional Image Generation", "Subject-driven Text-to-Image Generation via Apprenticeship Learning"], "answer_arxiv_id": ["2008.00951", "2208.01618", "2208.12242", "2212.04488", "2302.13848", "2303.09319", "2304.00186"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_dev_744"} +{"question": "Can you cite the research that showed the use of ControlNet in fine-tuning image diffusion models based on various secondary inputs?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "Sketch-Guided Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2211.13752"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_dev_745"} +{"question": "Could you provide me some works that have discussed uncertainty quantification in molecule property prediction?", "answer": ["Uncertainty Quantification using Neural Networks for Molecular Property Prediction"], "answer_arxiv_id": ["2005.10036"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_dev_746"} +{"question": "What studies have extended the zero-shot learning capability of CLIP to semantic segmentation?", "answer": ["DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation"], "answer_arxiv_id": ["2112.01518", "2212.03588"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_747"} +{"question": "What references detail the creation of the Large Language Models (LLMs)?", "answer": ["Language Models are Few-Shot Learners", "Galactica: A Large Language Model for Science", "LLaMA: Open and Efficient Foundation Language Models", "GLM-130B: An Open Bilingual Pre-trained Model"], "answer_arxiv_id": ["2005.14165", "2211.09085", "2302.13971", "2210.02414"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_dev_748"} +{"question": "What works involve training of DPMs with advanced sampling strategies to improve inference speed and reduce training costs?", "answer": ["Denoising Diffusion Implicit Models", "On Fast Sampling of Diffusion Probabilistic Models", "Noise Estimation for Generative Diffusion Models"], "answer_arxiv_id": ["2010.02502", "2106.00132", "2104.02600"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_dev_749"} +{"question": "Which research defined a faithful explanation as one that accurately represents the true reasoning process behind the model’s prediction?", "answer": ["Towards Faithfully Interpretable NLP Systems: How should we define and\n evaluate faithfulness?"], "answer_arxiv_id": ["2004.03685"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_750"} +{"question": "Can you give an example of a work that evaluates LLMs' reasoning ability using text generation instead of multiple choices?", "answer": ["Measuring Mathematical Problem Solving With the MATH Dataset"], "answer_arxiv_id": ["2103.03874"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_751"} +{"question": "Could you provide me some works that apply graph-based models for pose estimation?", "answer": ["Learning Skeletal Graph Neural Networks for Hard 3D Pose Estimation", "GLA-GCN: Global-local Adaptive Graph Convolutional Network for 3D Human\n Pose Estimation from Monocular Video", "Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action\n Recognition"], "answer_arxiv_id": ["2108.07181", "2307.05853", "1801.07455"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_752"} +{"question": "What studies provide general overviews on using computational models from NLP and machine learning to measure readability?", "answer": ["Trends, Limitations and Open Challenges in Automatic Readability\n Assessment Research"], "answer_arxiv_id": ["2105.00973"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_753"} +{"question": "In which works generating sparse graphs resulted in more computationally tractable solutions?", "answer": ["Neural Relational Inference for Interacting Systems", "Measuring and Relieving the Over-smoothing Problem for Graph Neural Networks from the Topological View"], "answer_arxiv_id": ["1802.04687", "1909.03211"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_dev_754"} +{"question": "Which studies adapt dynamic skeletal graphs with action-specific edges for pose estimation?", "answer": ["GLA-GCN: Global-local Adaptive Graph Convolutional Network for 3D Human\n Pose Estimation from Monocular Video", "Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action\n Recognition"], "answer_arxiv_id": ["2307.05853", "1801.07455"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_755"} +{"question": "What works covered the application of pseudo labeling?", "answer": ["DeepPanoContext: Panoramic 3D Scene Understanding with Holistic Scene\n Context Graph and Relation-based Optimization", "Domain Adaptive Semantic Segmentation with Self-Supervised Depth\n Estimation", "Curriculum Domain Adaptation for Semantic Segmentation of Urban Scenes"], "answer_arxiv_id": ["2108.10743", "2104.13613", "1707.09465"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_dev_756"} +{"question": "What works focus on improving privacy cost by assuming sub-Gaussian beauty of the underlying distribution in private mean estimation?", "answer": ["The Cost of Privacy: Optimal Rates of Convergence for Parameter Estimation with Differential Privacy", "Privately Learning High-Dimensional Distributions", "Average-Case Averages: Private Algorithms for Smooth Sensitivity and Mean Estimation", "CoinPress: Practical Private Mean and Covariance Estimation"], "answer_arxiv_id": ["1902.04495", "1805.00216", "1906.02830", "2006.06618"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_dev_757"} +{"question": "What works have been done to improve the optimization-based pipeline in novel view synthesis with sparse input views?", "answer": ["RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs", "SPARF: Neural Radiance Fields from Sparse and Noisy Poses", "DiffusioNeRF: Regularizing Neural Radiance Fields with Denoising\n Diffusion Models"], "answer_arxiv_id": ["2112.00724", "2211.11738", "2302.12231"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_dev_758"} +{"question": "Which works discuss the zero-violation approaches, where methods are initialized in the feasible region, and only updated in ways that are guaranteed not to leave the feasible region?", "answer": ["Learning Policies with Zero or Bounded Constraint Violation for Constrained MDPs"], "answer_arxiv_id": ["2106.02684"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_dev_759"} +{"question": "Can you provide me with some studies about the generation of robust counterfactuals?", "answer": ["Towards Robust and Reliable Algorithmic Recourse", "Consistent Counterfactuals for Deep Models", "Robust Counterfactual Explanations for Tree-Based Ensembles", "Formalising the Robustness of Counterfactual Explanations for Neural Networks"], "answer_arxiv_id": ["2102.13620", "2110.03109", "2207.02739", "2208.14878"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_dev_760"} +{"question": "What studies are focused on the KD strategy, aimed at training a smaller student model with the guidance of a teacher model?", "answer": ["Distilling the Knowledge in a Neural Network", "f-Divergence Minimization for Sequence-Level Knowledge Distillation", "MiniLLM: Knowledge Distillation of Large Language Models", "PANDA: Prompt Transfer Meets Knowledge Distillation for Efficient Model\n Adaptation", "Unified Instance and Knowledge Alignment Pretraining for Aspect-based\n Sentiment Analysis"], "answer_arxiv_id": ["1503.02531", "2307.15190v1", "2306.08543", "2208.10160", "2110.13398"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_761"} +{"question": "What research added supervision during training to bypass the impossibility of learning disentangled representations from independent and identically distributed (i.i.d) data?", "answer": ["Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations"], "answer_arxiv_id": ["1811.12359"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_762"} +{"question": "Are there any papers exploring disentanglement in diffusion models?", "answer": ["Diffusion Models already have a Semantic Latent Space", "Diffusion Autoencoders: Toward a Meaningful and Decodable Representation", "Uncovering the Disentanglement Capability in Text-to-Image Diffusion\n Models", "StyleDrop: Text-to-Image Generation in Any Style", "StyleAdapter: A Single-Pass LoRA-Free Model for Stylized Image\n Generation", "StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion\n Models"], "answer_arxiv_id": ["2210.10960", "2111.15640", "2212.08698", "2306.00983", "2309.01770", "2308.07863"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_763"} +{"question": "Which works introduced multi-task representation learning techniques in natural language domain?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_dev_764"} +{"question": "Which study was dedicated Multi-LLM Debate, which uses multiple LLM instances to propose and debate responses?", "answer": ["Improving Factuality and Reasoning in Language Models through Multiagent\n Debate"], "answer_arxiv_id": ["2305.14325"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_dev_765"} +{"question": "Any works are benchmarked on datasets with limited number of object instances for 6D Pose Estimation?", "answer": ["PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes"], "answer_arxiv_id": ["1711.00199"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_766"} +{"question": "Could you provide me studies about RNN-based approaches for short-term future prediction in 3D human pose forecasting?", "answer": ["Recurrent Network Models for Human Dynamics", "Structural-RNN: Deep Learning on Spatio-Temporal Graphs", "On human motion prediction using recurrent neural networks", "Structured Prediction Helps 3D Human Motion Modelling", "Action-Agnostic Human Pose Forecasting", "Imitation Learning for Human Pose Prediction", "A Neural Temporal Model for Human Motion Prediction", "QuaterNet: A Quaternion-based Recurrent Model for Human Motion"], "answer_arxiv_id": ["1508.00271", "1511.05298", "1705.02445", "1910.09070", "1810.09676", "1909.03449", "1809.03036", "1805.06485"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_dev_767"} +{"question": "What work specified the target styles in the instructions as constraints to improve controlled text generation?", "answer": ["Controlled Text Generation with Natural Language Instructions"], "answer_arxiv_id": ["2304.14293"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_768"} +{"question": "Could you give me studies that focus on two-stage models for diffusion models?", "answer": ["Image Super-Resolution via Iterative Refinement", "High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "Exploiting Diffusion Prior for Real-World Image Super-Resolution"], "answer_arxiv_id": ["2104.07636", "2112.10752", "2307.01952", "2305.07015"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_dev_769"} +{"question": "Which papers explore the utility of tool-calling in benchmarks for machine translation?", "answer": ["Estimating post-editing effort: a study on human judgements, task-based\n and reference-based metrics of MT quality", "MLQA: Evaluating Cross-lingual Extractive Question Answering"], "answer_arxiv_id": ["1910.06204", "1910.07475"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_dev_770"} +{"question": "Which studies can be mentioned as significant advancements in text-to-image models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_dev_771"} +{"question": "What papers are about translating natural language utterances into query language through semantic parsing?", "answer": ["Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs", "Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task", "Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning"], "answer_arxiv_id": ["2305.03111", "1809.08887", "1709.00103"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_dev_772"} +{"question": "Who first observed grokking for algorithmic datasets?", "answer": ["Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets"], "answer_arxiv_id": ["2201.02177"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_dev_773"} +{"question": "Can you name the studies propagating the use of 3D human models for sign language understanding and production tasks?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "Human Part-wise 3D Motion Context Learning for Sign Language Recognition"], "answer_arxiv_id": ["1904.05866", "2308.09305"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_dev_774"} +{"question": "What research involve the use of training trajectories in minimizing surrogate models during dataset distillation?", "answer": ["Dataset Distillation by Matching Training Trajectories", "DC-BENCH: Dataset Condensation Benchmark", "Minimizing the Accumulated Trajectory Error to Improve Dataset\n Distillation", "Scaling Up Dataset Distillation to ImageNet-1K with Constant Memory", "Dataset Distillation: A Comprehensive Review", "Towards Lossless Dataset Distillation via Difficulty-Aligned Trajectory\n Matching"], "answer_arxiv_id": ["2203.11932", "2207.09639", "2211.11004", "2211.10586", "2301.07014", "2310.05773"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_775"} +{"question": "Who employed the concept of acoustic tokens preserving all information of the audio with RVQ for raw audio reconstruction?", "answer": ["SoundStream: An End-to-End Neural Audio Codec", "High Fidelity Neural Audio Compression"], "answer_arxiv_id": ["2107.03312", "2210.13438"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_dev_776"} +{"question": "Could you provide me some works particularly related to quantifying reproducibility in Machine Learning?", "answer": ["A Systematic Review of Reproducibility Research in Natural Language Processing", "Are GANs Created Equal? A Large-Scale Study", "Deep Reinforcement Learning that Matters"], "answer_arxiv_id": ["2103.07929", "1711.10337", "1709.06560"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_dev_777"} +{"question": "Which researches utilize a symmetry deformation module to learn the reconstruction and compute dense correspondence in the context of learning on point clouds?", "answer": ["CorrNet3D: Unsupervised End-to-end Learning of Dense Correspondence for\n 3D Point Clouds"], "answer_arxiv_id": ["2012.15638"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_778"} +{"question": "Are there any papers that stated that the performance of in-context learning is sensitive to the input of pre-training language models?", "answer": ["Commonsense Knowledge Mining from Pretrained Models", "How Can We Know What Language Models Know?"], "answer_arxiv_id": ["1909.00505", "1911.12543"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_dev_779"} +{"question": "What studies explored the use of hypernetworks in meta-RL?", "answer": ["Hypernetworks in Meta-Reinforcement Learning", "HyperDynamics: Meta-Learning Object and Agent Dynamics with Hypernetworks", "Linear Representation Meta-Reinforcement Learning for Instant Adaptation", "Recomposing the Reinforcement Learning Building Blocks with Hypernetworks"], "answer_arxiv_id": ["2210.11348", "2103.09439v1", "2101.04750", "2106.06842"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_780"} +{"question": "Any papers introduced approaches that combine learning with classical graph planning for generalization across various planning domains?", "answer": ["Schema Networks: Zero-shot Transfer with a Generative Causal Model of Intuitive Physics", "Action Schema Networks: Generalised Policies with Deep Learning", "Generalized Planning With Deep Reinforcement Learning"], "answer_arxiv_id": ["1706.04317", "1709.04271", "2005.02305"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_dev_781"} +{"question": "Which works proposed to extend Large Language Models to other modalities, such as audio?", "answer": ["Listen, Think, and Understand", "SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal\n Conversational Abilities"], "answer_arxiv_id": ["2305.10790", "2305.11000"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_782"} +{"question": "What study proposed the refinement of avatar generation through coarse-to-fine and multi-box training for higher-quality avatar?", "answer": ["AvatarCraft: Transforming Text into Neural Human Avatars with Parameterized Shape and Pose Control"], "answer_arxiv_id": ["2303.17606"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_dev_783"} +{"question": "What studies propose to train such model on large-scale image-text data, enabling it to complete various instructions about images?", "answer": ["Improved Baselines with Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2310.03744", "2304.10592", "2304.15010", "2304.14178"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_784"} +{"question": "Which papers argue for a soft-constraint approach to Constrained Reinforcement Learning (CRL)?", "answer": ["Balancing Constraints and Rewards with Meta-Gradient D4PG", "An empirical investigation of the challenges of real-world reinforcement learning"], "answer_arxiv_id": ["2010.06324", "2003.11881"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_dev_785"} +{"question": "Which papers proposed random sketching operators used in solving overdetermined least squares problems?", "answer": ["Sketching as a Tool for Numerical Linear Algebra"], "answer_arxiv_id": ["1411.4357"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_dev_786"} +{"question": "Which research leveraged the memory of user’s feedback to generate prompt for LLMs?", "answer": ["Memory-assisted prompt editing to improve GPT-3 after deployment"], "answer_arxiv_id": ["2201.06009"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_787"} +{"question": "Which papers extended BERT paradigm to areas like point cloud, audio, and video perception?", "answer": ["Masked Autoencoders for Point Cloud Self-supervised Learning", "AST: Audio Spectrogram Transformer", "VideoMAE: Masked Autoencoders are Data-Efficient Learners for\n Self-Supervised Video Pre-Training", "Meta-Transformer: A Unified Framework for Multimodal Learning"], "answer_arxiv_id": ["2203.06604", "2104.01778", "2203.12602", "2307.10802"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_dev_788"} +{"question": "What works proposed regularization methods refining parameters or features without explicitly maintaining source knowledge?", "answer": ["Low-shot Visual Recognition by Shrinking and Hallucinating Features", "Regularizing CNN Transfer Learning with Randomised Regression"], "answer_arxiv_id": ["1606.02819", "1908.05997"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_dev_789"} +{"question": "What works tried to amalgamate the search query and all the candidates together as input for retrieval tasks?", "answer": ["Large Language Models are Effective Text Rankers with Pairwise Ranking\n Prompting"], "answer_arxiv_id": ["2306.17563"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_dev_790"} +{"question": "Any studies about the difficulty in tuning the rate of the second timescale in GTD-style approaches?", "answer": ["Bayesian Bellman Operators"], "answer_arxiv_id": ["2106.05012"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_791"} +{"question": "Could you tell me the papers which demonstrated that heuristically filtered data improves T5?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1910.10683"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_dev_792"} +{"question": "What research work introduced the PMD algorithm that was strictly limited to the tabular setting?", "answer": ["On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2201.07443"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_793"} +{"question": "Which work introduced the concept of oracle complexity separation?", "answer": ["Oracle Complexity Separation in Convex Optimization"], "answer_arxiv_id": ["2002.02706"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_dev_794"} +{"question": "Could you provide the studies about empirical development of federated learning in large-scale deep learning?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "Parallel training of DNNs with Natural Gradient and Parameter Averaging", "Experiments on Parallel Training of Deep Neural Network using Model Averaging"], "answer_arxiv_id": ["1602.05629", "1410.7455", "1507.01239"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_dev_795"} +{"question": "Could you provide me some works that studied the design of U-Nets and their connection to wavelets?", "answer": ["Multi-level Wavelet-CNN for Image Restoration"], "answer_arxiv_id": ["1805.07071"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_796"} +{"question": "What studies achieved local alignment by exploiting the fine-grained relation between visual objects and textual words?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "UNIMO: Towards Unified-Modal Understanding and Generation via\n Cross-Modal Contrastive Learning", "Product1M: Towards Weakly Supervised Instance-Level Product Retrieval\n via Cross-modal Pretraining", "ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision", "FILIP: Fine-grained Interactive Language-Image Pre-Training", "Improving Joint Learning of Chest X-Ray and Radiology Report by Word\n Region Alignment", "Multi-Granularity Cross-modal Alignment for Generalized Medical Visual\n Representation Learning"], "answer_arxiv_id": ["1909.11740", "2004.06165", "2012.15409", "2107.14572", "2102.03334", "2111.07783", "2109.01949", "2210.06044"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_dev_797"} +{"question": "Could you provide me some studies where DPMs were used for text-to-image synthesis?", "answer": ["DiVAE : Photorealistic Images Synthesis with Denoising Diffusion Decoder", "Text-Guided Synthesis of Artistic Images with Retrieval-Augmented Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2206.00386", "2207.13038", "2205.11487"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_dev_798"} +{"question": "Which papers discuss the improvements in performance across a wide range of NLP tasks due to the Transformer architecture?", "answer": ["Attention Is All You Need", "Training language models to follow instructions with human feedback", "Constitutional AI: Harmlessness from AI Feedback", "PaLM: Scaling Language Modeling with Pathways", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "LLaMA: Open and Efficient Foundation Language Models", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer"], "answer_arxiv_id": ["1706.03762", "2203.02155", "2212.08073", "2204.02311", "2307.09288", "2302.13971", "1910.10683"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_799"} +{"question": "Which works discusses about exploiting additional structures found in large classes of non-monotone VIPs?", "answer": ["The Complexity of Constrained Min-Max Optimization", "Fast Extra Gradient Methods for Smooth Structured Nonconvex-Nonconcave Minimax Problems"], "answer_arxiv_id": ["2009.09623", "2106.02326"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_800"} +{"question": "Which works are mentioned in the context of benchmarks for assessing multimodal capabilities?", "answer": ["Inter-GPS: Interpretable Geometry Problem Solving with Formal Language\n and Symbolic Reasoning", "GeoQA: A Geometric Question Answering Benchmark Towards Multimodal\n Numerical Reasoning", "UniGeo: Unifying Geometry Logical Reasoning via Reformulating\n Mathematical Expression", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science\n Question Answering", "MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning\n Benchmark for Expert AGI", "CMMMU: A Chinese Massive Multi-discipline Multimodal Understanding\n Benchmark", "CMMU: A Benchmark for Chinese Multi-modal Multi-type Question\n Understanding and Reasoning", "MathVista: Evaluating Mathematical Reasoning of Foundation Models in\n Visual Contexts"], "answer_arxiv_id": ["2105.04165", "2105.14517", "2212.02746", "2209.09513", "2311.16502", "2401.11944", "2401.14011", "2310.02255"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_dev_801"} +{"question": "Can you provide studies about perturbation methods like RISE and SHAP?", "answer": ["RISE: Randomized Input Sampling for Explanation of Black-box Models", "A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1806.07421", "1705.07874"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_802"} +{"question": "What papers provide evidence of semantic information encoded into the activations of a neural network?", "answer": ["Implicit Representations of Meaning in Neural Language Models", "Simpler Context-Dependent Logical Forms via Model Projections"], "answer_arxiv_id": ["2106.00737", "1606.05378"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_dev_803"} +{"question": "Which works applied alignment approaches to capture the domain invariant characteristics of images?", "answer": ["CyCADA: Cycle-Consistent Adversarial Domain Adaptation", "Bidirectional Learning for Domain Adaptation of Semantic Segmentation", "Image to Image Translation for Domain Adaptation"], "answer_arxiv_id": ["1711.03213", "1904.10620", "1712.00479"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_dev_804"} +{"question": "Which papers address the use of generative models, such as normalizing flows, Variational Autoencoders (VAE), and diffusion models, for VR HMD?", "answer": ["FLAG: Flow-based 3D Avatar Generation from Sparse Observations", "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking\n Inputs with Diffusion Model"], "answer_arxiv_id": ["2203.05789", "2304.08577"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_805"} +{"question": "Who projected the first method for distinguishing the neurons’ ability based on the neuron’s activation value?", "answer": ["Finding Skill Neurons in Pre-trained Transformer-based Language Models"], "answer_arxiv_id": ["2211.07349"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_dev_806"} +{"question": "Which papers have studied algorithmic fairness, specifically with regards to improving model performance at a group level?", "answer": ["When Personalization Harms: Reconsidering the Use of Group Attributes in Prediction"], "answer_arxiv_id": ["2206.02058"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_dev_807"} +{"question": "What are the studies that used external knowledge for reference retrieval to detect hallucinations?", "answer": ["Generate rather than Retrieve: Large Language Models are Strong Context\n Generators", "Investigating the Factual Knowledge Boundary of Large Language Models\n with Retrieval Augmentation"], "answer_arxiv_id": ["2209.10063", "2307.11019"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_dev_808"} +{"question": "Which studies focused on detecting contamination in models without accessing their training data, primarily using output probabilities?", "answer": ["Detecting Pretraining Data from Large Language Models", "Proving Test Set Contamination in Black Box Language Models", "Investigating the Impact of Data Contamination of Large Language Models\n in Text-to-SQL Translation"], "answer_arxiv_id": ["2310.16789", "2310.17623", "2402.08100"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_809"} +{"question": "What works have made advancements in using deep neural networks for real-space electronic systems?", "answer": ["Solving Many-Electron Schrödinger Equation Using Deep Neural Networks", "Deep neural network solution of the electronic Schrödinger equation"], "answer_arxiv_id": ["1807.07014", "1909.08423"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_dev_810"} +{"question": "What are the studies that used various motion representations in prediction tasks?", "answer": ["Deja Vu: Motion Prediction in Static Images", "Im2Flow: Motion Hallucination from Static Images for Action Recognition", "Generating Videos with Scene Dynamics", "An Uncertain Future: Forecasting from Static Images using Variational\n Autoencoders", "Visual Dynamics: Stochastic Future Generation via Layered Cross\n Convolutional Networks"], "answer_arxiv_id": ["1803.06951", "1712.04109", "1609.02612", "1606.07873", "1807.09245"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_dev_811"} +{"question": "Which papers propose solutions to multi-dataset object detection challenges?", "answer": ["Object Detection with a Unified Label Space from Multiple Datasets", "Cross-dataset Training for Class Increasing Object Detection"], "answer_arxiv_id": ["2008.06614", "2001.04621"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_812"} +{"question": "What works focus on animating input source image through explicit or implicit image-based rendering according to motion derived from external sources?", "answer": ["First Order Motion Model for Image Animation", "Motion Representations for Articulated Animation", "Animating Arbitrary Objects via Deep Motion Transfer", "Latent Image Animator: Learning to Animate Images via Latent Space\n Navigation", "DreamPose: Fashion Image-to-Video Synthesis via Stable Diffusion"], "answer_arxiv_id": ["2003.00196", "2104.11280", "1812.08861", "2203.09043", "2304.06025"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_dev_813"} +{"question": "Could you provide me some studies about efficient exploration strategies in RL that are based on starting exploration from specific states?", "answer": ["Learning Montezuma’s Revenge from a Single Demonstration", "Playing Atari Games with Deep Reinforcement Learning and Human Checkpoint Replay", "DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills", "Overcoming Exploration in Reinforcement Learning with Demonstrations", "Reverse Curriculum Generation for Reinforcement Learning", "Solving the Rubik’s Cube Without Human Knowledge", "BaRC: Backward Reachability Curriculum for Robotic Reinforcement Learning", "Backplay: ‘Man muss immer umkehren’", "First return, then explore"], "answer_arxiv_id": ["1812.03381", "1607.05077", "1804.02717", "1709.10089", "1707.05300", "1805.07470", "1806.06161", "1807.06919", "2004.12919v6"], "source_meta": {"published_time": "20220405"}, "qid": "AutoScholarQuery_dev_814"} +{"question": "Which studies explored methods to combine multiple concepts or attributes in personalization?", "answer": ["SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "Multi-Concept Customization of Text-to-Image Diffusion", "Break-A-Scene: Extracting Multiple Concepts from a Single Image", "StyleDrop: Text-to-Image Generation in Any Style", "ProSpect: Prompt Spectrum for Attribute-Aware Personalization of\n Diffusion Models", "Concept Decomposition for Visual Exploration and Inspiration"], "answer_arxiv_id": ["2303.11305", "2212.04488", "2305.16311", "2306.00983", "2305.16225", "2305.18203"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_815"} +{"question": "Which paper demonstrates mBERT’s ability to learn multilingual representations, enabling cross-lingual transfer for languages with different scripts?", "answer": ["How multilingual is Multilingual BERT?"], "answer_arxiv_id": ["1906.01502"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_dev_816"} +{"question": "Could you mention some works that focused on temporal or geographical contexts that can change the answer to the same question? ", "answer": ["SituatedQA: Incorporating Extra-Linguistic Contexts into QA", "StreamingQA: A Benchmark for Adaptation to New Knowledge over Time in Question Answering Models"], "answer_arxiv_id": ["2109.06157", "2205.11388"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_dev_817"} +{"question": "Could you name some video datasets that provide 2D keypoints with temporal information?", "answer": ["PoseTrack: A Benchmark for Human Pose Estimation and Tracking"], "answer_arxiv_id": ["1710.10000"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_dev_818"} +{"question": "Which works are related to the study of partial client participation in the context of PFL?", "answer": ["On the Convergence of FedAvg on Non-IID Data", "Achieving ​ Linear ​ Speedup ​ with ​​ Partial ​​ Worker ​​ Participation ​​ in ​​ Non-IID ​​ Federated ​​ Learning", "A Unified Analysis of Federated Learning with Arbitrary Client Participation"], "answer_arxiv_id": ["1907.02189", "2101.11203", "2205.13648"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_819"} +{"question": "Which works focus on the development of multi-modal visual foundation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Florence: A New Foundation Model for Computer Vision", "Image as a Foreign Language: BEiT Pretraining for All Vision and\n Vision-Language Tasks", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.11432", "2208.10442", "2204.14198"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_820"} +{"question": "What papers have considered a similar optimization approach as used in TTOpt?", "answer": ["Tensor Train for Global Optimization Problems in Robotics"], "answer_arxiv_id": ["2206.05077"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_dev_821"} +{"question": "Any studies focusing on attribute-aware personalization?", "answer": ["ProSpect: Prompt Spectrum for Attribute-Aware Personalization of\n Diffusion Models"], "answer_arxiv_id": ["2305.16225"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_822"} +{"question": "Could you provide me any research about AUM which identifies data by computing the Area Under the Margin?", "answer": ["Identifying Mislabeled Data using the Area Under the Margin Ranking"], "answer_arxiv_id": ["2001.10528"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_dev_823"} +{"question": "Which research work highlighted the integration of elements such as emotion, semantics in ESC systems?", "answer": ["TransESC: Smoothing Emotional Support Conversation via Turn-Level State\n Transition"], "answer_arxiv_id": ["2305.03296"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_dev_824"} +{"question": "What studies incorporated adversarial training in self-supervised settings?", "answer": ["Robust Pre-Training by Adversarial Contrastive Learning", "Adversarial Self-Supervised Contrastive Learning", "Contrastive Learning with Adversarial Examples"], "answer_arxiv_id": ["2010.13337", "2006.07589", "2010.12050"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_dev_825"} +{"question": "Which works considered the impact of attacks on multi-modal auto-driving systems?", "answer": ["Fusion is Not Enough: Single Modal Attacks on Fusion Models for 3D\n Object Detection"], "answer_arxiv_id": ["2304.14614"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_826"} +{"question": "Could you list the works that performed self-adversarial training by dual-BN technique or used pretrained models to generate pseudo-labels?", "answer": ["Robust Pre-Training by Adversarial Contrastive Learning", "Adversarial Examples Improve Image Recognition", "When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning?", "Decoupled Adversarial Contrastive Learning for Self-supervised Adversarial Robustness"], "answer_arxiv_id": ["2010.13337", "1911.09665", "2111.01124", "2207.10899"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_dev_827"} +{"question": "What works have proposed decoupling the environmental exploration from navigation to the target?", "answer": ["Zero-shot Active Visual Search (ZAVIS): Intelligent Object Search for Robotic Assistants"], "answer_arxiv_id": ["2209.08803"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_828"} +{"question": "Which papers applied open-loop imitation learning to predict driving behaviors of other road users in autonomous driving?", "answer": ["ChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing the Worst", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction", "Multiple Futures Prediction", "Learning Lane Graph Representations for Motion Forecasting", "Scene Transformer: A unified architecture for predicting multiple agent trajectories", "DenseTNT: End-to-end Trajectory Prediction from Dense Goal Sets", "Wayformer: Motion Forecasting via Simple & Efficient Attention Networks"], "answer_arxiv_id": ["1812.03079", "1910.05449", "1911.00997", "2007.13732", "2106.08417", "2108.09640", "2207.05844"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_dev_829"} +{"question": "Which work has proposed replacing ODE solvers in the forward with a Taylor-Lagrange expansion to improve training and prediction times?", "answer": ["Taylor-Lagrange Neural Ordinary Differential Equations: Toward Fast Training and Evaluation of Neural ODEs"], "answer_arxiv_id": ["2201.05715"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_dev_830"} +{"question": "What works discuss the use of white-box knowledge distillation in language learning models?", "answer": ["Distilling the Knowledge in a Neural Network", "A Survey on Model Compression for Large Language Models"], "answer_arxiv_id": ["1503.02531", "2308.07633"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_831"} +{"question": "Which studies proposed activation functions that encourage unit-variance activations and gradients?", "answer": ["Self-Normalizing Neural Networks"], "answer_arxiv_id": ["1706.02515"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_dev_832"} +{"question": "What are the works or models related to synthetic face generation using 2D-based models?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Training Generative Adversarial Networks with Limited Data", "Alias-Free Generative Adversarial Networks"], "answer_arxiv_id": ["1812.04948", "2006.06676", "2106.12423"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_dev_833"} +{"question": "Any works that use Vision Transformers (ViT) in place of the traditional convolutional backbone?", "answer": ["Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers", "Segmenter: Transformer for Semantic Segmentation"], "answer_arxiv_id": ["2012.15840", "2105.05633"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_834"} +{"question": "What research in Natural Language Processing focuses on tasks like common-sense reasoning and sentiment analysis for Uncertainty Estimation?", "answer": ["Uncertainty Quantification with Pre-trained Language Models: A\n Large-Scale Empirical Analysis"], "answer_arxiv_id": ["2210.04714"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_835"} +{"question": "Which papers presented the concept of Counterfactual examples (CFEs) as minimal modifications required to elicit a different outcome?", "answer": ["Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR", "The Hidden Assumptions Behind Counterfactual Explanations and Principal\n Reasons"], "answer_arxiv_id": ["1711.00399v3", "1912.04930"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_dev_836"} +{"question": "What is the reference for studies that highlighted the overdependence of existing multimodal pretraining methods on well-aligned multimodal sample pairs/tuples?", "answer": ["Multimodal Learning with Transformers: A Survey"], "answer_arxiv_id": ["2206.06488"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_dev_837"} +{"question": "Which works try to determine the explicit adversarial distributions for adversarial attack?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Adversarial Distributional Training for Robust Deep Learning", "NATTACK: Learning the Distributions of Adversarial Examples for an\n Improved Black-Box Attack on Deep Neural Networks"], "answer_arxiv_id": ["1706.06083", "2002.05999", "1905.00441"], "source_meta": {"published_time": "20220924"}, "qid": "AutoScholarQuery_dev_838"} +{"question": "Could you give me examples of research that have investigated the memorization behavior of pre-trained LMs?", "answer": ["Extracting Training Data from Large Language Models", "Deduplicating Training Data Mitigates Privacy Risks in Language Models", "Deduplicating Training Data Makes Language Models Better", "Quantifying Memorization Across Neural Language Models"], "answer_arxiv_id": ["2012.07805", "2202.06539", "2107.06499", "2202.07646"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_839"} +{"question": "Could you provide me some works focused on simplifying models by mapping between SEM of different levels of abstractions?", "answer": ["Abstracting Causal Models"], "answer_arxiv_id": ["1812.03789"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_dev_840"} +{"question": "What were the initial works that discussed Generative Adversarial Networks (GANs)?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["2203.00667"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_dev_841"} +{"question": "What research papers have been reporting on demographically imbalanced datasets, particularly lacking representation of females and individuals with darker skin tones?", "answer": ["Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the ImageNet Hierarchy", "Understanding and Evaluating Racial Biases in Image Captioning", "Auditing ImageNet: Towards a Model-driven Framework for Annotating Demographic Attributes of Large-Scale Image Datasets", "Data and its (dis)contents: A survey of dataset development and use in machine learning research"], "answer_arxiv_id": ["1912.07726", "2106.08503", "1905.01347", "2012.05345"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_dev_842"} +{"question": "Which studies talked about distilling reasoning capabilities from a stronger teacher to a weaker student in student-teacher frameworks?", "answer": ["Teaching Small Language Models to Reason", "Specializing Smaller Language Models towards Multi-Step Reasoning", "Large Language Models Are Reasoning Teachers", "Orca: Progressive Learning from Complex Explanation Traces of GPT-4"], "answer_arxiv_id": ["2212.08410", "2301.12726", "2212.10071", "2306.02707v1"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_dev_843"} +{"question": "Can you cite a study that applied QAT-based KD for LLMs and considered the risk of overfitting?", "answer": ["Token-Scaled Logit Distillation for Ternary Weight Generative Language\n Models"], "answer_arxiv_id": ["2308.06744"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_dev_844"} +{"question": "What works leverage consistency-based or verification-based approach to improve the reasoning capacity of LLMs?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Explanation-aware Soft Ensemble Empowers Large Language Model In-context\n Learning", "Deductive Verification of Chain-of-Thought Reasoning", "Self-Refine: Iterative Refinement with Self-Feedback"], "answer_arxiv_id": ["2203.11171", "2311.07099", "2306.03872", "2303.17651"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_845"} +{"question": "Which research introduced MMDBs, which allow seamless querying of text and tables using SQL?", "answer": ["Towards Multi-Modal DBMSs for Seamless Querying of Texts and Tables"], "answer_arxiv_id": ["2304.13559"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_dev_846"} +{"question": "What studies propose methods to achieve face capture in the multi-view setup using novel neural BRDF?", "answer": ["NeuFace: Realistic 3D Neural Face Rendering from Multi-view Images"], "answer_arxiv_id": ["2303.14092"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_847"} +{"question": "Which studies focus on the evaluation of the reasoning ability of large language models (LLMs)?", "answer": ["Training Verifiers to Solve Math Word Problems", "Measuring Massive Multitask Language Understanding", "Measuring Mathematical Problem Solving With the MATH Dataset", "IfQA: A Dataset for Open-domain Question Answering under Counterfactual\n Presuppositions", "Reasoning or Reciting? Exploring the Capabilities and Limitations of\n Language Models Through Counterfactual Tasks", "Counterfactual VQA: A Cause-Effect Look at Language Bias", "Don't Just Assume; Look and Answer: Overcoming Priors for Visual\n Question Answering"], "answer_arxiv_id": ["2110.14168", "2009.03300", "2103.03874", "2305.14010", "2307.02477", "2006.04315v4", "1712.00377"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_848"} +{"question": "What are the papers on the proposal-based methods used for 3D instance segmentation?", "answer": ["GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in\n Point Cloud", "Learning Object Bounding Boxes for 3D Instance Segmentation on Point\n Clouds", "3D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentation", "Learning Gaussian Instance Segmentation in Point Clouds"], "answer_arxiv_id": ["1812.03320", "1906.01140", "2003.13867", "2007.09860"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_dev_849"} +{"question": "Which papers have demonstrated the disentanglement achieved by pre-trained GANs?", "answer": ["Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Alias-Free Generative Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["1809.11096", "2106.12423", "1812.04948", "1912.04958"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_850"} +{"question": "What studies proposed to adopt convolutional layers with masked kernels to accelerate the computation of Jacobian determinants?", "answer": ["Emerging Convolutions for Generative Normalizing Flows", "MaCow: Masked Convolutional Generative Flow"], "answer_arxiv_id": ["1901.11137", "1902.04208"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_851"} +{"question": "Can you point out the study wherein a generative data augmentation method was devised to simultaneously train GANs and classifiers?", "answer": ["A Bayesian Data Augmentation Approach for Learning Deep Models"], "answer_arxiv_id": ["1710.10564"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_dev_852"} +{"question": "Which works advocate CNNs as multi-scale architectures?", "answer": ["Multiscale Deep Equilibrium Models"], "answer_arxiv_id": ["2006.08656"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_dev_853"} +{"question": "What studies focused on imitation learning?", "answer": ["Generative Adversarial Imitation Learning", "A Divergence Minimization Perspective on Imitation Learning Methods", "Imitation Learning as f-Divergence Minimization"], "answer_arxiv_id": ["1606.03476", "1911.02256", "1905.12888"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_dev_854"} +{"question": "Which studies proposed novel-view synthesis approaches aggregating image features from aligned pixels without 3D representations?", "answer": ["Generalizable Patch-Based Neural Rendering", "Is Attention All That NeRF Needs?"], "answer_arxiv_id": ["2207.10662", "2207.13298"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_dev_855"} +{"question": "What work proposed utilizing a separate model sequentially for each channel via UAE in reconstruction-based techniques?", "answer": ["An Evaluation of Anomaly Detection and Diagnosis in Multivariate Time Series"], "answer_arxiv_id": ["2109.11428"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_dev_856"} +{"question": "Which research study introduced neural ODEs in CT model estimation?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_dev_857"} +{"question": "What research has been done that integrates the MVS approach and two-view stereo matching for geometry estimation?", "answer": ["GeoNeRF: Generalizing NeRF with Geometry Priors"], "answer_arxiv_id": ["2111.13539"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_dev_858"} +{"question": "What are the commonly used graph augmentation methods based on random modification of graph structures or features?", "answer": ["Inductive Representation Learning on Large Graphs", "GraphCrop: Subgraph Cropping for Graph Classification", "Graph Contrastive Learning with Augmentations", "DropEdge: Towards Deep Graph Convolutional Networks on Node Classification", "An Empirical Study of Graph Contrastive Learning"], "answer_arxiv_id": ["1706.02216", "2009.10564", "2010.13902", "1907.10903", "2109.01116"], "source_meta": {"published_time": "20220226"}, "qid": "AutoScholarQuery_dev_859"} +{"question": "Which works proposed faster sampling strategies for diffusion models?", "answer": ["Denoising Diffusion Implicit Models", "On Fast Sampling of Diffusion Probabilistic Models"], "answer_arxiv_id": ["2010.02502", "2106.00132"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_dev_860"} +{"question": "Which works used autoencoder or density-based models in AD?", "answer": ["Unsupervised Detection of Lesions in Brain MRI using constrained adversarial auto-encoders", "Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery"], "answer_arxiv_id": ["1806.04972", "1703.05921"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_861"} +{"question": "Which studies have directly used LLMs such as GPT-3 for computer vision tasks?", "answer": ["Visual Programming: Compositional visual reasoning without training", "ViperGPT: Visual Inference via Python Execution for Reasoning"], "answer_arxiv_id": ["2211.11559", "2303.08128"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_dev_862"} +{"question": "Can you point to the study that used TTOpt for optimizing the weights of neural network in reinforcement learning problems?", "answer": ["TTOpt: A Maximum Volume Quantized Tensor Train-based Optimization and its Application to Reinforcement Learning"], "answer_arxiv_id": ["2205.00293"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_dev_863"} +{"question": "Which works used variance reduction methods such as SARAH, SPIDER and STORM to improve computational complexity?", "answer": ["SARAH: A Novel Method for Machine Learning Problems Using Stochastic Recursive Gradient", "Spider: Near-Optimal Non-Convex Optimization via Stochastic Path Integrated Differential Estimator", "Momentum-Based Variance Reduction in Non-Convex SGD", "A Stochastic Composite Gradient Method with Incremental Variance Reduction", "Solving Stochastic Compositional Optimization is Nearly as Easy as Solving Stochastic Optimization", "An Online Method for A Class of Distributionally Robust Optimization with Non-Convex Objectives"], "answer_arxiv_id": ["1703.00102", "1807.01695", "1905.10018", "1906.10186", "2008.10847", "2006.10138"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_864"} +{"question": "Which studies have explored in-context learning where an LLM improves at a given task after being provided with task-relevant demonstrations?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_865"} +{"question": "Which works discussed the use of image augmentations to improve robustness of representations in RL?", "answer": ["Reinforcement Learning with Augmented Data", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Generalization in Reinforcement Learning by Soft Data Augmentation", "Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation"], "answer_arxiv_id": ["2004.14990", "2004.13649", "2011.13389", "2107.00644v2"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_866"} +{"question": "What papers offered proof for the lower bounds of SGD-RR?", "answer": ["How Good is SGD with Random Shuffling?", "Random Shuffling Beats SGD Only After Many Epochs on Ill-Conditioned Problems", "Closing the convergence gap of SGD without replacement", "Tighter Lower Bounds for Shuffling SGD: Random Permutations and Beyond"], "answer_arxiv_id": ["1908.00045", "2106.06880", "2002.10400", "2303.07160"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_867"} +{"question": "What works focus on the theoretical aspects of provable P-MTRL?", "answer": ["On the Theory of Transfer Learning: The Importance of Task Diversity", "Provable Meta-Learning of Linear Representations", "Few-Shot Learning via Learning the Representation, Provably", "Sample Efficient Linear Meta-Learning by Alternating Minimization", "Exploiting Shared Representations for Personalized Federated Learning", "Representation Learning Beyond Linear Prediction Functions"], "answer_arxiv_id": ["2006.11650", "2002.11684", "2002.09434", "2105.08306", "2102.07078", "2105.14989"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_dev_868"} +{"question": "What are the studies that employed alignment training strategy and directly optimized LLMs to produce factual statements?", "answer": ["Self-Instruct: Aligning Language Models with Self-Generated Instructions", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2212.10560", "2203.02155"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_dev_869"} +{"question": "Could you provide me some examples of works constructing an atlas of maps from ℝ2 to ℝn?", "answer": ["Neural Surface Maps", "Neural Jacobian Fields: Learning Intrinsic Mappings of Arbitrary Meshes"], "answer_arxiv_id": ["2103.16942", "2205.02904"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_dev_870"} +{"question": "What are some of the works that explored the use of generative models in developing novel molecules with variational autoencoder and reinforcement learning?", "answer": ["Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules", "Junction Tree Variational Autoencoder for Molecular Graph Generation", "Reinforced Genetic Algorithm for Structure-based Drug Design", "Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation", "Molecular De-Novo Design through Deep Reinforcement Learning"], "answer_arxiv_id": ["1610.02415", "1802.04364", "2211.16508", "1806.02473", "1704.07555"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_dev_871"} +{"question": "Which papers does the approach Viewset Diffusion come from?", "answer": ["Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data"], "answer_arxiv_id": ["2306.07881"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_dev_872"} +{"question": "What paper has proposed the idea of using environment interactions to learn the dynamics model?", "answer": ["Recurrent World Models Facilitate Policy Evolution"], "answer_arxiv_id": ["1809.01999"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_dev_873"} +{"question": "Which research proposes Differential Privacy (DP) and DPSGD algorithm for private stochastic gradient descent?", "answer": ["Deep Learning with Differential Privacy"], "answer_arxiv_id": ["1607.00133"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_874"} +{"question": "In what studies did researchers match the lower bounds of SGD-RR with its upper bounds for both strongly convex and general convex cases?", "answer": ["Tighter Lower Bounds for Shuffling SGD: Random Permutations and Beyond", "Random Reshuffling: Simple Analysis with Vast Improvements"], "answer_arxiv_id": ["2303.07160", "2006.05988"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_875"} +{"question": "What work proposed forgetting score as a difficulty-based metric for sample importance in CoreSet selection?", "answer": ["An Empirical Study of Example Forgetting during Deep Neural Network\n Learning"], "answer_arxiv_id": ["1812.05159"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_876"} +{"question": "Could you provide me some works about data-driven methods in VR HMD settings?", "answer": ["FLAG: Flow-based 3D Avatar Generation from Sparse Observations", "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking\n Inputs with Diffusion Model", "HMD-NeMo: Online 3D Avatar Motion Generation From Sparse Observations", "AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion\n Sensing", "Realistic Full-Body Tracking from Sparse Observations via Joint-Level\n Modeling"], "answer_arxiv_id": ["2203.05789", "2304.08577", "2308.11261", "2207.13784", "2308.08855"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_877"} +{"question": "Can you name the study that reported high success rates on few-shot learning tasks using a text-to-image generative model pre-trained on extensive external datasets?", "answer": ["Is synthetic data from generative models ready for image recognition?"], "answer_arxiv_id": ["2210.07574"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_dev_878"} +{"question": "Could you list the studies that proved the upper bounds of SGD-RR?", "answer": ["Why Random Reshuffling Beats Stochastic Gradient Descent", "Random Shuffling Beats SGD after Finite Epochs", "SGD without Replacement: Sharper Rates for General Smooth Convex Functions", "SGD with shuffling: optimal rates without component convexity and large epoch requirements", "Random Reshuffling: Simple Analysis with Vast Improvements"], "answer_arxiv_id": ["1510.08560", "1806.10077", "1903.01463", "2006.06946", "2006.05988"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_dev_879"} +{"question": "Which papers discuss empirical defenses that are used in defending against adversarial attacks?", "answer": ["Adversarial Training for Free!", "Fast is better than free: Revisiting adversarial training", "Towards Deep Learning Models Resistant to Adversarial Attacks", "On Single Source Robustness in Deep Fusion Models", "Defending Multimodal Fusion Models against Single-Source Adversaries", "Can audio-visual integration strengthen robustness under multimodal\n attacks?"], "answer_arxiv_id": ["1904.12843", "2001.03994", "1706.06083", "1906.04691", "2206.12714", "2104.02000"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_880"} +{"question": "What works propose regularization of the embedding space for regression?", "answer": ["Learning Probabilistic Ordinal Embeddings for Uncertainty-Aware Regression", "Improving Deep Regression with Ordinal Entropy", "Delving into Deep Imbalanced Regression", "RankSim: Ranking Similarity Regularization for Deep Imbalanced Regression"], "answer_arxiv_id": ["2103.13629", "2301.08915", "2102.09554", "2205.15236"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_dev_881"} +{"question": "Could you provide me some works about sequence-based generative models for protein sequence design?", "answer": ["Tranception: protein fitness prediction with autoregressive transformers and inference-time retrieval"], "answer_arxiv_id": ["2205.13760"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_dev_882"} +{"question": "Could you list some research works that studied Unsupervised Domain Adaptation (UDA) in the context of mitigating domain shift for semantic segmentation?", "answer": ["Category Anchor-Guided Unsupervised Domain Adaptation for Semantic\n Segmentation", "Domain Adaptation for Semantic Segmentation with Maximum Squares Loss", "DAFormer: Improving Network Architectures and Training Strategies for\n Domain-Adaptive Semantic Segmentation", "DADA: Depth-aware Domain Adaptation in Semantic Segmentation", "Unsupervised Model Adaptation for Continual Semantic Segmentation", "ADVENT: Adversarial Entropy Minimization for Domain Adaptation in\n Semantic Segmentation", "Unsupervised Intra-domain Adaptation for Semantic Segmentation through\n Self-Supervision", "Self-supervised Augmentation Consistency for Adapting Semantic\n Segmentation", "Uncertainty-Aware Deep Co-training for Semi-supervised Medical Image\n Segmentation", "A Good Student is Cooperative and Reliable: CNN-Transformer\n Collaborative Learning for Semantic Segmentation", "Transformer-CNN Cohort: Semi-supervised Semantic Segmentation by the\n Best of Both Students", "Distilling Efficient Vision Transformers from CNNs for Semantic\n Segmentation", "CLIP Is Also a Good Teacher: A New Learning Framework for Inductive\n Zero-shot Semantic Segmentation"], "answer_arxiv_id": ["1910.13049", "1909.13589", "2111.14887", "1904.01886", "2009.12518", "1811.12833", "2004.07703", "2105.00097", "2111.11629", "2307.12574", "2209.02178", "2310.07265", "2310.02296"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_dev_883"} +{"question": "In which references was ARA formulated as a ranking problem instead of a classification task?", "answer": ["A Neural Pairwise Ranking Model for Readability Assessment"], "answer_arxiv_id": ["2203.07450"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_dev_884"} +{"question": "Could you provide me some studies about unsupervised concept learning methods in concept-based models?", "answer": ["A Framework for Learning Ante-hoc Explainable Models via Concepts"], "answer_arxiv_id": ["2108.11761"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_dev_885"} +{"question": "What studies evaluated the reasoning ability of LLMs from several perspectives, including mathematical reasoning, common-sense reasoning, logical reasoning, and domain-specific reasoning?", "answer": ["A Survey on Evaluation of Large Language Models", "AGIEval: A Human-Centric Benchmark for Evaluating Foundation Models", "A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on\n Reasoning, Hallucination, and Interactivity", "GLoRE: Evaluating Logical Reasoning of Large Language Models"], "answer_arxiv_id": ["2307.03109", "2304.06364", "2302.04023", "2310.09107"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_dev_886"} +{"question": "Which papers explore data augmentation techniques to improve task performance in low-resource languages?", "answer": ["Neural Machine Translation of Rare Words with Subword Units", "NL-Augmenter: A Framework for Task-Sensitive Natural Language\n Augmentation"], "answer_arxiv_id": ["1508.07909v5", "2112.02721"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_dev_887"} +{"question": "What works exist on using large-scale datasets for vision-language pre-training?", "answer": ["Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts", "Scaling Up Vision-Language Pre-training for Image Captioning", "GIT: A Generative Image-to-text Transformer for Vision and Language", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2102.08981", "2111.12233", "2205.14100", "2102.05918"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_dev_888"} +{"question": "Could you provide examples of participatory data creation initiatives focused specifically on NLP?", "answer": ["Probability distributions generated by fractional diffusion equations"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_dev_889"} +{"question": "Could you provide me some works that attempted to address the problem of experimental design for causal discovery for continuous variables?", "answer": ["ABCD-Strategy: Budgeted Experimental Design for Targeted Causal Structure Discovery", "Optimal experimental design via Bayesian optimization: active causal structure learning for Gaussian process networks", "Active Bayesian Causal Inference", "Cost-Optimal Learning of Causal Graphs", "Active Invariant Causal Prediction: Experiment Selection through Stability", "On the Number of Experiments Sufficient and in the Worst Case Necessary to Identify All Causal Relations Among N Variables", "Experimental Design for Cost-Aware Learning of Causal Graphs", "A Unified Experiment Design Approach for Cyclic and Acyclic Causal Models", "Budgeted Experiment Design for Causal Structure Learning", "Trust Your $\\nabla$: Gradient-based Intervention Targeting for Causal Discovery", "Learning Neural Causal Models with Active Interventions"], "answer_arxiv_id": ["1902.10347", "1910.03962", "2206.02063", "1703.02645", "2006.05690v3", "1207.1389", "1810.11867", "2205.10083", "1709.03625", "2211.13715", "2109.02429"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_dev_890"} +{"question": "What works cover the topic of in-situ adjustment for accuracy-robustness in AI platforms?", "answer": ["Test time Adaptation through Perturbation Robustness", "Evaluating the Adversarial Robustness of Adaptive Test-time Defenses", "Towards Robust Neural Networks via Random Self-ensemble", "Once-for-All Adversarial Training: In-Situ Tradeoff between Robustness and Accuracy for Free"], "answer_arxiv_id": ["2110.10232", "2202.13711", "1712.00673", "2010.11828"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_891"} +{"question": "What research showcases the potential of pre-trained diffusion models in generating diverse multi-view images?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image", "Consistent-1-to-3: Consistent Image to 3D View Synthesis via\n Geometry-aware Diffusion Models"], "answer_arxiv_id": ["2303.11328", "2306.07881", "2309.03453", "2310.03020"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_dev_892"} +{"question": "What studies highlight a connection between BERT model activations and 3D color space?", "answer": ["Can Language Models Encode Perceptual Structure Without Grounding? A Case Study in Color"], "answer_arxiv_id": ["2109.06129"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_dev_893"} +{"question": "Which study showed a relationship between the diversity and fidelity of synthetic samples and test accuracies when applied to generative data augmentation?", "answer": ["Effective Data Augmentation with Multi-Domain Learning GANs"], "answer_arxiv_id": ["1912.11597"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_dev_894"} +{"question": "In which study did the author apply TTOpt to the QUBO problem?", "answer": ["Are Quantum Computers Practical Yet? A Case for Feature Selection in Recommender Systems using Tensor Networks"], "answer_arxiv_id": ["2205.04490"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_dev_895"} +{"question": "Which paper popularized sinusoidal networks for implicit modelling tasks?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_dev_896"} +{"question": "Could you give me an example of a study that adapts a voxelized structure to protein targets?", "answer": ["Generating 3D Molecules Conditional on Receptor Binding Sites with Deep Generative Models"], "answer_arxiv_id": ["2110.15200v2"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_897"} +{"question": "Could you provide me some works regarding GNNs used in accurately reproducing molecular force fields?", "answer": ["SchNet – a deep learning architecture for molecules and materials", "PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges"], "answer_arxiv_id": ["1712.06113", "1902.08408"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_dev_898"} +{"question": "Which papers discuss the integration of image-based vision models with Large Language Models for multimodal understanding?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "VideoChat: Chat-Centric Video Understanding"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2304.10592", "2304.08485", "2305.06500", "2305.06355"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_dev_899"} +{"question": "Which works have explored the application of user instructions to determine sentence relationships and fine-tune clusters?", "answer": ["ClusterLLM: Large Language Models as a Guide for Text Clustering"], "answer_arxiv_id": ["2305.14871"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_dev_900"} +{"question": "What research papers proposed the sampling-free uncertainty estimation methods?", "answer": ["Towards Maximizing the Representation Gap between In-Domain & Out-of-Distribution Examples", "Uncertainty-Aware Deep Classifiers using Generative Models", "Graph Posterior Network: Bayesian Predictive Uncertainty for Node Classification", "Lightweight Probabilistic Deep Networks", "Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness", "On Feature Collapse and Deep Kernel Learning for Single Forward Pass Uncertainty", "Uncertainty Estimation Using a Single Deep Deterministic Neural Network", "Uncertainty on Asynchronous Time Event Prediction", "Accurate Uncertainties for Deep Learning Using Calibrated Regression", "Confidence-Aware Learning for Deep Neural Networks", "Individual Calibration with Randomized Forecasting", "Distribution Calibration for Regression", "Intra Order-Preserving Functions for Calibration of Multi-Class Neural Networks", "Natural-Parameter Networks: A Class of Probabilistic Neural Networks", "Sampling-free Epistemic Uncertainty Estimation Using Approximated Variance Propagation", "Lightweight Probabilistic Deep Networks", "Probabilistic Backpropagation for Scalable Learning of Bayesian Neural Networks", "Graph Mixture Density Networks"], "answer_arxiv_id": ["2010.10474", "2006.04183", "2110.14012", "1805.11327v1", "2006.10108", "2102.11409", "2003.02037", "1911.05503", "1807.00263", "2007.01458", "2006.10288", "1905.06023", "2003.06820", "1611.00448", "1908.00598", "1805.11327v1", "1502.05336", "2012.03085"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_dev_901"} +{"question": "Could you provide me with some studies that employed distributed methods to improve scalability of large-scale Graph Neural Networks?", "answer": ["Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks", "Distributed Graph Neural Network Training: A Survey"], "answer_arxiv_id": ["1905.07953", "2211.00216"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_902"} +{"question": "What papers discus coupling of large language models with image captioning models to enhance image descriptions?", "answer": ["ChatGPT Asks, BLIP-2 Answers: Automatic Questioning Towards Enriched Visual Descriptions"], "answer_arxiv_id": ["2303.06594"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_903"} +{"question": "Could you provide examples of studies where the feed-forward designs are used by learning reliable representations from data?", "answer": ["MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "GeoNeRF: Generalizing NeRF with Geometry Priors", "Neural Rays for Occlusion-aware Image-based Rendering"], "answer_arxiv_id": ["2103.15595", "2111.13539", "2107.13421"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_dev_904"} +{"question": "Which papers focus on improving the computational efficiency of NeRF by using a hybrid representation?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "TensoRF: Tensorial Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "HexPlane: A Fast Representation for Dynamic Scenes"], "answer_arxiv_id": ["2112.07945", "2203.09517", "2201.05989", "2301.10241", "2301.09632"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_dev_905"} +{"question": "Which research proposed an information flow type system that could be used for automatic marginalization of discrete random variables?", "answer": ["Conditional independence by typing"], "answer_arxiv_id": ["2010.11887"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_dev_906"} +{"question": "Could you provide me some works that improved neural scene representations methods by optimizing implementations to take maximum advantage of acceleration hardware with efficient compressed field representations?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient\n Neural Field Rendering on Mobile Architectures", "TensoRF: Tensorial Radiance Fields", "Efficient Geometry-aware 3D Generative Adversarial Networks", "HexPlane: A Fast Representation for Dynamic Scenes", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance"], "answer_arxiv_id": ["2201.05989", "2208.00277", "2203.09517", "2112.07945", "2301.09632", "2301.10241"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_dev_907"} +{"question": "Which work is most closely related to the current research and provides an alternative proof of the equivalence between TV indistinguishability, replicability, and differential privacy?", "answer": ["Stability is Stable: Connections between Replicability, Privacy, and Adaptive Generalization"], "answer_arxiv_id": ["2303.12921"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_908"} +{"question": "What projects used this strategy to the high-order method?", "answer": ["Adaptive Third-Order Methods for Composite Convex Optimization", "High-order methods beyond the classical complexity bounds, II: inexact high-order proximal-point methods with segment search"], "answer_arxiv_id": ["2202.12730", "2109.12303"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_dev_909"} +{"question": "Could you provide me some studies that apply slot attention to visual question answering?", "answer": ["Language-Mediated, Object-Centric Representation Learning"], "answer_arxiv_id": ["2012.15814"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_dev_910"} +{"question": "What works explored the memory footprint and access pattern of voxel representations compared to point cloud representations?", "answer": ["Point-Voxel CNN for Efficient 3D Deep Learning"], "answer_arxiv_id": ["1907.03739"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_911"} +{"question": "Which work devises a pipeline for generating synthetic data for evaluating the compositional reasoning ability of VQA models?", "answer": ["CLEVR: A Diagnostic Dataset for Compositional Language and Elementary\n Visual Reasoning"], "answer_arxiv_id": ["1612.06890"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_dev_912"} +{"question": "Where can I find details about the ChatGLM architecture?", "answer": ["GLM-130B: An Open Bilingual Pre-trained Model"], "answer_arxiv_id": ["2210.02414"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_dev_913"} +{"question": "What studies provide bounds for linear TD in both the i. i. d. data setting and a correlated data setting?", "answer": ["A Finite Time Analysis of Temporal Difference Learning With Linear Function Approximation"], "answer_arxiv_id": ["1806.02450"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_914"} +{"question": "What papers focused on addressing extant limitations within LLMs evaluators such as factuality, interpretability and position bias?", "answer": ["FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long\n Form Text Generation", "Large Language Models are not Fair Evaluators"], "answer_arxiv_id": ["2305.14251", "2305.17926"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_dev_915"} +{"question": "Which works are related to motion generation based on 3D face coefficients?", "answer": ["Perceptual Conversational Head Generation with Regularized Driver and\n Enhanced Renderer", "Learning to Listen: Modeling Non-Deterministic Dyadic Facial Motion", "MFR-Net: Multi-faceted Responsive Listening Head Generation via\n Denoising Diffusion Model"], "answer_arxiv_id": ["2206.12837", "2204.08451", "2308.16635"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_dev_916"} +{"question": "What study showed RNNs to be a competitive baseline in meta-RL?", "answer": ["Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPs"], "answer_arxiv_id": ["2110.05038"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_917"} +{"question": "Can you name the studies which provide solutions to the high cost associated with neural network-based wave function models?", "answer": ["Solving the electronic Schrödinger equation for multiple nuclear geometries with weight-sharing deep neural networks", "Ab-Initio Potential Energy Surfaces by Pairing GNNs with Neural Wave Functions", "Sampling-free Inference for Ab-Initio Potential Energy Surface Networks"], "answer_arxiv_id": ["2105.08351", "2110.05064", "2205.14962"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_dev_918"} +{"question": "Which study incorporated self-attention and Gumbel subset sampling in point-based methodologies to enhance recognition tasks in Point Cloud processing?", "answer": ["Modeling Point Clouds with Self-Attention and Gumbel Subset Sampling"], "answer_arxiv_id": ["1904.03375"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_919"} +{"question": "Which works focus on denoising diffusion probabilistic models?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2105.05233", "2112.10752"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_dev_920"} +{"question": "Could you provide me some studies about unstructured pruning in deep learning?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks"], "answer_arxiv_id": ["1803.03635"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_dev_921"} +{"question": "Any works about leveraging a multi-task framework to generate NLEs and labels simultaneously?", "answer": ["Leakage-Adjusted Simulatability: Can Models Generate Non-Trivial\n Explanations of Their Behavior in Natural Language?"], "answer_arxiv_id": ["2010.04119"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_922"} +{"question": "Are there any studies that optimize the process of dataset condensation?", "answer": ["Optimizing Millions of Hyperparameters by Implicit Differentiation", "Dataset Meta-Learning from Kernel Ridge-Regression", "Dataset Distillation with Infinitely Wide Convolutional Networks", "On Implicit Bias in Overparameterized Bilevel Optimization", "Dataset Distillation using Neural Feature Regression", "Efficient Dataset Distillation using Random Feature Approximation", "Accelerating Dataset Distillation via Model Augmentation", "Scaling Up Dataset Distillation to ImageNet-1K with Constant Memory", "Dataset Distillation with Convexified Implicit Gradients"], "answer_arxiv_id": ["1911.02590", "2011.00050", "2107.13034", "2212.14032", "2206.00719", "2210.12067", "2212.06152", "2211.10586", "2302.06755v2"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_dev_923"} +{"question": "Which works are examples of pretrained Large Language Models?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models", "OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["2005.14165", "2302.13971", "2205.01068"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_dev_924"} +{"question": "Which studies trained across multiple vision tasks and solved them simultaneously?", "answer": ["Multi-Task Self-Training for Learning General Representations", "INTERN: A New Learning Paradigm Towards General Vision"], "answer_arxiv_id": ["2108.11353", "2111.08687"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_dev_925"} +{"question": "Could you mention some studies that focused on generating adversarial examples designed to induce LLMs to generate harmful or non-factual content?", "answer": ["Universal and Transferable Adversarial Attacks on Aligned Language\n Models", "In ChatGPT We Trust? Measuring and Characterizing the Reliability of\n ChatGPT"], "answer_arxiv_id": ["2307.15043", "2304.08979"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_dev_926"} +{"question": "Which studies have been conducted on diffusion models performing conditional image generation with guiding input channels?", "answer": ["Zero-Shot Text-to-Image Generation", "Image Super-Resolution via Iterative Refinement", "Palette: Image-to-Image Diffusion Models", "Pretraining is All You Need for Image-to-Image Translation", "Diffusion Autoencoders: Toward a Meaningful and Decodable Representation", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2102.12092", "2104.07636", "2111.05826", "2205.12952", "2111.15640", "2207.12598"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_dev_927"} +{"question": "Could you provide me a work that extends the decomposition of 3D feature volumes to 4D dynamic scenes?", "answer": ["HexPlane: A Fast Representation for Dynamic Scenes"], "answer_arxiv_id": ["2301.09632"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_dev_928"} +{"question": "What is a study that proposes diverse ensembles as a measure of sample importance?", "answer": ["Trivial or impossible -- dichotomous data difficulty masks model\n differences (on ImageNet and beyond)"], "answer_arxiv_id": ["2110.05922"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_929"} +{"question": "What work focused on learning a weighting over the tasks in A-MTRL?", "answer": ["Weighted Training for Cross-Task Learning"], "answer_arxiv_id": ["2105.14095"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_dev_930"} +{"question": "Which pieces of research output a dense representation of the object’s coordinate space?", "answer": ["Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimation", "DPOD: 6D Pose Object Detector and Refiner"], "answer_arxiv_id": ["1901.02970", "1902.11020"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_931"} +{"question": "Which work investigated the impact of weight oscillation on quantization-aware training, and rooted it in depth-wise convolution (DW-Conv) and batch normalization (BN) layers?", "answer": ["Overcoming Oscillations in Quantization-Aware Training"], "answer_arxiv_id": ["2203.11086"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_dev_932"} +{"question": "Any works about TracIn being a promising approach for training data attribution?", "answer": ["Estimating Training Data Influence by Tracing Gradient Descent"], "answer_arxiv_id": ["2002.08484"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_dev_933"} +{"question": "Could you provide me some works that allow flexible learning and combination of attributes from different concepts?", "answer": ["ProSpect: Prompt Spectrum for Attribute-Aware Personalization of\n Diffusion Models"], "answer_arxiv_id": ["2305.16225"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_dev_934"} +{"question": "Could you provide the research that proposed an open-book QA model conditions answer generation upon newly-retrieved documents?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2005.11401"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_dev_935"} +{"question": "Which papers discussed the use of audio as auxiliary inputs in image editing?", "answer": ["SadTalker: Learning Realistic 3D Motion Coefficients for Stylized\n Audio-Driven Single Image Talking Face Animation"], "answer_arxiv_id": ["2211.12194"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_dev_936"} +{"question": "What work presents the graphlet kernel and similar fixed pattern-based approaches which only count subgraphs up to size around 5?", "answer": ["Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting"], "answer_arxiv_id": ["2006.09252"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_dev_937"} +{"question": "What studies detailed the success of Mixture of Experts in the fields of computer vision and natural language processing?", "answer": ["Deep Mixture of Experts via Shallow Embedding", "Scaling Vision with Sparse Mixture of Experts", "Cross-token Modeling with Conditional Computation", "Outrageously Large Neural Networks: The Sparsely-Gated\n Mixture-of-Experts Layer", "Beyond Distillation: Task-level Mixture-of-Experts for Efficient\n Inference", "Switch Transformers: Scaling to Trillion Parameter Models with Simple\n and Efficient Sparsity", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts", "ST-MoE: Designing Stable and Transferable Sparse Expert Models"], "answer_arxiv_id": ["1806.01531", "2106.05974", "2109.02008", "1701.06538", "2110.03742", "2101.03961", "2112.06905", "2202.08906"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_dev_938"} +{"question": "Who introduced the Set Abstraction module in point-based methodologies for enhancing Point Cloud processing?", "answer": ["PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space"], "answer_arxiv_id": ["1706.02413"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_dev_939"} +{"question": "What works are about prompting, which refers to providing manual instructions or fine-tuning task-specific tokens for a desired behavior from large language models?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "How Can We Know What Language Models Know?", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts"], "answer_arxiv_id": ["2107.13586v1", "1911.12543", "2010.15980"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_dev_940"} +{"question": "What research papers have been published around privacy leakage in prompt-based learning?", "answer": ["Commonsense Knowledge Mining from Pretrained Models", "How Can We Know What Language Models Know?", "Language Models as Knowledge Bases?"], "answer_arxiv_id": ["1909.00505", "1911.12543", "1909.01066"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_941"} +{"question": "Which studies incurred linear convergence guarantees for NPG with log-linear policies by adding entropy regularization?", "answer": ["Linear Convergence of Entropy-Regularized Natural Policy Gradient with Linear Function Approximation"], "answer_arxiv_id": ["2106.04096v4"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_942"} +{"question": "Which researches have focused on HMT techniques using sparse signals from HMD or wearable IMU sensors?", "answer": ["TransPose: Real-time 3D Human Translation and Pose Estimation with Six\n Inertial Sensors", "Physical Inertial Poser (PIP): Physics-aware Real-time Human Motion\n Tracking from Sparse Inertial Sensors", "AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion\n Sensing", "FLAG: Flow-based 3D Avatar Generation from Sparse Observations", "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking\n Inputs with Diffusion Model"], "answer_arxiv_id": ["2105.04605", "2203.08528", "2207.13784", "2203.05789", "2304.08577"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_dev_943"} +{"question": "Which papers outline criteria for useful benchmarks for natural language understanding?", "answer": ["What Will it Take to Fix Benchmarking in Natural Language Understanding?"], "answer_arxiv_id": ["2104.02145"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_dev_944"} +{"question": "Could you provide me some works about the progress in graph neural networks that enhance the usage of 2D molecular graph representation?", "answer": ["Constrained Graph Variational Autoencoders for Molecule Design", "GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation", "Junction Tree Variational Autoencoder for Molecular Graph Generation", "Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation", "Optimization of Molecules via Deep Reinforcement Learning"], "answer_arxiv_id": ["1805.09076", "2001.09382", "1802.04364", "1806.02473", "1810.08678"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_dev_945"} +{"question": "Which papers propose models that reason about orientation uncertainty by predicting the parameters of von Mises, Fisher, and Bingham distributions?", "answer": ["Deep Directional Statistics: Pose Estimation with Uncertainty Quantification", "Probabilistic orientation estimation with matrix Fisher distributions", "Deep Bingham Networks: Dealing with Uncertainty and Ambiguity in Pose Estimation"], "answer_arxiv_id": ["1805.03430", "2006.09740v1", "2012.11002v1"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_dev_946"} +{"question": "Which studies contain pre-training of embeddings with privileged information in task-inference methods?", "answer": ["Decoupling Exploration and Exploitation for Meta-Reinforcement Learning without Sacrifices"], "answer_arxiv_id": ["2008.02790"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_dev_947"} +{"question": "What papers talk about style/data-augmentation in DG strategies?", "answer": ["Domain Generalization with MixStyle", "Uncertainty Modeling for Out-of-Distribution Generalization", "Exact Feature Distribution Matching for Arbitrary Style Transfer and Domain Generalization", "Learning to Generate Novel Domains for Domain Generalization"], "answer_arxiv_id": ["2104.02008", "2202.03958", "2203.07740", "2007.03304"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_dev_948"} +{"question": "What research discussed that the regret bounds couldn't improve beyond the guarantees for the original E2D method?", "answer": ["The Statistical Complexity of Interactive Decision Making"], "answer_arxiv_id": ["2112.13487v3"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_dev_949"} +{"question": "Which works explore the establishment of clustering pseudo labels for achieving end-to-end multi-view clustering?", "answer": ["Self-supervised Discriminative Feature Learning for Deep Multi-view\n Clustering"], "answer_arxiv_id": ["2103.15069"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_dev_950"} +{"question": "What studies designed multi-agent reinforcement learning methods to train agents for Nash equilibrium in common-payoff games?", "answer": ["V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL"], "answer_arxiv_id": ["2110.14555"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_dev_951"} +{"question": "Which studies discussed architectures for predicting drug properties from SMILES/SELFIES strings in drug discovery?", "answer": ["Self-Referencing Embedded Strings (SELFIES): A 100% robust molecular string representation"], "answer_arxiv_id": ["1905.13741"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_dev_952"} +{"question": "What work describes Trans-MM, an interpretation method for Transformer-based architectures?", "answer": ["Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformers"], "answer_arxiv_id": ["2103.15679"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_dev_953"} +{"question": "Which works divide the GEC task into detection and correction processes?", "answer": ["Compositional Sequence Labeling Models for Error Detection in Learner\n Writing", "Context is Key: Grammatical Error Detection with Contextual Word\n Representations"], "answer_arxiv_id": ["1607.06153", "1906.06593"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_dev_954"} +{"question": "What works proposed a framework that jointly infers logical forms and direct answers?", "answer": ["DecAF: Joint Decoding of Answers and Logical Forms for Question Answering over Knowledge Bases"], "answer_arxiv_id": ["2210.00063"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_dev_955"} +{"question": "Could you provide me some works that use recurrent neural networks and an attention mechanism for routing information in 'sequential slots'?", "answer": ["Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects", "R-SQAIR:Relational Sequential Attend, Infer, Repeat"], "answer_arxiv_id": ["1603.08575", "1806.01794", "1910.05231"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_956"} +{"question": "Which papers indicated that LLMs have recently advanced the state-of-the-art performance across many NLP tasks?", "answer": ["PaLM 2 Technical Report", "GPT-4 Technical Report", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2305.10403", "2303.08774", "2204.02311", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_957"} +{"question": "What studies are associated with the problem of finetuning the entire model in hierarchical classification, which requires substantial data and is computationally inefficient?", "answer": ["B-CNN: Branch Convolutional Neural Network for Hierarchical\n Classification", "Visual Tree Convolutional Neural Network in Image Classification", "Network of Experts for Large-Scale Image Categorization", "Solving Long-tailed Recognition with Deep Realistic Taxonomic Classifier"], "answer_arxiv_id": ["1709.09890", "1906.01536", "1604.06119", "2007.09898"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_dev_958"} +{"question": "Which work decomposes the problem of task automation into action phrase-extraction and grounding stages?", "answer": ["Mapping Natural Language Instructions to Mobile UI Action Sequences"], "answer_arxiv_id": ["2005.03776"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_dev_959"} +{"question": "What studies introduced templates for multimodal ICL to improve ICL performance?", "answer": ["Beyond Task Performance: Evaluating and Reducing the Flaws of Large\n Multimodal Models with In-Context Learning"], "answer_arxiv_id": ["2310.00647"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_960"} +{"question": "Which papers focused on 'blind source separation' in the context of Independent Component Analysis?", "answer": ["Disentangling Identifiable Features from Noisy Data with Structured Nonlinear ICA", "Nonlinear Independent Component Analysis for Principled Disentanglement in Unsupervised Deep Learning"], "answer_arxiv_id": ["2106.09620", "2303.16535"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_dev_961"} +{"question": "Could you provide me some works about the concept of data pruning in the context of sparse data selection?", "answer": ["An Empirical Study of Example Forgetting during Deep Neural Network Learning", "What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation", "Not All Samples Are Created Equal: Deep Learning with Importance Sampling", "Beyond neural scaling laws: beating power law scaling via data pruning"], "answer_arxiv_id": ["1812.05159", "2008.03703", "1803.00942v3", "2206.14486v6"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_dev_962"} +{"question": "Which works detailed an algorithm for approximate ridge leverage-score sampling for the Khatri-Rao product?", "answer": ["Leverage Score Sampling for Tensor Product Matrices in Input Sparsity Time"], "answer_arxiv_id": ["2202.04515"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_dev_963"} +{"question": "What studies used diffusion models in the context of video?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data"], "answer_arxiv_id": ["2210.02303", "2209.14792"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_dev_964"} +{"question": "What work originally proposed chain-of-thought (CoT) prompting for LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20240628"}, "qid": "AutoScholarQuery_dev_965"} +{"question": "What research works are based on the deep subspace multi-view clustering and deep graph multi-view clustering?", "answer": ["Deep Multimodal Subspace Clustering Networks"], "answer_arxiv_id": ["1804.06498"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_dev_966"} +{"question": "Which paper mentioned that the sum of all leverage scores is the rank of the matrix?", "answer": ["Sketching as a Tool for Numerical Linear Algebra"], "answer_arxiv_id": ["1411.4357"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_dev_967"} +{"question": "What papers discuss the role of textual prompts in delivering contextual information necessary for in-context learning?", "answer": ["Large Language Models are Human-Level Prompt Engineers"], "answer_arxiv_id": ["2211.01910"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_968"} +{"question": "Could you provide me some works about different framings of the sparse coding problem?", "answer": ["Convex Sparse Matrix Factorizations"], "answer_arxiv_id": ["0812.1869"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_dev_969"} +{"question": "Any studies about allocation algorithms for subadditive valuations that guarantee 1/2121/2-envy-freeness ex-ante?", "answer": ["Breaking the Envy Cycle: Best-of-Both-Worlds Guarantees for Subadditive Valuations"], "answer_arxiv_id": ["2304.03706"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_dev_970"} +{"question": "What research papers have developed free-form text conditioned generation for motion?", "answer": ["TEMOS: Generating diverse human motions from textual descriptions", "Synthesis of Compositional Animations from Textual Descriptions", "FLAME: Free-form Language-based Motion Synthesis & Editing"], "answer_arxiv_id": ["2204.14109", "2103.14675", "2209.00349"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_dev_971"} +{"question": "What studies consider multi-task surrogate models in the context of safe bo?", "answer": ["Safe Controller Optimization for Quadrotors with Gaussian Processes", "Bayesian Optimization with Safety Constraints: Safe and Automatic Parameter Tuning in Robotics", "Adaptive and Safe Bayesian Optimization in High Dimensions via One-Dimensional Subspaces", "Stagewise Safe Bayesian Optimization with Gaussian Processes"], "answer_arxiv_id": ["1509.01066", "1602.04450", "1902.03229", "1806.07555"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_972"} +{"question": "Could you provide me some works on SlipCover?", "answer": ["Structure Learning of Probabilistic Logic Programs by Searching the Clause Space"], "answer_arxiv_id": ["1309.2080"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_dev_973"} +{"question": "Which works have incorporated equivariances into neural network models?", "answer": ["On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "Building powerful and equivariant graph neural networks with structural message-passing"], "answer_arxiv_id": ["1802.03690", "2006.15107"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_dev_974"} +{"question": "What studies have utilized image sequences derived from videos for evaluation of MLLMs?", "answer": ["AutoEval-Video: An Automatic Benchmark for Assessing Large Vision\n Language Models in Open-Ended Video Question Answering"], "answer_arxiv_id": ["2311.14906"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_dev_975"} +{"question": "Which studies discussed 3D diffusion models related to point clouds?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "$PC^2$: Projection-Conditioned Point Cloud Diffusion for Single-Image 3D\n Reconstruction", "GECCO: Geometrically-Conditioned Point Diffusion Models"], "answer_arxiv_id": ["2104.03670", "2103.01458", "2302.10668", "2303.05916"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_dev_976"} +{"question": "Are there any research outputs revealing the ambiguous benefits of techniques aiming at enhancing the reasoning capability of LLMs?", "answer": ["Large Language Models Cannot Self-Correct Reasoning Yet"], "answer_arxiv_id": ["2310.01798"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_dev_977"} +{"question": "Which research proposed the Knowledge-Guided Policy Network?", "answer": ["KoGuN: Accelerating Deep Reinforcement Learning via Integrating Human Suboptimal Knowledge"], "answer_arxiv_id": ["2002.07418"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_dev_978"} +{"question": "What work develops the Functional Mirror Ascent Policy Gradient (FMA-PG)?", "answer": ["A general class of surrogate functions for stable and efficient reinforcement learning"], "answer_arxiv_id": ["2108.05828v5"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_979"} +{"question": "Which papers present the use of differentiable material point method simulation in co-optimizing soft robots?", "answer": ["ChainQueen: A Real-Time Differentiable Physical Simulator for Soft Robotics"], "answer_arxiv_id": ["1810.01054"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_dev_980"} +{"question": "What are some examples of studies that introduced Transformer techniques to improve computer vision systems?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "MaxViT: Multi-Axis Vision Transformer", "CrossFormer++: A Versatile Vision Transformer Hinging on Cross-scale Attention", "Glance-and-Gaze Vision Transformer"], "answer_arxiv_id": ["2010.11929", "2204.01697", "2303.06908", "2106.02277"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_dev_981"} +{"question": "Which research papers introduced prompts paradigm to language models?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2107.13586v1"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_dev_982"} +{"question": "What is the initial study that proposed adversarial training as a method for enhancing model robustness?", "answer": ["Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1412.6572"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_dev_983"} +{"question": "Which studies developed 3D pretraining methods for semantic and instance segmentation that use instance discrimination based on different camera views?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts"], "answer_arxiv_id": ["2007.10985", "2012.09165"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_dev_984"} +{"question": "What work provided data for evaluating model behavior in more naturalistic, 3D environments?", "answer": ["Physion: Evaluating Physical Prediction from Vision in Humans and Machines"], "answer_arxiv_id": ["2106.08261"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_dev_985"} +{"question": "Who proposed efficient variants of the SM loss called the SSM and FDSSM objectives?", "answer": ["Sliced Score Matching: A Scalable Approach to Density and Score Estimation", "Efficient Learning of Generative Models via Finite-Difference Score Matching"], "answer_arxiv_id": ["1905.07088", "2007.03317"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_dev_986"} +{"question": "Could you provide me some papers about the work proposing a unified multimodal encoder to align all modalities with language?", "answer": ["X-LLM: Bootstrapping Advanced Large Language Models by Treating\n Multi-Modalities as Foreign Languages", "ChatBridge: Bridging Modalities with Large Language Model as a Language\n Catalyst", "AnyMAL: An Efficient and Scalable Any-Modality Augmented Language Model", "PandaGPT: One Model To Instruction-Follow Them All", "ImageBind-LLM: Multi-modality Instruction Tuning"], "answer_arxiv_id": ["2305.04160", "2305.16103", "2309.16058", "2305.16355", "2309.03905"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_dev_987"} +{"question": "Which works focus on detecting LLM-generated text using classifiers?", "answer": ["ChatLog: Carefully Evaluating the Evolution of ChatGPT Across Time", "How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation,\n and Detection", "DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability\n Curvature"], "answer_arxiv_id": ["2304.14106", "2301.07597", "2301.11305"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_dev_988"} +{"question": "What papers show that LLM hidden states can effectively represent aspects of texts such as space, time, and honesty?", "answer": ["Language Models Represent Space and Time", "Representation Engineering: A Top-Down Approach to AI Transparency"], "answer_arxiv_id": ["2310.02207", "2310.01405"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_dev_989"} +{"question": "What works used the reparameterisation implied by unit scaling?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "Feature Learning in Infinite-Width Neural Networks", "Tensor Programs V: Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfer"], "answer_arxiv_id": ["1806.07572", "2011.14522v3", "2203.03466"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_dev_990"} +{"question": "Which work developed stochastic compositional gradient descent (SCGD) to solve the problem in Stochastic Compositional Optimization?", "answer": ["Stochastic Compositional Gradient Descent: Algorithms for Minimizing Compositions of Expected-Value Functions"], "answer_arxiv_id": ["1411.3803v1"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_dev_991"} +{"question": "What studies propose methods for Partial Model Personalization by training personalized feature extractors?", "answer": ["Exploiting Shared Representations for Personalized Federated Learning"], "answer_arxiv_id": ["2102.07078"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_dev_992"} +{"question": "What research works have taken on the challenge of improving efficiency and scalability of large-scale GNNs through sampling methods?", "answer": ["Inductive Representation Learning on Large Graphs", "FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling", "GraphSAINT: Graph Sampling Based Inductive Learning Method", "Layer-Dependent Importance Sampling for Training Deep and Large Graph Convolutional Networks", "GNNAutoScale: Scalable and Expressive Graph Neural Networks via Historical Embeddings", "GraphFM: Improving Large-Scale GNN Training via Feature Momentum"], "answer_arxiv_id": ["1706.02216", "1801.10247", "1907.04931", "1911.07323", "2106.05609", "2206.07161"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_dev_993"} +{"question": "Which studies discuss about the concept of adversarial robustness for classifiers?", "answer": ["Robustness of classifiers: from adversarial to random noise"], "answer_arxiv_id": ["1608.08967"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_dev_994"} +{"question": "Which work introduced the Fisher-Rao norm as a crucial geometric complexity measure?", "answer": ["Fisher-Rao Metric, Geometry, and Complexity of Neural Networks"], "answer_arxiv_id": ["1711.01530"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_dev_995"} +{"question": "What works have used the concept of dimensionality reduction in diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "simple diffusion: End-to-end diffusion for high resolution images", "Wavelet Score-Based Generative Modeling"], "answer_arxiv_id": ["2112.10752", "2301.11093", "2208.05003"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_dev_996"} +{"question": "What work has been done to establish sublinear convergence rates for two-layer neural PPO and NAC?", "answer": ["Neural Policy Gradient Methods: Global Optimality and Rates of Convergence", "Finite-time analysis of entropy-regularized neural natural actor-critic algorithm"], "answer_arxiv_id": ["1909.01150", "2206.00833"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_dev_997"} +{"question": "What papers study the frameworks for mitigating hallucinations?", "answer": ["Analyzing and Mitigating Object Hallucination in Large Vision-Language\n Models", "Mitigating Fine-Grained Hallucination by Fine-Tuning Large\n Vision-Language Models with Caption Rewrites", "Mitigating Object Hallucinations in Large Vision-Language Models through\n Visual Contrastive Decoding", "Scalable Prompt Generation for Semi-supervised Learning with Language\n Models", "Mitigating Hallucination in Visual Language Models with Visual\n Supervision", "Hallucination Augmented Contrastive Learning for Multimodal Large\n Language Model", "OPERA: Alleviating Hallucination in Multi-Modal Large Language Models\n via Over-Trust Penalty and Retrospection-Allocation", "HalluciDoctor: Mitigating Hallucinatory Toxicity in Visual Instruction\n Data", "Beyond Hallucinations: Enhancing LVLMs through Hallucination-Aware\n Direct Preference Optimization"], "answer_arxiv_id": ["2310.00754", "2312.01701", "2311.16922", "2302.09236", "2311.16479", "2312.06968", "2311.17911", "2311.13614", "2311.16839"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_dev_998"} +{"question": "Which works have utilized image features from frozen vision models for multi-modal language models?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2010.11929", "2103.00020"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_dev_999"} diff --git a/recipe/paper_search/inference/datasets/AutoScholarQuery/test.jsonl b/recipe/paper_search/inference/datasets/AutoScholarQuery/test.jsonl new file mode 100644 index 0000000..c9716e6 --- /dev/null +++ b/recipe/paper_search/inference/datasets/AutoScholarQuery/test.jsonl @@ -0,0 +1,1000 @@ +{"question": "Can you tell me some papers about hybrid architectures in reconstruction-based techniques?", "answer": ["Multivariate Time-series Anomaly Detection via Graph Attention Network"], "answer_arxiv_id": ["2009.02040"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_0"} +{"question": "Are there any studies that analysed the use of target networks for Deep Q-learning?", "answer": ["A Theoretical Analysis of Deep Q-Learning"], "answer_arxiv_id": ["1901.00137"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_test_1"} +{"question": "Any resources providing information about attempts to detect or calibrate biases automatically in peer reviews?", "answer": ["Uncovering Latent Biases in Text: Method and Application to Peer Review", "You Are the Best Reviewer of Your Own Papers: An Owner-Assisted Scoring Mechanism", "Your 2 is My 1, Your 3 is My 9: Handling Arbitrary Miscalibrations in Ratings", "Least Square Calibration for Peer Reviews"], "answer_arxiv_id": ["2010.15300", "2110.14802", "1806.05085", "2110.12607"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_test_2"} +{"question": "What papers are the foundation models for the Natural Language Processing (NLP) field based on?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Language Models are Few-Shot Learners", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["1810.04805", "2005.14165", "1910.10683", "2204.02311", "2302.13971"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_3"} +{"question": "Could you list the papers that explored identifying optimal interventions through sequential experimentation in causal bandits and causal reinforcement learning?", "answer": ["Causal Bandits: Learning Good Interventions via Causal Inference"], "answer_arxiv_id": ["1606.03203"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_4"} +{"question": "Could you provide me some studies that focused on white-box scenarios for cyber-security in machine learning?", "answer": ["Universal Adversarial Triggers for Attacking and Analyzing NLP"], "answer_arxiv_id": ["1908.07125"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_test_5"} +{"question": "Which papers generalize the coordinate definition of the field to cases where the parameters of a viewing ray are used?", "answer": ["Light Field Networks: Neural Scene Representations with\n Single-Evaluation Rendering", "Scene Representation Transformer: Geometry-Free Novel View Synthesis\n Through Set-Latent Scene Representations"], "answer_arxiv_id": ["2106.02634", "2111.13152"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_6"} +{"question": "Which works develop suitable approximations of the predictive distribution or parts of the integral for uncertainties in deep learning?", "answer": ["A Probabilistic U-Net for Segmentation of Ambiguous Images", "Stochastic Segmentation Networks: Modelling Spatially Correlated Aleatoric Uncertainty", "A Hierarchical Probabilistic U-Net for Modeling Multi-Scale Ambiguities", "PHiSeg: Capturing Uncertainty in Medical Image Segmentation"], "answer_arxiv_id": ["1806.05034", "2006.06015", "1905.13077", "1906.04045"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_test_7"} +{"question": "Which studies have proposed using voxel for spatial geometry and texture modeling in 3D scene representation?", "answer": ["3D ShapeNets: A Deep Representation for Volumetric Shapes", "3D-R2N2: A Unified Approach for Single and Multi-view 3D Object\n Reconstruction"], "answer_arxiv_id": ["1406.5670", "1604.00449"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_test_8"} +{"question": "Which studies present issues about the stationary distribution of rewards over contexts?", "answer": ["The K-Nearest Neighbour UCB algorithm for multi-armed bandits with covariates", "Nonparametric Stochastic Contextual Bandits", "Smooth Contextual Bandits: Bridging the Parametric and Non-differentiable Regret Regimes", "Randomized Allocation with Nonparametric Estimation for Contextual Multi-Armed Bandits with Delayed Rewards", "Self-Tuning Bandits over Unknown Covariate-Shifts", "Smoothness-Adaptive Contextual Bandits", "Transfer Learning for Contextual Multi-armed Bandits"], "answer_arxiv_id": ["1803.00316v1", "1801.01750", "1909.02553", "1902.00819", "2007.08584", "1910.09714", "2211.12612"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_test_9"} +{"question": "Which work first implemented token-level edit operation prediction in Seq2Edit methods?", "answer": ["Encode, Tag, Realize: High-Precision Text Editing"], "answer_arxiv_id": ["1909.01187"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_test_10"} +{"question": "Could you provide me a study about generating sign pose sequences from gloss sequences by employing VQ-VAE?", "answer": ["G2P-DDM: Generating Sign Pose Sequence from Gloss Sequence with Discrete\n Diffusion Model"], "answer_arxiv_id": ["2208.09141"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_test_11"} +{"question": "Who proposed source-free universal domain adaptation (SF-UniDA)?", "answer": ["UMAD: Universal Model Adaptation under Domain and Category Shift", "Upcycling Models under Domain and Category Shift"], "answer_arxiv_id": ["2112.08553", "2303.07110"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_12"} +{"question": "What works aim to study the policies or features that remain stable across the different training tasks?", "answer": ["Invariant Policy Optimization: Towards Stronger Generalization in Reinforcement Learning", "Instance-based Generalization in Reinforcement Learning", "Domain Adversarial Reinforcement Learning", "Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneck", "Decoupling Representation Learning from Reinforcement Learning", "Deep Reinforcement and InfoMax Learning"], "answer_arxiv_id": ["2006.01096", "2011.01089", "2102.07097", "1910.12911", "2009.08319", "2006.07217"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_test_13"} +{"question": "Could you provide me some works about fine-tuning LLMs to better response to visual instructions?", "answer": ["mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "Improved Baselines with Visual Instruction Tuning"], "answer_arxiv_id": ["2304.14178", "2310.03744"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_14"} +{"question": "Could you mention some works that classify unsupervised segmentation into two categories: clustering based on invariance and clustering using pre-trained models?", "answer": ["PiCIE: Unsupervised Semantic Segmentation using Invariance and\n Equivariance in Clustering", "Invariant Information Clustering for Unsupervised Image Classification\n and Segmentation", "Unsupervised Semantic Segmentation with Self-supervised Object-centric\n Representations", "ACSeg: Adaptive Conceptualization for Unsupervised Semantic Segmentation", "Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "NamedMask: Distilling Segmenters from Complementary Foundation Models"], "answer_arxiv_id": ["2103.17070", "1807.06653", "2207.05027", "2210.05944", "2203.08414", "2209.11228"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_test_15"} +{"question": "Could you provide me examples of the development of more sophisticated feature extractors that enhance Point Cloud processing?", "answer": ["PointConv: Deep Convolutional Networks on 3D Point Clouds", "Point Transformer", "Rethinking Network Design and Local Geometry in Point Cloud: A Simple\n Residual MLP Framework", "An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["1811.07246", "2012.09164", "2202.07123", "2010.11929"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_16"} +{"question": "What are the papers related to face reenactment, specifically aimed at transferring facial expressions and movements?", "answer": ["Structure-Aware Motion Transfer with Deformable Anchor Model", "Thin-Plate Spline Motion Model for Image Animation", "Latent Image Animator: Learning to Animate Images via Latent Space\n Navigation", "DPE: Disentanglement of Pose and Expression for General Video Portrait\n Editing"], "answer_arxiv_id": ["2204.05018", "2203.14367", "2203.09043", "2301.06281"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_test_17"} +{"question": "What papers propose the use of spatiotemporal transformer for BEV generation?", "answer": ["BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers"], "answer_arxiv_id": ["2203.17270"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_test_18"} +{"question": "Can you name some works that extend Global Descent for deep learning architectures?", "answer": ["A Convergence Theory for Deep Learning via Over-Parameterization", "Gradient Descent Finds Global Minima of Deep Neural Networks", "An Improved Analysis of Training Over-parameterized Deep Neural Networks"], "answer_arxiv_id": ["1811.03962", "1811.03804", "1906.04688"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_test_19"} +{"question": "Could you provide me large multimodal models (LMMs) references?", "answer": ["Visual Instruction Tuning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Qwen Technical Report"], "answer_arxiv_id": ["2304.08485", "2301.12597", "2306.15195", "2304.10592", "2309.16609v1"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_20"} +{"question": "Could you provide me studies about achieving local editing by involving semantic masks as intermediate representations?", "answer": ["FENeRF: Face Editing in Neural Radiance Fields", "IDE-3D: Interactive Disentangled Editing for High-Resolution 3D-aware\n Portrait Synthesis"], "answer_arxiv_id": ["2111.15490", "2205.15517"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_test_21"} +{"question": "Which works propose methods for feature matching by detecting and describing keypoints on images?", "answer": ["SuperPoint: Self-Supervised Interest Point Detection and Description", "D2-Net: A Trainable CNN for Joint Detection and Description of Local Features", "R2D2: Repeatable and Reliable Detector and Descriptor"], "answer_arxiv_id": ["1712.07629", "1905.03561", "1906.06195"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_22"} +{"question": "Which work first demonstrated the possibility of reconstructing accurate 3D full-body motion using only six IMUs?", "answer": ["Sparse Inertial Poser: Automatic 3D Human Pose Estimation from Sparse\n IMUs"], "answer_arxiv_id": ["1703.08014"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_23"} +{"question": "Could you provide me a work that extended the minimax method to deep neural networks?", "answer": ["Stochastic AUC Maximization with Deep Neural Networks"], "answer_arxiv_id": ["1908.10831"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_test_24"} +{"question": "Any works that have commented on the challenge of training the PRM due to expensive human-annotated datasets?", "answer": ["Solving math word problems with process- and outcome-based feedback", "Let's Verify Step by Step"], "answer_arxiv_id": ["2211.14275", "2305.20050"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_25"} +{"question": "Which papers are known for initially representing 3D scenes with a set of 3D Gaussians?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_26"} +{"question": "Which works proposed architectures for group equivariance in image classification?", "answer": ["Group Equivariant Convolutional Networks", "Steerable CNNs", "Group Equivariant Stand-Alone Self-Attention For Vision"], "answer_arxiv_id": ["1602.07576", "1612.08498", "2010.00977v2"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_test_27"} +{"question": "What papers mention the increased computational complexity and decreased utility due to DPSGD?", "answer": ["Deep Learning with Differential Privacy", "Differentially Private Learning Needs Better Features (or Much More Data)"], "answer_arxiv_id": ["1607.00133", "2011.11660"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_28"} +{"question": "In which studies has it been demonstrated that multi-modal models are vulnerable to adversarial attacks?", "answer": ["Fusion is Not Enough: Single Modal Attacks on Fusion Models for 3D\n Object Detection", "Towards Adversarial Attack on Vision-Language Pre-training Models", "Can audio-visual integration strengthen robustness under multimodal\n attacks?", "Fooling Vision and Language Models Despite Localization and Attention\n Mechanism", "Cycle-Consistency for Robust Visual Question Answering", "Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["2304.14614", "2206.09391", "2104.02000", "1709.08693", "1902.05660", "1412.6572"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_29"} +{"question": "Could you provide me studies that expound the impossibility of identifying latent factors for i.i.d. nonlinearly-dependent data without labels or assumptions about the data generating process?", "answer": ["Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations", "Variational Autoencoders and Nonlinear ICA: A Unifying Framework"], "answer_arxiv_id": ["1811.12359", "1907.04809"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_30"} +{"question": "Could you provide me an example where an open-source model was introduced for input-output unsafety detection for LLMs?", "answer": ["Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations"], "answer_arxiv_id": ["2312.06674"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_test_31"} +{"question": "What papers focused on source data estimation or self-training for pinhole images in the context of SFUDA?", "answer": ["Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentation", "Source-Free Open Compound Domain Adaptation in Semantic Segmentation", "Source-Free Domain Adaptation for Semantic Segmentation", "Source-Free Domain Adaptation for Image Segmentation"], "answer_arxiv_id": ["2108.11249", "2106.03422", "2103.16372", "2108.03152"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_test_32"} +{"question": "Can you provide some works about predicting the contact map, the distance map and/or the torsion angles between protein residues?", "answer": ["Accurate De Novo Prediction of Protein Contact Map by Ultra-Deep Learning Model"], "answer_arxiv_id": ["1609.00680v6"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_test_33"} +{"question": "What paper explored the application of VLMs, specifically CLIP, for BEV retrieval tasks?", "answer": ["BEV-TSR: Text-Scene Retrieval in BEV Space for Autonomous Driving"], "answer_arxiv_id": ["2401.01065"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_test_34"} +{"question": "Could you list research that demonstrated the advantages of Quantization-Aware Training (QAT), which can enable the model to learn better representations for low-bit weights?", "answer": ["LLM-QAT: Data-Free Quantization Aware Training for Large Language Models", "OmniQuant: Omnidirectionally Calibrated Quantization for Large Language\n Models", "PB-LLM: Partially Binarized Large Language Models", "BitNet: Scaling 1-bit Transformers for Large Language Models"], "answer_arxiv_id": ["2305.17888v1", "2308.13137", "2310.00034", "2310.11453"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_35"} +{"question": "What are the researches that have explored the application of Crypto-based Private Learning in privacy-preserving machine learning?", "answer": ["Privacy-Preserving Machine Learning with Fully Homomorphic Encryption\n for Deep Neural Network"], "answer_arxiv_id": ["2106.07229"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_36"} +{"question": "Any works that focus on augmenting sparse inputs with synthetically generated views?", "answer": ["Ray Priors through Reprojection: Improving Neural Radiance Fields for\n Novel View Extrapolation", "VM-NeRF: Tackling Sparsity in NeRF with View Morphing", "GeCoNeRF: Few-shot Neural Radiance Fields via Geometric Consistency"], "answer_arxiv_id": ["2205.05922", "2210.04214", "2301.10941"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_37"} +{"question": "Which work introduces Point-E, a language-guided DM?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts"], "answer_arxiv_id": ["2212.08751"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_test_38"} +{"question": "Which papers discuss the practical applicability of black-box and transfer-based threat model, and the related security and safety risks?", "answer": ["Practical Black-Box Attacks against Machine Learning", "Boosting Adversarial Attacks with Momentum"], "answer_arxiv_id": ["1602.02697", "1710.06081"], "source_meta": {"published_time": "20220924"}, "qid": "AutoScholarQuery_test_39"} +{"question": "What studies develop hierarchical models in relation to diffusion models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2205.11487", "2204.06125", "2106.15282"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_test_40"} +{"question": "What are the papers that analyze the limitations of simple random walks on the clique expansion of the hypergraph?", "answer": ["Hyper-SAGNN: a self-attention based graph neural network for hypergraphs", "Neural Predicting Higher-order Patterns in Temporal Networks"], "answer_arxiv_id": ["1911.02613", "2106.06039"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_test_41"} +{"question": "Which study explicitly determines and measures the faithfulness of explanations in LLMs?", "answer": ["Question Decomposition Improves the Faithfulness of Model-Generated\n Reasoning"], "answer_arxiv_id": ["2307.11768"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_42"} +{"question": "Which study argued on the difficulties of implementing a GAN-like procedure using the dual form of UOT?", "answer": ["Robust Optimal Transport with Applications in Generative Modeling and Domain Adaptation"], "answer_arxiv_id": ["2010.05862"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_43"} +{"question": "What studies deal with standard feature selection that selects the same subset of features for each data sample?", "answer": ["Feature Selection: A Data Perspective"], "answer_arxiv_id": ["1601.07996"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_test_44"} +{"question": "What works are related to the use of commonsense knowledge in Knowledge Graphs?", "answer": ["ConceptNet 5.5: An Open Multilingual Graph of General Knowledge"], "answer_arxiv_id": ["1612.03975"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_45"} +{"question": "What works discuss the lack of robustness in NLP benchmarks?", "answer": ["When Benchmarks are Targets: Revealing the Sensitivity of Large Language\n Model Leaderboards"], "answer_arxiv_id": ["2402.01781"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_46"} +{"question": "Which papers examined pretraining on scientific text corpora?", "answer": ["SciBERT: A Pretrained Language Model for Scientific Text", "Domain-Specific Language Model Pretraining for Biomedical Natural\n Language Processing", "BioBERT: a pre-trained biomedical language representation model for biomedical text mining", "ClinicalBERT: Modeling Clinical Notes and Predicting Hospital\n Readmission"], "answer_arxiv_id": ["1903.10676", "2007.15779", "1901.08746v4", "1904.05342"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_test_47"} +{"question": "Which studies apply model-agnostic meta learning (MAML) to deep anomaly detector models?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Few-Shot One-Class Classification via Meta-Learning", "Few-Shot Scene-Adaptive Anomaly Detection", "Few-shot Network Anomaly Detection via Cross-network Meta-learning"], "answer_arxiv_id": ["1703.03400", "2007.04146", "2007.07843", "2102.11165"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_48"} +{"question": "What works have proposed guidelines for documenting ML datasets?", "answer": ["Datasheets for Datasets"], "answer_arxiv_id": ["1803.09010"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_49"} +{"question": "Which papers focused on locally aligning fixed patches with textual words?", "answer": ["ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision", "FILIP: Fine-grained Interactive Language-Image Pre-Training", "Improving Joint Learning of Chest X-Ray and Radiology Report by Word\n Region Alignment", "Multi-Granularity Cross-modal Alignment for Generalized Medical Visual\n Representation Learning"], "answer_arxiv_id": ["2102.03334", "2111.07783", "2109.01949", "2210.06044"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_test_50"} +{"question": "What is the fundamental work on fully convolutional networks (FCNs) used for deep learning-based semantic segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_51"} +{"question": "Which works focused on ray-based rendering for novel view synthesis approach?", "answer": ["Stereo Radiance Fields (SRF): Learning View Synthesis for Sparse Views\n of Novel Scenes", "IBRNet: Learning Multi-View Image-Based Rendering", "Generalizable Patch-Based Neural Rendering", "Is Attention All That NeRF Needs?", "Explicit Correspondence Matching for Generalizable Neural Radiance\n Fields"], "answer_arxiv_id": ["2104.06935", "2102.13090", "2207.10662", "2207.13298", "2304.12294"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_test_52"} +{"question": "Which papers contribute to the advancement of model-based reinforcement learning through the study of the world model?", "answer": ["Recurrent World Models Facilitate Policy Evolution", "Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Learning to Fly via Deep Model-Based Reinforcement Learning", "Mastering Atari with Discrete World Models", "Mastering Diverse Domains through World Models", "Model Based Reinforcement Learning for Atari", "Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model"], "answer_arxiv_id": ["1809.01999", "1811.04551", "1912.01603", "2003.08876", "2010.02193", "2301.04104", "1903.00374", "1911.08265"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_53"} +{"question": "Could you provide me some studies proposing models for learning latent graphs?", "answer": ["Dynamic Graph CNN for Learning on Point Clouds", "Latent-Graph Learning for Disease Prediction", "Differentiable Graph Module (DGM) for Graph Convolutional Networks"], "answer_arxiv_id": ["1801.07829", "2003.13620", "2002.04999"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_test_54"} +{"question": "Which study proposed a method that works only on toy images of up to 333 objects on a black background?", "answer": ["Learning Object-Centric Video Models by Contrasting Sets"], "answer_arxiv_id": ["2011.10287"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_55"} +{"question": "Any work about applying re-reading prompt to improve reasoning tasks of LLM?", "answer": ["Re-Reading Improves Reasoning in Large Language Models"], "answer_arxiv_id": ["2309.06275"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_56"} +{"question": "What studies introduce the unsupervised disentanglement score called Distortion?", "answer": ["Analyzing the Latent Space of GAN through Local Dimension Estimation"], "answer_arxiv_id": ["2205.13182"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_57"} +{"question": "Which research leveraged large language models like GPT-3 to learn a proxy reward function while avoiding the need for many expert demonstrations?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_58"} +{"question": "What papers used a predefined set of names for enhancing cross-style transfer?", "answer": ["Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "Description-Driven Task-Oriented Dialog Modeling", "Larger language models do in-context learning differently"], "answer_arxiv_id": ["2202.12837", "2201.08904", "2303.03846v2"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_59"} +{"question": "Which studies have recently been working on the integration of visual perception and large language models?", "answer": ["Attention Is All You Need", "Language Models are Few-Shot Learners", "GPT-4 Technical Report", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["1706.03762", "2005.14165", "2303.08774", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_60"} +{"question": "What papers introduced the fast gradient sign method (FGSM) and the basic iterative method (BIM) for adversarial attacks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Adversarial examples in the physical world"], "answer_arxiv_id": ["1412.6572", "1607.02533"], "source_meta": {"published_time": "20220924"}, "qid": "AutoScholarQuery_test_61"} +{"question": "Any works talked about the use of meta-gradients to learn a combination of hyperparameters?", "answer": ["Bootstrapped Meta-Learning"], "answer_arxiv_id": ["2109.04504v2"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_62"} +{"question": "Are there any works that improve cost-effectiveness, performance, and data generation quality in the prompting framework of large language models?", "answer": ["ReWOO: Decoupling Reasoning from Observations for Efficient Augmented\n Language Models", "Reflexion: Language Agents with Verbal Reinforcement Learning", "MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action", "ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs", "ToolAlpaca: Generalized Tool Learning for Language Models with 3000\n Simulated Cases"], "answer_arxiv_id": ["2305.18323", "2303.11366", "2303.11381", "2307.16789", "2306.05301"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_test_63"} +{"question": "In which paper the term FPE was formalised for general function approximators?", "answer": ["Batch Policy Learning under Constraints"], "answer_arxiv_id": ["1903.08738"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_test_64"} +{"question": "Which works focus on modelling the annotator distribution?", "answer": ["PHiSeg: Capturing Uncertainty in Medical Image Segmentation", "A Hierarchical Probabilistic U-Net for Modeling Multi-Scale Ambiguities", "Uncertainty quantification in medical image segmentation with normalizing flows"], "answer_arxiv_id": ["1906.04045", "1905.13077", "2006.02683"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_test_65"} +{"question": "Which studies designed a siamese network framework using AlexNet for feature extraction in visual object tracking?", "answer": ["Fully-Convolutional Siamese Networks for Object Tracking"], "answer_arxiv_id": ["1606.09549"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_test_66"} +{"question": "What graph analysis model is tested in the benchmark?", "answer": ["Explainable Classification of Brain Networks via Contrast Subgraphs"], "answer_arxiv_id": ["2006.05176"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_test_67"} +{"question": "Any research focused on the memorization risks during the fine-tuning stage?", "answer": ["Memorization in NLP Fine-tuning Methods", "Do Language Models Plagiarize?"], "answer_arxiv_id": ["2205.12506", "2203.07618"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_test_68"} +{"question": "Could you provide me some studies about reducing the gradient misestimation by approximating discrete quantization with a differentiable function?", "answer": ["Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networks", "DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients"], "answer_arxiv_id": ["1908.05033", "1606.06160"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_test_69"} +{"question": "Could you provide me some works about optimizing batch processing for LLMs?", "answer": ["Batch Prompting: Efficient Inference with Large Language Model APIs", "TurboTransformers: An Efficient GPU Serving System For Transformer Models"], "answer_arxiv_id": ["2301.08721", "2010.05680"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_70"} +{"question": "Which study extended the capabilities of LLMs to the field of multi-modality?", "answer": ["Gemini: A Family of Highly Capable Multimodal Models"], "answer_arxiv_id": ["2312.11805v4"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_71"} +{"question": "What research has been done on finding optimal interventions using observational data?", "answer": ["Learning to search efficiently for causally near-optimal treatments"], "answer_arxiv_id": ["2007.00973"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_72"} +{"question": "What papers are about prototypical adaptation methods?", "answer": ["Bending Reality: Distortion-aware Transformers for Adapting to Panoramic\n Semantic Segmentation", "Behind Every Domain There is a Shift: Adapting Distortion-aware Vision\n Transformers for Panoramic Semantic Segmentation"], "answer_arxiv_id": ["2203.01452", "2207.11860"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_test_73"} +{"question": "Could you name the works that applied CLIP for zero-shot AD, scoring the anomalies by comparing the alignment of test images with the correct text of normal samples?", "answer": ["Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images"], "answer_arxiv_id": ["2205.11474"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_74"} +{"question": "What papers illustrate recent neural scene representations methods that try to optimize poses with differentiable rendering in Structure-from-Motion research?", "answer": ["BARF : Bundle-Adjusting Neural Radiance Fields", "Self-Calibrating Neural Radiance Fields"], "answer_arxiv_id": ["2104.06405", "2108.13826"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_75"} +{"question": "Could you provide me some works that investigate the interplay between weight loss landscape and adversarial robustness?", "answer": ["Adversarial Weight Perturbation Helps Robust Generalization", "Enhancing Adversarial Training with Second-Order Statistics of Weights"], "answer_arxiv_id": ["2004.05884", "2203.06020"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_76"} +{"question": "Which works employed a dynamic weighting transformer for integration in MMEA?", "answer": ["MEAformer: Multi-modal Entity Alignment Transformer for Meta Modality\n Hybrid"], "answer_arxiv_id": ["2212.14454"], "source_meta": {"published_time": "20240723"}, "qid": "AutoScholarQuery_test_77"} +{"question": "Which works have been conducted on memory methods for object navigation tasks?", "answer": ["SOON: Scenario Oriented Object Navigation with Graph-based Exploration"], "answer_arxiv_id": ["2103.17138"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_test_78"} +{"question": "Which study presents the use of synthetic captions for training BLIP and BLIP2 models?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models"], "answer_arxiv_id": ["2201.12086", "2301.12597"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_79"} +{"question": "Could you provide me some works about multi-agent debating frameworks?", "answer": ["Improving Factuality and Reasoning in Language Models through Multiagent\n Debate", "Encouraging Divergent Thinking in Large Language Models through\n Multi-Agent Debate", "ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate"], "answer_arxiv_id": ["2305.14325", "2305.19118", "2308.07201"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_test_80"} +{"question": "Which research provide examples of multimodal-conditional image synthesis systems?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Adding Conditional Control to Text-to-Image Diffusion Models", "GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2112.10752", "2211.01324", "2302.05543", "2301.07093"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_test_81"} +{"question": "Which studies showed successful results using group-level persona variables?", "answer": ["Jury Learning: Integrating Dissenting Voices into Machine Learning\n Models", "When the Majority is Wrong: Modeling Annotator Disagreement for\n Subjective Tasks"], "answer_arxiv_id": ["2202.02950", "2305.06626"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_82"} +{"question": "Could you provide me with works that discuss the problem of performance degradation when distilling larger LMs, especially when the student is of small scale?", "answer": ["Improved Knowledge Distillation via Teacher Assistant", "On the Efficacy of Knowledge Distillation", "Lifting the Curse of Capacity Gap in Distilling Language Models"], "answer_arxiv_id": ["1902.03393", "1910.01348", "2305.12129"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_83"} +{"question": "Could you provide me some works about generative methods for transferable adversarial attacks?", "answer": ["Generative Adversarial Perturbations", "Cross-Domain Transferability of Adversarial Perturbations", "On Generating Transferable Targeted Perturbations"], "answer_arxiv_id": ["1712.02328v3", "1905.11736", "2103.14641"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_test_84"} +{"question": "What is the key work on Trust Region Policy Optimization?", "answer": ["Trust Region Policy Optimization"], "answer_arxiv_id": ["1502.05477"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_85"} +{"question": "What works focus on spatial feature transformation for BEV feature generation?", "answer": ["PersFormer: 3D Lane Detection via Perspective Transformer and the\n OpenLane Benchmark"], "answer_arxiv_id": ["2203.11089"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_test_86"} +{"question": "What work used a modified VQ-GAN for isolated word sign language video generation?", "answer": ["Continuous 3D Multi-Channel Sign Language Production via Progressive\n Transformers and Mixture Density Networks"], "answer_arxiv_id": ["2103.06982"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_test_87"} +{"question": "What papers propose the use of FP8 for accelerated inference?", "answer": ["FP8 Quantization: The Power of the Exponent"], "answer_arxiv_id": ["2208.09225"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_test_88"} +{"question": "Who analysed the NTK spectrum for shallow ReLU networks under the uniform and nonuniform distributions?", "answer": ["Frequency Bias in Neural Networks for Input of Non-Uniform Density"], "answer_arxiv_id": ["2003.04560"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_test_89"} +{"question": "Which works explored the theoretical analysis of the NTK spectrum via random matrix theory?", "answer": ["Spectra of the Conjugate Kernel and Neural Tangent Kernel for Linear-Width Neural Networks"], "answer_arxiv_id": ["2005.11879"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_test_90"} +{"question": "Any research work about directly predicting CNN classifier accuracy by deriving distribution distance features between training and test images with a linear regression model?", "answer": ["Are Labels Always Necessary for Classifier Accuracy Evaluation?", "What Does Rotation Prediction Tell Us about Classifier Accuracy under Varying Testing Environments?"], "answer_arxiv_id": ["2007.02915", "2106.05961"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_test_91"} +{"question": "What works feature insightful discussions on preconditioning?", "answer": ["When Does Preconditioning Help or Hurt Generalization?", "Preconditioned Score-based Generative Models", "Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["2006.10732", "2302.06504", "1512.03385"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_92"} +{"question": "Which paper introduced Vector Quantized Variational Autoencoders (VQ-VAE)?", "answer": ["Neural Discrete Representation Learning"], "answer_arxiv_id": ["1711.00937"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_test_93"} +{"question": "Which research introduced a graph generation method for query structure prediction in parsing?", "answer": ["Formal Query Building with Query Structure Prediction for Complex Question Answering over Knowledge Base"], "answer_arxiv_id": ["2109.03614"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_94"} +{"question": "Could you provide some works about deep AD approaches that employ a self-supervised loss function to train the detector and score anomalies?", "answer": ["Deep Anomaly Detection Using Geometric Transformations", "Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty", "Learning and Evaluating Representations for Deep One-class Classification", "Classification-Based Anomaly Detection for General Data", "Neural Transformation Learning for Deep Anomaly Detection Beyond Images", "Detecting Anomalies within Time Series using Local Neural Transformations", "Deep Anomaly Detection under Labeling Budget Constraints"], "answer_arxiv_id": ["1805.10917", "1906.12340", "2011.02578", "2005.02359", "2103.16440", "2202.03944", "2302.07832v2"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_95"} +{"question": "What studies have proposed methods to facilitate better model and AI service documentation?", "answer": ["Model Cards for Model Reporting"], "answer_arxiv_id": ["1810.03993"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_96"} +{"question": "Which study offers a lightweight, subject-driven personalization for text-to-image diffusion models?", "answer": ["HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2307.06949"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_test_97"} +{"question": "What works present operators of tensor decomposition composed of fast Fourier / trigonometric transforms?", "answer": ["Faster Johnson-Lindenstrauss Transforms via Kronecker Products"], "answer_arxiv_id": ["1909.04801"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_test_98"} +{"question": "What paper describes the dataset MiniWoB++, where sequences of low-level UI commands describe multi-step tasks?", "answer": ["Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration"], "answer_arxiv_id": ["1802.08802"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_99"} +{"question": "Which paper proposed improving the anomaly score for reconstruction-based techniques via constructing a score that integrates reconstruction error and discriminator loss?", "answer": ["TranAD: Deep Transformer Networks for Anomaly Detection in Multivariate Time Series Data"], "answer_arxiv_id": ["2201.07284"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_100"} +{"question": "What work proposes to model speech based on HuBERT codes or semantic tokens?", "answer": ["Generative Spoken Language Modeling from Raw Audio", "Speech Resynthesis from Discrete Disentangled Self-Supervised\n Representations"], "answer_arxiv_id": ["2102.01192", "2104.00355"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_101"} +{"question": "What research works talk about using Inverse Propensity Score (IPS) and Self-Normalized IPS (SNIPS) methods to tackle selection bias on data?", "answer": ["Recommendations as Treatments: Debiasing Learning and Evaluation"], "answer_arxiv_id": ["1602.05352"], "source_meta": {"published_time": "20220510"}, "qid": "AutoScholarQuery_test_102"} +{"question": "What paper introduced the Segment Anything Model (SAM) which is a foundational model for image segmentation?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_103"} +{"question": "Which publications contain the principal components of latent space obtained by performing PCA as global meaningful perturbations?", "answer": ["GANSpace: Discovering Interpretable GAN Controls"], "answer_arxiv_id": ["2004.02546"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_104"} +{"question": "What studies introduced mask classification-based methods for instance-level segmentation?", "answer": ["End-to-End Object Detection with Transformers", "MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers"], "answer_arxiv_id": ["2005.12872", "2012.00759"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_105"} +{"question": "What studies extensively examined the MMD two-sample test?", "answer": ["Learning Deep Kernels for Exponential Family Densities", "Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy", "Fast Two-Sample Testing with Analytic Representations of Probability Measures", "Interpretable Distribution Features with Maximum Testing Power"], "answer_arxiv_id": ["1811.08357", "1611.04488", "1506.04725", "1605.06796"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_test_106"} +{"question": "What works are related to 3D Gaussian Splatting for 3D reconstruction?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis"], "answer_arxiv_id": ["2308.04079", "2308.09713"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_test_107"} +{"question": "Can you specify the studies about using prompt and fine-tuning techniques for adapting VLMs to a new downstream task?", "answer": ["Learning to Prompt for Vision-Language Models", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters"], "answer_arxiv_id": ["2109.01134", "2110.04544"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_test_108"} +{"question": "What works propose 2D-to-3D pose lifting task and regress the 3D keypoints based on a convolutional neural network from 2D keypoints?", "answer": ["A simple yet effective baseline for 3d human pose estimation"], "answer_arxiv_id": ["1705.03098"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_test_109"} +{"question": "What researches have created evaluation data for individual tasks in Indic languages?", "answer": ["IndicTrans2: Towards High-Quality and Accessible Machine Translation Models for all 22 Scheduled Indian Languages", "Mukhyansh: A Headline Generation Dataset for Indic Languages", "PMIndiaSum: Multilingual and Cross-lingual Headline Summarization for\n Languages in India", "Samanantar: The Largest Publicly Available Parallel Corpora Collection for 11 Indic Languages"], "answer_arxiv_id": ["2305.16307v3", "2311.17743", "2305.08828", "2104.05596v4"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_test_110"} +{"question": "What are some classic methods for learning latent graphs?", "answer": ["Learning Discrete Structures for Graph Neural Networks", "Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings", "Graph Structure Learning for Robust Graph Neural Networks"], "answer_arxiv_id": ["1903.11960", "2006.13009", "2005.10203"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_test_111"} +{"question": "Which paper introduced the method known as CoT prompting?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_112"} +{"question": "Which works assumed Gaussian noise in RGB space for pixel-wise uncertainty in the context of NeRF?", "answer": ["NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo\n Collections", "ActiveNeRF: Learning where to See with Uncertainty Estimation"], "answer_arxiv_id": ["2008.02268", "2209.08546"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_113"} +{"question": "What papers explored using diffusion models within the realm of reinforcement learning?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_test_114"} +{"question": "Could you provide me works that touch upon one-class classification approaches for video anomaly detection?", "answer": ["Old is Gold: Redefining the Adversarially Learned One-Class Classifier\n Training Paradigm", "Adversarially Learned One-Class Classifier for Novelty Detection", "Memorizing Normality to Detect Anomaly: Memory-augmented Deep\n Autoencoder for Unsupervised Anomaly Detection", "Object-centric Auto-encoders and Dummy Anomalies for Abnormal Event\n Detection in Video"], "answer_arxiv_id": ["2004.07657", "1802.09088", "1904.02639", "1812.04960"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_test_115"} +{"question": "Which studies have focused on the protein docking technique?", "answer": ["Independent SE(3)-Equivariant Models for End-to-End Rigid Protein Docking"], "answer_arxiv_id": ["2111.07786"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_116"} +{"question": "What are some works in vision that stress the importance of data selection in supervised or semi-supervised setting?", "answer": ["Beyond neural scaling laws: beating power law scaling via data pruning", "Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt", "Learning From Less Data: A Unified Data Subset Selection and Active Learning Framework for Computer Vision", "Glister: Generalization based Data Subset Selection for Efficient and Robust Learning", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training", "RETRIEVE: Coreset Selection for Efficient and Robust Semi-Supervised Learning", "Optimizing Data Usage via Differentiable Rewards", "Deep Learning on a Data Diet: Finding Important Examples Early in Training", "Coresets for Data-efficient Training of Machine Learning Models", "Selection via Proxy: Efficient Data Selection for Deep Learning", "Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["2206.14486v6", "2206.07137", "1901.01151", "2012.10630", "2103.00123", "2106.07760v2", "1911.10088", "2107.07075", "1906.01827", "1906.11829", "1708.00489"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_117"} +{"question": "What works adopted large language models (LLMs) for a cost-effective generation of Counterfactually Augmented Data (CAD)?", "answer": ["Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and\n Improving Models", "Generate Your Counterfactuals: Towards Controlled Counterfactual\n Generation for Text", "AutoCAD: Automatically Generating Counterfactuals for Mitigating\n Shortcut Learning", "CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generation", "Automatic Prompt Optimization with \"Gradient Descent\" and Beam Search", "DISCO: Distilling Counterfactuals with Large Language Models"], "answer_arxiv_id": ["2101.00288", "2012.04698", "2211.16202", "2210.04873", "2305.03495", "2212.10534"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_test_118"} +{"question": "Which studies demonstrated that vector arithmetic on latent space leads to the semantic arithmetic on the image space?", "answer": ["Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks", "Deep Feature Interpolation for Image Content Changes"], "answer_arxiv_id": ["1511.06434", "1611.05507"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_119"} +{"question": "Which research papers cover the challenges of covariate shift and causal confusion in behavior cloning?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "Causal Confusion in Imitation Learning"], "answer_arxiv_id": ["1011.0686v3", "1905.11979"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_test_120"} +{"question": "What papers focus on the concept of equivariance in Convolutional Neural Networks (CNNs)?", "answer": ["Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1602.07576"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_test_121"} +{"question": "What papers discuss the definitions and concepts of Pseudo-Global Stability in regards to Differential Privacy?", "answer": ["User-Level Private Learning via Correlated Sampling"], "answer_arxiv_id": ["2110.11208"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_122"} +{"question": "Which works focus on predicting model generalization error?", "answer": ["Are Labels Always Necessary for Classifier Accuracy Evaluation?", "Leveraging Unlabeled Data to Predict Out-of-Distribution Performance", "Predicting Out-of-Distribution Error with the Projection Norm", "On the Strong Correlation Between Model Invariance and Generalization", "Predicting with Confidence on Unseen Distributions", "What Does Rotation Prediction Tell Us about Classifier Accuracy under Varying Testing Environments?"], "answer_arxiv_id": ["2007.02915", "2201.04234", "2202.05834", "2207.07065", "2107.03315", "2106.05961"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_test_123"} +{"question": "Which study proposed the Subformer model with shared middle layers and embedding factorization?", "answer": ["Subformer: Exploring Weight Sharing for Parameter Efficiency in\n Generative Transformers"], "answer_arxiv_id": ["2101.00234"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_124"} +{"question": "What studies have proven Generation-Augmented Retrieval effective in question answering and passage retrieval?", "answer": ["Generation-Augmented Retrieval for Open-domain Question Answering", "Precise Zero-Shot Dense Retrieval without Relevance Labels", "Query2doc: Query Expansion with Large Language Models"], "answer_arxiv_id": ["2009.08553", "2212.10496", "2303.07678"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_test_125"} +{"question": "What paper introduces the Mirror Descent Modified Policy Iteration (MD-MPI) framework?", "answer": ["A Theory of Regularized Markov Decision Processes"], "answer_arxiv_id": ["1901.11275"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_126"} +{"question": "What papers provide linear convergence guarantees of NPG and PMD in softmax tabular policy settings by adding regularization?", "answer": ["Fast Global Convergence of Natural Policy Gradient Methods with Entropy Regularization", "Policy Mirror Descent for Regularized Reinforcement Learning: A Generalized Framework with Linear Convergence", "Policy Mirror Descent for Reinforcement Learning: Linear Convergence, New Sampling Complexity, and Generalized Problem Classes", "Homotopic Policy Mirror Descent: Policy Convergence, Implicit Regularization, and Improved Sample Complexity"], "answer_arxiv_id": ["2007.06558", "2105.11066", "2102.00135", "2201.09457v9"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_127"} +{"question": "Which papers established a basis of ARA using language models based on deep neural networks?", "answer": ["Supervised and Unsupervised Neural Approaches to Text Readability"], "answer_arxiv_id": ["1907.11779"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_128"} +{"question": "Which works proposed a simple approach based on the statistics of the dataset for response length prediction?", "answer": ["Fast Structured Decoding for Sequence Models"], "answer_arxiv_id": ["1910.11555"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_129"} +{"question": "Can you mention publications that implemented transformers in 2D-to-3D pose lifting?", "answer": ["3D Human Pose Estimation with Spatial and Temporal Transformers", "MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose\n Estimation in Video", "PoseFormerV2: Exploring Frequency Domain for Efficient and Robust 3D\n Human Pose Estimation"], "answer_arxiv_id": ["2103.10455", "2203.00859", "2303.17472"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_130"} +{"question": "Can you provide an example of a paper that presents self-tuning algorithms?", "answer": ["A Self-Tuning Actor-Critic Algorithm"], "answer_arxiv_id": ["2002.12928"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_131"} +{"question": "Which papers talk about using LLMs to interpret themselves or other ML models by providing numerical importances for their inputs?", "answer": ["Are Large Language Models Post Hoc Explainers?"], "answer_arxiv_id": ["2310.05797"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_132"} +{"question": "Could you provide me studies about knowledge distillation methods for semantic segmentation focusing on preserving structural semantic relations?", "answer": ["Knowledge Adaptation for Efficient Semantic Segmentation", "Channel-wise Knowledge Distillation for Dense Prediction", "Cross-Image Relational Knowledge Distillation for Semantic Segmentation"], "answer_arxiv_id": ["1903.04688v1", "2011.13256v4", "2204.06986"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_test_133"} +{"question": "What works have focused on evaluating API-use scenarios based on a separate LLM to assess the quality of examples?", "answer": ["ToolAlpaca: Generalized Tool Learning for Language Models with 3000\n Simulated Cases", "ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs"], "answer_arxiv_id": ["2306.05301", "2307.16789"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_test_134"} +{"question": "Any works about user-annotations based image animation?", "answer": ["iPOKE: Poking a Still Image for Controlled Stochastic Video Synthesis", "Stochastic Latent Residual Video Prediction", "DragNUWA: Fine-grained Control in Video Generation by Integrating Text,\n Image, and Trajectory", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "ControlVideo: Training-free Controllable Text-to-Video Generation", "Motion-Conditioned Diffusion Model for Controllable Video Synthesis"], "answer_arxiv_id": ["2107.02790", "2002.09219", "2308.08089", "2306.02018", "2305.13077", "2304.14404"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_test_135"} +{"question": "What works conducted studies on LegalBERT and CaseLaw-BERT that focused on the legal domain?", "answer": ["LEGAL-BERT: The Muppets straight out of Law School", "When Does Pretraining Help? Assessing Self-Supervised Learning for Law\n and the CaseHOLD Dataset"], "answer_arxiv_id": ["2010.02559", "2104.08671"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_test_136"} +{"question": "Are there any works that created datasets and systems for tasks related to information time-tracking?", "answer": ["Socially-Informed Timeline Generation for Complex Events", "Examining the State-of-the-Art in News Timeline Summarization"], "answer_arxiv_id": ["1606.05699", "2005.10107"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_test_137"} +{"question": "Could you provide me studies about anomaly detection in federated learning particularly related to network security?", "answer": ["A Secure Federated Learning Framework for 5G Networks"], "answer_arxiv_id": ["2005.05752"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_test_138"} +{"question": "What research focused on 3D instance segmentation with only 3D box annotation requirements?", "answer": ["Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using\n Bounding Boxes"], "answer_arxiv_id": ["2206.01203"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_test_139"} +{"question": "Are there any papers that investigate the need for an initially annotated fraction of data to bootstrap an active learning method?", "answer": ["Making Your First Choice: To Address Cold Start Problem in Vision Active Learning", "You Never Get a Second Chance To Make a Good First Impression: Seeding Active Learning for 3D Semantic Segmentation", "Cold-start Active Learning through Self-supervised Language Modeling", "Addressing the Item Cold-start Problem by Attribute-driven Active Learning"], "answer_arxiv_id": ["2210.02442", "2304.11762v2", "2010.09535", "1805.09023"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_test_140"} +{"question": "Can you provide works that evaluate RL agents by changing the initial states in the same environment?", "answer": ["Teacher algorithms for curriculum learning of Deep RL in continuously parameterized environments", "Sim-to-Real Transfer of Robotic Control with Dynamics Randomization", "Continual Reinforcement Learning with Complex Synapses"], "answer_arxiv_id": ["1910.07224", "1710.06537", "1802.07239"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_test_141"} +{"question": "Which research paper introduced the DenseCL method that applies contrastive learning on patches with highest similarity?", "answer": ["Dense Contrastive Learning for Self-Supervised Visual Pre-Training"], "answer_arxiv_id": ["2011.09157"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_test_142"} +{"question": "Which work outlines a way to calculate adversarial perturbations during training by first randomly perturbing the initial point then applying a single step of projected gradient descent?", "answer": ["Fast is better than free: Revisiting adversarial training"], "answer_arxiv_id": ["2001.03994"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_test_143"} +{"question": "Could you provide some works that discuss multimodal prompting methods?", "answer": ["Large Language Models are Zero-Shot Reasoners", "Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning\n by Large Language Models", "Better Zero-Shot Reasoning with Self-Adaptive Prompting", "Language Models are Few-Shot Learners", "Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "A Survey on In-context Learning", "Fairness-guided Few-shot Prompting for Large Language Models", "ExpertPrompting: Instructing Large Language Models to be Distinguished\n Experts", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Automatic Chain of Thought Prompting in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Graph of Thoughts: Solving Elaborate Problems with Large Language Models", "Beyond Chain-of-Thought, Effective Graph-of-Thought Reasoning in\n Language Models", "Boosting Logical Reasoning in Large Language Models through a New\n Framework: The Graph of Thought"], "answer_arxiv_id": ["2205.11916", "2305.04091", "2305.14106", "2005.14165", "2202.12837", "2301.00234", "2303.13217", "2305.14688", "2201.11903", "2210.03493", "2203.11171", "2305.10601", "2308.09687", "2305.16582", "2308.08614"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_test_144"} +{"question": "Which work proposes reweighting different activation channels based on the global context of the input samples?", "answer": ["Squeeze-and-Excitation Networks"], "answer_arxiv_id": ["1709.01507"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_test_145"} +{"question": "What works focused on MAML and its variants?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Sharp-MAML: Sharpness-Aware Model-Agnostic Meta Learning", "Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML", "Alpha MAML: Adaptive Model-Agnostic Meta-Learning", "Meta-Learning with Implicit Gradients"], "answer_arxiv_id": ["1703.03400", "2206.03996", "1909.09157", "1905.07435", "1909.04630"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_test_146"} +{"question": "What research papers have addressed issues of equivariance by customizing kernel designs in the context of kernel methods and Gaussian processes (GPs)?", "answer": ["GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Acceleration", "Deep Kernel Learning", "Kernel Identification Through Transformers"], "answer_arxiv_id": ["1809.11165", "1511.02222", "2106.08185"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_test_147"} +{"question": "What studies have leveraged extensive image-text pair datasets to broaden the detection vocabulary in Open-vocabulary detection?", "answer": ["Open-Vocabulary Object Detection Using Captions", "RegionCLIP: Region-based Language-Image Pretraining", "PromptDet: Towards Open-vocabulary Detection using Uncurated Images", "Grounded Language-Image Pre-training", "Learning Object-Language Alignments for Open-Vocabulary Object Detection", "DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via\n Word-Region Alignment"], "answer_arxiv_id": ["2011.10678", "2112.09106", "2203.16513", "2112.03857", "2211.14843", "2304.04514"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_148"} +{"question": "Could you mention research that addresses MMS approximations for suodular and subadditive valuations?", "answer": ["Approximation Algorithms for Maximin Fair Division"], "answer_arxiv_id": ["1703.01851"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_149"} +{"question": "Which works can you provide that are focused on creating evaluation data on Indic languages?", "answer": ["Towards Leaving No Indic Language Behind: Building Monolingual Corpora,\n Benchmark and Models for Indic Languages", "Naamapadam: A Large-Scale Named Entity Annotated Data for Indic\n Languages", "MASSIVE: A 1M-Example Multilingual Natural Language Understanding\n Dataset with 51 Typologically-Diverse Languages", "GLUECoS : An Evaluation Benchmark for Code-Switched NLP", "The Belebele Benchmark: a Parallel Reading Comprehension Dataset in 122\n Language Variants"], "answer_arxiv_id": ["2212.05409", "2212.10168", "2204.08582", "2004.12376", "2308.16884"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_test_150"} +{"question": "What work highlights the disconnect between benchmark results and real world impacts in NLP?", "answer": ["Machine Learning that Matters"], "answer_arxiv_id": ["1206.4656v1"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_151"} +{"question": "What work applied CLIP to dense prediction tasks?", "answer": ["DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting"], "answer_arxiv_id": ["2112.01518"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_152"} +{"question": "Which papers have proposed for extracting the specific style from reference images?", "answer": ["StyleAdapter: A Single-Pass LoRA-Free Model for Stylized Image\n Generation", "Domain Enhanced Arbitrary Image Style Transfer via Contrastive Learning", "StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion\n Models", "Inversion-Based Style Transfer with Diffusion Models", "StyleDrop: Text-to-Image Generation in Any Style"], "answer_arxiv_id": ["2309.01770", "2205.09542", "2308.07863", "2211.13203", "2306.00983"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_test_153"} +{"question": "What papers studied the construction of diffusion models for discrete categorical data using a categorical noise process?", "answer": ["Structured Denoising Diffusion Models in Discrete State-Spaces", "Argmax Flows and Multinomial Diffusion: Learning Categorical Distributions"], "answer_arxiv_id": ["2107.03006", "2102.05379"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_154"} +{"question": "Which works combine external knowledge from KGs into LLMs during the prompting stage?", "answer": ["Knowledge-Augmented Language Model Prompting for Zero-Shot Knowledge Graph Question Answering", "Knowledge-Driven CoT: Exploring Faithful Reasoning in LLMs for\n Knowledge-intensive Question Answering", "MindMap: Knowledge Graph Prompting Sparks Graph of Thoughts in Large\n Language Models", "Reasoning on Graphs: Faithful and Interpretable Large Language Model\n Reasoning", "Think-on-Graph: Deep and Responsible Reasoning of Large Language Model\n on Knowledge Graph"], "answer_arxiv_id": ["2306.04136v1", "2308.13259", "2308.09729", "2310.01061", "2307.07697"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_test_155"} +{"question": "What works have concentrated on generating free-form natural language explanations (NLEs) for justifying model predictions?", "answer": ["e-SNLI: Natural Language Inference with Natural Language Explanations", "WT5?! Training Text-to-Text Models to Explain their Predictions", "Teach Me to Explain: A Review of Datasets for Explainable Natural\n Language Processing"], "answer_arxiv_id": ["1812.01193", "2004.14546", "2102.12060"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_156"} +{"question": "What work proposes the intrinsic local dimension estimation scheme for the latent manifold through the robust rank estimate?", "answer": ["Analyzing the Latent Space of GAN through Local Dimension Estimation"], "answer_arxiv_id": ["2205.13182"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_157"} +{"question": "Which paper first used the term hypercolumn in the context of neural network features?", "answer": ["Hypercolumns for Object Segmentation and Fine-grained Localization"], "answer_arxiv_id": ["1411.5752"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_158"} +{"question": "Could you provide me some studies that used NeRFs for novel view synthesis and 3D scene reconstruction?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2003.08934", "2111.12077", "2201.05989"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_159"} +{"question": "Could you provide me some works that incorporate ideas from multicalibration to improve conditional coverage?", "answer": ["Practical Adversarial Multivalid Conformal Prediction"], "answer_arxiv_id": ["2206.01067"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_test_160"} +{"question": "Which papers have achieved progress in the field of graph contrastive learning?", "answer": ["Graph Contrastive Learning with Augmentations", "Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "Graph Contrastive Learning Automated", "Adversarial Graph Contrastive Learning with Information Regularization", "Bringing Your Own View: Graph Contrastive Learning without Prefabricated Data Augmentations"], "answer_arxiv_id": ["2010.13902", "2106.05819", "2106.07594", "2202.06491", "2201.01702"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_test_161"} +{"question": "Can you provide existing research that used contrastive methods in representation learning?", "answer": ["Deep Graph Infomax"], "answer_arxiv_id": ["1809.10341"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_test_162"} +{"question": "What works introduced methodologies involving the learning of kernels to detect illuminant chromaticity?", "answer": ["Convolutional Color Constancy", "Fast Fourier Color Constancy"], "answer_arxiv_id": ["1507.00410", "1611.07596"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_163"} +{"question": "Could you tell me what studies propose to bridge vision and language modalities through visual prompt generators?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Visual Instruction Tuning", "Language Is Not All You Need: Aligning Perception with Language Models"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2305.06500", "2304.08485", "2302.14045"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_164"} +{"question": "Could you provide me some studies about sign language translation and production using RWTH-Phoenix dataset?", "answer": ["Sign Language Transformers: Joint End-to-end Sign Language Recognition\n and Translation", "Progressive Transformers for End-to-End Sign Language Production"], "answer_arxiv_id": ["2003.13830", "2004.14874"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_165"} +{"question": "Which work talked about ODIN, a method where two networks jointly optimize the segmentation masks and representation of objects?", "answer": ["Object discovery and representation networks"], "answer_arxiv_id": ["2203.08777"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_166"} +{"question": "Which works proposed modifications for discrete-time models, such as recurrent neural networks to handle irregular time series data?", "answer": ["On the Properties of Neural Machine Translation: Encoder–Decoder Approaches"], "answer_arxiv_id": ["1409.1259"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_test_167"} +{"question": "Could you mention a study that generates diverse reasoning paths using CoT?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_test_168"} +{"question": "Could you provide me studies where the solution of the reverse-time SDE is theoretically derived?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_test_169"} +{"question": "Could you provide me some researches about extensions for population-based training?", "answer": ["Population Based Training of Neural Networks", "A Generalized Framework for Population Based Training"], "answer_arxiv_id": ["1711.09846", "1902.01894"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_170"} +{"question": "Any studies that focused on the problem of proving structured queries to knowledge base?", "answer": ["End-to-End Differentiable Proving", "Differentiable Reasoning on Large Knowledge Bases and Natural Language"], "answer_arxiv_id": ["1705.11040", "1912.10824"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_test_171"} +{"question": "Which paper improved the scene text editing performance by incorporating stroke-level information?", "answer": ["Exploring Stroke-Level Modifications for Scene Text Editing"], "answer_arxiv_id": ["2212.01982"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_test_172"} +{"question": "Which works implement the conceptual design of complex-valued activations for object binding in synchrony-based models?", "answer": ["Neuronal Synchrony in Complex-Valued Deep Networks", "Complex-Valued Autoencoders for Object Discovery"], "answer_arxiv_id": ["1312.6115", "2204.02075"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_173"} +{"question": "Which works have implemented different kernels at various regions of the DP pair for single-task based deblurring?", "answer": ["Defocus Deblurring Using Dual-Pixel Data", "Learning to Reduce Defocus Blur by Realistically Modeling Dual-Pixel\n Data", "Improving Single-Image Defocus Deblurring: How Dual-Pixel Images Help Through Multi-Task Learning"], "answer_arxiv_id": ["2005.00305", "2012.03255", "2108.05251v2"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_174"} +{"question": "Where does the ControlNet, a finetuning model for synthetic face generation, discussed?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_test_175"} +{"question": "Which works considered the physical world and social norms for affordance inference?", "answer": ["Learning to Act Properly: Predicting and Explaining Affordances from Images"], "answer_arxiv_id": ["1712.07576"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_test_176"} +{"question": "Can you mention some studies that have explored quantization as a means of compressing the parameters of LLMs for efficient inference?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale"], "answer_arxiv_id": ["2208.07339"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_177"} +{"question": "Could you provide me with some works on memorization study that involve finding training dataset documents related to the output?", "answer": ["Do Language Models Plagiarize?", "Large Language Models Struggle to Learn Long-Tail Knowledge", "Data Contamination: From Memorization to Exploitation"], "answer_arxiv_id": ["2203.07618", "2211.08411", "2203.08242"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_178"} +{"question": "Which references provided a dataset facilitating entity embedding through linked images?", "answer": ["MMKG: Multi-Modal Knowledge Graphs"], "answer_arxiv_id": ["1903.05485"], "source_meta": {"published_time": "20240723"}, "qid": "AutoScholarQuery_test_179"} +{"question": "What works have studied sampling for regret minimization in the framework of Unsupervised Environment Design?", "answer": ["Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Design", "Replay-Guided Adversarial Environment Design"], "answer_arxiv_id": ["2012.02096", "2110.02439"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_test_180"} +{"question": "Which works used middle-layer LLM outputs and inserted them into the VPG to identify differences between images?", "answer": ["Fine-tuning Multimodal LLMs to Follow Zero-shot Demonstrative\n Instructions"], "answer_arxiv_id": ["2308.04152"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_181"} +{"question": "Which works achieve high identity preservation in personalized generation through massive datasets and expensive hardware resources?", "answer": ["Face0: Instantaneously Conditioning a Text-to-Image Model on a Face", "Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning"], "answer_arxiv_id": ["2306.06638", "2307.11410"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_test_182"} +{"question": "Are there any studies about efficient selection of source tasks in NLP?", "answer": ["NLP From Scratch Without Large-Scale Pretraining: A Simple and Efficient Framework", "Task Compass: Scaling Multi-task Pre-training with Task Prefix"], "answer_arxiv_id": ["2111.04130", "2210.06277"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_test_183"} +{"question": "What works discuss transformer-based methods for 3D instance segmentation?", "answer": ["Mask3D: Mask Transformer for 3D Semantic Instance Segmentation", "Superpoint Transformer for 3D Scene Instance Segmentation"], "answer_arxiv_id": ["2210.03105", "2211.15766"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_test_184"} +{"question": "Which work explores model calibration for Uncertainty Estimation in the context of multiple-choice question answering?", "answer": ["How Can We Know When Language Models Know? On the Calibration of\n Language Models for Question Answering"], "answer_arxiv_id": ["2012.00955"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_185"} +{"question": "Could you provide me some studies that explored policy optimization via backpropagating through the dynamics model?", "answer": ["Model-Augmented Actor-Critic: Backpropagating through Paths"], "answer_arxiv_id": ["2005.08068"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_test_186"} +{"question": "Which study proposed a complete graph kernel based on homomorphism counts in the context of graph learning tasks?", "answer": ["Lovász Meets Weisfeiler and Leman"], "answer_arxiv_id": ["1802.08876"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_test_187"} +{"question": "Which study developed a Jacobi diffusion process for discrete data diffusion models?", "answer": ["Dirichlet Diffusion Score Model for Biological Sequence Generation"], "answer_arxiv_id": ["2305.10699"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_test_188"} +{"question": "What research introduced DPMs and linked the generative model to a denoising diffusion model?", "answer": ["Auto-Encoding Variational Bayes", "Generative Adversarial Nets", "Towards Building A Group-based Unsupervised Representation Disentanglement Framework", "Learning Disentangled Representation by Exploiting Pretrained Generative Models: A Contrastive Learning View", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality", "Gotta Go Fast When Generating Data with Score-Based Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["1312.6114", "1406.2661", "2102.10303", "2102.10543", "1503.03585", "2006.11239", "2102.09672", "2010.02502", "2202.05830", "2105.14080", "2201.06503", "2106.05931"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_test_189"} +{"question": "Which studies focused on structured neural networks embedding principles like equivariance, Euclidean symmetry and periodicity?", "answer": ["Group Equivariant Convolutional Networks", "Neural Descriptor Fields: \"SE\"⁢(3)-Equivariant Object Representations for Manipulation", "Periodic DMP formulation for Quaternion Trajectories"], "answer_arxiv_id": ["1602.07576", "2112.05124", "2110.10510"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_test_190"} +{"question": "Which study proposed the use of memorization as a metric in CoreSet selection?", "answer": ["What Neural Networks Memorize and Why: Discovering the Long Tail via\n Influence Estimation"], "answer_arxiv_id": ["2008.03703"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_191"} +{"question": "Which papers discussed Parameter-efficient fine-tuning (PEFT) methods for reducing storage requirements in large-scale PLMs?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["1902.00751", "2104.08691", "2101.00190"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_test_192"} +{"question": "What studies compare using a large web corpus versus Wikipedia?", "answer": ["Cloze-driven Pretraining of Self-attention Networks", "XLNet: Generalized Autoregressive Pretraining for Language Understanding"], "answer_arxiv_id": ["1903.07785", "1906.08237"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_193"} +{"question": "Are there any studies that discuss about learning 'subspace juntas', functions of an unknown low-dimensional subspace?", "answer": ["Structure from Local Optima: Learning Subspace Juntas via Higher Order PCA"], "answer_arxiv_id": ["1108.3329v3"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_test_194"} +{"question": "Which work attempted to explore the inner modeling of hierarchical Transformer-based backbone?", "answer": ["An Efficient Spatio-Temporal Pyramid Transformer for Action Detection"], "answer_arxiv_id": ["2207.10448"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_195"} +{"question": "Which works on deep learning are primarily used as losses to train generative models?", "answer": ["Towards Principled Methods for Training Generative Adversarial Networks"], "answer_arxiv_id": ["1701.04862"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_test_196"} +{"question": "Which references discussed the empirical observation of oversmoothing issue in attention-based GNNs such as GATs or transformers?", "answer": ["A Survey on Oversmoothing in Graph Neural Networks", "Revisiting Over-smoothing in BERT from the Perspective of Graph"], "answer_arxiv_id": ["2303.10993", "2202.08625"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_test_197"} +{"question": "Which researchers proposed altering the memory-computation trade-off of the neural architecture for improving computational speed in neural scene representations?", "answer": ["DeRF: Decomposed Radiance Fields", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "FastNeRF: High-Fidelity Neural Rendering at 200FPS", "Plenoxels: Radiance Fields without Neural Networks", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction"], "answer_arxiv_id": ["2011.12490", "2103.13744", "2103.10380", "2112.05131", "2111.11215"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_198"} +{"question": "What are the recent works that employed clustering for personalized Federated Learning?", "answer": ["Federated learning with hierarchical clustering of local updates to improve training on non-IID data", "Clustered Federated Learning: Model-Agnostic Distributed Multi-Task Optimization under Privacy Constraints", "TiFL: A Tier-based Federated Learning System"], "answer_arxiv_id": ["2004.11791v2", "1910.01991", "2001.09249"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_test_199"} +{"question": "What research demonstrated that the Felzenswalb algorithm can generate useful geometric segment clusters in the ScanNet dataset?", "answer": ["ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes"], "answer_arxiv_id": ["1702.04405"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_test_200"} +{"question": "Which studies showed the success of Diffusion Models in the field of image synthesis?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Variational Diffusion Models", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2105.05233", "2107.00630", "2006.11239"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_test_201"} +{"question": "Which researches focused on batchifying queries in few-shot settings for LLMs?", "answer": ["Batch Prompting: Efficient Inference with Large Language Model APIs"], "answer_arxiv_id": ["2301.08721"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_202"} +{"question": "What are the papers that applied zero/few-shot learning using large language models?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2204.02311"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_test_203"} +{"question": "Which works have achieved impressive results in the field of image generation using Diffusion Probabilistic Models?", "answer": ["Denoising Diffusion Probabilistic Models", "Diffusion Models in Vision: A Survey"], "answer_arxiv_id": ["2006.11239", "2209.04747"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_test_204"} +{"question": "Which work resulted in the creation of the BLOOM model and ROOTS corpus as part of open science community initiatives?", "answer": ["BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset"], "answer_arxiv_id": ["2211.05100", "2303.03915"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_205"} +{"question": "Could you detail any works about answering questions given counterfactual conditions?", "answer": ["IfQA: A Dataset for Open-domain Question Answering under Counterfactual\n Presuppositions", "WIQA: A dataset for \"What if...\" reasoning over procedural text", "TIMEDIAL: Temporal Commonsense Reasoning in Dialog"], "answer_arxiv_id": ["2305.14010", "1909.04739", "2106.04571"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_206"} +{"question": "Which works focused on leveraging the internal states of LLMs to study hallucinated content?", "answer": ["A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of\n LLMs by Validating Low-Confidence Generation", "Attention Satisfies: A Constraint-Satisfaction Lens on Factual Errors of\n Language Models", "The Internal State of an LLM Knows When It's Lying"], "answer_arxiv_id": ["2307.03987", "2309.15098", "2304.13734"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_test_207"} +{"question": "Can you provide references for grouping-based methods of 3D instance segmentation?", "answer": ["PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation", "Hierarchical Aggregation for 3D Instance Segmentation", "Instance Segmentation in 3D Scenes using Semantic Superpoint Tree\n Networks", "MaskGroup: Hierarchical Point Grouping and Masking for 3D Instance\n Segmentation", "SoftGroup for 3D Instance Segmentation on Point Clouds", "3D Instances as 1D Kernels", "ISBNet: a 3D Point Cloud Instance Segmentation Network with\n Instance-aware Sampling and Box-aware Dynamic Convolution"], "answer_arxiv_id": ["2004.01658", "2108.02350", "2108.07478", "2203.14662", "2203.01509", "2207.07372", "2303.00246"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_test_208"} +{"question": "In which studies has the focus been on translating natural language proofs into formal representations?", "answer": ["MiniF2F: a cross-system benchmark for formal Olympiad-level mathematics", "NaturalProofs: Mathematical Theorem Proving in Natural Language"], "answer_arxiv_id": ["2109.00110", "2104.01112"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_test_209"} +{"question": "Which works extend the concept of estimating a transformation between two point clouds into finding an ideal set of primitives to represent a shape collection?", "answer": ["Learning elementary structures for 3D shape generation and matching"], "answer_arxiv_id": ["1908.04725"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_test_210"} +{"question": "Which works achieved impressive results in scene understanding using multi-modal models?", "answer": ["EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action\n Recognition"], "answer_arxiv_id": ["1908.08498"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_211"} +{"question": "Who are the researchers that attempted to close the gap between QM calculations and ML potentials?", "answer": ["SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects", "OrbNet: Deep Learning for Quantum Chemistry Using Symmetry-Adapted Atomic-Orbital Features", "Finding Density Functionals with Machine Learning", "Ab-Initio Potential Energy Surfaces by Pairing GNNs with Neural Wave Functions", "Generalizing Neural Wave Functions", "Sampling-free Inference for Ab-Initio Potential Energy Surface Networks"], "answer_arxiv_id": ["2105.00304", "2007.08026", "1112.5441", "2110.05064", "2302.04168", "2205.14962"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_test_212"} +{"question": "Any references that applied DDPM based on PVCNNs on the point-voxel representation of 3D shapes?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion"], "answer_arxiv_id": ["2104.03670"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_test_213"} +{"question": "What papers discussed multi-modal models?", "answer": ["EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action\n Recognition", "Multi-Modal Fusion Transformer for End-to-End Autonomous Driving", "SNE-RoadSeg: Incorporating Surface Normal Information into Semantic\n Segmentation for Accurate Freespace Detection"], "answer_arxiv_id": ["1908.08498", "2104.09224", "2008.11351"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_214"} +{"question": "Which paper provides a thorough discussion on quantisation-aware training?", "answer": ["A White Paper on Neural Network Quantization"], "answer_arxiv_id": ["2106.08295"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_test_215"} +{"question": "What works propose the usage of vector-quantized variational autoencoder and point-based diffusion model in the context of LiDAR scene generation?", "answer": ["UltraLiDAR: Learning Compact Representations for LiDAR Completion and\n Generation", "Learning to Generate Realistic LiDAR Point Clouds"], "answer_arxiv_id": ["2311.01448", "2209.03954"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_test_216"} +{"question": "What papers proposed IL+RL methods that initialize policy search methods with policies trained via behavioral cloning?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations"], "answer_arxiv_id": ["1709.10087"], "source_meta": {"published_time": "20220405"}, "qid": "AutoScholarQuery_test_217"} +{"question": "Could you provide some studies about LiDAR perception techniques using active learning and domain adaptation methods?", "answer": ["Label-Efficient Point Cloud Semantic Segmentation: An Active Learning Approach", "A Survey on Deep Domain Adaptation for LiDAR Perception"], "answer_arxiv_id": ["2101.06931", "2106.02377"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_test_218"} +{"question": "Which studies used deep learning in vision tasks like classification and object detection?", "answer": ["Reconstruction of 3D Porous Media From 2D Slices"], "answer_arxiv_id": ["1901.10233"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_219"} +{"question": "Which works construct a generalized decoding model capable of predicting pixel-level segmentation and language tokens?", "answer": ["Generalized Decoding for Pixel, Image, and Language", "Segment Everything Everywhere All at Once"], "answer_arxiv_id": ["2212.11270", "2304.06718"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_220"} +{"question": "What works showed that tighter dynamic regret rates are possible in non-stationary multi-armed bandits?", "answer": ["A New Look at Dynamic Regret for Non-Stationary Stochastic Bandits", "Tracking Most Significant Arm Switches in Bandits"], "answer_arxiv_id": ["2201.06532", "2112.13838"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_test_221"} +{"question": "What work combines an original NeRF with an edited NeRF generated using text-based SDS?", "answer": ["Vox-E: Text-guided Voxel Editing of 3D Objects"], "answer_arxiv_id": ["2303.12048"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_222"} +{"question": "What works establish the first connection between replicability and differential privacy in the context of PAC learning?", "answer": ["User-Level Private Learning via Correlated Sampling"], "answer_arxiv_id": ["2110.11208"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_223"} +{"question": "Could you provide me some works that focus on specific applications such as dialogue, structured knowledge grounding, or chain-of-thought reasoning?", "answer": ["OpenAssistant Conversations -- Democratizing Large Language Model\n Alignment", "UnifiedSKG: Unifying and Multi-Tasking Structured Knowledge Grounding\n with Text-to-Text Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "The CoT Collection: Improving Zero-shot and Few-shot Learning of\n Language Models via Chain-of-Thought Fine-Tuning"], "answer_arxiv_id": ["2304.07327", "2201.05966", "2201.11903", "2305.14045"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_224"} +{"question": "What papers are about updating text generation metrics?", "answer": ["The GEM Benchmark: Natural Language Generation, its Evaluation and Metrics", "How Robust are Model Rankings : A Leaderboard Customization Approach for Equitable Evaluation", "Bidimensional Leaderboards: Generate and Evaluate Language Hand in Hand"], "answer_arxiv_id": ["2102.01672", "2106.05532", "2112.04139"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_test_225"} +{"question": "Which work provided upper and lower bounds on optimal regret based on a variant of the DEC?", "answer": ["Tight Guarantees for Interactive Decision Making with the Decision-Estimation Coefficient"], "answer_arxiv_id": ["2301.08215v1"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_226"} +{"question": "Which papers propose graph-based approaches for capturing longer-term dependencies in 3D human pose forecasting?", "answer": ["Learning Trajectory Dependencies for Human Motion Prediction", "MSR-GCN: Multi-Scale Residual Graph Convolution Networks for Human\n Motion Prediction", "Dynamic Multiscale Graph Neural Networks for 3D Skeleton-Based Human\n Motion Prediction", "Space-Time-Separable Graph Convolutional Network for Pose Forecasting", "Spatio-Temporal Gating-Adjacency GCN for Human Motion Prediction", "Multitask Non-Autoregressive Model for Human Motion Prediction", "Diverse Human Motion Prediction Guided by Multi-Level Spatial-Temporal\n Anchors"], "answer_arxiv_id": ["1908.05436", "2108.07152", "2003.08802", "2110.04573", "2203.01474", "2007.06426", "2302.04860"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_227"} +{"question": "Could you provide me with the works using unified maximum likelihood estimation and object retrieval to support various tasks?", "answer": ["Uni-Perceiver v2: A Generalist Model for Large-Scale Vision and\n Vision-Language Tasks", "Universal Instance Perception as Object Discovery and Retrieval"], "answer_arxiv_id": ["2211.09808", "2303.06674"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_228"} +{"question": "Which works classify samples as hard based on the presence of large gradient norm and large norm of error vectors?", "answer": ["Deep Learning on a Data Diet: Finding Important Examples Early in Training"], "answer_arxiv_id": ["2107.07075"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_test_229"} +{"question": "What work uses meta-learning methods in the field of knowledge editing?", "answer": ["Editing Factual Knowledge in Language Models", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2104.08164", "2110.11309"], "source_meta": {"published_time": "20230916"}, "qid": "AutoScholarQuery_test_230"} +{"question": "What are some studies that discuss non-linear front-door adjustment?", "answer": ["Kernel Methods for Causal Functions: Dose, Heterogeneous, and Incremental Response Curves"], "answer_arxiv_id": ["2010.04855"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_test_231"} +{"question": "What works apply the mentioned technique to the first-order methods?", "answer": ["Gradient Sliding for Composite Optimization", "Accelerated gradient sliding for structured convex optimization"], "answer_arxiv_id": ["1406.0919v2", "1609.04905"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_test_232"} +{"question": "What papers explored the combination of GANs and implicit neural representation for 3D-aware image synthesis?", "answer": ["GIRAFFE: Representing Scenes as Compositional Generative Neural Feature\n Fields", "StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation", "Efficient Geometry-aware 3D Generative Adversarial Networks", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis"], "answer_arxiv_id": ["2011.12100", "2112.11427", "2112.07945", "2012.00926"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_test_233"} +{"question": "Could you provide me some research that demonstrated zero-shot cross-lingual transfer?", "answer": ["A Neural Pairwise Ranking Model for Readability Assessment"], "answer_arxiv_id": ["2203.07450"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_234"} +{"question": "What papers discuss the application of non-Euclidean diffusion equation leading to a scheme with adaptive spatial derivatives?", "answer": ["Beltrami Flow and Neural Diffusion on Graphs"], "answer_arxiv_id": ["2110.09443"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_test_235"} +{"question": "Which paper used mutual information for invariant representation learning outside of RL?", "answer": ["Invariant Representations with Stochastically Quantized Neural Networks"], "answer_arxiv_id": ["2208.02656"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_236"} +{"question": "Which paper introduces DataComp, a benchmark for designing better pre-training datasets for CLIP?", "answer": ["DataComp: In search of the next generation of multimodal datasets"], "answer_arxiv_id": ["2304.14108"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_237"} +{"question": "What paper advances the representative Seq2Edit model, GECToR, by improving its multi-round correction?", "answer": ["Type-Driven Multi-Turn Corrections for Grammatical Error Correction"], "answer_arxiv_id": ["2203.09136"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_test_238"} +{"question": "Which works discuss provably efficient algorithms in the context of function approximation in linear mixture MDPs?", "answer": ["Provably Efficient Exploration in Policy Optimization", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Provably Efficient Reinforcement Learning for Discounted MDPs with Feature Mapping"], "answer_arxiv_id": ["1912.05830", "2012.08507", "2006.13165"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_test_239"} +{"question": "Could you provide me some critiques on benchmarks in NLP about annotation artifacts?", "answer": ["Annotation Artifacts in Natural Language Inference Data"], "answer_arxiv_id": ["1803.02324"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_240"} +{"question": "What studies presented method for selecting single target-state pair with stochastic batch acquisition in a BOED setting?", "answer": ["Interventions, Where and How? Experimental Design for Causal Models at Scale"], "answer_arxiv_id": ["2203.02016"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_test_241"} +{"question": "Which work introduced API specially designed for content moderation in vast amounts of data?", "answer": ["A Holistic Approach to Undesired Content Detection in the Real World"], "answer_arxiv_id": ["2208.03274"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_test_242"} +{"question": "What is the research paper that discusses instruction tuning?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_243"} +{"question": "Which work proposed the method of calculating the inner integral of Sampled Policy Gradients (SPG) using a Q-network?", "answer": ["Mean Actor-Critic"], "answer_arxiv_id": ["1709.00503"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_test_244"} +{"question": "Which work applied perturbations to word embeddings in the NLP domain?", "answer": ["Adversarial Training Methods for Semi-Supervised Text Classification"], "answer_arxiv_id": ["1605.07725"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_test_245"} +{"question": "What studies have considered extending beta diffusion to encompass the exponential family?", "answer": ["Variational Inference: A Review for Statisticians"], "answer_arxiv_id": ["1601.00670"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_test_246"} +{"question": "What papers explored metrics that either emphasize a single dimension or lack human relevance?", "answer": ["Asking and Answering Questions to Evaluate the Factual Consistency of\n Summaries", "GRADE: Automatic Graph-Enhanced Coherence Metric for Evaluating\n Open-Domain Dialogue Systems", "USR: An Unsupervised and Reference Free Evaluation Metric for Dialog\n Generation", "Towards a Unified Multi-Dimensional Evaluator for Text Generation"], "answer_arxiv_id": ["2004.04228", "2010.03994", "2005.00456", "2210.07197"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_247"} +{"question": "Are there any studies that provide convergence guarantees for PFL with the bounded gradients assumption on heterogeneous data?", "answer": ["On the Convergence of FedAvg on Non-IID Data"], "answer_arxiv_id": ["1907.02189"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_test_248"} +{"question": "Can you give examples of research that altered the parameter of the transition function (like gravity strength or the weight of the ball) for RL agent evaluation?", "answer": ["Assessing Generalization in Deep Reinforcement Learning", "Robust Adversarial Reinforcement Learning"], "answer_arxiv_id": ["1810.12282", "1703.02702"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_test_249"} +{"question": "What subsequent work improved the complexities of the method developed by bib.bib60?", "answer": ["Accelerating Stochastic Composition Optimization"], "answer_arxiv_id": ["1607.07329"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_250"} +{"question": "What works proposed methods based on the primary form of the Optimal Transport problem?", "answer": ["On Scalable and Efficient Computation of Large Scale Optimal Transport", "Large-Scale Optimal Transport via Adversarial Training with Cycle-Consistency"], "answer_arxiv_id": ["1905.00158", "2003.06635"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_251"} +{"question": "Could you provide me some works that focus on aligning the learning trajectories for emulating long-range training dynamics of real data when performing dataset condensation?", "answer": ["Dataset Distillation by Matching Training Trajectories"], "answer_arxiv_id": ["2203.11932"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_test_252"} +{"question": "Which studies succeeded in addressing regression tasks through deep learning?", "answer": ["MPIIGaze: Real-World Dataset and Deep Appearance-Based Gaze Estimation", "It’s Written All Over Your Face: Full-Face Appearance-Based Gaze Estimation"], "answer_arxiv_id": ["1711.09017", "1611.08860"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_test_253"} +{"question": "What works exist that provide corrective measures for issues concerning the RP gradient?", "answer": ["Gradients are Not All You Need"], "answer_arxiv_id": ["2111.05803"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_254"} +{"question": "What works have been done to extract answers from question-specific subgraphs generated with text corpora?", "answer": ["Open Domain Question Answering Using Early Fusion of Knowledge Bases and Text", "PullNet: Open Domain Question Answering with Iterative Retrieval on Knowledge Bases and Text"], "answer_arxiv_id": ["1809.00782", "1904.09537"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_255"} +{"question": "Which works propose methods to handle continuous treatments in the back-door adjustment?", "answer": ["Learning Counterfactual Representations for Estimating Individual Dose-Response Curves", "Nonparametric methods for doubly robust estimation of continuous treatment effects", "Double Debiased Machine Learning Nonparametric Inference with Continuous Treatments", "Automatic Debiased Machine Learning of Causal and Structural Effects"], "answer_arxiv_id": ["1902.00981", "1507.00747", "2004.03036", "1809.05224"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_test_256"} +{"question": "Can you identify any papers that analysed the use of target networks with linear function approximation, needed in theoretical properties of target networks?", "answer": ["Breaking the Deadly Triad with a Target Network"], "answer_arxiv_id": ["2101.08862"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_test_257"} +{"question": "Which works adapt diffusion models to condition generation on protein pockets?", "answer": ["Equivariant 3D-Conditional Diffusion Models for Molecular Linker Design", "Structure-based Drug Design with Equivariant Diffusion Models", "DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking"], "answer_arxiv_id": ["2210.05274", "2210.13695", "2210.01776"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_258"} +{"question": "What are some of the works that provide a detailed analysis of the stationary distribution of SGD iterations with a constant learning rate?", "answer": ["Stochastic Gradient Descent as Approximate Bayesian Inference"], "answer_arxiv_id": ["1704.04289"], "source_meta": {"published_time": "20220924"}, "qid": "AutoScholarQuery_test_259"} +{"question": "What works measure bias based on the difference in ground-truth object-group co-occurrences in the training set and test set co-occurrences predicted by a model?", "answer": ["Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints"], "answer_arxiv_id": ["1707.09457"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_test_260"} +{"question": "What works are there on incorporating interactive discussions and human collaboration into LLM-based evaluation methodologies?", "answer": ["ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate", "Collaborative Evaluation: Exploring the Synergy of Large Language Models\n and Humans for Open-ended Generation Evaluation"], "answer_arxiv_id": ["2308.07201", "2310.19740"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_261"} +{"question": "What works made advancements in achieving certified robustness with state-of-the-art randomized ablation?", "answer": ["Robustness Certificates for Sparse Adversarial Attacks by Randomized\n Ablation", "Almost Tight L0-norm Certified Robustness of Top-k Predictions against\n Adversarial Perturbations"], "answer_arxiv_id": ["1911.09272", "2011.07633"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_262"} +{"question": "Which works perform multi-view refinement with flow or dense features in Structure-from-Motion (SfM)?", "answer": ["Multi-View Optimization of Local Feature Geometry", "Pixel-Perfect Structure-from-Motion with Featuremetric Refinement"], "answer_arxiv_id": ["2003.08348", "2108.08291"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_263"} +{"question": "Could you give me examples of research achieving progress in sentiment analysis using multi-modal models?", "answer": ["Self-attention fusion for audiovisual emotion recognition with incomplete data", "MOSI: Multimodal Corpus of Sentiment Intensity and Subjectivity Analysis in Online Opinion Videos", "Gated Mechanism for Attention Based Multimodal Sentiment Analysis"], "answer_arxiv_id": ["2201.11095v1", "1606.06259v2", "2003.01043"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_264"} +{"question": "Which works propose techniques for improving the quality of image-text datasets in multimodal networks?", "answer": ["Filtering, Distillation, and Hard Negatives for Vision-Language Pre-Training", "Less is More: Removing Text-regions Improves CLIP Training Efficiency and Robustness", "T-MARS: Improving Visual Representations by Circumventing Text Feature Learning", "SemDeDup: Data-efficient learning at web-scale through semantic deduplication"], "answer_arxiv_id": ["2301.02280", "2305.05095", "2307.03132v2", "2303.09540"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_265"} +{"question": "In what works can I find large-scale unsupervised pre-training on unstructured text for multilingual corpora?", "answer": ["BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "What Language Model to Train if You Have One Million GPU Hours?", "The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset", "MADLAD-400: A Multilingual And Document-Level Large Audited Dataset", "LLM-powered Data Augmentation for Enhanced Cross-lingual Performance"], "answer_arxiv_id": ["2211.05100", "2210.15424", "2303.03915", "2309.04662", "2305.14288"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_266"} +{"question": "Which papers focus on broader applications of NeRF, including generative modeling, video synthesis, and scene editing?", "answer": ["GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images", "VoLux-GAN: A Generative Model for 3D Face Synthesis with HDRI Relighting", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Tensor4D : Efficient Neural 4D Decomposition for High-fidelity Dynamic\n Reconstruction and Rendering", "Neural Radiance Flow for 4D View Synthesis and Video Processing", "Editing Conditional Radiance Fields", "NeRF-Editing: Geometry Editing of Neural Radiance Fields"], "answer_arxiv_id": ["2209.11163", "2201.04873", "2011.13084", "2211.11610", "2012.09790", "2105.06466", "2205.04978"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_267"} +{"question": "What studies considered to initialise the vertices of the graphs with random labels in graph embedding?", "answer": ["The Surprising Power of Graph Neural Networks with Random Node Initialization", "Random Features Strengthen Graph Neural Networks"], "answer_arxiv_id": ["2010.01179v2", "2002.03155"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_test_268"} +{"question": "What studies removed the strong assumptions about knowledge of non-stationarity in non-contextual bandits?", "answer": ["A New Algorithm for Non-stationary Contextual Bandits: Efficient, Optimal, and Parameter-free"], "answer_arxiv_id": ["1902.00980v3"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_test_269"} +{"question": "What is the standout example of point-based methodologies that transformed the direct processing of Point Cloud?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation"], "answer_arxiv_id": ["1612.00593"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_270"} +{"question": "Could you provide me some works about studying a special case of fractionally subadditive valuations?", "answer": ["The Fair Division of Hereditary Set Systems"], "answer_arxiv_id": ["1812.09561"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_271"} +{"question": "Which works explored ad-hoc teamwork and zero-shot coordination in AI?", "answer": ["“Other-Play” for Zero-Shot Coordination", "A New Formalism, Method and Open Issues for Zero-Shot Coordination", "Equivariant Networks for Zero-Shot Coordination"], "answer_arxiv_id": ["2003.02979", "2106.06613", "2210.12124"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_272"} +{"question": "What works explored the constraint of Neural Differential Equations by expensive training and prediction times?", "answer": ["Augmented Neural ODEs", "Learning differential equations that are easy to solve", "STEER: Simple Temporal Regularization For Neural ODEs", "Opening the Blackbox: Accelerating Neural Differential Equations by Regularizing Internal Solver Heuristics"], "answer_arxiv_id": ["1904.01681", "2007.04504", "2006.10711", "2105.03918"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_test_273"} +{"question": "What work is the only public simulator that supports differentiable simulation for in-graph acceleration?", "answer": ["Imagining The Road Ahead: Multi-Agent Trajectory Prediction via Differentiable Simulation"], "answer_arxiv_id": ["2104.11212"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_test_274"} +{"question": "What studies work on body motion conditioned on text descriptions?", "answer": ["FLAME: Free-form Language-based Motion Synthesis & Editing", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE", "Synthesizing Long-Term Human Motions with Diffusion Models via Coherent\n Sampling", "TEMOS: Generating diverse human motions from textual descriptions", "Synthesis of Compositional Animations from Textual Descriptions"], "answer_arxiv_id": ["2209.00349", "2104.05670", "2308.01850", "2204.14109", "2103.14675"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_test_275"} +{"question": "Which study examines users' discomfort or concern due to the lack of responsibility in LLMs' recommendations for emotional support response?", "answer": ["The Typing Cure: Experiences with Large Language Model Chatbots for\n Mental Health Support"], "answer_arxiv_id": ["2401.14362"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_test_276"} +{"question": "Which reference is about the TorchGeo, a Python library for the integration of remote sensing datasets into the PyTorch deep learning ecosystem?", "answer": ["TorchGeo: Deep Learning With Geospatial Data"], "answer_arxiv_id": ["2111.08872"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_test_277"} +{"question": "Which studies have investigated factors like loss function, surrogate gradient estimation, and batch normalization that affect the learning behavior in direct training of SNNs?", "answer": ["Temporal Efficient Training of Spiking Neural Network via Gradient Re-weighting", "Membrane Potential Batch Normalization for Spiking Neural Networks"], "answer_arxiv_id": ["2202.11946", "2308.08359"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_test_278"} +{"question": "Could you provide me works that exemplified an asymmetric discussion mechanism with different LLMs and using a weighted voting mechanism?", "answer": ["ReConcile: Round-Table Conference Improves Reasoning via Consensus among\n Diverse LLMs"], "answer_arxiv_id": ["2309.13007"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_279"} +{"question": "What references proposed methods to discover the global semantic structure underlying the whole dataset, shed light on graph contrastive learning?", "answer": ["Self-supervised Graph-level Representation Learning with Local and Global Structure", "Omni-Granular Ego-Semantic Propagation for Self-Supervised Graph Representation Learning"], "answer_arxiv_id": ["2106.04113", "2205.15746"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_test_280"} +{"question": "Could you mention research studies that utilized features for minimizing the surrogate models in dataset distillation?", "answer": ["CAFE: Learning to Condense Dataset by Aligning Features"], "answer_arxiv_id": ["2203.01531"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_281"} +{"question": "Could you name the papers where the research focus on the L2 convergence rate for KRR, and it can be easily extended to [ℋ]γ,γ≥0 convergence rate?", "answer": ["Sobolev Norm Learning Rates for Regularized Least-Squares Algorithms", "Optimal Rates for Spectral Algorithms with Least-Squares Regression over Hilbert Spaces"], "answer_arxiv_id": ["1702.07254", "1801.06720"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_test_282"} +{"question": "What work used MCP when evaluating InstructGPT?", "answer": ["Can Large Language Models Reason about Medical Questions?"], "answer_arxiv_id": ["2207.08143"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_test_283"} +{"question": "What studies in medical VLMs use diffusion-based methods in report-to-CXR generation task?", "answer": ["RoentGen: Vision-Language Foundation Model for Chest X-ray Generation"], "answer_arxiv_id": ["2211.12737"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_test_284"} +{"question": "Which works initially proposed generating test cases and corresponding accurate assert statements with Transformer models?", "answer": ["Unit Test Case Generation with Transformers and Focal Context", "Generating Accurate Assert Statements for Unit Test Cases using Pretrained Transformers"], "answer_arxiv_id": ["2009.05617", "2009.05634"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_test_285"} +{"question": "Which paper carried out an extensive empirical study which shows significant improvement in model performance by implementing SWAG with multiple randomly initialized models?", "answer": ["Bayesian Deep Learning and a Probabilistic Perspective of Generalization"], "answer_arxiv_id": ["2002.08791"], "source_meta": {"published_time": "20220924"}, "qid": "AutoScholarQuery_test_286"} +{"question": "What works have developed algorithms to solve the sparse coding problem?", "answer": ["Online Learning for Matrix Factorization and Sparse Coding", "New Algorithms for Learning Incoherent and Overcomplete Dictionaries", "More algorithms for provable dictionary learning", "Simple, Efficient, and Neural Algorithms for Sparse Coding"], "answer_arxiv_id": ["0908.0050", "1308.6273", "1401.0579", "1503.00778"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_test_287"} +{"question": "What studies used concept activation vectors and multimodal models to annotate concepts for CBMs?", "answer": ["Post-hoc Concept Bottleneck Models", "Label-free Concept Bottleneck Models", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2205.15480", "2304.06129", "2103.00020"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_test_288"} +{"question": "Any works that discuss learning the GAN fingerprints towards image attribution?", "answer": ["Attributing Fake Images to GANs: Learning and Analyzing GAN Fingerprints"], "answer_arxiv_id": ["1811.08180"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_289"} +{"question": "What works used pretrained GAN generators and text encoders to optimize images based on textual prompts?", "answer": ["Paint by Word", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators"], "answer_arxiv_id": ["2103.10951", "2108.00946"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_test_290"} +{"question": "What research proposed GroundingSAM, a combination of GroundingDINO and SAM for generating segmentation masks?", "answer": ["Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection", "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2303.05499", "2401.14159"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_test_291"} +{"question": "What research is done about 3D diffusion models involving meshes?", "answer": ["3DGen: Triplane Latent Diffusion for Textured Mesh Generation", "MeshDiffusion: Score-based Generative 3D Mesh Modeling", "Controllable Mesh Generation Through Sparse Latent Point Diffusion\n Models"], "answer_arxiv_id": ["2303.05371", "2303.08133", "2303.07938"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_test_292"} +{"question": "Which papers deal with the utilization of context to rewrite the conversation into a standalone query in CQR models?", "answer": ["Conversational Question Reformulation via Sequence-to-Sequence\n Architectures and Pretrained Language Models", "ConvGQR: Generative Query Reformulation for Conversational Search"], "answer_arxiv_id": ["2004.01909", "2305.15645"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_test_293"} +{"question": "In what work was PoseNet’s innovative transfer learning first introduced?", "answer": ["PoseNet: A Convolutional Network for Real-Time 6-DOF Camera\n Relocalization"], "answer_arxiv_id": ["1505.07427"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_294"} +{"question": "Which studies describe model structures that implicitly generate reasoning processes?", "answer": ["Program Induction by Rationale Generation : Learning to Solve and\n Explain Algebraic Word Problems", "TAT-QA: A Question Answering Benchmark on a Hybrid of Tabular and\n Textual Content in Finance", "Answering Numerical Reasoning Questions in Table-Text Hybrid Contents\n with Graph-based Encoder and Tree-based Decoder", "Chaining Simultaneous Thoughts for Numerical Reasoning", "ELASTIC: Numerical Reasoning with Adaptive Symbolic Compiler"], "answer_arxiv_id": ["1705.04146", "2105.07624", "2209.07692", "2211.16482", "2210.10105"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_295"} +{"question": "What studies investigated phenomenon on invariant representations in the context of algorithmic fairness?", "answer": ["Censoring Representations with an Adversary", "Mitigating Unwanted Biases with Adversarial Learning", "Conditional Learning of Fair Representations"], "answer_arxiv_id": ["1511.05897", "1801.07593", "1910.07162"], "source_meta": {"published_time": "20201219"}, "qid": "AutoScholarQuery_test_296"} +{"question": "Any studies showcase the potential of non-attention architectures in language modeling?", "answer": ["Hungry Hungry Hippos: Towards Language Modeling with State Space Models", "Hyena Hierarchy: Towards Larger Convolutional Language Models", "RWKV: Reinventing RNNs for the Transformer Era", "Hierarchically Gated Recurrent Neural Network for Sequence Modeling"], "answer_arxiv_id": ["2212.14052", "2302.10866v3", "2305.13048", "2311.04823"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_test_297"} +{"question": "What papers proposed iterative methods for transferable adversarial attacks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Boosting Adversarial Attacks with Momentum", "Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks", "Enhancing the Transferability of Adversarial Attacks through Variance Tuning", "Improving Transferability of Adversarial Examples with Input Diversity", "On Improving Adversarial Transferability of Vision Transformers", "Cross-Modal Transferable Adversarial Attacks from Images to Videos"], "answer_arxiv_id": ["1412.6572", "1710.06081", "1908.06281", "2103.15571", "1803.06978", "2106.04169", "2112.05379"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_test_298"} +{"question": "Which papers proposed datasets for open domain question answering (QA) for English and other languages?", "answer": ["TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages", "XOR QA: Cross-lingual Open-Retrieval Question Answering", "MIA 2022 Shared Task: Evaluating Cross-lingual Open-Retrieval Question Answering for 16 Diverse Languages", "MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering", "Mr. TYDI: A Multi-lingual Benchmark for Dense Retrieval"], "answer_arxiv_id": ["2003.05002", "2010.11856v3", "2207.00758", "2007.15207", "2108.08787"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_test_299"} +{"question": "What research work focuses on a criterion, named difference of confidences (DoC), which estimates and reflects model accuracy?", "answer": ["Predicting with Confidence on Unseen Distributions"], "answer_arxiv_id": ["2107.03315"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_test_300"} +{"question": "What studies introduced the mask-reconstruction paradigm for unimodal pretraining?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_test_301"} +{"question": "Are there any studies that suggest the algorithm they propose cannot be directly applied to model-free reinforcement learning settings?", "answer": ["Unified Algorithms for RL with Decision-Estimation Coefficients: No-Regret, PAC, and Reward-Free Learning"], "answer_arxiv_id": ["2209.11745"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_302"} +{"question": "Are there any studies that used meta-learning for updating the knowledge in LLMs through varying their parameters?", "answer": ["Editable Neural Networks", "Fast Model Editing at Scale", "Editing Factual Knowledge in Language Models"], "answer_arxiv_id": ["2004.00345", "2110.11309", "2104.08164"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_303"} +{"question": "Which works were proposed that consider networks of weights matrices with bounded norms?", "answer": ["Spectrally-normalized margin bounds for neural networks", "Robust Large Margin Deep Neural Networks"], "answer_arxiv_id": ["1706.08498", "1605.08254"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_test_304"} +{"question": "Which study proposed the method PixSfM for feature-metric keypoint adjustment and bundle adjustment in SfM?", "answer": ["Pixel-Perfect Structure-from-Motion with Featuremetric Refinement"], "answer_arxiv_id": ["2108.08291"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_305"} +{"question": "What papers studied unbalanced optimal transport (UOT) using methods that estimate UOT potentials on discrete space?", "answer": ["Neural Unbalanced Optimal Transport via Cycle-Consistent Semi-Couplings", "On Unbalanced Optimal Transport: An Analysis of Sinkhorn Algorithm", "Unbalanced Optimal Transport through Non-negative Penalized Linear Regression", "Unbalanced minibatch Optimal Transport; applications to Domain Adaptation"], "answer_arxiv_id": ["2209.15621", "2002.03293", "2106.04145", "2103.03606"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_306"} +{"question": "What works used open-loop imitation learning for predicting the behavior of the ego vehicle in autonomous driving?", "answer": ["End to End Learning for Self-Driving Cars", "PRECOG: PREdiction Conditioned On Goals in Visual Multi-Agent Settings", "End-to-end Driving via Conditional Imitation Learning", "SafetyNet: Safe planning for real-world self-driving vehicles using machine-learned policies", "Learning by cheating"], "answer_arxiv_id": ["1604.07316", "1905.01296", "1710.02410", "2109.13602", "1912.12294"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_test_307"} +{"question": "Which works formulates medical VQA datasets based on MIMIC-CXR?", "answer": ["Interpretable Medical Image Visual Question Answering via Multi-Modal Relationship Graph Learning"], "answer_arxiv_id": ["2302.09636"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_test_308"} +{"question": "Can you name some studies that propose different metrics to prune networks at initialization?", "answer": ["Picking Winning Tickets Before Training by Preserving Gradient Flow", "Pruning neural networks without any data by iteratively conserving synaptic flow", "Progressive Skeletonization: Trimming more fat from a network at initialization", "PHEW : Constructing Sparse Networks that Learn Fast and Generalize Well Without Training Data", "Prospect Pruning: Finding Trainable Weights at Initialization using Meta-Gradients"], "answer_arxiv_id": ["2002.07376", "2006.05467", "2006.09081", "2010.11354", "2202.08132"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_test_309"} +{"question": "What papers reviewed when focusing on methods for quantifying aleatoric segmentation uncertainty?", "answer": ["A Probabilistic U-Net for Segmentation of Ambiguous Images", "Stochastic Segmentation Networks: Modelling Spatially Correlated Aleatoric Uncertainty"], "answer_arxiv_id": ["1806.05034", "2006.06015"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_test_310"} +{"question": "What papers outlined the studies on how to select the firing threshold to cover all the features in an ANN?", "answer": ["Theory and Tools for the Conversion of Analog to Spiking Convolutional Neural Networks"], "answer_arxiv_id": ["1612.04052"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_test_311"} +{"question": "What papers express key concerns about the validity of performance measurements obtained with various NLP benchmarks?", "answer": ["What Will it Take to Fix Benchmarking in Natural Language Understanding?"], "answer_arxiv_id": ["2104.02145"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_312"} +{"question": "Please provide papers advocating the use of scratchpads in Large Language Models.", "answer": ["Show Your Work: Scratchpads for Intermediate Computation with Language\n Models"], "answer_arxiv_id": ["2112.00114"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_test_313"} +{"question": "Which studies look at the convergence rates for PMD-type methods for Lipschitz and smooth policies?", "answer": ["Policy Optimization with Stochastic Mirror Descent", "Bregman Gradient Policy Optimization"], "answer_arxiv_id": ["1906.10462v5", "2106.12112"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_314"} +{"question": "What papers discuss mitigation methods for temporal adaptation?", "answer": ["Time-Aware Language Models as Temporal Knowledge Bases", "TemporalWiki: A Lifelong Benchmark for Training and Evaluating Ever-Evolving Language Models", "Towards Continual Knowledge Learning of Language Models", "Plug-and-Play Adaptation for Continuously-updated QA"], "answer_arxiv_id": ["2106.15110", "2204.14211", "2110.03215", "2204.12785"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_test_315"} +{"question": "What are the papers that introduce various learning objectives to tackle data heterogeneity in Federated Learning?", "answer": ["Federated Optimization in Heterogeneous Networks", "Federated Visual Classification with Real-World Data Distribution", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Model-Contrastive Federated Learning"], "answer_arxiv_id": ["1812.06127", "2003.08082", "1910.06378", "2103.16257"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_test_316"} +{"question": "Could you provide me some works that use complex-valued activations with real-valued weights and a gating mechanism?", "answer": ["Neuronal Synchrony in Complex-Valued Deep Networks", "Complex-Valued Autoencoders for Object Discovery"], "answer_arxiv_id": ["1312.6115", "2204.02075"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_317"} +{"question": "Which studies highlight the benefit of capturing long-distance relations in Graph Neural Networks (GNNs) by stacking more feature aggregation layers or unrolling various fixed point iterations?", "answer": ["Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "Implicit Graph Neural Networks", "Towards Deeper Graph Neural Networks", "Simple and Deep Graph Convolutional Networks", "Training Graph Neural Networks with 1000 Layers", "A Unified View on Graph Neural Networks as Graph Signal Denoising", "Interpreting and Unifying Graph Neural Networks with An Optimization Framework"], "answer_arxiv_id": ["1810.05997", "2009.06211", "2007.09296", "2007.02133", "2106.07476", "2010.01777", "2101.11859"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_test_318"} +{"question": "Could you mention some studies about learning-based MVS approaches?", "answer": ["Deep Stereo using Adaptive Thin Volume Representation with Uncertainty\n Awareness", "Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo\n Matching", "Cost Volume Pyramid Based Depth Inference for Multi-View Stereo"], "answer_arxiv_id": ["1911.12012", "1912.06378", "1912.08329"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_test_319"} +{"question": "Which works opted to incorporate the backbone into the network's training process to form an end-to-end TAD framework?", "answer": ["Learning Salient Boundary Feature for Anchor-free Temporal Action\n Localization", "TALLFormer: Temporal Action Localization with a Long-memory Transformer", "An Efficient Spatio-Temporal Pyramid Transformer for Action Detection", "Re^2TAL: Rewiring Pretrained Video Backbones for Reversible Temporal\n Action Localization"], "answer_arxiv_id": ["2103.13137", "2204.01680", "2207.10448", "2211.14053"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_320"} +{"question": "Which works are referred when discussing previous benchmarks for solving math word problems?", "answer": ["Program Induction by Rationale Generation : Learning to Solve and\n Explain Algebraic Word Problems", "MathQA: Towards Interpretable Math Word Problem Solving with\n Operation-Based Formalisms", "Training Verifiers to Solve Math Word Problems", "CMATH: Can Your Language Model Pass Chinese Elementary School Math Test?"], "answer_arxiv_id": ["1705.04146", "1905.13319", "2110.14168", "2306.16636"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_test_321"} +{"question": "Are there works that use image groupings and pairs for disentanglement?", "answer": ["Weakly Supervised Disentanglement with Guarantees", "Weakly-Supervised Disentanglement Without Compromises"], "answer_arxiv_id": ["1910.09772", "2002.02886"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_322"} +{"question": "Could you provide me some examples of research that discusses the application of data augmentations in the latent space?", "answer": ["FreeLB: Enhanced Adversarial Training for Natural Language Understanding", "AdvAug: Robust Adversarial Augmentation for Neural Machine Translation", "DoubleMix: Simple Interpolation-Based Data Augmentation for Text\n Classification", "Text Smoothing: Enhance Various Data Augmentation Methods on Text\n Classification Tasks", "Controlled Text Generation for Data Augmentation in Intelligent\n Artificial Agents"], "answer_arxiv_id": ["1909.11764", "2006.11834", "2209.05297", "2202.13840", "1910.03487"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_test_323"} +{"question": "Which works make a formal equivalence between differential privacy and replicability for finite domains?", "answer": ["Stability is Stable: Connections between Replicability, Privacy, and Adaptive Generalization"], "answer_arxiv_id": ["2303.12921"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_324"} +{"question": "Could you find any works that proposed methods applicable to online non-parametric regression tasks?", "answer": ["A Chaining Algorithm for Online Nonparametric Regression", "Online Isotonic Regression", "Online Forecasting of Total-Variation-bounded Sequences", "Total Variation Classes Beyond 1d: Minimax Rates, and the Limitations of Linear Smoothers"], "answer_arxiv_id": ["1502.07697v2", "1603.04190", "1906.03364", "1605.08400"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_325"} +{"question": "Which studies introduced text-to-SQL datasets associated with MIMIC-III and eICU?", "answer": ["EHRSQL: A Practical Text-to-SQL Benchmark for Electronic Health Records"], "answer_arxiv_id": ["2301.07695"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_test_326"} +{"question": "Are there any works that used hypernetworks in Meta-SL?", "answer": ["Meta-Learning with Latent Embedding Optimization", "Meta Networks"], "answer_arxiv_id": ["1807.05960", "1703.00837"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_test_327"} +{"question": "Which researches have covered the attributes of non-functional requirements in terms of time/space performance?", "answer": ["Learning Performance-Improving Code Edits", "WizardCoder: Empowering Code Large Language Models with Evol-Instruct"], "answer_arxiv_id": ["2302.07867", "2306.08568"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_test_328"} +{"question": "Could you provide examples of works about certified defenses focused on unimodal models?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing", "Certified Robustness for Top-k Predictions against Adversarial\n Perturbations via Randomized Smoothing", "Robustness Certificates for Sparse Adversarial Attacks by Randomized\n Ablation", "Certified Defenses for Adversarial Patches", "SAFER: A Structure-free Approach for Certified Robustness to Adversarial\n Word Substitutions", "Certified Robustness to Adversarial Examples with Differential Privacy", "PointGuard: Provably Robust 3D Point Cloud Classification", "Certified Robustness to Text Adversarial Attacks by Randomized [MASK]", "PatchCleanser: Certifiably Robust Defense against Adversarial Patches\n for Any Image Classifier", "MultiGuard: Provably Robust Multi-label Classification against\n Adversarial Examples", "TextGuard: Provable Defense against Backdoor Attacks on Text\n Classification", "PointCert: Point Cloud Classification with Deterministic Certified\n Robustness Guarantees"], "answer_arxiv_id": ["1902.02918", "1912.09899", "1911.09272", "2003.06693", "2005.14424", "1802.03471", "2103.03046", "2105.03743", "2108.09135", "2210.01111", "2311.11225", "2303.01959"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_329"} +{"question": "What are the early efforts that expanded traditional KG representation learning methods for single-modal knowledge graphs?", "answer": ["Image-embodied Knowledge Representation Learning"], "answer_arxiv_id": ["1609.07028"], "source_meta": {"published_time": "20240723"}, "qid": "AutoScholarQuery_test_330"} +{"question": "Could you provide me the work that incorporated PEFT-based layer adaptation to the shared attention and FFN modules in transformers?", "answer": ["EdgeFormer: A Parameter-Efficient Transformer for On-Device Seq2seq\n Generation"], "answer_arxiv_id": ["2202.07959"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_331"} +{"question": "What papers proposed domain adaptation (DA) methods?", "answer": ["Domain-Adversarial Training of Neural Networks", "A review of domain adaptation without target labels"], "answer_arxiv_id": ["1505.07818", "1901.05335"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_332"} +{"question": "Which studies proposed the use of a seq2seq architecture and a progressive transformer for Sign Language Production (SLP)?", "answer": ["Progressive Transformers for End-to-End Sign Language Production"], "answer_arxiv_id": ["2004.14874"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_test_333"} +{"question": "What work stored manual edits in a memory module to amend the output of LLMs?", "answer": ["Memory-Based Model Editing at Scale"], "answer_arxiv_id": ["2206.06520"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_334"} +{"question": "What researches have been conducted about privacy-preserving learning with learning invariant representations?", "answer": ["Minimax Filter: Learning to Preserve Privacy from Inference Attacks", "Privacy-preserving Neural Representations of Text", "Adversarial Learning of Privacy-Preserving and Task-Oriented Representations"], "answer_arxiv_id": ["1610.03577", "1808.09408", "1911.10143"], "source_meta": {"published_time": "20201219"}, "qid": "AutoScholarQuery_test_335"} +{"question": "What research papers have explored multimodal tasks within UI contexts?", "answer": ["VUT: Versatile UI Transformer for Multi-Modal Multi-Task User Interface\n Modeling", "UIBert: Learning Generic Multimodal Representations for UI Understanding", "ActionBert: Leveraging User Actions for Semantic Understanding of User\n Interfaces"], "answer_arxiv_id": ["2112.05692", "2107.13731", "2012.12350"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_test_336"} +{"question": "What studies focus on the different techniques utilized to fine-tune the pre-trained models?", "answer": ["Scaling Instruction-Finetuned Language Models", "Training language models to follow instructions with human feedback", "Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally\n Across Scales and Tasks"], "answer_arxiv_id": ["2210.11416", "2203.02155", "1902.00751", "2106.09685", "2101.00190", "2104.08691", "2110.07602"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_test_337"} +{"question": "Which works discussed the relation between sinusoidal networks and networks with Fourier feature transformations?", "answer": ["Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains"], "answer_arxiv_id": ["2006.10739"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_test_338"} +{"question": "Could you provide me some works that enhance unimodal language models with retrieval?", "answer": ["Improving language models by retrieving from trillions of tokens", "Few-shot Learning with Retrieval Augmented Language Models"], "answer_arxiv_id": ["2112.04426", "2208.03299"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_test_339"} +{"question": "Which works have utilized instructional videos for online learning, especially in the medical field?", "answer": ["A Dataset for Medical Instructional Video Classification and Question Answering"], "answer_arxiv_id": ["2201.12888"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_test_340"} +{"question": "Who designed attentional GNNs by discretizing parabolic diffusion-type PDEs?", "answer": ["GRAND: Graph Neural Diffusion"], "answer_arxiv_id": ["2106.10934"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_test_341"} +{"question": "Which studies investigated zero-shot NAS methods aiming to reduce search costs?", "answer": ["Zero-Cost Proxies for Lightweight NAS", "Neural Architecture Search without Training"], "answer_arxiv_id": ["2101.08134v2", "2006.04647"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_test_342"} +{"question": "Which papers approach the studies about adversarial attacks?", "answer": ["Intriguing properties of neural networks", "Evasion Attacks against Machine Learning at Test Time", "Towards Evaluating the Robustness of Neural Networks", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Reliable evaluation of adversarial robustness with an ensemble of\n diverse parameter-free attacks", "Obfuscated Gradients Give a False Sense of Security: Circumventing\n Defenses to Adversarial Examples"], "answer_arxiv_id": ["1312.6199", "1708.06131", "1608.04644", "1706.06083", "2003.01690", "1802.00420"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_343"} +{"question": "What papers give an account of the datasets developed for the vision-based 3D Semantic Occupancy Prediction?", "answer": ["SSCBench: A Large-Scale 3D Semantic Scene Completion Benchmark for Autonomous Driving", "OpenOccupancy: A Large Scale Benchmark for Surrounding Semantic\n Occupancy Perception", "Occ3D: A Large-Scale 3D Occupancy Prediction Benchmark for Autonomous\n Driving"], "answer_arxiv_id": ["2306.09001v3", "2303.03991", "2304.14365"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_test_344"} +{"question": "Which studies surpassed traditional image compression standards by using neural image compression methods?", "answer": ["Causal Contextual Prediction for Learned Image Compression", "Learned Image Compression with Mixed Transformer-CNN Architectures", "High-Fidelity Generative Image Compression"], "answer_arxiv_id": ["2011.09704", "2303.14978", "2006.09965"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_test_345"} +{"question": "Which research works have explored the organizational structures required to facilitate the development of research communities around under-represented languages?", "answer": ["AI4D -- African Language Program", "NLP for Ghanaian Languages"], "answer_arxiv_id": ["2104.02516", "2103.15475"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_346"} +{"question": "Which studies proposed a homotopy algorithm in the context of a response vector parameterized with a real value?", "answer": ["Fast Exact Conformalization of Lasso using Piecewise Linear Homotopy"], "answer_arxiv_id": ["1708.00427"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_test_347"} +{"question": "What works proposed kernel-based distillation methods?", "answer": ["Dataset Distillation using Neural Feature Regression", "Efficient Dataset Distillation Using Random Feature Approximation"], "answer_arxiv_id": ["2206.00719", "2210.12067"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_348"} +{"question": "What research allows direct NeRF editing by changing a reference image in 2D space?", "answer": ["SINE: Semantic-driven Image-based NeRF Editing with Prior-guided Editing\n Field"], "answer_arxiv_id": ["2303.13277"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_349"} +{"question": "What works focus on obtaining strong visual representations through self-supervised training on large-scale image data?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "Masked Autoencoders Are Scalable Vision Learners", "EVA: Exploring the Limits of Masked Visual Representation Learning at\n Scale"], "answer_arxiv_id": ["2104.14294", "2111.06377", "2211.07636"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_350"} +{"question": "Can you refer any works that are used rounding or merging strategies to produce long feature tracks?", "answer": ["ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformer"], "answer_arxiv_id": ["2208.14201"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_351"} +{"question": "What is the research proposing a two-step approach involving foreground-background separation and text spatial alignment for scene text editing?", "answer": ["Editing Text in the Wild"], "answer_arxiv_id": ["1908.03047"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_test_352"} +{"question": "Could you provide me some works about zero-shot classification?", "answer": ["Zero-Shot Learning - A Comprehensive Evaluation of the Good, the Bad and the Ugly"], "answer_arxiv_id": ["1707.00600"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_353"} +{"question": "What research extended the work of Zaheer et al. by using hierarchical clustering to obtain fine-grained pseudo-labels?", "answer": ["A Coarse-to-Fine Pseudo-Labeling (C2FPL) Framework for Unsupervised\n Video Anomaly Detection"], "answer_arxiv_id": ["2310.17650"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_test_354"} +{"question": "Which paper discussed about the retrieval augmented generation (RAG) solutions?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2005.11401"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_test_355"} +{"question": "Which paper uses an Iterative Dataset Update(IDU) strategy to edit NeRF’s image dataset?", "answer": ["Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions"], "answer_arxiv_id": ["2303.12789"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_356"} +{"question": "What studies showed that maximum distance from a dataset to its subset was an effective measure of risk for instance optimality?", "answer": ["Near Instance-Optimality in Differential Privacy"], "answer_arxiv_id": ["2005.10630"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_test_357"} +{"question": "Which studies reported that the performance of LLMs declines as the proportion of noise in the retrieval context increases?", "answer": ["Benchmarking Large Language Models in Retrieval-Augmented Generation", "Making Retrieval-Augmented Language Models Robust to Irrelevant Context", "NoMIRACL: Knowing When You Don't Know for Robust Multilingual\n Retrieval-Augmented Generation"], "answer_arxiv_id": ["2309.01431", "2310.01558", "2312.11361"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_test_358"} +{"question": "What works have used hypercolumns for tasks like keypoint detection, segmentation and semantic correspondence?", "answer": ["Hypercolumns for Object Segmentation and Fine-grained Localization", "Deep Layer Aggregation", "Hyperpixel Flow: Semantic Correspondence with Multi-layer Neural Features", "AnchorNet: A Weakly Supervised Network to Learn Geometry-sensitive Features For Semantic Matching", "Learning to Compose Hypercolumns for Visual Correspondence", "Neural Best-Buddies: Sparse Cross-Domain Correspondence"], "answer_arxiv_id": ["1411.5752", "1707.06484", "1908.06537", "1704.04749", "2007.10587", "1805.04140v2"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_359"} +{"question": "Which papers mentioned about automatic learning of prompts, often termed as 'Prompt Learning'?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2104.08691"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_test_360"} +{"question": "What paper uses dynamically-updatable tree-sketches in the context of Kronecker regression?", "answer": ["Dynamic Tensor Product Regression"], "answer_arxiv_id": ["2210.03961v2"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_test_361"} +{"question": "Which references mentioned that the estimation of a policy gradient in policy-gradient approaches requires more data than in their benchmarks?", "answer": ["Fast Context Adaptation via Meta-Learning", "A Survey of Meta-Reinforcement Learning"], "answer_arxiv_id": ["1810.03642", "2301.08028"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_test_362"} +{"question": "Can you provide some examples of studies that used differentiable simulation for efficient co-optimization of soft robots?", "answer": ["DiffAqua: A Differentiable Computational Design Pipeline for Soft Underwater Swimmers with Shape Interpolation"], "answer_arxiv_id": ["2104.00837"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_test_363"} +{"question": "Which studies have examined the Polyak-Lojasiewicz (PL) inequality, a generalization of strong-convexity?", "answer": ["Linear Convergence of Gradient and Proximal-Gradient Methods Under the Polyak-Łojasiewicz Condition"], "answer_arxiv_id": ["1608.04636"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_test_364"} +{"question": "Which works have explored self-consistency techniques for refining language models in post-hoc correction?", "answer": ["Language Models (Mostly) Know What They Know", "Self-Evaluation Improves Selective Generation in Large Language Models", "Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence\n Scores from Language Models Fine-Tuned with Human Feedback", "Self-Refine: Iterative Refinement with Self-Feedback", "Chain-of-Verification Reduces Hallucination in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2207.05221", "2312.09300", "2305.14975", "2303.17651", "2309.11495", "2203.11171"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_test_365"} +{"question": "Could you provide me some studies that attempted to mitigate bias found in in-context learning by utilizing outputs distribution obtained from content-free texts?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models", "Mitigating Label Biases for In-context Learning"], "answer_arxiv_id": ["2102.09690", "2305.19148"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_test_366"} +{"question": "What papers propose a reduction from list-global stability to pseudo-global stability via correlated sampling in finite domains?", "answer": ["User-Level Private Learning via Correlated Sampling"], "answer_arxiv_id": ["2110.11208"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_367"} +{"question": "Which works provide a categorization of existing graph OOD generalization methodologies?", "answer": ["Out-Of-Distribution Generalization on Graphs: A Survey"], "answer_arxiv_id": ["2202.07987"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_test_368"} +{"question": "What studies are about Mixture of Experts, a technique used to improve robustness and overall accuracy in ensemble learning?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated\n Mixture-of-Experts Layer", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts"], "answer_arxiv_id": ["1701.06538", "2112.06905"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_test_369"} +{"question": "Can you indicate which studies have investigated the trade-off between personalization and performance?", "answer": ["Operationalizing the Legal Principle of Data Minimization for Personalization", "Learning to Limit Data Collection via Scaling Laws: A Computational Interpretation for the Legal Principle of Data Minimization"], "answer_arxiv_id": ["2005.13718", "2107.08096"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_test_370"} +{"question": "Which paper proposed the ROAR algorithm for finding the closest and robust counterfactuals?", "answer": ["Towards Robust and Reliable Algorithmic Recourse"], "answer_arxiv_id": ["2102.13620"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_test_371"} +{"question": "Which studies deal with aligning visual features with pre-trained LLMs for multimodal comprehension tasks?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Image as a Foreign Language: BEiT Pretraining for All Vision and\n Vision-Language Tasks", "Visual Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "Language Is Not All You Need: Aligning Perception with Language Models", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2304.10592", "2208.10442", "2304.08485", "2304.14178", "2302.14045", "2301.12597", "2305.11175", "2305.03726"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_372"} +{"question": "Which studies have designed additional supervision signals or training processes for models in context of CoT reasoning?", "answer": ["Tailoring Self-Rationalizers with Multi-Reward Distillation", "Crystal: Introspective Reasoners Reinforced with Self-Feedback"], "answer_arxiv_id": ["2311.02805", "2310.04921"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_373"} +{"question": "What works present the intervention technique of steering model output?", "answer": ["Plug and Play Language Models: a Simple Approach to Controlled Text Generation", "Back to the Future: Unsupervised Backprop-based Decoding for Counterfactual and Abductive Commonsense Reasoning", "GeDi: Generative Discriminator guided Sequence Generation", "Diffusion-LM Improves Controllable Text Generation"], "answer_arxiv_id": ["1912.02164", "2010.05906", "2009.06367", "2205.14217"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_test_374"} +{"question": "What research papers focus on minimizing the surrogate models learned from synthetic and original datasets in dataset distillation using matching gradients?", "answer": ["Dataset Condensation with Gradient Matching", "Dataset Condensation via Efficient Synthetic-Data Parameterization", "Accelerating Dataset Distillation via Model Augmentation"], "answer_arxiv_id": ["2006.05929", "2205.14959v2", "2212.06152"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_375"} +{"question": "What works study the geometry of embeddings that decomposes the shifted pointwise mutual information matrix?", "answer": ["Towards Understanding Linear Word Analogies"], "answer_arxiv_id": ["1810.04882"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_test_376"} +{"question": "Are there any papers that use MIAs to assess whether a given data point was used to train an LLM?", "answer": ["Membership Inference Attacks From First Principles", "Membership Inference Attacks Against Machine Learning Models"], "answer_arxiv_id": ["2112.03570", "1610.05820"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_377"} +{"question": "What research directly trains sparse GANs from scratch?", "answer": ["Don’t Be So Dense: Sparse-to-Sparse GAN Training Without Sacrificing Performance"], "answer_arxiv_id": ["2203.02770"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_test_378"} +{"question": "Could you name any studies that focus on report-to-CXR generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Adapting Pretrained Vision-Language Foundational Models to Medical\n Imaging Domains", "RoentGen: Vision-Language Foundation Model for Chest X-ray Generation"], "answer_arxiv_id": ["2112.10752", "2210.04133", "2211.12737"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_test_379"} +{"question": "What paper provides a connection between the matrix mechanism and the line of work about generating differentially private synthetic data?", "answer": ["Graphical-model based estimation and inference for differential privacy"], "answer_arxiv_id": ["1901.09136"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_test_380"} +{"question": "What works employed human pose estimation using 3D shape rendering from multiple views?", "answer": ["Multi-View Multi-Person 3D Pose Estimation with Plane Sweep Stereo"], "answer_arxiv_id": ["2104.02273"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_test_381"} +{"question": "Which studies regarded Kernel ridge regression as a special kind of spectral regularization algorithm?", "answer": ["Regularization in kernel learning"], "answer_arxiv_id": ["1001.2094v1"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_test_382"} +{"question": "Which paper first proposed Graph Contrastive Learning (GCL) with random edge dropping and feature masking as data augmentations?", "answer": ["Deep Graph Contrastive Representation Learning"], "answer_arxiv_id": ["2006.04131"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_test_383"} +{"question": "What studies have suggested the usage of produced clusters for multi-document summarization?", "answer": ["Multi-News: a Large-Scale Multi-Document Summarization Dataset and\n Abstractive Hierarchical Model", "Multi-XScience: A Large-scale Dataset for Extreme Multi-document\n Summarization of Scientific Articles"], "answer_arxiv_id": ["1906.01749", "2010.14235"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_test_384"} +{"question": "Which research gave regret bounds for a E2D that incorporates randomized estimators, but not optimism?", "answer": ["Unified Algorithms for RL with Decision-Estimation Coefficients: No-Regret, PAC, and Reward-Free Learning"], "answer_arxiv_id": ["2209.11745"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_385"} +{"question": "What studies introduced strategies to improve the computational and parameter efficiency of transformers by refining parameter-sharing mechanisms?", "answer": ["Lessons on Parameter Sharing across Layers in Transformers"], "answer_arxiv_id": ["2104.06022"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_386"} +{"question": "Could you provide me with research papers about learning a disentangled representation for RL?", "answer": ["DARLA: Improving Zero-Shot Transfer in Reinforcement Learning", "Temporal Disentanglement of Representations for Improved Generalisation in Reinforcement Learning"], "answer_arxiv_id": ["1707.08475", "2207.05480"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_387"} +{"question": "Which studies treated Transformer-based backbones as a 'black box'? ", "answer": ["Faster-TAD: Towards Temporal Action Detection with Proposal Generation\n and Classification in a Unified Network", "TALLFormer: Temporal Action Localization with a Long-memory Transformer", "Re^2TAL: Rewiring Pretrained Video Backbones for Reversible Temporal\n Action Localization"], "answer_arxiv_id": ["2204.02674", "2204.01680", "2211.14053"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_388"} +{"question": "Which studies first used CLIP to optimize an underlying 3D representation in text-guided 3D generation?", "answer": ["Zero-Shot Text-Guided Object Generation with Dream Fields", "CLIP-Mesh: Generating textured meshes from text using pretrained image-text models", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2112.01455", "2203.13333", "2103.00020"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_test_389"} +{"question": "What papers mention the vulnerability of DNNs to common corruptions, random noises, and adversarial perturbations?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations", "A Study and Comparison of Human and Deep Learning Recognition Performance Under Visual Distortions", "Evasion attacks against machine learning at test time", "Intriguing properties of neural networks"], "answer_arxiv_id": ["1903.12261", "1705.02498", "1708.06131", "1312.6199"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_test_390"} +{"question": "Are there any studies regarding the creation of evaluation benchmarks for Natural Language Generation (NLG) on Indic languages?", "answer": ["IndicNLG Benchmark: Multilingual Datasets for Diverse NLG Tasks in Indic Languages"], "answer_arxiv_id": ["2203.05437v2"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_test_391"} +{"question": "Which work first initiated the efforts to manipulate Neural Radiance Fields (NeRFs)?", "answer": ["NeRF-Editing: Geometry Editing of Neural Radiance Fields"], "answer_arxiv_id": ["2205.04978"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_392"} +{"question": "Could you refer me to some studies that use score-based models for graph generation?", "answer": ["Permutation Invariant Graph Generation via Score-Based Generative Modeling", "Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations", "Score-Based Generative Modeling through Stochastic Differential Equations", "DiGress: Discrete Denoising diffusion for graph generation", "Diffusion Models for Graphs Benefit From Discrete State Spaces"], "answer_arxiv_id": ["2003.00638", "2202.02514", "2011.13456", "2209.14734", "2210.01549"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_test_393"} +{"question": "Could you provide some studies about unfaithful hallucination?", "answer": ["Truthful AI: Developing and governing AI that does not lie", "AI Deception: A Survey of Examples, Risks, and Potential Solutions", "Inference-Time Intervention: Eliciting Truthful Answers from a Language\n Model"], "answer_arxiv_id": ["2110.06674", "2308.14752", "2306.03341"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_test_394"} +{"question": "Any works that introduced a method of learning policy by simulating states via the dynamics model?", "answer": ["Model Based Reinforcement Learning for Atari"], "answer_arxiv_id": ["1903.00374"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_test_395"} +{"question": "Which research made a claim that oversmoothing is asymptotically inevitable in GATs?", "answer": ["Improving Graph Attention Networks with Large Margin-based Constraints"], "answer_arxiv_id": ["1910.11945"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_test_396"} +{"question": "Which work considers a different approach to limiting the amount of predicted information within learning-augmented paging?", "answer": ["Parsimonious Learning-Augmented Caching"], "answer_arxiv_id": ["2202.04262"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_test_397"} +{"question": "What work demonstrates the use of transformers in 3D pose estimation task?", "answer": ["Epipolar Transformers"], "answer_arxiv_id": ["2005.04551"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_398"} +{"question": "Are there any studies using a sequence-encoding VQ-VAE to learn the discrete codebook of listener motion?", "answer": ["Learning to Listen: Modeling Non-Deterministic Dyadic Facial Motion"], "answer_arxiv_id": ["2204.08451"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_test_399"} +{"question": "What paper presents a form of data augmentation for training CLIP models?", "answer": ["Improving CLIP Training with Language Rewrites"], "answer_arxiv_id": ["2305.20088"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_400"} +{"question": "Can you indicate some works that use 2D diffusion models to generate multi-view images then use them for 3D reconstruction with NeRF?", "answer": ["Novel View Synthesis with Diffusion Models", "Zero-1-to-3: Zero-shot One Image to 3D Object", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization"], "answer_arxiv_id": ["2210.04628", "2303.11328", "2306.07881", "2306.16928"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_401"} +{"question": "Which works propose consistency-based methods for detecting non-factual generations in LLM generated content?", "answer": ["Measuring and Improving Consistency in Pretrained Language Models", "Self-contradictory Hallucinations of Large Language Models: Evaluation,\n Detection and Mitigation", "How to Catch an AI Liar: Lie Detection in Black-Box LLMs by Asking\n Unrelated Questions", "SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for\n Generative Large Language Models", "LM vs LM: Detecting Factual Errors via Cross Examination", "The Internal State of an LLM Knows When It's Lying", "Chain-of-Verification Reduces Hallucination in Large Language Models", "Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation\n in Natural Language Generation", "Language Models (Mostly) Know What They Know", "Representation Engineering: A Top-Down Approach to AI Transparency", "Factcheck-Bench: Fine-Grained Evaluation Benchmark for Automatic\n Fact-checkers", "RARR: Researching and Revising What Language Models Say, Using Language\n Models", "FacTool: Factuality Detection in Generative AI -- A Tool Augmented\n Framework for Multi-Task and Multi-Domain Scenarios"], "answer_arxiv_id": ["2102.01017", "2305.15852", "2309.15840", "2303.08896", "2305.13281", "2304.13734", "2309.11495", "2302.09664", "2207.05221", "2310.01405", "2311.09000", "2210.08726", "2307.13528"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_test_402"} +{"question": "Can you list the works that followed the concept of neural fields for 3D scene and object representation?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Multiview Neural Surface Reconstruction by Disentangling Geometry and\n Appearance"], "answer_arxiv_id": ["2003.08934", "2003.09852"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_403"} +{"question": "Who uses similarities between word sense definitions to approximate human judgments on semantic proximity?", "answer": ["Interpretable Word Sense Representations via Definition Generation: The\n Case of Semantic Change Analysis"], "answer_arxiv_id": ["2305.11993"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_test_404"} +{"question": "Could you provide me some works about human-curated instruction datasets in languages outside of English?", "answer": ["Crosslingual Generalization through Multitask Finetuning", "M$^3$IT: A Large-Scale Dataset towards Multi-Modal Multilingual\n Instruction Tuning"], "answer_arxiv_id": ["2211.01786", "2306.04387"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_405"} +{"question": "Who are the pioneer researchers in providing different contexts for a single visual concept using multiple images as part of personalized visual content generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_test_406"} +{"question": "Are there any studies in sports video understanding which involves benchmarks for spatio-temporal reasoning?", "answer": ["UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild", "MultiSports: A Multi-Person Video Dataset of Spatio-Temporally Localized\n Sports Actions", "FineGym: A Hierarchical Video Dataset for Fine-grained Action\n Understanding", "SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos", "SportsMOT: A Large Multi-Object Tracking Dataset in Multiple Sports\n Scenes", "Social Adaptive Module for Weakly-supervised Group Activity Recognition", "A Hierarchical Deep Temporal Model for Group Activity Recognition"], "answer_arxiv_id": ["1212.0402", "2105.07404", "2004.06704", "1804.04527", "2304.05170", "2007.09470", "1511.06040"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_test_407"} +{"question": "Could you provide me some studies of the use of beam search on KGs using LLMs to dynamically extract the most relevant reasoning paths?", "answer": ["Think-on-Graph: Deep and Responsible Reasoning of Large Language Model\n on Knowledge Graph"], "answer_arxiv_id": ["2307.07697"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_test_408"} +{"question": "What research observed the detrimental effect of adding irrelevant noise to the context on model performance?", "answer": ["Adversarial Examples for Evaluating Reading Comprehension Systems", "Selection-Inference: Exploiting Large Language Models for Interpretable\n Logical Reasoning"], "answer_arxiv_id": ["1707.07328", "2205.09712"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_test_409"} +{"question": "What papers explored hallucination source and detection in Linguistically-Informed Language Models (LLMs)?", "answer": ["A Survey on Hallucination in Large Language Models: Principles,\n Taxonomy, Challenges, and Open Questions", "Survey of Hallucination in Natural Language Generation", "Siren's Song in the AI Ocean: A Survey on Hallucination in Large\n Language Models", "HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large\n Language Models"], "answer_arxiv_id": ["2311.05232", "2202.03629", "2309.01219", "2305.11747"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_test_410"} +{"question": "Could you provide me some works that provide convergence guarantees of stochastic methods for solving quasi-strongly monotone Variational inequality problems (VIPs)?", "answer": ["Stochastic Extragradient: General Analysis and Improved Rates", "On the Convergence of Single-Call Stochastic Extra-Gradient Methods"], "answer_arxiv_id": ["2111.08611v3", "1908.08465"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_411"} +{"question": "What references were made to works related to transformer-based models in hierarchical classification research?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Swin Transformer V2: Scaling Up Capacity and Resolution"], "answer_arxiv_id": ["2010.11929", "2103.14030", "2111.09883"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_test_412"} +{"question": "Which paper tackles the problem of registration of two scene fragments captured from two 3D viewpoints with low overlap?", "answer": ["ObjectMatch: Robust Registration using Canonical Object Correspondences"], "answer_arxiv_id": ["2212.01985"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_test_413"} +{"question": "Which studies have leveraged object template information for identifying potential object locations?", "answer": ["Incremental Class Discovery for Semantic Segmentation with RGBD Sensing"], "answer_arxiv_id": ["1907.10008"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_test_414"} +{"question": "Which paper is about the Transformers architecture that most recent large language models (LLM) are based on?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_test_415"} +{"question": "Which studies highlight that increased model capacity is needed to achieve robustness against adversarial examples?", "answer": ["Adversarial Robustness May Be at Odds With Simplicity"], "answer_arxiv_id": ["1901.00532"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_416"} +{"question": "Could you provide me some works that apply Gaussian corruptions to token-vector embeddings in diffusion models?", "answer": ["Self-conditioned Embedding Diffusion for Text Generation", "Continuous diffusion for categorical data"], "answer_arxiv_id": ["2211.04236", "2211.15089"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_417"} +{"question": "Could you list down studies that convert building footprint extraction into roof segmentation and roof-to-footprint offset estimation tasks?", "answer": ["Learning to Extract Building Footprints from Off-Nadir Aerial Images"], "answer_arxiv_id": ["2204.13637"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_test_418"} +{"question": "Which paper proposed a Mixed Integer Linear Programming based method to relax GNN certification?", "answer": ["Collective Robustness Certificates: Exploiting Interdependence in Graph Neural Networks"], "answer_arxiv_id": ["2302.02829"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_test_419"} +{"question": "What works discuss initializations from the optimization perspective?", "answer": ["Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification", "Mean Field Residual Networks: On the Edge of Chaos", "Tunable Efficient Unitary Neural Networks (EUNN) and their application to RNNs", "Revisiting Initialization of Neural Networks"], "answer_arxiv_id": ["1502.01852", "1712.08969", "1612.05231", "2004.09506v3"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_test_420"} +{"question": "In which work is explained how LLMs have also been trained on instruction datasets, meaning they can generate text using a custom instruction prompt?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_test_421"} +{"question": "Could you provide me studies that apply CoT prompting for multiple reasoning traces and diversifying reasoning paths?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Cross-lingual Prompting: Improving Zero-shot Chain-of-Thought Reasoning\n across Languages"], "answer_arxiv_id": ["2203.11171", "2310.14799"], "source_meta": {"published_time": "20240628"}, "qid": "AutoScholarQuery_test_422"} +{"question": "Which research papers employ recurrent attention for routing but not for inferring slots?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Genesis: Generative Scene Inference and Sampling with Object-Centric Latent Representations"], "answer_arxiv_id": ["1901.11390", "1907.13052"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_423"} +{"question": "Which papers talk about the upper bounds for general activation functions?", "answer": ["Universal Approximation with Deep Narrow Networks", "Minimum Width for Universal Approximation"], "answer_arxiv_id": ["1905.08539", "2006.08859"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_test_424"} +{"question": "Can you provide the references of studies that have employed diffusion models for text-driven image editing?", "answer": ["SEGA: Instructing Diffusion using Semantic Dimensions", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2301.12247", "2211.09800"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_test_425"} +{"question": "Which are the studies that have done efforts via multi-task learning in personalized Federated Learning?", "answer": ["Federated Multi-Task Learning", "Personalized Cross-Silo Federated Learning on Non-IID Data", "Federated Multi-Task Learning under a Mixture of Distributions"], "answer_arxiv_id": ["1705.10467", "2007.03797", "2108.10252v4"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_test_426"} +{"question": "What studies discuss the field of 'learning from human feedback'?", "answer": ["Neural Machine Translation by Jointly Learning to Align and Translate", "WebGPT: Browser-assisted question-answering with human feedback", "Training language models to follow instructions with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback", "Constitutional AI: Harmlessness from AI Feedback", "ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation", "Text-guided Image-and-Shape Editing and Generation: A Short Survey", "Aligning Text-to-Image Models using Human Feedback", "RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignment"], "answer_arxiv_id": ["1409.0473", "2112.09332", "2203.02155", "2204.05862", "2212.08073", "2304.05977", "2304.09244", "2302.12192", "2304.06767"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_test_427"} +{"question": "What work discusses the limitation of how answer generation can be updated based on retrievd documents?", "answer": ["Entity-Based Knowledge Conflicts in Question Answering"], "answer_arxiv_id": ["2109.05052"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_test_428"} +{"question": "What study proposed a hierarchical transformer architecture that unifies semantic tokens and stacked hierarchical acoustic tokens within one stage?", "answer": ["UniAudio: An Audio Foundation Model Toward Universal Audio Generation"], "answer_arxiv_id": ["2310.00704"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_429"} +{"question": "What are some existing methods proposed for Federated Domain Generalization?", "answer": ["FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Continuous Frequency Space", "Federated Domain Generalization for Image Recognition via Cross-Client Style Transfer"], "answer_arxiv_id": ["2103.06030", "2210.00912"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_test_430"} +{"question": "Which papers present locate-then-edit methods for knowledge editing?", "answer": ["Knowledge Neurons in Pretrained Transformers", "Locating and Editing Factual Associations in GPT", "Mass-Editing Memory in a Transformer"], "answer_arxiv_id": ["2104.08696", "2202.05262", "2210.07229"], "source_meta": {"published_time": "20230916"}, "qid": "AutoScholarQuery_test_431"} +{"question": "What are some works that have focused on how LLMs can be connected to visual foundation models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Language-based Action Concept Spaces Improve Video Self-Supervised\n Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models"], "answer_arxiv_id": ["2204.14198", "2307.10922", "2301.12597", "2304.08485", "2306.05424"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_test_432"} +{"question": "Could you provide me some works that reveal the limitations of LLMs' ability in providing emotional support?", "answer": ["Challenges of Large Language Models for Mental Health Counseling", "ChatGPT as a Therapist Assistant: A Suitability Study", "The Typing Cure: Experiences with Large Language Model Chatbots for\n Mental Health Support"], "answer_arxiv_id": ["2311.13857", "2304.09873", "2401.14362"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_test_433"} +{"question": "Could you provide me some studies about Neural Radiance Field (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_test_434"} +{"question": "Could you provide me some studies about imitation learning methods that doesn't strictly fit into IDM or GAIL based approaches?", "answer": ["Imitating Latent Policies from Observation", "Reinforcement Learning with Videos: Combining Offline Observations with Interaction"], "answer_arxiv_id": ["1805.07914", "2011.06507"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_435"} +{"question": "Which research proposed discrete adaptations of the diffusion model in the context of text generation?", "answer": ["Structured Denoising Diffusion Models in Discrete State-Spaces"], "answer_arxiv_id": ["2107.03006"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_test_436"} +{"question": "Could you provide me some works where human feedback was utilised to finetune large language models?", "answer": ["Neural Machine Translation by Jointly Learning to Align and Translate", "WebGPT: Browser-assisted question-answering with human feedback", "Training language models to follow instructions with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback", "Constitutional AI: Harmlessness from AI Feedback"], "answer_arxiv_id": ["1409.0473", "2112.09332", "2203.02155", "2204.05862", "2212.08073"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_test_437"} +{"question": "What papers study the integration of visual encoding modules, like ViT, with LLMs such as LLaMa?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2010.11929", "2302.13971"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_test_438"} +{"question": "Which study was the first to consider the problem of differentially private submodular maximization in the context of CPPP?", "answer": ["Differentially Private Combinatorial Optimization"], "answer_arxiv_id": ["0903.4510"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_test_439"} +{"question": "What research papers proposed using attention mechanisms to combine optical flow and deformable convolution for feature alignment?", "answer": ["Recurrent Video Restoration Transformer with Guided Deformable Attention", "Spatio-Temporal Deformable Attention Network for Video Deblurring", "BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation\n and Alignment", "Video Dehazing via a Multi-Range Temporal Alignment Network with\n Physical Prior"], "answer_arxiv_id": ["2206.02146", "2207.10852", "2104.13371", "2303.09757"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_test_440"} +{"question": "Which paper suggested that LLMs internally need to infer latent variables for better prediction in the context of in-context learning?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_441"} +{"question": "Are there any works on analytical reconstruction attacks which recursively reconstruct activation maps layer per layer?", "answer": ["R-GAP: Recursive Gradient Attack on Privacy"], "answer_arxiv_id": ["2010.07733"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_442"} +{"question": "What is the first work that proposed a question-answering dataset that requires the model to perform counterfactual reasoning?", "answer": ["IfQA: A Dataset for Open-domain Question Answering under Counterfactual\n Presuppositions"], "answer_arxiv_id": ["2305.14010"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_test_443"} +{"question": "Which works focus on neural architecture search (NAS) for discovering optimal network structures?", "answer": ["Global Convergence of MAML and Theory-Inspired Neural Architecture Search for Few-Shot Learning", "DARTS: Differentiable Architecture Search"], "answer_arxiv_id": ["2203.09137", "1806.09055"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_test_444"} +{"question": "Which works have used pre-computing or post-computing methods for feature aggregation in GNN models?", "answer": ["Simplifying Graph Convolutional Networks", "SIGN: Scalable Inception Graph Neural Networks", "Scalable and Adaptive Graph Neural Networks with Self-Label-Enhanced Training", "Graph Attention Multi-Layer Perceptron", "Scaling Graph Neural Networks with Approximate PageRank", "Combining Label Propagation and Simple Models out-performs Graph Neural Networks"], "answer_arxiv_id": ["1902.07153", "2004.11198", "2104.09376", "2206.04355", "2007.01570", "2010.13993"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_test_445"} +{"question": "Could you provide me some works about the problem of exploding variance of the RP gradients due to the chaotic nature of environments?", "answer": ["PIPPS: Flexible Model-Based Policy Search Robust to the Curse of Chaos", "Gradients are Not All You Need"], "answer_arxiv_id": ["1902.01240", "2111.05803"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_446"} +{"question": "What research introduced methods that adapt the training procedure of the classifier itself?", "answer": ["Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Flipout: Efficient Pseudo-Independent Weight Perturbations on Mini-Batches", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "When Does Label Smoothing Help?", "Transferable Calibration with Lower Bias and Variance in Domain Adaptation", "On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networks", "mixup: Beyond Empirical Risk Minimization", "Evidential Deep Learning to Quantify Classification Uncertainty", "Towards Trustworthy Predictions from Deep Neural Networks with Fast Adversarial Calibration", "Pitfalls of In-Domain Uncertainty Estimation and Ensembling in Deep Learning"], "answer_arxiv_id": ["1906.12340", "1506.02142", "1803.04386", "1612.01474", "1906.02629", "2007.08259", "1905.11001", "1710.09412", "1806.01768", "2012.10923", "2002.06470"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_test_447"} +{"question": "Which works look at achieving ϵ-optimal reward in non-convex settings?", "answer": ["Optimal Gradient-based Algorithms for Non-concave Bandit Optimization", "Optimal Stochastic Nonconvex Optimization with Bandit Feedback"], "answer_arxiv_id": ["2107.04518", "2103.16082"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_test_448"} +{"question": "Which work proposed the Mirror Learning algorithm?", "answer": ["Mirror Learning: A Unifying Framework of Policy Optimisation"], "answer_arxiv_id": ["2201.02373"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_449"} +{"question": "What works have explored video human-object interaction detection?", "answer": ["ST-HOI: A Spatial-Temporal Baseline for Human-Object Interaction\n Detection in Videos"], "answer_arxiv_id": ["2105.11731"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_test_450"} +{"question": "What are some representative works about graph embedding-based methods?", "answer": ["RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space", "Convolutional 2D Knowledge Graph Embeddings", "Complex Embeddings for Simple Link Prediction", "Holographic Embeddings of Knowledge Graphs", "kbgan: Adversarial Learning for Knowledge Graph Embeddings", "TuckER: Tensor Factorization for Knowledge Graph Completion", "Embedding Entities and Relations for Learning and Inference in Knowledge Bases", "Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings", "BoxE: A Box Embedding Model for Knowledge Base Completion", "Modeling Fine-Grained Entity Types with Box Embeddings"], "answer_arxiv_id": ["1902.10197", "1707.01476", "1606.06357", "1510.04935", "1711.04071", "1901.09590", "1412.6575", "2002.05969", "2007.06267v2", "2101.00345"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_451"} +{"question": "What are some works related to the Mean Teacher paradigm?", "answer": ["Self-supervised Augmentation Consistency for Adapting Semantic\n Segmentation", "DAFormer: Improving Network Architectures and Training Strategies for\n Domain-Adaptive Semantic Segmentation", "Prototypical Pseudo Label Denoising and Target Structure Learning for\n Domain Adaptive Semantic Segmentation", "End-to-End Semi-Supervised Object Detection with Soft Teacher", "Active Teacher for Semi-Supervised Object Detection", "Consistent-Teacher: Towards Reducing Inconsistent Pseudo-targets in Semi-supervised Object Detection", "Omni-DETR: Omni-Supervised Object Detection with Transformers", "ALWOD: Active Learning for Weakly-Supervised Object Detection", "Contrastive Mean Teacher for Domain Adaptive Object Detectors", "Cross-Domain Adaptive Teacher for Object Detection", "Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain\n Adaptation on Person Re-identification", "Exploiting Sample Uncertainty for Domain Adaptive Person\n Re-Identification", "Delving into Probabilistic Uncertainty for Unsupervised Domain Adaptive\n Person Re-Identification"], "answer_arxiv_id": ["2105.00097", "2111.14887", "2101.10979", "2106.09018", "2303.08348", "2209.01589v3", "2203.16089", "2309.07914", "2305.03034", "2111.13216", "2001.01526", "2012.08733", "2112.14025"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_test_452"} +{"question": "Could you provide me studies where they utilized self-supervision and few-shot prompting to address the scalability of large language model tool-use?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "TALM: Tool Augmented Language Models"], "answer_arxiv_id": ["2302.04761", "2205.12255"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_test_453"} +{"question": "What papers depict studies of sample complexity of posterior sampling in tabular settings?", "answer": ["Generalization and Exploration via Randomized Value Functions", "Deep Exploration via Randomized Value Functions", "Worst-Case Regret Bounds for Exploration via Randomized Value Functions"], "answer_arxiv_id": ["1402.0635", "1703.07608v5", "1906.02870"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_test_454"} +{"question": "Can you provide the studies that use additional guiding signals to incorporate explicit control in conditional diffusion models?", "answer": ["SpaText: Spatio-Textual Representation for Controllable Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2211.14305", "2112.10752", "2211.09800"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_test_455"} +{"question": "What research papers discuss memory-based methods as a knowledge editing paradigm?", "answer": ["Memory-Based Model Editing at Scale", "Transformer-Patcher: One Mistake worth One Neuron", "Calibrating Factual Knowledge in Pretrained Language Models", "Can We Edit Factual Knowledge by In-Context Learning?"], "answer_arxiv_id": ["2206.06520", "2301.09785", "2210.03329", "2305.12740"], "source_meta": {"published_time": "20230916"}, "qid": "AutoScholarQuery_test_456"} +{"question": "Any studies about generating adversarial examples in textual domains?", "answer": ["Adversarial Examples for Evaluating Reading Comprehension Systems", "Generating Natural Language Adversarial Examples", "Adversarial VQA: A New Benchmark for Evaluating the Robustness of VQA\n Models", "HotFlip: White-Box Adversarial Examples for Text Classification", "Universal Adversarial Triggers for Attacking and Analyzing NLP"], "answer_arxiv_id": ["1707.07328", "1804.07998", "2106.00245", "1712.06751", "1908.07125"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_457"} +{"question": "What work highlights computational challenges and memory issues in bi-level optimization-based methods?", "answer": ["Scaling Up Dataset Distillation to ImageNet-1K with Constant Memory"], "answer_arxiv_id": ["2211.10586"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_458"} +{"question": "Which works describe FL strategies for handling communication burden issues?", "answer": ["FedPAQ: A Communication-Efficient Federated Learning Method with Periodic Averaging and Quantization"], "answer_arxiv_id": ["1909.13014v4"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_test_459"} +{"question": "What papers discussed handcrafted prompts for specific tasks?", "answer": ["AutoPrompt: Eliciting Knowledge from Language Models with Automatically\n Generated Prompts"], "answer_arxiv_id": ["2010.15980"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_test_460"} +{"question": "What are some works that have advanced image editing methods achieving more accurate attribute modification textually?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models", "Negative-prompt Inversion: Fast Image Inversion for Editing with\n Text-guided Diffusion Models"], "answer_arxiv_id": ["2208.01626", "2211.09800", "2211.09794", "2305.16807"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_test_461"} +{"question": "Which research papers studied methods based on model mixture for Full Model Personalization?", "answer": ["Federated Learning of a Mixture of Global and Local Models", "Adaptive Personalized Federated Learning", "Three Approaches for Personalization with Applications to Federated\n Learning"], "answer_arxiv_id": ["2002.05516", "2003.13461", "2002.10619"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_test_462"} +{"question": "Can you mention studies that are trained on massive pairs of images and captions, enabling them to generate detailed image content conditioned on textual instruction?", "answer": ["Zero-Shot Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2204.06125", "2112.10752"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_463"} +{"question": "What studies further propagated the token-level edit operation approach proposed by LaserTagger?", "answer": ["Parallel Iterative Edit Models for Local Sequence Transduction", "GECToR -- Grammatical Error Correction: Tag, Not Rewrite"], "answer_arxiv_id": ["1910.02893", "2005.12592"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_test_464"} +{"question": "Which paper first introduced the concept of fully unsupervised anomaly detection?", "answer": ["Generative Cooperative Learning for Unsupervised Video Anomaly Detection"], "answer_arxiv_id": ["2203.03962"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_test_465"} +{"question": "Which study introduced a metric based on the sample influence score of the optimal empirical risk in CoreSet selection?", "answer": ["Data Pruning via Moving-one-Sample-out"], "answer_arxiv_id": ["2310.14664"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_466"} +{"question": "What papers study hallucination issues in MLLMs?", "answer": ["Evaluation and Analysis of Hallucination in Large Vision-Language Models"], "answer_arxiv_id": ["2308.15126"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_test_467"} +{"question": "What works learn disentangled representations from time-series data?", "answer": ["Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICA"], "answer_arxiv_id": ["1605.06336"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_468"} +{"question": "Could you provide some examples of diffusion models that involve different number of denoising steps and parameterization of transformation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2010.02502", "2206.00927", "2211.01095", "2202.09778"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_test_469"} +{"question": "Which works use consistency between model generated content and external information for factuality detection in LLM?", "answer": ["Factcheck-Bench: Fine-Grained Evaluation Benchmark for Automatic\n Fact-checkers", "RARR: Researching and Revising What Language Models Say, Using Language\n Models", "FacTool: Factuality Detection in Generative AI -- A Tool Augmented\n Framework for Multi-Task and Multi-Domain Scenarios"], "answer_arxiv_id": ["2311.09000", "2210.08726", "2307.13528"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_test_470"} +{"question": "Which papers discussed the use of gradient guidance to increase the occurrence of a desired attribute in discrete generative models?", "answer": ["Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space"], "answer_arxiv_id": ["1612.00005"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_471"} +{"question": "Which works experimentally proved that adding Gaussian noise during training increases adversarial robustness?", "answer": ["Robustness of classifiers: from adversarial to random noise", "Adversarial Examples Are a Natural Consequence of Test Error in Noise"], "answer_arxiv_id": ["1608.08967", "1901.10513"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_test_472"} +{"question": "Could you name some studies that have explored the influence of asking LLMs to respond as a particular person?", "answer": ["Meet Your Favorite Character: Open-domain Chatbot Mimicking Fictional Characters with only a Few Utterances", "CTRL: A Conditional Transformer Language Model for Controllable Generation"], "answer_arxiv_id": ["2204.10825", "1909.05858"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_473"} +{"question": "What papers provide a discussion or implementation of the method of 'prompting'?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Exploring Visual Prompts for Adapting Large-Scale Models", "Task Bias in Vision-Language Models"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2203.17274", "2212.04412"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_474"} +{"question": "Are there any studies that have applied large-scale pre-trained models to the UCDR task?", "answer": ["Episodic Training for Domain Generalization", "Semantic Data Augmentation based Distance Metric Learning for Domain\n Generalization", "Universal Cross-Domain Retrieval: Generalizing Across Classes and\n Domains"], "answer_arxiv_id": ["1902.00113", "2208.02803", "2108.08356"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_test_475"} +{"question": "What work is similar to the researcher's use of context distillation in knowledge editing?", "answer": ["A General Language Assistant as a Laboratory for Alignment"], "answer_arxiv_id": ["2112.00861"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_476"} +{"question": "Which studies focused on vision-based GUI navigation using GPT-4V?", "answer": ["GPT-4V in Wonderland: Large Multimodal Models for Zero-Shot Smartphone\n GUI Navigation", "ASSISTGUI: Task-Oriented Desktop Graphical User Interface Automation", "AppAgent: Multimodal Agents as Smartphone Users"], "answer_arxiv_id": ["2311.07562", "2312.13108", "2312.13771"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_test_477"} +{"question": "Could you provide me the works that introduced policy-based methods for performing adaptive experimentation?", "answer": ["Implicit Deep Adaptive Design: Policy–Based Experimental Design without Likelihoods", "Deep Adaptive Design: Amortizing Sequential Bayesian Experimental Design"], "answer_arxiv_id": ["2111.02329", "2103.02438"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_test_478"} +{"question": "Which works focus on developing more effective pretraining losses beyond the autoregressive or masked language modeling objectives?", "answer": ["XLNet: Generalized Autoregressive Pretraining for Language Understanding", "ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators", "UniLMv2: Pseudo-Masked Language Models for Unified Language Model Pre-Training"], "answer_arxiv_id": ["1906.08237", "2003.10555", "2002.12804"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_test_479"} +{"question": "Are there any works on multilingual ARA models?", "answer": ["Automatic Readability Assessment for Closely Related Languages"], "answer_arxiv_id": ["2305.13478"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_480"} +{"question": "Are there studies adapting popular uncertainty tools in Bayesian Uncertainty Estimation literature to generative LLMs?", "answer": ["Uncertainty Estimation in Autoregressive Structured Prediction"], "answer_arxiv_id": ["2002.07650"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_481"} +{"question": "Could you provide me some works about knowledge retrieval from prompts?", "answer": ["Label Words are Anchors: An Information Flow Perspective for\n Understanding In-Context Learning"], "answer_arxiv_id": ["2305.14160"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_482"} +{"question": "What studies have been conducted on multilingual Large Language Models (LLMs) with a focus on hundreds of languages?", "answer": ["mT5: A massively multilingual pre-trained text-to-text transformer", "GPT-4 Technical Report", "Gemini: A Family of Highly Capable Multimodal Models"], "answer_arxiv_id": ["2010.11934", "2303.08774", "2312.11805v4"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_test_483"} +{"question": "Could you provide me some works about visual pretraining that emphasized contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2104.14294"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_test_484"} +{"question": "What studies have focused on architectures that exhibit combinatorial generalization such as transformers, graph neural networks, and bilinear models?", "answer": ["Attention Is All You Need", "Combinatorial Optimization and Reasoning with Graph Neural Networks"], "answer_arxiv_id": ["1706.03762", "2102.09544"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_test_485"} +{"question": "What paper proposed the averaged stochastic approximation (NASA) to obtain a better rate for non-convex objectives?", "answer": ["A Single Time-Scale Stochastic Approximation Method for Nested Stochastic Optimization"], "answer_arxiv_id": ["1812.01094"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_486"} +{"question": "Which papers concentrate on decomposing complex questions into sub-questions?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models", "Successive Prompting for Decomposing Complex Questions"], "answer_arxiv_id": ["2205.10625", "2212.04092"], "source_meta": {"published_time": "20240628"}, "qid": "AutoScholarQuery_test_487"} +{"question": "Which paper proposed multi-channel equivariant graph networks?", "answer": ["Conditional Antibody Design as 3D Equivariant Graph Translation"], "answer_arxiv_id": ["2208.06073"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_488"} +{"question": "Can you name the papers where VeRA, the method that reduces the number of parameters with the help of random projections, is mentioned?", "answer": ["VeRA: Vector-based Random Matrix Adaptation"], "answer_arxiv_id": ["2310.11454"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_489"} +{"question": "Any studies about prompt fine-tuning for vision models to adapt image models from one image task to another?", "answer": ["Visual Prompt Tuning", "Adversarial Reprogramming of Neural Networks"], "answer_arxiv_id": ["2203.12119", "1806.11146"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_test_490"} +{"question": "Can you mention some sequence models proposing the use of multiple time scales, akin to the multi-rate mechanism in this study?", "answer": ["UnICORNN: A recurrent model for learning very long time dependencies", "Long Expressive Memory for Sequence Modeling"], "answer_arxiv_id": ["2103.05487", "2110.04744"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_test_491"} +{"question": "What works propose strategies for face capture that are more easily accessible and convenient for daily users?", "answer": ["AvatarMe: Realistically Renderable 3D Facial Reconstruction\n \"in-the-wild\"", "Relightify: Relightable 3D Faces from a Single Image via Diffusion\n Models", "Learning a 3D Morphable Face Reflectance Model from Low-cost Data", "A Morphable Face Albedo Model", "Learning Formation of Physically-Based Face Attributes", "FitMe: Deep Photorealistic 3D Morphable Model Avatars", "Practical Face Reconstruction via Differentiable Ray Tracing"], "answer_arxiv_id": ["2003.13845", "2305.06077", "2303.11686", "2004.02711", "2004.03458", "2305.09641", "2101.05356"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_492"} +{"question": "Which early works employed ConvNets and inverse perspective mapping (IPM) for mapping features from perspective view to BEV view?", "answer": ["Multi-View 3D Object Detection Network for Autonomous Driving", "Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by\n Implicitly Unprojecting to 3D", "Cross-view Semantic Segmentation for Sensing Surroundings"], "answer_arxiv_id": ["1611.07759", "2008.05711", "1906.03560"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_test_493"} +{"question": "What researches made use of auxiliary data such as human-annotated attribute information, text description or knowledge graph in Zero-Shot Learning (ZSL)?", "answer": ["Learning Deep Representations of Fine-grained Visual Descriptions", "Multi-Label Zero-Shot Learning with Structured Knowledge Graphs"], "answer_arxiv_id": ["1605.05395", "1711.06526"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_test_494"} +{"question": "What research papers prioritize salient weights for PTQ in LLMs?", "answer": ["SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight\n Compression", "AWQ: Activation-aware Weight Quantization for LLM Compression and\n Acceleration", "SqueezeLLM: Dense-and-Sparse Quantization"], "answer_arxiv_id": ["2306.03078", "2306.00978", "2306.07629"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_495"} +{"question": "Any studies showing that LLM hidden states can effectively represent a task defined by input-output pairs?", "answer": ["Function Vectors in Large Language Models"], "answer_arxiv_id": ["2310.15213"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_test_496"} +{"question": "What studies gave rise to prompt-based learning in LLMs?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "GPT-4 Technical Report", "OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["1810.04805", "2005.14165", "2303.08774", "2205.01068"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_497"} +{"question": "What studies apply slot attention for object discovery?", "answer": ["Object-Centric Learning with Slot Attention", "GENESIS: Generative Scene Inference and Sampling with Object-Centric\n Latent Representations", "Conditional Object-Centric Learning from Video", "Shepherding Slots to Objects: Towards Stable and Robust Object-Centric\n Learning"], "answer_arxiv_id": ["2006.15055", "1907.13052", "2111.12594", "2303.17842"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_498"} +{"question": "What works proposed method for experimental design for causal discovery in a non-BOED setting in the presence of cycles?", "answer": ["A Unified Experiment Design Approach for Cyclic and Acyclic Causal Models"], "answer_arxiv_id": ["2205.10083"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_test_499"} +{"question": "What study employs a contextual word retrieval task where the model is tasked with finding corresponding words and sentences across parallel corpora?", "answer": ["Multilingual Alignment of Contextual Word Representations"], "answer_arxiv_id": ["2002.03518"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_test_500"} +{"question": "Which work leverages graph generation as the training objective in generative self-supervised learning?", "answer": ["GPT-GNN: Generative Pre-Training of Graph Neural Networks"], "answer_arxiv_id": ["2006.15437"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_test_501"} +{"question": "Could you provide me some studies that propose non-parametric approaches to post-hoc calibration methods?", "answer": ["Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning", "Non-Parametric Calibration for Classification", "Dirichlet-based Gaussian Processes for Large-scale Calibrated Classification"], "answer_arxiv_id": ["2003.07329", "1906.04933v3", "1805.10915"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_test_502"} +{"question": "Could you provide me some studies about object hallucination in MLLMs?", "answer": ["Evaluating Object Hallucination in Large Vision-Language Models"], "answer_arxiv_id": ["2305.10355"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_test_503"} +{"question": "Which studies proposed methods for learning 3D-aware image and geometry generation with implicit neural radiance fields as generators?", "answer": ["GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis"], "answer_arxiv_id": ["2007.02442", "2012.00926"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_test_504"} +{"question": "Could you provide me some studies on global stability that contributed to the Probably Eventually Correct (PEC) learning model?", "answer": ["An Equivalence Between Private Classification and Online Prediction", "Sample-efficient proper PAC learning with approximate differential privacy"], "answer_arxiv_id": ["2003.00563", "2012.03893"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_505"} +{"question": "Which studies have aimed at generalization to underrepresented languages in LLM-based evaluation methodologies?", "answer": ["Are Large Language Model-based Evaluators the Solution to Scaling Up\n Multilingual Evaluation?"], "answer_arxiv_id": ["2309.07462"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_506"} +{"question": "What studies have aimed to capture functional requirements in software development by generating code-like outlines using in-context learning?", "answer": ["Self-planning Code Generation with Large Language Models", "Think Outside the Code: Brainstorming Boosts Large Language Models in\n Code Generation"], "answer_arxiv_id": ["2303.06689v3", "2305.10679"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_test_507"} +{"question": "Could you provide me some references that evaluate the ability of LMs to reason about emerging entities?", "answer": ["Mind the Gap: Assessing Temporal Generalization in Neural Language Models", "Time-Aware Language Models as Temporal Knowledge Bases", "RealTime QA: What’s the Answer Right Now?"], "answer_arxiv_id": ["2102.01951", "2106.15110", "2207.13332"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_508"} +{"question": "What papers are about utilizing machine-learned predictions for designing efficient offline algorithms?", "answer": ["Faster Matchings via Learned Duals", "Faster Fundamental Graph Algorithms via Learned Predictions", "Discrete-Convex-Analysis-Based Framework for Warm-Starting Algorithms with Predictions", "Learning-Augmented Maximum Flow"], "answer_arxiv_id": ["2107.09770", "2204.12055", "2205.09961", "2207.12911"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_test_509"} +{"question": "Could you provide me a work that introduced EL2N score as a measure of importance in CoreSet selection?", "answer": ["Deep Learning on a Data Diet: Finding Important Examples Early in\n Training"], "answer_arxiv_id": ["2107.07075"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_510"} +{"question": "Which works proposed reinforcement learning methods to tackle challenges of behavior cloning in autonomous driving?", "answer": ["Learning to Drive in a Day", "Navigating Occluded Intersections with Autonomous Vehicles using Deep Reinforcement Learning", "A Reinforcement Learning Based Approach for Automated Lane Change Maneuvers", "Imitation Is Not Enough: Robustifying Imitation with Reinforcement Learning for Challenging Driving Scenarios"], "answer_arxiv_id": ["1807.00412", "1705.01196", "1804.07871v1", "2212.11419"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_test_511"} +{"question": "Which paper involves the use of the Faster-RCNN model in Vision-Language Pre-training?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks"], "answer_arxiv_id": ["1506.01497"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_test_512"} +{"question": "What works examined training agents through reinforcement learning on the MiniWob web environment?", "answer": ["Reinforcement Learning on Web Interfaces Using Workflow-Guided\n Exploration", "Learning to Navigate the Web"], "answer_arxiv_id": ["1802.08802", "1812.09195"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_test_513"} +{"question": "What works proposed hierarchical BERT models designed for extractive summarization?", "answer": ["HIBERT: Document Level Pre-training of Hierarchical Bidirectional\n Transformers for Document Summarization", "Unsupervised Extractive Summarization by Pre-training Hierarchical\n Transformers", "HiStruct+: Improving Extractive Text Summarization with Hierarchical\n Structure Information"], "answer_arxiv_id": ["1905.06566", "2010.08242", "2203.09629"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_test_514"} +{"question": "Which work related to the researcher's work examines tasks like distilling a persona-conditioned language model?", "answer": ["Prompt Injection: Parameterization of Fixed Inputs"], "answer_arxiv_id": ["2206.11349"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_515"} +{"question": "Could you provide me studies focused on the grounding capabilities of LVLMs?", "answer": ["VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "MiniGPT-v2: large language model as a unified interface for\n vision-language multi-task learning"], "answer_arxiv_id": ["2305.11175", "2310.09478"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_test_516"} +{"question": "Which work is related to private prediction?", "answer": ["Privacy-preserving Prediction"], "answer_arxiv_id": ["1803.10266"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_test_517"} +{"question": "Which studies incorporated the use of diffusion models in their work?", "answer": ["Any-to-Any Generation via Composable Diffusion", "Emu: Generative Pretraining in Multimodality"], "answer_arxiv_id": ["2305.11846", "2307.05222"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_test_518"} +{"question": "What work proposed a method of retraining classifiers on 're-weighting' data to reduce reliance on spurious features?", "answer": ["Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations"], "answer_arxiv_id": ["2204.02937"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_test_519"} +{"question": "Could you provide me research that allows for controlled text-based scene editing by finetuning an image diffusion model?", "answer": ["DreamEditor: Text-Driven 3D Scene Editing with Neural Fields"], "answer_arxiv_id": ["2306.13455"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_520"} +{"question": "Could you provide me some work that applied transformer or its variants into TAD head?", "answer": ["An Empirical Study of End-to-End Temporal Action Detection", "Relaxed Transformer Decoders for Direct Action Proposal Generation", "ActionFormer: Localizing Moments of Actions with Transformers", "ReAct: Temporal Action Detection with Relational Queries"], "answer_arxiv_id": ["2204.02932", "2102.01894", "2202.07925", "2207.07097"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_521"} +{"question": "What papers have incorporated the use of NLEs in fields beyond NLP, such as in computer vision, medical field, and self-driving cars?", "answer": ["Grounding Visual Explanations", "From Recognition to Cognition: Visual Commonsense Reasoning", "Knowledge-Grounded Self-Rationalization via Extractive and Natural\n Language Explanations", "Explaining Chest X-ray Pathologies in Natural Language", "Textual Explanations for Self-Driving Vehicles"], "answer_arxiv_id": ["1807.09685", "1811.10830", "2106.13876", "2207.04343", "1807.11546"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_522"} +{"question": "Which research proposes to enhance the reasoning process in LLMs by framing thoughts as graphs?", "answer": ["Beyond Chain-of-Thought, Effective Graph-of-Thought Reasoning in\n Language Models"], "answer_arxiv_id": ["2305.16582"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_523"} +{"question": "Any studies arguing for centering language models’ evaluation on how models will be used in practice?", "answer": ["Rethinking Model Evaluation as Narrowing the Socio-Technical Gap"], "answer_arxiv_id": ["2306.03100"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_524"} +{"question": "Could you name some studies employing distributional semantic models such as count-based or Word2Vec approaches for the LSC task?", "answer": ["Cultural Shift or Linguistic Drift? Comparing Two Computational Measures\n of Semantic Change"], "answer_arxiv_id": ["1606.02821"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_test_525"} +{"question": "Which works addressed the issue of inadequate data quality in large corpora and extended cleaning efforts?", "answer": ["Towards a Cleaner Document-Oriented Multilingual Crawled Corpus"], "answer_arxiv_id": ["2201.06642"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_526"} +{"question": "Can you provide papers that discussed the concept of latent embeddings?", "answer": ["RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space", "Convolutional 2D Knowledge Graph Embeddings", "Complex Embeddings for Simple Link Prediction", "Holographic Embeddings of Knowledge Graphs", "kbgan: Adversarial Learning for Knowledge Graph Embeddings", "TuckER: Tensor Factorization for Knowledge Graph Completion", "Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings", "BoxE: A Box Embedding Model for Knowledge Base Completion", "Modeling Fine-Grained Entity Types with Box Embeddings"], "answer_arxiv_id": ["1902.10197", "1707.01476", "1606.06357", "1510.04935", "1711.04071", "1901.09590", "2002.05969", "2007.06267v2", "2101.00345"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_527"} +{"question": "What kinds of researches have been conducted using reinforcement learning to train policies for adaptive experimental design?", "answer": ["Optimizing Sequential Experimental Design with Deep Reinforcement Learning", "Policy-Based Bayesian Experimental Design for Non-Differentiable Implicit Models"], "answer_arxiv_id": ["2202.00821", "2203.04272v1"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_test_528"} +{"question": "Which research investigations in language finetuning consider data selection as crucial?", "answer": ["Selection via Proxy: Efficient Data Selection for Deep Learning", "Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt"], "answer_arxiv_id": ["1906.11829", "2206.07137"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_529"} +{"question": "What works covered methods that factor in agreement between positive pixel pairs for dense predictions?", "answer": ["Propagate Yourself: Exploring Pixel-Level Consistency for Unsupervised Visual Representation Learning", "Self-supervised Learning with Local Contrastive Loss for Detection and Semantic Segmentation"], "answer_arxiv_id": ["2011.10043", "2207.04398"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_test_530"} +{"question": "Could you name some research papers that utilized pre-trained text-to-image diffusion models and SMPL models for avatar generation?", "answer": ["DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via Diffusion Models", "AvatarCraft: Transforming Text into Neural Human Avatars with Parameterized Shape and Pose Control"], "answer_arxiv_id": ["2304.00916", "2303.17606"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_test_531"} +{"question": "What research proposed multi-plane images for novel-view synthesis?", "answer": ["Stereo Magnification: Learning View Synthesis using Multiplane Images"], "answer_arxiv_id": ["1805.09817"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_test_532"} +{"question": "What are some studies that have used data statistics, representations, logits, and embedding to avoid exposing privacy in Federated Learning?", "answer": ["XOR Mixup: Privacy-Preserving Data Augmentation for One-Shot Federated Learning", "Towards Fair Federated Learning with Zero-Shot Data Augmentation", "Cronus: Robust and Heterogeneous Collaborative Learning with Black-Box Knowledge Transfer", "No Fear of Heterogeneity: Classifier Calibration for Federated Learning with Non-IID Data", "FedProto: Federated Prototype Learning across Heterogeneous Clients"], "answer_arxiv_id": ["2006.05148", "2104.13417", "1912.11279v1", "2106.05001", "2105.00243"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_test_533"} +{"question": "What are the key works in the field of diffusion models which are a class of generative probabilistic models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Diffusion Models in Vision: A Survey", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2209.04747", "2105.05233", "2006.11239", "2102.09672"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_test_534"} +{"question": "Are there any works that reported that the volume of input knowledge for each query in ICL is constrained by the maximum input length of PLMs?", "answer": ["Structured Prompting: Scaling In-Context Learning to 1,000 Examples"], "answer_arxiv_id": ["2212.06713"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_test_535"} +{"question": "Can you name works that have used specific methodology like Android’s View Hierarchy, Regions of Interest, or screenshots for representing interfaces?", "answer": ["Spotlight: Mobile UI Understanding using Vision-Language Models with a\n Focus", "Reinforced UI Instruction Grounding: Towards a Generic UI Task\n Automation API"], "answer_arxiv_id": ["2209.14927", "2310.04716"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_test_536"} +{"question": "Which papers demonstrate that language models can map conceptual domains onto grounded world representations?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_test_537"} +{"question": "What are the task-informed models such as XL-LEXEME based on?", "answer": ["Interpretable Word Sense Representations via Definition Generation: The\n Case of Semantic Change Analysis"], "answer_arxiv_id": ["2305.11993"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_test_538"} +{"question": "Any studies that employ the Masked Language Model to obtain corrections in GEC tasks?", "answer": ["Felix: Flexible Text Editing Through Tagging and Insertion", "GEC-DePenD: Non-Autoregressive Grammatical Error Correction with Decoupled Permutation and Decoding"], "answer_arxiv_id": ["2003.10687", "2311.08191v1"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_test_539"} +{"question": "Could you provide me the works on finding architectures with high accuracy on clean examples not considering their robustness?", "answer": ["AdvRush: Searching for Adversarially Robust Neural Architectures", "Neural Architecture Design and Robustness: A Dataset"], "answer_arxiv_id": ["2108.01289", "2306.06712"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_test_540"} +{"question": "What does paper `bib.bib31` propose?", "answer": ["A Generalist Agent"], "answer_arxiv_id": ["2205.06175"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_test_541"} +{"question": "Who looked at prompting closed-source LLMs to leverage their reasoning and planning abilities for web tasks through in-context learning and self-refine?", "answer": ["Synapse: Trajectory-as-Exemplar Prompting with Memory for Computer\n Control", "Language Models can Solve Computer Tasks"], "answer_arxiv_id": ["2306.07863", "2303.17491"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_test_542"} +{"question": "Which research papers introduced initial vision-language pre-training models?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning", "Unified Vision-Language Pre-Training for Image Captioning and VQA", "Unifying Vision-and-Language Tasks via Text Generation", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision", "Large-Scale Adversarial Training for Vision-and-Language Representation\n Learning", "Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal\n Transformers", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "12-in-1: Multi-Task Vision and Language Representation Learning", "Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language\n Representation Learning", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision"], "answer_arxiv_id": ["1909.11740", "1909.11059", "2102.02779", "2004.06165", "2102.03334", "2006.06195", "2004.00849", "1908.02265", "1908.08530", "1912.02315", "2104.03135", "2108.10904"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_test_543"} +{"question": "What studies have explored incorporating additional conditioning, to generate images with precise control?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "Composer: Creative and Controllable Image Synthesis with Composable Conditions", "HumanSD: A Native Skeleton-Guided Diffusion Model for Human Image Generation"], "answer_arxiv_id": ["2302.05543", "2302.09778", "2304.04269"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_test_544"} +{"question": "Which studies primarily improve ACI by setting the learning rate adaptively?", "answer": ["Conformal Inference for Online Prediction with Arbitrary Distribution Shifts", "Adaptive Conformal Predictions for Time Series", "Improved Online Conformal Prediction via Strongly Adaptive Online Learning"], "answer_arxiv_id": ["2208.08401", "2202.07282", "2302.07869"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_test_545"} +{"question": "In what papers do the researchers use the dual potentials to recover the OT map?", "answer": ["Large-Scale Optimal Transport and Mapping Estimation"], "answer_arxiv_id": ["1711.02283"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_546"} +{"question": "What paper proposed an unsupervised global disentanglement score called Distortion?", "answer": ["Analyzing the Latent Space of GAN through Local Dimension Estimation"], "answer_arxiv_id": ["2205.13182"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_547"} +{"question": "Any works achieve higher accuracy on low-textured regions with the help of Transformer in semi-dense matching methods?", "answer": ["LoFTR: Detector-Free Local Feature Matching with Transformers", "ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformer", "MatchFormer: Interleaving Attention in Transformers for Feature Matching"], "answer_arxiv_id": ["2104.00680", "2208.14201", "2203.09645"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_548"} +{"question": "Which work is related to the theoretical foundations and implementation of Proximal Policy Optimization?", "answer": ["Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1707.06347v2"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_549"} +{"question": "What research papers utilized the embedding-based metrics which make use of PLM embeddings like BERT?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "BERTScore: Evaluating Text Generation with BERT", "MoverScore: Text Generation Evaluating with Contextualized Embeddings\n and Earth Mover Distance"], "answer_arxiv_id": ["1810.04805", "1904.09675", "1909.02622"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_550"} +{"question": "What etudies are on Transformer-based models for speech that have been used to test their brain alignment for speech-evoked brain activity?", "answer": ["Vector-Quantized Autoregressive Predictive Coding", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech\n Representations", "HuBERT: Self-Supervised Speech Representation Learning by Masked\n Prediction of Hidden Units", "Toward a realistic model of speech processing in the brain with\n self-supervised learning", "Self-supervised models of audio effectively explain human cortical\n responses to speech"], "answer_arxiv_id": ["2005.08392", "2006.11477", "2106.07447", "2206.01685", "2205.14252"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_551"} +{"question": "Could you name some research papers on exploratory attacks to achieve different attack goals such as link re-identification, property inference, membership inference, and model stealing?", "answer": ["Stealing Links from Graph Neural Networks", "Inference Attacks Against Graph Neural Networks", "Model Extraction Attacks on Graph Neural Networks: Taxonomy and Realization", "Quantifying Privacy Leakage in Graph Embedding"], "answer_arxiv_id": ["2005.02131", "2110.02631", "2010.12751v2", "2010.00906"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_552"} +{"question": "What studies introduce methods to reduce the learning time by reducing the communication overheads of Federated Learning?", "answer": ["Adaptive Gradient Sparsification for Efficient Federated Learning: An Online Learning Approach"], "answer_arxiv_id": ["2001.04756"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_test_553"} +{"question": "What works represent the approach of contrastive learning for image-text pre-training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "Florence: A New Foundation Model for Computer Vision"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.07991", "2111.11432"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_test_554"} +{"question": "Are there any methods using hierarchical Reinforcement Learning to decompose complex tasks into sub-tasks?", "answer": ["Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation", "The Option-Critic Architecture", "Near-Optimal Representation Learning for Hierarchical Reinforcement Learning", "Language as an Abstraction for Hierarchical Deep Reinforcement Learning", "Unsupervised Skill Discovery with Bottleneck Option Learning", "Toward Robust Long Range Policy Transfer"], "answer_arxiv_id": ["1604.06057", "1609.05140", "1810.01257", "1906.07343", "2106.14305", "2103.02957"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_test_555"} +{"question": "Can you give examples of research that aligns the source and target point clouds with an orientation estimation module before using a teacher-student model and a DGCNN backbone to find the correspondence?", "answer": ["SE-ORNet: Self-Ensembling Orientation-aware Network for Unsupervised\n Point Cloud Shape Correspondence"], "answer_arxiv_id": ["2304.05395"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_test_556"} +{"question": "What works discuss defense mechanisms that generate adversarially perturbed images to improve neural network's robustness?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "DeepFool: a simple and accurate method to fool deep neural networks"], "answer_arxiv_id": ["1412.6572", "1706.06083", "1511.04599"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_test_557"} +{"question": "In what studies LMMs directly reason over embedded visual features?", "answer": ["Visual Instruction Tuning", "Improved Baselines with Visual Instruction Tuning", "Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with\n Modality Collaboration", "MultiModal-GPT: A Vision and Language Model for Dialogue with Humans", "PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2304.08485", "2310.03744", "2204.14198", "2301.12597", "2305.06500", "2304.10592", "2304.14178", "2311.04257", "2305.04790", "2303.03378"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_test_558"} +{"question": "Could you give me some references about adversarial learning approach in domain adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks", "CyCADA: Cycle-Consistent Adversarial Domain Adaptation", "Progressive Domain Adaptation for Object Detection"], "answer_arxiv_id": ["1505.07818", "1711.03213", "1910.11319"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_559"} +{"question": "What works proposed methods to improve computational efficiency by reducing the number of times the underlying field needs to be evaluated in the context of neural scene representations?", "answer": ["Neural Sparse Voxel Fields", "Plenoxels: Radiance Fields without Neural Networks", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2007.11571", "2112.05131", "2103.14024", "2201.05989"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_560"} +{"question": "Which paper uses counterfactual explanations in explainable recommender systems?", "answer": ["Counterfactual Explanations for Neural Recommenders"], "answer_arxiv_id": ["2105.05008"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_test_561"} +{"question": "Can you name the methods that tackle the high-dimensional importance weight estimation problem?", "answer": ["Telescoping Density-Ratio Estimation", "Featurized Density Ratio Estimation", "Density Ratio Estimation via Infinitesimal Classification"], "answer_arxiv_id": ["2006.12204", "2107.02212", "2111.11010v2"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_562"} +{"question": "Which works have pioneered in text-editing as a category of generative data augmentation?", "answer": ["EDA: Easy Data Augmentation Techniques for Boosting Performance on Text\n Classification Tasks"], "answer_arxiv_id": ["1901.11196"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_test_563"} +{"question": "Which papers proposed an algorithm to handle non-stationary MDPs with linear mixture function approximation of both transitions and rewards?", "answer": ["Optimistic Policy Optimization is Provably Efficient in Non-stationary MDPs"], "answer_arxiv_id": ["2110.08984"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_test_564"} +{"question": "What studies discuss solutions involving interpolation between LR and RP gradients?", "answer": ["PIPPS: Flexible Model-Based Policy Search Robust to the Curse of Chaos", "Do Differentiable Simulators Give Better Policy Gradients?"], "answer_arxiv_id": ["1902.01240", "2202.00817"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_565"} +{"question": "Which works have implemented multimodal understanding and generative capacities across modalities?", "answer": ["ImageBind: One Embedding Space To Bind Them All", "Any-to-Any Generation via Composable Diffusion", "Generating Images with Multimodal Language Models", "NExT-GPT: Any-to-Any Multimodal LLM", "Emu: Generative Pretraining in Multimodality", "Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction\n Tuning"], "answer_arxiv_id": ["2305.05665v2", "2305.11846", "2305.17216", "2309.05519", "2307.05222", "2309.02591"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_566"} +{"question": "What research conducted studies on zero-shot tasks on images using foundation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Florence: A New Foundation Model for Computer Vision"], "answer_arxiv_id": ["2103.00020", "2205.01917", "2102.05918", "2111.11432"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_567"} +{"question": "What paper used shifted window attention to enable information propagation in the area of action recongnition?", "answer": ["Video Swin Transformer"], "answer_arxiv_id": ["2106.13230"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_568"} +{"question": "What works have explored the field of zero-shot segmentation recently?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "ReCo: Retrieve and Co-segment for Zero-shot Transfer", "Image Segmentation Using Text and Image Prompts", "Zero-Shot Semantic Segmentation", "DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic\n Segmentation Using Diffusion Models", "Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2401.14159", "2206.07045", "2112.10003", "1906.00817", "2303.11681", "2112.01071"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_test_569"} +{"question": "Could you provide me the research where local visual features aligned with textual concepts in CLIP were revealed?", "answer": ["Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2112.01071"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_570"} +{"question": "Could you provide me some works about localization strategies using region-proposal detector or a segmentation network in vision-language models?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_test_571"} +{"question": "Any research on probablistic personalized page rank (ProPPR) models?", "answer": ["Programming with Personalized PageRank: A Locally Groundable First-Order Probabilistic Logic"], "answer_arxiv_id": ["1305.2254"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_572"} +{"question": "Which studies offer solutions to deal with accuracy loss when a sub-model is sent to slower devices?", "answer": ["FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout", "HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients"], "answer_arxiv_id": ["2102.13451", "2010.01264"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_test_573"} +{"question": "What paper proposed SuperGLUE after seeing that many models were surpassing non-expert humans on GLUE?", "answer": ["SuperGLUE: A Stickier Benchmark for General-Purpose Language\n Understanding Systems"], "answer_arxiv_id": ["1905.00537"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_574"} +{"question": "Who is responsible for the best-known result for the MMS approximation in additive valuations, 3/4+3/3836?", "answer": ["Breaking the 3/4 Barrier for Approximate Maximin Share"], "answer_arxiv_id": ["2307.07304"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_575"} +{"question": "Which works pioneered the use of Hessian-based influence functions to understand how training data affects a model's prediction?", "answer": ["Understanding Black-box Predictions via Influence Functions"], "answer_arxiv_id": ["1703.04730"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_test_576"} +{"question": "What work has used self-supervised pre-training to design a flexible reward function by utilizing a broad dataset of human videos and a small dataset of robot videos?", "answer": ["Learning Generalizable Robotic Reward Functions from “In-The-Wild” Human Videos"], "answer_arxiv_id": ["2103.16817"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_577"} +{"question": "Could you provide some works where delayed feedback is studied in stochastic settings for UCB-based methods?", "answer": ["Bandit Learning with Delayed Impact of Actions", "Linear Bandits with Stochastic Delayed Feedback", "Stochastic bandits with arm-dependent delays"], "answer_arxiv_id": ["2002.10316", "1807.02089v3", "2006.10459"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_test_578"} +{"question": "Which papers discuss solutions to commonsense reasoning problems?", "answer": ["CommonsenseQA: A Question Answering Challenge Targeting Commonsense\n Knowledge", "CommonsenseQA 2.0: Exposing the Limits of AI through Gamification", "Cosmos QA: Machine Reading Comprehension with Contextual Commonsense\n Reasoning", "Abductive Commonsense Reasoning", "SocialIQA: Commonsense Reasoning about Social Interactions"], "answer_arxiv_id": ["1811.00937", "2201.05320", "1909.00277", "1908.05739", "1904.09728"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_579"} +{"question": "Could you name some works where diffusion models were applied to represent 3D scenes and motion sequences?", "answer": ["DiffRF: Rendering-Guided 3D Radiance Field Diffusion", "PhysDiff: Physics-Guided Human Motion Diffusion Model", "Human Motion Diffusion Model"], "answer_arxiv_id": ["2212.01206", "2212.02500", "2209.14916"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_test_580"} +{"question": "Where is the concept of Generative Adversarial Networks (GANs) introduced?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_test_581"} +{"question": "Any works about searching for the optimal scaling factor in the context of quantization-aware training?", "answer": ["PACT: Parameterized Clipping Activation for Quantized Neural Networks", "DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients", "Learned Step Size Quantization"], "answer_arxiv_id": ["1805.06085", "1606.06160", "1902.08153"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_test_582"} +{"question": "Who applied the mutual information in the context of maximising similarity between successive observations for representation learning?", "answer": ["Deep Reinforcement and InfoMax Learning"], "answer_arxiv_id": ["2006.07217"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_583"} +{"question": "Are there any studies that have explored teacher assistant-based and student-friendly distillation to alleviate the problem of performance degradation in larger LMs?", "answer": ["Improved Knowledge Distillation via Teacher Assistant", "On the Efficacy of Knowledge Distillation", "Decoupled Knowledge Distillation", "Lifting the Curse of Capacity Gap in Distilling Language Models"], "answer_arxiv_id": ["1902.03393", "1910.01348", "2203.08679", "2305.12129"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_584"} +{"question": "What works used data-driven approaches in AutoRL methods to learn various algorithmic components?", "answer": ["Meta-Learning via Learned Loss", "Meta-Gradient Reinforcement Learning with an Objective Discovered Online", "VeLO: Training Versatile Learned Optimizers by Scaling Up"], "answer_arxiv_id": ["1906.05374", "2007.08433", "2211.09760"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_585"} +{"question": "What works have modified compressed sensing approaches to include a cross-validation step?", "answer": ["Compressed Sensing with Cross Validation"], "answer_arxiv_id": ["0803.1845"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_test_586"} +{"question": "Are there works that investigate the implementation of gradient descent and stochastic gradient descent based on the Polyak-Lojasiewicz (PL) assumption?", "answer": ["Linear Convergence of Gradient and Proximal-Gradient Methods Under the Polyak-Łojasiewicz Condition", "SGD for Structured Nonconvex Functions: Learning Rates, Minibatching and Interpolation"], "answer_arxiv_id": ["1608.04636", "2006.10311v3"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_test_587"} +{"question": "Which work used counterfactual interventions to determine the unfaithfulness of explanations of a LLM's predictions?", "answer": ["Faithfulness Tests for Natural Language Explanations"], "answer_arxiv_id": ["2305.18029"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_588"} +{"question": "Which work was the first to propose Generation-Augmented Retrieval in question answering?", "answer": ["Generation-Augmented Retrieval for Open-domain Question Answering"], "answer_arxiv_id": ["2009.08553"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_test_589"} +{"question": "What studies discuss the introduction of large-scale continuous sign language datasets?", "answer": ["BSL-1K: Scaling up co-articulated sign language recognition using\n mouthing cues", "How2Sign: A Large-scale Multimodal Dataset for Continuous American Sign\n Language"], "answer_arxiv_id": ["2007.12131", "2008.08143"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_590"} +{"question": "What papers proposed architectural modifications to recurrence equations for irregular time series data?", "answer": ["Discrete Event, Continuous Time RNNs", "Recurrent Neural Networks for Multivariate Time Series with Missing Values", "Unitary Evolution Recurrent Neural Networks"], "answer_arxiv_id": ["1710.04110", "1606.01865", "1511.06464"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_test_591"} +{"question": "Could you provide me with the papers that proposed using graph homomorphism counts for feature embedding in learning tasks?", "answer": ["Graph Homomorphism Convolution"], "answer_arxiv_id": ["2005.01214"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_test_592"} +{"question": "Which research applied inversion attacks to models across different domains like computational genetics, computer vision, and NLP?", "answer": ["Information Leakage in Embedding Models"], "answer_arxiv_id": ["2004.00053"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_test_593"} +{"question": "Which works expanded on the contrastive methods of representation learning by integrating additional views into the mutual information maximization objective?", "answer": ["Contrastive Multi-View Representation Learning on Graphs", "Deep Graph Contrastive Representation Learning"], "answer_arxiv_id": ["2006.05582", "2006.04131"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_test_594"} +{"question": "Can you provide me studies about decomposing MOCO problems into a series of single-objective combinatorial optimization problems?", "answer": ["Pareto Set Learning for Expensive Multi-Objective Optimization", "Learning the Pareto Front with Hypernetworks", "Pareto Multi-Task Learning"], "answer_arxiv_id": ["2210.08495", "2010.04104", "1912.12854"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_test_595"} +{"question": "Which papers focused on the calibration of classifiers in foundation models for natural language processing?", "answer": ["How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answering", "Calibration of Pre-trained Transformers"], "answer_arxiv_id": ["2012.00955", "2003.07892"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_test_596"} +{"question": "Which works have been established for regret minimization in two types of MDPs under linear function approximation?", "answer": ["Reinforcement Learning in Feature Space: Matrix Bandit, Kernels, and Regret Bound", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Sample-Optimal Parametric Q-Learning Using Linearly Additive Features", "Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1905.10389", "2006.01107", "1902.04779", "1907.05388"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_test_597"} +{"question": "What works have focused on integrating constraints into sequential decision problems?", "answer": ["Constrained Policy Optimization", "Exploration-Exploitation in Constrained MDPs", "A Lyapunov-based Approach to Safe Reinforcement Learning", "Constrained Reinforcement Learning Has Zero Duality Gap"], "answer_arxiv_id": ["1705.10528", "2003.02189", "1805.07708", "1910.13393"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_test_598"} +{"question": "What work analysed VAE-based disentanglement techniques on correlated data?", "answer": ["On Disentangled Representations Learned from Correlated Data"], "answer_arxiv_id": ["2006.07886"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_599"} +{"question": "Could you provide me some studies that investigate where factual information is stored in transformers?", "answer": ["Knowledge Neurons in Pretrained Transformers", "Locating and Editing Factual Associations in GPT", "Mass-Editing Memory in a Transformer"], "answer_arxiv_id": ["2104.08696", "2202.05262", "2210.07229"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_600"} +{"question": "What publications discuss the usage of Bayesian neural network in enhancing the efficiency of FL?", "answer": ["Bayesian Nonparametric Federated Learning of Neural Networks", "Statistical Model Aggregation via Parameter Matching"], "answer_arxiv_id": ["1905.12022", "1911.00218"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_test_601"} +{"question": "What papers delved into the conformal prediction methods for drug property prediction?", "answer": ["Concepts and Applications of Conformal Prediction in Computational Drug Discovery"], "answer_arxiv_id": ["1908.03569v1"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_test_602"} +{"question": "What papers evaluated Generation-Augmented Retrieval in passage retrieval and fine-tuning?", "answer": ["Precise Zero-Shot Dense Retrieval without Relevance Labels", "Query2doc: Query Expansion with Large Language Models"], "answer_arxiv_id": ["2212.10496", "2303.07678"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_test_603"} +{"question": "Can you provide any works that adopted transition sampling for long-term generation of motion?", "answer": ["Synthesizing Long-Term Human Motions with Diffusion Models via Coherent\n Sampling"], "answer_arxiv_id": ["2308.01850"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_test_604"} +{"question": "Which work is the only one investigating training a hypernetwork end-to-end to arbitrarily modify the weights of a policy in meta-RL?", "answer": ["Hypernetworks in Meta-Reinforcement Learning"], "answer_arxiv_id": ["2210.11348"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_test_605"} +{"question": "Can you mention some studies that combined 3D Gaussian Splatting with diffusion models for efficient text-to-3D generation?", "answer": ["DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation", "Text-to-3D using Gaussian Splatting"], "answer_arxiv_id": ["2309.16653", "2309.16585"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_606"} +{"question": "Does there exist a work which demonstrates that the PAC-Bayes framework can't be used to derive distribution-free PAC learning bounds for classes that have infinite Littlestone dimension?", "answer": ["A Limitation of the PAC-Bayes Framework"], "answer_arxiv_id": ["2006.13508"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_607"} +{"question": "What works tackle the problem of designing data augmentations for graph classification?", "answer": ["Graph Data Augmentation for Graph Machine Learning: A Survey", "Data Augmentation for Deep Graph Learning: A Survey"], "answer_arxiv_id": ["2202.08871", "2202.08235"], "source_meta": {"published_time": "20220226"}, "qid": "AutoScholarQuery_test_608"} +{"question": "In what papers are the singular vectors of the first weight matrix proposed as global disentangled perturbations?", "answer": ["Closed-Form Factorization of Latent Semantics in GANs"], "answer_arxiv_id": ["2007.06600v4"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_609"} +{"question": "Which works propose alternate perspectives of robustness in explanations?", "answer": ["Issues with post-hoc counterfactual explanations: a discussion", "On the Robustness of Interpretability Methods"], "answer_arxiv_id": ["1906.04774v1", "1806.08049v1"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_test_610"} +{"question": "Which references explore reproducibility in optimization?", "answer": ["Reproducibility in Optimization: Theoretical Framework and Limits"], "answer_arxiv_id": ["2202.04598"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_611"} +{"question": "What research have used normalizing flows in molecular structure design?", "answer": ["Equivariant Flows: Exact Likelihood Generative Learning for Symmetric Densities", "E(n) Equivariant Normalizing Flows"], "answer_arxiv_id": ["2006.02425", "2105.09016"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_612"} +{"question": "Could you provide me some studies of zero-order methods applying strategies that the more expensive oracles are used infrequently?", "answer": ["Derivative-Free Method For Composite Optimization With Applications To Decentralized Distributed Optimization", "Oracle Complexity Separation in Convex Optimization", "One-Point Gradient-Free Methods for Composite Optimization with Applications to Distributed Optimization"], "answer_arxiv_id": ["1911.10645", "2002.02706", "2107.05951v2"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_test_613"} +{"question": "What work adds a buffer of demonstrations to the RL framework?", "answer": ["Overcoming Exploration in Reinforcement Learning with Demonstrations"], "answer_arxiv_id": ["1709.10089"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_test_614"} +{"question": "What works are amongst the most influential in relation to U-Net?", "answer": ["UNet++: A Nested U-Net Architecture for Medical Image Segmentation", "Attention U-Net: Learning Where to Look for the Pancreas", "3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation", "Denoising Diffusion Probabilistic Models", "nnU-Net: Self-adapting Framework for U-Net-Based Medical Image Segmentation", "A Probabilistic U-Net for Segmentation of Ambiguous Images", "A Variational U-Net for Conditional Appearance and Shape Generation", "Road Extraction by Deep Residual U-Net"], "answer_arxiv_id": ["1807.10165", "1804.03999", "1606.06650", "2006.11239", "1809.10486", "1806.05034", "1804.04694", "1711.10684"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_615"} +{"question": "What studies discuss the issue of undefined importance weights in disjoint source and target?", "answer": ["Understanding Self-Training for Gradual Domain Adaptation", "Connect, Not Collapse: Explaining Contrastive Learning for Unsupervised Domain Adaptation"], "answer_arxiv_id": ["2002.11361v1", "2204.00570v4"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_616"} +{"question": "What works emphasize that an LLM's acquired knowledge should mirror established facts?", "answer": ["Alignment for Honesty"], "answer_arxiv_id": ["2312.07000"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_test_617"} +{"question": "What work propose the task of emotional support conversation and release the corresponding dataset?", "answer": ["Towards Emotional Support Dialog Systems"], "answer_arxiv_id": ["2106.01144"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_test_618"} +{"question": "Which papers discussed the task of temporal or timeline summarization?", "answer": ["A Temporally Sensitive Submodularity Framework for Timeline Summarization"], "answer_arxiv_id": ["1810.07949"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_test_619"} +{"question": "Can you name the studies that provided allocation algorithms that guarantee ex-ante and EF1 ex-post for additive valuations?", "answer": ["Best of Both Worlds: Ex-Ante and Ex-Post Fairness in Resource Allocation"], "answer_arxiv_id": ["2005.14122"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_620"} +{"question": "What works used text-based language models to predict text-evoked and speech-evoked brain activity?", "answer": ["Interpreting and improving natural-language processing (in machines)\n with natural language-processing (in the brain)", "Inducing brain-relevant bias in natural language processing models", "Relating Simple Sentence Representations in Deep Neural Networks and the\n Brain", "Low-Dimensional Structure in the Space of Language Representations is\n Reflected in Brain Responses", "Neural Language Taskonomy: Which NLP Tasks are the most Predictive of\n fMRI Brain Activity?", "Language models and brain alignment: beyond word-level semantics and\n prediction", "Joint processing of linguistic properties in brains and language models"], "answer_arxiv_id": ["1905.11833", "1911.03268", "1906.11861", "2106.05426", "2205.01404", "2212.00596", "2212.08094"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_621"} +{"question": "Could you list the studies that discuss the strong correlation between the supervised disentanglement score and the global-basis-compatibility?", "answer": ["Analyzing the Latent Space of GAN through Local Dimension Estimation"], "answer_arxiv_id": ["2205.13182"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_622"} +{"question": "What works initially used a one-time retrieval method but faced knowledge omissions issues in handling knowledge-sensitive tasks?", "answer": ["Improving language models by retrieving from trillions of tokens", "Internet-augmented language models through few-shot prompting for\n open-domain question answering", "Atlas: Few-shot Learning with Retrieval Augmented Language Models"], "answer_arxiv_id": ["2112.04426", "2203.05115", "2208.03299"], "source_meta": {"published_time": "20240628"}, "qid": "AutoScholarQuery_test_623"} +{"question": "Can you provide scholarly works that discuss domain-specific knowledge in Knowledge Graphs?", "answer": ["What Disease does this Patient Have? A Large-scale Open Domain Question Answering Dataset from Medical Exams"], "answer_arxiv_id": ["2009.13081"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_624"} +{"question": "Are there any research papers studying VR HMD settings using public datasets such as AMASS?", "answer": ["AMASS: Archive of Motion Capture as Surface Shapes"], "answer_arxiv_id": ["1904.03278"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_625"} +{"question": "What research paper is related to and compared with the one in this section, where it considers CT model with latent variables, subsections, and an additional loss term for the integration error?", "answer": ["Continuous-time system identification with neural networks: model structures and fitting criteria"], "answer_arxiv_id": ["2006.02915"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_test_626"} +{"question": "Can you provide studies that used transformer-based architectures to tackle German sign language production?", "answer": ["Progressive Transformers for End-to-End Sign Language Production", "Sign Language Transformers: Joint End-to-end Sign Language Recognition\n and Translation"], "answer_arxiv_id": ["2004.14874", "2003.13830"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_627"} +{"question": "Are there any references showing that the chain-of-thoughts generated by LLMs are often arbitrary or contradictory?", "answer": ["Do Models Explain Themselves? Counterfactual Simulatability of Natural\n Language Explanations", "Chain-of-Verification Reduces Hallucination in Large Language Models"], "answer_arxiv_id": ["2307.08678", "2309.11495"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_628"} +{"question": "Which paper applied the Gumbel-Softmax relaxation within gradient-based BOED for contextual optimization?", "answer": ["Efficient Real-world Testing of Causal Decision Making via Bayesian Experimental Design for Contextual Optimisation"], "answer_arxiv_id": ["2207.05250"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_test_629"} +{"question": "Which papers used the BookCorpus, also known as the Toronto Books Corpus, for pretraining?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer"], "answer_arxiv_id": ["1810.04805", "1907.11692", "1910.10683"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_test_630"} +{"question": "Which papers discussed strategies to adjust the spatiotemporal resolution as a way to fine-tune TAD models?", "answer": ["TALLFormer: Temporal Action Localization with a Long-memory Transformer", "Re^2TAL: Rewiring Pretrained Video Backbones for Reversible Temporal\n Action Localization", "An Efficient Spatio-Temporal Pyramid Transformer for Action Detection"], "answer_arxiv_id": ["2204.01680", "2211.14053", "2207.10448"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_631"} +{"question": "What previous works have analyzed errors in model’s reasoning in black-box LLMs?", "answer": ["Measuring Faithfulness in Chain-of-Thought Reasoning", "Self-Contradictory Reasoning Evaluation and Detection", "Towards Understanding Chain-of-Thought Prompting: An Empirical Study of\n What Matters"], "answer_arxiv_id": ["2307.13702", "2311.09603", "2212.10001"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_632"} +{"question": "Could you point me to some papers that discuss the HiPPO framework and its applications in time-series modelling?", "answer": ["HiPPO: Recurrent Memory with Optimal Polynomial Projections", "Efficiently Modeling Long Sequences with Structured State Spaces"], "answer_arxiv_id": ["2008.07669", "2111.00396"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_test_633"} +{"question": "What research has been done in using Deep Neural Networks and graph representations with reinforcement learning in resolving propositional logic problems?", "answer": ["Automated proof synthesis for propositional logic with deep neural\n networks", "Automated Theorem Proving in Intuitionistic Propositional Logic by Deep\n Reinforcement Learning"], "answer_arxiv_id": ["1805.11799", "1811.00796"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_test_634"} +{"question": "Which paper first tackled the pose estimation of novel objects without CAD models using RGB images?", "answer": ["Gen6D: Generalizable Model-Free 6-DoF Object Pose Estimation from RGB\n Images"], "answer_arxiv_id": ["2204.10776"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_test_635"} +{"question": "Can you cite works that have automated QA in online courses?", "answer": ["Agent Smith: Teaching Question Answering to Jill Watson"], "answer_arxiv_id": ["2112.13677v2"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_test_636"} +{"question": "Which studies looked at robustness of MRL to distributional shifts?", "answer": ["Meta-Reinforcement Learning Robust to Distributional Shift via Model Identification and Experience Relabeling", "Distributionally Adaptive Meta Reinforcement Learning"], "answer_arxiv_id": ["2006.07178", "2210.03104"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_test_637"} +{"question": "Which studies have transferred the knowledge from pre-trained vision-language models to object detectors?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation"], "answer_arxiv_id": ["2104.13921"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_638"} +{"question": "Could you provide me some work where the concept of LoRA is proposed?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_639"} +{"question": "Any links to studies considering a single update at each client and focusing on random walks?", "answer": ["Incremental Stochastic Subgradient Algorithms for Convex Optimization", "Private Weighted Random Walk Stochastic Gradient Descent", "Walkman: A Communication-Efficient Random-Walk Algorithm for Decentralized Optimization", "Privacy Amplification by Decentralization"], "answer_arxiv_id": ["0806.1092", "2009.01790", "1804.06568", "2012.05326v4"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_test_640"} +{"question": "Which study introduced the Neural Tangent Kernel (NTK)?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_test_641"} +{"question": "Has anyone tried to generalize the feature embedding approach to adaptive features?", "answer": ["Learning Deep Features in Instrumental Variable Regression", "Deep Proxy Causal Learning and its Application to Confounded Bandit Policy Evaluation"], "answer_arxiv_id": ["2010.07154", "2106.03907"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_test_642"} +{"question": "Who proposed property-aware relation networks (PAR) in the context of few-shot learning?", "answer": ["Property-Aware Relation Networks for Few-Shot Molecular Property Prediction"], "answer_arxiv_id": ["2107.07994"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_test_643"} +{"question": "What study models reasoning procedures as BFS or DFS search on reasoning trees?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2305.10601"], "source_meta": {"published_time": "20240628"}, "qid": "AutoScholarQuery_test_644"} +{"question": "Which studies focus on evaluation of a targeted update in knowledge updating?", "answer": ["Locating and Editing Factual Associations in GPT", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2202.05262", "2110.11309"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_645"} +{"question": "What works describe the role of the Neural Tangent Kernel in the density of neural networks?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_test_646"} +{"question": "What studies analyze the effect of multiple regularizations employed in deep learning, like weight decay, early stopping, or drop-outs, on the generalization abilities?", "answer": ["Norm-Based Capacity Control in Neural Networks", "Dropout Rademacher Complexity of Deep Neural Networks"], "answer_arxiv_id": ["1503.00036", "1402.3811"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_test_647"} +{"question": "Could you provide me some studies that relate to early works on aligning text and image embeddings?", "answer": ["Deep Visual-Semantic Alignments for Generating Image Descriptions"], "answer_arxiv_id": ["1412.2306"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_test_648"} +{"question": "Which paper introduced joint differential privacy (JDP)?", "answer": ["Robust Mediators in Large Games"], "answer_arxiv_id": ["1512.02698v2"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_test_649"} +{"question": "Which research papers explored the implementation of state space models-based architectures?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces", "Diagonal State Spaces are as Effective as Structured State Spaces", "On the Parameterization and Initialization of Diagonal State Space\n Models"], "answer_arxiv_id": ["2111.00396", "2203.14343", "2206.11893"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_test_650"} +{"question": "Which papers propose goal-driven forecasting to predict goal locations for future human walking trajectories?", "answer": ["Long-term Human Motion Prediction with Scene Context"], "answer_arxiv_id": ["2007.03672"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_651"} +{"question": "Could you list the works exploring the methodologies to effectively utilize the inherent prior knowledge within CAD?", "answer": ["Counterfactually-Augmented SNLI Training Data Does Not Yield Better\n Generalization Than Unaugmented Data", "An Investigation of the (In)effectiveness of Counterfactually Augmented\n Data", "Unlock the Potential of Counterfactually-Augmented Data in\n Out-Of-Distribution Generalization"], "answer_arxiv_id": ["2010.04762", "2107.00753", "2310.06666"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_test_652"} +{"question": "Which papers discuss the concept of semantic meaning and algebraic structure of popular embeddings?", "answer": ["Towards Understanding Linear Word Analogies", "Linear Spaces of Meanings: Compositional Structures in Vision-Language Models", "Prompt Algebra for Task Composition"], "answer_arxiv_id": ["1810.04882", "2302.14383", "2306.00310"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_test_653"} +{"question": "Which works illustrate that transformer-based LLMs have been trained on diverse, large-scale datasets to simultaneously learn several language understanding tasks?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["1910.10683", "2210.11416"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_test_654"} +{"question": "What studies discuss techniques for data efficiency in deep learning?", "answer": ["Low-Shot Learning from Imaginary Data", "Hallucination Improves the Performance of Unsupervised Visual Representation Learning", "GenCo: An Auxiliary Generator from Contrastive Learning for Enhanced Few-Shot Learning in Remote Sensing"], "answer_arxiv_id": ["1801.05401", "2307.12168", "2307.14612"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_test_655"} +{"question": "Could you provide me some works that focus on improving the quality of input reconstruction?", "answer": ["Inverting Gradients - How easy is it to break privacy in federated learning?", "See through Gradients: Image Batch Recovery via GradInversion", "Gradient Inversion with Generative Image Prior"], "answer_arxiv_id": ["2003.14053", "2104.07586", "2110.14962"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_656"} +{"question": "Could you provide me some works about model averaging methods that are further related to federated learning and ensemble methods?", "answer": ["Federated learning with matched averaging", "Model Fusion via Optimal Transport", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time"], "answer_arxiv_id": ["2002.06440", "1910.05653", "2203.05482"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_test_657"} +{"question": "Are there studies focused on mitigation of hallucinations during the pre-training stage by the process of dataset curating and cleaning?", "answer": ["Diving Deep into Modes of Fact Hallucinations in Dialogue Systems", "HAGRID: A Human-LLM Collaborative Dataset for Generative\n Information-Seeking with Attribution", "Med-HALT: Medical Domain Hallucination Test for Large Language Models"], "answer_arxiv_id": ["2301.04449", "2307.16883", "2307.15343"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_test_658"} +{"question": "Could you provide examples of image-text datasets that have their own preprocessing techniques?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Combined Scaling for Zero-shot Transfer Learning", "Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts", "RedCaps: Web-curated image-text data created by the people, for the people", "LAION-5B: An open large-scale dataset for training next generation image-text models"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.10050", "2102.08981", "2111.11431", "2210.08402"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_659"} +{"question": "Which works consider the instruction-following abilities of LLMs?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2302.13971", "2307.09288"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_test_660"} +{"question": "Could you point out the studies which discusses the high variance problem of likelihood-ratio gradient?", "answer": ["High-Dimensional Continuous Control Using Generalized Advantage Estimation"], "answer_arxiv_id": ["1506.02438"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_661"} +{"question": "What papers have introduced single-camera methods for egocentric pose estimation?", "answer": ["Mo2Cap2: Real-time Mobile 3D Motion Capture with a Cap-mounted Fisheye\n Camera", "xR-EgoPose: Egocentric 3D Human Pose from an HMD Camera"], "answer_arxiv_id": ["1803.05959", "1907.10045"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_662"} +{"question": "Which works propose a novel framework or expand the idea to obtain complex and difficult instructions gradually?", "answer": ["WizardLM: Empowering Large Language Models to Follow Complex\n Instructions", "WizardCoder: Empowering Code Large Language Models with Evol-Instruct", "Textbooks Are All You Need"], "answer_arxiv_id": ["2304.12244", "2306.08568", "2306.11644"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_663"} +{"question": "Could you provide me some studies that have applied the concept of teacher-student network?", "answer": ["FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Distilling the Knowledge in a Neural Network", "Knowledge Distillation: A Survey", "Semi-supervised semantic segmentation needs strong, varied perturbations", "Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning", "PseudoSeg: Designing Pseudo Labels for Semantic Segmentation", "Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic\n Segmentation", "Unbiased Teacher for Semi-Supervised Object Detection", "Humble Teachers Teach Better Students for Semi-Supervised Object\n Detection", "Distilling Vision-Language Pre-training to Collaborate with\n Weakly-Supervised Temporal Action Localization", "End-to-End Semi-Supervised Object Detection with Soft Teacher"], "answer_arxiv_id": ["2001.07685v2", "1503.02531", "2006.05525", "1906.01916", "2110.05474", "2010.09713", "2208.09910", "2102.09480", "2106.10456", "2212.09335", "2106.09018"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_test_664"} +{"question": "What papers extended the methods to address pose estimation for novel objects or low-textured objects?", "answer": ["OnePose: One-Shot Object Pose Estimation without CAD Models", "OnePose++: Keypoint-Free One-Shot Object Pose Estimation without CAD\n Models"], "answer_arxiv_id": ["2205.12257", "2301.07673"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_test_665"} +{"question": "What studies propose generative adversarial networks for 3D shape generation?", "answer": ["Learning Representations and Generative Models for 3D Point Clouds", "3D Point Cloud Generative Adversarial Network Based on Tree Structured Graph Convolutions"], "answer_arxiv_id": ["1707.02392", "1905.06292"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_test_666"} +{"question": "What paper demonstrated that a model trained on synthetic captions can perform better than one trained on human-provided captions?", "answer": ["Is a Caption Worth a Thousand Images? A Controlled Study for Representation Learning"], "answer_arxiv_id": ["2207.07635"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_667"} +{"question": "Are there any studies that added a few trainable parameters representing new knowledge to LLMs, while keeping the original parameters frozen?", "answer": ["Calibrating Factual Knowledge in Pretrained Language Models", "Transformer-Patcher: One Mistake worth One Neuron", "Rank-One Editing of Encoder-Decoder Models", "Neural Knowledge Bank for Pretrained Transformers"], "answer_arxiv_id": ["2210.03329", "2301.09785", "2211.13317", "2208.00399"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_668"} +{"question": "Can you provide references where global non-convex optimization is applied in machine learning?", "answer": ["X-Armed Bandits", "Global optimization of Lipschitz functions"], "answer_arxiv_id": ["1001.4475", "1703.02628"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_test_669"} +{"question": "Which works are known for tackling various tasks simultaneously in the field of vision-language models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Flamingo: a Visual Language Model for Few-Shot Learning", "Towards General Purpose Vision Systems", "Class-agnostic Object Detection with Multi-modal Transformer"], "answer_arxiv_id": ["2103.00020", "2204.14198", "2104.00743", "2111.11430"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_test_670"} +{"question": "What datasets are used for knowledge editing?", "answer": ["Zero-Shot Relation Extraction via Reading Comprehension", "Locating and Editing Factual Associations in GPT", "MQuAKE: Assessing Knowledge Editing in Language Models via Multi-Hop\n Questions", "Eva-KELLM: A New Benchmark for Evaluating Knowledge Editing of LLMs"], "answer_arxiv_id": ["1706.04115", "2202.05262", "2305.14795", "2308.09954"], "source_meta": {"published_time": "20230916"}, "qid": "AutoScholarQuery_test_671"} +{"question": "What works proposed a method to learn a linear subspace for the disentanglement of written text from visual components?", "answer": ["Disentangling visual and written concepts in CLIP"], "answer_arxiv_id": ["2206.07835"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_672"} +{"question": "Could you provide me some studies about the use of case-based reasoning?", "answer": ["Case-Based Reasoning for Natural Language Queries over Knowledge Bases", "CBR-iKB: A Case-Based Reasoning Approach for Question Answering over Incomplete Knowledge Bases", "Knowledge Base Question Answering by Case-based Reasoning over Subgraphs"], "answer_arxiv_id": ["2104.08762", "2204.08554", "2202.10610"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_673"} +{"question": "Any existing research on generating the 3D human avatars with predefined parametric human templates?", "answer": ["AvatarGen: A 3D Generative Model for Animatable Human Avatars", "EVA3D: Compositional 3D Human Generation from 2D Image Collections", "Unsupervised Learning of Efficient Geometry-Aware Neural Articulated\n Representations", "Generative Neural Articulated Radiance Fields", "3D-Aware Semantic-Guided Generative Model for Human Synthesis"], "answer_arxiv_id": ["2211.14589", "2210.04888", "2204.08839", "2206.14314", "2112.01422"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_test_674"} +{"question": "What papers provided a theoratical analysis of the mechanism with respect to mean squared error of the noise?", "answer": ["Almost Tight Error Bounds on Differentially Private Continual Counting"], "answer_arxiv_id": ["2211.05006"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_test_675"} +{"question": "Which papers propose solutions for panel detection in comic understanding?", "answer": ["Object Detection for Comics using Manga109 Annotations"], "answer_arxiv_id": ["1803.08670"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_test_676"} +{"question": "What studies reported varying performance of LLMs when generating different types of reasoning processes?", "answer": ["Program of Thoughts Prompting: Disentangling Computation from Reasoning\n for Numerical Reasoning Tasks", "PAL: Program-aided Language Models", "Tab-CoT: Zero-shot Tabular Chain of Thought"], "answer_arxiv_id": ["2211.12588", "2211.10435", "2305.17812"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_677"} +{"question": "Which papers propose heuristic-based uncertainty metrics for generative Large Language Models (LLMs) considering machine translation?", "answer": ["Wat zei je? Detecting Out-of-Distribution Translations with Variational\n Transformers", "Unsupervised Quality Estimation for Neural Machine Translation"], "answer_arxiv_id": ["2006.08344", "2005.10608"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_678"} +{"question": "Which studies focus on 3D pose estimation task through a single RGB image input?", "answer": ["End-to-end Recovery of Human Shape and Pose", "Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the\n Loop"], "answer_arxiv_id": ["1712.06584", "1909.12828"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_test_679"} +{"question": "Which papers propose methods for audio-visual segmentation task?", "answer": ["Class-aware Sounding Objects Localization via Audiovisual Correspondence", "Discriminative Sounding Objects Localization via Self-supervised\n Audiovisual Matching", "Deep Multimodal Clustering for Unsupervised Audiovisual Learning", "Unsupervised Sound Localization via Iterative Contrastive Learning", "Localizing Visual Sounds the Hard Way", "Self-Supervised Predictive Learning: A Negative-Free Method for Sound Source Localization in Visual Scenes", "Exploiting Transformation Invariance and Equivariance for\n Self-supervised Sound Localisation", "Learning to Localize Sound Source in Visual Scenes", "Multiple Sound Sources Localization from Coarse to Fine", "Annotation-free Audio-Visual Segmentation"], "answer_arxiv_id": ["2112.11749", "2010.05466", "1807.03094", "2104.00315", "2104.02691", "2203.13412v1", "2206.12772", "1803.03849", "2007.06355", "2305.11019"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_test_680"} +{"question": "Which works are variants of the FedAvg algorithm aimed to resolve the minimization problem in federated learning?", "answer": ["Local SGD Converges Fast and Communicates Little", "Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning", "On the Linear Speedup Analysis of Communication Efficient Momentum SGD for Distributed Non-Convex Optimization"], "answer_arxiv_id": ["1805.09767", "1807.06629", "1905.03817"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_test_681"} +{"question": "What was the research that put forward a data augmentation-dependent method for contrastive learning for object-centric representations?", "answer": ["Towards Self-Supervised Learning of Global and Object-Centric Representations"], "answer_arxiv_id": ["2203.05997"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_682"} +{"question": "Which studies focus on table-based EHR question answering?", "answer": ["Question Answering for Complex Electronic Health Records Database using Unified Encoder-Decoder Architecture", "LeafAI: query generator for clinical cohort discovery rivaling a human programmer", "EHRSQL: A Practical Text-to-SQL Benchmark for Electronic Health Records", "Towards Understanding the Generalization of Medical Text-to-SQL Models and Datasets", "Text-to-SQL Generation for Question Answering on Electronic Medical Records"], "answer_arxiv_id": ["2111.14703", "2304.06203v2", "2301.07695", "2303.12898", "1908.01839"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_test_683"} +{"question": "Can you provide examples of studies about the methodology of studying games that gradually change over time?", "answer": ["Econometrics for Learning Agents", "Learning and Efficiency in Games with Dynamic Population", "An Experimental Evaluation of Regret-Based Econometrics", "Dynamic network congestion games"], "answer_arxiv_id": ["1505.00720", "1505.00391", "1605.03838", "2009.13632v1"], "source_meta": {"published_time": "20220928"}, "qid": "AutoScholarQuery_test_684"} +{"question": "Are there any studies where the shuffling-based method was applied to FL?", "answer": ["Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond", "On the Convergence of Federated Averaging with Cyclic Client Participation"], "answer_arxiv_id": ["2110.10342", "2302.03109v1"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_test_685"} +{"question": "Can you list the research works where the number of timesteps used to train SNNs have been reduced?", "answer": ["Enabling Deep Spiking Neural Networks with Hybrid Conversion and Spike Timing Dependent Backpropagation", "ESL-SNNs: An Evolutionary Structure Learning Strategy for Spiking Neural Networks", "Going Deeper With Directly-Trained Larger Spiking Neural Networks"], "answer_arxiv_id": ["2005.01807", "2306.03693", "2011.05280"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_test_686"} +{"question": "Could you mention the studies that focused on lifting 2D pre-trained models to create 3D models from textual prompts?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "ATT3D: Amortized Text-to-3D Object Synthesis", "ZeroAvatar: Zero-shot 3D Avatar Generation from a Single Image", "MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion", "Sparse3D: Distilling Multiview-Consistent Diffusion for Object\n Reconstruction from Sparse Views", "MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2211.10440", "2303.13873", "2305.16213", "2212.00774v1", "2306.07349", "2305.16411", "2307.01097", "2308.14078", "2308.16512"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_687"} +{"question": "Which studies have discussed a two-step approach for Temporal Action Detection (TAD)?", "answer": ["BMN: Boundary-Matching Network for Temporal Action Proposal Generation", "BSN: Boundary Sensitive Network for Temporal Action Proposal Generation", "G-TAD: Sub-Graph Localization for Temporal Action Detection", "Video Self-Stitching Graph Network for Temporal Action Localization"], "answer_arxiv_id": ["1907.09702", "1806.02964", "1911.11462", "2011.14598"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_688"} +{"question": "Could you provide me the study that constructed the emrKBQA dataset for patient-specific QA on MIMIC-III?", "answer": ["emrQA: A Large Corpus for Question Answering on Electronic Medical Records"], "answer_arxiv_id": ["1809.00732v1"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_test_689"} +{"question": "What research papers are provided for additive valuations with binary marginals?", "answer": ["A Probabilistic Approach to Voting, Allocation, Matching, and Coalition Formation", "Fair Division with Binary Valuations: One Rule to Rule Them All"], "answer_arxiv_id": ["2002.10171", "2007.06073"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_690"} +{"question": "What research studies propose changing the weight in convolutional layers according to input features using the attention mechanism?", "answer": ["CondConv: Conditionally Parameterized Convolutions for Efficient Inference", "Dynamic Neural Networks: A Survey"], "answer_arxiv_id": ["1904.04971", "2102.04906"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_test_691"} +{"question": "Can you inform me about papers which suggested that pretraining on domain-specific text can enhance language model performance on related tasks?", "answer": ["SciBERT: A Pretrained Language Model for Scientific Text", "Domain-Specific Language Model Pretraining for Biomedical Natural\n Language Processing", "LEGAL-BERT: The Muppets straight out of Law School", "BudgetLongformer: Can we Cheaply Pretrain a SotA Legal Language Model\n From Scratch?"], "answer_arxiv_id": ["1903.10676", "2007.15779", "2010.02559", "2211.17135"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_test_692"} +{"question": "Which papers have discussed Bound Propagation methods and analyzed the output bounds based on input bounds?", "answer": ["Semidefinite relaxations for certifying robustness to adversarial examples", "Provable Defenses against Adversarial Examples via the Convex Outer Adversarial Polytope", "Efficient Neural Network Robustness Certification with General Activation Functions", "Certifiable Robustness and Robust Training for Graph Convolutional Networks", "Collective Robustness Certificates: Exploiting Interdependence in Graph Neural Networks"], "answer_arxiv_id": ["1811.01057", "1711.00851", "1811.00866", "1906.12269", "2302.02829"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_test_693"} +{"question": "What studies present the application of the measurement theory concepts to NLP where desirable model capabilities are unobservable constructs?", "answer": ["Evaluating Evaluation Metrics: A Framework for Analyzing NLG Evaluation\n Metrics using Measurement Theory"], "answer_arxiv_id": ["2305.14889"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_test_694"} +{"question": "Any works about continual pretraining for encoder-decoder LMs?", "answer": ["KILM: Knowledge Injection into Encoder-Decoder Language Models", "Lifelong Pretraining: Continually Adapting Language Models to Emerging Corpora"], "answer_arxiv_id": ["2302.09170", "2110.08534"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_695"} +{"question": "Which study proposed to minimize a robust loss and find worst-case perturbation during training with projected gradient descent?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_test_696"} +{"question": "Which works have investigated the advantages of decoupling policy and value functions for generalization in RL?", "answer": ["Decoupling Value and Policy for Generalization in Reinforcement Learning", "Phasic Policy Gradient"], "answer_arxiv_id": ["2102.10330", "2009.04416"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_test_697"} +{"question": "Which work considered the notion of envy-freeness up to any item (EFX) in fractionally subadditive valuations?", "answer": ["Almost Envy-Freeness with General Valuations"], "answer_arxiv_id": ["1707.04769"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_698"} +{"question": "Which papers demonstrate story generation strategies for better coherency in story-telling using language models?", "answer": ["Plan-And-Write: Towards Better Automatic Storytelling"], "answer_arxiv_id": ["1811.05701"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_test_699"} +{"question": "Which studies utilized the concept of query for different applications like detection of different objects, video instance segmentation, multiple object tracking, and video object detection?", "answer": ["End-to-End Object Detection with Transformers", "MOTR: End-to-End Multiple-Object Tracking with Transformer", "TrackFormer: Multi-Object Tracking with Transformers", "End-to-End Video Instance Segmentation with Transformers"], "answer_arxiv_id": ["2005.12872", "2105.03247", "2101.02702", "2011.14503"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_test_700"} +{"question": "Can you list some research papers that adopted a similar formulation for advertising?", "answer": ["Send Mixed Signals – Earn More, Work Less", "Signaling Schemes for Revenue Maximization", "Screening with Persuasion"], "answer_arxiv_id": ["1202.1483", "1202.1590", "2212.03360v1"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_test_701"} +{"question": "What papers use LLM to retrieve reasoning paths from KGs based on relation 'plans' grounded by KGs?", "answer": ["Reasoning on Graphs: Faithful and Interpretable Large Language Model\n Reasoning"], "answer_arxiv_id": ["2310.01061"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_test_702"} +{"question": "What study focused on examining node classification performance of one-layer GATs on a random graph model?", "answer": ["Graph Attention Retrospective"], "answer_arxiv_id": ["2202.13060"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_test_703"} +{"question": "What works proposed variants on normalisation, which is complementary to unit scaling?", "answer": ["Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift", "Layer Normalization", "Proxy-Normalizing Activations to Match Batch Normalization while Removing Batch Dependence", "Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks"], "answer_arxiv_id": ["1502.03167", "1607.06450", "2106.03743", "1602.07868"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_test_704"} +{"question": "Which papers introduced algorithms regarding distributed optimization in a full participation setting using deterministic methods?", "answer": ["Communication Efficient Distributed Optimization using an Approximate Newton-type Method", "AIDE: Fast and Communication Efficient Distributed Optimization", "On Convergence of Distributed Approximate Newton Methods: Globalization, Sharper Bounds and Beyond", "Statistically Preconditioned Accelerated Gradient Method for Distributed Optimization", "An Accelerated Second-Order Method for Distributed Stochastic Optimization", "Newton Method over Networks is Fast up to the Statistical Precision"], "answer_arxiv_id": ["1312.7853", "1608.06879v1", "1908.02246", "2002.10726", "2103.14392", "2102.06780"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_test_705"} +{"question": "Which works used alignment approaches to capture the feature of domain invariant characteristics?", "answer": ["CyCADA: Cycle-Consistent Adversarial Domain Adaptation", "Progressive Feature Alignment for Unsupervised Domain Adaptation", "FCNs in the Wild: Pixel-level Adversarial and Constraint-based\n Adaptation", "Both Style and Distortion Matter: Dual-Path Unsupervised Domain\n Adaptation for Panoramic Semantic Segmentation"], "answer_arxiv_id": ["1711.03213", "1811.08585", "1612.02649", "2303.14360"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_test_706"} +{"question": "Which works extended the binary meta-learners in the CATE's estimation without a thorough theoretical analysis of their behaviour?", "answer": ["Uplift Modeling for Multiple Treatments with Cost Optimization"], "answer_arxiv_id": ["1908.05372"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_test_707"} +{"question": "Can you specify the papers that introduced novel pre-training tasks for transformer-based pre-trained models?", "answer": ["CodeBERT: A Pre-Trained Model for Programming and Natural Languages", "CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for\n Code Understanding and Generation", "UniXcoder: Unified Cross-Modal Pre-training for Code Representation"], "answer_arxiv_id": ["2002.08155", "2109.00859", "2203.03850"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_test_708"} +{"question": "Can you provide me some studies that use clustering or topic modeling to identify aspects in documents?", "answer": ["Modeling Online Reviews with Multi-grain Topic Models"], "answer_arxiv_id": ["0801.1063"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_709"} +{"question": "What studies utilized VGG, ResNet, LSTM, and customized loss functions to enhance vision tasks in deep learning?", "answer": ["Very Deep Convolutional Networks for Large-Scale Image Recognition", "Deep Residual Learning for Image Recognition", "Image-based localization using LSTMs for structured feature correlation"], "answer_arxiv_id": ["1409.1556", "1512.03385", "1611.07890"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_710"} +{"question": "What works are about data-driven LiDAR simulators?", "answer": ["LiDARsim: Realistic LiDAR Simulation by Leveraging the Real World", "Pattern-Aware Data Augmentation for LiDAR 3D Object Detection", "VISTA 2.0: An Open, Data-driven Simulator for Multimodal Sensing and\n Policy Learning for Autonomous Vehicles", "Learning Interactive Driving Policies via Data-driven Simulation"], "answer_arxiv_id": ["2006.09348", "2112.00050", "2111.12083", "2111.12137"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_test_711"} +{"question": "Can you specify some multimodal pretraining methods that require paired or interleaved data?", "answer": ["VLMo: Unified Vision-Language Pre-Training with\n Mixture-of-Modality-Experts", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "Image as a Foreign Language: BEiT Pretraining for All Vision and\n Vision-Language Tasks", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2111.02358", "2108.10904", "2208.10442", "2204.14198"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_test_712"} +{"question": "What references mention the use of generative adversarial networks (GANs) in recent advancements in reconstruction from fMRI techniques?", "answer": ["Mind Reader: Reconstructing complex images from brain activities", "Reconstructing Natural Scenes from fMRI Patterns using BigBiGAN", "Reconstruction of Perceived Images from fMRI Patterns and Semantic Brain\n Exploration using Instance-Conditioned GANs"], "answer_arxiv_id": ["2210.01769", "2001.11761", "2202.12692"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_test_713"} +{"question": "Which papers discuss the two mainstream approaches for multi-hop KGQA: Semantic Parsing and Information Retrieval?", "answer": ["A Survey on Complex Knowledge Base Question Answering: Methods, Challenges and Solutions"], "answer_arxiv_id": ["2105.11644"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_714"} +{"question": "Which publications introduced the maximum entropy exploration (maxEnt) framework?", "answer": ["Provably Efficient Maximum Entropy Exploration", "Task-Agnostic Exploration via Policy Gradient of a Non-Parametric State Entropy Estimate"], "answer_arxiv_id": ["1812.02690", "2007.04640"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_test_715"} +{"question": "What papers recently gave attention to maximum entropy policies in the context of reinforcement learning (RL)?", "answer": ["Behavior From the Void: Unsupervised Active Pre-Training", "APS: Active Pretraining with Successor Features", "Reinforcement Learning with Prototypical Representations", "State Entropy Maximization with Random Encoders for Efficient Exploration", "Provably Efficient Maximum Entropy Exploration", "Task-Agnostic Exploration via Policy Gradient of a Non-Parametric State Entropy Estimate"], "answer_arxiv_id": ["2103.04551", "2108.13956", "2102.11271v2", "2102.09430", "1812.02690", "2007.04640"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_test_716"} +{"question": "Which research papers have introduced more sophisticated network structures for the mapping between hazy and clear images in the context of image dehazing?", "answer": ["TransWeather: Transformer-based Restoration of Images Degraded by\n Adverse Weather Conditions", "MB-TaylorFormer: Multi-branch Efficient Transformer Expanded by Taylor\n Formula for Image Dehazing"], "answer_arxiv_id": ["2111.14813", "2308.14036"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_test_717"} +{"question": "Which papers discuss that Large Language Models (LLMs) memorize data both from their original large training corpora and smaller private datasets used for downstream tasks?", "answer": ["Quantifying Memorization Across Neural Language Models", "Preventing Verbatim Memorization in Language Models Gives a False Sense of Privacy", "How BPE Affects Memorization in Transformers", "Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Models", "Counterfactual Memorization in Neural Language Models", "Memorization in NLP Fine-tuning Methods"], "answer_arxiv_id": ["2202.07646", "2210.17546v3", "2110.02782", "2205.10770", "2112.12938", "2205.12506"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_718"} +{"question": "Which papers exploited feature-space/n-gram discrepancy measures for selecting data in domain adaptation setting?", "answer": ["Learning to select data for transfer learning with Bayesian Optimization"], "answer_arxiv_id": ["1707.05246"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_719"} +{"question": "Any works that employed a consistency-based method for confidence estimation?", "answer": ["Fine-tuning Language Models for Factuality"], "answer_arxiv_id": ["2311.08401"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_test_720"} +{"question": "Could you mention a study that proposed to decouple the bi-level optimization of dataset condensation?", "answer": ["Squeeze, Recover and Relabel: Dataset Condensation at ImageNet Scale\n From A New Perspective"], "answer_arxiv_id": ["2306.13092"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_721"} +{"question": "What research papers have been involved in the use of kernel fusion for efficient inference techniques in large language models (LLMs)?", "answer": ["FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness"], "answer_arxiv_id": ["2205.14135"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_722"} +{"question": "What works used Moore-Lewis selection in selection of examples?", "answer": ["Cynical Selection of Language Model Training Data", "Automatic Document Selection for Efficient Encoder Pretraining"], "answer_arxiv_id": ["1709.02279", "2210.10951"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_723"} +{"question": "Which work first introduced the wait-k policy for simultaneous text translation?", "answer": ["STACL: Simultaneous Translation with Implicit Anticipation and Controllable Latency using Prefix-to-Prefix Framework"], "answer_arxiv_id": ["1810.08398v5"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_test_724"} +{"question": "In which study did the researchers design the Poolingformer technique?", "answer": ["Poolingformer: Long Document Modeling with Pooling Attention"], "answer_arxiv_id": ["2105.04371"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_test_725"} +{"question": "Which works studied the offline total variation denoising problem?", "answer": ["Adaptive piecewise polynomial estimation via trend filtering", "Trend Filtering on Graphs", "Adaptive Risk Bounds in Univariate Total Variation Denoising and Trend Filtering"], "answer_arxiv_id": ["1304.2986", "1410.7690", "1702.05113"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_726"} +{"question": "Who initially introduced the Diffusion Generative Model?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_test_727"} +{"question": "What studies have applied machine learning models for connectivity analysis on brain networks?", "answer": ["Machine Learning Methods for Brain Network Classification: Application to Autism Diagnosis using Cortical Morphological Networks", "Explainable Classification of Brain Networks via Contrast Subgraphs"], "answer_arxiv_id": ["2004.13321", "2006.05176"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_test_728"} +{"question": "Could you provide me some works that involve the application of contrastive learning in diverse data modalities?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2002.05709", "2103.00020"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_test_729"} +{"question": "What work introduced a K-Planes decomposition technique designed to reconstruct static 3D scenes and dynamic 4D videos?", "answer": ["K-Planes: Explicit Radiance Fields in Space, Time, and Appearance"], "answer_arxiv_id": ["2301.10241"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_test_730"} +{"question": "In which papers did the researchers discuss the KRR in misspecified case?", "answer": ["Statistical Optimality of Stochastic Gradient Descent on Hard Learning Problems through Multiple Passes", "Sobolev Norm Learning Rates for Regularized Least-Squares Algorithms", "Analyzing the discrepancy principle for kernelized spectral filter learning algorithms", "Optimal Rates for Regularized Conditional Mean Embedding Learning"], "answer_arxiv_id": ["1805.10074", "1702.07254", "2004.08436v1", "2208.01711v3"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_test_731"} +{"question": "Which papers implemented neural networks like CNNs and RNNs to enhance co-embedding methods?", "answer": ["A ConvNet for the 2020s", "Deep Residual Learning for Image Recognition", "Going Deeper with Convolutions", "Very Deep Convolutional Networks for Large-Scale Image Recognition", "Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term\n Memory (LSTM) Network"], "answer_arxiv_id": ["2201.03545", "1512.03385", "1409.4842", "1409.1556", "1808.03314"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_732"} +{"question": "Which papers approached the problem of Source-free Unsupervised Domain Adaptation (SFUDA) in the absence of source domain data?", "answer": ["Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentation", "Model Adaptation: Historical Contrastive Learning for Unsupervised\n Domain Adaptation without Source Data"], "answer_arxiv_id": ["2108.11249", "2110.03374"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_test_733"} +{"question": "Could you provide me some works about training convolutional networks for 6D Pose Estimation?", "answer": ["Mask R-CNN"], "answer_arxiv_id": ["1703.06870"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_734"} +{"question": "Could you provide studies about ensemble methods for yielding pixel-wise uncertainty estimates?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1612.01474"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_test_735"} +{"question": "What papers propose certified defenses which conduct a layer-by-layer analysis to derive the certified robustness guarantee of an unimodal model?", "answer": ["Provable defenses against adversarial examples via the convex outer\n adversarial polytope", "On the Effectiveness of Interval Bound Propagation for Training\n Verifiably Robust Models", "Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks"], "answer_arxiv_id": ["1711.00851", "1810.12715", "1702.01135"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_736"} +{"question": "Could you provide me some studies that have addressed self-consistency in Large Language Models?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_test_737"} +{"question": "What datasets contain scenes where multiple people are performing various actions concurrently?", "answer": ["AVA: A Video Dataset of Spatio-temporally Localized Atomic Visual\n Actions", "The AVA-Kinetics Localized Human Actions Video Dataset", "MultiSports: A Multi-Person Video Dataset of Spatio-Temporally Localized\n Sports Actions"], "answer_arxiv_id": ["1705.08421", "2005.00214", "2105.07404"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_test_738"} +{"question": "What papers are about analyzing and understanding conventions in overcoming coordination problems?", "answer": ["Legible Normativity for AI Alignment: The Value of Silly Rules"], "answer_arxiv_id": ["1811.01267"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_739"} +{"question": "Could you give an example of research papers on techniques for distinguishing uncertain areas as the dimension of the goal space increases, in the field of curriculum goal generation?", "answer": ["Uncertainty-Aware Anticipation of Activities", "BeLFusion: Latent Diffusion for Behavior-Driven Human Motion Prediction"], "answer_arxiv_id": ["1908.09540", "2211.14304"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_740"} +{"question": "What are some studies about Operator Learning that leverage the Fourier transform?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations", "Factorized Fourier Neural Operators"], "answer_arxiv_id": ["2010.08895", "2111.13802"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_test_741"} +{"question": "What work proposed the eNCE method and identified its limitations?", "answer": ["Analyzing and Improving the Optimization Landscape of Noise-Contrastive Estimation"], "answer_arxiv_id": ["2110.11271"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_742"} +{"question": "Which publications use autoregressive models for molecular structure design?", "answer": ["Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules", "Generating equilibrium molecules with deep neural networks"], "answer_arxiv_id": ["1906.00957", "1810.11347"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_743"} +{"question": "Could you provide me some studies exploring to make Gaussian diffusion faster and more data efficient in training?", "answer": ["Patch Diffusion: Faster and More Data-Efficient Training of Diffusion Models", "Efficient Diffusion Training via Min-SNR Weighting Strategy", "Fast Sampling of Diffusion Models via Operator Learning", "Fast Diffusion Model"], "answer_arxiv_id": ["2304.12526", "2303.09556", "2211.13449", "2306.06991"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_test_744"} +{"question": "What research discusses the advantages of the PRM over the ORM in providing detailed feedback to enhance generators?", "answer": ["Fine-Grained Human Feedback Gives Better Rewards for Language Model\n Training"], "answer_arxiv_id": ["2306.01693"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_745"} +{"question": "What research studies constructed datasets targeting multiple image issues and proposed context schemes to better understand interleaved inputs?", "answer": ["MMICL: Empowering Vision-language Model with Multi-Modal In-Context\n Learning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2309.07915", "2305.03726"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_746"} +{"question": "Can you name some studies that tackled the objective mismatch issue in RLHF learning schemes?", "answer": ["The Alignment Ceiling: Objective Mismatch in Reinforcement Learning from\n Human Feedback"], "answer_arxiv_id": ["2311.00168"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_test_747"} +{"question": "Can you provide papers that discuss the important role of the noise distribution choice in success of Noise-contrastive estimation?", "answer": ["Generative Adversarial Nets", "Flow Contrastive Estimation of Energy-Based Models", "Analyzing and Improving the Optimization Landscape of Noise-Contrastive Estimation"], "answer_arxiv_id": ["1406.2661", "1912.00589", "2110.11271"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_748"} +{"question": "Which paper introduced the slot attention for object-centric learning?", "answer": ["Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["2006.15055"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_749"} +{"question": "Can you show me some papers that focus on improving sample efficiency with sub-optimal external policies?", "answer": ["KoGuN: Accelerating Deep Reinforcement Learning via Integrating Human Suboptimal Knowledge"], "answer_arxiv_id": ["2002.07418"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_test_750"} +{"question": "Which works studied the concept of co-training involving separate models to generate improved pseudolabels?", "answer": ["Co-teaching: Robust Training of Deep Neural Networks with Extremely\n Noisy Labels", "Co-training Improves Prompt-based Learning for Large Language Models"], "answer_arxiv_id": ["1804.06872", "2202.00828"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_test_751"} +{"question": "Can you provide references regarding data-driven approaches for stereo-matching?", "answer": ["A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation", "Unifying Flow, Stereo and Depth Estimation", "Pyramid Stereo Matching Network", "GA-Net: Guided Aggregation Net for End-to-end Stereo Matching", "A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation"], "answer_arxiv_id": ["1512.02134", "2211.05783", "1803.08669", "1904.06587", "1512.02134"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_test_752"} +{"question": "What papers focus on developing kernel methods for learning molecular potentials?", "answer": ["On representing chemical environments", "FCHL revisited: faster and more accurate quantum machine learning"], "answer_arxiv_id": ["1209.3140", "1909.01946"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_test_753"} +{"question": "What early works utilize CNN-based image translation in portrait and face relighting?", "answer": ["Deep Shading: Convolutional Neural Networks for Screen-Space Shading", "Deferred Neural Rendering: Image Synthesis using Neural Textures"], "answer_arxiv_id": ["1603.06078", "1904.12356"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_test_754"} +{"question": "Which papers discuss the application of specific criteria to remove weights in post-hoc pruning?", "answer": ["Dynamic Network Surgery for Efficient DNNs", "Learning to Prune Deep Neural Networks via Layer-wise Optimal Brain Surgeon", "Compressing Neural Networks using the Variational Information Bottleneck", "NISP: Pruning Networks using Neuron Importance Score Propagation", "Importance Estimation for Neural Network Pruning"], "answer_arxiv_id": ["1608.04493", "1705.07565", "1802.10399", "1711.05908", "1906.10771"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_test_755"} +{"question": "Can you identify any works that aimed to improve computationally efficient FL with personalized local models using quantization and model parameter decoupling?", "answer": ["QuPeD: Quantized Personalization via Distillation with Applications to Federated Learning", "Efficient Split-Mix Federated Learning for On-Demand and In-Situ Customization", "HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients", "Exploiting Shared Representations for Personalized Federated Learning", "Achieving Personalized Federated Learning with Sparse Local Models"], "answer_arxiv_id": ["2107.13892", "2203.09747", "2010.01264", "2102.07078", "2201.11380"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_test_756"} +{"question": "Which papers explore the application of parameter-efficient fine-tuning methods, such as Prefix-Tuning and LoRA?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2101.00190", "2106.09685"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_757"} +{"question": "Could you tell me about the studies that have proposed goal-driven clustering for personalising the grouping of text corpora?", "answer": ["Goal-Driven Explainable Clustering via Language Descriptions"], "answer_arxiv_id": ["2305.13749"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_test_758"} +{"question": "Are there any studies that used a few-shot manner to train the CDR model?", "answer": ["Few-Shot Conversational Dense Retrieval", "History-Aware Conversational Dense Retrieval"], "answer_arxiv_id": ["2105.04166", "2401.16659"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_test_759"} +{"question": "Can you list any studies that utilize differentiable logical rule learning", "answer": ["Embedding Entities and Relations for Learning and Inference in Knowledge Bases", "Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "DRUM: End-To-End Differentiable Rule Mining On Knowledge Graphs", "DeepPath: A Reinforcement Learning Method for Knowledge Graph Reasoning", "Variational Knowledge Graph Reasoning", "Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning", "Multi-Hop Knowledge Graph Reasoning with Reward Shaping", "M-Walk: Learning to Walk over Graphs using Monte Carlo Tree Search"], "answer_arxiv_id": ["1412.6575", "1702.08367", "1911.00055", "1707.06690", "1803.06581", "1711.05851", "1808.10568", "1802.04394"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_760"} +{"question": "What studies provide insight into provably efficient exploration techniques in RL?", "answer": ["Model-based Reinforcement Learning and the Eluder Dimension", "Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning", "Learning Near Optimal Policies with Low Inherent Bellman Error", "Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1406.1853", "2004.10019", "2106.04895", "2003.00153", "1907.05388"], "source_meta": {"published_time": "20220405"}, "qid": "AutoScholarQuery_test_761"} +{"question": "What work proposed a term-graph rewriting system for marginalizing a log joint density with conjugacy?", "answer": ["Autoconj: Recognizing and Exploiting Conjugacy Without a Domain-Specific Language"], "answer_arxiv_id": ["1811.11926"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_762"} +{"question": "What studies discuss about adversarial attacks as potential safety risks in machine learning models?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1706.06083", "1901.08573"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_test_763"} +{"question": "Which works used a simple peak detection algorithm to extract atomic coordinates from the generated voxel grids similar to the method used in the current study?", "answer": ["Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models"], "answer_arxiv_id": ["2010.08687"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_764"} +{"question": "Which work focuses on 'Cause on Tape' (CoT) explanations where the explanation precedes the answer?", "answer": ["Language Models Don't Always Say What They Think: Unfaithful\n Explanations in Chain-of-Thought Prompting"], "answer_arxiv_id": ["2305.04388"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_765"} +{"question": "Which works have introduced domain adversarial methods to learn domain-invariant embeddings across the source domain and the target domain in unsupervised graph domain adaption problem?", "answer": ["DANE: Domain Adaptive Network Embedding"], "answer_arxiv_id": ["1906.00684"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_test_766"} +{"question": "Which work first incorporated Differentiable Inductive Logic Programming to RL domain?", "answer": ["Neural Logic Reinforcement Learning"], "answer_arxiv_id": ["1904.10729"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_767"} +{"question": "What papers propose 3D pretraining methods utilizing local augmentations?", "answer": ["Self-Supervised Pretraining of 3D Features on any Point-Cloud"], "answer_arxiv_id": ["2101.02691"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_test_768"} +{"question": "Can you name the studies that looked at CVaR optimization using PG?", "answer": ["Optimizing the CVaR via Sampling", "EPOpt: Learning Robust Neural Network Policies Using Model Ensembles", "Learning Robust Options by Conditional Value at Risk Optimization"], "answer_arxiv_id": ["1404.3862", "1610.01283", "1905.09191"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_test_769"} +{"question": "Are there any studies about the policy update of AMPO to mirror descent algorithm based on value iteration and Bellman operators?", "answer": ["A Theory of Regularized Markov Decision Processes"], "answer_arxiv_id": ["1901.11275"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_770"} +{"question": "What papers have examined the effect of smaller design decisions like the loss function or policy regularization?", "answer": ["Revisiting Design Choices in Proximal Policy Optimization"], "answer_arxiv_id": ["2009.10897"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_771"} +{"question": "Any works that developed representations of statistical and causal dependencies between latent factors and auxiliary variables?", "answer": ["Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICA", "Weakly-Supervised Disentanglement Without Compromises", "Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style", "The Incomplete Rosetta Stone Problem: Identifiability Results for Multi-View Nonlinear ICA", "Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learning", "Variational Autoencoders and Nonlinear ICA: A Unifying Framework", "ICE-BeeM: Identifiable Conditional Energy-Based Deep Models Based on Nonlinear ICA", "Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding", "Contrastive Learning Inverts the Data Generating Process"], "answer_arxiv_id": ["1605.06336", "2002.02886", "2106.04619v4", "1905.06642", "1805.08651", "1907.04809", "2002.11537", "2007.10930", "2102.08850v4"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_772"} +{"question": "What studies have demonstrated the effectiveness of contrastive methods in learning useful representations for downstream tasks?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Representation Learning with Contrastive Predictive Coding", "Learning deep representations by mutual information estimation and maximization", "Learning Representations by Maximizing Mutual Information Across Views", "Contrastive Multiview Coding", "On Mutual Information Maximization for Representation Learning", "What Makes for Good Views for Contrastive Learning?", "Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere", "Representation Learning with Contrastive Predictive Coding", "Contrastive Learning Inverts the Data Generating Process", "Representation Learning with Contrastive Predictive Coding", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Estimating divergence functionals and the likelihood ratio by convex risk minimization"], "answer_arxiv_id": ["1807.03748", "2002.05709", "1911.05722", "2106.04156", "1807.03748", "1808.06670", "1906.00910", "1906.05849", "1907.13625", "2005.10243", "2005.10242", "1807.03748", "2102.08850v4", "1807.03748", "2106.04156", "0809.0853"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_773"} +{"question": "What research introduces competition-level problems integrating mathematical logic and background knowledge?", "answer": ["Measuring Mathematical Problem Solving With the MATH Dataset", "Mathematical Capabilities of ChatGPT", "Have LLMs Advanced Enough? A Challenging Problem Solving Benchmark For\n Large Language Models"], "answer_arxiv_id": ["2103.03874", "2301.13867", "2305.15074"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_test_774"} +{"question": "Could you provide examples of studies that use dense voxel grids in conjunction with shallow multilayer perceptrons to expedite 3D reconstruction?", "answer": ["Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "Improved Direct Voxel Grid Optimization for Radiance Fields\n Reconstruction", "HexPlane: A Fast Representation for Dynamic Scenes"], "answer_arxiv_id": ["2111.11215", "2206.05085", "2301.09632"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_775"} +{"question": "Could you provide me some studies about meta-learning in DG strategies?", "answer": ["Episodic Training for Domain Generalization", "Learning to Learn with Variational Information Bottleneck for Domain Generalization", "Learning to Generalize: Meta-Learning for Domain Generalization", "Learning to Generalize Unseen Domains via Memory-based Multi-Source Meta-Learning for Person Re-Identification"], "answer_arxiv_id": ["1902.00113", "2007.07645", "1710.03463", "2012.00417"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_test_776"} +{"question": "What research study discusses protein docking methods such as EquiDock without prior knowledge of the epitope and the paratope?", "answer": ["Independent SE(3)-Equivariant Models for End-to-End Rigid Protein Docking"], "answer_arxiv_id": ["2111.07786"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_777"} +{"question": "Could you provide me some papers that investigated adversarial perturbations in MRI and CT image reconstruction?", "answer": ["Solving Inverse Problems With Deep Neural Networks – Robustness Included?"], "answer_arxiv_id": ["2011.04268"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_test_778"} +{"question": "Can you provide some studies dealing with the strong quadratic cost in Optimal Transport?", "answer": ["Optimal transport mapping via input convex neural networks", "2-Wasserstein Approximation via Restricted Convex Potentials with Application to Improved Training for GANs", "Wasserstein-2 Generative Networks"], "answer_arxiv_id": ["1908.10962", "1902.07197", "1909.13082v4"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_test_779"} +{"question": "Which paper proposed a fully differentiable equivariant model that can predict coordinates of docked poses?", "answer": ["EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction"], "answer_arxiv_id": ["2202.05146"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_test_780"} +{"question": "Which studies have explored improved choices for the matrices in the continual counting context?", "answer": ["Almost Tight Error Bounds on Differentially Private Continual Counting"], "answer_arxiv_id": ["2211.05006"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_test_781"} +{"question": "Which works are about using dynamic computation graph by skipping different blocks based on samples?", "answer": ["SkipNet: Learning Dynamic Routing in Convolutional Networks", "BlockDrop: Dynamic Inference Paths in Residual Networks"], "answer_arxiv_id": ["1711.09485", "1711.08393"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_test_782"} +{"question": "Could you provide me some works that discuss improving model's robustness through data augmentation?", "answer": ["The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization", "AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "A simple way to make neural networks robust against diverse image corruptions", "AugMax: Adversarial Composition of Random Augmentations for Robust Training"], "answer_arxiv_id": ["2006.16241", "1912.02781", "2001.06057", "2110.13771"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_test_783"} +{"question": "What research has been done on the interaction between the Bregman projected policy class and the expected Lipschitz and smooth policies?", "answer": ["A general sample complexity analysis of vanilla policy gradient", "An Improved Analysis of (Variance-Reduced) Policy Gradient and Natural Policy Gradient Methods"], "answer_arxiv_id": ["2107.11433v5", "2211.07937"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_784"} +{"question": "Can you give me examples of LLM papers that use the MCP approach for evaluation?", "answer": ["Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2112.11446", "2203.15556"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_test_785"} +{"question": "In what papers were methods described that locate and edit the parameters and neurons in the LLMs in light of specific knowledge?", "answer": ["Locating and Editing Factual Associations in GPT", "Knowledge Neurons in Pretrained Transformers", "Mass-Editing Memory in a Transformer", "Editing a classifier by rewriting its prediction rules", "Transformer Feed-Forward Layers Build Predictions by Promoting Concepts\n in the Vocabulary Space"], "answer_arxiv_id": ["2202.05262", "2104.08696", "2210.07229", "2112.01008", "2203.14680"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_786"} +{"question": "What study proposed the method PatchFlow to solve the technical challenge of per-view detection of feature points in SfM?", "answer": ["Multi-View Optimization of Local Feature Geometry"], "answer_arxiv_id": ["2003.08348"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_787"} +{"question": "Which papers discuss the use of Visual Instruction Fine-tuning in Large Multimodal Models?", "answer": ["Visual Instruction Tuning", "SVIT: Scaling up Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485", "2307.04087"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_test_788"} +{"question": "What papers have employed use of persona prompts in LLMs?", "answer": ["Out of One, Many: Using Language Models to Simulate Human Samples", "Simulating Social Media Using Large Language Models to Evaluate\n Alternative News Feed Algorithms"], "answer_arxiv_id": ["2209.06899", "2310.05984"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_789"} +{"question": "Which work introduced the notion of (ρ,μ)-competitiveness?", "answer": ["Learning-Augmented Dynamic Power Management with Multiple States via New Ski Rental Bounds"], "answer_arxiv_id": ["2110.13116"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_test_790"} +{"question": "What research investigates higher-order grammatical feature representation across languages using probing classifiers trained on mBERT embeddings?", "answer": ["Deep Subjecthood: Higher-Order Grammatical Features in Multilingual BERT"], "answer_arxiv_id": ["2101.11043"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_test_791"} +{"question": "Which studies explored techniques in learning process design to address biases and instability in in-context learning?", "answer": ["Noisy Channel Language Model Prompting for Few-Shot Text Classification", "MetaICL: Learning to Learn In Context"], "answer_arxiv_id": ["2108.04106", "2110.15943"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_test_792"} +{"question": "Which works have studied the underlying representations in diffusion models and proposed using them for various downstream tasks?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image Translation", "Label-Efficient Semantic Segmentation with Diffusion Models"], "answer_arxiv_id": ["2211.12572", "2112.03126"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_793"} +{"question": "What studies proposed gradient-guided autoregressive models by using the decoder’s activations as a gradient-friendly latent space?", "answer": ["Plug and Play Language Models: a Simple Approach to Controlled Text Generation", "Fudge: Controlled Text Generation With Future Discriminators"], "answer_arxiv_id": ["1912.02164", "2104.05218"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_794"} +{"question": "Which research papers report on response-based distillation methods for object detection?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_test_795"} +{"question": "Are there any studies based on using pre-trained language model agents for role-play in text-based games?", "answer": ["Learning to Speak and Act in a Fantasy Text Adventure Game", "Pre-trained Language Models as Prior Knowledge for Playing Text-based\n Games", "Exploring Large Language Models for Communication Games: An Empirical\n Study on Werewolf"], "answer_arxiv_id": ["1903.03094", "2107.08408", "2309.04658"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_test_796"} +{"question": "Which works combined models used for ARA with tradition linguistic features?", "answer": ["Linguistic Features for Readability Assessment", "BERT Embeddings for Automatic Readability Assessment"], "answer_arxiv_id": ["2006.00377", "2106.07935"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_797"} +{"question": "Which paper distills pre-trained Stable Diffusion using Score Distillation Sampling (SDS) to extract a Neural Radiance Field (NeRF) from a given text prompt?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_798"} +{"question": "What papers focus on extending the signal-prediction framework to forecast aggregation settings?", "answer": ["The Wisdom of the Crowd and Higher-Order Beliefs"], "answer_arxiv_id": ["2102.02666"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_test_799"} +{"question": "Which works were pertinent in the development of the Large Multimodal Models?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "PaLM: Scaling Language Modeling with Pathways", "UL2: Unifying Language Learning Paradigms", "Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "CyCLIP: Cyclic Contrastive Language-Image Pretraining"], "answer_arxiv_id": ["1910.10683", "2204.02311", "2205.05131", "2103.00020", "2201.12086", "2205.14459"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_test_800"} +{"question": "Are there any articles proposing methods to evaluate the complexity of the hierarchical structure of a graph?", "answer": ["Structural Entropy Guided Graph Hierarchical Pooling", "A Simple yet Effective Method for Graph Classification"], "answer_arxiv_id": ["2206.13510", "2206.02404"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_test_801"} +{"question": "Which papers discuss the role of Transformers in the field of NLP?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_test_802"} +{"question": "Which papers describe the basic annotation approach in creating NLI data?", "answer": ["A large annotated corpus for learning natural language inference", "A Broad-Coverage Challenge Corpus for Sentence Understanding through\n Inference"], "answer_arxiv_id": ["1508.05326", "1704.05426"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_803"} +{"question": "Could you provide references about the PixelHelp dataset which includes task goals and step-by-step instructions for Android?", "answer": ["Mapping Natural Language Instructions to Mobile UI Action Sequences"], "answer_arxiv_id": ["2005.03776"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_804"} +{"question": "What studies utilize first-order logic to inspect and improve model's logical consistency in synthetic compositional reasoning tasks?", "answer": ["Logic-LM: Empowering Large Language Models with Symbolic Solvers for\n Faithful Logical Reasoning", "LINC: A Neurosymbolic Approach for Logical Reasoning by Combining\n Language Models with First-Order Logic Provers", "FaiRR: Faithful and Robust Deductive Reasoning over Natural Language", "Learning Symbolic Rules for Reasoning in Quasi-Natural Language"], "answer_arxiv_id": ["2305.12295", "2310.15164", "2203.10261", "2111.12038"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_805"} +{"question": "Which studies propose interpolation-based mixup methods for graph augmentations?", "answer": ["G-Mixup: Graph Data Augmentation for Graph Classification", "Graph Transplant: Node Saliency-Guided Graph Mixup with Local Structure Preservation"], "answer_arxiv_id": ["2202.07179", "2111.05639"], "source_meta": {"published_time": "20220226"}, "qid": "AutoScholarQuery_test_806"} +{"question": "What research provided theoretical studies on the convergence of the LocalSGDM algorithm?", "answer": ["On the Linear Speedup Analysis of Communication Efficient Momentum SGD for Distributed Non-Convex Optimization"], "answer_arxiv_id": ["1905.03817"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_test_807"} +{"question": "What works present GALOIS framework and its sketch setting?", "answer": ["GALOIS: Boosting Deep Reinforcement Learning via Generalizable Logic Synthesis"], "answer_arxiv_id": ["2205.13728"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_test_808"} +{"question": "Can you provide me some researches that support sensor simulation in the context of autonomous driving perception?", "answer": ["CARLA: An Open Urban Driving Simulator", "Block-NeRF: Scalable Large Scene Neural View Synthesis", "SUMMIT: A Simulator for Urban Driving in Massive Mixed Traffic", "VISTA 2.0: An Open, Data-driven Simulator for Multimodal Sensing and Policy Learning for Autonomous Vehicles"], "answer_arxiv_id": ["1711.03938", "2202.05263", "1911.04074", "2111.12083"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_test_809"} +{"question": "What are some of the recent research papers that achieved success in solving downstream language tasks using finetuning methods?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "ALBERT: A Lite BERT for Self-supervised Learning of Language Representations", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "XLNet: Generalized Autoregressive Pretraining for Language Understanding"], "answer_arxiv_id": ["1810.04805", "1909.11942", "1907.11692", "1906.08237"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_test_810"} +{"question": "Who proposed a GNN framework arising from a mixture of parabolic and hyperbolic PDEs on graphs with convolutional coupling operators?", "answer": ["PDE-GCN: Novel Architectures for Graph Neural Networks Motivated by Partial Differential Equations"], "answer_arxiv_id": ["2108.01938"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_test_811"} +{"question": "Which works have utilized gating in classical recurrent neural network (RNN) architectures such as LSTM and GRU?", "answer": ["Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation"], "answer_arxiv_id": ["1406.1078"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_test_812"} +{"question": "Can you provide papers that explored novel ways of training CNNs for FGVC?", "answer": ["Channel DropBlock: An Improved Regularization Method for Fine-Grained Visual Classification"], "answer_arxiv_id": ["2106.03432"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_813"} +{"question": "What studies proposed deep learning models for protein sequence design using structure-based generative models?", "answer": ["Protein Structure and Sequence Generation with Equivariant Denoising Diffusion Probabilistic Models", "Learning from Protein Structure with Geometric Vector Perceptrons", "Multi-level Protein Representation Learning for Blind Mutational Effect Prediction"], "answer_arxiv_id": ["2205.15019", "2009.01411", "2306.04899"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_test_814"} +{"question": "Which work proposed the model-free algorithm MOFFLE for low-nonnegative-rank MDPs in the field of reward-free reinforcement learning?", "answer": ["Model-free Representation Learning and Exploration in Low-rank MDPs"], "answer_arxiv_id": ["2102.07035"], "source_meta": {"published_time": "20220628"}, "qid": "AutoScholarQuery_test_815"} +{"question": "What papers utilize the Monte Carlo Dropout method in variational Bayesian methods for approximating the intractable integrals arising in Bayesian inference?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inference"], "answer_arxiv_id": ["1506.02142", "1506.02158"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_test_816"} +{"question": "Which studies extended the work of cage-based deformation in NeRF editing?", "answer": ["Interactive Geometry Editing of Neural Radiance Fields"], "answer_arxiv_id": ["2303.11537"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_817"} +{"question": "What studies discuss the training of Pre-trained Language Models(PLMs) for predicting masked words?", "answer": ["Mask More and Mask Later: Efficient Pre-training of Masked Language Models by Disentangling the [MASK] Token", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension", "MASS: Masked Sequence to Sequence Pre-training for Language Generation", "Unsupervised Cross-lingual Representation Learning at Scale", "Cross-lingual Language Model Pretraining"], "answer_arxiv_id": ["2211.04898", "1810.04805", "1907.11692", "1910.13461", "1905.02450", "1911.02116", "1901.07291"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_test_818"} +{"question": "What research focused on constructing a purpose-driven affordance dataset?", "answer": ["One-Shot Affordance Detection"], "answer_arxiv_id": ["2106.14747"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_test_819"} +{"question": "What works have been done in the field of image editing in computer vision?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "Expressive Text-to-Image Generation with Rich Text"], "answer_arxiv_id": ["2211.12572", "2304.06720"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_test_820"} +{"question": "Could you provide me some works that focus on source domain data estimation in the context of SFUDA?", "answer": ["Source-Free Domain Adaptation for Semantic Segmentation"], "answer_arxiv_id": ["2103.16372"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_test_821"} +{"question": "Which work combines the semantic and instance segmentation tasks effectively?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2107.06278", "2112.01527"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_822"} +{"question": "Which studies presented automatic machine learning (AutoML) approaches?", "answer": ["Auto-Pytorch: Multi-Fidelity MetaLearning for Efficient and Robust AutoDL"], "answer_arxiv_id": ["2006.13799"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_test_823"} +{"question": "Which papers discussed federated learning in the context of autonomous driving?", "answer": ["FedDrive: Generalizing Federated Learning to Semantic Segmentation in\n Autonomous Driving", "Deep Federated Learning for Autonomous Driving"], "answer_arxiv_id": ["2202.13670", "2110.05754"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_test_824"} +{"question": "Could you provide me studies that propose approaches to unite segmentation datasets from multiple domains?", "answer": ["MSeg: A Composite Dataset for Multi-domain Semantic Segmentation", "Multi-dataset Pretraining: A Unified Model for Semantic Segmentation"], "answer_arxiv_id": ["2112.13762", "2106.04121"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_825"} +{"question": "Which papers have investigated the simplicity bias in Deep Neural Networks (DNNs)?", "answer": ["SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data", "The Implicit Bias of Gradient Descent on Separable Data", "Implicit Bias of Gradient Descent on Linear Convolutional Networks", "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness", "The Origins and Prevalence of Texture Bias in Convolutional Neural Networks"], "answer_arxiv_id": ["1710.10174", "1710.10345", "1806.00468", "1811.12231", "1911.09071"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_test_826"} +{"question": "What works propose to utilize the MVS approach using cost volume for better occlusion handling in generalizable view synthesis?", "answer": ["MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "GeoNeRF: Generalizing NeRF with Geometry Priors", "Neural Rays for Occlusion-aware Image-based Rendering"], "answer_arxiv_id": ["2103.15595", "2111.13539", "2107.13421"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_test_827"} +{"question": "Which papers discuss earlier datasets focused on sign language recognition using isolated signs?", "answer": ["MS-ASL: A Large-Scale Data Set and Benchmark for Understanding American\n Sign Language", "Word-level Deep Sign Language Recognition from Video: A New Large-scale\n Dataset and Methods Comparison"], "answer_arxiv_id": ["1812.01053", "1910.11006"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_828"} +{"question": "Any works that propose a view synthesis framework using two images with small overlapped regions?", "answer": ["Learning to Render Novel Views from Wide-Baseline Stereo Pairs"], "answer_arxiv_id": ["2304.08463"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_test_829"} +{"question": "Which papers proposed a multi-stage language model for TTS with phonemes as input and acoustic tokens as output?", "answer": ["Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers"], "answer_arxiv_id": ["2301.02111"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_830"} +{"question": "Any studies about enhancing PLMs through pretraining on domain-relevant documents?", "answer": ["BARTScore: Evaluating Generated Text as Text Generation"], "answer_arxiv_id": ["2106.11520"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_831"} +{"question": "What research works give a gradient-based approach to interpretability methods?", "answer": ["Layer-wise Relevance Propagation for Neural Networks with Local\n Renormalization Layers", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1604.00825", "1703.01365"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_832"} +{"question": "Which works have utilized the relationship between multiple images and between network layers in FGVC?", "answer": ["Cross-X Learning for Fine-Grained Visual Categorization"], "answer_arxiv_id": ["1909.04412"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_833"} +{"question": "Which research proposed the Conditional Mutual Information (CMI) approach for learning disentangled representations?", "answer": ["Disentanglement and Generalization Under Correlation Shifts"], "answer_arxiv_id": ["2112.14754"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_834"} +{"question": "Could you provide me some studies about delayed sampling which uses automatic marginalization to improve inference?", "answer": ["Delayed Sampling and Automatic Rao-Blackwellization of Probabilistic Programs", "Automated learning with a probabilistic programming language: Birch", "Pyro: Deep Universal Probabilistic Programming", "Functional Tensors for Probabilistic Programming", "Tensor Variable Elimination for Plated Factor Graphs", "Reactive Probabilistic Programming", "Semi-Symbolic Inference for Efficient Streaming Probabilistic Programming"], "answer_arxiv_id": ["1708.07787v2", "1810.01539", "1810.09538", "1910.10775v2", "1902.03210", "1908.07563v2", "2209.07490v2"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_835"} +{"question": "Can you name some studies that use Recurrent Neural Network or Graph Neural Network for improving the geometry shapes of building extraction results?", "answer": ["Topological Map Extraction from Overhead Images", "PolyWorld: Polygonal Building Extraction with Graph Neural Networks in\n Satellite Images"], "answer_arxiv_id": ["1812.01497", "2111.15491"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_test_836"} +{"question": "What studies examined the performance issues in machine translation models?", "answer": ["Analyzing Uncertainty in Neural Machine Translation"], "answer_arxiv_id": ["1803.00047"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_test_837"} +{"question": "Which studies are about leveraging demonstrations into the policy-update steps of Reinforcement Learning?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations", "Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewards", "Overcoming Exploration in Reinforcement Learning with Demonstrations", "Reinforced Imitation: Sample Efficient Deep Reinforcement Learning for Map-less Navigation by Leveraging Prior Demonstrations", "Integrating Behavior Cloning and Reinforcement Learning for Improved Performance in Dense and Sparse Reward Environments"], "answer_arxiv_id": ["1709.10087", "1707.08817", "1709.10089", "1805.07095", "1910.04281"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_test_838"} +{"question": "Can you cite the research papers that investigated the use of six IMUs for full-body motion estimation?", "answer": ["Sparse Inertial Poser: Automatic 3D Human Pose Estimation from Sparse\n IMUs", "Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time", "TransPose: Real-time 3D Human Translation and Pose Estimation with Six\n Inertial Sensors", "Physical Inertial Poser (PIP): Physics-aware Real-time Human Motion\n Tracking from Sparse Inertial Sensors"], "answer_arxiv_id": ["1703.08014", "1810.04703", "2105.04605", "2203.08528"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_839"} +{"question": "Which work marked a significant development in Vision-Language Pre-training (VLP) and has trained encoders on a large amount of data?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_840"} +{"question": "What papers are about contextualized models for encoding word meaning in the LSC task?", "answer": ["Explaining and Improving BERT Performance on Lexical Semantic Change\n Detection"], "answer_arxiv_id": ["2103.07259"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_test_841"} +{"question": "Could you mention any works that proposed the concept of learning EBM by using a ConvNet as the energy function?", "answer": ["A Theory of Generative ConvNet"], "answer_arxiv_id": ["1602.03264v3"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_842"} +{"question": "Which investigation proposes a method that mitigates the issue of model bias and generalization in zero-/few-shot anomaly detection?", "answer": ["Catching Both Gray and Black Swans: Open-set Supervised Anomaly\n Detection", "Explicit Boundary Guided Semi-Push-Pull Contrastive Learning for\n Supervised Anomaly Detection"], "answer_arxiv_id": ["2203.14506", "2207.01463"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_test_843"} +{"question": "Who introduced the approach of dataset condensation that leverages gradient-based hyper-parameter optimization?", "answer": ["Dataset Distillation"], "answer_arxiv_id": ["1811.10959"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_test_844"} +{"question": "What works have used frame averaging to produce equivariant output from non-equivariant architecture backbones?", "answer": ["Frame Averaging for Invariant and Equivariant Network Design", "Frame Averaging for Equivariant Shape Space Learning", "FAENet: Frame Averaging Equivariant GNN for Materials Modeling"], "answer_arxiv_id": ["2110.03336", "2112.01741", "2305.05577"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_test_845"} +{"question": "What papers focus on HMT techniques using depth sensors?", "answer": ["Real-time RGBD-based Extended Body Pose Estimation", "RobustFusion: Robust Volumetric Performance Reconstruction under\n Human-object Interactions from Monocular RGBD Stream", "DoubleFusion: Real-time Capture of Human Performances with Inner Body\n Shapes from a Single Depth Sensor"], "answer_arxiv_id": ["2103.03663", "2104.14837", "1804.06023"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_846"} +{"question": "What study modifies questions and answers by counterfactual presupposition in VQAv2 for a new challenging scenario for complementary MLLMs?", "answer": ["Making the V in VQA Matter: Elevating the Role of Image Understanding in\n Visual Question Answering"], "answer_arxiv_id": ["1612.00837"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_test_847"} +{"question": "Can you mention any work about Out-of-distribution (OOD) generalization on graphs?", "answer": ["Out-Of-Distribution Generalization on Graphs: A Survey", "Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training Data"], "answer_arxiv_id": ["2202.07987", "2108.01099"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_test_848"} +{"question": "Can you tell me about the studies that propose tuning methods for CLIP?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_test_849"} +{"question": "Which research papers focus on addressing label-shift scenarios with open-set domain adaptation (OSDA), partial-set domain adaptation (PDA), and open-partial-set domain adaptation (OPDA)?", "answer": ["Open Set Domain Adaptation by Backpropagation", "Partial Transfer Learning with Selective Adversarial Networks", "Learning to Transfer Examples for Partial Domain Adaptation", "OVANet: One-vs-All Network for Universal Domain Adaptation"], "answer_arxiv_id": ["1804.10427", "1707.07901", "1903.12230", "2104.03344"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_850"} +{"question": "What research studies use hard pseudolabels from teachers to train student models in the outcontext of low-resource semi-supervised sequence generation?", "answer": ["Sequence-Level Knowledge Distillation", "Is GPT-3 a Good Data Annotator?", "GPT3Mix: Leveraging Large-scale Language Models for Text Augmentation", "Want To Reduce Labeling Cost? GPT-3 Can Help", "ChatGPT Outperforms Crowd-Workers for Text-Annotation Tasks", "Large Language Models Are Reasoning Teachers"], "answer_arxiv_id": ["1606.07947", "2212.10450", "2104.08826", "2108.13487", "2303.15056", "2212.10071"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_test_851"} +{"question": "What papers dealt with the classical transduction setting, a special case of semi-supervised learning?", "answer": ["Learning by Transduction"], "answer_arxiv_id": ["1301.7375v1"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_test_852"} +{"question": "Which papers introduce an asymmetric mechanism within the Multi-Agent Debate framework?", "answer": ["Encouraging Divergent Thinking in Large Language Models through\n Multi-Agent Debate", "ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate"], "answer_arxiv_id": ["2305.19118", "2308.07201"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_test_853"} +{"question": "Which works applied data augmentation strategy or generative adversarial networks (GANs) in Domain Generalization (DG) to handle the domain shift?", "answer": ["Towards Recognizing Unseen Categories in Unseen Domains", "Learning to Generate Novel Domains for Domain Generalization"], "answer_arxiv_id": ["2007.12256", "2007.03304"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_test_854"} +{"question": "Are there any references that proposed the first general gradient inversion method?", "answer": ["Deep Leakage from Gradients"], "answer_arxiv_id": ["1906.08935"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_855"} +{"question": "Can you give examples of studies that have provided physics questions in multiple-choice format and require multistep reasoning?", "answer": ["Measuring Massive Multitask Language Understanding", "C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for\n Foundation Models", "Have LLMs Advanced Enough? A Challenging Problem Solving Benchmark For\n Large Language Models"], "answer_arxiv_id": ["2009.03300", "2305.08322", "2305.15074"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_test_856"} +{"question": "Are there any studies on vision-language models that train to generate text autoregressively?", "answer": ["Long-term Recurrent Convolutional Networks for Visual Recognition and Description", "Show and Tell: A Neural Image Caption Generator", "Unifying Vision-and-Language Tasks via Text Generation", "Answer-Me: Multi-Task Learning for Generalization to Many Question-Answering Tasks"], "answer_arxiv_id": ["1411.4389", "1411.4555", "2102.02779", "2205.00949"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_test_857"} +{"question": "Which work leveraged orthogonal transformations to avoid the direct computation of Jacobian determinants?", "answer": ["Improving Variational Auto-Encoders using Householder Flow"], "answer_arxiv_id": ["1611.09630"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_858"} +{"question": "What studies contribute to formality in cross-style learning?", "answer": ["Dear Sir or Madam, May I introduce the GYAFC Dataset: Corpus, Benchmarks\n and Metrics for Formality Style Transfer"], "answer_arxiv_id": ["1803.06535"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_859"} +{"question": "Can you provide some references that introduce code filtering strategies where functional correctness prediction is achieved without code execution?", "answer": ["Fault-Aware Neural Code Rankers", "LEVER: Learning to Verify Language-to-Code Generation with Execution", "Coder Reviewer Reranking for Code Generation"], "answer_arxiv_id": ["2206.03865", "2302.08468", "2211.16490"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_test_860"} +{"question": "Could you provide examples of research employ additional convolutional modules to learn a hierarchical feature space?", "answer": ["B-CNN: Branch Convolutional Neural Network for Hierarchical\n Classification", "Visual Tree Convolutional Neural Network in Image Classification"], "answer_arxiv_id": ["1709.09890", "1906.01536"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_test_861"} +{"question": "Could you provide me some studies that focus on modelling the user's state alongside the strategies in ESC systems?", "answer": ["Improving Multi-turn Emotional Support Dialogue Generation with\n Lookahead Strategy Planning", "Knowledge-enhanced Memory Model for Emotional Support Conversation"], "answer_arxiv_id": ["2210.04242", "2310.07700"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_test_862"} +{"question": "Are there any studies on decoupling the modeling of the environment using a semantic map from the end-to-end network?", "answer": ["Learning To Explore Using Active Neural SLAM"], "answer_arxiv_id": ["2004.05155"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_test_863"} +{"question": "Could you provide the study that reevaluated the methods of [bib.bibx36] and reported small robustness gains compared to adversarial training?", "answer": ["MagNet and “Efficient Defenses Against Adversarial Attacks” are Not Robust to Adversarial Examples"], "answer_arxiv_id": ["1711.08478"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_test_864"} +{"question": "Which papers discuss multi-agent autonomous driving simulators which use real driving data to initialize scenarios and logged behavior?", "answer": ["Nocturne: a scalable driving benchmark for bringing multi-agent learning one step closer to the real world", "MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learning", "nuPlan: A closed-loop ML-based planning benchmark for autonomous vehicles"], "answer_arxiv_id": ["2206.09889", "2109.12674", "2106.11810"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_test_865"} +{"question": "What studies used cross-entropy method for both optimization and sampling in standard RL?", "answer": ["Efficient Risk-Averse Reinforcement Learning"], "answer_arxiv_id": ["2205.05138"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_test_866"} +{"question": "What work can recover Lipschitz contextual bandits, despite having suboptimal dynamic regret bounds?", "answer": ["A Kernel-Based Approach to Non-Stationary Reinforcement Learning in Metric Spaces"], "answer_arxiv_id": ["2007.05078v2"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_test_867"} +{"question": "What works are about text/speech balloon detection in manga?", "answer": ["Object Detection for Comics using Manga109 Annotations", "COO: Comic Onomatopoeia Dataset for Recognizing Arbitrary or Truncated\n Texts"], "answer_arxiv_id": ["1803.08670", "2207.04675"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_test_868"} +{"question": "What works have explored prompting and style conversion for generative data augmentation in low-resource NLP?", "answer": ["ZeroGen: Efficient Zero-shot Learning via Dataset Generation", "PromptMix: A Class Boundary Augmentation Method for Large Language Model\n Distillation", "Style Transfer as Data Augmentation: A Case Study on Named Entity\n Recognition"], "answer_arxiv_id": ["2202.07922", "2310.14192", "2210.07916"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_test_869"} +{"question": "Can you cite studies showcasing the effectiveness of LLM-powered data augmentation in cross-lingual commonsense reasoning?", "answer": ["LLM-powered Data Augmentation for Enhanced Cross-lingual Performance"], "answer_arxiv_id": ["2305.14288"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_870"} +{"question": "Could you provide me the work that conducted analysis on the embedding layer of mT5 and XLM-R?", "answer": ["Hyperpolyglot LLMs: Cross-Lingual Interpretability in Token Embeddings"], "answer_arxiv_id": ["2311.18034"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_test_871"} +{"question": "Which works use Graph Neural Networks and Recurrent Neural Networks to update encodings in temporal graph learning?", "answer": ["Structured Sequence Modeling with Graph Convolutional Recurrent Networks", "T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction", "ROLAND: Graph Learning Framework for Dynamic Graphs", "CS-TGN: Community Search via Temporal Graph Neural Networks", "Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks", "Anomaly Detection in Multiplex Dynamic Networks: from Blockchain Security to Brain Disease Prediction"], "answer_arxiv_id": ["1612.07659", "1811.05320", "2208.07239", "2303.08964", "1908.01207", "2211.08378"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_test_872"} +{"question": "Which papers solved classification and detection problems in LiDAR perception using deep learning?", "answer": ["Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline", "Benchmarking and Analyzing Point Cloud Classification under Corruptions", "PointCLIP: Point Cloud Understanding by CLIP", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detection", "Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving"], "answer_arxiv_id": ["2106.05304", "2202.03377", "2112.02413", "1812.05784", "2102.00463", "1906.06310"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_test_873"} +{"question": "Could you provide the work that developed an off-policy algorithm capable of computing the inner integral analytically?", "answer": ["Expected Policy Gradients for Reinforcement Learning"], "answer_arxiv_id": ["1801.03326v2"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_test_874"} +{"question": "What works propose to make use of loss functions to handle known symmetries in object pose prediction?", "answer": ["PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes", "Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimation"], "answer_arxiv_id": ["1711.00199", "1901.02970"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_875"} +{"question": "Which studies evaluated the generalization of an RL agent by training and testing them in totally different environments?", "answer": ["Quantifying Generalization in Reinforcement Learning", "Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Design"], "answer_arxiv_id": ["1812.02341", "2012.02096"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_test_876"} +{"question": "Which work showed that the ranking of attention scores computed by a GAT layer is unconditioned on the query node?", "answer": ["How Attentive are Graph Attention Networks?"], "answer_arxiv_id": ["2105.14491"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_test_877"} +{"question": "What studies introduced a strategy for normalizing multi-modal attributes to ensure consistency across modalities?", "answer": ["Attribute-Consistent Knowledge Graph Representation Learning for\n Multi-Modal Entity Alignment"], "answer_arxiv_id": ["2304.01563"], "source_meta": {"published_time": "20240723"}, "qid": "AutoScholarQuery_test_878"} +{"question": "Which papers have created datasets to support research in human behavior understanding?", "answer": ["Towards Automatic Learning of Procedures from Web Instructional Videos", "UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild"], "answer_arxiv_id": ["1703.09788", "1212.0402"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_test_879"} +{"question": "What work demonstrated the faster learning and better generalization when NTK-target alignment is high?", "answer": ["What can linearized neural networks actually say about generalization?"], "answer_arxiv_id": ["2106.06770"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_test_880"} +{"question": "What research has been conducted on Memory-efficient training methods for reducing the memory footprint during the training process?", "answer": ["Reversible Vision Transformers", "Reformer: The Efficient Transformer", "Training Deep Nets with Sublinear Memory Cost"], "answer_arxiv_id": ["2302.04869", "2001.04451", "1604.06174"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_test_881"} +{"question": "What research focuses on distilling reasoning processes?", "answer": ["Learning by Distilling Context"], "answer_arxiv_id": ["2209.15189"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_test_882"} +{"question": "What research focuses on manipulating visual features to target the issue of image-text isolation?", "answer": ["InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Cheap and Quick: Efficient Vision-Language Instruction Tuning for Large\n Language Models"], "answer_arxiv_id": ["2305.06500", "2305.15023"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_883"} +{"question": "Which studies focused on bottom-up methods in instance segmentation in 3D perception?", "answer": ["OccuSeg: Occupancy-aware 3D Instance Segmentation", "Hierarchical Aggregation for 3D Instance Segmentation", "3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans", "Language-Grounded Indoor 3D Semantic Segmentation in the Wild", "Instance Segmentation in 3D Scenes using Semantic Superpoint Tree\n Networks"], "answer_arxiv_id": ["2003.06537v3", "2108.02350", "1812.07003", "2204.07761", "2108.07478"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_test_884"} +{"question": "Any work discusses unigram similarity metrics related to the downstream performance?", "answer": ["Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks"], "answer_arxiv_id": ["2004.10964"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_885"} +{"question": "What papers incorporate rationalization methodologies into supervised NLI models to enhance their resilience against adversarial datasets?", "answer": ["Can Rationalization Improve Robustness?", "Supervising Model Attention with Human Explanations for Robust Natural\n Language Inference"], "answer_arxiv_id": ["2204.11790", "2104.08142"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_886"} +{"question": "What studies provide solutions for feature matching in low-textured regions using dense or semi-dense matching methods?", "answer": ["Learning Accurate Dense Correspondences and When to Trust Them", "Neighbourhood Consensus Networks", "Dual-Resolution Correspondence Networks", "LoFTR: Detector-Free Local Feature Matching with Transformers", "Quadtree Attention for Vision Transformers", "ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformer", "MatchFormer: Interleaving Attention in Transformers for Feature Matching"], "answer_arxiv_id": ["2101.01710", "1810.10510", "2006.08844", "2104.00680", "2201.02767", "2208.14201", "2203.09645"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_887"} +{"question": "Who has defined simplicity bias based on the number of linear components to define a decision boundary and studied its effects?", "answer": ["The Pitfalls of Simplicity Bias in Neural Networks"], "answer_arxiv_id": ["2006.07710"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_test_888"} +{"question": "Which works proposed the extension of 3D Gaussian Splatting technique to dynamic scenes?", "answer": ["4D Gaussian Splatting for Real-Time Dynamic Scene Rendering", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis", "Real-time Photorealistic Dynamic Scene Representation and Rendering with\n 4D Gaussian Splatting"], "answer_arxiv_id": ["2310.08528", "2308.09713", "2310.10642"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_889"} +{"question": "Any works discussed about the introduction of additional layers into the model architecture instead of updating a large number of model parameters?", "answer": ["Differentially Private Fine-tuning of Language Models"], "answer_arxiv_id": ["2110.06500"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_890"} +{"question": "Which publications discuss HMT methods utilizing ego-centric views?", "answer": ["UnrealEgo: A New Dataset for Robust Egocentric 3D Human Motion Capture", "Scene-aware Egocentric 3D Human Pose Estimation", "Estimating Egocentric 3D Human Pose in Global Space", "Ego-Body Pose Estimation via Ego-Head Pose Estimation"], "answer_arxiv_id": ["2208.01633", "2212.11684", "2104.13454", "2212.04636"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_891"} +{"question": "Which studies focus on the integration of contrastive signals in dataset condensation?", "answer": ["Dataset Condensation with Contrastive Signals"], "answer_arxiv_id": ["2202.02916"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_test_892"} +{"question": "Which studies focused on using synthetic data to create new datasets or augment existing ones?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "Playing for Data: Ground Truth from Computer Games", "VisDA: The Visual Domain Adaptation Challenge", "CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning", "Structured3D: A Large Photo-realistic Dataset for Structured 3D Modeling", "ThreeDWorld: A Platform for Interactive Multi-Modal Physical Simulation"], "answer_arxiv_id": ["1504.06852", "1608.02192v1", "1710.06924", "1612.06890", "1908.00222", "2007.04954"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_893"} +{"question": "Which works employed contrastive learning for graph representation learning?", "answer": ["Deep Graph Infomax", "InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization", "Graph Contrastive Learning with Augmentations", "Deep Graph Contrastive Representation Learning", "GCC: Graph Contrastive Coding for Graph Neural Network Pre-Training"], "answer_arxiv_id": ["1809.10341", "1908.01000", "2010.13902", "2006.04131", "2006.09963"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_test_894"} +{"question": "What papers leverage pre-trained object detection models to extract image regional features offline for training multi-modal transformers?", "answer": ["Large-Scale Adversarial Training for Vision-and-Language Representation\n Learning"], "answer_arxiv_id": ["2006.06195"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_test_895"} +{"question": "Which study formally introduced reward-free exploration for tabular MDP?", "answer": ["Reward-Free Exploration for Reinforcement Learning"], "answer_arxiv_id": ["2002.02794"], "source_meta": {"published_time": "20220628"}, "qid": "AutoScholarQuery_test_896"} +{"question": "What research proposed the gradient-sliding method for addressing separated structure in optimization problems?", "answer": ["Gradient Sliding for Composite Optimization"], "answer_arxiv_id": ["1406.0919v2"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_test_897"} +{"question": "Any studies that work on alignment of LLMs evaluators to human evaluation standards?", "answer": ["Calibrating LLM-Based Evaluator"], "answer_arxiv_id": ["2309.13308"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_test_898"} +{"question": "Which works offer end-to-end methods for multimodal Language Models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BEiT: BERT Pre-Training of Image Transformers", "Image as a Foreign Language: BEiT Pretraining for All Vision and\n Vision-Language Tasks", "Visual Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "PaLI: A Jointly-Scaled Multilingual Language-Image Model"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2305.06500", "2201.12086", "2106.08254", "2208.10442", "2304.08485", "2304.14178", "2304.10592", "2303.16199", "2305.03726", "2202.03052", "2209.06794"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_test_899"} +{"question": "Could you tell me the works related to representer point selection for estimating training data influence?", "answer": ["Representer Point Selection for Explaining Deep Neural Networks", "Adapting and Evaluating Influence-Estimation Methods for Gradient-Boosted Decision Trees"], "answer_arxiv_id": ["1811.09720", "2205.00359"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_900"} +{"question": "Which works are responsible for the introduction of unsupervised models for perceptual grouping?", "answer": ["On the Binding Problem in Artificial Neural Networks"], "answer_arxiv_id": ["2012.05208"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_901"} +{"question": "Can you name some examples of projects that integrated machine learning, particularly LLMs, into automated theorem proving?", "answer": ["Learning to Reason in Large Theories without Imitation", "Constructions in combinatorics via neural networks", "LeanDojo: Theorem Proving with Retrieval-Augmented Language Models", "Generative Language Modeling for Automated Theorem Proving", "Proof Artifact Co-training for Theorem Proving with Language Models", "NaturalProofs: Mathematical Theorem Proving in Natural Language", "Draft, Sketch, and Prove: Guiding Formal Theorem Provers with Informal\n Proofs"], "answer_arxiv_id": ["1905.10501", "2104.14516", "2306.15626", "2009.03393", "2102.06203", "2104.01112", "2210.12283"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_test_902"} +{"question": "Could you provide me some works centered on multilingual machine translation?", "answer": ["Beyond English-Centric Multilingual Machine Translation"], "answer_arxiv_id": ["2010.11125"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_903"} +{"question": "What studies have used techniques like residual structure, skip connection, and dropout in basic CNN frameworks for image restoration?", "answer": ["Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Plug-and-Play Image Restoration with Deep Denoiser Prior", "Image Super-Resolution Using Very Deep Residual Channel Attention Networks", "Residual Dense Network for Image Restoration", "Reflash Dropout in Image Super-Resolution"], "answer_arxiv_id": ["1511.04587", "2008.13751", "1807.02758", "1812.10477", "2112.12089"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_test_904"} +{"question": "Are there studies which focus on refining generated code through iterative reviews and improvements based on execution results?", "answer": ["Reflexion: Language Agents with Verbal Reinforcement Learning"], "answer_arxiv_id": ["2303.11366"], "source_meta": {"published_time": "20240802"}, "qid": "AutoScholarQuery_test_905"} +{"question": "Which works have been done on representation learning on hypergraphs?", "answer": ["Hypergraph Neural Networks", "A Survey on Hyperlink Prediction"], "answer_arxiv_id": ["1809.09401", "2207.02911v1"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_test_906"} +{"question": "Which studies showed that the fine-tuning of large vision-language foundational models with a few examples from the target dataset can enhance performance?", "answer": ["Conditional Prompt Learning for Vision-Language Models", "Learning to Prompt for Vision-Language Models", "Visual Prompt Tuning", "MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2203.05557", "2109.01134", "2203.12119", "2210.03117"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_test_907"} +{"question": "Could you provide me some examples of works that introduced a parallel iterative routing?", "answer": ["Capsules with Inverted Dot-Product Attention Routing"], "answer_arxiv_id": ["2002.04764"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_test_908"} +{"question": "Which papers extended the ideas from non-stationary multi-armed bandits to various contextual bandit settings?", "answer": ["Learning Contextual Bandits in a Non-stationary Environment", "Efficient Contextual Bandits in Non-stationary Worlds", "A New Algorithm for Non-stationary Contextual Bandits: Efficient, Optimal, and Parameter-free", "Non-stationary Reinforcement Learning without Prior Knowledge: An Optimal Black-box Approach"], "answer_arxiv_id": ["1805.09365", "1708.01799v4", "1902.00980v3", "2102.05406"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_test_909"} +{"question": "What papers discussed techniques on disentanglement based on Variation Autoencoder (VAE)?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_910"} +{"question": "Which papers cover the convergence of PFL on homogeneous data?", "answer": ["Tighter Theory for Local SGD on Identical and Heterogeneous Data"], "answer_arxiv_id": ["1909.04746v4"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_test_911"} +{"question": "What papers made advancements with hybrid pose-based methods in deep learning?", "answer": ["Camera Relocalization by Computing Pairwise Relative Poses Using\n Convolutional Neural Network"], "answer_arxiv_id": ["1707.09733"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_912"} +{"question": "What prior works discussed random walk-based unsupervised representation learning methods?", "answer": ["node2vec: Scalable Feature Learning for Networks"], "answer_arxiv_id": ["1607.00653"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_test_913"} +{"question": "Which research provides learning methods for object-level anomaly detection during meta-training?", "answer": ["Few-Shot One-Class Classification via Meta-Learning"], "answer_arxiv_id": ["2007.04146"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_914"} +{"question": "Which papers proposed the Preference Ranking Optimization as an alternative to PPO?", "answer": ["Preference Ranking Optimization for Human Alignment"], "answer_arxiv_id": ["2306.17492"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_test_915"} +{"question": "Any works about combining NeRFs by inserting objects into pre-existing NeRF scenes?", "answer": ["Control-NeRF: Editable Feature Volumes for Scene Rendering and\n Manipulation"], "answer_arxiv_id": ["2204.10850"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_916"} +{"question": "What studies require an additional boundedness assumption when fρ* falls into a less-smooth interpolation space?", "answer": ["Optimal Convergence for Distributed Learning with Stochastic Gradient Methods and Spectral Algorithms", "Sobolev Norm Learning Rates for Regularized Least-Squares Algorithms", "Optimal Learning Rates for Regularized Least-Squares with a Fourier Capacity Condition", "Optimal Rates for Regularized Conditional Mean Embedding Learning"], "answer_arxiv_id": ["1801.07226", "1702.07254", "2204.07856v4", "2208.01711v3"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_test_917"} +{"question": "Which research papers adopted a 3D-Unet architecture to produce video volumes directly from an input image?", "answer": ["Stochastic Adversarial Video Prediction", "Animate-A-Story: Storytelling with Retrieval-Augmented Video Generation", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "Stochastic Image-to-Video Synthesis using cINNs", "MCVD: Masked Conditional Video Diffusion for Prediction, Generation, and\n Interpolation", "Diffusion Models for Video Prediction and Infilling"], "answer_arxiv_id": ["1804.01523", "2307.06940", "2307.04725", "2105.04551", "2205.09853", "2206.07696"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_test_918"} +{"question": "Who first explored the food bank problem in the context of additive valuations?", "answer": ["Online Fair Division: analysing a Food Bank problem"], "answer_arxiv_id": ["1502.07571"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_919"} +{"question": "Which paper proposed an enhanced Key-Value Memory neural network for Information Retrieval-based methods?", "answer": ["Key-Value Memory Networks for Directly Reading Documents"], "answer_arxiv_id": ["1606.03126"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_920"} +{"question": "What works studied enabling more robust operation for certain tasks through recovering explicit localization information in model representations?", "answer": ["Perceptual Grouping in Contrastive Vision-Language Models"], "answer_arxiv_id": ["2210.09996"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_test_921"} +{"question": "Are there any existing datasets that concentrate on the novel use of word meanings and omit conventional examples?", "answer": ["Metaphorical Polysemy Detection: Conventional Metaphor meets Word Sense\n Disambiguation"], "answer_arxiv_id": ["2212.08395"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_test_922"} +{"question": "Which papers present the use of ordinary and partial differential equations in designing, interpreting, and analyzing graph machine learning architectures?", "answer": ["Discrete and Continuous Deep Residual Learning Over Graphs", "Graph Neural Ordinary Differential Equations", "Continuous Graph Neural Networks"], "answer_arxiv_id": ["1911.09554", "1911.07532", "1912.00967"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_test_923"} +{"question": "Which papers focus on distributed optimization using stochastic methods through client sampling?", "answer": ["Faster federated optimization under second-order similarity"], "answer_arxiv_id": ["2209.02257"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_test_924"} +{"question": "What work explores learning an anomaly detector in the feature space with a category-agnostic model?", "answer": ["Registration based Few-Shot Anomaly Detection"], "answer_arxiv_id": ["2207.07361"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_925"} +{"question": "Which papers propose first-order methods for efficiently solving min-max optimization problems in Weak Minty Variational Inequalities?", "answer": ["The Complexity of Constrained Min-Max Optimization", "Efficient Methods for Structured Nonconvex-Nonconcave Min-Max Optimization", "Escaping limit cycles: Global convergence for constrained nonconvex-nonconcave minimax problems", "Fast Extra Gradient Methods for Smooth Structured Nonconvex-Nonconcave Minimax Problems", "Solving Nonconvex-Nonconcave Min-Max Problems exhibiting Weak Minty Solutions", "Solving stochastic weak Minty variational inequalities without increasing batch size"], "answer_arxiv_id": ["2009.09623", "2011.00364", "2302.09831", "2106.02326", "2201.12247", "2302.09029"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_926"} +{"question": "Could you tell me which papers introduced RL from AI Feedback training approach in the context of RLHF?", "answer": ["Constitutional AI: Harmlessness from AI Feedback"], "answer_arxiv_id": ["2212.08073"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_test_927"} +{"question": "Which works propose methods for character re-identification in manga?", "answer": ["Unsupervised Manga Character Re-identification via Face-body and\n Spatial-temporal Associated Clustering", "Identity-Aware Semi-Supervised Learning for Comic Character\n Re-Identification"], "answer_arxiv_id": ["2204.04621", "2308.09096"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_test_928"} +{"question": "Which papers discuss the RP gradient based on the reparameterization trick?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_929"} +{"question": "Which work showed that simple classifiers can detect images created by a single category of networks?", "answer": ["FaceForensics++: Learning to Detect Manipulated Facial Images"], "answer_arxiv_id": ["1901.08971"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_930"} +{"question": "Which research introduced acoustic tokens into semantic token modeling and proposed a multi-stage generative framework in speech language models?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation"], "answer_arxiv_id": ["2209.03143"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_test_931"} +{"question": "What studies proposed Stochastic Gradient Descent Ascent (SGDA) algorithms to address stochastic minimax optimization problems?", "answer": ["A Single-Loop Smoothed Gradient Descent-Ascent Algorithm for Nonconvex-Concave Min-Max Problems", "On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems", "Single-Timescale Stochastic Nonconvex-Concave Optimization for Smooth Nonlinear TD Learning", "Optimal Epoch Stochastic Gradient Descent Ascent Methods for Min-Max Optimization"], "answer_arxiv_id": ["2010.15768", "1906.00331", "2008.10103", "2002.05309"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_test_932"} +{"question": "Could you provide me the studies of part correspondence?", "answer": ["Learning Local Shape Descriptors from Part Correspondences With\n Multi-view Convolutional Networks"], "answer_arxiv_id": ["1706.04496"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_test_933"} +{"question": "Which studies fall under the category of sparse-point annotation methods in weakly supervised 3D instance segmentation?", "answer": ["Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts", "PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding"], "answer_arxiv_id": ["2012.09165", "2007.10985"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_test_934"} +{"question": "Any studies about anomaly score improvement in reconstruction-based techniques by combining forecasting error and reconstruction probability?", "answer": ["Multivariate Time-series Anomaly Detection via Graph Attention Network"], "answer_arxiv_id": ["2009.02040"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_test_935"} +{"question": "Can you provide works that used feature-based distillation methods in object detection?", "answer": ["FitNets: Hints for Thin Deep Nets", "Distilling Object Detectors with Fine-grained Feature Imitation", "Distilling Object Detectors via Decoupled Features"], "answer_arxiv_id": ["1412.6550", "1906.03609", "2103.14475"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_test_936"} +{"question": "Which papers provide a comprehensive overview of transfer learning?", "answer": ["A Comprehensive Survey on Transfer Learning"], "answer_arxiv_id": ["1911.02685"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_test_937"} +{"question": "Could you tell me some studies that implemented honesty-based fine-tuning and trained LLMs to admit limitations?", "answer": ["Alignment for Honesty"], "answer_arxiv_id": ["2312.07000"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_test_938"} +{"question": "Could you tell me which paper proposed the fixed pre-decision module to bridge the gap between SimulMT and SimulST?", "answer": ["SimulMT to SimulST: Adapting Simultaneous Text Translation to End-to-End Simultaneous Speech Translation"], "answer_arxiv_id": ["2011.02048"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_test_939"} +{"question": "Which papers discuss the use of barriers or immediate switching between the objective and constraint in CMDP framework?", "answer": ["IPO: Interior-point Policy Optimization under Constraints", "CRPO: A New Approach for Safe Reinforcement Learning with Convergence Guarantee"], "answer_arxiv_id": ["1910.09615", "2011.05869"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_940"} +{"question": "Which work modified the algorithm to get the Probabilistic Serial fractional outcome while showing a weak notion of efficiency with a simpler proof?", "answer": ["A Probabilistic Approach to Voting, Allocation, Matching, and Coalition Formation"], "answer_arxiv_id": ["2002.10171"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_test_941"} +{"question": "Which work introduced the use of orthogonality in feature space to encourage inter-class separation and intra-class clustering?", "answer": ["Orthogonal Projection Loss", "On orthogonality and learning recurrent networks with long term dependencies"], "answer_arxiv_id": ["2103.14021", "1702.00071"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_test_942"} +{"question": "What paper employs self and cross-reconstruction modules to learn discriminative and smooth representations and uses DGCNN for learning the per-point feature embeddings?", "answer": ["DPC: Unsupervised Deep Point Correspondence via Cross and Self\n Construction"], "answer_arxiv_id": ["2110.08636"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_test_943"} +{"question": "Could you mention some studies that contributed to remarkable text-to-3D generation results?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Fantasia3D: Disentangling Geometry and Appearance for High-quality Text-to-3D Content Creation", "DreamTime: An Improved Optimization Strategy for Text-to-3D Content Creation"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2303.13873", "2306.12422"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_test_944"} +{"question": "What studies advanced diffusion probabilistic models to generate high-resolution and diverse images?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2112.10741", "2112.10752", "2205.11487"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_945"} +{"question": "Which works have sought to detect important skill neurons in performing a specific task?", "answer": ["Task-Specific Skill Localization in Fine-tuned Language Models", "Finding Skill Neurons in Pre-trained Transformer-based Language Models", "Task-specific Compression for Multi-task Language Models using\n Attribution-based Pruning"], "answer_arxiv_id": ["2302.06600", "2211.07349", "2205.04157"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_test_946"} +{"question": "Which studies have proposed methods to detect skill neurons within already trained models using an inference-based method?", "answer": ["Finding Skill Neurons in Pre-trained Transformer-based Language Models", "Task-specific Compression for Multi-task Language Models using\n Attribution-based Pruning"], "answer_arxiv_id": ["2211.07349", "2205.04157"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_test_947"} +{"question": "Which papers investigated CLIP-based classifiers and their performance in open-granularity classification?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Conditional Prompt Learning for Vision-Language Models", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2103.00020", "2203.05557", "2109.01134"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_test_948"} +{"question": "What studies have produced abundant equivariant neural networks?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "Directional Message Passing for Molecular Graphs", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "E(n) Equivariant Graph Neural Networks"], "answer_arxiv_id": ["1802.08219", "2003.03123", "2006.10503", "2102.09844"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_949"} +{"question": "Which studies used transformers and diffusion models for creating high-fidelity images from text?", "answer": ["VQGAN-CLIP: Open Domain Image Generation and Editing with Natural\n Language Guidance", "CogView2: Faster and Better Text-to-Image Generation via Hierarchical\n Transformers", "Denoising Diffusion Probabilistic Models", "Blended Diffusion for Text-driven Editing of Natural Images", "Text2LIVE: Text-Driven Layered Image and Video Editing", "Prompt-to-Prompt Image Editing with Cross Attention Control", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "More Control for Free! Image Synthesis with Semantic Diffusion Guidance", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2204.08583", "2204.14217", "2006.11239", "2111.14818", "2204.02491", "2208.01626", "2110.02711", "2112.05744", "2112.10741"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_test_950"} +{"question": "What research works have explored how to address the susceptibility of current large language models to certain errors?", "answer": ["Internet-Augmented Dialogue Generation", "LaMDA: Language Models for Dialog Applications", "PAL: Program-aided Language Models", "Toolformer: Language Models Can Teach Themselves to Use Tools"], "answer_arxiv_id": ["2107.07566", "2201.08239", "2211.10435", "2302.04761"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_test_951"} +{"question": "Are there any works that have further strengthened text embedders?", "answer": ["Text Embeddings by Weakly-Supervised Contrastive Pre-training", "C-Pack: Packaged Resources To Advance General Chinese Embedding", "Large Dual Encoders Are Generalizable Retrievers"], "answer_arxiv_id": ["2212.03533", "2309.07597", "2112.07899"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_test_952"} +{"question": "Could you provide some works that discussed the problem of selling information in economics and computer science?", "answer": ["Optimal Mechanisms for Selling Information", "Selling Information Through Consulting", "How to Sell Information Optimally: an Algorithmic Study", "Optimal Pricing of Information", "Is Selling Complete Information (Approximately) Optimal?", "Optimal Advertising for Information Products", "Selling Data to an Agent with Endogenous Information"], "answer_arxiv_id": ["1204.5519", "1907.04397v3", "2011.14570", "2102.13289", "2202.09013", "2002.10045v5", "2103.05788v4"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_test_953"} +{"question": "Which studies approached the combination of Imitation learning and Reinforcement learning by treating RL as a sequence modelling problem and train an autoregressive model using offline data?", "answer": ["Online Decision Transformer", "Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["2202.05607", "2106.02039", "2106.01345"], "source_meta": {"published_time": "20220405"}, "qid": "AutoScholarQuery_test_954"} +{"question": "Could you provide me studies that improved the performance of small-scale models by fine-tuning them using reasoning processes generated by LLMs?", "answer": ["Training Verifiers to Solve Math Word Problems", "Large Language Models Are Reasoning Teachers", "SCOTT: Self-Consistent Chain-of-Thought Distillation", "Teaching Small Language Models to Reason"], "answer_arxiv_id": ["2110.14168", "2212.10071", "2305.01879", "2212.08410"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_955"} +{"question": "Could you provide me some works that show asymmetric quantization methods outperform their symmetric counterparts?", "answer": ["AFPQ: Asymmetric Floating Point Quantization for LLMs"], "answer_arxiv_id": ["2311.01792"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_956"} +{"question": "Could you provide me some works focusing on video visual relation detection using datasets like AG, VidVRD and VidOR?", "answer": ["Hollywood in Homes: Crowdsourcing Data Collection for Activity\n Understanding"], "answer_arxiv_id": ["1604.01753"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_test_957"} +{"question": "Could you mention a study that proposed plain diffusion Transformer architecture to learn the denoising diffusion process on latent patches?", "answer": ["Scalable Diffusion Models with Transformers"], "answer_arxiv_id": ["2212.09748"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_test_958"} +{"question": "What works discusses the combination of the KD and model quantization method to achieve high compression ratios?", "answer": ["TernaryBERT: Distillation-aware Ultra-low Bit BERT", "Understanding and Improving Knowledge Distillation for\n Quantization-Aware Training of Large Transformer Encoders"], "answer_arxiv_id": ["2009.12812", "2211.11014"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_959"} +{"question": "Which research first studied a local minimax risk for instance optimality in differential privacy?", "answer": ["Near Instance-Optimality in Differential Privacy"], "answer_arxiv_id": ["2005.10630"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_test_960"} +{"question": "Are there any works related to the framework of detector-free Structure-from-Motion?", "answer": ["Structure-from-Motion using Dense CNN Features with Keypoint Relocalization", "OnePose++: Keypoint-Free One-Shot Object Pose Estimation without CAD Models"], "answer_arxiv_id": ["1805.03879", "2301.07673"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_961"} +{"question": "Which work discusses the generation of CDRs on a single chain in regards to pipeline-based antibody design?", "answer": ["Iterative Refinement Graph Neural Network for Antibody Sequence-Structure Co-design"], "answer_arxiv_id": ["2110.04624"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_962"} +{"question": "Which research works presented improvements on the top of EDM?", "answer": ["MDM: Molecular Diffusion Model for 3D Molecule Generation", "Diffusion-based Molecule Generationwith Informative Prior Bridges", "MiDi: Mixed Graph and 3D Denoising Diffusion for Molecule Generation", "Geometric Latent Diffusion Models for 3D Molecule Generation"], "answer_arxiv_id": ["2209.05710", "2209.00865", "2302.09048", "2305.01140"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_test_963"} +{"question": "Could you provide me some works that have developed RL, behavioral cloning, and LLM-based models on the web front?", "answer": ["Learning to Navigate the Web", "Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration", "A Data-Driven Approach for Learning to Control Computers"], "answer_arxiv_id": ["1812.09195", "1802.08802", "2202.08137"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_964"} +{"question": "Which papers introduced factorization along spatial and temporal dimensions on the granularity of the encoder?", "answer": ["ViViT: A Video Vision Transformer", "Is Space-Time Attention All You Need for Video Understanding?"], "answer_arxiv_id": ["2103.15691", "2102.05095"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_965"} +{"question": "Can you mention some studies that address the issue of diversity in single-imgae 3D generation, especially in face generation or starting from text for 3D generation?", "answer": ["Generating Diverse 3D Reconstructions from a Single Occluded Face Image", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation"], "answer_arxiv_id": ["2112.00879", "2305.16213"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_966"} +{"question": "Can you provide a study that effectively tackles domain adaptation and active domain adaptation tasks for 3D semantic segmentation?", "answer": ["UniDA3D: Unified Domain Adaptive 3D Semantic Segmentation Pipeline"], "answer_arxiv_id": ["2212.10390"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_test_967"} +{"question": "Which study proposed to use rooted homomorphism counts as node features in a graph neural network (GNN)?", "answer": ["Graph Neural Networks with Local Graph Parameters"], "answer_arxiv_id": ["2106.06707"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_test_968"} +{"question": "Which studies used feature decomposition to enhance feature alignment in domain adaptation and domain generalization?", "answer": ["Decompose to Adapt: Cross-domain Object Detection via Feature\n Disentanglement", "Learning to Balance Specificity and Invariance for In and Out of Domain\n Generalization", "Efficient Domain Generalization via Common-Specific Low-Rank\n Decomposition", "Modality-Agnostic Debiasing for Single Domain Generalization"], "answer_arxiv_id": ["2201.01929", "2008.12839", "2003.12815", "2303.07123"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_test_969"} +{"question": "Could you provide me some studies that focus on the impact of noisy information on retrieval-augmented generation?", "answer": ["Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language\n Models", "Making Retrieval-Augmented Language Models Robust to Irrelevant Context", "Benchmarking Large Language Models in Retrieval-Augmented Generation"], "answer_arxiv_id": ["2311.09210", "2310.01558", "2309.01431"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_test_970"} +{"question": "What studies tackled the challenge of missing data within MMKGs with adversarial methods?", "answer": ["Embedding Multimodal Relational Data for Knowledge Base Completion"], "answer_arxiv_id": ["1809.01341"], "source_meta": {"published_time": "20240723"}, "qid": "AutoScholarQuery_test_971"} +{"question": "What papers utilized the strategy of transforming categorical data into a continuous space and then applying Gaussian diffusion?", "answer": ["Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning", "Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2208.04202", "2203.17003"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_test_972"} +{"question": "Name a research work that involves the use of neural networks as alternative function classes in BO methods?", "answer": ["Scalable Bayesian Optimization Using Deep Neural Networks"], "answer_arxiv_id": ["1502.05700"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_test_973"} +{"question": "Which studies explored the effect of the choice of the in-context samples on the performance of in-context learning?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?"], "answer_arxiv_id": ["2101.06804"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_test_974"} +{"question": "Could you mention any research that was focused on applying CLIP to downstream tasks?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2110.04544", "2111.03930"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_975"} +{"question": "What are the studies where the RL agents are evaluated by seeing if they can quickly adapt from one task to another?", "answer": ["A Simple Neural Attentive Meta-Learner", "CARLA: An Open Urban Driving Simulator", "BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning", "Language Conditioned Imitation Learning over Unstructured Data"], "answer_arxiv_id": ["1707.03141v3", "1711.03938", "1810.08272", "2005.07648"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_test_976"} +{"question": "Which works focused on the evaluation of RL agents by changing the surfaces of the objects in the environment?", "answer": ["Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Natural Environment Benchmarks for Reinforcement Learning", "Investigating Generalisation in Continuous Deep Reinforcement Learning", "Sim-to-Real via Sim-to-Sim: Data-efficient Robotic Grasping via Randomized-to-Canonical Adaptation Networks"], "answer_arxiv_id": ["1703.06907", "1811.06032", "1902.07015", "1812.07252"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_test_977"} +{"question": "Could you cite the works where multilingual LLMs were evaluated on individual tasks such as Translation, Question-Answering, Summarization, and Reasoning?", "answer": ["On the Cross-lingual Transferability of Monolingual Representations", "TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages", "XOR QA: Cross-lingual Open-Retrieval Question Answering", "XL-Sum: Large-Scale Multilingual Abstractive Summarization for 44\n Languages", "CrossSum: Beyond English-Centric Cross-Lingual Summarization for 1,500+\n Language Pairs", "Language Models are Multilingual Chain-of-Thought Reasoners", "XCOPA: A Multilingual Dataset for Causal Commonsense Reasoning"], "answer_arxiv_id": ["1910.11856", "2003.05002v1", "2010.11856", "2106.13822", "2112.08804", "2210.03057", "2005.00333"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_test_978"} +{"question": "Can you provide the references that propose techniques for enhancing hierarchical classification consistency?", "answer": ["Unified Vision and Language Prompt Learning", "MaPLe: Multi-modal Prompt Learning", "LASP: Text-to-Text Optimization for Language-Aware Soft Prompting of\n Vision & Language Models"], "answer_arxiv_id": ["2210.07225", "2210.03117", "2210.01115"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_test_979"} +{"question": "Could you provide some references focusing on achieving ϵ-stationary points under non-convex settings?", "answer": ["Derivative-free optimization methods"], "answer_arxiv_id": ["1904.11585"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_test_980"} +{"question": "What studies have been conducted to predict future actions from a sequence of observed actions?", "answer": ["Uncertainty-Aware Anticipation of Activities", "When will you do what? - Anticipating Temporal Occurrences of Activities", "Anticipating human actions by correlating past with the future with\n Jaccard similarity measures"], "answer_arxiv_id": ["1908.09540", "1804.00892", "2105.12414"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_981"} +{"question": "Are there works that demonstrate attempts to make images unlearnable or uneditable to handle unauthorized data usage issues in diffusion models?", "answer": ["Glaze: Protecting Artists from Style Mimicry by Text-to-Image Models", "Raising the Cost of Malicious AI-Powered Image Editing"], "answer_arxiv_id": ["2302.04222", "2302.06588"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_test_982"} +{"question": "Can you list the research that introduced the Mirror Descent Policy Optimization?", "answer": ["Mirror Descent Policy Optimization"], "answer_arxiv_id": ["2005.09814"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_test_983"} +{"question": "Which research projects looked into misclassification issues of black-box classifiers where human text was mistaken as LLM-generated?", "answer": ["Can AI-Generated Text be Reliably Detected?"], "answer_arxiv_id": ["2303.11156"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_984"} +{"question": "Could you provide examples of the works on building structured and interpretable models such as Hamiltonian Neural Networks (HNN), Lagrangian Neural Networks (LNN), and Sparse Identification of Nonlinear Dynamics (SINDy)?", "answer": ["Hamiltonian Neural Networks", "Lagrangian Neural Networks"], "answer_arxiv_id": ["1906.01563", "2003.04630"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_test_985"} +{"question": "Did any researchers create a synchrony-based model with no reliance on explicit supervision for grouping?", "answer": ["Neuronal Synchrony in Complex-Valued Deep Networks", "Complex-Valued Autoencoders for Object Discovery"], "answer_arxiv_id": ["1312.6115", "2204.02075"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_986"} +{"question": "Who proposed analytic methods for gradient inversion?", "answer": ["R-GAP: Recursive Gradient Attack on Privacy", "Rethinking Privacy Preserving Deep Learning: How to Evaluate and Thwart Privacy Attacks"], "answer_arxiv_id": ["2010.07733", "2006.11601"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_987"} +{"question": "Are there any research related to meta-learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["1703.03400"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_988"} +{"question": "Could you provide me some studies that detected hallucinations by analyzing the relationship between input prompts and the LLM's output responses?", "answer": ["Exploring the Relationship between LLM Hallucinations and Prompt\n Linguistic Nuances: Readability, Formality, and Concreteness", "SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for\n Generative Large Language Models", "LLM Lies: Hallucinations are not Bugs, but Features as Adversarial\n Examples"], "answer_arxiv_id": ["2309.11064", "2303.08896", "2310.01469"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_test_989"} +{"question": "Which papers mentioned the application of transformer-based pre-trained models in code search?", "answer": ["GraphCodeBERT: Pre-training Code Representations with Data Flow", "Soft-Labeled Contrastive Pre-training for Function-level Code\n Representation"], "answer_arxiv_id": ["2009.08366", "2210.09597"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_test_990"} +{"question": "Could you provide references that discuss alternative approaches to the matrix mechanism that reduce the variance by adding bias?", "answer": ["A Simple and Practical Algorithm for Differentially Private Data Release", "Leveraging Public Data for Practical Private Query Release", "Differentially Private Query Release Through Adaptive Projection", "Dual Query: Practical Private Query Release for High Dimensional Data", "AIM: An Adaptive and Iterative Mechanism for Differentially Private Synthetic Data", "Iterative Methods for Private Synthetic Data: Unifying Framework and New Methods", "New Oracle-Efficient Algorithms for Private Synthetic Data Release", "PrivSyn: Differentially Private Data Synthesis"], "answer_arxiv_id": ["1012.4763", "2102.08598v2", "2103.06641", "1402.1526", "2201.12677", "2106.07153", "2007.05453", "2012.15128v1"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_test_991"} +{"question": "What papers are about using synthetic data to improve the domain-generalization of NLI models?", "answer": ["Training Question Answering Models From Synthetic Data", "Generate, Annotate, and Learn: NLP with Synthetic Text", "QAmeleon: Multilingual QA with Only 5 Examples", "Synthetic Data Generation with Large Language Models for Text\n Classification: Potential and Limitations"], "answer_arxiv_id": ["2002.09599", "2106.06168", "2211.08264", "2310.07849"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_992"} +{"question": "Which works describe the advancements in 3D reconstruction and novel view synthesis with NeRF?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo\n Collections", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural\n Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2008.02268", "2103.13415", "2103.15595", "2307.11335"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_993"} +{"question": "What works evaluated the applicability of MAML and its variants to graph neural networks (GNNs)?", "answer": ["Meta-Learning GNN Initializations for Low-Resource Molecular Property Prediction"], "answer_arxiv_id": ["2003.05996"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_test_994"} +{"question": "What papers discuss client-side local distillation to transfer global knowledge to local models in generic FL?", "answer": ["Data-Free Knowledge Distillation for Heterogeneous Federated Learning"], "answer_arxiv_id": ["2105.10056"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_test_995"} +{"question": "Which papers describe the analysis of the semantic property of intermediate latent space by its local geometry?", "answer": ["Do Not Escape From the Manifold: Discovering the Local Coordinates on the Latent Space of GANs"], "answer_arxiv_id": ["2106.06959"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_test_996"} +{"question": "What are the examples of research done on SLAM methods in the context of object navigation tasks?", "answer": ["Object Goal Navigation using Goal-Oriented Semantic Exploration", "PONI: Potential Functions for ObjectGoal Navigation with Interaction-free Learning"], "answer_arxiv_id": ["2007.00643", "2201.10029"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_test_997"} +{"question": "What works discuss the performance degradation in NLI models when presented with adversarial datasets?", "answer": ["Stress Test Evaluation for Natural Language Inference", "Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural\n Language Inference", "Adversarial NLI: A New Benchmark for Natural Language Understanding", "An Empirical Study on Model-agnostic Debiasing Strategies for Robust\n Natural Language Inference"], "answer_arxiv_id": ["1806.00692", "1902.01007", "1910.14599", "2010.03777"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_998"} +{"question": "What studies have focused on training language like models with privacy guarantees?", "answer": ["Large-Scale Differentially Private BERT"], "answer_arxiv_id": ["2108.01624"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_999"} diff --git a/recipe/paper_search/inference/datasets/AutoScholarQuery/test_lt_5.jsonl b/recipe/paper_search/inference/datasets/AutoScholarQuery/test_lt_5.jsonl new file mode 100644 index 0000000..b23cddc --- /dev/null +++ b/recipe/paper_search/inference/datasets/AutoScholarQuery/test_lt_5.jsonl @@ -0,0 +1,112 @@ +{"question": "What papers are the foundation models for the Natural Language Processing (NLP) field based on?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Language Models are Few-Shot Learners", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["1810.04805", "2005.14165", "1910.10683", "2204.02311", "2302.13971"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_3"} +{"question": "Which studies present issues about the stationary distribution of rewards over contexts?", "answer": ["The K-Nearest Neighbour UCB algorithm for multi-armed bandits with covariates", "Nonparametric Stochastic Contextual Bandits", "Smooth Contextual Bandits: Bridging the Parametric and Non-differentiable Regret Regimes", "Randomized Allocation with Nonparametric Estimation for Contextual Multi-Armed Bandits with Delayed Rewards", "Self-Tuning Bandits over Unknown Covariate-Shifts", "Smoothness-Adaptive Contextual Bandits", "Transfer Learning for Contextual Multi-armed Bandits"], "answer_arxiv_id": ["1803.00316v1", "1801.01750", "1909.02553", "1902.00819", "2007.08584", "1910.09714", "2211.12612"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_test_9"} +{"question": "What works aim to study the policies or features that remain stable across the different training tasks?", "answer": ["Invariant Policy Optimization: Towards Stronger Generalization in Reinforcement Learning", "Instance-based Generalization in Reinforcement Learning", "Domain Adversarial Reinforcement Learning", "Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneck", "Decoupling Representation Learning from Reinforcement Learning", "Deep Reinforcement and InfoMax Learning"], "answer_arxiv_id": ["2006.01096", "2011.01089", "2102.07097", "1910.12911", "2009.08319", "2006.07217"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_test_13"} +{"question": "Could you mention some works that classify unsupervised segmentation into two categories: clustering based on invariance and clustering using pre-trained models?", "answer": ["PiCIE: Unsupervised Semantic Segmentation using Invariance and\n Equivariance in Clustering", "Invariant Information Clustering for Unsupervised Image Classification\n and Segmentation", "Unsupervised Semantic Segmentation with Self-supervised Object-centric\n Representations", "ACSeg: Adaptive Conceptualization for Unsupervised Semantic Segmentation", "Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "NamedMask: Distilling Segmenters from Complementary Foundation Models"], "answer_arxiv_id": ["2103.17070", "1807.06653", "2207.05027", "2210.05944", "2203.08414", "2209.11228"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_test_15"} +{"question": "Could you provide me large multimodal models (LMMs) references?", "answer": ["Visual Instruction Tuning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Qwen Technical Report"], "answer_arxiv_id": ["2304.08485", "2301.12597", "2306.15195", "2304.10592", "2309.16609v1"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_20"} +{"question": "In which studies has it been demonstrated that multi-modal models are vulnerable to adversarial attacks?", "answer": ["Fusion is Not Enough: Single Modal Attacks on Fusion Models for 3D\n Object Detection", "Towards Adversarial Attack on Vision-Language Pre-training Models", "Can audio-visual integration strengthen robustness under multimodal\n attacks?", "Fooling Vision and Language Models Despite Localization and Attention\n Mechanism", "Cycle-Consistency for Robust Visual Question Answering", "Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["2304.14614", "2206.09391", "2104.02000", "1709.08693", "1902.05660", "1412.6572"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_29"} +{"question": "Which works focused on ray-based rendering for novel view synthesis approach?", "answer": ["Stereo Radiance Fields (SRF): Learning View Synthesis for Sparse Views\n of Novel Scenes", "IBRNet: Learning Multi-View Image-Based Rendering", "Generalizable Patch-Based Neural Rendering", "Is Attention All That NeRF Needs?", "Explicit Correspondence Matching for Generalizable Neural Radiance\n Fields"], "answer_arxiv_id": ["2104.06935", "2102.13090", "2207.10662", "2207.13298", "2304.12294"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_test_52"} +{"question": "Which papers contribute to the advancement of model-based reinforcement learning through the study of the world model?", "answer": ["Recurrent World Models Facilitate Policy Evolution", "Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Learning to Fly via Deep Model-Based Reinforcement Learning", "Mastering Atari with Discrete World Models", "Mastering Diverse Domains through World Models", "Model Based Reinforcement Learning for Atari", "Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model"], "answer_arxiv_id": ["1809.01999", "1811.04551", "1912.01603", "2003.08876", "2010.02193", "2301.04104", "1903.00374", "1911.08265"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_53"} +{"question": "Which studies have recently been working on the integration of visual perception and large language models?", "answer": ["Attention Is All You Need", "Language Models are Few-Shot Learners", "GPT-4 Technical Report", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["1706.03762", "2005.14165", "2303.08774", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_60"} +{"question": "Are there any works that improve cost-effectiveness, performance, and data generation quality in the prompting framework of large language models?", "answer": ["ReWOO: Decoupling Reasoning from Observations for Efficient Augmented\n Language Models", "Reflexion: Language Agents with Verbal Reinforcement Learning", "MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action", "ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs", "ToolAlpaca: Generalized Tool Learning for Language Models with 3000\n Simulated Cases"], "answer_arxiv_id": ["2305.18323", "2303.11366", "2303.11381", "2307.16789", "2306.05301"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_test_63"} +{"question": "Could you provide some works about deep AD approaches that employ a self-supervised loss function to train the detector and score anomalies?", "answer": ["Deep Anomaly Detection Using Geometric Transformations", "Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty", "Learning and Evaluating Representations for Deep One-class Classification", "Classification-Based Anomaly Detection for General Data", "Neural Transformation Learning for Deep Anomaly Detection Beyond Images", "Detecting Anomalies within Time Series using Local Neural Transformations", "Deep Anomaly Detection under Labeling Budget Constraints"], "answer_arxiv_id": ["1805.10917", "1906.12340", "2011.02578", "2005.02359", "2103.16440", "2202.03944", "2302.07832v2"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_test_95"} +{"question": "What are some works in vision that stress the importance of data selection in supervised or semi-supervised setting?", "answer": ["Beyond neural scaling laws: beating power law scaling via data pruning", "Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt", "Learning From Less Data: A Unified Data Subset Selection and Active Learning Framework for Computer Vision", "Glister: Generalization based Data Subset Selection for Efficient and Robust Learning", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training", "RETRIEVE: Coreset Selection for Efficient and Robust Semi-Supervised Learning", "Optimizing Data Usage via Differentiable Rewards", "Deep Learning on a Data Diet: Finding Important Examples Early in Training", "Coresets for Data-efficient Training of Machine Learning Models", "Selection via Proxy: Efficient Data Selection for Deep Learning", "Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["2206.14486v6", "2206.07137", "1901.01151", "2012.10630", "2103.00123", "2106.07760v2", "1911.10088", "2107.07075", "1906.01827", "1906.11829", "1708.00489"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_test_117"} +{"question": "What works adopted large language models (LLMs) for a cost-effective generation of Counterfactually Augmented Data (CAD)?", "answer": ["Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and\n Improving Models", "Generate Your Counterfactuals: Towards Controlled Counterfactual\n Generation for Text", "AutoCAD: Automatically Generating Counterfactuals for Mitigating\n Shortcut Learning", "CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generation", "Automatic Prompt Optimization with \"Gradient Descent\" and Beam Search", "DISCO: Distilling Counterfactuals with Large Language Models"], "answer_arxiv_id": ["2101.00288", "2012.04698", "2211.16202", "2210.04873", "2305.03495", "2212.10534"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_test_118"} +{"question": "Which works focus on predicting model generalization error?", "answer": ["Are Labels Always Necessary for Classifier Accuracy Evaluation?", "Leveraging Unlabeled Data to Predict Out-of-Distribution Performance", "Predicting Out-of-Distribution Error with the Projection Norm", "On the Strong Correlation Between Model Invariance and Generalization", "Predicting with Confidence on Unseen Distributions", "What Does Rotation Prediction Tell Us about Classifier Accuracy under Varying Testing Environments?"], "answer_arxiv_id": ["2007.02915", "2201.04234", "2202.05834", "2207.07065", "2107.03315", "2106.05961"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_test_123"} +{"question": "Any works about user-annotations based image animation?", "answer": ["iPOKE: Poking a Still Image for Controlled Stochastic Video Synthesis", "Stochastic Latent Residual Video Prediction", "DragNUWA: Fine-grained Control in Video Generation by Integrating Text,\n Image, and Trajectory", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "ControlVideo: Training-free Controllable Text-to-Video Generation", "Motion-Conditioned Diffusion Model for Controllable Video Synthesis"], "answer_arxiv_id": ["2107.02790", "2002.09219", "2308.08089", "2306.02018", "2305.13077", "2304.14404"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_test_135"} +{"question": "Could you provide some works that discuss multimodal prompting methods?", "answer": ["Large Language Models are Zero-Shot Reasoners", "Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning\n by Large Language Models", "Better Zero-Shot Reasoning with Self-Adaptive Prompting", "Language Models are Few-Shot Learners", "Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "A Survey on In-context Learning", "Fairness-guided Few-shot Prompting for Large Language Models", "ExpertPrompting: Instructing Large Language Models to be Distinguished\n Experts", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Automatic Chain of Thought Prompting in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Graph of Thoughts: Solving Elaborate Problems with Large Language Models", "Beyond Chain-of-Thought, Effective Graph-of-Thought Reasoning in\n Language Models", "Boosting Logical Reasoning in Large Language Models through a New\n Framework: The Graph of Thought"], "answer_arxiv_id": ["2205.11916", "2305.04091", "2305.14106", "2005.14165", "2202.12837", "2301.00234", "2303.13217", "2305.14688", "2201.11903", "2210.03493", "2203.11171", "2305.10601", "2308.09687", "2305.16582", "2308.08614"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_test_144"} +{"question": "What works focused on MAML and its variants?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Sharp-MAML: Sharpness-Aware Model-Agnostic Meta Learning", "Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML", "Alpha MAML: Adaptive Model-Agnostic Meta-Learning", "Meta-Learning with Implicit Gradients"], "answer_arxiv_id": ["1703.03400", "2206.03996", "1909.09157", "1905.07435", "1909.04630"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_test_146"} +{"question": "What studies have leveraged extensive image-text pair datasets to broaden the detection vocabulary in Open-vocabulary detection?", "answer": ["Open-Vocabulary Object Detection Using Captions", "RegionCLIP: Region-based Language-Image Pretraining", "PromptDet: Towards Open-vocabulary Detection using Uncurated Images", "Grounded Language-Image Pre-training", "Learning Object-Language Alignments for Open-Vocabulary Object Detection", "DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via\n Word-Region Alignment"], "answer_arxiv_id": ["2011.10678", "2112.09106", "2203.16513", "2112.03857", "2211.14843", "2304.04514"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_test_148"} +{"question": "Which works can you provide that are focused on creating evaluation data on Indic languages?", "answer": ["Towards Leaving No Indic Language Behind: Building Monolingual Corpora,\n Benchmark and Models for Indic Languages", "Naamapadam: A Large-Scale Named Entity Annotated Data for Indic\n Languages", "MASSIVE: A 1M-Example Multilingual Natural Language Understanding\n Dataset with 51 Typologically-Diverse Languages", "GLUECoS : An Evaluation Benchmark for Code-Switched NLP", "The Belebele Benchmark: a Parallel Reading Comprehension Dataset in 122\n Language Variants"], "answer_arxiv_id": ["2212.05409", "2212.10168", "2204.08582", "2004.12376", "2308.16884"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_test_150"} +{"question": "Which papers have proposed for extracting the specific style from reference images?", "answer": ["StyleAdapter: A Single-Pass LoRA-Free Model for Stylized Image\n Generation", "Domain Enhanced Arbitrary Image Style Transfer via Contrastive Learning", "StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion\n Models", "Inversion-Based Style Transfer with Diffusion Models", "StyleDrop: Text-to-Image Generation in Any Style"], "answer_arxiv_id": ["2309.01770", "2205.09542", "2308.07863", "2211.13203", "2306.00983"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_test_153"} +{"question": "Which works combine external knowledge from KGs into LLMs during the prompting stage?", "answer": ["Knowledge-Augmented Language Model Prompting for Zero-Shot Knowledge Graph Question Answering", "Knowledge-Driven CoT: Exploring Faithful Reasoning in LLMs for\n Knowledge-intensive Question Answering", "MindMap: Knowledge Graph Prompting Sparks Graph of Thoughts in Large\n Language Models", "Reasoning on Graphs: Faithful and Interpretable Large Language Model\n Reasoning", "Think-on-Graph: Deep and Responsible Reasoning of Large Language Model\n on Knowledge Graph"], "answer_arxiv_id": ["2306.04136v1", "2308.13259", "2308.09729", "2310.01061", "2307.07697"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_test_155"} +{"question": "Which papers have achieved progress in the field of graph contrastive learning?", "answer": ["Graph Contrastive Learning with Augmentations", "Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "Graph Contrastive Learning Automated", "Adversarial Graph Contrastive Learning with Information Regularization", "Bringing Your Own View: Graph Contrastive Learning without Prefabricated Data Augmentations"], "answer_arxiv_id": ["2010.13902", "2106.05819", "2106.07594", "2202.06491", "2201.01702"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_test_161"} +{"question": "Could you tell me what studies propose to bridge vision and language modalities through visual prompt generators?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Visual Instruction Tuning", "Language Is Not All You Need: Aligning Perception with Language Models"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2305.06500", "2304.08485", "2302.14045"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_test_164"} +{"question": "What research introduced DPMs and linked the generative model to a denoising diffusion model?", "answer": ["Auto-Encoding Variational Bayes", "Generative Adversarial Nets", "Towards Building A Group-based Unsupervised Representation Disentanglement Framework", "Learning Disentangled Representation by Exploiting Pretrained Generative Models: A Contrastive Learning View", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality", "Gotta Go Fast When Generating Data with Score-Based Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["1312.6114", "1406.2661", "2102.10303", "2102.10543", "1503.03585", "2006.11239", "2102.09672", "2010.02502", "2202.05830", "2105.14080", "2201.06503", "2106.05931"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_test_189"} +{"question": "Which researchers proposed altering the memory-computation trade-off of the neural architecture for improving computational speed in neural scene representations?", "answer": ["DeRF: Decomposed Radiance Fields", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "FastNeRF: High-Fidelity Neural Rendering at 200FPS", "Plenoxels: Radiance Fields without Neural Networks", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction"], "answer_arxiv_id": ["2011.12490", "2103.13744", "2103.10380", "2112.05131", "2111.11215"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_198"} +{"question": "Can you provide references for grouping-based methods of 3D instance segmentation?", "answer": ["PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation", "Hierarchical Aggregation for 3D Instance Segmentation", "Instance Segmentation in 3D Scenes using Semantic Superpoint Tree\n Networks", "MaskGroup: Hierarchical Point Grouping and Masking for 3D Instance\n Segmentation", "SoftGroup for 3D Instance Segmentation on Point Clouds", "3D Instances as 1D Kernels", "ISBNet: a 3D Point Cloud Instance Segmentation Network with\n Instance-aware Sampling and Box-aware Dynamic Convolution"], "answer_arxiv_id": ["2004.01658", "2108.02350", "2108.07478", "2203.14662", "2203.01509", "2207.07372", "2303.00246"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_test_208"} +{"question": "Who are the researchers that attempted to close the gap between QM calculations and ML potentials?", "answer": ["SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects", "OrbNet: Deep Learning for Quantum Chemistry Using Symmetry-Adapted Atomic-Orbital Features", "Finding Density Functionals with Machine Learning", "Ab-Initio Potential Energy Surfaces by Pairing GNNs with Neural Wave Functions", "Generalizing Neural Wave Functions", "Sampling-free Inference for Ab-Initio Potential Energy Surface Networks"], "answer_arxiv_id": ["2105.00304", "2007.08026", "1112.5441", "2110.05064", "2302.04168", "2205.14962"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_test_212"} +{"question": "Which papers propose graph-based approaches for capturing longer-term dependencies in 3D human pose forecasting?", "answer": ["Learning Trajectory Dependencies for Human Motion Prediction", "MSR-GCN: Multi-Scale Residual Graph Convolution Networks for Human\n Motion Prediction", "Dynamic Multiscale Graph Neural Networks for 3D Skeleton-Based Human\n Motion Prediction", "Space-Time-Separable Graph Convolutional Network for Pose Forecasting", "Spatio-Temporal Gating-Adjacency GCN for Human Motion Prediction", "Multitask Non-Autoregressive Model for Human Motion Prediction", "Diverse Human Motion Prediction Guided by Multi-Level Spatial-Temporal\n Anchors"], "answer_arxiv_id": ["1908.05436", "2108.07152", "2003.08802", "2110.04573", "2203.01474", "2007.06426", "2302.04860"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_test_227"} +{"question": "In what works can I find large-scale unsupervised pre-training on unstructured text for multilingual corpora?", "answer": ["BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "What Language Model to Train if You Have One Million GPU Hours?", "The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset", "MADLAD-400: A Multilingual And Document-Level Large Audited Dataset", "LLM-powered Data Augmentation for Enhanced Cross-lingual Performance"], "answer_arxiv_id": ["2211.05100", "2210.15424", "2303.03915", "2309.04662", "2305.14288"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_test_266"} +{"question": "Which papers focus on broader applications of NeRF, including generative modeling, video synthesis, and scene editing?", "answer": ["GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images", "VoLux-GAN: A Generative Model for 3D Face Synthesis with HDRI Relighting", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Tensor4D : Efficient Neural 4D Decomposition for High-fidelity Dynamic\n Reconstruction and Rendering", "Neural Radiance Flow for 4D View Synthesis and Video Processing", "Editing Conditional Radiance Fields", "NeRF-Editing: Geometry Editing of Neural Radiance Fields"], "answer_arxiv_id": ["2209.11163", "2201.04873", "2011.13084", "2211.11610", "2012.09790", "2105.06466", "2205.04978"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_267"} +{"question": "What studies work on body motion conditioned on text descriptions?", "answer": ["FLAME: Free-form Language-based Motion Synthesis & Editing", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE", "Synthesizing Long-Term Human Motions with Diffusion Models via Coherent\n Sampling", "TEMOS: Generating diverse human motions from textual descriptions", "Synthesis of Compositional Animations from Textual Descriptions"], "answer_arxiv_id": ["2209.00349", "2104.05670", "2308.01850", "2204.14109", "2103.14675"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_test_275"} +{"question": "Which studies describe model structures that implicitly generate reasoning processes?", "answer": ["Program Induction by Rationale Generation : Learning to Solve and\n Explain Algebraic Word Problems", "TAT-QA: A Question Answering Benchmark on a Hybrid of Tabular and\n Textual Content in Finance", "Answering Numerical Reasoning Questions in Table-Text Hybrid Contents\n with Graph-based Encoder and Tree-based Decoder", "Chaining Simultaneous Thoughts for Numerical Reasoning", "ELASTIC: Numerical Reasoning with Adaptive Symbolic Compiler"], "answer_arxiv_id": ["1705.04146", "2105.07624", "2209.07692", "2211.16482", "2210.10105"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_test_295"} +{"question": "What papers proposed iterative methods for transferable adversarial attacks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Boosting Adversarial Attacks with Momentum", "Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks", "Enhancing the Transferability of Adversarial Attacks through Variance Tuning", "Improving Transferability of Adversarial Examples with Input Diversity", "On Improving Adversarial Transferability of Vision Transformers", "Cross-Modal Transferable Adversarial Attacks from Images to Videos"], "answer_arxiv_id": ["1412.6572", "1710.06081", "1908.06281", "2103.15571", "1803.06978", "2106.04169", "2112.05379"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_test_298"} +{"question": "Which papers proposed datasets for open domain question answering (QA) for English and other languages?", "answer": ["TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages", "XOR QA: Cross-lingual Open-Retrieval Question Answering", "MIA 2022 Shared Task: Evaluating Cross-lingual Open-Retrieval Question Answering for 16 Diverse Languages", "MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering", "Mr. TYDI: A Multi-lingual Benchmark for Dense Retrieval"], "answer_arxiv_id": ["2003.05002", "2010.11856v3", "2207.00758", "2007.15207", "2108.08787"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_test_299"} +{"question": "What works used open-loop imitation learning for predicting the behavior of the ego vehicle in autonomous driving?", "answer": ["End to End Learning for Self-Driving Cars", "PRECOG: PREdiction Conditioned On Goals in Visual Multi-Agent Settings", "End-to-end Driving via Conditional Imitation Learning", "SafetyNet: Safe planning for real-world self-driving vehicles using machine-learned policies", "Learning by cheating"], "answer_arxiv_id": ["1604.07316", "1905.01296", "1710.02410", "2109.13602", "1912.12294"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_test_307"} +{"question": "Can you name some studies that propose different metrics to prune networks at initialization?", "answer": ["Picking Winning Tickets Before Training by Preserving Gradient Flow", "Pruning neural networks without any data by iteratively conserving synaptic flow", "Progressive Skeletonization: Trimming more fat from a network at initialization", "PHEW : Constructing Sparse Networks that Learn Fast and Generalize Well Without Training Data", "Prospect Pruning: Finding Trainable Weights at Initialization using Meta-Gradients"], "answer_arxiv_id": ["2002.07376", "2006.05467", "2006.09081", "2010.11354", "2202.08132"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_test_309"} +{"question": "Which studies highlight the benefit of capturing long-distance relations in Graph Neural Networks (GNNs) by stacking more feature aggregation layers or unrolling various fixed point iterations?", "answer": ["Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "Implicit Graph Neural Networks", "Towards Deeper Graph Neural Networks", "Simple and Deep Graph Convolutional Networks", "Training Graph Neural Networks with 1000 Layers", "A Unified View on Graph Neural Networks as Graph Signal Denoising", "Interpreting and Unifying Graph Neural Networks with An Optimization Framework"], "answer_arxiv_id": ["1810.05997", "2009.06211", "2007.09296", "2007.02133", "2106.07476", "2010.01777", "2101.11859"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_test_318"} +{"question": "Could you provide me some examples of research that discusses the application of data augmentations in the latent space?", "answer": ["FreeLB: Enhanced Adversarial Training for Natural Language Understanding", "AdvAug: Robust Adversarial Augmentation for Neural Machine Translation", "DoubleMix: Simple Interpolation-Based Data Augmentation for Text\n Classification", "Text Smoothing: Enhance Various Data Augmentation Methods on Text\n Classification Tasks", "Controlled Text Generation for Data Augmentation in Intelligent\n Artificial Agents"], "answer_arxiv_id": ["1909.11764", "2006.11834", "2209.05297", "2202.13840", "1910.03487"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_test_323"} +{"question": "Could you provide examples of works about certified defenses focused on unimodal models?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing", "Certified Robustness for Top-k Predictions against Adversarial\n Perturbations via Randomized Smoothing", "Robustness Certificates for Sparse Adversarial Attacks by Randomized\n Ablation", "Certified Defenses for Adversarial Patches", "SAFER: A Structure-free Approach for Certified Robustness to Adversarial\n Word Substitutions", "Certified Robustness to Adversarial Examples with Differential Privacy", "PointGuard: Provably Robust 3D Point Cloud Classification", "Certified Robustness to Text Adversarial Attacks by Randomized [MASK]", "PatchCleanser: Certifiably Robust Defense against Adversarial Patches\n for Any Image Classifier", "MultiGuard: Provably Robust Multi-label Classification against\n Adversarial Examples", "TextGuard: Provable Defense against Backdoor Attacks on Text\n Classification", "PointCert: Point Cloud Classification with Deterministic Certified\n Robustness Guarantees"], "answer_arxiv_id": ["1902.02918", "1912.09899", "1911.09272", "2003.06693", "2005.14424", "1802.03471", "2103.03046", "2105.03743", "2108.09135", "2210.01111", "2311.11225", "2303.01959"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_test_329"} +{"question": "What studies focus on the different techniques utilized to fine-tune the pre-trained models?", "answer": ["Scaling Instruction-Finetuned Language Models", "Training language models to follow instructions with human feedback", "Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally\n Across Scales and Tasks"], "answer_arxiv_id": ["2210.11416", "2203.02155", "1902.00751", "2106.09685", "2101.00190", "2104.08691", "2110.07602"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_test_337"} +{"question": "Which papers approach the studies about adversarial attacks?", "answer": ["Intriguing properties of neural networks", "Evasion Attacks against Machine Learning at Test Time", "Towards Evaluating the Robustness of Neural Networks", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Reliable evaluation of adversarial robustness with an ensemble of\n diverse parameter-free attacks", "Obfuscated Gradients Give a False Sense of Security: Circumventing\n Defenses to Adversarial Examples"], "answer_arxiv_id": ["1312.6199", "1708.06131", "1608.04644", "1706.06083", "2003.01690", "1802.00420"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_343"} +{"question": "What works have used hypercolumns for tasks like keypoint detection, segmentation and semantic correspondence?", "answer": ["Hypercolumns for Object Segmentation and Fine-grained Localization", "Deep Layer Aggregation", "Hyperpixel Flow: Semantic Correspondence with Multi-layer Neural Features", "AnchorNet: A Weakly Supervised Network to Learn Geometry-sensitive Features For Semantic Matching", "Learning to Compose Hypercolumns for Visual Correspondence", "Neural Best-Buddies: Sparse Cross-Domain Correspondence"], "answer_arxiv_id": ["1411.5752", "1707.06484", "1908.06537", "1704.04749", "2007.10587", "1805.04140v2"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_test_359"} +{"question": "Which works have explored self-consistency techniques for refining language models in post-hoc correction?", "answer": ["Language Models (Mostly) Know What They Know", "Self-Evaluation Improves Selective Generation in Large Language Models", "Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence\n Scores from Language Models Fine-Tuned with Human Feedback", "Self-Refine: Iterative Refinement with Self-Feedback", "Chain-of-Verification Reduces Hallucination in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2207.05221", "2312.09300", "2305.14975", "2303.17651", "2309.11495", "2203.11171"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_test_365"} +{"question": "Which studies deal with aligning visual features with pre-trained LLMs for multimodal comprehension tasks?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Image as a Foreign Language: BEiT Pretraining for All Vision and\n Vision-Language Tasks", "Visual Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "Language Is Not All You Need: Aligning Perception with Language Models", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2304.10592", "2208.10442", "2304.08485", "2304.14178", "2302.14045", "2301.12597", "2305.11175", "2305.03726"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_372"} +{"question": "Could you refer me to some studies that use score-based models for graph generation?", "answer": ["Permutation Invariant Graph Generation via Score-Based Generative Modeling", "Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations", "Score-Based Generative Modeling through Stochastic Differential Equations", "DiGress: Discrete Denoising diffusion for graph generation", "Diffusion Models for Graphs Benefit From Discrete State Spaces"], "answer_arxiv_id": ["2003.00638", "2202.02514", "2011.13456", "2209.14734", "2210.01549"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_test_393"} +{"question": "Which works propose consistency-based methods for detecting non-factual generations in LLM generated content?", "answer": ["Measuring and Improving Consistency in Pretrained Language Models", "Self-contradictory Hallucinations of Large Language Models: Evaluation,\n Detection and Mitigation", "How to Catch an AI Liar: Lie Detection in Black-Box LLMs by Asking\n Unrelated Questions", "SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for\n Generative Large Language Models", "LM vs LM: Detecting Factual Errors via Cross Examination", "The Internal State of an LLM Knows When It's Lying", "Chain-of-Verification Reduces Hallucination in Large Language Models", "Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation\n in Natural Language Generation", "Language Models (Mostly) Know What They Know", "Representation Engineering: A Top-Down Approach to AI Transparency", "Factcheck-Bench: Fine-Grained Evaluation Benchmark for Automatic\n Fact-checkers", "RARR: Researching and Revising What Language Models Say, Using Language\n Models", "FacTool: Factuality Detection in Generative AI -- A Tool Augmented\n Framework for Multi-Task and Multi-Domain Scenarios"], "answer_arxiv_id": ["2102.01017", "2305.15852", "2309.15840", "2303.08896", "2305.13281", "2304.13734", "2309.11495", "2302.09664", "2207.05221", "2310.01405", "2311.09000", "2210.08726", "2307.13528"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_test_402"} +{"question": "Are there any studies in sports video understanding which involves benchmarks for spatio-temporal reasoning?", "answer": ["UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild", "MultiSports: A Multi-Person Video Dataset of Spatio-Temporally Localized\n Sports Actions", "FineGym: A Hierarchical Video Dataset for Fine-grained Action\n Understanding", "SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos", "SportsMOT: A Large Multi-Object Tracking Dataset in Multiple Sports\n Scenes", "Social Adaptive Module for Weakly-supervised Group Activity Recognition", "A Hierarchical Deep Temporal Model for Group Activity Recognition"], "answer_arxiv_id": ["1212.0402", "2105.07404", "2004.06704", "1804.04527", "2304.05170", "2007.09470", "1511.06040"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_test_407"} +{"question": "What studies discuss the field of 'learning from human feedback'?", "answer": ["Neural Machine Translation by Jointly Learning to Align and Translate", "WebGPT: Browser-assisted question-answering with human feedback", "Training language models to follow instructions with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback", "Constitutional AI: Harmlessness from AI Feedback", "ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation", "Text-guided Image-and-Shape Editing and Generation: A Short Survey", "Aligning Text-to-Image Models using Human Feedback", "RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignment"], "answer_arxiv_id": ["1409.0473", "2112.09332", "2203.02155", "2204.05862", "2212.08073", "2304.05977", "2304.09244", "2302.12192", "2304.06767"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_test_427"} +{"question": "What are some works that have focused on how LLMs can be connected to visual foundation models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Language-based Action Concept Spaces Improve Video Self-Supervised\n Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models"], "answer_arxiv_id": ["2204.14198", "2307.10922", "2301.12597", "2304.08485", "2306.05424"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_test_432"} +{"question": "Could you provide me some works where human feedback was utilised to finetune large language models?", "answer": ["Neural Machine Translation by Jointly Learning to Align and Translate", "WebGPT: Browser-assisted question-answering with human feedback", "Training language models to follow instructions with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback", "Constitutional AI: Harmlessness from AI Feedback"], "answer_arxiv_id": ["1409.0473", "2112.09332", "2203.02155", "2204.05862", "2212.08073"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_test_437"} +{"question": "Which works have used pre-computing or post-computing methods for feature aggregation in GNN models?", "answer": ["Simplifying Graph Convolutional Networks", "SIGN: Scalable Inception Graph Neural Networks", "Scalable and Adaptive Graph Neural Networks with Self-Label-Enhanced Training", "Graph Attention Multi-Layer Perceptron", "Scaling Graph Neural Networks with Approximate PageRank", "Combining Label Propagation and Simple Models out-performs Graph Neural Networks"], "answer_arxiv_id": ["1902.07153", "2004.11198", "2104.09376", "2206.04355", "2007.01570", "2010.13993"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_test_445"} +{"question": "What research introduced methods that adapt the training procedure of the classifier itself?", "answer": ["Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Flipout: Efficient Pseudo-Independent Weight Perturbations on Mini-Batches", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "When Does Label Smoothing Help?", "Transferable Calibration with Lower Bias and Variance in Domain Adaptation", "On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networks", "mixup: Beyond Empirical Risk Minimization", "Evidential Deep Learning to Quantify Classification Uncertainty", "Towards Trustworthy Predictions from Deep Neural Networks with Fast Adversarial Calibration", "Pitfalls of In-Domain Uncertainty Estimation and Ensembling in Deep Learning"], "answer_arxiv_id": ["1906.12340", "1506.02142", "1803.04386", "1612.01474", "1906.02629", "2007.08259", "1905.11001", "1710.09412", "1806.01768", "2012.10923", "2002.06470"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_test_447"} +{"question": "What are some representative works about graph embedding-based methods?", "answer": ["RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space", "Convolutional 2D Knowledge Graph Embeddings", "Complex Embeddings for Simple Link Prediction", "Holographic Embeddings of Knowledge Graphs", "kbgan: Adversarial Learning for Knowledge Graph Embeddings", "TuckER: Tensor Factorization for Knowledge Graph Completion", "Embedding Entities and Relations for Learning and Inference in Knowledge Bases", "Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings", "BoxE: A Box Embedding Model for Knowledge Base Completion", "Modeling Fine-Grained Entity Types with Box Embeddings"], "answer_arxiv_id": ["1902.10197", "1707.01476", "1606.06357", "1510.04935", "1711.04071", "1901.09590", "1412.6575", "2002.05969", "2007.06267v2", "2101.00345"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_451"} +{"question": "What are some works related to the Mean Teacher paradigm?", "answer": ["Self-supervised Augmentation Consistency for Adapting Semantic\n Segmentation", "DAFormer: Improving Network Architectures and Training Strategies for\n Domain-Adaptive Semantic Segmentation", "Prototypical Pseudo Label Denoising and Target Structure Learning for\n Domain Adaptive Semantic Segmentation", "End-to-End Semi-Supervised Object Detection with Soft Teacher", "Active Teacher for Semi-Supervised Object Detection", "Consistent-Teacher: Towards Reducing Inconsistent Pseudo-targets in Semi-supervised Object Detection", "Omni-DETR: Omni-Supervised Object Detection with Transformers", "ALWOD: Active Learning for Weakly-Supervised Object Detection", "Contrastive Mean Teacher for Domain Adaptive Object Detectors", "Cross-Domain Adaptive Teacher for Object Detection", "Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain\n Adaptation on Person Re-identification", "Exploiting Sample Uncertainty for Domain Adaptive Person\n Re-Identification", "Delving into Probabilistic Uncertainty for Unsupervised Domain Adaptive\n Person Re-Identification"], "answer_arxiv_id": ["2105.00097", "2111.14887", "2101.10979", "2106.09018", "2303.08348", "2209.01589v3", "2203.16089", "2309.07914", "2305.03034", "2111.13216", "2001.01526", "2012.08733", "2112.14025"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_test_452"} +{"question": "Any studies about generating adversarial examples in textual domains?", "answer": ["Adversarial Examples for Evaluating Reading Comprehension Systems", "Generating Natural Language Adversarial Examples", "Adversarial VQA: A New Benchmark for Evaluating the Robustness of VQA\n Models", "HotFlip: White-Box Adversarial Examples for Text Classification", "Universal Adversarial Triggers for Attacking and Analyzing NLP"], "answer_arxiv_id": ["1707.07328", "1804.07998", "2106.00245", "1712.06751", "1908.07125"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_457"} +{"question": "Could you provide some examples of diffusion models that involve different number of denoising steps and parameterization of transformation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2010.02502", "2206.00927", "2211.01095", "2202.09778"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_test_469"} +{"question": "What works propose strategies for face capture that are more easily accessible and convenient for daily users?", "answer": ["AvatarMe: Realistically Renderable 3D Facial Reconstruction\n \"in-the-wild\"", "Relightify: Relightable 3D Faces from a Single Image via Diffusion\n Models", "Learning a 3D Morphable Face Reflectance Model from Low-cost Data", "A Morphable Face Albedo Model", "Learning Formation of Physically-Based Face Attributes", "FitMe: Deep Photorealistic 3D Morphable Model Avatars", "Practical Face Reconstruction via Differentiable Ray Tracing"], "answer_arxiv_id": ["2003.13845", "2305.06077", "2303.11686", "2004.02711", "2004.03458", "2305.09641", "2101.05356"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_test_492"} +{"question": "What papers have incorporated the use of NLEs in fields beyond NLP, such as in computer vision, medical field, and self-driving cars?", "answer": ["Grounding Visual Explanations", "From Recognition to Cognition: Visual Commonsense Reasoning", "Knowledge-Grounded Self-Rationalization via Extractive and Natural\n Language Explanations", "Explaining Chest X-ray Pathologies in Natural Language", "Textual Explanations for Self-Driving Vehicles"], "answer_arxiv_id": ["1807.09685", "1811.10830", "2106.13876", "2207.04343", "1807.11546"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_test_522"} +{"question": "Can you provide papers that discussed the concept of latent embeddings?", "answer": ["RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space", "Convolutional 2D Knowledge Graph Embeddings", "Complex Embeddings for Simple Link Prediction", "Holographic Embeddings of Knowledge Graphs", "kbgan: Adversarial Learning for Knowledge Graph Embeddings", "TuckER: Tensor Factorization for Knowledge Graph Completion", "Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings", "BoxE: A Box Embedding Model for Knowledge Base Completion", "Modeling Fine-Grained Entity Types with Box Embeddings"], "answer_arxiv_id": ["1902.10197", "1707.01476", "1606.06357", "1510.04935", "1711.04071", "1901.09590", "2002.05969", "2007.06267v2", "2101.00345"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_527"} +{"question": "What are some studies that have used data statistics, representations, logits, and embedding to avoid exposing privacy in Federated Learning?", "answer": ["XOR Mixup: Privacy-Preserving Data Augmentation for One-Shot Federated Learning", "Towards Fair Federated Learning with Zero-Shot Data Augmentation", "Cronus: Robust and Heterogeneous Collaborative Learning with Black-Box Knowledge Transfer", "No Fear of Heterogeneity: Classifier Calibration for Federated Learning with Non-IID Data", "FedProto: Federated Prototype Learning across Heterogeneous Clients"], "answer_arxiv_id": ["2006.05148", "2104.13417", "1912.11279v1", "2106.05001", "2105.00243"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_test_533"} +{"question": "What are the key works in the field of diffusion models which are a class of generative probabilistic models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Diffusion Models in Vision: A Survey", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2209.04747", "2105.05233", "2006.11239", "2102.09672"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_test_534"} +{"question": "Which research papers introduced initial vision-language pre-training models?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning", "Unified Vision-Language Pre-Training for Image Captioning and VQA", "Unifying Vision-and-Language Tasks via Text Generation", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision", "Large-Scale Adversarial Training for Vision-and-Language Representation\n Learning", "Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal\n Transformers", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "12-in-1: Multi-Task Vision and Language Representation Learning", "Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language\n Representation Learning", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision"], "answer_arxiv_id": ["1909.11740", "1909.11059", "2102.02779", "2004.06165", "2102.03334", "2006.06195", "2004.00849", "1908.02265", "1908.08530", "1912.02315", "2104.03135", "2108.10904"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_test_543"} +{"question": "What etudies are on Transformer-based models for speech that have been used to test their brain alignment for speech-evoked brain activity?", "answer": ["Vector-Quantized Autoregressive Predictive Coding", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech\n Representations", "HuBERT: Self-Supervised Speech Representation Learning by Masked\n Prediction of Hidden Units", "Toward a realistic model of speech processing in the brain with\n self-supervised learning", "Self-supervised models of audio effectively explain human cortical\n responses to speech"], "answer_arxiv_id": ["2005.08392", "2006.11477", "2106.07447", "2206.01685", "2205.14252"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_551"} +{"question": "Are there any methods using hierarchical Reinforcement Learning to decompose complex tasks into sub-tasks?", "answer": ["Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation", "The Option-Critic Architecture", "Near-Optimal Representation Learning for Hierarchical Reinforcement Learning", "Language as an Abstraction for Hierarchical Deep Reinforcement Learning", "Unsupervised Skill Discovery with Bottleneck Option Learning", "Toward Robust Long Range Policy Transfer"], "answer_arxiv_id": ["1604.06057", "1609.05140", "1810.01257", "1906.07343", "2106.14305", "2103.02957"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_test_555"} +{"question": "In what studies LMMs directly reason over embedded visual features?", "answer": ["Visual Instruction Tuning", "Improved Baselines with Visual Instruction Tuning", "Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with\n Modality Collaboration", "MultiModal-GPT: A Vision and Language Model for Dialogue with Humans", "PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2304.08485", "2310.03744", "2204.14198", "2301.12597", "2305.06500", "2304.10592", "2304.14178", "2311.04257", "2305.04790", "2303.03378"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_test_558"} +{"question": "Which works have implemented multimodal understanding and generative capacities across modalities?", "answer": ["ImageBind: One Embedding Space To Bind Them All", "Any-to-Any Generation via Composable Diffusion", "Generating Images with Multimodal Language Models", "NExT-GPT: Any-to-Any Multimodal LLM", "Emu: Generative Pretraining in Multimodality", "Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction\n Tuning"], "answer_arxiv_id": ["2305.05665v2", "2305.11846", "2305.17216", "2309.05519", "2307.05222", "2309.02591"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_test_566"} +{"question": "What works have explored the field of zero-shot segmentation recently?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "ReCo: Retrieve and Co-segment for Zero-shot Transfer", "Image Segmentation Using Text and Image Prompts", "Zero-Shot Semantic Segmentation", "DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic\n Segmentation Using Diffusion Models", "Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2401.14159", "2206.07045", "2112.10003", "1906.00817", "2303.11681", "2112.01071"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_test_569"} +{"question": "Which papers discuss solutions to commonsense reasoning problems?", "answer": ["CommonsenseQA: A Question Answering Challenge Targeting Commonsense\n Knowledge", "CommonsenseQA 2.0: Exposing the Limits of AI through Gamification", "Cosmos QA: Machine Reading Comprehension with Contextual Commonsense\n Reasoning", "Abductive Commonsense Reasoning", "SocialIQA: Commonsense Reasoning about Social Interactions"], "answer_arxiv_id": ["1811.00937", "2201.05320", "1909.00277", "1908.05739", "1904.09728"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_test_579"} +{"question": "What works are amongst the most influential in relation to U-Net?", "answer": ["UNet++: A Nested U-Net Architecture for Medical Image Segmentation", "Attention U-Net: Learning Where to Look for the Pancreas", "3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation", "Denoising Diffusion Probabilistic Models", "nnU-Net: Self-adapting Framework for U-Net-Based Medical Image Segmentation", "A Probabilistic U-Net for Segmentation of Ambiguous Images", "A Variational U-Net for Conditional Appearance and Shape Generation", "Road Extraction by Deep Residual U-Net"], "answer_arxiv_id": ["1807.10165", "1804.03999", "1606.06650", "2006.11239", "1809.10486", "1806.05034", "1804.04694", "1711.10684"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_test_615"} +{"question": "What works used text-based language models to predict text-evoked and speech-evoked brain activity?", "answer": ["Interpreting and improving natural-language processing (in machines)\n with natural language-processing (in the brain)", "Inducing brain-relevant bias in natural language processing models", "Relating Simple Sentence Representations in Deep Neural Networks and the\n Brain", "Low-Dimensional Structure in the Space of Language Representations is\n Reflected in Brain Responses", "Neural Language Taskonomy: Which NLP Tasks are the most Predictive of\n fMRI Brain Activity?", "Language models and brain alignment: beyond word-level semantics and\n prediction", "Joint processing of linguistic properties in brains and language models"], "answer_arxiv_id": ["1905.11833", "1911.03268", "1906.11861", "2106.05426", "2205.01404", "2212.00596", "2212.08094"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_621"} +{"question": "Could you provide examples of image-text datasets that have their own preprocessing techniques?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Combined Scaling for Zero-shot Transfer Learning", "Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts", "RedCaps: Web-curated image-text data created by the people, for the people", "LAION-5B: An open large-scale dataset for training next generation image-text models"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.10050", "2102.08981", "2111.11431", "2210.08402"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_659"} +{"question": "Could you provide me some studies that have applied the concept of teacher-student network?", "answer": ["FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Distilling the Knowledge in a Neural Network", "Knowledge Distillation: A Survey", "Semi-supervised semantic segmentation needs strong, varied perturbations", "Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning", "PseudoSeg: Designing Pseudo Labels for Semantic Segmentation", "Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic\n Segmentation", "Unbiased Teacher for Semi-Supervised Object Detection", "Humble Teachers Teach Better Students for Semi-Supervised Object\n Detection", "Distilling Vision-Language Pre-training to Collaborate with\n Weakly-Supervised Temporal Action Localization", "End-to-End Semi-Supervised Object Detection with Soft Teacher"], "answer_arxiv_id": ["2001.07685v2", "1503.02531", "2006.05525", "1906.01916", "2110.05474", "2010.09713", "2208.09910", "2102.09480", "2106.10456", "2212.09335", "2106.09018"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_test_664"} +{"question": "Any existing research on generating the 3D human avatars with predefined parametric human templates?", "answer": ["AvatarGen: A 3D Generative Model for Animatable Human Avatars", "EVA3D: Compositional 3D Human Generation from 2D Image Collections", "Unsupervised Learning of Efficient Geometry-Aware Neural Articulated\n Representations", "Generative Neural Articulated Radiance Fields", "3D-Aware Semantic-Guided Generative Model for Human Synthesis"], "answer_arxiv_id": ["2211.14589", "2210.04888", "2204.08839", "2206.14314", "2112.01422"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_test_674"} +{"question": "Which papers propose methods for audio-visual segmentation task?", "answer": ["Class-aware Sounding Objects Localization via Audiovisual Correspondence", "Discriminative Sounding Objects Localization via Self-supervised\n Audiovisual Matching", "Deep Multimodal Clustering for Unsupervised Audiovisual Learning", "Unsupervised Sound Localization via Iterative Contrastive Learning", "Localizing Visual Sounds the Hard Way", "Self-Supervised Predictive Learning: A Negative-Free Method for Sound Source Localization in Visual Scenes", "Exploiting Transformation Invariance and Equivariance for\n Self-supervised Sound Localisation", "Learning to Localize Sound Source in Visual Scenes", "Multiple Sound Sources Localization from Coarse to Fine", "Annotation-free Audio-Visual Segmentation"], "answer_arxiv_id": ["2112.11749", "2010.05466", "1807.03094", "2104.00315", "2104.02691", "2203.13412v1", "2206.12772", "1803.03849", "2007.06355", "2305.11019"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_test_680"} +{"question": "Which studies focus on table-based EHR question answering?", "answer": ["Question Answering for Complex Electronic Health Records Database using Unified Encoder-Decoder Architecture", "LeafAI: query generator for clinical cohort discovery rivaling a human programmer", "EHRSQL: A Practical Text-to-SQL Benchmark for Electronic Health Records", "Towards Understanding the Generalization of Medical Text-to-SQL Models and Datasets", "Text-to-SQL Generation for Question Answering on Electronic Medical Records"], "answer_arxiv_id": ["2111.14703", "2304.06203v2", "2301.07695", "2303.12898", "1908.01839"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_test_683"} +{"question": "Could you mention the studies that focused on lifting 2D pre-trained models to create 3D models from textual prompts?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "ATT3D: Amortized Text-to-3D Object Synthesis", "ZeroAvatar: Zero-shot 3D Avatar Generation from a Single Image", "MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion", "Sparse3D: Distilling Multiview-Consistent Diffusion for Object\n Reconstruction from Sparse Views", "MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2211.10440", "2303.13873", "2305.16213", "2212.00774v1", "2306.07349", "2305.16411", "2307.01097", "2308.14078", "2308.16512"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_test_687"} +{"question": "Which papers have discussed Bound Propagation methods and analyzed the output bounds based on input bounds?", "answer": ["Semidefinite relaxations for certifying robustness to adversarial examples", "Provable Defenses against Adversarial Examples via the Convex Outer Adversarial Polytope", "Efficient Neural Network Robustness Certification with General Activation Functions", "Certifiable Robustness and Robust Training for Graph Convolutional Networks", "Collective Robustness Certificates: Exploiting Interdependence in Graph Neural Networks"], "answer_arxiv_id": ["1811.01057", "1711.00851", "1811.00866", "1906.12269", "2302.02829"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_test_693"} +{"question": "Which papers introduced algorithms regarding distributed optimization in a full participation setting using deterministic methods?", "answer": ["Communication Efficient Distributed Optimization using an Approximate Newton-type Method", "AIDE: Fast and Communication Efficient Distributed Optimization", "On Convergence of Distributed Approximate Newton Methods: Globalization, Sharper Bounds and Beyond", "Statistically Preconditioned Accelerated Gradient Method for Distributed Optimization", "An Accelerated Second-Order Method for Distributed Stochastic Optimization", "Newton Method over Networks is Fast up to the Statistical Precision"], "answer_arxiv_id": ["1312.7853", "1608.06879v1", "1908.02246", "2002.10726", "2103.14392", "2102.06780"], "source_meta": {"published_time": "20230415"}, "qid": "AutoScholarQuery_test_705"} +{"question": "What papers recently gave attention to maximum entropy policies in the context of reinforcement learning (RL)?", "answer": ["Behavior From the Void: Unsupervised Active Pre-Training", "APS: Active Pretraining with Successor Features", "Reinforcement Learning with Prototypical Representations", "State Entropy Maximization with Random Encoders for Efficient Exploration", "Provably Efficient Maximum Entropy Exploration", "Task-Agnostic Exploration via Policy Gradient of a Non-Parametric State Entropy Estimate"], "answer_arxiv_id": ["2103.04551", "2108.13956", "2102.11271v2", "2102.09430", "1812.02690", "2007.04640"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_test_716"} +{"question": "Which papers discuss that Large Language Models (LLMs) memorize data both from their original large training corpora and smaller private datasets used for downstream tasks?", "answer": ["Quantifying Memorization Across Neural Language Models", "Preventing Verbatim Memorization in Language Models Gives a False Sense of Privacy", "How BPE Affects Memorization in Transformers", "Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Models", "Counterfactual Memorization in Neural Language Models", "Memorization in NLP Fine-tuning Methods"], "answer_arxiv_id": ["2202.07646", "2210.17546v3", "2110.02782", "2205.10770", "2112.12938", "2205.12506"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_test_718"} +{"question": "Which papers implemented neural networks like CNNs and RNNs to enhance co-embedding methods?", "answer": ["A ConvNet for the 2020s", "Deep Residual Learning for Image Recognition", "Going Deeper with Convolutions", "Very Deep Convolutional Networks for Large-Scale Image Recognition", "Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term\n Memory (LSTM) Network"], "answer_arxiv_id": ["2201.03545", "1512.03385", "1409.4842", "1409.1556", "1808.03314"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_test_732"} +{"question": "Can you provide references regarding data-driven approaches for stereo-matching?", "answer": ["A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation", "Unifying Flow, Stereo and Depth Estimation", "Pyramid Stereo Matching Network", "GA-Net: Guided Aggregation Net for End-to-end Stereo Matching", "A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation"], "answer_arxiv_id": ["1512.02134", "2211.05783", "1803.08669", "1904.06587", "1512.02134"], "source_meta": {"published_time": "20240421"}, "qid": "AutoScholarQuery_test_752"} +{"question": "Which papers discuss the application of specific criteria to remove weights in post-hoc pruning?", "answer": ["Dynamic Network Surgery for Efficient DNNs", "Learning to Prune Deep Neural Networks via Layer-wise Optimal Brain Surgeon", "Compressing Neural Networks using the Variational Information Bottleneck", "NISP: Pruning Networks using Neuron Importance Score Propagation", "Importance Estimation for Neural Network Pruning"], "answer_arxiv_id": ["1608.04493", "1705.07565", "1802.10399", "1711.05908", "1906.10771"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_test_755"} +{"question": "Can you identify any works that aimed to improve computationally efficient FL with personalized local models using quantization and model parameter decoupling?", "answer": ["QuPeD: Quantized Personalization via Distillation with Applications to Federated Learning", "Efficient Split-Mix Federated Learning for On-Demand and In-Situ Customization", "HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients", "Exploiting Shared Representations for Personalized Federated Learning", "Achieving Personalized Federated Learning with Sparse Local Models"], "answer_arxiv_id": ["2107.13892", "2203.09747", "2010.01264", "2102.07078", "2201.11380"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_test_756"} +{"question": "Can you list any studies that utilize differentiable logical rule learning", "answer": ["Embedding Entities and Relations for Learning and Inference in Knowledge Bases", "Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "DRUM: End-To-End Differentiable Rule Mining On Knowledge Graphs", "DeepPath: A Reinforcement Learning Method for Knowledge Graph Reasoning", "Variational Knowledge Graph Reasoning", "Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning", "Multi-Hop Knowledge Graph Reasoning with Reward Shaping", "M-Walk: Learning to Walk over Graphs using Monte Carlo Tree Search"], "answer_arxiv_id": ["1412.6575", "1702.08367", "1911.00055", "1707.06690", "1803.06581", "1711.05851", "1808.10568", "1802.04394"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_test_760"} +{"question": "What studies provide insight into provably efficient exploration techniques in RL?", "answer": ["Model-based Reinforcement Learning and the Eluder Dimension", "Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning", "Learning Near Optimal Policies with Low Inherent Bellman Error", "Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1406.1853", "2004.10019", "2106.04895", "2003.00153", "1907.05388"], "source_meta": {"published_time": "20220405"}, "qid": "AutoScholarQuery_test_761"} +{"question": "Any works that developed representations of statistical and causal dependencies between latent factors and auxiliary variables?", "answer": ["Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICA", "Weakly-Supervised Disentanglement Without Compromises", "Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style", "The Incomplete Rosetta Stone Problem: Identifiability Results for Multi-View Nonlinear ICA", "Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learning", "Variational Autoencoders and Nonlinear ICA: A Unifying Framework", "ICE-BeeM: Identifiable Conditional Energy-Based Deep Models Based on Nonlinear ICA", "Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding", "Contrastive Learning Inverts the Data Generating Process"], "answer_arxiv_id": ["1605.06336", "2002.02886", "2106.04619v4", "1905.06642", "1805.08651", "1907.04809", "2002.11537", "2007.10930", "2102.08850v4"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_772"} +{"question": "What studies have demonstrated the effectiveness of contrastive methods in learning useful representations for downstream tasks?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Representation Learning with Contrastive Predictive Coding", "Learning deep representations by mutual information estimation and maximization", "Learning Representations by Maximizing Mutual Information Across Views", "Contrastive Multiview Coding", "On Mutual Information Maximization for Representation Learning", "What Makes for Good Views for Contrastive Learning?", "Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere", "Representation Learning with Contrastive Predictive Coding", "Contrastive Learning Inverts the Data Generating Process", "Representation Learning with Contrastive Predictive Coding", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Estimating divergence functionals and the likelihood ratio by convex risk minimization"], "answer_arxiv_id": ["1807.03748", "2002.05709", "1911.05722", "2106.04156", "1807.03748", "1808.06670", "1906.00910", "1906.05849", "1907.13625", "2005.10243", "2005.10242", "1807.03748", "2102.08850v4", "1807.03748", "2106.04156", "0809.0853"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_test_773"} +{"question": "In what papers were methods described that locate and edit the parameters and neurons in the LLMs in light of specific knowledge?", "answer": ["Locating and Editing Factual Associations in GPT", "Knowledge Neurons in Pretrained Transformers", "Mass-Editing Memory in a Transformer", "Editing a classifier by rewriting its prediction rules", "Transformer Feed-Forward Layers Build Predictions by Promoting Concepts\n in the Vocabulary Space"], "answer_arxiv_id": ["2202.05262", "2104.08696", "2210.07229", "2112.01008", "2203.14680"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_test_786"} +{"question": "Which works were pertinent in the development of the Large Multimodal Models?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "PaLM: Scaling Language Modeling with Pathways", "UL2: Unifying Language Learning Paradigms", "Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "CyCLIP: Cyclic Contrastive Language-Image Pretraining"], "answer_arxiv_id": ["1910.10683", "2204.02311", "2205.05131", "2103.00020", "2201.12086", "2205.14459"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_test_800"} +{"question": "What studies discuss the training of Pre-trained Language Models(PLMs) for predicting masked words?", "answer": ["Mask More and Mask Later: Efficient Pre-training of Masked Language Models by Disentangling the [MASK] Token", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension", "MASS: Masked Sequence to Sequence Pre-training for Language Generation", "Unsupervised Cross-lingual Representation Learning at Scale", "Cross-lingual Language Model Pretraining"], "answer_arxiv_id": ["2211.04898", "1810.04805", "1907.11692", "1910.13461", "1905.02450", "1911.02116", "1901.07291"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_test_818"} +{"question": "Which papers have investigated the simplicity bias in Deep Neural Networks (DNNs)?", "answer": ["SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data", "The Implicit Bias of Gradient Descent on Separable Data", "Implicit Bias of Gradient Descent on Linear Convolutional Networks", "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness", "The Origins and Prevalence of Texture Bias in Convolutional Neural Networks"], "answer_arxiv_id": ["1710.10174", "1710.10345", "1806.00468", "1811.12231", "1911.09071"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_test_826"} +{"question": "Could you provide me some studies about delayed sampling which uses automatic marginalization to improve inference?", "answer": ["Delayed Sampling and Automatic Rao-Blackwellization of Probabilistic Programs", "Automated learning with a probabilistic programming language: Birch", "Pyro: Deep Universal Probabilistic Programming", "Functional Tensors for Probabilistic Programming", "Tensor Variable Elimination for Plated Factor Graphs", "Reactive Probabilistic Programming", "Semi-Symbolic Inference for Efficient Streaming Probabilistic Programming"], "answer_arxiv_id": ["1708.07787v2", "1810.01539", "1810.09538", "1910.10775v2", "1902.03210", "1908.07563v2", "2209.07490v2"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_test_835"} +{"question": "Which studies are about leveraging demonstrations into the policy-update steps of Reinforcement Learning?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations", "Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewards", "Overcoming Exploration in Reinforcement Learning with Demonstrations", "Reinforced Imitation: Sample Efficient Deep Reinforcement Learning for Map-less Navigation by Leveraging Prior Demonstrations", "Integrating Behavior Cloning and Reinforcement Learning for Improved Performance in Dense and Sparse Reward Environments"], "answer_arxiv_id": ["1709.10087", "1707.08817", "1709.10089", "1805.07095", "1910.04281"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_test_838"} +{"question": "What research studies use hard pseudolabels from teachers to train student models in the outcontext of low-resource semi-supervised sequence generation?", "answer": ["Sequence-Level Knowledge Distillation", "Is GPT-3 a Good Data Annotator?", "GPT3Mix: Leveraging Large-scale Language Models for Text Augmentation", "Want To Reduce Labeling Cost? GPT-3 Can Help", "ChatGPT Outperforms Crowd-Workers for Text-Annotation Tasks", "Large Language Models Are Reasoning Teachers"], "answer_arxiv_id": ["1606.07947", "2212.10450", "2104.08826", "2108.13487", "2303.15056", "2212.10071"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_test_851"} +{"question": "Which works use Graph Neural Networks and Recurrent Neural Networks to update encodings in temporal graph learning?", "answer": ["Structured Sequence Modeling with Graph Convolutional Recurrent Networks", "T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction", "ROLAND: Graph Learning Framework for Dynamic Graphs", "CS-TGN: Community Search via Temporal Graph Neural Networks", "Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks", "Anomaly Detection in Multiplex Dynamic Networks: from Blockchain Security to Brain Disease Prediction"], "answer_arxiv_id": ["1612.07659", "1811.05320", "2208.07239", "2303.08964", "1908.01207", "2211.08378"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_test_872"} +{"question": "Which papers solved classification and detection problems in LiDAR perception using deep learning?", "answer": ["Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline", "Benchmarking and Analyzing Point Cloud Classification under Corruptions", "PointCLIP: Point Cloud Understanding by CLIP", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detection", "Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving"], "answer_arxiv_id": ["2106.05304", "2202.03377", "2112.02413", "1812.05784", "2102.00463", "1906.06310"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_test_873"} +{"question": "Which studies focused on bottom-up methods in instance segmentation in 3D perception?", "answer": ["OccuSeg: Occupancy-aware 3D Instance Segmentation", "Hierarchical Aggregation for 3D Instance Segmentation", "3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans", "Language-Grounded Indoor 3D Semantic Segmentation in the Wild", "Instance Segmentation in 3D Scenes using Semantic Superpoint Tree\n Networks"], "answer_arxiv_id": ["2003.06537v3", "2108.02350", "1812.07003", "2204.07761", "2108.07478"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_test_884"} +{"question": "What studies provide solutions for feature matching in low-textured regions using dense or semi-dense matching methods?", "answer": ["Learning Accurate Dense Correspondences and When to Trust Them", "Neighbourhood Consensus Networks", "Dual-Resolution Correspondence Networks", "LoFTR: Detector-Free Local Feature Matching with Transformers", "Quadtree Attention for Vision Transformers", "ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformer", "MatchFormer: Interleaving Attention in Transformers for Feature Matching"], "answer_arxiv_id": ["2101.01710", "1810.10510", "2006.08844", "2104.00680", "2201.02767", "2208.14201", "2203.09645"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_test_887"} +{"question": "Which studies focused on using synthetic data to create new datasets or augment existing ones?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "Playing for Data: Ground Truth from Computer Games", "VisDA: The Visual Domain Adaptation Challenge", "CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning", "Structured3D: A Large Photo-realistic Dataset for Structured 3D Modeling", "ThreeDWorld: A Platform for Interactive Multi-Modal Physical Simulation"], "answer_arxiv_id": ["1504.06852", "1608.02192v1", "1710.06924", "1612.06890", "1908.00222", "2007.04954"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_test_893"} +{"question": "Which works employed contrastive learning for graph representation learning?", "answer": ["Deep Graph Infomax", "InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization", "Graph Contrastive Learning with Augmentations", "Deep Graph Contrastive Representation Learning", "GCC: Graph Contrastive Coding for Graph Neural Network Pre-Training"], "answer_arxiv_id": ["1809.10341", "1908.01000", "2010.13902", "2006.04131", "2006.09963"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_test_894"} +{"question": "Which works offer end-to-end methods for multimodal Language Models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BEiT: BERT Pre-Training of Image Transformers", "Image as a Foreign Language: BEiT Pretraining for All Vision and\n Vision-Language Tasks", "Visual Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "PaLI: A Jointly-Scaled Multilingual Language-Image Model"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2305.06500", "2201.12086", "2106.08254", "2208.10442", "2304.08485", "2304.14178", "2304.10592", "2303.16199", "2305.03726", "2202.03052", "2209.06794"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_test_899"} +{"question": "Can you name some examples of projects that integrated machine learning, particularly LLMs, into automated theorem proving?", "answer": ["Learning to Reason in Large Theories without Imitation", "Constructions in combinatorics via neural networks", "LeanDojo: Theorem Proving with Retrieval-Augmented Language Models", "Generative Language Modeling for Automated Theorem Proving", "Proof Artifact Co-training for Theorem Proving with Language Models", "NaturalProofs: Mathematical Theorem Proving in Natural Language", "Draft, Sketch, and Prove: Guiding Formal Theorem Provers with Informal\n Proofs"], "answer_arxiv_id": ["1905.10501", "2104.14516", "2306.15626", "2009.03393", "2102.06203", "2104.01112", "2210.12283"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_test_902"} +{"question": "What studies have used techniques like residual structure, skip connection, and dropout in basic CNN frameworks for image restoration?", "answer": ["Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Plug-and-Play Image Restoration with Deep Denoiser Prior", "Image Super-Resolution Using Very Deep Residual Channel Attention Networks", "Residual Dense Network for Image Restoration", "Reflash Dropout in Image Super-Resolution"], "answer_arxiv_id": ["1511.04587", "2008.13751", "1807.02758", "1812.10477", "2112.12089"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_test_904"} +{"question": "Which research papers adopted a 3D-Unet architecture to produce video volumes directly from an input image?", "answer": ["Stochastic Adversarial Video Prediction", "Animate-A-Story: Storytelling with Retrieval-Augmented Video Generation", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "Stochastic Image-to-Video Synthesis using cINNs", "MCVD: Masked Conditional Video Diffusion for Prediction, Generation, and\n Interpolation", "Diffusion Models for Video Prediction and Infilling"], "answer_arxiv_id": ["1804.01523", "2307.06940", "2307.04725", "2105.04551", "2205.09853", "2206.07696"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_test_918"} +{"question": "Which papers propose first-order methods for efficiently solving min-max optimization problems in Weak Minty Variational Inequalities?", "answer": ["The Complexity of Constrained Min-Max Optimization", "Efficient Methods for Structured Nonconvex-Nonconcave Min-Max Optimization", "Escaping limit cycles: Global convergence for constrained nonconvex-nonconcave minimax problems", "Fast Extra Gradient Methods for Smooth Structured Nonconvex-Nonconcave Minimax Problems", "Solving Nonconvex-Nonconcave Min-Max Problems exhibiting Weak Minty Solutions", "Solving stochastic weak Minty variational inequalities without increasing batch size"], "answer_arxiv_id": ["2009.09623", "2011.00364", "2302.09831", "2106.02326", "2201.12247", "2302.09029"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_test_926"} +{"question": "What studies advanced diffusion probabilistic models to generate high-resolution and diverse images?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2112.10741", "2112.10752", "2205.11487"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_test_945"} +{"question": "Which studies used transformers and diffusion models for creating high-fidelity images from text?", "answer": ["VQGAN-CLIP: Open Domain Image Generation and Editing with Natural\n Language Guidance", "CogView2: Faster and Better Text-to-Image Generation via Hierarchical\n Transformers", "Denoising Diffusion Probabilistic Models", "Blended Diffusion for Text-driven Editing of Natural Images", "Text2LIVE: Text-Driven Layered Image and Video Editing", "Prompt-to-Prompt Image Editing with Cross Attention Control", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "More Control for Free! Image Synthesis with Semantic Diffusion Guidance", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2204.08583", "2204.14217", "2006.11239", "2111.14818", "2204.02491", "2208.01626", "2110.02711", "2112.05744", "2112.10741"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_test_950"} +{"question": "Could you provide some works that discussed the problem of selling information in economics and computer science?", "answer": ["Optimal Mechanisms for Selling Information", "Selling Information Through Consulting", "How to Sell Information Optimally: an Algorithmic Study", "Optimal Pricing of Information", "Is Selling Complete Information (Approximately) Optimal?", "Optimal Advertising for Information Products", "Selling Data to an Agent with Endogenous Information"], "answer_arxiv_id": ["1204.5519", "1907.04397v3", "2011.14570", "2102.13289", "2202.09013", "2002.10045v5", "2103.05788v4"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_test_953"} +{"question": "Could you cite the works where multilingual LLMs were evaluated on individual tasks such as Translation, Question-Answering, Summarization, and Reasoning?", "answer": ["On the Cross-lingual Transferability of Monolingual Representations", "TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages", "XOR QA: Cross-lingual Open-Retrieval Question Answering", "XL-Sum: Large-Scale Multilingual Abstractive Summarization for 44\n Languages", "CrossSum: Beyond English-Centric Cross-Lingual Summarization for 1,500+\n Language Pairs", "Language Models are Multilingual Chain-of-Thought Reasoners", "XCOPA: A Multilingual Dataset for Causal Commonsense Reasoning"], "answer_arxiv_id": ["1910.11856", "2003.05002v1", "2010.11856", "2106.13822", "2112.08804", "2210.03057", "2005.00333"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_test_978"} +{"question": "Could you provide references that discuss alternative approaches to the matrix mechanism that reduce the variance by adding bias?", "answer": ["A Simple and Practical Algorithm for Differentially Private Data Release", "Leveraging Public Data for Practical Private Query Release", "Differentially Private Query Release Through Adaptive Projection", "Dual Query: Practical Private Query Release for High Dimensional Data", "AIM: An Adaptive and Iterative Mechanism for Differentially Private Synthetic Data", "Iterative Methods for Private Synthetic Data: Unifying Framework and New Methods", "New Oracle-Efficient Algorithms for Private Synthetic Data Release", "PrivSyn: Differentially Private Data Synthesis"], "answer_arxiv_id": ["1012.4763", "2102.08598v2", "2103.06641", "1402.1526", "2201.12677", "2106.07153", "2007.05453", "2012.15128v1"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_test_991"} +{"question": "Which works describe the advancements in 3D reconstruction and novel view synthesis with NeRF?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo\n Collections", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural\n Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2008.02268", "2103.13415", "2103.15595", "2307.11335"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_test_993"} diff --git a/recipe/paper_search/inference/datasets/AutoScholarQuery/train.jsonl b/recipe/paper_search/inference/datasets/AutoScholarQuery/train.jsonl new file mode 100644 index 0000000..336fbd2 --- /dev/null +++ b/recipe/paper_search/inference/datasets/AutoScholarQuery/train.jsonl @@ -0,0 +1,33551 @@ +{"question": "What works are related to the field of image retrieval?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning"], "answer_arxiv_id": ["1909.11740"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_0"} +{"question": "Could you provide me some works employs image patches and superpixels in region-based methods for semantic segmentation?", "answer": ["CEREALS – Cost-Effective REgion-based Active Learning for Semantic Segmentation", "Reinforced Active Learning for Image Segmentation", "MetaBox+: A new Region Based Active Learning Method for Semantic Segmentation using Priority Maps", "ViewAL: Active Learning With Viewpoint Entropy for Semantic Segmentation"], "answer_arxiv_id": ["1810.09726", "2002.06583", "2010.01884", "1911.11789"], "source_meta": {"published_time": "20230917"}, "qid": "AutoScholarQuery_train_1"} +{"question": "Could you provide me some studies that proposed hierarchical neural models to capture spatio-temporal features in sign videos?", "answer": ["TSPNet: Hierarchical Feature Learning via Temporal Semantic Pyramid for Sign Language Translation", "Sign Language Translation with Hierarchical Spatio-Temporal Graph Neural Network"], "answer_arxiv_id": ["2010.05468", "2111.07258"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_2"} +{"question": "Which works are focused on online unsupervised skill discovery for hierarchical RL?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function", "Dynamics-Aware Unsupervised Discovery of Skills", "Learning Latent Plans from Play", "Efficient Exploration via State Marginal Matching"], "answer_arxiv_id": ["1802.06070", "1907.01657", "1903.01973", "1906.05274"], "source_meta": {"published_time": "20220819"}, "qid": "AutoScholarQuery_train_3"} +{"question": "Could you give me examples of research that developed datasets for molecular force field prediction?", "answer": ["Machine Learning of Accurate Energy-Conserving Molecular Force Fields", "Accurate global machine learning force fields for molecules with hundreds of atoms"], "answer_arxiv_id": ["1611.04678", "2209.14865"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_4"} +{"question": "Which papers propose methods to accelerate the generation process in diffusion models?", "answer": ["Progressive Distillation for Fast Sampling of Diffusion Models", "Score-Based Generative Modeling with Critically-Damped Langevin\n Diffusion", "Tackling the Generative Learning Trilemma with Denoising Diffusion GANs", "Denoising Diffusion Implicit Models", "Consistency Models"], "answer_arxiv_id": ["2202.00512", "2112.07068", "2112.07804", "2010.02502", "2303.01469"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_5"} +{"question": "Can you name some studies that proposed solutions to the challenges of satellite imagery?", "answer": ["Foreground-Aware Relation Network for Geospatial Object Segmentation in\n High Spatial Resolution Remote Sensing Imagery", "PointFlow: Flowing Semantics Through Points for Aerial Image\n Segmentation"], "answer_arxiv_id": ["2011.09766", "2103.06564"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_6"} +{"question": "Which follow-up works bake the resulting surface geometry into a mesh that is further optimized and simplified?", "answer": ["BakedSDF: Meshing Neural SDFs for Real-Time View Synthesis", "VMesh: Hybrid Volume-Mesh Representation for Efficient View Synthesis"], "answer_arxiv_id": ["2302.14859", "2303.16184"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_7"} +{"question": "Which studies have focused on nonstationary RL using value-based methods, specifically Upper Confidence Bound (UCB) based algorithms?", "answer": ["Reinforcement Learning for Non-Stationary Markov Decision Processes: The Blessing of (More) Optimism", "Efficient Learning in Non-Stationary Linear Markov Decision Processes", "Nonstationary Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["2006.14389", "2010.12870", "2010.04244v3"], "source_meta": {"published_time": "20230810"}, "qid": "AutoScholarQuery_train_8"} +{"question": "What works demonstrated the advancements in T2I synthesis brought by autoregressive transformers?", "answer": ["Taming Transformers for High-Resolution Image Synthesis", "Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["2012.09841", "2102.12092"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_9"} +{"question": "Can you provide references that conducted experiments on both indoor and outdoor datasets for 3D object detection?", "answer": ["EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection", "EPNet++: Cascade Bi-directional Fusion for Multi-Modal 3D Object Detection", "Multimodal Token Fusion for Vision Transformers", "ImVoxelNet: Image to Voxels Projection for Monocular and Multi-View General-Purpose 3D Object Detection"], "answer_arxiv_id": ["2007.08856", "2112.11088", "2204.08721", "2106.01178"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_10"} +{"question": "Which paper explained the success of SAM via using a PAC-Bayes generalization bound?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization"], "answer_arxiv_id": ["2010.01412"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_11"} +{"question": "What works developed auto-regressive and diffusion models for text-to-image (T2I) generation?", "answer": ["CogView: Mastering Text-to-Image Generation via Transformers", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2105.13290", "2112.10741", "2111.14822", "2205.11487"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_12"} +{"question": "What works about semi-supervised learning (SSL) pertain to consistency regularization methods?", "answer": ["Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning", "Unsupervised Data Augmentation for Consistency Training"], "answer_arxiv_id": ["1606.04586", "1904.12848"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_13"} +{"question": "Any works using depthwise convolution with LKs to enhance model efficiency?", "answer": ["A ConvNet for the 2020s"], "answer_arxiv_id": ["2201.03545"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_14"} +{"question": "Which studies use point cloud representations for high-resolution geometries modelling?", "answer": ["A Point Set Generation Network for 3D Object Reconstruction from a Single Image", "Learning Representations and Generative Models for 3D Point Clouds"], "answer_arxiv_id": ["1612.00603", "1707.02392"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_15"} +{"question": "Any studies on application of these methods in drug discovery and healthcare?", "answer": ["Biological Sequence Design with GFlowNets"], "answer_arxiv_id": ["2203.04115"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_16"} +{"question": "Any research on adaptive optimizers helping with potential discrepancies in network growing techniques?", "answer": ["Adam: A Method for Stochastic Optimization", "Domain-independent Dominance of Adaptive Methods", "Large Batch Optimization for Deep Learning: Training BERT in 76 minutes"], "answer_arxiv_id": ["1412.6980v9", "1912.01823", "1904.00962"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_17"} +{"question": "Which research work proposed a framework that applies the masked prediction idea for either speech, NLP, or CV?", "answer": ["data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language"], "answer_arxiv_id": ["2202.03555"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_18"} +{"question": "Which works proposed the idea of sharpness-aware minimization (SAM)?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization"], "answer_arxiv_id": ["2010.01412"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19"} +{"question": "Which papers are talking about structure-based methods and regularization-based methods for Class-Incremental Learning problem?", "answer": ["Lifelong Learning with Dynamically Expandable Networks", "Overcoming catastrophic forgetting in neural networks", "Memory Aware Synapses: Learning what (not) to forget"], "answer_arxiv_id": ["1708.01547", "1612.00796", "1711.09601"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_20"} +{"question": "Can you provide any studies that focus on extending the context of language models without training?", "answer": ["Parallel Context Windows for Large Language Models"], "answer_arxiv_id": ["2212.10947"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_21"} +{"question": "What papers develop techniques for strengthening LP-based relaxations in conservative methods for robustness certification?", "answer": ["Evaluating the Robustness of Neural Networks: An Extreme Value Theory Approach", "Towards Fast Computation of Certified Robustness for ReLU Networks", "Efficient Neural Network Robustness Certification with General Activation Functions"], "answer_arxiv_id": ["1801.10578", "1804.09699", "1811.00866"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_22"} +{"question": "What studies developed text-only-training zero-shot IC methods by mapping the visual feature to the text feature?", "answer": ["DeCap: Decoding CLIP Latents for Zero-Shot Captioning via Text-Only\n Training", "Text-Only Training for Image Captioning using Noise-Injected CLIP", "Transferable Decoding with Visual Entities for Zero-Shot Image\n Captioning"], "answer_arxiv_id": ["2303.03032", "2211.00575", "2307.16525"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_23"} +{"question": "Which papers focus on the development of comprehensive theories for equivariant 3D learning with respect to rigid transformations?", "answer": ["On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "A General Theory of Equivariant CNNs on Homogeneous Spaces", "Unified Fourier-based Kernel and Nonlinearity Design for Equivariant Networks on Homogeneous Spaces"], "answer_arxiv_id": ["1802.03690", "1811.02017", "2206.08362"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_24"} +{"question": "What are some of the papers that talk about the method of prompt tuning as a Parameter Efficient FineTuning(PEFT) method?", "answer": ["Learning How to Ask: Querying LMs with Mixtures of Soft Prompts", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2104.06599", "2104.08691", "2101.00190"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_25"} +{"question": "Could you provide me some studies on referring expression comprehension?", "answer": ["Modeling Context in Referring Expressions", "Generation and Comprehension of Unambiguous Object Descriptions", "PhraseCut: Language-based Image Segmentation in the Wild"], "answer_arxiv_id": ["1608.00272", "1511.02283", "2008.01187"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_26"} +{"question": "Is there any research using MAML for meta-learning on a distribution of tasks?", "answer": ["Universal linguistic inductive biases via meta-learning", "Modeling rapid language learning by distilling Bayesian priors into\n artificial neural networks", "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["2006.16324", "2305.14701", "1703.03400"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_27"} +{"question": "Which studies use an RM for on-policy reinforcement learning?", "answer": ["Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["1909.08593"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_28"} +{"question": "Any studies employing attention-based propagation operator with a tanh/cosine activation function in GNNs?", "answer": ["Two Sides of the Same Coin: Heterophily and Oversmoothing in Graph Convolutional Neural Networks"], "answer_arxiv_id": ["2102.06462"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_29"} +{"question": "Any works about studying biological cells, bacteria, tissue types, and material structures through microscopic image classification?", "answer": ["Methods for Segmentation and Classification of Digital Microscopy Tissue Images"], "answer_arxiv_id": ["1810.13230v2"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_30"} +{"question": "Can you list the papers that studied finetuning the global model to generate personalized models?", "answer": ["Think Locally, Act Globally: Federated Learning with Local and Global Representations", "FedBABU: Toward Enhanced Representation for Federated Image Classification", "Exploiting Shared Representations for Personalized Federated Learning"], "answer_arxiv_id": ["2001.01523", "2106.06042", "2102.07078"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_31"} +{"question": "Which researches indicate that inner-loops in decoupled optimization procedures can be potentially slow and need a block-barrier instruction?", "answer": ["An Optimal Algorithm for Decentralized Finite Sum Optimization"], "answer_arxiv_id": ["2005.10675"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_32"} +{"question": "What study introduced a fully automatic dataset generation tool to promote the use of sewing patterns in deep learning?", "answer": ["Generating Datasets of 3D Garments with Sewing Patterns"], "answer_arxiv_id": ["2109.05633"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_33"} +{"question": "What works focus on models that extract scalar representations from the atoms’ positions?", "answer": ["Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties", "SchNet – a deep learning architecture for molecules and materials", "Directional Message Passing for Molecular Graphs", "Spherical Message Passing for 3D Molecular Graphs", "GemNet: Universal Directional Graph Neural Networks for Molecules", "GemNet-OC: Developing Graph Neural Networks for Large and Diverse Molecular Simulation Datasets"], "answer_arxiv_id": ["1710.10324", "1712.06113", "2003.03123", "2102.05013", "2106.08903v10", "2204.02782"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_34"} +{"question": "What papers discuss the parametric explanation method of Interpretability and feature selection in GNNs?", "answer": ["GNNExplainer: Generating Explanations for Graph Neural Networks", "Parameterized Explainer for Graph Neural Network", "XGNN: Towards Model-Level Explanations of Graph Neural Networks", "PGM-Explainer: Probabilistic Graphical Model Explanations for Graph Neural Networks"], "answer_arxiv_id": ["1903.03894", "2011.04573", "2006.02587", "2010.05788"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_35"} +{"question": "Which works proposed adversarial approaches for prompt learning to mitigate biases in pre-trained VLMs?", "answer": ["A Prompt Array Keeps the Bias Away: Debiasing Vision-Language Models\n with Adversarial Learning"], "answer_arxiv_id": ["2203.11933"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_36"} +{"question": "Which research looks into the similarity of music and linguistic structures?", "answer": ["Learning Music Helps You Read: Using Transfer to Study Linguistic\n Structure in Language Models"], "answer_arxiv_id": ["2004.14601"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_37"} +{"question": "Which papers are about utilizing optical-flow for video enhancement to handle the temporal consistency problem?", "answer": ["Video Enhancement with Task-Oriented Flow", "FlowNet: Learning Optical Flow with Convolutional Networks", "DVDnet: A Fast Network for Deep Video Denoising", "Optical Flow Estimation using a Spatial Pyramid Network", "FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks", "Recurrent Back-Projection Network for Video Super-Resolution"], "answer_arxiv_id": ["1711.09078", "1504.06852", "1906.11890", "1611.00850", "1612.01925", "1903.10128"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_38"} +{"question": "Which works looked at the relationship between dynamical systems and machine learning?", "answer": ["PDE-Net: Learning PDEs from Data", "Learning Sparse Dynamical Systems from a Single Sample Trajectory"], "answer_arxiv_id": ["1710.09668", "1904.09396"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_39"} +{"question": "Any works about investigating the phenomenon of the singularity in the conditional score as noise vanishes in finite dimensions?", "answer": ["Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation", "Score-Based Generative Modeling with Critically-Damped Langevin Diffusion"], "answer_arxiv_id": ["2106.05527", "2112.07068"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_40"} +{"question": "Any works about LLMs easily being hypnotized to generate harmful content?", "answer": ["DeepInception: Hypnotize Large Language Model to Be Jailbreaker"], "answer_arxiv_id": ["2311.03191"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_41"} +{"question": "Could you provide me some works that extended TD to control setting?", "answer": ["Finite-Sample Analysis for SARSA with Linear Function Approximation"], "answer_arxiv_id": ["1902.02234"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_42"} +{"question": "Can you list some studies that learn a time-varying deformation of 3D points into a static canonical scene?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Video", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2011.12948", "2012.12247", "2106.13228v2"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_43"} +{"question": "What are some works that implemented the concept of maximizing feature similarity in self-supervised learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Exploring Simple Siamese Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations", "Supervised Contrastive Learning"], "answer_arxiv_id": ["1911.05722", "2011.10566", "2006.07733", "2002.05709", "2104.14548", "2004.11362"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_44"} +{"question": "What papers have focused on machine learning approaches to variable selection in MILP solvers?", "answer": ["Exact Combinatorial Optimization with Graph Convolutional Neural Networks", "Hybrid Models for Learning to Branch", "Parameterizing Branch-and-Bound Search Trees to Learn Branching Policies"], "answer_arxiv_id": ["1906.01629", "2006.15212", "2002.05120"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_45"} +{"question": "What works discuss white-box ensembles where the ensemble logits are calculated by averaging the corresponding logits of the constituent classifiers?", "answer": ["On the Certified Robustness for Ensemble Models and Beyond", "Enhancing Certifiable Robustness via a Deep Model Ensemble", "Enhancing Certified Robustness via Smoothed Weighted Ensembling"], "answer_arxiv_id": ["2107.10873", "1910.14655v1", "2005.09363"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_46"} +{"question": "Which studies address the issue of reducing extrapolation error by using value pessimism about unseen actions?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Fisher Divergence Critic Regularization"], "answer_arxiv_id": ["2006.04779", "2103.08050"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_47"} +{"question": "Which works conduct implicit policy regularization using variants of importance sampling under iterative methods in offline RL?", "answer": ["OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation", "Off-Policy Policy Gradient with State Distribution Correction", "AlgaeDICE: Policy Gradient from Arbitrary Experience"], "answer_arxiv_id": ["2106.10783", "1904.08473", "1912.02074"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_48"} +{"question": "What studies showed that edge-independent models cannot reproduce the desired statistics of the target network?", "answer": ["On the Power of Edge Independent Graph Models"], "answer_arxiv_id": ["2111.00048"], "source_meta": {"published_time": "20230506"}, "qid": "AutoScholarQuery_train_49"} +{"question": "What works have been done on applying Gradient checkpointing?", "answer": ["Training Deep Nets with Sublinear Memory Cost"], "answer_arxiv_id": ["1604.06174"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_50"} +{"question": "Can you mention the work that recently improved on the dimension dependence in bandit convex optimization with memory?", "answer": ["Online Nonstochastic Model-Free Reinforcement Learning"], "answer_arxiv_id": ["2305.17552"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_51"} +{"question": "Which works have demonstrated that orthogonality can help plain Recurrent Neural Networks achieve long term memory?", "answer": ["Unitary Evolution Recurrent Neural Networks", "Full-Capacity Unitary Recurrent Neural Networks", "Orthogonal Recurrent Neural Networks with Scaled Cayley Transform", "On orthogonality and learning recurrent networks with long term dependencies", "Cheap Orthogonal Constraints in Neural Networks: A Simple Parametrization of the Orthogonal and Unitary Group"], "answer_arxiv_id": ["1511.06464", "1611.00035", "1707.09520", "1702.00071", "1901.08428"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_52"} +{"question": "What papers have studied trade-offs between learning rate and batch size in stochastic optimization methods?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "Don’t Decay the Learning Rate, Increase the Batch Size"], "answer_arxiv_id": ["1609.04836", "1711.00489"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_53"} +{"question": "Which papers discuss about the use of cross-attention and additional objective functions for image-text matching?", "answer": ["ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision", "Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "Vision-Language Pre-Training with Triple Contrastive Learning", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2102.03334", "2107.07651", "2202.10401", "2201.12086"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_54"} +{"question": "What studies fall into the category of contrastive learning in SSL approaches?", "answer": ["Learning deep representations by mutual information estimation and maximization", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1808.06670", "1807.03748"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_55"} +{"question": "Could you provide some studies that introduced text conditions to guide action generation?", "answer": ["TIPS: Text-Induced Pose Synthesis"], "answer_arxiv_id": ["2207.11718"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_56"} +{"question": "Could you provide me some studies that implement vanilla FT by updating the whole parameters when discussing backdoor fine-tuning methods?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks", "Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution"], "answer_arxiv_id": ["2002.05709", "2103.14030", "1805.12185", "2202.10054"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_57"} +{"question": "Any work that hypothesized that all neural networks of a certain architecture trained on the same dataset are linear mode connected?", "answer": ["The Role of Permutation Invariance in Linear Mode Connectivity of Neural Networks"], "answer_arxiv_id": ["2110.06296"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_58"} +{"question": "What are some works that employed the concept of sparse MoE in the field of vision?", "answer": ["Scaling Vision with Sparse Mixture of Experts", "Deep Mixture of Experts via Shallow Embedding"], "answer_arxiv_id": ["2106.05974", "1806.01531"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_59"} +{"question": "What works have established multimodal web-scale datasets and pipelines?", "answer": ["LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs", "LAION-5B: An open large-scale dataset for training next generation image-text models", "Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with Text"], "answer_arxiv_id": ["2111.02114", "2210.08402", "2304.06939"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_60"} +{"question": "What are some research papers that talk about motion-captured (mocap) datasets for co-speech animation?", "answer": ["ZeroEGGS: Zero-shot Example-based Gesture Generation from Speech", "Capture, Learning, and Synthesis of 3D Speaking Styles", "MeshTalk: 3D Face Animation from Speech using Cross-Modality\n Disentanglement", "EmoTalk: Speech-Driven Emotional Disentanglement for 3D Face Animation", "BEAT: A Large-Scale Semantic and Emotional Multi-Modal Dataset for\n Conversational Gestures Synthesis", "ARKitScenes: A Diverse Real-World Dataset For 3D Indoor Scene\n Understanding Using Mobile RGB-D Data"], "answer_arxiv_id": ["2209.07556", "1905.03079", "2104.08223", "2303.11089", "2203.05297", "2111.08897"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_61"} +{"question": "What recent research works have developed novel large language model prompting techniques?", "answer": ["Emergent Abilities of Large Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models"], "answer_arxiv_id": ["2206.07682", "2201.11903", "2205.10625"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_62"} +{"question": "What works propose designing a decoder that supports good decomposition in the context of object-centric learning?", "answer": ["Illiterate DALL-E Learns to Compose", "Object-Centric Slot Diffusion", "SlotDiffusion: Object-Centric Generative Modeling with Diffusion Models"], "answer_arxiv_id": ["2110.11405", "2303.10834", "2305.11281"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_63"} +{"question": "What works have been done on Transformer quantization specifically with 8-bit?", "answer": ["Efficient 8-Bit Quantization of Transformer Neural Machine Language Translation Model", "Fully Quantized Transformer for Machine Translation", "Q8BERT: Quantized 8Bit BERT"], "answer_arxiv_id": ["1906.00532", "1910.10485", "1910.06188"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_64"} +{"question": "Which studies propose training a diffusion model inside the lower-dimensional latent space of auto-encoder for better generation efficiency?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_65"} +{"question": "Could you give me some examples of studies proposing approaches for lower-level problem with multiple solutions?", "answer": ["A Generic First-Order Algorithmic Framework for Bi-Level Programming Beyond Lower-Level Singleton", "Improved Bilevel Model: Fast and Optimal Algorithm with Theoretical Guarantee", "A Value-Function-based Interior-point Method for Non-convex Bi-level Optimization", "BOME! Bilevel Optimization Made Easy: A Simple First-Order Approach", "On Penalty-based Bilevel Gradient Descent Method", "Towards Gradient-based Bilevel Optimization with Non-convex Followers and Beyond"], "answer_arxiv_id": ["2006.04045", "2009.00690", "2106.07991", "2209.08709", "2302.05185", "2110.00455"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_66"} +{"question": "Which studies provided a guarantee for the case of zero bias and bounded distributions in ReLU regression problem analysis?", "answer": ["Agnostic Learning of a Single Neuron with Gradient Descent"], "answer_arxiv_id": ["2005.14426"], "source_meta": {"published_time": "20220804"}, "qid": "AutoScholarQuery_train_67"} +{"question": "Which works have used diffusion models in the field of decision making as a powerful policy class?", "answer": ["Is Conditional Generative Modeling all you need for Decision-Making?", "Diffusion Policy: Visuomotor Policy Learning via Action Diffusion", "Diffusion Policies as an Expressive Policy Class for Offline\n Reinforcement Learning", "Efficient Diffusion Policies for Offline Reinforcement Learning", "Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2211.15657", "2303.04137", "2208.06193", "2305.20081", "2205.09991"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_68"} +{"question": "Can you name studies that focused on generic CAD modelling by fully reconstructing the model?", "answer": ["Point2Cyl: Reverse Engineering 3D Objects from Point Clouds to Extrusion Cylinders", "ComplexGen: CAD Reconstruction by B-Rep Chain Complex Generation"], "answer_arxiv_id": ["2112.09329v2", "2205.14573"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_69"} +{"question": "What research has utilized the planning capabilities of LLMs for task-solving?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation\n Models", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face"], "answer_arxiv_id": ["2302.04761", "2303.04671", "2303.17580"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_70"} +{"question": "Could you provide me some works about the study of Graph Neural Networks?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "Inductive Representation Learning on Large Graphs", "Representation Learning on Graphs with Jumping Knowledge Networks", "Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "How Powerful are Graph Neural Networks?", "Heterogeneous Temporal Graph Neural Network"], "answer_arxiv_id": ["1609.02907", "1710.10903", "1706.02216", "1806.03536v2", "1810.05997", "1810.00826", "2110.13889"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_71"} +{"question": "What papers tackle a diverse array of temporal localization tasks within a single framework?", "answer": ["UnLoc: A Unified Framework for Video Localization Tasks", "UniVTG: Towards Unified Video-Language Temporal Grounding"], "answer_arxiv_id": ["2308.11062", "2307.16715"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_72"} +{"question": "What research papers worked on dedicating language to human motion through text-to-motion and action-to-motion synthesis?", "answer": ["Language2Pose: Natural Language Grounded Pose Forecasting", "Synthesis of Compositional Animations from Textual Descriptions", "TEMOS: Generating diverse human motions from textual descriptions", "SINC: Spatial Composition of 3D Human Motions for Simultaneous Action\n Generation", "Action2Motion: Conditioned Generation of 3D Human Motions", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE"], "answer_arxiv_id": ["1907.01108", "2103.14675", "2204.14109", "2304.10417", "2007.15240", "2104.05670"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_73"} +{"question": "Any works carried out reinforcements using Temporal Difference learning to credit past state-action pairs for the current reward?", "answer": ["Expected Eligibility Traces"], "answer_arxiv_id": ["2007.01839"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_74"} +{"question": "Could you list out the papers that proposed recent methods relying on depth maps in novel view synthesis?", "answer": ["DINER: Depth-aware Image-based NEural Radiance fields", "Depth-supervised NeRF: Fewer Views and Faster Training for Free", "SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis"], "answer_arxiv_id": ["2211.16630", "2107.02791", "2303.16196"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_75"} +{"question": "Which papers initiated the development of diffusion models?", "answer": ["Deep Generative Stochastic Networks Trainable by Backprop", "Markov Chain Monte Carlo and Variational Inference: Bridging the Gap", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1306.1091", "1410.6460", "1503.03585"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_76"} +{"question": "What works proposed the fine-tuning methods for Large Language Models (LLMs)?", "answer": ["Extending Context Window of Large Language Models via Positional\n Interpolation", "Giraffe: Adventures in Expanding Context Lengths in LLMs", "YaRN: Efficient Context Window Extension of Large Language Models"], "answer_arxiv_id": ["2306.15595", "2308.10882", "2309.00071"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_77"} +{"question": "What works proposed fast ensembling methods that collect ensemble members on the mode-connecting-paths?", "answer": ["Snapshot Ensembles: Train 1, get M for free", "Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs"], "answer_arxiv_id": ["1704.00109", "1802.10026"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_78"} +{"question": "Could you list the works that studied gradient descent's implicit regularization in classification problems?", "answer": ["The Implicit Bias of Gradient Descent on Separable Data", "Iterative regularization in classification via hinge loss diagonal descent"], "answer_arxiv_id": ["1710.10345", "2212.12675"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_79"} +{"question": "Could you provide me some works that guide the image generation process using large scale text-image datasets and strong language understandings?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Language Models are Few-Shot Learners", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding"], "answer_arxiv_id": ["2204.06125", "2102.12092", "2205.11487", "2005.14165", "1910.10683", "1810.04805"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_80"} +{"question": "What studies are there about continually changing datasets used in the continual learning literature?", "answer": ["CORe50: a New Dataset and Benchmark for Continuous Object Recognition", "Are We Ready for Service Robots? The OpenLORIS-Scene Datasets for Lifelong SLAM", "SODA10M: A Large-Scale 2D Self/Semi-Supervised Object Detection Dataset for Autonomous Driving", "Scalability in Perception for Autonomous Driving: Waymo Open Dataset", "BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning"], "answer_arxiv_id": ["1705.03550", "1911.05603", "2106.11118", "1912.04838", "1805.04687"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_81"} +{"question": "Are there any works that dealt with the class-imbalance problem of pseudo-labeling in SSL?", "answer": ["Distribution Aligning Refinery of Pseudo-label for Imbalanced Semi-supervised Learning", "CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learning"], "answer_arxiv_id": ["2007.08844", "2102.09559"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_82"} +{"question": "Which papers proposed to identify OOD data by using the minimum distance from the class centers?", "answer": ["A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks"], "answer_arxiv_id": ["1807.03888"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_83"} +{"question": "What papers established forms of Robust PCA (RPCA) and Dual PCA (DPCA)?", "answer": ["On the Robust PCA and Weiszfeld's Algorithm", "Dual Principal Component Pursuit"], "answer_arxiv_id": ["1902.04292", "1510.04390"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_84"} +{"question": "Which papers detail the extension of backdoor attacks to other data domains or learning paradigms?", "answer": ["Hidden Backdoors in Human-Centric Language Models", "BadNL: Backdoor Attacks against NLP Models with Semantic-preserving Improvements", "Few-Shot Backdoor Attacks on Visual Object Tracking", "BadEncoder: Backdoor Attacks to Pre-trained Encoders in Self-Supervised Learning"], "answer_arxiv_id": ["2105.00164", "2006.01043v2", "2201.13178", "2108.00352"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_85"} +{"question": "Which studies have heavily focused on the generalization capabilities of overparameterized models on in-distribution data using conventional machine learning tools?", "answer": ["Norm-Based Capacity Control in Neural Networks", "Exploring Generalization in Deep Learning", "Implicit Regularization in Deep Learning", "Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data", "Spectrally-normalized margin bounds for neural networks", "Non-vacuous Generalization Bounds at the ImageNet Scale: a PAC-Bayesian Compression Approach", "Generalization bounds for deep convolutional neural networks", "Deterministic PAC-Bayesian generalization bounds for deep networks via generalizing noise-resilience"], "answer_arxiv_id": ["1503.00036", "1706.08947", "1709.01953", "1703.11008", "1706.08498", "1804.05862", "1905.12600", "1905.13344"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_86"} +{"question": "Which research works showed that differential privacy is also useful as tool for ensuring generalization in settings where the queries are chosen adaptively?", "answer": ["Preserving Statistical Validity in Adaptive Data Analysis", "Algorithmic Stability for Adaptive Data Analysis"], "answer_arxiv_id": ["1411.2664", "1511.02513"], "source_meta": {"published_time": "20210620"}, "qid": "AutoScholarQuery_train_87"} +{"question": "Which research work considers the gradient from prediction to feature inputs in context of explanations?", "answer": ["Visualizing and Understanding Convolutional Networks", "Towards better understanding of gradient-based attribution methods for Deep Neural Networks"], "answer_arxiv_id": ["1311.2901", "1711.06104"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_88"} +{"question": "What are the approaches for addressing the misalignment between the warped clothing and the human body?", "answer": ["Towards Photo-Realistic Virtual Try-On by Adaptively\n Generating$\\leftrightarrow$Preserving Image Content", "High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled\n Conditions", "GP-VTON: Towards General Purpose Virtual Try-on via Collaborative\n Local-Flow Global-Parsing Learning", "VITON-HD: High-Resolution Virtual Try-On via Misalignment-Aware\n Normalization", "Do Not Mask What You Do Not Need to Mask: a Parser-Free Virtual Try-On", "Parser-Free Virtual Try-on via Distilling Appearance Flows"], "answer_arxiv_id": ["2003.05863", "2206.14180", "2303.13756", "2103.16874", "2007.02721", "2103.04559"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_89"} +{"question": "What are the papers on mesh-based methods for 3D asset generation?", "answer": ["CLIP-Actor: Text-Driven Recommendation and Stylization for Animating\n Human Meshes", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "Text2Tex: Text-driven Texture Synthesis via Diffusion Models"], "answer_arxiv_id": ["2206.04382", "2112.03221", "2303.11396"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_90"} +{"question": "Which papers proposed diffusion models for graphs based on Gaussian noise?", "answer": ["Permutation Invariant Graph Generation via Score-Based Generative Modeling", "Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations"], "answer_arxiv_id": ["2003.00638", "2202.02514"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_91"} +{"question": "Is there any research that has used model-based methods to formalize definitions of non-additive interactions?", "answer": ["Predictive learning via rule ensembles", "Detecting Statistical Interactions From Neural Network Weights", "Feature Interaction Interpretability: A Case for Explaining Ad-Recommendation Systems via Neural Interaction Detection", "Does my multimodal model learn cross-modal interactions? It’s harder to tell than you might think!"], "answer_arxiv_id": ["0811.1679", "1705.04977", "2006.10966", "2010.06572"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_92"} +{"question": "Can you name some works that focused on achieving global style transformations of 3D scenes?", "answer": ["NeRF-Art: Text-Driven Neural Radiance Fields Stylization", "Stylizing 3D Scene via Implicit Representation and HyperNetwork", "Learning to Stylize Novel Views", "StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D\n Mutual Learning", "SNeRF: Stylized Neural Implicit Representations for 3D Scenes", "ARF: Artistic Radiance Fields", "PaletteNeRF: Palette-based Color Editing for NeRFs"], "answer_arxiv_id": ["2212.08070", "2105.13016", "2105.13509", "2205.12183", "2207.02363", "2206.06360", "2212.12871"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_93"} +{"question": "Could you provide me some studies that used imitation learning methods and achieved good performance on offline reinforcement learning benchmarks?", "answer": ["Online Decision Transformer"], "answer_arxiv_id": ["2202.05607"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_94"} +{"question": "What papers implemented unit quaternion in rotation regression?", "answer": ["Geometric Loss Functions for Camera Pose Regression with Deep Learning", "PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization", "PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes"], "answer_arxiv_id": ["1704.00390v2", "1505.07427", "1711.00199"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_95"} +{"question": "What are the papers that discuss 2D-based methods in talking head synthesis?", "answer": ["A Lip Sync Expert Is All You Need for Speech to Lip Generation In The\n Wild", "Talking Face Generation by Adversarially Disentangled Audio-Visual\n Representation", "StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via\n Pre-trained StyleGAN"], "answer_arxiv_id": ["2008.10010", "1807.07860", "2203.04036"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_96"} +{"question": "What research papers have proposed the use of residual learning for out-of-distribution detection?", "answer": ["ViM: Out-Of-Distribution with Virtual-logit Matching"], "answer_arxiv_id": ["2203.10807"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_97"} +{"question": "What are some research papers that apply causality in OOD problems?", "answer": ["Variational Inference: A Review for Statisticians", "Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1601.00670", "1312.6114"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_98"} +{"question": "Which studies proposed training-free efficiency enhancement schemes?", "answer": ["Neuron Merging: Compensating for Pruned Neurons", "Data-free parameter pruning for Deep Neural Networks", "RED : Looking for Redundancies for Data-Free Structured Compression of\n Deep Neural Networks", "A Fast Post-Training Pruning Framework for Transformers", "Adaptive Token Sampling For Efficient Vision Transformers", "Token Merging: Your ViT But Faster", "Zero-TPrune: Zero-Shot Token Pruning through Leveraging of the Attention\n Graph in Pre-Trained Transformers"], "answer_arxiv_id": ["2010.13160", "1507.06149", "2105.14797", "2204.09656", "2111.15667", "2210.09461", "2305.17328"], "source_meta": {"published_time": "20240508"}, "qid": "AutoScholarQuery_train_99"} +{"question": "Could you tell me about the works that address deterministic multi-person 3D motion forecasting?", "answer": ["Socially and Contextually Aware Human Motion and Pose Forecasting", "TRiPOD: Human Trajectory and Pose Dynamics Forecasting in the Wild", "Multi-Person Extreme Motion Prediction", "Multi-Person 3D Motion Prediction with Multi-Range Transformers"], "answer_arxiv_id": ["2007.06843", "2104.04029", "2105.08825", "2111.12073"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_100"} +{"question": "Could you provide me studies about Control-guided image generation?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "UniControl: A Unified Diffusion Model for Controllable Visual Generation\n In the Wild"], "answer_arxiv_id": ["2302.05543", "2305.11147"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_101"} +{"question": "Any studies about a class-incremental method without using source training data?", "answer": ["Class-Incremental Domain Adaptation"], "answer_arxiv_id": ["2008.01389"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_102"} +{"question": "Which research works utilized diffusion-based models in the field of Co-speech Gesture Generation?", "answer": ["Listen, Denoise, Action! Audio-Driven Motion Synthesis with Diffusion\n Models", "Taming Diffusion Models for Audio-Driven Co-Speech Gesture Generation", "DiffuseStyleGesture: Stylized Audio-Driven Co-Speech Gesture Generation\n with Diffusion Models"], "answer_arxiv_id": ["2211.09707", "2303.09119", "2305.04919"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_103"} +{"question": "What paper pioneered the generative commonsense reasoning task format?", "answer": ["CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning"], "answer_arxiv_id": ["1911.03705"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_104"} +{"question": "What works have proposed pre-training models for identifying urban villages through a master-slave framework?", "answer": ["A Contextual Master-Slave Framework on Urban Region Graph for Urban Village Detection"], "answer_arxiv_id": ["2211.14633"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_105"} +{"question": "What were some approaches taken to handle missing data in RNNs?", "answer": ["Recurrent Neural Networks for Multivariate Time Series with Missing Values"], "answer_arxiv_id": ["1606.01865"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_106"} +{"question": "Could you name a few works that used unsupervised 2D body keypoints to extract motion representations?", "answer": ["Animating Arbitrary Objects via Deep Motion Transfer", "First Order Motion Model for Image Animation", "Motion Representations for Articulated Animation"], "answer_arxiv_id": ["1812.08861", "2003.00196", "2104.11280"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_107"} +{"question": "What works mentioned using convolutions in transformer blocks for vision transformers?", "answer": ["CvT: Introducing Convolutions to Vision Transformers", "CMT: Convolutional Neural Networks Meet Vision Transformers", "ViTAE: Vision Transformer Advanced by Exploring Intrinsic Inductive Bias"], "answer_arxiv_id": ["2103.15808", "2107.06263", "2106.03348"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_108"} +{"question": "Could you provide works related to system optimizations to improve inference speed for transforming LLMs?", "answer": ["LightSeq: A High Performance Inference Library for Transformers", "Efficiently Scaling Transformer Inference", "TurboTransformers: An Efficient GPU Serving System For Transformer\n Models", "DeepSpeed Inference: Enabling Efficient Inference of Transformer Models\n at Unprecedented Scale"], "answer_arxiv_id": ["2010.13887", "2211.05102", "2010.05680", "2207.00032"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_109"} +{"question": "Could you provide me studies about training neural networks using unlabeled video data for motion estimation?", "answer": ["Back to Basics: Unsupervised Learning of Optical Flow via Brightness Constancy and Motion Smoothness", "Occlusion Aware Unsupervised Learning of Optical Flow", "DDFlow: Learning Optical Flow with Unlabeled Data Distillation", "What Matters in Unsupervised Optical Flow", "Learning Pixel Trajectories with Multiscale Contrastive Random Walks"], "answer_arxiv_id": ["1608.05842", "1711.05890", "1902.09145", "2006.04902", "2201.08379"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_110"} +{"question": "Which studies can you provide that involve tailored methods in classification scoring?", "answer": ["Identifying Mislabeled Data using the Area Under the Margin Ranking", "Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamics", "Data-IQ: Characterizing subgroups with heterogeneous outcomes in tabular data", "An Empirical Study of Example Forgetting during Deep Neural Network Learning", "Characterizing Datapoints via Second-Split Forgetting"], "answer_arxiv_id": ["2001.10528", "2009.10795", "2210.13043", "1812.05159", "2210.15031"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_111"} +{"question": "Could you provide me some works about data augmentation through latent space exploration?", "answer": ["Dataset Augmentation in Feature Space"], "answer_arxiv_id": ["1702.05538"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_112"} +{"question": "Which papers have proposed the technique of prompt tuning while working on Few-Shot Learning with Foundation Models?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Prompt-aligned Gradient for Prompt Tuning", "MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2205.14865", "2210.03117"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_113"} +{"question": "Any works about RL algorithms that use measures like Conditional Value at Risk, Entropic risk measure and others?", "answer": ["Policy Gradient for Coherent Risk Measures", "Policy Gradients with Variance Related Risk Criteria", "A policy gradient approach for optimization of smooth risk measures", "Algorithms for CVaR Optimization in MDPs"], "answer_arxiv_id": ["1502.03919", "1206.6404", "2202.11046", "1406.3339"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_114"} +{"question": "Can you tell me the studies that have used datasets like Vimeo90K or REDS for video restoration tasks?", "answer": ["Video Enhancement with Task-Oriented Flow"], "answer_arxiv_id": ["1711.09078"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_115"} +{"question": "Which papers provide an approach to labelling videos with actions using a curated, action-labelled dataset?", "answer": ["Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos", "Reinforcement Learning with Videos: Combining Offline Observations with Interaction"], "answer_arxiv_id": ["2206.11795", "2011.06507"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_116"} +{"question": "What papers focused on the use of instruction tuning and alignment in the adaptation of Large Language Models (LLMs) to new tasks?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2109.01652", "2203.02155"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_117"} +{"question": "Could you provide me some works about solutions for the prior hole problem?", "answer": ["VAE with a VampPrior", "Variational Inference with Normalizing Flows", "Normalizing Flows for Probabilistic Modeling and Inference", "Self-Supervised Variational Auto-Encoders", "Learning Hierarchical Priors in VAEs", "BIVA: A Very Deep Hierarchy of Latent Variables for Generative Modeling", "Missing Data Imputation and Acquisition with Deep Hierarchical Models and Hamiltonian Monte Carlo", "LION: Latent Point Diffusion Models for 3D Shape Generation"], "answer_arxiv_id": ["1705.07120", "1505.05770", "1912.02762", "2010.02014", "1905.04982", "1902.02102", "2202.04599", "2210.06978"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_118"} +{"question": "What papers used VLMs for zero-shot visual recognition?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Open-Vocabulary Object Detection Using Captions", "Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "Learning to Detect and Segment for Open Vocabulary Object Detection"], "answer_arxiv_id": ["2103.00020", "2011.10678", "2104.13921", "2212.12130"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_119"} +{"question": "Can you list some works that study the batch acquisition setting in Bayesian active learning and optimization?", "answer": ["BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning", "Batched Gaussian Process Bandit Optimization via Determinantal Point Processes"], "answer_arxiv_id": ["1906.08158", "1611.04088"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_120"} +{"question": "Which paper introduced Self-Refine, an iterative framework for self-refinement to autonomously improve generation?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback"], "answer_arxiv_id": ["2303.17651"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_121"} +{"question": "What work utilises self-generated supervisory labels from data themselves to efficiently learn target representations in medical image segmentation tasks?", "answer": ["Deep Adversarial Training for Multi-Organ Nuclei Segmentation in Histopathology Images", "CXR Segmentation by AdaIN-based Domain Adaptation and Knowledge Distillation"], "answer_arxiv_id": ["1810.00236", "2104.05892"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_122"} +{"question": "What works have raised concerns regarding the sensitivity of standard DRO to outliers?", "answer": ["Fairness Without Demographics in Repeated Loss Minimization", "Does Distributionally Robust Supervised Learning Give Robust Classifiers?", "Generalized Resilience and Robust Statistics"], "answer_arxiv_id": ["1806.08010", "1611.02041", "1909.08755"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_123"} +{"question": "What references first showed the impossibility to learn with single policy coverage in offline multi-agent Markov games?", "answer": ["When is Offline Two-Player Zero-Sum Markov Game Solvable?"], "answer_arxiv_id": ["2201.03522"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_124"} +{"question": "What papers discuss the concept of diffusion models in the context of motion modeling?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Planning with Diffusion for Flexible Behavior Synthesis", "Trace and Pace: Controllable Pedestrian Animation via Guided Trajectory\n Diffusion", "Human Motion Diffusion Model", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "EDGE: Editable Dance Generation From Music", "MoFusion: A Framework for Denoising-Diffusion-based Motion Synthesis", "Diffusion-based Generation, Optimization, and Planning in 3D Scenes", "Resolving 3D Human Pose Ambiguities with 3D Scene Constraints"], "answer_arxiv_id": ["1503.03585", "2205.09991", "2304.01893", "2209.14916", "2208.15001", "2211.10658", "2212.04495", "2301.06015", "1908.06963"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_125"} +{"question": "Can you name some research papers focusing on the understanding aspect of multimodal data?", "answer": ["Otter: A Multi-Modal Model with In-Context Instruction Tuning", "MiniGPT-5: Interleaved Vision-and-Language Generation via Generative\n Vokens"], "answer_arxiv_id": ["2305.03726", "2310.02239"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_126"} +{"question": "What research efforts have been made to reduce the size of model weights in LLM?", "answer": ["ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers", "LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers", "Rethinking the Role of Scale for In-Context Learning: An Interpretability-based Case Study at 66 Billion Scale", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models"], "answer_arxiv_id": ["2206.01861", "2208.07339", "2210.17323", "2212.09095", "2211.10438"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_127"} +{"question": "What are the papers about marginal motion prediction?", "answer": ["Implicit Latent Variable Model for Scene-Consistent Motion Forecasting", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction", "The Trajectron: Probabilistic Multi-Agent Trajectory Modeling With Dynamic Spatiotemporal Graphs", "DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents", "Learning Lane Graph Representations for Motion Forecasting"], "answer_arxiv_id": ["2007.12036", "1910.05449", "1810.05993", "1704.04394", "2007.13732"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_128"} +{"question": "What research work derives from the pioneering study of disentangled texture and shape representation learning for motion magnification?", "answer": ["Learning-based Video Motion Magnification"], "answer_arxiv_id": ["1804.02684"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_129"} +{"question": "Which paper reviewed the task of estimating quantum functions using classical shadow?", "answer": ["Predicting Many Properties of a Quantum System from Very Few Measurements"], "answer_arxiv_id": ["2002.08953"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_130"} +{"question": "What studies discuss low-precision quantization in architectures?", "answer": ["Low-bit Quantization of Neural Networks for Efficient Inference", "Learning to Quantize Deep Networks by Optimizing Quantization Intervals\n with Task Loss", "Pareto-Optimal Quantized ResNet Is Mostly 4-bit", "PROFIT: A Novel Training Method for sub-4-bit MobileNet Models"], "answer_arxiv_id": ["1902.06822", "1808.05779", "2105.03536", "2008.04693"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_131"} +{"question": "What research introduce the concepts of latent diffusion models and cascaded diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2112.10752", "2106.15282"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_132"} +{"question": "What papers documented regret analysis of expected improvememnt in the noiseless and frequentist setting?", "answer": ["Convergence rates of efficient global optimization algorithms"], "answer_arxiv_id": ["1101.3501v3"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_133"} +{"question": "Could you provide me some studies conducted on Decision Tree Distillation?", "answer": ["NDT: Neual Decision Tree Towards Fully Functioned Neural Graph", "Distilling a Neural Network Into a Soft Decision Tree", "Adaptive Neural Trees", "Attention Convolutional Binary Neural Tree for Fine-Grained Visual Categorization"], "answer_arxiv_id": ["1712.05934", "1711.09784", "1807.06699", "1909.11378"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_134"} +{"question": "What works provide neural approximators but use fully autoregressive architectures, resulting in slow sampling, like AQF and NAF?", "answer": ["Autoregressive Quantile Flows for Predictive Uncertainty Estimation", "Neural Autoregressive Flows"], "answer_arxiv_id": ["2112.04643", "1804.00779"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_135"} +{"question": "In which study is the WebQA benchmark, an open-domain multi-modal question answering benchmark, introduced?", "answer": ["WebQA: Multihop and Multimodal QA"], "answer_arxiv_id": ["2109.00590"], "source_meta": {"published_time": "20220901"}, "qid": "AutoScholarQuery_train_136"} +{"question": "Which research papers delve into the the application of neural implicit representation with SLAM?", "answer": ["iMAP: Implicit Mapping and Positioning in Real-Time", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of\n Signed Distance Fields", "Vox-Fusion: Dense Tracking and Mapping with Voxel-based Neural Implicit\n Representation", "In-Place Scene Labelling and Understanding with Implicit Scene\n Representation", "vMAP: Vectorised Object Mapping for Neural Field SLAM", "Neural Implicit Dense Semantic SLAM"], "answer_arxiv_id": ["2103.12352", "2112.12130", "2211.11704", "2210.15858", "2103.15875", "2302.01838v2", "2304.14560"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_137"} +{"question": "What papers focus on video-first learning through transformer-based video architectures?", "answer": ["ViViT: A Video Vision Transformer", "Is Space-Time Attention All You Need for Video Understanding?", "Video Swin Transformer"], "answer_arxiv_id": ["2103.15691", "2102.05095", "2106.13230"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_138"} +{"question": "Which works propose to enhance the representation power of NAT by introducing latent variables?", "answer": ["Theory and Experiments on Vector Quantized Autoencoders", "Latent-Variable Non-Autoregressive Neural Machine Translation with Deterministic Inference Using a Delta Posterior"], "answer_arxiv_id": ["1805.11063", "1908.07181"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_139"} +{"question": "Any works that utilize vision and sound for cross-modal SSL?", "answer": ["Self-Supervised Learning by Cross-Modal Audio-Video Clustering", "Look, Listen and Learn", "Objects that Sound", "Self-labelling via simultaneous clustering and representation learning", "SoundNet: Learning Sound Representations from Unlabeled Video", "Learning Representations from Audio-Visual Spatial Alignment", "Ambient Sound Provides Supervision for Visual Learning", "Self-Supervised Learning of Audio-Visual Objects from Video"], "answer_arxiv_id": ["1911.12667", "1705.08168", "1712.06651", "1911.05371", "1610.09001", "2011.01819", "1608.07017", "2008.04237"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_140"} +{"question": "Could you name the papers that introduced attention-based memory methods that have supplemented Recurrent Neural Networks in many applications?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_141"} +{"question": "Which works derived the expressions for mixture GFlowNet policies and classifier-guided GFlowNet policies analogous to those derived for diffusion models?", "answer": ["Non-Denoising Forward-Time Diffusions", "Q", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2312.14589", "1611.08152", "1503.03585", "2105.05233"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_142"} +{"question": "What studies discuss the application of disentangled representation learning in computer vision and natural language processing?", "answer": ["Diverse Image-to-Image Translation via Disentangled Representations", "Generating Sentences from Disentangled Syntactic and Semantic Spaces"], "answer_arxiv_id": ["1808.00948", "1907.05789"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_143"} +{"question": "Can you name some studies that focused on deep learning uncertainty, exploring to combine deep learning with Bayesian probability theory?", "answer": ["Bayesian Neural Networks: An Introduction and Survey"], "answer_arxiv_id": ["2006.12024"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_144"} +{"question": "Which papers represent the first attempts to address personalization in text-conditioned image generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_145"} +{"question": "Which studies developed model stitching as an approach for efficient reutilization of pre-trained models?", "answer": ["Understanding image representations by measuring their equivariance and\n equivalence", "Revisiting Model Stitching to Compare Neural Representations", "Similarity and Matching of Neural Network Representations", "Deep Model Reassembly", "Stitchable Neural Networks"], "answer_arxiv_id": ["1411.5908", "2106.07682", "2110.14633", "2210.17409", "2302.06586"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_146"} +{"question": "Are there any works that introduced models with latent representations identified by deterministic hard interventions?", "answer": ["Interventional Causal Representation Learning"], "answer_arxiv_id": ["2209.11924"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_147"} +{"question": "What works proposed learning an ideal invariant classifier on top of the representation space in the field of domain generalization?", "answer": ["Invariant Risk Minimization"], "answer_arxiv_id": ["1907.02893"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_148"} +{"question": "What research demonstrates a comprehensive approach to assessing the quality of images by leveraging content and image quality-aware features?", "answer": ["Re-IQA: Unsupervised Learning for Image Quality Assessment in the Wild"], "answer_arxiv_id": ["2304.00451"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_149"} +{"question": "Are there any studies on large model prompting that worked on pre-trained vision-language models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Learning Transferable Visual Models From Natural Language Supervision", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations"], "answer_arxiv_id": ["2204.14198", "2103.00020", "1908.08530"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_150"} +{"question": "Any works about extracting human hand poses from videos and using RL/IL for visuomotor control tasks?", "answer": ["DexMV: Imitation Learning for Dexterous Manipulation from Human Videos", "Robotic Telekinesis: Learning a Robotic Hand Imitator by Watching Humans on YouTube", "VideoDex: Learning Dexterity from Internet Videos", "Learning to Imitate Object Interactions from Internet Videos", "Masked Visual Pre-training for Motor Control", "Where are we in the search for an Artificial Visual Cortex for Embodied Intelligence?"], "answer_arxiv_id": ["2108.05877", "2202.10448", "2212.04498", "2211.13225", "2203.06173", "2303.18240"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_151"} +{"question": "What works used ImageNet pre-training for visual RL tasks?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "Pre-Trained Image Encoder for Generalizable Visual Reinforcement Learning"], "answer_arxiv_id": ["2107.03380", "2212.08860"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_152"} +{"question": "Which studies attempted to optimize a deformation field to predict the displacement of the scene across time for dynamic scene reconstruction?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields", "Neural Radiance Flow for 4D View Synthesis and Video Processing", "Neural 3D Video Synthesis from Multi-view Video", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View\n Synthesis of a Dynamic Scene From Monocular Video", "Editable Free-viewpoint Video Using a Layered Neural Representation", "Fast Dynamic Radiance Fields with Time-Aware Neural Voxels", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2011.12948", "2012.09790", "2103.02597", "2012.12247", "2104.14786", "2205.15285", "2106.13228v2"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_153"} +{"question": "What studies detailed the use of deep learning in software testing?", "answer": ["Improving Automatic Source Code Summarization via Deep Reinforcement\n Learning"], "answer_arxiv_id": ["1811.07234"], "source_meta": {"published_time": "20230716"}, "qid": "AutoScholarQuery_train_154"} +{"question": "What recent works integrated textual conditioning within denoising steps in the context of text-to-image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers", "ERNIE-ViLG 2.0: Improving Text-to-Image Diffusion Model with Knowledge-Enhanced Mixture-of-Denoising-Experts", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2211.01324", "2210.15257", "2205.11487"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_155"} +{"question": "Could you tell me some researches that utilized auxiliary information such as optical flow, motion segmentation and etc. for object-centric representation learning?", "answer": ["Conditional Object-Centric Learning from Video", "Discovering Objects that Can Move", "SAVi++: Towards End-to-End Object-Centric Learning from Real-World\n Videos"], "answer_arxiv_id": ["2111.12594", "2203.10159", "2206.07764"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_156"} +{"question": "Which works discuss applying a learned policy for structured prediction in the context of iterative decoding?", "answer": ["Search-based Structured Prediction"], "answer_arxiv_id": ["0907.0786"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_157"} +{"question": "Which works demonstrated that scaling up transformer-based language models improve performance on NLP tasks?", "answer": ["Attention Is All You Need", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1706.03762", "1810.04805", "2005.14165"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_158"} +{"question": "What research proposes a CNN-LSTM framework for content-aware layout generation?", "answer": ["PosterLayout: A New Benchmark and Approach for Content-aware\n Visual-Textual Presentation Layout"], "answer_arxiv_id": ["2303.15937"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_159"} +{"question": "Which publication introduced the ABC dataset, containing geometric models defined by parametric surfaces and their decompositions into patches?", "answer": ["ABC: A Big CAD Model Dataset For Geometric Deep Learning"], "answer_arxiv_id": ["1812.06216"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_160"} +{"question": "Which studies employed the approach of treating the accuracy of the LFs as parameters then using iterative methods?", "answer": ["Constrained Labeling for Weakly Supervised Learning"], "answer_arxiv_id": ["2009.07360"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_161"} +{"question": "What researches have focused on fine-grained classification in the area of part detection?", "answer": ["Part-based R-CNNs for Fine-grained Category Detection"], "answer_arxiv_id": ["1407.3867"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_162"} +{"question": "What works stemmed from the Adapter in the field of parameter-efficient fine-tuning?", "answer": ["AdapterFusion: Non-Destructive Task Composition for Transfer Learning", "AdapterHub: A Framework for Adapting Transformers"], "answer_arxiv_id": ["2005.00247", "2007.07779"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_163"} +{"question": "Any studies highlight the significance of continual learning for Language Models?", "answer": ["An Empirical Study of Catastrophic Forgetting in Large Language Models\n During Continual Fine-tuning", "Fine-tuned Language Models are Continual Learners"], "answer_arxiv_id": ["2308.08747", "2205.12393"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_164"} +{"question": "Which works utilize sketch-based methods in the decoder side of text-to-SQL tasks?", "answer": ["A Comprehensive Exploration on WikiSQL with Table-Aware Word Contextualization", "SQLNet: Generating Structured Queries From Natural Language Without Reinforcement Learning", "Improving Text-to-SQL with Schema Dependency Learning"], "answer_arxiv_id": ["1902.01069", "1711.04436v1", "2103.04399"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_165"} +{"question": "Any studies covering the topic of non-smooth weakly convex optimization?", "answer": ["Stochastic model-based minimization of weakly convex functions", "Proximally Guided Stochastic Subgradient Method for Nonsmooth, Nonconvex Problems", "Efficiency of minimizing compositions of convex functions and smooth maps", "Stochastic Methods for Composite and Weakly Convex Optimization Problems"], "answer_arxiv_id": ["1803.06523", "1707.03505", "1605.00125", "1703.08570v3"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_166"} +{"question": "Could you provide me some studies about methods that use quantiles of the most recent errors for prediction interval?", "answer": ["Conformal prediction for time series"], "answer_arxiv_id": ["2010.09107"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_167"} +{"question": "Are there any studies highlighting the oppression and misconceptions about sign languages?", "answer": ["Including Signed Languages in Natural Language Processing"], "answer_arxiv_id": ["2105.05222"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_168"} +{"question": "What work analyzes the relationship between the learning of different frequencies and the robustness of a DNN?", "answer": ["A Fourier Perspective on Model Robustness in Computer Vision"], "answer_arxiv_id": ["1906.08988"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_169"} +{"question": "What studies propose data augmentation techniques for achieving cross-lingual transfer?", "answer": ["Consistency Regularization for Cross-Lingual Fine-Tuning"], "answer_arxiv_id": ["2106.08226"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_170"} +{"question": "Which studies demonstrate that traditional methods for unsupervised domain adaptation do not effectively address geographical domain shift?", "answer": ["Can domain adaptation make object recognition work for everyone?", "GeoNet: Benchmarking Unsupervised Adaptation across Geographies"], "answer_arxiv_id": ["2204.11122", "2303.15443"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_171"} +{"question": "Which works led to techniques for high-fidelity customizations in subject-driven image generation using diffusion models?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2208.01618", "2210.09276"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_172"} +{"question": "Which studies do approximation of self-attention with a low-rank decomposition?", "answer": ["SOFT: Softmax-free Transformer with Linear Complexity", "Nystr\\\"omformer: A Nystr\\\"om-Based Algorithm for Approximating\n Self-Attention"], "answer_arxiv_id": ["2110.11945", "2102.03902"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_173"} +{"question": "What works have studied scheduling with testing where there are unknown processing requirements of a job?", "answer": ["An Adversarial Model for Scheduling with Testing", "Explorable Uncertainty in Scheduling with Non-Uniform Testing Times"], "answer_arxiv_id": ["1709.02592v3", "2009.13316v2"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_174"} +{"question": "What are the papers that have made significant progress in human-robot handovers?", "answer": ["Object Handovers: a Review for Robotics", "Object-Independent Human-to-Robot Handovers using Real Time Robotic\n Vision"], "answer_arxiv_id": ["2007.12952", "2006.01797"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_175"} +{"question": "What papers originally proposed diffusion models for image synthesis?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2105.05233"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_176"} +{"question": "Are there any works regarding the application of diffusion models in image variation tasks?", "answer": ["Versatile Diffusion: Text, Images and Variations All in One Diffusion Model"], "answer_arxiv_id": ["2211.08332v4"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_177"} +{"question": "Can you indicate studies where Transformer model was used for imagery data processing?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_178"} +{"question": "Which works discuss the role of LLMs in abstractive summarization research?", "answer": ["Summarization is (Almost) Dead"], "answer_arxiv_id": ["2309.09558"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_179"} +{"question": "Could you mention research papers that focus on the role and effectiveness of layer normalization in transformer architecture?", "answer": ["On Layer Normalization in the Transformer Architecture", "Understanding and Improving Layer Normalization"], "answer_arxiv_id": ["2002.04745", "1911.07013"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_180"} +{"question": "What papers have studied the training dynamics of overparametrized neural networks with random initialization?", "answer": ["SGD Learns the Conjugate Kernel Class of the Network", "Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Data", "Gradient Descent Provably Optimizes Over-parameterized Neural Networks", "Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "On Lazy Training in Differentiable Programming"], "answer_arxiv_id": ["1702.08503", "1808.01204", "1810.02054", "1811.03804", "1811.03962", "1901.08584", "1812.07956"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_181"} +{"question": "Which works implement Spatial Propagation Network in 2D based depth completion methods?", "answer": ["Non-Local Spatial Propagation Network for Depth Completion", "RigNet++: Semantic Assisted Repetitive Image Guided Network for Depth\n Completion"], "answer_arxiv_id": ["2007.10042", "2309.00655"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_182"} +{"question": "Which papers have proposed generalizable methods to decompose the object’s geometry from hand-object interaction (HOI) and encode it as a condition for generative models?", "answer": ["A Skeleton-Driven Neural Occupancy Representation for Articulated Hands", "Grasping Field: Learning Implicit Representations for Human Grasps", "CAMS: CAnonicalized Manipulation Spaces for Category-Level Functional\n Hand-Object Manipulation Synthesis", "ContactGen: Generative Contact Modeling for Grasp Generation", "Contact2Grasp: 3D Grasp Synthesis via Hand-Object Contact Constraint"], "answer_arxiv_id": ["2109.11399", "2008.04451", "2303.15469", "2310.03740", "2210.09245"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_183"} +{"question": "What is the work where FedProx regularizes ℓ2-distance between local and global models?", "answer": ["Federated Optimization in Heterogeneous Networks"], "answer_arxiv_id": ["1812.06127"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_184"} +{"question": "Could you provide me some research that worked on aligning the generated 3D skeleton-based gesturing avatars with the audio input?", "answer": ["QPGesture: Quantization-Based and Phase-Guided Motion Matching for\n Natural Speech-Driven Gesture Generation"], "answer_arxiv_id": ["2305.11094"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_185"} +{"question": "Who discusses a connection between EOT projection problem and maximum-likelihood deconvolution?", "answer": ["Entropic optimal transport is maximum-likelihood deconvolution"], "answer_arxiv_id": ["1809.05572"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_186"} +{"question": "What work modifies model workflow design to accommodate bias factors in language model detoxification?", "answer": ["Measuring and Reducing Gendered Correlations in Pre-trained Models", "Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based\n Bias in NLP", "ADEPT: A DEbiasing PrompT Framework"], "answer_arxiv_id": ["2010.06032", "2103.00453", "2211.05414"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_187"} +{"question": "Which research papers primarily focus on large-scale indoor point cloud datasets?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts", "Self-Supervised Pretraining of 3D Features on any Point-Cloud"], "answer_arxiv_id": ["2007.10985", "2012.09165", "2101.02691"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_188"} +{"question": "What works are about regular group convolutions, which represent the input signal in terms of scalar fields on a group and rely on a discretization of the group space?", "answer": ["Group Equivariant Convolutional Networks", "On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "Roto-Translation Covariant Convolutional Networks for Medical Image Analysis", "B-Spline CNNs on Lie Groups", "Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data"], "answer_arxiv_id": ["1602.07576", "1802.03690", "1804.03393", "1909.12057", "2002.12880"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_189"} +{"question": "What is the work that introduced the concept of Zero-Shot Super-Resolution (ZSSR)?", "answer": ["\"Zero-Shot\" Super-Resolution using Deep Internal Learning"], "answer_arxiv_id": ["1712.06087"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_190"} +{"question": "What research has been conducted on the use of recalibration methods to correct the miscalibration of credible intervals?", "answer": ["Accurate Uncertainties for Deep Learning Using Calibrated Regression"], "answer_arxiv_id": ["1807.00263"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_191"} +{"question": "In what work is group theory used as a formalism to axiomatically define symmetries?", "answer": ["Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["2104.13478"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_192"} +{"question": "What research works deal with layout generation conditioned on element relationships?", "answer": ["Constrained Graphic Layout Generation via Latent Optimization", "Neural Design Network: Graphic Layout Generation with Constraints"], "answer_arxiv_id": ["2108.00871", "1912.09421"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_193"} +{"question": "What papers propose reweighting as a method for conformal prediction under distribution shift?", "answer": ["Conformal Prediction Under Covariate Shift", "Distribution-free uncertainty quantification for classification under label shift", "Conformalized Survival Analysis", "Conformal Prediction Beyond Exchangeability"], "answer_arxiv_id": ["1904.06019", "2103.03323", "2103.09763", "2202.13415"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_194"} +{"question": "Please provide references that have used the concept of self-training for weakly-supervised object detection methods?", "answer": ["Multiple Instance Detection Network with Online Instance Classifier\n Refinement", "C-MIL: Continuation Multiple Instance Learning for Weakly Supervised\n Object Detection", "Instance-aware, Context-focused, and Memory-efficient Weakly Supervised\n Object Detection"], "answer_arxiv_id": ["1704.00138", "1904.05647", "2004.04725"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_195"} +{"question": "Can you provide me studies about the Lagrangian-based approach in safe reinforcement learning?", "answer": ["Risk-Constrained Reinforcement Learning with Percentile Risk Criteria", "Constrained Policy Optimization via Bayesian World Models"], "answer_arxiv_id": ["1512.01629", "2201.09802"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_196"} +{"question": "What papers discuss the Inverse Clue Task for contrastive pre-training in dense retrieval?", "answer": ["Latent Retrieval for Weakly Supervised Open Domain Question Answering", "Pre-training Tasks for Embedding-based Large-scale Retrieval"], "answer_arxiv_id": ["1906.00300", "2002.03932"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_197"} +{"question": "Which papers use the greedy algorithm to generate adversarial examples?", "answer": ["Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment", "TextBugger: Generating Adversarial Text Against Real-world Applications", "A Strong Baseline for Query Efficient Attacks in a Black Box Setting"], "answer_arxiv_id": ["1907.11932", "1812.05271", "2109.04775"], "source_meta": {"published_time": "20240202"}, "qid": "AutoScholarQuery_train_198"} +{"question": "Which studies showed the use of single images as input into neural networks for creating strand-based hair models?", "answer": ["NeuralHDHair: Automatic High-fidelity Hair Modeling from a Single Image\n Using Implicit Neural Representations", "HairStep: Transfer Synthetic to Real Using Strand and Depth Maps for\n Single-View 3D Hair Modeling", "HairNet: Single-View Hair Reconstruction using Convolutional Neural\n Networks"], "answer_arxiv_id": ["2205.04175", "2303.02700", "1806.07467"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_199"} +{"question": "What paper proposed the notion of 'privacy-preserving prediction' to ensure privacy while making model predictions?", "answer": ["Privacy-preserving Prediction"], "answer_arxiv_id": ["1803.10266"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_200"} +{"question": "Which researches involved pre-trained LLMs that generate plans in language for RL systems?", "answer": ["Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents"], "answer_arxiv_id": ["2201.07207"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_201"} +{"question": "Which works discussed the approach of effectively incorporating target attribute information into the input of language models for controlled text generation?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Controllable Natural Language Generation with Contrastive Prefixes", "Focused Prefix Tuning for Controllable Text Generation"], "answer_arxiv_id": ["2101.00190", "2104.08691", "2202.13257", "2306.00369"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_202"} +{"question": "Which work proposes the concept of sequential annealing ABC based on the prior distribution?", "answer": ["A Simulated Annealing Approach to Approximate Bayes Computations"], "answer_arxiv_id": ["1208.2157"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_203"} +{"question": "Any works about the training of Language Models on massive data without group structure?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "The Pile: An 800GB Dataset of Diverse Text for Language Modeling", "Scaling Laws for Neural Language Models"], "answer_arxiv_id": ["1910.10683", "2101.00027", "2001.08361"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_204"} +{"question": "Could you list out some studies that focus on graph clustering using graph neural networks?", "answer": ["Attributed Graph Clustering: A Deep Attentional Embedding Approach", "Spectral Clustering with Graph Neural Networks for Graph Pooling", "Graph Clustering with Graph Neural Networks"], "answer_arxiv_id": ["1906.06532", "1907.00481", "2006.16904v3"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_205"} +{"question": "In what research was the concept of counterfactual memorization introduced adapting a notion from label memorization?", "answer": ["Does Learning Require Memorization? A Short Tale about a Long Tail"], "answer_arxiv_id": ["1906.05271"], "source_meta": {"published_time": "20211224"}, "qid": "AutoScholarQuery_train_206"} +{"question": "What models provide conversational interaction with human users?", "answer": ["Training language models to follow instructions with human feedback", "Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on Self-Chat Data"], "answer_arxiv_id": ["2203.02155", "2304.01196"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_207"} +{"question": "Which papers initially focused on answer-aware QG with factoid answers before shifting to natural, information-seeking questions?", "answer": ["A Dataset of Information-Seeking Questions and Answers Anchored in\n Research Papers"], "answer_arxiv_id": ["2105.03011"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_208"} +{"question": "What works make contributions to object shape reconstruction using detection and multi-view optimization?", "answer": ["FroDO: From Detections to 3D Objects"], "answer_arxiv_id": ["2005.05125"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_209"} +{"question": "What works have proposed various benchmarks of high-resource images?", "answer": ["ImageNet Large Scale Visual Recognition Challenge", "Microsoft COCO: Common Objects in Context", "The Cityscapes Dataset for Semantic Urban Scene Understanding", "Deep Learning Face Attributes in the Wild", "Visual Genome: Connecting Language and Vision Using Crowdsourced Dense\n Image Annotations"], "answer_arxiv_id": ["1409.0575", "1405.0312", "1604.01685v2", "1411.7766", "1602.07332"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_210"} +{"question": "Which works leverage CLIP for open-vocabulary semantic segmentation?", "answer": ["DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "Extract Free Dense Labels from CLIP", "Semantic Segmentation In-the-Wild Without Seeing Any Segmentation Examples", "Decoupling Zero-Shot Semantic Segmentation", "Image Segmentation Using Text and Image Prompts", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels"], "answer_arxiv_id": ["2112.01518", "2112.01071", "2112.03185", "2112.07910", "2112.10003", "2112.12143"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_211"} +{"question": "What works have explored view-invariance in action recognition?", "answer": ["Recognizing Actions in Videos from Unseen Viewpoints"], "answer_arxiv_id": ["2103.16516"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_212"} +{"question": "Are there any benchmarks that provide coded data such as ICD diagnosis codes?", "answer": ["EHRSHOT: An EHR Benchmark for Few-Shot Evaluation of Foundation Models"], "answer_arxiv_id": ["2307.02028"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_213"} +{"question": "Which works presents RAIN, allowing pre-trained LLMs to evaluate model outputs for AI safety?", "answer": ["RAIN: Your Language Models Can Align Themselves without Finetuning"], "answer_arxiv_id": ["2309.07124"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_214"} +{"question": "What studies adopted the 'sampling-evaluation' framework in tensor network structure search (TN-SS)?", "answer": ["Adaptive Learning of Tensor Network Structures", "Alternating Local Enumeration (TnALE): Solving Tensor Network Structure\n Search with Fewer Evaluations"], "answer_arxiv_id": ["2008.05437", "2304.12875"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_215"} +{"question": "What papers have contributed to the field of optical flow estimation in computer vision?", "answer": ["PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume", "Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimation"], "answer_arxiv_id": ["1709.02371", "1904.05290"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_216"} +{"question": "Can you give examples of research analyzing the implicit regularization and the incremental learning of gradient flow in hierarchical tensor decomposition?", "answer": ["Implicit Regularization in Tensor Factorization", "Implicit Regularization in Hierarchical Tensor Factorization and Deep Convolutional Neural Networks"], "answer_arxiv_id": ["2102.09972", "2201.11729"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_217"} +{"question": "Which papers focused on training vision-language models using contrastive losses to learn alignment between images and text?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_218"} +{"question": "Which works support the idea that LLMs can automatically generate prompts for themselves?", "answer": ["Large Language Models are Human-Level Prompt Engineers", "Automatic Chain of Thought Prompting in Large Language Models", "Generate rather than Retrieve: Large Langu-age Models are Strong Context Generators"], "answer_arxiv_id": ["2211.01910", "2210.03493", "2209.10063"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_219"} +{"question": "What papers are about the developments of hybrid neural-symbolic methods?", "answer": ["End-to-End Differentiable Proving", "Abstract Spatial-Temporal Reasoning via Probabilistic Abduction and Execution", "Neural-Symbolic Recursive Machine for Systematic Generalization", "A Minimalist Dataset for Systematic Generalization of Perception, Syntax, and Semantics"], "answer_arxiv_id": ["1705.11040", "2103.14230", "2210.01603v2", "2103.01403"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_220"} +{"question": "Which paper originally proposed Retrieval Augmented Generation (RAG)?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2005.11401"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_221"} +{"question": "What works extended neuro-symbolic frameworks to temporal reasoning tasks?", "answer": ["Grounding Physical Concepts of Objects and Events Through Dynamic Visual Reasoning"], "answer_arxiv_id": ["2103.16564"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_222"} +{"question": "What papers discussed the use of feed-forward networks in 3D scene synthesis?", "answer": ["Deep Generative Modeling for Scene Synthesis via Hybrid Representations"], "answer_arxiv_id": ["1808.02084"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_223"} +{"question": "Could you provide me some research that utilizes explicit animation methods to warp the source image to the target?", "answer": ["Animating Arbitrary Objects via Deep Motion Transfer", "First Order Motion Model for Image Animation", "Motion Representations for Articulated Animation", "Thin-Plate Spline Motion Model for Image Animation", "PIRenderer: Controllable Portrait Image Generation via Semantic Neural\n Rendering", "Fast Bi-layer Neural Synthesis of One-Shot Realistic Head Avatars", "One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing", "Implicit Warping for Animation with Image Sets", "FSGAN: Subject Agnostic Face Swapping and Reenactment"], "answer_arxiv_id": ["1812.08861", "2003.00196", "2104.11280", "2203.14367", "2109.08379", "2008.10174", "2011.15126", "2210.01794", "1908.05932"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_224"} +{"question": "Which papers discuss the conversion of visual information into text as a function of large multimodal models?", "answer": ["Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language", "Language Models with Image Descriptors are Strong Few-Shot\n Video-Language Learners", "PromptCap: Prompt-Guided Task-Aware Image Captioning"], "answer_arxiv_id": ["2204.00598", "2205.10747", "2211.09699"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_225"} +{"question": "Can you name the research papers that have used variational approaches in the study of FRL?", "answer": ["The Variational Fair Autoencoder", "Invariant Representations without Adversarial Training", "Learning Fair Representation via Distributional Contrastive Disentanglement"], "answer_arxiv_id": ["1511.00830", "1805.09458", "2206.08743"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_226"} +{"question": "Which paper uses an encoder-decoder model for completing missing parts caused by a single viewpoint?", "answer": ["Unsupervised Point Cloud Pre-Training via Occlusion Completion"], "answer_arxiv_id": ["2010.01089"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_227"} +{"question": "What work applied and adapted CLIP to downstream tasks using labeled data?", "answer": ["Learning to Prompt for Vision-Language Models", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters"], "answer_arxiv_id": ["2109.01134", "2110.04544"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_228"} +{"question": "What is the paper that introduced the MGSM dataset?", "answer": ["Language Models are Multilingual Chain-of-Thought Reasoners"], "answer_arxiv_id": ["2210.03057"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_229"} +{"question": "Which studies simplified environment parameters such as action duration or block break time to overcome the problem of item scarcity in Minecraft?", "answer": ["Align-RUDDER: Learning From Few Demonstrations by Reward Redistribution", "Mastering Diverse Domains through World Models"], "answer_arxiv_id": ["2009.14108", "2301.04104"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_230"} +{"question": "What works used graph diffusion to rewire the input graph, improving long-range connectivity for the GNN?", "answer": ["Diffusion Improves Graph Learning"], "answer_arxiv_id": ["1911.05485"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_231"} +{"question": "What studies investigate the problem of finding Nash equilibria/saddle points in convex-concave and nonconvex-concave problems?", "answer": ["Near-Optimal Algorithms for Minimax Optimization", "A Unified Analysis of Extra-gradient and Optimistic Gradient Methods for Saddle Point Problems: Proximal Point Approach", "Efficient Algorithms for Smooth Minimax Optimization", "Hybrid Block Successive Approximation for One-Sided Non-Convex Min-Max Problems: Algorithms and Applications", "Solving a Class of Non-Convex Min-Max Games Using Iterative First Order Methods", "An accelerated inexact proximal point method for solving nonconvex-concave min-max problems", "On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems"], "answer_arxiv_id": ["2002.02417", "1901.08511", "1907.01543", "1902.08294", "1902.08297", "1905.13433", "1906.00331"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_232"} +{"question": "Could you provide me a study that generated a hierarchical dataset for agents to learn from?", "answer": ["Skill Induction and Planning with Latent Language"], "answer_arxiv_id": ["2110.01517"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_233"} +{"question": "Are there any other recent, inspiring methods in cross-episodic learning?", "answer": ["In-context Reinforcement Learning with Algorithm Distillation", "Human-Timescale Adaptation in an Open-Ended Task Space"], "answer_arxiv_id": ["2210.14215", "2301.07608"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_234"} +{"question": "Which papers studied the connection between the flatness of minima and model generalization?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "Exploring Generalization in Deep Learning", "Sharpness-Aware Minimization for Efficiently Improving Generalization", "Penalizing Gradient Norm for Efficiently Improving Generalization in Deep Learning"], "answer_arxiv_id": ["1609.04836", "1706.08947", "2010.01412", "2202.03599"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_235"} +{"question": "What source discussed a method where both images and partial point clouds are accepted as inputs?", "answer": ["RevealNet: Seeing Behind Objects in RGB-D Scans"], "answer_arxiv_id": ["1904.12012"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_236"} +{"question": "Could you provide me studies, that utilized self-supervised models in the efforts to find the best performing pre-trained model for each brain region of interest (ROI)?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "An Empirical Study of Training Self-Supervised Vision Transformers", "DINOv2: Learning Robust Visual Features without Supervision", "PatchGame: Learning to Signal Mid-level Patches in Referential Games", "CoCoNets: Continuous Contrastive 3D Scene Representations"], "answer_arxiv_id": ["2111.06377", "2104.02057", "2304.07193", "2111.01785", "2104.03851"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_237"} +{"question": "Are there studies about encouraging LLMs to produce each reasoning step one at a time?", "answer": ["ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language"], "answer_arxiv_id": ["2012.13048"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_238"} +{"question": "What are some of the federated learning models that have been proposed recently?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "Federated Optimization in Heterogeneous Networks", "Personalized Federated Learning with Moreau Envelopes", "Personalized Federated Learning via Variational Bayesian Inference"], "answer_arxiv_id": ["1602.05629", "1812.06127", "2006.08848", "2206.07977"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_239"} +{"question": "Can you tell me what work was first on deep learning method for SAT instance generation?", "answer": ["G2SAT: Learning to Generate SAT Formulas"], "answer_arxiv_id": ["1910.13445"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_240"} +{"question": "What studies have manually crafted experiential prompts to provide textual experience to LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Better Zero-Shot Reasoning with Role-Play Prompting"], "answer_arxiv_id": ["2201.11903", "2308.07702"], "source_meta": {"published_time": "20240712"}, "qid": "AutoScholarQuery_train_241"} +{"question": "Could you provide me some studies that use synthetic data by rendering human models?", "answer": ["Learning from Synthetic Humans", "VirtualPose: Learning Generalizable 3D Human Pose Models from Virtual\n Data", "AGORA: Avatars in Geography Optimized for Regression Analysis"], "answer_arxiv_id": ["1701.01370", "2207.09949", "2104.14643"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_242"} +{"question": "Which research provides a detailed survey of the results in the area of feature learning in the presence of spurious correlations?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_243"} +{"question": "Which work conducts direct training of the Transformer within the SNN framework?", "answer": ["Spikformer: When Spiking Neural Network Meets Transformer"], "answer_arxiv_id": ["2209.15425"], "source_meta": {"published_time": "20240717"}, "qid": "AutoScholarQuery_train_244"} +{"question": "What paper discusses the computational complexity of algorithms using G-optimal design?", "answer": ["Learning with Good Feature Representations in Bandits and in RL with a Generative Model"], "answer_arxiv_id": ["1911.07676"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_245"} +{"question": "Which studies have been conducted in long-form text generation, specifically in story generation?", "answer": ["Strategies for Structuring Story Generation", "MEGATRON-CNTRL: Controllable Story Generation with External Knowledge\n Using Large-Scale Language Models"], "answer_arxiv_id": ["1902.01109", "2010.00840"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_246"} +{"question": "Are there any works about image decomposition-based enhancement for low-light-image enhancement?", "answer": ["Deep Retinex Decomposition for Low-Light Enhancement"], "answer_arxiv_id": ["1808.04560"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_247"} +{"question": "What works explored ways to address imbalance in large-vocabulary detectors using a change in loss or self-training with weakly labeled data?", "answer": ["Seesaw Loss for Long-Tailed Instance Segmentation", "Equalization Loss for Long-Tailed Object Recognition", "Equalization Loss v2: A New Gradient Balance Approach for Long-tailed\n Object Detection", "Long-tail Detection with Effective Class-Margins", "Probabilistic two-stage detection", "Simple Copy-Paste is a Strong Data Augmentation Method for Instance\n Segmentation", "MosaicOS: A Simple and Effective Use of Object-Centric Images for\n Long-Tailed Object Detection", "Rethinking Pre-training and Self-training"], "answer_arxiv_id": ["2008.10032", "2003.05176", "2012.08548", "2301.09724", "2103.07461", "2012.07177", "2102.08884", "2006.06882"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_248"} +{"question": "Which works focused on automatically identifying slices of data on which classifiers perform poorly?", "answer": ["Domino: Discovering Systematic Errors with Cross-Modal Embeddings", "Distilling Model Failures as Directions in Latent Space", "Adaptive Testing of Computer Vision Models", "Discovering Bugs in Vision Models using Off-the-shelf Image Generation and Captioning", "Identification of Systematic Errors of Image Classifiers on Rare Subgroups", "Diagnosing and Rectifying Vision Models using Language"], "answer_arxiv_id": ["2203.14960", "2206.14754", "2212.02774", "2208.08831", "2303.05072", "2302.04269"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_249"} +{"question": "Could you provide studies that used GAN for enhancing the quality of compressed images?", "answer": ["IEGAN: Multi-purpose Perceptual Quality Image Enhancement Using\n Generative Adversarial Network", "Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure\n Synthetic Data"], "answer_arxiv_id": ["1811.09134", "2107.10833"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_250"} +{"question": "Which studies are dedicated to selecting or reweighting training instances to reduce the negative effect of corrupted examples?", "answer": ["MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks\n on Corrupted Labels", "Learning to Reweight Examples for Robust Deep Learning", "Rethinking Importance Weighting for Deep Learning under Distribution\n Shift"], "answer_arxiv_id": ["1712.05055", "1803.09050", "2006.04662"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_251"} +{"question": "What is the earliest work that studied symmetries in deep learning architectures?", "answer": ["Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1602.07576"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_252"} +{"question": "Which papers discuss the use of Swin Transformer in OCR-free methods for VDU?", "answer": ["OCR-free Document Understanding Transformer", "End-to-end Document Recognition and Understanding with Dessurt"], "answer_arxiv_id": ["2111.15664", "2203.16618"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_253"} +{"question": "Which works utilize activation regularization between the target and student models and interpolation between generated examples for training?", "answer": ["ENHANCING DATA-FREE ADVERSARIAL DISTILLATION WITH ACTIVATION REGULARIZATION AND VIRTUAL INTERPOLATION"], "answer_arxiv_id": ["2102.11638"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_254"} +{"question": "Could you provide me some studies that utilize volume rendering techniques for novel view image generation?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "IBRNet: Learning Multi-View Image-Based Rendering"], "answer_arxiv_id": ["2012.02190", "2102.13090"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_255"} +{"question": "Can you point out work focusing on training specialized models on a single task and modality such as predicting masked RGB pixels?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "SiT: Self-supervised vIsion Transformer", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: a Simple Framework for Masked Image Modeling", "Are Large-scale Datasets Necessary for Self-Supervised Pre-training?"], "answer_arxiv_id": ["2010.11929", "2104.03602", "2111.06377", "2111.09886", "2112.10740"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_256"} +{"question": "What is the pioneering work for learning on raw point sets as input data for the tasks of classification, part segmentation, and semantic segmentation?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation"], "answer_arxiv_id": ["1612.00593"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_257"} +{"question": "Can you provide works that have used generative models as priors for image super-resolution using an iterative process?", "answer": ["Compressed Sensing using Generative Models", "Trumpets: Injective Flows for Inference and Inverse Problems"], "answer_arxiv_id": ["1703.03208", "2102.10461"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_258"} +{"question": "Could you provide some studies that used pose-estimates to build maps in the field of embodied AI?", "answer": ["Learning Exploration Policies for Navigation", "Learning To Explore Using Active Neural SLAM", "Occupancy Anticipation for Efficient Exploration and Navigation"], "answer_arxiv_id": ["1903.01959", "2004.05155", "2008.09285"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_259"} +{"question": "Which works considered coarse-grained pruning methods such as filter-wise or layer-wise pruning?", "answer": ["Pruning Filters for Efficient ConvNets", "ThiNet: A Filter Level Pruning Method for Deep Neural Network Compression", "Learning Structured Sparsity in Deep Neural Networks"], "answer_arxiv_id": ["1608.08710", "1707.06342", "1608.03665"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_260"} +{"question": "Which studies focus on overcoming the convergence issue in query-based image segmentation?", "answer": ["OneFormer: One Transformer to Rule Universal Image Segmentation"], "answer_arxiv_id": ["2211.06220"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_261"} +{"question": "Which works delved into Prompt Optimization techniques for image generation?", "answer": ["Optimizing Prompts for Text-to-Image Generation", "Promptify: Text-to-Image Generation through Interactive Prompt\n Exploration with Large Language Models"], "answer_arxiv_id": ["2212.09611", "2304.09337"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_262"} +{"question": "What paper introduced the Neural Radiance Fields (NeRF) representing scenes based on a continuous volumetric function?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_263"} +{"question": "Could you provide some studies on mitigating the catastrophic forgetting of old knowledge in continual segmentation?", "answer": ["Modeling the Background for Incremental Learning in Semantic\n Segmentation", "PLOP: Learning without Forgetting for Continual Semantic Segmentation", "Continual Semantic Segmentation with Automatic Memory Sample Selection"], "answer_arxiv_id": ["2002.00718", "2011.11390", "2304.05015"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_264"} +{"question": "Could you provide me the study that provided a perturbation analysis of NC to study 'inexact collapse'?", "answer": ["Perturbation Analysis of Neural Collapse"], "answer_arxiv_id": ["2210.16658"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_265"} +{"question": "What studies have improved text-to-image models using large-scale auto-regressive models?", "answer": ["Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers"], "answer_arxiv_id": ["2102.12092", "2105.13290"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_266"} +{"question": "Which works proposed to limit the deviation from the behavior policy by using an explicit density model in offline RL?", "answer": ["Behavior Regularized Offline Reinforcement Learning", "Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "EMaQ: Expected-Max Q-Learning Operator for Simple Yet Effective Offline and Online RL"], "answer_arxiv_id": ["1911.11361", "1812.02900", "1906.00949", "2007.11091"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_267"} +{"question": "What papers worked on learning a general representation across multiple and possibly unseen tasks and environments?", "answer": ["Domain Generalization: A Survey", "In Search of Lost Domain Generalization", "Wilds: A Benchmark of in-the-Wild Distribution Shifts", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time", "Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization", "A Fine-Grained Analysis on Distribution Shift", "Domain Generalization via Invariant Feature Representation"], "answer_arxiv_id": ["2103.02503", "2007.01434", "2012.07421", "2203.05482", "2107.04649", "2110.11328", "1301.2115"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_268"} +{"question": "What papers proposed Dropout-based methods such as Monte Carlo Dropout for uncertainty quantification?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.02142"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_269"} +{"question": "Which studies propose diffusion models for high quality image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_270"} +{"question": "What is the title of the work that presented the SubsetSearch algorithm in the context of local causal graph discovery?", "answer": ["Subset verification and search algorithms for causal DAGs"], "answer_arxiv_id": ["2301.03180"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_271"} +{"question": "Which works proposed multi-task train-once design in image segmentation?", "answer": ["OneFormer: One Transformer to Rule Universal Image Segmentation"], "answer_arxiv_id": ["2211.06220"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_272"} +{"question": "What studies have used selfies for video stabilization?", "answer": ["Real-Time Selfie Video Stabilization"], "answer_arxiv_id": ["2009.02007"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_273"} +{"question": "Which work employs a token-wise conditional computation strategy to reduce the overall computation cost?", "answer": ["CoLT5: Faster Long-Range Transformers with Conditional Computation"], "answer_arxiv_id": ["2303.09752v3"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_274"} +{"question": "Which research work first utilized pre-trained vision-and-language transformers for navigating agents?", "answer": ["Improving Vision-and-Language Navigation with Image-Text Pairs from the Web"], "answer_arxiv_id": ["2004.14973"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_275"} +{"question": "Which works explored prompt learning for efficient and lightweight video understanding?", "answer": ["Prompting Visual-Language Models for Efficient Video Understanding"], "answer_arxiv_id": ["2112.04478"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_276"} +{"question": "Could you provide me some works that capture spatiotemporal dynamics directly from fMRI time series?", "answer": ["BolT: Fused Window Transformers for fMRI Time Series Analysis"], "answer_arxiv_id": ["2205.11578"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_277"} +{"question": "What works reframe prompts by decomposing complex task instructions into simpler ones?", "answer": ["Decomposed Prompting: A Modular Approach for Solving Complex Tasks"], "answer_arxiv_id": ["2210.02406"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_278"} +{"question": "Which paper introduced the robust ability of gating to generate long timescales and address exploding and vanishing gradients problem (EVGP)?", "answer": ["On the difficulty of training Recurrent Neural Networks", "Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation"], "answer_arxiv_id": ["1211.5063", "1406.1078"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_279"} +{"question": "What research papers propose or discuss watermarking techniques for LLM-generated texts?", "answer": ["Tracing Text Provenance via Context-Aware Lexical Substitution", "Frustratingly Easy Edit-based Linguistic Steganography with a Masked\n Language Model", "A Watermark for Large Language Models", "Undetectable Watermarks for Language Models", "Can AI-Generated Text be Reliably Detected?", "Three Bricks to Consolidate Watermarks for Large Language Models", "Provable Robust Watermarking for AI-Generated Text"], "answer_arxiv_id": ["2112.07873", "2104.09833", "2301.10226", "2306.09194", "2303.11156", "2308.00113", "2306.17439"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_280"} +{"question": "What studies used hand-crafted representations to transform raw event data into 2D grid-shaped feature maps", "answer": ["Event-based Vision meets Deep Learning on Steering Prediction for Self-driving Cars", "EV-FlowNet: Self-Supervised Optical Flow Estimation for Event-based Cameras", "Unsupervised Event-based Learning of Optical Flow, Depth, and Egomotion"], "answer_arxiv_id": ["1804.01310v1", "1802.06898", "1812.08156"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_281"} +{"question": "Which papers introduced regret-based procedures for finding correlated and coarse correlated equilibria in multiplayer games?", "answer": ["Beyond Time-Average Convergence: Near-Optimal Uncoupled Online Learning via Clairvoyant Multiplicative Weights Update"], "answer_arxiv_id": ["2111.14737"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_282"} +{"question": "What papers deal with the application of influence functions?", "answer": ["R"], "answer_arxiv_id": ["1210.6589"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_283"} +{"question": "What studies have done theoretical analysis on how transformers learn the spatial structure of image-type datasets?", "answer": ["Vision Transformers provably learn spatial structure"], "answer_arxiv_id": ["2210.09221"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_284"} +{"question": "Could you provide me some studies about the generalization properties in adaptive data analysis?", "answer": ["Preserving Statistical Validity in Adaptive Data Analysis", "Calibrating Noise to Variance in Adaptive Data Analysis", "Algorithmic Stability for Adaptive Data Analysis", "A New Analysis of Differential Privacy’s Generalization Guarantees", "Generalization in Adaptive Data Analysis and Holdout Reuse", "Generalization for Adaptively-chosen Estimators via Stable Median", "The Limits of Post-Selection Generalization"], "answer_arxiv_id": ["1411.2664", "1712.07196", "1511.02513", "1909.03577", "1506.02629", "1706.05069", "1806.06100"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_285"} +{"question": "Are there any examples of studies demonstrating the integration of neural solvers and classical solvers to reduce numerical error?", "answer": ["Learning data driven discretizations for partial differential equations", "Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers", "Machine learning accelerated computational fluid dynamics"], "answer_arxiv_id": ["1808.04930", "2007.00016", "2102.01010"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_286"} +{"question": "What works established the Atari 100K benchmark for evaluating sample-efficiency in RL?", "answer": ["Model Based Reinforcement Learning for Atari"], "answer_arxiv_id": ["1903.00374"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_287"} +{"question": "Which works mention that language models still require fine-tuning for specific downstream tasks?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning", "VIOLET : End-to-End Video-Language Transformers with Masked Visual-token Modeling", "Decoupling the Role of Data, Attention, and Losses in Multimodal Transformers", "Hero: Hierarchical Encoder for Video+Language Omni-representation Pre-training", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "FLAVA: A Foundational Language And Vision Alignment Model", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "VideoBERT: A Joint Model for Video and Language Representation Learning", "LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "UFO: A UniFied TransfOrmer for Vision-Language Representation Learning", "POUF: Prompt-oriented unsupervised fine-tuning for large pre-trained models"], "answer_arxiv_id": ["1909.11740", "2111.12681", "2102.00529", "2005.00200", "2004.06165", "1908.02265", "2112.04482", "1908.08530", "1904.01766", "1908.07490", "2111.10023", "2305.00350"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_288"} +{"question": "Which works focus on neural relational inference over temporal sequences in the context of Graph Neural Networks?", "answer": ["Neural Relational Inference for Interacting Systems", "Roto-translated Local Coordinate Frames For Interacting Dynamical Systems"], "answer_arxiv_id": ["1802.04687", "2110.14961"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_289"} +{"question": "Could you provide me with researches that set the programming problem in an RL framework using an actor-critic setup to debug programs?", "answer": ["CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning"], "answer_arxiv_id": ["2207.01780"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_290"} +{"question": "Is there a research that discusses the first efficient, low-regret algorithm for online nonstochastic control under the assumption that the system is controllable?", "answer": ["Black-Box Control for Linear Dynamical Systems"], "answer_arxiv_id": ["2007.06650"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_291"} +{"question": "Which papers studied the enhancement of the realism of predictions based on variational autoencoders?", "answer": ["Stochastic Variational Video Prediction", "MoCoGAN: Decomposing Motion and Content for Video Generation", "Stochastic Latent Residual Video Prediction"], "answer_arxiv_id": ["1710.11252", "1707.04993", "2002.09219"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_292"} +{"question": "Can you provide me with works that used MetaBBO-SR which uses black box optimizers at both the meta and low level to enhance optimization performance?", "answer": ["Discovering Evolution Strategies via Meta-Black-Box Optimization", "Discovering Attention-Based Genetic Algorithms via Meta-Black-Box Optimization"], "answer_arxiv_id": ["2211.11260", "2304.03995"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_293"} +{"question": "Could you provide me some studies that explored mixture models in Generative Adversarial Networks (GANs)?", "answer": ["Gaussian Mixture Generative Adversarial Networks for Diverse Datasets, and the Unsupervised Clustering of Images"], "answer_arxiv_id": ["1808.10356"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_294"} +{"question": "Can you provide papers that use the metric Fŕechet Inception Distance (FID) to assess the quality of generated images?", "answer": ["GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash\n Equilibrium"], "answer_arxiv_id": ["1706.08500"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_295"} +{"question": "Which research achieve excellent results in zero-shot image recognition and open-vocabulary object detection?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Simple Open-Vocabulary Object Detection with Vision Transformers"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2205.06230"], "source_meta": {"published_time": "20240412"}, "qid": "AutoScholarQuery_train_296"} +{"question": "What research has indicated that modern overparameterized models easily overfit when applied to improve the worst-group performance?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "An investigation of why overparameterization exacerbates spurious correlations"], "answer_arxiv_id": ["1911.08731", "2005.04345"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_297"} +{"question": "Any works about the significance of cut selection in modern MILP solvers?", "answer": ["Theoretical challenges towards cutting-plane selection", "Reinforcement Learning for Integer Programming: Learning to Cut"], "answer_arxiv_id": ["1805.02782", "1906.04859"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_298"} +{"question": "Could you provide me some works that examined the spectrum of CK with more general input data?", "answer": ["A Random Matrix Approach to Neural Networks", "Spectra of the Conjugate Kernel and Neural Tangent Kernel for Linear-Width Neural Networks"], "answer_arxiv_id": ["1702.05419", "2005.11879"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_299"} +{"question": "Can you provide me examples of research where language models are used for decision making in conversation models?", "answer": ["Improving alignment of dialogue agents via targeted human judgements", "A Simple Language Model for Task-Oriented Dialogue"], "answer_arxiv_id": ["2209.14375", "2005.00796"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_300"} +{"question": "What are the papers that worked on enhancing multi-step reasoning part of language models either by fine-tuning?", "answer": ["Solving Quantitative Reasoning Problems with Language Models"], "answer_arxiv_id": ["2206.14858"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_301"} +{"question": "Could you provide me some works where they design domain-specific languages executable on text?", "answer": ["Database Reasoning Over Text"], "answer_arxiv_id": ["2106.01074"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_302"} +{"question": "Could you provide me some examples of optimization-based methods in few-shot learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Piecewise classifier mappings: Learning fine-grained learners for novel categories with few examples"], "answer_arxiv_id": ["1703.03400", "1805.04288"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_303"} +{"question": "Which research papers develop and discuss the concepts of matrix factorization techniques?", "answer": ["Sparse Matrix Decompositions and Graph Characterizations"], "answer_arxiv_id": ["1111.6845"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_304"} +{"question": "Which works discuss about how transliteration substantially improves the performance of neural machine translation for low-resource languages?", "answer": ["A Universal Parent Model for Low-Resource Neural Machine Translation\n Transfer", "On Romanization for Model Transfer Between Scripts in Neural Machine\n Translation"], "answer_arxiv_id": ["1909.06516", "2009.14824"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_305"} +{"question": "Could you provide me some studies that apply the per-sample gradient clipping for good utility?", "answer": ["Deep Learning with Differential Privacy", "Large Language Models Can Be Strong Differentially Private Learners", "Large Scale Transfer Learning for Differentially Private Image Classification", "Toward Training at ImageNet Scale with Differential Privacy", "Unlocking High-Accuracy Differentially Private Image Classification through Scale"], "answer_arxiv_id": ["1607.00133", "2110.05679", "2205.02973", "2201.12328", "2204.13650"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_306"} +{"question": "Which works are related to using GNN for aggregating neighborhood information for each graph snapshot as a part of dynamic graph neural networks?", "answer": ["Discrete-time Temporal Network Embedding via Implicit Hierarchical Learning in Hyperbolic Space", "Hyperbolic Variational Graph Neural Network for Modeling Dynamic Graphs", "Variational Graph Recurrent Neural Networks", "Structured Sequence Modeling with Graph Convolutional Recurrent Networks"], "answer_arxiv_id": ["2107.03767", "2104.02228", "1908.09710", "1612.07659"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_307"} +{"question": "Could you suggest some references on structured pruning methods?", "answer": ["Arxiv"], "answer_arxiv_id": ["2004.12380"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_308"} +{"question": "Are there any studies that proposed a continuous 6D representation?", "answer": ["On the Continuity of Rotation Representations in Neural Networks"], "answer_arxiv_id": ["1812.07035"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_309"} +{"question": "Any studies that showcase the use of image prompts for object customization?", "answer": ["ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models"], "answer_arxiv_id": ["2302.13848", "2305.14720", "2308.06721"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_310"} +{"question": "Could you provide me some references about the surrogate training method in deep learning for SNNs?", "answer": ["Spatio-Temporal Backpropagation for Training High-performance Spiking Neural Networks", "SLAYER: Spike Layer Error Reassignment in Time"], "answer_arxiv_id": ["1706.02609", "1810.08646"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_311"} +{"question": "Can you name the paper that proved linear convergence of gradient descent for linear networks without bottlenecks?", "answer": ["A Convergence Analysis of Gradient Descent for Deep Linear Neural Networks"], "answer_arxiv_id": ["1810.02281"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_312"} +{"question": "What papers are about GAN inversion techniques?", "answer": ["Pivotal Tuning for Latent-based Editing of Real Images", "Learning Detailed Radiance Manifolds for High-Fidelity and 3D-Consistent\n Portrait Synthesis from Monocular Image"], "answer_arxiv_id": ["2106.05744", "2211.13901"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_313"} +{"question": "Could you provide me some works about improving feature representations to enhance algorithmic performance in reinforcement learning?", "answer": ["A Geometric Perspective on Optimal Representations for Reinforcement Learning"], "answer_arxiv_id": ["1901.11530"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_314"} +{"question": "What works have used contrastive learning in terms of molecular substructures to understand protein structure similarity and functionality?", "answer": ["Contrastive Representation Learning for 3D Protein Structures"], "answer_arxiv_id": ["2205.15675"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_315"} +{"question": "Could you provide me some papers which propose methods for calibrating the dynamics of the source domain in the context of online dynamics adaptation?", "answer": ["Closing the Sim-to-Real Loop: Adapting Simulation Randomization with Real World Experience", "BayesSim: adaptive domain randomization via probabilistic inference for robotics simulators", "Auto-Tuned Sim-to-Real Transfer", "Data-efficient Domain Randomization with Bayesian Optimization", "An Imitation from Observation Approach to Transfer Learning with Dynamics Mismatch", "Policy Transfer across Visual and Dynamics Domain Gaps via Iterative Grounding", "When to Trust Your Model: Model-Based Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning", "Off-Dynamics Reinforcement Learning: Training for Transfer with Domain Classifiers"], "answer_arxiv_id": ["1810.05687", "1906.01728", "2104.07662", "2003.02471", "2008.01594", "2107.00339", "1906.08253", "2005.05951", "2006.13916"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_316"} +{"question": "Could you mention any work that deals with the VT2VT summarization?", "answer": ["Align and Attend: Multimodal Summarization with Dual Contrastive Losses", "VideoXum: Cross-modal Visual and Textural Summarization of Videos"], "answer_arxiv_id": ["2303.07284", "2303.12060"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_317"} +{"question": "Could you provide a study that enhanced dynamic neural rendering by incorporating depth information?", "answer": ["T\\\"oRF: Time-of-Flight Radiance Fields for Dynamic Scene View Synthesis"], "answer_arxiv_id": ["2109.15271"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_318"} +{"question": "Which studies use an optimal transport kernel GP for Bayes Optimization over molecular graphs?", "answer": ["ChemBO: Bayesian Optimization of Small Organic Molecules with Synthesizable Recommendations"], "answer_arxiv_id": ["1908.01425v2"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_319"} +{"question": "Which work proposes to use depth for adapting segmentation models to new data domains?", "answer": ["Domain Adaptive Semantic Segmentation with Self-Supervised Depth\n Estimation"], "answer_arxiv_id": ["2104.13613"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_320"} +{"question": "What works does the researcher refer to regarding bounds for non-Lipschitz losses?", "answer": ["Fine-Grained Analysis of Stability and Generalization for Stochastic Gradient Descent"], "answer_arxiv_id": ["2006.08157"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_321"} +{"question": "Which works proposed 'target-aware' objectives leading to higher performance in tabular tasks?", "answer": ["Revisiting Pretraining Objectives for Tabular Deep Learning", "TransTab: Learning Transferable Tabular Transformers Across Tables"], "answer_arxiv_id": ["2207.03208", "2205.09328"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_322"} +{"question": "What studies propose learning good similarity metrics from data to overcome the challenge of choosing distance functions?", "answer": ["Two Simple Ways to Learn Individual Fairness Metrics from Data"], "answer_arxiv_id": ["2006.11439"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_323"} +{"question": "What papers talks about that adaptive algorithms can achieve order-optimal rates without knowing problem parameters in nonconvex optimization?", "answer": ["AdaGrad stepsizes: Sharp convergence over nonconvex landscapes", "UniXGrad: A Universal, Adaptive Algorithm with Optimal Guarantees for Constrained Optimization"], "answer_arxiv_id": ["1806.01811", "1910.13857"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_324"} +{"question": "What papers discuss methods to improve generation of rare concepts in text-to-image models?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Design Guidelines for Prompt Engineering Text-to-Image Generative Models", "A very preliminary analysis of DALL-E 2", "DiffusionDB: A Large-scale Prompt Gallery Dataset for Text-to-Image Generative Models"], "answer_arxiv_id": ["2105.05233", "2207.12598", "2112.10741", "2205.11487", "2109.06977", "2204.13807v2", "2210.14896"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_325"} +{"question": "What work proposes to finetune a GAN with an inverted latent code for personalized generation of images?", "answer": ["Pivotal Tuning for Latent-based Editing of Real Images"], "answer_arxiv_id": ["2106.05744"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_326"} +{"question": "What works propose neural animatable implicit representations to combine neural implicit representations with an explicit human template?", "answer": ["Neural Actor: Neural Free-view Synthesis of Human Actors with Pose Control", "H-NeRF: Neural Radiance Fields for Rendering and Temporal Reconstruction of Humans in Motion", "Neural Articulated Radiance Field", "Editable Free-Viewpoint Video using a Layered Neural Representation", "Neural Human Performer: Learning Generalizable Radiance Fields for Human Performance Rendering", "Neural Novel Actor: Learning a Generalized Animatable Neural Representation for Human Actors", "Structured Local Radiance Fields for Human Avatar Modeling", "ARAH: Animatable Volume Rendering of Articulated Human SDFs", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular Video", "TAVA: Template-free Animatable Volumetric Actors", "DANBO: Disentangled Articulated Neural Body Representations via Graph Neural Networks", "NeuMan: Neural Human Radiance Field from a Single Video", "Learning Motion-Dependent Appearance for High-Fidelity Rendering of Dynamic Humans from a Single Camera", "Neural Image-based Avatars: Generalizable Radiance Fields for Human Avatar Modeling"], "answer_arxiv_id": ["2106.02019", "2110.13746", "2104.03110", "2104.14786", "2109.07448", "2208.11905", "2203.14478", "2210.10036v1", "2201.04127", "2206.08929", "2205.01666", "2203.12575", "2203.12780", "2304.04897"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_327"} +{"question": "Can you tell me which papers have built Transformer agents trained at scale for robotic manipulation tasks?", "answer": ["VIMA: General Robot Manipulation with Multimodal Prompts", "RT-1: Robotics Transformer for Real-World Control at Scale"], "answer_arxiv_id": ["2210.03094", "2212.06817"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_328"} +{"question": "Could you provide me some studies that improve training signals for scaling in object-centric learning?", "answer": ["Illiterate DALL-E Learns to Compose", "Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos", "Object-Centric Slot Diffusion", "SlotDiffusion: Object-Centric Generative Modeling with Diffusion Models"], "answer_arxiv_id": ["2110.11405", "2205.14065", "2303.10834", "2305.11281"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_329"} +{"question": "Any works about controlling pretrained diffusion models with additional information?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_330"} +{"question": "What studies propose automatic proof generation with a neurosymbolic framework?", "answer": ["LINC: A Neurosymbolic Approach for Logical Reasoning by Combining Language Models with First-Order Logic Provers"], "answer_arxiv_id": ["2310.15164v2"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_331"} +{"question": "What papers applied GFlowNets for biological molecule and sequence design, causal structure learning, and robust combinatorial optimization?", "answer": ["Biological Sequence Design with GFlowNets", "Multi-Objective GFlowNets", "Bayesian Structure Learning with Generative Flow Networks", "Bayesian learning of Causal Structure and Mechanisms with GFlowNets and Variational Bayes", "Robust Scheduling with GFlowNets"], "answer_arxiv_id": ["2203.04115", "2210.12765", "2202.13903", "2211.02763", "2302.05446"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_332"} +{"question": "Which works mention that modern visual backbones provide consistent semantic correspondences in the feature space?", "answer": ["Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "Exploring Cross-Image Pixel Contrast for Semantic Segmentation", "Rethinking Semantic Segmentation: A Prototype View", "Point-Level Region Contrast for Object Detection Pre-Training"], "answer_arxiv_id": ["2203.08414", "2101.11939", "2203.15102", "2202.04639"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_333"} +{"question": "What works are in the category of Hierarchical RL methods where the language represents the hierarchy?", "answer": ["Hierarchical and Interpretable Skill Acquisition in Multi-task Reinforcement Learning", "Language as an Abstraction for Hierarchical Deep Reinforcement Learning", "A Persistent Spatial Semantic Representation for High-level Natural Language Instruction Execution", "Hierarchical Task Learning from Language Instructions with Unified Transformers and Self-Monitoring", "Fast and Precise: Adjusting Planning Horizon with Adaptive Subgoal Search"], "answer_arxiv_id": ["1712.07294", "1906.07343", "2107.05612", "2106.03427", "2206.00702"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_334"} +{"question": "What studies are about the automatic generation of prompts in large language models?", "answer": ["AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts", "Making Pre-trained Language Models Better Few-shot Learners", "Large Language Models are Human-Level Prompt Engineers"], "answer_arxiv_id": ["2010.15980", "2012.15723", "2211.01910"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_335"} +{"question": "Are there any references discussing the model’s learning of its own positional encoding with the use of causal masking?", "answer": ["Transformer Language Models without Positional Encodings Still Learn Positional Information"], "answer_arxiv_id": ["2203.16634"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_336"} +{"question": "What work introduced the Whac-A-Mole dilemma for multiple shortcuts?", "answer": ["A Whac-A-Mole Dilemma: Shortcuts Come in Multiples Where Mitigating One\n Amplifies Others"], "answer_arxiv_id": ["2212.04825"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_337"} +{"question": "Can you list some works about representation learning in egocentric videos, which is facilitated by the introduction of the Ego4d dataset?", "answer": ["Egocentric Video-Language Pretraining", "Learning Video Representations from Large Language Models", "Helping Hands: An Object-Aware Ego-Centric Video Recognition Model", "HierVL: Learning Hierarchical Video-Language Embeddings"], "answer_arxiv_id": ["2206.01670", "2212.04501v1", "2308.07918", "2301.02311"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_338"} +{"question": "Which study used a program interpreter in 'Program of Thoughts' method for CoT prompting?", "answer": ["Program of Thoughts Prompting: Disentangling Computation from Reasoning\n for Numerical Reasoning Tasks"], "answer_arxiv_id": ["2211.12588"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_339"} +{"question": "In what study was Edge of Stability (EoS) first formalized through empirical study?", "answer": ["Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability"], "answer_arxiv_id": ["2103.00065"], "source_meta": {"published_time": "20230709"}, "qid": "AutoScholarQuery_train_340"} +{"question": "Which studies utilize self-attention to gain both cross-modal and intra-modal attention?", "answer": ["Multi-modal Transformer for Video Retrieval", "Attention Bottlenecks for Multimodal Fusion", "Everything at Once – Multi-modal Fusion Transformer for Video Retrieval", "Multimodal Token Fusion for Vision Transformers"], "answer_arxiv_id": ["2007.10639", "2107.00135", "2112.04446", "2204.08721"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_341"} +{"question": "What work addresses the shortcomings of the serial computing characteristics of the recurrent structure by discarding the recurrent structure and employing convolutional layers?", "answer": ["Neural Machine Translation in Linear Time", "Convolutional Sequence to Sequence Learning"], "answer_arxiv_id": ["1610.10099", "1705.03122v3"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_342"} +{"question": "What works introduce cutout, a method that randomly masks out square regions within the image?", "answer": ["Improved Regularization of Convolutional Neural Networks with Cutout"], "answer_arxiv_id": ["1708.04552"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_343"} +{"question": "Could you list the studies that developed new frameworks for learning directly from geometric data after the emergence of robust mesh feature extractors?", "answer": ["Unsupervised Learning of Robust Spectral Shape Matching", "Learning Multi-resolution Functional Maps with Spectral Attention for\n Robust Shape Matching", "Unsupervised Deep Multi-Shape Matching", "Deep Orientation-Aware Functional Maps: Tackling Symmetry Issues in\n Shape Matching"], "answer_arxiv_id": ["2304.14419v1", "2210.06373", "2207.09610", "2204.13453"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_344"} +{"question": "Which work extends RWS by using filtering SMC to approximate posterior expectations instead of SNIS?", "answer": ["Neural Adaptive Sequential Monte Carlo"], "answer_arxiv_id": ["1506.03338"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_345"} +{"question": "Which research papers discuss constraining the learned policy to the behavior policy used to collect the dataset to address the distributional shift in offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "1906.00949", "1911.11361"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_346"} +{"question": "What are some of the original works on direct image synthesis in text-guided image generation?", "answer": ["Generative Adversarial Text to Image Synthesis", "Learning What and Where to Draw", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks", "StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks", "AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks", "Controllable Text-to-Image Generation", "ManiGAN: Text-Guided Image Manipulation", "MirrorGAN: Learning Text-to-image Generation by Redescription", "Cycle-Consistent Inverse GAN for Text-to-Image Synthesis"], "answer_arxiv_id": ["1605.05396", "1610.02454", "1612.03242", "1710.10916", "1711.10485", "1909.07083", "1912.06203", "1903.05854", "2108.01361"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_347"} +{"question": "What works utilize the approach of attention map for concept and style manipulation?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image Translation"], "answer_arxiv_id": ["2208.01626", "2211.12572"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_348"} +{"question": "Could you provide me some works showing the connections between FedAvg and Reptile?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "On First-Order Meta-Learning Algorithms"], "answer_arxiv_id": ["1602.05629", "1803.02999"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_349"} +{"question": "Could you provide me some studies that tested the idea of using local losses to increase potential for asynchrony in the training procedure?", "answer": ["Decoupled Greedy Learning of CNNs for Synchronous and Asynchronous Distributed Learning", "Greedy Layerwise Learning Can Scale to ImageNet", "Putting An End to End-to-End: Gradient-Isolated Learning of Representations", "Training Neural Networks with Local Error Signals", "Interlocking Backpropagation: Improving depthwise model-parallelism"], "answer_arxiv_id": ["2106.06401", "1812.11446", "1905.11786", "1901.06656", "2010.04116"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_350"} +{"question": "What studies discuss the differentiation of Delaunay triangulation by introducing weighting strategies?", "answer": ["Differentiable Surface Triangulation"], "answer_arxiv_id": ["2109.10695"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_351"} +{"question": "What works focus on improving LB?", "answer": ["Local Branching Relaxation Heuristics for Integer Linear Programs"], "answer_arxiv_id": ["2212.08183"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_352"} +{"question": "What are some references that discuss the use of convolutions in Generative Adversarial Networks (GANs)?", "answer": ["Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks"], "answer_arxiv_id": ["1511.06434"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_353"} +{"question": "Which papers provide sufficient conditions on a data distribution implying it is learnable by certain neural networks?", "answer": ["Globally Optimal Gradient Descent for a ConvNet with Gaussian Inputs", "Error bounds for approximations with deep ReLU networks", "A Provably Correct Algorithm for Deep Learning that Actually Works", "Improved Learning of One-hidden-layer Convolutional Neural Networks with Overlaps", "Gradient Descent Learns One-hidden-layer CNN: Don’t be Afraid of Spurious Local Minima", "End-to-end Learning of a Convolutional Neural Network via Deep Tensor Decomposition", "Exponential Convergence of the Deep Neural Network Approximation for Analytic Functions"], "answer_arxiv_id": ["1702.07966", "1610.01145", "1803.09522", "1805.07798", "1712.00779", "1805.06523", "1807.00297"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_354"} +{"question": "What research demonstrated that a restarted online gradient descent algorithm can achieve dynamic regret bound for convex function and bound for strongly convex function under noisy feedback?", "answer": ["Non-stationary Stochastic Optimization"], "answer_arxiv_id": ["1307.5449"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_355"} +{"question": "Could you provide me some references that analyzed how misspecified likelihood models affect the accuracy of the inferred quantity when performing Bayesian inference in the context of inverse problems?", "answer": ["On the Brittleness of Bayesian Inference", "On the Local Lipschitz Stability of Bayesian Inverse Problems"], "answer_arxiv_id": ["1308.6306", "1906.07120"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_356"} +{"question": "Which papers discuss about improving the memory efficiency of activations through 'Activation compressed training'?", "answer": ["ActNN: Reducing Training Memory Footprint via 2-Bit Activation Compressed Training"], "answer_arxiv_id": ["2104.14129"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_357"} +{"question": "Which papers described the synthetic approach of constructing hallucination datasets by purposely triggering Language Models to produce spurious responses?", "answer": ["HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large\n Language Models"], "answer_arxiv_id": ["2305.11747"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_358"} +{"question": "Could you provide me some datasets used for referring expression comprehension (REC)?", "answer": ["Modeling Context in Referring Expressions", "Generation and Comprehension of Unambiguous Object Descriptions", "Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations", "PhraseCut: Language-based Image Segmentation in the Wild"], "answer_arxiv_id": ["1608.00272", "1511.02283", "1602.07332", "2008.01187"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_359"} +{"question": "What works are related to 'in-context learning' in large language models?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_360"} +{"question": "What works utilize models with unimodal encoders followed by cross-attention fusion?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BridgeTower: Building Bridges Between Encoders in Vision-Language\n Representation Learning", "Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual\n Concepts", "mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal\n Skip-connections", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "mPLUG-2: A Modularized Multi-modal Foundation Model Across Text, Image\n and Video", "X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks"], "answer_arxiv_id": ["2201.12086", "2206.08657", "2111.08276", "2205.12005", "2107.07651", "2302.00402", "2211.12402"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_361"} +{"question": "Which papers addressed subgraph matching by only modeling graph structure?", "answer": ["R"], "answer_arxiv_id": ["1210.6589"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_362"} +{"question": "Can you provide references where significant improvements in self-supervision on large indoor and outdoor datasets using contrastive self-supervision is discussed?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Self-Supervised Pretraining of 3D Features on any Point-Cloud", "ProposalContrast: Unsupervised Pre-training for LiDAR-based 3D Object\n Detection"], "answer_arxiv_id": ["2007.10985", "2101.02691", "2207.12654"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_363"} +{"question": "Any works on addressing the information leakage problem in CBMs?", "answer": ["Promises and Pitfalls of Black-Box Concept Learning Models", "Do Concept Bottleneck Models Learn As Intended?"], "answer_arxiv_id": ["2106.13314", "2105.04289"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_364"} +{"question": "Which research works extended empowerment to high-dimensional image space by using a non-parametric nearest neighbor to estimate entropy?", "answer": ["Behavior From the Void: Unsupervised Active Pre-Training", "Reinforcement Learning with Prototypical Representations", "State Entropy Maximization with Random Encoders for Efficient Exploration"], "answer_arxiv_id": ["2103.04551", "2102.11271v2", "2102.09430"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_365"} +{"question": "Could you provide me some papers about designing loss functions and manipulating gradients to improve the optimization process of multi-task learning?", "answer": ["Learning Multiple Dense Prediction Tasks from Partially Annotated Data", "Contrastive Multi-Task Dense Prediction", "Auto-Lambda: Disentangling Dynamic Task Relationships", "Robust Learning Through Cross-Task Consistency", "Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics", "Conflict-Averse Gradient Descent for Multi-task Learning", "Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign\n Dropout", "GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep\n Multitask Networks", "Gradient Vaccine: Investigating and Improving Multi-task Optimization in\n Massively Multilingual Models", "Gradient Surgery for Multi-Task Learning"], "answer_arxiv_id": ["2111.14893", "2307.07934", "2202.03091v2", "2006.04096", "1705.07115v3", "2110.14048", "2010.06808", "1711.02257", "2010.05874", "2001.06782"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_366"} +{"question": "Which studies attempted to recover long-term statistics by learning a discrete-time stochastic reduced-order system?", "answer": ["Data-based stochastic model reduction for the Kuramoto–Sivashinsky equation"], "answer_arxiv_id": ["1509.09279"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_367"} +{"question": "Which papers looked into the steganography issue of CycleGAN and proposed solutions?", "answer": ["CycleGAN, a Master of Steganography", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "Geometry-Consistent Generative Adversarial Networks for One-Sided\n Unsupervised Domain Mapping", "A study of the effect of JPG compression on adversarial images", "Exploring Patch-wise Semantic Relation for Contrastive Learning in\n Image-to-Image Translation Tasks", "FUN-SIS: a Fully UNsupervised approach for Surgical Instrument\n Segmentation"], "answer_arxiv_id": ["1712.02950", "1703.10593", "1809.05852", "1608.00853", "2203.01532", "2202.08141"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_368"} +{"question": "What works introduced reactiveness and decodability as conditions that promote sample-efficient learning of POMDPs?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "Provable Reinforcement Learning with a Short-Term Memory"], "answer_arxiv_id": ["1610.09512v2", "2202.03983"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_369"} +{"question": "Could you provide me some examples of research but focused on generating layouts using Transformer-based methods?", "answer": ["LayoutTransformer: Layout Generation and Completion with Self-attention", "BLT: Bidirectional Layout Transformer for Controllable Layout Generation"], "answer_arxiv_id": ["2006.14615", "2112.05112"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_370"} +{"question": "Can you provide some studies highlighting the issues with the implementation of deep learning frameworks in State Space Models?", "answer": ["TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems", "MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems"], "answer_arxiv_id": ["1603.04467v2", "1512.01274"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_371"} +{"question": "Which papers attempted modeling Super Resolution (SR) in the wavelet domain?", "answer": ["Perception-Distortion Balanced ADMM Optimization for Single-Image\n Super-Resolution"], "answer_arxiv_id": ["2208.03324"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_372"} +{"question": "Which studies have based their work on supervised learning methods in document-level EAE?", "answer": ["Prompt for Extraction? PAIE: Prompting Argument Interaction for Event\n Argument Extraction", "Document-level Event Extraction via Heterogeneous Graph-based\n Interaction Model with a Tracker", "Multi-Sentence Argument Linking", "Document-Level Event Role Filler Extraction using Multi-Granularity\n Contextualized Encoding"], "answer_arxiv_id": ["2202.12109", "2105.14924", "1911.03766", "2005.06579"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_373"} +{"question": "Any works about utilizing an approach related to Noether’s theorem for relating continuous symmetries to dynamics of conserved quantities?", "answer": ["Noether’s Learning Dynamics: Role of Symmetry Breaking in Neural Networks"], "answer_arxiv_id": ["2105.02716"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_374"} +{"question": "Which works proposed to use meshes in 3D representation?", "answer": ["AtlasNet: A Papier-M\\^ach\\'e Approach to Learning 3D Surface Generation", "Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images"], "answer_arxiv_id": ["1802.05384", "1804.01654"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_375"} +{"question": "Which papers are related to the building of large multimodal models based on large language models?", "answer": ["The All-Seeing Project: Towards Panoptic Visual Recognition and\n Understanding of the Open World", "MultiModal-GPT: A Vision and Language Model for Dialogue with Humans", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "BuboGPT: Enabling Visual Grounding in Multi-Modal LLMs", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Ferret: Refer and Ground Anything Anywhere at Any Granularity", "Position-Enhanced Visual Instruction Tuning for Multimodal Large\n Language Models", "Kosmos-2: Grounding Multimodal Large Language Models to the World", "What Matters in Training a GPT4-Style Language Model with Multimodal\n Inputs?", "Visual Instruction Tuning with Polite Flamingo", "Sparkles: Unlocking Chats Across Multiple Images for Multimodal\n Instruction-Following Models", "SVIT: Scaling up Visual Instruction Tuning", "DeepSpeed-VisualChat: Multi-Round Multi-Image Interleave Chat via Multi-Modal Causal Attention"], "answer_arxiv_id": ["2308.01907", "2305.04790", "2305.03726", "2304.14178", "2307.08581", "2304.15010", "2209.06794", "2307.03601", "2306.15195", "2310.07704", "2308.13437", "2306.14824", "2307.02469", "2307.01003", "2308.16463", "2307.04087", "2309.14327v3"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_376"} +{"question": "Which works focus on evaluating the performance of prediction models on real-world datasets?", "answer": ["Human Motion Trajectory Prediction: A Survey"], "answer_arxiv_id": ["1905.06113"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_377"} +{"question": "Which methods have been proposed for learning to solve RPM-like problems?", "answer": ["Measuring abstract reasoning in neural networks", "Improving Generalization for Abstract Reasoning Tasks Using Disentangled Feature Representations", "Are Disentangled Representations Helpful for Abstract Visual Reasoning?", "Learning Perceptual Inference by Contrasting", "Abstract Reasoning with Distracting Features", "Abstract Diagrammatic Reasoning with Multiplex Graph Networks", "The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning", "Scale-Localized Abstract Reasoning", "Stratified Rule-Aware Network for Abstract Visual Reasoning", "Effective Abstract Reasoning with Dual-Contrast Network"], "answer_arxiv_id": ["1807.04225", "1811.04784", "1905.12506", "1912.00086", "1912.00569", "2006.11197", "2007.04212", "2009.09405", "2002.06838", "2205.13720"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_378"} +{"question": "Could you provide me some works about bandit algorithms for hierarchical models?", "answer": ["Meta-Thompson Sampling", "No Regrets for Learning the Prior in Bandits", "Bayesian decision-making under misspecified priors with applications to meta-learning", "Metadata-based Multi-Task Bandits with Bayesian Hierarchical Models", "Hierarchical Bayesian Bandits", "Metalearning Linear Bandits by Prior Update"], "answer_arxiv_id": ["2102.06129", "2107.06196", "2107.01509", "2108.06422v1", "2111.06929", "2107.05320v2"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_379"} +{"question": "What papers introduced the concept of 3D Semantic Scene Completion (SSC)?", "answer": ["Semantic Scene Completion from a Single Depth Image"], "answer_arxiv_id": ["1611.08974"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_380"} +{"question": "What are the works that use transformers in reinforcement learning?", "answer": ["A Survey on Transformers in Reinforcement Learning", "Stabilizing Transformers for Reinforcement Learning", "Transformers are Meta-Reinforcement Learners"], "answer_arxiv_id": ["2301.03044", "1910.06764", "2206.06614"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_381"} +{"question": "What works study the emergence of behaviors not directly specified by the objective of the task?", "answer": ["Emergent Complexity via Multi-Agent Competition", "Emergent Tool Use From Multi-Agent Autocurricula", "The AI Economist: Improving Equality and Productivity with AI-Driven Tax Policies", "Open-Ended Learning Leads to Generally Capable Agents", "Human-Timescale Adaptation in an Open-Ended Task Space"], "answer_arxiv_id": ["1710.03748", "1909.07528v2", "2004.13332", "2107.12808", "2301.07608"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_382"} +{"question": "What works achieved impressive self-supervised performance and robustness in monocular depth estimation?", "answer": ["Unsupervised Monocular Depth Estimation with Left-Right Consistency", "Digging Into Self-Supervised Monocular Depth Estimation"], "answer_arxiv_id": ["1609.03677", "1806.01260"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_383"} +{"question": "In what papers is causal reasoning included in broad benchmarks for language understanding?", "answer": ["SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems"], "answer_arxiv_id": ["1905.00537"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_384"} +{"question": "Can you name the works that study private mean estimation for distributions with bounded moments?", "answer": ["Privacy and Statistical Risk: Formalisms and Minimax Bounds", "Private Mean Estimation of Heavy-Tailed Distributions", "Robust and differentially private mean estimation", "Propose, Test, Release: Differentially private estimation with high probability"], "answer_arxiv_id": ["1412.4451", "2002.09464", "2102.09159", "2002.08774"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_385"} +{"question": "Which works propose applying differentiable augmentation to the generator and discriminator of GANs to avoid expensive data collection?", "answer": ["Differentiable Augmentation for Data-Efficient GAN Training"], "answer_arxiv_id": ["2006.10738"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_386"} +{"question": "Which research papers have proposed deep-learning based point-cloud descriptors?", "answer": ["The Perfect Match: 3D Point Cloud Matching with Smoothed Densities", "Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors", "3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions"], "answer_arxiv_id": ["1811.06879", "2202.11660", "1603.08182"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_387"} +{"question": "What works attempted to improve the CTC model’s prediction by conditioning on previously generated tokens in the non-autoregressive framework?", "answer": ["Mask CTC: Non-Autoregressive End-to-End ASR with CTC and Mask Predict", "Imputer: Sequence Modelling via Imputation and Dynamic Programming"], "answer_arxiv_id": ["2005.08700", "2002.08926"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_388"} +{"question": "Which research papers discuss scaling up Posterior Sampling for Reinforcement Learning to non-tabular settings?", "answer": ["Model-based Reinforcement Learning for Continuous Control with Posterior Sampling"], "answer_arxiv_id": ["2012.09613"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_389"} +{"question": "Which study found that LLMs can ground high-level tasks to a set of actionable steps in structured synthetic environments?", "answer": ["Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents"], "answer_arxiv_id": ["2201.07207"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_390"} +{"question": "Could you provide me some efforts done to represent corpora and models for underrepresented languages?", "answer": ["SERENGETI: Massively Multilingual Language Models for Africa", "IndicTrans2: Towards High-Quality and Accessible Machine Translation Models for all 22 Scheduled Indian Languages"], "answer_arxiv_id": ["2212.10785", "2305.16307v3"], "source_meta": {"published_time": "20230909"}, "qid": "AutoScholarQuery_train_391"} +{"question": "Which studies propose neural heuristics for MOCOPs that adopt a neighborhood-based parameter-transfer strategy?", "answer": ["MODRL/D-AM: Multiobjective Deep Reinforcement Learning Algorithm Using Decomposition and Attention Model for Multiobjective Optimization"], "answer_arxiv_id": ["2002.05484"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_392"} +{"question": "Could you offer some references on designing learning/bandit algorithms to optimize reserve prices in auctions?", "answer": ["Dynamic Reserve Prices for Repeated Auctions: Learning from Bids", "Learning Simple Auctions", "Learning Algorithms for Second-Price Auctions with Reserve", "Learning Multi-item Auctions with (or without) Samples", "Incentive-aware Contextual Pricing with Non-parametric Market Noise", "Dynamic Incentive-aware Learning: Robust Pricing in Contextual Auctions"], "answer_arxiv_id": ["2002.07331v1", "1604.03171", "1310.5665", "1709.00228", "1911.03508v3", "2002.11137v1"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_393"} +{"question": "What research works have been made to apply diffusion models in image generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2006.11239", "2011.13456", "2105.05233"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_394"} +{"question": "Do there exist any works that follow intervention-based RL where neither student nor teacher policy fully control the agent?", "answer": ["DisCoRL: Continual Reinforcement Learning via Policy Distillation", "Policy Distillation"], "answer_arxiv_id": ["1907.05855", "1511.06295"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_395"} +{"question": "What dataset introduced enhanced annotations by replacing bounding boxes with object segmentation masks?", "answer": ["Panoptic Scene Graph Generation"], "answer_arxiv_id": ["2207.11247"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_396"} +{"question": "What researches exist around FQE methods that extrapolate policy returns from approximated Q-functions?", "answer": ["Bootstrapping Fitted Q-Evaluation for Off-Policy Inference", "Batch Policy Learning under Constraints", "Statistical Bootstrapping for Uncertainty Estimation in Off-Policy Evaluation"], "answer_arxiv_id": ["2102.03607", "1903.08738", "2007.13609"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_397"} +{"question": "What are the major research papers discuss the trainability issue of a randomly initialization network in the context of network pruning?", "answer": ["A Signal Propagation Perspective for Pruning Neural Networks at Initialization", "A Gradient Flow Framework For Analyzing Network Pruning", "Connectivity Matters: Neural Network Pruning Through the Lens of Effective Sparsity"], "answer_arxiv_id": ["1906.06307", "2009.11839", "2107.02306"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_398"} +{"question": "Can you provide an example of a work that modified RLHF method using pure RL for training LLMs with human feedback in an online manner?", "answer": ["Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback"], "answer_arxiv_id": ["2204.05862"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_399"} +{"question": "Which studies try to augment the LLM with external information for question-answering and conversational tasks?", "answer": ["Check Your Facts and Try Again: Improving Large Language Models with External Knowledge and Automated Feedback", "Toolformer: Language Models Can Teach Themselves to Use Tools", "Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questions", "Knowledge-in-Context: Towards Knowledgeable Semi-Parametric Language Models"], "answer_arxiv_id": ["2302.12813", "2302.04761", "2212.10509", "2210.16433"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_400"} +{"question": "Which works discuss disentangled representation learning in the context of improving downstream transfer?", "answer": ["Representation Learning: A Review and New Perspectives", "Recent Advances in Autoencoder-Based Representation Learning", "Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations", "Open World Compositional Zero-Shot Learning", "Revisiting Training Strategies and Generalization Performance in Deep Metric Learning", "Disentanglement and Generalization Under Correlation Shifts"], "answer_arxiv_id": ["1206.5538", "1812.05069", "1811.12359", "2101.12609", "2002.08473", "2112.14754"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_401"} +{"question": "Who initially proposed the concept of Video language grounding?", "answer": ["Localizing Moments in Video with Natural Language"], "answer_arxiv_id": ["1708.01641"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_402"} +{"question": "Could you provide me some recent studies that focus on adding randomness to training data through generative models such as GANs and diffusion models?", "answer": ["Improving Robustness using Generated Data", "Better Diffusion Models Further Improve Adversarial Training"], "answer_arxiv_id": ["2110.09468", "2302.04638"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_403"} +{"question": "Which study proposes the use of learnable filters for adaptive mining of frequency forgery clues?", "answer": ["Thinking in Frequency: Face Forgery Detection by Mining Frequency-aware Clues"], "answer_arxiv_id": ["2007.09355"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_404"} +{"question": "Are there any research papers that explored zero-shot image classification using CLIP?", "answer": ["Learning to Prompt for Vision-Language Models", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "AutoCLIP: Auto-tuning Zero-Shot Classifiers for Vision-Language Models", "CHiLS: Zero-Shot Image Classification with Hierarchical Label Sets"], "answer_arxiv_id": ["2109.01134", "2110.04544", "2309.16414", "2302.02551"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_405"} +{"question": "Could you provide me some works that have achieved excellent object segmentation accuracy in identifying 3D objects from complex scenes?", "answer": ["3D Bounding Box Estimation Using Deep Learning and Geometry", "VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds"], "answer_arxiv_id": ["1612.00496", "1711.06396", "1906.01140"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_406"} +{"question": "Could you provide me some research that utilizes Linear Blending Skinning for data-driven 3D pose transfer?", "answer": ["Skeleton-Aware Networks for Deep Motion Retargeting", "Skeleton-free Pose Transfer for Stylized 3D Characters"], "answer_arxiv_id": ["2005.05732", "2208.00790"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_407"} +{"question": "Are there any works that attempt to build benchmarks for evaluating various detectors under different datasets?", "answer": ["Towards Benchmarking and Evaluating Deepfake Detection"], "answer_arxiv_id": ["2203.02115"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_408"} +{"question": "Any works related to k-server problem have involved analysis of switching costs?", "answer": ["Online Optimization with Predictions and Non-convex Losses"], "answer_arxiv_id": ["1911.03827"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_409"} +{"question": "What studies utilized transformer models in object detection tasks?", "answer": ["End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2005.12872", "2010.04159"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_410"} +{"question": "Which papers discuss the estimation of CATE for an intervention using structured information?", "answer": ["GraphITE: Estimating Individual Effects of Graph-structured Treatments", "Causal Effect Inference for Structured Treatments"], "answer_arxiv_id": ["2009.14061", "2106.01939"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_411"} +{"question": "Are there any papers about the variant of the split CP called Conformal Quantile Regression (CQR)?", "answer": ["Conformalized Quantile Regression"], "answer_arxiv_id": ["1905.03222"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_412"} +{"question": "Could you provide me some studies about object-centric learning for unsupervised semantic segmentation?", "answer": ["Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "DeepCut: Unsupervised Segmentation using Graph Neural Networks Clustering", "FreeSOLO: Learning to Segment Objects without Annotations"], "answer_arxiv_id": ["2203.08414", "2212.05853", "2202.12181v2"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_413"} +{"question": "Could you provide me with the studies that work on adjusting the loss influence across different classes?", "answer": ["Equalization Loss for Long-Tailed Object Recognition", "Equalization Loss v2: A New Gradient Balance Approach for Long-tailed\n Object Detection", "Dealing with Cross-Task Class Discrimination in Online Continual\n Learning"], "answer_arxiv_id": ["2003.05176", "2012.08548", "2305.14657"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_414"} +{"question": "Which studies used Kullback-Leibler divergence and Task2Vec to represent asymmetric distances between datasets?", "answer": ["Task2Vec: Task Embedding for Meta-Learning"], "answer_arxiv_id": ["1902.03545"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_415"} +{"question": "Which works extend the capabilities of INR to handle images and facilitate interactive virtual scene editing and content creation?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "Decomposing NeRF for Editing via Feature Field Distillation"], "answer_arxiv_id": ["2012.02190", "2205.15585"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_416"} +{"question": "What works proposed deterministic and stochastic algorithms for nonconvex-concave minimax problems?", "answer": ["Solving a Class of Non-Convex Min-Max Games Using Iterative First Order Methods", "Efficient Algorithms for Smooth Minimax Optimization", "An accelerated inexact proximal point method for solving nonconvex-concave min-max problems"], "answer_arxiv_id": ["1902.08297", "1907.01543", "1905.13433"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_417"} +{"question": "What are some recent test-time adaptation methods?", "answer": ["Tent: Fully Test-Time Adaptation by Entropy Minimization", "Generalization on Unseen Domains via Inference-time Label-Preserving Target Projections", "Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "Learning to Generalize across Domains on Single Test Samples", "Test-time Fourier Style Calibration for Domain Generalization"], "answer_arxiv_id": ["2006.10726", "2103.01134", "1909.13231", "2202.08045", "2205.06427"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_418"} +{"question": "What papers suggest optimizing both calibration and accuracy in neural networks by adding a calibration regularizer?", "answer": ["Soft Calibration Objectives for Neural Networks"], "answer_arxiv_id": ["2108.00106"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_419"} +{"question": "Can you cite works relevant to the domain-invariant training type of domain adaptation?", "answer": ["Return of Frustratingly Easy Domain Adaptation", "Deep CORAL: Correlation Alignment for Deep Domain Adaptation", "Deep Domain Confusion: Maximizing for Domain Invariance", "Learning Transferable Features with Deep Adaptation Networks", "Domain-Adversarial Training of Neural Networks", "Conditional Adversarial Domain Adaptation", "Adversarial Discriminative Domain Adaptation", "Generate To Adapt: Aligning Domains using Generative Adversarial Networks"], "answer_arxiv_id": ["1511.05547", "1607.01719", "1412.3474", "1502.02791", "1505.07818", "1705.10667", "1702.05464", "1704.01705"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_420"} +{"question": "Which works focus on the concept of identifiability in causal representation learning?", "answer": ["Variational Autoencoders and Nonlinear ICA: A Unifying Framework", "Underspecification Presents Challenges for Credibility in Modern Machine Learning", "Posterior Collapse and Latent Variable Non-identifiability"], "answer_arxiv_id": ["1907.04809", "2011.03395", "2301.00537"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_421"} +{"question": "Could you provide me the research that devises a stochastic parameter-free method that acquires high probability convergence guarantee and optimal rate for noiseless convex and smooth objective?", "answer": ["Making SGD Parameter-Free"], "answer_arxiv_id": ["2205.02160"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_422"} +{"question": "Which recent works use Transformer to represent policies in modular RL, aiming to overcome the problem of message-passing in complex morphologies?", "answer": ["My Body is a Cage: the Role of Morphology in Graph-Based Incompatible Control", "MetaMorph: Learning Universal Controllers with Transformers", "Low-Rank Modular Reinforcement Learning via Muscle Synergy"], "answer_arxiv_id": ["2010.01856", "2203.11931", "2210.15479"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_423"} +{"question": "What prior studies have used Generative Adversarial Networks (GANs) in the field of text-to-image generation?", "answer": ["Generative Adversarial Networks", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked\n Generative Adversarial Networks", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "MirrorGAN: Learning Text-to-image Generation by Redescription"], "answer_arxiv_id": ["2203.00667", "1612.03242", "1711.10485", "1903.05854"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_424"} +{"question": "What research papers have used adaption of Code-LLMs for program repair?", "answer": ["Unified Pre-training for Program Understanding and Generation", "CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generation"], "answer_arxiv_id": ["2103.06333", "2109.00859"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_425"} +{"question": "What papers have contributed to building more accurate models of sketch classification using deep-learning architectures, particularly Convolutional Neural Networks and Recurrent Neural Networks and have solved challenges including partial sketch classification and sketch progression incorporation?", "answer": ["A Neural Representation of Sketch Drawings"], "answer_arxiv_id": ["1704.03477v4"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_426"} +{"question": "Could you list me some papers that mention the concept of 'relative representations'?", "answer": ["Do better ImageNet classifiers assess perceptual similarity better?", "Human alignment of neural network representations", "Harmonizing the object recognition strategies of deep neural networks with humans", "Relative representations enable zero-shot latent space communication"], "answer_arxiv_id": ["2203.04946", "2211.01201", "2211.04533", "2209.15430"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_427"} +{"question": "Could you name the popular real-world video datasets for which TAP-Vid has released annotations?", "answer": ["The 2017 DAVIS Challenge on Video Object Segmentation", "The Kinetics Human Action Video Dataset"], "answer_arxiv_id": ["1704.00675", "1705.06950"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_428"} +{"question": "What studies proposed maximal mutual information-based methods for intrinsic behavioral learning?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function", "Fast Task Inference with Variational Intrinsic Successor Features", "APS: Active Pretraining with Successor Features", "Dynamics-Aware Unsupervised Discovery of Skills"], "answer_arxiv_id": ["1802.06070", "1906.05030", "2108.13956", "1907.01657"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_429"} +{"question": "Can you provide me some research that was focused on the generation of range-azimuth-Doppler (RAD) tensor?", "answer": ["CARRADA Dataset: Camera and Automotive Radar with Range-Angle-Doppler Annotations", "RADDet: Range-Azimuth-Doppler based Radar Object Detection for Dynamic Road Users"], "answer_arxiv_id": ["2005.01456", "2105.00363"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_430"} +{"question": "Which works propose modeling correlations between 2D keypoints and 3D hand and object poses using cross attention?", "answer": ["Keypoint Transformer: Solving Joint Identification in Challenging Hands\n and Object Interactions for Accurate 3D Pose Estimation"], "answer_arxiv_id": ["2104.14639"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_431"} +{"question": "Which works have tried to emulate symbolic reasoning through a differentiable component in neuro-symbolic learning?", "answer": ["Neural Arithmetic Logic Units", "Stochastic Optimization of Sorting Networks via Continuous Relaxations", "SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver", "Enforcing Policy Feasibility Constraints through Differentiable Projection for Energy Optimization"], "answer_arxiv_id": ["1808.00508", "1903.08850", "1905.12149", "2105.08881"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_432"} +{"question": "What papers proposed methodologies to decode the intensity, density or the cumulative hazard function from the history in neural TPPs?", "answer": ["Intensity-Free Learning of Temporal Point Processes", "Deep Mixture Point Processes: Spatio-temporal Event Prediction with Rich Contextual Information", "UNIPoint: Universally Approximating Point Processes Intensities", "CAUSE: Learning Granger Causality from Event Sequences using Attribution Methods", "Fully Neural Network based Model for General Temporal Point Processes", "Exploring Generative Neural Temporal Point Process"], "answer_arxiv_id": ["1909.12127", "1906.08952", "2007.14082", "2002.07906", "1905.09690", "2208.01874"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_433"} +{"question": "What research works have used pre-collected data generated by behavior policy for training in offline RL methods?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning", "Off-Policy Deep Reinforcement Learning without Exploration"], "answer_arxiv_id": ["2006.04779", "2110.06169", "1812.02900"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_434"} +{"question": "Could you provide me some papers that relate the optimal learning rate to the effective width of neural networks?", "answer": ["The Effect of Network Width on Stochastic Gradient Descent and Generalization: an Empirical Study"], "answer_arxiv_id": ["1905.03776"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_435"} +{"question": "What papers used curiosity or model error to provide novelty in exploration bonuses?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction", "Self-Supervised Exploration via Disagreement", "Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "VIME: Variational Information Maximizing Exploration", "Surprise-Based Intrinsic Motivation for Deep Reinforcement Learning", "MIMEx: Intrinsic Rewards from Masked Input Modeling"], "answer_arxiv_id": ["1705.05363", "1906.04161", "1507.00814", "1605.09674", "1703.01732", "2305.08932"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_436"} +{"question": "What are the studies where Transformers are introduced for the vision community?", "answer": ["Attention Is All You Need", "Non-local Neural Networks", "Stand-Alone Self-Attention in Vision Models", "Local Relation Networks for Image Recognition"], "answer_arxiv_id": ["1706.03762", "1711.07971", "1906.05909", "1904.11491"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_437"} +{"question": "What studies have enriched source code in a graph-structured manner for machine learning?", "answer": ["Learning to Represent Programs with Graphs", "GraphCodeBERT: Pre-training Code Representations with Data Flow"], "answer_arxiv_id": ["1711.00740", "2009.08366"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_438"} +{"question": "Any works about convergence analyses on other variants of PG?", "answer": ["On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift", "On the Global Convergence Rates of Softmax Policy Gradient Methods"], "answer_arxiv_id": ["1908.00261", "2005.06392"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_439"} +{"question": "Can you name the research that criticizes the idea of defining or measuring unlearning success based on indistinguishability from retrain-from-scratch?", "answer": ["On the Necessity of Auditable Algorithmic Definitions for Machine Unlearning"], "answer_arxiv_id": ["2110.11891v2"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_440"} +{"question": "Could you provide the paper where the researchers found that adding a gradient norm of the loss function can help the optimizer find flat local minima?", "answer": ["Penalizing Gradient Norm for Efficiently Improving Generalization in Deep Learning"], "answer_arxiv_id": ["2202.03599"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_441"} +{"question": "Can you provide references that analyze the optimal complexity of Stochastic Gradient Descent in the stochastic setting?", "answer": ["Stochastic First- and Zeroth-order Methods for Nonconvex Stochastic Programming", "Lower Bounds for Non-Convex Stochastic Optimization"], "answer_arxiv_id": ["1309.5549", "1912.02365"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_442"} +{"question": "Which research work used autoencoder to reduce pixel-level redundancy in image synthesis?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_443"} +{"question": "What works have considered using generative training data for image classification and pre-training?", "answer": ["Is synthetic data from generative models ready for image recognition?", "Fake it till you make it: Learning transferable representations from synthetic ImageNet clones"], "answer_arxiv_id": ["2210.07574", "2212.08420v2"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_444"} +{"question": "Which studies present a reward-transition distribution shift induced by task-dependent data collection as a challenge in offline meta-RL?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables"], "answer_arxiv_id": ["1703.03400", "1903.08254"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_445"} +{"question": "What research works have been done to improve the training speed on policy gradient methods?", "answer": ["Asynchronous Methods for Deep Reinforcement Learning", "Reinforcement Learning through Asynchronous Advantage Actor-Critic on a GPU", "IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures", "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference", "Efficient Parallel Methods for Deep Reinforcement Learning", "Emergence of Locomotion Behaviours in Rich Environments", "DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion Frames"], "answer_arxiv_id": ["1602.01783", "1611.06256", "1802.01561", "1910.06591", "1705.04862", "1707.02286", "1911.00357"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_446"} +{"question": "Which study proposed a quality focal loss as a joint representation of the IoU score and classification score?", "answer": ["Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection"], "answer_arxiv_id": ["2006.04388"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_447"} +{"question": "What are the works on representation learning in reinforcement learning?", "answer": ["Time-Contrastive Networks: Self-Supervised Learning from Video", "Deep Reinforcement and InfoMax Learning", "Curiosity-driven Exploration by Self-supervised Prediction", "Dream to Control: Learning Behaviors by Latent Imagination", "Learning Representations for Pixel-based Control: What Matters and Why?"], "answer_arxiv_id": ["1704.06888", "2006.07217", "1705.05363", "1912.01603", "2111.07775"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_448"} +{"question": "Which models blend a vision encoder with a resampler and a cross-gated attention layer for vision-text cross-attention and instruction optimization?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2204.14198", "2305.03726"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_449"} +{"question": "Which work utilizes Slot Attention with Neural Radiance Fields for disentangling object representations?", "answer": ["Unsupervised Discovery of Object Radiance Fields", "Decomposing 3D Scenes into Objects via Unsupervised Volume Segmentation"], "answer_arxiv_id": ["2107.07905", "2104.01148"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_450"} +{"question": "What research studies have been inspired by the right to be forgotten in machine learning literature?", "answer": ["Making AI Forget You: Data Deletion in Machine Learning", "Formalizing Data Deletion in the Context of the Right to be Forgotten", "Certified Data Removal from Machine Learning Models", "Machine Unlearning", "Approximate Data Deletion from Machine Learning Models", "Descent-to-Delete: Gradient-Based Methods for Machine Unlearning"], "answer_arxiv_id": ["1907.05012", "2002.10635", "1911.03030", "1912.03817v3", "2002.10077", "2007.02923"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_451"} +{"question": "What works have used the Constructive Solid Geometry (CSG) representation for reconstructing CAD shapes with program synthesis?", "answer": ["CSGNet: Neural Shape Parser for Constructive Solid Geometry", "Write, Execute, Assess: Program Synthesis with a REPL", "Learning to Infer and Execute 3D Shape Programs"], "answer_arxiv_id": ["1712.08290", "1906.04604", "1901.02875"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_452"} +{"question": "Could you point out some works that delve into the concept of transfer learning in reinforcement learning?", "answer": ["Transfer Learning in Deep Reinforcement Learning: A Survey"], "answer_arxiv_id": ["2009.07888"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_453"} +{"question": "What studies have learned general representations of touch through self-supervision?", "answer": ["Touch and Go: Learning from Human-Collected Vision and Touch", "Self-Supervised Visuo-Tactile Pretraining to Locate and Follow Garment\n Features"], "answer_arxiv_id": ["2211.12498", "2209.13042"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_454"} +{"question": "Could you provide me with papers that proposed to replace the computational oracle with more tractable estimators?", "answer": ["A Free Lunch from the Noise: Provable and Practical Exploration for Representation Learning", "Making Linear MDPs Practical via Contrastive Representation Learning", "Contrastive UCB: Provably Efficient Contrastive Self-Supervised Learning in Online Reinforcement Learning"], "answer_arxiv_id": ["2111.11485", "2207.07150", "2207.14800"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_455"} +{"question": "Could you tell me about the works on abstractive summarization?", "answer": ["BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension", "PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive\n Summarization"], "answer_arxiv_id": ["1910.13461", "1912.08777"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_456"} +{"question": "Could you mention some works that use machine learning to estimate algebraic multigrid (AMG) parameters?", "answer": ["Learning Algebraic Multigrid Using Graph Neural Networks", "Learning to Optimize Multigrid PDE Solvers"], "answer_arxiv_id": ["2003.05744", "1902.10248"], "source_meta": {"published_time": "20220522"}, "qid": "AutoScholarQuery_train_457"} +{"question": "Can you name some works which studied the trade-off between local DP and accuracy?", "answer": ["Privacy Aware Learning", "RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response", "Collecting Telemetry Data Privately"], "answer_arxiv_id": ["1210.2085", "1407.6981", "1712.01524"], "source_meta": {"published_time": "20230911"}, "qid": "AutoScholarQuery_train_458"} +{"question": "Could you provide me with the studies using 3D tokenizer to quantize videos in various video generation tasks?", "answer": ["MAGVIT: Masked Generative Video Transformer"], "answer_arxiv_id": ["2212.05199"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_459"} +{"question": "What research proposed to learn a non-linear network that combines all losses into a single coherent objective function?", "answer": ["Auxiliary Learning by Implicit Differentiation"], "answer_arxiv_id": ["2007.02693"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_460"} +{"question": "Can you mention some work that has been done on data augmentation-based methods in unsupervised anomaly detectors?", "answer": ["CutPaste: Self-Supervised Learning for Anomaly Detection and\n Localization"], "answer_arxiv_id": ["2104.04015"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_461"} +{"question": "Which papers explored RAG for image synthesis?", "answer": ["Text-Guided Synthesis of Artistic Images with Retrieval-Augmented\n Diffusion Models", "ReMoDiffuse: Retrieval-Augmented Motion Diffusion Model", "Re-Imagen: Retrieval-Augmented Text-to-Image Generator", "Retrieval-Augmented Multimodal Language Modeling", "KNN-Diffusion: Image Generation via Large-Scale Retrieval", "Label-Retrieval-Augmented Diffusion Models for Learning from Noisy\n Labels"], "answer_arxiv_id": ["2207.13038", "2304.01116", "2209.14491", "2211.12561", "2204.02849", "2305.19518"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_462"} +{"question": "What studies involve parameter-factorization-based methods in the concept of PEFT?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "DyLoRA: Parameter Efficient Tuning of Pre-trained Models using Dynamic\n Search-Free Low-Rank Adaptation", "One-for-All: Generalized LoRA for Parameter-Efficient Fine-tuning"], "answer_arxiv_id": ["2106.09685", "2210.07558", "2306.07967"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_463"} +{"question": "Which research work proposed the first MAB algorithm robust to corruptions with a regret C times worse than regret in the stochastic setting?", "answer": ["Stochastic bandits robust to adversarial corruptions"], "answer_arxiv_id": ["1803.09353"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_464"} +{"question": "Could you provide me some studies that leveraged large-scale pretrained vision-language models for improving open-vocabulary performance?", "answer": ["Open-Vocabulary Object Detection Using Captions", "Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "RegionCLIP: Region-based Language-Image Pretraining", "Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model", "Detecting Twenty-thousand Classes using Image-level Supervision"], "answer_arxiv_id": ["2011.10678", "2104.13921", "2112.09106", "2203.14940", "2201.02605"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_465"} +{"question": "What papers propose methods of frame or segment-level classification to generate proposals for temporal action detection?", "answer": ["CDC: Convolutional-De-Convolutional Networks for Precise Temporal Action\n Localization in Untrimmed Videos", "Temporal Activity Detection in Untrimmed Videos with Recurrent Neural\n Networks", "Temporal Gaussian Mixture Layer for Videos", "Proposal-Free Temporal Action Detection via Global Segmentation Mask\n Learning"], "answer_arxiv_id": ["1703.01515", "1608.08128", "1803.06316", "2207.06580"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_466"} +{"question": "Could you provide me some studies that use gradient surgery or task weighting to reduce gradient conflicts across tasks?", "answer": ["Gradient Surgery for Multi-Task Learning", "Gradient Vaccine: Investigating and Improving Multi-task Optimization in Massively Multilingual Models", "RotoGrad: Gradient Homogenization in Multitask Learning"], "answer_arxiv_id": ["2001.06782", "2010.05874", "2103.02631"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_467"} +{"question": "What works first used adversarial attacks on machine-generated text detectors?", "answer": ["Red Teaming Language Model Detectors with Language Models"], "answer_arxiv_id": ["2305.19713"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_468"} +{"question": "What research has introduced the concept of black-box attacks, where the attacker has no knowledge of the underlying model?", "answer": ["Sentence Embedding Leaks More Information than You Expect: Generative\n Embedding Inversion Attack to Recover the Whole Sentence"], "answer_arxiv_id": ["2305.03010"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_469"} +{"question": "What works proposed the popular approach knowledge graph embedding (KGE) in Knowledge graph completion?", "answer": ["Complex Embeddings for Simple Link Prediction", "Convolutional 2D Knowledge Graph Embeddings", "RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space", "TuckER: Tensor Factorization for Knowledge Graph Completion", "Low-Dimensional Hyperbolic Knowledge Graph Embeddings", "Modeling Heterogeneous Hierarchies with Relation-specific Hyperbolic Cones"], "answer_arxiv_id": ["1606.06357", "1707.01476", "1902.10197", "1901.09590", "2005.00545", "2110.14923"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_470"} +{"question": "Which studies show that pretrained code models can perform well in code competitions that are challenging to programmers?", "answer": ["PaLM: Scaling Language Modeling with Pathways", "Evaluating Large Language Models Trained on Code", "Program Synthesis with Large Language Models", "Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2204.02311", "2107.03374", "2108.07732", "2203.07814"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_471"} +{"question": "Which paper proposes to finetune StyleGAN using around 100 face images to obtain a personalized generative prior?", "answer": ["MyStyle: A Personalized Generative Prior"], "answer_arxiv_id": ["2203.17272"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_472"} +{"question": "Which research endeavored to develop negative-example-free methods in the context of general contrastive learning?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2006.07733", "2011.10566"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_473"} +{"question": "Can you name the study that reported Elo scores that appear to scale as a log of environment frames for humanoid agents playing football?", "answer": ["From Motor Control to Team Play in Simulated Humanoid Football"], "answer_arxiv_id": ["2105.12196"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_474"} +{"question": "Could you provide me some works about ensemble methods in order to boost performances by combining multiple networks?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Calibrated ensembles can mitigate accuracy tradeoffs under distribution shift", "Invariant Training 2D-3D Joint Hard Samples for Few-Shot Point Cloud Recognition", "Robust fine-tuning of zero-shot models", "Averaging Weights Leads to Wider Optima and Better Generalization", "Distilling the Knowledge in a Neural Network", "Ensemble Distillation for Robust Model Fusion in Federated Learning"], "answer_arxiv_id": ["1612.01474", "2207.08977", "2308.09694", "2109.01903", "1803.05407", "1503.02531", "2006.07242"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_475"} +{"question": "What research has proposed a 3D heatmap volume representation to utilize a powerful 3D-CNN model?", "answer": ["Revisiting Skeleton-based Action Recognition"], "answer_arxiv_id": ["2104.13586"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_476"} +{"question": "What papers focus on deterministic verification of geometric robustness of neural networks and remarked that they are exceedingly slow?", "answer": ["Towards Verifying Robustness of Neural Networks Against A Family of Semantic Perturbations"], "answer_arxiv_id": ["1912.09533"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_477"} +{"question": "Could you provide me some studies focused on learning a mapping between function spaces, often referred as the neural operator?", "answer": ["Neural Operator: Learning Maps Between Function Spaces", "Clifford Neural Layers for PDE Modeling", "Neural Operator: Graph Kernel Network for Partial Differential Equations", "Lie Point Symmetry Data Augmentation for Neural PDE Solvers", "Fourier Neural Operator for Parametric Partial Differential Equations", "DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators", "MIONet: Learning multiple-input operators via tensor product", "Multiwavelet-based Operator Learning for Differential Equations", "Choose a Transformer: Fourier or Galerkin", "GNOT: A General Neural Operator Transformer for Operator Learning", "ViTO: Vision Transformer-Operator", "A comprehensive and fair comparison of two neural operators (with practical extensions) based on FAIR data"], "answer_arxiv_id": ["2108.08481v6", "2209.04934", "2003.03485", "2202.07643", "2010.08895", "1910.03193", "2202.06137", "2109.13459", "2105.14995", "2302.14376", "2303.08891", "2111.05512"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_478"} +{"question": "Which papers presented the distributed Information Bottleneck approach for interpretability?", "answer": ["The Distributed Information Bottleneck reveals the explanatory structure of complex systems"], "answer_arxiv_id": ["2204.07576"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_479"} +{"question": "Can you mention some works that generate soft labels using the output of trained models?", "answer": ["Truth Discovery in Sequence Labels from Crowds"], "answer_arxiv_id": ["2109.04470"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_480"} +{"question": "Could you provide me some works that combined Local SGD with other methods to improve generalization?", "answer": ["Don’t Use Large Mini-Batches, Use Local SGD", "Extrapolation for Large-batch Training in Deep Learning"], "answer_arxiv_id": ["1808.07217", "2006.05720"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_481"} +{"question": "What works have been done on low-rank compression of weight matrices in Natural Language Processing (NLP) transformers?", "answer": ["Exploring extreme parameter compression for pre-trained language models", "Language model compression with weighted low-rank factorization"], "answer_arxiv_id": ["2205.10036", "2207.00112v1"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_482"} +{"question": "Are there any studies on mitigating popularity bias for distillation that is specific to recommender systems?", "answer": ["Unbiased Knowledge Distillation for Recommendation"], "answer_arxiv_id": ["2211.14729"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_483"} +{"question": "What research has been carried out on compressing the generator network for generative tasks?", "answer": ["Information-Theoretic GAN Compression with Variational Energy-based\n Model", "Discriminator-Cooperated Feature Map Distillation for GAN Compression", "Online Multi-Granularity Distillation for GAN Compression", "Revisiting Discriminator in GAN Compression: A Generator-discriminator\n Cooperative Compression Scheme", "Learning Efficient GANs for Image Translation via Differentiable Masks\n and co-Attention Distillation"], "answer_arxiv_id": ["2303.16050", "2212.14169", "2108.06908", "2110.14439", "2011.08382"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_484"} +{"question": "Which papers have used a homeomorphic Variational Autoencoder (VAE) to perform tasks in an unsupervised manner?", "answer": ["Explorations in Homeomorphic Variational Auto-Encoding"], "answer_arxiv_id": ["1807.04689"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_485"} +{"question": "What works employed episodic training regime in meta-learning systems?", "answer": ["Matching Networks for One Shot Learning"], "answer_arxiv_id": ["1606.04080"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_486"} +{"question": "Who proposed using physics informed neural networks or PINNs for nonlinear hyperbolic conservation laws?", "answer": ["wPINNs: Weak Physics informed neural networks for approximating entropy solutions of hyperbolic conservation laws"], "answer_arxiv_id": ["2207.08483"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_487"} +{"question": "Which research papers addresses the field of using cached attention for better understanding of long contexts?", "answer": ["Memorizing Transformers", "Augmenting Language Models with Long-Term Memory"], "answer_arxiv_id": ["2203.08913", "2306.07174"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_488"} +{"question": "Could you provide me some studies where the student network has arbitrary width and both its layers are trained?", "answer": ["Over-Parameterization Exponentially Slows Down Gradient Descent for Learning a Single Neuron"], "answer_arxiv_id": ["2302.10034"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_489"} +{"question": "What is the study that redefined the offline RL problem as a context-conditioned sequential problem, following the UDRL framework?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Offline Reinforcement Learning as One Big Sequence Modeling Problem"], "answer_arxiv_id": ["2106.01345", "2106.02039"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_490"} +{"question": "What works considered ELBOs defined by Hamiltonian Monte Carlo?", "answer": ["Energy-Inspired Models: Learning with Sampler-Induced Distributions"], "answer_arxiv_id": ["1910.14265"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_491"} +{"question": "What papers deal with the concept of filtration learning in a supervised way for graph data?", "answer": ["Graph Filtration Learning", "Topological Graph Neural Networks", "GEFL: Extended Filtration Learning for Graph Classification"], "answer_arxiv_id": ["1905.10996", "2102.07835v4", "2406.02732v1"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_492"} +{"question": "What papers establish regret guarantees for Thompson sampling algorithms?", "answer": ["Analysis of Thompson Sampling for the multi-armed bandit problem", "Thompson Sampling: An Asymptotically Optimal Finite Time Analysis", "Finite-Time Regret of Thompson Sampling Algorithms for Exponential Family Multi-Armed Bandits"], "answer_arxiv_id": ["1111.1797", "1205.4217", "2206.03520"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_493"} +{"question": "What research has been conducted on cooperative communication among agents?", "answer": ["Multi-Agent Cooperation and the Emergence of (Natural) Language", "Learning Multiagent Communication with Backpropagation", "Multiagent Bidirectionally-Coordinated Nets: Emergence of Human-level Coordination in Learning to Play StarCraft Combat Games"], "answer_arxiv_id": ["1612.07182", "1605.07736", "1703.10069v4"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_494"} +{"question": "What studies are about balancing task-related losses to overcome the limitations of multi-task learning?", "answer": ["Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics", "GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep\n Multitask Networks", "Gradient Adversarial Training of Neural Networks"], "answer_arxiv_id": ["1705.07115v3", "1711.02257", "1806.08028"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_495"} +{"question": "What research concentrated on studying artifacts in the frequency domain?", "answer": ["Detecting and Simulating Artifacts in GAN Fake Images"], "answer_arxiv_id": ["1907.06515"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_496"} +{"question": "What studies have discussed the applicaiton of spectral graph theory in machine learning?", "answer": ["SpectralNet: Spectral Clustering using Deep Neural Networks"], "answer_arxiv_id": ["1801.01587"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_497"} +{"question": "Can you name the studies that showed improvements using supervised contrastive frameworks?", "answer": ["Locality and Compositionality in Zero-Shot Learning", "Supervised Contrastive Learning"], "answer_arxiv_id": ["1912.12179", "2004.11362"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_498"} +{"question": "Could you provide me some studies about architecture-based methods to incorporate new classes of data in incremental learning?", "answer": ["Lifelong Learning with Dynamically Expandable Networks", "Progressive Neural Networks", "DER: Dynamically Expandable Representation for Class Incremental Learning", "Incremental Learning via Rate Reduction"], "answer_arxiv_id": ["1708.01547", "1606.04671", "2103.16788", "2011.14593"], "source_meta": {"published_time": "20220211"}, "qid": "AutoScholarQuery_train_499"} +{"question": "Could you name the research works that locate the most relevant 3D target objects in a raw point cloud scene given by the query text descriptions?", "answer": ["ScanRefer: 3D Object Localization in RGB-D Scans using Natural Language", "InstanceRefer: Cooperative Holistic Understanding for Visual Grounding\n on Point Clouds through Instance Multi-level Contextual Referring", "Free-form Description Guided 3D Visual Graph Network for Object\n Grounding in Point Cloud"], "answer_arxiv_id": ["1912.08830", "2103.01128", "2103.16381"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_500"} +{"question": "What RL framework minimizes the complexity of requirements and avoids additional abstractions?", "answer": ["CleanRL: High-quality Single-file Implementations of Deep Reinforcement Learning Algorithms"], "answer_arxiv_id": ["2111.08819"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_501"} +{"question": "Could you provide the references that first formalized the extractive approach in legal summarization?", "answer": ["EUR-Lex-Sum: A Multi- and Cross-lingual Dataset for Long-form\n Summarization in the Legal Domain"], "answer_arxiv_id": ["2210.13448"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_502"} +{"question": "Which papers were considered as recent advances in Transformer-based LLMs?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_503"} +{"question": "Which works propose the Gumbel-Softmax gradient estimator?", "answer": ["Categorical Reparameterization with Gumbel-Softmax", "The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables"], "answer_arxiv_id": ["1611.01144", "1611.00712"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_504"} +{"question": "Are there any methods that have tried to solve the inconsistency problem when matching multiple views?", "answer": ["ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformer"], "answer_arxiv_id": ["2208.14201"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_505"} +{"question": "In which papers did the researchers propose to cluster program surface forms?", "answer": ["Natural Language to Code Translation with Execution", "Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2204.11454", "2203.07814"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_506"} +{"question": "What works discuss model-based tools such as n-gram language models perplexities and ML classifier-based approaches?", "answer": ["CCNet: Extracting High Quality Monolingual Datasets from Web Crawl Data", "The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1911.00359", "2303.03915", "2005.14165"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_507"} +{"question": "Could you provide some works that use graph neural networks to aggregate information for retrosynthetic planning?", "answer": ["RetroGraph: Retrosynthetic Planning with Graph Search"], "answer_arxiv_id": ["2206.11477"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_508"} +{"question": "What works discuss the existence of adversarial examples from a data feature perspective?", "answer": ["Adversarial Examples Are Not Bugs, They Are Features"], "answer_arxiv_id": ["1905.02175"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_509"} +{"question": "What studies use deep transformers for inter-modal interactions in Vision-Language models?", "answer": ["ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks"], "answer_arxiv_id": ["2102.03334", "2201.12086", "2208.10442"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_510"} +{"question": "Can you list some studies that focused on scaling the Transformers for video foundation models to utilize the flexibility for multi-modal tasks?", "answer": ["MERLOT: Multimodal Neural Script Knowledge Models", "VideoMAE: Masked Autoencoders are Data-Efficient Learners for\n Self-Supervised Video Pre-Training", "VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking", "Unmasked Teacher: Towards Training-Efficient Video Foundation Models", "InternVideo: General Video Foundation Models via Generative and\n Discriminative Learning"], "answer_arxiv_id": ["2106.02636", "2203.12602", "2303.16727v2", "2303.16058", "2212.03191"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_511"} +{"question": "What are some studies about using large kernels in Global Convolution Networks and Local Relation Networks?", "answer": ["Large Kernel Matters -- Improve Semantic Segmentation by Global\n Convolutional Network", "Local Relation Networks for Image Recognition"], "answer_arxiv_id": ["1703.02719", "1904.11491"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_512"} +{"question": "Which papers have studied label memorization in machine learning theory?", "answer": ["A Closer Look at Memorization in Deep Networks", "Does Learning Require Memorization? A Short Tale about a Long Tail", "What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation"], "answer_arxiv_id": ["1706.05394", "1906.05271", "2008.03703"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_513"} +{"question": "Which studies indicated a critical learning regime or important phase in neural network training?", "answer": ["Critical Learning Periods in Deep Networks", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Gradient Descent Happens in a Tiny Subspace"], "answer_arxiv_id": ["1711.08856", "1803.03635", "1812.04754"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_514"} +{"question": "Which works introduced the technique of variance reduction in the finite-sum setting for accelerating convex optimization?", "answer": ["A Stochastic Gradient Method with an Exponential Convergence Rate for Finite Training Sets", "Stochastic Dual Coordinate Ascent Methods for Regularized Loss Minimization", "Optimization with First-Order Surrogate Functions", "SAGA: A Fast Incremental Gradient Method With Support for Non-Strongly Convex Composite Objectives"], "answer_arxiv_id": ["1202.6258", "1209.1873", "1305.3120v1", "1407.0202"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_515"} +{"question": "What are the documents tied to the notion of 'OOD' or 'anomaly'?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Deep Anomaly Detection with Outlier Exposure"], "answer_arxiv_id": ["1610.02136", "1812.04606"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_516"} +{"question": "Which study adopts self-supervised contrastive learning to deal with unlabeled data collected from edge devices?", "answer": ["Divergence-aware Federated Self-Supervised Learning"], "answer_arxiv_id": ["2204.04385"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_517"} +{"question": "Are there literature reviews on DRO?", "answer": ["Data-Driven Distributionally Robust Optimization Using the Wasserstein Metric: Performance Guarantees and Tractable Reformulations"], "answer_arxiv_id": ["1505.05116"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_518"} +{"question": "Which papers proposed techniques that are based on low-rank approximation of the kernel matrix, like Nyström method?", "answer": ["Revisiting the Nystrom Method for Improved Large-Scale Machine Learning", "Scalable Kernel K-Means Clustering with Nyström Approximation: Relative-Error Bounds"], "answer_arxiv_id": ["1303.1849v2", "1706.02803"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_519"} +{"question": "Which papers discuss obtaining representations that mimic the forward and/or backward passes within neural architectures?", "answer": ["A Generic Graph-based Neural Architecture Encoding Scheme for\n Predictor-based NAS"], "answer_arxiv_id": ["2004.01899"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_520"} +{"question": "What collected high-quality data for Supervised Fine Tuning of Large Language Models?", "answer": ["Learning From Mistakes Makes LLM Better Reasoner", "MetaMath: Bootstrap Your Own Mathematical Questions for Large Language\n Models", "MAmmoTH: Building Math Generalist Models through Hybrid Instruction\n Tuning", "WizardMath: Empowering Mathematical Reasoning for Large Language Models\n via Reinforced Evol-Instruct"], "answer_arxiv_id": ["2310.20689", "2309.12284", "2309.05653", "2308.09583"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_521"} +{"question": "Which paper first proposed the task of 3D part assembly?", "answer": ["Generative 3D Part Assembly via Dynamic Graph Learning"], "answer_arxiv_id": ["2006.07793"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_522"} +{"question": "Which study introduced self-critical training, a policy gradient method that baselines the REINFORCE gradient estimator?", "answer": ["Self-critical Sequence Training for Image Captioning"], "answer_arxiv_id": ["1612.00563"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_523"} +{"question": "Can you list some studies that utilize the neighboring-based methods in contrastive learning?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations", "Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction"], "answer_arxiv_id": ["2006.09882", "2104.14548", "2102.10106"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_524"} +{"question": "Could you list some works in the field of Language concept bottleneck models (CBM)?", "answer": ["Language in a Bottle: Language Model Guided Concept Bottlenecks for\n Interpretable Image Classification", "Label-Free Concept Bottleneck Models", "Visual Classification via Description from Large Language Models", "Post-hoc Concept Bottleneck Models", "Learning Concise and Descriptive Attributes for Visual Recognition", "Do Vision-Language Pretrained Models Learn Composable Primitive\n Concepts?"], "answer_arxiv_id": ["2211.11158", "2304.06129", "2210.07183", "2205.15480", "2308.03685", "2203.17271"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_525"} +{"question": "Which research papers focused on zero-shot methods for image-caption matching tasks?", "answer": ["CPT: Colorful Prompt Tuning for Pre-trained Vision-Language Models", "ReCLIP: A Strong Zero-Shot Baseline for Referring Expression Comprehension", "When Does Label Smoothing Help?"], "answer_arxiv_id": ["2109.11797", "2204.05991", "1906.02629"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_526"} +{"question": "What works propose the optimization of the non-rigid deformation upon SMPL from a monocular RGB or RGB-D video?", "answer": ["Video Based Reconstruction of 3D People Models", "Detailed Human Avatars from Monocular Video", "Dynamic Surface Function Networks for Clothed Human Bodies", "LaplacianFusion: Detailed 3D Clothed-Human Body Reconstruction"], "answer_arxiv_id": ["1803.04758", "1808.01338", "2104.03978", "2302.14251"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_527"} +{"question": "What research efforts have used residual block in SR methods?", "answer": ["Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network", "Enhanced Deep Residual Networks for Single Image Super-Resolution"], "answer_arxiv_id": ["1609.04802", "1707.02921"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_528"} +{"question": "Could you provide me some works about transformer adoption?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Learning Transferable Visual Models From Natural Language Supervision", "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2103.14030", "2103.00020", "2401.14159"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_529"} +{"question": "What papers discuss worst-case regret bounds for bandits with bounded rewards?", "answer": ["Kullback–Leibler upper confidence bounds for optimal sequential allocation", "A minimax and asymptotically optimal algorithm for stochastic bandits", "A minimax and asymptotically optimal algorithm for stochastic bandits", "KL-UCB-Switch: Optimal Regret Bounds for Stochastic Bandits from Both a Distribution-Dependent and a Distribution-Free Viewpoints", "Further Optimal Regret Bounds for Thompson Sampling", "Finite-Time Regret of Thompson Sampling Algorithms for Exponential Family Multi-Armed Bandits"], "answer_arxiv_id": ["1210.1136", "1702.07211", "1702.07211", "1805.05071", "1209.3353", "2206.03520"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_530"} +{"question": "What studies are examples of using open-ended questions as a part of objective evaluations?", "answer": ["TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for\n Reading Comprehension"], "answer_arxiv_id": ["1705.03551"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_531"} +{"question": "Which papers proposed more complex objectives than the generally used evidence lower bound (ELBO) in variational inference methods?", "answer": ["Importance Weighted Autoencoders", "Importance Weighting and Variational Inference", "Tighter Variational Bounds are Not Necessarily Better", "The Thermodynamic Variational Objective"], "answer_arxiv_id": ["1509.00519", "1808.09034", "1802.04537", "1907.00031"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_532"} +{"question": "Which research papers focus on auto-labeling solutions?", "answer": ["Data Programming: Creating Large Training Sets, Quickly", "Snorkel: Rapid Training Data Creation with Weak Supervision", "Fast and Three-rious: Speeding Up Weak Supervision with Triplet Methods"], "answer_arxiv_id": ["1605.07723", "1711.10160", "2002.11955"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_533"} +{"question": "Which papers contributed in the generation of text-controlled images using Text-to-Image (T2I) diffusion models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "LAION-5B: An open large-scale dataset for training next generation\n image-text models"], "answer_arxiv_id": ["2112.10741", "2205.11487", "2210.08402"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_534"} +{"question": "What work provides inspiration for U-FaTE through the use of closed-form solver and a universal dependence measure?", "answer": ["On Characterizing the Trade-off in Invariant Representation Learning"], "answer_arxiv_id": ["2109.03386"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_535"} +{"question": "What studies introduced non-autoregressive (NAR) transformers in generative models?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "Structured Denoising Diffusion Models in Discrete State-Spaces"], "answer_arxiv_id": ["2106.08254", "2111.06377", "2107.03006v3"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_536"} +{"question": "What studies examined memory retention capabilities of large language models?", "answer": ["Extracting Training Data from Large Language Models", "Quantifying Memorization Across Neural Language Models", "Counterfactual Memorization in Neural Language Models", "How BPE Affects Memorization in Transformers", "Understanding Unintended Memorization in Federated Learning", "The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks", "Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Models"], "answer_arxiv_id": ["2012.07805", "2202.07646", "2112.12938", "2110.02782", "2006.07490", "1802.08232", "2205.10770"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_537"} +{"question": "What papers introduced the use of a patch discriminator to help the generator learn details?", "answer": ["Precomputed Real-Time Texture Synthesis with Markovian Generative Adversarial Networks", "Image-to-Image Translation with Conditional Adversarial Networks", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks"], "answer_arxiv_id": ["1604.04382", "1611.07004", "1703.10593"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_538"} +{"question": "Are there recent methods addressing the computational limitations of Shapley-based data valuation ", "answer": ["LAVA: Data Valuation without Pre-Specified Learning Algorithms"], "answer_arxiv_id": ["2305.00054"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_539"} +{"question": "In what studies does the researcher refine the holistic visual features by learning a visual-semantic projection in generative ZSL?", "answer": ["Multi-modal Cycle-consistent Generalized Zero-Shot Learning", "Latent Embedding Feedback and Discriminative Features for Zero-Shot Classification", "FREE: Feature Refinement for Generalized Zero-Shot Learning"], "answer_arxiv_id": ["1808.00136", "2003.07833", "2107.13807"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_540"} +{"question": "Can you mention some studies that examined bias in ROC story cloze?", "answer": ["The Effect of Different Writing Tasks on Linguistic Style: A Case Study\n of the ROC Story Cloze Task"], "answer_arxiv_id": ["1702.01841"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_541"} +{"question": "What works are related to the improvement of the Neural Compression technique by reduced encoding times through meta-learning?", "answer": ["Implicit Neural Representations for Image Compression", "COIN++: Neural Compression Across Modalities", "Meta-Learning Sparse Compression Networks"], "answer_arxiv_id": ["2112.04267v2", "2201.12904", "2205.08957"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_542"} +{"question": "What recent works have discussed diffusion models and their effectiveness in image generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Generative Modeling by Estimating Gradients of the Data Distribution", "Improved Techniques for Training Score-Based Generative Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Implicit Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "Variational Diffusion Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "1907.05600", "2006.09011", "2011.13456", "2010.02502", "2206.00927", "2107.00630", "2206.00364", "2204.06125", "2112.10752"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_543"} +{"question": "Could you provide me some works discussing issues like over-smoothing and over-squashing in graph neural networks?", "answer": ["Graph Neural Networks Exponentially Lose Expressive Power for Node Classification", "A Note on Over-Smoothing for Graph Neural Networks", "Revisiting Graph Neural Networks: All We Have is Low-Pass Filters", "Understanding over-squashing and bottlenecks on graphs via curvature"], "answer_arxiv_id": ["1905.10947", "2006.13318", "1905.09550", "2111.14522"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_544"} +{"question": "Which researches investigated model behavior by learning differentiable masks over the input?", "answer": ["How do Decisions Emerge across Layers in Neural Models? Interpretation\n with Differentiable Masking"], "answer_arxiv_id": ["2004.14992"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_545"} +{"question": "Which paper provides theoretical guarantees of a small gap depending on the convexity along the training trajectory in the context of implicit regularization?", "answer": ["Continuous vs. Discrete Optimization of Deep Neural Networks"], "answer_arxiv_id": ["2107.06608"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_546"} +{"question": "Which works utilized instruction tuning for pre-trained multimodal large language models?", "answer": ["Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2304.08485", "2305.06500", "2305.03726"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_547"} +{"question": "What works deal with applying token-based generative transformer models to audio?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation", "High Fidelity Neural Audio Compression"], "answer_arxiv_id": ["2209.03143", "2210.13438"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_548"} +{"question": "What papers contributed to the development of earlier-generation Vision-and-Language (VL) models using rigorous and extensive pretraining with contrastive losses?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "GIT: A Generative Image-to-text Transformer for Vision and Language", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal\n Skip-connections", "TAP: Text-Aware Pre-training for Text-VQA and Text-Caption"], "answer_arxiv_id": ["2201.12086", "2205.14100", "2107.07651", "2202.03052", "2205.12005", "2012.04638"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_549"} +{"question": "What studies addressed the issue of data scarcity in 3D foundation models?", "answer": ["LERF: Language Embedded Radiance Fields", "OpenScene: 3D Scene Understanding with Open Vocabularies", "PLA: Language-Driven Open-Vocabulary 3D Scene Understanding", "Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Models", "CLIP-FO3D: Learning Free Open-world 3D Scene Representations from 2D Dense CLIP", "PartSLIP: Low-Shot Part Segmentation for 3D Point Clouds via Pretrained Image-Language Models", "RegionPLC: Regional Point-Language Contrastive Learning for Open-World 3D Scene Understanding", "ConceptFusion: Open-set Multimodal 3D Mapping"], "answer_arxiv_id": ["2303.09553", "2211.15654", "2211.16312", "2207.11514", "2303.04748", "2212.01558", "2304.00962", "2302.07241"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_550"} +{"question": "What research explained emergence by assuming a piece-wise power law functional form?", "answer": ["Broken Neural Scaling Laws"], "answer_arxiv_id": ["2210.14891"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_551"} +{"question": "Which papers introduced the Knowledge Distillation (KD) method to federated learning to aid the issues arising due to resource variation?", "answer": ["Cronus: Robust and Heterogeneous Collaborative Learning with Black-Box Knowledge Transfer", "Distillation-Based Semi-Supervised Federated Learning for Communication-Efficient Collaborative Training with Non-IID Private Data"], "answer_arxiv_id": ["1912.11279v1", "2008.06180"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_552"} +{"question": "Which works tackled zero-shot 3D segmentation or detection for both indoor and outdoor scenes?", "answer": ["Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Models", "ConceptFusion: Open-set Multimodal 3D Mapping", "RegionPLC: Regional Point-Language Contrastive Learning for Open-World 3D Scene Understanding", "Open-Vocabulary Point-Cloud Object Detection without 3D Annotation", "OpenScene: 3D Scene Understanding with Open Vocabularies", "CLIP2: Contrastive Language-Image-Point Pretraining from Real-World Point Cloud Data", "Joint Representation Learning for Text and 3D Point Cloud", "Language-Grounded Indoor 3D Semantic Segmentation in the Wild", "CLIP-FO3D: Learning Free Open-world 3D Scene Representations from 2D Dense CLIP", "LERF: Language Embedded Radiance Fields", "CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP", "LidarCLIP or: How I Learned to Talk to Point Clouds"], "answer_arxiv_id": ["2207.11514", "2302.07241", "2304.00962", "2304.00788v2", "2211.15654", "2303.12417", "2301.07584", "2204.07761", "2303.04748", "2303.09553", "2301.04926", "2212.06858"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_553"} +{"question": "Can you list the studies that explain the crucial effect of data augmentation on the generalization of contrastive learning?", "answer": ["Towards the Generalization of Contrastive Self-Supervised Learning"], "answer_arxiv_id": ["2111.00743"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_554"} +{"question": "What research simulates a universal Turing machine with a RAM-augmented LLM?", "answer": ["Memory Augmented Large Language Models are Computationally Universal"], "answer_arxiv_id": ["2301.04589"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_555"} +{"question": "What studies attempt to optimize vector-quantized networks (VQNs) using straight-through estimation?", "answer": ["Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation", "PyTorch: An Imperative Style, High-Performance Deep Learning Library"], "answer_arxiv_id": ["1308.3432", "1912.01703"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_556"} +{"question": "Which research papers have investigated ways of reducing the computational cost of standard scaled-dot product attention?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_557"} +{"question": "What papers propose a framework unifying the formalization of multiple generation processes under the assumption of instance-independence in PLL?", "answer": ["Learning with Proper Partial Labels"], "answer_arxiv_id": ["2112.12303v2"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_558"} +{"question": "Could you provide me research papers that suggested using a large perturbation budget to mitigate robust overfitting?", "answer": ["Strength-Adaptive Adversarial Training", "CFA: Class-wise Calibrated Fair Adversarial Training"], "answer_arxiv_id": ["2210.01288", "2303.14460"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_559"} +{"question": "Which methods are most similar to our approach in the aspect of NeRFs for closed-loop simulation?", "answer": ["MARS: An Instance-aware, Modular and Realistic Simulator for Autonomous\n Driving", "UniSim: A Neural Closed-Loop Sensor Simulator"], "answer_arxiv_id": ["2307.15058", "2308.01898"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_560"} +{"question": "Could you provide me some works that generate trajectories consisting of states and actions with an unconditional diffusion model?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_561"} +{"question": "Which work provides a problem-dependent regret bound with strong assumptions on ergodicity of the MDP?", "answer": ["Variance-Aware Regret Bounds for Undiscounted Reinforcement Learning in MDPs"], "answer_arxiv_id": ["1803.01626"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_562"} +{"question": "Which works focused on task clustering in multi-task learning to identify which tasks should be learnt together?", "answer": ["Taskonomy: Disentangling Task Transfer Learning", "Variational Multi-Task Learning with Gumbel-Softmax Priors", "Efficiently Identifying Task Groupings for Multi-Task Learning"], "answer_arxiv_id": ["1804.08328", "2111.05323", "2109.04617"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_563"} +{"question": "What papers focus on high efficiency attention mechanism to deal with motion-induced blurring?", "answer": ["MAXIM: Multi-Axis MLP for Image Processing", "Uformer: A General U-Shaped Transformer for Image Restoration", "Restormer: Efficient Transformer for High-Resolution Image Restoration", "Stripformer: Strip Transformer for Fast Image Deblurring", "Efficient and Explicit Modelling of Image Hierarchies for Image\n Restoration"], "answer_arxiv_id": ["2201.02973", "2106.03106", "2111.09881", "2204.04627", "2303.00748"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_564"} +{"question": "Could you provide me some studies about generating 3D consistent scenes?", "answer": ["Persistent Nature: A Generative Model of Unbounded 3D Worlds", "InfiniCity: Infinite-Scale City Synthesis", "SceneDreamer: Unbounded 3D Scene Generation from 2D Image Collections"], "answer_arxiv_id": ["2303.13515", "2301.09637", "2302.01330"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_565"} +{"question": "Which paper presented a benchmark for algorithmic tasks?", "answer": ["The CLRS Algorithmic Reasoning Benchmark"], "answer_arxiv_id": ["2205.15659"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_566"} +{"question": "Are there studies that have proposed metrics to measure stereotyped biases in language models?", "answer": ["Semantics derived automatically from language corpora contain human-like biases", "Theory-Grounded Measurement of U.S. Social Stereotypes in English\n Language Models", "Intrinsic Bias Metrics Do Not Correlate with Application Bias"], "answer_arxiv_id": ["1608.07187v4", "2206.11684", "2012.15859v5"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_567"} +{"question": "What are the works that followed the pipeline of symbolic knowledge distillation?", "answer": ["Symbolic Knowledge Distillation: from General Language Models to\n Commonsense Models", "Referee: Reference-Free Sentence Summarization with Sharper\n Controllability through Symbolic Knowledge Distillation", "I2D2: Inductive Knowledge Distillation with NeuroLogic and\n Self-Imitation", "NovaCOMET: Open Commonsense Foundation Models with Symbolic Knowledge\n Distillation"], "answer_arxiv_id": ["2110.07178", "2210.13800", "2212.09246", "2312.05979"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_568"} +{"question": "Which works have evaluated state-of-the-art methods in selection?", "answer": ["Domino: Discovering Systematic Errors with Cross-Modal Embeddings", "Selection via Proxy: Efficient Data Selection for Deep Learning"], "answer_arxiv_id": ["2203.14960", "1906.11829"], "source_meta": {"published_time": "20220720"}, "qid": "AutoScholarQuery_train_569"} +{"question": "Is there any work using MLP and MLP-Mixer as operators in knowledge graph reasoning?", "answer": ["Neural Methods for Logical Reasoning Over Knowledge Graphs"], "answer_arxiv_id": ["2209.14464"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_570"} +{"question": "Any studies deploying a Q-learning variant with variance reduction that achieves optimal regret in RL?", "answer": ["Breaking the Sample Complexity Barrier to Regret-Optimal Model-Free Reinforcement Learning"], "answer_arxiv_id": ["2110.04645"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_571"} +{"question": "What research speeds up voxelization and introduces depth supervision?", "answer": ["BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird’s-Eye View Representation", "BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detection"], "answer_arxiv_id": ["2205.13542", "2206.10092"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_572"} +{"question": "Any works on super-resolution interpolation while considering edge feature extraction?", "answer": ["Super-Resolution by Predicting Offsets: An Ultra-Efficient\n Super-Resolution Network for Rasterized Images"], "answer_arxiv_id": ["2210.04198"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_573"} +{"question": "Is there any comprehensive survey on fair allocation of indivisible items?", "answer": ["Fair Division of Indivisible Goods: A Survey"], "answer_arxiv_id": ["2202.07551v2"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_574"} +{"question": "What research works have studied the gradient-based meta-optimization in the field of meta-optimization?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "On First-Order Meta-Learning Algorithms", "EvoGrad: Efficient Gradient-Based Meta-Learning and Hyperparameter Optimization"], "answer_arxiv_id": ["1703.03400", "1803.02999", "2106.10575"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_575"} +{"question": "What is the seminal work that investigated the online caching problem with predictions?", "answer": ["Competitive caching with machine learned advice"], "answer_arxiv_id": ["1802.05399"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_576"} +{"question": "What work gives a comprehensive review of diffusion models for visual computing?", "answer": ["State of the Art on Diffusion Models for Visual Computing"], "answer_arxiv_id": ["2310.07204"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_577"} +{"question": "Can you give some examples of the papers that studied the impact objectives have on the distributions of representations in Modern Multi-Sketch Self-Supervision Learning?", "answer": ["Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere"], "answer_arxiv_id": ["2005.10242"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_578"} +{"question": "Which papers propose to minimize the distance between the final goals and curriculum goals using Euclidean distance metric?", "answer": ["Exploration via Hindsight Goal Generation"], "answer_arxiv_id": ["1906.04279"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_579"} +{"question": "Which papers presented a vision-language joint learning method for one-shot imitation learning?", "answer": ["BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning"], "answer_arxiv_id": ["2202.02005"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_580"} +{"question": "Which work demonstrated that equivariant ConvNets can learn to break equivariance when it's beneficial to the task at hand?", "answer": ["Using and Abusing Equivariance"], "answer_arxiv_id": ["2308.11316"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_581"} +{"question": "What works leveraged dropout in the test phase for uncertainty estimation?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.02142"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_582"} +{"question": "What is the study that designed the Differentiable Transformation Attack?", "answer": ["DTA: Physical Camouflage Attacks using Differentiable Transformation\n Network"], "answer_arxiv_id": ["2203.09831"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_583"} +{"question": "Which works use sampling based on importance weights allowing the training data to align with the distribution of high-quality corpora such as Wikipedia?", "answer": ["Data Selection for Language Models via Importance Resampling"], "answer_arxiv_id": ["2302.03169"], "source_meta": {"published_time": "20240709"}, "qid": "AutoScholarQuery_train_584"} +{"question": "What paper extends the particle-based deep models to mesh-based ones to improve the scalability of graph-based surrogate models?", "answer": ["Learning Mesh-Based Simulation with Graph Networks"], "answer_arxiv_id": ["2010.03409"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_585"} +{"question": "What studies provide comprehensive discussions about tuning choices for LD?", "answer": ["On the Anatomy of MCMC-Based Maximum Likelihood Learning of Energy-Based Models"], "answer_arxiv_id": ["1903.12370"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_586"} +{"question": "Which works propose automatic interventions in natural language processing such as back-translation and context augmentation?", "answer": ["Improving Neural Machine Translation Models with Monolingual Data", "Good-Enough Compositional Data Augmentation", "Contextual Augmentation: Data Augmentation by Words with Paradigmatic Relations"], "answer_arxiv_id": ["1511.06709", "1904.09545", "1805.06201"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_587"} +{"question": "Which studies focused on memory reduction on second moments using memory-efficient methods?", "answer": ["Adafactor: Adaptive Learning Rates with Sublinear Memory Cost", "Memory-Efficient Adaptive Optimization", "Extreme Tensoring for Low-Memory Preconditioning"], "answer_arxiv_id": ["1804.04235", "1901.11150", "1902.04620"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_588"} +{"question": "Could you provide me some works about cSG-MCMC and MC dropout methods as representative BNN models?", "answer": ["Cyclical Stochastic Gradient MCMC for Bayesian Deep Learning", "Bayesian Neural Network Priors Revisited", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Variational Dropout and the Local Reparameterization Trick"], "answer_arxiv_id": ["1902.03932", "2102.06571", "1506.02142", "1506.02557"], "source_meta": {"published_time": "20211227"}, "qid": "AutoScholarQuery_train_589"} +{"question": "Which studies proposed methods to generate text accompanied by attribution in Long-Form Question Answering?", "answer": ["LaMDA: Language Models for Dialog Applications", "Teaching language models to support answers with verified quotes", "Enabling Large Language Models to Generate Text with Citations", "SEMQA: Semi-Extractive Multi-Source Question Answering"], "answer_arxiv_id": ["2201.08239", "2203.11147", "2305.14627", "2311.04886"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_590"} +{"question": "Could you provide some studies that focused on heuristic quantum machine learning approaches for clustering?", "answer": ["Quantum Clustering with k-Means: a Hybrid Approach"], "answer_arxiv_id": ["2212.06691v2"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_591"} +{"question": "What papers showcased the unprecedented prowess of diffusion models in image generation and editing?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "RAPHAEL: Text-to-Image Generation via Large Mixture of Diffusion Paths", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2112.10752", "2205.11487", "2307.01952", "2204.06125", "2211.01324", "2305.18295", "2112.10741"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_592"} +{"question": "Could you provide me some studies that prove that adversarially robust training encourages invertibility in standard feedforward ANNs?", "answer": ["Adversarial Robustness as a Prior for Learned Representations"], "answer_arxiv_id": ["1906.00945"], "source_meta": {"published_time": "20230826"}, "qid": "AutoScholarQuery_train_593"} +{"question": "Could you provide me the work that proved SGD achieving linear convergence when the loss function is strongly convex?", "answer": ["Stochastic Gradient Descent, Weighted Sampling, and the Randomized Kaczmarz algorithm"], "answer_arxiv_id": ["1310.5715"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_594"} +{"question": "Any papers that work on the use of latent diffusion in molecule generation?", "answer": ["Geometric Latent Diffusion Models for 3D Molecule Generation"], "answer_arxiv_id": ["2305.01140"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_595"} +{"question": "Who studied Q-learning with optimism for online RL?", "answer": ["Provably Efficient Q-Learning with Low Switching Cost", "Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition", "Breaking the Sample Complexity Barrier to Regret-Optimal Model-Free Reinforcement Learning"], "answer_arxiv_id": ["1905.12849", "2004.10019", "2110.04645"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_596"} +{"question": "Which studies focus on post-hoc text detection methods that use perplexity-based features?", "answer": ["Origin Tracing and Detecting of LLMs"], "answer_arxiv_id": ["2304.14072"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_597"} +{"question": "Which papers scrutinize a suite of VQA models using synthetic images depicting simple 3D shapes?", "answer": ["CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning"], "answer_arxiv_id": ["1612.06890"], "source_meta": {"published_time": "20220618"}, "qid": "AutoScholarQuery_train_598"} +{"question": "Which papers mention about Vision-Language Transformers such as BLIP?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2103.00020", "2205.01917", "2208.10442", "2201.12086", "1810.04805", "2010.11929"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_599"} +{"question": "What papers discuss imposing fairness constraints that require each arm to be pulled a pre-specified fraction of times?", "answer": ["Fairness in Learning: Classic and Contextual Bandits", "Achieving Fairness in the Stochastic Multi-armed Bandit Problem", "My Fair Bandit: Distributed Learning of Max-Min Fairness with Multi-player Bandits", "Fair Algorithms for Multi-Agent Multi-Armed Bandits"], "answer_arxiv_id": ["1605.07139", "1907.10516", "2002.09808", "2007.06699"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_600"} +{"question": "Any works about momentum-based variance reduction methods in the context of problems with only stochastic gradient?", "answer": ["A Hybrid Variance-Reduced Method for Decentralized Stochastic Non-Convex Optimization", "GT-STORM: Taming Sample, Communication, and Memory Complexities in Decentralized Non-Convex Learning"], "answer_arxiv_id": ["2102.06752", "2105.01231"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_601"} +{"question": "Which works focus on accelerating pretraining with fixed, low resolution in image modelling?", "answer": ["DeiT III: Revenge of the ViT", "Swin Transformer V2: Scaling Up Capacity and Resolution"], "answer_arxiv_id": ["2204.07118", "2111.09883"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_602"} +{"question": "Which works explore structured latent priors for the Variational Autoencoder?", "answer": ["Independent Subspace Analysis for Unsupervised Learning of Disentangled Representations"], "answer_arxiv_id": ["1909.05063v1"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_603"} +{"question": "Which paper demonstrates that modern neural networks are overly confident?", "answer": ["Q"], "answer_arxiv_id": ["1611.08152"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_604"} +{"question": "Which papers first introduced Conditional Neural Processes (CNP) and its latent variable variant?", "answer": ["Conditional Neural Processes", "Neural Processes"], "answer_arxiv_id": ["1807.01613", "1807.01622"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_605"} +{"question": "Are there any papers that focus on improving language model performance using a self-consistency strategy?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20221215"}, "qid": "AutoScholarQuery_train_606"} +{"question": "What works discussed an extension of the algorithm which can be seen as a simulated annealing method for non-monotone suodular maximization?", "answer": ["Submodular Maximization by Simulated Annealing"], "answer_arxiv_id": ["1007.1632"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_607"} +{"question": "Could you point out the works that try to implement zero-shot novel view synthesis by fine-tuning pre-trained diffusion models?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object", "High-Resolution Image Synthesis with Latent Diffusion Models", "Objaverse: A Universe of Annotated 3D Objects"], "answer_arxiv_id": ["2303.11328", "2112.10752", "2212.08051"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_608"} +{"question": "Which works discuss the use of discrete diffusion methods on voxel space for outdoor scene generation?", "answer": ["Argmax Flows and Multinomial Diffusion: Learning Categorical\n Distributions", "Diffusion Probabilistic Models for Scene-Scale 3D Categorical Data"], "answer_arxiv_id": ["2102.05379", "2301.00527v1"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_609"} +{"question": "What studies measure the generalization gap between training and test accuracy with only training data?", "answer": ["Computing the Testing Error without a Testing Set", "Predicting the Generalization Gap in Deep Networks with Margin Distributions", "Exploring Generalization in Deep Learning", "Predicting Neural Network Accuracy from Weights", "Towards Task and Architecture-Independent Generalization Gap Predictors", "Heavy-Tailed Universality Predicts Trends in Test Accuracies for Very Large Pre-Trained Deep Neural Networks"], "answer_arxiv_id": ["2005.00450", "1810.00113", "1706.08947", "2002.11448", "1906.01550v1", "1901.08278"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_610"} +{"question": "Could you provide some works about the zero-shot capabilities of vision language models (VLMs)?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_611"} +{"question": "Which research papers discuss repairing input data to break dependence on sensitive attributes?", "answer": ["An algorithm for removing sensitive information: application to race-independent recidivism prediction", "Certifying and removing disparate impact"], "answer_arxiv_id": ["1703.04957", "1412.3756"], "source_meta": {"published_time": "20201201"}, "qid": "AutoScholarQuery_train_612"} +{"question": "Any studies applied theories of canonical forms in machine learning that can handle ambiguities?", "answer": ["ConDor: Self-Supervised Canonicalization of 3D Pose for Partial Shapes", "Canonical Fields: Self-Supervised Learning of Pose-Canonicalized Neural Fields", "Canonical Capsules: Self-Supervised Capsules in Canonical Pose", "Equivariance with Learned Canonicalization Functions"], "answer_arxiv_id": ["2201.07788", "2212.02493v3", "2012.04718", "2211.06489"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_613"} +{"question": "Which work is regarded as a state-of-the-art approach in open VQA and visual captioning tasks?", "answer": ["CoCa: Contrastive Captioners are Image-Text Foundation Models"], "answer_arxiv_id": ["2205.01917"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_614"} +{"question": "What research showed that model performance drops substantially when evaluated on minimally-edited contrast example sets authored by human experts?", "answer": ["Evaluating Models’ Local Decision Boundaries via Contrast Sets"], "answer_arxiv_id": ["2004.02709"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_615"} +{"question": "What studies give examples of how methods have been developed to augment pre-trained models with context in natural language processing (NLP)?", "answer": ["Reading Wikipedia to Answer Open-Domain Questions"], "answer_arxiv_id": ["1704.00051"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_616"} +{"question": "Which studies proposed performing approximate posterior inference using MFVI in weight space for a subset of layers in Transformers?", "answer": ["Bayesian Layers: A Module for Neural Network Uncertainty", "Bayesian Transformer Language Models for Speech Recognition"], "answer_arxiv_id": ["1812.03973", "2102.04754"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_617"} +{"question": "Could you provide some studies where sketches are used in downstream tasks like saliency detection, augmented reality, among others?", "answer": ["Sketch2Saliency: Learning to Detect Salient Objects from Human Drawings", "Structure-Aware 3D VR Sketch to 3D Shape Retrieval", "3D VR Sketch Guided 3D Shape Prototyping and Exploration", "Sketch-based Medical Image Retrieval", "What Can Human Sketches Do for Object Detection?", "Doodle It Yourself: Class Incremental Learning by Drawing a Few Sketches"], "answer_arxiv_id": ["2303.11502", "2209.09043", "2306.10830", "2303.03633", "2303.15149", "2203.14843"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_618"} +{"question": "What papers have studied large pre-trained vision-language models with static images and videos?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Valley: Video Assistant with Large Language model Enhanced abilitY"], "answer_arxiv_id": ["2301.12597", "2304.10592", "2306.07207"], "source_meta": {"published_time": "20240515"}, "qid": "AutoScholarQuery_train_619"} +{"question": "Could you give examples of research using contrastive learning for self-supervised representation learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Self-Supervised Learning of Pretext-Invariant Representations", "Representation Learning with Contrastive Predictive Coding", "Emerging Properties in Self-Supervised Vision Transformers", "Dense Contrastive Learning for Self-Supervised Visual Pre-Training", "Self-Supervised Visual Representation Learning with Semantic Grouping"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2006.07733", "1912.01991", "1807.03748", "2104.14294", "2011.09157", "2205.15288"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_620"} +{"question": "Which research introduced Slot attention for processing images and videos?", "answer": ["Multi-Object Representation Learning with Iterative Variational Inference", "MONet: Unsupervised Scene Decomposition and Representation", "Genesis: Generative Scene Inference and Sampling with Object-Centric Latent Representations", "SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition", "Unsupervised Foreground Extraction via Deep Region Competition", "Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "Tagger: Deep Unsupervised Perceptual Grouping", "Unsupervised Learning of Compositional Energy Concepts", "Object-Centric Learning with Slot Attention", "Illiterate DALL-E Learns to Compose", "SIMONe: View-Invariant, Temporally-Abstracted Object Representations via Unsupervised Video Decomposition", "Neural Expectation Maximization", "Conditional Object-Centric Learning from Video", "SCALOR: Generative World Models with Scalable Object Representations", "Entity Abstraction in Visual Model-Based Reinforcement Learning", "Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects"], "answer_arxiv_id": ["1903.00450", "1901.11390", "1907.13052", "2001.02407", "2110.15497", "1603.08575", "1606.06724", "2111.03042", "2006.15055", "2110.11405", "2106.03849", "1708.03498", "2111.12594", "1910.02384", "1910.12827", "1806.01794"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_621"} +{"question": "In the field of pose estimation, which works majorly focused on inferring pose from sparse frames?", "answer": ["PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in\n Cluttered Scenes", "CosyPose: Consistent multi-view multi-object 6D pose estimation"], "answer_arxiv_id": ["1711.00199", "2008.08465"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_622"} +{"question": "Which works present different frameworks to mitigate degree bias in Graph Neural Networks?", "answer": ["Cold Brew: Distilling Graph Node Representations with Incomplete or Missing Neighborhoods"], "answer_arxiv_id": ["2111.04840"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_623"} +{"question": "Which studies are about continuous prompt optimization?", "answer": ["Learning How to Ask: Querying LMs with Mixtures of Soft Prompts", "GPT Understands, Too", "The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2104.06599", "2103.10385", "2104.08691"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_624"} +{"question": "What works explored the use of enforcing invertibility between output and latent codes to achieve diversity in the conditional setting?", "answer": ["Toward Multimodal Image-to-Image Translation"], "answer_arxiv_id": ["1711.11586"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_625"} +{"question": "Which works have designed a visual attention based on the concept of biomimetic vision modeling?", "answer": ["Focal Self-attention for Local-Global Interactions in Vision\n Transformers"], "answer_arxiv_id": ["2107.00641"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_626"} +{"question": "Which works deal with zero-shot text-to-image generators in the context of prior-based 3D reconstruction?", "answer": ["Zero-Shot Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["2102.12092", "2112.10752", "2204.06125", "2205.11487", "2211.01324", "2203.13131"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_627"} +{"question": "Could you give examples of research related to ACE such as Bayesian Optimization for Likelihood-free Inference (BOLFI) or error-guided LFI-MCMC?", "answer": ["Bayesian Optimization for Likelihood-Free Inference of Simulator-Based Statistical Models", "Error-guided likelihood-free MCMC"], "answer_arxiv_id": ["1501.03291v3", "2010.06735"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_628"} +{"question": "Could you mention some studies that utilized multitasking as a strategy in neurosymbolic programming?", "answer": ["Houdini: Lifelong Learning as Program Synthesis"], "answer_arxiv_id": ["1804.00218"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_629"} +{"question": "Which papers used RNNs for reconstructing DS from measured time series?", "answer": ["Identifying nonlinear dynamical systems with multiple time scales and long-range dependencies", "Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI", "Backpropagation Algorithms and Reservoir Computing in Recurrent Neural Networks for the Forecasting of Complex Spatiotemporal Dynamics", "R", "Using Machine Learning to Anticipate Tipping Points and Extrapolate to Post-Tipping Dynamics of Non-Stationary Dynamical Systems"], "answer_arxiv_id": ["1910.03471", "1902.07186v2", "1910.05266", "1210.6589", "2207.00521"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_630"} +{"question": "What are some papers on the application of large-kernel ConvNets on downstream tasks?", "answer": ["More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using\n Sparsity", "LargeKernel3D: Scaling up Kernels in 3D Sparse CNNs", "LKD-Net: Large Kernel Convolution Network for Single Image Dehazing", "Large Kernel Distillation Network for Efficient Single Image Super-Resolution"], "answer_arxiv_id": ["2207.03620", "2206.10555", "2209.01788", "2407.14340v1"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_631"} +{"question": "What studies have been done on the application of weight-sharing in dynamic neural networks?", "answer": ["Dynamic Neural Networks: A Survey", "SkipNet: Learning Dynamic Routing in Convolutional Networks", "Fractional Skipping: Towards Finer-Grained Dynamic CNN Inference"], "answer_arxiv_id": ["2102.04906", "1711.09485", "2001.00705"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_632"} +{"question": "What papers study the direct method estimator which is a type of OPE estimator in reinforcement learning?", "answer": ["Batch Policy Learning under Constraints", "Accountable Off-Policy Evaluation With Kernel Bellman Statistics", "Estimating Dynamic Treatment Regimes in Mobile Health Using V-learning", "Bootstrapping Fitted Q-Evaluation for Off-Policy Inference", "Off-Policy Estimation of Long-Term Average Outcomes with Applications to Mobile Health", "On Well-posedness and Minimax Optimal Rates of Nonparametric Q-function Estimation in Off-policy Evaluation"], "answer_arxiv_id": ["1903.08738", "2008.06668", "1611.03531v2", "2102.03607", "1912.13088", "2201.06169v3"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_633"} +{"question": "Could you provide me some works that have addressed the problem of synthetic image detection?", "answer": ["Do GANs leave artificial fingerprints?", "FaceForensics++: Learning to Detect Manipulated Facial Images", "Learning Self-Consistency for Deepfake Detection", "DIRE for Diffusion-Generated Image Detection", "Leveraging Frequency Analysis for Deep Fake Image Recognition", "BiHPF: Bilateral High-Pass Filters for Robust Deepfake Detection", "FrePGAN: Robust Deepfake Detection Using Frequency-level Perturbations", "Towards Universal Fake Image Detectors that Generalize Across Generative\n Models", "Attributing Fake Images to GANs: Learning and Analyzing GAN Fingerprints", "CNN-generated images are surprisingly easy to spot... for now", "Thinking in Frequency: Face Forgery Detection by Mining Frequency-aware\n Clues"], "answer_arxiv_id": ["1812.11842", "1901.08971", "2012.09311", "2303.09295", "2003.08685", "2109.00911", "2202.03347", "2302.10174", "1811.08180", "1912.11035", "2007.09355"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_634"} +{"question": "What works rely on Language and Vision-Integrated Language Models (LVLMs) to provide Natural Language Feedback (NLF)?", "answer": ["Re3: Generating Longer Stories With Recursive Reprompting and Revision", "Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback", "RL4F: Generating Natural Language Feedback with Reinforcement Learning\n for Repairing Model Outputs"], "answer_arxiv_id": ["2210.06774", "2204.05862", "2305.08844"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_635"} +{"question": "Which study showed the practicality of generating rationales by LLMs?", "answer": ["STaR: Bootstrapping Reasoning With Reasoning"], "answer_arxiv_id": ["2203.14465"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_636"} +{"question": "What is the work in which Soft Q-learning introduces entropy regularization in RL?", "answer": ["Reinforcement Learning with Deep Energy-Based Policies"], "answer_arxiv_id": ["1702.08165"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_637"} +{"question": "What works learn a NeRF conditioned on an expression vector from monocular videos?", "answer": ["Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar\n Reconstruction"], "answer_arxiv_id": ["2012.03065"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_638"} +{"question": "What are some studies on training class-agnostic motion prediction models in a weakly-supervised or self-supervised manner?", "answer": ["Self-Supervised Pillar Motion Learning for Autonomous Driving"], "answer_arxiv_id": ["2104.08683"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_639"} +{"question": "Which paper found a gap in the convergence analysis of the Adam optimization algorithm and constructed counter-examples where Adam didn't converge?", "answer": ["On the convergence of Adam and Beyond"], "answer_arxiv_id": ["1904.09237"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_640"} +{"question": "What papers have explored optimal neural architectures to directly estimate dense pixel values for image enhancement?", "answer": ["DSLR-Quality Photos on Mobile Devices with Deep Convolutional Networks", "DeepLPF: Deep Local Parametric Filters for Image Enhancement", "CURL: Neural Curve Layers for Global Image Enhancement"], "answer_arxiv_id": ["1704.02470", "2003.13985", "1911.13175"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_641"} +{"question": "Which papers discussed shilling attacks in the context of injected fake profiles?", "answer": ["Assessing the Impact of a User-Item Collaborative Attack on Class of\n Users", "Tensor Trust: Interpretable Prompt Injection Attacks from an Online Game"], "answer_arxiv_id": ["1908.07968", "2311.01011"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_642"} +{"question": "What papers focus on multi-codebook quantization techniques that approximate the datapoint as some aggregate of the codebook element it was assigned to per-codebook?", "answer": ["Accelerating Large-Scale Inference with Anisotropic Vector Quantization"], "answer_arxiv_id": ["1908.10396"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_643"} +{"question": "Which studies involve powerful LLMs such as ChatGPT used in video-language models?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_644"} +{"question": "What works found that skip connections promote flat minima?", "answer": ["Visualizing the Loss Landscape of Neural Nets"], "answer_arxiv_id": ["1712.09913"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_645"} +{"question": "What works provide direction for distinguishing between identity and style in the context of domain adaptation?", "answer": ["StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "Mind the Gap: Domain Gap Control for Single Shot Domain Adaptation for\n Generative Adversarial Networks", "DynaGAN: Dynamic Few-shot Adaptation of GANs to Multiple Domains"], "answer_arxiv_id": ["2108.00946", "2110.08398", "2211.14554"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_646"} +{"question": "Which research applies TRs to focus the search on promising regions of the search space in the context of combinatorial and high-dimensional Bayesian optimization?", "answer": ["Think Global and Act Local: Bayesian Optimisation over High-Dimensional Categorical and Mixed Search Spaces"], "answer_arxiv_id": ["2102.07188"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_647"} +{"question": "Can you mention any works that introduced a bias loss during training which explicitly encourages the model to attend to the entities and phrases in the dictionary?", "answer": ["End-to-end contextual asr based on posterior distribution adaptation for\n hybrid ctc/attention system"], "answer_arxiv_id": ["2202.09003"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_648"} +{"question": "Could you provide me research that incorporated the prediction of tangent planes in higher dimensional curved manifolds for solving MMC?", "answer": ["Riemannian Multi-Manifold Modeling"], "answer_arxiv_id": ["1410.0095"], "source_meta": {"published_time": "20210728"}, "qid": "AutoScholarQuery_train_649"} +{"question": "Which paper studied stochastic regime with adversarial corruptions in prediction with expert advice (PEA) and what were their findings?", "answer": ["Prediction with Corrupted Expert Advice"], "answer_arxiv_id": ["2002.10286"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_650"} +{"question": "Which supervised methods use frame pairs to train discriminative instance embeddings for object tracking?", "answer": ["Quasi-Dense Similarity Learning for Multiple Object Tracking", "In Defense of Online Models for Video Instance Segmentation", "Universal Instance Perception as Object Discovery and Retrieval", "Tracking Every Thing in the Wild", "MOTR: End-to-End Multiple-Object Tracking with Transformer", "MOTRv2: Bootstrapping End-to-End Multi-Object Tracking by Pretrained\n Object Detectors", "Unifying Short and Long-Term Tracking with Graph Hierarchies", "TrackFormer: Multi-Object Tracking with Transformers", "Towards Real-Time Multi-Object Tracking", "Towards Grand Unification of Object Tracking"], "answer_arxiv_id": ["2006.06664", "2207.10661", "2303.06674", "2207.12978v1", "2105.03247", "2211.09791", "2212.03038", "2101.02702", "1909.12605", "2207.07078"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_651"} +{"question": "What researches have adapted the environment parameters in response to the agent’s performance when generating curricula?", "answer": ["Teacher algorithms for curriculum learning of Deep RL in continuously parameterized environments", "Teacher-Student Curriculum Learning", "Self-Paced Contextual Reinforcement Learning", "Self-Paced Context Evaluation for Contextual Reinforcement Learning"], "answer_arxiv_id": ["1910.07224", "1707.00183", "1910.02826v1", "2106.05110"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_652"} +{"question": "Could you provide examples of approaches that achieve structured and equivariant representations by learning equivariance?", "answer": ["Group Equivariant Convolutional Networks", "Scale-Equivariant Steerable Networks", "Enabling Equivariance for Arbitrary Lie Groups", "Quantised Transforming Auto-Encoders: Achieving Equivariance to Arbitrary Transformations in Deep Networks", "Offset equivariant networks and their applications"], "answer_arxiv_id": ["1602.07576", "1910.11093", "2111.08251", "2111.12873", "2207.00292"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_653"} +{"question": "What papers introduced conditional variational autoencoder (CVAE) into their RL algorithms?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction"], "answer_arxiv_id": ["1812.02900", "1906.00949"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_654"} +{"question": "What studies suggest that automatic image quality and image-text alignment metrics found to be inconsistent with human judgment in text-to-image generation tasks?", "answer": ["CogView2: Faster and Better Text-to-Image Generation via Hierarchical Transformers", "Toward Verifiable and Reproducible Human Evaluation for Text-to-Image Generation", "On Aliased Resizing and Surprising Subtleties in GAN Evaluation"], "answer_arxiv_id": ["2204.14217", "2304.01816", "2104.11222"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_655"} +{"question": "What are the research studies on methods to handle CD-FSC with large domain gaps?", "answer": ["Cross-Domain Few-Shot Classification via Adversarial Task Augmentation", "Channel Importance Matters in Few-Shot Image Classification", "Self-Supervision Can Be a Good Few-Shot Learner", "Self-training For Few-shot Transfer Across Extreme Task Differences", "Dynamic Distillation Network for Cross-Domain Few-Shot Recognition with Unlabeled Data"], "answer_arxiv_id": ["2104.14385", "2206.08126", "2207.09176", "2010.07734", "2106.07807"], "source_meta": {"published_time": "20231104"}, "qid": "AutoScholarQuery_train_656"} +{"question": "Could you tell me about the studies that introduced the spatio-temporal attention mechanism into their networks to transform pre-trained Text-to-Image model to the temporal dimension?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data", "MagicVideo: Efficient Video Generation With Latent Diffusion Models"], "answer_arxiv_id": ["2209.14792", "2211.11018"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_657"} +{"question": "What studies proposed the use of only positive pairs in SSL and applied techniques such as momentum networks or stop-gradient?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2006.07733", "2011.10566"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_658"} +{"question": "Which work proposes a modified objective using an unbounded preference mapping function to mitigate overfitting on deterministic preferences in the dataset?", "answer": ["A General Theoretical Paradigm to Understand Learning from Human\n Preferences"], "answer_arxiv_id": ["2310.12036"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_659"} +{"question": "Which works have studied the length generalization on synthetic reasoning tasks for pretrained large language models?", "answer": ["Exploring Length Generalization in Large Language Models", "Unveiling Transformers with LEGO: a synthetic reasoning task"], "answer_arxiv_id": ["2207.04901", "2206.04301"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_660"} +{"question": "Are there any research works that improved factual correctness and reasoning accuracy via multi-agent debate?", "answer": ["Improving Factuality and Reasoning in Language Models through Multiagent\n Debate"], "answer_arxiv_id": ["2305.14325"], "source_meta": {"published_time": "20230716"}, "qid": "AutoScholarQuery_train_661"} +{"question": "What studies focus on scalarization-based approach in MOBO?", "answer": ["A Flexible Framework for Multi-Objective Bayesian Optimization using Random Scalarizations", "Random Hypervolume Scalarizations for Provable Multi-Objective Black Box Optimization"], "answer_arxiv_id": ["1805.12168", "2006.04655"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_662"} +{"question": "Any works where the LLM acts as both a selection and inference module to produce explanations?", "answer": ["Selection-Inference: Exploiting Large Language Models for Interpretable\n Logical Reasoning"], "answer_arxiv_id": ["2205.09712"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_663"} +{"question": "Could you provide me some works where they introduced smoothed density value and reconstruction in patch-based methods?", "answer": ["PPFNet: Global Context Aware Local Features for Robust 3D Point Matching", "The Perfect Match: 3D Point Cloud Matching with Smoothed Densities", "Distinctive 3D local deep descriptors"], "answer_arxiv_id": ["1802.02669", "1811.06879", "2009.00258"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_664"} +{"question": "What are some studies that explore the use of LLMs like GPT-4 for generating visual instruction data?", "answer": ["GPT-4 Technical Report", "Visual Instruction Tuning", "LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "SVIT: Scaling up Visual Instruction Tuning"], "answer_arxiv_id": ["2303.08774", "2304.08485", "2306.17107v2", "2304.10592", "2307.04087"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_665"} +{"question": "Are there any studies about adapter-based methods?", "answer": ["Parameter-Efficient Transfer Learning for NLP"], "answer_arxiv_id": ["1902.00751"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_666"} +{"question": "What work generalized the policy elimination algorithm and introduced APEVE?", "answer": ["Sample-Efficient Reinforcement Learning with loglog(T) Switching Cost"], "answer_arxiv_id": ["2202.06385"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_667"} +{"question": "Which papers have utilized image reconstruction in representation learning for reinforcement learning?", "answer": ["Embed to Control: A Locally Linear Latent Dynamics Model for Control from Raw Images", "From Pixels to Torques: Policy Learning with Deep Dynamical Models"], "answer_arxiv_id": ["1506.07365", "1502.02251"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_668"} +{"question": "What research highlighted the challenges for reinforcement learning in crafting items in Minecraft's technology tree?", "answer": ["Align-RUDDER: Learning From Few Demonstrations by Reward Redistribution", "Mastering Diverse Domains through World Models"], "answer_arxiv_id": ["2009.14108", "2301.04104"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_669"} +{"question": "Could you provide me some studies about enabling interactions with scenes in the field of motion generation?", "answer": ["Learning joint reconstruction of hands and manipulated objects", "Capturing and Inferring Dense Full-Body Human-Scene Contact", "Generating Continual Human Motion in Diverse 3D Scenes"], "answer_arxiv_id": ["1904.05767", "2206.09553", "2304.02061"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_670"} +{"question": "Could you provide me with research that adopts LLMs for robotic action planning?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2204.01691", "2303.03378v1"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_671"} +{"question": "Could you list some papers where contrastive learning is improved by creating extra views through mixing up?", "answer": ["Hard Negative Mixing for Contrastive Learning"], "answer_arxiv_id": ["2010.01028v2"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_672"} +{"question": "Can you cite the studies that analyzed the estimator under various low-rank and sparsity assumptions of reshapings of the sufficient statistics into a tensor?", "answer": ["A Computationally Efficient Method for Learning Exponential Family Distributions"], "answer_arxiv_id": ["2110.15397v1"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_673"} +{"question": "Could you provide me any examples of papers that explored the use of discrete prompts in language models?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_674"} +{"question": "What are the main studies about the use of orthogonalized convolutions?", "answer": ["Skew Orthogonal Convolutions", "Orthogonalizing Convolutional Layers with the Cayley Transform", "A Dynamical System Perspective for Lipschitz Neural Networks", "A Unified Algebraic Perspective on Lipschitz Neural Networks"], "answer_arxiv_id": ["2105.11417", "2104.07167", "2110.12690", "2303.03169"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_675"} +{"question": "Could you provide me with some studies that discuss DDPM-alike Diffusion Models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Probabilistic Models", "Structured Denoising Diffusion Models in Discrete State-Spaces", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "Efficient Diffusion Training via Min-SNR Weighting Strategy", "Diffusion Models in Vision: A Survey", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2102.09672", "2006.11239", "2107.03006", "2201.09865", "2303.09556", "2209.04747", "2108.02938"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_676"} +{"question": "Which work applied this estimator to the GLB algorithm and results in a heavy-tailed GLB algorithm?", "answer": ["Scalable Generalized Linear Bandits: Online Computation and Hashing"], "answer_arxiv_id": ["1706.00136"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_677"} +{"question": "What works found that the Chain-of-Thought method leads to poor performance in certain kinds of reasonings tasks?", "answer": ["Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them"], "answer_arxiv_id": ["2210.09261v1"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_678"} +{"question": "What studies reported on joint training with language-only and vision-and-language instruction data to prevent catastrophic forgetting of language knowledge?", "answer": ["mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "Improved Baselines with Visual Instruction Tuning"], "answer_arxiv_id": ["2304.14178", "2310.03744"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_679"} +{"question": "What works found that feature visualizations are helpful but not more than highly activating natural exemplars?", "answer": ["Exemplary Natural Images Explain CNN Activations Better than State-of-the-Art Feature Visualization", "How Well do Feature Visualizations Support Causal Understanding of CNN Activations?"], "answer_arxiv_id": ["2010.12606", "2106.12447"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_680"} +{"question": "Which studies propose a graph-based approach for image segmentation that gradually refines segmentation results?", "answer": ["Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs", "DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs", "Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials", "Learning Deep Structured Models"], "answer_arxiv_id": ["1412.7062", "1606.00915", "1210.5644", "1407.2538"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_681"} +{"question": "Could you provide me some works on evaluating large language models' capacity to avoid producing toxic content?", "answer": ["RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models"], "answer_arxiv_id": ["2009.11462v2"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_682"} +{"question": "What studies made assumptions on mixture Gaussian data for domain generalization?", "answer": ["Heterogeneous Risk Minimization"], "answer_arxiv_id": ["2105.03818"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_683"} +{"question": "What work does evidence that LLMs have weakness in multi-hop reasoning?", "answer": ["Towards Reasoning in Large Language Models: A Survey"], "answer_arxiv_id": ["2212.10403"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_684"} +{"question": "What are some optimization-based methods that are lagging behind the quality of text-to-image models?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "Magic3D: High-Resolution Text-to-3D Content Creation", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "MVDream: Multi-view Diffusion for 3D Generation", "DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1", "2211.07600", "2211.10440", "2303.13873", "2305.16213", "2308.16512", "2309.16653"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_685"} +{"question": "Which works propose ethical regulations to rectify algorithms in AI Fairness?", "answer": ["Equality of Opportunity in Supervised Learning", "Achieving Fairness at No Utility Cost via Data Reweighing with Influence", "Deep Clustering based Fair Outlier Detection", "Robust Fair Clustering: A Novel Fairness Attack and Defense Framework"], "answer_arxiv_id": ["1610.02413", "2202.00787", "2106.05127", "2210.01953"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_686"} +{"question": "Which papers discuss that how all robust features become less useful in case of directed attacks?", "answer": ["Adversarial Examples Are Not Bugs, They Are Features", "Adversarial Perturbations Are Not So Weird: Entanglement of Robust and Non-Robust Features in Neural Network Classifiers"], "answer_arxiv_id": ["1905.02175", "2102.05110"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_687"} +{"question": "Which research papers highlighted the phenomena known as 'optimistic rates' in generalization theory?", "answer": ["Arxiv"], "answer_arxiv_id": ["2004.12380"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_688"} +{"question": "Can you provide any works related to the decomposition method in AI assistance?", "answer": ["Supervising strong learners by amplifying weak experts", "Recursively Summarizing Books with Human Feedback"], "answer_arxiv_id": ["1810.08575", "2109.10862"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_689"} +{"question": "Which papers studied the linear interpolation between the network at initialization and the network after training?", "answer": ["Qualitatively Characterizing Neural Network Optimization Problems", "Revisiting “Qualitatively Characterizing Neural Network Optimization Problems”", "Analyzing Monotonic Linear Interpolation in Neural Network Loss Landscapes"], "answer_arxiv_id": ["1412.6544", "2012.06898", "2104.11044v2"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_690"} +{"question": "Which studies have tackled open-domain dialogue generation through the use of cue words?", "answer": ["Sequence to Backward and Forward Sequences: A Content-Introducing Approach to Generative Short-Text Conversation", "Generating Multiple Diverse Responses for Short-Text Conversation", "Thinking Clearly, Talking Fast: Concept-Guided Non-Autoregressive Generation for Open-Domain Dialogue Systems"], "answer_arxiv_id": ["1607.00970", "1811.05696", "2109.04084"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_691"} +{"question": "Which works have applied Graph Neural Networks to various graph learning tasks?", "answer": ["A Comprehensive Survey on Graph Neural Networks", "How Powerful are Graph Neural Networks?", "Semi-Supervised Classification with Graph Convolutional Networks", "Inductive Representation Learning on Large Graphs", "Graph Attention Networks", "Data Augmentation for Deep Graph Learning: A Survey", "Learning Strong Graph Neural Networks with Weak Information"], "answer_arxiv_id": ["1901.00596", "1810.00826", "1609.02907", "1706.02216", "1710.10903", "2202.08235", "2305.18457"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_692"} +{"question": "Are there any papers focusing on population-based hyperparameter tuning of RL?", "answer": ["Population Based Training of Neural Networks"], "answer_arxiv_id": ["1711.09846"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_693"} +{"question": "In what research was the first visual prompting framework introduced that utilised inpainting with discrete tokens on images?", "answer": ["Visual Prompting via Image Inpainting"], "answer_arxiv_id": ["2209.00647"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_694"} +{"question": "Can you provide some works that discussed meta-learning for few-shot learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Matching Networks for One Shot Learning", "Prototypical Networks for Few-shot Learning"], "answer_arxiv_id": ["1703.03400", "1606.04080", "1703.05175"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_695"} +{"question": "Which work showed the link between network approximation error, network size, weight bounds to the training sample size in sparse DNN?", "answer": ["Consistent Sparse Deep Learning: Theory and Computation"], "answer_arxiv_id": ["2102.13229"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_696"} +{"question": "Could you provide me some works that indicate that convolutions can be effective for sequence modeling", "answer": ["CKConv: Continuous Kernel Convolution For Sequential Data", "What Makes Convolutional Models Great on Long Sequence Modeling?"], "answer_arxiv_id": ["2102.02611", "2210.09298"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_697"} +{"question": "Which papers talk about extending stationary Gaussian process kernels to the entire isometry group of Euclidean spaces?", "answer": ["Equivariant Learning of Stochastic Fields: Gaussian Processes and Steerable Conditional Neural Processes"], "answer_arxiv_id": ["2011.12916"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_698"} +{"question": "Which papers propose describing human motion with pre-defined action labels?", "answer": ["MotionBERT: A Unified Perspective on Learning Human Motion\n Representations"], "answer_arxiv_id": ["2210.06551"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_699"} +{"question": "Can you mention some notable works that used transformers for behavior learning?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Offline Reinforcement Learning as One Big Sequence Modeling Problem"], "answer_arxiv_id": ["2106.01345", "2106.02039"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_700"} +{"question": "Could you provide me some studies about learning the volumetric field between pairs of meshed objects without collisions?", "answer": ["MeshODE: A Robust and Scalable Framework for Mesh Deformation"], "answer_arxiv_id": ["2005.11617"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_701"} +{"question": "Which work formulates a knapsack problem for channel pruning with an explicit FLOPs constraint?", "answer": ["Knapsack Pruning with Inner Distillation"], "answer_arxiv_id": ["2002.08258"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_702"} +{"question": "What are some key works analyzing formal linguistic abilities in pre-learned models (PLMs)?", "answer": ["A Primer in BERTology: What we know about how BERT works", "When Do You Need Billions of Words of Pretraining Data?", "Construction Grammar and Language Models", "Language Models as Knowledge Bases?"], "answer_arxiv_id": ["2002.12327", "2011.04946", "2308.13315v2", "1909.01066"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_703"} +{"question": "Could you tell me about the work done on creating manual or semi-manual prompts and/or tuning LMs?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization"], "answer_arxiv_id": ["2110.08207"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_704"} +{"question": "What researches exist on learning contextual policies in classical adaptive control literature?", "answer": ["Adaptive Control: Algorithms, Analysis and Applications"], "answer_arxiv_id": ["2406.07073v1"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_705"} +{"question": "Are there works that progressed from static to dynamic human-scene interactions?", "answer": ["Synthesizing Long-Term 3D Human Motion and Interaction in 3D Scenes", "Stochastic Scene-Aware Motion Prediction", "Compositional Human-Scene Interaction Synthesis with Semantic Control"], "answer_arxiv_id": ["2012.05522", "2108.08284", "2207.12824"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_706"} +{"question": "Which works introduced latent variables in a commonsense knowledge graph?", "answer": ["Diversifying Content Generation for Commonsense Reasoning with Mixture of Knowledge Graph Experts"], "answer_arxiv_id": ["2203.07285v1"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_707"} +{"question": "Which work was the pioneering one that formulated image generation as a diffusion process?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_708"} +{"question": "Could you provide me with some works that discussed the instance of representation collapse in self-supervised learning?", "answer": ["Understanding Dimensional Collapse in Contrastive Self-supervised\n Learning"], "answer_arxiv_id": ["2110.09348"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_709"} +{"question": "What works are about generating music from visual or/and motion data?", "answer": ["Foley Music: Learning to Generate Music from Videos", "Video Background Music Generation with Controllable Music Transformer", "Dance2Music: Automatic Dance-driven Music Generation", "Quantized GAN for Complex Music Generation from Dance Videos"], "answer_arxiv_id": ["2007.10984", "2111.08380", "2107.06252", "2204.00604"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_710"} +{"question": "Which studies have focused on the lack of recovery guarantees of robust formulations in graph learning from stationary signals?", "answer": ["Characterization and Inference of Graph Diffusion Processes from Observations of Stationary Signals", "Joint Inference of Multiple Graphs from Matrix Polynomials"], "answer_arxiv_id": ["1605.02569", "2010.08120"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_711"} +{"question": "Which research studies introduced the utilization of CNNs in the field of computer vision and image denoising?", "answer": ["Very Deep Convolutional Networks for Large-Scale Image Recognition", "Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1409.1556", "1512.03385"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_712"} +{"question": "What studies used data augmentation methods, such as cropping, flipping, and rotating, to train RL algorithms?", "answer": ["Reinforcement Learning with Augmented Data"], "answer_arxiv_id": ["2004.14990"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_713"} +{"question": "Which methods propose optimizations to both the NeRF and input poses using embeddings?", "answer": ["NeRF--: Neural Radiance Fields Without Known Camera Parameters"], "answer_arxiv_id": ["2102.07064"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_714"} +{"question": "Do we have any research on directly parameterizing Gaussian processes or Neural Networks to satisfy Partial Differential Equations (PDEs)?", "answer": ["Algorithmic Linearly Constrained Gaussian Processes", "Linearly constrained Gaussian processes", "Linearly Constrained Neural Networks"], "answer_arxiv_id": ["1801.09197", "1703.00787", "2002.01600"], "source_meta": {"published_time": "20220718"}, "qid": "AutoScholarQuery_train_715"} +{"question": "Which research used human-like data augmentation to improve models' shape bias?", "answer": ["The Origins and Prevalence of Texture Bias in Convolutional Neural Networks"], "answer_arxiv_id": ["1911.09071"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_716"} +{"question": "What works imposed adaptive Huber regression to handle homogeneous offline heavy-tailed noise?", "answer": ["Adaptive Huber Regression"], "answer_arxiv_id": ["1706.06991v2"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_717"} +{"question": "Which papers have considered harmonic clustering of simplices?", "answer": ["A Notion of Harmonic Clustering in Simplicial Complexes", "Spectral Detection of Simplicial Communities via Hodge Laplacians"], "answer_arxiv_id": ["1910.07247", "2108.06547"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_718"} +{"question": "Which research dives into mask-based conformal risk control procedures in image regressions?", "answer": ["What’s Behind the Mask: Estimating Uncertainty in Image-to-Image Problems"], "answer_arxiv_id": ["2211.15211"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_719"} +{"question": "What works adopted recurrent models with input attention?", "answer": ["DRAW: A Recurrent Neural Network For Image Generation"], "answer_arxiv_id": ["1502.04623"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_720"} +{"question": "What study proposed a game protocol that allows simultaneous queries in MAB?", "answer": ["Most Correlated Arms Identification"], "answer_arxiv_id": ["1404.5903"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_721"} +{"question": "What research papers have proposed and developed the structured state space model for sequences and its variants?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces", "Simplified State Space Layers for Sequence Modeling", "Liquid Structural State-Space Models", "Diagonal State Spaces are as Effective as Structured State Spaces"], "answer_arxiv_id": ["2111.00396", "2208.04933", "2209.12951", "2203.14343"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_722"} +{"question": "Could you provide me with research that uses variational approaches for approximate inference in BNNs?", "answer": ["Weight Uncertainty in Neural Networks", "Multiplicative Normalizing Flows for Variational Bayesian Neural Networks"], "answer_arxiv_id": ["1505.05424", "1703.01961"], "source_meta": {"published_time": "20210719"}, "qid": "AutoScholarQuery_train_723"} +{"question": "Could you provide me some literature on the topic of language-conditioned vision-based robotic manipulation systems?", "answer": ["Language Conditioned Imitation Learning over Unstructured Data", "Language-Conditioned Imitation Learning for Robot Manipulation Tasks", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning", "Interactive Language: Talking to Robots in Real Time", "CLIPort: What and Where Pathways for Robotic Manipulation", "RT-1: Robotics Transformer for Real-World Control at Scale", "Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation", "Instruction-driven history-aware policies for robotic manipulations", "CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Tasks"], "answer_arxiv_id": ["2005.07648", "2010.12083", "2204.01691", "2202.02005", "2210.06407", "2109.12098", "2212.06817", "2209.05451", "2209.04899", "2112.03227"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_724"} +{"question": "What works have empirically shown the existence and exacerbation of 'accuracy disparity' in robust classifiers?", "answer": ["RobustBench: a standardized adversarial robustness benchmark", "Robustness May Be at Odds with Fairness: An Empirical Study on Class-wise Accuracy"], "answer_arxiv_id": ["2010.09670", "2010.13365v2"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_725"} +{"question": "Can you provide me with research that used CLIP for generating pseudo-labels?", "answer": ["Semantic Segmentation In-the-Wild Without Seeing Any Segmentation Examples", "A Closer Look at Self-training for Zero-Label Semantic Segmentation"], "answer_arxiv_id": ["2112.03185", "2104.11692"], "source_meta": {"published_time": "20230930"}, "qid": "AutoScholarQuery_train_726"} +{"question": "Which research critically studies the connection between GNNs and Local and Congest models for distributed computation and translates lower bounds for Congest into size lower bounds for GNNs?", "answer": ["What Graph Neural Networks Cannot Learn: Depth vs Width"], "answer_arxiv_id": ["1907.03199"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_727"} +{"question": "Can you mention some studies that focus on learning features for image matching across appearance changes?", "answer": ["Fine-Grained Segmentation Networks: Self-Supervised Segmentation for Improved Long-Term Visual Localization", "Same Features, Different Day: Weakly Supervised Feature Learning for Seasonal Invariance"], "answer_arxiv_id": ["1908.06387", "2003.13431"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_728"} +{"question": "Any works that propose a staged method for expanding the hidden size of features in progressive training?", "answer": ["Staged Training for Transformer Language Models"], "answer_arxiv_id": ["2203.06211"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_729"} +{"question": "What works introduced VDNs (Value-Decomposition Networks) and QMIX?", "answer": ["Value-Decomposition Networks For Cooperative Multi-Agent Learning", "QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1706.05296", "1803.11485"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_730"} +{"question": "What are some research papers discussing the design of typical augmentations in computer vision?", "answer": ["Big Self-Supervised Models are Strong Semi-Supervised Learners"], "answer_arxiv_id": ["2006.10029"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_731"} +{"question": "Which works have achieved progress in image synthesis using diffusion models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "Collaborative Diffusion for Multi-Modal Face Generation and Editing", "ReVersion: Diffusion-Based Relation Inversion from Images"], "answer_arxiv_id": ["2112.10741", "2111.14822", "2205.11487", "2112.10752", "2307.01952", "2304.10530", "2303.13495"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_732"} +{"question": "What references provide early results in individual regret minimization in two-player zero-sum games?", "answer": ["Let’s be Honest: An Optimal No-Regret Framework for Zero-Sum Games"], "answer_arxiv_id": ["1802.04221"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_733"} +{"question": "Could you provide me studies where neural networks were trained at the cellular level to enhance the performance of the cell graph?", "answer": ["Hierarchical Graph Representations in Digital Pathology"], "answer_arxiv_id": ["2102.11057"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_734"} +{"question": "Which papers study the problem of tensor network-permutation selection(TN-PS)?", "answer": ["Qubit seriation: Improving data-model alignment using spectral ordering", "One-dimensional Tensor Network Recovery"], "answer_arxiv_id": ["2211.15978", "2207.10665v3"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_735"} +{"question": "What papers used various 3D representations for text-to-3D generative models?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "Rodin: A Generative Model for Sculpting 3D Digital Avatars Using\n Diffusion", "Shap-E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2212.08751", "2212.04493", "2212.06135", "2305.02463"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_736"} +{"question": "Could you provide me some of the research about ANN methods that utilize quantization and space partition methods?", "answer": ["Learning Space Partitions for Nearest Neighbor Search", "Accelerating Large-Scale Inference with Anisotropic Vector Quantization"], "answer_arxiv_id": ["1901.08544", "1908.10396"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_737"} +{"question": "Who proposed a fully sparse-convolutional approach for point instance segmentation more recently?", "answer": ["Top-Down Beats Bottom-Up in 3D Instance Segmentation"], "answer_arxiv_id": ["2302.02871"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_738"} +{"question": "Which studies have focused on the grounding capabilities of MLLMs?", "answer": ["Ferret: Refer and Ground Anything Anywhere at Any Granularity", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Kosmos-2: Grounding Multimodal Large Language Models to the World", "Pink: Unveiling the Power of Referential Comprehension for Multi-modal\n LLMs"], "answer_arxiv_id": ["2310.07704", "2306.15195", "2306.14824", "2310.00582"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_739"} +{"question": "What works have focused on simplifications of the inverse rendering problem in controlled settings?", "answer": ["Shape, Illumination, and Reflectance from Shading"], "answer_arxiv_id": ["2010.03592"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_740"} +{"question": "Which researchers used image-to-image translation to create labeled target-like source data to train the network?", "answer": ["CyCADA: Cycle-Consistent Adversarial Domain Adaptation"], "answer_arxiv_id": ["1711.03213"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_741"} +{"question": "Which papers introduced supervised learning methods reliant on PCA?", "answer": ["Unsupervised and Supervised Principal Component Analysis: Tutorial"], "answer_arxiv_id": ["1906.03148"], "source_meta": {"published_time": "20211101"}, "qid": "AutoScholarQuery_train_742"} +{"question": "Which works aim to improve the efficiency of IRL algorithms by learning Q functions and then differencing them across sequential states to extract a reward function?", "answer": ["IQ-Learn: Inverse soft-Q Learning for Imitation"], "answer_arxiv_id": ["2106.12142"], "source_meta": {"published_time": "20230326"}, "qid": "AutoScholarQuery_train_743"} +{"question": "Any works about the empirical study where models trained on code perform fairly well at common sense reasoning?", "answer": ["Language Models of Code are Few-Shot Commonsense Learners"], "answer_arxiv_id": ["2210.07128"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_744"} +{"question": "Which papers described the typical pipeline of feature matching?", "answer": ["SuperPoint: Self-Supervised Interest Point Detection and Description", "D2-Net: A Trainable CNN for Joint Detection and Description of Local\n Features", "R2D2: Repeatable and Reliable Detector and Descriptor"], "answer_arxiv_id": ["1712.07629", "1905.03561", "1906.06195"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_745"} +{"question": "Which works discuss text-to-video generation models?", "answer": ["GODIVA: Generating Open-DomaIn Videos from nAtural Descriptions", "NÜWA: Visual Synthesis Pre-training for Neural visUal World creAtion", "NUWA-Infinity: Autoregressive over Autoregressive Generation for Infinite Visual Synthesis", "NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generation", "CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformers", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models", "Video Diffusion Models"], "answer_arxiv_id": ["2104.14806", "2111.12417", "2207.09814", "2303.12346", "2205.15868", "2209.14792", "2304.08818", "2210.02303", "2204.03458"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_746"} +{"question": "What papers discussed the challenges when using text-to-image models, where users have to carefully select and compose sentences to achieve a certain visual style?", "answer": ["A Taxonomy of Prompt Modifiers for Text-To-Image Generation"], "answer_arxiv_id": ["2204.13988"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_747"} +{"question": "Which work developed the denoising diffusion probabilistic model (DDPM)?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_748"} +{"question": "What are the studies that approximate full gradient statistics matrix using Kronecker factored preconditioners to reduce time and memory complexity?", "answer": ["Shampoo: Preconditioned Stochastic Tensor Optimization", "Scalable Second Order Optimization for Deep Learning"], "answer_arxiv_id": ["1802.09568", "2002.09018"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_749"} +{"question": "What work showed that neural networks can suffer from extreme simplicity bias and rely on simple spurious features while ignoring the core features?", "answer": ["The Pitfalls of Simplicity Bias in Neural Networks"], "answer_arxiv_id": ["2006.07710"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_750"} +{"question": "What are some feature and feature space-based methods for few-shot class incremental learning?", "answer": ["Forward Compatible Few-Shot Class-Incremental Learning", "Few-Shot Class-Incremental Learning from an Open-Set Perspective", "Few-Shot Class-Incremental Learning by Sampling Multi-Phase Tasks", "Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by\n Finding Flat Minima", "Subspace Regularizers for Few-Shot Class Incremental Learning"], "answer_arxiv_id": ["2203.06953", "2208.00147", "2203.17030", "2111.01549", "2110.07059"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_751"} +{"question": "What studies focused on the representation learning in block MDPs?", "answer": ["Provably efficient RL with Rich Observations via Latent State Decoding", "Kinematic State Abstraction and Provably Efficient Rich-Observation Reinforcement Learning"], "answer_arxiv_id": ["1901.09018", "1911.05815"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_752"} +{"question": "What research introduced an end-to-end trainable stereo matching framework?", "answer": ["A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation"], "answer_arxiv_id": ["1512.02134"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_753"} +{"question": "What studies used retrieved neighbors from an external database as conditional information to train diffusion models?", "answer": ["Semi-Parametric Neural Image Synthesis"], "answer_arxiv_id": ["2204.11824"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_754"} +{"question": "Which paper proposed a retrieval-based method for AI-text detection that includes the collection of historical responses from language models?", "answer": ["Paraphrasing evades detectors of AI-generated text, but retrieval is an\n effective defense"], "answer_arxiv_id": ["2303.13408"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_755"} +{"question": "Which works propose that language models should serve as reliable knowledge bases?", "answer": ["Language Models as Knowledge Bases?"], "answer_arxiv_id": ["1909.01066"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_756"} +{"question": "What works used CNN-based and Transformer-based models for video action recognition?", "answer": ["Convolutional Two-Stream Network Fusion for Video Action Recognition", "The Kinetics Human Action Video Dataset", "SlowFast Networks for Video Recognition", "GTA: Global Temporal Attention for Video Action Understanding", "Rethinking Resolution in the Context of Efficient Video Recognition", "Is Space-Time Attention All You Need for Video Understanding?", "Multiscale Vision Transformers", "Video Swin Transformer"], "answer_arxiv_id": ["1604.06573", "1705.06950", "1812.03982", "2012.08510", "2209.12797", "2102.05095", "2104.11227", "2106.13230"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_757"} +{"question": "What studies displayed notable advancements in Diffusion Models (DMs) for image generation using classifier guidance and classifier-free guidance (CFG) methods?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2011.13456", "2006.11239", "2105.05233", "2207.12598"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_758"} +{"question": "What studies focused on surface matching for code evaluation?", "answer": ["CodeXGLUE: A Machine Learning Benchmark Dataset for Code Understanding and Generation", "Learning to Mine Aligned Code and Natural Language Pairs from Stack Overflow", "MCoNaLa: A Benchmark for Code Generation from Multiple Natural Languages", "CodeSearchNet Challenge Evaluating the State of Semantic Code Search"], "answer_arxiv_id": ["2102.04664", "1805.08949", "2203.08388", "1909.09436"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_759"} +{"question": "Could you provide me some works about formalizing the first generation process for PLL?", "answer": ["Provably Consistent Partial-Label Learning"], "answer_arxiv_id": ["2007.08929"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_760"} +{"question": "Which studies propose methods of re-balancing the number of samples in tackling the problem of long-tailed data?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition", "Learning Fast Sample Re-weighting Without Reward Data"], "answer_arxiv_id": ["1910.09217", "2109.03216"], "source_meta": {"published_time": "20221230"}, "qid": "AutoScholarQuery_train_761"} +{"question": "What studies have been conducted on domain generalization?", "answer": ["Adversarial Learning for Zero-shot Domain Adaptation", "In Search of Lost Domain Generalization", "Wilds: A Benchmark of in-the-Wild Distribution Shifts"], "answer_arxiv_id": ["2009.05214", "2007.01434", "2012.07421"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_762"} +{"question": "What papers have established finite-sample efficiency for MARL using a generative model?", "answer": ["Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity", "Minimax-Optimal Multi-Agent RL in Markov Games With a Generative Model"], "answer_arxiv_id": ["2007.07461", "2208.10458"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_763"} +{"question": "Which works are available on 2D molecular graph self-supervised learning?", "answer": ["Strategies for Pre-training Graph Neural Networks", "InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization", "Graph Contrastive Learning with Augmentations", "N-Gram Graph: Simple Unsupervised Representation for Graphs, with Applications to Molecules", "GPT-GNN: Generative Pre-Training of Graph Neural Networks"], "answer_arxiv_id": ["1905.12265", "1908.01000", "2010.13902", "1806.09206", "2006.15437"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_764"} +{"question": "Which work is about the successful application of latent Dirichlet Allocation models for various applications in media?", "answer": ["Score-based Generative Modeling in Latent Space", "LION: Latent Point Diffusion Models for 3D Shape Generation", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2106.05931", "2210.06978", "2112.10752"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_765"} +{"question": "Which works have investigated the comparison of models using their internal representations?", "answer": ["SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability", "Insights on representational similarity in neural networks with canonical correlation", "Deconfounded Representation Similarity for Comparison of Neural Networks", "Similarity of Neural Network Representations Revisited", "Convergent Learning: Do different neural networks learn the same representations?", "Revisit Similarity of Neural Network Representations From Graph Perspective", "Similarity and Matching of Neural Network Representations", "Revisiting Model Stitching to Compare Neural Representations"], "answer_arxiv_id": ["1706.05806", "1806.05759", "2202.00095", "1905.00414", "1511.07543", "2111.11165", "2110.14633", "2106.07682"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_766"} +{"question": "Are there any works that tried to solve the OT problems using splitting methods as an alternative to the Sinkhorn algorithm?", "answer": ["Accelerated Bregman Primal-Dual methods applied to Optimal Transport and Wasserstein Barycenter problems", "A Fast and Accurate Splitting Method for Optimal Transport: Analysis and Implementation"], "answer_arxiv_id": ["2203.00802", "2110.11738"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_767"} +{"question": "Which studies have used conditional Generative Adversarial Networks (GANs) for image translation tasks?", "answer": ["DualGAN: Unsupervised Dual Learning for Image-to-Image Translation", "Unsupervised Image-to-Image Translation Networks", "High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs", "Semantic Image Synthesis with Spatially-Adaptive Normalization"], "answer_arxiv_id": ["1704.02510", "1703.00848", "1711.11585", "1903.07291"], "source_meta": {"published_time": "20230802"}, "qid": "AutoScholarQuery_train_768"} +{"question": "What are the works that introduced adapter approaches which are methods for transfer learning?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Towards a Unified View of Parameter-Efficient Transfer Learning"], "answer_arxiv_id": ["1902.00751", "2110.04366"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_769"} +{"question": "Which works investigate the use of generative models such as GAN and Diffusion models for entire image synthesis in deepfake generation?", "answer": ["Progressive Growing of GANs for Improved Quality, Stability, and\n Variation", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["1710.10196", "1812.04948", "2006.11239", "2010.02502", "2112.10752"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_770"} +{"question": "Can you provide research that have used representation from multiview video using time contrastive learning?", "answer": ["Time-Contrastive Networks: Self-Supervised Learning from Video"], "answer_arxiv_id": ["1704.06888"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_771"} +{"question": "What are the recent works involving 3D-AD methods?", "answer": ["Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors", "Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection", "Asymmetric Student-Teacher Networks for Industrial Anomaly Detection", "Multimodal Industrial Anomaly Detection via Hybrid Fusion", "Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detection"], "answer_arxiv_id": ["2202.11660", "2203.05550", "2210.07829", "2303.00601", "2303.13194v1"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_772"} +{"question": "Which research papers employed techniques for uncertainty quantification in physics-informed learning?", "answer": ["Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems", "Adversarial Uncertainty Quantification in Physics-Informed Neural Networks", "Physics-Informed Generative Adversarial Networks for Stochastic Differential Equations", "Wasserstein Generative Adversarial Uncertainty Quantification in Physics-Informed Neural Networks", "B-PINNs: Bayesian Physics-Informed Neural Networks for Forward and Inverse PDE Problems with Noisy Data"], "answer_arxiv_id": ["1809.08327", "1811.04026", "1811.02033", "2108.13054", "2003.06097"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_773"} +{"question": "What are some studies where they incorporated symmetries into deep neural network architectures?", "answer": ["Equivariance Through Parameter-Sharing", "On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups"], "answer_arxiv_id": ["1702.08389", "1802.03690", "2104.09459"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_774"} +{"question": "What papers address the topic of in-domain generalization behaviors when initializing from a single pretrained model?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_775"} +{"question": "Which works are about automatically populating static 3D scenes with 3D humans?", "answer": ["Populating 3D Scenes by Learning Human-Scene Interaction", "Generating 3D People in Scenes without People", "PLACE: Proximity Learning of Articulation and Contact in 3D Environments"], "answer_arxiv_id": ["2012.11581", "1912.02923", "2008.05570"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_776"} +{"question": "Which studies analyzed the feature properties of MRP in SSL analysis?", "answer": ["Predicting What You Already Know Helps: Provable Self-Supervised Learning", "How to Understand Masked Autoencoders"], "answer_arxiv_id": ["2008.01064", "2202.03670"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_777"} +{"question": "Which papers proposed recurrent-based models in deterministic predictive models?", "answer": ["Convolutional LSTM Network: A Machine Learning Approach for\n Precipitation Nowcasting", "Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model", "Temporal Attention Unit: Towards Efficient Spatiotemporal Predictive\n Learning"], "answer_arxiv_id": ["1506.04214", "1706.03458", "2206.12126"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_778"} +{"question": "What are some examples of datasets that include human motion data along with audio signals?", "answer": ["AI Choreographer: Music Conditioned 3D Dance Generation with AIST++", "Transflower: probabilistic autoregressive dance generation with\n multimodal attention"], "answer_arxiv_id": ["2101.08779", "2106.13871"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_779"} +{"question": "Which paper used trajectory sampling with probabilistic dynamics models in Reinforcement Learning?", "answer": ["Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models"], "answer_arxiv_id": ["1805.12114"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_780"} +{"question": "Which works focused on action representation using manually-crafted attributes in zero-shot video action recognition?", "answer": ["Zero-Shot Activity Recognition with Verb Attribute Induction"], "answer_arxiv_id": ["1707.09468"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_781"} +{"question": "What studies discuss the use of large language models (LLMs) as an interface to solving complex reasoning tasks with tools?", "answer": ["Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language", "Binding Language Models in Symbolic Languages", "Toolformer: Language Models Can Teach Themselves to Use Tools", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face", "Chameleon: Plug-and-Play Compositional Reasoning with Large Language\n Models"], "answer_arxiv_id": ["2204.00598", "2210.02875", "2302.04761", "2303.17580", "2304.09842"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_782"} +{"question": "Which research uses GNN to predict whether a query is a child of an anchor concept?", "answer": ["TaxoExpan: Self-supervised Taxonomy Expansion with Position-Enhanced\n Graph Neural Network"], "answer_arxiv_id": ["2001.09522"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_783"} +{"question": "What researches provided other methods for the covariate shift adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks", "A One-step Approach to Covariate Shift Adaptation", "Near-Optimal Linear Regression under Distribution Shift"], "answer_arxiv_id": ["1505.07818", "2007.04043", "2106.12108"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_784"} +{"question": "In which paper the researcher formulate classic MAML as BLO problem?", "answer": ["Meta-Learning with Implicit Gradients"], "answer_arxiv_id": ["1909.04630"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_785"} +{"question": "What research has been done to understand the implicit bias of gradient descent dynamics in non-linear neural networks?", "answer": ["Implicit Bias of Gradient Descent for Wide Two-layer Neural Networks Trained with the Logistic Loss"], "answer_arxiv_id": ["2002.04486"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_786"} +{"question": "What papers can provide information about transformers in the field of segmentation?", "answer": ["Attention Is All You Need", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["1706.03762", "2010.11929"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_787"} +{"question": "What are the examples of two-stage detectors in object detection task?", "answer": ["Fast R-CNN", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks", "Feature Pyramid Networks for Object Detection", "R-FCN: Object Detection via Region-based Fully Convolutional Networks"], "answer_arxiv_id": ["1504.08083", "1506.01497", "1612.03144", "1605.06409"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_788"} +{"question": "Which papers have utilized auxiliary task gradients and measured their similarity with the main task gradients to find beneficial auxiliary tasks?", "answer": ["Adapting Auxiliary Losses Using Gradient Similarity", "Auxiliary Task Reweighting for Minimum-data Learning"], "answer_arxiv_id": ["1812.02224", "2010.08244"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_789"} +{"question": "Which work focused on augmenting a policy gradient algorithm with demonstration data to solve dexterous manipulation tasks?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations"], "answer_arxiv_id": ["1709.10087"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_790"} +{"question": "What studies exploited point-supervised weakly-supervised oriented object detection approaches?", "answer": ["Training object class detectors with click supervision", "UFO$^2$: A Unified Framework towards Omni-supervised Object Detection", "Points as Queries: Weakly Semi-supervised Object Detection by Points", "Point-to-Box Network for Accurate Object Detection via Single Point\n Supervision", "Mapping Degeneration Meets Label Evolution: Learning Infrared Small\n Target Detection with Single Point Supervision", "Learning Remote Sensing Object Detection with Single Point Supervision", "Pointly-Supervised Panoptic Segmentation", "Point2Mask: Point-supervised Panoptic Segmentation via Optimal Transport", "What's the Point: Semantic Segmentation with Point Supervision", "Pointly-Supervised Instance Segmentation", "Object Localization under Single Coarse Point Supervision", "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["1704.06189", "2010.10804", "2104.07434", "2207.06827", "2304.01484", "2305.14141", "2210.13950", "2308.01779", "1506.02106", "2104.06404", "2203.09338", "2401.14159"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_791"} +{"question": "Which work argues that additions made to work around shortcut learning problem can make optimization very challenging?", "answer": ["Rich Feature Construction for the Optimization-Generalization Dilemma"], "answer_arxiv_id": ["2203.15516"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_792"} +{"question": "Which studies worked on the expansion of functional map-based methods for accommodating partial shapes?", "answer": ["Non-Rigid Puzzles", "Partial Functional Correspondence"], "answer_arxiv_id": ["2011.13076", "1506.05274"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_793"} +{"question": "Which papers are categorized as knowledge-based methods in unsuperised RL, targeting states that maximize the agent’s model error or uncertainty?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction", "Exploration by Random Network Distillation"], "answer_arxiv_id": ["1705.05363", "1810.12894"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_794"} +{"question": "What work employed a fixed green list to improve the robustness of KGW against paraphrase attacks and editing attacks?", "answer": ["Provable Robust Watermarking for AI-Generated Text"], "answer_arxiv_id": ["2306.17439"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_795"} +{"question": "What papers proposed the exploration of the zero-shot ability of CLIP in open-domain tasks?", "answer": ["CLIP-Art: Contrastive Pre-training for Fine-Grained Art Classification", "Tip-Adapter: Training-free Adaption of CLIP for Few-shot Classification", "Zero-Shot Out-of-Distribution Detection Based on the Pre-trained Model\n CLIP", "Open-Vocabulary Multi-Label Classification via Multi-Modal Knowledge\n Transfer", "CLIP-ReID: Exploiting Vision-Language Model for Image Re-Identification\n without Concrete Text Labels", "A Simple Framework for Open-Vocabulary Segmentation and Detection", "Learning to Prompt for Open-Vocabulary Object Detection with\n Vision-Language Model", "EdaDet: Open-Vocabulary Object Detection Using Early Dense Alignment", "Robust Region Feature Synthesizer for Zero-Shot Object Detection", "WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Side Adapter Network for Open-Vocabulary Semantic Segmentation"], "answer_arxiv_id": ["2204.14244", "2207.09519", "2109.02748", "2207.01887", "2211.13977", "2303.08131", "2203.14940", "2309.01151", "2201.00103", "2303.14814", "2210.04150", "2302.12242"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_796"} +{"question": "Are there any studies about using Codex, CodeGen, and InCoder in zero-shot code generation?", "answer": ["Evaluating Large Language Models Trained on Code", "InCoder: A Generative Model for Code Infilling and Synthesis"], "answer_arxiv_id": ["2107.03374", "2204.05999"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_797"} +{"question": "What works have studied 2D amodal representation through a supervised learning paradigm?", "answer": ["Semantic Amodal Segmentation", "Learning to See the Invisible: End-to-End Trainable Amodal Instance\n Segmentation"], "answer_arxiv_id": ["1509.01329", "1804.08864"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_798"} +{"question": "Could you provide references for different models of Graph Neural Networks?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "How Powerful are Graph Neural Networks?", "Graph Attention Networks"], "answer_arxiv_id": ["1609.02907", "1810.00826", "1710.10903"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_799"} +{"question": "What works utilize discrete coordinate tokens to encode spatial information in LVLMs?", "answer": ["Pix2seq: A Language Modeling Framework for Object Detection", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Kosmos-2: Grounding Multimodal Large Language Models to the World"], "answer_arxiv_id": ["2109.10852", "2202.03052", "2206.08916", "2306.15195", "2306.14824"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_800"} +{"question": "Which works are related to diffusion models for videos?", "answer": ["MCVD: Masked Conditional Video Diffusion for Prediction, Generation, and Interpolation", "Video Diffusion Models", "Diffusion Probabilistic Modeling for Video Generation", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Flexible Diffusion Modeling of Long Videos"], "answer_arxiv_id": ["2205.09853", "2204.03458", "2203.09481", "2209.14792", "2205.11495"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_801"} +{"question": "Could you name some research that strive for joint learning in dense video captioning?", "answer": ["Jointly Localizing and Describing Events for Dense Video Captioning", "End-to-end Dense Video Captioning as Sequence Generation", "Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense\n Video Captioning", "End-to-End Dense Video Captioning with Masked Transformer"], "answer_arxiv_id": ["1804.08274", "2204.08121", "2302.14115", "1804.00819"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_802"} +{"question": "Could you provide some works about prompt-based approaches for few-shot learning?", "answer": ["Language Models are Few-Shot Learners", "Making Pre-trained Language Models Better Few-shot Learners", "Knowledgeable Prompt-tuning: Incorporating Knowledge into Prompt Verbalizer for Text Classification", "Cutting Down on Prompts and Parameters: Simple Few-Shot Learning with Language Models", "Noisy Channel Language Model Prompting for Few-Shot Text Classification", "Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference", "Few-Shot Text Generation with Natural Language Instructions", "Improving and Simplifying Pattern Exploiting Training"], "answer_arxiv_id": ["2005.14165", "2012.15723", "2108.02035", "2106.13353", "2108.04106", "2001.07676", "2012.11926", "2103.11955"], "source_meta": {"published_time": "20221106"}, "qid": "AutoScholarQuery_train_803"} +{"question": "What research introduces a novel reconstruction and calibration framework for enhancing the performance of quantized diffusion models?", "answer": ["Temporal Dynamic Quantization for Diffusion Models", "Towards Accurate Post-training Quantization for Diffusion Models"], "answer_arxiv_id": ["2306.02316", "2305.18723"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_804"} +{"question": "Which study identifies useful 'skills' for news summarization and developed a pre-training task accordingly?", "answer": ["Does Pretraining for Summarization Require Knowledge Transfer?"], "answer_arxiv_id": ["2109.04953"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_805"} +{"question": "Which works related to extracting invariant geometric features from atomic positions using GNNs?", "answer": ["PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges", "Directional Message Passing for Molecular Graphs", "Spherical Message Passing for 3D Graph Networks", "Rotation Invariant Graph Neural Networks using Spin Convolutions", "Learning 3D Representations of Molecular Chirality with Invariance to Bond Rotations"], "answer_arxiv_id": ["1902.08408", "2003.03123", "2102.05013v5", "2106.09575", "2110.04383"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_806"} +{"question": "Could you name a work that uses matrix Fisher distribution on SO​(3) over rotation matrices for deep rotation regression?", "answer": ["Probabilistic orientation estimation with matrix Fisher distributions"], "answer_arxiv_id": ["2006.09740v1"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_807"} +{"question": "Could you provide me any paper that proposed a physics-based motion projection module to instill the laws of physics into the denoising diffusion process for motion generation?", "answer": ["PhysDiff: Physics-Guided Human Motion Diffusion Model"], "answer_arxiv_id": ["2212.02500"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_808"} +{"question": "Could you provide me some studies that focus on adjusting cross-entropy loss to AT setting?", "answer": ["Rethinking Softmax Cross-Entropy Loss for Adversarial Robustness", "Boosting Adversarial Training with Hypersphere Embedding"], "answer_arxiv_id": ["1905.10626", "2002.08619"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_809"} +{"question": "Which references have reported the usage of knowledge distillation in embedding-based methods for anomaly detectors?", "answer": ["Anomaly Detection via Reverse Distillation from One-Class Embedding", "Exploring Plain ViT Reconstruction for Multi-class Unsupervised Anomaly\n Detection"], "answer_arxiv_id": ["2201.10703", "2312.07495"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_810"} +{"question": "What papers focus on conjunctive logical queries?", "answer": ["Embedding Logical Queries on Knowledge Graphs"], "answer_arxiv_id": ["1806.01445"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_811"} +{"question": "Any researches investigate multi-scale feature extraction module in the field of image segmentation?", "answer": ["DeepLab: Semantic Image Segmentation with Deep Convolutional Nets,\n Atrous Convolution, and Fully Connected CRFs", "Suppress and Balance: A Simple Gated Network for Salient Object\n Detection", "A Simple Pooling-Based Design for Real-Time Salient Object Detection", "Pyramid Scene Parsing Network"], "answer_arxiv_id": ["1606.00915", "2007.08074", "1904.09569", "1612.01105"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_812"} +{"question": "Which research papers discuss Speech Self-supervised Learning (SSL) in the context of speech-text multimodal LLM?", "answer": ["Self-Supervised Speech Representation Learning: A Review", "SUPERB: Speech processing Universal PERformance Benchmark", "SUPERB-SG: Enhanced Speech processing Universal PERformance Benchmark\n for Semantic and Generative Capabilities", "On the Utility of Self-supervised Models for Prosody-related Tasks"], "answer_arxiv_id": ["2205.10643", "2105.01051", "2203.06849", "2210.07185"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_813"} +{"question": "Which study proposes to maximize the mutual information between the input data and crowdsourced labels?", "answer": ["Max-MIG: an Information Theoretic Approach for Joint Learning from\n Crowds"], "answer_arxiv_id": ["1905.13436"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_814"} +{"question": "Any works about fabric manipulation using learning visual correspondence?", "answer": ["Learning Dense Visual Correspondences in Simulation to Smooth and Fold\n Real Fabrics"], "answer_arxiv_id": ["2003.12698"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_815"} +{"question": "What works launched perturbation attack and patch attacks in MDE?", "answer": ["Monocular Depth Estimators: Vulnerabilities and Attacks"], "answer_arxiv_id": ["2005.14302"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_816"} +{"question": "Are there any research works about IPM GANs that estimate the Wasserstein-1 metric?", "answer": ["Improved Training of Wasserstein GANs"], "answer_arxiv_id": ["1704.00028"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_817"} +{"question": "What works utilize symbolic (non-parametric) identifiability methods for point identification?", "answer": ["What Counterfactuals Can Be Tested", "Nested Counterfactual Identification from Arbitrary Surrogate Experiments"], "answer_arxiv_id": ["1206.5294", "2107.03190"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_818"} +{"question": "Who initially introduced the concept of Relative Position Encoding (RPE)?", "answer": ["Self-Attention with Relative Position Representations"], "answer_arxiv_id": ["1803.02155"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_819"} +{"question": "What papers have shown that computing the coarse correlated equilibrium that is Markov stationary can be PPAD-hard for infinite-horizon discounted Markov games?", "answer": ["The Complexity of Markov Equilibrium in Stochastic Games", "The Complexity of Infinite-Horizon General-Sum Stochastic Games"], "answer_arxiv_id": ["2204.03991", "2204.04186"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_820"} +{"question": "Could you identify works focusing on lighting decomposition for appearance editing in the context of 3D scene editing?", "answer": ["Neural Inverse Rendering of an Indoor Scene from a Single Image", "Neural-PIL: Neural Pre-Integrated Lighting for Reflectance Decomposition", "NeRFactor: Neural Factorization of Shape and Reflectance Under an Unknown Illumination", "DIB-R++: Learning to Predict Lighting and Material with a Hybrid Differentiable Renderer"], "answer_arxiv_id": ["1901.02453", "2110.14373", "2106.01970", "2111.00140"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_821"} +{"question": "Can you provide works that derive stability results for graph classifiers under isomorphism invariant optimal transport distances?", "answer": ["Tree Mover’s Distance: Bridging Graph Metrics and Stability of Graph Neural Networks", "The Weisfeiler-Lehman Distance: Reinterpretation and Connection with GNNs"], "answer_arxiv_id": ["2210.01906", "2302.00713"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_822"} +{"question": "Can you name research papers focusing on causal inference for natural language understanding?", "answer": ["Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimates", "Challenges of Using Text Classifiers for Causal Inference"], "answer_arxiv_id": ["2005.00649", "1810.00956"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_823"} +{"question": "What papers have integrated deep learning with the traditional geometry framework for improving camera tracking and mapping?", "answer": ["DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras", "CodeSLAM - Learning a Compact, Optimisable Representation for Dense\n Visual SLAM", "SceneCode: Monocular Dense Semantic Reconstruction using Learned Encoded\n Scene Representations", "NodeSLAM: Neural Object Descriptors for Multi-View Shape Reconstruction"], "answer_arxiv_id": ["2108.10869", "1804.00874", "1903.06482", "2004.04485"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_824"} +{"question": "Which works prefer to process batches of event data via an intermediat representation and use the FEN or FIT methods?", "answer": ["High Speed and High Dynamic Range Video with an Event Camera", "EV-FlowNet: Self-Supervised Optical Flow Estimation for Event-based Cameras"], "answer_arxiv_id": ["1906.07165", "1802.06898"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_825"} +{"question": "What studies create synthetic datasets featuring clothed humans?", "answer": ["The Power of Points for Modeling Humans in Clothing", "TailorNet: Predicting Clothing in 3D as a Function of Human Pose, Shape\n and Garment Style", "CLOTH3D: Clothed 3D Humans", "BEDLAM: A Synthetic Dataset of Bodies Exhibiting Detailed Lifelike\n Animated Motion", "Towards Multi-Layered 3D Garments Animation"], "answer_arxiv_id": ["2109.01137", "2003.04583", "1912.02792", "2306.16940", "2305.10418"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_826"} +{"question": "What are the recent studies in multi-view 3D object detection methods?", "answer": ["Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by\n Implicitly Unprojecting to 3D", "BEVDet: High-performance Multi-camera 3D Object Detection in\n Bird-Eye-View", "BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers", "FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection", "PETR: Position Embedding Transformation for Multi-View 3D Object\n Detection"], "answer_arxiv_id": ["2008.05711", "2112.11790", "2203.17270", "2104.10956", "2203.05625"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_827"} +{"question": "What works have utilized meta learning as an approach to domain generalization?", "answer": ["Efficient Domain Generalization via Common-Specific Low-Rank\n Decomposition", "Learning to Learn with Variational Information Bottleneck for Domain\n Generalization", "Meta-Learning for Domain Generalization in Semantic Parsing"], "answer_arxiv_id": ["2003.12815", "2007.07645", "2010.11988"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_828"} +{"question": "Which work uses the cross-attention mechanism for conditioning in Stable Diffusion model?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_829"} +{"question": "Can you name a work that leverages an image prior to bridge multi-modal embedding spaces in text-to-image generation?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2204.06125"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_830"} +{"question": "Are there research papers on the use of partial autoregression in CARGAN for improvement in pitch and periodicity accuracy?", "answer": ["Chunked Autoregressive GAN for Conditional Waveform Synthesis", "WaveFlow: A Compact Flow-based Model for Raw Audio"], "answer_arxiv_id": ["2110.10139", "1912.01219"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_831"} +{"question": "Which papers developed large language models for code generation?", "answer": ["Evaluating Large Language Models Trained on Code", "A Systematic Evaluation of Large Language Models of Code", "InCoder: A Generative Model for Code Infilling and Synthesis", "Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2107.03374", "2202.13169", "2204.05999", "2203.07814"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_832"} +{"question": "Which papers present physics-based approaches for garment draping?", "answer": ["Deep Physics-aware Inference of Cloth Deformation for Monocular Human Performance Capture"], "answer_arxiv_id": ["2011.12866"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_833"} +{"question": "Which works focus on leveraging visual augmentation via image editing to train more generalizable policies in robotic tasks?", "answer": ["GenAug: Retargeting behaviors to unseen situations via Generative Augmentation", "CACTI: A Framework for Scalable Multi-Task Multi-Scene Visual Imitation Learning", "Scaling Robot Learning with Semantically Imagined Experience", "Understanding Domain Randomization for Sim-to-real Transfer"], "answer_arxiv_id": ["2302.06671", "2212.05711", "2302.11550", "2110.03239"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_834"} +{"question": "What is the reference for the study where model soups were first introduced?", "answer": ["Model soups: averaging weights of multiple fine-tuned models improves\n accuracy without increasing inference time"], "answer_arxiv_id": ["2203.05482"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_835"} +{"question": "What study combines adversarial training with larger perturbation radii and a ReLU-stability regularization based on the Box relaxation?", "answer": ["IBP Regularization for Verified Adversarial Robustness via Branch-and-Bound"], "answer_arxiv_id": ["2206.14772v2"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_836"} +{"question": "Could you provide me some works about generative self-supervised learning methods?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["2106.08254", "2111.06377", "1810.04805"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_837"} +{"question": "What studies are related to sampling impulse responses at multiple places to synthesize the RIR?", "answer": ["Few-Shot Audio-Visual Learning of Environment Acoustics", "Learning Neural Acoustic Fields"], "answer_arxiv_id": ["2206.04006", "2204.00628"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_838"} +{"question": "What works analyze the standard embodied question-answering frameworks?", "answer": ["Embodied Question Answering", "IQA: Visual Question Answering in Interactive Environments", "Multi-Target Embodied Question Answering"], "answer_arxiv_id": ["1711.11543", "1712.03316", "1904.04686"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_839"} +{"question": "Which works provide theoretical guarantees for continuous-time RL with linear systems?", "answer": ["Convergence and sample complexity of gradient methods for the model-free linear quadratic regulator problem", "Logarithmic Regret for Episodic Continuous-Time Linear-Quadratic Reinforcement Learning over a Finite-Time Horizon"], "answer_arxiv_id": ["1912.11899", "2006.15316"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_840"} +{"question": "Are there any studies about clean-label backdoor attacks?", "answer": ["Sleeper Agent: Scalable Hidden Trigger Backdoors for Neural Networks Trained from Scratch", "Untargeted Backdoor Watermark: Towards Harmless and Stealthy Dataset Copyright Protection"], "answer_arxiv_id": ["2106.08970", "2210.00875"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_841"} +{"question": "What works proposed predicting electron and phonon DOS based on features of crystalline materials?", "answer": ["Direct prediction of phonon density of states with Euclidean neural networks", "Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "Clebsch–Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network", "3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data"], "answer_arxiv_id": ["2009.05163", "1802.08219", "1806.09231", "1807.02547"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_842"} +{"question": "Could you provide me with some works that employed markerless mocap systems with multi-view cameras to capture 3D human motion?", "answer": ["Single-Shot Multi-Person 3D Pose Estimation From Monocular RGB", "Multi-Person Extreme Motion Prediction"], "answer_arxiv_id": ["1712.03453", "2105.08825"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_843"} +{"question": "What studies have leveraged the prior knowledge to guide implicit function representation to enhance reconstruction accuracy?", "answer": ["PaMIR: Parametric Model-Conditioned Implicit Representation for Image-based Human Reconstruction", "ICON: Implicit Clothed humans Obtained from Normals", "ECON: Explicit Clothed humans Optimized via Normal integration", "Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing"], "answer_arxiv_id": ["2007.03858", "2112.09127", "2212.07422", "2204.08906v1"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_844"} +{"question": "What work introduces the method DeCap used in zero-shot captioning?", "answer": ["DeCap: Decoding CLIP Latents for Zero-Shot Captioning via Text-Only\n Training"], "answer_arxiv_id": ["2303.03032"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_845"} +{"question": "Which works demonstrate the use of feature maps for tasks such as few-shot semantic segmentation?", "answer": ["Semantic Segmentation with Generative Models: Semi-Supervised Learning and Strong Out-of-Domain Generalization", "DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort", "Linear Semantics in Generative Adversarial Networks", "Repurposing GANs for One-shot Semantic Part Segmentation", "SG-One: Similarity Guidance Network for One-Shot Semantic Segmentation", "Label-Efficient Semantic Segmentation with Diffusion Models"], "answer_arxiv_id": ["2104.05833", "2104.06490", "2104.00487", "2103.04379", "1810.09091", "2112.03126"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_846"} +{"question": "Which papers consider compensation for computation, energy, and other costs related to training as incentives for data sharing in machine learning?", "answer": ["Incentive Mechanisms for Federated Learning: From Economic and Game Theoretic Perspective", "Task-Load-Aware Game-Theoretic Framework for Wireless Federated Learning"], "answer_arxiv_id": ["2111.11850", "2205.13888"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_847"} +{"question": "Which studies have adopted the Wasserstein distance for imitation learning or skill discovery?", "answer": ["Primal Wasserstein Imitation Learning", "Watch and Match: Supercharging Imitation with Regularized Optimal Transport", "Wasserstein Adversarial Imitation Learning", "Wasserstein Distance Maximizing Intrinsic Control", "Cross-Domain Imitation Learning via Optimal Transport"], "answer_arxiv_id": ["2006.04678", "2206.15469", "1906.08113", "2110.15331", "2110.03684"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_848"} +{"question": "Could you provide me studies explicitly estimating the optical flow in temporal sliding-window based VSR methods?", "answer": ["Real-Time Video Super-Resolution with Spatio-Temporal Networks and\n Motion Compensation", "Detail-revealing Deep Video Super-resolution"], "answer_arxiv_id": ["1611.05250", "1704.02738"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_849"} +{"question": "What recent techniques in 3D style transfer were used to ensure coherent stylization across images rendered from multiple viewpoints?", "answer": ["Learning to Stylize Novel Views", "3D Photo Stylization: Learning to Generate Stylized Novel Views from a\n Single Image", "StyleMesh: Style Transfer for Indoor 3D Scene Reconstructions", "StyleRF: Zero-shot 3D Style Transfer of Neural Radiance Fields", "SNeRF: Stylized Neural Implicit Representations for 3D Scenes", "ARF: Artistic Radiance Fields", "Transforming Radiance Field with Lipschitz Network for Photorealistic 3D\n Scene Stylization", "Unified Implicit Neural Stylization", "Locally Stylized Neural Radiance Fields", "Stylizing 3D Scene via Implicit Representation and HyperNetwork", "Ref-NPR: Reference-Based Non-Photorealistic Radiance Fields for\n Controllable Scene Stylization"], "answer_arxiv_id": ["2105.13509", "2112.00169", "2112.01530", "2303.10598", "2207.02363", "2206.06360", "2303.13232", "2204.01943", "2309.10684", "2105.13016", "2212.02766"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_850"} +{"question": "Could you provide some research works that on multi-domain generalization constraints conditional covariance shifts?", "answer": ["RobustNet: Improving Domain Generalization in Urban-Scene Segmentation\n via Instance Selective Whitening"], "answer_arxiv_id": ["2103.15597"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_851"} +{"question": "What are the studies that map images to the token input space of the LLMs considering images as 'foreign languages'?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models", "Language Quantized AutoEncoders: Towards Unsupervised Text-Image\n Alignment", "SPAE: Semantic Pyramid AutoEncoder for Multimodal Generation with Frozen\n LLMs"], "answer_arxiv_id": ["2106.13884", "2302.00902", "2306.17842"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_852"} +{"question": "Which work first introduced the approachability task for VMDPs?", "answer": ["Reinforcement Learning with Convex Constraints"], "answer_arxiv_id": ["1906.09323"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_853"} +{"question": "Are there any studies showing superior performance across diverse natural language processing tasks due to the application of Large Language Models (LLMs)?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models", "OPT: Open Pre-trained Transformer Language Models", "PaLM: Scaling Language Modeling with Pathways", "Tool Learning with Foundation Models"], "answer_arxiv_id": ["2005.14165", "2302.13971", "2205.01068", "2204.02311", "2304.08354"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_854"} +{"question": "What studies suggest using normalizing flow as a method for 3D generation?", "answer": ["PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "Discrete Point Flow Networks for Efficient Point Cloud Generation", "SoftFlow: Probabilistic Framework for Normalizing Flow on Manifolds"], "answer_arxiv_id": ["1906.12320", "2007.10170", "2006.04604"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_855"} +{"question": "Could you mention which work proposed the amplitude spectrum transformation in the feature space for OCDA?", "answer": ["Amplitude Spectrum Transformation for Open Compound Domain Adaptive Semantic Segmentation"], "answer_arxiv_id": ["2202.04287"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_856"} +{"question": "Are there any studies that introduced a coarse-to-fine strategy in goal generation?", "answer": ["From Coarse to Fine: Robust Hierarchical Localization at Large Scale", "NeuMap: Neural Coordinate Mapping by Auto-Transdecoder for Camera\n Localization"], "answer_arxiv_id": ["1812.03506", "2211.11177"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_857"} +{"question": "Could you provide some examples of papers that implement top-down approaches in multi-person human pose estimation?", "answer": ["Mask R-CNN", "Deep High-Resolution Representation Learning for Human Pose Estimation", "ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation"], "answer_arxiv_id": ["1703.06870", "1902.09212", "2204.12484"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_858"} +{"question": "Which papers expands on the context of linear contextual bandits in regards to corruption-robust bandits and RL?", "answer": ["Stochastic Linear Bandits Robust to Adversarial Attacks", "Robust Stochastic Linear Contextual Bandits Under Adversarial Attacks", "Adapting to Misspecification in Contextual Bandits", "Achieving Near Instance-Optimality and Minimax-Optimality in Stochastic and Adversarial Linear Bandits Simultaneously", "Linear Contextual Bandits with Adversarial Corruptions", "Robust Lipschitz Bandits to Adversarial Corruptions"], "answer_arxiv_id": ["2007.03285", "2106.02978v3", "2107.05745v1", "2102.05858", "2110.12615", "2305.18543"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_859"} +{"question": "Which research works introduced text-guided image editing based on GAN models?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["1812.04948"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_860"} +{"question": "Who provided the generalization guarantees for WDRO models with the radius scaling as 1/n in case of linear models?", "answer": ["Regularization via Mass Transportation"], "answer_arxiv_id": ["1710.10016"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_861"} +{"question": "What works discuss the Constrained Markov Decision Process (CMDP) with a known safe baseline policy?", "answer": ["Learning Policies with Zero or Bounded Constraint Violation for Constrained MDPs", "Exploration-Exploitation in Constrained MDPs"], "answer_arxiv_id": ["2106.02684", "2003.02189"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_862"} +{"question": "Which papers discuss the replacement of the manual design of prompt templates with automatic search or learning in few-shot learning?", "answer": ["Prototypical Verbalizer for Prompt-based Few-shot Tuning", "WARP: Word-level Adversarial ReProgramming", "The Power of Scale for Parameter-Efficient Prompt Tuning", "GPT Understands, Too", "Differentiable Prompt Makes Pre-trained Language Models Better Few-shot Learners", "Factual Probing Is [MASK]: Learning vs. Learning to Recall"], "answer_arxiv_id": ["2203.09770", "2101.00121", "2104.08691", "2103.10385", "2108.13161", "2104.05240"], "source_meta": {"published_time": "20221106"}, "qid": "AutoScholarQuery_train_863"} +{"question": "Which work proves that random 2D CNNs could generate image correspondences sufficient to supervise a 3D network for the unsupervised registration problem?", "answer": ["Bootstrap Your Own Correspondences"], "answer_arxiv_id": ["2106.00677"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_864"} +{"question": "Are there any studies that critique model-generated text based on given reference text?", "answer": ["Self-critiquing models for assisting human evaluators", "Prometheus: Inducing Fine-grained Evaluation Capability in Language\n Models", "INSTRUCTSCORE: Explainable Text Generation Evaluation with Finegrained\n Feedback"], "answer_arxiv_id": ["2206.05802v2", "2310.08491", "2305.14282"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_865"} +{"question": "Which work transformed the Transformer architecture into MetaFormer by abstracting self-attention?", "answer": ["MetaFormer Is Actually What You Need for Vision"], "answer_arxiv_id": ["2111.11418"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_866"} +{"question": "Could you provide me the paper that introduced TVART?", "answer": ["Time-varying Autoregression with Low Rank Tensors"], "answer_arxiv_id": ["1905.08389"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_867"} +{"question": "Can you list some works that have explored the location of memorization in the model layers?", "answer": ["Deep Learning Through the Lens of Example Difficulty", "On the geometry of generalization and memorization in deep neural networks"], "answer_arxiv_id": ["2106.09647", "2105.14602"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_868"} +{"question": "Which papers introduced unsupervised model selection and discussed model selection for a new target domain?", "answer": ["Ranking Models in Unlabeled New Environments"], "answer_arxiv_id": ["2108.10310"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_869"} +{"question": "What papers describe graph-based generative models for molecules?", "answer": ["Constrained Graph Variational Autoencoders for Molecule Design", "Data-Efficient Graph Grammar Learning for Molecular Generation"], "answer_arxiv_id": ["1805.09076", "2203.08031"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_870"} +{"question": "What works explored image style transfer, which aims to render the semantic content under different styles?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion\n Models", "NoisyTwins: Class-Consistent and Diverse Image Generation through\n StyleGANs", "Pastiche Master: Exemplar-Based High-Resolution Portrait Style Transfer"], "answer_arxiv_id": ["1812.04948", "2308.07863", "2304.05866", "2203.13248"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_871"} +{"question": "Which studies focus on automatically generating CAD programs for purposes like reverse engineering and sketch-based modelling?", "answer": ["Fusion 360 Gallery: A Dataset and Environment for Programmatic CAD\n Construction from Human Design Sequences", "CAPRI-Net: Learning Compact CAD Shapes with Adaptive Primitive Assembly", "ShapeAssembly: Learning to Generate Programs for 3D Shape Structure\n Synthesis", "DeepCAD: A Deep Generative Network for Computer-Aided Design Models", "Hierarchical Neural Coding for Controllable CAD Model Generation"], "answer_arxiv_id": ["2010.02392", "2104.05652", "2009.08026", "2105.09492", "2307.00149"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_872"} +{"question": "Can you name papers that provided broader introductions to the subject of spurious correlations in deep learning?", "answer": ["Shortcut Learning in Deep Neural Networks", "Change is Hard: A Closer Look at Subpopulation Shift"], "answer_arxiv_id": ["2004.07780", "2302.12254"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_873"} +{"question": "Which works discussed modeling complex continuous-time phenomena by developing specialized neural architectures?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations", "Neural Flows: Efficient Alternative to Neural ODEs"], "answer_arxiv_id": ["2010.08895", "2110.13040"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_874"} +{"question": "Which papers are 2D generative models that have learned a wide range of visual concepts?", "answer": ["Zero-Shot Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_875"} +{"question": "What are the works that combined neural representations with inverse rendering?", "answer": ["Neural Reflectance Fields for Appearance Acquisition", "NeRV: Neural Reflectance and Visibility Fields for Relighting and View\n Synthesis", "PhySG: Inverse Rendering with Spherical Gaussians for Physics-based\n Material Editing and Relighting", "NeRD: Neural Reflectance Decomposition from Image Collections", "Neural-PIL: Neural Pre-Integrated Lighting for Reflectance Decomposition", "NeRFactor: Neural Factorization of Shape and Reflectance Under an\n Unknown Illumination", "Modeling Indirect Illumination for Inverse Rendering", "Shape, Light, and Material Decomposition from Images using Monte Carlo\n Rendering and Denoising", "IRON: Inverse Rendering by Optimizing Neural SDFs and Materials from\n Photometric Images", "TensoIR: Tensorial Inverse Rendering"], "answer_arxiv_id": ["2008.03824", "2012.03927", "2104.00674", "2012.03918", "2110.14373", "2106.01970", "2204.06837", "2206.03380", "2204.02232", "2304.12461"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_876"} +{"question": "Which papers discussed the powerful ability of LLMs on downstream generative and reasoning tasks?", "answer": ["Generative Speech Recognition Error Correction with Large Language\n Models and Task-Activating Prompting", "LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of\n Large Language Models", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention"], "answer_arxiv_id": ["2309.15649", "2304.01933", "2303.16199"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_877"} +{"question": "What works proposed verbalize-based and consistency-based methods to detect hallucinations in LLMs?", "answer": ["Can LLMs Express Their Uncertainty? An Empirical Evaluation of\n Confidence Elicitation in LLMs", "SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for\n Generative Large Language Models"], "answer_arxiv_id": ["2306.13063", "2303.08896"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_878"} +{"question": "Can you provide studies that use Structure from Motion (SfM) for large-scale annotation?", "answer": ["LabelFusion: A Pipeline for Generating Ground Truth Labels for Real RGBD\n Data of Cluttered Scenes"], "answer_arxiv_id": ["1707.04796"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_879"} +{"question": "Which studies reframed the problem of text-driven motion generation as a next-index prediction task?", "answer": ["T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations"], "answer_arxiv_id": ["2301.06052"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_880"} +{"question": "What studies provide examples of gradient-free methods used in quantum optimization, such as SPSA, COBYLA, and Bayesian optimization?", "answer": ["Variational quantum algorithm with information sharing", "Stochastic Gradient Line Bayesian Optimization for Efficient Noise-Robust Optimization of Parameterized Quantum Circuits"], "answer_arxiv_id": ["2103.16161", "2111.07952"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_881"} +{"question": "Could you provide me some sources about the use of sharpness for predicting generalization in deep learning?", "answer": ["Fantastic Generalization Measures and Where to Find Them", "Sharpness-Aware Minimization for Efficiently Improving Generalization"], "answer_arxiv_id": ["1912.02178", "2010.01412"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_882"} +{"question": "Has research been conducted on the time difference method in relation to CVaR?", "answer": ["Risk-Sensitive and Robust Decision-Making: a CVaR Optimization Approach"], "answer_arxiv_id": ["1506.02188"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_883"} +{"question": "Can you list some datasets for visual document understanding, including layout analysis and document classification?", "answer": ["FUNSD: A Dataset for Form Understanding in Noisy Scanned Documents", "Evaluation of Deep Convolutional Nets for Document Image Classification and Retrieval", "PubLayNet: largest dataset ever for document layout analysis", "DocBank: A Benchmark Dataset for Document Layout Analysis", "DocParser: Hierarchical Document Structure Parsing from Renderings"], "answer_arxiv_id": ["1905.13538", "1502.07058", "1908.07836", "2006.01038", "1911.01702"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_884"} +{"question": "Which works propose to edit the contents of dynamic NeRFs?", "answer": ["Dyn-E: Local Appearance Editing of Dynamic Neural Radiance Fields"], "answer_arxiv_id": ["2307.12909"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_885"} +{"question": "Which papers focus on learning methods that utilize spatial coordinates to predict RIRs under room acoustics?", "answer": ["Deep Impulse Responses: Estimating and Parameterizing Filters with Deep\n Networks", "IR-GAN: Room Impulse Response Generator for Far-field Speech Recognition"], "answer_arxiv_id": ["2202.03416", "2010.13219"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_886"} +{"question": "Can you provide me references about deriving predictive risks in neural networks?", "answer": ["Generalisation error in learning with random features and the hidden manifold model"], "answer_arxiv_id": ["2002.09339"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_887"} +{"question": "Which works introduced deep convolutional networks for efficient dense predictions in the field of low-latency segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_888"} +{"question": "What research has been done on test-time training?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "Tent: Fully Test-Time Adaptation by Entropy Minimization", "MEMO: Test Time Robustness via Adaptation and Augmentation", "Self-Supervised Aggregation of Diverse Experts for Test-Agnostic Long-Tailed Recognition"], "answer_arxiv_id": ["1909.13231", "2006.10726", "2110.09506", "2107.09249"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_889"} +{"question": "Which papers have explored the use of sketching in attention scheme inspired regression?", "answer": ["Solving Regularized Exp, Cosh and Sinh Regression Problems", "Attention Scheme Inspired Softmax Regression", "The Closeness of In-Context Learning and Weight Shifting for Softmax Regression", "An Iterative Algorithm for Rescaled Hyperbolic Functions Regression", "An Over-parameterized Exponential Regression"], "answer_arxiv_id": ["2303.15725", "2304.10411", "2304.13276", "2305.00660", "2303.16504"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_890"} +{"question": "Can you point out some references about ensemble learning techniques?", "answer": ["Distilling the Knowledge in a Neural Network", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift", "Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning", "Packed-Ensembles for efficient uncertainty estimation"], "answer_arxiv_id": ["1503.02531", "1612.01474", "1906.02530", "2012.09816", "2210.09184"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_891"} +{"question": "Are there any works that selected subsets for supervised learning based on gradient norm?", "answer": ["Not All Samples Are Created Equal: Deep Learning with Importance Sampling", "Coresets for Data-efficient Training of Machine Learning Models", "Adaptive Second Order Coresets for Data-efficient Machine Learning", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training"], "answer_arxiv_id": ["1803.00942v3", "1906.01827", "2207.13887", "2103.00123"], "source_meta": {"published_time": "20230218"}, "qid": "AutoScholarQuery_train_892"} +{"question": "Which studies demonstrate the usage of side-channel communication to exploit long-tailed distribution in order to increase jailbreak success rates?", "answer": ["GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via Cipher", "Multilingual Jailbreak Challenges in Large Language Models", "Low-Resource Languages Jailbreak GPT-4", "Use of LLMs for Illicit Purposes: Threats, Prevention Measures, and\n Vulnerabilities", "Exploiting Programmatic Behavior of LLMs: Dual-Use Through Standard\n Security Attacks"], "answer_arxiv_id": ["2308.06463", "2310.06474", "2310.02446", "2308.12833", "2302.05733"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_893"} +{"question": "Can you mention studies that proposed backdoor mitigation methods in the context of backdoor attacks?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural\n Networks", "Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks\n Without an Accuracy Tradeoff", "DeepSweep: An Evaluation Framework for Mitigating DNN Backdoor Attacks\n using Data Augmentation", "FLIP: A Provable Defense Framework for Backdoor Mitigation in Federated\n Learning", "Neural Attention Distillation: Erasing Backdoor Triggers from Deep\n Neural Networks", "Anti-Backdoor Learning: Training Clean Models on Poisoned Data", "Training with More Confidence: Mitigating Injected and Natural Backdoors\n During Training", "Constrained Optimization with Dynamic Bound-scaling for Effective\n NLPBackdoor Defense"], "answer_arxiv_id": ["1805.12185", "2011.09527", "2012.07006", "2210.12873", "2101.05930", "2110.11571", "2202.06382", "2202.05749"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_894"} +{"question": "Could you provide me a study examines how common cost indicators of machine learning models can contradict each other?", "answer": ["The Efficiency Misnomer"], "answer_arxiv_id": ["2110.12894"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_895"} +{"question": "Which work seeks to identify common information in both views without an objective to retain the unique information?", "answer": ["Learning Robust Representations via Multi-View Information Bottleneck"], "answer_arxiv_id": ["2002.07017"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_896"} +{"question": "What work provides theoretical insights into the feature learning process of deep networks?", "answer": ["Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning"], "answer_arxiv_id": ["2012.09816"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_897"} +{"question": "Could you provide me studies about models using UniVL-DR, Clip4Cir, and UniIR, which encode image and text separately using encoders from CLIP or BLIP?", "answer": ["Universal Vision-Language Dense Retrieval: Learning A Unified\n Representation Space for Multi-Modal Retrieval", "Composed Image Retrieval using Contrastive Learning and Task-oriented\n CLIP-based Features", "UniIR: Training and Benchmarking Universal Multimodal Information\n Retrievers"], "answer_arxiv_id": ["2209.00179", "2308.11485", "2311.17136"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_898"} +{"question": "Which research focuses on active learning from human teachers?", "answer": ["Learning Perceptual Concepts by Bootstrapping from Human Queries"], "answer_arxiv_id": ["2111.05251v3"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_899"} +{"question": "Which works have applied manually crafted rules to reduce the token scores to document scores in the retrieval-augmented language models?", "answer": ["Relevance-guided Supervision for OpenQA with ColBERT", "Retrieval as Attention: End-to-end Learning of Retrieval and Reading within a Single Transformer", "ColBERTv2: Effective and Efficient Retrieval via Lightweight Late Interaction"], "answer_arxiv_id": ["2007.00814", "2212.02027", "2112.01488"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_900"} +{"question": "Which works have utilized the Segment Anything Model in the field of medical imaging?", "answer": ["Segment Anything Model for Medical Image Analysis: an Experimental Study", "Segment Anything in Medical Images", "Medical SAM Adapter: Adapting Segment Anything Model for Medical Image\n Segmentation"], "answer_arxiv_id": ["2304.10517", "2304.12306", "2304.12620"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_901"} +{"question": "Could you provide me some studies that design networks for speeding up segmentation algorithms?", "answer": ["ICNet for Real-Time Semantic Segmentation on High-Resolution Images"], "answer_arxiv_id": ["1704.08545"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_902"} +{"question": "Do you know about any improvements over the 'metadata normalization' (MDN)?", "answer": ["A Penalty Approach for Normalizing Feature Distributions to Build Confounder-Free Models"], "answer_arxiv_id": ["2207.04607"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_903"} +{"question": "What study is based on contrastive representation learning?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning"], "answer_arxiv_id": ["2004.04136"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_904"} +{"question": "Which research introduced DAIR-V2X dataset family for V2X datasets?", "answer": ["DAIR-V2X: A Large-Scale Dataset for Vehicle-Infrastructure Cooperative\n 3D Object Detection"], "answer_arxiv_id": ["2204.05575"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_905"} +{"question": "Which studies have expanded the textual vocabulary with quantized image coordinates for image tasks using autoregressive modeling?", "answer": ["Pix2seq: A Language Modeling Framework for Object Detection", "A Unified Sequence Interface for Vision Tasks"], "answer_arxiv_id": ["2109.10852", "2206.07669"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_906"} +{"question": "What paper mentioned using all four techniques to decrease both the latency and memory consumption of multi-head attention?", "answer": ["FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness"], "answer_arxiv_id": ["2205.14135"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_907"} +{"question": "Do you have any studies that focused on improving the multilingual reasoning abilities of large language models?", "answer": ["Breaking Language Barriers in Multilingual Mathematical Reasoning:\n Insights and Observations"], "answer_arxiv_id": ["2310.20246"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_908"} +{"question": "Can you name the studies that hand pick dimension jump points to increase the resolution of images?", "answer": ["Subspace Diffusion Generative Models", "Dimensionality-Varying Diffusion Process"], "answer_arxiv_id": ["2205.01490", "2211.16032"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_909"} +{"question": "Could you provide me studies about spatially partitioning a large scene into several individually trained sub-NeRFs?", "answer": ["Block-NeRF: Scalable Large Scene Neural View Synthesis", "Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual Fly-Throughs"], "answer_arxiv_id": ["2202.05263", "2112.10703"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_910"} +{"question": "Are there any works that developed algorithms based on stochastic proximal methods, demonstrating their improved robustness?", "answer": ["The importance of better models in stochastic optimization"], "answer_arxiv_id": ["1903.08619v1"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_911"} +{"question": "Which works discussed the implementation of alpha matte refinement under the encoder-decoder framework during the era of deep learning?", "answer": ["Deep Image Matting"], "answer_arxiv_id": ["1703.03872v3"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_912"} +{"question": "What research has been done concerning learning UDF from 3D point clouds for non-watertight models?", "answer": ["Neural Unsigned Distance Fields for Implicit Function Learning", "Learning Anchored Unsigned Distance Functions with Gradient Direction\n Alignment for Single-view Garment Reconstruction", "CAP-UDF: Learning Unsigned Distance Functions Progressively from Raw\n Point Clouds with Consistency-Aware Field Optimization"], "answer_arxiv_id": ["2010.13938", "2108.08478", "2210.02757"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_913"} +{"question": "Which researches explore the concept of AdaTS with the use of a separate VAE and MLP for producing an adaptive temperature?", "answer": ["Sample-dependent Adaptive Temperature Scaling for Improved Calibration"], "answer_arxiv_id": ["2207.06211"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_914"} +{"question": "What research works adopted the BEV space to align multi-frame features in Transformer-based approaches for 3D object detection?", "answer": ["BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers", "BEVDet: High-performance Multi-camera 3D Object Detection in\n Bird-Eye-View"], "answer_arxiv_id": ["2203.17270", "2112.11790"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_915"} +{"question": "Which studies discuss the utility of language modeling approaches in audio for generating audio given an input description?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation", "MusicLM: Generating Music From Text"], "answer_arxiv_id": ["2209.03143", "2301.11325"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_916"} +{"question": "Which studies proposed to optimize the number of prototypes?", "answer": ["ProtoPShare: Prototype Sharing for Interpretable Image Classification\n and Similarity Discovery", "Interpretable Image Classification with Differentiable Prototypes\n Assignment"], "answer_arxiv_id": ["2011.14340", "2112.02902"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_917"} +{"question": "Are there any studies regarding the interaction of entropic optimal transport with deep generative modelling?", "answer": ["Learning normalizing flows from Entropy-Kantorovich potentials", "Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling"], "answer_arxiv_id": ["2006.06033", "2106.01357"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_918"} +{"question": "In what study do the authors perform a hyperparameter search separately on each client?", "answer": ["FLoRA: Single-shot Hyper-parameter Optimization for Federated Learning"], "answer_arxiv_id": ["2112.08524"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_919"} +{"question": "What studies have enabled joint estimation of radiance fields and camera poses via gradient descent?", "answer": ["NeRF-⁣-: Neural Radiance Fields Without Known Camera Parameters", "iNeRF: Inverting Neural Radiance Fields for Pose Estimation", "BARF : Bundle-Adjusting Neural Radiance Fields"], "answer_arxiv_id": ["2102.07064", "2012.05877", "2104.06405"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_920"} +{"question": "Which studies work on end-to-end manipulation approaches that learn a direct mapping from RGB images to a robot action?", "answer": ["Transferring End-to-End Visuomotor Control from Simulation to Real World\n for a Multi-Stage Task", "Sim-to-Real Reinforcement Learning for Deformable Object Manipulation", "QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic\n Manipulation", "Learning to Manipulate Deformable Objects without Demonstrations"], "answer_arxiv_id": ["1707.02267", "1806.07851", "1806.10293", "1910.13439"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_921"} +{"question": "Which studies have focused on recovering high-fidelity lighting environments for augmented reality applications?", "answer": ["Learning to Predict Indoor Illumination from a Single Image", "Neural Illumination: Lighting Prediction for Indoor Environments", "Deep Outdoor Illumination Estimation", "DeepLight: Learning Illumination for Unconstrained Mobile Mixed Reality", "Learning Illumination from Diverse Portraits"], "answer_arxiv_id": ["1704.00090", "1906.07370", "1611.06403", "1904.01175", "2008.02396"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_922"} +{"question": "Could you provide me some studies about structure-wise data augmentation in graph representation learning?", "answer": ["Graph Contrastive Learning with Augmentations", "Data Augmentation for Graph Neural Networks"], "answer_arxiv_id": ["2010.13902", "2006.06830"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_923"} +{"question": "Any works about overall data augmentation methods used in user modeling and sequential recommendation?", "answer": ["Contrastive Learning for Sequential Recommendation"], "answer_arxiv_id": ["2010.14395"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_924"} +{"question": "What work employs neural planes to parameterize a 3D scene?", "answer": ["K-Planes: Explicit Radiance Fields in Space, Time, and Appearance"], "answer_arxiv_id": ["2301.10241"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_925"} +{"question": "Which paper used the built Visual Context Tree (VCTree) graph structure for visual question answering?", "answer": ["Learning to Compose Dynamic Tree Structures for Visual Contexts"], "answer_arxiv_id": ["1812.01880"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_926"} +{"question": "Which works proposed the use of the Fisher Information Matrix derived from a 'probe network' for dataset representation?", "answer": ["Task2Vec: Task Embedding for Meta-Learning", "Universal Statistics of Fisher Information in Deep Neural Networks: Mean\n Field Approach"], "answer_arxiv_id": ["1902.03545", "1806.01316"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_927"} +{"question": "What works tackle the INT4 weight quantization for transformers?", "answer": ["The case for 4-bit precision: k-bit Inference Scaling Laws", "GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers"], "answer_arxiv_id": ["2212.09720", "2210.17323"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_928"} +{"question": "What are some studies related to video question answering?", "answer": ["TVQA: Localized, Compositional Video Question Answering", "STAR: A Benchmark for Situated Reasoning in Real-World Videos", "NExT-QA:Next Phase of Question-Answering to Explaining Temporal Actions", "From Representation to Reasoning: Towards both Evidence and Commonsense\n Reasoning for Video Question-Answering", "AGQA: A Benchmark for Compositional Spatio-Temporal Reasoning", "Self-Chained Image-Language Model for Video Localization and Question\n Answering", "ComPhy: Compositional Physical Reasoning of Objects and Events from\n Videos"], "answer_arxiv_id": ["1809.01696", "2405.09711v1", "2105.08276", "2205.14895", "2103.16002", "2305.06988", "2205.01089"], "source_meta": {"published_time": "20240515"}, "qid": "AutoScholarQuery_train_929"} +{"question": "What studies have attempted to generalize classical ideas of active learning to different losses?", "answer": ["Importance Weighted Active Learning"], "answer_arxiv_id": ["0812.4952"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_930"} +{"question": "What are some of the research papers that focus on 2D super-resolution configurations in turbulent flows?", "answer": ["Super-resolution reconstruction of turbulent flows with machine learning"], "answer_arxiv_id": ["1811.11328"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_931"} +{"question": "Can you provide me some works that proposed constrained formulations for Domain Generalization(DG)?", "answer": ["Model-Based Domain Generalization", "Towards Principled Disentanglement for Domain Generalization"], "answer_arxiv_id": ["2102.11436", "2111.13839"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_932"} +{"question": "Which works utilized the large-scale pretrained vision-language model for OOD detection?", "answer": ["Delving into Out-of-Distribution Detection with Vision-Language Representations", "Zero-Shot Out-of-Distribution Detection Based on the Pre-trained Model CLIP"], "answer_arxiv_id": ["2211.13445", "2109.02748"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_933"} +{"question": "What studies have focused on learning Nash equilibrium in a sample-efficient manner using value-based methods?", "answer": ["Provable Self-Play Algorithms for Competitive Reinforcement Learning", "Near-Optimal Reinforcement Learning with Self-Play", "V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL", "Decentralized Q-Learning in Zero-sum Markov Games", "Learning Zero-Sum Simultaneous-Move Markov Games Using Function Approximation and Correlated Equilibrium"], "answer_arxiv_id": ["2002.04017", "2006.12007", "2110.14555", "2106.02748", "2002.07066"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_934"} +{"question": "Which works include a method for detecting self-contradictions or factual errors in responses to enhance quality?", "answer": ["LM vs LM: Detecting Factual Errors via Cross Examination", "Self-contradictory Hallucinations of Large Language Models: Evaluation,\n Detection and Mitigation"], "answer_arxiv_id": ["2305.13281", "2305.15852"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_935"} +{"question": "Could you provide me some works that focus on selecting valuable features and propose various loss functions for knowledge distillation in object detection?", "answer": ["Distilling Object Detectors via Decoupled Features", "Channel-wise Knowledge Distillation for Dense Prediction", "Focal and Global Knowledge Distillation for Detectors", "Masked Distillation with Receptive Tokens"], "answer_arxiv_id": ["2103.14475", "2011.13256v4", "2111.11837", "2205.14589"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_936"} +{"question": "Which works showcase the transition from white-box models to black-box APIs in the field of natural language processing?", "answer": ["Language Models are Few-Shot Learners", "GPT-4 Technical Report"], "answer_arxiv_id": ["2005.14165", "2303.08774"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_937"} +{"question": "Could you provide some references that introduced transformer-based models for object detection?", "answer": ["End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2005.12872", "2010.04159"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_938"} +{"question": "What are some of the examples of tasks where molecular LLMs have been used, as seen in scientific research?", "answer": ["MolXPT: Wrapping Molecules with Text for Generative Pre-training", "Mol-Instructions: A Large-Scale Biomolecular Instruction Dataset for Large Language Models", "MolCA: Molecular Graph-Language Modeling with Cross-Modal Projector and\n Uni-Modal Adapter"], "answer_arxiv_id": ["2305.10688", "2306.08018v5", "2310.12798"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_939"} +{"question": "Which works discovered reductions from solving common-payoff games to solving belief MDPs?", "answer": ["Decentralized Stochastic Control with Partial History Sharing: A Common Information Approach"], "answer_arxiv_id": ["1209.1695"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_940"} +{"question": "Can you give me examples of research that trains reinforcement learning algorithms directly in the conservative latent space?", "answer": ["PLAS: Latent Action Space for Offline Reinforcement Learning", "Latent-Variable Advantage-Weighted Policy Optimization for Offline RL", "Let Offline RL Flow: Training Conservative Agents in the Latent Space of Normalizing Flows"], "answer_arxiv_id": ["2011.07213", "2203.08949v1", "2211.11096"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_941"} +{"question": "Which works proposed the use of resampling strategies to improve a model’s robustness to dataset bias?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness", "Automatic Shortcut Removal for Self-Supervised Representation Learning", "Latent Adversarial Debiasing: Mitigating Collider Bias in Deep Neural Networks", "Learning Debiased and Disentangled Representations for Semantic Segmentation"], "answer_arxiv_id": ["1811.12231", "2002.08822", "2011.11486", "2111.00531"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_942"} +{"question": "Could you provide me some works about knowledge distillation?", "answer": ["Self-Distillation Amplifies Regularization in Hilbert Space", "Knowledge Distillation as Semiparametric Inference", "Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["2002.05715v3", "2104.09732", "1503.02531"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_943"} +{"question": "What work uses online learning algorithm for the ranking problem from user interactions rather than logged data?", "answer": ["Reinforcement Online Learning to Rank with Unbiased Reward Shaping"], "answer_arxiv_id": ["2201.01534"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_944"} +{"question": "What work proposed using neural network architecture to model the joint distribution over observed variables?", "answer": ["Learning Functional Causal Models with Generative Neural Networks"], "answer_arxiv_id": ["1709.05321v3"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_945"} +{"question": "Which papers are used in the CapEval1k dataset to provide captions?", "answer": ["Self-critical Sequence Training for Image Captioning", "Bottom-Up and Top-Down Attention for Image Captioning and Visual\n Question Answering", "Attention on Attention for Image Captioning"], "answer_arxiv_id": ["1612.00563", "1707.07998", "1908.06954"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_946"} +{"question": "Could you provide me some studies on manually-crafted hallucination datasets that entail manual identification?", "answer": ["FreshLLMs: Refreshing Large Language Models with Search Engine\n Augmentation", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2310.03214", "2005.11401"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_947"} +{"question": "Which research paper evaluates the adversarial robustness of classic attribute-based zero-shot models?", "answer": ["How Robust are Discriminatively Trained Zero-Shot Learning Models?"], "answer_arxiv_id": ["2201.10972"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_948"} +{"question": "What research assumes that the player with the maximal expected reward would get the reward when a collision occurs in MPMAB model?", "answer": ["Competing Bandits in Matching Markets", "Bandit Learning in Decentralized Matching Markets", "Learning Equilibria in Matching Markets from Bandit Feedback"], "answer_arxiv_id": ["1906.05363", "2012.07348", "2108.08843"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_949"} +{"question": "What works have proposed generating discrete variables in tabular data by Bayesian networks and decision trees?", "answer": ["Generating Synthetic but Plausible Healthcare Record Datasets"], "answer_arxiv_id": ["1807.01514"], "source_meta": {"published_time": "20221008"}, "qid": "AutoScholarQuery_train_950"} +{"question": "Which works proposed FedAvg, an effective architecture for Federated Learning?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized\n Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_951"} +{"question": "Which studies developed algorithms for certification of deterministic neural networks (NNs)?", "answer": ["Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks", "Towards Fast Computation of Certified Robustness for ReLU Networks", "Provable Defenses against Adversarial Examples via the Convex Outer Adversarial Polytope", "Lagrangian Decomposition for Neural Network Verification"], "answer_arxiv_id": ["1702.01135", "1804.09699", "1711.00851", "2002.10410"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_952"} +{"question": "Which studies scaled the text encoder and made other architectural changes in T2I models?", "answer": ["SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis"], "answer_arxiv_id": ["2307.01952"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_953"} +{"question": "What previous works utilized rationales on social reasoning?", "answer": ["What Makes it Ok to Set a Fire? Iterative Self-distillation of Contexts\n and Rationales for Disambiguating Defeasible Social and Moral Situations", "Social Bias Frames: Reasoning about Social and Power Implications of\n Language"], "answer_arxiv_id": ["2310.15431", "1911.03891"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_train_954"} +{"question": "What is the reference that first applied vector quantized-VAE to learn the mutual mapping between human motions and discrete tokens?", "answer": ["TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of\n 3D Human Motions and Texts"], "answer_arxiv_id": ["2207.01696"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_955"} +{"question": "What research has been conducted on the impact of different ways to obtain and use pseudolabels on the final predictions of the model?", "answer": ["Dash: Semi-Supervised Learning with Dynamic Thresholding", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "Label Propagation for Deep Semi-supervised Learning"], "answer_arxiv_id": ["2109.00650", "2110.08263", "1904.04717"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_956"} +{"question": "What are some studies that construct edges through the spatial position of cells for their cell graph representations in tissue?", "answer": ["CGC-Net: Cell Graph Convolutional Network for Grading of Colorectal\n Cancer Histology Images", "Pathomic Fusion: An Integrated Framework for Fusing Histopathology and\n Genomic Features for Cancer Diagnosis and Prognosis"], "answer_arxiv_id": ["1909.01068", "1912.08937"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_957"} +{"question": "What works provided polynomial complexity results for POMDPs where transitions are deterministic but stochastic emissions", "answer": ["Sample-Efficient Reinforcement Learning of Undercomplete POMDPs"], "answer_arxiv_id": ["2006.12484"], "source_meta": {"published_time": "20220624"}, "qid": "AutoScholarQuery_train_958"} +{"question": "Which work proposed the method of learning a prompt on source tasks and then using it for the target task in the context of prompt tuning?", "answer": ["SPoT: Better Frozen Model Adaptation through Soft Prompt Transfer"], "answer_arxiv_id": ["2110.07904"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_959"} +{"question": "Could you provide me some works that rely on adding exploration bonuses for computational efficiency?", "answer": ["Minimax Regret Bounds for Reinforcement Learning", "PC-PG: Policy Cover Directed Exploration for Provable Policy Gradient Learning", "Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition"], "answer_arxiv_id": ["1703.05449", "2007.08459", "2004.10019"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_960"} +{"question": "Can you point me to the work that reported better results from cross-modal contrastive learning than within-modal training?", "answer": ["Audio-Visual Instance Discrimination with Cross-Modal Agreement"], "answer_arxiv_id": ["2004.12943"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_961"} +{"question": "Which papers proposed the methods of employing SAM on NeRF-rendered images and using neural field to render SAM feature maps?", "answer": ["Segment Anything in 3D with Radiance Fields", "Interactive Segment Anything NeRF with Feature Imitation"], "answer_arxiv_id": ["2304.12308", "2305.16233"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_962"} +{"question": "Which studies involve the Fourier neural operator?", "answer": ["Seismic wave propagation and inversion with Neural Operators"], "answer_arxiv_id": ["2108.05421"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_963"} +{"question": "Could you provide me some studies that explored better token routing methods for Mixture-of-Experts (MoE) models?", "answer": ["Hash Layers For Large Sparse Models", "BASE Layers: Simplifying Training of Large, Sparse Models", "Mixture-of-Experts with Expert Choice Routing"], "answer_arxiv_id": ["2106.04426", "2103.16716", "2202.09368"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_964"} +{"question": "Could you provide me some studies that have focused on solving classic graph mining tasks in a hypergraph setting?", "answer": ["Finding Bipartite Components in Hypergraphs"], "answer_arxiv_id": ["2205.02771v1"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_965"} +{"question": "Could you name the papers that proposed personalized models in federated learning?", "answer": ["Federated Learning with Personalization Layers", "Exploiting Shared Representations for Personalized Federated Learning", "Think Locally, Act Globally: Federated Learning with Local and Global Representations", "CD2-pFed: Cyclic Distillation-guided Channel Decoupling for Model Personalization in Federated Learning", "Personalized Federated Learning using Hypernetworks", "FedTP: Federated Learning by Transformer Personalization", "Layer-wised Model Aggregation for Personalized Federated Learning", "On Bridging Generic and Personalized Federated Learning for Image Classification", "FedProc: Prototypical Contrastive Federated Learning on Non-IID data"], "answer_arxiv_id": ["1912.00818v1", "2102.07078", "2001.01523", "2204.03880", "2103.04628", "2211.01572", "2205.03993", "2107.00778", "2109.12273"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_966"} +{"question": "Which papers have contributed to the knowledge of mixtures of time series and trajectories in various domains?", "answer": ["Toeplitz Inverse Covariance-Based Clustering of Multivariate Time Series Data"], "answer_arxiv_id": ["1706.03161"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_967"} +{"question": "Could you provide me some work that aim to address the Janus problems in text-to-3D generation?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation", "Re-imagine the Negative Prompt Algorithm: Transform 2D Diffusion into\n 3D, alleviate Janus problem and Beyond", "Text-to-3D using Gaussian Splatting"], "answer_arxiv_id": ["2308.16512", "2304.04968", "2309.16585"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_968"} +{"question": "What works leveraged large-scale image-caption pairs in LVLMs to transform the image features into the embedding space of LLMs for modality alignment?", "answer": ["Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Knowledge-augmented Few-shot Visual Relation Detection"], "answer_arxiv_id": ["2304.08485", "2304.10592", "2204.14198", "2301.12597", "2303.05342"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_969"} +{"question": "Which studies suggested the importance of large kernels for semantic segmentation?", "answer": ["Rethinking the Inception Architecture for Computer Vision"], "answer_arxiv_id": ["1512.00567"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_970"} +{"question": "What recent approaches have demonstrated the ability to predict optical flow directly using deep neural networks?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks", "PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume", "LiteFlowNet: A Lightweight Convolutional Neural Network for Optical Flow\n Estimation", "Accurate Optical Flow via Direct Cost Volume Processing"], "answer_arxiv_id": ["1504.06852", "1612.01925", "1709.02371", "1805.07036", "1704.07325"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_971"} +{"question": "Which papers establish the optimality of natural actor critic in both on-policy and off-policy settings?", "answer": ["Neural Policy Gradient Methods: Global Optimality and Rates of Convergence", "An Improved Analysis of (Variance-Reduced) Policy Gradient and Natural Policy Gradient Methods", "Finite Sample Analysis of Two-Time-Scale Natural Actor-Critic Algorithm", "Finite-Sample Analysis of Off-Policy Natural Actor-Critic Algorithm", "Finite-Sample Analysis of Off-Policy Natural Actor-Critic with Linear Function Approximation"], "answer_arxiv_id": ["1909.01150", "2211.07937", "2101.10506", "2102.09318", "2105.12540"], "source_meta": {"published_time": "20211104"}, "qid": "AutoScholarQuery_train_972"} +{"question": "What is the reference that discusses non-isotropic Gaussian diffusion processes in the context of lossy compression?", "answer": ["Lossy Compression with Gaussian Diffusion"], "answer_arxiv_id": ["2206.08889"], "source_meta": {"published_time": "20220912"}, "qid": "AutoScholarQuery_train_973"} +{"question": "What works illustrate that the gradient domination structure can be globally and efficiently optimized by (stochastic) optimization algorithms?", "answer": ["Linear Convergence of Gradient and Proximal-Gradient Methods Under the Polyak-Łojasiewicz Condition", "A Two-Time-Scale Stochastic Optimization Framework with Applications in Control and Reinforcement Learning", "SGD for Structured Nonconvex Functions: Learning Rates, Minibatching and Interpolation"], "answer_arxiv_id": ["1608.04636", "2109.14756", "2006.10311v3"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_974"} +{"question": "What are the studies in which fine-grained labels are used to pseudo labels image-text data?", "answer": ["Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2112.03857"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_975"} +{"question": "Can you provide examples of diffusion models used in text-to-image generation?", "answer": ["Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Diffusion Models Beat GANs on Image Synthesis", "Adding Conditional Control to Text-to-Image Diffusion Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2006.11239", "2112.10752", "2105.05233", "2302.05543", "2208.12242"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_976"} +{"question": "Can you mention some papers utilize contrastive learning in image processing?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["2006.07733", "2002.05709", "1807.03748"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_977"} +{"question": "Which study evaluated the impact of Multi-Task Learning (MTL) on the robustness and performances of depth tasks?", "answer": ["Improved Noise and Attack Robustness for Semantic Segmentation by Using Multi-Task Training with Self-Supervised Depth Estimation"], "answer_arxiv_id": ["2004.11072"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_978"} +{"question": "What studies could provide insights into Physics-Informed Neural Networks?", "answer": ["Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations"], "answer_arxiv_id": ["1801.06637"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_979"} +{"question": "What papers have proposed deep-learning-based NVS methods that only use a single image during test time?", "answer": ["Deep3D: Fully Automatic 2D-to-3D Video Conversion with Deep\n Convolutional Neural Networks", "Single-View View Synthesis with Multiplane Images", "pixelNeRF: Neural Radiance Fields from One or Few Images", "MINE: Towards Continuous Depth MPI with NeRF for Novel View Synthesis", "Behind the Scenes: Density Fields for Single View Reconstruction", "Consistent View Synthesis with Pose-Guided Diffusion Models", "SceneRF: Self-Supervised Monocular 3D Scene Reconstruction with Radiance\n Fields", "VQ3D: Learning a 3D-Aware Generative Model on ImageNet"], "answer_arxiv_id": ["1604.03650", "2004.11364", "2012.02190", "2103.14910", "2301.07668", "2303.17598", "2212.02501", "2302.06833"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_980"} +{"question": "What studies provide an end-to-end analysis of diffusion models?", "answer": ["Score Approximation, Estimation and Distribution Recovery of Diffusion Models on Low-Dimensional Data", "Diffusion Models are Minimax Optimal Distribution Estimators"], "answer_arxiv_id": ["2302.07194", "2303.01861"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_981"} +{"question": "Could you provide me some papers that implement post-training adjustments to better manage the increased relative positional distances in longer sequences?", "answer": ["PoSE: Efficient Context Window Extension of LLMs via Positional\n Skip-wise Training", "YaRN: Efficient Context Window Extension of Large Language Models", "Extending Context Window of Large Language Models via Positional\n Interpolation"], "answer_arxiv_id": ["2309.10400", "2309.00071", "2306.15595"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_982"} +{"question": "Which works directly establish correspondence between a pair of point clouds in non-rigid point cloud matching?", "answer": ["DPC: Unsupervised Deep Point Correspondence via Cross and Self Construction", "Lepard: Learning partial point cloud matching in rigid and deformable scenes", "CorrNet3D: Unsupervised End-to-end Learning of Dense Correspondence for 3D Point Clouds"], "answer_arxiv_id": ["2110.08636", "2111.12591", "2012.15638"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_983"} +{"question": "Which studies are about creating similar samples to the existing ones to enhance model generalization?", "answer": ["MixText: Linguistically-Informed Interpolation of Hidden Space for Semi-Supervised Text Classification", "Large Language Models Can Self-Improve", "Neural Data Augmentation via Example Extrapolation", "Distantly-Supervised Named Entity Recognition with Noise-Robust Learning and Language Model Augmented Self-Training", "Adversarial Training Methods for Semi-Supervised Text Classification", "Unsupervised Data Augmentation for Consistency Training"], "answer_arxiv_id": ["2004.12239", "2210.11610", "2102.01335", "2109.05003", "1605.07725", "1904.12848"], "source_meta": {"published_time": "20221106"}, "qid": "AutoScholarQuery_train_984"} +{"question": "What research has been conducted on state space models?", "answer": ["Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layers", "HiPPO: Recurrent Memory with Optimal Polynomial Projections"], "answer_arxiv_id": ["2110.13985", "2008.07669"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_985"} +{"question": "Are there any works which inspired normalizing weights by their spectral norm?", "answer": ["Spectral Normalization for Generative Adversarial Networks"], "answer_arxiv_id": ["1802.05957"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_986"} +{"question": "What studies discussed the merits of adopting appropriate modules for different input data or target tasks in the context of model capacity and learning difficulty?", "answer": ["Modular Networks: Learning to Decompose Neural Computation", "Combining Modular Skills in Multitask Learning"], "answer_arxiv_id": ["1811.05249", "2202.13914"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_987"} +{"question": "What studies focus on gender biases associated with pronoun and coreferences?", "answer": ["Toward Gender-Inclusive Coreference Resolution", "Mind the GAP: A Balanced Corpus of Gendered Ambiguous Pronouns", "Gender Bias in Coreference Resolution: Evaluation and Debiasing Methods", "Gender Bias in Coreference Resolution"], "answer_arxiv_id": ["1910.13913", "1810.05201", "1804.06876", "1804.09301"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_988"} +{"question": "Can you list some studies showing that transformers can implement optimization algorithms across their layers?", "answer": ["What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "What learning algorithm is in-context learning? Investigations with linear models", "Transformers Learn In-Context by Gradient Descent"], "answer_arxiv_id": ["2208.01066", "2211.15661", "2212.07677"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_989"} +{"question": "Could you provide me some studies about facilitating semantic segmentation using the language-driven modeling paradigm?", "answer": ["Language-driven Semantic Segmentation", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "Decoupling Zero-Shot Semantic Segmentation"], "answer_arxiv_id": ["2201.03546", "2112.01518", "2112.07910"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_990"} +{"question": "What work explores the use of diffusion models for single-step decision-making in offline RL?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_991"} +{"question": "Where is the combined channel and spatial attention employed?", "answer": ["CBAM: Convolutional Block Attention Module"], "answer_arxiv_id": ["1807.06521"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_992"} +{"question": "Could you point out the studies that examine the algorithmic power of transformers?", "answer": ["Statistically Meaningful Approximation: a Case Study on Approximating Turing Machines with Transformers", "Looped Transformers as Programmable Computers", "What learning algorithm is in-context learning? Investigations with linear models", "In-context Learning and Induction Heads"], "answer_arxiv_id": ["2107.13163", "2301.13196v1", "2211.15661", "2209.11895v1"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_993"} +{"question": "What works have applied diffusion models successfully to various control tasks?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis", "SE(3)-DiffusionFields: Learning smooth cost functions for joint grasp and motion optimization through diffusion", "Is Conditional Generative Modeling all you need for Decision-Making?", "Diffusion Policy: Visuomotor Policy Learning via Action Diffusion", "AdaptDiffuser: Diffusion Models as Adaptive Self-evolving Planners"], "answer_arxiv_id": ["2205.09991", "2209.03855", "2211.15657", "2303.04137v5", "2302.01877"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_994"} +{"question": "Which paper introduced 3D Gaussian Splatting as a method for 3D scene representation?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_995"} +{"question": "Could you tell me about studies that explore using latent state information during training to improve sample efficiency?", "answer": ["Asymmetric Actor Critic for Image-Based Robot Learning", "Agile Autonomous Driving using End-to-End Deep Imitation Learning", "Learning by cheating", "Robust Asymmetric Learning in POMDPs"], "answer_arxiv_id": ["1710.06542", "1709.07174", "1912.12294", "2012.15566v3"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_996"} +{"question": "Which paper introduced a multi-scale deformable self/cross-attention scheme in DETR?", "answer": ["Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2010.04159"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_997"} +{"question": "Any works addressing the heterogeneity in federated learning by adopting knowledge distillation?", "answer": ["Data-Free Knowledge Distillation for Heterogeneous Federated Learning", "FedBE: Making Bayesian Model Ensemble Applicable to Federated Learning", "Ensemble Distillation for Robust Model Fusion in Federated Learning"], "answer_arxiv_id": ["2105.10056", "2009.01974", "2006.07242"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_998"} +{"question": "What papers have analyzed the finite-time convergence of distributed actor-critic algorithms?", "answer": ["Sample and Communication-Efficient Decentralized Actor-Critic Algorithms with Finite-Time Analysis", "Towards Understanding Asynchronous Advantage Actor-critic: Convergence and Linear Speedup"], "answer_arxiv_id": ["2109.03699v2", "2012.15511"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_999"} +{"question": "Could you provide me some studies that force the use of incomplete modalities during training or testing in the context of flexible-modal FAS?", "answer": ["Flexible-Modal Face Anti-Spoofing: A Benchmark", "Visual Prompt Flexible-Modal Face Anti-Spoofing", "Rethinking Vision Transformer and Masked Autoencoder in Multimodal Face\n Anti-Spoofing"], "answer_arxiv_id": ["2202.08192", "2307.13958", "2302.05744"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_1000"} +{"question": "What research have attempted to learn 3D priors from Internet images and videos?", "answer": ["Implicit Mesh Reconstruction from Unannotated Image Collections", "Unsupervised Learning of Probably Symmetric Deformable 3D Objects from\n Images in the Wild", "De-rendering the World's Revolutionary Artefacts", "Articulation-aware Canonical Surface Mapping", "Self-supervised Single-view 3D Reconstruction via Semantic Consistency", "Shelf-Supervised Mesh Prediction in the Wild", "Pre-train, Self-train, Distill: A simple recipe for Supersizing 3D\n Reconstruction", "Learning Category-Specific Mesh Reconstruction from Image Collections", "DOVE: Learning Deformable 3D Objects by Watching Videos", "MagicPony: Learning Articulated 3D Animals in the Wild", "Farm3D: Learning Articulated 3D Animals by Distilling 2D Diffusion"], "answer_arxiv_id": ["2007.08504", "1911.11130", "2104.03954", "2004.00614", "2003.06473", "2102.06195", "2204.03642", "1803.07549", "2107.10844", "2211.12497", "2304.10535"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_1001"} +{"question": "Which studies explored the networked structure in multi-agent RL where either the communication or interaction had some networked structure?", "answer": ["Fully Decentralized Multi-Agent Reinforcement Learning with Networked Agents", "Scalable Reinforcement Learning for Multi-Agent Networked Systems", "Scalable and Sample Efficient Distributed Policy Gradient Algorithms in Multi-Agent Networked Systems", "Global Convergence of Localized Policy Iteration in Networked Multi-Agent Reinforcement Learning", "Convergence Rates for Localized Actor-Critic in Networked Markov Potential Games"], "answer_arxiv_id": ["1802.08757v2", "1912.02906", "2212.06357", "2211.17116v1", "2303.04865"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_1002"} +{"question": "Any studies highlighting the application of few-shot learning in neural network architecture search?", "answer": ["SMASH: One-Shot Model Architecture Search through HyperNetworks"], "answer_arxiv_id": ["1708.05344"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_1003"} +{"question": "What papers parameterize the scene through a set of 3D Gaussians inspired by NeRF's volume rendering formula?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering", "Approximate Differentiable Rendering with Algebraic Surfaces", "Flexible Techniques for Differentiable Rendering with 3D Gaussians", "VoGE: A Differentiable Volume Renderer using Gaussian Ellipsoids for\n Analysis-by-Synthesis"], "answer_arxiv_id": ["2308.04079", "2207.10606", "2308.14737", "2205.15401"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_1004"} +{"question": "Could you provide me some papers on LLMs striking a balance between utility and safety?", "answer": ["Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language\n Models that Follow Instructions"], "answer_arxiv_id": ["2309.07875"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_1005"} +{"question": "What studies are known for the development of various diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "1907.05600", "2006.11239"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_1006"} +{"question": "What works use web queries or models pretrained on general world knowledge due to the expensive nature of obtaining task specific textual knowledge?", "answer": ["Learning to Query Internet Text for Informing Reinforcement Learning Agents", "Playing Text-Based Games with Common Sense", "Embodied BERT: A Transformer Model for Embodied, Language-guided Visual Task Completion", "Inner Monologue: Embodied Reasoning through Planning with Language Models", "LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Models"], "answer_arxiv_id": ["2205.13079", "2012.02757", "2108.04927", "2207.05608", "2212.04088"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_1007"} +{"question": "Could you provide me studies that use LLM-based agents in the domain of software engineering and scientific inquiry?", "answer": ["ChatDev: Communicative Agents for Software Development", "Emergent autonomous scientific research capabilities of large language models"], "answer_arxiv_id": ["2307.07924", "2304.05332v1"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_1008"} +{"question": "Which papers suggest that each trained deep learning model is a sample from a Gaussian Process posterior?", "answer": ["Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent", "Deep Neural Networks as Gaussian Processes", "Scaling Limits of Wide Neural Networks with Weight Sharing: Gaussian Process Behavior, Gradient Independence, and Neural Tangent Kernel Derivation"], "answer_arxiv_id": ["1902.06720", "1711.00165", "1902.04760"], "source_meta": {"published_time": "20220611"}, "qid": "AutoScholarQuery_train_1009"} +{"question": "What works concluded that the quality of the model’s self-generated feedback is bounded by its existing knowledge and abilities?", "answer": ["GPT-4 Doesn't Know It's Wrong: An Analysis of Iterative Prompting for Reasoning Problems", "A Closer Look at the Self-Verification Abilities of Large Language\n Models in Logical Reasoning"], "answer_arxiv_id": ["2310.12397v1", "2311.07954"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_1010"} +{"question": "Which works implemented model or knowledge distillation techniques to reduce architecture complexity of UNet?", "answer": ["SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two\n Seconds"], "answer_arxiv_id": ["2306.00980"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_1011"} +{"question": "Which works initially discussed diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Improved Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2102.09672", "2011.13456"], "source_meta": {"published_time": "20240508"}, "qid": "AutoScholarQuery_train_1012"} +{"question": "Which papers have discussed and validated the infilling task in the terms of code-related tasks?", "answer": ["InCoder: A Generative Model for Code Infilling and Synthesis", "SantaCoder: don't reach for the stars!", "StarCoder: may the source be with you!", "Code Llama: Open Foundation Models for Code", "Efficient Training of Language Models to Fill in the Middle"], "answer_arxiv_id": ["2204.05999", "2301.03988", "2305.06161", "2308.12950", "2207.14255"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_1013"} +{"question": "What studies are on generating adversarial samples by optimizing the loss function in image attacks?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1706.06083", "1412.6572"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_1014"} +{"question": "Which papers have proposed reconstruction-based methods for unsupervised anomaly detection and localization?", "answer": ["Skip-GANomaly: Skip Connected and Adversarially Trained Encoder-Decoder\n Anomaly Detection", "Deep Autoencoding Models for Unsupervised Anomaly Segmentation in Brain\n MR Images", "Unsupervised Anomaly Detection with Generative Adversarial Networks to\n Guide Marker Discovery", "VT-ADL: A Vision Transformer Network for Image Anomaly Detection and\n Localization", "Inpainting Transformer for Anomaly Detection"], "answer_arxiv_id": ["1901.08954", "1804.04488", "1703.05921", "2104.10036", "2104.13897"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_1015"} +{"question": "Which study presents the PatchFlow method?", "answer": ["Multi-View Optimization of Local Feature Geometry"], "answer_arxiv_id": ["2003.08348"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_1016"} +{"question": "What research papers are dedicated to developing unsupervised anomaly detection methods?", "answer": ["COPOD: Copula-Based Outlier Detection", "ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions"], "answer_arxiv_id": ["2009.09463", "2201.00382"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_1017"} +{"question": "Which research work has presented heteroscedastic models for quantifying uncertainty in image segmentation?", "answer": ["Stochastic Segmentation Networks: Modelling Spatially Correlated Aleatoric Uncertainty"], "answer_arxiv_id": ["2006.06015"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1018"} +{"question": "Could you provide me some studies about optimization-based models in text-to-image generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion", "SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "Key-Locked Rank One Editing for Text-to-Image Personalization", "Break-A-Scene: Extracting Multiple Concepts from a Single Image"], "answer_arxiv_id": ["2208.01618", "2208.12242", "2212.04488", "2303.11305", "2305.01644", "2305.16311"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_1019"} +{"question": "Which paper can be seen as a generalization of ABC, covering the whole space of possible loss functions in Bayesian inference methods?", "answer": ["A primer on PAC-Bayesian learning"], "answer_arxiv_id": ["1901.05353"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_1020"} +{"question": "Could you provide me a research that proposes differentiable data structures, such as stacks or queues?", "answer": ["Learning to Transduce with Unbounded Memory"], "answer_arxiv_id": ["1506.02516"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_1021"} +{"question": "What papers discussed the condition under which federated optimization exhibits convergence guarantees?", "answer": ["On the Convergence of FedAvg on Non-IID Data"], "answer_arxiv_id": ["1907.02189"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_1022"} +{"question": "What works are related to the offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems"], "answer_arxiv_id": ["1812.02900", "2005.01643"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_1023"} +{"question": "Which studies have discussed the challenge of adapting to label shift?", "answer": ["Detecting and Correcting for Label Shift with Black Box Predictors", "Regularized Learning for Domain Adaptation under Label Shifts", "Maximum Likelihood with Bias-Corrected Calibration is Hard-To-Beat at Label Shift Adaptation", "A Unified View of Label Shift Estimation"], "answer_arxiv_id": ["1802.03916", "1903.09734", "1901.06852", "2003.07554"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_1024"} +{"question": "What research work reduces the number of sampling steps in DDIM by exploring a non-Markovian process related to neural ODEs?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_1025"} +{"question": "Which research works have analyzed the performance of two-layer neural tangent kernel models?", "answer": ["Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "The Dynamics of Gradient Descent for Overparametrized Neural Networks", "On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models"], "answer_arxiv_id": ["1901.08584", "2105.06569v1", "2103.05243"], "source_meta": {"published_time": "20230409"}, "qid": "AutoScholarQuery_train_1026"} +{"question": "What works focus on Koopman autoencoders which are related to classical learning methods?", "answer": ["Deep learning for universal linear embeddings of nonlinear dynamics", "Learning Deep Neural Network Representations for Koopman Operators of Nonlinear Dynamical Systems", "Learning Compositional Koopman Operators for Model-Based Control", "Forecasting Sequential Data Using Consistent Koopman Autoencoders"], "answer_arxiv_id": ["1712.09707", "1708.06850", "1910.08264", "2003.02236"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_1027"} +{"question": "Which work introduced a multi-resolution spectrogram discriminator for generating audio with sharp spectrograms?", "answer": ["UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generation"], "answer_arxiv_id": ["2106.07889"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_1028"} +{"question": "Can you list the research about multi-layer Large Language Model (LLM) systems?", "answer": ["Decomposed Prompting: A Modular Approach for Solving Complex Tasks", "AI Chains: Transparent and Controllable Human-AI Interaction by Chaining Large Language Model Prompts", "Language Model Cascades", "Language as a Latent Variable: Discrete Generative Models for Sentence Compression"], "answer_arxiv_id": ["2210.02406", "2110.01691", "2207.10342", "1609.07317"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_1029"} +{"question": "What works demonstrated an improvement in performance through the distillation of knowledge from the reader to the retriever?", "answer": ["Distilling Knowledge from Reader to Retriever for Question Answering"], "answer_arxiv_id": ["2012.04584"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_1030"} +{"question": "Could you provide me some studies on non-linear motion models in motion tracking?", "answer": ["Multiple Object Tracking by Flowing and Fusing", "DEFT: Detection Embeddings for Tracking", "MotionTrack: Learning Motion Predictor for Multiple Object Tracking"], "answer_arxiv_id": ["2001.11180", "2102.02267", "2306.02585"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_1031"} +{"question": "Any research focusing on object-centric scene editing in 3D Scene Editing of Radiance Fields?", "answer": ["DreamEditor: Text-Driven 3D Scene Editing with Neural Fields"], "answer_arxiv_id": ["2306.13455"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_1032"} +{"question": "What is an example of work on Text-to-Image (T2I) style transfer?", "answer": ["Inversion-Based Style Transfer with Diffusion Models"], "answer_arxiv_id": ["2211.13203"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1033"} +{"question": "Which papers focused on examining the language bias within large language models?", "answer": ["Lost in Translation: Large Language Models in Non-English Content Analysis", "Low-Resource Languages Jailbreak GPT-4"], "answer_arxiv_id": ["2306.07377v1", "2310.02446"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_1034"} +{"question": "What work developed the currently most powerful 3D GNN architecture for crystals and achieved the best crystal property prediction performance?", "answer": ["Periodic Graph Transformers for Crystal Material Property Prediction"], "answer_arxiv_id": ["2209.11807"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1035"} +{"question": "What work has been done in training diffusion models in pixel space for text-to-image generation?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2112.10741", "2205.11487"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1036"} +{"question": "Could you provide me some papers that focused on maximizing the mutual information between representations of different views of the same object?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Contrastive Multiview Coding", "Learning Representations by Maximizing Mutual Information Across Views"], "answer_arxiv_id": ["1807.03748", "2002.05709", "1906.05849", "1906.00910"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_1037"} +{"question": "Which works considered situations of 'D=1' and with challenging 'pi' distribution?", "answer": ["Unbiased Multilevel Monte Carlo methods for intractable distributions: MLMC meets MCMC"], "answer_arxiv_id": ["2204.04808"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_1038"} +{"question": "What papers discuss storing learnable parameters using data structures such as voxel grids, octrees, and hash tables?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Variable Bitrate Neural Fields", "Fourier PlenOctrees for Dynamic Radiance Field Rendering in Real-time", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "F2-NeRF: Fast Neural Radiance Field Training with Free Camera Trajectories"], "answer_arxiv_id": ["2112.05131", "2111.11215", "2103.14024", "2206.07707", "2202.08614", "2201.05989", "2303.15951"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_1039"} +{"question": "Which paper observed that LLM-based evaluators tend to prefer responses generated by themselves?", "answer": ["G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment"], "answer_arxiv_id": ["2303.16634"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_1040"} +{"question": "Can you tell me about some research that has simplified the problem of integrating modalities in one model by focusing on a small set of universal objectives?", "answer": ["CoCa: Contrastive Captioners are Image-Text Foundation Models", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks"], "answer_arxiv_id": ["2205.01917", "2208.10442"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_1041"} +{"question": "Which study utilizes the mixture of von Mises distributions over Euler angles using Biternion networks?", "answer": ["Deep Directional Statistics: Pose Estimation with Uncertainty Quantification"], "answer_arxiv_id": ["1805.03430"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_1042"} +{"question": "Which paper studies tileable image synthesis by manipulating latent spaces in pre-trained GANs?", "answer": ["SeamlessGAN: Self-Supervised Synthesis of Tileable Texture Maps"], "answer_arxiv_id": ["2201.05120"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_1043"} +{"question": "Which works have explored depth estimation from sequential video frames?", "answer": ["Unsupervised Learning of Monocular Depth Estimation and Visual Odometry\n with Deep Feature Reconstruction", "Self-supervised Learning with Geometric Constraints in Monocular Video:\n Connecting Flow, Depth, and Camera", "Feature-metric Loss for Self-supervised Learning of Depth and Egomotion", "Semantically-Guided Representation Learning for Self-Supervised\n Monocular Depth", "Fine-grained Semantics-aware Representation Enhancement for\n Self-supervised Monocular Depth Estimation"], "answer_arxiv_id": ["1803.03893", "1907.05820", "2007.10603", "2002.12319", "2108.08829"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_1044"} +{"question": "What research papers have extensively studied about distributed training algorithms for nonconvex minimax under a serverless decentralized setting?", "answer": ["Decentralized Local Stochastic Extra-Gradient for Variational Inequalities", "PRECISION: Decentralized Constrained Min-Max Learning with Low Communication and Sample Complexities", "Decentralized Riemannian Algorithm for Nonconvex Minimax Problems"], "answer_arxiv_id": ["2106.08315", "2303.02532", "2302.03825"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_1045"} +{"question": "What papers have implemented and improved the self-supervised pre-training strategy SimCLR for visual recognition problems?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Improved Baselines with Momentum Contrastive Learning", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "2003.04297", "1911.05722"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_1046"} +{"question": "Which publications are about the domain of customized image generation?", "answer": ["Key-Locked Rank One Editing for Text-to-Image Personalization", "A Neural Space-Time Representation for Text-to-Image Personalization", "Imagic: Text-Based Real Image Editing with Diffusion Models", "MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models", "AnyDoor: Zero-shot Object-level Image Customization"], "answer_arxiv_id": ["2305.01644", "2305.15391", "2210.09276", "2304.08465v1", "2302.12228", "2307.09481"], "source_meta": {"published_time": "20240522"}, "qid": "AutoScholarQuery_train_1047"} +{"question": "What works used Neural Turing Machines for augmenting memory?", "answer": ["Neural Turing Machines"], "answer_arxiv_id": ["1410.5401"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_1048"} +{"question": "What papers have contributed to the design of suitable architectural search spaces?", "answer": ["EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", "Searching for MobileNetV3", "Kaleidoscope: An Efficient, Learnable Representation For All Structured Linear Maps", "Rethinking Neural Operations for Diverse Tasks"], "answer_arxiv_id": ["1905.11946", "1905.02244", "2012.14966", "2103.15798"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_1049"} +{"question": "What works proposed 'Predict, then Optimize' (SPO) methods as alternatives for task-based model learning?", "answer": ["Generalization Bounds in the Predict-then-Optimize Framework", "Smart Predict-and-Optimize for Hard Combinatorial Optimization Problems", "Active Learning in the Predict-then-Optimize Framework: A Margin-Based Approach"], "answer_arxiv_id": ["1905.11488v3", "1911.10092", "2305.06584v1"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_1050"} +{"question": "What works conducted pretraining of Vision-language models on large and noisy multi-modal datasets?", "answer": ["LAION-5B: An open large-scale dataset for training next generation\n image-text models", "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs"], "answer_arxiv_id": ["2210.08402", "2111.02114"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_1051"} +{"question": "Can you mention some researches that focus on the problem of overfitting to training sources because of unique frequency-level artifacts in generated images?", "answer": ["FrePGAN: Robust Deepfake Detection Using Frequency-level Perturbations"], "answer_arxiv_id": ["2202.03347"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_1052"} +{"question": "Could you name the works about the techniques of machine unlearning that avoid full dataset access", "answer": ["Data Redaction from Pre-trained GANs"], "answer_arxiv_id": ["2206.14389"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_1053"} +{"question": "Could you provide me some works about volume networks?", "answer": ["RGBD Based Dimensional Decomposition Residual Network for 3D Semantic\n Scene Completion", "Cascaded Context Pyramid for Full-Resolution 3D Semantic Scene\n Completion"], "answer_arxiv_id": ["1903.00620", "1908.00382"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_1054"} +{"question": "Could you provide me some works about prior-free matting?", "answer": ["Bridging Composite and Real: Towards End-to-end Deep Image Matting", "MODNet: Real-Time Trimap-Free Portrait Matting via Objective Decomposition", "Semantic Human Matting", "Privacy-Preserving Portrait Matting", "U2-Net: Going Deeper with Nested U-Structure for Salient Object Detection"], "answer_arxiv_id": ["2010.16188v3", "2011.11961", "1809.01354", "2104.14222", "2005.09007"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_1055"} +{"question": "What studies provide concept-based explanations for sequential decision making in relation to state preconditions and action costs?", "answer": ["Bridging the Gap: Providing Post-Hoc Symbolic Explanations for Sequential Decision-Making Problems with Inscrutable Representations"], "answer_arxiv_id": ["2002.01080"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_1056"} +{"question": "Which works have attempted to boost the performance of SAM by exploring its geometry, minimizing surrogate gap, or accelerating its training speed?", "answer": ["ASAM: Adaptive Sharpness-Aware Minimization for Scale-Invariant Learning of Deep Neural Networks", "Fisher SAM: Information Geometry and Sharpness Aware Minimisation", "Surrogate Gap Minimization Improves Sharpness-Aware Training", "Sharpness-Aware Training for Free", "Towards Efficient and Scalable Sharpness-Aware Minimization"], "answer_arxiv_id": ["2102.11600", "2206.04920", "2203.08065", "2205.14083", "2203.02714"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_1057"} +{"question": "Could you provide me some studies about traditional methods of mitigating forgetting by implementing parameter importance computation?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Memory Aware Synapses: Learning what (not) to forget"], "answer_arxiv_id": ["1612.00796", "1711.09601"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_1058"} +{"question": "Which works proposed watermark-based methods for AI-text detection?", "answer": ["A Watermark for Large Language Models"], "answer_arxiv_id": ["2301.10226"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_1059"} +{"question": "Which papers have made advancements in enhancing the rendering quality of the NeRF framework?", "answer": ["NeRF++: Analyzing and Improving Neural Radiance Fields", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural\n Radiance Fields", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2010.07492", "2103.13415", "2111.12077", "2307.11335", "2304.06706"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_1060"} +{"question": "What benchmark incorporated CoLA as a subtask and soon surpassed by language models?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language\n Understanding", "SuperGLUE: A Stickier Benchmark for General-Purpose Language\n Understanding Systems"], "answer_arxiv_id": ["1804.07461", "1905.00537"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_1061"} +{"question": "Could you name some previous works which have attempted at developing summary scoring systems?", "answer": ["Hallucinated but Factual! Inspecting the Factuality of Hallucinations in\n Abstractive Summarization", "Benchmarking Generation and Evaluation Capabilities of Large Language\n Models for Instruction Controllable Summarization", "LongEval: Guidelines for Human Evaluation of Faithfulness in Long-form\n Summarization"], "answer_arxiv_id": ["2109.09784", "2311.09184", "2301.13298"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_1062"} +{"question": "Which works demonstrate that instruction tuning improves performance on new, unseen tasks?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization", "Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2110.08207", "2109.01652"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_1063"} +{"question": "Which works proposed and demonstrated the feasibility of backdoor attacks on instruction-tuned LLMs?", "answer": ["On the Exploitability of Instruction Tuning", "Poisoning Language Models During Instruction Tuning", "Backdooring Instruction-Tuned Large Language Models with Virtual Prompt\n Injection"], "answer_arxiv_id": ["2306.17194", "2305.00944", "2307.16888"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_1064"} +{"question": "Which works have focussed on mitigating spherical distortions in monocular 360 depth estimation with single projection input?", "answer": ["ODE-CNN: Omnidirectional Depth Extension Networks", "OmniDepth: Dense Depth Estimation for Indoors Spherical Panoramas"], "answer_arxiv_id": ["2007.01475", "1807.09620"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_1065"} +{"question": "Which studies show the benefits of increasing batch size in contrastive learning?", "answer": ["Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup"], "answer_arxiv_id": ["2101.06983"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_1066"} +{"question": "What studies introduced the Wang–Landau algorithm in order to drive the sampler to explore under-explored energy regions?", "answer": ["Molecular Dynamics in the Multicanonical Ensemble: Equivalence of Wang–Landau Sampling, Statistical Temperature Molecular Dynamics, and Metadynamics", "A histogram-free multicanonical Monte Carlo algorithm for the basis expansion of density of states"], "answer_arxiv_id": ["1401.6184", "1707.07049"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_1067"} +{"question": "In what studies are CoT capabilities distilled into smaller models, to improve their performance in specific tasks?", "answer": ["Teaching Small Language Models to Reason", "Large Language Models Are Reasoning Teachers"], "answer_arxiv_id": ["2212.08410", "2212.10071"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_1068"} +{"question": "Which paper introduced a Normal-Inverse-Gamma (NIG) distribution to characterize the uncertainty in image matting?", "answer": ["Accurate Uncertainties for Deep Learning Using Calibrated Regression"], "answer_arxiv_id": ["1807.00263"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_1069"} +{"question": "Which work establishes FedPR, a prompt-tuning-based GFL method?", "answer": ["Learning Federated Visual Prompt in Null Space for MRI Reconstruction"], "answer_arxiv_id": ["2303.16181"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_1070"} +{"question": "What papers introduced categorizations concerning the type of challenge a prompt involves in text-to-image generation?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2205.11487", "2206.10789"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_1071"} +{"question": "Which studies came up with innovative voxel pooling strategies?", "answer": ["Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by\n Implicitly Unprojecting to 3D", "SA-BEV: Generating Semantic-Aware Bird's-Eye-View Feature for Multi-view\n 3D Object Detection"], "answer_arxiv_id": ["2008.05711", "2307.11477"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_1072"} +{"question": "What studies take into account cross-device localization?", "answer": ["ADVIO: An authentic dataset for visual-inertial odometry", "LaMAR: Benchmarking Localization and Mapping for Augmented Reality", "Benchmarking 6DOF Outdoor Visual Localization in Changing Conditions", "Large-scale Localization Datasets in Crowded Indoor Spaces", "Long-term Visual Localization with Mobile Sensors"], "answer_arxiv_id": ["1807.09828v1", "2210.10770", "1707.09092", "2105.08941", "2304.07691"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1073"} +{"question": "Which works are about modular reasoning approaches where the LM is used within a reasoning algorithm?", "answer": ["LAMBADA: Backward Chaining for Automated Reasoning in Natural Language", "Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning", "Iteratively Prompt Pre-trained Language Models for Chain of Thought", "Decomposed Prompting: A Modular Approach for Solving Complex Tasks"], "answer_arxiv_id": ["2212.13894", "2205.09712", "2203.08383", "2210.02406"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_1074"} +{"question": "Which papers investigated the knowledge-grounded dialogue generation task?", "answer": ["Wizard of Wikipedia: Knowledge-Powered Conversational agents", "MultiDoc2Dial: Modeling Dialogues Grounded in Multiple Documents"], "answer_arxiv_id": ["1811.01241", "2109.12595"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_1075"} +{"question": "What research helped identify semantic leakage as a consequence of improper mapping between syntactic and visual binding?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "DALL-E 2 Fails to Reliably Capture Common Syntactic Processes", "DALLE-2 is Seeing Double: Flaws in Word-to-Concept Mapping in Text2Image Models"], "answer_arxiv_id": ["2204.06125", "2210.12889v2", "2210.10606"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_1076"} +{"question": "What studies developed a single-path MBConv-based search space for Neural Architecture Search?", "answer": ["BigNAS: Scaling Up Neural Architecture Search with Big Single-Stage Models", "AttentiveNAS: Improving Neural Architecture Search via Attentive Sampling", "EfficientBERT: Progressively Searching Multilayer Perceptron via Warm-up Knowledge Distillation", "Cream of the Crop: Distilling Prioritized Paths For One-Shot Neural Architecture Search", "Searching by Generating: Flexible and Efficient One-Shot NAS with Architecture Generator", "FBNetV5: Neural Architecture Search for Multiple Tasks in One Run", "Efficient Architecture Search for Diverse Tasks"], "answer_arxiv_id": ["2003.11142", "2011.09011", "2109.07222", "2010.15821", "2103.07289", "2111.10007", "2204.07554"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1077"} +{"question": "What are the works that tackled the computational intractability of active learning algorithm selection for large datasets with expensive models?", "answer": ["Dynamic Ensemble Active Learning: A Non-Stationary Bandit with Expert Advice"], "answer_arxiv_id": ["1810.07778"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_1078"} +{"question": "What work introduced CoOp and CoCoOp in prompt engineering?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2109.01134", "2203.05557"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_1079"} +{"question": "What studies worked on designing algorithms for computing OT plans for weak OT?", "answer": ["Neural Optimal Transport", "Kernel Neural Optimal Transport"], "answer_arxiv_id": ["2201.12220", "2205.15269"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_1080"} +{"question": "Could you provide me some works about adaptivity of regret bounds in online learning?", "answer": ["Strongly Adaptive Online Learning", "Improved Strongly Adaptive Online Learning using Coin Betting"], "answer_arxiv_id": ["1502.07073", "1610.04578"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_1081"} +{"question": "What works discuss maintaining image properties in tasks such as super-resolution using GANs and additional similarity losses?", "answer": ["Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks"], "answer_arxiv_id": ["1703.10593"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1082"} +{"question": "What are some examples of traditional post-hoc explanation methods that use gradient-based techniques?", "answer": ["Not Just a Black Box: Learning Important Features Through Propagating\n Activation Differences", "SmoothGrad: removing noise by adding noise", "Visualizing and Understanding Convolutional Networks", "Full-Gradient Representation for Neural Network Visualization", "Axiomatic Attribution for Deep Networks", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based\n Localization"], "answer_arxiv_id": ["1605.01713", "1706.03825", "1311.2901", "1905.00780", "1703.01365", "1610.02391"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_1083"} +{"question": "Which works focused on adopting orthogonal learning in Catastrophic Forgetting (CL)?", "answer": ["Flattening Sharpness for Dynamic Gradient Projection Memory Benefits\n Continual Learning", "Gradient Projection Memory for Continual Learning", "Orthogonal Gradient Descent for Continual Learning", "Gradient Episodic Memory for Continual Learning"], "answer_arxiv_id": ["2110.04593", "2103.09762", "1910.07104", "1706.08840"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_1084"} +{"question": "Could you provide me with the research that treats user profiles as a time-varying variable?", "answer": ["Differentially-Private Federated Linear Bandits"], "answer_arxiv_id": ["2010.11425"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_1085"} +{"question": "Are there any works that showed positive results such as learning ReLU regression problems with bounded and well-spread features?", "answer": ["Agnostic Learning of a Single Neuron with Gradient Descent"], "answer_arxiv_id": ["2005.14426"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_1086"} +{"question": "How have DPMs been used for likelihood estimation according to recent literature?", "answer": ["Maximum Likelihood Training of Score-Based Diffusion Models", "Variational Diffusion Models", "Maximum Likelihood Training for Score-Based Diffusion ODEs by High-Order Denoising Score Matching", "Improved Techniques for Maximum Likelihood Estimation for Diffusion ODEs"], "answer_arxiv_id": ["2101.09258", "2107.00630", "2206.08265", "2305.03935"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_1087"} +{"question": "Which works extended matrix mechanisms to the adaptive streaming setting?", "answer": ["Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams", "Multi-Epoch Matrix Factorization Mechanisms for Private Machine Learning"], "answer_arxiv_id": ["2202.08312", "2211.06530"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_1088"} +{"question": "What research provided a solution to the gradient backpropagation problem in the backward pass?", "answer": ["Projective Manifold Gradient Layer for Deep Rotation Regression"], "answer_arxiv_id": ["2110.11657"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_1089"} +{"question": "Could you provide me with studies on the Learning to Rank problem and its applications?", "answer": ["Ranking Distillation: Learning Compact Ranking Models With High Performance for Recommender System"], "answer_arxiv_id": ["1809.07428"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_1090"} +{"question": "Which works started using text prompts to control the generated image in the field of image editing?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2112.10741", "2204.06125"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_1091"} +{"question": "Which papers mentioned that FLOPs do not accurately represent the model’s speed or latency?", "answer": ["An Energy and GPU-Computation Efficient Backbone Network for Real-Time\n Object Detection", "ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture\n Design"], "answer_arxiv_id": ["1904.09730", "1807.11164"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_1092"} +{"question": "Could you provide me some studies that help support the superficial alignment hypothesis?", "answer": ["LIMA: Less Is More for Alignment"], "answer_arxiv_id": ["2305.11206"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_1093"} +{"question": "Which papers use graphons for analyzing graph signal processing and MPNNs?", "answer": ["Graphon Signal Processing", "Convergence and Stability of Graph Convolutional Networks on Large Random Graphs", "Transferability of Graph Neural Networks: an Extended Graphon Approach"], "answer_arxiv_id": ["2003.05030", "2006.01868", "2109.10096"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_1094"} +{"question": "Could you provide me a work that implemented the interior point method on Riemannian manifolds?", "answer": ["Interior-point methods on manifolds: theory and applications"], "answer_arxiv_id": ["2303.04771"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_1095"} +{"question": "Which papers focused on generative model-based augmentation for 3D data?", "answer": ["Improved Adversarial Systems for 3D Object Generation and Reconstruction", "Learning Representations and Generative Models for 3D Point Clouds"], "answer_arxiv_id": ["1707.09557", "1707.02392"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_1096"} +{"question": "What publications have explored the use of even high-order algebraic shapes in their research?", "answer": ["QuadricSLAM: Dual Quadrics from Object Detections as Landmarks in\n Object-oriented SLAM", "Simultaneous Localisation and Mapping with Quadric Surfaces"], "answer_arxiv_id": ["1804.04011", "2203.08040"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_1097"} +{"question": "What papers have participated in initiatives to design effective automated evaluators for the quality of texts generated by language models?", "answer": ["QAFactEval: Improved QA-Based Factual Consistency Evaluation for\n Summarization", "Towards a Unified Multi-Dimensional Evaluator for Text Generation", "Is ChatGPT a Good NLG Evaluator? A Preliminary Study"], "answer_arxiv_id": ["2112.08542", "2210.07197", "2303.04048"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_train_1098"} +{"question": "Which research publications handle post-training compression scenarios?", "answer": ["Up or Down? Adaptive Rounding for Post-Training Quantization", "Optimal Brain Compression: A Framework for Accurate Post-Training\n Quantization and Pruning"], "answer_arxiv_id": ["2004.10568", "2208.11580"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_1099"} +{"question": "Could you provide me the works that have proposed diffusion-based image editing methods?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing", "Guided Image Synthesis via Initial Image Editing in Diffusion Model", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Zero-shot Image-to-Image Translation", "MagicMix: Semantic Mixing with Diffusion Models", "DragonDiffusion: Enabling Drag-style Manipulation on Diffusion Models", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Text2LIVE: Text-Driven Layered Image and Video Editing"], "answer_arxiv_id": ["2208.01626", "2304.08465v1", "2305.03382", "2210.09276", "2302.03027", "2210.16056", "2307.02421", "2211.12572", "2211.09800", "2108.01073", "2204.02491"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_1100"} +{"question": "Could you mention some studies about pretraining behavior policies to guide downstream policies training?", "answer": ["BRAC+: Improved Behavior Regularized Actor Critic for Offline Reinforcement Learning"], "answer_arxiv_id": ["2110.00894"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_1101"} +{"question": "What papers propose techniques such as subword mapping, transliteration, leveraging lexical overlap, vocabulary clustering and reallocation, continued or language-adaptive pretraining, and adaptation via bilingual lexica?", "answer": ["Subword Mapping and Anchoring across Languages", "Overlap-based Vocabulary Generation Improves Cross-lingual Transfer Among Related Languages", "Improving Multilingual Models with Language-Clustered Vocabularies", "How to Adapt Your Pretrained Multilingual Model to 1600 Languages", "Expanding Pretrained Models to Thousands More Languages via Lexicon-based Adaptation"], "answer_arxiv_id": ["2109.04556", "2203.01976v2", "2010.12777", "2106.02124", "2203.09435"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_1102"} +{"question": "Could you provide me studies that present arguments for adjusting load and temperature to mitigate overfitting in order to encourage models to exhibit Heavy-Tailed Self-Regularization (HT-SR)?", "answer": ["Rethinking generalization requires revisiting old ideas: statistical mechanics approaches and complex learning behavior"], "answer_arxiv_id": ["1710.09553"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_1103"} +{"question": "What studies demonstrated the use of decomposed convolutional filters for stochastic image generation tasks?", "answer": ["Stochastic Conditional Generative Networks with Basis Decomposition"], "answer_arxiv_id": ["1909.11286"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_1104"} +{"question": "Are there any research papers that explore the idea of competence-based intrinsic motivation and unsupervised skill discovery methods in RL?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function", "Dynamics-Aware Unsupervised Discovery of Skills"], "answer_arxiv_id": ["1802.06070", "1907.01657"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_1105"} +{"question": "What paper proposes inserting new cross-attention layers into the LLM for injection of visual features?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1106"} +{"question": "Which research works have contributed to the advancement in vector quantization and diffusion modeling for text-to-image generation?", "answer": ["Zero-Shot Text-to-Image Generation", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2102.12092", "2112.10741", "2112.10752", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_1107"} +{"question": "Could you provide me some research about neural networks exploiting sequential data in SBR tasks?", "answer": ["Session-based Recommendations with Recurrent Neural Networks"], "answer_arxiv_id": ["1511.06939"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_1108"} +{"question": "What studies tackle semantic segmentation without the use of human annotations?", "answer": ["PiCIE: Unsupervised Semantic Segmentation using Invariance and\n Equivariance in Clustering", "Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "Invariant Information Clustering for Unsupervised Image Classification\n and Segmentation", "Leveraging Hidden Positives for Unsupervised Semantic Segmentation"], "answer_arxiv_id": ["2103.17070", "2203.08414", "1807.06653", "2303.15014"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_1109"} +{"question": "Could you provide me a work introduced a new family of CNN for online classification of videos and developed a more comprehensive approach to data caching?", "answer": ["MoViNets: Mobile Video Networks for Efficient Video Recognition"], "answer_arxiv_id": ["2103.11511"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_1110"} +{"question": "Could you provide me studies that investigated extrapolation in the context of learning equations of control systems?", "answer": ["Extrapolation and learning equations", "Learning Equations for Extrapolation and Control"], "answer_arxiv_id": ["1610.02995", "1806.07259"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_1111"} +{"question": "What examples of research fall under the category of Coupled GNNs, which intertwine feature transformation and propagation within each layer?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Inductive Representation Learning on Large Graphs", "Graph Attention Networks"], "answer_arxiv_id": ["1609.02907", "1706.02216", "1710.10903"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_1112"} +{"question": "Which works applied regression in monocular mocap methods?", "answer": ["End-to-end Recovery of Human Shape and Pose", "Learning 3D Human Dynamics from Video", "VIBE: Video Inference for Human Body Pose and Shape Estimation", "Neural Descent for Visual 3D Human Pose and Shape"], "answer_arxiv_id": ["1712.06584", "1812.01601", "1912.05656", "2008.06910"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1113"} +{"question": "What is the reference for the recently published method TTOpt that is strongly related to the problem of finding the extreme entry value within a tensor?", "answer": ["TTOpt: A Maximum Volume Quantized Tensor Train-based Optimization and its Application to Reinforcement Learning"], "answer_arxiv_id": ["2205.00293"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_1114"} +{"question": "Which works have used CNN for token embedding and projection in vision transformers?", "answer": ["CvT: Introducing Convolutions to Vision Transformers"], "answer_arxiv_id": ["2103.15808"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_1115"} +{"question": "What literature projects the original point clouds to intermediate regular grid structures such as voxels?", "answer": ["PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection"], "answer_arxiv_id": ["1912.13192"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_1116"} +{"question": "Are there any works about JAMs in the context of inherently interpretable models?", "answer": ["Learning to Explain: An Information-Theoretic Perspective on Model Interpretation"], "answer_arxiv_id": ["1802.07814"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_1117"} +{"question": "Could you provide me some research that has utilized non-parametric tests to detect memorization or exact data copying in generative models?", "answer": ["An empirical study on evaluation metrics of generative adversarial networks", "Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs", "A Non-Parametric Test to Detect Data-Copying in Generative Models"], "answer_arxiv_id": ["1806.07755", "1706.02633", "2004.05675v1"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_1118"} +{"question": "What papers have used a context encoder as a pretext for self-supervised learning for speech?", "answer": ["Representation Learning with Contrastive Predictive Coding", "wav2vec: Unsupervised Pre-training for Speech Recognition", "Generative Pre-training for Speech with Autoregressive Predictive Coding", "vq-wav2vec: Self-Supervised Learning of Discrete Speech Representations", "Deep Contextualized Acoustic Representations For Semi-Supervised Speech Recognition", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations", "MOCKINGJAY: UNSUPERVISED SPEECH REPRESENTATION LEARNING WITH DEEP BIDIRECTIONAL TRANSFORMER ENCODERS"], "answer_arxiv_id": ["1807.03748", "1904.05862", "1910.12607", "1910.05453", "1912.01679", "2006.11477", "1910.12638"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_1119"} +{"question": "Can you list the works that introduced code generation tasks in multiple languages?", "answer": ["Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2203.07814"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_1120"} +{"question": "What works reviewed the Euler-Lagrange equation from a Finsler perspective?", "answer": ["Generalized Nonlinear and Finsler Geometry for Robotics"], "answer_arxiv_id": ["2010.14745"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_1121"} +{"question": "Can you provide some papers which involved ELECTRA as a part of their studies?", "answer": ["ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators"], "answer_arxiv_id": ["2003.10555"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_1122"} +{"question": "Could you provide me some studies that relied on Generative Adversarial Networks (GANs) for video synthesis?", "answer": ["Generating Videos with Dynamics-aware Implicit Generative Adversarial\n Networks", "StyleGAN-V: A Continuous Video Generator with the Price, Image Quality\n and Perks of StyleGAN2", "CogVideo: Large-scale Pretraining for Text-to-Video Generation via\n Transformers", "Learning to Forecast and Refine Residual Motion for Image-to-Video\n Generation", "Generating Videos with Scene Dynamics", "MoStGAN-V: Video Generation with Temporal Motion Styles", "A Good Image Generator Is What You Need for High-Resolution Video\n Synthesis"], "answer_arxiv_id": ["2202.10571", "2112.14683", "2205.15868", "1807.09951", "1609.02612", "2304.02777", "2104.15069"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_1123"} +{"question": "In what papers the universal approximation property of deep residual fully convolutional networks is explored from the perspective of dynamical systems?", "answer": ["On the Universal Approximation Property of Deep Fully Convolutional Neural Networks"], "answer_arxiv_id": ["2211.14047"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_1124"} +{"question": "What paper presents a lightweight model that facilitates NLI by segmenting input text into sentence units?", "answer": ["SummaC: Re-Visiting NLI-based Models for Inconsistency Detection in\n Summarization"], "answer_arxiv_id": ["2111.09525"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_train_1125"} +{"question": "Could you provide me papers that describe the setting when a smooth, leaky ReLU activation function was trained using the logistic instead of the hinge loss?", "answer": ["Benign Overfitting without Linearity: Neural Network Classifiers Trained by Gradient Descent for Noisy Linear Data"], "answer_arxiv_id": ["2202.05928"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_1126"} +{"question": "Which research papers pertain to source-free domain adaptation utilities like black-box, online, continual, and universal adaptation?", "answer": ["Unsupervised Domain Adaptation of Black-Box Source Models", "Casting a BAIT for Offline and Online Source-free Domain Adaptation", "Continual Test-Time Domain Adaptation", "Universal Source-Free Domain Adaptation"], "answer_arxiv_id": ["2101.02839", "2010.12427", "2203.13591", "2004.04393"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_1127"} +{"question": "Can you provide examples of research that proposed GAN models for high-quality image synthesis?", "answer": ["Progressive Growing of GANs for Improved Quality, Stability, and Variation", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["1710.10196", "1809.11096", "1812.04948", "1912.04958"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_1128"} +{"question": "Could you cite the research that explains how every SDE can be converted into an ODE with the same marginal distribution?", "answer": ["Denoising Diffusion Implicit Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2010.02502", "2011.13456"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_1129"} +{"question": "Which study proposed using gradient flows of f-divergences to refine fake samples in deep generative models?", "answer": ["Deep Generative Learning via Variational Gradient Flow"], "answer_arxiv_id": ["1901.08469"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_1130"} +{"question": "What research have proposed improving RNNs in tasks such as language processing, computer vision, time-series analysis, and speech recognition?", "answer": ["Hierarchical Multiscale Recurrent Neural Networks"], "answer_arxiv_id": ["1609.01704"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_1131"} +{"question": "What papers have focused on the field of arbitrary-scale generative super resolution?", "answer": ["Local Implicit Normalizing Flow for Arbitrary-Scale Image\n Super-Resolution"], "answer_arxiv_id": ["2303.05156"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_1132"} +{"question": "What studies used Transformer backbone to regress full-body motion directly from three 6D trackers?", "answer": ["AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion\n Sensing", "Realistic Full-Body Tracking from Sparse Observations via Joint-Level\n Modeling", "HMD-NeMo: Online 3D Avatar Motion Generation From Sparse Observations"], "answer_arxiv_id": ["2207.13784", "2308.08855", "2308.11261"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_1133"} +{"question": "What research constructs a static preference dataset with responses obtained from the SFT model ordered by JM?", "answer": ["SLiC-HF: Sequence Likelihood Calibration with Human Feedback"], "answer_arxiv_id": ["2305.10425"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_1134"} +{"question": "Which works have contributed to the progress of autonomous agents?", "answer": ["WebArena: A Realistic Web Environment for Building Autonomous Agents", "Voyager: An Open-Ended Embodied Agent with Large Language Models", "Generative Agents: Interactive Simulacra of Human Behavior", "Humanoid Agents: Platform for Simulating Human-like Generative Agents", "PromptAgent: Strategic Planning with Language Models Enables\n Expert-level Prompt Optimization"], "answer_arxiv_id": ["2307.13854", "2305.16291", "2304.03442", "2310.05418", "2310.16427"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_1135"} +{"question": "Which works proposed approaches for the layer-wise compression problem?", "answer": ["SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot", "A Simple and Effective Pruning Approach for Large Language Models", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large\n Language Models", "SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight\n Compression"], "answer_arxiv_id": ["2301.00774", "2306.11695", "2211.10438", "2306.03078"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_1136"} +{"question": "Which studies propose how to handle dynamic objects in self-supervised monocular depth estimation?", "answer": ["Learning to Fuse Monocular and Multi-view Cues for Multi-frame Depth\n Estimation in Dynamic Scenes", "Digging Into Self-Supervised Monocular Depth Estimation", "Disentangling Object Motion and Occlusion for Unsupervised Multi-frame\n Monocular Depth", "Kick Back & Relax: Learning to Reconstruct the World by Watching SlowTV", "Adaptive Fusion of Single-View and Multi-View Depth for Autonomous\n Driving"], "answer_arxiv_id": ["2304.08993", "1806.01260", "2203.15174", "2307.10713", "2403.07535"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_1137"} +{"question": "Which studies implemented neural networks with mesh-specific architecture design in the context of neural PDE solvers?", "answer": ["Towards Multi-spatiotemporal-scale Generalized PDE Modeling", "Message Passing Neural PDE Solvers", "Guaranteed Conservation of Momentum for Learning Particle-based Fluid Dynamics", "Learning the Solution Operator of Boundary Value Problems using Graph Neural Networks", "Deep Learning for Reduced Order Modelling and Efficient Temporal Evolution of Fluid Simulations", "Towards physics-informed deep learning for turbulent flow prediction", "Deep Learning Methods for Reynolds-Averaged Navier-Stokes Simulations of Airfoil Flows", "EAGLE: Large-scale Learning of Turbulent Fluid Dynamics with Mesh Transformers"], "answer_arxiv_id": ["2209.15616v2", "2202.03376", "2210.06036", "2206.14092v2", "2107.04556", "1911.08655", "1810.08217", "2302.10803"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_1138"} +{"question": "Which studies discuss the concept of maximizing and minimizing agreement between augmented views in contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2002.05709", "2103.03230", "2006.07733", "2011.10566"], "source_meta": {"published_time": "20230218"}, "qid": "AutoScholarQuery_train_1139"} +{"question": "Which research proposed a simplified interaction between hidden states in traditional recurrent neural networks?", "answer": ["Simple Recurrent Units for Highly Parallelizable Recurrence"], "answer_arxiv_id": ["1709.02755"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_1140"} +{"question": "Have there been any near-neighbor search studies using graph-based methods?", "answer": ["Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs", "Optimization of Indexing Based on k-Nearest Neighbor Graph for Proximity Search in High-dimensional Data"], "answer_arxiv_id": ["1603.09320v4", "1810.07355"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_1141"} +{"question": "Can you point out the works focused on accelerating NeRF by reducing the complexity of MLP?", "answer": ["DeRF: Decomposed Radiance Fields"], "answer_arxiv_id": ["2011.12490"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_1142"} +{"question": "What papers discussed the person-specific avatars generation by incorporating a 3DMM in the neural implicit representation?", "answer": ["Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar\n Reconstruction", "RigNeRF: Fully Controllable Neural 3D Portraits", "Instant Volumetric Head Avatars", "I M Avatar: Implicit Morphable Head Avatars from Videos", "PointAvatar: Deformable Point-based Head Avatars from Videos", "Mixture of Volumetric Primitives for Efficient Neural Rendering", "GaussianAvatars: Photorealistic Head Avatars with Rigged 3D Gaussians"], "answer_arxiv_id": ["2012.03065", "2206.06481", "2211.12499", "2112.07471", "2212.08377", "2103.01954", "2312.02069"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_1143"} +{"question": "Could you provide me with studies about diffusional probabilistic models?", "answer": ["Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Improved Denoising Diffusion Probabilistic Models", "Zero-Shot Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "Shifted Diffusion for Text-to-image Generation", "Scalable Diffusion Models with Transformers"], "answer_arxiv_id": ["2006.11239", "2105.05233", "2102.09672", "2102.12092", "2112.10752", "2211.15388", "2212.09748"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_1144"} +{"question": "Which papers discussed the methodology of prompt-tuning within the context of Parameter Efficient Transfer Learning (PETL)?", "answer": ["Language Models are Few-Shot Learners", "Template-Based Named Entity Recognition Using BART", "Visual Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "GPT Understands, Too", "Language Models as Knowledge Bases?", "Learning Transferable Visual Models From Natural Language Supervision", "Conditional Prompt Learning for Vision-Language Models", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2005.14165", "2106.01760", "2203.12119", "2101.00190", "2103.10385", "1909.01066", "2103.00020", "2203.05557", "2109.01134"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_1145"} +{"question": "Which work established polynomial time and sample efficient methods for learning the mean and variance of a univariate Gaussian?", "answer": ["Finite Sample Differentially Private Confidence Intervals"], "answer_arxiv_id": ["1711.03908"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_1146"} +{"question": "Which works use DMs in text-to-image generation?", "answer": ["eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2211.01324", "2204.06125", "2302.05543"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_1147"} +{"question": "Could you list the studies that worked on enhancing large language models with human feedback or external knowledge?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback", "Check Your Facts and Try Again: Improving Large Language Models with\n External Knowledge and Automated Feedback"], "answer_arxiv_id": ["2112.09332", "2302.12813"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_1148"} +{"question": "What work improved the efficiency of the S4 and S4D models and introduced the S5 model?", "answer": ["Simplified State Space Layers for Sequence Modeling"], "answer_arxiv_id": ["2208.04933"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_1149"} +{"question": "Which research paper introduced a state-of-the-art algorithm for document retrieval named ColBERT?", "answer": ["ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT"], "answer_arxiv_id": ["2004.12832"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_1150"} +{"question": "Can you mention some studies that extract important sentences to construct the desired timeline?", "answer": ["A Temporally Sensitive Submodularity Framework for Timeline Summarization", "Analyzing Evolving Stories in News Articles"], "answer_arxiv_id": ["1810.07949", "1703.08593"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_1151"} +{"question": "What papers designed AlpacaEval and MT-Bench to collect open-ended questions across different domains?", "answer": ["AlpacaFarm: A Simulation Framework for Methods that Learn from Human\n Feedback", "Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena"], "answer_arxiv_id": ["2305.14387", "2306.05685"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_1152"} +{"question": "What papers focused on model-based representation learning in low-rank MDPs?", "answer": ["FLAMBE: Structural Complexity and Representation Learning of Low Rank MDPs", "Representation Learning for Online and Offline RL in Low-rank MDPs", "Spectral Decomposition Representation for Reinforcement Learning"], "answer_arxiv_id": ["2006.10814", "2110.04652", "2208.09515"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_1153"} +{"question": "Could you provide me the study that used CSBMs to analyze attention-based GNNs?", "answer": ["Graph Attention Retrospective"], "answer_arxiv_id": ["2202.13060"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_1154"} +{"question": "Could you provide me some studies that proposed sample selection methods to address the noisy correspondence problem?", "answer": ["Noisy Correspondence Learning with Meta Similarity Correction"], "answer_arxiv_id": ["2304.06275"], "source_meta": {"published_time": "20230819"}, "qid": "AutoScholarQuery_train_1155"} +{"question": "What studies are designed to learn higher levels of knowledge from graphs by increasing the number of GCN layers?", "answer": ["Representation Learning on Graphs with Jumping Knowledge Networks"], "answer_arxiv_id": ["1806.03536v2"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_1156"} +{"question": "What works are about prediction of sequences that can fold into given backbone structures?,", "answer": ["Learning from Protein Structure with Geometric Vector Perceptrons"], "answer_arxiv_id": ["2009.01411"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_1157"} +{"question": "Are there any studies exploring free-form interleaved vision-text chatting in early LMMs?", "answer": ["Otter: A Multi-Modal Model with In-Context Instruction Tuning", "Gemini: A Family of Highly Capable Multimodal Models", "DeepSpeed-VisualChat: Multi-Round Multi-Image Interleave Chat via\n Multi-Modal Causal Attention"], "answer_arxiv_id": ["2305.03726", "2312.11805v4", "2309.14327"], "source_meta": {"published_time": "20240617"}, "qid": "AutoScholarQuery_train_1158"} +{"question": "Which studies applied minimax optimization in adversarial robustness?", "answer": ["On Evaluating Adversarial Robustness", "Certified Adversarial Robustness via Randomized Smoothing", "Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuning", "Robust Pre-Training by Adversarial Contrastive Learning"], "answer_arxiv_id": ["1902.06705", "1902.02918", "2003.12862", "2010.13337"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_1159"} +{"question": "What research works successfully applied diffusion-based generative models in text-based image editing?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "DiffEdit: Diffusion-based semantic image editing with mask guidance", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "Prompt-to-Prompt Image Editing with Cross Attention Control", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Zero-shot Image-to-Image Translation", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2208.12242", "2210.11427", "2211.12572", "2208.01626", "2108.01073", "2302.03027", "2211.09800"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_1160"} +{"question": "Which papers studied semantic image synthesis at pixel level that accurately mirrors user intentions?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs", "Diversity-sensitive Conditional Generative Adversarial Networks", "Learning to Predict Layout-to-image Conditional Convolutions for Semantic Image Synthesis", "Dual Attention GANs for Semantic Image Synthesis", "Image Synthesis via Semantic Composition", "Semantic Image Synthesis with Spatially-Adaptive Normalization", "SEAN: Image Synthesis with Semantic Region-Adaptive Normalization", "You Only Need Adversarial Supervision for Semantic Image Synthesis", "Diverse Semantic Image Synthesis via Probability Distribution Modeling"], "answer_arxiv_id": ["1611.07004", "1711.11585", "1901.09024", "1910.06809", "2008.13024", "2109.07053", "1903.07291", "1911.12861", "2012.04781", "2103.06878"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_1161"} +{"question": "Could you provide me a paper about the 𝙶𝚕𝚘𝚋𝚊𝚕𝙳𝚒𝚛𝚎𝚌𝚝𝚒𝚘𝚗𝙶𝚕𝚘𝚋𝚊𝚕𝙳𝚒𝚛𝚎𝚌𝚝𝚒𝚘𝚗 method in StyleCLIP?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery"], "answer_arxiv_id": ["2103.17249"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_1162"} +{"question": "Have there been previous studies on Task Incremental Learning in federated continual learning?", "answer": ["Federated Continual Learning with Weighted Inter-client Transfer"], "answer_arxiv_id": ["2003.03196"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_1163"} +{"question": "Which papers explore the task of LiDAR and camera self-calibration?", "answer": ["CMRNet: Camera to LiDAR-Map Registration", "CalibNet: Geometrically Supervised Extrinsic Calibration using 3D Spatial Transformer Networks", "CFNet: LiDAR-Camera Registration Using Calibration Flow Network"], "answer_arxiv_id": ["1906.10109", "1803.08181", "2104.11907"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_1164"} +{"question": "Is there any work out there like GroundingDino or GLIP wherein REC or PG networks function as open-vocabulary detectors?", "answer": ["Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection", "GLIPv2: Unifying Localization and Vision-Language Understanding", "Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2303.05499", "2206.05836", "2112.03857"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1165"} +{"question": "Which studies follow the two-stage method for HOI (Human Object Interaction) detection?", "answer": ["Learning Human-Object Interactions by Graph Parsing Neural Networks", "No-Frills Human-Object Interaction Detection: Factorization, Layout Encodings, and Training Techniques", "Pose-aware Multi-level Feature Network for Human Object Interaction Detection", "Transferable Interactiveness Knowledge for Human-Object Interaction Detection", "VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions", "Contextual Heterogeneous Graph Network for Human-Object Interaction Detection", "DRG: Dual Relation Graph for Human-Object Interaction Detection", "Visual Compositional Learning for Human-Object Interaction Detection", "HOI Analysis: Integrating and Decomposing Human-Object Interaction", "Detecting Human-Object Interactions with Action Co-occurrence Priors", "Detailed 2D-3D Joint Representation for Human-Object Interaction", "Cascaded Human-Object Interaction Recognition", "Spatially Conditioned Graphs for Detecting Human–Object Interactions", "Detecting and Recognizing Human-Object Interactions"], "answer_arxiv_id": ["1808.07962", "1811.05967", "1909.08453", "1811.08264", "2003.05541", "2010.10001", "2008.11714", "2007.12407", "2010.16219", "2007.08728", "2004.08154", "2003.04262", "2012.06060", "1704.07333"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_1166"} +{"question": "Can you provide some references about learning feasible state-action pairs to distinguish them from infeasible ones?", "answer": ["Learning Constraints from Demonstrations", "Maximum Likelihood Constraint Inference for Inverse Reinforcement Learning", "Maximum Likelihood Constraint Inference from Stochastic Demonstrations"], "answer_arxiv_id": ["1812.07084", "1909.05477", "2102.12554"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_1167"} +{"question": "Can you list out research papers that follows the idea of CLIP-BART by freezing large language models to utilize the in-context learning ability of LLMs?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models", "MAGMA – Multimodal Augmentation of Generative Models through Adapter-based Finetuning", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2106.13884", "2112.05253", "2204.14198"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_1168"} +{"question": "Could you cite a work that is about combining model parameters from the weights of task combinations in ATL?", "answer": ["ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning"], "answer_arxiv_id": ["2212.01378"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1169"} +{"question": "Which papers introduced improvements to NeRF for producing photo-realistic 3D representations?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields", "Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance\n Fields", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "NeuS2: Fast Learning of Neural Implicit Surfaces for Multi-view\n Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "Neuralangelo: High-Fidelity Neural Surface Reconstruction", "F$^{2}$-NeRF: Fast Neural Radiance Field Training with Free Camera\n Trajectories"], "answer_arxiv_id": ["2103.13415", "2111.12077", "2304.06706", "2112.03907", "2106.10689", "2212.05231", "2106.12052", "2306.03092", "2303.15951"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1170"} +{"question": "What papers propose the use of Directed Acyclic Graphs (DAGs) in Graph Neural Networks (GNNs) by implementing message passing across nodes in a sequential manner conforming to the topological order?", "answer": ["D-VAE: A Variational Autoencoder for Directed Acyclic Graphs", "Directed Acyclic Graph Neural Networks"], "answer_arxiv_id": ["1904.11088", "2101.07965"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_1171"} +{"question": "Could you provide me references where DeepONet has been applied to solve various problems?", "answer": ["DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators", "Operator learning for predicting multiscale bubble growth dynamics", "DeepM&Mnet: Inferring the electroconvection multiphysics fields based on operator approximation by neural networks"], "answer_arxiv_id": ["1910.03193", "2012.12816", "2009.12935"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_1172"} +{"question": "Could you provide me some papers about developing invertible mapping structures?", "answer": ["Density estimation using Real NVP", "Masked Autoregressive Flow for Density Estimation"], "answer_arxiv_id": ["1605.08803", "1705.07057"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_1173"} +{"question": "Are there any papers that evaluate the current benchmarks used in multi-task and transferability studies?", "answer": ["TransNAS-Bench-101: Improving Transferability and Generalizability of Cross-Task Neural Architecture Search"], "answer_arxiv_id": ["2105.11871"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1174"} +{"question": "What works follow the 'first-separation-then-encoding' paradigm in graph-level representations under distribution shifts?", "answer": ["Learning Causally Invariant Representations for Out-of-Distribution Generalization on Graphs", "Debiasing Graph Neural Networks via Learning Disentangled Causal Substructure", "Causal Attention for Interpretable and Generalizable Graph Classification"], "answer_arxiv_id": ["2202.05441", "2209.14107", "2112.15089"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_1175"} +{"question": "Could you provide me some works on generating descriptions for videos?", "answer": ["Open-book Video Captioning with Retrieve-Copy-Generate Network", "Spatio-Temporal Graph for Video Captioning with Knowledge Distillation", "Hierarchical Modular Network for Video Captioning", "SwinBERT: End-to-End Transformers with Sparse Attention for Video\n Captioning", "MART: Memory-Augmented Recurrent Transformer for Coherent Video\n Paragraph Captioning"], "answer_arxiv_id": ["2103.05284", "2003.13942", "2111.12476", "2111.13196", "2005.05402"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_1176"} +{"question": "Which studies have used self-normalized concentration for breakthroughs in differential privacy applications?", "answer": ["Brownian Noise Reduction: Maximizing Privacy Subject to Accuracy Constraints", "Fully Adaptive Composition in Differential Privacy"], "answer_arxiv_id": ["2206.07234", "2203.05481"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_1177"} +{"question": "Which works focus on training methods for inducing interpretable mechanisms in neural networks?", "answer": ["Inducing Causal Structure for Interpretable Neural Networks", "Causal Distillation for Language Models", "Causal Proxy Models for Concept-based Model Explanations", "Inducing Character-level Structure in Subword-based Language Models with Type-level Interchange Intervention Training"], "answer_arxiv_id": ["2112.00826v2", "2112.02505", "2209.14279", "2212.09897"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_1178"} +{"question": "Which researchers have explored the injection of backdoors in the Federated Learning model?", "answer": ["Data Poisoning Attacks Against Federated Learning Systems", "How To Backdoor Federated Learning", "Local Model Poisoning Attacks to Byzantine-Robust Federated Learning", "Back to the Drawing Board: A Critical Evaluation of Poisoning Attacks on Production Federated Learning"], "answer_arxiv_id": ["2007.08432", "1807.00459", "1911.11815", "2108.10241"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_1179"} +{"question": "Which works have focused on Absolute Pose Regressors (APRs) and end-to-end learning-based methods for direct pose regression?", "answer": ["PoseNet: A Convolutional Network for Real-Time 6-DOF Camera\n Relocalization", "Geometric Loss Functions for Camera Pose Regression with Deep Learning", "Learning Multi-Scene Absolute Pose Regression with Transformers", "Image-based Localization using Hourglass Networks", "Learning Multi-Scene Absolute Pose Regression with Transformers", "Direct-PoseNet: Absolute Pose Regression with Photometric Consistency", "DFNet: Enhance Absolute Pose Regression with Direct Feature Matching", "Geometry-Aware Learning of Maps for Camera Localization", "CoordiNet: uncertainty-aware pose regressor for reliable vehicle\n localization"], "answer_arxiv_id": ["1505.07427", "1704.00390v2", "2103.11468", "1703.07971", "2103.11468", "2104.04073", "2204.00559", "1712.03342", "2103.10796"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1180"} +{"question": "Which study proposed a training-free sampling method for approximately solving traveling salesman problems?", "answer": ["Diffusion models as plug-and-play priors"], "answer_arxiv_id": ["2206.09012"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_1181"} +{"question": "Which papers focus on the topic of neural architecture search?", "answer": ["A Survey on Neural Architecture Search", "Neural Architecture Search: A Survey", "Neural Architecture Search with Reinforcement Learning"], "answer_arxiv_id": ["1905.01392", "1808.05377", "1611.01578"], "source_meta": {"published_time": "20230113"}, "qid": "AutoScholarQuery_train_1182"} +{"question": "Could you provide me some works that developed benchmarks for document review or case summarization?", "answer": ["CUAD: An Expert-Annotated NLP Dataset for Legal Contract Review", "Multi-LexSum: Real-World Summaries of Civil Rights Lawsuits at Multiple Granularities", "Legal Case Document Summarization: Extractive and Abstractive Methods and their Evaluation"], "answer_arxiv_id": ["2103.06268", "2206.10883", "2210.07544"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_1183"} +{"question": "Could you propose some studies about the application of MetaRL?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning", "Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables", "Decoupling Exploration and Exploitation for Meta-Reinforcement Learning without Sacrifices"], "answer_arxiv_id": ["1703.03400", "1910.08348", "1903.08254", "2008.02790"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_1184"} +{"question": "Which study aligns a frozen visual encoder with a frozen LLM?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2304.10592"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_1185"} +{"question": "What studies are about agents interacting with a web-grounded environment?", "answer": ["Mind2Web: Towards a Generalist Agent for the Web", "WebShop: Towards Scalable Real-World Web Interaction with Grounded\n Language Agents"], "answer_arxiv_id": ["2306.06070", "2207.01206"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_1186"} +{"question": "Could you provide me some studies about methods for identifying critical weights in network such as magnitude-based criterion, geometry median, Taylor expansions and BN-based?", "answer": ["Filter Pruning via Geometric Median for Deep Convolutional Neural\n Networks Acceleration", "Importance Estimation for Neural Network Pruning", "Learning Efficient Convolutional Networks through Network Slimming"], "answer_arxiv_id": ["1811.00250", "1906.10771", "1708.06519"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_1187"} +{"question": "What works are there on medical AQA under the Objective Structured Assessment of Technical Skill (OSATS) system?", "answer": ["Video and Accelerometer-Based Motion Analysis for Automated Surgical\n Skills Assessment"], "answer_arxiv_id": ["1702.07772"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_1188"} +{"question": "Which works were involved with the development of the reverse denoising process in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_1189"} +{"question": "What papers proposed methods for out of distribution detection that use density estimation to pose the identification of covariate shift as anomaly detection?", "answer": ["Likelihood Ratios for Out-of-Distribution Detection"], "answer_arxiv_id": ["1906.02845"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_1190"} +{"question": "Which works utilized explicit labels in TTS models to generate diverse speech that matches the prompt?", "answer": ["Speech Resynthesis from Discrete Disentangled Self-Supervised\n Representations", "EXPRESSO: A Benchmark and Analysis of Discrete Expressive Speech\n Resynthesis"], "answer_arxiv_id": ["2104.00355", "2308.05725"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_1191"} +{"question": "What are some studies about application of Graph Transformers in various graph-level tasks like molecular property prediction, image classification, and human interaction recognition?", "answer": ["A Generalization of Transformer Networks to Graphs", "Rethinking Graph Transformers with Spectral Attention", "Benchmarking Graphormer on Large-Scale Molecular Modeling Datasets", "Global Self-Attention as a Replacement for Graph Convolution", "Recipe for a General, Powerful, Scalable Graph Transformer", "Self-Supervised Graph Transformer on Large-Scale Molecular Data", "A graph-transformer for whole slide image classification", "IGFormer: Interaction Graph Transformer for Skeleton-based Human\n Interaction Recognition"], "answer_arxiv_id": ["2012.09699", "2106.03893", "2203.04810", "2108.03348", "2205.12454", "2007.02835", "2205.09671", "2207.12100"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_1192"} +{"question": "Which research empirically verified the robustness of Vit to patch shuffling compared to convolutional networks?", "answer": ["Intriguing Properties of Vision Transformers"], "answer_arxiv_id": ["2105.10497"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_1193"} +{"question": "What studies deal with multilingual generation benchmarks with focus on cross-lingual summarization?", "answer": ["MLSUM: The Multilingual Summarization Corpus", "WikiLingua: A New Benchmark Dataset for Cross-Lingual Abstractive\n Summarization"], "answer_arxiv_id": ["2004.14900", "2010.03093"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_1194"} +{"question": "What are the early works that studied about different 3D representations in the 3D Generation field?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "3D Shape Induction from 2D Views of Multiple Objects", "Improved Adversarial Systems for 3D Object Generation and Reconstruction", "Escaping Plato's Cave: 3D Shape From Adversarial Rendering", "Inverse Graphics GAN: Learning to Generate 3D Shapes from Unstructured\n 2D Data", "Learning Representations and Generative Models for 3D Point Clouds", "StructureNet: Hierarchical Graph Networks for 3D Shape Generation", "PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "Image GANs meet Differentiable Rendering for Inverse Graphics and\n Interpretable 3D Neural Rendering", "Deep Marching Tetrahedra: a Hybrid Representation for High-Resolution 3D\n Shape Synthesis", "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images", "Learning Implicit Fields for Generative Shape Modeling", "Occupancy Networks: Learning 3D Reconstruction in Function Space"], "answer_arxiv_id": ["1610.07584", "1612.05872", "1707.09557", "1811.11606", "2002.12674", "1707.02392", "1908.00575", "1906.12320", "2010.09125", "2111.04276", "2209.11163", "1812.02822", "1812.03828"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_1195"} +{"question": "Which works implement a simple framework with quality-preserving augmentations to learn representational data that works well with synthetic and realistic distortions in image/video quality assessment?", "answer": ["Image Quality Assessment using Contrastive Learning"], "answer_arxiv_id": ["2110.13266"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_1196"} +{"question": "Which paper proposed RevNet which allow back propagation without saving intermediate activations?", "answer": ["The Reversible Residual Network: Backpropagation Without Storing Activations"], "answer_arxiv_id": ["1707.04585"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_1197"} +{"question": "What works use language model to automatically creating the knowledge bases?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_1198"} +{"question": "Which research papers demonstrate the utilization of ConvRNNs in various domains like biomedical, robotics, traffic modelling and weather forecasting?", "answer": ["Parallel Multi-Dimensional LSTM, With Application to Fast Biomedical Volumetric Image Segmentation", "Unsupervised Learning for Physical Interaction through Video Prediction", "Non-Local ConvLSTM for Video Compression Artifact Reduction", "FACLSTM: ConvLSTM with Focused Attention for Scene Text Recognition", "Removing Rain in Videos: A Large-scale Database and A Two-stream ConvLSTM Approach"], "answer_arxiv_id": ["1506.07452", "1605.07157", "1910.12286", "1904.09405", "1906.02526"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_1199"} +{"question": "Any works about the application of knowledge-enhanced methods in fields other than NLP?", "answer": ["Re-Imagen: Retrieval-Augmented Text-to-Image Generator", "Retrieval-Augmented Reinforcement Learning"], "answer_arxiv_id": ["2209.14491", "2202.08417"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_1200"} +{"question": "Could you give me some studies about Top Two algorithms?", "answer": ["Top Two Algorithms Revisited", "Information-Directed Selection for Top-Two Algorithms"], "answer_arxiv_id": ["2206.05979", "2205.12086"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_1201"} +{"question": "Which work provides the first algorithm for general list-decodable covariance estimation?", "answer": ["List-decodable Covariance Estimation"], "answer_arxiv_id": ["2206.10942"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_1202"} +{"question": "What multi-task datasets are relatively small and have only a few tasks?", "answer": ["Microsoft COCO: Common Objects in Context", "The Cityscapes Dataset for Semantic Urban Scene Understanding"], "answer_arxiv_id": ["1405.0312", "1604.01685v2"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_1203"} +{"question": "What works have been done on understanding epistemic uncertainty at a scale?", "answer": ["Aleatoric and Epistemic Uncertainty in Machine Learning: An Introduction\n to Concepts and Methods"], "answer_arxiv_id": ["1910.09457"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_1204"} +{"question": "Among studies about pruning MRF label space, which ones predicted a unimodal search range which is susceptible to local optimum?", "answer": ["DeepPruner: Learning Efficient Stereo Matching via Differentiable\n PatchMatch", "CFNet: Cascade and Fused Cost Volume for Robust Stereo Matching", "Deep Stereo using Adaptive Thin Volume Representation with Uncertainty\n Awareness"], "answer_arxiv_id": ["1909.05845", "2104.04314", "1911.12012"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_1205"} +{"question": "What research deals with randomized smoothing technique in adversarial defense?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing"], "answer_arxiv_id": ["1902.02918"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_1206"} +{"question": "What studies have developed OOD methods that can be implemented on the Rotated Faster R-CNN framework?", "answer": ["Gliding vertex on the horizontal bounding box for multi-oriented object\n detection", "ReDet: A Rotation-equivariant Detector for Aerial Object Detection", "Oriented R-CNN for Object Detection"], "answer_arxiv_id": ["1911.09358", "2103.07733", "2108.05699"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_1207"} +{"question": "Which papers report that large language models don't naturally follow human intents well from pre-training?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_1208"} +{"question": "Which studies use a two-time scale update rule (TTUR) to push GANs’ training process towards converging to a local Nash equilibrium?", "answer": ["GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium"], "answer_arxiv_id": ["1706.08500"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_1209"} +{"question": "Could you give me examples of research that proposed techniques for imputing missing labels using feedback from human experts?", "answer": ["Learning under selective labels in the presence of expert consistency"], "answer_arxiv_id": ["1807.00905"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_1210"} +{"question": "What research first proposed the concept of disentangled variational auto-encoders?", "answer": ["Bayesian Representation Learning with Oracle Constraints"], "answer_arxiv_id": ["1506.05011"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_1211"} +{"question": "Which studies have focused on key challenges raised by Federated Learning such as data/system heterogeneity?", "answer": ["FedCor: Correlation-Based Active Client Selection Strategy for\n Heterogeneous Federated Learning"], "answer_arxiv_id": ["2103.13822"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_1212"} +{"question": "Which works proposed the Federated learning (FL) technique?", "answer": ["Federated Machine Learning: Concept and Applications", "Combating Data Imbalances in Federated Semi-supervised Learning with\n Dual Regulators", "Communication-Efficient Learning of Deep Networks from Decentralized\n Data"], "answer_arxiv_id": ["1902.04885", "2307.05358", "1602.05629"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1213"} +{"question": "Can you mention some papers addressing diacritization within a multi-task setup, jointly modeling diacritization with relevant tasks such as POS tagging?", "answer": ["A Multitask Learning Approach for Diacritic Restoration"], "answer_arxiv_id": ["2006.04016"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_train_1214"} +{"question": "Which studies first researched rematerialization in the context of compiler optimization and automatic differentiation with checkpointing?", "answer": ["Survey on Combinatorial Register Allocation and Instruction Scheduling", "Combinatorial Register Allocation and Instruction Scheduling"], "answer_arxiv_id": ["1409.7628", "1804.02452v5"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_1215"} +{"question": "What efforts were made in early text-to-3D techniques?", "answer": ["Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and Text-to-Image Diffusion Models", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "CLIP-Forge: Towards Zero-Shot Text-to-Shape Generation", "CLIP-Mesh: Generating textured meshes from text using pretrained image-text models", "Zero-Shot Text-Guided Object Generation with Dream Fields", "TANGO: Text-driven Photorealistic and Robust 3D Stylization via Lighting Decomposition"], "answer_arxiv_id": ["2212.14704", "2205.08535", "2112.03221", "2110.02624", "2203.13333", "2112.01455", "2210.11277"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_1216"} +{"question": "In what paper the researcher considered case when an extra observation is available after taking the actions?", "answer": ["The Statistical Complexity of Interactive Decision Making"], "answer_arxiv_id": ["2112.13487v3"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_1217"} +{"question": "What works have developed methods to refine estimated 3D human meshes using 2D keypoints?", "answer": ["Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a\n Single Image", "Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the\n Loop", "On the Calibration of Human Pose Estimation"], "answer_arxiv_id": ["1607.08128", "1909.12828", "2311.17105"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_1218"} +{"question": "What research papers deal with In-Context Learning?", "answer": ["Language Models are Few-Shot Learners", "MetaICL: Learning to Learn In Context", "Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science\n Question Answering"], "answer_arxiv_id": ["2005.14165", "2110.15943", "2202.12837", "2107.13586v1", "2209.09513"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1219"} +{"question": "Could you provide the datasets that are used for title generation in video datasets?", "answer": ["Title Generation for User Generated Videos"], "answer_arxiv_id": ["1608.07068"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_1220"} +{"question": "Which research works refined LoRA weight-space using Singular Value Decomposition (SVD) and orthogonal incomplete basis?", "answer": ["SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2303.11305", "2307.06949"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1221"} +{"question": "Could you provide me some studies about the results of scaling the size of vision transformers?", "answer": ["Scaling Vision Transformers"], "answer_arxiv_id": ["2106.04560"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_1222"} +{"question": "Which paper suggests a model that leverages screenshot and image pairs with a screen parsing pre-training task that converts webpage screenshots to HTML text", "answer": ["Pix2Struct: Screenshot Parsing as Pretraining for Visual Language\n Understanding"], "answer_arxiv_id": ["2210.03347"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_1223"} +{"question": "Could you provide me some studies on camouflaging objects in a scene through re-texturing?", "answer": ["GANmouflage: 3D Object Nondetection with Texture Fields"], "answer_arxiv_id": ["2201.07202"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1224"} +{"question": "What papers propose methods to attain multi-view consistency in images?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation", "Instant3D: Fast Text-to-3D with Sparse-View Generation and Large\n Reconstruction Model", "EfficientDreamer: High-Fidelity and Robust 3D Creation via\n Orthogonal-view Diffusion Prior", "Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2308.16512", "2311.06214", "2308.13223", "2303.11328"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_1225"} +{"question": "Could you provide me some studies about convergence to global Nash equilibria in minmax optimization where the objective satisfies a two-sided PL condition?", "answer": ["Faster Single-loop Algorithms for Minimax Optimization without Strong Concavity", "Convergence Rates of Two-Time-Scale Gradient Descent-Ascent Dynamics for Solving Nonconvex Min-Max Problems"], "answer_arxiv_id": ["2112.05604", "2112.09579"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_1226"} +{"question": "Which studies have utilized image-text contrastive learning models for 2D open-vocabulary segmentation?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm", "SLIP: Self-supervision meets Language-Image Pre-training", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2102.05918", "2110.05208", "2112.12750", "2103.00020"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_1227"} +{"question": "Can you cite any studies about scene graphs for vision understanding?", "answer": ["Scene Graph Generation by Iterative Message Passing", "Neural Motifs: Scene Graph Parsing with Global Context", "Bridging Knowledge Graphs to Generate Scene Graphs"], "answer_arxiv_id": ["1701.02426", "1711.06640", "2001.02314"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_1228"} +{"question": "What datasets in sRGB domain are available for moiré removal?", "answer": ["Moiré Photo Restoration Using Multiresolution Convolutional Neural Networks"], "answer_arxiv_id": ["1805.02996"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_1229"} +{"question": "What studies design risk-consistent methods in the literature of learning with label noise?", "answer": ["Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach", "Classification with Noisy Labels by Importance Reweighting"], "answer_arxiv_id": ["1609.03683", "1411.7718"], "source_meta": {"published_time": "20220204"}, "qid": "AutoScholarQuery_train_1230"} +{"question": "What are some of the studies that have produced histopathology vision-language datasets?", "answer": ["Multiple Instance Captioning: Learning Representations from Histopathology Textbooks and Articles", "PathVQA: 30000+ Questions for Medical Visual Question Answering"], "answer_arxiv_id": ["2103.05121", "2003.10286"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_1231"} +{"question": "Which paper proposed the usage of a multi-resolution attention module in FS-PCS?", "answer": ["Few-Shot Point Cloud Semantic Segmentation via Contrastive\n Self-Supervision and Multi-Resolution Attention"], "answer_arxiv_id": ["2302.10501"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_1232"} +{"question": "What studies contributed to JAX-based environments providing continuous control tasks?", "answer": ["Brax - A Differentiable Physics Engine for Large Scale Rigid Body Simulation"], "answer_arxiv_id": ["2106.13281"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_1233"} +{"question": "Which methods were presented in prior literature for privacy-preserving action recognition?", "answer": ["Learning to Anonymize Faces for Privacy Preserving Action Detection", "Learning Privacy Preserving Encodings through Adversarial Training", "Privacy-Preserving Deep Action Recognition: An Adversarial Learning Framework and A New Dataset", "SPAct: Self-supervised Privacy Preservation for Action Recognition"], "answer_arxiv_id": ["1803.11556", "1802.05214", "1906.05675", "2203.15205"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_1234"} +{"question": "What papers proposed to manipulate the cross-attention map for text-driven editing on images?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing"], "answer_arxiv_id": ["2208.01626", "2304.08465v1"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_1235"} +{"question": "Which studies incorporated substructure counting to the initial node features to enhance the expressiveness of GNNs?", "answer": ["Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting", "Graph Neural Networks with Local Graph Parameters"], "answer_arxiv_id": ["2006.09252", "2106.06707"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_1236"} +{"question": "What papers focus on learning more object-centric representations in egocentric computer vision?", "answer": ["Transformed ROIs for Capturing Visual Transformations in Videos", "Object-Region Video Transformers", "Hand-Object Interaction Reasoning"], "answer_arxiv_id": ["2106.03162", "2110.06915", "2201.04906"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_1237"} +{"question": "What papers study the concept of a routing mechanism in sparse mixture of experts?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated\n Mixture-of-Experts Layer"], "answer_arxiv_id": ["1701.06538"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_1238"} +{"question": "Where can I find works on memory-based methods for continual learning?", "answer": ["Continual Learning with Deep Generative Replay", "iCaRL: Incremental Classifier and Representation Learning", "Gradient Episodic Memory for Continual Learning", "On Tiny Episodic Memories in Continual Learning", "Gradient based sample selection for online continual learning", "Dark Experience for General Continual Learning: a Strong, Simple Baseline", "Rehearsal revealed: The limits and merits of revisiting samples in continual learning", "Rainbow Memory: Continual Learning with a Memory of Diverse Samples", "A Multi-Head Model for Continual Learning via Out-of-Distribution Replay", "A simple but strong baseline for online continual learning: Repeated Augmented Rehearsal", "Memory Replay with Data Compression for Continual Learning", "Saliency Guided Experience Packing for Replay in Continual Learning"], "answer_arxiv_id": ["1705.08690", "1611.07725", "1706.08840", "1902.10486", "1903.08671", "2004.07211", "2104.07446", "2103.17230", "2208.09734", "2209.13917v2", "2202.06592", "2109.04954"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_1239"} +{"question": "Could you tell me some studies on data selection in active learning focusing on the importance of diversity or coverage?", "answer": ["Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds", "Batch Active Learning at Scale", "Active Bias: Training More Accurate Neural Networks by Emphasizing High Variance Samples"], "answer_arxiv_id": ["1906.03671", "2107.14263", "1704.07433"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_1240"} +{"question": "What works in fprintfocus on adaptive training from the original distribution with modified architecture designs?", "answer": ["Training Convolutional Networks with Noisy Labels", "Graph Convolutional Label Noise Cleaner: Train a Plug-and-play Action\n Classifier for Anomaly Detection"], "answer_arxiv_id": ["1406.2080", "1903.07256"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_1241"} +{"question": "What works have developed pre-trained multi-modal models that are robust to spurious correlations?", "answer": ["Contrastive Adapters for Foundation Model Group Robustness", "Mitigating Spurious Correlations in Multi-modal Models during\n Fine-tuning"], "answer_arxiv_id": ["2207.07180", "2304.03916"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_1242"} +{"question": "Which works focus on the integration of low and high-level features in an unsupervised manner for GIQA?", "answer": ["Re-IQA: Unsupervised Learning for Image Quality Assessment in the Wild"], "answer_arxiv_id": ["2304.00451"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_1243"} +{"question": "What works deal with personalization in federated learning via a hypernetwork?", "answer": ["Personalized Federated Learning using Hypernetworks"], "answer_arxiv_id": ["2103.04628"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_1244"} +{"question": "Could you provide me some works where the IB principle was used to study the information processing capacity of the brain?", "answer": ["A mathematical theory of semantic development in deep neural networks", "Opening the black box of Deep Neural Networks via Information"], "answer_arxiv_id": ["1810.10531", "1703.00810"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_1245"} +{"question": "Can you provide me some research that are about panoramic localization?", "answer": ["LDL: Line Distance Functions for Panoramic Localization"], "answer_arxiv_id": ["2308.13989"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_1246"} +{"question": "Which studies have shown a tight last-iterate convergence for OGDA?", "answer": ["Tight Last-Iterate Convergence of the Extragradient and the Optimistic Gradient Descent-Ascent Algorithm for Constrained Monotone Variational Inequalities"], "answer_arxiv_id": ["2204.09228"], "source_meta": {"published_time": "20220619"}, "qid": "AutoScholarQuery_train_1247"} +{"question": "Which studies proposed the unsupervised models trained in a data-free regime?", "answer": ["DGM: A deep learning algorithm for solving partial differential equations", "Learning Incompressible Fluid Dynamics from Scratch - Towards Fast, Differentiable Fluid Models that Generalize"], "answer_arxiv_id": ["1708.07469", "2006.08762"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_1248"} +{"question": "Which research works have implemented distributed versions of the PPO algorithm?", "answer": ["Emergence of Locomotion Behaviours in Rich Environments", "DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion Frames"], "answer_arxiv_id": ["1707.02286", "1911.00357"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_1249"} +{"question": "Could you provide me some papers that enrich the training degradation space through hand-crafted synthesisation for blind image super-resolution?", "answer": ["Designing a Practical Degradation Model for Deep Blind Image\n Super-Resolution", "Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure\n Synthetic Data", "Denoising Diffusion Probabilistic Models for Robust Image\n Super-Resolution in the Wild"], "answer_arxiv_id": ["2103.14006", "2107.10833", "2302.07864"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_1250"} +{"question": "What studies are related to transferring policies using goal-conditioned policy?", "answer": ["Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning", "Distral: Robust Multitask Reinforcement Learning"], "answer_arxiv_id": ["1706.05064", "1707.04175"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_1251"} +{"question": "What study improved the performance by replacing the clustering with farthest-point sampling?", "answer": ["PointInst3D: Segmenting 3D Instances by Points"], "answer_arxiv_id": ["2204.11402"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_1252"} +{"question": "Could you list the studies that worked on state-of-the-art generative models such as GANs?", "answer": ["Analyzing and Improving the Image Quality of StyleGAN", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Alias-Free Generative Adversarial Networks", "Taming Transformers for High-Resolution Image Synthesis"], "answer_arxiv_id": ["1912.04958", "1812.04948", "2106.12423", "2012.09841"], "source_meta": {"published_time": "20210503"}, "qid": "AutoScholarQuery_train_1253"} +{"question": "Could you point me to studies that have considered the convergence of gradient decent using the Neural Tangent Hierarchy?", "answer": ["Dynamics of Deep Neural Networks and Neural Tangent Hierarchy"], "answer_arxiv_id": ["1909.08156"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_1254"} +{"question": "Could you provide me with the paper that introduced Rev-ViT, an extension of the reversible CNN block to the reversible Transformer block?", "answer": ["Reversible Vision Transformers"], "answer_arxiv_id": ["2302.04869"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_1255"} +{"question": "Which works propose batch constrained offline RL methods?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration"], "answer_arxiv_id": ["1812.02900"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_1256"} +{"question": "Could you give me some examples of papers that focus on the novel-view acoustic synthesis task?", "answer": ["Novel-View Acoustic Synthesis", "AV-NeRF: Learning Neural Fields for Real-World Audio-Visual Scene\n Synthesis", "Novel-View Acoustic Synthesis from 3D Reconstructed Rooms", "Sound Localization from Motion: Jointly Learning Sound Direction and\n Camera Rotation"], "answer_arxiv_id": ["2301.08730", "2302.02088", "2310.15130", "2303.11329"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_1257"} +{"question": "What research proposed to capture and analyze the interaction of visual concepts contributing cooperatively to prediction outcomes in CNNs?", "answer": ["Discovering and Explaining the Representation Bottleneck of DNNs"], "answer_arxiv_id": ["2111.06236"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_1258"} +{"question": "What research paper shows the use of user-provided masks to localize text-based editing?", "answer": ["Paint by Word"], "answer_arxiv_id": ["2103.10951"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_1259"} +{"question": "Can you list the studies that showed the ESBN and CoRelNet architectures depended on the presence of pre-segmented objects as input?", "answer": ["Emergent Symbols through Binding in External Memory", "On Neural Architecture Inductive Biases for Relational Tasks"], "answer_arxiv_id": ["2012.14601", "2206.05056"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_1260"} +{"question": "What works discuss federated learning, continual learning, and their intersection?", "answer": ["Federated and continual learning for classification tasks in a society of devices", "Federated Continual Learning with Weighted Inter-client Transfer", "A distillation-based approach integrating continual learning and federated learning for pervasive services", "Federated Reconnaissance: Efficient, Distributed, Class-Incremental Learning", "Federated Class-Incremental Learning"], "answer_arxiv_id": ["2006.07129", "2003.03196", "2109.04197", "2109.00150", "2203.11473"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_1261"} +{"question": "What are some examples of models that pair a vision encoder with multilayer perceptron (MLP) for visual information processing in MLLMs?", "answer": ["Visual Instruction Tuning", "Improved Baselines with Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485", "2310.03744"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_1262"} +{"question": "Could you provide me some studies where contrastive learning has been used for self-supervised learning tasks?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Contrastive Multiview Coding", "CLEAR: Contrastive Learning for Sentence Representation", "SimCSE: Simple Contrastive Learning of Sentence Embeddings"], "answer_arxiv_id": ["2002.05709", "1911.05722", "1906.05849", "2012.15466", "2104.08821"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_1263"} +{"question": "Are there any works about the digital twins of articulated objects?", "answer": ["A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representation", "Ditto: Building Digital Twins of Articulated Objects from Interaction", "CARTO: Category and Joint Agnostic Reconstruction of ARTiculated Objects", "Ditto in the House: Building Articulation Models of Indoor Scenes through Interactive Perception"], "answer_arxiv_id": ["2104.07645", "2202.08227", "2303.15782", "2302.01295"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_1264"} +{"question": "What studies use continuous normalizing flows with the regularization in the entropy regularized unbalanced optimal transport loss?", "answer": ["TrajectoryNet: A Dynamic Optimal Transport Network for Modeling Cellular Dynamics", "Manifold Interpolating Optimal-Transport Flows for Trajectory Inference"], "answer_arxiv_id": ["2002.04461", "2206.14928"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_1265"} +{"question": "What are some examples of prior studies that constructed large-scale image-text datasets sourced from the web?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Learning Transferable Visual Models From Natural Language Supervision", "WIT: Wikipedia-based Image Text Dataset for Multimodal Multilingual Machine Learning"], "answer_arxiv_id": ["2112.10752", "2103.00020", "2103.01913"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_1266"} +{"question": "Could you provide me some studies about linear Transformers for finding the approximation of self-attention?", "answer": ["Flowformer: Linearizing Transformers with Conservation Flows", "Fast Transformers with Clustered Attention", "Rethinking Attention with Performers", "cosFormer: Rethinking Softmax in Attention"], "answer_arxiv_id": ["2202.06258", "2007.04825", "2009.14794", "2202.08791"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_1267"} +{"question": "What previous studies dealt with feature skew data heterogeneity in federated learning?", "answer": ["FedBN: Federated Learning on Non-IID Features via Local Batch Normalization", "Personalized Federated Learning with Adaptive Batchnorm for Healthcare"], "answer_arxiv_id": ["2102.07623", "2112.00734"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_1268"} +{"question": "Which methods use lightweight geometric features for pose estimation in visual SLAM?", "answer": ["ORB-SLAM: a Versatile and Accurate Monocular SLAM System", "ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D\n Cameras", "ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial\n and Multi-Map SLAM"], "answer_arxiv_id": ["1502.00956", "1610.06475", "2007.11898"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_1269"} +{"question": "Which non-deep learning-based NiSID methods utilize prior hypotheses and statistical laws?", "answer": ["Nighttime Dehazing with a Synthetic Benchmark"], "answer_arxiv_id": ["2008.03864"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_1270"} +{"question": "Which works introduced the Self-attention architecture for tabular data?", "answer": ["TabNet: Attentive Interpretable Tabular Learning", "SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training"], "answer_arxiv_id": ["1908.07442", "2106.01342"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_1271"} +{"question": "Could you provide me some works that involve instance-level tasks such as referring expression comprehension and part detection?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "Open-Vocabulary DETR with Conditional Matching", "Grounded Language-Image Pre-training", "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection"], "answer_arxiv_id": ["2104.13921", "2203.11876", "2112.03857", "2303.05499"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_1272"} +{"question": "Which papers on deep learning primitives optimization discuss reducing data movement through the technique of checkpointing?", "answer": ["A Graph Theoretic Framework of Recomputation Algorithms for Memory-Efficient Backpropagation"], "answer_arxiv_id": ["1905.11722"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_1273"} +{"question": "What work has proven the non-asymptotic convergence to the constrained Nash equilibrium by adding built-in exploration mechanisms?", "answer": ["Provably Efficient Generalized Lagrangian Policy Optimization for Safe Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["2306.00212"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_1274"} +{"question": "Are there any studies on evolution-evaluation approaches in Automated Feature Transformation?", "answer": ["Group-wise Reinforcement Feature Generation for Optimal and Explainable Representation Space Reconstruction", "Feature Engineering for Predictive Modeling using Reinforcement Learning", "Traceable Automatic Feature Transformation via Cascading Actor-Critic Agents", "Traceable Group-Wise Self-Optimizing Feature Transformation Learning: A Dual Optimization Perspective"], "answer_arxiv_id": ["2205.14526", "1709.07150", "2212.13402", "2306.16893"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_1275"} +{"question": "Which works are about addressing the limitations of PIPs in point tracking?", "answer": ["MFT: Long-Term Tracking of Every Pixel", "TAP-Vid: A Benchmark for Tracking Any Point in a Video", "TAPIR: Tracking Any Point with per-frame Initialization and temporal\n Refinement", "Tracking Everything Everywhere All at Once", "PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point\n Tracking"], "answer_arxiv_id": ["2305.12998", "2211.03726", "2306.08637", "2306.05422", "2307.15055"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_1276"} +{"question": "Which works proposed methods dealing with the challenges of long planning horizons in RL?", "answer": ["Is Long Horizon Reinforcement Learning More Difficult Than Short Horizon Reinforcement Learning?", "Is Reinforcement Learning More Difficult Than Bandits? A Near-optimal Algorithm Escaping the Curse of Horizon", "Nearly Horizon-Free Offline Reinforcement Learning", "Stochastic Shortest Path: Minimax, Parameter-Free and Towards Horizon-Free Regret", "Settling the Horizon-Dependence of Sample Complexity in Reinforcement Learning", "Horizon-Free Reinforcement Learning in Polynomial Time: the Power of Stationary Policies", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs"], "answer_arxiv_id": ["2005.00527", "2009.13503", "2103.14077", "2104.11186", "2111.00633", "2203.12922v2", "2205.11507"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_1277"} +{"question": "Which papers propose the use of state abstraction in model-based RL tasks?", "answer": ["Exploration by Random Network Distillation"], "answer_arxiv_id": ["1810.12894"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_1278"} +{"question": "What work created AutoTinyBERT, a pre-trained model created using one-shot NAS?", "answer": ["AutoTinyBERT: Automatic Hyper-parameter Optimization for Efficient Pre-trained Language Models"], "answer_arxiv_id": ["2107.13686"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_1279"} +{"question": "Which works develop rotation equivariant descriptors using group equivariant learning?", "answer": ["You Only Hypothesize Once: Point Cloud Registration with\n Rotation-equivariant Descriptors"], "answer_arxiv_id": ["2109.00182"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_1280"} +{"question": "Which studies demonstrate that incorporating symmetries of the data in machine learning models can improve their data-efficiency and ability to generalize?", "answer": ["Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1602.07576"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_1281"} +{"question": "What research suggested the use of a convolutional neural network (CNN) to reconstruct turbulence-corrupted video sequence?", "answer": ["Atmospheric turbulence removal using convolutional neural network"], "answer_arxiv_id": ["1912.11350"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_1282"} +{"question": "Can you provide references that have tried to make inference using IPs?", "answer": ["Functional Variational Bayesian Neural Networks"], "answer_arxiv_id": ["1903.05779"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_1283"} +{"question": "Could you provide works on the statistical analysis and application of the Sinkhorn Divergence?", "answer": ["Limit Theorems for Entropic Optimal Transport Maps and the Sinkhorn Divergence", "Weak limits of entropy regularized Optimal Transport; potentials, plans and divergences", "An improved central limit theorem and fast convergence rates for entropic transportation costs"], "answer_arxiv_id": ["2207.08683", "2207.07427", "2204.09105"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_1284"} +{"question": "Could you provide me a study that suggests SGD with label noise has an inherent regularizer penalizing sharp minimizers?", "answer": ["Label Noise SGD Provably Prefers Flat Global Minimizers"], "answer_arxiv_id": ["2106.06530"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_1285"} +{"question": "What research proposed a complex clipped SGD algorithm with momentum under the same noise assumption?", "answer": ["High-probability bounds for Non-Convex Stochastic Optimization with Heavy Tails"], "answer_arxiv_id": ["2106.14343"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_1286"} +{"question": "What study considers relative positional embedding with conditional positive definite kernel?", "answer": ["KERPLE: Kernelized Relative Positional Embedding for Length Extrapolation"], "answer_arxiv_id": ["2205.09921"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_1287"} +{"question": "Which studies have evolved MLLMs to handle more complex Visual Language (VL) tasks?", "answer": ["G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment"], "answer_arxiv_id": ["2303.16634"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_1288"} +{"question": "Which studies address safety-constrained reinforcement learning using the CMDP framework?", "answer": ["Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learning"], "answer_arxiv_id": ["2108.06266"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_1289"} +{"question": "Which papers construct the foundation of Grounded SAM?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection"], "answer_arxiv_id": ["2104.14294", "1810.04805", "2303.05499"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_1290"} +{"question": "What are the studies that provide supporting evidence through the use of rationales?", "answer": ["ERASER: A Benchmark to Evaluate Rationalized NLP Models", "Unification-based Reconstruction of Multi-hop Explanations for Science\n Questions"], "answer_arxiv_id": ["1911.03429", "2004.00061"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_1291"} +{"question": "Which paper investigated the consequences for adversarial examples of the implicit bias to margin maximisation in parameter space?", "answer": ["Adversarial Reprogramming Revisited"], "answer_arxiv_id": ["2206.03466"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_1292"} +{"question": "Can you mention some recent diffusion methods that focus on using more conditions from a reference image in text-to-image generation?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing"], "answer_arxiv_id": ["2302.05543", "2308.06721", "2305.14720"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1293"} +{"question": "What works proposed the model distillation without reliance on the training corpus?", "answer": ["Prompting to Distill: Boosting Data-Free Knowledge Distillation via Reinforced Prompt", "Adversarial Self-Supervised Data-Free Distillation for Text Classification", "Towards Zero-Shot Knowledge Distillation for Natural Language Processing"], "answer_arxiv_id": ["2205.07523", "2010.04883", "2012.15495"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_1294"} +{"question": "Can you provide examples of recent studies that achieved impressive results on object-centric data using diffusion models?", "answer": ["3D-aware Image Generation using 2D Diffusion Models", "Novel View Synthesis with Diffusion Models", "ZeroNVS: Zero-Shot 360-Degree View Synthesis from a Single Image", "Zero-1-to-3: Zero-shot One Image to 3D Object", "SparseFusion: Distilling View-conditioned Diffusion for 3D\n Reconstruction", "RenderDiffusion: Image Diffusion for 3D Reconstruction, Inpainting and\n Generation", "DiffRF: Rendering-Guided 3D Radiance Field Diffusion"], "answer_arxiv_id": ["2303.17905", "2210.04628", "2310.17994", "2303.11328", "2212.00792", "2211.09869", "2212.01206"], "source_meta": {"published_time": "20240626"}, "qid": "AutoScholarQuery_train_1295"} +{"question": "What are some studies on end-to-end trained models that directly map input to actions in visual navigation?", "answer": ["Reinforcement Learning with Unsupervised Auxiliary Tasks", "Learning to Navigate in Complex Environments"], "answer_arxiv_id": ["1611.05397", "1611.03673"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_1296"} +{"question": "Can you cite a publication that discusses the use of risk-informed statistics and epistemic risk measures to address value overestimation in RL?", "answer": ["Epistemic Risk-Sensitive Reinforcement Learning"], "answer_arxiv_id": ["1906.06273"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_1297"} +{"question": "What research has been done in the area of 3D human reconstruction using Neural Radiance Fields (NeRFs)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "H-NeRF: Neural Radiance Fields for Rendering and Temporal Reconstruction\n of Humans in Motion", "NeuMan: Neural Human Radiance Field from a Single Video", "PoseVocab: Learning Joint-structured Pose Embeddings for Human Avatar\n Modeling", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "HumanRF: High-Fidelity Neural Radiance Fields for Humans in Motion", "SHERF: Generalizable Human NeRF from a Single Image", "One-shot Implicit Animatable Avatars with Model-based Priors"], "answer_arxiv_id": ["2003.08934", "2110.13746", "2203.12575", "2304.13006", "2201.04127", "2305.06356", "2303.12791", "2212.02469"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_1298"} +{"question": "What works suggest that pixel or feature difference loss functions like L1 or L2 norm and Cosine distance are effective in image transformation tasks?", "answer": ["The Perception-Distortion Tradeoff", "The Unreasonable Effectiveness of Deep Features as a Perceptual Metric", "Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "Loss Functions for Neural Networks for Image Processing"], "answer_arxiv_id": ["1711.06077", "1801.03924", "1603.08155", "1511.08861"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_1299"} +{"question": "What studies provided user control over the generation process, while preserving design intent?", "answer": ["Sketch2CAD: Sequential CAD Modeling by Sketching in Context", "Zero-shot CAD Program Re-Parameterization for Interactive Manipulation", "PLay: Parametrically Conditioned Layout Generation using Latent Diffusion", "SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebooks"], "answer_arxiv_id": ["2009.04927", "2306.03217", "2301.11529", "2207.04632"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_1300"} +{"question": "Do any studies suggest that EBMs do not require an explicit neural network for sample generation?", "answer": ["Implicit Generation and Generalization in Energy-Based Models"], "answer_arxiv_id": ["1903.08689"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_1301"} +{"question": "Could you provide me with studies that utilize 3D Morphable Model in audio-driven talking face generation?", "answer": ["Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From\n Single Image to Image Set"], "answer_arxiv_id": ["1903.08527"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1302"} +{"question": "Could you name the studies that utilized the attention mechanism to amplify subtle differences in the shallow layers for improving artifact detection performance?", "answer": ["On the Detection of Digital Face Manipulation", "Multi-attentional Deepfake Detection"], "answer_arxiv_id": ["1910.01717", "2103.02406"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_1303"} +{"question": "What are some papers about using diffusion models for image processing issues like image inpainting, super-resolution, deblurring and colorization?", "answer": ["Improving Diffusion Models for Inverse Problems using Manifold\n Constraints", "Diffusion Posterior Sampling for General Noisy Inverse Problems"], "answer_arxiv_id": ["2206.00941", "2209.14687"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_1304"} +{"question": "What research introduced equivariant GANs for group-invariant distribution learning?", "answer": ["Differentiable Augmentation for Data-Efficient GAN Training", "Group Equivariant Generative Adversarial Networks", "Structure-preserving GANs"], "answer_arxiv_id": ["2006.10738", "2005.01683", "2202.01129"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_1305"} +{"question": "What papers introduced replay-based continual learning methods?", "answer": ["Experience Replay for Continual Learning", "New Insights on Reducing Abrupt Representation Change in Online\n Continual Learning", "Dark Experience for General Continual Learning: a Strong, Simple\n Baseline", "Dealing with Cross-Task Class Discrimination in Online Continual\n Learning", "Online Prototype Learning for Online Continual Learning", "On Tiny Episodic Memories in Continual Learning", "Episodic Memory in Lifelong Language Learning"], "answer_arxiv_id": ["1811.11682", "2104.05025", "2004.07211", "2305.14657", "2308.00301", "1902.10486", "1906.01076"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_1306"} +{"question": "Can you list some works on generating entire demonstrations using Language Learning Models in In-Context Learning?", "answer": ["Self-Generated In-Context Learning: Leveraging Auto-regressive Language\n Models as a Demonstration Generator", "Thought Propagation: An Analogical Approach to Complex Reasoning with\n Large Language Models", "Self-ICL: Zero-Shot In-Context Learning with Self-Generated\n Demonstrations"], "answer_arxiv_id": ["2206.08082", "2310.03965", "2305.15035"], "source_meta": {"published_time": "20240712"}, "qid": "AutoScholarQuery_train_1307"} +{"question": "What studies have used meta-learning in inferring the effective return of trajectories?", "answer": ["Meta-Gradient Reinforcement Learning", "Beyond Exponentially Discounted Sum: Automatic Learning of Return Function", "Meta-Learning via Learned Loss", "What Can Learned Intrinsic Rewards Capture?"], "answer_arxiv_id": ["1805.09801", "1905.11591", "1906.05374v4", "1912.05500"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_1308"} +{"question": "Which work still requires a time-consuming optimization using SDS loss for 3D reconstruction?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2308.16512"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_1309"} +{"question": "Which papers present a study on the effects of communication topology on consensus rates in decentralized learning?", "answer": ["Tackling Data Heterogeneity: A New Unified Framework for Decentralized SGD with Sample-induced Topology", "Revisiting Optimal Convergence Rate for Smooth and Non-convex Stochastic Decentralized Optimization", "Communication-Efficient Topologies for Decentralized Learning with $O(1)$ Consensus Rate", "Beyond spectral gap: The role of the topology in decentralized learning"], "answer_arxiv_id": ["2207.03730", "2210.07863", "2210.07881v2", "2206.03093"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_1310"} +{"question": "Which works utilized curriculum learning to adaptively select simpler tasks?", "answer": ["Learning by Playing – Solving Sparse Reward Tasks from Scratch", "Teacher-Student Curriculum Learning", "Learning Curriculum Policies for Reinforcement Learning"], "answer_arxiv_id": ["1802.10567", "1707.00183", "1812.00285"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_1311"} +{"question": "Which papers propose learning latent semantic concepts from data for machine learning interpretability?", "answer": ["A Framework for Learning Ante-hoc Explainable Models via Concepts", "Neural Prototype Trees for Interpretable Fine-grained Image Recognition", "Deformable ProtoPNet: An Interpretable Image Classifier Using Deformable Prototypes", "Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictions", "On Completeness-aware Concept-Based Explanations in Deep Neural Networks"], "answer_arxiv_id": ["2108.11761", "2012.02046", "2111.15000", "1710.04806", "1910.07969"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_1312"} +{"question": "Which research pieces covered Prompt Tuning that validated prepending several learnable tokens into the input sequence of each Transformer layer?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "PPT: Pre-trained Prompt Tuning for Few-shot Learning", "Visual Prompt Tuning", "Conditional Prompt Learning for Vision-Language Models", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "MaPLe: Multi-modal Prompt Learning", "PLOT: Prompt Learning with Optimal Transport for Vision-Language Models", "Multimodal Few-Shot Learning with Frozen Language Models", "Prompting Visual-Language Models for Efficient Video Understanding", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science\n Question Answering", "Multimodal Chain-of-Thought Reasoning in Language Models"], "answer_arxiv_id": ["2101.00190", "2104.08691", "2109.04332", "2203.12119", "2203.05557", "2112.01518", "2210.03117", "2210.01253", "2106.13884", "2112.04478", "2209.09513", "2302.00923"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_1313"} +{"question": "Could you provide me the research proposing M-FAC?", "answer": ["M-FAC: Efficient Matrix-Free Approximations of Second-Order Information"], "answer_arxiv_id": ["2107.03356"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_1314"} +{"question": "Could you provide me papers about studying linguistic calibration on conversational models?", "answer": ["Reducing conversational agents’ overconfidence through linguistic calibration"], "answer_arxiv_id": ["2012.14983"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_1315"} +{"question": "Which works have been done on speech representation learning with the application of vector-quantization?", "answer": ["Vector-Quantized Autoregressive Predictive Coding"], "answer_arxiv_id": ["2005.08392"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_1316"} +{"question": "Has there been any theoretical analysis on sharpness along the GD trajectory in a two-layer linear network setting?", "answer": ["Analyzing Sharpness along GD Trajectory: Progressive Sharpening and Edge of Stability"], "answer_arxiv_id": ["2207.12678"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_1317"} +{"question": "Which studies have demonstrated impressive performances in the field of contrastive learning with semantic segmentation?", "answer": ["Supervised Contrastive Learning", "Exploring Cross-Image Pixel Contrast for Semantic Segmentation", "Multi-scale and Cross-scale Contrastive Learning for Semantic\n Segmentation"], "answer_arxiv_id": ["2004.11362", "2101.11939", "2203.13409"], "source_meta": {"published_time": "20240416"}, "qid": "AutoScholarQuery_train_1318"} +{"question": "Could you provide me some studies that have attempted to reconstruct objects under occlusion?", "answer": ["VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "3D-R2N2: A Unified Approach for Single and Multi-view 3D Object\n Reconstruction", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation"], "answer_arxiv_id": ["1711.06396", "1604.00449", "1901.05103"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_1319"} +{"question": "Which papers discussed the usage of structured data or unstructured data in Knowledge-based VQA Systems?", "answer": ["Out of the Box: Reasoning with Graph Convolution Nets for Factual Visual Question Answering", "Multi-Modal Answer Validation for Knowledge-Based VQA", "KRISP: Integrating Implicit and Symbolic Knowledge for Open-Domain Knowledge-Based VQA", "Multi-Modal Answer Validation for Knowledge-Based VQA", "KAT: A Knowledge Augmented Transformer for Vision-and-Language"], "answer_arxiv_id": ["1811.00538", "2103.12248", "2012.11014", "2103.12248", "2112.08614"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_1320"} +{"question": "What are some of the studies that devise evaluation metrics based on semantic similarity and likelihood of PLMs?", "answer": ["MoverScore: Text Generation Evaluating with Contextualized Embeddings\n and Earth Mover Distance", "BERTScore: Evaluating Text Generation with BERT", "BARTScore: Evaluating Generated Text as Text Generation", "InfoLM: A New Metric to Evaluate Summarization & Data2Text Generation", "CTRLEval: An Unsupervised Reference-Free Metric for Evaluating\n Controlled Text Generation", "GPTScore: Evaluate as You Desire"], "answer_arxiv_id": ["1909.02622", "1904.09675", "2106.11520", "2112.01589", "2204.00862", "2302.04166"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1321"} +{"question": "Could you cite works that created datasets with images and data generated using generative adversarial networks?", "answer": ["GANerated Hands for Real-Time 3D Hand Tracking from Monocular RGB"], "answer_arxiv_id": ["1712.01057"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_1322"} +{"question": "What work was inspired to rethink parameter-efficient fine-tuning from a design perspective?", "answer": ["On Network Design Spaces for Visual Recognition"], "answer_arxiv_id": ["1905.13214"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_1323"} +{"question": "Could you provide a study that reduced the constraint violation to O(1) by adding slackness to the primal-dual algorithms for CMDPs?", "answer": ["Learning Policies with Zero or Bounded Constraint Violation for Constrained MDPs"], "answer_arxiv_id": ["2106.02684"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_1324"} +{"question": "Which research revealed the susceptibility of deep learning-based SR methods to adversarial attacks?", "answer": ["Evaluating Robustness of Deep Image Super-Resolution against Adversarial\n Attacks"], "answer_arxiv_id": ["1904.06097"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_1325"} +{"question": "Which works revealed that data augmentation and consistency regularization can improve generalization and feature learning in supervised learning?", "answer": ["Sample Efficiency of Data Augmentation Consistency Regularization", "Data Augmentation as Feature Manipulation"], "answer_arxiv_id": ["2202.12230", "2203.01572"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_1326"} +{"question": "Can you provide the references for the two-tower framework in cross-modal retrieval that independently maps images and texts into a joint feature space?", "answer": ["Learning the Best Pooling Strategy for Visual Semantic Embedding", "VSE++: Improving Visual-Semantic Embeddings with Hard Negatives", "Dual-Path Convolutional Image-Text Embeddings with Instance Loss"], "answer_arxiv_id": ["2011.04305", "1707.05612", "1711.05535"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_1327"} +{"question": "Could you provide me some papers about using DDIM inversion for text-driven image-to-image translation?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "Prompt-to-Prompt Image Editing with Cross Attention Control", "DiffEdit: Diffusion-based semantic image editing with mask guidance"], "answer_arxiv_id": ["2211.12572", "2208.01626", "2210.11427"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_1328"} +{"question": "Which works are related to the extension of standard reinforcement learning methods using hierarchical reinforcement learning?", "answer": ["The Option-Critic Architecture", "Near-Optimal Representation Learning for Hierarchical Reinforcement Learning"], "answer_arxiv_id": ["1609.05140", "1810.01257"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_1329"} +{"question": "What frameworks are available for running workflows on clouds or supercomputers?", "answer": ["IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures", "Distributed Prioritized Experience Replay", "RLlib: Abstractions for Distributed Reinforcement Learning"], "answer_arxiv_id": ["1802.01561", "1803.00933", "1712.09381"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_1330"} +{"question": "Could you list some Large Multimodal Models that demonstrated effectiveness in detecting hateful memes?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Towards Language Models That Can See: Computer Vision Through the LENS\n of Natural Language"], "answer_arxiv_id": ["2204.14198", "2305.06500", "2306.16410"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_1331"} +{"question": "What researches have learned representations of actions on-the-fly for black-box optimization and path planning setting?", "answer": ["Learning Search Space Partition for Black-box Optimization using Monte Carlo Tree Search"], "answer_arxiv_id": ["2007.00708"], "source_meta": {"published_time": "20220822"}, "qid": "AutoScholarQuery_train_1332"} +{"question": "Could you provide me some works about the theoretical foundation of adversarial domain adaptation?", "answer": ["Generative Adversarial Nets", "A survey on domain adaptation theory: learning bounds and theoretical guarantees", "Bridging Theory and Algorithm for Domain Adaptation"], "answer_arxiv_id": ["1406.2661", "2004.11829", "1904.05801"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_1333"} +{"question": "What works build upon the recent success of language models?", "answer": ["Generative Language Modeling for Automated Theorem Proving", "Formal Mathematics Statement Curriculum Learning", "HyperTree Proof Search for Neural Theorem Proving"], "answer_arxiv_id": ["2009.03393", "2202.01344", "2205.11491"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_1334"} +{"question": "Where was the Diffusion Probabilistic Models (DPMs) first proposed?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_1335"} +{"question": "Which study proposes the selection of good in-context examples based on their semantic closeness to query sentences?", "answer": ["What Makes Good In-Context Examples for GPT-3?"], "answer_arxiv_id": ["2101.06804"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_1336"} +{"question": "Could you provide me studies that applied state-of-the-art score-based and denoising diffusion models in generative models?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "Image Super-Resolution via Iterative Refinement", "Palette: Image-to-Image Diffusion Models", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "Denoising Diffusion Restoration Models", "Score-Based Diffusion Models as Principled Priors for Inverse Imaging", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "Diffusion Posterior Sampling for General Noisy Inverse Problems"], "answer_arxiv_id": ["1907.05600", "2006.11239", "2011.13456", "2104.07636", "2111.05826", "2201.09865", "2201.11793", "2304.11751", "2212.00490", "2209.14687"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_1337"} +{"question": "Which work proposes a practical algorithm for threshold based selective classification on deep neural networks?", "answer": ["Selective Classification for Deep Neural Networks"], "answer_arxiv_id": ["1705.08500"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_1338"} +{"question": "What studies focus on reducing the computational burden during model training using Parameter-Efficient Fine-Tuning methods?", "answer": ["Sensitivity-Aware Visual Parameter-Efficient Fine-Tuning"], "answer_arxiv_id": ["2303.08566"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_1339"} +{"question": "What works made advancement in text-to-image diffusion models that stimulated the interest in data augmentation with synthetically-generated images?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10741", "2112.10752", "2205.11487", "2204.06125"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_1340"} +{"question": "Are there any studies that have applied linearized models to predict training speed?", "answer": ["Predicting Training Time Without Training"], "answer_arxiv_id": ["2008.12478"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_1341"} +{"question": "What research proposed a generic trajectory following reward in cross-morphology imitation learning?", "answer": ["DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills", "To Follow or not to Follow: Selective Imitation Learning from Observations"], "answer_arxiv_id": ["1804.02717", "1912.07670"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_1342"} +{"question": "Which papers proposed using bootstrap ensemble for enabling CP on time series?", "answer": ["Ensemble Conformalized Quantile Regression for Probabilistic Time Series Forecasting"], "answer_arxiv_id": ["2202.08756"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_1343"} +{"question": "Which work shows impressive results with single-step generation in student model training?", "answer": ["InstaFlow: One Step is Enough for High-Quality Diffusion-Based\n Text-to-Image Generation"], "answer_arxiv_id": ["2309.06380"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_1344"} +{"question": "What papers proposed to reduce the distance between pre-edit and post-edit parameters in fine-tuning-based model editors?", "answer": ["Modifying Memories in Transformer Models"], "answer_arxiv_id": ["2012.00363"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_1345"} +{"question": "Could you provide me with studies that derived similar results for ASGD, RMSprop and Adam?", "answer": ["Stochastic modified equations for the asynchronous stochastic gradient descent", "On the SDEs and Scaling Rules for Adaptive Gradient Algorithms"], "answer_arxiv_id": ["1805.08244", "2205.10287"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_1346"} +{"question": "Which works describe information directed sampling in reinforcement learning?", "answer": ["Information-Theoretic Confidence Bounds for Reinforcement Learning", "Reinforcement Learning, Bit by Bit", "Regret Bounds for Information-Directed Reinforcement Learning", "Surprise-Based Intrinsic Motivation for Deep Reinforcement Learning", "Wasserstein Dependency Measure for Representation Learning", "An Information-Theoretic Approach to Minimax Regret in Partial Monitoring", "Connections Between Mirror Descent, Thompson Sampling and the Information Ratio"], "answer_arxiv_id": ["1911.09724", "2103.04047", "2206.04640", "1703.01732", "1903.11780", "1902.00470", "1905.11817"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_1347"} +{"question": "Which paper does this research build upon by introducing a new structure learning algorithm based on sequential Monte Carlo sampling?", "answer": ["Structure Discovery in Nonparametric Regression through Compositional Kernel Search"], "answer_arxiv_id": ["1302.4922v4"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_1348"} +{"question": "What works have explored the dynamics of Sharpness-Aware Minimization (SAM) for convex quadratics?", "answer": ["The Dynamics of Sharpness-Aware Minimization: Bouncing Across Ravines and Drifting Towards Wide Minima"], "answer_arxiv_id": ["2210.01513"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_1349"} +{"question": "What studies explored the lottery ticket hypothesis in context of pruning in transformers?", "answer": ["The Lottery Ticket Hypothesis for Pre-trained BERT Networks", "EarlyBERT: Efficient BERT Training via Early-bird Lottery Tickets", "When BERT Plays the Lottery, All Tickets Are Winning"], "answer_arxiv_id": ["2007.12223", "2101.00063", "2005.00561"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_1350"} +{"question": "Which works propose the use of InfoNCE loss in Contrastive SSL algorithms?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1911.05722", "2002.05709"], "source_meta": {"published_time": "20211101"}, "qid": "AutoScholarQuery_train_1351"} +{"question": "Which study focuses on manipulating appearance by controlling color palette weights in NeRF editing?", "answer": ["PaletteNeRF: Palette-based Appearance Editing of Neural Radiance Fields", "PaletteNeRF: Palette-based Color Editing for NeRFs"], "answer_arxiv_id": ["2212.10699", "2212.12871"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_1352"} +{"question": "What studies use techniques to reduce modality gap in gloss-free SLT?", "answer": ["Sign Language Translation from Instructional Videos", "Open-Domain Sign Language Translation Learned from Online Video", "Gloss Attention for Gloss-free Sign Language Translation", "Gloss-Free End-to-End Sign Language Translation", "Gloss-free Sign Language Translation: Improving from Visual-Language\n Pretraining", "YouTube-ASL: A Large-Scale, Open-Domain American Sign Language-English\n Parallel Corpus"], "answer_arxiv_id": ["2304.06371", "2205.12870", "2307.07361", "2305.12876", "2307.14768", "2306.15162"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_1353"} +{"question": "What research applied the idea of merging candidates generated by LLMs to MT?", "answer": ["An Empirical Study of Translation Hypothesis Ensembling with Large\n Language Models"], "answer_arxiv_id": ["2310.11430"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_1354"} +{"question": "What studies make the assumption that the transition kernel in linear MDPs is a linear combination of basis transition probability functions?", "answer": ["Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs", "Variance-Dependent Regret Bounds for Linear Bandits and Reinforcement Learning: Adaptivity and Computational Efficiency"], "answer_arxiv_id": ["1910.10597", "2006.01107", "2006.01107", "2012.08507", "2205.11507", "2302.10371"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1355"} +{"question": "What are some works on generating audio conditioned on text?", "answer": ["AudioGen: Textually Guided Audio Generation", "Diffsound: Discrete Diffusion Model for Text-to-sound Generation", "Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion\n Models", "Make-An-Audio 2: Temporal-Enhanced Text-to-Audio Generation", "Text-to-Audio Generation using Instruction-Tuned LLM and Latent\n Diffusion Model", "CLIPSonic: Text-to-Audio Synthesis with Unlabeled Videos and Pretrained\n Language-Vision Models"], "answer_arxiv_id": ["2209.15352", "2207.09983", "2301.12661", "2305.18474", "2304.13731", "2306.09635"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1356"} +{"question": "Which sources explain how one can leverage self-supervised techniques to generate pseudo-labels in order to use traditional supervised learning rejection methods?", "answer": ["Self-Supervised Anomaly Detection: A Survey and Outlook", "SSD: A Unified Framework for Self-Supervised Outlier Detection", "Anomaly Detection in Video via Self-Supervised and Multi-Task Learning", "CutPaste: Self-Supervised Learning for Anomaly Detection and Localization"], "answer_arxiv_id": ["2205.05173", "2103.12051", "2011.07491", "2104.04015"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_1357"} +{"question": "Which research is an early effort in exploring better controllability on diffusion models beyond the prompt-based image editing?", "answer": ["Diffusion Self-Guidance for Controllable Image Generation"], "answer_arxiv_id": ["2306.00986"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_1358"} +{"question": "Are there any works focusing on the pixel-wise unit vectors pointing to 2D projections of a set of 3D keypoints, such as PVNet?", "answer": ["PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation"], "answer_arxiv_id": ["1812.11788"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_1359"} +{"question": "What works constructed the Hamiltonian matrices datasets for the study of Hamiltonian matrices for molecules?", "answer": ["A deep neural network for molecular wave functions in quasi-atomic minimal basis representation", "Efficient and Equivariant Graph Networks for Predicting Quantum Hamiltonian"], "answer_arxiv_id": ["2005.06979", "2306.04922"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_1360"} +{"question": "What studies used linguistic observations to map from phoneme to viseme sequences?", "answer": ["LRS3-TED: a large-scale dataset for visual speech recognition"], "answer_arxiv_id": ["1809.00496"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1361"} +{"question": "Which paper introduced the FVAE metric for model evaluation when the generative process of a dataset is given?", "answer": ["Disentangling by Factorising"], "answer_arxiv_id": ["1802.05983"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_1362"} +{"question": "Is there any study proposes a contrastive strategy based on the meta-path for sequence based modelling?", "answer": ["MERIt: Meta-Path Guided Contrastive Learning for Logical Reasoning"], "answer_arxiv_id": ["2203.00357"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_1363"} +{"question": "Are there any papers that study emergent generalization to new language tasks within large language models?", "answer": ["Language Models are Few-Shot Learners", "Emergent Abilities of Large Language Models"], "answer_arxiv_id": ["2005.14165", "2206.07682"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_1364"} +{"question": "Could you provide me some empirical studies supporting the idea that effective neural networks tend to learn similar representations for semantically similar data?", "answer": ["Relative representations enable zero-shot latent space communication", "ASIF: Coupled Data Turns Unimodal Models to Multimodal without Training", "How do Variational Autoencoders Learn? Insights from Representational Similarity", "Representation Topology Divergence: a Method for Comparing Neural Network Representations", "Word translation without parallel data", "Understanding image representations by measuring their equivariance and equivalence", "Low-Dimensional Structure in the Space of Language Representations is Reflected in Brain Responses", "Representation Learning: A Review and New Perspectives", "The Geometry of Multilingual Language Model Representations"], "answer_arxiv_id": ["2209.15430", "2210.01738", "2205.08399", "2201.00058", "1710.04087", "1411.5908", "2106.05426", "1206.5538", "2205.10964"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_1365"} +{"question": "Which research articles are related to studying unstructured sparsity?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding"], "answer_arxiv_id": ["1510.00149"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_1366"} +{"question": "What studies suggest that other sample-based models may suffer from mode collapse?", "answer": ["Generative Moment Matching Networks"], "answer_arxiv_id": ["1502.02761"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_1367"} +{"question": "What studies are there that discuss identifying causal structures with a fixed budget of experiments?", "answer": ["Budgeted Experiment Design for Causal Structure Learning"], "answer_arxiv_id": ["1709.03625"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_1368"} +{"question": "Could you show me some research on Source-Free Domain Adaptation?", "answer": ["Do We Really Need to Access the Source Data? Source Hypothesis Transfer\n for Unsupervised Domain Adaptation", "Universal Source-Free Domain Adaptation", "Generalized Source-free Domain Adaptation"], "answer_arxiv_id": ["2002.08546", "2004.04393", "2108.01614"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_1369"} +{"question": "What literature has discussed feature distribution in the context of domain adaptation?", "answer": ["Adversarial Discriminative Domain Adaptation"], "answer_arxiv_id": ["1702.05464"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_1370"} +{"question": "What is the study that transforms pose estimation into a classification problem in the context of direct methods?", "answer": ["SSD-6D: Making RGB-based 3D detection and 6D pose estimation great again"], "answer_arxiv_id": ["1711.10006"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_1371"} +{"question": "Are there any works that proposed a global and local mixture consistency loss and head-tail soft label re-weighted loss?", "answer": ["Global and Local Mixture Consistency Cumulative Learning for Long-tailed\n Visual Recognitions"], "answer_arxiv_id": ["2305.08661"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_1372"} +{"question": "Are there any studies about the gains from scaling model size?", "answer": ["Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2112.11446", "2201.11990", "2204.02311"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_1373"} +{"question": "What papers propose aligning the task losses magnitudes by rescaling them based on the task uncertainty?", "answer": ["Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics"], "answer_arxiv_id": ["1705.07115"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_1374"} +{"question": "Are there any studies that used vector quantization and diffusion for learning 3D-aware generative models?", "answer": ["VQ3D: Learning a 3D-Aware Generative Model on ImageNet", "3D-aware Image Generation using 2D Diffusion Models"], "answer_arxiv_id": ["2302.06833", "2303.17905"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_1375"} +{"question": "Could you tell me about research that applied NeRF to fuse multi-view features?", "answer": ["LERF: Language Embedded Radiance Fields", "Decomposing NeRF for Editing via Feature Field Distillation", "3D Concept Learning and Reasoning from Multi-View Images", "OpenMask3D: Open-Vocabulary 3D Instance Segmentation", "Neural Implicit Vision-Language Feature Fields"], "answer_arxiv_id": ["2303.09553", "2205.15585", "2303.11327", "2306.13631", "2303.10962"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_1376"} +{"question": "Which papers have studied membership inference and utilized some score function like loss for this purpose?", "answer": ["Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "White-box vs Black-box: Bayes Optimal Strategies for Membership Inference"], "answer_arxiv_id": ["1709.01604", "1908.11229"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_1377"} +{"question": "Which research papers introduced belief over the model parameter space of clients in Decentralized Federated Learning (DFL)?", "answer": ["Peer-to-Peer Federated Learning on Graphs"], "answer_arxiv_id": ["1901.11173"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_1378"} +{"question": "Which papers investigate budget-management in second-price auctions?", "answer": ["Multiplicative Pacing Equilibria in Auction Markets"], "answer_arxiv_id": ["1706.07151v5"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_1379"} +{"question": "Which studies especially focus on the generative approach in SSL, specifically masked language models?", "answer": ["Language Models are Few-Shot Learners", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["2005.14165", "1810.04805"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_1380"} +{"question": "Are there any studies on manipulation problems involving fluids?", "answer": ["Graph-Structured Visual Imitation", "Self-supervised Transparent Liquid Segmentation for Robotic Pouring", "Rethinking Optimization with Differentiable Simulation from a Global Perspective", "SPNets: Differentiable Fluid Dynamics for Deep Neural Networks", "ToolFlowNet: Robotic Manipulation with Tools via Predicting Tool Flow from Point Clouds"], "answer_arxiv_id": ["1907.05518", "2203.01538v1", "2207.00167", "1806.06094", "2211.09006v1"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_1381"} +{"question": "Which papers prior to this focussed on zooming techniques applied only to non-ImageNet, fine-grained classifications?", "answer": ["Deformable ProtoPNet: An Interpretable Image Classifier Using Deformable Prototypes", "Looking for the Devil in the Details: Learning Trilinear Attention Sampling Network for Fine-grained Image Recognition", "Visual correspondence-based explanations improve AI robustness and human-AI team accuracy", "Looking for the Devil in the Details: Learning Trilinear Attention Sampling Network for Fine-grained Image Recognition", "Efficient Classification of Very Large Images with Tiny Objects", "Learning to Downsample for Segmentation of Ultra-High Resolution Images"], "answer_arxiv_id": ["2111.15000", "1903.06150", "2208.00780", "1903.06150", "2106.02694", "2109.11071"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_1382"} +{"question": "What research is closely similar to ours in the concept of learning to combine the probability distributions of multiple language models?", "answer": ["CombLM: Adapting Black-Box Language Models through Small Fine-Tuned\n Models"], "answer_arxiv_id": ["2305.16876"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_1383"} +{"question": "Are there any studies that focus on differentially private stochastic optimization?", "answer": ["Adapting to Function Difficulty and Growth Conditions in Private Optimization", "Private Stochastic Convex Optimization with Optimal Rates", "Stability of Stochastic Gradient Descent on Nonsmooth Convex Losses", "Private Stochastic Convex Optimization: Optimal Rates in Linear Time", "Improved Rates for Differentially Private Stochastic Convex Optimization with Heavy-Tailed Data", "Private Convex Optimization via Exponential Mechanism", "ReSQueing Parallel and Private Stochastic Convex Optimization", "Private Convex Optimization in General Norms"], "answer_arxiv_id": ["2108.02391", "1908.09970", "2006.06914", "2005.04763", "2106.01336", "2203.00263", "2301.00457", "2207.08347"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_1384"} +{"question": "Which works have employed advanced generative models for automatic colorization?", "answer": ["Learning Diverse Image Colorization", "Unsupervised Diverse Colorization via Generative Adversarial Networks", "UniColor: A Unified Framework for Multi-Modal Colorization with Transformer", "ChromaGAN: Adversarial Picture Colorization with Semantic Class Distribution", "PalGAN: Image Colorization with Palette Generative Adversarial Networks", "Guided Image Generation with Conditional Invertible Neural Networks", "Colorization Transformer"], "answer_arxiv_id": ["1612.01958", "1702.06674", "2209.11223", "1907.09837", "2210.11204", "1907.02392", "2102.04432"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_1385"} +{"question": "What works labeled distinctive factors in existing datasets for the evaluation of model robustness?", "answer": ["ImageNet-X: Understanding Model Mistakes with Factor of Variation Annotations"], "answer_arxiv_id": ["2211.01866"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_1386"} +{"question": "What are some studies focusing on revising the surrogate in the context of Perturbative availability poison?", "answer": ["Self-Ensemble Protection: Training Checkpoints Are Good Data Protectors"], "answer_arxiv_id": ["2211.12005"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_1387"} +{"question": "What papers discuss the extension of GANs from image generation to video generation?", "answer": ["Adversarial Video Generation on Complex Datasets", "Latent Neural Differential Equations for Video Generation", "Lower Dimensional Kernels for Video Discriminators", "StyleVideoGAN: A Temporal Generative Model using a Pretrained StyleGAN", "StyleGAN-V: A Continuous Video Generator with the Price, Image Quality\n and Perks of StyleGAN2"], "answer_arxiv_id": ["1907.06571", "2011.03864", "1912.08860v1", "2107.07224", "2112.14683"], "source_meta": {"published_time": "20230826"}, "qid": "AutoScholarQuery_train_1388"} +{"question": "Could you provide me with some works that applies score-based generative models for image translation?", "answer": ["UNIT-DDPM: UNpaired Image Translation with Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2104.05358"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_1389"} +{"question": "Can you provide me some surveys that provide more details on EA?", "answer": ["A Benchmark and Comprehensive Survey on Knowledge Graph Entity Alignment via Representation Learning"], "answer_arxiv_id": ["2103.15059v5"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_1390"} +{"question": "Do you know any studies that employed diffusion models in image inpainting tasks?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2112.10752", "2205.11487"], "source_meta": {"published_time": "20231224"}, "qid": "AutoScholarQuery_train_1391"} +{"question": "What research proposed use of transformer architecture for generation of individual neuron's representations?", "answer": ["Seeing the forest and the tree: Building representations of both individual and collective dynamics with transformers"], "answer_arxiv_id": ["2206.06131"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_1392"} +{"question": "Could you provide the work that primarily investigate optimal stopping problems?", "answer": ["Cutting Your Losses: Learning Fault-Tolerant Control and Optimal Stopping under Adverse Risk"], "answer_arxiv_id": ["1902.05045"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_1393"} +{"question": "What research identified patterns in real-world hypergraphs?", "answer": ["Structural Patterns and Generative Models of Real-world Hypergraphs"], "answer_arxiv_id": ["2006.07060"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_1394"} +{"question": "Could you provide me an example of studies that designed customized neural architectures for periodic learning?", "answer": ["Multi-Task Temporal Shift Attention Networks for On-Device Contactless Vitals Measurement", "Counting Out Time: Class Agnostic Video Repetition Counting in the Wild"], "answer_arxiv_id": ["2006.03790", "2006.15418"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_1395"} +{"question": "What studies explore the method of ratio matching for learning EBMs?", "answer": ["Interpretation and Generalization of Score Matching"], "answer_arxiv_id": ["1205.2629v1"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_1396"} +{"question": "What papers discuss the current attempts in IRL algorithms to account for possible irrationalities in the expert?", "answer": ["Learning the Preferences of Ignorant, Inconsistent Agents", "Where Do You Think You’re Going?: Inferring Beliefs about Dynamics from Behavior", "On the Feasibility of Learning, Rather than Assuming, Human Biases for Reward Inference"], "answer_arxiv_id": ["1512.05832v1", "1805.08010", "1906.09624"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_1397"} +{"question": "Which work stands as representative for text conditioned image generation with high efficiency and competitive quality?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_1398"} +{"question": "What works have formalized the paradigm of reward-free reinforcement learning (RL), providing both upper and lower bounds on the sample complexity?", "answer": ["Reward-Free Exploration for Reinforcement Learning"], "answer_arxiv_id": ["2002.02794"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_1399"} +{"question": "Could you provide studies that explored the latent reasoning of LLMs using single-hop reasoning tasks?", "answer": ["Locating and Editing Factual Associations in GPT", "Dissecting Recall of Factual Associations in Auto-Regressive Language\n Models", "Identifying Linear Relational Concepts in Large Language Models", "Linearity of Relation Decoding in Transformer Language Models"], "answer_arxiv_id": ["2202.05262", "2304.14767", "2311.08968", "2308.09124"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_1400"} +{"question": "Could you provide me some studies stating that learning performance of individual problems do not exactly align with the transport cost?", "answer": ["Do Neural Optimal Transport Solvers Work? A Continuous Wasserstein-2 Benchmark"], "answer_arxiv_id": ["2106.01954"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_1401"} +{"question": "Which works focus on comparing generalization and dynamical properties of neural networks in both linear and non-linear regimes?", "answer": ["Deep learning versus kernel learning: an empirical study of loss landscape geometry and the time evolution of the Neural Tangent Kernel", "Disentangling feature and lazy training in deep neural networks", "Geometric compression of invariant manifolds in neural nets", "Limitations of the NTK for Understanding Generalization in Deep Learning", "Learning sparse features can lead to overfitting in neural networks"], "answer_arxiv_id": ["2010.15110", "1906.08034", "2007.11471", "2206.10012", "2206.12314"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_1402"} +{"question": "What studies talk about latent diffusion models used for creating illusions?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1403"} +{"question": "Which research papers delved into assessing which tokens receive the most attention in Transformers?", "answer": ["What Does BERT Look At? An Analysis of BERT’s Attention", "A Primer in BERTology: What We Know About How BERT Works", "BERTology Meets Biology: Interpreting Attention in Protein Language Models"], "answer_arxiv_id": ["1906.04341", "2002.12327", "2006.15222"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_1404"} +{"question": "Which works studied the effect of step size on the sharpness along the optimization trajectory of the Edge of Stability (EoS)?", "answer": ["On the relation between the sharpest directions of DNN loss and the SGD step length", "The break-even point on optimization trajectories of deep neural networks"], "answer_arxiv_id": ["1807.05031", "2002.09572"], "source_meta": {"published_time": "20230709"}, "qid": "AutoScholarQuery_train_1405"} +{"question": "Which work studies Average Individual Fairness by regulating the average error rate for individuals?", "answer": ["Average Individual Fairness: Algorithms, Generalization and Experiments"], "answer_arxiv_id": ["1905.10607"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_1406"} +{"question": "Which work utilizes pre-defined SPARQL templates to solve questions step-by-step in an agent-based method?", "answer": ["AgentBench: Evaluating LLMs as Agents"], "answer_arxiv_id": ["2308.03688"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_1407"} +{"question": "Which work proves the best-iterate convergence for the same range of ρ as in the best-known result?", "answer": ["Solving Nonconvex-Nonconcave Min-Max Problems exhibiting Weak Minty Solutions"], "answer_arxiv_id": ["2201.12247"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_1408"} +{"question": "What papers are about the use of StyleGAN to perform image editing through a text-based interface?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.17249", "2103.00020"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_1409"} +{"question": "What research demonstrates that gradient-based techniques for MTO often perform on par with the less costly approach of directly optimizing the average of the task losses?", "answer": ["Do Current Multi-Task Optimization Methods in Deep Learning Even Help?", "In Defense of the Unitary Scalarization for Deep Multi-Task Learning"], "answer_arxiv_id": ["2209.11379", "2201.04122"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_1410"} +{"question": "Can you mention some studies that use a limited number of tokens, like ToMe in which the method progressively merges similar tokens layer-by-layer?", "answer": ["Token Merging: Your ViT But Faster"], "answer_arxiv_id": ["2210.09461"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_1411"} +{"question": "Are there any works about Transformer models that use point-wise attention in time series forecasting?", "answer": ["Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecasting", "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting"], "answer_arxiv_id": ["1907.00235", "2106.13008"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_1412"} +{"question": "What is the basis of the UMIC metric used for evaluating image captioning models?", "answer": ["UMIC: An Unreferenced Metric for Image Captioning via Contrastive\n Learning"], "answer_arxiv_id": ["2106.14019"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_1413"} +{"question": "Which papers extend the NeRF formulation for novel LiDAR view synthesis?", "answer": ["LiDAR-NeRF: Novel LiDAR View Synthesis via Neural Radiance Fields", "NeRF-LiDAR: Generating Realistic LiDAR Point Clouds with Neural Radiance\n Fields", "Neural LiDAR Fields for Novel View Synthesis", "NeuralPCI: Spatio-temporal Neural Field for 3D Point Cloud Multi-frame\n Non-linear Interpolation", "PC-NeRF: Parent-Child Neural Radiance Fields under Partial Sensor Data\n Loss in Autonomous Driving Environments"], "answer_arxiv_id": ["2304.10406", "2304.14811", "2305.01643v2", "2303.15126", "2310.00874"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1414"} +{"question": "What are earlier studies on the mean-field analysis of gradient descent algorithms for implementing neural networks?", "answer": ["A Mean Field View of the Landscape of Two-Layer Neural Networks", "Mean Field Analysis of Neural Networks: A Law of Large Numbers", "On the Global Convergence of Gradient Descent for Over-parameterized Models using Optimal Transport", "On the Banach spaces associated with multi-layer ReLU networks Function representation, approximation theory and gradient descent dynamics", "A mean-field limit for certain deep neural networks", "Mean Field Analysis of Deep Neural Networks", "Mean Field Limit of the Learning Dynamics of Multilayer Neural Networks"], "answer_arxiv_id": ["1804.06561", "1805.01053", "1805.09545", "2007.15623", "1906.00193", "1903.04440", "1902.02880"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_1415"} +{"question": "What research proposed DoLa as a decoding strategy that aims to mitigate hallucinations in MLLMs?", "answer": ["DoLa: Decoding by Contrasting Layers Improves Factuality in Large\n Language Models"], "answer_arxiv_id": ["2309.03883"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1416"} +{"question": "Could you provide me with some studies that conducted a human study on code generation tools?", "answer": ["Grounded Copilot: How Programmers Interact with Code-Generating Models", "In-IDE Code Generation from Natural Language: Promise and Challenges"], "answer_arxiv_id": ["2206.15000", "2101.11149"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_1417"} +{"question": "What research describe the usage of Latent space methods for optimization?", "answer": ["Auto-Encoding Variational Bayes", "Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules", "High Dimensional Bayesian Optimization via Supervised Dimension Reduction", "Sample-Efficient Optimization in the Latent Space of Deep Generative Models via Weighted Retraining", "Good practices for Bayesian Optimization of high dimensional structured spaces", "High-Dimensional Bayesian Optimisation with Variational Autoencoders and Deep Metric Learning"], "answer_arxiv_id": ["1312.6114", "1610.02415", "1907.08953v1", "2006.09191", "2012.15471v2", "2106.03609v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1418"} +{"question": "Which studies view adversarial training as an effective way to improve the adversarial robustness of deep learning models?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_1419"} +{"question": "What studies employed transliteration to a common script during the pretraining phase to enable cross-lingual transfer?", "answer": ["Exploiting Language Relatedness for Low Web-Resource Language Model\n Adaptation: An Indic Languages Study", "Role of Language Relatedness in Multilingual Fine-tuning of Language\n Models: A Case Study in Indo-Aryan Languages", "Does Transliteration Help Multilingual Language Modeling?", "Romanization-based Large-scale Adaptation of Multilingual Language\n Models"], "answer_arxiv_id": ["2106.03958", "2109.10534", "2201.12501", "2304.08865"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_1420"} +{"question": "Which papers examined the ability of LLMs to make rational decisions in game-theoretic scenarios?", "answer": ["Can Large Language Models Serve as Rational Players in Game Theory? A\n Systematic Analysis"], "answer_arxiv_id": ["2312.05488"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1421"} +{"question": "Which paper proposes removing cross-correlation across feature vector embeddings from Siamese networks?", "answer": ["Barlow Twins: Self-Supervised Learning via Redundancy Reduction"], "answer_arxiv_id": ["2103.03230"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_1422"} +{"question": "What studies attempted to quantify data difficulty by calculating the 'margin' from the decision boundary in the context of robust learning?", "answer": ["Geometry-aware Instance-reweighted Adversarial Training"], "answer_arxiv_id": ["2010.01736"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_1423"} +{"question": "What papers look into Domain Adaptation as a part of Transfer Learning?", "answer": ["Unsupervised Domain Adaptation by Backpropagation", "Maximum Classifier Discrepancy for Unsupervised Domain Adaptation", "Bridging Theory and Algorithm for Domain Adaptation", "Transferable Semantic Augmentation for Domain Adaptation"], "answer_arxiv_id": ["1409.7495", "1712.02560", "1904.05801", "2103.12562"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_1424"} +{"question": "Could you tell me about the works focusing on the sensitivity of deep RL algorithms to stochasticity and hyperparameters, and the extreme variability of results across seeds?", "answer": ["Deep Reinforcement Learning that Matters", "Deep Reinforcement Learning at the Edge of the Statistical Precipice"], "answer_arxiv_id": ["1709.06560", "2108.13264"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_1425"} +{"question": "Can you provide some works that optimize NeRF in latent spaces?", "answer": ["Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures"], "answer_arxiv_id": ["2211.07600"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_1426"} +{"question": "Which works are considered fundamental in the field of Large Language Models (LLMs)?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2005.14165", "2302.13971"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1427"} +{"question": "What research papers discuss the extension of the diffusion framework to various domains?", "answer": ["Diffusion-GAN: Training GANs with Diffusion", "Noise2Music: Text-conditioned Music Generation with Diffusion Models", "DreamFusion: Text-to-3D using 2D Diffusion", "Re-imagine the Negative Prompt Algorithm: Transform 2D Diffusion into 3D, alleviate Janus problem and Beyond", "Diffusion-LM Improves Controllable Text Generation", "NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with 360° Views", "CARD: Classification and Regression Diffusion Models", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning"], "answer_arxiv_id": ["2206.02262", "2302.03917", "2209.14988", "2304.04968", "2205.14217", "2211.16431", "2206.07275", "2208.06193"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_1428"} +{"question": "Which papers have achieved success in diffusion models in different data modalities?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_1429"} +{"question": "Which papers talk about Multimodal Large Language Models (MLLMs) and the techniques they use for aligning vision and language modalities?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Visual Instruction Tuning", "Kosmos-2: Grounding Multimodal Large Language Models to the World", "Florence-2: Advancing a Unified Representation for a Variety of Vision\n Tasks"], "answer_arxiv_id": ["2204.14198", "2303.16199", "2301.12597", "2305.06500", "2304.08485", "2306.14824", "2311.06242"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_1430"} +{"question": "Which papers contain information on direct methods for offline policy evaluation?", "answer": ["Batch Policy Learning under Constraints"], "answer_arxiv_id": ["1903.08738"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_1431"} +{"question": "What work presents a plug-and-play distillation approach utilizing LoRA to reduce the computation cost in inference time of diffusion models?", "answer": ["LCM-LoRA: A Universal Stable-Diffusion Acceleration Module"], "answer_arxiv_id": ["2311.05556"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_1432"} +{"question": "Which works have shown that gradient-based optimization with a differentiable renderer (DiffDRR) is faster and more robust than gradient-free methods for intensity-based 2D/3D registration?", "answer": ["Fast Auto-Differentiable Digitally Reconstructed Radiographs for Solving\n Inverse Problems in Intraoperative Imaging"], "answer_arxiv_id": ["2208.12737"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_1433"} +{"question": "What are the studies that propose 4D representation for dynamic implicit shapes?", "answer": ["LoRD: Local 4D Implicit Representation for High-Fidelity Dynamic Human Modeling", "Learning Compositional Representation for 4D Captures with Neural ODE", "Learning Parallel Dense Correspondence from Spatio-Temporal Descriptors for Efficient and Robust 4D Reconstruction", "CaSPR: Learning Canonical Spatiotemporal Point Cloud Representations"], "answer_arxiv_id": ["2208.08622", "2103.08271", "2103.16341", "2008.02792"], "source_meta": {"published_time": "20230716"}, "qid": "AutoScholarQuery_train_1434"} +{"question": "Which papers mention the use of MiniGPT-4 and LLaVA in M-IT?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning"], "answer_arxiv_id": ["2304.10592", "2304.08485"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_1435"} +{"question": "What papers showcased the method of knowledge distillation by guiding a compact network to mimic the output of a large network?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20231104"}, "qid": "AutoScholarQuery_train_1436"} +{"question": "Can you provide me with examples of work that use bi-directional GRU and Transformer architectures to solve in-betweening and super-resolution tasks in neural motion generation?", "answer": ["Recurrent Transition Networks for Character Locomotion", "Convolutional Autoencoders for Human Motion Infilling", "Robust Motion In-betweening", "Single-Shot Motion Completion with Transformer"], "answer_arxiv_id": ["1810.02363v5", "2010.11531", "2102.04942", "2103.00776"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_1437"} +{"question": "What research papers discuss the estimation of transferability through various metrics such as negative conditional entropy, log expectation, marginalized likelihood, PAC-Bayesian bound, and mutual information?", "answer": ["Transferability and Hardness of Supervised Classification Tasks", "LEEP: A New Measure to Evaluate Transferability of Learned Representations", "LogME: Practical Assessment of Pre-trained Models for Transfer Learning", "Ranking and Tuning Pre-trained Models: A New Paradigm for Exploiting Model Hubs", "PACTran: PAC-Bayesian Metrics for Estimating the Transferability of Pretrained Models to Classification Tasks", "Frustratingly Easy Transferability Estimation"], "answer_arxiv_id": ["1908.08142", "2002.12462", "2102.11005", "2110.10545", "2203.05126", "2106.09362"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_1438"} +{"question": "Can you list the papers that investigated the shared structure of tasks, meta-learning, or generative models to boost the capabilities of continual reinforcement learning agents?", "answer": ["Reinforced Continual Learning", "Continual and Multi-Task Architecture Search", "Reset-Free Lifelong Learning with Skill-Space Planning", "Unicorn: Continual learning with a universal, off-policy agent", "Block Contextual MDPs for Continual Learning", "Meta-Learning Representations for Continual Learning", "Meta-learnt priors slow down catastrophic forgetting in neural networks", "Learning to Continually Learn", "Online Fast Adaptation and Knowledge Accumulation: a New Approach to Continual Learning", "Learning to Learn without Forgetting By Maximizing Transfer and Minimizing Interference", "Online Meta-Critic Learning for Off-Policy Actor-Critic Methods", "PowerPlay: Training an Increasingly General Problem Solver by Continually Searching for the Simplest Still Unsolvable Problem", "Continual Learning with Deep Generative Replay", "Pseudo-Rehearsal: Achieving Deep Reinforcement Learning without Catastrophic Forgetting"], "answer_arxiv_id": ["1805.12369", "1906.05226", "2012.03548", "1802.08294", "2110.06972", "1905.12588", "1909.04170", "2002.09571", "2003.05856v3", "1810.11910", "2003.05334", "1112.5309", "1705.08690", "1812.02464"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_1439"} +{"question": "In the context of machine learning models' vulnerability to privacy attacks, which papers discuss membership inference attacks?", "answer": ["Membership Inference Attacks Against Machine Learning Models"], "answer_arxiv_id": ["1610.05820"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_1440"} +{"question": "Which studies talk about variance-aware bounds in bandits?", "answer": ["Improved Regret Analysis for Variance-Adaptive Linear Bandits and Horizon-Free Linear Mixture MDPs", "Improved Variance-Aware Confidence Sets for Linear Bandits and Linear Mixture MDP", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Variance-Dependent Regret Bounds for Linear Bandits and Reinforcement Learning: Adaptivity and Computational Efficiency"], "answer_arxiv_id": ["2111.03289", "2101.12745", "2012.08507", "2302.10371"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_1441"} +{"question": "What research proposed the encoding of reciprocal information of a query image into a vector and computation of similarities using Jaccard distance?", "answer": ["Re-ranking Person Re-identification with k-reciprocal Encoding"], "answer_arxiv_id": ["1701.08398"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_1442"} +{"question": "Can you name the studies that involve the Generalized Referring Expression Segmentation (GRES) task, which supports multi-target and empty-target scenarios?", "answer": ["GRES: Generalized Referring Expression Segmentation"], "answer_arxiv_id": ["2306.00968"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_1443"} +{"question": "Can you provide some work that focus on finding primitives in 3D data?", "answer": ["MonteBoxFinder: Detecting and Filtering Primitives to Fit a Noisy Point Cloud"], "answer_arxiv_id": ["2207.14268"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_1444"} +{"question": "Which work introduces variational information distillation in order to maximize mutual information between intermediate feature maps in super resolution tasks?", "answer": ["Learning with Privileged Information for Efficient Image Super-Resolution"], "answer_arxiv_id": ["2007.07524"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_1445"} +{"question": "What study proposed a maximum-likelihood approach for inference task in phylogenetic trees?", "answer": ["Learning phylogenetic trees as hyperbolic point configurations"], "answer_arxiv_id": ["2104.11430"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_1446"} +{"question": "Can you give me examples of work that utilize the data replay method for incremental learning in the image domain?", "answer": ["Gradient Episodic Memory for Continual Learning", "iCaRL: Incremental Classifier and Representation Learning"], "answer_arxiv_id": ["1706.08840", "1611.07725"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_1447"} +{"question": "What was the first work introducing a model-based method for the online episodic setting in RL?", "answer": ["Minimax Regret Bounds for Reinforcement Learning"], "answer_arxiv_id": ["1703.05449"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_1448"} +{"question": "What methods have been proposed to augment tasks with complicated structures?", "answer": ["FlipDA: Effective and Robust Data Augmentation for Few-Shot Learning"], "answer_arxiv_id": ["2108.06332"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_1449"} +{"question": "What studies used the French National Math Exam, Hungarian National High School Exam, and GHOST for assessing LLMs’ reasoning proficiency?", "answer": ["Mathematical Capabilities of ChatGPT"], "answer_arxiv_id": ["2301.13867"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_1450"} +{"question": "What papers are about accelerating the generation process of the diffusion process in image generation?", "answer": ["Denoising Diffusion Implicit Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2010.02502", "2202.09778"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_1451"} +{"question": "Could you tell me about some works that used Test-Time Adaptation methods in Test-Time adaptive techniques?", "answer": ["Tent: Fully Test-Time Adaptation by Entropy Minimization", "MEMO: Test Time Robustness via Adaptation and Augmentation"], "answer_arxiv_id": ["2006.10726", "2110.09506"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_1452"} +{"question": "Which paper originally proposed Transformers?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_1453"} +{"question": "Could you provide me some studies that employed evolutionary computation in discovery of activation functions?", "answer": ["Evolutionary Optimization of Deep Learning Activation Functions", "Discovering Parametric Activation Functions", "Evolving Normalization-Activation Layers", "The Quest for the Golden Activation Function"], "answer_arxiv_id": ["2002.07224", "2006.03179", "2004.02967", "1808.00783v1"], "source_meta": {"published_time": "20230113"}, "qid": "AutoScholarQuery_train_1454"} +{"question": "Which works use an auto-encoding approach to mask and reconstruct key substructures of molecular graphs for pretraining?", "answer": ["Strategies for Pre-training Graph Neural Networks", "N-Gram Graph: Simple Unsupervised Representation for Graphs, with Applications to Molecules"], "answer_arxiv_id": ["1905.12265", "1806.09206"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_1455"} +{"question": "What are some well-known results related to linear setting?", "answer": ["Variance-Dependent Regret Bounds for Linear Bandits and Reinforcement Learning: Adaptivity and Computational Efficiency"], "answer_arxiv_id": ["2302.10371"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_1456"} +{"question": "What research papers discussed about the smoothness of loss landscape of policy optimization?", "answer": ["On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift", "On the Convergence Rates of Policy Gradient Methods", "Global Convergence of Policy Gradient Methods for the Linear Quadratic Regulator"], "answer_arxiv_id": ["1908.00261", "2201.07443", "1801.05039"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_1457"} +{"question": "Which papers explored the idea of LLMs prompting themselves or other LLMs?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models", "Agent Instructs Large Language Models to be General Zero-Shot Reasoners", "MathPrompter: Mathematical Reasoning using Large Language Models", "Self-Refine: Iterative Refinement with Self-Feedback", "System 2 Attention (is something you might need too)"], "answer_arxiv_id": ["2205.10625", "2310.03710", "2303.05398", "2303.17651", "2311.11829"], "source_meta": {"published_time": "20240705"}, "qid": "AutoScholarQuery_train_1458"} +{"question": "Which works contributed to the recent trend in video diffusion models that use factorizing space and time?", "answer": ["Video Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "MCVD: Masked Conditional Video Diffusion for Prediction, Generation, and\n Interpolation"], "answer_arxiv_id": ["2204.03458", "2210.02303", "2209.14792", "2211.11018", "2304.08818", "2205.09853"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_1459"} +{"question": "Which work theorizes that gradient descent introduces an additional term regularizing the norm of gradients?", "answer": ["Implicit Gradient Regularization"], "answer_arxiv_id": ["2009.11162"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_1460"} +{"question": "Which research papers have proposed the use of a deterministic model to generate an ensemble forecast by perturbing its initial conditions?", "answer": ["FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators"], "answer_arxiv_id": ["2202.11214"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_1461"} +{"question": "What research articles are about semi-supervised methods of image segmentation?", "answer": ["Semi-Supervised Semantic Segmentation with High- and Low-level\n Consistency", "Improving Semantic Segmentation via Self-Training"], "answer_arxiv_id": ["1908.05724", "2004.14960"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_1462"} +{"question": "Which studies adopt an EMA teacher-student structure for continual test time adaptation?", "answer": ["Continual Test-Time Domain Adaptation", "Energy-Based Test Sample Adaptation for Domain Generalization", "A Probabilistic Framework for Lifelong Test-Time Adaptation"], "answer_arxiv_id": ["2203.13591", "2302.11215", "2212.09713"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_1463"} +{"question": "What studies have modified model aggregation to address the non-IID challenge in Federated Learning?", "answer": ["Ensemble Distillation for Robust Model Fusion in Federated Learning", "Clustered Sampling: Low-Variance and Improved Representativity for Clients Selection in Federated Learning", "FedBE: Making Bayesian Model Ensemble Applicable to Federated Learning", "Federated learning with matched averaging"], "answer_arxiv_id": ["2006.07242", "2105.05883", "2009.01974", "2002.06440"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_1464"} +{"question": "What work combined an episodic novelty module and a lifelong novelty module for intrinsic reward generation?", "answer": ["Never Give Up: Learning Directed Exploration Strategies"], "answer_arxiv_id": ["2002.06038"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_1465"} +{"question": "Which works have explored masked generation of content?", "answer": ["MaskGIT: Masked Generative Image Transformer"], "answer_arxiv_id": ["2202.04200"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_1466"} +{"question": "Which studies incorporate the use of EBM as the policy representation in their approach?", "answer": ["Reinforcement Learning with Deep Energy-Based Policies"], "answer_arxiv_id": ["1702.08165"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_1467"} +{"question": "What is the work that explores different search algorithms such as breadth-first and depth-first searches?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2305.10601"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_1468"} +{"question": "What research proposed the use of spectral decomposition of dense DINO features?", "answer": ["Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization"], "answer_arxiv_id": ["2205.07839"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_1469"} +{"question": "What studies focus on local explanation methods within the realm of post-hoc methods in Explainable AI?", "answer": ["How to Explain Individual Classification Decisions", "A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["0912.1128", "1705.07874"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_1470"} +{"question": "Can you name some works that extended the concept of diffusion to 3D head generation?", "answer": ["Rodin: A Generative Model for Sculpting 3D Digital Avatars Using\n Diffusion", "StyleAvatar3D: Leveraging Image-Text Diffusion Models for High-Fidelity\n 3D Avatar Generation"], "answer_arxiv_id": ["2212.06135", "2305.19012"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1471"} +{"question": "Which studies have used extrusion-based techniques in deep-learning-based CAD shape generation?", "answer": ["DeepCAD: A Deep Generative Network for Computer-Aided Design Models", "SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebooks", "ExtrudeNet: Unsupervised Inverse Sketch-and-Extrude for Shape Parsing"], "answer_arxiv_id": ["2105.09492", "2207.04632", "2209.15632"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_1472"} +{"question": "Which work proposed a mechanism to reward agents for having causal influences on others' policies in MARL using mutual information?", "answer": ["Social Influence as Intrinsic Motivation for Multi-Agent Deep Reinforcement Learning"], "answer_arxiv_id": ["1810.08647"], "source_meta": {"published_time": "20230319"}, "qid": "AutoScholarQuery_train_1473"} +{"question": "Can you list the studies that localized concepts to directions in a latent space?", "answer": ["Efficient Estimation of Word Representations in Vector Space", "Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)", "Towards Automatic Concept-based Explanations", "Invertible Concept-based Explanations for CNN Models with Non-negative Concept Activation Vectors", "Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces", "Rewriting a Deep Generative Model", "Editing a classifier by rewriting its prediction rules", "Transformer Feed-Forward Layers Are Key-Value Memories"], "answer_arxiv_id": ["1301.3781", "1711.11279", "1902.03129", "2006.15417", "2212.14855", "2007.15646", "2112.01008", "2012.14913"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_1474"} +{"question": "Which study provides improved results on text tasks by using a mixture of MLE and Reinforce to optimize the non-differentiable rewards?", "answer": ["Sequence Level Training with Recurrent Neural Networks"], "answer_arxiv_id": ["1511.06732"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_1475"} +{"question": "Which research papers found that there are obvious visual artifacts in the GAN-generated images and detected these images from the frequency view?", "answer": ["Detecting and Simulating Artifacts in GAN Fake Images", "Leveraging Frequency Analysis for Deep Fake Image Recognition", "Thinking in Frequency: Face Forgery Detection by Mining Frequency-aware\n Clues"], "answer_arxiv_id": ["1907.06515", "2003.08685", "2007.09355"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_1476"} +{"question": "Can you list studies that propose to leverage context for enriching learned image representations?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Coherent Semantic Attention for Image Inpainting", "Local Relation Networks for Image Recognition", "CAE v2: Context Autoencoder with CLIP Target", "Context Autoencoder for Self-Supervised Representation Learning"], "answer_arxiv_id": ["1505.05192", "1905.12384", "1904.11491", "2211.09799", "2202.03026"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_1477"} +{"question": "What research related to concerns regarding the potential malicious usage of generated images in the context of deep generative models?", "answer": ["Countering Malicious DeepFakes: Survey, Battleground, and Horizon"], "answer_arxiv_id": ["2103.00218"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_1478"} +{"question": "What studies showcased the use of sparse MoEs in speech processing?", "answer": ["SpeechMoE2: MIXTURE-OF-EXPERTS MODEL WITH IMPROVED ROUTING"], "answer_arxiv_id": ["2111.11831"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_1479"} +{"question": "Which works proposed to fuse features from different scales together for image demoiréing?", "answer": ["Moiré Photo Restoration Using Multiresolution Convolutional Neural Networks", "Towards Efficient and Scale-Robust Ultra-High-Definition Image Demoiréing"], "answer_arxiv_id": ["1805.02996", "2207.09935"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_1480"} +{"question": "What recent works demonstrated that PbRL can significantly improve the performance of large-scale language models by finetuning them with human feedback?", "answer": ["Learning to summarize from human feedback", "Recursively Summarizing Books with Human Feedback", "WebGPT: Browser-assisted question-answering with human feedback", "Training language models to follow instructions with human feedback", "Is Reinforcement Learning (Not) for Natural Language Processing: Benchmarks, Baselines, and Building Blocks for Natural Language Policy Optimization", "GPT-4 Technical Report"], "answer_arxiv_id": ["2009.01325", "2109.10862", "2112.09332", "2203.02155", "2210.01241", "2303.08774"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1481"} +{"question": "Which studies are foundational in combining GANs and CLIP for text-based image editing?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "TediGAN: Text-Guided Diverse Face Image Generation and Manipulation", "CLIP2StyleGAN: Unsupervised Extraction of StyleGAN Edit Directions"], "answer_arxiv_id": ["2103.17249", "2108.00946", "2012.03308", "2112.05219"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_1482"} +{"question": "Can you name the reference that defines the minimum verification number which denotes the size of a minimum size verifying set for any DAG G?", "answer": ["Verification and search algorithms for causal DAGs"], "answer_arxiv_id": ["2206.15374"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_1483"} +{"question": "Could you provide me with some studies that focused on defining measures of task affinity for multi-task learning?", "answer": ["Efficiently Identifying Task Groupings for Multi-Task Learning"], "answer_arxiv_id": ["2109.04617"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_1484"} +{"question": "Which studies apply pre-trained 2D foundation models to 3D tasks such as point cloud classification?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP", "Image2Point: 3D Point-Cloud Understanding with 2D Image Pretrained\n Models"], "answer_arxiv_id": ["2112.02413", "2106.04180"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_1485"} +{"question": "Are there any papers about distribution interpolation and its impact on generalizability for pre-training tasks?", "answer": ["Generating Synthetic Datasets by Interpolating along Generalized Geodesics"], "answer_arxiv_id": ["2306.06866"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_1486"} +{"question": "Which papers have proposed the idea of learning point dynamics directly from vision?", "answer": ["RoboCraft: Learning to See, Simulate, and Shape Elasto-Plastic Objects with Graph Networks"], "answer_arxiv_id": ["2205.02909"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_1487"} +{"question": "Could you provide me with some studies that proposed fine-grained bounds in the finite-sum setting?", "answer": ["Stochastic Variance Reduction Methods for Saddle-Point Problems", "Sharper Rates for Separable Minimax and Finite Sum Optimization via Primal-Dual Extragradient Methods"], "answer_arxiv_id": ["1605.06398", "2202.04640"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_1488"} +{"question": "Which works classified as utilizing auto-encoders in unsupervised AD approaches that analyze RGB images?", "answer": ["Improving Unsupervised Defect Segmentation by Applying Structural\n Similarity to Autoencoders", "Divide-and-Assemble: Learning Block-wise Memory for Unsupervised Anomaly\n Detection", "Self-Supervised Predictive Convolutional Attentive Block for Anomaly\n Detection"], "answer_arxiv_id": ["1807.02011", "2107.13118", "2111.09099"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_1489"} +{"question": "Which research papers have contributed to improved performance of ViTs?", "answer": ["Training data-efficient image transformers & distillation through attention", "How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers", "DeiT III: Revenge of the ViT"], "answer_arxiv_id": ["2012.12877", "2106.10270", "2204.07118"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_1490"} +{"question": "What works aimed to address the limitation of computational expenses of per-scene optimization in neural implicit representation?", "answer": ["MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo", "IBRNet: Learning Multi-View Image-Based Rendering", "Point-NeRF: Point-based Neural Radiance Fields", "pixelNeRF: Neural Radiance Fields from One or Few Images", "FWD: Real-time Novel View Synthesis with Forward Warping and Depth"], "answer_arxiv_id": ["2103.15595", "2102.13090", "2201.08845", "2012.02190", "2206.08355"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_1491"} +{"question": "Which methods are known for having a common approach of computing trajectory derivatives by solving an auxiliary affine-quadratic OC problem?", "answer": ["Differentiable MPC for End-to-end Planning and Control", "Pontryagin Differentiable Programming: An End-to-End Learning and Control Framework", "Safe Pontryagin Differentiable Programming"], "answer_arxiv_id": ["1810.13400", "1912.12970v5", "2105.14937"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_1492"} +{"question": "Which papers propose the concept of maximizing behavior diversity in unsupervised pre-training?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function"], "answer_arxiv_id": ["1802.06070"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_1493"} +{"question": "Which work studied the impact of fine-tuning BERT with different random seeds on neural network generalization in NLP?", "answer": ["BERTs of a feather do not generalize together: Large variability in\n generalization across models with similar test set performance"], "answer_arxiv_id": ["1911.02969"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_1494"} +{"question": "What works propose improvements to model update schemes or structures to tackle performance degeneration in non-IID settings in decentralized learning?", "answer": ["Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data", "The Non-IID Data Quagmire of Decentralized Machine Learning", "Decentralized Clustering and Linking by Networked Agents", "Cross-Gradient Aggregation for Decentralized Learning from Non-IID Data"], "answer_arxiv_id": ["2102.04761", "1910.00189v2", "1610.09112v1", "2103.02051"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_1495"} +{"question": "What are the studies that utilized contrastive learning for pre-training point cloud features?", "answer": ["Self-Contrastive Learning with Hard Negative Sampling for Self-supervised Point Cloud Learning", "Spatio-temporal Self-Supervised Representation Learning for 3D Point Clouds", "Info3D: Representation Learning on 3D Objects using Mutual Information Maximization and Contrastive Learning", "PointContrast: Unsupervised Pre-training for 3D Point Cloud Understanding", "Self-Supervised Pretraining of 3D Features on any Point-Cloud", "CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understanding", "Self-supervised Modal and View Invariant Feature Learning", "Disentangling Object Motion and Occlusion for Unsupervised Multi-frame Monocular Depth", "Advancing Self-supervised Monocular Depth Learning with Sparse LiDAR"], "answer_arxiv_id": ["2107.01886", "2109.00179", "2006.02598", "2007.10985", "2101.02691", "2203.00680", "2005.14169", "2203.15174", "2109.09628"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_1496"} +{"question": "Which studies have drawn a parallel between ResNets and Euler’s method?", "answer": ["Beyond Finite Layer Neural Networks: Bridging Deep Architectures and Numerical Differential Equations", "Stable Architectures for Deep Neural Networks", "Deep Neural Networks Motivated by Partial Differential Equations", "Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1710.10121", "1705.03341", "1804.04272", "1806.07366"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_1497"} +{"question": "What studies have proposed numerous forms of calibration in classification?", "answer": ["Reliability, Sufficiency, and the Decomposition of Proper Scores", "Outcome Indistinguishability", "On Fairness and Calibration", "Local Calibration: Metrics and Recalibration"], "answer_arxiv_id": ["0806.0813", "2011.13426", "1709.02012", "2102.10809"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_1498"} +{"question": "Which method leverages pre-defined token probabilities as relevance scores at inference time?", "answer": ["Document Ranking with a Pretrained Sequence-to-Sequence Model"], "answer_arxiv_id": ["2003.06713"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_1499"} +{"question": "What studies classify neurons based on their class selectivity?", "answer": ["Understanding Trained CNNs by Indexing Neuron Selectivity"], "answer_arxiv_id": ["1702.00382"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_1500"} +{"question": "Can you name studies that utilized a hierarchical VAE model for learning both distributed and symbolic representations in object-centric generative modelling?", "answer": ["Generative Neurosymbolic Machines"], "answer_arxiv_id": ["2010.12152"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1501"} +{"question": "What papers proposed nonconformity scores designed for classification?", "answer": ["Uncertainty Sets for Image Classifiers using Conformal Prediction", "Classification with Valid and Adaptive Coverage"], "answer_arxiv_id": ["2009.14193", "2006.02544"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_1502"} +{"question": "Which papers proposed video-compatible language learning models?", "answer": ["VideoChat: Chat-Centric Video Understanding", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models"], "answer_arxiv_id": ["2305.06355", "2306.02858", "2306.05424"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1503"} +{"question": "What is the only method that attempts 3D instance segmentation using 2D foundation models?", "answer": ["OpenMask3D: Open-Vocabulary 3D Instance Segmentation"], "answer_arxiv_id": ["2306.13631"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_1504"} +{"question": "What works studied the struggles of VLMs in understanding complex multi-modal prompts?", "answer": ["MMICL: Empowering Vision-language Model with Multi-Modal In-Context\n Learning"], "answer_arxiv_id": ["2309.07915"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_1505"} +{"question": "Could you provide some papers where machine learning techniques were used to create realistic RIRs?", "answer": ["FAST-RIR: Fast neural diffuse room impulse response generator"], "answer_arxiv_id": ["2110.04057"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_1506"} +{"question": "What studies introduced story synthesis as the task of story visualization?", "answer": ["StoryGAN: A Sequential Conditional GAN for Story Visualization", "Improving Generation and Evaluation of Visual Stories via Semantic\n Consistency", "Integrating Visuospatial, Linguistic and Commonsense Structure into\n Story Visualization", "Word-Level Fine-Grained Story Visualization", "Character-Centric Story Visualization via Visual Planning and Token\n Alignment", "StoryDALL-E: Adapting Pretrained Text-to-Image Transformers for Story\n Continuation", "Synthesizing Coherent Story with Auto-Regressive Latent Diffusion Models", "NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generation", "TaleCrafter: Interactive Story Visualization with Multiple Characters"], "answer_arxiv_id": ["1812.02784", "2105.10026", "2110.10834", "2208.02341", "2210.08465", "2209.06192", "2211.10950", "2303.12346", "2305.18247"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_1507"} +{"question": "Which work proposed a randomized algorithm that leverages smoothness to achieve O(T2/3) expected regret?", "answer": ["Faster Projection-free Online Learning"], "answer_arxiv_id": ["2001.11568"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_1508"} +{"question": "Any research addressing the inpainting problem via approximation of the posterior sampling?", "answer": ["Diffusion Posterior Sampling for General Noisy Inverse Problems"], "answer_arxiv_id": ["2209.14687"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_1509"} +{"question": "Are there any works proposing adaptive and layerwise training methods for avoiding barren plateaus?", "answer": ["An adaptive variational algorithm for exact molecular simulations on a quantum computer", "Layerwise learning for quantum neural networks"], "answer_arxiv_id": ["1812.11173", "2006.14904"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_1510"} +{"question": "What work proposed to replace the original classifier for each client with a fixed ETF classifier to address the data heterogeneity issue in FL?", "answer": ["No Fear of Classifier Biases: Neural Collapse Inspired Federated\n Learning with Synthetic and Fixed Classifier"], "answer_arxiv_id": ["2303.10058"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_1511"} +{"question": "Could you provide me some works about policy gradient with rates to global optimality for tabular and softmax policies?", "answer": ["Global Optimality Guarantees For Policy Gradient Methods", "On the Global Convergence Rates of Softmax Policy Gradient Methods"], "answer_arxiv_id": ["1906.01786", "2005.06392"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_1512"} +{"question": "Could you provide some studies about learning fair models under restricted settings?", "answer": ["Trade-offs and Guarantees of Adversarial Representation Learning for\n Information Obfuscation", "On the Global Optima of Kernelized Adversarial Representation Learning", "On Characterizing the Trade-off in Invariant Representation Learning"], "answer_arxiv_id": ["1906.07902", "1910.07423", "2109.03386"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_1513"} +{"question": "Could you provide me some works exploring sentence fusion in large text units?", "answer": ["Bridging Continuous and Discrete Spaces: Interpretable Sentence\n Representation Learning via Compositional Operations"], "answer_arxiv_id": ["2305.14599"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_1514"} +{"question": "Could you provide me some works that use model-based approaches for reward maximization via planning?", "answer": ["Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Control", "Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models", "Learning Latent Dynamics for Planning from Pixels", "Deep Dynamics Models for Learning Dexterous Manipulation"], "answer_arxiv_id": ["1812.00568", "1805.12114", "1811.04551", "1909.11652"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_1515"} +{"question": "Which works consider knowledge as features extracted from different layers of neural networks?", "answer": ["Paying More Attention to Attention: Improving the Performance of\n Convolutional Neural Networks via Attention Transfer", "Paraphrasing Complex Network: Network Compression via Factor Transfer", "ALP-KD: Attention-Based Layer Projection for Knowledge Distillation"], "answer_arxiv_id": ["1612.03928", "1802.04977", "2012.14022"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_1516"} +{"question": "Could you provide me some examples of the recent work on dealing with spurious correlation in the context of Few-Shot Learning?", "answer": ["Interventional Few-Shot Learning", "Unsupervised Domain Adaptation via Structurally Regularized Deep Clustering", "Cycle Self-Training for Domain Adaptation", "Make the U in UDA Matter: Invariant Consistency Learning for\n Unsupervised Domain Adaptation"], "answer_arxiv_id": ["2009.13000", "2003.08607v1", "2103.03571", "2309.12742"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_1517"} +{"question": "Which works are about improving natural language processing performance by incorporating external knowledge?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Retrieval-Enhanced Contrastive Vision-Text Models"], "answer_arxiv_id": ["2005.11401", "2306.07196"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_1518"} +{"question": "Can you provide me some work on acoustic-style matching of audio?", "answer": ["Unsupervised Speech Decomposition via Triple Information Bottleneck", "Speech Resynthesis from Discrete Disentangled Self-Supervised Representations", "One-shot voice conversion for style transfer based on speaker adaptation"], "answer_arxiv_id": ["2004.11284", "2104.00355", "2111.12277"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_1519"} +{"question": "Which work proposed Denoising Diffusion Implicit Models (DDIMs) to reduce the iteration numbers from the original Denoising Diffusion Probabilistic Models (DDPMs)?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_1520"} +{"question": "Are there any recent research works based on automatic hierarchy learning in the scope of visual contents?", "answer": ["DetCo: Unsupervised Contrastive Learning for Object Detection", "UP-DETR: Unsupervised Pre-training for Object Detection with Transformers", "End-to-End Object Detection with Adaptive Clustering Transformer", "GroupViT: Semantic Segmentation Emerges from Text Supervision", "Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["2102.04803", "2011.09094", "2011.09315", "2202.11094", "2006.15055"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_1521"} +{"question": "Which works proposed new approaches and benchmarks for learning general-purpose representations of sound?", "answer": ["Contrastive Learning of General-Purpose Audio Representations", "BYOL for Audio: Self-Supervised Learning for General-Purpose Audio\n Representation"], "answer_arxiv_id": ["2010.10915", "2103.06695"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_1522"} +{"question": "Which papers have proposed variants of layer normalization configurations for Transformers?", "answer": ["DeepNet: Scaling Transformers to 1,000 Layers", "NormFormer: Improved Transformer Pretraining with Extra Normalization", "Understanding the Difficulty of Training Transformers"], "answer_arxiv_id": ["2203.00555", "2110.09456", "2004.08249"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_1523"} +{"question": "Which studies aim to select the best model by utilizing prior model performances or meta-features of a dataset?", "answer": ["Simple Algorithm Portfolio for SAT", "MetaGL: Evaluation-Free Selection of Graph Learning Models via Meta-Learning"], "answer_arxiv_id": ["1107.0268", "2206.09280"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_1524"} +{"question": "What paper involves separating communicative and non-communicative decision-making using environment actions via minimum entropy coupling in the context of multi-agent reinforcement learning?", "answer": ["Communicating via Markov Decision Processes"], "answer_arxiv_id": ["2107.08295"], "source_meta": {"published_time": "20230319"}, "qid": "AutoScholarQuery_train_1525"} +{"question": "Are there any papers on the semantic enrichment of subject representation in LMs?", "answer": ["Dissecting Recall of Factual Associations in Auto-Regressive Language\n Models"], "answer_arxiv_id": ["2304.14767"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1526"} +{"question": "Could you provide me some studies about improving the efficiency of decoders using recurrence or cross-attention?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Block-Recurrent Transformers", "General-purpose, long-context autoregressive modeling with Perceiver AR"], "answer_arxiv_id": ["1901.02860", "2203.07852", "2202.07765"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_1527"} +{"question": "Are there any studies that optimized the policy using DiCE through re-weighting behavior cloning?", "answer": ["Offline Reinforcement Learning with Realizability and Single-policy Concentrability", "Optimal Conservative Offline RL with General Function Approximation via Augmented Lagrangian", "AlgaeDICE: Policy Gradient from Arbitrary Experience", "OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation"], "answer_arxiv_id": ["2202.04634", "2211.00716", "1912.02074", "2106.10783"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_1528"} +{"question": "Could you provide me some studies that show the need for LLMs to be grounded in knowledge to generate truthful answers?", "answer": ["UnifiedSKG: Unifying and Multi-Tasking Structured Knowledge Grounding with Text-to-Text Language Models", "WebGPT: Browser-assisted question-answering with human feedback"], "answer_arxiv_id": ["2201.05966", "2112.09332"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_1529"} +{"question": "What works utilized natural language to improve RL and explored language abstraction for exploration?", "answer": ["ELLA: Exploration through Learned Language Abstraction", "Semantic Exploration from Language Abstractions and Pretrained Representations", "Improving Intrinsic Exploration with Language Abstractions"], "answer_arxiv_id": ["2103.05825", "2204.05080", "2202.08938"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_1530"} +{"question": "Could you provide me with studies that inspect the notion of sharpness that SAM regulates?", "answer": ["How Does Sharpness-Aware Minimization Minimize Sharpness?"], "answer_arxiv_id": ["2211.05729"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_1531"} +{"question": "Which works proposed faster solvers for the generation speed of diffusion models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in\n Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Elucidating the Design Space of Diffusion-Based Generative Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds", "Fast Sampling of Diffusion Models with Exponential Integrator", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic\n Models", "UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2102.09672", "2010.02502", "2201.06503", "2011.13456", "2206.00364v2", "2202.09778", "2204.13902", "2206.00927", "2211.01095", "2302.04867v4"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1532"} +{"question": "Can you name some studies that inferred beliefs, actions, or instructions as an observer of agent behavior?", "answer": ["Machine Theory of Mind", "BOSS: A Benchmark for Human Belief Prediction in Object-context\n Scenarios"], "answer_arxiv_id": ["1802.07740", "2206.10665"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_1533"} +{"question": "What papers have discussed score-based generative models in the context of graph generation?", "answer": ["Permutation Invariant Graph Generation via Score-Based Generative Modeling", "Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations"], "answer_arxiv_id": ["2003.00638", "2202.02514"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_1534"} +{"question": "Which works have contributed to developing metrics for active learning sample selection?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Minimum-Margin Active Learning", "Adaptive Region-Based Active Learning"], "answer_arxiv_id": ["1708.00489", "1906.00025", "2002.07348"], "source_meta": {"published_time": "20200624"}, "qid": "AutoScholarQuery_train_1535"} +{"question": "Could you provide me some works about pseudo-labeling based learning in semi-supervised learning?", "answer": ["FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["2001.07685v2"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_1536"} +{"question": "What works use machine learning to improve online algorithm performance?", "answer": ["Competitive caching with machine learned advice", "Online Computation with Untrusted Advice", "Optimal Robustness-Consistency Trade-offs for Learning-Augmented Online Algorithms", "(Optimal) Online Bipartite Matching with Degree Information", "Online metric algorithms with untrusted predictions", "Customizing ML Predictions for Online Algorithms"], "answer_arxiv_id": ["1802.05399", "1905.05655", "2010.11443", "2110.11439", "2003.02144v3", "2205.08715"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_1537"} +{"question": "Could you provide some studies that have explored the stochastic bandit setting with heavy-tailed payoffs?", "answer": ["A Finite Time Analysis of Temporal Difference Learning With Linear Function Approximation"], "answer_arxiv_id": ["1806.02450"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_1538"} +{"question": "Which papers have discussed feature disentanglement in image domain?", "answer": ["Learning Disentangled Semantic Representation for Domain Adaptation", "Graph Domain Adaptation: A Generative View", "DIVA: Domain Invariant Variational Autoencoders"], "answer_arxiv_id": ["2012.11807", "2106.07482", "1905.10427"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_1539"} +{"question": "What studies have worked on the development of safer algorithms for learning and exploration in the context of deep RL?", "answer": ["Smoothing Policies and Safe Policy Gradients"], "answer_arxiv_id": ["1905.03231"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_1540"} +{"question": "What research proposes a fine-tuning approach of the diffusion model in the context of subject-driven text-to-image generation?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_1541"} +{"question": "What studies demonstrate the potential of encouraging non-toxic language by manipulating the coefficients of FF neurons in LLMs?", "answer": ["Transformer Feed-Forward Layers Build Predictions by Promoting Concepts\n in the Vocabulary Space"], "answer_arxiv_id": ["2203.14680"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_1542"} +{"question": "Could you point to studies about the applications and popularity of conformal prediction in deep learning?", "answer": ["Learning Optimal Conformal Classifiers", "Training Uncertainty-Aware Classifiers with Conformalized Deep Learning", "Predictive Inference with Feature Conformal Prediction", "Classification with Valid and Adaptive Coverage", "Least Ambiguous Set-Valued Classifiers with Bounded Error Levels"], "answer_arxiv_id": ["2110.09192", "2205.05878", "2210.00173", "2006.02544", "1609.00451"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_1543"} +{"question": "What research involved in the development of FactorVAE architecture?", "answer": ["Disentangling by Factorising"], "answer_arxiv_id": ["1802.05983"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_1544"} +{"question": "Which researchers proposed the text-driven NeRF editing method named Iterative Dataset Update?", "answer": ["Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions"], "answer_arxiv_id": ["2303.12789"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_1545"} +{"question": "Which works have studied fair representation in the context of fairness for algorithms and machine learning?", "answer": ["Fair Clustering Through Fairlets", "Fair Algorithms for Clustering", "On the cost of essentially fair clusterings", "Scalable Fair Clustering", "Clustering without Over-Representation", "Fair Representation Clustering with Several Protected Classes"], "answer_arxiv_id": ["1802.05733", "1901.02393", "1811.10319", "1902.03519", "1905.12753", "2202.01391"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_1546"} +{"question": "Can you identify the research that proposes a multiwavelet-based method for compressing operator kernels?", "answer": ["Multiwavelet-based Operator Learning for Differential Equations"], "answer_arxiv_id": ["2109.13459"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_1547"} +{"question": "Who developed distortion metrics through a contrastive learning approach?", "answer": ["Lossy Compression for Lossless Prediction"], "answer_arxiv_id": ["2106.10800"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_1548"} +{"question": "What papers discuss the method of addressing the limitations of MAP decoding by unbiased sampling?", "answer": ["Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural\n Machine Translation", "Understanding the Properties of Minimum Bayes Risk Decoding in Neural\n Machine Translation", "Sampling-Based Approximations to Minimum Bayes Risk Decoding for Neural\n Machine Translation"], "answer_arxiv_id": ["2005.10283", "2105.08504", "2108.04718"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_1549"} +{"question": "In which work was evidence of power-law scaling of performance with compute found by measuring the performance of AlphaZero agents.", "answer": ["Scaling Scaling Laws with Board Games"], "answer_arxiv_id": ["2104.03113"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_1550"} +{"question": "Can you list studies that utilized interpolation techniques for image generation?", "answer": ["Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?", "Interpreting the Latent Space of GANs for Semantic Face Editing", "Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation", "In-Domain GAN Inversion for Real Image Editing", "Smooth image-to-image translations with latent space interpolations", "Progressive Growing of GANs for Improved Quality, Stability, and Variation", "Fast and Robust Shortest Paths on Manifolds Learned from Data", "Latent Space Oddity: on the Curvature of Deep Generative Models", "A prior-based approximate latent Riemannian metric"], "answer_arxiv_id": ["1904.03189", "1907.10786", "2008.00951", "2004.00049", "2210.00841", "1710.10196", "1901.07229", "1710.11379", "2103.05290"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_1551"} +{"question": "Which papers discuss the task of music recommendation based on video attributes?", "answer": ["Cross-Modal Music-Video Recommendation: A Study of Design Choices", "It's Time for Artistic Correspondence in Music and Video", "Audio-Visual Embedding for Cross-Modal MusicVideo Retrieval through\n Supervised Deep CCA", "Cross-modal Variational Auto-encoder for Content-based Micro-video\n Background Music Recommendation"], "answer_arxiv_id": ["2104.14799", "2206.07148", "1908.03744", "2107.07268"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_1552"} +{"question": "Who first formalized unsupervised environment design (UED) and introduced the PAIRED algorithm?", "answer": ["Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Design"], "answer_arxiv_id": ["2012.02096"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_1553"} +{"question": "Which work recovers the classical mirror-prox method with Optimistic online mirror descent?", "answer": ["Optimization, Learning, and Games with Predictable Sequences"], "answer_arxiv_id": ["1311.1869"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_1554"} +{"question": "Could you provide me some works about the goal of 3D part assembly which is to predict a translation vector and a rotation vector for each part?", "answer": ["Generative 3D Part Assembly via Dynamic Graph Learning", "RGL-NET: A Recurrent Graph Learning framework for Progressive Part\n Assembly", "3D Part Assembly Generation with Instance Encoded Transformer", "Score-PA: Score-based 3D Part Assembly"], "answer_arxiv_id": ["2006.07793", "2107.12859", "2207.01779", "2309.04220"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_1555"} +{"question": "Which works are about language-supervised vision representation learning?", "answer": ["Learning Visual Features from Large Weakly Supervised Data", "Learning Visual N-Grams from Web Data", "YFCC100M: The New Data in Multimedia Research", "Learning Visual Representations with Caption Annotations", "VirTex: Learning Visual Representations from Textual Annotations", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1511.02251", "1612.09161v2", "1503.01817", "2008.01392", "2006.06666", "2103.00020"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_1556"} +{"question": "What work further optimized the implementation of Flash Attention and accelerated the speed of model computation?", "answer": ["FlashAttention-2: Faster Attention with Better Parallelism and Work\n Partitioning"], "answer_arxiv_id": ["2307.08691"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_1557"} +{"question": "Could you mention the works which used Gaussian mixture models for differentiating between clean and noisy labels?", "answer": ["DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "FINE Samples for Learning with Noisy Labels", "Sample Prior Guided Robust Model Learning to Suppress Noisy Labels"], "answer_arxiv_id": ["2002.07394", "2102.11628", "2112.01197"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_1558"} +{"question": "What works demonstrate that transformers cannot robustly model noncounter-free regular languages even when allowing infinite precision?", "answer": ["Theoretical Limitations of Self-Attention in Neural Sequence Models"], "answer_arxiv_id": ["1906.06755"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_1559"} +{"question": "Which work proposed the use of accelerations of particles/meshes to numerical update the respective positions in trying to solve PDEs?", "answer": ["Learning to Simulate Complex Physics with Graph Networks", "Learning Mesh-Based Simulation with Graph Networks", "Boundary Graph Neural Networks for 3D Simulations"], "answer_arxiv_id": ["2002.09405", "2010.03409", "2106.11299"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_1560"} +{"question": "What paper proposes the use of weight modulation method for model fingerprinting?", "answer": ["Responsible Disclosure of Generative Models Using Scalable\n Fingerprinting"], "answer_arxiv_id": ["2012.08726"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_1561"} +{"question": "What research papers propose heuristics to calibrate the probabilistic predictions of KGE models ex-post?", "answer": ["Probability Calibration for Knowledge Graph Embedding Models"], "answer_arxiv_id": ["1912.10000"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_1562"} +{"question": "Any work about turning a LLaMA into an instruction-following model by inserting adapters into the transformer?", "answer": ["LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model"], "answer_arxiv_id": ["2304.15010"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_1563"} +{"question": "What studies worked on analyzing the generalization properties of neural network approximations to PDE solutions?", "answer": ["A Priori Generalization Analysis of the Deep Ritz Method for Solving High Dimensional Elliptic Equations", "A Priori Generalization Error Analysis of Two-Layer Neural Networks for Solving High Dimensional Schrödinger Eigenvalue Problems", "Estimates on the generalization error of Physics Informed Neural Networks (PINNs) for approximating PDEs."], "answer_arxiv_id": ["2101.01708", "2105.01228", "2006.16144"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_1564"} +{"question": "Could you provide me some works that use weakly-supervised pre-training by using hashtag or text descriptions?", "answer": ["Exploring the Limits of Weakly Supervised Pretraining", "Revisiting Weakly Supervised Pre-Training of Visual Perception Models", "Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["1805.00932", "2201.08371", "2103.00020", "2102.05918"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_1565"} +{"question": "What research papers have fine-tuned 1D string generation in molecular design using RL and deep neural networks?", "answer": ["Molecular De-Novo Design through Deep Reinforcement Learning"], "answer_arxiv_id": ["1704.07555"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_1566"} +{"question": "What work is cited for the development and explanation of Denoising Diffusion Implicit Models (DDIMs)?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1567"} +{"question": "Could you provide some works that propose change detection frameworks for anomaly detection in videos?", "answer": ["A Discriminative Framework for Anomaly Detection in Large Videos", "Unmasking the abnormal events in video", "Self-trained Deep Ordinal Regression for End-to-End Video Anomaly\n Detection"], "answer_arxiv_id": ["1609.08938", "1705.08182", "2003.06780"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_1568"} +{"question": "In what researchers propose methods for improving the sampling efficiency of diffusion models by developing numerical solvers?", "answer": ["Denoising Diffusion Implicit Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic\n Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in\n Diffusion Probabilistic Models", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2010.02502", "2206.00927", "2211.01095", "2201.06503", "2206.00364v2"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_1569"} +{"question": "Can you name some works that studied the phenomenon known as 'neural collapse'?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning training"], "answer_arxiv_id": ["2008.08186"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_1570"} +{"question": "Which papers proposed data-matching methods in dataset distillation?", "answer": ["Dataset Condensation with Distribution Matching", "Dataset Condensation with Gradient Matching", "Dataset Condensation with Differentiable Siamese Augmentation", "Delving into Effective Gradient Matching for Dataset Condensation", "Dataset Condensation via Efficient Synthetic-Data Parameterization", "Dataset Distillation by Matching Training Trajectories", "Minimizing the Accumulated Trajectory Error to Improve Dataset Distillation", "Scaling Up Dataset Distillation to ImageNet-1K with Constant Memory", "Dataset Distillation via Factorization", "Remember the Past: Distilling Datasets into Addressable Memories for Neural Networks", "Dataset Condensation with Latent Space Knowledge Factorization and Sharing"], "answer_arxiv_id": ["2110.04181", "2006.05929", "2102.08259", "2208.00311", "2205.14959v2", "2203.11932", "2211.11004", "2211.10586", "2210.16774", "2206.02916", "2208.10494"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_1571"} +{"question": "Which studies focused on leveraging unlabeled data in zero-shot MDE?", "answer": ["Rethinking Pre-training and Self-training", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "End-to-End Semi-Supervised Object Detection with Soft Teacher", "Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic\n Segmentation"], "answer_arxiv_id": ["2006.06882", "2001.07685v2", "2106.09018", "2208.09910"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_1572"} +{"question": "What papers are about joint text, layout and images for document pre-training?", "answer": ["LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document\n Understanding", "LayoutLMv3: Pre-training for Document AI with Unified Text and Image\n Masking", "SelfDoc: Self-Supervised Document Representation Learning", "StructuralLM: Structural Pre-training for Form Understanding", "StrucTexT: Structured Text Understanding with Multi-Modal Transformers", "DocFormer: End-to-End Transformer for Document Understanding", "Unified Pretraining Framework for Document Understanding", "LiLT: A Simple yet Effective Language-Independent Layout Transformer for\n Structured Document Understanding", "XYLayoutLM: Towards Layout-Aware Multimodal Networks For Visually-Rich\n Document Understanding", "BROS: A Pre-trained Language Model Focusing on Text and Layout for\n Better Key Information Extraction from Documents", "StrucTexTv2: Masked Visual-Textual Prediction for Document Image\n Pre-training", "ERNIE-Layout: Layout Knowledge Enhanced Pre-training for Visually-rich\n Document Understanding", "GeoLayoutLM: Geometric Pre-training for Visual Information Extraction", "Vision Grid Transformer for Document Layout Analysis"], "answer_arxiv_id": ["2012.14740", "2204.08387", "2106.03331", "2105.11210", "2108.02923", "2106.11539", "2204.10939", "2202.13669", "2203.06947", "2108.04539", "2303.00289", "2210.06155", "2304.10759", "2308.14978"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_1573"} +{"question": "Which study developed ViL-BERT for vision-language tasks?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks"], "answer_arxiv_id": ["1908.02265"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_1574"} +{"question": "Which papers introduced patch-level representation learning for semantic segmentation?", "answer": ["Unsupervised Learning of Dense Visual Representations", "Dense Contrastive Learning for Self-Supervised Visual Pre-Training", "Propagate Yourself: Exploring Pixel-Level Consistency for Unsupervised\n Visual Representation Learning"], "answer_arxiv_id": ["2011.05499", "2011.09157", "2011.10043"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_1575"} +{"question": "Can you provide studies that discuss the complexities of SSL algorithms with different types of losses and optimization tricks?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Exploring Simple Siamese Representation Learning", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "2103.03230", "2105.04906", "2002.05709", "2011.10566", "1911.05722"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_1576"} +{"question": "Any previous studies have used a cross-modal transformer to re-synthesize the audio?", "answer": ["Visual Acoustic Matching"], "answer_arxiv_id": ["2202.06875"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_1577"} +{"question": "What papers discussed about development of API-based methods that integrate vision APIs with LLMs for solving vision-centric tasks?", "answer": ["Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation\n Models", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face", "MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action", "ViperGPT: Visual Inference via Python Execution for Reasoning", "GPT4Tools: Teaching Large Language Model to Use Tools via\n Self-instruction", "ControlLLM: Augment Language Models with Tools by Searching on Graphs"], "answer_arxiv_id": ["2303.04671", "2303.17580", "2303.11381", "2303.08128", "2305.18752", "2310.17796"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_1578"} +{"question": "What works discuss gradually increasing the window size during pre-training?", "answer": ["XGen-7B Technical Report"], "answer_arxiv_id": ["2309.03450"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_1579"} +{"question": "Which work demonstrates that implicit updates give rise to regret guarantees that depend on the temporal variability of the losses?", "answer": ["Temporal Variability in Implicit Online Learning"], "answer_arxiv_id": ["2006.07503"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_1580"} +{"question": "Could you provide me some studies about vision-language models like CLIP for downstream computer vision tasks?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Attention Is All You Need", "IMITATE: Clinical Prior Guided Hierarchical Vision-Language Pre-training", "CLIP4Caption: CLIP for Video Caption", "CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "CRIS: CLIP-Driven Referring Image Segmentation", "Image Segmentation Using Text and Image Prompts"], "answer_arxiv_id": ["2103.00020", "1706.03762", "2310.07355", "2110.06615", "2106.11097", "2103.17249", "2111.15174", "2112.10003"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_1581"} +{"question": "Which studies use tensor networks as theoretical models for studying deep learning?", "answer": ["On the Expressive Power of Deep Learning: A Tensor Analysis", "Tensorial Mixture Models", "Convolutional Rectifier Networks as Generalized Tensor Decompositions", "Inductive Bias of Deep Convolutional Networks through Pooling Geometry", "On the Expressive Power of Overlapping Architectures of Deep Learning", "Boosting Dilated Convolutional Networks with Mixed Tensor Decompositions", "Deep Learning and Quantum Entanglement: Fundamental Connections with Implications to Network Design", "Expressive power of recurrent neural networks", "Quantum Entanglement in Deep Learning Architectures", "Generalized Tensor Models for Recurrent Neural Networks", "On the Ability of Graph Neural Networks to Model Interactions Between Vertices"], "answer_arxiv_id": ["1509.05009", "1610.04167", "1603.00162", "1605.06743", "1703.02065", "1703.06846v3", "1704.01552v2", "1711.00811", "1803.09780", "1901.10801", "2211.16494"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_1582"} +{"question": "Which works are about the semantic matching models?", "answer": ["Embedding Entities and Relations for Learning and Inference in Knowledge\n Bases", "Complex Embeddings for Simple Link Prediction"], "answer_arxiv_id": ["1412.6575", "1606.06357"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_1583"} +{"question": "Could you provide me with some studies about train-time calibration methods?", "answer": ["A Stitch in Time Saves Nine: A Train-Time Regularizing Loss for Improved Neural Network Calibration", "The Devil is in the Margin: Margin-based Label Smoothing for Network Calibration"], "answer_arxiv_id": ["2203.13834", "2111.15430"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_1584"} +{"question": "What works have extensively studied on scheduling problems with machine-learned advice?", "answer": ["Learning Augmented Energy Minimization via Speed Scaling", "A Novel Prediction Setup for Online Speed-Scaling"], "answer_arxiv_id": ["2010.11629", "2112.03082"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_1585"} +{"question": "What study examined the use of modulating activations of autoencoders for image compression?", "answer": ["Variable Rate Deep Image Compression with Modulated Autoencoder"], "answer_arxiv_id": ["1912.05526"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_1586"} +{"question": "What papers investigate whether reducing catastrophic forgetting is a good objective for continual learning research or should the community shift focus on the forward transfer?", "answer": ["Continual World: A Robotic Benchmark For Continual Reinforcement Learning"], "answer_arxiv_id": ["2105.10919"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_1587"} +{"question": "Any works that improved results of efficient algorithms for linear MDPs in online settings?", "answer": ["First-Order Regret in Reinforcement Learning with Linear Function Approximation: A Robust Estimation Approach", "Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes"], "answer_arxiv_id": ["2112.03432", "2206.11489", "2212.06132"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_1588"} +{"question": "Who defines some specific operations, such as addition and subtraction, to combine existing LoRAs?", "answer": ["Composing Parameter-Efficient Modules with Arithmetic Operations"], "answer_arxiv_id": ["2306.14870"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_1589"} +{"question": "Any research about emergent abilities in deep neural networks?", "answer": ["Are Emergent Abilities of Large Language Models a Mirage?", "The Quantization Model of Neural Scaling"], "answer_arxiv_id": ["2304.15004", "2303.13506v3"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_1590"} +{"question": "Which works have enabled non-humanoid motion learning from a single animation?", "answer": ["Single Motion Diffusion"], "answer_arxiv_id": ["2302.05905"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_1591"} +{"question": "Which works showed that majority voting based on the execution results is effective for candidate selection in code generation?", "answer": ["Competition-Level Code Generation with AlphaCode", "Training Verifiers to Solve Math Word Problems"], "answer_arxiv_id": ["2203.07814", "2110.14168"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_1592"} +{"question": "Could you provide me with studies that proposed to solve ad hoc teamwork in an open environment?", "answer": ["Towards Open Ad Hoc Teamwork Using Graph-based Policy Learning"], "answer_arxiv_id": ["2006.10412"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_1593"} +{"question": "Which works showed that gradient descent can learn single index models when the link function is Lipschitz and monotonic?", "answer": ["Efficient Learning of Generalized Linear and Single Index Models with Isotonic Regression"], "answer_arxiv_id": ["1104.2018"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_1594"} +{"question": "What works studied the ability of neural networks to learn single or multi-index models?", "answer": ["Learning Single-Index Models with Shallow Neural Networks", "Neural Networks can Learn Representations with Gradient Descent", "High-dimensional Asymptotics of Feature Learning: How One Gradient Step Improves the Representation", "The merged-staircase property: a necessary and nearly sufficient condition for SGD learning of sparse functions on two-layer neural networks", "SGD learning on neural networks: leap complexity and saddle-to-saddle dynamics"], "answer_arxiv_id": ["2210.15651", "2206.15144", "2205.01445", "2202.08658", "2302.11055"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_1595"} +{"question": "Any works about implementing LSTM, GRUs, and two model-based memory models?", "answer": ["Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPs"], "answer_arxiv_id": ["2110.05038"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_1596"} +{"question": "Which studies employ neural rendering to separate scenes into objects and background either without or with weak supervisory signals?", "answer": ["NeRF-SOS: Any-View Self-supervised Object Segmentation on Complex Scenes", "FiG-NeRF: Figure-Ground Neural Radiance Fields for 3D Object Category Modelling", "Unsupervised Discovery of Object Radiance Fields", "NeuralDiff: Segmenting 3D objects that move in egocentric videos", "Neural Groundplans: Persistent Neural Scene Representations from a Single Image", "Neural Scene Graphs for Dynamic Scenes", "LaTeRF: Label and Text Driven Object Radiance Fields", "D2NeRF: Self-Supervised Decoupling of Dynamic and Static Objects from a Monocular Video"], "answer_arxiv_id": ["2209.08776", "2104.08418", "2107.07905", "2110.09936", "2207.11232", "2011.10379", "2207.01583", "2205.15838"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_1597"} +{"question": "What works proposed different techniques to address the problem of inferring dynamical systems based on noisy or incomplete observations?", "answer": ["Data-driven discovery of partial differential equations", "Machine Learning of Linear Differential Equations using Gaussian Processes", "Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations", "Physics-Informed Generative Adversarial Networks for Stochastic Differential Equations", "NSFnets (Navier-Stokes Flow nets): Physics-informed neural networks for the incompressible Navier-Stokes equations", "Physics Informed Neural Fields for Smoke Reconstruction with Sparse Data"], "answer_arxiv_id": ["1609.06401", "1701.02440", "1708.00588", "1811.02033", "2003.06496", "2206.06577"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_1598"} +{"question": "Which works proposed point cloud generation techniques?", "answer": ["Point Cloud GAN", "3D Point Cloud Generative Adversarial Network Based on Tree Structured\n Graph Convolutions", "Learning Representations and Generative Models for 3D Point Clouds", "PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "Learning Gradient Fields for Shape Generation", "3D Shape Generation and Completion through Point-Voxel Diffusion", "LION: Latent Point Diffusion Models for 3D Shape Generation", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts"], "answer_arxiv_id": ["1810.05795", "1905.06292", "1707.02392", "1906.12320", "2008.06520", "2104.03670", "2210.06978", "2212.08751"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_1599"} +{"question": "Which works propose achieving group fairness using fair regularization and adversarial debias methods?", "answer": ["Fair Mixup: Fairness via Interpolation", "Learning to Pivot with Adversarial Networks"], "answer_arxiv_id": ["2103.06503", "1611.01046"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_1600"} +{"question": "What studies exist that apply masked autoencoders to learn tactile representations directly from tactile inputs?", "answer": ["Learn from Incomplete Tactile Data: Tactile Representation Learning with\n Masked Autoencoders"], "answer_arxiv_id": ["2307.07358"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_1601"} +{"question": "Which works discuss distilling the knowledge of data distribution to other models?", "answer": ["A Comprehensive Survey on Knowledge Distillation of Diffusion Models"], "answer_arxiv_id": ["2304.04262"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_1602"} +{"question": "What works have shown good performance by using consistency losses between pseudo-labels of different inputs?", "answer": ["Interpolation Consistency Training for Semi-Supervised Learning", "SimPLE: Similar Pseudo Label Exploitation for Semi-Supervised Classification"], "answer_arxiv_id": ["1903.03825", "2103.16725"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_1603"} +{"question": "What are some studies that have previously attempted hindsight optimization?", "answer": ["Scalable Deep Reinforcement Learning for Ride-Hailing"], "answer_arxiv_id": ["2009.14679"], "source_meta": {"published_time": "20220713"}, "qid": "AutoScholarQuery_train_1604"} +{"question": "Which paper on Transformers shows it can universally approximate arbitrary continuous sequence-to-sequence functions?", "answer": ["Are Transformers universal approximators of sequence-to-sequence functions?"], "answer_arxiv_id": ["1912.10077"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_1605"} +{"question": "Which works proposed intrinsic motivation-based exploration by adding an additional bonus, called intrinsic reward, to the extrinsic reward?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "Surprise-Based Intrinsic Motivation for Deep Reinforcement Learning", "Exploration by Random Network Distillation"], "answer_arxiv_id": ["1606.01868", "1703.01732", "1810.12894"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_1606"} +{"question": "Which works used Influence Functions as a tool for tracing model behaviour on examples to the training examples?", "answer": ["Understanding Black-box Predictions via Influence Functions", "FastIF: Scalable Influence Functions for Efficient Model Interpretation and Debugging"], "answer_arxiv_id": ["1703.04730", "2012.15781"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_1607"} +{"question": "Which studies proposed a method for a robust ICL setup and what is their approach?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?", "Learning To Retrieve Prompts for In-Context Learning"], "answer_arxiv_id": ["2101.06804", "2112.08633"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_1608"} +{"question": "What paper introduced ORIL, a method to learn reward function and use it to relabel offline trajectories?", "answer": ["Offline Learning from Demonstrations and Unlabeled Experience"], "answer_arxiv_id": ["2011.13885"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_1609"} +{"question": "Which works obtained polynomial sample complexity results under observability assumptions?", "answer": ["Sample-Efficient Reinforcement Learning of Undercomplete POMDPs", "When Is Partially Observable Reinforcement Learning Not Scary?"], "answer_arxiv_id": ["2006.12484", "2204.08967"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_1610"} +{"question": "Are there any papers related to image-based methods in Active Learning for semantic segmentation?", "answer": ["Suggestive Annotation: A Deep Active Learning Framework for Biomedical Image Segmentation", "Variational Adversarial Active Learning", "Suggestive Annotation of Brain Tumour Images with Gradient-guided Sampling"], "answer_arxiv_id": ["1706.04737", "1904.00370", "2006.14984"], "source_meta": {"published_time": "20230917"}, "qid": "AutoScholarQuery_train_1611"} +{"question": "What papers introduced randomness in the input space to produce more diverse outputs for voting in CoT reasoning?", "answer": ["Rationale-Augmented Ensembles in Language Models"], "answer_arxiv_id": ["2207.00747"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_1612"} +{"question": "Which works proposed specific learning objectives for disentangling factors or properties of the environment?", "answer": ["Causal Curiosity: RL Agents Discovering Self-supervised Experiments for Causal Representation Learning", "Independently Controllable Factors"], "answer_arxiv_id": ["2010.03110", "1708.01289"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_1613"} +{"question": "What works discuss the use of Transfer Learning, Self-Supervised Learning and Zero-Shot Learning techniques in the domain of audio processing?", "answer": ["Contrastive learning of general-purpose audio representations", "Towards Learning a Universal Non-Semantic Representation of Speech", "BYOL for Audio: Self-Supervised Learning for General-Purpose Audio Representation", "Masked Autoencoders that Listen", "HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations", "CLAP : Learning Audio Concepts From Natural Language Supervision", "MuLan: A Joint Embedding of Music Audio and Natural Language", "Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation", "WavCaps: A ChatGPT-Assisted Weakly-Labelled Audio Captioning Dataset for Audio-Language Multimodal Research", "Describing Emotions with acoustic property prompts for Speech Emotion Recognition", "NaRLE: Natural Language Models using Reinforcement Learning with Emotion Feedback"], "answer_arxiv_id": ["2010.10915", "2002.12764", "2103.06695", "2207.06405", "2106.07447", "2006.11477", "2206.04769", "2208.12415", "2211.06687", "2303.17395", "2211.07737", "2110.02148"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_1614"} +{"question": "Which studies have discussed the use of LLMs such as GPT-4 for action planning in interactive tasks?", "answer": ["On Grounded Planning for Embodied Tasks with Language Models", "Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents", "LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Models", "Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents"], "answer_arxiv_id": ["2209.00465v3", "2201.07207", "2212.04088", "2302.01560v2"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_1615"} +{"question": "What research has been done to generate 3D representations that adhere to a text prompt?", "answer": ["Understanding Pure CLIP Guidance for Voxel Grid NeRF Models", "TextDeformer: Geometry Manipulation using Text Guidance", "Zero-Shot Text-Guided Object Generation with Dream Fields", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "ShapeCrafter: A Recursive Text-Conditioned 3D Shape Generation Model", "CLIP-Mesh: Generating textured meshes from text using pretrained\n image-text models", "LION: Latent Point Diffusion Models for 3D Shape Generation", "HyperFields: Towards Zero-Shot Generation of NeRFs from Text"], "answer_arxiv_id": ["2209.15172", "2304.13348", "2112.01455", "2112.03221", "2207.09446", "2203.13333", "2210.06978", "2310.17075"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_1616"} +{"question": "What papers studied visual search and exploration in Image Retrieval by engaging human feedback?", "answer": ["Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language Feedback", "Dialog-based Interactive Image Retrieval"], "answer_arxiv_id": ["1905.12794", "1805.00145"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_1617"} +{"question": "Can you provide me some studies that used unsupervised losses such as agreement among multiple classifiers and data augmentation for OOD error prediction?", "answer": ["Assessing Generalization of SGD via Disagreement", "Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach", "What Does Rotation Prediction Tell Us about Classifier Accuracy under Varying Testing Environments?"], "answer_arxiv_id": ["2106.13799", "1705.07086", "2106.05961"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_1618"} +{"question": "Could you provide me some studies about the empirical examination of the contrast between rich and lazy networks?", "answer": ["Deep learning versus kernel learning: an empirical study of loss landscape geometry and the time evolution of the Neural Tangent Kernel", "What can linearized neural networks actually say about generalization?", "Disentangling feature and lazy training in deep neural networks"], "answer_arxiv_id": ["2010.15110", "2106.06770", "1906.08034"], "source_meta": {"published_time": "20221223"}, "qid": "AutoScholarQuery_train_1619"} +{"question": "What works employed SEMs to mitigate the issues of Granger causality?", "answer": ["DYNOTEARS: Structure Learning from Time-Series Data", "DAGs with NO TEARS: Continuous Optimization for Structure Learning"], "answer_arxiv_id": ["2002.00498v2", "1803.01422"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_1620"} +{"question": "Which works initiated high-dimensional image generation in generative modeling?", "answer": ["Generative Adversarial Networks", "Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1406.2661", "1312.6114"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_1621"} +{"question": "Can you indicate which papers discuss computing methods for the entropic OT plan between discrete distributions with large support sizes?", "answer": ["Fast Computation of Wasserstein Barycenters"], "answer_arxiv_id": ["1310.4375"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_1622"} +{"question": "Are there works that deal with offline policy evaluation or offline hyperparameter selection?", "answer": ["Hyperparameter Selection for Offline Reinforcement Learning", "Active Offline Policy Selection", "Autoregressive Dynamics Models for Offline Policy Evaluation and Optimization", "Benchmarks for Deep Off-Policy Evaluation"], "answer_arxiv_id": ["2007.09055", "2106.10251", "2104.13877", "2103.16596"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_1623"} +{"question": "Are there any citations for applying diffusion models in text-to-3D computer vision tasks?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Text-To-4D Dynamic Scene Generation"], "answer_arxiv_id": ["2209.14988", "2301.11280"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_1624"} +{"question": "Could you provide me with references on extending PAC theory via strategic VC analysis?", "answer": ["PAC-Learning for Strategic Classification"], "answer_arxiv_id": ["2012.03310"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_1625"} +{"question": "What work discusses a more efficient reinforcement learning method for NCO model training?", "answer": ["Neural Combinatorial Optimization with Reinforcement Learning"], "answer_arxiv_id": ["1611.09940"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_1626"} +{"question": "What research works focused on label flipping as a type of attack in targeted data poisoning?", "answer": ["Label Sanitization against Label Flipping Poisoning Attacks"], "answer_arxiv_id": ["1803.00992"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_1627"} +{"question": "What papers present a method to decouple foreground things and background stuff with separate decoders to alleviate granularity discrepancy?", "answer": ["A Simple Framework for Open-Vocabulary Segmentation and Detection", "Hierarchical Open-vocabulary Universal Image Segmentation"], "answer_arxiv_id": ["2303.08131", "2307.00764"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1628"} +{"question": "What papers redesigned the architecture of intrinsic-curiosity-module (ICM) using a variational auto-encoder?", "answer": ["Intrinsic Reward Driven Imitation Learning via Generative Model", "Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["2006.15061", "1312.6114"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_1629"} +{"question": "What research presents datasets like ABO, GSO and OmniObjects3D that improved the texture quality of their CAD models?", "answer": ["ABO: Dataset and Benchmarks for Real-World 3D Object Understanding", "Google Scanned Objects: A High-Quality Dataset of 3D Scanned Household Items", "OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic Perception, Reconstruction and Generation"], "answer_arxiv_id": ["2110.06199", "2204.11918", "2301.07525"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_1630"} +{"question": "What works have proposed synthesizing examples for minority groups using GAN for data augmentation?", "answer": ["Model Patching: Closing the Subgroup Performance Gap with Data Augmentation"], "answer_arxiv_id": ["2008.06775"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_1631"} +{"question": "Are there any works that found popular methods of uncertainty estimation inadequate for providing uncertainty estimates on OOD samples?", "answer": ["Uncertainty Estimation Using a Single Deep Deterministic Neural Network", "The Peril of Popular Deep Learning Uncertainty Estimation Methods"], "answer_arxiv_id": ["2003.02037", "2112.05000"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_1632"} +{"question": "Could you provide me some studies about optimizing adapters for downstream tasks by maintaining a set of adapters and combine them during inference?", "answer": ["K-Adapter: Infusing Knowledge into Pre-Trained Models with Adapters", "AdapterFusion: Non-Destructive Task Composition for Transfer Learning", "MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer", "AdaMix: Mixture-of-Adaptations for Parameter-efficient Model Tuning"], "answer_arxiv_id": ["2002.01808", "2005.00247", "2005.00052", "2205.12410"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_1633"} +{"question": "Which study introduces Imagic, a method that optimizes a pretrained text-to-image diffusion model in image-to-image translation?", "answer": ["Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2210.09276"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_1634"} +{"question": "Which works introduce general text embedding models?", "answer": ["Unsupervised Dense Information Retrieval with Contrastive Learning", "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", "Text and Code Embeddings by Contrastive Pre-Training", "C-Pack: Packaged Resources To Advance General Chinese Embedding", "BGE M3-Embedding: Multi-Lingual, Multi-Functionality, Multi-Granularity\n Text Embeddings Through Self-Knowledge Distillation"], "answer_arxiv_id": ["2112.09118", "1908.10084", "2201.10005", "2309.07597", "2402.03216"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_1635"} +{"question": "What works have been done in contrastive learning of visual and graph representations?", "answer": ["Learning deep representations by mutual information estimation and maximization", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Graph Contrastive Learning with Augmentations"], "answer_arxiv_id": ["1808.06670", "1911.05722", "2002.05709", "2010.13902"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_1636"} +{"question": "Could you mention papers that propose offline test methods in OMRL?", "answer": ["Multi-task Batch Reinforcement Learning with Metric Learning", "FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization"], "answer_arxiv_id": ["1909.11373", "2010.01112"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_1637"} +{"question": "Which works propose benchmarks for multi-task language understanding that also include tasks like reading comprehension and natural language inference?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding", "SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems"], "answer_arxiv_id": ["1804.07461", "1905.00537"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_1638"} +{"question": "What study developed Sparse Inertial Poser for offline joint angle reconstruction?", "answer": ["Sparse Inertial Poser: Automatic 3D Human Pose Estimation from Sparse\n IMUs"], "answer_arxiv_id": ["1703.08014"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_1639"} +{"question": "Could you tell me more about the studies which use multiple positives for combating noisy anchor-positive alignment?", "answer": ["End-to-End Learning of Visual Representations from Uncurated Instructional Videos"], "answer_arxiv_id": ["1912.06430"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_1640"} +{"question": "Could you provide some works that introduced new methods by considering heterogeneity caused by class imbalance?", "answer": ["Re-thinking Federated Active Learning based on Inter-class Diversity", "Knowledge-Aware Federated Active Learning with Non-IID Data"], "answer_arxiv_id": ["2303.12317", "2211.13579"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_1641"} +{"question": "Are there any studies that estimated the value of individual data instances with a measure known as consistency score or C-score?", "answer": ["Characterizing Structural Regularities of Labeled Data in Overparameterized Models"], "answer_arxiv_id": ["2002.03206"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_1642"} +{"question": "Which papers provide insight into in-context learning, including its theoretical foundations?", "answer": ["Language Models are Few-Shot Learners", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "Large Language Models are Zero-Shot Reasoners", "Beyond the Imitation Game: Quantifying and extrapolating the\n capabilities of language models", "Larger language models do in-context learning differently", "Why Can GPT Learn In-Context? Language Models Implicitly Perform\n Gradient Descent as Meta-Optimizers", "What learning algorithm is in-context learning? Investigations with\n linear models", "Transformers as Algorithms: Generalization and Stability in In-context\n Learning", "Trained Transformers Learn Linear Models In-Context"], "answer_arxiv_id": ["2005.14165", "2107.13586v1", "2205.11916", "2206.04615", "2303.03846v2", "2212.10559", "2211.15661", "2301.07067", "2306.09927"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_1643"} +{"question": "Can you provide me with some studies that presented significant innovations in network architectures for monocular depth estimation?", "answer": ["From Big to Small: Multi-Scale Local Planar Guidance for Monocular Depth Estimation", "Vision Transformers for Dense Prediction", "MonoViT: Self-Supervised Monocular Depth Estimation with a Vision Transformer", "Lite-Mono: A Lightweight CNN and Transformer Architecture for Self-Supervised Monocular Depth Estimation", "Self-supervised Monocular Trained Depth Estimation using Self-attention and Discrete Disparity Volume"], "answer_arxiv_id": ["1907.10326", "2103.13413", "2208.03543", "2211.13202", "2003.13951"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_1644"} +{"question": "Could you provide me some studies that propose a dynamic parameter update strategy for different rank settings in LoRA?", "answer": ["DyLoRA: Parameter Efficient Tuning of Pre-trained Models using Dynamic\n Search-Free Low-Rank Adaptation"], "answer_arxiv_id": ["2210.07558"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1645"} +{"question": "Which paper first introduces Generative Adversarial Network (GAN)?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["2203.00667"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_1646"} +{"question": "What research utilized knowledge-enhanced methods for comprehending long context and improving generation quality?", "answer": ["FiD-Light: Efficient and Effective Retrieval-Augmented Text Generation", "In-Context Retrieval-Augmented Language Models", "Atlas: Few-shot Learning with Retrieval Augmented Language Models", "Relational Memory Augmented Language Models"], "answer_arxiv_id": ["2209.14290", "2302.00083", "2208.03299", "2201.09680"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_1647"} +{"question": "Could you list some studies that provided RL environments for combinatorial optimization problems?", "answer": ["Jumanji: a Diverse Suite of Scalable Reinforcement Learning Environments in JAX"], "answer_arxiv_id": ["2306.09884"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_1648"} +{"question": "What studies focus on estimating the expected calibration errors (ECE) for different datasets?", "answer": ["Calibrated Language Model Fine-Tuning for In- and Out-of-Distribution Data", "How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answering", "Teaching models to express their uncertainty in words"], "answer_arxiv_id": ["2010.11506", "2012.00955", "2205.14334"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_1649"} +{"question": "Are there any papers where priors have been used in the multi-task literature to guide exploration?", "answer": ["Accelerating Reinforcement Learning with Learned Skill Priors", "Information asymmetry in KL-regularized RL", "Keep Doing What Worked: Behavioral Modelling Priors for Offline Reinforcement Learning", "Distral: Robust Multitask Reinforcement Learning"], "answer_arxiv_id": ["2010.11944", "1905.01240", "2002.08396", "1707.04175"], "source_meta": {"published_time": "20220120"}, "qid": "AutoScholarQuery_train_1650"} +{"question": "Which papers discussed utilizing an estimate of clean sample at each reverse step for modifying the sampling process during image super-resolution tasks", "answer": ["Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "Denoising Diffusion Restoration Models"], "answer_arxiv_id": ["2212.00490", "2201.09865", "2201.11793"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_1651"} +{"question": "What studies demonstrate LLM agents reacting and performing tasks without predefined explicit instructions?", "answer": ["On the Creativity of Large Language Models"], "answer_arxiv_id": ["2304.00008"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_1652"} +{"question": "Which work provides the cue-conflict datasets illustrating robustness against confusing textures that are misaligned with the correct object class?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness"], "answer_arxiv_id": ["1811.12231"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_1653"} +{"question": "Which study mentioned alleviating the latent code problem by choosing a beta value smaller than one?", "answer": ["Fixing a Broken ELBO"], "answer_arxiv_id": ["1711.00464"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_1654"} +{"question": "Could you provide some studies investigating negative texts for general vision & language models?", "answer": ["Teaching Structured Vision&Language Concepts to Vision&Language Models", "CREPE: Can Vision-Language Foundation Models Reason Compositionally?", "When and why vision-language models behave like bags-of-words, and what\n to do about it?", "Winoground: Probing Vision and Language Models for Visio-Linguistic\n Compositionality", "FOIL it! Find One mismatch between Image and Language caption"], "answer_arxiv_id": ["2211.11733", "2212.07796", "2210.01936", "2204.03162", "1705.01359"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_1655"} +{"question": "Could you provide me studies about Superpixel Segmentation using neural networks?", "answer": ["Superpixels: An Evaluation of the State-of-the-Art", "Superpixel Sampling Networks", "Superpixel Segmentation with Fully Convolutional Networks", "Learning the Superpixel in a Non-iterative and Lifelong Manner", "Superpixel Segmentation with Fully Convolutional Networks"], "answer_arxiv_id": ["1612.01601v3", "1807.10174", "2003.12929", "2103.10681", "2003.12929"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_1656"} +{"question": "Which papers discuss vision-language pre-training models with impressive zero-shot transfer abilities?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Florence: A New Foundation Model for Computer Vision"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.11432"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_1657"} +{"question": "Could you provide me some works about multi-task LLMs?", "answer": ["OPT-IML: Scaling Language Model Instruction Meta Learning through the Lens of Generalization", "Finetuned Language Models Are Zero-Shot Learners", "Scaling Instruction-Finetuned Language Models", "Multitask Prompted Training Enables Zero-Shot Task Generalization"], "answer_arxiv_id": ["2212.12017", "2109.01652", "2210.11416", "2110.08207"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_1658"} +{"question": "Which works focus on the utilization of bidirectional language models for obtaining better context-sensitive representation?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "1907.11692"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_1659"} +{"question": "Could you provide me some studies about attention weights in deep learning models?", "answer": ["Visualizing Attention in Transformer-Based Language Representation Models"], "answer_arxiv_id": ["1904.02679"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1660"} +{"question": "Could you provide some papers that leverage spherical harmonics and tensor product to construct equivariant layers for 3D molecular graph representations?", "answer": ["e3nn: Euclidean Neural Networks", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs", "MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields"], "answer_arxiv_id": ["2207.09453", "2006.10503", "2206.11990", "2206.07697"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_1661"} +{"question": "What paper showed that the minimization of convex terms for identifying simultaneously structured objects allows for only near-optimal sampling rates?", "answer": ["Simultaneously Structured Models with Application to Sparse and Low-rank Matrices"], "answer_arxiv_id": ["1212.3753"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_1662"} +{"question": "Which works do the research draw upon to make some theoretical comparisons between DPM-Solver-v3 and existing diffusion ODE solvers that are based on exponential integrators?", "answer": ["Fast Sampling of Diffusion Models with Exponential Integrator", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models", "UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2204.13902", "2206.00927", "2211.01095", "2302.04867"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_1663"} +{"question": "Could you tell me about some works that improved the challenging training process of the Lipschitzness properties architecture?", "answer": ["Boosting the Certified Robustness of L-infinity Distance Nets"], "answer_arxiv_id": ["2110.06850"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_1664"} +{"question": "Could you provide the papers that detail that Row-Hammer Attack method?", "answer": ["TBT: Targeted Neural Network Attack with Bit Trojan"], "answer_arxiv_id": ["1909.05193"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_1665"} +{"question": "Can you name the studies that focus on hand-object interactions as a prevailing aspect in ego-videos for human-worn cameras?", "answer": ["Scaling Egocentric Vision: The EPIC-KITCHENS Dataset", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["1804.02748", "2110.07058"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_1666"} +{"question": "Are there any works that used bivariate EI and MUC while considering the relationship between 𝒙t(1) and 𝒙t(2)?", "answer": ["Preferential Bayesian Optimization", "Efficient Exploration in Binary and Preferential Bayesian Optimization"], "answer_arxiv_id": ["1704.03651", "2110.09361v1"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_1667"} +{"question": "Which works include diffusion model into the encoder-decoder structure through cross-attention mechanisms?", "answer": ["SeqDiffuSeq: Text Diffusion with Encoder-Decoder Transformers", "Text Generation with Diffusion Language Models: A Pre-training Approach with Continuous Paragraph Denoise", "DiNoiSer: Diffused Conditional Sequence Learning by Manipulating Noises"], "answer_arxiv_id": ["2212.10325", "2212.11685", "2302.10025"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_1668"} +{"question": "Could you provide me some works about graph construction details, post-construction adjustments, and query-time parameters for beam search in the context of graph approaches?", "answer": ["Fast Approximate Nearest Neighbor Search With The Navigating Spreading-out Graph", "Optimization of Indexing Based on k-Nearest Neighbor Graph for Proximity Search in High-dimensional Data"], "answer_arxiv_id": ["1707.00143", "1810.07355"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_1669"} +{"question": "What papers were proposed methods to sample viewports along a fixed direction?", "answer": ["Perceptual Quality Assessment of Omnidirectional Images as Moving Camera Videos"], "answer_arxiv_id": ["2005.10547"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_1670"} +{"question": "What papers came up with concatenate-style temporal modeling?", "answer": ["CLIP2Video: Mastering Video-Text Retrieval via Image CLIP"], "answer_arxiv_id": ["2106.11097"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_1671"} +{"question": "What are the papers that focus on one-class classification methods for Anomaly Detection?", "answer": ["Patch SVDD: Patch-level SVDD for Anomaly Detection and Segmentation", "Classification-Based Anomaly Detection for General Data", "Deep One-Class Classification via Interpolated Gaussian Descriptor", "Deep Semi-Supervised Anomaly Detection"], "answer_arxiv_id": ["2006.16067", "2005.02359", "2101.10043", "1906.02694"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1672"} +{"question": "Could you provide me some works about the evaluation of factuality in summarization?", "answer": ["On Faithfulness and Factuality in Abstractive Summarization"], "answer_arxiv_id": ["2005.00661"], "source_meta": {"published_time": "20220801"}, "qid": "AutoScholarQuery_train_1673"} +{"question": "Which research studied the use of different projection heads for each augmentation as part of SSL?", "answer": ["What Should Not Be Contrastive in Contrastive Learning"], "answer_arxiv_id": ["2008.05659"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_1674"} +{"question": "Could you provide some research that used hybrid representations for learning disentangled 3D objects?", "answer": ["EyeNeRF: A Hybrid Representation for Photorealistic Synthesis, Animation\n and Relighting of Human Eyes", "Hybrid Mesh-neural Representation for 3D Transparent Object Reconstruction", "Structural Causal 3D Reconstruction", "Capturing and Animation of Body and Clothing from Monocular Video", "Learning Disentangled Avatars with Hybrid 3D Representations", "Text-Guided Generation and Editing of Compositional 3D Avatars"], "answer_arxiv_id": ["2206.08428", "2203.12613v3", "2207.10156", "2210.01868", "2309.06441", "2309.07125"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1675"} +{"question": "Can you find me a paper that discusses resolving the complexity issues of Neural ODEs with better numerical solvers?", "answer": ["Hypersolvers: Toward Fast Continuous-Depth Models"], "answer_arxiv_id": ["2007.09601"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_1676"} +{"question": "What works helped in devising the Federated Learning method?", "answer": ["A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protection", "Communication-Efficient Learning of Deep Networks from Decentralized Data", "Federated Optimization in Heterogeneous Networks", "Ensemble Distillation for Robust Model Fusion in Federated Learning", "FedAUX: Leveraging Unlabeled Auxiliary Data in Federated Learning", "Data-Free Knowledge Distillation for Heterogeneous Federated Learning", "Architecture Agnostic Federated Learning for Neural Networks"], "answer_arxiv_id": ["1907.09693", "1602.05629", "1812.06127", "2006.07242", "2102.02514", "2105.10056", "2202.07757"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_1677"} +{"question": "Can you mention any works that have used transformer blocks for direct processing of points in point-based methods?", "answer": ["Point Transformer", "Point Transformer V2: Grouped Vector Attention and Partition-based\n Pooling"], "answer_arxiv_id": ["2012.09164", "2210.05666"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_1678"} +{"question": "What are some works on interactive retrieval in the context of fashion image retrieval?", "answer": ["Composed Image Retrieval with Text Feedback via Multi-grained\n Uncertainty Regularization", "Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language\n Feedback", "Fashion++: Minimal Edits for Outfit Improvement"], "answer_arxiv_id": ["2211.07394", "1905.12794", "1904.09261"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_1679"} +{"question": "What papers have provided shared tasks to encourage centralized evaluation and competition?", "answer": ["Findings of the The RuATD Shared Task 2022 on Artificial Text Detection\n in Russian"], "answer_arxiv_id": ["2206.01583"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_1680"} +{"question": "Could you cite the works that developed unique network architectures or training methodologies for deep image segmentation?", "answer": ["Encoder-Decoder with Atrous Separable Convolution for Semantic Image\n Segmentation"], "answer_arxiv_id": ["1802.02611"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_1681"} +{"question": "Which research proposed methods storing samples of past tasks to mitigate forgetting of previous knowledge?", "answer": ["Learning to Learn without Forgetting By Maximizing Transfer and Minimizing Interference", "Online Continual Learning with Maximally Interfered Retrieval", "Dark Experience for General Continual Learning: a Strong, Simple Baseline", "iCaRL: Incremental Classifier and Representation Learning"], "answer_arxiv_id": ["1810.11910", "1908.04742", "2004.07211", "1611.07725"], "source_meta": {"published_time": "20230409"}, "qid": "AutoScholarQuery_train_1682"} +{"question": "Which benchmark has been introduced recently that converts a suite of tasks requiring structured knowledge into text-to-text format?", "answer": ["UnifiedSKG: Unifying and Multi-Tasking Structured Knowledge Grounding with Text-to-Text Language Models"], "answer_arxiv_id": ["2201.05966"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_1683"} +{"question": "Which research found that for linear networks with arbitrarily rank-constrained function space, the squared error loss is special in the sense that it ensures the non-existence of non-global local minima?", "answer": ["Pure and Spurious Critical Points: a Geometric Study of Linear Networks"], "answer_arxiv_id": ["1910.01671"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_1684"} +{"question": "In which studies diffusion models were used to generate trajectories of state-action pairs?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_1685"} +{"question": "What are some studies that employed learning-based methods to solve puzzles with only shifted pieces?", "answer": ["GANzzle: Reframing jigsaw puzzle solving as a retrieval task using a\n generative mental image", "Deepzzle: Solving Visual Jigsaw Puzzles with Deep Learning andShortest\n Path Optimization", "JigsawGAN: Auxiliary Learning for Solving Jigsaw Puzzles with Generative\n Adversarial Networks", "Positional Diffusion: Ordering Unordered Sets with Diffusion\n Probabilistic Models"], "answer_arxiv_id": ["2207.05634", "2005.12548", "2101.07555", "2303.11120"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_1686"} +{"question": "Could you provide me some studies using LiDAR for scene flow estimation?", "answer": ["FlowNet3D: Learning Scene Flow in 3D Point Clouds", "PV-RAFT: Point-Voxel Correlation Fields for Scene Flow Estimation of\n Point Clouds", "Self-Supervised 3D Scene Flow Estimation Guided by Superpoints", "SCOOP: Self-Supervised Correspondence and Optimization-Based Scene Flow"], "answer_arxiv_id": ["1806.01411", "2012.00987", "2305.02528", "2211.14020"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_1687"} +{"question": "Which studies have used contrastive-image-text losses or auto-regressive generation losses in their vision and language pre-training strategies?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "GIT: A Generative Image-to-text Transformer for Vision and Language", "CoCa: Contrastive Captioners are Image-Text Foundation Models"], "answer_arxiv_id": ["2103.00020", "2201.12086", "2107.07651", "2108.10904", "2205.14100", "2205.01917"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_1688"} +{"question": "What studies discuss the vulnerability of graph contrastive learning to adversarial attacks?", "answer": ["Unsupervised Graph Poisoning Attack via Contrastive Loss Back-propagation"], "answer_arxiv_id": ["2201.07986"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_1689"} +{"question": "What studies examined the scaling law in the correlation between model states and human neural data?", "answer": ["Scaling laws for language encoding models in fMRI", "Roles of Scaling and Instruction Tuning in Language Perception: Model\n vs. Human Attention"], "answer_arxiv_id": ["2305.11863", "2310.19084"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_1690"} +{"question": "Can you suggest some studies providing a reference-free or quality estimation metrics for Automatic MT?", "answer": ["TransQuest: Translation Quality Estimation with Cross-lingual\n Transformers", "CometKiwi: IST-Unbabel 2022 Submission for the Quality Estimation Shared\n Task"], "answer_arxiv_id": ["2011.01536", "2209.06243"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_1691"} +{"question": "Who provide a comprehensive review of works on fair allocation and online fair allocation problems?", "answer": ["Fair Division of Indivisible Goods: A Survey", "Online Fair Division: A Survey"], "answer_arxiv_id": ["2202.07551v2", "1911.09488"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_1692"} +{"question": "What works in the field of 3D scene texturing have applied 2D style transfer techniques to 3D domain?", "answer": ["Controlling Perceptual Factors in Neural Style Transfer", "Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "UPST-NeRF: Universal Photorealistic Style Transfer of Neural Radiance\n Fields for 3D Scene", "Stylizing 3D Scene via Implicit Representation and HyperNetwork", "StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D\n Mutual Learning", "ARF: Artistic Radiance Fields", "Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions", "StyleMesh: Style Transfer for Indoor 3D Scene Reconstructions"], "answer_arxiv_id": ["1611.07865", "1603.08155", "2208.07059", "2105.13016", "2205.12183", "2206.06360", "2303.12789", "2112.01530"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_1693"} +{"question": "What specific research adopted the approach of integrating dictionaries into other ASR systems?", "answer": ["Contextual RNN-T For Open Domain ASR", "Two Stage Contextual Word Filtering for Context bias in Unified\n Streaming and Non-streaming Transducer", "CIF-based Collaborative Decoding for End-to-end Contextual Speech\n Recognition"], "answer_arxiv_id": ["2006.03411", "2301.06735", "2012.09466"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_1694"} +{"question": "Which studies initially derived DVS datasets from pre-existing image classification datasets?", "answer": ["Converting Static Image Datasets to Spiking Neuromorphic Datasets Using Saccades"], "answer_arxiv_id": ["1507.07629"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_1695"} +{"question": "What studies about the estimation for non-linear systems involved a known and smooth nonlinearity?", "answer": ["Active Learning for Nonlinear System Identification with Guarantees", "Non-asymptotic and Accurate Learning of Nonlinear Dynamical Systems", "Learning nonlinear dynamical systems from a single trajectory"], "answer_arxiv_id": ["2006.10277", "2002.08538", "2004.14681"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_1696"} +{"question": "Could you provide me papers that propose imbalanced SSL methods utilizing consistency regularization?", "answer": ["Temporal Ensembling for Semi-Supervised Learning", "Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "MixMatch: A Holistic Approach to Semi-Supervised Learning", "ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring", "Self-training with Noisy Student improves ImageNet classification", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Unsupervised Data Augmentation for Consistency Training", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "Dash: Semi-Supervised Learning with Dynamic Thresholding"], "answer_arxiv_id": ["1610.02242v3", "1703.01780", "1905.02249", "1911.09785", "1911.04252", "2001.07685", "1904.12848", "2110.08263", "2109.00650"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_1697"} +{"question": "What research papers discuss the choice of neuro-symbolic methods to combine language models and symbolic reasoning?", "answer": ["Dynamic Neuro-Symbolic Knowledge Graph Construction for Zero-shot Commonsense Question Answering", "NeuralLog: Natural Language Inference with Joint Neural and Logical Reasoning", "Harnessing Deep Neural Networks with Logic Rules", "Maieutic Prompting: Logically Consistent Reasoning with Recursive Explanations", "A Logic-Driven Framework for Consistency of Neural Models"], "answer_arxiv_id": ["1911.03876", "2105.14167", "1603.06318", "2205.11822", "1909.00126"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_1698"} +{"question": "What research studies have extended the guidance for diffusion models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations"], "answer_arxiv_id": ["2103.00020", "2207.06635"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_1699"} +{"question": "Which papers have focused on automatically synthesizing programs for string transformation?", "answer": ["RobustFill: Neural Program Learning under Noisy I/O", "Latent Programmer: Discrete Latent Codes for Program Synthesis", "Hierarchical Neural Program Synthesis"], "answer_arxiv_id": ["1703.07469", "2012.00377", "2303.06018"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1700"} +{"question": "Which papers have shown that language models are vulnerable to adversarial prompts?", "answer": ["On the Robustness of ChatGPT: An Adversarial and Out-of-distribution\n Perspective", "Dr ChatGPT, tell me what I want to hear: How prompt knowledge impacts\n health answer correctness"], "answer_arxiv_id": ["2302.12095", "2302.13793"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_1701"} +{"question": "Which research papers reported designing powerful pre-trained models with ViT-based backbone?", "answer": ["Training data-efficient image transformers & distillation through\n attention", "BEiT: BERT Pre-Training of Image Transformers", "DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object\n Detection"], "answer_arxiv_id": ["2012.12877", "2106.08254", "2203.03605"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_1702"} +{"question": "Which paper introduced the concept of task vectors?", "answer": ["Editing Models with Task Arithmetic"], "answer_arxiv_id": ["2212.04089"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_1703"} +{"question": "Which studies have focused on clean label attacks in case of data poisoning?", "answer": ["MetaPoison: Practical General-purpose Clean-label Data Poisoning", "Witches’ Brew: Industrial Scale Data Poisoning via Gradient Matching"], "answer_arxiv_id": ["2004.00225", "2009.02276"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_1704"} +{"question": "What research uses common data collected by self-driving cars that only acquire noisy and sparse LiDAR points?", "answer": ["nuScenes: A multimodal dataset for autonomous driving", "Scalability in Perception for Autonomous Driving: Waymo Open Dataset"], "answer_arxiv_id": ["1903.11027", "1912.04838"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_1705"} +{"question": "Which early works aimed to manipulate the resulting images of DMs by replacing latent variables?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2108.02938"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_1706"} +{"question": "Which studies overcome the limitation of needing an auxiliary outlier dataset by exploring the synthesis of virtual outliers in the feature space?", "answer": ["VOS: Learning What You Don’t Know by Virtual Outlier Synthesis", "Non-parametric Outlier Synthesis"], "answer_arxiv_id": ["2202.01197", "2303.02966"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_1707"} +{"question": "What research papers discuss including the reject option in the model and optimizing it during the learning phase?", "answer": ["Binary Classification with Bounded Abstention Rate", "SafePredict: A Meta-Algorithm for Machine Learning That Uses Refusals to Guarantee Correctness"], "answer_arxiv_id": ["1905.09561", "1708.06425"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_1708"} +{"question": "Any works that explore SSL in the pretext tasks of jigsaw puzzles and image colorization?", "answer": ["Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles", "Learning Representations for Automatic Colorization"], "answer_arxiv_id": ["1603.09246v3", "1603.06668"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_1709"} +{"question": "Which references used interactions to discuss the difference in information gained by humans and machine learning models?", "answer": ["Discovering and Explaining the Representation Bottleneck of DNNs"], "answer_arxiv_id": ["2111.06236"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_1710"} +{"question": "Which research works have generalized the heat map-based approach for solving large-scale TSP instances?", "answer": ["Generalize a Small Pre-trained Model to Arbitrarily Large TSP Instances", "DIMES: A Differentiable Meta Solver for Combinatorial Optimization Problems"], "answer_arxiv_id": ["2012.10658", "2210.04123"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_1711"} +{"question": "Which papers introduced the concept of knowledge distillation?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_1712"} +{"question": "Could you provide some studies that have introduced reasoning knowledge for comprehensive attention to the user states in the context of dialogue-level models?", "answer": ["CEM: Commonsense-aware Empathetic Response Generation", "CASE: Aligning Coarse-to-Fine Cognition and Affection for Empathetic\n Response Generation", "Improving Empathetic Dialogue Generation by Dynamically Infusing\n Commonsense Knowledge"], "answer_arxiv_id": ["2109.05739", "2208.08845", "2306.04657"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_1713"} +{"question": "Can you provide examples of research works that focus on generating the segmentation mask of the whole object as part of amodal segmentation?", "answer": ["Semantic Amodal Segmentation", "Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers"], "answer_arxiv_id": ["1509.01329", "2103.12340"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_1714"} +{"question": "What research introduced filtering strategies for distinguishing machine-generated and human-written texts?", "answer": ["Time Travel in LLMs: Tracing Data Contamination in Large Language Models", "Stop Uploading Test Data in Plain Text: Practical Strategies for\n Mitigating Data Contamination by Evaluation Benchmarks"], "answer_arxiv_id": ["2308.08493", "2305.10160"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1715"} +{"question": "Which articles focus on the dynamic nature of language causing temporal misalignment?", "answer": ["Time Masking for Temporal Language Models", "Temporal Adaptation of BERT and Performance on Downstream Document\n Classification: Insights from Social Media", "TimeLMs: Diachronic Language Models from Twitter", "Temporal Effects on Pre-trained Models for Language Processing Tasks", "Dynamic Language Models for Continuously Evolving Content"], "answer_arxiv_id": ["2110.06366", "2104.08116", "2202.03829", "2111.12790", "2106.06297"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_1716"} +{"question": "What are the studies proposing to enhance the awareness of heterogeneity in federated self-supervised learning?", "answer": ["Collaborative Unsupervised Visual Representation Learning from\n Decentralized Data", "Divergence-aware Federated Self-Supervised Learning"], "answer_arxiv_id": ["2108.06492", "2204.04385"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_1717"} +{"question": "Which works studied online optimization problems with bandit problems and knapsack constraints?", "answer": ["Bandits with Knapsacks", "Linear Contextual Bandits with Knapsacks"], "answer_arxiv_id": ["1305.2545", "1507.06738"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_1718"} +{"question": "What studies explore human modelling that leverage estimated human pose priors for reconstructing dynamic humans with complex motions?", "answer": ["Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video"], "answer_arxiv_id": ["2012.15838", "2201.04127"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_1719"} +{"question": "What papers describe the use of code-like structures to convert tasks into Python graphs with the Codex LLM for dealing with structured commonsense tasks?", "answer": ["Language Models of Code are Few-Shot Commonsense Learners"], "answer_arxiv_id": ["2210.07128"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_1720"} +{"question": "What are some advanced semantic segmentation models based on the Transformer architecture?", "answer": ["SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2105.15203", "2012.15840", "2112.01527"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_1721"} +{"question": "What works observed the possibility of additional acceleration in finite sum processes when each local function is a sum of elementary functions?", "answer": ["An Optimal Algorithm for Decentralized Finite Sum Optimization", "Dual-Free Stochastic Decentralized Optimization with Variance Reduction", "A principled framework for the design and analysis of token algorithms"], "answer_arxiv_id": ["2005.10675", "2006.14384", "2205.15015"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_1722"} +{"question": "Which studies advocate for partially trained models in distingushing between machine-generated and human-written texts?", "answer": ["Smaller Language Models are Better Black-box Machine-Generated Text\n Detectors"], "answer_arxiv_id": ["2305.09859"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_1723"} +{"question": "What are some works that propose methods for determining precise editing regions while avoiding unnecessary modifications to non-editing regions in a 3D scene?", "answer": ["Vox-E: Text-guided Voxel Editing of 3D Objects", "DreamEditor: Text-Driven 3D Scene Editing with Neural Fields", "RePaint-NeRF: NeRF Editting via Semantic Masks and Diffusion Models"], "answer_arxiv_id": ["2303.12048", "2306.13455", "2306.05668"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1724"} +{"question": "Could you mention some studies that effectively utilized compositional modules to mitigate the negative impact of gradient conflict in model training?", "answer": ["Neural Module Networks", "Learning Modular Neural Network Policies for Multi-Task and Multi-Robot Transfer", "Routing Networks: Adaptive Selection of Non-linear Functions for Multi-Task Learning", "Multi-Task Reinforcement Learning with Soft Modularization"], "answer_arxiv_id": ["1511.02799v4", "1609.07088", "1711.01239", "2003.13661"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_1725"} +{"question": "Are there any works that compare noisy gradient descent with or without momentum?", "answer": ["On the insufficiency of existing momentum schemes for Stochastic Optimization", "Dynamics of Stochastic Momentum Methods on Large-scale, Quadratic Models"], "answer_arxiv_id": ["1803.05591v2", "2106.03696"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_1726"} +{"question": "What papers are dedicated to designing sample efficient algorithms for online RL?", "answer": ["Minimax Regret Bounds for Reinforcement Learning", "Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learning", "Non-Asymptotic Gap-Dependent Regret Bounds for Tabular MDPs", "Towards Tractable Optimism in Model-Based Reinforcement Learning", "Beyond No Regret: Instance-Dependent PAC Reinforcement Learning"], "answer_arxiv_id": ["1703.05449", "1703.07710", "1905.03814", "2006.11911", "2108.02717"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_1727"} +{"question": "What works proposed a bit-pruning MPQ strategy that achieves high compression rate?", "answer": ["BSQ: Exploring Bit-Level Sparsity for Mixed-Precision Neural Network\n Quantization"], "answer_arxiv_id": ["2102.10462"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1728"} +{"question": "Which study introduced the notion of Self-consistency in decision-making by sampling many rationales?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_1729"} +{"question": "What models have been used for image editing tasks that involve manipulating a set of user provided images using text?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_1730"} +{"question": "Which papers are literature reviews about isolated sign language recognition (ISLR)?", "answer": ["Quantitative Survey of the State of the Art in Sign Language Recognition"], "answer_arxiv_id": ["2008.09918v2"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_1731"} +{"question": "Who conducted research about applying reward penalties for state-action pairs with high uncertainty in modifying the learnt MDP?", "answer": ["MOReL: Model-Based Offline Reinforcement Learning", "Revisiting Design Choices in Offline Model-Based Reinforcement Learning", "MOPO: Model-based Offline Policy Optimization"], "answer_arxiv_id": ["2005.05951", "2110.04135", "2005.13239"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_1732"} +{"question": "Which studies have utilized diffusion-based models for image generation from text?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2204.06125"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_1733"} +{"question": "Which paper introduced-equivariant learning for point clouds?", "answer": ["On the Universality of Rotation Equivariant Point Cloud Networks"], "answer_arxiv_id": ["2010.02449"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_1734"} +{"question": "Which papers used text-conditioned image generative models for open-vocabulary image understanding tasks such as classification, detection and segmentation?", "answer": ["Your Diffusion Model is Secretly a Zero-Shot Classifier", "Text-to-Image Diffusion Models are Zero-Shot Classifiers", "OVTrack: Open-Vocabulary Multiple Object Tracking", "DiffusionDet: Diffusion Model for Object Detection", "Open-vocabulary Object Segmentation with Diffusion Models", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models", "Diffuse, Attend, and Segment: Unsupervised Zero-Shot Segmentation using\n Stable Diffusion", "SLiMe: Segment Like Me", "DiffusionSeg: Adapting Diffusion Towards Unsupervised Object Discovery", "Ref-Diff: Zero-shot Referring Image Segmentation with Generative Models", "Diffusion Model is Secretly a Training-free Open Vocabulary Semantic\n Segmenter", "From Text to Mask: Localizing Entities Using the Attention of\n Text-to-Image Diffusion Models", "What the DAAM: Interpreting Stable Diffusion Using Cross Attention"], "answer_arxiv_id": ["2303.16203", "2303.15233", "2304.08408", "2211.09788", "2301.05221", "2303.04803", "2308.12469", "2309.03179", "2303.09813", "2308.16777", "2309.02773", "2309.04109", "2210.04885"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_1735"} +{"question": "Could you provide me some studies about architectural methods in Continual Learning?", "answer": ["Progressive Neural Networks", "Progress & Compress: A scalable framework for continual learning", "PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning", "Piggyback: Adapting a Single Network to Multiple Tasks by Learning to\n Mask Weights", "Supermasks in Superposition", "Lifelong Reinforcement Learning with Modulating Masks", "Sharing Lifelong Reinforcement Learning Knowledge via Modulating Masks"], "answer_arxiv_id": ["1606.04671", "1805.06370", "1711.05769", "1801.06519", "2006.14769", "2212.11110", "2305.10997"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_1736"} +{"question": "Is there any dataset focusing on visual relationships?", "answer": ["Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations"], "answer_arxiv_id": ["1602.07332"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_1737"} +{"question": "Which work proposed an approach for compressing and accelerating BNNs through the use of a Minimum Spanning Tree?", "answer": ["MST-compression: Compressing and Accelerating Binary Neural Networks\n with Minimum Spanning Tree"], "answer_arxiv_id": ["2308.13735"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_1738"} +{"question": "Which papers inspired the current work with their focus on the evaluation of natural language explanations?", "answer": ["Enhancing Ethical Explanations of Large Language Models through\n Iterative Symbolic Refinement", "Do Natural Language Explanations Represent Valid Logical Arguments?\n Verifying Entailment in Explainable NLI Gold Standards", "Teach Me to Explain: A Review of Datasets for Explainable Natural\n Language Processing", "A Survey on Explainability in Machine Reading Comprehension", "Explaining Answers with Entailment Trees", "e-SNLI: Natural Language Inference with Natural Language Explanations"], "answer_arxiv_id": ["2402.00745", "2105.01974", "2102.12060", "2010.00389", "2104.08661", "1812.01193"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_1739"} +{"question": "What are some of the transformer-based methods in Single Image Super Resolution?", "answer": ["SwinIR: Image Restoration Using Swin Transformer", "Restormer: Efficient Transformer for High-Resolution Image Restoration", "Uformer: A General U-Shaped Transformer for Image Restoration"], "answer_arxiv_id": ["2108.10257", "2111.09881", "2106.03106"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_1740"} +{"question": "What work is an improved version of active search that updates a subset of parameters?", "answer": ["Efficient Active Search for Combinatorial Optimization Problems"], "answer_arxiv_id": ["2106.05126"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_1741"} +{"question": "Which works propose methods of learning from observations (LfO) or examples?", "answer": ["Behavioral Cloning from Observation", "Replacing Rewards with Examples: Example-Based Policy Search via Recursive Classification"], "answer_arxiv_id": ["1805.01954", "2103.12656"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_1742"} +{"question": "Which papers have explored unified representation for both generation and perception tasks in foundation models?", "answer": ["Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning", "SegGPT: Segmenting Everything In Context", "InstructDiffusion: A Generalist Modeling Interface for Vision Tasks"], "answer_arxiv_id": ["2212.02499", "2304.03284", "2309.03895"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1743"} +{"question": "What papers mention the latent spaces of StyleGAN used for manipulation like 𝒲 and 𝒮?", "answer": ["StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation"], "answer_arxiv_id": ["2011.12799"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_1744"} +{"question": "What works have utilized Generative Adversarial Networks (GANs) to synthesize extra training samples?", "answer": ["Data Augmentation Generative Adversarial Networks", "DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort", "BigDatasetGAN: Synthesizing ImageNet with Pixel-wise Annotations", "Generative Models as a Data Source for Multiview Representation Learning"], "answer_arxiv_id": ["1711.04340", "2104.06490", "2201.04684", "2106.05258"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_1745"} +{"question": "Could you provide me some studies about autoregressive retrieval for document retrieval?", "answer": ["Transformer Memory as a Differentiable Search Index", "A Neural Corpus Indexer for Document Retrieval", "Autoregressive Search Engines: Generating Substrings as Document Identifiers"], "answer_arxiv_id": ["2202.06991", "2206.02743", "2204.10628"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_1746"} +{"question": "What study combined the Discrete Vortex Method with neural networks?", "answer": ["Neural vortex method: from finite Lagrangian particles to infinite dimensional Eulerian dynamics"], "answer_arxiv_id": ["2006.04178"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_1747"} +{"question": "Could you provide me with studies where hierarchical information is leveraged during training in hierarchical classification?", "answer": ["Exploring Hierarchical Graph Representation for Large-Scale Zero-Shot Image Classification"], "answer_arxiv_id": ["2203.01386"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_1748"} +{"question": "Can you provide some datasets for multi-vehicle cooperative perception research?", "answer": ["V2X-Sim: Multi-Agent Collaborative Perception Dataset and Benchmark for Autonomous Driving", "OPV2V: An Open Benchmark Dataset and Fusion Pipeline for Perception with Vehicle-to-Vehicle Communication", "V2V4Real: A Real-world Large-scale Dataset for Vehicle-to-Vehicle Cooperative Perception"], "answer_arxiv_id": ["2202.08449", "2109.07644", "2303.07601"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_1749"} +{"question": "Could you provide me some studies about volume-based methods for 3D asset generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Zero-Shot Text-Guided Object Generation with Dream Fields", "AvatarCraft: Transforming Text into Neural Human Avatars with\n Parameterized Shape and Pose Control", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "DreamWaltz: Make a Scene with Complex 3D Animatable Avatars", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "DreamTime: An Improved Optimization Strategy for Diffusion-Guided 3D\n Generation"], "answer_arxiv_id": ["2209.14988", "2112.01455", "2303.17606", "2205.08535", "2304.00916", "2305.12529", "2305.16213", "2211.07600", "2212.00774v1", "2306.12422"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_1750"} +{"question": "Can you provide some works that replaced only one part of the PDE with a neural network?", "answer": ["Learning data driven discretizations for partial differential equations", "Augmenting Physical Models with Deep Networks for Complex Dynamics Forecasting", "Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers"], "answer_arxiv_id": ["1808.04930", "2010.04456", "2007.00016"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_1751"} +{"question": "Any references about substantial datasets and methods for aerial-only research in View-homogeneous ReID?", "answer": ["Person Re-identification in Aerial Imagery", "UAV-Human: A Large Benchmark for Human Behavior Understanding with\n Unmanned Aerial Vehicles", "Rotation Invariant Transformer for Recognizing Object in UAVs"], "answer_arxiv_id": ["1908.05024", "2104.00946", "2311.02559"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_1752"} +{"question": "Are there any works that apply projection maintenance data structure to empirical risk minimization?", "answer": ["Solving Empirical Risk Minimization in the Current Matrix Multiplication Time"], "answer_arxiv_id": ["1905.04447"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_1753"} +{"question": "Can you mention any works that have proposed diffusion models for compression?", "answer": ["Autoregressive Diffusion Models", "Lossy Compression with Gaussian Diffusion"], "answer_arxiv_id": ["2110.02037", "2206.08889"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_train_1754"} +{"question": "Which studies have introduced the use of metric learning for training few-shot learning models?", "answer": ["Prototypical Networks for Few-shot Learning"], "answer_arxiv_id": ["1703.05175"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_1755"} +{"question": "Can you point out the studies where curriculum learning is used with auxiliary rewards?", "answer": ["Reinforcement Learning with Unsupervised Auxiliary Tasks", "Situational Fusion of Visual Representation for Visual Navigation"], "answer_arxiv_id": ["1611.05397", "1908.09073"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_1756"} +{"question": "What are some non-contrastive learning (non-CL) methods that use information maximization techniques in self-supervised learning?", "answer": ["Whitening for Self-Supervised Representation Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning"], "answer_arxiv_id": ["2007.06346", "2103.03230", "2105.04906"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_1757"} +{"question": "Could you provide me with studies about Polyak stepsize and its extensions in the stochastic setting?", "answer": ["L4: Practical loss-based stepsize adaptation for deep learning", "Training Neural Networks for and by Interpolation", "Stochastic Gradient Descent with Polyak’s learning rate"], "answer_arxiv_id": ["1802.05074", "1906.05661", "1903.08688"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_1758"} +{"question": "Which papers demonstrated recent uses of soft prompting for large-scale vision-language models?", "answer": ["Learning to Prompt for Vision-Language Models", "A Good Prompt Is Worth Millions of Parameters: Low-resource Prompt-based Learning for Vision-Language Models", "Prompting Visual-Language Models for Efficient Video Understanding", "Domain Adaptation via Prompt Learning", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2109.01134", "2110.08484", "2112.04478", "2202.06687", "2203.05557"], "source_meta": {"published_time": "20220407"}, "qid": "AutoScholarQuery_train_1759"} +{"question": "What studies introduce mathematical problems that come with diagrams as part of visual context?", "answer": ["DVQA: Understanding Data Visualizations via Question Answering", "IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoning", "Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoning", "UniGeo: Unifying Geometry Logical Reasoning via Reformulating Mathematical Expression"], "answer_arxiv_id": ["1801.08163", "2110.13214", "2105.04165", "2212.02746"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_1760"} +{"question": "What publications explored modeling aleatoric uncertainty in classification tasks using a Dirichlet prior or post-training with calibrated loss?", "answer": ["Predictive Uncertainty Estimation via Prior Networks", "Reverse KL-Divergence Training of Prior Networks: Improved Uncertainty and Adversarial Robustness", "Evidential Deep Learning to Quantify Classification Uncertainty", "On Calibration of Modern Neural Networks", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration"], "answer_arxiv_id": ["1802.10501", "1905.13472", "1806.01768", "1706.04599", "1910.12656"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_1761"} +{"question": "What study used the NeRF model's flexibility to learn a descriptor field in addition to the radiance and density information of the map for high precision localization?", "answer": ["CROSSFIRE: Camera Relocalization On Self-Supervised Features from an\n Implicit Representation"], "answer_arxiv_id": ["2303.04869"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_1762"} +{"question": "Which paper's analysis applies to their β-EB-TC algorithm whose empirical stopping times is order of magnitude larger than its competitors for δ=0.01?", "answer": ["Top Two Algorithms Revisited"], "answer_arxiv_id": ["2206.05979"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_1763"} +{"question": "What research addresses the challenge of significant data imbalance in Graph Anomaly Detection (GAD)?", "answer": ["Rethinking Graph Neural Networks for Anomaly Detection"], "answer_arxiv_id": ["2205.15508"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_1764"} +{"question": "Which paper introduced the transformer structure to the object detection task?", "answer": ["End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2005.12872"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_1765"} +{"question": "Which works proposed diverse designs to effectively encode semantic information for semantic segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation", "Learning Deconvolution Network for Semantic Segmentation", "SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image\n Segmentation", "U-Net: Convolutional Networks for Biomedical Image Segmentation", "DeepLab: Semantic Image Segmentation with Deep Convolutional Nets,\n Atrous Convolution, and Fully Connected CRFs", "Multi-Scale Context Aggregation by Dilated Convolutions", "ParseNet: Looking Wider to See Better", "Pyramid Scene Parsing Network", "ICNet for Real-Time Semantic Segmentation on High-Resolution Images", "Dual Attention Network for Scene Segmentation", "CCNet: Criss-Cross Attention for Semantic Segmentation", "Asymmetric Non-local Neural Networks for Semantic Segmentation", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Semi-supervised Semantic Segmentation with Directional Context-aware\n Consistency", "Learning Context-aware Classifier for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038", "1505.04366", "1511.00561", "1505.04597", "1606.00915", "1511.07122", "1506.04579", "1612.01105", "1704.08545", "1809.02983", "1811.11721", "1908.07678", "2107.06278", "2106.14133", "2303.11633"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_1766"} +{"question": "In what papers did they dynamically remove 'redundant' edges according to the downstream task performance on the current sparsified structure?", "answer": ["Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings", "Graph Structure Learning for Robust Graph Neural Networks", "Learning to Drop: Robust Graph Neural Network via Topological Denoising"], "answer_arxiv_id": ["2006.13009", "2005.10203", "2011.07057"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_1767"} +{"question": "Which study introduced another point-based representation for modeling and animating humans wearing loose garments?", "answer": ["Neural Point-based Shape Modeling of Humans in Challenging Clothing"], "answer_arxiv_id": ["2209.06814"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_1768"} +{"question": "Which paper introduced the canonical design presented in SimCLR?", "answer": ["Exploring Contrastive Learning in Human Activity Recognition for Healthcare"], "answer_arxiv_id": ["2011.11542"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_1769"} +{"question": "Which works have conducted empirical studies on ensembles adapted from different SSL models?", "answer": ["No One Representation to Rule Them All: Overlapping Features of Training Methods"], "answer_arxiv_id": ["2110.12899"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_1770"} +{"question": "Which study proposed a solution that demands a driving video for pose modulation?", "answer": ["Pose-Controllable Talking Face Generation by Implicitly Modularized\n Audio-Visual Representation"], "answer_arxiv_id": ["2104.11116"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_1771"} +{"question": "What paper proposed a two-stage fine-tuning solution to align GPT-3 to answer instructions according to human preferences?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_1772"} +{"question": "Which papers are related to the development of transformer-based segmentation models?", "answer": ["Segmenter: Transformer for Semantic Segmentation", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers", "Semantic Segmentation by Early Region Proxy", "K-Net: Towards Unified Image Segmentation"], "answer_arxiv_id": ["2105.05633", "2012.15840", "2203.14043", "2106.14855"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1773"} +{"question": "Which papers discuss the advancement of Reinforcement Learning from human feedback in the context of language models?", "answer": ["Deep Reinforcement Learning from Human Preferences", "Reward learning from human preferences and demonstrations in Atari", "Training language models to follow instructions with human feedback", "WebGPT: Browser-assisted question-answering with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback"], "answer_arxiv_id": ["1706.03741", "1811.06521", "2203.02155", "2112.09332", "2204.05862"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_1774"} +{"question": "Which research works have focused on safe RL intending to optimize the return-based objective function while constraining the safety-related cost function?", "answer": ["Quantile Constrained Reinforcement Learning: A Reinforcement Learning Framework Constraining Outage Probability"], "answer_arxiv_id": ["2211.15034"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_1775"} +{"question": "What are some works that address the parameter-efficiency problem in Transformer-based approaches?", "answer": ["ALBERT: A Lite BERT for Self-supervised Learning of Language\n Representations"], "answer_arxiv_id": ["1909.11942"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_1776"} +{"question": "Which research increased the performance of diffusion models by operating within an autoencoder’s latent space?", "answer": ["Vector Quantized Diffusion Model for Text-to-Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2111.14822", "2112.10752"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_1777"} +{"question": "Are there any papers that examine post-processing approaches to achieve fairness?", "answer": ["Equality of Opportunity in Supervised Learning"], "answer_arxiv_id": ["1610.02413"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_1778"} +{"question": "Could you provide me some studies about maintaining electric flows in graphs?", "answer": ["Faster Maxflow via Improved Dynamic Spectral Vertex Sparsifiers"], "answer_arxiv_id": ["2112.00722v1"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_1779"} +{"question": "Could you provide some examples of previous research where Kashin’s representation is explored for communication efficiency and for local differential privacy?", "answer": ["Expanding the Reach of Federated Learning by Reducing Client Resource Requirements", "Uncertainty Principle for Communication Compression in Distributed and Federated Learning and the Search for an Optimal Compressor", "Statistical Query Algorithms for Mean Vector Estimation and Stochastic Convex Optimization"], "answer_arxiv_id": ["1812.07210", "2002.08958", "1512.09170"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_1780"} +{"question": "Which works are focused on forecasting future motion given context?", "answer": ["Long-term Human Motion Prediction with Scene Context", "Contact-aware Human Motion Forecasting"], "answer_arxiv_id": ["2007.03672", "2210.03954"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_1781"} +{"question": "Which dataset is used for long-tailed detection?", "answer": ["LVIS: A Dataset for Large Vocabulary Instance Segmentation"], "answer_arxiv_id": ["1908.03195"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_1782"} +{"question": "Could you provide me some works focusing on a series of benchmark evaluations focused around factuality and reasoning, but largely neglect open-ended instruction following abilities?", "answer": ["Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2210.11416"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_1783"} +{"question": "Could you provide some studies using learned coordinate frames for point clouds?", "answer": ["Equivariant Point Cloud Analysis via Learning Orientations for Message Passing", "SE(3) Equivariant Graph Neural Networks with Complete Local Frames"], "answer_arxiv_id": ["2203.14486", "2110.14811"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_1784"} +{"question": "Which papers introduced the use of discrete Variational Autoencoders (VAEs) in Text-to-image (T2I) synthesis?", "answer": ["Generating Diverse High-Fidelity Images with VQ-VAE-2", "Taming Transformers for High-Resolution Image Synthesis"], "answer_arxiv_id": ["1906.00446", "2012.09841"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1785"} +{"question": "Who first provided the equivalence between the log-marginal likelihood and exhaustive cross-validation?", "answer": ["On the marginal likelihood and cross-validation"], "answer_arxiv_id": ["1905.08737"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_1786"} +{"question": "Which papers applied regularization techniques to improve generalization under the settings of label noise?", "answer": ["Wasserstein Adversarial Regularization for learning with label noise", "Simple and Effective Regularization Methods for Training on Noisily Labeled Data with Generalization Guarantee", "Peer Loss Functions: Learning from Noisy Labels without Knowing Noise Rates", "Understanding and Improving Early Stopping for Learning with Noisy Labels", "Early-Learning Regularization Prevents Memorization of Noisy Labels", "A Second-Order Approach to Learning with Instance-Dependent Label Noise", "Robust Training under Label Noise by Over-parameterization"], "answer_arxiv_id": ["1904.03936", "1905.11368", "1910.03231", "2106.15853", "2007.00151", "2012.11854", "2202.14026"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_1787"} +{"question": "Could you mention some studies that look into Transformers' ability for in-context learning?", "answer": ["Language Models are Few-Shot Learners", "What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "What learning algorithm is in-context learning? Investigations with linear models", "Transformers Learn In-Context by Gradient Descent", "An Explanation of In-context Learning as Implicit Bayesian Inference", "Transformers as Statisticians: Provable In-Context Learning with In-Context Algorithm Selection", "In-context Learning and Induction Heads"], "answer_arxiv_id": ["2005.14165", "2208.01066", "2211.15661", "2212.07677", "2111.02080", "2306.04637", "2209.11895"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_1788"} +{"question": "What work incorporated the usage of Matryoshka Representations for adaptive representations in ANNS?", "answer": ["Matryoshka Representation Learning"], "answer_arxiv_id": ["2205.13147"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_1789"} +{"question": "Can you provide works that represent a scene as a composition of static objects given their 2D masks?", "answer": ["Learning Object-Compositional Neural Radiance Field for Editable Scene Rendering", "Nerflets: Local Radiance Fields for Efficient Structure-Aware 3D Scene Representation from 2D Supervision"], "answer_arxiv_id": ["2109.01847", "2303.03361"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_1790"} +{"question": "What are some of the references that gave theoretical explanations for the phenomenon of mode connectivity?", "answer": ["Topology and Geometry of Half-Rectified Network Optimization", "Spurious Valleys in Two-layer Neural Network Optimization Landscapes", "On Connected Sublevel Sets in Deep Learning", "A Note on Connectivity of Sublevel Sets in Deep Learning", "Explaining Landscape Connectivity of Low-cost Solutions for Multilayer Nets", "Landscape Connectivity and Dropout Stability of SGD Solutions for Over-parameterized Neural Networks", "When Are Solutions Connected in Deep Networks?"], "answer_arxiv_id": ["1611.01540", "1802.06384v4", "1901.07417", "2101.08576", "1906.06247v2", "1912.10095", "2102.09671"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_1791"} +{"question": "What papers discussed quality-of-service harms in applications?", "answer": ["The Measure and Mismeasure of Fairness", "Modeling Techniques for Machine Learning Fairness: A Survey"], "answer_arxiv_id": ["1808.00023", "2111.03015"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_1792"} +{"question": "What research has been done to recover 6D object pose from crowded scene or densely packed objects from RGB-D input?", "answer": ["Scene-level Pose Estimation for Multiple Instances of Densely Packed Objects", "Recovering 6D Object Pose and Predicting Next-Best-View in the Crowd"], "answer_arxiv_id": ["1910.04953", "1512.07506"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_1793"} +{"question": "Which papers proposed advanced NMS to mitigate overdependence on IoU?", "answer": ["Adaptive NMS: Refining Pedestrian Detection in a Crowd", "Visibility Guided NMS: Efficient Boosting of Amodal Object Detection in Crowded Traffic Scenes", "Relation Networks for Object Detection", "Learning non-maximum suppression"], "answer_arxiv_id": ["1904.03629", "2006.08547", "1711.11575", "1705.02950"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_1794"} +{"question": "What papers made progress in semantic segmentation by leveraging the powerful convolutional features of classification networks?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20211214"}, "qid": "AutoScholarQuery_train_1795"} +{"question": "What research incorporates a scenario where group labels are not known and need to be inferred from the data in the context of group robustness?", "answer": ["Just Train Twice: Improving Group Robustness without Training Group Information", "Increasing Robustness to Spurious Correlations using Forgettable Examples", "Towards Debiasing NLU Models from Unknown Biases", "Environment Inference for Invariant Learning", "A Too-Good-to-be-True Prior to Reduce Shortcut Reliance", "Correct-n-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlations"], "answer_arxiv_id": ["2107.09044", "1911.03861", "2009.12303", "2010.07249", "2102.06406v3", "2203.01517"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_1796"} +{"question": "Could you provide studies that combined the approach with hierarchical priors on likelihood parameters to infer a family of distributions?", "answer": ["Predictive Uncertainty Estimation via Prior Networks", "Deep Evidential Regression"], "answer_arxiv_id": ["1802.10501", "1910.02600"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_1797"} +{"question": "Which papers set the foundation of using Generative Adversarial Network (GAN) in text-to-image (T2I) generation models?", "answer": ["Generative Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661", "1809.11096", "1812.04948"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_1798"} +{"question": "Could you provide me some works discussing considerations of using a subset of the training samples to discretize the reproduction space?", "answer": ["Neural Estimation of the Rate-Distortion Function With Applications to Operational Source Coding"], "answer_arxiv_id": ["2204.01612"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_1799"} +{"question": "What studies have proven that computing equilibria is PPAD-complete even for three/two-player general-sum normal-form games?", "answer": ["Settling the Complexity of Computing Two-Player Nash Equilibria"], "answer_arxiv_id": ["0704.1678"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_1800"} +{"question": "Which study introduces a method of perturbing and aggregating predictions from multiple variations of an input prompt to identify adversarial inputs?", "answer": ["SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks"], "answer_arxiv_id": ["2310.03684"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_1801"} +{"question": "In what papers the image-based relevance annotations are used to identify relevant cue objects in a model’s visual input feature space?", "answer": ["Taking a HINT: Leveraging Explanations to Make Vision and Language\n Models More Grounded", "Self-Critical Reasoning for Robust Visual Question Answering", "Visual Grounding Methods for VQA are Working for the Wrong Reasons!", "VisFIS: Visual Feature Importance Supervision with\n Right-for-the-Right-Reason Objectives"], "answer_arxiv_id": ["1902.03751", "1905.09998", "2004.05704", "2206.11212"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_1802"} +{"question": "Which research papers focus on the topic of Accuracy-on-the-Line?", "answer": ["Do ImageNet Classifiers Generalize to ImageNet?", "Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization"], "answer_arxiv_id": ["1902.10811", "2107.04649"], "source_meta": {"published_time": "20220220"}, "qid": "AutoScholarQuery_train_1803"} +{"question": "Which papers provided a lower bound for auditing Rényi differential privacy?", "answer": ["Lower Bounds for Rényi Differential Privacy in a Black-Box Setting"], "answer_arxiv_id": ["2212.04739"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_1804"} +{"question": "Could you provide me some works that utilized the task of solving jigsaw puzzles in self-supervised learning?", "answer": ["Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles"], "answer_arxiv_id": ["1603.09246v3"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_1805"} +{"question": "Could you provide the work that expanded the reversible network scope from CNNs to Transformers?", "answer": ["Reversible Vision Transformers"], "answer_arxiv_id": ["2302.04869"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_1806"} +{"question": "What works introduced reconstruction-based methods in text-based anomaly detection?", "answer": ["Towards Total Recall in Industrial Anomaly Detection"], "answer_arxiv_id": ["2106.08265"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_1807"} +{"question": "Which works proposed that LLM can utilize its own feedback signal to refine itself?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback", "Teaching Large Language Models to Self-Debug", "Reflexion: Language Agents with Verbal Reinforcement Learning"], "answer_arxiv_id": ["2303.17651", "2304.05128v2", "2303.11366"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_1808"} +{"question": "What are the works about guidance in diffusion models, such as the classifier-free guidance and diffusion in the latent space?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2207.12598", "2112.10741", "2112.10752"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_1809"} +{"question": "Could you provide me some studies that tried to approximate the dot product between query and key matrix by exploiting the low-rank structure of it?", "answer": ["Linformer: Self-Attention with Linear Complexity", "Reformer: The Efficient Transformer", "Nyströmformer: A Nyström-based Algorithm for Approximating Self-Attention", "Big Bird: Transformers for Longer Sequences", "Generating Long Sequences with Sparse Transformers", "Axial Attention In Multidimensional Transformers"], "answer_arxiv_id": ["2006.04768", "2001.04451", "2102.03902", "2007.14062", "1904.10509", "1912.12180"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_1810"} +{"question": "Which papers introduced quantization approaches in communication-efficient algorithms?", "answer": ["signSGD: Compressed Optimisation for Non-Convex Problems", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "TernGrad: Ternary Gradients to Reduce Communication in Distributed Deep Learning"], "answer_arxiv_id": ["1802.04434", "1610.02132", "1705.07878"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_1811"} +{"question": "Which studies were conducted on coreset selection?", "answer": ["Gradient based sample selection for online continual learning", "End-to-End Incremental Learning", "iCaRL: Incremental Classifier and Representation Learning", "Active Learning for Convolutional Neural Networks: A Core-Set Approach", "An Empirical Study of Example Forgetting during Deep Neural Network\n Learning"], "answer_arxiv_id": ["1903.08671", "1807.09536", "1611.07725", "1708.00489", "1812.05159"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_1812"} +{"question": "Whose work involved the learning of latent variable distribution in the context of Neural Processes?", "answer": ["Neural Processes"], "answer_arxiv_id": ["1807.01622"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_1813"} +{"question": "Which research rigorously evaluated variational dropout, l0 regularization, and GMP on large-scale tasks?", "answer": ["The State of Sparsity in Deep Neural Networks"], "answer_arxiv_id": ["1902.09574"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_1814"} +{"question": "Could you provide me the study which talks about the Stochastic Inference Network (SIN)?", "answer": ["Variational Message Passing with Structured Inference Networks"], "answer_arxiv_id": ["1803.05589"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_1815"} +{"question": "Can you provide me with sources that discuss online batch selection methods?", "answer": ["Online Batch Selection for Faster Training of Neural Networks", "Not All Samples Are Created Equal: Deep Learning with Importance Sampling", "Accelerating Deep Learning by Focusing on the Biggest Losers"], "answer_arxiv_id": ["1511.06343", "1803.00942v3", "1910.00762v1"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_1816"} +{"question": "Which publication pointed out that publicly deployed cloud EaaS APIs are vulnerable to imitation attacks?", "answer": ["StolenEncoder: Stealing Pre-trained Encoders in Self-supervised Learning"], "answer_arxiv_id": ["2201.05889"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_1817"} +{"question": "Can you provide research that applies NAS techniques in face recognition systems?", "answer": ["Neural Architecture Search for Deep Face Recognition"], "answer_arxiv_id": ["1904.09523"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_1818"} +{"question": "Which research papers show that generalization depends on the alignment between pairwise data relations with augmentations and downstream task?", "answer": ["Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods"], "answer_arxiv_id": ["2205.11508"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_1819"} +{"question": "What works present the application of deep learning methods in MRI reconstruction?", "answer": ["Results of the 2020 fastMRI Challenge for Machine Learning MR Image\n Reconstruction", "Fill the K-Space and Refine the Image: Prompting for Dynamic and\n Multi-Contrast MRI Reconstruction", "HUMUS-Net: Hybrid unrolled multi-scale network architecture for\n accelerated MRI reconstruction"], "answer_arxiv_id": ["2012.06318", "2309.13839", "2203.08213"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_1820"} +{"question": "Which paper highlights an asymmetric cross-entropy loss (ER-ACE) on incoming samples as a noticeable approach?", "answer": ["New Insights on Reducing Abrupt Representation Change in Online Continual Learning"], "answer_arxiv_id": ["2104.05025"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_1821"} +{"question": "Which works discuss recalibration methods and prediction sets?", "answer": ["On Calibration of Modern Neural Networks", "A Tutorial on Conformal Prediction", "Uncertainty Sets for Image Classifiers using Conformal Prediction"], "answer_arxiv_id": ["1706.04599", "0706.3188", "2009.14193"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_1822"} +{"question": "What works have been done to estimate egocentric 3D human poses from body-worn cameras?", "answer": ["Ego-Body Pose Estimation via Ego-Head Pose Estimation", "Seeing Invisible Poses: Estimating 3D Body Pose from Egocentric Video", "You2Me: Inferring Body Pose in Egocentric Video via First and Second\n Person Interactions", "Ego-Pose Estimation and Forecasting as Real-Time PD Control", "Dynamics-Regulated Kinematic Policy for Egocentric Pose Estimation"], "answer_arxiv_id": ["2212.04636", "1603.07763", "1904.09882", "1906.03173", "2106.05969"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_1823"} +{"question": "What papers have discussed the issue of only editing attention heads while ignoring FFN module during the inference-time?", "answer": ["Inference-Time Intervention: Eliciting Truthful Answers from a Language\n Model", "Transformer Feed-Forward Layers Are Key-Value Memories", "Inspecting and Editing Knowledge Representations in Language Models", "Emergent World Representations: Exploring a Sequence Model Trained on a\n Synthetic Task"], "answer_arxiv_id": ["2306.03341", "2012.14913", "2304.00740", "2210.13382"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_1824"} +{"question": "Which research introduced the Influence Function (IF), a method to quantify the impact of each training instance on a model’s prediction?", "answer": ["Understanding Black-box Predictions via Influence Functions"], "answer_arxiv_id": ["1703.04730"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_1825"} +{"question": "What research relates to the utilization of diffusion models in HSI generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "1907.05600", "2006.11239"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_1826"} +{"question": "Which research papers discussed the generation of toxic content through LLMs?", "answer": ["Toxicity in ChatGPT: Analyzing Persona-assigned Language Models"], "answer_arxiv_id": ["2304.05335"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_1827"} +{"question": "Which paper addresses the importance of language understanding in diffusion models by using a frozen T5 encoder?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2205.11487"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_1828"} +{"question": "What papers are about providing explanations for unsupervised models?", "answer": ["RELAX: Representation Learning Explainability", "Label-Free Explainability for Unsupervised Models"], "answer_arxiv_id": ["2112.10161", "2203.01928"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_1829"} +{"question": "Which references described Transformer-based large language models?", "answer": ["GPT-4 Technical Report", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2303.08774", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_1830"} +{"question": "Which works are related to the notion of strong adaptivity in dynamic OCO on bounded domains?", "answer": ["Strongly Adaptive Online Learning", "Dynamic Regret of Strongly Adaptive Methods", "Optimal Dynamic Regret in Exp-Concave Online Learning", "Optimal Dynamic Regret in Proper Online Learning with Strongly Convex Losses and Beyond"], "answer_arxiv_id": ["1502.07073", "1701.07570", "2104.11824", "2201.08905"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_1831"} +{"question": "Could you provide me some works which developed hateful memes dataset for low-resource languages?", "answer": ["BanglaAbuseMeme: A Dataset for Bengali Abusive Meme Classification", "Multimodal Hate Speech Detection from Bengali Memes and Texts"], "answer_arxiv_id": ["2310.11748", "2204.10196"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_1832"} +{"question": "What are some works that achieved fine-grained image editing based on the feature correspondence?", "answer": ["DragDiffusion: Harnessing Diffusion Models for Interactive Point-based\n Image Editing", "DragonDiffusion: Enabling Drag-style Manipulation on Diffusion Models"], "answer_arxiv_id": ["2306.14435", "2307.02421"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_1833"} +{"question": "Which research has studied the gradient Langevin dynamics in the context of SGD?", "answer": ["On the Noisy Gradient Descent that Generalizes as SGD", "The Anisotropic Noise in Stochastic Gradient Descent: Its Behavior of Escaping from Sharp Minima and Regularization Effects"], "answer_arxiv_id": ["1906.07405", "1803.00195"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_1834"} +{"question": "Can you mention the main studies that followed an optimization workflow to update 3D representations?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "Magic3D: High-Resolution Text-to-3D Content Creation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1", "2211.10440", "2211.07600", "2303.12789", "2303.13873"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_1835"} +{"question": "Which work introduces GCN for missing features (GCNMF) by adapting graph convolutional networks (GCN) to graphs that contain missing node features?", "answer": ["Graph Convolutional Networks for Graphs Containing Missing Features"], "answer_arxiv_id": ["2007.04583"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_1836"} +{"question": "What works introduce advanced numerical SDE or ODE solvers to speed up the backward process in diffusion models?", "answer": ["Gotta Go Fast When Generating Data with Score-Based Models", "GENIE: Higher-Order Denoising Diffusion Solvers", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Elucidating the Design Space of Diffusion-Based Generative Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2105.14080", "2210.05475", "2011.13456", "2206.00364v2", "2202.09778"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_1837"} +{"question": "What works use 3D CAD models for monocular 3D object detection?", "answer": ["AutoShape: Real-Time Shape-Aware Monocular 3D Object Detection"], "answer_arxiv_id": ["2108.11127"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_1838"} +{"question": "Which papers discuss self-training approaches proposed for semi-supervised learning?", "answer": ["Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring"], "answer_arxiv_id": ["1703.01780", "2001.07685v2", "1911.09785"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_1839"} +{"question": "Could you mention the study that proposed a framework called 'FewshotQA' aiming to save the effort of pretraining the model on a large-scale corpus?", "answer": ["FewshotQA: A simple framework for few-shot learning of question\n answering tasks using pre-trained text-to-text models"], "answer_arxiv_id": ["2109.01951"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_1840"} +{"question": "Which paper developed a method for fine-tuning CLIP by optimizing a set of continuous prompt vectors for few-shot image recognition?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_1841"} +{"question": "Can you name the works which attempted to replace Transformers and their attention mechanism with SSMs?", "answer": ["Long Range Language Modeling via Gated State Spaces", "Mega: Moving Average Equipped Gated Attention"], "answer_arxiv_id": ["2206.13947", "2209.10655"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_1842"} +{"question": "In which studies the Decision Transformer has been used?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "The Quality-Diversity Transformer: Generating Behavior-Conditioned Trajectories with Decision Transformers"], "answer_arxiv_id": ["2106.01345", "2303.16207"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_1843"} +{"question": "Who studied fractional delegations in liquid democracy?", "answer": ["Resolving multi-proxy transitive vote delegation"], "answer_arxiv_id": ["1412.4039"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_1844"} +{"question": "What are some works that applied linear models in genomics, reinforcement learning, and pandemic modelling?", "answer": ["Anti-Concentrated Confidence Bonuses for Scalable Exploration"], "answer_arxiv_id": ["2110.11202"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_1845"} +{"question": "Could you provide the studies that proposed to use shallow networks to aggregate information in local contexts defined as character n-grams, byte patches or character blocks?", "answer": ["CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language\n Representation", "MEGABYTE: Predicting Million-byte Sequences with Multiscale Transformers", "Charformer: Fast Character Transformers via Gradient-based Subword\n Tokenization"], "answer_arxiv_id": ["2103.06874", "2305.07185", "2106.12672"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_1846"} +{"question": "Can you cite works that mentioned the Overkill phenomenon but did not conduct in-depth analysis?", "answer": ["Mistral 7B"], "answer_arxiv_id": ["2310.06825"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_1847"} +{"question": "Can you name studies that have proposed to learn distribution over constraints with Bayesian approaches?", "answer": ["Uncertainty-Aware Constraint Learning for Adaptive Safe Motion Planning from Demonstrations"], "answer_arxiv_id": ["2011.04141"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_1848"} +{"question": "Could you provide some studies that applied diffusion process in the latent space to enhance the efficiency of diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_1849"} +{"question": "Could you name the first work that adopted GCN as the feature extractor for skeleton-based action recognition?", "answer": ["Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition"], "answer_arxiv_id": ["1801.07455"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_1850"} +{"question": "What studies focus on the improvement of text descriptions of classes through prompt tuning?", "answer": ["MaPLe: Multi-modal Prompt Learning", "Prompt-aligned Gradient for Prompt Tuning", "Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language\n Models", "Decorate the Newcomers: Visual Domain Prompt for Continual Test Time\n Adaptation", "Visual Prompt Tuning"], "answer_arxiv_id": ["2210.03117", "2205.14865", "2209.07511", "2212.04145", "2203.12119"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_1851"} +{"question": "Who popularized the concept of Forward Gradient?", "answer": ["Gradients without Backpropagation"], "answer_arxiv_id": ["2202.08587"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1852"} +{"question": "Which papers consider 2D image or 3D voxels in projection-based methods for point cloud semantic segmentation?", "answer": ["SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation", "Virtual Multi-view Fusion for 3D Semantic Segmentation", "Progressive LiDAR Adaptation for Road Detection", "Deep Projective 3D Semantic Segmentation", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "O-CNN: Octree-based Convolutional Neural Networks for 3D Shape Analysis", "SEGCloud: Semantic Segmentation of 3D Point Clouds", "OccuSeg: Occupancy-aware 3D Instance Segmentation", "3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "Submanifold Sparse Convolutional Networks", "VoxSegNet: Volumetric CNNs for Semantic Part Segmentation of 3D Shapes"], "answer_arxiv_id": ["2004.01803", "2007.13138", "1904.01206", "1705.03428", "1904.08755", "1712.01537", "1710.07563", "2003.06537v3", "1711.10275", "1706.01307", "1809.00226"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_1853"} +{"question": "Can you list the studies that focus on entropy regularization in multi-player zero-sum games?", "answer": ["Exploration-Exploitation in Multi-Agent Competition: Convergence with Bounded Rationality", "Asynchronous Gradient Play in Zero-Sum Multi-agent Games"], "answer_arxiv_id": ["2106.12928", "2211.08980v1"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_1854"} +{"question": "Which papers proposed a general form of measure-preserving dynamics that can be utilized to construct deterministic Gibbs samplers?", "answer": ["Deterministic Gibbs Sampling via Ordinary Differential Equations"], "answer_arxiv_id": ["2106.10188v1"], "source_meta": {"published_time": "20220516"}, "qid": "AutoScholarQuery_train_1855"} +{"question": "Could you provide me some studies that use fully convolutional networks to predict semantic masks?", "answer": ["Multi-Oriented Text Detection with Fully Convolutional Networks", "Real-time Scene Text Detection with Differentiable Binarization", "Real-Time Scene Text Detection with Differentiable Binarization and\n Adaptive Scale Fusion", "Shape Robust Text Detection with Progressive Scale Expansion Network", "TextSnake: A Flexible Representation for Detecting Text of Arbitrary\n Shapes"], "answer_arxiv_id": ["1604.04018", "1911.08947", "2202.10304", "1903.12473", "1807.01544"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_1856"} +{"question": "Can you name some studies that employed LLMs as agents operating in text domains?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face", "Toolformer: Language Models Can Teach Themselves to Use Tools", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2303.17651", "2303.17580", "2302.04761", "2210.03629"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_1857"} +{"question": "Could you provide me some research that designs more complicated bridge networks to compress or adaptively select visual information?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2204.14198", "2304.10592", "2301.12597", "2305.06500", "2305.03726", "2304.14178"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_1858"} +{"question": "Any works about state-of-the-art ISLR methods' low recognition performance?", "answer": ["Word-level Deep Sign Language Recognition from Video: A New Large-scale Dataset and Methods Comparison", "OpenHands: Making Sign Language Recognition Accessible with Pose-based Pretrained Models across Languages"], "answer_arxiv_id": ["1910.11006", "2110.05877v1"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_1859"} +{"question": "What work analyzes locally-elastic stochastic differential equations and shows the emergence of NC in their solutions?", "answer": ["Imitating Deep Learning Dynamics via Locally Elastic Stochastic Differential Equations"], "answer_arxiv_id": ["2110.05960"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_1860"} +{"question": "What papers have released datasets including self-reported gender and age for fairness annotations?", "answer": ["Towards Measuring Fairness in AI: the Casual Conversations Dataset"], "answer_arxiv_id": ["2104.02821"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_1861"} +{"question": "Can you list some studies that use autoregressive denoising approach in SSL?", "answer": ["Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2111.06377"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_1862"} +{"question": "What studies have deciphered model behavior by examining attention patterns?", "answer": ["What Does BERT Look At? An Analysis of BERT's Attention"], "answer_arxiv_id": ["1906.04341"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_1863"} +{"question": "What studies offer solutions for the optimization issues of weight-sharing NAS through regularization approaches?", "answer": ["K-shot NAS: Learnable Weight-Sharing for NAS with K-shot Supernets", "Powering One-shot Topological NAS with Stabilized Share-parameter Proxy", "You Only Search Once: Single Shot Neural Architecture Search via Direct Sparse Optimization"], "answer_arxiv_id": ["2106.06442", "2005.10511", "1811.01567"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_1864"} +{"question": "Which works extend the computational complexity problem setting to more practical scenarios such as local access?", "answer": ["Efficient Local Planning with Linear Function Approximation"], "answer_arxiv_id": ["2108.05533"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_1865"} +{"question": "Could you point out works that discuss how Transfer Learning accounts for the diversity of geometries and discretizations?", "answer": ["Mosaic Flows: A Transferable Deep Learning Framework for Solving PDEs on Unseen Domains", "Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios", "Deep transfer operator learning for partial differential equations under conditional shift", "Multi-fidelity prediction of fluid flow and temperature field based on transfer learning using Fourier Neural Operator", "Transfer learning based multi-fidelity physics informed deep neural network", "Transfer Learning on Multi-Fidelity Data"], "answer_arxiv_id": ["2104.10873", "2205.07731", "2204.09810", "2304.06972v1", "2005.10614v2", "2105.00856"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_1866"} +{"question": "What papers have studied methods that attempted to approximate object-centric representations?", "answer": ["The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning", "Scale-Localized Abstract Reasoning"], "answer_arxiv_id": ["2007.04212", "2009.09405"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_1867"} +{"question": "What is a study that investigates the best source mix for datasets?", "answer": ["The Flan Collection: Designing Data and Methods for Effective Instruction Tuning", "OPT-IML: Scaling Language Model Instruction Meta Learning through the Lens of Generalization"], "answer_arxiv_id": ["2301.13688", "2212.12017"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_1868"} +{"question": "What studies follow the Gaussian diffusion process of DDPM and leverage an additional image encoder to extract image features as condition to generate masks?", "answer": ["SegDiff: Image Segmentation with Diffusion Probabilistic Models", "MedSegDiff: Medical Image Segmentation with Diffusion Probabilistic Model", "MedSegDiff-V2: Diffusion based Medical Image Segmentation with Transformer", "A Generalist Framework for Panoptic Segmentation of Images and Videos", "DDP: Diffusion Model for Dense Visual Prediction"], "answer_arxiv_id": ["2112.00390", "2211.00611", "2301.11798", "2210.06366", "2303.17559"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_1869"} +{"question": "What works supported the findings that deep neural networks (DNNs) are vulnerable to both artificially-induced adversarial attacks and naturally occurring, non-adversarial corruptions?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["1903.12261"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_1870"} +{"question": "Which papers investigated architectures and training frameworks to enhance audio-visual source separation?", "answer": ["The Sound of Pixels", "Looking to Listen at the Cocktail Party: A Speaker-Independent Audio-Visual Model for Speech Separation", "Learning to Separate Object Sounds by Watching Unlabeled Video", "Recursive Visual Sound Separation Using Minus-Plus Net", "Cyclic Co-Learning of Sounding Object Visual Grounding and Sound Separation", "Into the Wild with AudioScope: Unsupervised Audio-Visual Separation of On-Screen Sounds"], "answer_arxiv_id": ["1804.03160", "1804.03619", "1804.01665", "1908.11602", "2104.02026", "2011.01143"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_1871"} +{"question": "What works improved SSC through added modalities?", "answer": ["S3CNet: A Sparse Semantic Scene Completion Network for LiDAR Point\n Clouds", "Anisotropic Convolutional Networks for 3D Semantic Scene Completion"], "answer_arxiv_id": ["2012.09242", "2004.02122"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_1872"} +{"question": "What studies introduced an adversarial benchmark including three white-box attacks and one black-box attack?", "answer": ["Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks"], "answer_arxiv_id": ["2003.01690"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_1873"} +{"question": "Could you provide me some studies that use finite difference instead of automatic differentiation to avoid double propagation?", "answer": ["Human Performance Modeling and Rendering via Neural Animated Mesh", "Neuralangelo: High-Fidelity Neural Surface Reconstruction"], "answer_arxiv_id": ["2209.08468", "2306.03092"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_1874"} +{"question": "Which papers conduct the comparative research between Mini-batch SGD and Local SGD?", "answer": ["Don’t Use Large Mini-Batches, Use Local SGD", "Is Local SGD Better than Minibatch SGD?", "Minibatch vs Local SGD for Heterogeneous Distributed Learning", "Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond"], "answer_arxiv_id": ["1808.07217", "2002.07839", "2006.04735", "2110.10342"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_1875"} +{"question": "What are some notable datasets for face-swapping approach in deepfake technology?", "answer": ["DeeperForensics-1.0: A Large-Scale Dataset for Real-World Face Forgery Detection"], "answer_arxiv_id": ["2001.03024"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_1876"} +{"question": "Which works introduce the reparameterization trick with the Gumbel-softmax distribution for continuous relaxation in discrete optimization?", "answer": ["Categorical Reparameterization with Gumbel-Softmax", "The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables"], "answer_arxiv_id": ["1611.01144", "1611.00712"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_1877"} +{"question": "What studies have been conducted on the expressivity of CPWA NNs in terms of the number of affine regions?", "answer": ["On the number of response regions of deep feedforward networks with piecewise linear activations", "On the Number of Linear Regions of Deep Neural Networks", "Representation Benefits of Deep Feedforward Networks", "Benefits of depth in neural networks", "On the Expressive Power of Deep Neural Networks", "Bounding and Counting Linear Regions of Deep Neural Networks", "Complexity of Linear Regions in Deep Networks", "Locally Linear Attributes of ReLU Neural Networks"], "answer_arxiv_id": ["1312.6098", "1402.1869", "1509.08101", "1602.04485", "1606.05336", "1711.02114v4", "1901.09021", "2012.01940"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1878"} +{"question": "Which studies concentrated on achieving parallel decoding process for entropy models?", "answer": ["Checkerboard Context Model for Efficient Learned Image Compression", "Channel-wise Autoregressive Entropy Models for Learned Image Compression", "ELIC: Efficient Learned Image Compression with Unevenly Grouped\n Space-Channel Contextual Adaptive Coding"], "answer_arxiv_id": ["2103.15306", "2007.08739", "2203.10886"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_1879"} +{"question": "What research has been done on diffusion models for 3D data focusing on single object shapes?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "A Conditional Point Diffusion-Refinement Paradigm for 3D Point Cloud\n Completion", "LION: Latent Point Diffusion Models for 3D Shape Generation", "CLIP-Forge: Towards Zero-Shot Text-to-Shape Generation", "CLIP-Sculptor: Zero-Shot Generation of High-Fidelity and Diverse Shapes\n from Natural Language", "Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2104.03670", "2103.01458", "2112.03530", "2210.06978", "2110.02624", "2211.01427", "2212.14704"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_1880"} +{"question": "What studies measured the curiosity of state based on the error of prediction of the output of the environment model?", "answer": ["Surprise-Based Intrinsic Motivation for Deep Reinforcement Learning", "Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models"], "answer_arxiv_id": ["1703.01732", "1507.00814"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_1881"} +{"question": "What papers addressed sparsification approaches in communication-efficient algorithms?", "answer": ["Sparsified SGD with Memory", "Sparse Communication for Distributed Gradient Descent", "On the Discrepancy between the Theoretical Analysis and Practical Implementations of Compressed Communication for Distributed Deep Learning", "ScaleCom: Scalable Sparsified Gradient Compression for Communication-Efficient Distributed Training"], "answer_arxiv_id": ["1809.07599", "1704.05021", "1911.08250", "2104.11125"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_1882"} +{"question": "What research presented the Semantic Entropy (SE) to tackle the difficulty in uncertainty quantification?", "answer": ["Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation\n in Natural Language Generation"], "answer_arxiv_id": ["2302.09664"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_1883"} +{"question": "Which studies introduced long-horizon planning and reasoning in large language models?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "On the Paradox of Learning to Reason from Data", "ReAct: Synergizing Reasoning and Acting in Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2204.01691", "2205.11502", "2210.03629", "2305.10601"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_1884"} +{"question": "What research works are about pretraining for E2E ST?", "answer": ["Pre-training on high-resource speech recognition improves low-resource speech-to-text translation", "Analyzing ASR pretraining for low-resource speech-to-text translation", "Stacked Acoustic-and-Textual Encoding: Integrating the Pre-trained Models into Speech Translation Encoders", "End-to-end Speech Translation via Cross-modal Progressive Training", "Curriculum Pre-training for End-to-End Speech Translation", "Lightweight Adapter Tuning for Multilingual Speech Translation"], "answer_arxiv_id": ["1809.01431", "1910.10762", "2105.05752v2", "2104.10380", "2004.10093", "2106.01463"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_1885"} +{"question": "Which works designed knowledge-aware training objectives by incorporating KG entities and relations into the training data?", "answer": ["ERNIE: Enhanced Language Representation with Informative Entities", "ERNIE 3.0: Large-scale Knowledge Enhanced Pre-training for Language\n Understanding and Generation"], "answer_arxiv_id": ["1905.07129", "2107.02137"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_1886"} +{"question": "Could you provide me some examples of works about the use of boosting methods in post-processing?", "answer": ["Multiaccuracy: Black-Box Post-Processing for Fairness in Classification"], "answer_arxiv_id": ["1805.12317"], "source_meta": {"published_time": "20201201"}, "qid": "AutoScholarQuery_train_1887"} +{"question": "What papers proposed backflow and Jastrow factors for the introduction of electron correlation effects?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_1888"} +{"question": "Could you provide me some works that obtained transferability results for spectral graph convolution networks?", "answer": ["Transferability of Spectral Graph Convolutional Neural Networks"], "answer_arxiv_id": ["1907.12972"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_1889"} +{"question": "Could you tell me some works that focused on the last-iterate convergence of the dynamics in finite action zero-sum games?", "answer": ["Adaptive Learning in Continuous Games: Optimal Regret Bounds and Convergence to Nash Equilibrium"], "answer_arxiv_id": ["2104.12761"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_1890"} +{"question": "Could you provide me some studies on prompting-based approaches that leverage the instruction-following abilities of LLMs?", "answer": ["Language Models (Mostly) Know What They Know", "Large Language Models Cannot Self-Correct Reasoning Yet", "Knowledge Card: Filling LLMs' Knowledge Gaps with Plug-in Specialized\n Language Models", "Can ChatGPT Defend its Belief in Truth? Evaluating LLM Reasoning via\n Debate", "Getting MoRE out of Mixture of Language Model Reasoning Experts"], "answer_arxiv_id": ["2207.05221", "2310.01798", "2305.09955", "2305.13160", "2305.14628"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_1891"} +{"question": "Any work about pretraining a vision encoder using discriminative objectives?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["1911.05722", "2002.05709", "2006.07733", "2104.14294"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_1892"} +{"question": "What papers are on the subject of using 2D diffusion models in 3D generation?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2112.10752"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_1893"} +{"question": "Which papers discuss Bayesian Neural Networks (BNNs) for probabilistic predictions?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer\n Vision?"], "answer_arxiv_id": ["1703.04977"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_1894"} +{"question": "Could you provide me some studies who treat prompt engineering as a retrieval problem?", "answer": ["Learning To Retrieve Prompts for In-Context Learning"], "answer_arxiv_id": ["2112.08633"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_1895"} +{"question": "In what papers did the researchers address the inefficiency of fine-tuning steps in subject-driven image generation?", "answer": ["Pivotal Tuning for Latent-based Editing of Real Images"], "answer_arxiv_id": ["2106.05744"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_1896"} +{"question": "Which works study fairness-specific shifts handling group changes, particularly correlated with fair training?", "answer": ["Does enforcing fairness mitigate biases caused by subpopulation shift?", "Transferring Fairness under Distribution Shifts via Fair Consistency Regularization"], "answer_arxiv_id": ["2011.03173", "2206.12796"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_1897"} +{"question": "Could you provide me information about research papers that have been conducted to expand large language models to the visual modality?", "answer": ["Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models"], "answer_arxiv_id": ["2303.04671"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_1898"} +{"question": "Could you name the work that demonstrates a mathematical correspondence between proper loss functions and associated pooling methods in probabilistic opinion pooling?", "answer": ["From Proper Scoring Rules to Max-Min Optimal Forecast Aggregation"], "answer_arxiv_id": ["2102.07081"], "source_meta": {"published_time": "20220222"}, "qid": "AutoScholarQuery_train_1899"} +{"question": "Which papers consider the utilization of published classifiers as instruments in an offline setting in the context of causal strategic learning?", "answer": ["Strategic Instrumental Variable Regression: Recovering Causal Relationships From Strategic Responses"], "answer_arxiv_id": ["2107.05762"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_1900"} +{"question": "What works have focused on achieving a consistent estimate of target accuracy that requires target domain calibration?", "answer": ["Assessing Generalization of SGD via Disagreement", "Predicting with Confidence on Unseen Distributions"], "answer_arxiv_id": ["2106.13799", "2107.03315"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_1901"} +{"question": "Could you give me research papers that proposed improvements for the performance of SNIP at high sparsity ratios?", "answer": ["Progressive Skeletonization: Trimming more fat from a network at initialization", "A Signal Propagation Perspective for Pruning Neural Networks at Initialization"], "answer_arxiv_id": ["2006.09081", "1906.06307"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_1902"} +{"question": "Are there any papers that extended the polynomial methods for learning GMM's parameters to the robust setting?", "answer": ["Settling the Robust Learnability of Mixtures of Gaussians"], "answer_arxiv_id": ["2011.03622"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_1903"} +{"question": "What papers provide overviews on the understanding of nonconvex optimization for low-rank matrix estimation?", "answer": ["Nonconvex Optimization Meets Low-Rank Matrix Factorization: An Overview"], "answer_arxiv_id": ["1809.09573"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_1904"} +{"question": "What studies have proposed the use of Neural temporal point processes (NTPPs) and which of them have utilized RNNs for this?", "answer": ["The Neural Hawkes Process: A Neurally Self-Modulating Multivariate Point Process"], "answer_arxiv_id": ["1612.09328"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_1905"} +{"question": "Could you provide me some papers showing that every Gaussian approximation corresponds to the true posterior of a surrogate regression problem?", "answer": ["Approximate Inference Turns Deep Networks into Gaussian Processes", "Dual Parameterization of Sparse Variational Gaussian Processes"], "answer_arxiv_id": ["1906.01930", "2111.03412v2"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_1906"} +{"question": "Which work demonstrated that fine-tuning CLIP image-text encoders without any specialized modules can efficiently generate video representations?", "answer": ["Fine-tuned CLIP Models are Efficient Video Learners"], "answer_arxiv_id": ["2212.03640"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_1907"} +{"question": "What advances have been made in training paradigms for LVLMs?", "answer": ["PaLI-X: On Scaling up a Multilingual Vision and Language Model", "Training language models to follow instructions with human feedback", "Aligning Large Multimodal Models with Factually Augmented RLHF", "Silkie: Preference Distillation for Large Visual Language Models"], "answer_arxiv_id": ["2305.18565", "2203.02155", "2309.14525", "2312.10665"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_1908"} +{"question": "Which work talked about the problem of computing a coarse correlated equilibrium comprised of stationary Markov policies in a general-sum infinite horizon Markov game?", "answer": ["The Complexity of Markov Equilibrium in Stochastic Games"], "answer_arxiv_id": ["2204.03991"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_1909"} +{"question": "What works have been done on multi-modal retrieval tasks?", "answer": ["ManyModalQA: Modality Disambiguation and QA over Diverse Inputs", "MultiModalQA: complex question answering over text, tables and images", "WebQA: Multihop and Multimodal QA"], "answer_arxiv_id": ["2001.08034", "2104.06039", "2109.00590"], "source_meta": {"published_time": "20220901"}, "qid": "AutoScholarQuery_train_1910"} +{"question": "Which studies were in the field of contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Supervised Contrastive Learning", "Representation Learning with Contrastive Predictive Coding", "On Variational Bounds of Mutual Information"], "answer_arxiv_id": ["2002.05709", "2004.11362", "1807.03748", "1905.06922"], "source_meta": {"published_time": "20221110"}, "qid": "AutoScholarQuery_train_1911"} +{"question": "What works established a T2/3 worst-case regret in the convex cost setting of LQG control?", "answer": ["Improper Learning for Non-Stochastic Control", "Regret Minimization in Partially Observable Linear Quadratic Control"], "answer_arxiv_id": ["2001.09254", "2002.00082"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_1912"} +{"question": "Could you mention some researches in understanding how Differential Privacy aids in privacy protection?", "answer": ["Provable Defense against Privacy Leakage in Federated Learning from Representation Perspective", "Bayesian Framework for Gradient Leakage", "Understanding Training-Data Leakage from Gradients in Neural Networks for Image Classification", "Evaluating Gradient Inversion Attacks and Defenses in Federated Learning"], "answer_arxiv_id": ["2012.06043", "2111.04706", "2111.10178", "2112.00059"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_1913"} +{"question": "Which work made progress with CFR-type algorithms and achieved a faster average-iterate convergence rate in EFGs?", "answer": ["Stable-Predictive Optimistic Counterfactual Regret Minimization"], "answer_arxiv_id": ["1902.04982"], "source_meta": {"published_time": "20220619"}, "qid": "AutoScholarQuery_train_1914"} +{"question": "Could you provide me some references developing efficient online RL methods for POMDPs?", "answer": ["A PAC RL Algorithm for Episodic POMDPs", "PAC Reinforcement Learning with Rich Observations", "Sample-Efficient Reinforcement Learning of Undercomplete POMDPs", "Sublinear Regret for Learning POMDPs", "Online Learning for Unknown Partially Observable MDPs", "Provable Reinforcement Learning with a Short-Term Memory", "When Is Partially Observable Reinforcement Learning Not Scary?"], "answer_arxiv_id": ["1605.08062v2", "1602.02722", "2006.12484", "2107.03635v4", "2102.12661", "2202.03983", "2204.08967"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_1915"} +{"question": "What papers specify neural network models based on statistical interactions?", "answer": ["Explaining Explanations: Axiomatic Feature Interactions for Deep Networks", "How does this interaction affect me? Interpretable attribution for feature interactions", "Detecting Statistical Interactions From Neural Network Weights"], "answer_arxiv_id": ["2002.04138", "2006.10965", "1705.04977"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_1916"} +{"question": "Are there any studies that use various sampling and memory layout optimizations for real-time HD rendering?", "answer": ["MERF: Memory-Efficient Radiance Fields for Real-time View Synthesis in\n Unbounded Scenes"], "answer_arxiv_id": ["2302.12249"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_1917"} +{"question": "Which works modify the bi-encoder retriever to be equipped with mBERT in the open-QA system?", "answer": ["XOR QA: Cross-lingual Open-Retrieval Question Answering", "One Question Answering Model for Many Languages with Cross-lingual Dense Passage Retrieval"], "answer_arxiv_id": ["2010.11856v3", "2107.11976"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_1918"} +{"question": "Which papers discuss using 2D diffusion models for generating high-fidelity and diversified 2D content from text prompts?", "answer": ["Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2112.10752", "2202.00512"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1919"} +{"question": "Which work is most related to the hybrid architecture design with voxel grids and a shallow MLP?", "answer": ["Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction"], "answer_arxiv_id": ["2111.11215"], "source_meta": {"published_time": "20220826"}, "qid": "AutoScholarQuery_train_1920"} +{"question": "Can you provide examples of research demonstrating the benefits of augmenting datasets with high-quality content?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2005.14165", "2302.13971"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1921"} +{"question": "Which papers are about understanding scene generation, particularly in indoor settings using scene graphs?", "answer": ["Image Generation from Scene Graphs", "DiffuScene: Denoising Diffusion Models for Generative Indoor Scene\n Synthesis"], "answer_arxiv_id": ["1804.01622", "2303.14207"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_1922"} +{"question": "Which work proposed a novel approach for causal inference that uses score matching algorithms?", "answer": ["Score matching enables causal discovery of nonlinear additive noise models"], "answer_arxiv_id": ["2203.04413"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_1923"} +{"question": "What are the studies that applied score-based diffusion models to 3D molecular and material structure generation?", "answer": ["Learning Gradient Fields for Molecular Conformation Generation"], "answer_arxiv_id": ["2105.03902"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_1924"} +{"question": "What studies explore Gaussian diffusion for multivariate time-series imputation and forecasting?", "answer": ["Diffusion-based Time Series Imputation and Forecasting with Structured State Space Models", "Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting"], "answer_arxiv_id": ["2208.09399", "2101.12072"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_1925"} +{"question": "Which works propose a mask-guided method for localized image editing?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images"], "answer_arxiv_id": ["2111.14818"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_1926"} +{"question": "Could you provide studies that have shown improved bounds in model-free RL when the estimation-to-decisions (E2D) algorithm is combined with optimistic estimation?", "answer": ["Feel-Good Thompson Sampling for Contextual Bandits and Reinforcement Learning"], "answer_arxiv_id": ["2110.00871"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_1927"} +{"question": "Which studies have used masked image modeling as the pretraining task in the context of generative approaches?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: a Simple Framework for Masked Image Modeling"], "answer_arxiv_id": ["2010.11929", "2106.08254", "2111.06377", "2111.09886"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_1928"} +{"question": "Are there any works about machine-learning algorithms for estimating room impulse response?", "answer": ["Learning Neural Acoustic Fields", "Few-Shot Audio-Visual Learning of Environment Acoustics", "MESH2IR: Neural Acoustic Impulse Response Generator for Complex 3D\n Scenes", "FAST-RIR: Fast neural diffuse room impulse response generator", "Neural Acoustic Context Field: Rendering Realistic Room Impulse Response\n With Neural Fields", "Listen2Scene: Interactive material-aware binaural sound propagation for\n reconstructed 3D scenes"], "answer_arxiv_id": ["2204.00628", "2206.04006", "2205.09248", "2110.04057", "2309.15977", "2302.02809"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_1929"} +{"question": "Which papers use ZeroScope in their research, a model in video diffusion?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object", "Objaverse-XL: A Universe of 10M+ 3D Objects"], "answer_arxiv_id": ["2303.11328", "2307.05663"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_1930"} +{"question": "Could you provide me some studies about architectures with the same order of GroupNorm and residual connections?", "answer": ["MetaFormer Is Actually What You Need for Vision"], "answer_arxiv_id": ["2111.11418"], "source_meta": {"published_time": "20211123"}, "qid": "AutoScholarQuery_train_1931"} +{"question": "Which research papers discuss the optimization of geodesic convex functions and the challenges associated with it?", "answer": ["First-order Methods for Geodesically Convex Optimization"], "answer_arxiv_id": ["1602.06053"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_1932"} +{"question": "Which works proposed confidence calibration scaling methods?", "answer": ["Non-Parametric Calibration for Classification", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration"], "answer_arxiv_id": ["1906.04933", "1910.12656"], "source_meta": {"published_time": "20220215"}, "qid": "AutoScholarQuery_train_1933"} +{"question": "Which paper introduced NetHack as a testbed for RL agents?", "answer": ["The NetHack Learning Environment"], "answer_arxiv_id": ["2006.13760"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_1934"} +{"question": "Could you provide me some studies about deterministic Markov Chain Monte Carlo (MCMC)?", "answer": ["Driving Markov chain Monte Carlo with a dependent random stream", "Hamiltonian Dynamics with Non-Newtonian Momentum for Rapid Sampling", "Deterministic Gibbs Sampling via Ordinary Differential Equations"], "answer_arxiv_id": ["1204.3187", "2111.02434", "2106.10188v1"], "source_meta": {"published_time": "20220516"}, "qid": "AutoScholarQuery_train_1935"} +{"question": "Are there any studies that focus specifically on audio-visual DOAE tasks?", "answer": ["Multi-target DoA estimation with an audio-visual fusion mechanism", "Deep Learning Based Audio-Visual Multi-Speaker DOA Estimation Using Permutation-Free Loss Function"], "answer_arxiv_id": ["2105.06107", "2210.14581"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_1936"} +{"question": "What papers have proposed the use of Language Models for Autonomous Driving?", "answer": ["Drive Anywhere: Generalizable End-to-end Autonomous Driving with\n Multi-modal Foundation Models", "GPT-Driver: Learning to Drive with GPT", "DriveGPT4: Interpretable End-to-end Autonomous Driving via Large\n Language Model", "MotionLM: Multi-Agent Motion Forecasting as Language Modeling"], "answer_arxiv_id": ["2310.17642", "2310.01415", "2310.01412", "2309.16534"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_1937"} +{"question": "Can you name some research papers that study about the generalization of Open-set domain adaptation allowing unknown classes to exist in both source and target domains?", "answer": ["Universal Domain Adaptation through Self-Supervision"], "answer_arxiv_id": ["2002.07953"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_1938"} +{"question": "Could you provide me information on the studies involving the human-authored benchmark FOLIO for multistep deductive reasoning?", "answer": ["FOLIO: Natural Language Reasoning with First-Order Logic"], "answer_arxiv_id": ["2209.00840"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_1939"} +{"question": "Which works incorporated explicit geometric constraints into the diffusion models?", "answer": ["LDM3D: Latent Diffusion Model for 3D", "SweetDreamer: Aligning Geometric Priors in 2D Diffusion for Consistent\n Text-to-3D", "Wonder3D: Single Image to 3D using Cross-Domain Diffusion", "HumanNorm: Learning Normal Diffusion Model for High-quality and\n Realistic 3D Human Generation"], "answer_arxiv_id": ["2305.10853", "2310.02596", "2310.15008", "2310.01406"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_1940"} +{"question": "Which studies applied weights sharing and multiplexing to reduce model parameters in vision transformer?", "answer": ["MiniViT: Compressing Vision Transformers with Weight Multiplexing"], "answer_arxiv_id": ["2204.07154"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_1941"} +{"question": "What are some studies regarding emergent reasoning in recurrent neural sequence models?", "answer": ["Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation", "Deep contextualized word representations", "Universal Language Model Fine-tuning for Text Classification"], "answer_arxiv_id": ["1609.08144", "1802.05365", "1801.06146"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_1942"} +{"question": "What works use deep generative learning and conditional sampling for multivariate time series imputation?", "answer": ["GP-VAE: Deep Probabilistic Time Series Imputation", "Latent ODEs for Irregularly-Sampled Time Series", "Neural Controlled Differential Equations for Irregular Time Series", "Scalable Gradients for Stochastic Differential Equations", "Continuous Latent Process Flows"], "answer_arxiv_id": ["1907.04155", "1907.03907", "2005.08926", "2001.01328v6", "2106.15580"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_1943"} +{"question": "What works focused on transferring pre-trained vision transformers into vision-language domains using adapters?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2301.12597", "2304.15010", "2304.10592"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_1944"} +{"question": "Can you identify works that have focused on real paired event denoising data?", "answer": ["Event Probability Mask (EPM) and Event Denoising Convolutional Neural\n Network (EDnCNN) for Neuromorphic Cameras", "Neuromorphic Camera Denoising using Graph Neural Network-driven\n Transformers"], "answer_arxiv_id": ["2003.08282", "2112.09685"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_1945"} +{"question": "What works introduced joint vision-language pre-training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "Combined Scaling for Zero-shot Transfer Learning", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Sigmoid Loss for Language Image Pre-Training", "Equivariant Similarity for Vision-Language Foundation Models"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.07991", "2111.10050", "2108.10904", "2202.03052", "2201.12086", "2205.01917", "2303.15343v4", "2303.14465"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_1946"} +{"question": "Could you mention the studies that optimized the form of the rate functional via gradient descent on the parameters of a variational measure parameterized by a neural network?", "answer": ["Neural Estimation of the Rate-Distortion Function With Applications to Operational Source Coding"], "answer_arxiv_id": ["2204.01612"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_1947"} +{"question": "What studies have been conducted on instance discrimination task by using a strategy of contrastive learning?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap your own latent: A new approach to self-supervised Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster\n Assignments", "Masked Siamese Networks for Label-Efficient Learning", "Adaptive Soft Contrastive Learning", "MaskCon: Masked Contrastive Learning for Coarse-Labelled Dataset"], "answer_arxiv_id": ["2104.14294", "2002.05709", "1911.05722", "2006.07733", "2006.09882", "2204.07141", "2207.11163", "2303.12756"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_1948"} +{"question": "What works have made approaches to learn useful representations via data augmentations in the area of in-domain representation learning?", "answer": ["Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Reinforcement Learning with Augmented Data"], "answer_arxiv_id": ["2004.13649", "2004.14990"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_1949"} +{"question": "What papers have tried to validate the robustness of SAM under more challenging scenarios, such as medical images, camouflaged objects, and pose estimation?", "answer": ["Segment Anything in Medical Images", "Customized Segment Anything Model for Medical Image Segmentation", "Can SAM Segment Anything? When SAM Meets Camouflaged Object Detection", "Segment Anything Model (SAM) Meets Glass: Mirror and Transparent Objects\n Cannot Be Easily Detected", "SAM-6D: Segment Anything Model Meets Zero-Shot 6D Object Pose Estimation"], "answer_arxiv_id": ["2304.12306", "2304.13785", "2304.04709", "2305.00278", "2311.15707"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_1950"} +{"question": "Any works about the Marvel model which is trained specifically for multi-modal document retrieval?", "answer": ["WebQA: Multihop and Multimodal QA", "Universal Vision-Language Dense Retrieval: Learning A Unified\n Representation Space for Multi-Modal Retrieval"], "answer_arxiv_id": ["2109.00590", "2209.00179"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_1951"} +{"question": "Which papers showcased the advantage of attention-based models over recurrent neural networks in bAbI tasks?", "answer": ["End-To-End Memory Networks", "Tracking the World State with Recurrent Entity Networks"], "answer_arxiv_id": ["1503.08895", "1612.03969"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_1952"} +{"question": "What works utilize the bi-level optimization in tasks like neural architecture search, meta-learning and hyperparameter optimization?", "answer": ["DARTS: Differentiable Architecture Search", "Meta-Learning with Implicit Gradients", "Optimizing Millions of Hyperparameters by Implicit Differentiation", "Stability and Generalization of Bilevel Programming in Hyperparameter Optimization"], "answer_arxiv_id": ["1806.09055", "1909.04630", "1911.02590", "2106.04188"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_1953"} +{"question": "Which researches focused on very large scale training runs?", "answer": ["GPT-NeoX-20B: An Open-Source Autoregressive Language Model"], "answer_arxiv_id": ["2204.06745"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_1954"} +{"question": "Could you name the studies that focused on penalizing the geodesic distortion of the predicted maps in shape matching?", "answer": ["Cyclic Functional Mapping: Self-supervised correspondence between non-isometric deformable shapes"], "answer_arxiv_id": ["1912.01249"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1955"} +{"question": "What work is a representative structure used in low-level vision that employs downsampling and upsampling?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation"], "answer_arxiv_id": ["1505.04597"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_1956"} +{"question": "Can you list the works that explore and anticipate the object affordance in human-object interaction?", "answer": ["Learning to Act Properly: Predicting and Explaining Affordances from\n Images", "AffordanceNet: An End-to-End Deep Learning Approach for Object\n Affordance Detection", "Grounded Human-Object Interaction Hotspots from Video", "Learning Visual Affordance Grounding from Demonstration Videos", "Learning Affordance Grounding from Exocentric Images", "One-Shot Object Affordance Detection in the Wild"], "answer_arxiv_id": ["1712.07576", "1709.07326", "1812.04558", "2108.05675", "2203.09905", "2108.03658"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_1957"} +{"question": "What are some studies that have used probing approach to discover targeted neurons in LLMs?", "answer": ["Finding Neurons in a Haystack: Case Studies with Sparse Probing", "Probing Classifiers: Promises, Shortcomings, and Advances"], "answer_arxiv_id": ["2305.01610", "2102.12452"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_1958"} +{"question": "Which papers used reinforcement learning-based NAS strategies and one-shot NAS methods for efficient architecture search in face recognition?", "answer": ["PC-DARTS: Partial Channel Connections for Memory-Efficient Architecture Search"], "answer_arxiv_id": ["1907.05737"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_1959"} +{"question": "Can you provide a reference where the study focused on unsupervised concept discovery and concept composition methods for energy-based models?", "answer": ["Unsupervised Learning of Compositional Energy Concepts"], "answer_arxiv_id": ["2111.03042"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_1960"} +{"question": "Are there any works that demonstrate preprocessing methods often inadequate compared to in-processing and post-processing methods?", "answer": ["Retiring Adult: New Datasets for Fair Machine Learning"], "answer_arxiv_id": ["2108.04884"], "source_meta": {"published_time": "20220916"}, "qid": "AutoScholarQuery_train_1961"} +{"question": "What works introduce adaptive data augmentation techniques to enhance semi-supervised semantic segmentation?", "answer": ["Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning", "Augmentation Matters: A Simple-yet-Effective Approach to Semi-supervised Semantic Segmentation"], "answer_arxiv_id": ["2110.05474", "2212.04976"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_1962"} +{"question": "Which works studied the generation of images, videos, and 3D-scenes within text-guided multimodal systems?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Zero-Shot Text-to-Image Generation", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "VideoFusion: Decomposed Diffusion Models for High-Quality Video Generation", "Shap⋅E: Generating Conditional 3D Implicit Functions", "DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2102.12092", "2209.14792", "2303.08320", "2305.02463", "2209.14988", "2211.10440"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_1963"} +{"question": "What studies have focused on learning generative models with DMs via score matching?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Interpretation and Generalization of Score Matching", "Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "1205.2629v1", "1907.05600", "2011.13456", "2006.11239", "2102.09672"], "source_meta": {"published_time": "20220611"}, "qid": "AutoScholarQuery_train_1964"} +{"question": "What papers are about injecting KG or knowledge base data into LLM prompts?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Graph Reasoning for Question Answering with Triplet Retrieval"], "answer_arxiv_id": ["2005.11401", "2305.18742"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_1965"} +{"question": "Any studies on Transformers extension for object detection tasks?", "answer": ["End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2005.12872"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_1966"} +{"question": "Which works discuss the enhancement of LLM's capacity to follow natural language instructions through instruction-based fine-tuning?", "answer": ["Deep Reinforcement Learning from Human Preferences", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["1706.03741", "2203.02155"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_1967"} +{"question": "Which work proposes an efficient hashing-based implementation of 'selective softmax'?", "answer": ["Accelerated Training for Massive Classification via Dynamic Class Selection"], "answer_arxiv_id": ["1801.01687v1"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_1968"} +{"question": "Any works about applying RLHF for text summarization?", "answer": ["Learning to summarize from human feedback"], "answer_arxiv_id": ["2009.01325"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_1969"} +{"question": "Which papers proposed methods to eliminate the effect of data heterogeneity on the convergence rate?", "answer": ["NEXT: In-Network Nonconvex Optimization", "Achieving Geometric Convergence for Distributed Optimization over Time-Varying Graphs", "Distributed Stochastic Gradient Tracking Methods", "Momentum Tracking: Momentum Acceleration for Decentralized Deep Learning on Heterogeneous Data", "Distributed heavy-ball: A generalization and acceleration of first-order methods with gradient tracking", "A Decentralized Proximal-Gradient Method with Network Independent Step-sizes and Separated Convergence Rates", "RelaySum for Decentralized Deep Learning on Heterogeneous Data", "Exact Diffusion for Distributed Optimization and Learning — Part I: Algorithm Development"], "answer_arxiv_id": ["1602.00591", "1607.03218v3", "1805.11454", "2209.15505", "1808.02942", "1704.07807", "2110.04175", "1702.05122"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_1970"} +{"question": "Could you provide me some studies about introducing multiple levels of hierarchies for skill learning or planning?", "answer": ["Self-Consistent Trajectory Autoencoder: Hierarchical Reinforcement Learning with Trajectory Embeddings", "Data-Efficient Hierarchical Reinforcement Learning", "The Option Keyboard Combining Skills in Reinforcement Learning", "Composing Task-Agnostic Policies with Deep Reinforcement Learning", "Reinforcement Learning with Competitive Ensembles of Information-Constrained Primitives", "Learning Transferable Motor Skills with Hierarchical Latent Mixture Policies"], "answer_arxiv_id": ["1806.02813", "1805.08296", "2106.13105", "1905.10681", "1906.10667", "2112.05062"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_1971"} +{"question": "Which works report about enforcing control over desired attributes at inference time without updating the PLM parameters?", "answer": ["Plug and Play Language Models: a Simple Approach to Controlled Text Generation", "GeDi: Generative Discriminator guided Sequence Generation", "Controlled Text Generation as Continuous Optimization with Multiple Constraints", "DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts", "A Plug-and-Play Method for Controlled Text Generation", "Fudge: Controlled Text Generation With Future Discriminators"], "answer_arxiv_id": ["1912.02164", "2009.06367", "2108.01850", "2105.03023", "2109.09707", "2104.05218"], "source_meta": {"published_time": "20221106"}, "qid": "AutoScholarQuery_train_1972"} +{"question": "Could you provide me studies about training diffusion models in the latent space?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2204.06125"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_1973"} +{"question": "Which papers studied single-item single-bidder posted-price auctions with approximate bid predictions?", "answer": ["Revenue Optimization with Approximate Bid Predictions"], "answer_arxiv_id": ["1706.04732"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_1974"} +{"question": "Are there any works that used feature maps and gradients as general representations over images for forgery detection?", "answer": ["Towards Universal Fake Image Detectors that Generalize Across Generative\n Models"], "answer_arxiv_id": ["2302.10174"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_1975"} +{"question": "Which works showcase the application of MAE in the video domain?", "answer": ["Masked Autoencoders As Spatiotemporal Learners", "VideoMAE: Masked Autoencoders are Data-Efficient Learners for\n Self-Supervised Video Pre-Training", "VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking"], "answer_arxiv_id": ["2205.09113", "2203.12602", "2303.16727v2"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_1976"} +{"question": "Can you tell me some studies that focused on quantifying interactions between input variables in DNNs?", "answer": ["Beyond Word Importance: Contextual Decomposition to Extract Interactions from LSTMs", "Hierarchical interpretations for neural network predictions", "Towards Hierarchical Importance Attribution: Explaining Compositional Semantics for Neural Sequence Models", "Explaining Explanations: Axiomatic Feature Interactions for Deep Networks"], "answer_arxiv_id": ["1801.05453", "1806.05337", "1911.06194", "2002.04138"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_1977"} +{"question": "Can you list some papers that center on the use of neural radiance fields models?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2003.08934", "2209.14988"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_1978"} +{"question": "What works about LLMs methods that learn from feedback and demonstrate enhanced reasoning potential?", "answer": ["Reflexion: Language Agents with Verbal Reinforcement Learning", "Inner Monologue: Embodied Reasoning through Planning with Language\n Models"], "answer_arxiv_id": ["2303.11366", "2207.05608"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_1979"} +{"question": "What researches disrupt the output of the generative models by adding perturbations to the training data?", "answer": ["Disrupting Deepfakes: Adversarial Attacks Against Conditional Image\n Translation Networks and Facial Manipulation Systems"], "answer_arxiv_id": ["2003.01279"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_1980"} +{"question": "Which study designed a text-to-image association test to measure the implicit stereotypes?", "answer": ["T2IAT: Measuring Valence and Stereotypical Biases in Text-to-Image\n Generation"], "answer_arxiv_id": ["2306.00905"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_1981"} +{"question": "Which works explores private empirical risk minimization?", "answer": ["Differentially Private Empirical Risk Minimization", "Private Empirical Risk Minimization Beyond the Worst Case: The Effect of the Constraint Set Geometry", "Bolt-on Differential Privacy for Scalable Stochastic Gradient Descent-based Analytics"], "answer_arxiv_id": ["0912.0071v5", "1411.5417", "1606.04722"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_1982"} +{"question": "What datasets or benchmarks have been used to test the capability of large language models to understand and process long text?", "answer": ["LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding"], "answer_arxiv_id": ["2308.14508v2"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_1983"} +{"question": "Could you provide me some studies about the use of embedding layers in natural language processing?", "answer": ["Distributed Representations of Words and Phrases and their Compositionality", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1310.4546", "1810.04805"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_1984"} +{"question": "Which studies implement rule-based approaches for filtering in the data processing pipeline?", "answer": ["CCNet: Extracting High Quality Monolingual Datasets from Web Crawl Data", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "mT5: A massively multilingual pre-trained text-to-text transformer", "The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset"], "answer_arxiv_id": ["1911.00359", "1910.10683", "2010.11934", "2303.03915"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_1985"} +{"question": "What are the works that proposed methods for measuring uncertainty and diversity of samples for generic active learning?", "answer": ["Active Learning by Feature Mixing", "PAL : Pretext-based Active Learning", "Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Hierarchical Subquery Evaluation for Active Learning on a Graph"], "answer_arxiv_id": ["2203.07034", "2010.15947", "1708.00489", "1504.08219"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_1986"} +{"question": "What are the available 3D anomaly detection datasets?", "answer": ["The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localization", "The Eyecandies Dataset for Unsupervised Multimodal Anomaly Detection and Localization"], "answer_arxiv_id": ["2112.09045", "2210.04570"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_1987"} +{"question": "Who proposed a gradient highway unit to mitigate the gradient vanishing?", "answer": ["PredRNN++: Towards A Resolution of the Deep-in-Time Dilemma in Spatiotemporal Predictive Learning"], "answer_arxiv_id": ["1804.06300"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_1988"} +{"question": "Which works have contributed significant advancements toward 2D and 3D scene understanding in computer vision?", "answer": ["Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation", "Point Transformer V2: Grouped Vector Attention and Partition-based Pooling", "Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation", "PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation", "Point-to-Voxel Knowledge Distillation for LiDAR Semantic Segmentation", "RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR Point Cloud Segmentation", "(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Network", "Rethinking Range View Representation for LiDAR Segmentation", "Human-centric Scene Understanding for 3D Large-scale Scenarios", "UniSeg: A Unified Multi-Modal LiDAR Segmentation Network and the OpenPCSeg Codebase", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Segmenter: Transformer for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "Vision Transformers for Dense Prediction", "InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions", "Hierarchical multi-scale attention for semantic segmentation", "CRIS: CLIP-Driven Referring Image Segmentation", "Robo3D: Towards Robust and Reliable 3D Perception against Corruptions", "PR-Net: Preference Reasoning for Personalized Video Highlight Detection"], "answer_arxiv_id": ["2011.10033", "2210.05666", "1911.10194", "1612.00593", "2011.10033", "2206.02099", "2103.12978", "2102.04530v1", "2303.05367", "2307.14392", "2309.05573", "2107.06278", "2105.05633", "2112.01527", "2103.13413", "2211.05778", "2005.10821", "2111.15174", "2303.17597", "2109.01799"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_1989"} +{"question": "Could you list papers that have expanded the application of adversarial training to areas beyond the traditional vision applications?", "answer": ["Fast is better than free: Revisiting adversarial training", "Theoretically Principled Trade-off between Robustness and Accuracy", "Unlabeled Data Improves Adversarial Robustness", "Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuning"], "answer_arxiv_id": ["2001.03994", "1901.08573", "1905.13736", "2003.12862"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_1990"} +{"question": "Who has given a nonconformity score based on quantile regression?", "answer": ["Conformalized Quantile Regression"], "answer_arxiv_id": ["1905.03222"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_1991"} +{"question": "Could you provide me some studies indicating combining fine-tuning and prompt engineering could deliver orthogonal benefits?", "answer": ["True Few-Shot Learning with Language Models", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2105.11447", "2203.02155"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_1992"} +{"question": "What studies proposed predicting correspondence labels to recast the margin of triplet ranking loss as a soft margin?", "answer": ["BiCro: Noisy Correspondence Rectification for Multi-modality Data via Bi-directional Cross-modal Similarity Consistency"], "answer_arxiv_id": ["2303.12419v2"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_1993"} +{"question": "What are the studies that suggested Diffusion SCMs (Diff-SCMs) for high-fidelity counterfactuals?", "answer": ["Diffusion Causal Models for Counterfactual Estimation", "What is Healthy? Generative Counterfactual Diffusion for Lesion Localization"], "answer_arxiv_id": ["2202.10166", "2207.12268"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_1994"} +{"question": "What work introduces a new framework for partial multi-view learning to estimate missing data?", "answer": ["Deep Partial Multi-View Learning"], "answer_arxiv_id": ["2011.06170"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_1995"} +{"question": "Which papers studied the sixth-order polynomial scaled by an exponential function?", "answer": ["STay-ON-the-Ridge: Guaranteed Convergence to Local Minimax Equilibrium in Nonconvex-Nonconcave Games", "On Solving Minimax Optimization Locally: A Follow-the-Ridge Approach"], "answer_arxiv_id": ["2210.09769v1", "1910.07512"], "source_meta": {"published_time": "20221226"}, "qid": "AutoScholarQuery_train_1996"} +{"question": "What paper worked with MaskFormer for the fine-tuning of the segmentation branch?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation"], "answer_arxiv_id": ["2107.06278"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_1997"} +{"question": "Could you tell me about the research papers that handle the straggler problem in asynchronous horizontal federated learning?", "answer": ["Asynchronous Federated Optimization", "Asynchronous Federated Learning with Reduced Number of Rounds and with Differential Privacy from Less Aggregated Gaussian Noise", "Stragglers Are Not Disaster: A Hybrid Federated Learning Algorithm with Delayed Gradients", "Papaya: Practical, Private, and Scalable Federated Learning", "FedAT: A High-Performance and Communication-Efficient Federated Learning System with Asynchronous Tiers", "Federated Learning with Buffered Asynchronous Aggregation"], "answer_arxiv_id": ["1903.03934", "2007.09208", "2102.06329", "2111.04877v2", "2010.05958", "2106.06639v4"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_1998"} +{"question": "What studies talk about improving contrastive learning using adversarial training?", "answer": ["AdCo: Adversarial Contrast for Efficient Learning of Unsupervised Representations from Self-Trained Negative Adversaries"], "answer_arxiv_id": ["2011.08435"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_1999"} +{"question": "Which papers explore decentralized optimization methods?", "answer": ["On the Linear Speedup Analysis of Communication Efficient Momentum SGD for Distributed Non-Convex Optimization", "A Unified Theory of Decentralized SGD with Changing Topology and Local Updates"], "answer_arxiv_id": ["1905.03817", "2003.10422"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_2000"} +{"question": "Which works proposed specialized datasets for precipitation-event forecasting?", "answer": ["IowaRain: A Statewide Rain Event Dataset Based on Weather Radars and Quantitative Precipitation Estimation"], "answer_arxiv_id": ["2107.03432"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_2001"} +{"question": "Can you name the studies that examined the use of hypernetworks in the context of PDEs?", "answer": ["On the Modularity of Hypernetworks", "Implicit Neural Representations with Periodic Activation Functions", "Hypernetwork functional image representation", "HyperPINN: Learning parameterized differential equations with physics-informed hypernetworks", "Neural Implicit Flow: a mesh-agnostic dimensionality reduction paradigm of spatio-temporal data"], "answer_arxiv_id": ["2002.10006", "2006.09661", "1902.10404", "2111.01008", "2204.03216"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_2002"} +{"question": "Which studies explored the benefits of recognizing and rewarding subgoals in reinforcement learning?", "answer": ["Computational Benefits of Intermediate Rewards for Goal-Reaching Policy\n Learning"], "answer_arxiv_id": ["2107.03961"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_2003"} +{"question": "Could you provide me some studies that introduced trainable stitching layers for combining different networks?", "answer": ["Understanding image representations by measuring their equivariance and equivalence", "Revisiting Model Stitching to Compare Neural Representations"], "answer_arxiv_id": ["1411.5908", "2106.07682"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_2004"} +{"question": "Which document gives a comprehensive review on backdoor attacks?", "answer": ["Backdoor Learning: A Survey"], "answer_arxiv_id": ["2007.08745"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_2005"} +{"question": "Which works have been done on contrastive language-image pretraining?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_2006"} +{"question": "What are some works proposing infrastructure for trading data as a commodity in machine learning?", "answer": ["Data Shapley: Equitable Valuation of Data for Machine Learning", "A Marketplace for Data: An Algorithmic Solution"], "answer_arxiv_id": ["1904.02868", "1805.08125"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2007"} +{"question": "Which studies have extended diffusion-based models to various domains and achieved state-of-the-art results?", "answer": ["iEdit: Localised Text-guided Image Editing with Weak Supervision", "InterHandGen: Two-Hand Interaction Generation via Cascaded Reverse\n Diffusion", "LLDiffusion: Learning Degradation Representations in Diffusion Models\n for Low-Light Image Enhancement"], "answer_arxiv_id": ["2305.05947", "2403.17422", "2307.14659"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_2008"} +{"question": "Which study adds a marginal temporal Gaussian distribution into the origin 3D Gaussians?", "answer": ["Real-time Photorealistic Dynamic Scene Representation and Rendering with\n 4D Gaussian Splatting"], "answer_arxiv_id": ["2310.10642"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_2009"} +{"question": "Could you list the studies on Image-3D Generation that further improved the generation quality by combining Zero-1-to-3 with RealFusion?", "answer": ["Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors", "Consistent123: Improve Consistency for One Image to 3D Object Synthesis"], "answer_arxiv_id": ["2306.17843", "2310.08092"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_2010"} +{"question": "Are there any works that present LM toxicity after examining the content generated by language models?", "answer": ["Universal Adversarial Triggers for Attacking and Analyzing NLP", "Recipes for building an open-domain chatbot", "Detoxifying Language Models Risks Marginalizing Minority Voices"], "answer_arxiv_id": ["1908.07125", "2004.13637", "2104.06390"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_2011"} +{"question": "Which publications are about the application of deep learning in point cloud upsampling?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "KPConv: Flexible and Deformable Convolution for Point Clouds", "PU-Net: Point Cloud Upsampling Network", "Patch-based Progressive 3D Point Set Upsampling", "PU-GAN: a Point Cloud Upsampling Adversarial Network", "Point Cloud Upsampling via Disentangled Refinement", "PU-GCN: Point Cloud Upsampling using Graph Convolutional Networks", "PU-EVA: An Edge Vector based Approximation Solution for Flexible-scale\n Point Cloud Upsampling", "PUGeo-Net: A Geometry-centric Network for 3D Point Cloud Upsampling", "Neural Points: Point Cloud Representation with Neural Fields for\n Arbitrary Upsampling", "Grad-PU: Arbitrary-Scale Point Cloud Upsampling via Gradient Descent\n with Learned Distance Functions"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1904.08755", "1904.08889", "1801.06761", "1811.11286", "1907.10844", "2106.04779", "1912.03264", "2204.10750", "2002.10277", "2112.04148", "2304.11846"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_2012"} +{"question": "Could you provide me some works that discussed the guarantees of neural networks in the NTK regime?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Data", "Gradient Descent Provably Optimizes Over-parameterized Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Gradient Descent Finds Global Minima of Deep Neural Networks", "An Improved Analysis of Training Over-parameterized Deep Neural Networks", "On the Convergence Rate of Training Recurrent Neural Networks", "Learning and Generalization in Overparameterized Neural Networks, Going Beyond Two Layers", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "On Exact Computation with an Infinitely Wide Neural Net", "Generalization Bounds of Stochastic Gradient Descent for Wide and Deep Neural Networks", "Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networks", "How Much Over-parameterization Is Sufficient to Learn Deep ReLU Networks?"], "answer_arxiv_id": ["1806.07572", "1808.01204", "1810.02054", "1811.03962", "1811.03804", "1906.04688", "1810.12065", "1811.04918", "1901.08584", "1904.11955", "1905.13210", "1909.12292", "1911.12360"], "source_meta": {"published_time": "20210825"}, "qid": "AutoScholarQuery_train_2013"} +{"question": "Could you provide me some studies about explicit updating of backward weights in backprop?", "answer": ["How Auto-Encoders Could Provide Credit Assignment in Deep Networks via Target Propagation", "Difference Target Propagation", "Assessing the Scalability of Biologically-Motivated Deep Learning Algorithms and Architectures"], "answer_arxiv_id": ["1407.7906", "1412.7525", "1807.04587"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_2014"} +{"question": "Which works relied on object tracking to provide instance supervision for 3D panoptic segmentation?", "answer": ["Panoptic Neural Fields: A Semantic Object-Aware Neural Scene\n Representation"], "answer_arxiv_id": ["2205.04334"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_2015"} +{"question": "Which studies introduced edge-based methods such as CAWN and GraphMixer in the context of temporal graph methods?", "answer": ["Inductive Representation Learning in Temporal Networks via Causal Anonymous Walks", "Do We Really Need Complicated Model Architectures For Temporal Networks?"], "answer_arxiv_id": ["2101.05974", "2302.11636"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_2016"} +{"question": "Which articles discuss the theoretical properties of solving a different class of PDEs?", "answer": ["Generic bounds on the approximation error for physics-informed (and) operator learning", "A Priori Generalization Error Analysis of Two-Layer Neural Networks for Solving High Dimensional Schrödinger Eigenvalue Problems", "A Priori Generalization Analysis of the Deep Ritz Method for Solving High Dimensional Elliptic Equations", "On the Representation of Solutions to Elliptic PDEs in Barron Spaces"], "answer_arxiv_id": ["2205.11393", "2105.01228", "2101.01708", "2106.07539"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_2017"} +{"question": "Which studies proposed composing diffusion models by adding their scores?", "answer": ["Compositional Visual Generation with Composable Diffusion Models", "Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMC"], "answer_arxiv_id": ["2206.01714", "2302.11552"], "source_meta": {"published_time": "20220928"}, "qid": "AutoScholarQuery_train_2018"} +{"question": "Which research works focus on using color cues for image matting?", "answer": ["Sparse Coding for Alpha Matting"], "answer_arxiv_id": ["1604.02898"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_2019"} +{"question": "Which work proposed a no-regret policy using a primal-dual framework for fair online resource allocation policies?", "answer": ["Enabling Long-term Fairness in Dynamic Resource Allocation"], "answer_arxiv_id": ["2208.05898"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_2020"} +{"question": "Which works looked into domain generalization by assuming access to several related datasets?", "answer": ["Domain Generalization via Invariant Feature Representation"], "answer_arxiv_id": ["1301.2115"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_2021"} +{"question": "Could you provide me some studies about pre-training on image-text pairs?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_2022"} +{"question": "Could you provide me the work that acted similarly by optimizing the ColBERT in the form of PLAID?", "answer": ["PLAID: An Efficient Engine for Late Interaction Retrieval"], "answer_arxiv_id": ["2205.09707"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_2023"} +{"question": "What is the research that explored dense depth map reconstruction with polarization?", "answer": ["Polarimetric Monocular Dense Mapping Using Relative Deep Depth Prior"], "answer_arxiv_id": ["2102.05212"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_2024"} +{"question": "What research papers investigated the in-context learning capabilities of Transformer models?", "answer": ["Language Models are Few-Shot Learners", "An Explanation of In-context Learning as Implicit Bayesian Inference", "What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "Transformers Learn In-Context by Gradient Descent", "What learning algorithm is in-context learning? Investigations with linear models", "Looped Transformers as Programmable Computers"], "answer_arxiv_id": ["2005.14165", "2111.02080", "2208.01066", "2212.07677", "2211.15661", "2301.13196v1"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_2025"} +{"question": "What paper defined an intervention which induces an interventional graph where all incoming arcs to vertices are removed?", "answer": ["On the Number of Experiments Sufficient and in the Worst Case Necessary to Identify All Causal Relations Among N Variables"], "answer_arxiv_id": ["1207.1389v1"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2026"} +{"question": "Are there any works that highlight how model-based RL performs on benchmarks like the Deepmind Control Suite and Atari?", "answer": ["Dream to Control: Learning Behaviors by Latent Imagination", "Model Based Reinforcement Learning for Atari", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1912.01603", "1903.00374", "2010.02193"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_2027"} +{"question": "Can you list the articles applying data augmentation consistency regularization as a semi-supervised learning tool in medical imaging?", "answer": ["Semi-Supervised Medical Image Segmentation via Learning Consistency under Transformations", "Data augmentation using learned transformations for one-shot medical image segmentation", "Transformation-consistent Self-ensembling Model for Semi-supervised Medical Image Segmentation", "SSMD: Semi-Supervised Medical Image Detection with Adaptive Consistency and Heterogeneous Perturbation", "An Embarrassingly Simple Consistency Regularization Method for Semi-Supervised Medical Image Segmentation"], "answer_arxiv_id": ["1911.01218", "1902.09383", "1903.00348", "2106.01544", "2202.00677"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_2028"} +{"question": "Can you reference work that describes DINO's state-of-the-art results using a publicly available backbone and public datasets only?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2103.14030"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_2029"} +{"question": "Which works used continuous learnable RPEs represented as an MLP?", "answer": ["Point Transformer"], "answer_arxiv_id": ["2012.09164"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_2030"} +{"question": "What studies fine-tuned code LLms with high-quality instruction fine-tuning datasets?", "answer": ["WizardCoder: Empowering Code Large Language Models with Evol-Instruct", "Magicoder: Empowering Code Generation with OSS-Instruct", "Textbooks Are All You Need", "OctoPack: Instruction Tuning Code Large Language Models"], "answer_arxiv_id": ["2306.08568", "2312.02120", "2306.11644", "2308.07124"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_2031"} +{"question": "Which research papers show the usage of subtitles as text sources for pre-training of large-scale video datasets?", "answer": ["HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips", "Advancing High-Resolution Video-Language Representation with Large-Scale Video Transcriptions"], "answer_arxiv_id": ["1906.03327", "2111.10337"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_train_2032"} +{"question": "Which research papers designed LLM-based agents for user interface interaction?", "answer": ["Language Models can Solve Computer Tasks"], "answer_arxiv_id": ["2303.17491"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_2033"} +{"question": "Which work discovered that combining synthetic annotated datasets with real data can significantly improve the performance of instance detection?", "answer": ["Cut, Paste and Learn: Surprisingly Easy Synthesis for Instance Detection"], "answer_arxiv_id": ["1708.01642"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_2034"} +{"question": "What works discussed differential privacy in the context of memorization?", "answer": ["Deep Learning with Differential Privacy", "Large-Scale Differentially Private BERT"], "answer_arxiv_id": ["1607.00133", "2108.01624"], "source_meta": {"published_time": "20220215"}, "qid": "AutoScholarQuery_train_2035"} +{"question": "Which works utilize the idea of mask classification for instance segmentation?", "answer": ["InstanceCut: from Edges to Instances with MultiCut", "Path Aggregation Network for Instance Segmentation", "MaskLab: Instance Segmentation by Refining Object Detection with Semantic and Direction Features", "Cascade R-CNN: Delving into High Quality Object Detection", "YOLACT Real-time Instance Segmentation", "Hybrid Task Cascade for Instance Segmentation", "Conditional Convolutions for Instance Segmentation", "SOLOv2: Dynamic and Fast Instance Segmentation", "DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution"], "answer_arxiv_id": ["1611.08272", "1803.01534", "1712.04837", "1712.00726", "1904.02689", "1901.07518", "2003.05664", "2003.10152", "2006.02334"], "source_meta": {"published_time": "20230804"}, "qid": "AutoScholarQuery_train_2036"} +{"question": "Could you provide me some works about the cross-questioning paradigm for assessing factuality and relevance of generations?", "answer": ["Asking and Answering Questions to Evaluate the Factual Consistency of\n Summaries", "FEQA: A Question Answering Evaluation Framework for Faithfulness\n Assessment in Abstractive Summarization", "Towards Question-Answering as an Automatic Metric for Evaluating the\n Content Quality of a Summary"], "answer_arxiv_id": ["2004.04228", "2005.03754", "2010.00490"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_2037"} +{"question": "Could you provide me some works that have proposed an attention-guided network for enhancing low-light images?", "answer": ["EnlightenGAN: Deep Light Enhancement without Paired Supervision"], "answer_arxiv_id": ["1906.06972"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_2038"} +{"question": "What studies have examined stochastic decentralized methods for problems with a nonsmooth term?", "answer": ["Convergence of a Multi-Agent Projected Stochastic Gradient Algorithm for Non-Convex Optimization", "Distributed Stochastic Consensus Optimization with Momentum for Nonconvex Nonsmooth Problems", "A Stochastic Proximal Gradient Framework for Decentralized Non-Convex Composite Optimization: Topology-Independent Sample Complexity and Communication Efficiency", "Proximal Stochastic Recursive Momentum Methods for Nonconvex Composite Decentralized Optimization"], "answer_arxiv_id": ["1107.2526", "2011.05082", "2110.01594v1", "2211.11954"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_2039"} +{"question": "Which papers proposed new object representations for addressing the issue of arbitrary orientations in remote sensing images?", "answer": ["Gliding vertex on the horizontal bounding box for multi-oriented object\n detection", "Oriented RepPoints for Aerial Object Detection", "Oriented Object Detection in Aerial Images with Box Boundary-Aware\n Vectors", "Dense Label Encoding for Boundary Discontinuity Free Rotation Detection"], "answer_arxiv_id": ["1911.09358", "2105.11111", "2008.07043", "2011.09670"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_2040"} +{"question": "What are some studies related to model editing that have focused on intervening on the latent space of neural networks?", "answer": ["Editable Neural Networks", "Rewriting a Deep Generative Model", "Editing a classifier by rewriting its prediction rules"], "answer_arxiv_id": ["2004.00345", "2007.15646", "2112.01008"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_2041"} +{"question": "What research works demonstrated that gradient descent and its variants can avoid saddle points and converge to a second-order stationary point?", "answer": ["First-order methods Almost Always Avoid Saddle points: The case of Vanishing Step-sizes", "How to Escape Saddle Points Efficiently", "Sharp Analysis for Nonconvex SGD Escaping from Saddle Points", "Escaping Saddles with Stochastic Gradients"], "answer_arxiv_id": ["1906.07772", "1703.00887", "1902.00247", "1803.05999"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_2042"} +{"question": "What works have utilized the Zen-score to approximate the gradient with respect to featuremaps and measure the complexity of neural networks?", "answer": ["Zen-NAS: A Zero-Shot NAS for High-Performance Image Recognition"], "answer_arxiv_id": ["2102.01063"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_2043"} +{"question": "Which papers showed statistical, robustness, and computational guarantees for sliced Wasserstein distances?", "answer": ["Statistical, Robustness, and Computational Guarantees for Sliced Wasserstein Distances"], "answer_arxiv_id": ["2210.09160"], "source_meta": {"published_time": "20220927"}, "qid": "AutoScholarQuery_train_2044"} +{"question": "Which papers use frame-by-frame target object segmentation in referring video segmentation?", "answer": ["Video Object Segmentation with Language Referring Expressions", "MAttNet: Modular Attention Network for Referring Expression\n Comprehension"], "answer_arxiv_id": ["1803.08006", "1801.08186"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_2045"} +{"question": "Which method uses mixture density networks to estimate a distribution over 3D keypoints conditioned on observed 2D keypoint locations?", "answer": ["Generating Multiple Hypotheses for 3D Human Pose Estimation with Mixture\n Density Network"], "answer_arxiv_id": ["1904.05547"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_2046"} +{"question": "What work is most similar to ours in the context of text-based video generation using latent diffusion model?", "answer": ["LLM-grounded Video Diffusion Models"], "answer_arxiv_id": ["2309.17444"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2047"} +{"question": "Which works study SSL with linear models and kernel methods?", "answer": ["Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods", "Joint Embedding Self-Supervised Learning in the Kernel Regime", "The SSL Interplay: Augmentations, Inductive Bias, and Generalization"], "answer_arxiv_id": ["2205.11508", "2209.14884", "2302.02774"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_2048"} +{"question": "Which research uses a visual-semantic graph and applies a graph neural network for label refinement?", "answer": ["Webly Supervised Image Classification with Metadata: Automatic Noisy Label Correction via Visual-Semantic Graph"], "answer_arxiv_id": ["2010.05864"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_2049"} +{"question": "What work enhanced usability of LLM by fine-tuning with detailed image descriptions?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2304.10592"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2050"} +{"question": "What research work proposed the use of encoder-only models with a masked language modeling objective for code language models?", "answer": ["CodeBERT: A Pre-Trained Model for Programming and Natural Languages", "GraphCodeBERT: Pre-training Code Representations with Data Flow", "Learning and Evaluating Contextual Embedding of Source Code"], "answer_arxiv_id": ["2002.08155", "2009.08366", "2001.00059"], "source_meta": {"published_time": "20220626"}, "qid": "AutoScholarQuery_train_2051"} +{"question": "Which work does the researcher feel is similar to their approach of using relaxed symmetries for differentiable search over architectural structures?", "answer": ["DARTS: Differentiable Architecture Search"], "answer_arxiv_id": ["1806.09055"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_2052"} +{"question": "What processes are proposed in AutoShape and MonoJSG for 3D detection?", "answer": ["AutoShape: Real-Time Shape-Aware Monocular 3D Object Detection", "MonoJSG: Joint Semantic and Geometric Cost Volume for Monocular 3D\n Object Detection"], "answer_arxiv_id": ["2108.11127", "2203.08563"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_2053"} +{"question": "Could you provide me some works about object-centric approaches in dynamics prediction?", "answer": ["Learning Visual Predictive Models of Physics for Playing Billiards", "A Compositional Object-Based Approach to Learning Physical Dynamics", "Compositional Video Prediction"], "answer_arxiv_id": ["1511.07404", "1612.00341", "1908.08522"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_2054"} +{"question": "Which papers combined camera and microphone signals to estimate entire floor plans?", "answer": ["Audio-Visual Floorplan Reconstruction"], "answer_arxiv_id": ["2012.15470"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_2055"} +{"question": "What are the significant works in the field of category-level pose estimation?", "answer": ["FS-Net: Fast Shape-based Network for Category-Level 6D Object Pose\n Estimation with Decoupled Rotation Mechanism", "GPV-Pose: Category-level Object Pose Estimation via Geometry-guided\n Point-wise Voting", "Shape Prior Deformation for Categorical 6D Object Pose and Size\n Estimation", "CPS++: Improving Class-level 6D Pose and Shape Estimation From Monocular\n Images With Self-Supervised Learning", "CenterSnap: Single-Shot Multi-Object 3D Shape Reconstruction and\n Categorical 6D Pose and Size Estimation", "Learning Canonical Shape Space for Category-Level 6D Object Pose and\n Size Estimation", "DualPoseNet: Category-level 6D Object Pose and Size Estimation Using\n Dual Pose Network with Refined Learning of Pose Consistency"], "answer_arxiv_id": ["2103.07054", "2203.07918", "2007.08454", "2003.05848", "2203.01929", "2001.09322", "2103.06526"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_2056"} +{"question": "What studies demonstrate that pretrained Language Models (LMs) can generalize to tasks with different modalities?", "answer": ["Pretrained Transformers As Universal Computation Engines"], "answer_arxiv_id": ["2103.05247"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2057"} +{"question": "Could you mention studies about quantization challenges for large tensor like in CLIP ViT-Huge networks?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale"], "answer_arxiv_id": ["2208.07339"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_2058"} +{"question": "Which studies extended Net2Net to transformers?", "answer": ["bert2BERT: Towards Reusable Pretrained Language Models"], "answer_arxiv_id": ["2110.07143"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_2059"} +{"question": "What papers deal with prototype-based learning in various learning setups, including few-shot, zero-shot and unsupervised learning?", "answer": ["Prototypical Networks for Few-shot Learning", "Attribute Prototype Network for Zero-Shot Learning"], "answer_arxiv_id": ["1703.05175", "2008.08290"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_2060"} +{"question": "What papers applied the concept of aligning models with human intentions to train virtual robots or Atari games?", "answer": ["Deep reinforcement learning from human preferences", "Reward learning from human preferences and demonstrations in Atari"], "answer_arxiv_id": ["1706.03741", "1811.06521"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_2061"} +{"question": "Which works studied implicit bias for regression tasks?", "answer": ["(S)GD over Diagonal Linear Networks: Implicit Regularisation, Large Stepsizes and Edge of Stability", "Saddle-to-Saddle Dynamics in Diagonal Linear Networks"], "answer_arxiv_id": ["2302.08982v2", "2304.00488"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_2062"} +{"question": "What works focus on dynamically selecting important tokens for different inputs in terms of ViT pruning?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "IA-RED$^2$: Interpretability-Aware Redundancy Reduction for Vision\n Transformers"], "answer_arxiv_id": ["2106.02034v2", "2106.12620"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_2063"} +{"question": "Which papers proposed other types of scene representation methods such as point clouds or anisotropic 3D Gaussians?", "answer": ["Point-NeRF: Point-based Neural Radiance Fields", "3D Gaussian Splatting for Real-Time Radiance Field Rendering", "Mixture of Volumetric Primitives for Efficient Neural Rendering"], "answer_arxiv_id": ["2201.08845", "2308.04079", "2103.01954"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2064"} +{"question": "Could you give me works which propose large-scale language or multimodal models based on existing MoE architectures?", "answer": ["M6: A Chinese Multimodal Pretrainer", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts"], "answer_arxiv_id": ["2103.00823", "2112.06905"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_2065"} +{"question": "What studies developed solutions to implicit layers that can enforce equality constraints between the layer’s input and output?", "answer": ["Implicit Deep Learning"], "answer_arxiv_id": ["1908.06315"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2066"} +{"question": "Are there any papers that discuss sequential editing?", "answer": ["Do Language Models Have Beliefs? Methods for Detecting, Updating, and Visualizing Model Beliefs", "Transformer-Patcher: One Mistake worth One Neuron"], "answer_arxiv_id": ["2111.13654", "2301.09785"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_2067"} +{"question": "Any work about ResNets being viewed as dynamic transport systems?", "answer": ["Optimal Unsupervised Domain Translation", "A Principle of Least Action for the Training of Neural Networks"], "answer_arxiv_id": ["1906.01292v1", "2009.08372"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_2068"} +{"question": "Who proposed the idea of SpotTune regarding the choice of weights to fine-tune?", "answer": ["SpotTune: Transfer Learning through Adaptive Fine-tuning"], "answer_arxiv_id": ["1811.08737"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_2069"} +{"question": "What research has investigated the use of optical flow for video stabilization?", "answer": ["Out-of-boundary View Synthesis Towards Full-Frame Video Stabilization", "DUT: Learning Video Stabilization by Simply Watching Unstable Videos", "Deep Iterative Frame Interpolation for Full-frame Video Stabilization", "Hybrid Neural Fusion for Full-frame Video Stabilization"], "answer_arxiv_id": ["2108.09041", "2011.14574", "1909.02641", "2102.06205"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_2070"} +{"question": "What papers use AvatarCLIP for text-guided 3D human generation?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars"], "answer_arxiv_id": ["2205.08535"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_2071"} +{"question": "Which research works have shown that scaling the number of training tasks, prompts per task, and size of the LM helps in improving zero-shot task generalization performance?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization", "Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks"], "answer_arxiv_id": ["2110.08207", "2204.07705v3"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_2072"} +{"question": "What researchers have greatly influenced the text-to-image content with the help of Large language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Language Models are Few-Shot Learners", "GPT-4 Technical Report"], "answer_arxiv_id": ["1810.04805", "2005.14165", "2303.08774"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_2073"} +{"question": "What researchers have tried to estimate mutual information in high dimensional feature spaces in SSL?", "answer": ["Mutual Information Neural Estimation"], "answer_arxiv_id": ["1801.04062"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_2074"} +{"question": "Any studies about the introduction of physics-constrained loss for high-dimensional surrogate modeling?", "answer": ["Physics-Constrained Deep Learning for High-dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data"], "answer_arxiv_id": ["1901.06314"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_2075"} +{"question": "What works have used weak supervision for geometric matching and semantic correspondence learning?", "answer": ["DGC-Net: Dense Geometric Correspondence Network", "Dual-Resolution Correspondence Networks", "Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions", "RANSAC-Flow: generic two-stage image alignment", "GLU-Net: Global-Local Universal Network for Dense Flow and Correspondences", "Proposal Flow", "End-to-end weakly-supervised semantic alignment", "Neighbourhood Consensus Networks", "Dynamic Context Correspondence Network for Semantic Alignment", "Warp Consistency for Unsupervised Learning of Dense Correspondences"], "answer_arxiv_id": ["1810.08393", "2006.08844", "2004.10566", "2004.01526", "1912.05524", "1511.05065", "1712.06861", "1810.10510", "1909.03444", "2104.03308"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_2076"} +{"question": "Which studies explored integrating Large Language Models (LLMs) into vision-language models?", "answer": ["Multimodal Large Language Models: A Survey"], "answer_arxiv_id": ["2311.13165"], "source_meta": {"published_time": "20240630"}, "qid": "AutoScholarQuery_train_2077"} +{"question": "Can you share references that discuss backdoor removal methods like Fine-tuning or Fine-pruning?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks"], "answer_arxiv_id": ["1805.12185"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_2078"} +{"question": "What work proposed that the graph-level functions (and in the context of SC, k-forms) should remain invariant to both sign and basis (either of orientation or of space)?", "answer": ["Sign and Basis Invariant Networks for Spectral Graph Representation Learning"], "answer_arxiv_id": ["2202.13013"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_2079"} +{"question": "What are some papers where model-based reinforcement learning has shown progress in many domains?", "answer": ["Model Based Reinforcement Learning for Atari", "MOPO: Model-based Offline Policy Optimization", "Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models"], "answer_arxiv_id": ["1903.00374", "2005.13239", "1805.12114"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2080"} +{"question": "What papers investigate the use of Structural Causal Models (SCMs) for data augmentation in continuous spaces?", "answer": ["Sample-Efficient Reinforcement Learning via Counterfactual-Based Data Augmentation"], "answer_arxiv_id": ["2012.09092"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_2081"} +{"question": "What works have applied the CUT Loss to StyleGAN and Diffusion models for image translation and style transfer?", "answer": ["One-Shot Adaptation of GAN in Just One CLIP", "Zero-Shot Contrastive Loss for Text-Guided Diffusion Image Style\n Transfer", "Diffusion-based Image Translation using Disentangled Style and Content\n Representation"], "answer_arxiv_id": ["2203.09301", "2303.08622", "2209.15264"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2082"} +{"question": "Which papers propose long-term tracking methods?", "answer": ["Siam R-CNN: Visual Tracking by Re-Detection", "Learning Target Candidate Association to Keep Track of What Not to Track", "Learning Spatio-Temporal Transformer for Visual Tracking"], "answer_arxiv_id": ["1911.12836", "2103.16556", "2103.17154"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2083"} +{"question": "Any studies that have shown that well-trained deep networks can also leak private samples?", "answer": ["Reconstructing Training Data from Trained Neural Networks"], "answer_arxiv_id": ["2206.07758"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_2084"} +{"question": "Any works about layered depth images and the prediction of an extra depth channel?", "answer": ["Deep Multi Depth Panoramas for View Synthesis", "Worldsheet: Wrapping the World in a 3D Sheet for View Synthesis from a\n Single Image", "Stereo Magnification with Multi-Layer Images", "Self-improving Multiplane-to-layer Images for Novel View Synthesis"], "answer_arxiv_id": ["2008.01815", "2012.09854", "2201.05023", "2210.01602"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_2085"} +{"question": "What works utilized 3D CNN for voxel processing in voxel-based methods?", "answer": ["VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "PointPillars: Fast Encoders for Object Detection from Point Clouds"], "answer_arxiv_id": ["1711.06396", "1812.05784"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_2086"} +{"question": "Which studies introduce the Gumbel-Max SCM for the dynamics of an arbitrary discrete POMPD?", "answer": ["Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models"], "answer_arxiv_id": ["1905.05824"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_2087"} +{"question": "Which work develops an embodied multi-modal language model to resolve a broad range of tasks in robotic planning?", "answer": ["PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2303.03378"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_2088"} +{"question": "Can you provide the research that developed sharp recovery guarantees of alternating minimization for generalized rank-one matrix sensing?", "answer": ["Alternating minimization for generalized rank one matrix sensing: Sharp predictions from a random initialization"], "answer_arxiv_id": ["2207.09660v1"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_2089"} +{"question": "Could you provide me with some works that incorporate volume rendering to make GANs 3D-aware?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_2090"} +{"question": "Could you mention studies about multi-objective multi-task RL that learns a joint policy by sharing parameters?", "answer": ["Learning to Push by Grasping: Using multiple tasks for effective learning"], "answer_arxiv_id": ["1609.09025"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_2091"} +{"question": "Are there any studies that have revealed the vulnerabilities of existing detectors in deepfake text detection?", "answer": ["Can AI-Generated Text be Reliably Detected?", "Paraphrasing evades detectors of AI-generated text, but retrieval is an\n effective defense"], "answer_arxiv_id": ["2303.11156", "2303.13408"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_2092"} +{"question": "Can you list studies that applied diffusion models in various applications?", "answer": ["Diffusion Models: A Comprehensive Survey of Methods and Applications", "Diffusion Models in Vision: A Survey", "A Survey on Generative Diffusion Model"], "answer_arxiv_id": ["2209.00796", "2209.04747", "2209.02646"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_2093"} +{"question": "Any studies focused on modeling deformation explicitly in novel view synthesis for dynamic scenes?", "answer": ["Neural Radiance Flow for 4D View Synthesis and Video Processing", "Dynamic View Synthesis from Dynamic Monocular Video", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Space-time Neural Irradiance Fields for Free-Viewpoint Video", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "STaR: Self-supervised Tracking and Reconstruction of Rigid Objects in Motion with Neural Rendering"], "answer_arxiv_id": ["2012.09790", "2105.06468", "2011.13084", "2011.12950", "2011.13961", "2101.01602"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2094"} +{"question": "Could you give me examples of conferences that used randomization to evaluate the benefits of double-blind reviewing?", "answer": ["A Large Scale Randomized Controlled Trial on Herding in Peer-Review Discussions"], "answer_arxiv_id": ["2011.15083"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_2095"} +{"question": "What works aim to find invariant representation from different training environments for out-of-distribution (OOD) generalization?", "answer": ["Invariant Risk Minimization", "Out-of-Distribution Generalization via Risk Extrapolation (REx)", "Invariant Risk Minimization Games"], "answer_arxiv_id": ["1907.02893", "2003.00688v5", "2002.04692"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2096"} +{"question": "What research proposed the use of Combinatorial Brain Surgeon (CBS) for addressing issues arising from pruning multiple weights?", "answer": ["The Combinatorial Brain Surgeon: Pruning Weights That Cancel One Another in Neural Networks"], "answer_arxiv_id": ["2203.04466v3"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_2097"} +{"question": "Which studies initially applied DDPM to unconditional point cloud generation?", "answer": ["Diffusion Probabilistic Models for 3D Point Cloud Generation"], "answer_arxiv_id": ["2103.01458"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_2098"} +{"question": "What papers discuss the use of machine learning methods to improve the MILP solver performance?", "answer": ["Exact Combinatorial Optimization with Graph Convolutional Neural Networks"], "answer_arxiv_id": ["1906.01629"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_2099"} +{"question": "Can you mention some research about drag-based editing with diffusion models?", "answer": ["DragDiffusion: Harnessing Diffusion Models for Interactive Point-based\n Image Editing", "DragonDiffusion: Enabling Drag-style Manipulation on Diffusion Models"], "answer_arxiv_id": ["2306.14435", "2307.02421"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2100"} +{"question": "In what study was an exploration algorithm with inflated UCB proposed?", "answer": ["Bridging the gap between regret minimization and best arm identification, with application to A/B tests"], "answer_arxiv_id": ["1810.04088v2"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_2101"} +{"question": "Could you provide me some references that concentrated on pixel-level data augmentation through mixing images using pixel-wise weighted averages?", "answer": ["mixup: Beyond Empirical Risk Minimization", "Manifold Mixup: Better Representations by Interpolating Hidden States", "PixMix: Dreamlike Pictures Comprehensively Improve Safety Measures"], "answer_arxiv_id": ["1710.09412", "1806.05236", "2112.05135"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_2102"} +{"question": "Which paper proposed the Structured Diffusion method to address the issue of missing entities and semantic leakage of attributes?", "answer": ["Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesis"], "answer_arxiv_id": ["2212.05032"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2103"} +{"question": "Could you provide me some studies that propose to learn a conditional generative model via the cVAE framework for aleatoric uncertainty estimation in semantic segmentation?", "answer": ["A Probabilistic U-Net for Segmentation of Ambiguous Images", "Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1806.05034", "1312.6114"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_2104"} +{"question": "Could you provide some researches that introduced additional control signals for image editing or controllable image generation?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "Palette: Image-to-Image Diffusion Models", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2302.05543", "2111.05826", "2211.12572"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_2105"} +{"question": "What papers extend the GNN-based models to particle-based graph surrogate models?", "answer": ["Learning to Simulate Complex Physics with Graph Networks"], "answer_arxiv_id": ["2002.09405"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_2106"} +{"question": "Can you provide works that studied local or fairness ideals for correlation clustering?", "answer": ["Fair Correlation Clustering"], "answer_arxiv_id": ["2002.02274"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2107"} +{"question": "Which works focus on the adaptation of agents to unseen environment dynamics?", "answer": ["Policy Transfer with Strategy Optimization", "Learning to Learn How to Learn: Self-Adaptive Visual Navigation using Meta-Learning", "Environment Probing Interaction Policies", "Unsupervised Domain Adaptation for Visual Navigation", "Self-Supervised Deep Visual Odometry with Online Adaptation", "Learning Agile Robotic Locomotion Skills by Imitating Animals", "Rapidly Adaptable Legged Robots via Evolutionary Meta-Learning", "Learning High-Speed Flight in the Wild", "Context is Everything: Implicit Identification for Dynamics Adaptation", "Neural-Fly Enables Rapid Learning for Agile Flight in Strong Winds", "RMA: Rapid Motor Adaptation for Legged Robots", "Adapting Rapid Motor Adaptation for Bipedal Robots", "Legged Locomotion in Challenging Terrains using Egocentric Vision"], "answer_arxiv_id": ["1810.05751", "1812.00971", "1907.11740", "2010.14543", "2005.06136", "2004.00784", "2003.01239", "2110.05113", "2203.05549", "2205.06908", "2107.04034", "2205.15299", "2211.07638"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_2108"} +{"question": "What work proposed training on multi-step PGD adversaries and found that this method leads to robust training loss in wide neural networks?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_2109"} +{"question": "What works have contributed to the recent breakthroughs in text-to-image models primarily driven by advances in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Diffusion Models: A Comprehensive Survey of Methods and Applications"], "answer_arxiv_id": ["2006.11239", "2011.13456", "2209.00796"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2110"} +{"question": "What work described the Stealthy and Physical-Object-Oriented attack which relies on a printable adversarial patch?", "answer": ["Physical Attack on Monocular Depth Estimation with Optimal Adversarial\n Patches"], "answer_arxiv_id": ["2207.04718"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_2111"} +{"question": "Could you provide the works that established convergence in high probability of AdaGrad-Norm without further restrictive assumptions?", "answer": ["AdaGrad stepsizes: Sharp convergence over nonconvex landscapes", "The Power of Adaptivity in SGD: Self-Tuning Step Sizes with Unbounded Gradients and Affine Variance"], "answer_arxiv_id": ["1806.01811", "2202.05791"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_2112"} +{"question": "Could you provide me the study that discussed the performance of the wake-sleep algorithm for tasks involving stochastic branching?", "answer": ["Revisiting Reweighted Wake-Sleep for Models with Stochastic Control Flow"], "answer_arxiv_id": ["1805.10469"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_2113"} +{"question": "What works have introduced the concept of 'Data Shapley' to assess the value of individual data instances?", "answer": ["Data Shapley: Equitable Valuation of Data for Machine Learning", "Efficient computation and analysis of distributional Shapley values"], "answer_arxiv_id": ["1904.02868v2", "2007.01357"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_2114"} +{"question": "Which paper introduced DAGFormer designed for directed acyclic graphs?", "answer": ["Transformers over Directed Acyclic Graphs"], "answer_arxiv_id": ["2210.13148"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_2115"} +{"question": "What research paper provides a solution to reduce the weight shift during continual learning by introducing an additional surrogate loss?", "answer": ["Continual Learning Through Synaptic Intelligence"], "answer_arxiv_id": ["1703.04200"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_2116"} +{"question": "Which studies discuss the use of StyleGAN in image editing?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN", "Alias-Free Generative Adversarial Networks"], "answer_arxiv_id": ["1812.04948", "1912.04958", "2106.12423"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_2117"} +{"question": "Are there any works which introduced exploration of the denoiser through an encoded clean version of its own target?", "answer": ["Diffusion-Based Representation Learning", "Scheduled denoising autoencoders"], "answer_arxiv_id": ["2105.14257", "1406.3269"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_2118"} +{"question": "Could you name studies focusing on the evaluation of Large Language Models in code generation?", "answer": ["Evaluating Large Language Models Trained on Code", "Program Synthesis with Large Language Models", "ClassEval: A Manually-Crafted Benchmark for Evaluating LLMs on\n Class-level Code Generation"], "answer_arxiv_id": ["2107.03374", "2108.07732", "2308.01861"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_2119"} +{"question": "Can you provide references to recent studies aiming to improve Adam optimizers?", "answer": ["AdaX: Adaptive Gradient Descent with Exponential Long Term Memory", "AdaFamily: A family of Adam-like adaptive gradient methods"], "answer_arxiv_id": ["2004.09740", "2203.01603"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_2120"} +{"question": "Which works have advanced the training of multi-modal chatbots by generating an instruction-tuning dataset from image captions?", "answer": ["Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model"], "answer_arxiv_id": ["2304.08485", "2304.10592", "2304.15010"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2121"} +{"question": "Could you provide me with the work that proposed an instance-conditioned variational auto-encoder for routing problems?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_2122"} +{"question": "What works have attempted to incorporate scene context to model diverse visual concepts for SGG?", "answer": ["Neural Motifs: Scene Graph Parsing with Global Context", "Learning to Compose Dynamic Tree Structures for Visual Contexts", "Bridging Knowledge Graphs to Generate Scene Graphs", "RU-Net: Regularized Unrolling Network for Scene Graph Generation"], "answer_arxiv_id": ["1711.06640", "1812.01880", "2001.02314", "2205.01297"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_2123"} +{"question": "Any works about using volunteer computing as an alternative in HPC environment?", "answer": ["Distributed Deep Learning Using Volunteer Computing-Like Paradigm"], "answer_arxiv_id": ["2103.08894"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_2124"} +{"question": "Could you provide me some examples of Transformer-based models in natural language processing?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Language Models are Few-Shot Learners", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "2005.14165", "1910.13461", "1907.11692"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_2125"} +{"question": "Which work shows that large language models forget examples seen early in training?", "answer": ["Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Models"], "answer_arxiv_id": ["2205.10770"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_2126"} +{"question": "What work found a correlation between the number of decoder parameters in autoregressive Transformers and task performance?", "answer": ["LiteTransformerSearch: Training-free Neural Architecture Search for Efficient Language Models"], "answer_arxiv_id": ["2203.02094"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_2127"} +{"question": "Could you provide some references associated with the synthesis of human dance motions from audio signals?", "answer": ["Listen, Denoise, Action! Audio-Driven Motion Synthesis with Diffusion\n Models", "EDGE: Editable Dance Generation From Music", "Pretrained Diffusion Models for Unified Human Motion Synthesis", "MoFusion: A Framework for Denoising-Diffusion-based Motion Synthesis"], "answer_arxiv_id": ["2211.09707", "2211.10658", "2212.02837", "2212.04495"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_2128"} +{"question": "Which pre-trained models have been used for tasks such as zero-shot classification, image editing, open-world segmentation, and 3D classification?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Zero-Shot Out-of-Distribution Detection Based on the Pre-trained Model\n CLIP", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "Blended Diffusion for Text-driven Editing of Natural Images", "ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation", "CRIS: CLIP-Driven Referring Image Segmentation", "PointCLIP: Point Cloud Understanding by CLIP"], "answer_arxiv_id": ["2103.00020", "2109.02748", "2111.07991", "2103.17249", "2111.14818", "2212.03588", "2111.15174", "2112.02413"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_2129"} +{"question": "Which papers discuss the use of score-based diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "1907.05600", "2006.11239"], "source_meta": {"published_time": "20220912"}, "qid": "AutoScholarQuery_train_2130"} +{"question": "Could you provide me some references about methods based on data augmentation for dealing with known spurious attributes?", "answer": ["Explaining the Efficacy of Counterfactually Augmented Data", "Explaining the Efficacy of Counterfactually Augmented Data"], "answer_arxiv_id": ["2010.02114", "2010.02114"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_2131"} +{"question": "Which paper demonstrated that learning ReLU regression is NP-hard without distribution assumption?", "answer": ["Tight Hardness Results for Training Depth-2 ReLU Networks"], "answer_arxiv_id": ["2011.13550v1"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_2132"} +{"question": "What studies have contributed to the history of non-convex constrained optimization?", "answer": ["Optimality of orders one to three and beyond: characterization and evaluation complexity in constrained nonconvex optimization"], "answer_arxiv_id": ["1705.07285"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2133"} +{"question": "What papers focus on altering material perception in images using Generative Adversarial Networks (GANs)?", "answer": ["In-the-wild Material Appearance Editing using Perceptual Attributes"], "answer_arxiv_id": ["2302.03619"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_2134"} +{"question": "What research works focus on integrating an adapter or projection layer to align the embedding spaces of various modal encoders with the text embedding space of the LLM?", "answer": ["NExT-GPT: Any-to-Any Multimodal LLM", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2309.05519", "2304.10592"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_2135"} +{"question": "What works discuss using hidden test cases in test-driven development?", "answer": ["Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2203.07814"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_2136"} +{"question": "Which works state that Mixup helps with robust representation learning and alleviates overconfident problems?", "answer": ["mixup: Beyond Empirical Risk Minimization", "On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networks", "UMix: Improving Importance Weighting for Subpopulation Shift via Uncertainty-Aware Mixup", "Reweighted Mixup for Subpopulation Shift"], "answer_arxiv_id": ["1710.09412", "1905.11001", "2209.08928", "2304.04148"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_2137"} +{"question": "Could you provide me some studies that utilized YouTube videos for large-scale pre-training in the video game Minecraft?", "answer": ["MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledge", "Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos"], "answer_arxiv_id": ["2206.08853", "2206.11795"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_2138"} +{"question": "Which research papers presented the concept of using short videos of around 5-10 seconds in spatio-temporal grounding study?", "answer": ["Weakly-Supervised Spatio-Temporally Grounding Natural Sentence in Video"], "answer_arxiv_id": ["1906.02549"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_2139"} +{"question": "Which papers initially employed prompt learning to devise additional text instructions for large-scale NLP models?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2107.13586v1"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2140"} +{"question": "Could you provide some studies about modeling molecules as 3D objects?", "answer": ["3DMolNet: A Generative Network for Molecular Structures", "Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules", "E(n) Equivariant Normalizing Flows", "Generating valid Euclidean distance matrices", "Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2010.06477", "1906.00957", "2105.09016", "1910.03131", "2203.17003"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2141"} +{"question": "What study proposed ConVIRT to contrast the radiograph features with latent embeddings of sentences in radiology reports?", "answer": ["Contrastive Learning of Medical Visual Representations from Paired Images and Text"], "answer_arxiv_id": ["2010.00747"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2142"} +{"question": "Any research papers that focus on modeling the spatial correspondence between pose and appearance?", "answer": ["Progressive Pose Attention Transfer for Person Image Generation", "XingGAN for Person Image Generation", "Dense Intrinsic Appearance Flow for Human Pose Transfer", "Deep Image Spatial Transformation for Person Image Generation", "Liquid Warping GAN: A Unified Framework for Human Motion Imitation,\n Appearance Transfer and Novel View Synthesis", "Neural Texture Extraction and Distribution for Controllable Person Image\n Synthesis"], "answer_arxiv_id": ["1904.03349", "2007.09278", "1903.11326", "2003.00696", "1909.12224", "2204.06160"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_2143"} +{"question": "Can you name some works that includes a database search in the text generation process while working on retrieval-augmented language models?", "answer": ["A Retrieve-and-Edit Framework for Predicting Structured Outputs", "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT", "Baleen: Robust Multi-Hop Reasoning at Scale via Condensed Retrieval", "REALM: Retrieval-Augmented Language Model Pre-Training", "Improving language models by retrieving from trillions of tokens", "Few-shot Learning with Retrieval Augmented Language Models"], "answer_arxiv_id": ["1812.01194", "2004.12832", "2101.00436", "2002.08909", "2112.04426", "2208.03299"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_2144"} +{"question": "What are the papers that have applied self-supervised learning to facilitate RL research?", "answer": ["Auto-Encoding Variational Bayes", "Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["1312.6114", "1807.03748", "2002.05709", "1911.05722", "2006.07733"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2145"} +{"question": "Which research papers have obtained results for universal dynamic regret in OCO with memory?", "answer": ["Non-stationary Online Learning with Memory and Non-stochastic Control", "Revisiting Smoothed Online Learning"], "answer_arxiv_id": ["2102.03758", "2102.06933"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_2146"} +{"question": "Are there any studies on PDE-based learning where the model itself is a continuous diffusion process described by a differential equation?", "answer": ["GRAND: Graph Neural Diffusion", "PDE-GCN: Novel Architectures for Graph Neural Networks Motivated by Partial Differential Equations"], "answer_arxiv_id": ["2106.10934", "2108.01938"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_2147"} +{"question": "Any works about motion-centric video customization?", "answer": ["MotionDirector: Motion Customization of Text-to-Video Diffusion Models", "LAMP: Learn A Motion Pattern for Few-Shot-Based Video Generation"], "answer_arxiv_id": ["2310.08465", "2310.10769"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_2148"} +{"question": "Could you provide me some works that enhance translation capabilities of LLMs in the machine translation field?", "answer": ["No Language Left Behind: Scaling Human-Centered Machine Translation"], "answer_arxiv_id": ["2207.04672v3"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_2149"} +{"question": "Could you provide me some works about creating naturalistic patches using the learned image manifold of pre-trained GANs?", "answer": ["TnT Attacks! Universal Naturalistic Adversarial Patches Against Deep\n Neural Network Systems", "Feasibility of Inconspicuous GAN-generated Adversarial Patches against\n Object Detection"], "answer_arxiv_id": ["2111.09999", "2207.07347"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_2150"} +{"question": "Which papers have used image-conditioned methods to condition NeRF?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "Vision Transformer for NeRF-Based View Synthesis from a Single Input Image"], "answer_arxiv_id": ["2012.02190", "2207.05736"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_2151"} +{"question": "Which works discussed the application of diffusion models to downstream image synthesis tasks?", "answer": ["DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "Towards Practical Plug-and-Play Diffusion Models", "Diffusion models as plug-and-play priors"], "answer_arxiv_id": ["2110.02711", "2108.01073", "2212.05973", "2206.09012"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2152"} +{"question": "Can you provide works that apply QUD to other fields?", "answer": ["Elaborative Simplification as Implicit Questions Under Discussion", "A Question Answering Framework for Decontextualizing User-facing\n Snippets from Scientific Documents"], "answer_arxiv_id": ["2305.10387", "2305.14772"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_2153"} +{"question": "Could you tell me about the works that highlighted the importance of the initial period of learning in models of biological and artificial learning?", "answer": ["Critical Learning Periods in Deep Networks"], "answer_arxiv_id": ["1711.08856"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_2154"} +{"question": "Which research work proposes discriminative kernels by applying the clustering method?", "answer": ["DyCo3D: Robust Instance Segmentation of 3D Point Clouds through Dynamic\n Convolution"], "answer_arxiv_id": ["2011.13328"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_2155"} +{"question": "Could you provide me an example of a work that has recently tackled the removal of artifacts from any pre-trained NeRF in a post-processing step?", "answer": ["Nerfbusters: Removing Ghostly Artifacts from Casually Captured NeRFs"], "answer_arxiv_id": ["2304.10532"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_2156"} +{"question": "Which work optimizes the 3D generated mesh by dual normal maps?", "answer": ["Chupa: Carving 3D Clothed Humans from Skinned Shape Priors using 2D\n Diffusion Probabilistic Models"], "answer_arxiv_id": ["2305.11870"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_2157"} +{"question": "Can you list some studies about using pseudo-labels to leverage unlabeled data?", "answer": ["Weakly Supervised Semantic Segmentation for Large-Scale Point Cloud", "SQN: Weakly-Supervised Semantic Segmentation of Large-Scale 3D Point Clouds", "One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation", "Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs", "SSPC-Net: Semi-supervised Semantic 3D Point Cloud Segmentation Network"], "answer_arxiv_id": ["2212.04744", "2104.04891", "2104.02246", "1711.09869", "2104.07861"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_2158"} +{"question": "Which works have embedded semantic data from segmentation networks into 3D spaces?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene\n Representation", "Panoptic Lifting for 3D Scene Understanding with Neural Fields"], "answer_arxiv_id": ["2103.15875", "2212.09802"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_2159"} +{"question": "What research used anchor trajectories for predicting the offset in trajectory refinement?", "answer": ["Bootstrap Motion Forecasting With Self-Consistent Constraints"], "answer_arxiv_id": ["2204.05859"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_2160"} +{"question": "Any study discussing CyCTR which introduces a cycle-consistent attention mechanism in few-shot segmentation?", "answer": ["Few-Shot Segmentation via Cycle-Consistent Transformer"], "answer_arxiv_id": ["2106.02320"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_2161"} +{"question": "What works demonstrated that ETF is only applicable in a DNNs classification framework when the dimension of the space is greater or equal to the number of lines minus one?", "answer": ["Equiangular Lines and Spherical Codes in Euclidean Space", "A Geometric Analysis of Neural Collapse with Unconstrained Features"], "answer_arxiv_id": ["1606.06620", "2105.02375"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_2162"} +{"question": "Is there any concurrent work that drew a strong connection between over-squashing and Effective Resistance?", "answer": ["Understanding Oversquashing in GNNs through the Lens of Effective Resistance"], "answer_arxiv_id": ["2302.06835"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_2163"} +{"question": "Which papers propose learning latent visual-semantic relevance/similarities for cross-modal retrieval?", "answer": ["VSE++: Improving Visual-Semantic Embeddings with Hard Negatives", "Stacked Cross Attention for Image-Text Matching", "Similarity Reasoning and Filtration for Image-Text Matching", "Learning the Best Pooling Strategy for Visual Semantic Embedding", "CyCLIP: Cyclic Contrastive Language-Image Pretraining"], "answer_arxiv_id": ["1707.05612", "1803.08024", "2101.01368", "2011.04305", "2205.14459"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2164"} +{"question": "What work converts self-attention in diffusion models into a mutual and mask-guided self-attention strategy, enabling pose transformation?", "answer": ["MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing"], "answer_arxiv_id": ["2304.08465v1"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_2165"} +{"question": "Could you provide me some works that used a spatially-varying parameter to control the softness of the surface in 3D GANs?", "answer": ["Adaptive Shells for Efficient Neural Radiance Field Rendering", "Volume Rendering of Neural Implicit Surfaces"], "answer_arxiv_id": ["2311.10091", "2106.12052"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_2166"} +{"question": "What research utilizes knowledge distillation and replayed examples in Continual Learning?", "answer": ["iCaRL: Incremental Classifier and Representation Learning", "Revisiting Distillation and Incremental Classifier Learning"], "answer_arxiv_id": ["1611.07725", "1807.02802"], "source_meta": {"published_time": "20230326"}, "qid": "AutoScholarQuery_train_2167"} +{"question": "Are there any unsupervised pre-training methods for 3D object detection?", "answer": ["GD-MAE: Generative Decoder for MAE Pre-training on LiDAR Point Clouds", "ProposalContrast: Unsupervised Pre-training for LiDAR-based 3D Object\n Detection", "Robust Moving Objects Detection in Lidar Data Exploiting Visual Cues", "Learning to Detect Mobile Objects from LiDAR Scans Without Labels", "Towards Unsupervised Object Detection From LiDAR Point Clouds"], "answer_arxiv_id": ["2212.03010", "2207.12654", "1609.09267", "2203.15882", "2311.02007"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_2168"} +{"question": "What works tried to fuse the goal image with the intermediate feature maps of observation encoder using an affine transformation proposed by FiLM?", "answer": ["BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning", "Using Both Demonstrations and Language Instructions to Efficiently Learn Robotic Tasks"], "answer_arxiv_id": ["2202.02005", "2210.04476"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_2169"} +{"question": "What recent researches have utilized weight averaging which is motivated by mode connectivity?", "answer": ["On Using Very Large Target Vocabulary for Neural Machine Translation", "Regularizing and Optimizing LSTM Language Models", "Exploring Mode Connectivity for Pre-trained Language Models", "Linear Mode Connectivity in Multitask and Continual Learning"], "answer_arxiv_id": ["1412.2007", "1708.02182", "2210.14102", "2010.04495"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_2170"} +{"question": "What papers suggested novel weight learning methods for general Q-functions?", "answer": ["Double Reinforcement Learning for Efficient Off-Policy Evaluation in Markov Decision Processes", "Minimax Weight and Q-Function Learning for Off-Policy Evaluation"], "answer_arxiv_id": ["1908.08526", "1910.12809"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_2171"} +{"question": "Where can I find studies on the use of skill discovery during online unsupervised pretraining in RL?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function"], "answer_arxiv_id": ["1802.06070"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2172"} +{"question": "What are the works where grouping of points is based on predicted semantics and object centers?", "answer": ["PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation", "3D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentation", "Hierarchical Aggregation for 3D Instance Segmentation", "Instance Segmentation in 3D Scenes using Semantic Superpoint Tree\n Networks", "OccuSeg: Occupancy-aware 3D Instance Segmentation"], "answer_arxiv_id": ["2004.01658", "2003.13867", "2108.02350", "2108.07478", "2003.06537v3"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_2173"} +{"question": "Could you provide me some works involving fine-tuning the diffusion model for specific subjects?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_2174"} +{"question": "Could you provide me some works that focus on explicitly seeking a trade-off between the global model and the local models in PFL?", "answer": ["Federated Learning of a Mixture of Global and Local Models", "Think Locally, Act Globally: Federated Learning with Local and Global\n Representations", "Personalized Federated Learning through Local Memorization", "On Bridging Generic and Personalized Federated Learning for Image\n Classification"], "answer_arxiv_id": ["2002.05516", "2001.01523", "2111.09360", "2107.00778"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_2175"} +{"question": "What papers have explored more complex degradation models to approximate the real-world degradations in the Real-ISR?", "answer": ["Designing a Practical Degradation Model for Deep Blind Image\n Super-Resolution", "Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure\n Synthetic Data"], "answer_arxiv_id": ["2103.14006", "2107.10833"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_2176"} +{"question": "What are some examples of marker-less datasets that capture more complex poses for 3D HPS?", "answer": ["Monocular 3D Human Pose Estimation In The Wild Using Improved CNN\n Supervision", "AI Choreographer: Music Conditioned 3D Dance Generation with AIST++"], "answer_arxiv_id": ["1611.09813", "2101.08779"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2177"} +{"question": "Could you provide the research that proposed the functional form (y−ϵ∞)/((ϵ0−y)a)=b​xc in scaling laws?", "answer": ["Revisiting Neural Scaling Laws in Language and Vision"], "answer_arxiv_id": ["2209.06640"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_2178"} +{"question": "Could you provide some papers that discuss further developments on the line of research brought by 'bib.bib102' and 'bib.bib126'?", "answer": ["Balancing Stability and Plasticity through Advanced Null Space in Continual Learning", "Continual Learning with Recursive Gradient Optimization", "TRGP: Trust Region Gradient Projection for Continual Learning"], "answer_arxiv_id": ["2207.12061", "2201.12522", "2202.02931"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_2179"} +{"question": "What works provide datasets with longer videos of 1-5 minutes?", "answer": ["Towards Automatic Learning of Procedures from Web Instructional Videos", "Dense-Captioning Events in Videos", "Multimodal Pretraining for Dense Video Captioning"], "answer_arxiv_id": ["1703.09788", "1705.00754", "2011.11760"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_2180"} +{"question": "Which studies make use of lossy compression, such as through capacity-contrained variational auto-encoders, to distinguish semantic information in images?", "answer": ["Towards Conceptual Compression"], "answer_arxiv_id": ["1604.08772"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_2181"} +{"question": "Which works engage in the study of two-player zero-sum Markov games under the generative model?", "answer": ["Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity", "Minimax-Optimal Multi-Agent RL in Markov Games With a Generative Model"], "answer_arxiv_id": ["2007.07461", "2208.10458"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_2182"} +{"question": "Are there any research papers about the use of model editing techniques for hallucination mitigation during the training stage?", "answer": ["Elastic Weight Removal for Faithful and Abstractive Dialogue Generation"], "answer_arxiv_id": ["2303.17574"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_2183"} +{"question": "What works are about enhancing network efficiency through feature re-use?", "answer": ["GhostNet: More Features from Cheap Operations", "Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks"], "answer_arxiv_id": ["1911.11907", "2303.03667"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_2184"} +{"question": "Which state-of-the-art TTS models are known for producing speech quality nearly indistinguishable from human speech?", "answer": ["FastSpeech 2: Fast and High-Quality End-to-End Text to Speech", "NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level\n Quality"], "answer_arxiv_id": ["2006.04558", "2205.04421"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_2185"} +{"question": "Could you provide me some studies that examined what navigation agents learn about their environments?", "answer": ["What do navigation agents learn about their environment?"], "answer_arxiv_id": ["2206.08500"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2186"} +{"question": "What project conducted re-weighting to specifically target data heterogeneity?", "answer": ["Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization"], "answer_arxiv_id": ["2007.07481"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2187"} +{"question": "Which papers have utilized Transformers for tasks in computer vision?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2010.11929", "2103.14030"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_2188"} +{"question": "Which work develops the HAIM framework for fine-tuning on multimodal data?", "answer": ["Integrated multimodal artificial intelligence framework for healthcare applications"], "answer_arxiv_id": ["2202.12998v4"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_2189"} +{"question": "What studies are about using polarization properties of light and deep learning for specular reflection removal?", "answer": ["Separating Reflection and Transmission Images in the Wild", "Polarized Reflection Removal with Perfect Alignment in the Wild"], "answer_arxiv_id": ["1712.02099", "2003.12789"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_2190"} +{"question": "Which papers proposed methods of federated learning?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2191"} +{"question": "What are the works that discuss the problems of real-world datasets that have noisy/incomplete attributes for each image or sensitive information?", "answer": ["Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations", "Deep Learning Face Attributes in the Wild"], "answer_arxiv_id": ["1602.07332", "1411.7766v3"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_2192"} +{"question": "What research have adopted the concept of using coarse-grained features to adjust fine-grained predictions in image segmentation?", "answer": ["SPGNet: Semantic Prediction Guidance for Scene Parsing", "Bottom-up Instance Segmentation using Deep Higher-Order CRFs", "Higher Order Conditional Random Fields in Deep Neural Networks"], "answer_arxiv_id": ["1908.09798", "1609.02583v1", "1511.08119v4"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_2193"} +{"question": "Could you provide me some works related to SSL from the causality and data-generating processes perspective?", "answer": ["Contrastive Learning Inverts the Data Generating Process", "Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style", "What Makes for Good Views for Contrastive Learning?", "Representation Learning via Invariant Causal Mechanisms", "Chaos is a Ladder: A New Theoretical Understanding of Contrastive Learning via Augmentation Overlap"], "answer_arxiv_id": ["2102.08850", "2106.04619", "2005.10243", "2010.07922", "2203.13457"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_2194"} +{"question": "Which works proposed the most widely used open-vocabulary approaches such as BPE, WordPiece and UnigramLM?", "answer": ["Neural Machine Translation of Rare Words with Subword Units", "Byte Pair Encoding is Suboptimal for Language Model Pretraining"], "answer_arxiv_id": ["1508.07909", "2004.03720"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_2195"} +{"question": "Which papers proposed attack methods under the pointwise lpsubscript𝑙𝑝l_{p}-bounded image distortion threat model?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1412.6572", "1706.06083", "1608.04644"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_2196"} +{"question": "What are the studies that use ideas from RL to train GANs?", "answer": ["SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient", "IRGAN: A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models", "RL-GAN-Net: A Reinforcement Learning Agent Controlled GAN Network for Real-Time Point Cloud Shape Completion"], "answer_arxiv_id": ["1609.05473", "1705.10513v2", "1904.12304"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_2197"} +{"question": "Any research on learning a domain-specific and domain-agnostic visual prompt for adaptation?", "answer": ["Decorate the Newcomers: Visual Domain Prompt for Continual Test Time\n Adaptation"], "answer_arxiv_id": ["2212.04145"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2198"} +{"question": "What works propose dense retrieval methods which use dense embedding vectors to represent queries and documents?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "1907.11692"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_2199"} +{"question": "What work is recognized as the backbone model of their method to represent 3-dimensional equivariant features?", "answer": ["TorchMD-NET: Equivariant Transformers for Neural Network based Molecular Potentials"], "answer_arxiv_id": ["2202.02541"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_2200"} +{"question": "Which papers present methods that can capture actor interactions using an additional heavyweight module?", "answer": ["Asynchronous Interaction Aggregation for Action Detection", "Actor-Context-Actor Relation Network for Spatio-Temporal Action\n Localization"], "answer_arxiv_id": ["2004.07485", "2006.07976"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_2201"} +{"question": "What studies proposed algorithms to reduce the inside algorithm complexity from cubic to linear?", "answer": ["R2D2: Recursive Transformer based on Differentiable Tree for\n Interpretable Hierarchical Language Modeling", "Fast-R2D2: A Pretrained Recursive Neural Network based on Pruned CKY for\n Grammar Induction and Text Representation", "Augmenting Transformers with Recursively Composed Multi-grained\n Representations"], "answer_arxiv_id": ["2107.00967", "2203.00281", "2309.16319"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_2202"} +{"question": "Which papers propose transformation methods to project 2D image features to 3D BEV space for object detection?", "answer": ["Orthographic Feature Transform for Monocular 3D Object Detection", "Lift, Splat, Shoot: Encoding Images from Arbitrary Camera Rigs by Implicitly Unprojecting to 3D"], "answer_arxiv_id": ["1811.08188", "2008.05711"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_2203"} +{"question": "Could you provide me some works outlining solutions to handle instance-dependent label noise?", "answer": ["Part-dependent Label Noise: Towards Instance-dependent Label Noise", "Confidence Scores Make Instance-dependent Label-noise Learning Possible", "Clusterability as an Alternative to Anchor Points When Learning with Noisy Labels", "Instance-dependent Label-noise Learning under a Structural Causal Model"], "answer_arxiv_id": ["2006.07836", "2001.03772", "2102.05291", "2109.02986"], "source_meta": {"published_time": "20220204"}, "qid": "AutoScholarQuery_train_2204"} +{"question": "Which datasets are used to solve applications such as architectural style classification?", "answer": ["WikiChurches: A Fine-Grained Dataset of Architectural Styles with\n Real-World Challenges"], "answer_arxiv_id": ["2108.06959"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_2205"} +{"question": "What are some works that applied the analysis by synthesis approach on different problems like face recognition, pose estimation, 3D reconstruction, and semantic image editing?", "answer": ["3D-Aware Neural Body Fitting for Occlusion Robust 3D Human Pose\n Estimation", "Robust Category-Level 6D Pose Estimation with Coarse-to-Fine Rendering\n of Neural Features", "Humans as Light Bulbs: 3D Human Reconstruction from Thermal Reflection", "Shadows Shed Light on 3D Objects", "Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?", "Landscape Learning for Neural Network Inversion", "In-Domain GAN Inversion for Real Image Editing"], "answer_arxiv_id": ["2308.10123", "2209.05624", "2305.01652", "2206.08990", "1904.03189", "2206.09027", "2004.00049"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_2206"} +{"question": "Which works made significant advancements in model-based RL tasks with low-dimensional state spaces?", "answer": ["Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models", "Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model"], "answer_arxiv_id": ["1805.12114", "1911.08265"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_2207"} +{"question": "Can you provide references for works about the extension to multilingual versions of pretrained language models?", "answer": ["Unsupervised Cross-lingual Representation Learning at Scale", "Larger-Scale Transformers for Multilingual Masked Language Modeling", "DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with\n Gradient-Disentangled Embedding Sharing", "mT5: A massively multilingual pre-trained text-to-text transformer", "ByT5: Towards a token-free future with pre-trained byte-to-byte models", "Multilingual Denoising Pre-training for Neural Machine Translation"], "answer_arxiv_id": ["1911.02116", "2105.00572", "2111.09543", "2010.11934", "2105.13626", "2001.08210"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_2208"} +{"question": "Which papers employ inpainting-based approaches for seamless stitching in panorama generation?", "answer": ["TEXTure: Text-Guided Texturing of 3D Shapes", "Text2Tex: Text-driven Texture Synthesis via Diffusion Models", "Blended Diffusion for Text-driven Editing of Natural Images", "Blended Latent Diffusion"], "answer_arxiv_id": ["2302.01721", "2303.11396", "2111.14818", "2206.02779"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_2209"} +{"question": "Which work first formulates the online clustering of bandit (CB) problem?", "answer": ["Online Clustering of Bandits"], "answer_arxiv_id": ["1401.8257"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_2210"} +{"question": "Which papers discussed the potential of improving pretrained language model performance through cleverly chosen prompts?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2201.11903", "2205.10625", "2205.11916"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_2211"} +{"question": "What work achieves multi-view consistency by attending multi-view features with camera projection in diffusion models?", "answer": ["MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion"], "answer_arxiv_id": ["2307.01097"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_2212"} +{"question": "Which work studied the performance of GP-UCB in the frequentist setting?", "answer": ["On Kernelized Multi-armed Bandits"], "answer_arxiv_id": ["1704.00445"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_2213"} +{"question": "Which papers apply heteroscedastic models to uncertainty quantification in classification?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "A Simple Probabilistic Method for Deep Classification under Input-Dependent Label Noise", "Correlated Input-Dependent Label Noise in Large-Scale Image Classification", "Epistemic Neural Networks", "Deep Classifiers with Label Noise Modeling and Distance Awareness", "Plex: Towards Reliability Using Pretrained Large Model Extensions"], "answer_arxiv_id": ["1703.04977", "2003.06778v3", "2105.10305", "2107.08924", "2110.02609", "2207.07411"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2214"} +{"question": "Which recent initiatives exhibited limitations in citation annotation and evaluation methods?", "answer": ["WebCPM: Interactive Web Search for Chinese Long-form Question Answering", "WebGLM: Towards An Efficient Web-Enhanced Question Answering System with\n Human Preferences", "Enabling Large Language Models to Generate Text with Citations"], "answer_arxiv_id": ["2305.06849", "2306.07906", "2305.14627"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_2215"} +{"question": "Could you provide me some studies where the deep learning is applied for discriminative matching in Multi-view stereo (MVS) methods?", "answer": ["MVSNet: Depth Inference for Unstructured Multi-view Stereo", "Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching", "PatchmatchNet: Learned Multi-View Patchmatch Stereo", "Rethinking Depth Estimation for Multi-View Stereo: A Unified Representation"], "answer_arxiv_id": ["1804.02505", "1912.06378", "2012.01411", "2201.01501"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_2216"} +{"question": "Could you provide me some studies about membership inference which concern querying the model on similar or related examples to the target point?", "answer": ["Revisiting Membership Inference Under Realistic Assumptions", "Membership Leakage in Label-Only Exposures"], "answer_arxiv_id": ["2005.10881", "2007.15528"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_2217"} +{"question": "What works use hierarchical transformers to downsample activations in intermediate layers for increasing the context length of transformers?", "answer": ["Hierarchical Transformers Are More Efficient Language Models"], "answer_arxiv_id": ["2110.13711"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_2218"} +{"question": "Which papers relied on strong assumptions such as isotropy and multiplicative EMA?", "answer": ["Understanding Self-Supervised Learning Dynamics without Contrastive Pairs"], "answer_arxiv_id": ["2102.06810"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_2219"} +{"question": "What studies have applied few-shot learning in object detection tasks?", "answer": ["Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector", "Hallucination Improves Few-Shot Object Detection"], "answer_arxiv_id": ["1908.01998", "2105.01294"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_2220"} +{"question": "Any works about empirical Multi-Agent Reinforcement Learning (MARL) where information-sharing is discussed?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments", "Counterfactual Multi-Agent Policy Gradients", "Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1706.02275", "1705.08926", "2003.08839"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_2221"} +{"question": "What studies mentioned the problems of 'hallucination' in machine learning?", "answer": ["Mutual Information Alleviates Hallucinations in Abstractive\n Summarization", "The Factual Inconsistency Problem in Abstractive Text Summarization: A\n Survey", "A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on\n Reasoning, Hallucination, and Interactivity"], "answer_arxiv_id": ["2210.13210", "2104.14839", "2302.04023"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_2222"} +{"question": "Which research papers state that the model size should be scaled uniformly with the data size in language modeling?", "answer": ["Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2203.15556"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_2223"} +{"question": "What research attributes the success of transformer models to the scaling effect?", "answer": ["Scaling Laws for Neural Language Models", "Scaling Laws for Autoregressive Generative Modeling", "Scaling Laws for Generative Mixed-Modal Language Models"], "answer_arxiv_id": ["2001.08361", "2010.14701", "2301.03728"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_2224"} +{"question": "Which work proposed a window-level sparse Transformer?", "answer": ["SparseViT: Revisiting Activation Sparsity for Efficient High-Resolution\n Vision Transformer"], "answer_arxiv_id": ["2303.17605"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_2225"} +{"question": "Could you mention studies about coupling multiple samples and exploiting their dependencies for further variance reduction?", "answer": ["CARMS: Categorical-Antithetic-REINFORCE Multi-Sample Gradient Estimator", "ARMS: Antithetic-REINFORCE-Multi-Sample Gradient for Binary Variables", "DisARM: An Antithetic Gradient Estimator for Binary Latent Variables", "Coupled Gradient Estimators for Discrete Latent Variables", "ARM: Augment-REINFORCE-Merge Gradient for Stochastic Binary Networks", "ARSM: Augment-REINFORCE-Swap-Merge Estimator for Gradient Backpropagation Through Categorical Variables", "Probabilistic Best Subset Selection via Gradient-Based Optimization"], "answer_arxiv_id": ["2110.14002", "2105.14141", "2006.10680", "2106.08056", "1807.11143", "1905.01413", "2006.06448"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_2226"} +{"question": "Which work proposed to improve the corruption matrix using a clean set of data?", "answer": ["Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe\n Noise"], "answer_arxiv_id": ["1802.05300"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_2227"} +{"question": "Which papers utilized large-scale pretraining for grounding actions temporally through text-to-frame similarity matching?", "answer": ["End-to-End Learning of Visual Representations from Uncurated\n Instructional Videos", "Multimodal Clustering Networks for Self-supervised Learning from\n Unlabeled Videos"], "answer_arxiv_id": ["1912.06430", "2104.12671"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_2228"} +{"question": "Which studies have showed interest in 3D-language learning, including recognition, localization, question-answer and general conversation?", "answer": ["ShapeNet: An Information-Rich 3D Model Repository", "ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "3D ShapeNets: A Deep Representation for Volumetric Shapes", "Objaverse: A Universe of Annotated 3D Objects", "ScanRefer: 3D Object Localization in RGB-D Scans using Natural Language", "3DRP-Net: 3D Relative Position-aware Network for 3D Visual Grounding", "ScanQA: 3D Question Answering for Spatial Scene Understanding", "SQA3D: Situated Question Answering in 3D Scenes", "Chat-3D: Data-efficiently Tuning Large Language Model for Universal Dialogue of 3D Scenes", "3D-LLM: Injecting the 3D World into Large Language Models"], "answer_arxiv_id": ["1512.03012", "1702.04405v2", "1406.5670v3", "2212.08051", "1912.08830", "2307.13363", "2112.10482", "2210.07474", "2308.08769", "2307.12981"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_2229"} +{"question": "Which papers studied the choice of the collocation points in the discretization of PINN losses?", "answer": ["DeepXDE: A deep learning library for solving differential equations", "Efficient Training of Physics-Informed Neural Networks via Importance Sampling", "Investigating molecular transport in the human brain from MRI with physics-informed neural networks", "Respecting causality is all you need for training physics-informed neural networks", "A comprehensive study of non-adaptive and residual-based adaptive sampling for physics-informed neural networks"], "answer_arxiv_id": ["1907.04502", "2104.12325", "2205.02592", "2203.07404", "2207.10289"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_2230"} +{"question": "Any works about recently proposed deep learning-based schemes in image steganography?", "answer": ["SSGAN: Secure Steganography Based on Generative Adversarial Networks", "SteganoGAN: High Capacity Image Steganography with GANs", "HiDDeN: Hiding Data With Deep Networks"], "answer_arxiv_id": ["1707.01613v4", "1901.03892v2", "1807.09937"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2231"} +{"question": "What is the source paper that introduced the concept of realism and rarity scores for evaluating individual samples from generative models?", "answer": ["Improved Precision and Recall Metric for Assessing Generative Models", "Rarity Score : A New Metric to Evaluate the Uncommonness of Synthesized\n Images"], "answer_arxiv_id": ["1904.06991", "2206.08549"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_2232"} +{"question": "Any works about the properties of 1-Lipschitz neural nets, particularly, they are not subject to exploding nor vanishing gradients?", "answer": ["Preventing Gradient Attenuation in Lipschitz Constrained Convolutional Networks"], "answer_arxiv_id": ["1911.00937"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_2233"} +{"question": "What study proposed an implicit approach to fit a proxy SCM on the data using generative adversarial networks?", "answer": ["CausalGAN: Learning Causal Implicit Generative Models with Adversarial Training"], "answer_arxiv_id": ["1709.02023"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2234"} +{"question": "Could you list the research papers that explored audio to dance, text to motion, or audio to gestures?", "answer": ["EDGE: Editable Dance Generation From Music", "Human Motion Diffusion Model", "Listen, Denoise, Action! Audio-Driven Motion Synthesis with Diffusion\n Models", "GestureDiffuCLIP: Gesture Diffusion Model with CLIP Latents", "Talking Head Generation with Probabilistic Audio-to-Visual Diffusion\n Priors", "LivelySpeaker: Towards Semantic-Aware Co-Speech Gesture Generation"], "answer_arxiv_id": ["2211.10658", "2209.14916", "2211.09707", "2303.14613", "2212.04248", "2309.09294"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_2235"} +{"question": "What study proposed Feature propagation (FP) to reconstruct missing features by diffusing known features?", "answer": ["On the Unreasonable Effectiveness of Feature Propagation in Learning on Graphs with Missing Node Features"], "answer_arxiv_id": ["2111.12128"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2236"} +{"question": "Which studies have integrated unsupervised pretraining and data selection?", "answer": ["PT4AL: Using Self-Supervised Pretext Tasks for Active Learning", "Low-Budget Active Learning via Wasserstein Distance: An Integer Programming Approach", "Unsupervised Selective Labeling for More Effective Semi-Supervised Learning", "Active Finetuning: Exploiting Annotation Budget in the Pretraining-Finetuning Paradigm"], "answer_arxiv_id": ["2201.07459", "2106.02968", "2110.03006", "2303.14382"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_2237"} +{"question": "Could you give me examples of works that propose benchmarks for XAI using synthetic data?", "answer": ["OpenXAI: Towards a Transparent Evaluation of Model Explanations"], "answer_arxiv_id": ["2206.11104v5"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_2238"} +{"question": "Which papers discussed state-of-the-art unsupervised VAE baselines for disentangling learning representations?", "answer": ["Disentangling by Factorising", "Isolating Sources of Disentanglement in Variational Autoencoders", "Variational Inference of Disentangled Latent Concepts from Unlabeled Observations"], "answer_arxiv_id": ["1802.05983", "1802.04942", "1711.00848"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_2239"} +{"question": "What works utilized teachers like CLIP to distill patch representations under the MIM pipeline?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizers", "A Unified View of Masked Image Modeling", "MVP: Multimodality-guided Visual Pre-training", "EVA: Exploring the Limits of Masked Visual Representation Learning at Scale"], "answer_arxiv_id": ["2103.00020", "2208.06366", "2210.10615", "2203.05175", "2211.07636"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_2240"} +{"question": "What are the researches on video synthesis extension with DMs?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models", "Video Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation", "Structure and Content-Guided Video Synthesis with Diffusion Models"], "answer_arxiv_id": ["2210.02303", "2204.03458", "2209.14792", "2303.08320", "2302.03011"], "source_meta": {"published_time": "20240626"}, "qid": "AutoScholarQuery_train_2241"} +{"question": "Which papers considered a detailed causal modeling on the college admission system and analyzed previously proposed causal fairness notions?", "answer": ["Counterfactual Risk Assessments, Evaluation, and Fairness", "Principal Fairness for Human and Algorithmic Decision-Making"], "answer_arxiv_id": ["1909.00066", "2005.10400"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_2242"} +{"question": "Which papers employed rationales in explaining the predictions of neural models?", "answer": ["Measuring Association Between Labels and Free-Text Rationales"], "answer_arxiv_id": ["2010.12762"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_train_2243"} +{"question": "Which studies have improved model performance of text-to-image models by various network architectures and training pipelines?", "answer": ["Scaling up GANs for Text-to-Image Synthesis", "Generating High Fidelity Images with Subscale Pixel Networks and\n Multidimensional Upscaling", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Elucidating the Design Space of Diffusion-Based Generative Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2303.05511", "1812.01608", "2206.10789", "2206.00364v2", "1503.03585", "1907.05600", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_2244"} +{"question": "In what research was point-based ray tracing replaced with cone tracing to tackle aliasing?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields"], "answer_arxiv_id": ["2103.13415"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_2245"} +{"question": "Which papers are focused on using graphical models for post-segmentation refinement in high-quality segmentation?", "answer": ["Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials", "Semantic Segmentation Refinement by Monte Carlo Region Growing of High Confidence Detections"], "answer_arxiv_id": ["1210.5644", "1802.07789"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_2246"} +{"question": "Which research focus on domain-invariant learning methods in Domain Generalization?", "answer": ["Deep CORAL: Correlation Alignment for Deep Domain Adaptation", "Generalizing Across Domains via Cross-Gradient Training", "Causality Inspired Representation Learning for Domain Generalization", "Modality-Agnostic Debiasing for Single Domain Generalization", "Decompose, Adjust, Compose: Effective Normalization by Playing with\n Frequency for Domain Generalization", "DomainDrop: Suppressing Domain-Sensitive Channels for Domain\n Generalization"], "answer_arxiv_id": ["1607.01719", "1804.10745", "2203.14237", "2303.07123", "2303.02328", "2308.10285"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_2247"} +{"question": "What research exists on reinforcement learning applications of approximate symmetries?", "answer": ["Residual Pathway Priors for Soft Equivariance Constraints"], "answer_arxiv_id": ["2112.01388"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_2248"} +{"question": "Could you provide me some studies addressing visual grounding of 3D scenes?", "answer": ["Scan2Cap: Context-aware Dense Captioning in RGB-D Scans", "Spatiality-guided Transformer for 3D Dense Captioning on Point Clouds", "X-Trans2Cap: Cross-Modal Knowledge Transfer using Transformer for 3D Dense Captioning", "ScanEnts3D: Exploiting Phrase-to-3D-Object Correspondences for Improved Visio-Linguistic Models in 3D Scenes", "ScanQA: 3D Question Answering for Spatial Scene Understanding", "SQA3D: Situated Question Answering in 3D Scenes", "Towards Explainable 3D Grounded Visual Question Answering: A New Benchmark and Strong Baseline"], "answer_arxiv_id": ["2012.02206", "2204.10688", "2203.00843v3", "2212.06250", "2112.10482", "2210.07474", "2209.12028"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_2249"} +{"question": "Could you provide the references that consider data sharing between market competitors?", "answer": ["MarS-FL: Enabling Competitors to Collaborate in Federated Learning"], "answer_arxiv_id": ["2110.13464v2"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_2250"} +{"question": "Could you provide me some works that employ uncertainty in edge detection?", "answer": ["The Treasure Beneath Multiple Annotations: An Uncertainty-aware Edge\n Detector"], "answer_arxiv_id": ["2303.11828"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_2251"} +{"question": "In which studies are analysed the LLM's failure in privacy by leaking private information?", "answer": ["Extracting Training Data from Large Language Models"], "answer_arxiv_id": ["2012.07805"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_2252"} +{"question": "Which empirical works discuss the setting of online Reinforcement Learning with access to logged data?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations", "Overcoming Exploration in Reinforcement Learning with Demonstrations", "Deep Q-learning from Demonstrations", "Efficient Online Reinforcement Learning with Offline Data", "Cal-QL: Calibrated Offline RL Pre-Training for Efficient Online Fine-Tuning", "Adaptive Policy Learning for Offline-to-Online Reinforcement Learning"], "answer_arxiv_id": ["1709.10087", "1709.10089", "1704.03732", "2302.02948", "2303.05479", "2303.07693"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_2253"} +{"question": "Could you provide me some works that reduce MO-HPO into a single-objective one through scalarization techniques?", "answer": ["Random Hypervolume Scalarizations for Provable Multi-Objective Black Box Optimization", "A Flexible Framework for Multi-Objective Bayesian Optimization using Random Scalarizations"], "answer_arxiv_id": ["2006.04655", "1805.12168"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_2254"} +{"question": "Could you provide studies about developing generative models of neural networks?", "answer": ["Hyper-Representations as Generative Models: Sampling Unseen Neural Network Weights", "Learning to Learn with Generative Models of Neural Network Checkpoints"], "answer_arxiv_id": ["2209.14733", "2209.12892"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_2255"} +{"question": "What works exist about one-step RL methods that apply a single step of policy improvement to the behavioral policy?", "answer": ["Offline RL Without Off-Policy Evaluation", "Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning"], "answer_arxiv_id": ["2106.08909", "1910.00177"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_2256"} +{"question": "Which studies have analyzed the phenomena of shaping user preferences and stereotyping in recommender systems?", "answer": ["The Stereotyping Problem in Collaboratively Filtered Recommender Systems"], "answer_arxiv_id": ["2106.12622"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_2257"} +{"question": "Which papers are identified as the foundation for the concept of object-centric learning, meaning that reasoning over a small number of objects is more efficient than over a feature map?", "answer": ["Systematic Evaluation of Causal Discovery in Visual Model Based Reinforcement Learning", "Better Set Representations For Relational Reasoning"], "answer_arxiv_id": ["2107.00848", "2003.04448"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2258"} +{"question": "What papers introduce restricting the parameter-updates to a low-dimensional subspace?", "answer": ["Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tuning", "Compacter: Efficient Low-Rank Hypercomplex Adapter Layers"], "answer_arxiv_id": ["2012.13255", "2106.04647"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2259"} +{"question": "What is the work that proposes leveraging the technique of prompting for downstream tasks of large language models (LLMs)?", "answer": ["Foundation Models for Decision Making: Problems, Methods, and\n Opportunities"], "answer_arxiv_id": ["2303.04129"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_2260"} +{"question": "Which studies have used the Frank-Wolfe method and its variants for classification?", "answer": ["Block-Coordinate Frank-Wolfe Optimization for Structural SVMs", "Frank-Wolfe algorithm for learning SVM-type multi-category classifiers"], "answer_arxiv_id": ["1207.4747", "2008.08894"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2261"} +{"question": "What research works explored fairness problem in machine translation?", "answer": ["Evaluating Gender Bias in Machine Translation"], "answer_arxiv_id": ["1906.00591"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_2262"} +{"question": "What work constructs an INR generator by combining a Bayesian NN with Gaussian weight priors?", "answer": ["Function-space Inference with Sparse Implicit Processes"], "answer_arxiv_id": ["2110.07618"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_2263"} +{"question": "What works introduced the concept of the commonly used concentrability coefficient for all policy in a policy class?", "answer": ["Information-Theoretic Considerations in Batch Reinforcement Learning"], "answer_arxiv_id": ["1905.00360"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_2264"} +{"question": "What works used analyzing pixel-to-pixel correspondences for avoiding the loss of spatial structure in Few-Shot Segmentation?", "answer": ["Hierarchical Dense Correlation Distillation for Few-Shot Segmentation", "Dense Cross-Query-and-Support Attention Weighted Mask Aggregation for\n Few-Shot Segmentation", "Simpler is Better: Few-shot Semantic Segmentation with Classifier Weight\n Transformer", "Prior Guided Feature Enrichment Network for Few-Shot Segmentation"], "answer_arxiv_id": ["2303.14652", "2207.08549", "2108.03032", "2008.01449"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2265"} +{"question": "Could you provide me some works that continued to scale the kernel size?", "answer": ["Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs", "More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using Sparsity"], "answer_arxiv_id": ["2203.06717", "2207.03620"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2266"} +{"question": "Can you name some works which used image augmentation techniques to achieve generalization in image-based RL tasks?", "answer": ["Generalization in Reinforcement Learning by Soft Data Augmentation", "Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation", "Look where you look! Saliency-guided Q-networks for generalization in visual Reinforcement Learning", "Reinforcement Learning with Augmented Data", "Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels"], "answer_arxiv_id": ["2011.13389", "2107.00644v2", "2209.09203", "2004.14990", "2107.00644v2", "2004.13649"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_2267"} +{"question": "Which research papers focus on the use of irreducible representations of the SO(3) group in equivariant methods?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "Cormorant: Covariant Molecular Neural Networks", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "Geometric and Physical Quantities improve E(3) Equivariant Message Passing", "MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields", "So3krates: Equivariant attention for interactions on arbitrary length-scales in molecular systems"], "answer_arxiv_id": ["1802.08219", "1906.04015", "2006.10503", "2110.02905", "2206.07697", "2205.14276"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_2268"} +{"question": "Which papers proposed benchmarks that assess models' ability to generate programs given surrounding program context beyond the target program?", "answer": ["Mapping Language to Code in Programmatic Context", "JuICe: A Large Scale Distantly Supervised Dataset for Open Domain Context-based Code Generation"], "answer_arxiv_id": ["1808.09588", "1910.02216"], "source_meta": {"published_time": "20220325"}, "qid": "AutoScholarQuery_train_2269"} +{"question": "What are some works that generate music from text?", "answer": ["Jukebox: A Generative Model for Music"], "answer_arxiv_id": ["2005.00341"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_2270"} +{"question": "What studies have proposed methods based on worst-group optimization when spurious attributes are known?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_2271"} +{"question": "What works talk about the adoption of BERT and Faster-RCNN for extracting visual and text features in VL pre-training?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks"], "answer_arxiv_id": ["1810.04805", "1506.01497"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_2272"} +{"question": "Could you provide me with some studies about using projectors for adversarial attacks in physical world?", "answer": ["Optical Adversarial Attack", "Adversarial Color Projection: A Projector-based Physical Attack to DNNs"], "answer_arxiv_id": ["2108.06247", "2209.09652v2"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_2273"} +{"question": "Which studies enhance SR performance by incorporating channel attention mechanisms?", "answer": ["Single Image Super-Resolution via a Holistic Attention Network", "Image Super-Resolution Using Very Deep Residual Channel Attention\n Networks"], "answer_arxiv_id": ["2008.08767", "1807.02758"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_2274"} +{"question": "Could you provide me some works about learning high-order model derivatives?", "answer": ["GENIE: Higher-Order Denoising Diffusion Solvers"], "answer_arxiv_id": ["2210.05475"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_2275"} +{"question": "Where can I find research demonstrating the effectiveness of segmenting objects in 3D with minimal user input?", "answer": ["Neural Volumetric Object Selection"], "answer_arxiv_id": ["2205.14929"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_2276"} +{"question": "Which works were inspired by the success of 2D generation and built the autoencoder architecture for NVS?", "answer": ["Multi-view 3D Models from Single Images with a Convolutional Network", "Transformation-Grounded Image Generation Network for Novel 3D View\n Synthesis", "SynSin: End-to-end View Synthesis from a Single Image", "Explicit Correspondence Matching for Generalizable Neural Radiance\n Fields"], "answer_arxiv_id": ["1511.06702", "1703.02921", "1912.08804", "2304.12294"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2277"} +{"question": "Could you provide me some works about diffusion models for generating images?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2112.10752"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_2278"} +{"question": "Which work introduces Bayesian loss with point-wise supervision as an alternative to L2 loss?", "answer": ["Bayesian Loss for Crowd Count Estimation with Point Supervision"], "answer_arxiv_id": ["1908.03684v1"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_2279"} +{"question": "Which works discuss parameter-free methods in the online learning literature?", "answer": ["Coin Betting and Parameter-Free Online Learning", "No-Regret Algorithms for Unconstrained Online Convex Optimization", "Parameter-free Stochastic Optimization of Variationally Coherent Functions", "Training Deep Networks without Learning Rates Through Coin Betting", "Lipschitz Adaptivity with Multiple Learning Rates in Online Learning", "Lipschitz and Comparator-Norm Adaptivity in Online Learning", "Normalized Gradients for All"], "answer_arxiv_id": ["1602.04128", "1211.2260", "2102.00236", "1705.07795", "1902.10797v2", "2002.12242", "2308.05621"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_2280"} +{"question": "What paper proposed using a cosine similarity classifier instead of a dot-product classifier for handling biased classifier weights issue?", "answer": ["Dynamic Few-Shot Visual Learning without Forgetting"], "answer_arxiv_id": ["1804.09458"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_2281"} +{"question": "What works use Central Limit Theorem to approximate the distribution of Y=∑i=1kYi as Gaussian distribution?", "answer": ["Deep Learning with Gaussian Differential Privacy", "Analytical Composition of Differential Privacy via the Edgeworth Accountant"], "answer_arxiv_id": ["1911.11607", "2206.04236"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_2282"} +{"question": "Could you recommend studies that explore proficiency of LLMs such as ChatGPT in law exams and technical aspects of GPT-4?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_2283"} +{"question": "What papers focus on second-order methods, particularly Hessian-based methods, for post-training sparsification?", "answer": ["Learning to Prune Deep Neural Networks via Layer-wise Optimal Brain Surgeon"], "answer_arxiv_id": ["1705.07565"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_2284"} +{"question": "What studies have applied the IB principle in machine learning tasks such as unsupervised learning, classification and clustering?", "answer": ["Deep Learning and the Information Bottleneck Principle", "Deep Variational information bottleneck", "Information Dropout: Learning Optimal Representations Through Noisy Computation", "Multi-view Semantic Consistency based Information Bottleneck for Clustering"], "answer_arxiv_id": ["1503.02406", "1612.00410", "1611.01353v3", "2303.00002"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_2285"} +{"question": "What work improved the generation speed of the diffusion model by proposing the Denoising Diffusion Implicit Models (DDIM)?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_2286"} +{"question": "Could you tell me some studies that offer runtime improvements under sparsity assumption?", "answer": ["Quantum algorithms for zero-sum games"], "answer_arxiv_id": ["1904.03180"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_2287"} +{"question": "Which research papers are there on designing V-learning algorithms to find CCE in multi-agent general-sum games?", "answer": ["Near-Optimal Reinforcement Learning with Self-Play", "V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL"], "answer_arxiv_id": ["2006.12007", "2110.14555"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_2288"} +{"question": "What paper uses an autoregressive conditional Variational Autoencoder (VAE) to construct a latent space for human movements?", "answer": ["Character Controllers Using Motion VAEs"], "answer_arxiv_id": ["2103.14274"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_2289"} +{"question": "Could you list the voxel-based approaches used for LiDAR semantic segmentation?", "answer": ["3D ShapeNets: A Deep Representation for Volumetric Shapes", "Volumetric and Multi-View CNNs for Object Classification on 3D Data", "3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1406.5670", "1604.03265", "1711.10275", "1904.08755"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_2290"} +{"question": "What are some studies adopting variational autoencoders in the advancement of 3D generation?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_2291"} +{"question": "Could you point out studies that discuss the challenges in adopting the full covariance matrix due to the difficulties in Monte-Carlo based KL approximation?", "answer": ["A Rotated Hyperbolic Wrapped Normal Distribution for Hierarchical Representation Learning"], "answer_arxiv_id": ["2205.13371"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2292"} +{"question": "Can you mention some works that discuss techniques Used for providing a certificate for l2subscript perturbations in Randomized Smoothing?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing"], "answer_arxiv_id": ["1902.02918"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_2293"} +{"question": "What are the research papers exploring the integration of using tools or models in large language models?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "ViperGPT: Visual Inference via Python Execution for Reasoning", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "TaskMatrix.AI: Completing Tasks by Connecting Foundation Models with Millions of APIs", "Tool Learning with Foundation Models"], "answer_arxiv_id": ["2302.04761", "2303.08128", "2303.04671", "2303.16434", "2304.08354"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_2294"} +{"question": "Which studies adopted reinforcement learning or reward function to fine-tune diffusion models?", "answer": ["Training Diffusion Models with Reinforcement Learning", "DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion\n Models", "ImageReward: Learning and Evaluating Human Preferences for Text-to-Image\n Generation"], "answer_arxiv_id": ["2305.13301", "2305.16381", "2304.05977"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_2295"} +{"question": "Could you provide studies that used adversarial test sets?", "answer": ["Intriguing properties of neural networks", "Adversarial Diversity and Hard Positive Generation", "Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning", "Adversarial examples in the physical world", "Improving Transferability of Adversarial Examples with Input Diversity", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1312.6199", "1605.01775", "1704.03976", "1607.02533", "1803.06978", "1706.06083"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_2296"} +{"question": "Could you give me examples of research proposing the development of scale-equivariant CNN?", "answer": ["Locally Scale-Invariant Convolutional Neural Networks", "Scale Steerable Filters for Locally Scale-Invariant Convolutional Neural Networks", "B-Spline CNNs on Lie Groups", "Deep Scale-spaces: Equivariance Over Scale", "DISCO: accurate Discrete Scale Convolutions", "Scale-Equivariant Steerable Networks", "Scaling-Translation-Equivariant Networks with Decomposed Convolutional Filters"], "answer_arxiv_id": ["1412.5104", "1906.03861v1", "1909.12057", "1905.11697", "2106.02733", "1910.11093", "1909.11193"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_2297"} +{"question": "Which works considered the bounded bilinear rank assumption in the study of RL with general function approximation?", "answer": ["Bilinear Classes: A Structural Framework for Provable Generalization in RL"], "answer_arxiv_id": ["2103.10897"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_2298"} +{"question": "Could you provide me the resources that extended the Neural Radiance Fields (NeRF) to enable capturing a dynamic moving humans?", "answer": ["Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "Vid2Actor: Free-viewpoint Animatable Person Synthesis from Video in the\n Wild", "Neural Actor: Neural Free-view Synthesis of Human Actors with Pose\n Control", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "NeuMan: Neural Human Radiance Field from a Single Video", "Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via\n Self-supervised Scene Decomposition", "Capturing and Animation of Body and Clothing from Monocular Video", "KeypointNeRF: Generalizing Image-based Volumetric Avatars using Relative\n Spatial Encoding of Keypoints"], "answer_arxiv_id": ["2012.15838", "2012.12884", "2106.02019", "2201.04127", "2203.12575", "2302.11566", "2210.01868", "2205.04992"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_2299"} +{"question": "Who explored more generalized Policy Distillation architectures?", "answer": ["Generalized Decision Transformer for Offline Hindsight Information Matching"], "answer_arxiv_id": ["2111.10364"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_2300"} +{"question": "Which works discuss the concept of positive transfer, in which learning in one modality improves performance in another?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_2301"} +{"question": "What papers developed pruning metrics that score data points for data selection?", "answer": ["Deep Learning on a Data Diet: Finding Important Examples Early in Training"], "answer_arxiv_id": ["2107.07075"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_train_2302"} +{"question": "Were there any attempts to find a tractable formulation for the Wasserstein gradient with spatial discretization?", "answer": ["Discretization of functionals involving the Monge-Ampère operator", "Convergence of Entropic Schemes for Optimal Transport and Gradient Flows"], "answer_arxiv_id": ["1408.4536", "1512.02783v2"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_2303"} +{"question": "Which research proposed using exemplars in CIL for iCaRL knowledge distillation regularization, finetuning, or bias correction?", "answer": ["iCaRL: Incremental Classifier and Representation Learning", "Gradient Episodic Memory for Continual Learning", "Large Scale Incremental Learning", "End-to-End Incremental Learning"], "answer_arxiv_id": ["1611.07725", "1706.08840", "1905.13260", "1807.09536"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_2304"} +{"question": "Which papers touched on the subject of text-to-image diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2305"} +{"question": "Which research handles the problem of off-policy evaluation under tabular MDP?", "answer": ["Asymptotically Efficient Off-Policy Evaluation for Tabular Reinforcement Learning"], "answer_arxiv_id": ["2001.10742"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_2306"} +{"question": "What works have been done in the area of rematerialization in machine learning, especially in training large DNN on GPU?", "answer": ["A Graph Theoretic Framework of Recomputation Algorithms for Memory-Efficient Backpropagation", "Training Deep Nets with Sublinear Memory Cost", "Dynamic Tensor Rematerialization", "Sequential Aggregation and Rematerialization: Distributed Full-batch Training of Graph Neural Networks on Large Graphs", "GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism"], "answer_arxiv_id": ["1905.11722", "1604.06174", "2006.09616", "2111.06483v3", "1811.06965v5"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_2307"} +{"question": "Which studies have contributed to image-text contrastive learning for VLM pretraining?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual\n Concepts", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "FLAVA: A Foundational Language And Vision Alignment Model", "Sigmoid Loss for Language Image Pre-Training", "EVA-CLIP: Improved Training Techniques for CLIP at Scale"], "answer_arxiv_id": ["2102.05918", "2111.08276", "2205.01917", "2112.04482", "2303.15343v4", "2303.15389"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2308"} +{"question": "What work utilized SAM's segmentation results in weakly supervised segmentation tasks?", "answer": ["Segment Anything Model (SAM) Enhanced Pseudo Labels for Weakly\n Supervised Semantic Segmentation", "Token Contrast for Weakly-Supervised Semantic Segmentation", "SSA: Semantic Structure Aware Inference for Weakly Pixel-Wise Dense\n Predictions without Cost"], "answer_arxiv_id": ["2305.05803", "2303.01267", "2111.03392"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2309"} +{"question": "Can you provide some works related to layout completion and refinement?", "answer": ["LayoutTransformer: Layout Generation and Completion with Self-attention", "Auto Completion of User Interface Layout Design Using Transformer-Based Tree Decoders"], "answer_arxiv_id": ["2006.14615", "2001.05308"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_2310"} +{"question": "What studies used deep learning models for graph generation?", "answer": ["A Systematic Survey on Deep Generative Models for Graph Generation"], "answer_arxiv_id": ["2007.06686"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_2311"} +{"question": "What recent researches have revealed risks associated with fine-tuning of LLMs?", "answer": ["Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models", "Removing RLHF Protections in GPT-4 via Fine-Tuning", "Exploiting Novel GPT-4 APIs"], "answer_arxiv_id": ["2310.02949", "2311.05553", "2312.14302"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_2312"} +{"question": "Which works made progress in vision language models for remote sensing tasks, such as image captioning, zero-shot classification and visual question answering?", "answer": ["From Easy to Hard: Learning Language-guided Curriculum for Visual\n Question Answering on Remote Sensing Data"], "answer_arxiv_id": ["2205.03147"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_2313"} +{"question": "Which study handles distribution shifts by retraining a forecasting model with training data from a non-uniform adaptive sampling?", "answer": ["Adaptive Sampling for Probabilistic Forecasting under Distribution Shift"], "answer_arxiv_id": ["2302.11870"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_2314"} +{"question": "Which works treat text as treatment in causal inference with text variables?", "answer": ["Causal Effects of Linguistic Properties", "Challenges of Using Text Classifiers for Causal Inference", "How to Make Causal Inferences Using Texts", "When do Words Matter? Understanding the Impact of Lexical Choice on Audience Perception using Individual Treatment Effect Estimation", "The effect of wording on message propagation: Topic- and author-controlled natural experiments on Twitter"], "answer_arxiv_id": ["2010.12919", "1810.00956", "1802.02163", "1811.04890", "1405.1438"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2315"} +{"question": "Which papers are about federated learning methods that utilize public data for model training?", "answer": ["Ensemble Distillation for Robust Model Fusion in Federated Learning", "FedMD: Heterogenous Federated Learning via Model Distillation", "FedGH: Heterogeneous Federated Learning with Generalized Global Header"], "answer_arxiv_id": ["2006.07242", "1910.03581", "2303.13137"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_2316"} +{"question": "Which works utilize generative models for full body motion reconstruction?", "answer": ["FLAG: Flow-based 3D Avatar Generation from Sparse Observations", "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking\n Inputs with Diffusion Model", "BoDiffusion: Diffusing Sparse Observations for Full-Body Human Motion\n Synthesis"], "answer_arxiv_id": ["2203.05789", "2304.08577", "2304.11118"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_2317"} +{"question": "Can you name any studies that enhanced spatio-temporal control in generating videos by integrating guidance from depth maps or target motion?", "answer": ["Structure and Content-Guided Video Synthesis with Diffusion Models", "Motion-Conditioned Diffusion Model for Controllable Video Synthesis", "LaMD: Latent Motion Diffusion for Video Generation", "VideoComposer: Compositional Video Synthesis with Motion Controllability"], "answer_arxiv_id": ["2302.03011", "2304.14404", "2304.11603", "2306.02018"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2318"} +{"question": "Can you provide me with some papers about neurosymbolic learning?", "answer": ["The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences from Natural Supervision", "Closed Loop Neural-Symbolic Learning via Integrating Neural Perception, Grammar Parsing, and Symbolic Reasoning", "Modular Framework for Visuomotor Language Grounding", "Control Regularization for Reduced Variance Reinforcement Learning", "Imitation-Projected Programmatic Reinforcement Learning", "Learning Differentiable Programs with Admissible Neural Heuristics", "Discovering Symbolic Models from Deep Learning with Inductive Biases", "PIGLeT: Language Grounding Through Neuro-Symbolic Interaction in a 3D World"], "answer_arxiv_id": ["1904.12584", "2006.06649v2", "2109.02161", "1905.05380", "1907.05431", "2007.12101", "2006.11287", "2106.00188"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_2319"} +{"question": "Which works propose optimization techniques for test-time adaptation?", "answer": ["Tent: Fully Test-Time Adaptation by Entropy Minimization", "Improving Test-Time Adaptation via Shift-agnostic Weight Regularization and Nearest Source Prototypes", "Continual Test-Time Domain Adaptation", "Contrastive Test-Time Adaptation"], "answer_arxiv_id": ["2006.10726", "2207.11707", "2203.13591", "2204.10377"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_2320"} +{"question": "Could you provide me with research papers talking about image editing models that use an input mask as additional input?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images", "Imagen Editor and EditBench: Advancing and Evaluating Text-Guided Image\n Inpainting", "SmartBrush: Text and Shape Guided Object Inpainting with Diffusion Model"], "answer_arxiv_id": ["2111.14818", "2212.06909", "2212.05034"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_2321"} +{"question": "What paper developed ViT-VQGAN by extending VQGAN to the Transformer architecture?", "answer": ["Vector-quantized Image Modeling with Improved VQGAN"], "answer_arxiv_id": ["2110.04627"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_2322"} +{"question": "Could you give me examples of research that has contributed to the theoretical understanding of multi-modal learning?", "answer": ["A Survey on Multi-view Learning", "Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning", "What Makes Multi-modal Learning Better than Single (Provably)", "Modality Competition: What Makes Joint Training of Multi-modal Network Fail in Deep Learning? (Provably)", "Does my multimodal model learn cross-modal interactions? It’s harder to tell than you might think!"], "answer_arxiv_id": ["1304.5634", "2012.09816", "2106.04538", "2203.12221", "2010.06572"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_2323"} +{"question": "What works suggested the method of instruction tuning for making better language models?", "answer": ["MetaICL: Learning to Learn In Context", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2110.15943", "2203.02155"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_2324"} +{"question": "Could you give me examples of studies that rely on dense depth estimates in 3D detection?", "answer": ["Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object\n Detection for Autonomous Driving", "Rethinking Pseudo-LiDAR Representation"], "answer_arxiv_id": ["1812.07179", "2008.04582"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_2325"} +{"question": "Could you provide me with some works that applied quantization to GLMs to minimize inference costs?", "answer": ["AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration", "LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "Outlier Suppression+: Accurate quantization of large language models by equivalent and optimal shifting and scaling", "ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers"], "answer_arxiv_id": ["2306.00978", "2208.07339", "2304.09145", "2206.01861"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_2326"} +{"question": "Which datasets were collected for a small number of categories when learning 3D models?", "answer": ["BANMo: Building Animatable 3D Neural Models from Many Casual Videos", "DOVE: Learning Deformable 3D Objects by Watching Videos", "Common Pets in 3D: Dynamic New-View Synthesis of Real-Life Deformable\n Categories"], "answer_arxiv_id": ["2112.12761", "2107.10844", "2211.03889"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_2327"} +{"question": "Could you provide me with the research that proposed using an MLP as the projector and finetuned on academic instruction datasets to achieve state-of-the-art performance on various benchmarks?", "answer": ["Improved Baselines with Visual Instruction Tuning"], "answer_arxiv_id": ["2310.03744"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_2328"} +{"question": "Could you tell me which paper defined a general framework for learning game-theoretic solution concepts from samples?", "answer": ["A Learning Framework for Distribution-Based Game-Theoretic Solution Concepts"], "answer_arxiv_id": ["1903.08322"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_2329"} +{"question": "Can you provide some studies that showed the limited calibration and capability in Out-Of-Distribution (OOD) detection of the Trained Mixture Classifier (TMC)?", "answer": ["Uncertainty Estimation for Multi-view Data: The Power of Seeing the Whole Picture"], "answer_arxiv_id": ["2210.02676"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_2330"} +{"question": "Which paper established Neural Radiance Fields (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_2331"} +{"question": "Which paper introduced the concept of Attention Sinks?", "answer": ["Efficient Streaming Language Models with Attention Sinks"], "answer_arxiv_id": ["2309.17453"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_2332"} +{"question": "Are there any works which distilled CLIP image features into a NeRF representation for 3D segmentation?", "answer": ["TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2203.09517"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_2333"} +{"question": "What paper provides an in-depth survey on the expressive power of GNNs?", "answer": ["Weisfeiler and Leman go Machine Learning: The Story so far"], "answer_arxiv_id": ["2112.09992"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_2334"} +{"question": "Which works proposed modeling skin reflectance using a Microfacet BRDF with diffuse and specular components?", "answer": ["S2F2: Self-Supervised High Fidelity Face Reconstruction from Monocular\n Image", "Towards High Fidelity Monocular Face Reconstruction with Rich\n Reflectance using Self-supervised Learning and Ray Tracing", "Practical Face Reconstruction via Differentiable Ray Tracing"], "answer_arxiv_id": ["2203.07732", "2103.15432", "2101.05356"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_2335"} +{"question": "What works provide privacy filters for Gaussian DP?", "answer": ["Individual Privacy Accounting with Gaussian Differential Privacy"], "answer_arxiv_id": ["2209.15596"], "source_meta": {"published_time": "20220310"}, "qid": "AutoScholarQuery_train_2336"} +{"question": "What are some papers that discuss metrics such as the Inception Score (IS) and Frechet Inception Distance (FID) used for measuring image quality in automated text-to-image (T2I) evaluation?", "answer": ["Improved Techniques for Training GANs", "GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium"], "answer_arxiv_id": ["1606.03498", "1706.08500"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_2337"} +{"question": "Which study employed text rewriting techniques with GPT3.5 and Bard for integrating text augmentations?", "answer": ["Improving CLIP Training with Language Rewrites"], "answer_arxiv_id": ["2305.20088"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_2338"} +{"question": "Which works studied non-stationarity introduced by a primal-dual formulation of distinct problem classes into min-max games?", "answer": ["Reward is Enough for Convex MDPs", "ReLOAD: Reinforcement Learning with Optimistic Ascent-Descent for Last-Iterate Convergence in Constrained MDPs"], "answer_arxiv_id": ["2106.00661", "2302.01275"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_2339"} +{"question": "What are some works that discuss one-shot pruning methods in deep learning?", "answer": ["WoodFisher: Efficient Second-Order Approximation for Neural Network Compression", "The Combinatorial Brain Surgeon: Pruning Weights That Cancel One Another in Neural Networks"], "answer_arxiv_id": ["2004.14340", "2203.04466v3"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_2340"} +{"question": "Could you provide me some deep-learning based approaches in Granger causality for time series discovery?", "answer": ["Amortized Causal Discovery: Learning to Infer Causal Graphs from Time-Series Data", "Neural Additive Vector Autoregression Models for Causal Discovery in Time Series", "seq2graph: Discovering Dynamic Dependencies from Multivariate Time Series with Multi-level Attention"], "answer_arxiv_id": ["2006.10833", "2010.09429", "1812.04448"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_2341"} +{"question": "Could you provide me some studies where ML methods have been applied to improve LNS?", "answer": ["Learning to Perform Local Rewriting for Combinatorial Optimization", "Neural Large Neighborhood Search for the Capacitated Vehicle Routing Problem", "Learning to Delegate for Large-scale Vehicle Routing"], "answer_arxiv_id": ["1810.00337", "1911.09539", "2107.04139"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_2342"} +{"question": "Which research papers propose to seek flat minima by injecting noise into the optimizers?", "answer": ["The Anisotropic Noise in Stochastic Gradient Descent: Its Behavior of Escaping from Sharp Minima and Regularization Effects", "Anticorrelated Noise Injection for Improved Generalization", "Low-Pass Filtering SGD for Recovering Flat Optima in the Deep Learning Optimization Landscape"], "answer_arxiv_id": ["1803.00195", "2202.02831", "2201.08025v2"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_2343"} +{"question": "What forms of self-supervised learning (SSL) techniques are most similar methodologically to the representation learning method utilized by the researchers?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Efficient Visual Pretraining with Contrastive Detection", "Emerging Properties in Self-Supervised Vision Transformers", "Deep Clustering for Unsupervised Learning of Visual Features", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2002.05709", "2006.09882", "2103.10957", "2104.14294", "1807.05520", "2006.07733", "2011.10566", "2111.06377"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_2344"} +{"question": "Are there any studies that base offline RL algorithms on the idea of constraining the current policy distribution to be close to the dataset’s policy distribution?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Conservative Q-Learning for Offline Reinforcement Learning", "Behavior Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "2006.04779", "1911.11361"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_2345"} +{"question": "What study demonstrated a major step forward by distilling unsupervised features into discrete semantic labels with the DINO backbone?", "answer": ["Unsupervised Semantic Segmentation by Distilling Feature Correspondences"], "answer_arxiv_id": ["2203.08414"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_2346"} +{"question": "What studies focus on maximizing the entropy of the data collected, thus being recognized as data-based methods in unsupervised RL?", "answer": ["Provably Efficient Maximum Entropy Exploration", "Reinforcement Learning with Prototypical Representations", "Behavior From the Void: Unsupervised Active Pre-Training"], "answer_arxiv_id": ["1812.02690", "2102.11271v2", "2103.04551"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_2347"} +{"question": "What works introduced the idea of using conservative Q-learning to penalize out-of-distribution actions?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning"], "answer_arxiv_id": ["2006.04779"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2348"} +{"question": "What studies renovated Transformer-based models for time series forecasting?", "answer": ["Reformer: The Efficient Transformer", "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting", "Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting"], "answer_arxiv_id": ["2001.04451", "2106.13008", "2012.07436"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2349"} +{"question": "What studies have benchmarked popular semi-supervised learning algorithms using a unified codebase?", "answer": ["FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo\n Labeling"], "answer_arxiv_id": ["2110.08263"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_2350"} +{"question": "Can you provide the reference for the method that extended MaskFormer with learnable queries, deformable multi-scale attention in the decoder, and a masked cross-attention?", "answer": ["Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2112.01527"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_2351"} +{"question": "Could you provide me with the studies about Diversity-Aware Meta Visual Prompting that transfers a network to another target dataset with varying representation distribution?", "answer": ["Diversity-Aware Meta Visual Prompting"], "answer_arxiv_id": ["2303.08138"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_2352"} +{"question": "What papers present the best self-supervised methods using vision transformers as their backbones?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "An Empirical Study of Training Self-Supervised Vision Transformers", "Attention Is All You Need", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2104.14294", "2104.02057", "1706.03762", "2010.11929"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_2353"} +{"question": "What studies have utilized state-similarity metrics for the representation learning problem?", "answer": ["MICo: Improved representations via sampling-based state similarity for Markov decision processes", "Learning Invariant Representations for Reinforcement Learning without Reconstruction", "Contrastive Behavioral Similarity Embeddings for Generalization in Reinforcement Learning"], "answer_arxiv_id": ["2106.08229", "2006.10742", "2101.05265"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_2354"} +{"question": "What works proposed pre-trained vision-language models?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "Emerging Properties in Self-Supervised Vision Transformers", "DINOv2: Learning Robust Visual Features without Supervision", "Learning Transferable Visual Models From Natural Language Supervision", "Revisiting Weakly Supervised Pre-Training of Visual Perception Models", "Align and Prompt: Video-and-Language Pre-training with Entity Prompts", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2002.05709", "1908.08530", "2104.14294", "2304.07193", "2103.00020", "2201.08371", "2112.09583", "2201.12086"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_2355"} +{"question": "Any work on generating panoramic images by employing a latent diffusion model?", "answer": ["PanoGen: Text-Conditioned Panoramic Environment Generation for\n Vision-and-Language Navigation"], "answer_arxiv_id": ["2305.19195"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_2356"} +{"question": "Could you provide me some studies about translating natural language problems to general SQL programs?", "answer": ["Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task", "Picard: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models"], "answer_arxiv_id": ["1809.08887", "2109.05093"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_2357"} +{"question": "What work quantifies a gain for specific regularizers in the field of implicit updates for FTRL?", "answer": ["A Unified View of Regularized Dual Averaging and Mirror Descent with Implicit Updates"], "answer_arxiv_id": ["1009.3240"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2358"} +{"question": "Which works tested the GPT-3.5 model on the AGENDA dataset?", "answer": ["Evaluating Generative Models for Graph-to-Text Generation"], "answer_arxiv_id": ["2307.14712"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_2359"} +{"question": "Which works demonstrated remarkable performance in representing complex geometric structures by learning implicit functions for 3D shapes?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "Convolutional Occupancy Networks", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Implicit Geometric Regularization for Learning Shapes"], "answer_arxiv_id": ["1812.03828", "2003.04618", "1901.05103", "2002.10099"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_2360"} +{"question": "Which research papers address the safety aspects in Inverse Reinforcement Learning?", "answer": ["Learning Robust Rewards with Adversarial Inverse Reinforcement Learning", "Robust Inverse Reinforcement Learning under Transition Dynamics Mismatch", "Learner-aware Teaching: Inverse Reinforcement Learning with Preferences and Constraints", "Inverse Constrained Reinforcement Learning", "X-MEN: Guaranteed XOR-Maximum Entropy Constrained Inverse Reinforcement Learning"], "answer_arxiv_id": ["1710.11248", "2007.01174", "1906.00429", "2011.09999", "2203.11842"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2361"} +{"question": "What works are related to using movMF formulations in image segmentation methods?", "answer": ["Prototype Mixture Models for Few-shot Semantic Segmentation", "SegSort: Segmentation by Discriminative Sorting of Segments"], "answer_arxiv_id": ["2008.03898", "1910.06962"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_2362"} +{"question": "What papers pretrain the model on execution results of random expressions?", "answer": ["Reasoning Like Program Executors"], "answer_arxiv_id": ["2201.11473"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_2363"} +{"question": "What recent studies investigate the idea of constraining the search to expressions that exhibit properties such as compositionality, additivity, and generalized symmetry in SR?", "answer": ["AI Feynman: a Physics-Inspired Method for Symbolic Regression", "AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularity"], "answer_arxiv_id": ["1905.11481", "2006.10782"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_2364"} +{"question": "Which research studies attempted to update the generative model of the latent space within Latent space Bayesian optimization?", "answer": ["Local Latent Space Bayesian Optimization over Structured Inputs"], "answer_arxiv_id": ["2201.11872"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_2365"} +{"question": "Which studies propose the use of standard language models, for instance, GPT in the field of symbolic regression?", "answer": ["SymbolicGPT: A Generative Transformer Model for Symbolic Regression"], "answer_arxiv_id": ["2106.14131"], "source_meta": {"published_time": "20231230"}, "qid": "AutoScholarQuery_train_2366"} +{"question": "What works focus on understanding the working mechanism of KD?", "answer": ["Towards Understanding Knowledge Distillation", "On the Efficacy of Knowledge Distillation", "Understanding and Improving Knowledge Distillation", "Co-advise: Cross Inductive Bias Distillation", "TinyMIM: An Empirical Study of Distilling MIM Pre-trained Models"], "answer_arxiv_id": ["2105.13093", "1910.01348", "2002.03532", "2106.12378", "2301.01296"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_2367"} +{"question": "Which research discusses the impact of changing the random seed on prediction incompatibility in classification tasks?", "answer": ["An Empirical Analysis of Backward Compatibility in Machine Learning Systems"], "answer_arxiv_id": ["2008.04572"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_2368"} +{"question": "Can you give examples of research papers that disentangled both segmentation and motion estimation for object articulation and multi-body scenes?", "answer": ["MultiBodySync: Multi-Body Segmentation and Motion Estimation via 3D Scan Synchronization", "Dynamic 3D Scene Analysis by Point Cloud Accumulation", "Banana: Banach Fixed-Point Network for Pointcloud Segmentation with\n Inter-Part Equivariance", "OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point\n Clouds"], "answer_arxiv_id": ["2101.06605v3", "2207.12394v1", "2305.16314", "2210.04458"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_2369"} +{"question": "Are there any studies that have developed biologically inspired adversarial defenses using predictive or sparse coding techniques?", "answer": ["Recent Advances in Adversarial Training for Adversarial Robustness"], "answer_arxiv_id": ["2102.01356"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_2370"} +{"question": "Which studies are focused on the strategy of deletion in text simplification as commonly taken by professional editors?", "answer": ["Discourse Level Factors for Sentence Deletion in Text Simplification"], "answer_arxiv_id": ["1911.10384"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_2371"} +{"question": "Which studies provided real object datasets, typically captured in studio setups with constrained lighting?", "answer": ["Objaverse: A Universe of Annotated 3D Objects", "Multi-View Photometric Stereo: A Robust Solution and Benchmark Dataset for Spatially Varying Isotropic Materials", "ReLight My NeRF: A Dataset for Novel View Synthesis and Relighting of Real World Objects"], "answer_arxiv_id": ["2212.08051", "2001.06659", "2304.10448"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_2372"} +{"question": "Can you provide works incorporating motion information in ZSVOS, focusing on significant performance improvements?", "answer": ["Hierarchical Feature Alignment Network for Unsupervised Video Object\n Segmentation", "Treating Motion as Option to Reduce Motion Dependency in Unsupervised\n Video Object Segmentation"], "answer_arxiv_id": ["2207.08485", "2209.03138"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_2373"} +{"question": "What research focuses on causal models in video?", "answer": ["Online Action Detection", "MoViNets: Mobile Video Networks for Efficient Video Recognition", "Real-time Online Video Detection with Temporal Smoothing Transformers", "Streaming Video Model", "Online Real-time Multiple Spatiotemporal Action Localisation and\n Prediction", "Tracking without bells and whistles", "Towards Real-Time Multi-Object Tracking", "Towards Streaming Perception", "MOT16: A Benchmark for Multi-Object Tracking", "The 2019 DAVIS Challenge on VOS: Unsupervised Multi-Object Segmentation"], "answer_arxiv_id": ["1604.06506", "2103.11511", "2209.09236", "2303.17228", "1611.08563", "1903.05625", "1909.12605", "2005.10420", "1603.00831", "1905.00737"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_2374"} +{"question": "Are there any works on transformers in offline model-based learning?", "answer": ["Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Masked Autoencoding for Scalable and Generalizable Decision Making"], "answer_arxiv_id": ["2106.02039", "2211.12740"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2375"} +{"question": "What research has been conducted that employs self-attention networks over all tokens from both modalities while training VLMs?", "answer": ["ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "FLAVA: A Foundational Language And Vision Alignment Model", "UNIMO: Towards Unified-Modal Understanding and Generation via\n Cross-Modal Contrastive Learning", "UNITER: UNiversal Image-TExt Representation Learning"], "answer_arxiv_id": ["2102.03334", "2004.06165", "2112.04482", "2012.15409", "1909.11740"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_2376"} +{"question": "What studies extended the results for algorithms such as SGD with heavy ball momentum or adaptive algorithms like AMSGrad and AdaGrad?", "answer": ["Convergence of a Stochastic Gradient Method with Momentum for Non-Smooth Non-Convex Optimization"], "answer_arxiv_id": ["2002.05466"], "source_meta": {"published_time": "20220329"}, "qid": "AutoScholarQuery_train_2377"} +{"question": "What are some key studies addressing exploration and exploitation trade-off in the bandit problem?", "answer": ["Online Stochastic Linear Optimization under One-bit Feedback"], "answer_arxiv_id": ["1509.07728"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_2378"} +{"question": "Could you provide me with the name of the work that extended previous adversarial imitation learning solution based on latent samples?", "answer": ["Visual Adversarial Imitation Learning using Variational Models"], "answer_arxiv_id": ["2107.08829"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_2379"} +{"question": "Can you provide a few works that have explored the temporal reasoning capability of Language Learning Models (LLMs) with time-sensitive Question Answering (QA) datasets?", "answer": ["SituatedQA: Incorporating Extra-Linguistic Contexts into QA", "Improving Time Sensitivity for Question Answering over Temporal\n Knowledge Graphs", "Towards Benchmarking and Improving the Temporal Reasoning Capability of\n Large Language Models"], "answer_arxiv_id": ["2109.06157", "2203.00255", "2306.08952"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_2380"} +{"question": "What works have utilized diffusion models in various image restoration tasks?", "answer": ["JPEG Artifact Correction using Denoising Diffusion Restoration Models", "Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "Inversion by Direct Iteration: An Alternative to Denoising Diffusion for Image Restoration"], "answer_arxiv_id": ["2209.11888", "2201.11793", "2212.00490", "2303.11435"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_2381"} +{"question": "Which work allowed pseudo-3D rotations but required access to 3D models to create a dataset for fine-tuning a controlNet backend?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2303.11328"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_2382"} +{"question": "Which research introduced Consistency Models, a single-step generative approach that learns from a pre-trained diffusion model?", "answer": ["Consistency Models"], "answer_arxiv_id": ["2303.01469"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_2383"} +{"question": "Could you reference the studies that have applied diffusion models to generate audio from text and vision prompts?", "answer": ["AudioLDM: Text-to-Audio Generation with Latent Diffusion Models", "Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models"], "answer_arxiv_id": ["2301.12503", "2301.12661"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_2384"} +{"question": "What study introduces the application of attention to sets?", "answer": ["Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks"], "answer_arxiv_id": ["1810.00825"], "source_meta": {"published_time": "20220826"}, "qid": "AutoScholarQuery_train_2385"} +{"question": "Any studies discussing how kernel alignments between kernels and training labels accelerate training?", "answer": ["A Theory of Neural Tangent Kernel Alignment and Its Influence on Training"], "answer_arxiv_id": ["2105.14301"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_2386"} +{"question": "Which study proposed the creation of unit tests for Python function completion problems?", "answer": ["CodeT: Code Generation with Generated Tests"], "answer_arxiv_id": ["2207.10397"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_2387"} +{"question": "Which papers optimize RL algorithm components via meta-gradients in PMO?", "answer": ["Meta-Gradient Reinforcement Learning", "Meta-Gradient Reinforcement Learning with an Objective Discovered Online"], "answer_arxiv_id": ["1805.09801", "2007.08433"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_2388"} +{"question": "Which papers are about cross encoders that encode queries and documents together in document retrieval?", "answer": ["Multi-Stage Document Ranking with BERT", "RocketQA: An Optimized Training Approach to Dense Passage Retrieval for Open-Domain Question Answering", "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT", "Sparse, Dense, and Attentional Representations for Text Retrieval"], "answer_arxiv_id": ["1910.14424", "2010.08191", "2004.12832", "2005.00181"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_2389"} +{"question": "Which works propose neural learning approaches to predict primitive decomposition from a collection of shapes?", "answer": ["Learning Shape Abstractions by Assembling Volumetric Primitives", "Superquadrics Revisited: Learning 3D Shape Parsing beyond Cuboids", "Learning Unsupervised Hierarchical Part Decomposition of 3D Objects from a Single RGB Image", "Primitive-based Shape Abstraction via Nonparametric Bayesian Inference", "Supervised Fitting of Geometric Primitives to 3D Point Clouds", "CPFN: Cascaded Primitive Fitting Networks for High-Resolution Point Clouds", "Learning elementary structures for 3D shape generation and matching", "Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networks", "Learnable Earth Parser: Discovering 3D Prototypes in Aerial Scans"], "answer_arxiv_id": ["1612.00404", "1904.09970", "2004.01176", "2203.14714", "1811.08988", "2109.00113v2", "1908.04725", "2103.10429v1", "2304.09704"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_2390"} +{"question": "Which work utilized template deformation for 3D caricature auto-decoder?", "answer": ["Deep Deformable 3D Caricatures with Learned Shape Control"], "answer_arxiv_id": ["2207.14593"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_2391"} +{"question": "Could you provide me some papers that have attempted to improve model efficiency by uniformly sampling video sequences?", "answer": ["P-STMO: Pre-Trained Spatial Temporal Many-to-One Model for 3D Human Pose\n Estimation", "DeciWatch: A Simple Baseline for 10x Efficient 2D and 3D Pose Estimation", "Uplift and Upsample: Efficient 3D Human Pose Estimation with Uplifting\n Transformers"], "answer_arxiv_id": ["2203.07628", "2203.08713", "2210.06110"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_2392"} +{"question": "What research papers focus on automatic frequency tuning through the use of nonlinearity activation functions?", "answer": ["Implicit Neural Representations with Periodic Activation Functions", "Beyond Periodicity: Towards a Unifying Framework for Activations in\n Coordinate-MLPs", "WIRE: Wavelet Implicit Neural Representations"], "answer_arxiv_id": ["2006.09661", "2111.15135", "2301.05187"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_2393"} +{"question": "Could you provide me some studies that unified detection datasets and image-text datasets through region-text matching?", "answer": ["GLIPv2: Unifying Localization and Vision-Language Understanding", "Learning Object-Language Alignments for Open-Vocabulary Object Detection", "DetCLIP: Dictionary-Enriched Visual-Concept Paralleled Pre-training for\n Open-world Detection", "DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via\n Word-Region Alignment"], "answer_arxiv_id": ["2206.05836", "2211.14843", "2209.09407", "2304.04514"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_2394"} +{"question": "What papers focused on neurons for localization analysis?", "answer": ["Are Neural Nets Modular? Inspecting Functional Modularity Through Differentiable Weight Masks", "Sparse Interventions in Language Models with Differentiable Masking", "Learning to Generate Reviews and Discovering Sentiment", "Interpreting Deep Visual Representations via Network Dissection", "The emergence of number and syntax units in LSTM language models", "Understanding the Role of Individual Units in a Deep Neural Network", "Causal Mediation Analysis for Interpreting Neural NLP: The Case of Gender Bias", "Compositional Explanations of Neurons", "Knowledge Neurons in Pretrained Transformers", "Mechanisms for Handling Nested Dependencies in Neural-Network Language Models and Humans", "An Interpretability Illusion for BERT", "Natural Language Descriptions of Deep Visual Features", "Local Relighting of Real Scenes", "Finding Skill Neurons in Pre-trained Transformer-based Language Models"], "answer_arxiv_id": ["2010.02066", "2112.06837", "1704.01444", "1711.05611", "1903.07435", "2009.05041", "2004.12265", "2006.14032", "2104.08696", "2006.11098v2", "2104.07143", "2201.11114", "2207.02774", "2211.07349"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_2395"} +{"question": "Which studies use COUNT-based intrinsic reward for efficient exploration in reinforcement learning?", "answer": ["#Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning"], "answer_arxiv_id": ["1611.04717"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_2396"} +{"question": "Which works employed bias labels to discourage the model from learning specific bias features?", "answer": ["Learning Not to Learn: Training Deep Neural Networks with Biased Data", "EnD: Entangling and Disentangling deep representations for bias\n correction", "Distributionally Robust Neural Networks for Group Shifts: On the\n Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1812.10352", "2103.02023", "1911.08731"], "source_meta": {"published_time": "20240430"}, "qid": "AutoScholarQuery_train_2397"} +{"question": "Which studies are about knowledge distillation?", "answer": ["Distilling the Knowledge in a Neural Network", "Deep Mutual Learning", "Online Knowledge Distillation via Multi-branch Diversity Enhancement", "Decoupled Knowledge Distillation", "Curriculum Temperature for Knowledge Distillation", "Knowledge Distillation with the Reused Teacher Classifier", "Relational Knowledge Distillation", "Mutual Contrastive Learning for Visual Representation Learning"], "answer_arxiv_id": ["1503.02531", "1706.00384", "2010.00795", "2203.08679", "2211.16231", "2203.14001", "1904.05068", "2104.12565"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_2398"} +{"question": "What researches utilize the Diffusion model to estimate the motion prior in the denoising process?", "answer": ["Denoising Diffusion Probabilistic Models", "Hierarchical Integration Diffusion Model for Realistic Image Deblurring", "Generative Diffusion Prior for Unified Image Restoration and Enhancement", "DiffIR: Efficient Diffusion Model for Image Restoration"], "answer_arxiv_id": ["2006.11239", "2305.12966", "2304.01247", "2303.09472"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_2399"} +{"question": "Where can I find information on the use of the Open Assistant dataset for RLHF training?", "answer": ["OpenAssistant Conversations - Democratizing Large Language Model Alignment"], "answer_arxiv_id": ["2304.07327"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_2400"} +{"question": "What work proposed GeoDiff, a diffusion model that predicts molecular conformations from a molecular graph?", "answer": ["GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation"], "answer_arxiv_id": ["2203.02923"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_2401"} +{"question": "Which studies have explored audio-visual learning to learn audio-visual correspondence from videos?", "answer": ["SoundNet: Learning Sound Representations from Unlabeled Video", "Ambient Sound Provides Supervision for Visual Learning", "Look, Listen and Learn", "Cooperative Learning of Audio and Video Models from Self-Supervised Synchronization", "Learning to Localize Sound Source in Visual Scenes", "The Sound of Pixels", "The Sound of Motions", "Music Gesture for Visual Sound Separation", "Learning Representations from Audio-Visual Spatial Alignment", "Robust Audio-Visual Instance Discrimination", "Audio-Visual Instance Discrimination with Cross-Modal Agreement", "DiffAVA: Personalized Text-to-Audio Generation with Visual Alignment", "A Unified Audio-Visual Learning Framework for Localization, Separation, and Recognition", "Audio-Visual Class-Incremental Learning", "Class-Incremental Grouping Network for Continual Audio-Visual Learning"], "answer_arxiv_id": ["1610.09001", "1608.07017", "1705.08168", "1807.00230", "1803.03849v1", "1804.03160", "1904.05979", "2004.09476", "2011.01819", "2103.15916", "2004.12943", "2305.12903", "2305.19458", "2308.11073", "2309.05281"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_2402"} +{"question": "What works use knowledge distillation to enhance task transfer in continual reinforcement learning?", "answer": ["Policy Distillation", "Learning without Forgetting", "Progress & Compress: A scalable framework for continual learning", "Policy Consolidation for Continual Reinforcement Learning", "DisCoRL: Continual Reinforcement Learning via Policy Distillation"], "answer_arxiv_id": ["1511.06295", "1606.09282", "1805.06370", "1902.00255", "1907.05855"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_2403"} +{"question": "Which works studied how diffusion models utilize the learned score functions to estimate the data distribution?", "answer": ["Sampling is as easy as learning the score: theory for diffusion models with minimal data assumptions", "Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling", "Convergence for score-based generative modeling with polynomial complexity", "Convergence of score-based generative modeling for general data distributions"], "answer_arxiv_id": ["2209.11215", "2106.01357", "2206.06227", "2209.12381"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_2404"} +{"question": "Could you provide me some papers where machine learning has been used for mapping in radar systems?", "answer": ["See Through Smoke: Robust Indoor Mapping with Low-cost mmWave Radar"], "answer_arxiv_id": ["1911.00398"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_2405"} +{"question": "Which paper is known to make attempts in extracting sequential actions and dealing with partial non-sequential actions with LLMs?", "answer": ["Leveraging pre-trained language models for conversational information\n seeking from text"], "answer_arxiv_id": ["2204.03542"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_2406"} +{"question": "What are some contributions in context modeling to refine the audio-visual features?", "answer": ["Active Speakers in Context", "How to Design a Three-Stage Architecture for Audio-Visual Active Speaker\n Detection in the Wild", "Is Someone Speaking? Exploring Long-term Temporal Features for\n Audio-visual Active Speaker Detection", "MAAS: Multi-modal Assignation for Active Speaker Detection", "End-to-End Active Speaker Detection"], "answer_arxiv_id": ["2005.09812", "2106.03932", "2107.06592", "2101.03682", "2203.14250"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_2407"} +{"question": "Could you provide me some studies about chaotic behaviors in learning dynamics?", "answer": ["Fractal Structure and Generalization Properties of Stochastic Optimization Algorithms", "Gradients are Not All You Need", "On the difficulty of training Recurrent Neural Networks", "PIPPS: Flexible Model-Based Policy Search Robust to the Curse of Chaos", "Do Differentiable Simulators Give Better Policy Gradients?"], "answer_arxiv_id": ["2106.04881", "2111.05803", "1211.5063", "1902.01240", "2202.00817"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_2408"} +{"question": "What work studied random cumulants?", "answer": ["The Value-Improvement Path: Towards Better Representations for Reinforcement Learning", "On The Effect of Auxiliary Tasks on Representation Dynamics"], "answer_arxiv_id": ["2006.02243", "2102.13089v1"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_2409"} +{"question": "Which paper used parameters such as sleeve length or chest circumference to automate pattern design?", "answer": ["Learning a Shared Shape Space for Multimodal Garment Design", "Fully Convolutional Graph Neural Networks for Parametric Virtual Try-On"], "answer_arxiv_id": ["1806.11335", "2009.04592"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_2410"} +{"question": "What works can be categorized under decoupling methods in imbalanced learning?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition", "Long-Tailed Recognition via Weight Balancing"], "answer_arxiv_id": ["1910.09217", "2203.14197"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_2411"} +{"question": "Which studies utilize complex physics simulation steps or feature line estimation for free-flowing garment handling?", "answer": ["Deep Fashion3D: A Dataset and Benchmark for 3D Garment Reconstruction\n from Single Images", "Registering Explicit to Implicit: Towards High-Fidelity Garment mesh\n Reconstruction from Single Images"], "answer_arxiv_id": ["2003.12753", "2203.15007"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_2412"} +{"question": "Which works analysed generalization for two layer networks using techniques from random matrix theory and statistical mechanics?", "answer": ["Universality Laws for High-Dimensional Learning with Random Features", "The Neural Tangent Kernel in High Dimensions: Triple Descent and a Multi-Scale Theory of Generalization", "A Precise Performance Analysis of Learning with Random Features", "Understanding Double Descent Requires a Fine-Grained Bias-Variance Decomposition", "Generalisation error in learning with random features and the hidden manifold model", "Double Trouble in Double Descent: Bias and Variance(s) in the Lazy Regime", "Triple descent and the Two Kinds of Overfitting: Where & Why do they Appear?"], "answer_arxiv_id": ["2009.07669", "2008.06786", "2008.11904", "2011.03321", "2002.09339", "2003.01054", "2006.03509"], "source_meta": {"published_time": "20221223"}, "qid": "AutoScholarQuery_train_2413"} +{"question": "Which works pertained to safe, anytime-valid inference, specifically, confidence sequences?", "answer": ["Estimating means of bounded random variables by betting"], "answer_arxiv_id": ["2010.09686v7"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_2414"} +{"question": "Could you give an example of research that transforms imbalanced regression to classification?", "answer": ["Uniformity in Heterogeneity:Diving Deep into Count Interval Partition\n for Crowd Counting"], "answer_arxiv_id": ["2107.12619"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2415"} +{"question": "What papers are focused on problems that come with the use of language modality in multimodal video understanding?", "answer": ["HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips"], "answer_arxiv_id": ["1906.03327"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_2416"} +{"question": "What studies assessed LLMs’ proficiency in solving complex tasks by analyzing their responses?", "answer": ["Identifying the Risks of LM Agents with an LM-Emulated Sandbox", "R-Judge: Benchmarking Safety Risk Awareness for LLM Agents"], "answer_arxiv_id": ["2309.15817v2", "2401.10019"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_2417"} +{"question": "Which papers use the non-tuning paradigm in generative LLM-based approaches?", "answer": ["Is ChatGPT a Good Recommender? A Preliminary Study", "Uncovering ChatGPT's Capabilities in Recommender Systems", "Large Language Model Augmented Narrative Driven Recommendations", "Rethinking the Evaluation for Conversational Recommendation in the Era\n of Large Language Models", "Large Language Models are Zero-Shot Rankers for Recommender Systems"], "answer_arxiv_id": ["2304.10149", "2305.02182", "2306.02250", "2305.13112", "2305.08845"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_2418"} +{"question": "Can you name the studies focused on explicitly modeling false/hard negative samples in training to improve contrastive learning?", "answer": ["Max-Margin Contrastive Learning", "Hard Negative Mixing for Contrastive Learning", "Contrastive Learning with Hard Negative Samples"], "answer_arxiv_id": ["2112.11450", "2010.01028v2", "2010.04592"], "source_meta": {"published_time": "20220716"}, "qid": "AutoScholarQuery_train_2419"} +{"question": "Could you cite some papers that proposed constraining the policy outputs in offline reinforcement learning?", "answer": ["A Minimalist Approach to Offline Reinforcement Learning", "Revisiting the Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["2106.06860", "2305.09836"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_2420"} +{"question": "What studies in the field of implicit volumetric representations have proposed to encode scene geometry as the density field?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2103.13415", "2111.11215", "2201.05989", "2111.12077", "2304.06706"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_2421"} +{"question": "Which work evaluates the robustness of KGW against paraphrase attacks and copy-paste attacks?", "answer": ["On the Reliability of Watermarks for Large Language Models"], "answer_arxiv_id": ["2306.04634"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_2422"} +{"question": "Can you name some studies about post-processing purification as a backdoor defense strategy?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks", "Adversarial Neuron Pruning Purifies Backdoored Deep Models", "Data-free Backdoor Removal based on Channel Lipschitzness"], "answer_arxiv_id": ["1805.12185", "2110.14430", "2208.03111"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_2423"} +{"question": "What works are noteworthy in 3D anomaly detection?", "answer": ["Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors", "Back to the Feature: Classical 3D Features are (Almost) All You Need for\n 3D Anomaly Detection", "Asymmetric Student-Teacher Networks for Industrial Anomaly Detection", "Multimodal Industrial Anomaly Detection via Hybrid Fusion", "Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detection", "EasyNet: An Easy Network for 3D Industrial Anomaly Detection"], "answer_arxiv_id": ["2202.11660", "2203.05550", "2210.07829", "2303.00601", "2303.13194v1", "2307.13925"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_2424"} +{"question": "Which study investigated the emergence of object recognition in children?", "answer": ["The developmental trajectory of object recognition robustness: children are like small adults but unlike big deep neural networks"], "answer_arxiv_id": ["2205.10144"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_2425"} +{"question": "Which research papers propose to use BERT for raw text pre-processing in a multi-modal learning context?", "answer": ["VisualBERT: A Simple and Performant Baseline for Vision and Language", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "UNITER: UNiversal Image-TExt Representation Learning"], "answer_arxiv_id": ["1908.03557", "2004.06165", "1909.11740"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_2426"} +{"question": "What works have altered model architectures, inference procedures, and training objectives to address the infill of text with standard left-to-right language models?", "answer": ["CM3: A Causal Masked Multimodal Model of the Internet", "Insertion Transformer: Flexible Sequence Generation via Insertion Operations", "Reflective Decoding: Beyond Unidirectional Generation with Off-the-Shelf Language Models", "HTLM: Hyper-Text Pre-Training and Prompting of Language Models"], "answer_arxiv_id": ["2201.07520", "1902.03249", "2010.08566", "2107.06955"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_2427"} +{"question": "Could you provide me some works that used contrastive learning for representation learning in the realm of RL?", "answer": ["Temporal Difference Learning for Model Predictive Control", "Data-Efficient Reinforcement Learning with Self-Predictive Representations", "Learning Temporally-Consistent Representations for Data-Efficient Reinforcement Learning"], "answer_arxiv_id": ["2203.04955", "2007.05929", "2110.04935"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2428"} +{"question": "Which works use graph contrastive learning that treats different parts of a graph as positive pairs and construct negative examples from a corrupted graph?", "answer": ["Strategies for Pre-training Graph Neural Networks", "Sub-graph Contrast for Scalable Self-Supervised Graph Representation Learning", "Graph Representation Learning via Graphical Mutual Information Maximization", "InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization"], "answer_arxiv_id": ["1905.12265", "2009.10273", "2002.01169", "1908.01000"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_2429"} +{"question": "Any works about program synthesis approaches that condition on natural language descriptions?", "answer": ["Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task", "Learning to Mine Aligned Code and Natural Language Pairs from Stack Overflow", "SPoC: Search-based Pseudocode to Code", "Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["1809.08887", "1805.08949", "1906.04908", "2107.03374"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_2430"} +{"question": "Any works about the use of INRs in cross-model media representation or compression?", "answer": ["ObjectFolder 2.0: A Multisensory Object Dataset for Sim2Real Transfer", "Implicit Neural Representations for Image Compression"], "answer_arxiv_id": ["2204.02389", "2112.04267v2"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_2431"} +{"question": "Can you list the works that integrated diffusion models into style transfer?", "answer": ["ArtFusion: Controllable Arbitrary Style Transfer using Dual Conditional\n Latent Diffusion Models", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2306.09330", "2006.11239"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2432"} +{"question": "Which work first proposed Adaptive Computation Time (ACT) algorithm for dynamic halting in recurrent neural networks?", "answer": ["Adaptive Computation Time for Recurrent Neural Networks"], "answer_arxiv_id": ["1603.08983"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2433"} +{"question": "Could you provide me the studies about GCL algorithms with augmentations?", "answer": ["Deep Graph Infomax", "Graph Contrastive Learning with Augmentations", "Graph Contrastive Learning with Adaptive Augmentation", "Large-Scale Representation Learning on Graphs via Bootstrapping", "From Canonical Correlation Analysis to Self-supervised Graph Neural Networks"], "answer_arxiv_id": ["1809.10341", "2010.13902", "2010.14945", "2102.06514", "2106.12484"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_2434"} +{"question": "Which research works first attempted to combine ae models with gps?", "answer": ["Gaussian Process Prior Variational Autoencoders"], "answer_arxiv_id": ["1810.11738"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_2435"} +{"question": "What study used a gradient-guided method to select small-scale data for continual pre-training?", "answer": ["Understanding In-Context Learning via Supportive Pretraining Data"], "answer_arxiv_id": ["2306.15091"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_2436"} +{"question": "What papers discuss use of graph neural networks for neural network building blocks?", "answer": ["Neural Operator: Graph Kernel Network for Partial Differential Equations", "Multipole Graph Neural Operator for Parametric Partial Differential Equations"], "answer_arxiv_id": ["2003.03485", "2006.09535"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_2437"} +{"question": "Which studies developed publicly available PGC databases to investigate the effects of spatial resolution and frame rate variations on video quality?", "answer": ["Subjective and Objective Quality Assessment of High Frame Rate Videos", "A Subjective and Objective Study of Space-Time Subsampled Video Quality"], "answer_arxiv_id": ["2007.11634", "2102.00088v1"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_2438"} +{"question": "Which studies proposed datasets for photometric stereo?", "answer": ["Multi-View Photometric Stereo: A Robust Solution and Benchmark Dataset\n for Spatially Varying Isotropic Materials", "LUCES: A Dataset for Near-Field Point Light Source Photometric Stereo", "From Shading to Local Shape"], "answer_arxiv_id": ["2001.06659", "2104.13135", "1310.2916"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_2439"} +{"question": "What papers proposed replay-based methods for incremental learning?", "answer": ["A continual learning survey: Defying forgetting in classification tasks", "A Comprehensive Study of Class Incremental Learning Algorithms for\n Visual Tasks", "Large Scale Incremental Learning", "PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning", "iCaRL: Incremental Classifier and Representation Learning", "Gradient Episodic Memory for Continual Learning", "Efficient Lifelong Learning with A-GEM", "Learning to Learn without Forgetting by Maximizing Transfer and\n Minimizing Interference"], "answer_arxiv_id": ["1909.08383", "2011.01844", "1905.13260", "2004.13513", "1611.07725", "1706.08840", "1812.00420", "1810.11910"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_2440"} +{"question": "Which works explore the connection between information-theoretic bounds and algorithmic stability?", "answer": ["A PAC-Bayesian Analysis of Randomized Learning with Application to Stochastic Gradient Descent", "PAC-Bayes bounds for stable algorithms with instance-dependent priors", "On Generalization Error Bounds of Noisy Gradient Methods for Non-Convex Learning", "Reasoning About Generalization via Conditional Mutual Information", "Information-theoretic generalization bounds for black-box learning algorithms", "Stability Based Generalization Bounds for Exponential Family Langevin Dynamics", "On Leave-One-Out Conditional Mutual Information For Generalization"], "answer_arxiv_id": ["1709.06617", "1806.06827", "1902.00621", "2001.09122", "2110.01584", "2201.03064", "2207.00581"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_2441"} +{"question": "What studies exist on the occlusion problem in the context of manipulation of 3D articulated objects?", "answer": ["Graspness Discovery in Clutters for Fast and Accurate Grasp Detection", "Robotic Pick-and-Place of Novel Objects in Clutter with Multi-Affordance Grasping and Cross-Domain Image Matching", "Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenes"], "answer_arxiv_id": ["2406.11142v1", "1710.01330", "2103.14127"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_2442"} +{"question": "Which works have proposed architectures to preserve invariant and equivariant properties for transformations in 3D molecular representations?", "answer": ["SchNet: A continuous-filter convolutional neural network for modeling quantum interactions", "Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "Directional Message Passing for Molecular Graphs", "Spherical Message Passing for 3D Molecular Graphs", "ComENet: Towards Complete and Efficient Message Passing for 3D Molecular Graphs", "LieTransformer: Equivariant Self-Attention for Lie Groups", "Benchmarking Graphormer on Large-Scale Molecular Modeling Datasets"], "answer_arxiv_id": ["1706.08566", "1802.08219", "2003.03123", "2102.05013", "2206.08515", "2012.10885", "2203.04810"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_2443"} +{"question": "Could you provide me some studies about self-supervised learning methods focused on masked image modeling (MIM)?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "Context Autoencoder for Self-Supervised Representation Learning", "Masked Autoencoders Are Scalable Vision Learners", "BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizers", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "MVP: Multimodality-guided Visual Pre-training", "SimMIM: a Simple Framework for Masked Image Modeling", "iBOT : Image BERT Pre-Training with Online Tokenizer"], "answer_arxiv_id": ["2106.08254", "2202.03026", "2111.06377", "2208.06366", "2112.09133", "2203.05175", "2111.09886", "2111.07832"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_2444"} +{"question": "Who developed the single-loop two-timescale GDA method that improved the complexity?", "answer": ["Convergence Rates of Two-Time-Scale Gradient Descent-Ascent Dynamics for Solving Nonconvex Min-Max Problems"], "answer_arxiv_id": ["2112.09579"], "source_meta": {"published_time": "20221226"}, "qid": "AutoScholarQuery_train_2445"} +{"question": "Could you provide me some works that have utilized machine learning for decoding of cognitive processes and diagnosing mental health disorders?", "answer": ["ASD-DiagNet: A hybrid learning approach for detection of Autism Spectrum Disorder using fMRI data"], "answer_arxiv_id": ["1904.07577"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_2446"} +{"question": "Could you provide me some studies about graph-transformer methods implemented in 3D human pose estimation?", "answer": ["Pose-Oriented Transformer with Uncertainty-Guided Refinement for\n 2D-to-3D Human Pose Estimation", "DiffPose: Toward More Reliable 3D Pose Estimation"], "answer_arxiv_id": ["2302.07408", "2211.16940"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_2447"} +{"question": "Any works utilizing deep learning to find approximate optimal auction?", "answer": ["Certifying Strategyproof Auction Networks", "PreferenceNet: Encoding Human Preferences in Auction Design with Deep Learning", "A Permutation-Equivariant Neural Network Architecture For Auction Design", "A Context-Integrated Transformer-Based Neural Network for Auction Design", "Optimal-er Auctions through Attention"], "answer_arxiv_id": ["2006.08742", "2106.03215", "2003.01497", "2201.12489", "2202.13110"], "source_meta": {"published_time": "20230520"}, "qid": "AutoScholarQuery_train_2448"} +{"question": "Could you provide me some works that use feature removal strategies specifically for image data?", "answer": ["Visualizing and Understanding Convolutional Networks", "Interpretable Explanations of Black Boxes by Meaningful Perturbation", "Understanding Deep Networks via Extremal Perturbations and Smooth Masks", "Real Time Image Saliency for Black Box Classifiers"], "answer_arxiv_id": ["1311.2901", "1704.03296", "1910.08485", "1705.07857"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_2449"} +{"question": "What works developed learning algorithms with non-asymptotic convergence guarantees in terms of the average of value functions?", "answer": ["Exploration-Exploitation in Constrained MDPs", "Provably Efficient Safe Exploration via Primal-Dual Policy Optimization", "Constrained Upper Confidence Reinforcement Learning", "Upper Confidence Primal-Dual Reinforcement Learning for CMDP with Adversarial Loss", "Constrained episodic reinforcement learning in concave-convex and knapsack settings", "Provably Efficient Algorithms for Multi-Objective Competitive RL", "Learning Policies with Zero or Bounded Constraint Violation for Constrained MDPs", "A Sample-Efficient Algorithm for Episodic Finite-Horizon MDP with Constraints", "A Simple Reward-free Approach to Constrained Reinforcement Learning", "DOPE: Doubly Optimistic and Pessimistic Exploration for Safe Reinforcement Learning", "Provably Efficient Model-Free Constrained RL with Linear Function Approximation", "Provably Efficient Model-Free Algorithms for Non-stationary CMDPs", "A Best-of-Both-Worlds Algorithm for Constrained MDPs with Long-Term Constraints", "Safe Posterior Sampling for Constrained MDPs with Bounded Constraint Violation"], "answer_arxiv_id": ["2003.02189", "2003.00534", "2001.09377", "2003.00660", "2006.05051", "2102.03192", "2106.02684", "2009.11348", "2107.05216", "2112.00885", "2206.11889", "2303.05733", "2304.14326", "2301.11547"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_2450"} +{"question": "Which works fine-tuned network weights or a subset of diffusion model layers in an effort to reconstruct the target concept in customization tasks?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2208.12242", "2212.04488"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_2451"} +{"question": "Which researchers studied the effects of stochasticity in Stochastic Gradient Descent (SGD) on generalisation?", "answer": ["A Variational Analysis of Stochastic Gradient Algorithms", "Train longer, generalize better: closing the generalization gap in large batch training of neural networks", "Stochastic gradient descent performs variational inference, converges to limit cycles for deep networks", "An Alternative View: When Does SGD Escape Local Minima?"], "answer_arxiv_id": ["1602.02666", "1705.08741", "1710.11029", "1802.06175"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_2452"} +{"question": "Are there works about implementing adaptive noise scheduling as a strategy to speed up the sampling in diffusion models?", "answer": ["Noise Estimation for Generative Diffusion Models"], "answer_arxiv_id": ["2104.02600"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_2453"} +{"question": "What research argues that the contrastive loss is implicitly doing SNE with 'positive' pairs constructed from data augmentation?", "answer": ["Your Contrastive Learning Is Secretly Doing Stochastic Neighbor Embedding"], "answer_arxiv_id": ["2205.14814"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_2454"} +{"question": "Which works introduced GNN-based GAD methods and improved detection performance in recent years?", "answer": ["A Comprehensive Survey on Graph Anomaly Detection with Deep Learning"], "answer_arxiv_id": ["2106.07178"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_2455"} +{"question": "Which papers are about Fourier neural operator (FNO)", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations"], "answer_arxiv_id": ["2010.08895"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2456"} +{"question": "In open-set SSL, what are the works that propose techniques to alleviate the influence of Out-of-Distribution (OOD) data?", "answer": ["Open-World Semi-Supervised Learning", "Harnessing Out-Of-Distribution Examples via Augmenting Content and Style"], "answer_arxiv_id": ["2102.03526", "2207.03162"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_2457"} +{"question": "Which paper introduced 3D Gaussians that provide accurate selection and manipulation of editing areas?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_2458"} +{"question": "Could you provide me with some studies on methods that operate on the sphere but are not rotation-equivariant?", "answer": ["Spherical CNNs on Unstructured Grids"], "answer_arxiv_id": ["1901.02039"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_2459"} +{"question": "Can you mention some studies about learning to reverse different types of corruptions?", "answer": ["Generative Modelling With Inverse Heat Dissipation", "Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise", "Soft Diffusion Score Matching For General Corruptions"], "answer_arxiv_id": ["2206.13397v7", "2208.09392", "2209.05442"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_2460"} +{"question": "Are there any recent studies proposing repository-level code generation frameworks and benchmarks?", "answer": ["CoCoMIC: Code Completion By Jointly Modeling In-file and Cross-file Context", "RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation"], "answer_arxiv_id": ["2212.10007", "2303.12570"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_2461"} +{"question": "Can you demonstrate some works that apply SGD algorithms to the 2-port model?", "answer": ["Exponential Graph is Provably Efficient for Decentralized Deep Training"], "answer_arxiv_id": ["2110.13363"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2462"} +{"question": "Can you provide some works about Generalized Category Discovery (GCD)?", "answer": ["Generalized Category Discovery", "Open-World Semi-Supervised Learning"], "answer_arxiv_id": ["2201.02609", "2102.03526"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_2463"} +{"question": "Can you provide some studies about unsupervised and self-supervised learning methods in the field of Computer Vision?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: a Simple Framework for Masked Image Modeling", "BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizers", "Stare at What You See: Masked Image Modeling without Reconstruction"], "answer_arxiv_id": ["1911.05722", "2003.04297", "2002.05709", "2106.08254", "2111.06377", "2111.09886", "2208.06366", "2211.08887"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_2464"} +{"question": "What papers aim to maximize mutual information (MI) between positive samples in the theoretical understanding of contrastive SSL?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Learning Representations by Maximizing Mutual Information Across Views", "Learning deep representations by mutual information estimation and maximization", "Contrastive Multiview Coding", "What Makes for Good Views for Contrastive Learning?", "On Mutual Information Maximization for Representation Learning"], "answer_arxiv_id": ["1807.03748", "1906.00910", "1808.06670", "1906.05849", "2005.10243", "1907.13625"], "source_meta": {"published_time": "20211101"}, "qid": "AutoScholarQuery_train_2465"} +{"question": "Which paper inspired the design of a prior-guide large window transformer (PLWformer)?", "answer": ["SRFormer: Permuted Self-Attention for Single Image Super-Resolution"], "answer_arxiv_id": ["2303.09735"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_2466"} +{"question": "Can you list some works in Hyperparameter Optimization for Federated Learning that learn one set of hyperparameters for all clients?", "answer": ["Evaluation of Hyperparameter-Optimization Approaches in an Industrial Federated Learning System", "Federated Hyperparameter Tuning: Challenges, Baselines, and Connections to Weight-Sharing"], "answer_arxiv_id": ["2110.08202", "2106.04502"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_2467"} +{"question": "What work uses multi-layer perceptrons as a trilinear filter to produce a voxel grid of temporal features for the input representation of asynchronous event streams?", "answer": ["End-to-End Learning of Representations for Asynchronous Event-Based Data"], "answer_arxiv_id": ["1904.08245"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_2468"} +{"question": "Which research employ gradient updates to refine rollouts prior to taking each real action?", "answer": ["Policy Gradient Search: Online Planning and Expert Iteration without Search Trees", "Model-Based Planning with Discrete and Continuous Actions", "Scalable Online Planning via Reinforcement Learning Fine-Tuning"], "answer_arxiv_id": ["1904.03646", "1705.07177", "2109.15316"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_2469"} +{"question": "What are some works that focused on using human feedback for learning metrics in specific tasks?", "answer": ["Adaptively Learning the Crowd Kernel", "Cost-Effective HITs for Relative Similarity Comparisons"], "answer_arxiv_id": ["1105.1033", "1404.3291"], "source_meta": {"published_time": "20220826"}, "qid": "AutoScholarQuery_train_2470"} +{"question": "Could you provide some studies using Transformer model for 3D part assembly?", "answer": ["3D Part Assembly Generation with Instance Encoded Transformer", "Neural Shape Mating: Self-Supervised Object Assembly with Adversarial\n Shape Priors"], "answer_arxiv_id": ["2207.01779", "2205.14886"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2471"} +{"question": "Can you provide studies that incorporate explanations within the model architecture in understanding complex models?", "answer": ["Concept Bottleneck Models"], "answer_arxiv_id": ["2007.04612"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_train_2472"} +{"question": "Could you provide me some works about composing multiple large language models or models with symbolic functions for multi-step reasoning?", "answer": ["Maieutic Prompting: Logically Consistent Reasoning with Recursive Explanations", "Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning", "Measuring and Narrowing the Compositionality Gap in Language Models", "TALM: Tool Augmented Language Models", "PAL: Program-aided Language Models", "Toolformer: Language Models Can Teach Themselves to Use Tools"], "answer_arxiv_id": ["2205.11822", "2205.09712", "2210.03350", "2205.12255", "2211.10435", "2302.04761"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_2473"} +{"question": "What works extended the Bayesian Optimization (BO) algorithms to the batch-wise calculation of an objective function?", "answer": ["Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization", "Parallel Bayesian Optimization of Multiple Noisy Objectives with Expected Hypervolume Improvement"], "answer_arxiv_id": ["2006.05078", "2105.08195"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_2474"} +{"question": "Which papers have used mutual information optimization to maximize the estimated mutual information lower bound?", "answer": ["MINE: Mutual Information Neural Estimation", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1801.04062", "1807.03748"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_2475"} +{"question": "What studies aim to address the lack of diversity in GAN-generated samples by ensembling GANs?", "answer": ["MGAN: Training Generative Adversarial Nets with Multiple Generators"], "answer_arxiv_id": ["1708.02556"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_2476"} +{"question": "What studies indicate that ImageNet-trained models perform worse on distorted images compared to humans?", "answer": ["A Study and Comparison of Human and Deep Learning Recognition Performance Under Visual Distortions", "On the Limitation of Convolutional Neural Networks in Recognizing Negative Images", "Generalisation in humans and deep neural networks"], "answer_arxiv_id": ["1705.02498", "1703.06857v2", "1808.08750v3"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_2477"} +{"question": "Is there any research that provided an insight into the exponentially weighted integral structure of the solution trajectory?", "answer": ["DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "Fast Sampling of Diffusion Models with Exponential Integrator"], "answer_arxiv_id": ["2206.00927", "2204.13902"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_2478"} +{"question": "Are there any studies that have extensively explored differentiable rasterization in terms of differentiable rendering?", "answer": ["Neural 3D Mesh Renderer", "Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning"], "answer_arxiv_id": ["1711.07566", "1904.01786"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_2479"} +{"question": "What study modified the model by removing LSTM and FC layers and replacing the convolutional layers with depth-wise convolutions?", "answer": ["Efficient Low-Latency Speech Enhancement with Mobile Audio Streaming Networks"], "answer_arxiv_id": ["2008.07244"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_2480"} +{"question": "Which papers have used faster implicit representations based on voxel grids for 3D reconstruction?", "answer": ["Fast and Explicit Neural View Synthesis", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data"], "answer_arxiv_id": ["2107.05775", "2306.07881"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_2481"} +{"question": "What works show that the best model according to traditional metrics does not always result in the best performance on downstream tasks?", "answer": ["Learning to Evaluate Perception Models Using Planner-Centric Metrics", "Control-Aware Prediction Objectives for Autonomous Driving"], "answer_arxiv_id": ["2004.08745", "2204.13319"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_2482"} +{"question": "Which papers discuss controlling the norm of parameters in network initialization?", "answer": ["All you need is a good init", "NeST: A Neural Network Synthesis Tool Based on a Grow-and-Prune Paradigm", "Splitting Steepest Descent for Growing Neural Architectures"], "answer_arxiv_id": ["1511.06422", "1711.02017", "1910.02366"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_2483"} +{"question": "What researches have been performed for simulating clothes on human bodies?", "answer": ["PBNS: Physically Based Neural Simulator for Unsupervised Garment Pose\n Space Deformation", "SNUG: Self-Supervised Neural Dynamic Garments", "HOOD: Hierarchical Graphs for Generalized Modelling of Clothing Dynamics", "Neural Cloth Simulation", "Self-Supervised Collision Handling via Generative 3D Garment Models for\n Virtual Try-On"], "answer_arxiv_id": ["2012.11310", "2204.02219", "2212.07242", "2212.11220", "2105.06462"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_2484"} +{"question": "Which work introduced a weakly supervised method for 3D LiDAR object detection?", "answer": ["Weakly Supervised 3D Object Detection from Lidar Point Cloud"], "answer_arxiv_id": ["2007.11901"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_2485"} +{"question": "What works have proposed solutions for generating diverse variations of given image exemplars while preserving semantic content and visual quality, specifically using neural style transfer?", "answer": ["Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization", "Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "Neural Neighbor Style Transfer"], "answer_arxiv_id": ["1703.06868", "1603.08155", "2203.13215"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2486"} +{"question": "Which work formulated the semantic segmentation as an encoder-decoder architecture for the first time?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_2487"} +{"question": "What papers about the robustness of AUC have been made?", "answer": ["Large-scale Robust Deep AUC Maximization: A New Surrogate Loss and Empirical Studies on Medical Image Classification", "AdAUC: End-to-end Adversarial AUC Optimization Against Long-tail Problems"], "answer_arxiv_id": ["2012.03173", "2206.12169"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_2488"} +{"question": "What studies are about training a CNN to generate multi-plane images and rendering novel views through alpha-compositing?", "answer": ["Stereo Magnification: Learning View Synthesis using Multiplane Images", "Extreme View Synthesis", "DeepView: View Synthesis with Learned Gradient Descent", "Local Light Field Fusion: Practical View Synthesis with Prescriptive\n Sampling Guidelines"], "answer_arxiv_id": ["1805.09817", "1812.04777", "1906.07316", "1905.00889"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_2489"} +{"question": "Which studies utilized differential privacy in the field of natural language processing?", "answer": ["Differential Privacy for Text Analytics via Natural Text Sanitization", "SynTF: Synthetic and Differentially Private Term Frequency Vectors for Privacy-Preserving Text Mining"], "answer_arxiv_id": ["2106.01221v1", "1805.00904"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_2490"} +{"question": "Which research works focused on improving model performance by averaging weights across different checkpoints or different runs?", "answer": ["Averaging Weights Leads to Wider Optima and Better Generalization", "Stochastic Weight Averaging in Parallel: Large-Batch Training That Generalizes Well", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time"], "answer_arxiv_id": ["1803.05407", "2001.02312", "2203.05482"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_2491"} +{"question": "Which research works explored the dynamics adaptation given an offline dataset collected in the target domain?", "answer": ["DARA: Dynamics-Aware Reward Augmentation in Offline Reinforcement Learning", "When to Trust Your Simulator: Dynamics-Aware Hybrid Offline-and-Online Reinforcement Learning", "The Difficulty of Passive Learning in Deep Reinforcement Learning"], "answer_arxiv_id": ["2203.06662", "2206.13464", "2110.14020"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_2492"} +{"question": "Which works have been conducted on region-level image annotation where fine-grained masks are automatically generated?", "answer": ["Kosmos-2: Grounding Multimodal Large Language Models to the World", "Grounded Language-Image Pre-training", "Language Is Not All You Need: Aligning Perception with Language Models", "The All-Seeing Project: Towards Panoptic Visual Recognition and\n Understanding of the Open World", "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2306.14824", "2112.03857", "2302.14045", "2308.01907", "2401.14159"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_2493"} +{"question": "What work concentrates on learning the singular values of the SVD decomposition of weight matrices?", "answer": ["SVDiff: Compact Parameter Space for Diffusion Fine-Tuning"], "answer_arxiv_id": ["2303.11305"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_2494"} +{"question": "What papers introduces a unified form for the temporal understanding capability of language models?", "answer": ["TRAM: Benchmarking Temporal Reasoning for Large Language Models"], "answer_arxiv_id": ["2310.00835"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_2495"} +{"question": "Which works showcase the outstanding adaptability of LLMs on downstream NLP tasks without any parameter tuning?", "answer": ["MetaICL: Learning to Learn In Context", "Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?", "Generative Speech Recognition Error Correction with Large Language Models and Task-Activating Prompting"], "answer_arxiv_id": ["2110.15943", "2202.12837", "2309.15649"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_2496"} +{"question": "What work used the learned similarity and label consistency to identify and discard data with noisy labels?", "answer": ["Iterative Learning with Open-set Noisy Labels"], "answer_arxiv_id": ["1804.00092"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2497"} +{"question": "Could you tell me about a paper that uses a non-Markovian forward noising to speed up the sampling of DMs?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20220429"}, "qid": "AutoScholarQuery_train_2498"} +{"question": "Which papers have been presented on the theoretical model of hand-object interaction generation?", "answer": ["Grasping Field: Learning Implicit Representations for Human Grasps", "A Skeleton-Driven Neural Occupancy Representation for Articulated Hands", "Hand-Object Contact Consistency Reasoning for Human Grasps Generation", "Grasp'D: Differentiable Contact-rich Grasp Synthesis for Multi-fingered\n Hands", "Diffusion-based Generation, Optimization, and Planning in 3D Scenes", "Physics-Based Dexterous Manipulations with Estimated Hand Poses and\n Residual Reinforcement Learning"], "answer_arxiv_id": ["2008.04451", "2109.11399", "2104.03304", "2208.12250", "2301.06015", "2008.03285"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_2499"} +{"question": "Any works showcasing the application of episodic control in multi-agent tasks accompanied by curiosity-based exploration and model-based reinforcement learning?", "answer": ["Episodic Multi-Agent Reinforcement Learning with Curiosity-Driven Exploration", "Model-Based Episodic Memory Induces Dynamic Hybrid Controls"], "answer_arxiv_id": ["2111.11032", "2111.02104"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_2500"} +{"question": "What works are about the more general language-based object detection task?", "answer": ["OmniLabel: A Challenging Benchmark for Language-Based Object Detection", "Described Object Detection: Liberating Object Detection with Flexible\n Expressions"], "answer_arxiv_id": ["2304.11463", "2307.12813"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_2501"} +{"question": "Who extended the investigation of oversquashing by considering the same type of rewiring but using different notions of discrete curvature?", "answer": ["Rewiring Networks for Graph Neural Network Training Using Discrete Geometry"], "answer_arxiv_id": ["2207.08026"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_2502"} +{"question": "Are there any works that has accelerated MAML by adopting first-order approximation on the gradient estimation?", "answer": ["On First-Order Meta-Learning Algorithms"], "answer_arxiv_id": ["1803.02999"], "source_meta": {"published_time": "20230108"}, "qid": "AutoScholarQuery_train_2503"} +{"question": "Which papers utilized prior data to inform behavioral priors or skills?", "answer": ["OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning", "Behavior Priors for Efficient Reinforcement Learning", "Learning and Retrieval from Prior Data for Skill-based Imitation Learning"], "answer_arxiv_id": ["2010.13611", "2010.14274", "2210.11435"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_2504"} +{"question": "Are there any notable works that focus on automating matrix engineering to enhance the LoRA structure?", "answer": ["FacT: Factor-Tuning for Lightweight Adaptation on Vision Transformer"], "answer_arxiv_id": ["2212.03145"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_2505"} +{"question": "Could you provide me with the research studies that introduce Diffusion Models (DMs) as the state of the art for Image Generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2112.10741", "2112.10752", "2105.05233"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_2506"} +{"question": "What studies propose artifact correction approach through latent code manipulation based on a binary linear classifier in GANs?", "answer": ["Interpreting the Latent Space of GANs for Semantic Face Editing"], "answer_arxiv_id": ["1907.10786"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_2507"} +{"question": "What are some recent studies that have used statistical mechanics to analyze and improve learning?", "answer": ["Comparing Dynamics: Deep Neural Networks versus Glassy Systems", "Optimal Errors and Phase Transitions in High-Dimensional Generalized Linear Models", "Beyond neural scaling laws: beating power law scaling via data pruning"], "answer_arxiv_id": ["1803.06969", "1708.03395", "2206.14486v6"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_2508"} +{"question": "Could you provide details of the study where the Algorithm distillation conditions the model on multiple trajectories?", "answer": ["In-context Reinforcement Learning with Algorithm Distillation"], "answer_arxiv_id": ["2210.14215"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2509"} +{"question": "Which papers propose IR-based methods for few-shot KBQA where entities are iteratively accessed until finding an answer?", "answer": ["StructGPT: A General Framework for Large Language Model to Reason over\n Structured Data", "Think-on-Graph: Deep and Responsible Reasoning of Large Language Model\n on Knowledge Graph"], "answer_arxiv_id": ["2305.09645", "2307.07697"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_2510"} +{"question": "Could you provide me studies that introduced and adapted MCL in deep learning settings?", "answer": ["Why M Heads are Better than One: Training a Diverse Ensemble of Deep Networks", "Stochastic Multiple Choice Learning for Training Diverse Deep Ensembles"], "answer_arxiv_id": ["1511.06314", "1606.07839"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_2511"} +{"question": "Which works have leveraged the Lie group structure of the manifold to define a parametrization of the flow?", "answer": ["Normalizing Flows on Tori and Spheres"], "answer_arxiv_id": ["2002.02428"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2512"} +{"question": "What study used convolutions in the patch embedding process and the feed-forward network of ViT?", "answer": ["Incorporating Convolution Designs into Visual Transformers"], "answer_arxiv_id": ["2103.11816"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_2513"} +{"question": "What works show that real datasets have issues with spurious correlations and biases?", "answer": ["Counterfactual VQA: A Cause-Effect Look at Language Bias", "SwapMix: Diagnosing and Regularizing the Over-Reliance on Visual Context in Visual Question Answering", "Counterfactual VQA: A Cause-Effect Look at Language Bias", "Analyzing the Behavior of Visual Question Answering Models", "Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering", "Roses are Red, Violets are Blue… But Should VQA expect Them To?", "How Transferable are Reasoning Patterns in VQA?"], "answer_arxiv_id": ["2006.04315v4", "2204.02285", "2006.04315v4", "1606.07356", "1612.00837", "2006.05121", "2104.03656"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_2514"} +{"question": "Can you name the studies that proposed a group activity recognition network jointly trained with the person action recognition network?", "answer": ["Learning Actor Relation Graphs for Group Activity Recognition", "Convolutional Relational Machine for Group Activity Recognition", "Joint Learning of Social Groups, Individuals Action and Sub-group\n Activities in Videos", "Actor-Transformers for Group Activity Recognition", "Dual-AI: Dual-path Actor Interaction Learning for Group Activity\n Recognition", "Hunting Group Clues with Transformers for Social Group Activity\n Recognition", "COMPOSER: Compositional Reasoning of Group Activity in Videos with\n Keypoint-Only Modality"], "answer_arxiv_id": ["1904.10117", "1904.03308", "2007.02632", "2003.12737", "2204.02148", "2207.05254", "2112.05892"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_2515"} +{"question": "Which studies elaborated on different client and environment settings in the context of federated linear bandits?", "answer": ["Differentially-Private Federated Linear Bandits", "Federated Linear Contextual Bandits", "Asynchronous Upper Confidence Bound Algorithms for Federated Linear Bandits", "A Simple and Provably Efficient Algorithm for Asynchronous Federated Contextual Linear Bandits"], "answer_arxiv_id": ["2010.11425", "2110.14177", "2110.01463", "2207.03106"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_2516"} +{"question": "Which studies analyze modularity within neural networks using learned binary masks?", "answer": ["Are Neural Nets Modular? Inspecting Functional Modularity Through Differentiable Weight Masks"], "answer_arxiv_id": ["2010.02066"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_2517"} +{"question": "Which work proposed the setting of OCDA for handling unlabeled compound target domain and open domain?", "answer": ["Open Compound Domain Adaptation"], "answer_arxiv_id": ["1909.03403"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_2518"} +{"question": "Which papers discuss hybrid models that scale up PCs by integrating them with neural networks (NNs)?", "answer": ["HyperSPNs: Compact and Expressive Probabilistic Circuits"], "answer_arxiv_id": ["2112.00914"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_2519"} +{"question": "What studies involve the COCO Caption dataset which contributed in the development of the Image-Caption Task?", "answer": ["Microsoft COCO Captions: Data Collection and Evaluation Server"], "answer_arxiv_id": ["1504.00325"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_2520"} +{"question": "Could you provide examples of studies that examined customization involving multiple concepts in T2I models?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion", "SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "Key-Locked Rank One Editing for Text-to-Image Personalization"], "answer_arxiv_id": ["2212.04488", "2303.11305", "2305.01644"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_2521"} +{"question": "Which works discuss the practice of pre-training with Transformer for Natural Language Processing tasks?", "answer": ["Attention Is All You Need", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1706.03762", "1810.04805", "2005.14165"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_2522"} +{"question": "Which studies incorporate extracted visual features directly through linear layers into pre-trained models to enhance the perceptual capacities of LLMs?", "answer": ["PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2303.03378"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2523"} +{"question": "Could you provide me some works on post-processing defenses for backdoor attacks?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks", "Data-free Backdoor Removal based on Channel Lipschitzness", "Adversarial Neuron Pruning Purifies Backdoored Deep Models", "Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks", "Adversarial Unlearning of Backdoors via Implicit Hypergradient"], "answer_arxiv_id": ["1805.12185", "2208.03111", "2110.14430", "2101.05930", "2110.03735"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_2524"} +{"question": "Can you name the studies that have simplified deep neural networks as last-layer features and classifiers with proper constraints and proven that ETF emerges under the cross-entropy loss?", "answer": ["Disentangling Trainability and Generalization in Deep Neural Networks", "A Geometric Analysis of Neural Collapse with Unconstrained Features", "An unconstrained layer-peeled perspective on neural collapse", "Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training"], "answer_arxiv_id": ["1912.13053", "2105.02375", "2110.02796", "2101.12699"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_2525"} +{"question": "Which papers introduced statistical methods to quantify the robustness of a BNN?", "answer": ["Statistical Guarantees for the Robustness of Bayesian Neural Networks", "Bayesian Inference with Certifiable Adversarial Robustness"], "answer_arxiv_id": ["1903.01980", "2102.05289v2"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_2526"} +{"question": "Which studies discuss the identifiability of causal structure in the context of available interventional data?", "answer": ["Characterization and Greedy Learning of Interventional Markov Equivalence Classes of Directed Acyclic Graphs", "Characterizing and Learning Equivalence Classes of Causal DAGs under Interventions", "Permutation-Based Causal Structure Learning with Unknown Intervention Targets"], "answer_arxiv_id": ["1104.2808", "1802.06310", "1910.09007"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_2527"} +{"question": "What studies have expanded the scope from object-centric reconstruction to larger indoor environments?", "answer": ["Learning to Explore using Active Neural SLAM", "Uncertainty-driven Planner for Exploration and Navigation", "Active Neural Mapping"], "answer_arxiv_id": ["2004.05155", "2202.11907", "2308.16246"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_2528"} +{"question": "What studies aim to collect commonsense questions and statements that are visually grounded and geographically diverse in vision-and-language tasks?", "answer": ["Broaden the Vision: Geo-Diverse Visual Commonsense Reasoning", "Visually Grounded Reasoning across Languages and Cultures"], "answer_arxiv_id": ["2109.06860", "2109.13238v2"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_2529"} +{"question": "What studies have utilized late fusion architectures, where inputs from each modality are processed independently and fused together in later layers?", "answer": ["Learning Two-Branch Neural Networks for Image-Text Matching Tasks", "Latent Normalizing Flows for Many-to-Many Cross-Domain Mappings"], "answer_arxiv_id": ["1704.03470", "2002.06661"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_2530"} +{"question": "Which studies have used the logit lens to decode intermediate token representations from transformer models?", "answer": ["Eliciting Latent Predictions from Transformers with the Tuned Lens"], "answer_arxiv_id": ["2303.08112"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_2531"} +{"question": "What papers discuss the use of neural implicit representations in SLAM using RGBD images?", "answer": ["GO-SLAM: Global Optimization for Consistent 3D Instant Reconstruction", "Multi-Modal Neural Radiance Field for Monocular Dense SLAM with a Light-Weight ToF Sensor"], "answer_arxiv_id": ["2309.02436", "2308.14383"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_2532"} +{"question": "Which papers propose techniques for personalizing Text-to-Image (T2I) models that result in higher subject fidelity and versatile text based recontextualization of a given subject?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_2533"} +{"question": "Which paper studied the combination of different scale information in graph learning?", "answer": ["Representation Learning on Graphs with Jumping Knowledge Networks"], "answer_arxiv_id": ["1806.03536v2"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_2534"} +{"question": "Which works propose a method for coarse-to-fine curriculum learning?", "answer": ["Coarse-to-Fine Curriculum Learning"], "answer_arxiv_id": ["2106.04072"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_2535"} +{"question": "Could you provide some examples of works where active learning and Bayesian updates were used to recover biological networks?", "answer": ["ABCD-Strategy: Budgeted Experimental Design for Targeted Causal Structure Discovery"], "answer_arxiv_id": ["1902.10347v1"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2536"} +{"question": "Can you mention the works that model dynamics as an image translation problem?", "answer": ["Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting", "Stochastic Video Generation with a Learned Prior", "Stochastic Adversarial Video Prediction"], "answer_arxiv_id": ["1506.04214", "1802.07687", "1804.01523"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_2537"} +{"question": "What works proposed sampling goals from the least visited areas using a parametric density model on the visited states?", "answer": ["Skew-Fit: State-Covering Self-Supervised Reinforcement Learning"], "answer_arxiv_id": ["1903.03698"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_2538"} +{"question": "What papers introduce PointNet and PointNet++ for reducing high-dimensional unordered points into fixed-length feature vectors?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space"], "answer_arxiv_id": ["1612.00593", "1706.02413"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_2539"} +{"question": "Which work proposed a method that tightens a subset of variable bounds by using an off-the-shelf branch and bound solver?", "answer": ["Solving Mixed Integer Programs Using Neural Networks"], "answer_arxiv_id": ["2012.13349"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_2540"} +{"question": "Could you provide me some works about clustering high-dimensional data using spectral clustering methods?", "answer": ["A Tighter Analysis of Spectral Clustering, and Beyond", "A Tutorial on Spectral Clustering", "Spectral Clustering with Graph Neural Networks for Graph Pooling", "A Tighter Analysis of Spectral Clustering, and Beyond"], "answer_arxiv_id": ["2208.01724v1", "0711.0189", "1907.00481", "2208.01724v1"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_2541"} +{"question": "What works explore the method of learning to interpolate between available views in the 3D space to estimate the scene geometry?", "answer": ["MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse\n Views"], "answer_arxiv_id": ["2103.15595", "2206.05737"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_2542"} +{"question": "Are there any research papers that have used both behavioral testing and interventions in their counterfactual benchmarks?", "answer": ["Rigorously Assessing Natural Language Explanations of Neurons"], "answer_arxiv_id": ["2309.10312"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2543"} +{"question": "Could you provide me with some works that explored the existence of competing subnetworks as a part of explaining the Grokking phenomenon?", "answer": ["A Tale of Two Circuits: Grokking as Competition of Sparse and Dense\n Subnetworks", "Explaining grokking through circuit efficiency"], "answer_arxiv_id": ["2303.11873", "2309.02390"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_2544"} +{"question": "What work propose a learning-based schema to solve localization with only geometry information?", "answer": ["Is Geometry Enough for Matching in Visual Localization?"], "answer_arxiv_id": ["2203.12979"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2545"} +{"question": "Which works have driven the advent of FL algorithms through variance reduction techniques?", "answer": ["SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "From Deterioration to Acceleration: A Calibration Approach to Rehabilitating Step Asynchronism in Federated Optimization", "Variance Reduced Local SGD with Lower Communication Complexity", "Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning", "Bias-Variance Reduced Local SGD for Less Heterogeneous Federated Learning", "Linear Convergence in Federated Learning: Tackling Client Heterogeneity and Sparse Gradients"], "answer_arxiv_id": ["1910.06378", "2112.09355", "1912.12844", "2008.03606", "2102.03198", "2102.07053"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_2546"} +{"question": "Which papers focused on using Reinforcement Learning in solving combinatorial optimization problems?", "answer": ["Learning Combinatorial Optimization Algorithms over Graphs", "Reinforcement Learning for Solving the Vehicle Routing Problem", "Learning to Perform Local Rewriting for Combinatorial Optimization", "GLSearch: Maximum Common Subgraph Detection via Learning to Search"], "answer_arxiv_id": ["1704.01665", "1802.04240", "1810.00337", "2002.03129"], "source_meta": {"published_time": "20201216"}, "qid": "AutoScholarQuery_train_2547"} +{"question": "Which study gave a rate for T-steps Adaptive SGD, assuming uniformly bounded gradients?", "answer": ["AdaGrad stepsizes: Sharp convergence over nonconvex landscapes"], "answer_arxiv_id": ["1806.01811"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_2548"} +{"question": "What is the study proposed the chain-of-thought prompting approach?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_2549"} +{"question": "What papers suggested to use lower order linear multi-step method for warming start?", "answer": ["Fast Sampling of Diffusion Models with Exponential Integrator"], "answer_arxiv_id": ["2204.13902"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2550"} +{"question": "What research provided a framework for computing valid p-values for DNN-based image segmentation results?", "answer": ["Quantifying Statistical Significance of Neural Network-based Image Segmentation by Selective Inference"], "answer_arxiv_id": ["2010.01823"], "source_meta": {"published_time": "20230106"}, "qid": "AutoScholarQuery_train_2551"} +{"question": "Could you tell me about the researches related to attention-based editing works in the field of image generation?", "answer": ["Null-text Inversion for Editing Real Images using Guided Diffusion Models", "Zero-shot Image-to-Image Translation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Prompt-to-Prompt Image Editing with Cross Attention Control"], "answer_arxiv_id": ["2211.09794", "2302.03027", "2303.09535", "2208.01626"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2552"} +{"question": "Which works studied improving the performance of language models with fewer labeled instances in the context of Active Learning?", "answer": ["A Survey of Active Learning for Natural Language Processing", "Towards Computationally Feasible Deep Active Learning", "Active Learning by Acquiring Contrastive Examples"], "answer_arxiv_id": ["2210.10109", "2205.03598", "2109.03764"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_2553"} +{"question": "What works proposed a progressive approach to build and train large-scale scene models?", "answer": ["BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale Scene Rendering"], "answer_arxiv_id": ["2112.05504"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_2554"} +{"question": "Which works leverage a diffusion model and introduce modules to refine both visual–textual and textual–textual presentations?", "answer": ["Relation-Aware Diffusion Model for Controllable Poster Layout Generation"], "answer_arxiv_id": ["2306.09086"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_2555"} +{"question": "What works employ self-supervised learning to obtain state representations?", "answer": ["The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "R3M: A Universal Visual Representation for Robot Manipulation", "Masked Visual Pre-training for Motor Control"], "answer_arxiv_id": ["2203.03580", "2203.12601", "2203.06173"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_2556"} +{"question": "Could you provide me some works that investigated the failure of ERM models in high-consequence applications?", "answer": ["Deep Learning Face Attributes in the Wild", "Fair prediction with disparate impact: A study of bias in recidivism prediction instruments", "Hidden Stratification Causes Clinically Meaningful Failures in Machine Learning for Medical Imaging"], "answer_arxiv_id": ["1411.7766v3", "1610.07524", "1909.12475"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_2557"} +{"question": "Which studies discuss the use of deep neural networks in accelerating dynamics forecasting?", "answer": ["PDE-Net: Learning PDEs from Data", "Fourier Neural Operator for Parametric Partial Differential Equations", "Neural Solvers for Fast and Accurate Numerical Optimal Control", "FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators", "Meta-SysId: A Meta-Learning Approach for Simultaneous Identification and Prediction", "Transform Once: Efficient Operator Learning in Frequency Domain", "MAgNet: Mesh Agnostic Neural PDE Solver"], "answer_arxiv_id": ["1710.09668", "2010.08895", "2203.08072", "2202.11214", "2206.00694", "2211.14453v1", "2210.05495"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_2558"} +{"question": "What method is cited for clustering data by enforcing consistency between the assigned clusters of augmented views?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments"], "answer_arxiv_id": ["2006.09882"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_2559"} +{"question": "Which earlier works are focused on designing various attention mechanisms for semantic segmentation?", "answer": ["Non-local Neural Networks", "CCNet: Criss-Cross Attention for Semantic Segmentation"], "answer_arxiv_id": ["1711.07971", "1811.11721"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_2560"} +{"question": "Which works proposed the creation of random features for the neural tangent kernel?", "answer": ["Scaling Neural Tangent Kernels via Sketching and Random Features"], "answer_arxiv_id": ["2106.07880"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_2561"} +{"question": "Which paper demonstrated the effectiveness of few-shot prompting in a variety of tasks?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2107.13586"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_2562"} +{"question": "Could you provide me some works about encouraging generalization by conditioning policies on low dimensional feature representations shared across different tasks?", "answer": ["Visual Reinforcement Learning with Imagined Goals", "Skew-Fit: State-Covering Self-Supervised Reinforcement Learning", "Contextual Imagined Goals for Self-Supervised Robotic Learning", "The Variational Bandwidth Bottleneck: Stochastic Evaluation on an Information Budget", "Discovering and Achieving Goals via World Models"], "answer_arxiv_id": ["1807.04742", "1903.03698", "1910.11670", "2004.11935", "2110.09514"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_2563"} +{"question": "Which work provided the first large-scale, real-world 3D laneline dataset?", "answer": ["PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark"], "answer_arxiv_id": ["2203.11089"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_2564"} +{"question": "Can you show me some studies where the learning of image embeddings is done in a supervised manner on crowd-labeled datasets?", "answer": ["Deep Residual Learning for Image Recognition", "Densely Connected Convolutional Networks", "Squeeze-and-Excitation Networks", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["1512.03385", "1608.06993", "1709.01507", "2010.11929"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_2565"} +{"question": "Which papers examine the two-timescale/asymmetric gradient descent/ascent methods under the PŁcondition?", "answer": ["Solving a Class of Non-Convex Min-Max Games Using Iterative First Order Methods", "Global Convergence and Variance-Reduced Optimization for a Class of Nonconvex-Nonconcave Minimax Problems"], "answer_arxiv_id": ["1902.08297", "2002.09621"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_2566"} +{"question": "Can you list some studies that tackle the problem of Domain Adaptive Segmentation (DAS)?", "answer": ["Learning to Adapt Structured Output Space for Semantic Segmentation", "ADVENT: Adversarial Entropy Minimization for Domain Adaptation in\n Semantic Segmentation", "Domain Adaptation for Semantic Segmentation with Maximum Squares Loss", "FCNs in the Wild: Pixel-level Adversarial and Constraint-based\n Adaptation", "Significance-aware Information Bottleneck for Domain Adaptive Semantic\n Segmentation", "Domain Adaptive Video Segmentation via Temporal Pseudo Supervision", "CyCADA: Cycle-Consistent Adversarial Domain Adaptation"], "answer_arxiv_id": ["1802.10349", "1811.12833", "1909.13589", "1612.02649", "1904.00876", "2207.02372", "1711.03213"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_2567"} +{"question": "Could you provide me with the reference that generalized the non-convex SGD approach?", "answer": ["Non-Convex SGD Learns Halfspaces with Adversarial Label Noise"], "answer_arxiv_id": ["2006.06742"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_2568"} +{"question": "Which studies proposed the two named models, VL-BART and CLIP-BART?", "answer": ["Unifying Vision-and-Language Tasks via Text Generation", "VL-Adapter: Parameter-Efficient Transfer Learning for Vision-and-Language Tasks"], "answer_arxiv_id": ["2102.02779", "2112.06825"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_2569"} +{"question": "What research used PAC-Bayes to study the generalization properties of stochastic reconstruction models?", "answer": ["On PAC-Bayesian reconstruction guarantees for VAEs"], "answer_arxiv_id": ["2202.11455v1"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_2570"} +{"question": "Could you provide me with some literature about learning to create new examples that preserve the properties of original graphs?", "answer": ["Local Augmentation for Graph Neural Networks", "Graph Rationalization with Environment-based Augmentations", "Robust Optimization as Data Augmentation for Large-scale Graphs", "G-Mixup: Graph Data Augmentation for Graph Classification", "Automated Data Augmentations for Graph Classification"], "answer_arxiv_id": ["2109.03856", "2206.02886", "2010.09891v3", "2202.07179v2", "2202.13248"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_2571"} +{"question": "What studies use integrated gradient (IG) attribution for neural prediction interpretation?", "answer": ["Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1703.01365"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_2572"} +{"question": "Can you provide some research papers making use of masked autoencoding?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "VideoMAE: Masked Autoencoders are Data-Efficient Learners for\n Self-Supervised Video Pre-Training", "VIOLET : End-to-End Video-Language Transformers with Masked Visual-token\n Modeling", "An Empirical Study of End-to-End Video-Language Transformers with Masked\n Visual Modeling", "SMAUG: Sparse Masked Autoencoder for Efficient Video-Language\n Pre-training", "SimVTP: Simple Video Text Pre-training with Masked Autoencoders"], "answer_arxiv_id": ["2111.06377", "2203.12602", "2111.12681", "2209.01540", "2211.11446", "2212.03490"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2573"} +{"question": "Can you name some popular and effective methods for generating adversarial examples?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Evaluating the Robustness of Neural Networks", "Robustness via curvature regularization, and vice versa", "Boosting Adversarial Attacks with Momentum"], "answer_arxiv_id": ["1412.6572", "1608.04644", "1811.09716", "1710.06081"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_2574"} +{"question": "Which papers discussed the usage of manually designed cloze-style prompts in probing knowledge from pre-trained language models?", "answer": ["Knowledge Neurons in Pretrained Transformers"], "answer_arxiv_id": ["2104.08696"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_2575"} +{"question": "Which paper developed DRRN for learning agents within text environment?", "answer": ["Deep Reinforcement Learning with a Natural Language Action Space"], "answer_arxiv_id": ["1511.04636"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_2576"} +{"question": "What are some examples of works that utilized online RL methods combined with language models for natural language tasks?", "answer": ["Sequence Level Training with Recurrent Neural Networks", "Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation", "A Deep Reinforced Model for Abstractive Summarization", "Learning to Extract Coherent Summary via Deep Reinforcement Learning"], "answer_arxiv_id": ["1511.06732", "1609.08144", "1705.04304", "1804.07036"], "source_meta": {"published_time": "20220605"}, "qid": "AutoScholarQuery_train_2577"} +{"question": "Could you provide me some studies about sentence-level grounding with complex semantics?", "answer": ["ScanRefer: 3D Object Localization in RGB-D Scans using Natural Language", "Scan2Cap: Context-aware Dense Captioning in RGB-D Scans"], "answer_arxiv_id": ["1912.08830", "2012.02206"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_2578"} +{"question": "Which studies have performed shape matching by training a network for segmenting each vertex on the source shape?", "answer": ["Geometric deep learning on graphs and manifolds using mixture model CNNs", "Learning shape correspondence with anisotropic convolutional neural networks", "Geodesic convolutional neural networks on Riemannian manifolds", "Multi-directional Geodesic Neural Networks via Equivariant Convolution"], "answer_arxiv_id": ["1611.08402v3", "1605.06437", "1501.06297", "1810.02303"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_2579"} +{"question": "Who provided the solution of the equivariance constraint given by spherical harmonics modulated by an arbitrary continuous radial function?", "answer": ["3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data"], "answer_arxiv_id": ["1807.02547"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_2580"} +{"question": "Which research papers utilized NeRF for mimicking the rendering equation?", "answer": ["Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance\n Fields"], "answer_arxiv_id": ["2112.03907"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_2581"} +{"question": "What references discuss the challenges of image editing with generative models?", "answer": ["Learning Diverse Image Colorization", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks", "Cut-and-Paste Object Insertion by Enabling Deep Image Prior for Reshading", "Deep Image Prior", "Semantic Image Synthesis with Spatially-Adaptive Normalization"], "answer_arxiv_id": ["1612.01958", "1703.10593", "2010.05907", "1711.10925", "1903.07291"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_2582"} +{"question": "Are there any works about 2D image inpainting and outpainting?", "answer": ["Generative Image Inpainting with Contextual Attention", "In-N-Out: Towards Good Initialization for Inpainting and Outpainting", "In&Out : Diverse Image Outpainting via GAN Inversion"], "answer_arxiv_id": ["1801.07892", "2106.13953", "2104.00675"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_2583"} +{"question": "Could you provide me some studies that approach zero-shot DST by leveraging extra dialogue corpora in similar domains?", "answer": ["Improving Limited Labeled Dialogue State Tracking with Self-Supervision", "Zero-Shot Dialogue State Tracking via Cross-Task Transfer", "Multi-Task Pre-Training for Plug-and-Play Task-Oriented Dialogue System"], "answer_arxiv_id": ["2010.13920", "2109.04655", "2109.14739"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_2584"} +{"question": "What paper discusses the impracticality of reformulating the regret through the expectation of the geometric mean?", "answer": ["Fairness and Welfare Quantification for Regret in Multi-Armed Bandits"], "answer_arxiv_id": ["2205.13930"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_2585"} +{"question": "What studies focus on extracting the cross-domain knowledge for unlabeled target data from single source models?", "answer": ["Learning Part Segmentation through Unsupervised Domain Adaptation from\n Synthetic Vehicles"], "answer_arxiv_id": ["2103.14098"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_2586"} +{"question": "Could you provide me some research where autoregressive models have been used to enhance motion generation quality?", "answer": ["AttT2M: Text-Driven Human Motion Generation with Multi-Perspective\n Attention Mechanism", "MotionGPT: Human Motion as a Foreign Language"], "answer_arxiv_id": ["2309.00796", "2306.14795"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_2587"} +{"question": "Which papers explore structure learning over latent variables in the presence of interventional data?", "answer": ["CITRIS: Causal Identifiability from Temporal Intervened Sequences", "Contrastive Learning Inverts the Data Generating Process", "Weakly Supervised Representation Learning with Sparse Perturbations", "Score-based Causal Representation Learning with Interventions", "Interventional Causal Representation Learning", "Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2202.03169", "2102.08850v4", "2206.01101", "2301.08230", "2209.11924", "2202.06856"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_2588"} +{"question": "Any studies about the Network Dissection (NetDissect) method within global explainability methods in Explainable AI?", "answer": ["Network Dissection: Quantifying Interpretability of Deep Visual Representations", "GAN Dissection: Visualizing and Understanding Generative Adversarial Networks"], "answer_arxiv_id": ["1704.05796", "1811.10597"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_2589"} +{"question": "Could you give me some research studies that introduced advanced image-text alignments method using cross-attention layers?", "answer": ["Vision-Language Pre-Training with Triple Contrastive Learning", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models"], "answer_arxiv_id": ["2202.10401", "2201.12086", "2301.12597"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2590"} +{"question": "Which works discuss the need for high-quality datasets in the process of fine-tuning, specifically instruction tuning, in language models?", "answer": ["LIMA: Less Is More for Alignment"], "answer_arxiv_id": ["2305.11206"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_2591"} +{"question": "Are there any studies that use the prediction error as the intrinsic reward for exploration in curiosity-driven exploration methodology?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models"], "answer_arxiv_id": ["1507.00814"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_2592"} +{"question": "Which studies focused on the theoretical properties of modern graph-based nearest neighbor data structures?", "answer": ["Graph-based time–space trade-offs for approximate near neighbors", "Graph-based Nearest Neighbor Search: From Practice to Theory"], "answer_arxiv_id": ["1712.03158", "1907.00845"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_2593"} +{"question": "What research effort extends the idea from bib.bib30 to analyze vanilla Stochastic Gradient Descent?", "answer": ["Information-Theoretic Generalization Bounds for Stochastic Gradient Descent"], "answer_arxiv_id": ["2102.00931"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_2594"} +{"question": "Which work introduces a human evaluation framework and instruction prompt tuning for the use of large language models (LLMs) in clinical research?", "answer": ["Large Language Models Encode Clinical Knowledge"], "answer_arxiv_id": ["2212.13138"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_2595"} +{"question": "What are the studies that played an indispensable role in most dialogue tasks?", "answer": ["DailyDialog: A Manually Labelled Multi-turn Dialogue Dataset", "Filling the Gap of Utterance-aware and Speaker-aware Representation for Multi-turn Dialogue"], "answer_arxiv_id": ["1710.03957", "2009.06504"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_2596"} +{"question": "Could you provide me some works about improving uncertainty estimation in the context of self-training?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Single-Model Uncertainties for Deep Learning", "Deep Evidential Regression"], "answer_arxiv_id": ["1506.02142", "1811.00908", "1910.02600"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_2597"} +{"question": "Which works improved modeling the interactions between person features for group activity recognition using Transformer?", "answer": ["Actor-Transformers for Group Activity Recognition", "Dual-AI: Dual-path Actor Interaction Learning for Group Activity\n Recognition", "Hunting Group Clues with Transformers for Social Group Activity\n Recognition", "COMPOSER: Compositional Reasoning of Group Activity in Videos with\n Keypoint-Only Modality"], "answer_arxiv_id": ["2003.12737", "2204.02148", "2207.05254", "2112.05892"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_2598"} +{"question": "Which works are related to Dynamic Neural Radiance Field?", "answer": ["Neural 3D Video Synthesis from Multi-view Video", "Space-time Neural Irradiance Fields for Free-Viewpoint Video", "DynIBaR: Neural Dynamic Image-Based Rendering", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View\n Synthesis of a Dynamic Scene From Monocular Video", "Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "Flow supervision for Deformable NeRF", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Dynamic View Synthesis from Dynamic Monocular Video", "Neural Trajectory Fields for Dynamic Novel View Synthesis", "Robust Dynamic Radiance Fields", "Fast Dynamic Radiance Fields with Time-Aware Neural Voxels", "HexPlane: A Fast Representation for Dynamic Scenes", "HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling"], "answer_arxiv_id": ["2103.02597", "2011.12950", "2211.11082", "2011.13961", "2012.12247", "2011.12948", "2106.13228v2", "2303.16333", "2011.13084", "2105.06468", "2105.05994", "2301.02239", "2205.15285", "2301.09632", "2301.02238"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_2599"} +{"question": "Which works utilized element-wise multiplication for feature aggregation in network design?", "answer": ["Focal Modulation Networks", "Visual Attention Network", "Conv2Former: A Simple Transformer-Style ConvNet for Visual Recognition", "HorNet: Efficient High-Order Spatial Interactions with Recursive Gated\n Convolutions", "Scale-Aware Modulation Meet Transformer", "MogaNet: Multi-order Gated Aggregation Network", "Video-FocalNets: Spatio-Temporal Focal Modulation for Video Action\n Recognition", "SPANet: Frequency-balancing Token Mixer using Spectral Pooling\n Aggregation Modulation"], "answer_arxiv_id": ["2203.11926", "2202.09741", "2211.11943", "2207.14284", "2307.08579", "2211.03295", "2307.06947", "2308.11568"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_2600"} +{"question": "Which papers focus on learning manipulation affordance for articulated objects?", "answer": ["Synergies Between Affordance and Geometry: 6-DoF Grasp Detection via Implicit Representations", "Learning Dexterous Grasping with Object-Centric Visual Affordances", "3D AffordanceNet: A Benchmark for Visual Object Affordance Understanding", "Where2Act: From Pixels to Actions for Articulated 3D Objects", "VAT-Mart: Learning Visual Action Trajectory Proposals for Manipulating 3D ARTiculated Objects", "Learning Agent-Aware Affordances for Closed-Loop Interaction with Articulated Objects.", "End-to-End Affordance Learning for Robotic Manipulation", "Grounding 3D Object Affordance from 2D Interactions in Images", "PartAfford: Part-level Affordance Discovery from 3D Objects", "STRAP: Structured Object Affordance Segmentation with Point Supervision"], "answer_arxiv_id": ["2104.01542", "2009.01439", "2103.16397", "2101.02692", "2106.14440", "2209.05802", "2209.12941", "2303.10437", "2202.13519", "2304.08492v1"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_2601"} +{"question": "Which studies considered learning invariant embeddings for noise-invariance in speaker recognition?", "answer": ["Within-sample Variability-invariant loss for robust speaker recognition under noisy environments", "Robust speaker recognition using unsupervised adversarial invariance"], "answer_arxiv_id": ["2002.00924", "1911.00940"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_2602"} +{"question": "List some research which are related to dynamic assortment optimization?", "answer": ["Dynamic Assortment Optimization with Changing Contextual Information"], "answer_arxiv_id": ["1810.13069"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_2603"} +{"question": "What papers represent recent developments in extending the capabilities of T2I synthesis models?", "answer": ["Prompt-Free Diffusion: Taking \"Text\" out of Text-to-Image Diffusion\n Models", "Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion Models", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models"], "answer_arxiv_id": ["2305.16223", "2305.16322", "2308.06721"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2604"} +{"question": "Which papers discussed Extreme multi-label learning (XML) approaches for vulnerable library identification?", "answer": ["CHRONOS: Time-Aware Zero-Shot Identification of Libraries from\n Vulnerability Reports"], "answer_arxiv_id": ["2301.03944"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_2605"} +{"question": "What researches proposed to discover meaningful directions of diffusion models in the bottleneck of denoising networks?", "answer": ["Diffusion Models already have a Semantic Latent Space", "Unsupervised Discovery of Semantic Latent Directions in Diffusion Models", "DisDiff: Unsupervised Disentanglement of Diffusion Probabilistic Models", "Latent Space Editing in Transformer-Based Flow Matching"], "answer_arxiv_id": ["2210.10960v2", "2302.12469", "2301.13721", "2312.10825"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_2606"} +{"question": "Which paper set the formal benchmarks for the OOD detection problem and proposed to use softmax prediction probability as a baseline method?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks"], "answer_arxiv_id": ["1610.02136"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_2607"} +{"question": "What works handle the problem of negative transfer in multi-task learning?", "answer": ["UberNet: Training a `Universal' Convolutional Neural Network for Low-,\n Mid-, and High-Level Vision using Diverse Datasets and Limited Memory"], "answer_arxiv_id": ["1609.02132"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_2608"} +{"question": "Could you tell me about the works that let expand SE​(3) equivariance to part level?", "answer": ["Rotationally Equivariant 3D Object Detection", "EFEM: Equivariant Neural Field Expectation Maximization for 3D Object Segmentation Without Scene Supervision", "Self-Supervised Category-Level Articulated Object Pose Estimation with Part-Level SE(3) Equivariance"], "answer_arxiv_id": ["2204.13630", "2303.15440", "2302.14268"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_2609"} +{"question": "Who first proposed the diffusion model to fit complex distributions?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_2610"} +{"question": "Which studies discuss the use of DeepMind Lab in the context of 3D simulation libraries?", "answer": ["DeepMind Lab"], "answer_arxiv_id": ["1612.03801"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_2611"} +{"question": "Could you provide the work that discussed achieving online distillation in a single pass by using Consistency models?", "answer": ["Consistency Models"], "answer_arxiv_id": ["2303.01469"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2612"} +{"question": "Could you mention some studies that use pre-trained image diffusion models to achieve video editing?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "ControlVideo: Training-free Controllable Text-to-Video Generation", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "Pix2Video: Video Editing using Image Diffusion"], "answer_arxiv_id": ["2212.11565", "2306.07954", "2303.09535", "2303.13439", "2305.13077", "2306.02018", "2303.12688"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_2613"} +{"question": "Which works focus on traditional tasks like sentiment detection or classification in the context of financial NLP benchmark datasets?", "answer": ["Good Debt or Bad Debt: Detecting Semantic Orientations in Economic Texts", "Impact of News on the Commodity Market: Dataset and Results"], "answer_arxiv_id": ["1307.5336", "2009.04202"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_2614"} +{"question": "Could you provide me some studies about image captioning?", "answer": ["VinVL: Revisiting Visual Representations in Vision-Language Models", "Scaling Up Vision-Language Pre-training for Image Captioning", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2101.00529", "2111.12233", "2201.12086"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_2615"} +{"question": "Any works on predicting local semantic inter-object relationships and building a graph of objects?", "answer": ["Learning 3D Semantic Scene Graphs from 3D Indoor Reconstructions", "Exploiting Edge-Oriented Reasoning for 3D Point-based Scene Graph\n Analysis", "SceneGraphFusion: Incremental 3D Scene Graph Prediction from RGB-D\n Sequences", "Incremental 3D Semantic Scene Graph Prediction from RGB Sequences", "VL-SAT: Visual-Linguistic Semantics Assisted Training for 3D Semantic\n Scene Graph Prediction in Point Cloud", "Lang3DSG: Language-based contrastive pre-training for 3D Scene Graph\n prediction"], "answer_arxiv_id": ["2004.03967", "2103.05558", "2103.14898", "2305.02743", "2303.14408", "2310.16494"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_2616"} +{"question": "Which paper is considered as the most well-known work of using Language-based Learning Models (LLMs) for reasoning?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_2617"} +{"question": "Could you provide some research papers that propose improvements to Neural Processes (NPs) for meta learning?", "answer": ["Conditional Neural Processes", "Neural Processes", "Attentive Neural Processes", "Convolutional Conditional Neural Processes", "Meta-Learning Stationary Stochastic Process Prediction with Convolutional Neural Processes"], "answer_arxiv_id": ["1807.01613", "1807.01622", "1901.05761", "1910.13556", "2007.01332"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_2618"} +{"question": "Which studies explored the use of side information in the context of online learning?", "answer": ["Bandits with Side Observations: Bounded vs. Logarithmic Regret", "Leveraging Initial Hints for Free in Stochastic Linear Bandits"], "answer_arxiv_id": ["1807.03558", "2203.04274v1"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_2619"} +{"question": "What are some works about certified defenses that provide robustness guarantees?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing", "Certified Robustness to Adversarial Examples with Differential Privacy", "Denoised Smoothing: A Provable Defense for Pretrained Classifiers", "Boosting Randomized Smoothing with Variance Reduced Classifiers", "Efficient Neural Network Robustness Certification with General Activation Functions", "Certified Defenses against Adversarial Examples", "Semidefinite relaxations for certifying robustness to adversarial examples", "A Convex Relaxation Barrier to Tight Robustness Verification of Neural Networks", "Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Neural Network Robustness Verification"], "answer_arxiv_id": ["1902.02918", "1802.03471", "2003.01908", "2106.06946", "1811.00866", "1801.09344", "1811.01057", "1902.08722", "2103.06624"], "source_meta": {"published_time": "20221101"}, "qid": "AutoScholarQuery_train_2620"} +{"question": "What studies discuss the influence of LLMs on text and graph mining in social networks?", "answer": ["User Modeling in the Era of Large Language Models: Current Research and Future Directions", "Large Language Models on Graphs: A Comprehensive Survey"], "answer_arxiv_id": ["2312.11518v2", "2312.02783"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_2621"} +{"question": "Which works propose using Backward Compatible Training (BCT) for the alignment of models?", "answer": ["Towards Backward-Compatible Representation Learning"], "answer_arxiv_id": ["2003.11942"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_2622"} +{"question": "What is the study that focused on examining gender bias with respect to occupations?", "answer": ["Gender Bias in Coreference Resolution"], "answer_arxiv_id": ["1804.09301"], "source_meta": {"published_time": "20230913"}, "qid": "AutoScholarQuery_train_2623"} +{"question": "Which publications remark on the attention mechanism's ability to model long-range interactions?", "answer": ["Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention", "Rethinking Attention with Performers", "FMMformer: Efficient and Flexible Transformer via Decomposed Near-field and Far-field Attention"], "answer_arxiv_id": ["2006.16236", "2009.14794", "2108.02347"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_2624"} +{"question": "What research papers have explored the learning bias of the MLP networks to low-frequency signals in Implicit Neural Representation?", "answer": ["On the Spectral Bias of Neural Networks", "Frequency Principle: Fourier Analysis Sheds Light on Deep Neural Networks"], "answer_arxiv_id": ["1806.08734", "1901.06523"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_2625"} +{"question": "Which research introduced a (1+ε)-approximation to the BMF problem with rank-k factors, yet their algorithm uses doubly exponential time?", "answer": ["Approximation Schemes for Low-Rank Binary Matrix Approximation Problems"], "answer_arxiv_id": ["1807.07156v1"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_2626"} +{"question": "Which papers proposed a multi-sampling strategy in the conical frustum instead of the camera ray to counter aliasing effect on neural representations?", "answer": ["Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2304.06706"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_2627"} +{"question": "What studies have shown how language can be beneficial in educational settings?", "answer": ["Show or Tell? Demonstration is More Robust to Changes in Shared Perception than Explanation"], "answer_arxiv_id": ["2012.09035"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_2628"} +{"question": "What is the research that provided evidence suggesting infilling capabilities can impact the quality of generation in code generation models?", "answer": ["Code Llama: Open Foundation Models for Code"], "answer_arxiv_id": ["2308.12950"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_2629"} +{"question": "Which research uses a plugged semantic group module for open-vocabulary semantic segmentation, learning the segmentation masks from text supervision?", "answer": ["GroupViT: Semantic Segmentation Emerges from Text Supervision"], "answer_arxiv_id": ["2202.11094"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_2630"} +{"question": "What works propose to append a smaller neural network to the original model and accelerate adaptation by only training the attachment?", "answer": ["LST: Ladder Side-Tuning for Parameter and Memory Efficient Transfer Learning"], "answer_arxiv_id": ["2206.06522"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_2631"} +{"question": "What is the paper that introduced the Denoising Diffusion Probabilistic Model (DDPM)?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_2632"} +{"question": "Which research demonstrates that labeled samples can bolster the convergence speed of Expectation-Maximization within the context of the study?", "answer": ["On the Semi-supervised Expectation Maximization"], "answer_arxiv_id": ["2211.00537"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2633"} +{"question": "Which works have re-implemented the CUDA programming due to it not supporting the double propagation deduced by the eikonal regularization?", "answer": ["PermutoSDF: Fast Multi-View Reconstruction with Implicit Surfaces using Permutohedral Lattices", "NeuS2: Fast Learning of Neural Implicit Surfaces for Multi-view\n Reconstruction"], "answer_arxiv_id": ["2211.12562v2", "2212.05231"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_2634"} +{"question": "Which works contributed in designing rehearsal-based methods for continual learning?", "answer": ["Continual Classification Learning Using Generative Models", "iCaRL: Incremental Classifier and Representation Learning", "Selective Experience Replay for Lifelong Learning", "Continual Detection Transformer for Incremental Object Detection", "Experience Replay for Continual Learning", "Continual Learning with Deep Generative Replay"], "answer_arxiv_id": ["1810.10612", "1611.07725", "1802.10269", "2304.03110", "1811.11682", "1705.08690"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_2635"} +{"question": "Can you name some studies that use the idea of identifying and training only a subset of all model parameters for parameter-efficient fine-tuning?", "answer": ["Masking as an Efficient Alternative to Finetuning for Pretrained Language Models", "Parameter-Efficient Transfer Learning with Diff Pruning", "Piggyback: Adapting a Single Network to Multiple Tasks by Learning to Mask Weights", "Training Neural Networks with Fixed Sparse Masks", "BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-models"], "answer_arxiv_id": ["2004.12406", "2012.07463", "1801.06519", "2111.09839", "2106.10199"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_2636"} +{"question": "Can you name some papers that propose the concept of using multi-hop MPNNs as a form of local graph rewiring?", "answer": ["Shortest Path Networks for Graph Property Prediction", "MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing", "N-GCN: Multi-scale Graph Convolution for Semi-supervised Node Classification", "k-hop graph neural networks", "Nested Graph Neural Networks"], "answer_arxiv_id": ["2206.01003", "1905.00067", "1802.08888", "1907.06051", "2110.13197"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_2637"} +{"question": "Which research proposed SH and HB as effective randomized policies for multi-fidelity HPO that use early stopping of configurations?", "answer": ["Non-stochastic Best Arm Identification and Hyperparameter Optimization", "Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization", "BOHB: Robust and Efficient Hyperparameter Optimization at Scale"], "answer_arxiv_id": ["1502.07943", "1603.06560", "1807.01774"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_2638"} +{"question": "Are there any works which used graph convolutional neural networks and multi-head attention for multi-agent environments?", "answer": ["Graph Convolutional Reinforcement Learning"], "answer_arxiv_id": ["1810.09202"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_2639"} +{"question": "What papers have been published in the area of open-ended learning which aims to continually discover and approach objectives?", "answer": ["Open-Ended Learning Leads to Generally Capable Agents", "Open-Ended Reinforcement Learning with Neural Reward Functions"], "answer_arxiv_id": ["2107.12808", "2202.08266"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_2640"} +{"question": "What works exist about reusing raw data and retrieval-based methods in transfer learning?", "answer": ["Improving language models by retrieving from trillions of tokens", "WebGPT: Browser-assisted question-answering with human feedback", "Latent Retrieval for Weakly Supervised Open Domain Question Answering", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "ERNIE 3.0: Large-scale Knowledge Enhanced Pre-training for Language Understanding and Generation", "Retrieval-Augmented Reinforcement Learning"], "answer_arxiv_id": ["2112.04426", "2112.09332", "1906.00300", "2005.11401", "2107.02137", "2202.08417"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_2641"} +{"question": "Any works about employing pre-trained models for local or global alignments?", "answer": ["Text-Based Person Search with Limited Data", "CLIP-Driven Fine-grained Text-Image Person Re-identification", "Cross-Modal Implicit Relation Reasoning and Aligning for Text-to-Image\n Person Retrieval"], "answer_arxiv_id": ["2110.10807", "2210.10276", "2303.12501"], "source_meta": {"published_time": "20230819"}, "qid": "AutoScholarQuery_train_2642"} +{"question": "Can you mention some of the research papers that used deep learning-based methods for chip placement?", "answer": ["Sample-Efficient Reinforcement Learning via Conservative Model-Based Actor-Critic", "Efficient Exploration in Resource-Restricted Reinforcement Learning"], "answer_arxiv_id": ["2112.10504", "2212.06988"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_2643"} +{"question": "Which studies used LLMs to generate feedback for general domains?", "answer": ["GPTScore: Evaluate as You Desire", "Check Your Facts and Try Again: Improving Large Language Models with External Knowledge and Automated Feedback", "Re3: Generating Longer Stories With Recursive Reprompting and Revision"], "answer_arxiv_id": ["2302.04166", "2302.12813", "2210.06774"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_2644"} +{"question": "Which works have analyzed the equilibrium of content providers in the context of algorithm-curated platforms?", "answer": ["A Game-Theoretic Approach to Recommendation Systems with Strategic Content Providers", "From Recommendation Systems to Facility Location Games", "Supply-Side Equilibria in Recommender Systems", "Modeling Content Creator Incentives on Algorithm-Curated Platforms"], "answer_arxiv_id": ["1806.00955", "1809.02931", "2206.13489", "2206.13102"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2645"} +{"question": "Which papers handled safety in RL by minimizing the expected episode-wise violation?", "answer": ["Convergent Policy Optimization for Safe Reinforcement Learning", "Safe Reinforcement Learning in Constrained Markov Decision Processes", "Upper Confidence Primal-Dual Reinforcement Learning for CMDP with Adversarial Loss", "Exploration-Exploitation in Constrained MDPs", "Safe Reinforcement Learning via Curriculum Induction", "Learning in Markov Decision Processes under Constraints", "A Sample-Efficient Algorithm for Episodic Finite-Horizon MDP with Constraints", "Provably Efficient Safe Exploration via Primal-Dual Policy Optimization"], "answer_arxiv_id": ["1910.12156", "2008.06626", "2003.00660", "2003.02189", "2006.12136", "2002.12435", "2009.11348", "2003.00534"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_2646"} +{"question": "What is the research that introduced a point-based representation to reconstruct high-frequency details in facial attributes?", "answer": ["PointAvatar: Deformable Point-based Head Avatars from Videos"], "answer_arxiv_id": ["2212.08377"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_2647"} +{"question": "Which work introduced a combination of parametrizations in diffusion models to improve image quality?", "answer": ["Dynamic Dual-Output Diffusion Models"], "answer_arxiv_id": ["2203.04304"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2648"} +{"question": "What are some benchmarks specifically designed for the data science domain?", "answer": ["CERT: Continual Pre-Training on Sketches for Library-Oriented Code\n Generation", "Training and Evaluating a Jupyter Notebook Data Science Assistant", "DS-1000: A Natural and Reliable Benchmark for Data Science Code\n Generation"], "answer_arxiv_id": ["2206.06888", "2201.12901", "2211.11501"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2649"} +{"question": "Which works discuss the implementation of invertibility by designing dedicated architectures in the context of deep learning applications?", "answer": ["i-RevNet: Deep Invertible Networks", "Q"], "answer_arxiv_id": ["1802.07088", "1611.08152"], "source_meta": {"published_time": "20230826"}, "qid": "AutoScholarQuery_train_2650"} +{"question": "What work proposed formulating text classification tasks as a textual entailment problem?", "answer": ["Benchmarking Zero-shot Text Classification: Datasets, Evaluation and Entailment Approach"], "answer_arxiv_id": ["1909.00161"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_2651"} +{"question": "What papers make use of various algebraic structures to represent relational patterns in knowledge graphs?", "answer": ["Complex Embeddings for Simple Link Prediction", "RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space", "TorusE: Knowledge Graph Embedding on a Lie Group", "Quaternion Knowledge Graph Embeddings"], "answer_arxiv_id": ["1606.06357", "1902.10197", "1711.05435", "1904.10281"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_2652"} +{"question": "Which works show that large pretrained models are more robust to distributions shift?", "answer": ["Pretrained Transformers Improve Out-of-Distribution Robustness", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2004.06100", "2103.00020"], "source_meta": {"published_time": "20220220"}, "qid": "AutoScholarQuery_train_2653"} +{"question": "Which works developed Vision Transformers (ViTs) that split images into non-overlapping patches or tokens?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Training data-efficient image transformers & distillation through attention"], "answer_arxiv_id": ["2010.11929", "2012.12877"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_2654"} +{"question": "Could you tell me what works have addressed the nonlinear modeling of concepts?", "answer": ["Concept Whitening for Interpretable Image Recognition", "Concept Activation Regions: A Generalized Framework For Concept-Based Explanations"], "answer_arxiv_id": ["2002.01650", "2209.11222"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_2655"} +{"question": "What paper applies the principle of aligning 2D and 3D for object-level point cloud?", "answer": ["CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D\n Point Cloud Understanding"], "answer_arxiv_id": ["2203.00680"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2656"} +{"question": "Which papers discuss the direct matching methods used to adapt a model across domains in adversarial domain adaptation?", "answer": ["Deep Domain Confusion: Maximizing for Domain Invariance", "Deep CORAL: Correlation Alignment for Deep Domain Adaptation", "Moment Matching for Multi-Source Domain Adaptation", "Unsupervised Multi-Target Domain Adaptation Through Knowledge Distillation", "Continuously Indexed Domain Adaptation"], "answer_arxiv_id": ["1412.3474", "1607.01719", "1812.01754", "2007.07077", "2007.01807"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_2657"} +{"question": "Which are the studies pursuing differentiable subset selection in the context of Differentiable Subset Sampling?", "answer": ["Categorical Reparameterization with Gumbel-Softmax", "The Concrete Distribution: A Continuous Relaxation of Discrete Random\n Variables", "Stochastic Beams and Where to Find Them: The Gumbel-Top-k Trick for\n Sampling Sequences Without Replacement", "Differentiable Patch Selection for Image Recognition"], "answer_arxiv_id": ["1611.01144", "1611.00712", "1903.06059", "2104.03059"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_2658"} +{"question": "What works introduced neuro-symbolic methods for grounding in 3D scenes?", "answer": ["3D Concept Grounding on Neural Fields", "NS3D: Neuro-Symbolic Grounding of 3D Objects and Relations"], "answer_arxiv_id": ["2207.06403", "2303.13483"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_2659"} +{"question": "What papers are about the training of the model in methods incorporating spatial and stylistic control while generating images from text?", "answer": ["GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2301.07093"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2660"} +{"question": "What work have been done on distillation for recommender systems?", "answer": ["DE-RRD: A Knowledge Distillation Framework for Recommender System", "Bidirectional Distillation for Top-K Recommender System", "A Generic Network Compression Framework for Sequential Recommender Systems", "Unbiased Knowledge Distillation for Recommendation"], "answer_arxiv_id": ["2012.04357", "2106.02870v1", "2004.13139", "2211.14729"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_2661"} +{"question": "Are there any studies that integrate images for a more comprehensive perspective in point cloud pre-training?", "answer": ["PiMAE: Point Cloud and Image Interactive Masked Autoencoders for 3D Object Detection", "CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP", "Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Data"], "answer_arxiv_id": ["2303.08129", "2301.04926", "2203.16258"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_2662"} +{"question": "Which papers used supervised features for localizing single objects in an image?", "answer": ["Large-Scale Unsupervised Object Discovery", "Toward Unsupervised, Multi-Object Discovery in Large-Scale Image Collections"], "answer_arxiv_id": ["2106.06650", "2007.02662"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_2663"} +{"question": "Which works focused on improving event representation in dense video captioning?", "answer": ["Dense-Captioning Events in Videos", "Streamlined Dense Video Captioning", "A Better Use of Audio-Visual Cues: Dense Video Captioning with Bi-modal\n Transformer"], "answer_arxiv_id": ["1705.00754", "1904.03870", "2005.08271"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_2664"} +{"question": "Could you provide me some works related to generative models for learning representations?", "answer": ["Auto-Encoding Variational Bayes", "Generative Adversarial Nets", "Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks", "Adversarially Learned Inference", "Adversarial Feature Learning", "Large Scale Adversarial Representation Learning"], "answer_arxiv_id": ["1312.6114", "1406.2661", "1511.06434", "1606.00704", "1605.09782", "1907.02544"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_2665"} +{"question": "What work shared a similar focus to the study context by accelerating the AudioLM with a non-autoregressive decoding scheme?", "answer": ["SoundStorm: Efficient Parallel Audio Generation"], "answer_arxiv_id": ["2305.09636"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_2666"} +{"question": "Which works illustrate the decomposition of long-term occupancy matrix for representation learning?", "answer": ["Learning One Representation to Optimize All Rewards", "A Novel Stochastic Gradient Descent Algorithm for Learning Principal Subspaces"], "answer_arxiv_id": ["2103.07945", "2212.04025v1"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_2667"} +{"question": "Which papers provide details about methods similar to EFT and CLIFF, which update the human mesh recovery estimation network weight with 2D reprojection loss?", "answer": ["Exemplar Fine-Tuning for 3D Human Model Fitting Towards In-the-Wild 3D\n Human Pose Estimation", "CLIFF: Carrying Location Information in Full Frames into Human Pose and\n Shape Estimation"], "answer_arxiv_id": ["2004.03686", "2208.00571"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_2668"} +{"question": "What examples of Gaussian Processes (GP) libraries are mentioned in the paper?", "answer": ["GPf low: A Gaussian process library using TensorFlow", "A Framework for Interdomain and Multioutput Gaussian Processes", "GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Acceleration"], "answer_arxiv_id": ["1610.08733", "2003.01115", "1809.11165"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_2669"} +{"question": "What studies delve into the generation of adversarial examples in a black-box setting?", "answer": ["ZOO: Zeroth Order Optimization Based Black-box Attacks to Deep Neural Networks without Training Substitute Models", "Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models"], "answer_arxiv_id": ["1708.03999", "1712.04248"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_2670"} +{"question": "In which study was the AutoAttack, an ensemble of attacks based on PGD, proposed?", "answer": ["Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks"], "answer_arxiv_id": ["2003.01690"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_2671"} +{"question": "What works discuss doubly-robust approaches in the context of offline policy evaluation?", "answer": ["Doubly Robust Policy Evaluation and Optimization", "Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning", "Doubly Robust Off-policy Value Evaluation for Reinforcement Learning", "More Robust Doubly Robust Off-policy Evaluation"], "answer_arxiv_id": ["1503.02834", "1604.00923", "1511.03722", "1802.03493"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2672"} +{"question": "Can you point me to some papers where bird’s-eye view is generated by using 3D geometry-driven inductive biases such as unprojection into a volume?", "answer": ["Efficient and Robust 2D-to-BEV Representation Learning via Geometry-guided Kernel Transformer", "Orthographic Feature Transform for Monocular 3D Object Detection"], "answer_arxiv_id": ["2206.04584", "1811.08188"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_2673"} +{"question": "Which works perform image editing by traversing latent representation of StyleGAN?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN", "Training Generative Adversarial Networks with Limited Data"], "answer_arxiv_id": ["1812.04948", "1912.04958", "2006.06676"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_2674"} +{"question": "Could you provide me some papers where importance sampling was used to assist with off-environment learning in the domain of MARL?", "answer": ["Stabilising Experience Replay for Deep Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1702.08887"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_2675"} +{"question": "Which papers introduced the concept of depth-wise convolution for network efficiency?", "answer": ["MobileNets: Efficient Convolutional Neural Networks for Mobile Vision\n Applications", "MobileNetV2: Inverted Residuals and Linear Bottlenecks"], "answer_arxiv_id": ["1704.04861", "1801.04381"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_2676"} +{"question": "What papers extended the posterior regularization methods to deep generative models?", "answer": ["Deep Generative Models with Learnable Knowledge Constraints", "Learning Implicit Generative Models by Teaching Explicit Ones", "Amortized Inference Regularization", "Multi-objects Generation with Amortized Structural Regularization"], "answer_arxiv_id": ["1806.09764", "1807.03870v3", "1805.08913", "1906.03923"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_2677"} +{"question": "Which datasets are chosen in the work for providing diverse range of questions and answers?", "answer": ["TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension"], "answer_arxiv_id": ["1705.03551"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_2678"} +{"question": "Where has the combination of blueprint strategies with subgame solving led to state-of-the art performance?", "answer": ["DeepStack: Expert-Level Artificial Intelligence in Heads-Up No-Limit Poker", "Human-Level Performance in No-Press Diplomacy via Equilibrium Search", "Finding Friend and Foe in Multi-Agent Games"], "answer_arxiv_id": ["1701.01724", "2010.02923", "1906.02330"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_2679"} +{"question": "What studies represent the dual-stream category of VLP?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Filtering, Distillation, and Hard Negatives for Vision-Language\n Pre-Training"], "answer_arxiv_id": ["2103.00020", "2107.07651", "2102.05918", "2301.02280"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_2680"} +{"question": "Can you mention the studies focusing on improvements to BERT in compute settings comparable to the original BERT?", "answer": ["The MultiBerts: Bert Reproductions for Robustness Analysis"], "answer_arxiv_id": ["2106.16163"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_2681"} +{"question": "Who proposed SwinIR, a work building on transformer models?", "answer": ["SwinIR: Image Restoration Using Swin Transformer"], "answer_arxiv_id": ["2108.10257"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_2682"} +{"question": "Can you provide research that proposes progressive distillation algorithms to reduce the number of denoising steps in the DDPMs?", "answer": ["Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2202.00512"], "source_meta": {"published_time": "20230814"}, "qid": "AutoScholarQuery_train_2683"} +{"question": "Can you list the work that applied an extra adversarial loss to improve the unimodal Gaussian distribution matching?", "answer": ["Adversarial score matching and improved sampling for image generation"], "answer_arxiv_id": ["2009.05475"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_2684"} +{"question": "Can you cite research works that focus on a pruning-and-growing strategy for Vision Transformers?", "answer": ["Chasing Sparsity in Vision Transformers: An End-to-End Exploration"], "answer_arxiv_id": ["2106.04533"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_2685"} +{"question": "Can you mention a work that attempted generalization of neural networks to hyperbolic space?", "answer": ["Hyperbolic Neural Networks++"], "answer_arxiv_id": ["2006.08210"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_2686"} +{"question": "What articles suggest unrolling the model at training time to introduce a loss term on the multi-step predictions?", "answer": ["FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators", "Forecasting Global Weather with Graph Neural Networks", "GraphCast: Learning skillful medium-range global weather forecasting", "Predicting physics in mesh-reduced space with temporal attention"], "answer_arxiv_id": ["2202.11214", "2202.07575", "2212.12794", "2201.09113"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_2687"} +{"question": "What papers discussed reducing the inter-client variance to neutralize inconsistent update across clients in federated learning?", "answer": ["Variance Reduced Local SGD with Lower Communication Complexity", "Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning", "FedDANE: A Federated Newton-Type Method", "FedCM: Federated Learning with Client-level Momentum", "STEM: A Stochastic Two-Sided Momentum Algorithm Achieving Near-Optimal\n Sample and Communication Complexities for Federated Learning", "Faster Non-Convex Federated Learning via Global and Local Momentum", "Federated Learning via Posterior Averaging: A New Perspective and Practical Algorithms", "SlowMo: Improving Communication-Efficient Distributed SGD with Slow\n Momentum", "Measuring the Effects of Non-Identical Data Distribution for Federated\n Visual Classification", "Adaptive Federated Optimization"], "answer_arxiv_id": ["1912.12844", "2008.03606", "2001.01920", "2106.10874", "2106.10435", "2012.04061", "2010.05273v4", "1910.00643", "1909.06335", "2003.00295"], "source_meta": {"published_time": "20220110"}, "qid": "AutoScholarQuery_train_2688"} +{"question": "Can you highlight studies that explore the application of substructure related methods in various domains?", "answer": ["Junction Tree Variational Autoencoder for Molecular Graph Generation", "Hierarchical Graph-to-Graph Translation for Molecules", "Convolutional Networks on Graphs for Learning Molecular Fingerprints"], "answer_arxiv_id": ["1802.04364", "1907.11223", "1509.09292"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_2689"} +{"question": "Which works have improved the original AT formulation to balance the trade-off between standard and robust accuracy?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy", "Attacks Which Do Not Kill Training Make Adversarial Learning Stronger"], "answer_arxiv_id": ["1901.08573", "2002.11242"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_2690"} +{"question": "Which works proposed self-supervised learning where pretrained network is used by predicting rotation angle of the image?", "answer": ["Unsupervised Representation Learning by Predicting Image Rotations"], "answer_arxiv_id": ["1803.07728"], "source_meta": {"published_time": "20231104"}, "qid": "AutoScholarQuery_train_2691"} +{"question": "What study introduced a closed-loop and decentralized motion planner to avoid collision in dual-arm manipulation?", "answer": ["Learning a Decentralized Multi-arm Motion Planner"], "answer_arxiv_id": ["2011.02608"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_2692"} +{"question": "Which works have focused on interpreting the interrelationships between various actions and procedural events in instructional videos?", "answer": ["Cross-task weakly supervised learning from instructional videos"], "answer_arxiv_id": ["1903.08225"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_2693"} +{"question": "Which works proposed a recurrent framework utilizing optical flow for VSR?", "answer": ["Frame-Recurrent Video Super-Resolution"], "answer_arxiv_id": ["1801.04590"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_2694"} +{"question": "Can you list the studies that used vision transformers incorporating attention mechanisms to predict input-adaptive attention values?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Neighborhood Attention Transformer"], "answer_arxiv_id": ["2010.11929", "2103.14030", "2204.07143"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_2695"} +{"question": "Can you cite examples of studies which have proved global convergence of mean-field Langevin dynamics?", "answer": ["A Mean Field View of the Landscape of Two-Layer Neural Networks", "Mean-Field Langevin Dynamics : Exponential Convergence and Annealing", "Convex Analysis of the Mean Field Langevin Dynamics", "Mean-Field Langevin Dynamics and Energy Landscape of Neural Networks", "On the Global Convergence of Gradient Descent for Over-parameterized Models using Optimal Transport"], "answer_arxiv_id": ["1804.06561", "2202.01009", "2201.10469", "1905.07769", "1805.09545"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_2696"} +{"question": "Any works that focused on landmarks selection for high-level policy in hierarchical planning?", "answer": ["Landmark-Guided Subgoal Generation in Hierarchical Reinforcement Learning"], "answer_arxiv_id": ["2110.13625"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_2697"} +{"question": "What benchmark datasets were presented for LLMs in the context of VLM?", "answer": ["Learn to Explain: Multimodal Reasoning via Thought Chains for Science\n Question Answering", "MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning\n Benchmark for Expert AGI", "M3Exam: A Multilingual, Multimodal, Multilevel Benchmark for Examining\n Large Language Models"], "answer_arxiv_id": ["2209.09513", "2311.16502", "2306.05179"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_2698"} +{"question": "Could you provide me some work about text-guided audio generation systems in continuous space?", "answer": ["Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models", "AudioLDM: Text-to-Audio Generation with Latent Diffusion Models"], "answer_arxiv_id": ["2301.12661", "2301.12503"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_2699"} +{"question": "What is the original paper that introduced the concepts for DGNs typically used today, particularly those based on the message passing paradigm?", "answer": ["Neural Message Passing for Quantum Chemistry"], "answer_arxiv_id": ["1704.01212"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_2700"} +{"question": "Could you provide me with some studies that focused on data-driven discovery of physical laws in machine learning?", "answer": ["Data-driven discovery of partial differential equations", "Discovering Nonlinear PDEs from Scarce Data with Physics-encoded Learning", "DL-PDE: Deep-learning based data-driven discovery of partial differential equations from discrete and noisy data", "Model selection for dynamical systems via sparse regression and information criteria", "Any equation is a forest: Symbolic genetic algorithm for discovering open-form partial differential equations (SGA-PDE)", "DLGA-PDE: Discovery of PDEs with incomplete candidate library via combination of deep learning and genetic algorithm", "PDE-Net 2.0: Learning PDEs from Data with A Numeric-Symbolic Hybrid Deep Network", "PDE-Net: Learning PDEs from Data", "Numerical Gaussian Processes for Time-dependent and Non-linear Partial Differential Equations", "Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations"], "answer_arxiv_id": ["1609.06401", "2201.12354", "1908.04463v2", "1701.01773", "2106.11927v1", "2001.07305", "1812.04426", "1710.09668", "1703.10230", "1708.00588"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_2701"} +{"question": "Are there any studies about automatically generating suggestions for alternative responses in the context of peer counseling?", "answer": ["Human-AI Collaboration Enables More Empathic Conversations in Text-based\n Peer-to-Peer Mental Health Support", "Helping the Helper: Supporting Peer Counselors via AI-Empowered Practice\n and Feedback"], "answer_arxiv_id": ["2203.15144", "2305.08982"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_2702"} +{"question": "What works propose predicting the relative position of paired patches from the same image in self-supervised learning?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction"], "answer_arxiv_id": ["1505.05192"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_2703"} +{"question": "What papers demonstrate the use of model-based RL on tasks with high-dimensional input?", "answer": ["Recurrent World Models Facilitate Policy Evolution"], "answer_arxiv_id": ["1809.01999"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_2704"} +{"question": "What research mentioned orthogonal transformations as examples in identifying causal sources?", "answer": ["Contrastive Learning Inverts the Data Generating Process"], "answer_arxiv_id": ["2102.08850v4"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_2705"} +{"question": "What works use Large Vision-Language Models (LVLMs) to perform real-world tasks such as tool-using, web browsing, and autonomous driving?", "answer": ["LLaVA-Plus: Learning to Use Tools for Creating Multimodal Agents", "GPT-4V(ision) is a Generalist Web Agent, if Grounded", "DriveGPT4: Interpretable End-to-end Autonomous Driving via Large\n Language Model"], "answer_arxiv_id": ["2311.05437", "2401.01614", "2310.01412"], "source_meta": {"published_time": "20240630"}, "qid": "AutoScholarQuery_train_2706"} +{"question": "Can you specify some works that applied Gaussian processes to estimate treatment effects over time in the presence of irregular samples?", "answer": ["A Bayesian Nonparametric Approach for Estimating Individualized Treatment-Response Curves", "Reliable Decision Support using Counterfactual Models", "Treatment-Response Models for Counterfactual Reasoning with Continuous-time, Continuous-valued Interventions"], "answer_arxiv_id": ["1608.05182", "1703.10651", "1704.02038"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_2707"} +{"question": "Could you provide me studies that discussed backpropagation, and its uses and applications?", "answer": ["Automatic Differentiation in Machine Learning: a Survey", "Stein Unbiased GrAdient estimator of the Risk (SUGAR) for multiple parameter selection", "Truncated Back-propagation for Bilevel Optimization"], "answer_arxiv_id": ["1502.05767", "1405.1164v2", "1810.10667"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_2708"} +{"question": "Which works model facial details as displacements maps in 3D face reconstruction from RGB inputs?", "answer": ["Photo-Realistic Facial Details Synthesis from Single Image", "Learning an Animatable Detailed 3D Face Model from In-The-Wild Images", "A Hierarchical Representation Network for Accurate and Detailed Face\n Reconstruction from In-The-Wild Images"], "answer_arxiv_id": ["1903.10873", "2012.04012", "2302.14434"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_2709"} +{"question": "Can you mention some research papers on semi-dense methods for SLAM?", "answer": ["Direct Sparse Odometry"], "answer_arxiv_id": ["1607.02565"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_2710"} +{"question": "What works discuss about the group shifts in Distributionally Robust Optimization?", "answer": ["Distributionally Robust Language Modeling", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1909.02060", "1911.08731"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_2711"} +{"question": "Which study discussed the optimal number of layers for node prediction tasks in the context of the oversmoothing phenomenon?", "answer": ["Not too little, not too much: a theoretical analysis of graph (over)smoothing"], "answer_arxiv_id": ["2205.12156"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_2712"} +{"question": "Which papers have proposed generative approaches using diffusion models for 3D scene generation?", "answer": ["Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision", "Rodin: A Generative Model for Sculpting 3D Digital Avatars Using\n Diffusion"], "answer_arxiv_id": ["2306.11719", "2212.06135"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_2713"} +{"question": "What research has been conducted on assessing large language models' ability to generate coherent thought chains in reasoning tasks?", "answer": ["Chain-of-Thought Hub: A Continuous Effort to Measure Large Language\n Models' Reasoning Performance", "ThoughtSource: A central hub for large language model reasoning data"], "answer_arxiv_id": ["2305.17306", "2301.11596v5"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_2714"} +{"question": "Could you refer some works that introduced the federated multi-target domain adaptation problem?", "answer": ["Federated Multi-Target Domain Adaptation"], "answer_arxiv_id": ["2108.07792"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_2715"} +{"question": "What works proposed a two-stage framework for depth estimation that estimates affine-invariant depth and then upgrades it to metric depth?", "answer": ["Learning to Recover 3D Scene Shape from a Single Image"], "answer_arxiv_id": ["2012.09365"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2716"} +{"question": "Can you name the papers that make the distribution shift bounded by constraining the differences between learned policy and behavior policy?", "answer": ["Behavior Regularized Offline Reinforcement Learning", "Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction"], "answer_arxiv_id": ["1911.11361", "1812.02900", "1906.00949"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_2717"} +{"question": "Which works used adapters for domain adaptation in the field of computer vision?", "answer": ["Learning multiple visual domains with residual adapters", "Efficient parametrization of multi-domain deep neural networks"], "answer_arxiv_id": ["1705.08045", "1803.10082"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_2718"} +{"question": "What large language models contributed significantly to revolutionize the NLP research?", "answer": ["Language Models are Few-Shot Learners", "OPT: Open Pre-trained Transformer Language Models", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model"], "answer_arxiv_id": ["2005.14165", "2205.01068", "2211.05100"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_2719"} +{"question": "What studies focus on deriving reaction-time measures from neural networks as a proxy for computational cost?", "answer": ["Adaptive Computation Time for Recurrent Neural Networks"], "answer_arxiv_id": ["1603.08983"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_2720"} +{"question": "Could you name any works that have improved DreamFusion using a two-stage optimization framework?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation"], "answer_arxiv_id": ["2211.10440"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2721"} +{"question": "What works are studying 3D generation in the form of polygon meshes?", "answer": ["GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images", "Learning Category-Specific Mesh Reconstruction from Image Collections", "PolyGen: An Autoregressive Generative Model of 3D Meshes", "DeepMetaHandles: Learning Deformation Meta-Handles of 3D Meshes with\n Biharmonic Coordinates", "Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images", "MeshDiffusion: Score-based Generative 3D Mesh Modeling"], "answer_arxiv_id": ["2209.11163", "1803.07549", "2002.10880", "2102.09105", "1804.01654", "2303.08133"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_2722"} +{"question": "Which works utilise neural network architecture for causal modeling?", "answer": ["The Causal-Neural Connection: Expressiveness, Learnability, and Inference", "Learning Functional Causal Models with Generative Neural Networks"], "answer_arxiv_id": ["2107.00793", "1709.05321v3"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_2723"} +{"question": "What papers include a complex short-range part with higher-order geometric information in their multiple message sums approach?", "answer": ["Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures"], "answer_arxiv_id": ["2011.07457"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_2724"} +{"question": "Can you mention some works that introduce a fusion method based on a mixed attention mechanism?", "answer": ["RGB-T Tracking Based on Mixed Attention"], "answer_arxiv_id": ["2304.04264v4"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_2725"} +{"question": "Which paper proposed a method to turn a classifier into a conditional generative model?", "answer": ["Traditional Classification Neural Networks are Good Generators: They are Competitive with DDPMs and GANs"], "answer_arxiv_id": ["2211.14794"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_2726"} +{"question": "Which studies propose a frequentist definition for reconstruction error?", "answer": ["Bounding Training Data Reconstruction in Private (Deep) Learning"], "answer_arxiv_id": ["2201.12383"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_2727"} +{"question": "Can you list studies that used ViT based masked image modeling for computer vision tasks?", "answer": ["iBOT : Image BERT Pre-Training with Online Tokenizer", "BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: a Simple Framework for Masked Image Modeling", "Masked Siamese Networks for Label-Efficient Learning"], "answer_arxiv_id": ["2111.07832", "2106.08254", "2111.06377", "2111.09886", "2204.07141"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_2728"} +{"question": "Is there any work that investigated the role of data in the pre-training stage and analyzed what information is encoded into the rewind point by the dense network training?", "answer": ["Lottery Tickets on a Data Diet: Finding Initializations with Sparse Trainable Networks"], "answer_arxiv_id": ["2206.01278"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_2729"} +{"question": "Which papers cover depth estimation from single-image captures?", "answer": ["Unsupervised Monocular Depth Estimation with Left-Right Consistency", "3D Packing for Self-Supervised Monocular Depth Estimation"], "answer_arxiv_id": ["1609.03677", "1905.02693"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_2730"} +{"question": "Which research papers have introduced question answering tasks on video?", "answer": ["TVQA: Localized, Compositional Video Question Answering", "LEMMA: A Multi-view Dataset for LEarning Multi-agent Multi-task Activities", "EgoTaskQA: Understanding Human Tasks in Egocentric Videos", "AGQA: A Benchmark for Compositional Spatio-Temporal Reasoning", "STAR: A Benchmark for Situated Reasoning in Real-World Videos", "Episodic Memory Question Answering"], "answer_arxiv_id": ["1809.01696", "2007.15781", "2210.03929", "2103.16002", "2405.09711", "2205.01652"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_2731"} +{"question": "What studies utilized the Single Shape Fréchet Inception Distance in deep learning-based methods?", "answer": ["Learning to Generate 3D Shapes from a Single Example"], "answer_arxiv_id": ["2208.02946"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_2732"} +{"question": "Could you cite any papers that explored techniques like ReAct, DFS tree search, or self-consistency to address errors in large language models?", "answer": ["ReAct: Synergizing Reasoning and Acting in Language Models", "ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2210.03629", "2307.16789", "2203.11171"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2733"} +{"question": "Which studies have utilized multi-scale information in convolutional neural networks and transformers?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Deep Residual Learning for Image Recognition", "Focal Modulation Networks", "Feature Pyramid Networks for Object Detection", "Multiscale Vision Transformers", "Remote Sensing Change Detection With Transformers Trained from Scratch", "Mask R-CNN", "Transformers in Remote Sensing: A Survey"], "answer_arxiv_id": ["2103.14030", "1512.03385", "2203.11926", "1612.03144", "2104.11227", "2304.06710", "1703.06870", "2209.01206"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_2734"} +{"question": "Which papers propose hybrid DR as a solution to address the bias and variance issues of DM and IPS?", "answer": ["Doubly Robust Policy Evaluation and Optimization", "Optimal and Adaptive Off-policy Evaluation in Contextual Bandits", "More Robust Doubly Robust Off-policy Evaluation", "Doubly robust off-policy evaluation with shrinkage"], "answer_arxiv_id": ["1503.02834", "1612.01205", "1802.03493", "1907.09623"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_2735"} +{"question": "Which works show that self-supervised learning can train large-scale language models and provide powerful representations for downstream natural language processing tasks?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20220202"}, "qid": "AutoScholarQuery_train_2736"} +{"question": "Could you give me examples of studies that received a lot of attention for their usage of the one-inclusion graph algorithm?", "answer": ["The One-Inclusion Graph Algorithm is not Always Optimal", "A Theory of Universal Learning", "Multiclass Learnability Beyond the PAC Framework: Universal Rates and Partial Concept Classes", "A Theory of PAC Learnability under Transformation Invariances", "A Theory of PAC Learnability of Partial Concept Classes", "A Characterization of List Learnability"], "answer_arxiv_id": ["2212.09270", "2011.04483", "2210.02297", "2202.07552", "2107.08444v2", "2211.04956v2"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_2737"} +{"question": "In which studies was parameter and neuron pruning used for model compression?", "answer": ["Pruning Filters for Efficient ConvNets"], "answer_arxiv_id": ["1608.08710"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_2738"} +{"question": "Could you provide me some studies using deformable convolution-based methods to learn offsets from compressed frames for VQE?", "answer": ["MFQE 2.0: A New Approach for Multi-frame Quality Enhancement on\n Compressed Video"], "answer_arxiv_id": ["1902.09707"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_2739"} +{"question": "Which researches built datasets for multi-view pedestrian detection?", "answer": ["Multiview Detection with Feature Perspective Transformation"], "answer_arxiv_id": ["2007.07247"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_2740"} +{"question": "Could you mention some studies in the domain of self-supervised time series representation learning?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Time-Series Representation Learning via Temporal and Contextual Contrasting", "TS2Vec: Towards Universal Representation of Time Series"], "answer_arxiv_id": ["1807.03748", "2106.14112", "2106.10466"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_2741"} +{"question": "Which papers used the Covariance Matrix Adaptation Evolutionary Strategy (CMA-ES) for single-objective optimization?", "answer": ["The CMA Evolution Strategy: A Tutorial"], "answer_arxiv_id": ["1604.00772"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_2742"} +{"question": "Which papers proposed the pipeline-based methods to reduce hallucinations in chart-to-summary generation?", "answer": ["Chart-to-Text: A Large-Scale Benchmark for Chart Summarization"], "answer_arxiv_id": ["2203.06486"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_2743"} +{"question": "Which work provided the first breakthrough in data-free quantization?", "answer": ["Data-Free Quantization Through Weight Equalization and Bias Correction"], "answer_arxiv_id": ["1906.04721"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_2744"} +{"question": "What papers used the model’s own future scores for expanding a set of candidates for future steps?", "answer": ["On the Depth between Beam Search and Exhaustive Search for Text\n Generation"], "answer_arxiv_id": ["2308.13696"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_2745"} +{"question": "Could you name some research papers that introduced tasks for controllable timbre generation?", "answer": ["Conditional Generation of Audio from Video via Foley Analogies", "VarietySound: Timbre-Controllable Video to Sound Generation via\n Unsupervised Information Disentanglement"], "answer_arxiv_id": ["2304.08490", "2211.10666"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_2746"} +{"question": "Could you provide me some works that learn the symmetry transformations on inputs and latent features end-to-end with the task function?", "answer": ["Meta-learning Symmetries by Reparameterization", "Automatic Symmetry Discovery with Lie Algebra Convolutional Network"], "answer_arxiv_id": ["2007.02933", "2109.07103"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_2747"} +{"question": "What studies describe the use of Zernike-based methods in atmospheric turbulence simulation?", "answer": ["Real-Time Dense Field Phase-to-Space Simulation of Imaging through\n Atmospheric Turbulence", "Accelerating Atmospheric Turbulence Simulation via Learned\n Phase-to-Space Transform", "Simulating Anisoplanatic Turbulence by Sampling Inter-modal and\n Spatially Correlated Zernike Coefficients"], "answer_arxiv_id": ["2210.06713", "2107.11627", "2004.11210"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_2748"} +{"question": "What works have been referenced on the topic of retrieval-augmented generation in language modeling?", "answer": ["Improving language models by retrieving from trillions of tokens", "Generalization through Memorization: Nearest Neighbor Language Models", "REPLUG: Retrieval-Augmented Black-Box Language Models"], "answer_arxiv_id": ["2112.04426", "1911.00172", "2301.12652"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_2749"} +{"question": "What papers have approached Dropout Variational Inference by proposing Bayesian Dropout inference methods?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Variational Dropout and the Local Reparameterization Trick"], "answer_arxiv_id": ["1506.02142", "1506.02557"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_2750"} +{"question": "Could you tell me any papers that showcase algorithms known to be computationally intractable?", "answer": ["Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms"], "answer_arxiv_id": ["2102.00815"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_2751"} +{"question": "Which work put forward the idea of using unsupervised retrieval to get candidate examples, with a top selection made using a supervised prompt retriever to boost downstream performance?", "answer": ["Learning To Retrieve Prompts for In-Context Learning"], "answer_arxiv_id": ["2112.08633"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_2752"} +{"question": "Which works focused on video generation using off-the-shelf image diffusion models?", "answer": ["FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation", "TokenFlow: Consistent Diffusion Features for Consistent Video Editing", "FLATTEN: optical FLow-guided ATTENtion for consistent text-to-video\n editing"], "answer_arxiv_id": ["2303.09535", "2212.11565", "2306.07954", "2307.10373", "2310.05922"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_2753"} +{"question": "What are some studies that employ dual-stream neural networks for processing global and local features?", "answer": ["Polar Transformer Networks", "Attention for Fine-Grained Categorization", "Glance and Focus: a Dynamic Approach to Reducing Spatial Redundancy in Image Classification", "Learning Scene Gist with Convolutional Neural Networks to Improve Object Recognition"], "answer_arxiv_id": ["1709.01889", "1412.7054", "2010.05300", "1803.01967v2"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_2754"} +{"question": "Can you list some works that focus on third-person imitation learning and view-invariant visual representations?", "answer": ["Third-Person Imitation Learning", "Third-Person Visual Imitation Learning via Decoupled Hierarchical Controller", "Self-Supervised Disentangled Representation Learning for Third-Person Imitation Learning", "3D-OES: Viewpoint-Invariant Object-Factorized Environment Simulators", "Unsupervised Learning of Visual 3D Keypoints for Control", "3D Neural Scene Representations for Visuomotor Control", "Visual Reinforcement Learning with Self-Supervised 3D Representations", "Reinforcement Learning with Neural Radiance Fields"], "answer_arxiv_id": ["1703.01703", "1911.09676", "2108.01069", "2011.06464", "2106.07643", "2107.04004", "2210.07241", "2206.01634"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_2755"} +{"question": "Which research assumes the attacker can access both training procedures and samplers and applies correction terms on DDPM and DDIM to launch the attack?", "answer": ["TrojDiff: Trojan Attacks on Diffusion Models with Diverse Targets"], "answer_arxiv_id": ["2303.05762"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_2756"} +{"question": "Could you provide me some works on spectral-based hypergraph neural networks similar to applying GNNs on clique expansions?", "answer": ["Hypergraph Neural Networks", "Hypergraph Convolution and Hypergraph Attention", "Hypergraph Convolutional Networks via Equivalency between Hypergraphs and Undirected Graphs"], "answer_arxiv_id": ["1809.09401", "1901.08150", "2203.16939v3"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_2757"} +{"question": "Could you provide me some studies dealing with category-level object pose estimation?", "answer": ["Normalized Object Coordinate Space for Category-Level 6D Object Pose and\n Size Estimation", "Learning Canonical Shape Space for Category-Level 6D Object Pose and\n Size Estimation", "Category-Level 6D Object Pose and Size Estimation using Self-Supervised\n Deep Prior Deformation Networks"], "answer_arxiv_id": ["1901.02970", "2001.09322", "2207.05444"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_2758"} +{"question": "What works propose context-free grammars for action detection and recognition?", "answer": ["Weakly supervised learning of actions from transcripts", "Weakly Supervised Action Learning with RNN based Fine-to-coarse Modeling", "NeuralNetwork-Viterbi: A Framework for Weakly Supervised Video Learning"], "answer_arxiv_id": ["1610.02237", "1703.08132", "1805.06875"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2759"} +{"question": "Which paper present the popular DP algorithm, DP-SGD?", "answer": ["Deep Learning with Differential Privacy"], "answer_arxiv_id": ["1607.00133"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_2760"} +{"question": "What are the Knowledge distillation methods for Anomaly Detection?", "answer": ["Anomaly Detection via Reverse Distillation from One-Class Embedding", "Uninformed Students: Student-Teacher Anomaly Detection with\n Discriminative Latent Embeddings", "Multiresolution Knowledge Distillation for Anomaly Detection", "Student-Teacher Feature Pyramid Matching for Anomaly Detection", "Anomaly Detection under Distribution Shift", "DeSTSeg: Segmentation Guided Denoising Student-Teacher for Anomaly\n Detection"], "answer_arxiv_id": ["2201.10703", "1911.02357", "2011.11108", "2103.04257", "2303.13845", "2211.11317"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_2761"} +{"question": "What research work is most relevant to incorporating geometry priors with epipolar geometries?", "answer": ["SparseFusion: Distilling View-conditioned Diffusion for 3D\n Reconstruction"], "answer_arxiv_id": ["2212.00792"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_2762"} +{"question": "Can you provide studies that investigate empirical notions of consistency and diversity, and variations in dataset scales?", "answer": ["How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers"], "answer_arxiv_id": ["2106.10270"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_2763"} +{"question": "What studies focused on the issue of lack of marginal preservation when using batch couplings for generative modeling?", "answer": ["Minibatch optimal transport distances; analysis and applications", "Improving Mini-batch Optimal Transport via Partial Transportation"], "answer_arxiv_id": ["2101.01792", "2108.09645"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_2764"} +{"question": "What papers propose to introduce penalties for nonfactual responses aiming to mitigate hallucination in reinforcement learning from human feedback?", "answer": ["Training language models to follow instructions with human feedback", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2203.02155", "2307.09288"], "source_meta": {"published_time": "20240225"}, "qid": "AutoScholarQuery_train_2765"} +{"question": "Which paper introduced the Low-Rank Adaptation (LoRA) technique?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_2766"} +{"question": "Which works suggest that deep neural networks forget hard samples after pruning?", "answer": ["What Do Compressed Deep Neural Networks Forget?", "Self-Damaging Contrastive Learning"], "answer_arxiv_id": ["1911.05248", "2106.02990"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_2767"} +{"question": "Which paper originally introduced the concept of Neural Radiance Fields (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_2768"} +{"question": "What studies have explored decomposition techniques like question reduction, iterative prompting, and chaining the steps?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "SeqZero: Few-shot Compositional Semantic Parsing with Sequential Prompts and Zero-shot Models", "AI Chains: Transparent and Controllable Human-AI Interaction by Chaining Large Language Model Prompts"], "answer_arxiv_id": ["2205.10625", "2205.07381", "2110.01691"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_2769"} +{"question": "What papers discussed factorizing value functions across interactions?", "answer": ["Mean Field Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1802.05438"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_2770"} +{"question": "Which papers are about the diffusion-based methods achieving significantly superior results in text-to-image research?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Composer: Creative and Controllable Image Synthesis with Composable\n Conditions"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2112.10752", "2112.10741", "2211.01324", "2302.09778"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_2771"} +{"question": "Which research papers proposed methods for alleviating the oversquashing problem in GNNs?", "answer": ["On the Bottleneck of Graph Neural Networks and its Practical Implications", "Understanding over-squashing and bottlenecks on graphs via curvature"], "answer_arxiv_id": ["2006.05205", "2111.14522"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_2772"} +{"question": "What studies have explored the feasibility of combining simulation in computer vision tasks?", "answer": ["Object Detection Using Sim2Real Domain Randomization for Robotic Applications", "Towards Inclusive HRI: Using Sim2Real to Address Underrepresentation in Emotion Expression Recognition", "CAD2RL: Real Single-Image Flight Without a Single Real Image"], "answer_arxiv_id": ["2208.04171v2", "2208.07472", "1611.04201"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_2773"} +{"question": "Which studies focused on the practicability of E2E approach and the study of the network structure of Transformer for ST?", "answer": ["Listen and Translate: A Proof of Concept for End-to-End Speech-to-Text Translation", "Sequence-to-Sequence Models Can Directly Translate Foreign Speech", "Fluent Translations from Disfluent Speech in End-to-End Speech Translation", "A Comparative Study on End-to-End Speech to Text Translation"], "answer_arxiv_id": ["1612.01744", "1703.08581", "1906.00556", "1911.08870"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_2774"} +{"question": "What papers are about the mitigation of data heterogeneity issues in DSGD?", "answer": ["A Unified Theory of Decentralized SGD with Changing Topology and Local Updates", "On the Influence of Bias-Correction on Distributed Stochastic Optimization"], "answer_arxiv_id": ["2003.10422", "1903.10956"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2775"} +{"question": "Could you provide me some studies about text-to-image setting using diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2112.10752", "2205.11487"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2776"} +{"question": "Could you provide examples of works that have simplified and interpreted GCN from a spatial perspective?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks"], "answer_arxiv_id": ["1609.02907"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_2777"} +{"question": "What work details the adoption of the Vision Transformer (ViT) model?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_2778"} +{"question": "What studies have utilized layer-wise relevance propagation for weakly-supervised grounding?", "answer": ["Explaining NonLinear Classification Decisions with Deep Taylor\n Decomposition"], "answer_arxiv_id": ["1512.02479"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2779"} +{"question": "What methods only adjust the model predictions using a group annotated held-out set?", "answer": ["Distributionally Robust Post-hoc Classifiers under Prior Shifts"], "answer_arxiv_id": ["2309.08825"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_2780"} +{"question": "Which work studies the impact of different pre-training set-ups on distribution shift robustness?", "answer": ["An Empirical Study on Distribution Shift Robustness From the Perspective of Pre-Training and Data Augmentation"], "answer_arxiv_id": ["2205.12753"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_2781"} +{"question": "What research has improved concatenation based conditioning through the use of local features?", "answer": ["Mixing-Denoising Generalizable Occupancy Networks", "Learning Deep Implicit Functions for 3D Shapes with Dynamic Code Clouds", "Local Deep Implicit Functions for 3D Shape", "PatchNets: Patch-Based Generalizable Deep Implicit 3D Shape Representations", "Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes", "Convolutional Occupancy Networks", "Implicit Feature Networks for Texture Completion from Partial 3D Data", "Local Implicit Grid Representations for 3D Scenes", "Points2Surf Learning Implicit Surfaces from Point Clouds"], "answer_arxiv_id": ["2311.12125", "2203.14048", "1912.06126", "2008.01639", "2101.10994", "2003.04618", "2009.09458", "2003.08981", "2007.10453"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_2782"} +{"question": "What studies propose Federated Distillation techniques?", "answer": ["Communication-Efficient On-Device Machine Learning: Federated\n Distillation and Augmentation under Non-IID Private Data"], "answer_arxiv_id": ["1811.11479"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_2783"} +{"question": "What studies focus on the issue of factual errors over long contexts in deep NLP models?", "answer": ["Sharp Nearby, Fuzzy Far Away: How Neural Language Models Use Context", "Long Range Arena: A Benchmark for Efficient Transformers", "LongT5: Efficient Text-To-Text Transformer for Long Sequences", "Survey of Hallucination in Natural Language Generation"], "answer_arxiv_id": ["1805.04623", "2011.04006", "2112.07916", "2202.03629"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2784"} +{"question": "Can you provide research that studied the CPE phenomenon arising in isolation of data curation, data augmentation, and misspecified priors?", "answer": ["Disentangling the Roles of Curation, Data-Augmentation and the Prior in the Cold Posterior Effect"], "answer_arxiv_id": ["2106.06596"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_2785"} +{"question": "Are there any works that used UPop as a structured pruning approach?", "answer": ["Reducing Transformer Depth on Demand with Structured Dropout", "Are Sixteen Heads Really Better than One?", "Vision Transformer Pruning", "Vision Transformer Slimming: Multi-Dimension Searching in Continuous Optimization Space"], "answer_arxiv_id": ["1909.11556", "1905.10650", "2104.08500", "2201.00814"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_2786"} +{"question": "Which studies explore NeRF in the frequency space?", "answer": ["BARF: Bundle-Adjusting Neural Radiance Fields", "Nerfies: Deformable Neural Radiance Fields", "HF-NeuS: Improved Surface Reconstruction Using High-Frequency Details", "FreeNeRF: Improving Few-shot Neural Rendering with Free Frequency\n Regularization", "WaveNeRF: Wavelet-based Generalizable Neural Radiance Fields"], "answer_arxiv_id": ["2104.06405", "2011.12948", "2206.07850", "2303.07418", "2308.04826"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_2787"} +{"question": "What paper discusses the improvement of spatial and temporal resolution through latent space upsampling?", "answer": ["AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "MagicVideo: Efficient Video Generation With Latent Diffusion Models"], "answer_arxiv_id": ["2307.04725", "2306.02018", "2211.11018"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_2788"} +{"question": "In what studies researchers used model-based RL in the domain of sequential decision-making problems?", "answer": ["Simplifying Model-based RL: Learning Representations, Latent-space Models, and Policies with One Objective"], "answer_arxiv_id": ["2209.08466"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_2789"} +{"question": "Which works have dealt with bad predictions due to occlusions?", "answer": ["Learning to Look around Objects for Top-View Representations of Outdoor\n Scenes"], "answer_arxiv_id": ["1803.10870"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_2790"} +{"question": "Which papers propose the use of recurrent transformer models?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Recurrent Memory Transformer"], "answer_arxiv_id": ["1901.02860", "2207.06881"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2791"} +{"question": "What are some works that focused on studying the minimizer of the training error and finding the optimal smoothing parameter for label smoothing?", "answer": ["An Investigation of how Label Smoothing Affects Generalization"], "answer_arxiv_id": ["2010.12648"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_2792"} +{"question": "What research proposes an update based on the asymmetric part of the game Hessian obtained from its Helmholtz decomposition?", "answer": ["The Mechanics of n-Player Differentiable Games"], "answer_arxiv_id": ["1802.05642v2"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_2793"} +{"question": "Could you provide me with research that used standard membership inference attacks for evaluating different privacy analysis algorithms?", "answer": ["Evaluating Differentially Private Machine Learning in Practice"], "answer_arxiv_id": ["1902.08874"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_2794"} +{"question": "Which papers study generating entity descriptions from Wikipedia infotables?", "answer": ["Neural Text Generation from Structured Data with Application to the\n Biography Domain"], "answer_arxiv_id": ["1603.07771"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_2795"} +{"question": "What papers focused on backdoor attack defense methods focusing on backdoor detection?", "answer": ["Rethinking the Reverse-engineering of Trojan Triggers", "Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs", "Detecting AI Trojans Using Meta Neural Analysis", "Defending Neural Backdoors via Generative Distribution Modeling", "NeuronInspect: Detecting Backdoors in Neural Networks via Output\n Explanations", "Backdoor Scanning for Deep Neural Networks through K-Arm Optimization", "Practical Detection of Trojan Neural Networks: Data-Limited and\n Data-Free Cases", "BEAGLE: Forensics of Deep Learning Backdoor Attack for Better Defense", "Detecting Backdoors in Pre-trained Encoders"], "answer_arxiv_id": ["2210.15127", "1906.10842", "1910.03137", "1910.04749", "1911.07399", "2102.05123", "2007.15802", "2301.06241", "2303.15180"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_2796"} +{"question": "Are there any papers on Score Distillation Sampling (SDS) in text-to-3D generative models?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation"], "answer_arxiv_id": ["2211.10440", "2212.00774v1", "2303.13873", "2305.16213"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_2797"} +{"question": "Which paper investigates adversarial training on long-tailed datasets?", "answer": ["Adversarial Robustness under Long-Tailed Distribution"], "answer_arxiv_id": ["2104.02703"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_2798"} +{"question": "Which papers discuss vehicle-side datasets used in autonomous driving?", "answer": ["nuScenes: A multimodal dataset for autonomous driving", "Scalability in Perception for Autonomous Driving: Waymo Open Dataset"], "answer_arxiv_id": ["1903.11027", "1912.04838"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_2799"} +{"question": "Which papers propose smoother noise methods that induce time correlation by keeping perturbation parameters constant for multiple steps?", "answer": ["Parameter Space Noise for Exploration", "Smooth Exploration for Robotic Reinforcement Learning", "Noisy Networks for Exploration"], "answer_arxiv_id": ["1706.01905", "2005.05719", "1706.10295"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2800"} +{"question": "Are there any works that extend causal convolution to streaming videos?", "answer": ["Massively Parallel Video Networks", "Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "MoViNets: Mobile Video Networks for Efficient Video Recognition"], "answer_arxiv_id": ["1806.03863", "1901.02860", "2103.11511"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_2801"} +{"question": "Which work proposes aligning MLLM semantics with the diffusion image decoder via image-caption pair training?", "answer": ["M-VADER: A Model for Diffusion with Multimodal Context"], "answer_arxiv_id": ["2212.02936"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_2802"} +{"question": "Can you name studies that have identified preference shaping of users induced by a recommendation algorithm?", "answer": ["Towards Psychologically-Grounded Dynamic Preference Models"], "answer_arxiv_id": ["2208.01534"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_2803"} +{"question": "What works have assessed memory in the context of Transformers in RL?", "answer": ["Stabilizing Transformers for Reinforcement Learning", "Deep Transformer Q-Networks for Partially Observable Reinforcement Learning", "TransDreamer: Reinforcement Learning with Transformer World Models", "POPGym: Benchmarking Partially Observable Reinforcement Learning"], "answer_arxiv_id": ["1910.06764", "2206.01078", "2202.09481", "2303.01859"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_2804"} +{"question": "Which works assumed their decoder distributions as Student-t and asymmetric Laplace distributions to mitigate the zero-variance problem?", "answer": ["Deep Quantile Regression for Uncertainty Estimation in Unsupervised and Supervised Lesion Detection"], "answer_arxiv_id": ["2109.09374"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_2805"} +{"question": "Which papers propose strategies to achieve Zero-shot coordination by training using diverse populations of strategies?", "answer": ["Collaborating with Humans without Human Data", "Maximum Entropy Population-Based Training for Zero-Shot Human-AI Coordination"], "answer_arxiv_id": ["2110.08176", "2112.11701"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_2806"} +{"question": "What is the closest research to this particular work that uses optimal transport theory to find lower bounds on multi-class classification?", "answer": ["The multimarginal optimal transport formulation of adversarial multiclass classification"], "answer_arxiv_id": ["2204.12676"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_2807"} +{"question": "Which work discusses rich feature learning and its impact on OOD performance?", "answer": ["Rich Feature Construction for the Optimization-Generalization Dilemma", "Diverse Weight Averaging for Out-of-Distribution Generalization", "Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization"], "answer_arxiv_id": ["2203.15516", "2205.09739", "2110.10832"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_2808"} +{"question": "Could you provide the reference to the study which proposed an efficient alternative to render the radiance field using 3D Gaussian splatting?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_2809"} +{"question": "What studies are about input coordinate embeddings being utilized in implicit neural representations?", "answer": ["Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2006.10739", "2201.05989"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_2810"} +{"question": "What papers worked with RALMs aiming primarily to reduce hallucination and increase verifiability using the generation citations?", "answer": ["ASQA: Factoid Questions Meet Long-Form Answers", "QAMPARI: An Open-domain Question Answering Benchmark for Questions with\n Many Answers from Multiple Paragraphs", "ELI5: Long Form Question Answering"], "answer_arxiv_id": ["2204.06092", "2205.12665", "1907.09190"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_2811"} +{"question": "What papers incorporated the realization of image-point alignment by projecting point clouds to 2D depth images in 3D?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP"], "answer_arxiv_id": ["2112.02413"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_2812"} +{"question": "What papers explored 8 bit training for convolutional neural networks?", "answer": ["Training Deep Neural Networks with 8-bit Floating Point Numbers", "Towards Unified INT8 Training for Convolutional Neural Network", "DKM: Differentiable K-Means Clustering Layer for Neural Network Compression"], "answer_arxiv_id": ["1812.08011", "1912.12607", "2108.12659v4"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_2813"} +{"question": "Could you provide some references where 3D convolutions on a 4D cost volume were used in building stereo matching networks?", "answer": ["End-to-End Learning of Geometry and Context for Deep Stereo Regression", "Pyramid Stereo Matching Network", "Learning Depth with Convolutional Spatial Propagation Network", "Adaptive Unimodal Cost Volume Filtering for Deep Stereo Matching", "GA-Net: Guided Aggregation Net for End-to-end Stereo Matching", "Attention Concatenation Volume for Accurate and Efficient Stereo\n Matching"], "answer_arxiv_id": ["1703.04309", "1803.08669", "1810.02695", "1909.03751", "1904.06587", "2203.02146"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_2814"} +{"question": "Can you name the studies that connected object-centric representation learning with unsupervised contrastive learning?", "answer": ["InfoSeg: Unsupervised Semantic Image Segmentation with Mutual Information Maximization", "Self-Supervised Visual Representation Learning with Semantic Grouping", "Rethinking Alignment and Uniformity in Unsupervised Image Semantic Segmentation"], "answer_arxiv_id": ["2110.03477", "2205.15288", "2211.14513"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_2815"} +{"question": "Could you provide me some works that further developed attribution methods?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based\n Localization", "Axiomatic Attribution for Deep Networks", "SmoothGrad: removing noise by adding noise"], "answer_arxiv_id": ["1610.02391", "1703.01365", "1706.03825"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_2816"} +{"question": "Which paper optimized the darknet backbone structure used in YOLO series?", "answer": ["YOLOv4: Optimal Speed and Accuracy of Object Detection"], "answer_arxiv_id": ["2004.10934"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_2817"} +{"question": "Can you list some research papers that built databases of whole-body interactions with daily objects and promoted human-object interactions?", "answer": ["A Whole-Body Pose Taxonomy for Loco-Manipulation Tasks", "GRAB: A Dataset of Whole-Body Human Grasping of Objects", "SAGA: Stochastic Whole-Body Grasping with Contact", "TOCH: Spatio-Temporal Object-to-Hand Correspondence for Motion\n Refinement", "GOAL: Generating 4D Whole-Body Motion for Hand-Object Grasping"], "answer_arxiv_id": ["1503.06839", "2008.11200", "2112.10103", "2205.07982", "2112.11454"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_2818"} +{"question": "Could you provide the references for recent developments in image generation and contrastive learning methodologies?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Improved Denoising Diffusion Probabilistic Models", "Taming Transformers for High-Resolution Image Synthesis", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2105.05233", "2102.09672", "2012.09841", "2002.05709"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_2819"} +{"question": "What paper uses the self-training paradigm and corresponding unlabeled data to bridge the gap in adversarial robustness's sample complexity?", "answer": ["Unlabeled Data Improves Adversarial Robustness"], "answer_arxiv_id": ["1905.13736"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_2820"} +{"question": "What works have been conducted previously that propose an onboarding procedure involving participants describing different regions of the data space where AI made significant mistakes?", "answer": ["Teaching Humans When To Defer to a Classifier via Exemplars"], "answer_arxiv_id": ["2111.11297"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_2821"} +{"question": "Which studies introduced a dataset for generating descriptions of laptops and TVs?", "answer": ["Multi-domain Neural Network Language Generation for Spoken Dialogue\n Systems"], "answer_arxiv_id": ["1603.01232"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_2822"} +{"question": "Who conducted research on the width necessary for a ReLU network to fit any continuous real-valued function?", "answer": ["The Expressive Power of Neural Networks: A View from the Width", "Approximating Continuous Functions by ReLU Nets of Minimal Width"], "answer_arxiv_id": ["1709.02540", "1710.11278"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_2823"} +{"question": "Which papers explored gradient flow with particular parametrization converging to min-ℓ1-norm solution when using small initialization and min-ℓ2-norm solution when using large initialization?", "answer": ["Kernel and Rich Regimes in Overparametrized Models", "On the Implicit Bias of Initialization Shape: Beyond Infinitesimal Mirror Descent", "A unifying view on implicit bias in training linear neural networks"], "answer_arxiv_id": ["2002.09277", "2102.09769", "2010.02501"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_2824"} +{"question": "What studies address the update of factual information causing temporal misalignment?", "answer": ["Dynamic Benchmarking of Masked Language Models on Temporal Concept Drift\n with Multiple Views", "Towards Continual Knowledge Learning of Language Models", "TemporalWiki: A Lifelong Benchmark for Training and Evaluating\n Ever-Evolving Language Models", "Mind the Gap: Assessing Temporal Generalization in Neural Language\n Models", "Time-Aware Language Models as Temporal Knowledge Bases", "StreamingQA: A Benchmark for Adaptation to New Knowledge over Time in\n Question Answering Models"], "answer_arxiv_id": ["2302.12297", "2110.03215", "2204.14211", "2102.01951", "2106.15110", "2205.11388"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_2825"} +{"question": "Can you list studies that deal with skinning and rigging implicit surface representations?", "answer": ["ARAH: Animatable Volume Rendering of Articulated Human SDFs", "TAVA: Template-free Animatable Volumetric Actors"], "answer_arxiv_id": ["2210.10036v1", "2206.08929"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_2826"} +{"question": "Which study used sequential LSTM models for fusing hand and object features?", "answer": ["Joint Hand-object 3D Reconstruction from a Single Image with\n Cross-branch Feature Fusion"], "answer_arxiv_id": ["2006.15561"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_2827"} +{"question": "Which studies have been motivated by transforming conventional RL as a supervised learning problem?", "answer": ["Training Agents using Upside-Down Reinforcement Learning", "Planning from Pixels using Inverse Dynamics Models", "Masked Autoencoding for Scalable and Generalizable Decision Making", "Uni​[MASK]: Unified Inference in Sequential Decision Problems", "Decision Transformer: Reinforcement Learning via Sequence Modeling", "In-context Reinforcement Learning with Algorithm Distillation"], "answer_arxiv_id": ["1912.02877", "2012.02419", "2211.12740", "2211.10869", "2106.01345", "2210.14215"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2828"} +{"question": "Which studies made efforts towards depth prior-free few-shot optimization?", "answer": ["Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis", "RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from Sparse Inputs", "InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Rendering"], "answer_arxiv_id": ["2104.00677", "2112.00724", "2112.15399"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_2829"} +{"question": "Is there research applying DMs for 3D molecule generation for applications such as target drug generation and protein design?", "answer": ["Equivariant Diffusion for Molecule Generation in 3D", "Diffusion-based Molecule Generationwith Informative Prior Bridges", "Geometric Latent Diffusion Models for 3D Molecule Generation", "DiffBP: Generative Diffusion of 3D Molecules for Target Protein Binding", "Protein Structure and Sequence Generation with Equivariant Denoising Diffusion Probabilistic Models", "Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem"], "answer_arxiv_id": ["2203.17003", "2209.00865", "2305.01140", "2211.11214", "2205.15019", "2206.04119"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2830"} +{"question": "Which work proposed incorporating an exploration component in evaluations and adding a crossover operator to recombine hyperparameter vectors?", "answer": ["Regularized Evolutionary Population-Based Training"], "answer_arxiv_id": ["2002.04225"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_2831"} +{"question": "What studies are there about empirical exploration of forgetting?", "answer": ["Amnesiac Machine Learning"], "answer_arxiv_id": ["2010.10981"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_2832"} +{"question": "Which works proposed to adapt multiple shooting (MS) to neural-network-based models and large data regimes?", "answer": ["Learning Dynamical Systems from Noisy Sensor Measurements using Multiple Shooting"], "answer_arxiv_id": ["2106.11712"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_2833"} +{"question": "Can you inform me about the works that employ CKA to compare vision transformers with convolutional neural networks?", "answer": ["Do Vision Transformers See Like Convolutional Neural Networks?"], "answer_arxiv_id": ["2108.08810"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_2834"} +{"question": "Can you name some studies that investigated symmetries admitted by different ReLU networks?", "answer": ["Model Reconstruction from Model Explanations", "Reverse-Engineering Deep ReLU Networks", "Cryptanalytic Extraction of Neural Network Models"], "answer_arxiv_id": ["1807.05185", "1910.00744", "2003.04884"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_2835"} +{"question": "Could you give me references where a pre-trained visual model was adapted to different domains by adding a few prompt parameters?", "answer": ["Visual Prompt Tuning", "LPT: Long-tailed Prompt Tuning for Image Classification", "Exploring Visual Prompts for Adapting Large-Scale Models", "E^2VPT: An Effective and Efficient Approach for Visual Prompt Tuning"], "answer_arxiv_id": ["2203.12119", "2210.01033", "2203.17274", "2307.13770"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_2836"} +{"question": "Which studies examined the best-iterate convergence under ρ-star-negative comonotonicity and L-Lipschitzness?", "answer": ["Efficient Methods for Structured Nonconvex-Nonconcave Min-Max Optimization", "Escaping limit cycles: Global convergence for constrained nonconvex-nonconcave minimax problems", "Solving Nonconvex-Nonconcave Min-Max Problems exhibiting Weak Minty Solutions"], "answer_arxiv_id": ["2011.00364", "2302.09831", "2201.12247"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_2837"} +{"question": "Could you tell me which research identified the identifiability issue in unsupervised disentangled representation learning?", "answer": ["Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations"], "answer_arxiv_id": ["1811.12359"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2838"} +{"question": "Are there any papers introducing controllable generative frameworks for instantiation?", "answer": ["Penguins Don't Fly: Reasoning about Generics through Instantiations and\n Exceptions"], "answer_arxiv_id": ["2205.11658"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_2839"} +{"question": "Which research papers developed the so-called foundation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "PaLM-E: An Embodied Multimodal Language Model", "GPT-4 Technical Report"], "answer_arxiv_id": ["2103.00020", "2204.14198", "2301.12597", "2303.03378", "2303.08774"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_train_2840"} +{"question": "Which studies focused on generative models for producing high-quality 2D images of clothed humans?", "answer": ["HumanGAN: A Generative Model of Humans Images", "StyleGAN-Human: A Data-Centric Odyssey of Human Generation"], "answer_arxiv_id": ["2103.06902", "2204.11823"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_2841"} +{"question": "Which works used parametric fitting techniques on the edges of CAD models to recover their geometric features?", "answer": ["PIE-NET: Parametric Inference of Point Cloud Edges", "PC2WF: 3D Wireframe Reconstruction from Raw Point Clouds", "SepicNet: Sharp Edges Recovery by Parametric Inference of Curves in 3D\n Shapes", "SHARP Challenge 2023: Solving CAD History and pArameters Recovery from\n Point clouds and 3D scans. Overview, Datasets, Metrics, and Baselines"], "answer_arxiv_id": ["2007.04883", "2103.02766", "2304.06531", "2308.15966"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2842"} +{"question": "Could you provide some studies that advocate multiple rounds of full-shot decoding together with an encoder for intra-antibody context, and a separate encoder for external interactions?", "answer": ["Conditional Antibody Design as 3D Equivariant Graph Translation"], "answer_arxiv_id": ["2208.06073"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2843"} +{"question": "Could you provide me some studies that did not find empirical support for the claim that IF approximate the Leave-Some-Out Retraining effect on the loss in deep neural networks?", "answer": ["Influence Functions in Deep Learning Are Fragile", "Revisiting Methods for Finding Influential Examples"], "answer_arxiv_id": ["2006.14651", "2111.04683"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2844"} +{"question": "Could you list some papers that dealt with embedding pixel-aligned feature vectors from technologies such as LSeg or DINO into NeRF frameworks?", "answer": ["Decomposing NeRF for Editing via Feature Field Distillation", "NeRF-SOS: Any-View Self-supervised Object Segmentation on Complex Scenes", "LERF: Language Embedded Radiance Fields", "Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D\n Image Representations"], "answer_arxiv_id": ["2205.15585", "2209.08776", "2303.09553", "2209.03494"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_2845"} +{"question": "Which works have focused on developing L2O's generalization ability for different optimization tasks?", "answer": ["Hyperparameter Tuning is All You Need for LISTA"], "answer_arxiv_id": ["2110.15900"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_2846"} +{"question": "Could you tell me some studies that have improved 3D point representations by exploiting pixel-point alignments for distillation or contrastive learning?", "answer": ["Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Data", "CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP"], "answer_arxiv_id": ["2203.16258", "2301.04926"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_2847"} +{"question": "Can you provide any research papers that discussed unsupervised methods for text editing?", "answer": ["CGMH: Constrained Sentence Generation by Metropolis-Hastings Sampling", "Unsupervised Paraphrasing by Simulated Annealing"], "answer_arxiv_id": ["1811.10996", "1909.03588"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_2848"} +{"question": "What papers have proposed and used WikiSplit, a corpus created for the Split and Rephrase task?", "answer": ["Learning To Split and Rephrase From Wikipedia Edit History"], "answer_arxiv_id": ["1808.09468"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_2849"} +{"question": "What papers covered the domain adaptation for vision tasks?", "answer": ["Domain Adaptive Faster R-CNN for Object Detection in the Wild", "Moment Matching for Multi-Source Domain Adaptation", "ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation"], "answer_arxiv_id": ["1803.03243", "1812.01754", "1811.12833"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_2850"} +{"question": "Could you provide me some research papers on bridging modality gap for E2E ST?", "answer": ["MAESTRO: Matched Speech Text Representations through Modality Matching", "End-to-End Speech Translation with Knowledge Distillation", "Bridging the Gap between Pre-Training and Fine-Tuning for End-to-End Speech Translation", "Improving Speech Translation by Understanding and Learning from the Auxiliary Text Translation Task", "STEMM: Self-learning with Speech-text Manifold Mixup for Speech Translation", "Cross-modal Contrastive Learning for Speech Translation", "WACO: Word-Aligned Contrastive Learning for Speech Translation", "Fused Acoustic and Text Encoding for Multimodal Bilingual Pretraining and Speech Translation", "Unified Speech-Text Pre-training for Speech Translation and Recognition", "SpeechUT: Bridging Speech and Text with Hidden-Unit for Encoder-Decoder Based Speech-Text Pre-training", "mSLAM: Massively multilingual joint pre-training for speech and text", "Mu2SLAM: Multitask, Multilingual Speech and Language Models", "MAESTRO: Matched Speech Text Representations through Modality Matching", "Google USM: Scaling Automatic Speech Recognition Beyond 100 Languages"], "answer_arxiv_id": ["2204.03409", "1904.08075", "1909.07575", "2107.05782", "2203.10426", "2205.02444", "2212.09359", "2102.05766", "2204.05409", "2210.03730", "2202.01374", "2212.09553", "2204.03409", "2303.01037"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_2851"} +{"question": "What studies utilized the hints method in the process of paraphrase data collection?", "answer": ["Directed Diversity: Leveraging Language Embedding Distances for Collective Creativity in Crowd Ideation"], "answer_arxiv_id": ["2101.06030v1"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_2852"} +{"question": "What work provides a computational complexity and communication complexity analysis for DPOSG?", "answer": ["A Decentralized Parallel Algorithm for Training Generative Adversarial Nets"], "answer_arxiv_id": ["1910.12999"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_2853"} +{"question": "Do you know any studies that proposed the use of embedding of video and language to guide each other?", "answer": ["CLIP-Adapter: Better Vision-Language Models with Feature Adapters"], "answer_arxiv_id": ["2110.04544"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_2854"} +{"question": "Could you provide me some studies that focus on landmark detection for facial analysis?", "answer": ["Unsupervised Learning of Object Landmarks through Conditional Image\n Generation", "Sparse Local Patch Transformer for Robust Face Alignment and Landmarks\n Inherent Relation Learning", "Learning Spatial-Temporal Implicit Neural Representations for\n Event-Guided Video Super-Resolution"], "answer_arxiv_id": ["1806.07823", "2203.06541", "2303.13767"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_2855"} +{"question": "What works extended the approach of iterative pose refinement through incremental updates?", "answer": ["DeepIM: Deep Iterative Matching for 6D Pose Estimation", "CosyPose: Consistent multi-view multi-object 6D pose estimation"], "answer_arxiv_id": ["1804.00175", "2008.08465"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_2856"} +{"question": "What studies focus on large-scale text pre-training on attention-based models for vision-language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["1810.04805", "2205.01068"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_2857"} +{"question": "Could you provide some research works where Wikipedia edits were used as domain-specific feedback?", "answer": ["PEER: A Collaborative Language Model"], "answer_arxiv_id": ["2208.11663"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_2858"} +{"question": "What studies use contrastive learning to align visual features with language representations in Vision-language models?", "answer": ["UNIMO: Towards Unified-Modal Understanding and Generation via Cross-Modal Contrastive Learning", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2012.15409", "2103.00020"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_2859"} +{"question": "Could you provide me some studies about infilling in language models?", "answer": ["Enabling Language Models to Fill in the Blanks", "Efficient Training of Language Models to Fill in the Middle"], "answer_arxiv_id": ["2005.05339", "2207.14255"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_2860"} +{"question": "What studies propose using a Gaussian variational posterior or the reparameterization trick in the field of Variational Inference?", "answer": ["Auto-Encoding Variational Bayes", "Variational Dropout and the Local Reparameterization Trick"], "answer_arxiv_id": ["1312.6114", "1506.02557"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_2861"} +{"question": "Can you specify the studies investigated robotic applications as test beds for constraint inference?", "answer": ["Constrained Inverse Optimal Control with Application to a Human Manipulation Task", "Learning Parametric Constraints in High Dimensions from Demonstrations", "Uncertainty-Aware Constraint Learning for Adaptive Safe Motion Planning from Demonstrations", "Inverse Constrained Reinforcement Learning"], "answer_arxiv_id": ["1812.11600", "1910.03477", "2011.04141", "2011.09999"], "source_meta": {"published_time": "20220620"}, "qid": "AutoScholarQuery_train_2862"} +{"question": "Which works proposed methods to reduce communication costs in FL through data compression techniques such as quantization, sketching, and selective parameter sending?", "answer": ["Tuning nanoscale adhesive contact behavior to a near ideal Hertzian state via graphene coverage", "Adaptive Communication Strategies to Achieve the Best Error-Runtime Trade-off in Local-Update SGD", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Cost-Effective Federated Learning Design", "Faster Adaptive Federated Learning", "Federated Learning: Strategies for Improving Communication Efficiency", "Towards Federated Learning at Scale: System Design", "Communication Efficient Federated Learning with Adaptive Quantization", "FedHM: Efficient Federated Learning for Heterogeneous Models via Low-rank Factorization", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "Communication-efficient Distributed SGD with Sketching", "SplitFed: When Federated Learning Meets Split Learning", "Adaptive Gradient Sparsification for Efficient Federated Learning: An Online Learning Approach", "Federated Select: A Primitive for Communication- and Memory-Efficient Federated Learning"], "answer_arxiv_id": ["2011.00705v2", "1810.08313", "1910.06378", "2012.08336", "2212.00974", "1610.05492", "1902.01046v2", "2104.06023", "2111.14655", "1610.02132", "1903.04488", "2004.12088", "2001.04756", "2208.09432"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_2863"} +{"question": "Could you give me examples of studies that employ contrastive learning methodologies for representation learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "1911.05722"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_2864"} +{"question": "What projects have been proposed recently to allow users to control the generated image using a variety of modalities?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_2865"} +{"question": "What are the studies that tried to remove the simulation components in Riemannian generative models?", "answer": ["Moser Flow: Divergence-based Generative Modeling on Manifolds", "Matching Normalizing Flows and Probability Paths on Manifolds"], "answer_arxiv_id": ["2108.08052", "2207.04711"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_2866"} +{"question": "In what studies did authors develop multi-modal LLMs in the video modality?", "answer": ["Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "Valley: Video Assistant with Large Language model Enhanced abilitY"], "answer_arxiv_id": ["2306.02858", "2306.07207"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_2867"} +{"question": "Which work proposes sending distances between data points and centroids to the server in Federated Clustering?", "answer": ["Secure Federated Clustering"], "answer_arxiv_id": ["2205.15564"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_2868"} +{"question": "What is the cited paper in which the authors propose SCONES to recover the entire conditional distribution of the OT plan?", "answer": ["Score-based Generative Neural Networks for Large-Scale Optimal Transport"], "answer_arxiv_id": ["2110.03237"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_2869"} +{"question": "Could you provide me some studies that tried to solve fairness under various distribution shifts?", "answer": ["Robust Fairness under Covariate Shift", "Transferring Fairness under Distribution Shifts via Fair Consistency Regularization", "Fairness Violations and Mitigation under Covariate Shift"], "answer_arxiv_id": ["2010.05166", "2206.12796", "1911.00677"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_2870"} +{"question": "Which paper proposed the state-of-the-art video-based method for human motion capture?", "answer": ["Visibility Aware Human-Object Interaction Tracking from Single RGB\n Camera"], "answer_arxiv_id": ["2303.16479"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_2871"} +{"question": "What studies obtained lower bounds for learning depth-333 networks with Gaussian inputs using statistical query (SQ) algorithms?", "answer": ["Hardness of Noise-Free Learning for Two-Hidden-Layer Neural Networks"], "answer_arxiv_id": ["2202.05258"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_2872"} +{"question": "Could you provide me some studies that focused on the notion of disagreement?", "answer": ["Let’s Agree to Agree: Neural Networks Share Classification Order on Real Datasets", "Detecting Errors and Estimating Accuracy on Unlabeled Data with Self-training Ensembles", "Assessing Generalization of SGD via Disagreement", "Distributional Generalization: A New Kind of Generalization", "Agreement-on-the-Line: Predicting the Performance of Neural Networks under Distribution Shift", "Task Discovery: Finding the Tasks that Neural Networks Generalize on", "When Deep Classifiers Agree: Analyzing Correlations between Learning Order and Image Statistics"], "answer_arxiv_id": ["1905.10854", "2106.15728", "2106.13799", "2009.08092", "2206.13089", "2212.00261", "2105.08997"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_2873"} +{"question": "Which studies use LLMs like GPT-4 to assess the response quality of MLLMs in subjective evaluations?", "answer": ["Visual Instruction Tuning", "TouchStone: Evaluating Vision-Language Models by Language Models", "MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities"], "answer_arxiv_id": ["2304.08485", "2308.16890", "2308.02490"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_2874"} +{"question": "Could you tell me which papers introduced a CLIP-induced score to regularizes the generated process of simCTG in text-only-training zero-shot IC methods?", "answer": ["Language Models Can See: Plugging Visual Controls in Text Generation", "ZeroGen: Zero-shot Multimodal Controllable Text Generation with Multiple\n Oracles"], "answer_arxiv_id": ["2205.02655", "2306.16649"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_2875"} +{"question": "What work is most similar to the research in terms of proposition of a Bayesian joint model for the sequences of treatments and outcomes?", "answer": ["Personalized Dynamic Treatment Regimes in Continuous Time: A Bayesian Approach for Optimizing Clinical Decisions with Timing"], "answer_arxiv_id": ["2007.04155"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_2876"} +{"question": "What studies have found the success of in-context learning mostly depends on training distribution, prompt text structure, and label examples?", "answer": ["Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?", "Data Distributional Properties Drive Emergent In-Context Learning in Transformers"], "answer_arxiv_id": ["2202.12837", "2205.05055"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_2877"} +{"question": "What are the works that learned a mapping between two domains for IL with domain shift?", "answer": ["Learn what matters: cross-domain imitation learning with task-relevant embeddings"], "answer_arxiv_id": ["2209.12093"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_2878"} +{"question": "What are some studies that propose using options for decision-making in hierarchical RL?", "answer": ["The Option-Critic Architecture", "Compositional Planning Using Optimal Option Models"], "answer_arxiv_id": ["1609.05140", "1206.6473"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2879"} +{"question": "Which papers have used 2D CNNs for action recognition?", "answer": ["Temporal Segment Networks for Action Recognition in Videos", "Temporal Relational Reasoning in Videos", "TSM: Temporal Shift Module for Efficient Video Understanding", "Gate-Shift-Fuse for Video Action Recognition", "MoViNets: Mobile Video Networks for Efficient Video Recognition"], "answer_arxiv_id": ["1705.02953", "1711.08496", "1811.08383", "2203.08897", "2103.11511"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_2880"} +{"question": "Which work shifts the study focus from standard accuracy to robustness in the context of ConvNeXt?", "answer": ["A ConvNet for the 2020s"], "answer_arxiv_id": ["2201.03545"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_2881"} +{"question": "Which research proposes the Go-explore method in the area of exploration in RL?", "answer": ["First return, then explore"], "answer_arxiv_id": ["2004.12919"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_2882"} +{"question": "What papers studied the trade-off between accuracy and robustness?", "answer": ["A Closer Look at Accuracy vs. Robustness"], "answer_arxiv_id": ["2003.02460"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_2883"} +{"question": "Could you provide me a study showing how hierarchical Semantic IDs can improve model generalization?", "answer": ["Better Generalization with Semantic IDs: A Case study in Ranking for Recommendations"], "answer_arxiv_id": ["2306.08121"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_2884"} +{"question": "Which studies have done text-to-image synthesis using diffusion-based models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2885"} +{"question": "With respect to the diffusion model, can you provide some studies about utilizing a diffusion prior in image super-resolution?", "answer": ["Image Super-Resolution via Iterative Refinement", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic\n Models", "Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "Exploiting Diffusion Prior for Real-World Image Super-Resolution"], "answer_arxiv_id": ["2104.07636", "2104.14951", "2201.11793", "2212.00490", "2305.07015"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_2886"} +{"question": "Could you provide me with some research papers that developed visual LLMs?", "answer": ["Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2304.08485", "2304.10592"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_2887"} +{"question": "Which work proposed the scheme that is a key comparison for the researcher’s own work?", "answer": ["Distributed Prioritized Experience Replay"], "answer_arxiv_id": ["1803.00933"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_2888"} +{"question": "Any works about the ExGrad method as an approach to OOD detection?", "answer": ["How Useful are Gradients for OOD Detection Really?"], "answer_arxiv_id": ["2205.10439"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_2889"} +{"question": "Which research works developed the three-stage detection and tracking framework for tackling VQL tasks?", "answer": ["Know Your Surroundings: Exploiting Scene Information for Object Tracking", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2003.11014", "2110.07058"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2890"} +{"question": "Which research works use vision-language joint embedding space such as CLIP, or text-conditional generative models, like a text-to-image diffusion model, for measuring similarity as losses in 3D asset generation?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2103.00020", "2112.10752", "2205.11487"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_2891"} +{"question": "What is the work that accomplished one-shot human avatar reconstruction with 5 stages of VQA caption?", "answer": ["TeCH: Text-guided Reconstruction of Lifelike Clothed Humans"], "answer_arxiv_id": ["2308.08545"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_2892"} +{"question": "Can you give examples of research studying how pruning impacts neural network generalization?", "answer": ["The Generalization-Stability Tradeoff In Neural Network Pruning", "Pruning’s Effect on Generalization Through the Lens of Training and Regularization"], "answer_arxiv_id": ["1906.03728", "2210.13738"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_2893"} +{"question": "What works propose variants of CNN aiming for equivariance for Lie groups?", "answer": ["Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data"], "answer_arxiv_id": ["2002.12880"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_2894"} +{"question": "What are some optimization-based methods that model keypoint locations as multi-modal probability distributions?", "answer": ["The Alignment of the Spheres: Globally-Optimal Spherical Mixture Alignment for Camera Pose Estimation"], "answer_arxiv_id": ["1812.01232v2"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_2895"} +{"question": "What research papers suggested penalties that approximate the l0 penalty?", "answer": ["A unified approach to model selection and sparse recovery using regularized least squares"], "answer_arxiv_id": ["0905.3573v2"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_2896"} +{"question": "In which publications did they first develop systems based on a discrete diffusion model for text-to-audio aims?", "answer": ["Diffsound: Discrete Diffusion Model for Text-to-sound Generation", "Vector Quantized Diffusion Model for Text-to-Image Synthesis"], "answer_arxiv_id": ["2207.09983", "2111.14822"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_2897"} +{"question": "Which methods in the field of neural-symbolic methods require the elaborate design of functions to be used in corresponding task and the calibration of corresponding neural modules?", "answer": ["Neural Module Networks", "Learning to Reason: End-to-End Module Networks for Visual Question Answering", "Neural Modular Control for Embodied Question Answering", "Neural Module Networks for Reasoning over Text", "Text Modular Networks: Learning to Decompose Tasks in the Language of Existing Models", "Break It Down: A Question Understanding Benchmark"], "answer_arxiv_id": ["1511.02799v4", "1704.05526", "1810.11181", "1912.04971", "2009.00751", "2001.11770"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_2898"} +{"question": "Which works utilized mixture learning with variational objectives outside the VAE literature?", "answer": ["Automatic Differentiation Variational Inference", "Thompson Sampling via Local Uncertainty"], "answer_arxiv_id": ["1603.00788", "1910.13673"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2899"} +{"question": "Could you provide research examples for the use of these algorithms in modern applications like movie recommendation system?", "answer": ["Reconfiguration Problems on Submodular Functions"], "answer_arxiv_id": ["2111.14030"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_2900"} +{"question": "Which papers developed variants of Neural Radiance Fields (NeRF)?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw\n Images", "RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs", "Robust e-NeRF: NeRF from Sparse & Noisy Events under Non-Uniform Motion", "F$^{2}$-NeRF: Fast Neural Radiance Field Training with Free Camera\n Trajectories", "NoPe-NeRF: Optimising Neural Radiance Field with No Pose Prior", "NeRF-DS: Neural Radiance Fields for Dynamic Specular Objects"], "answer_arxiv_id": ["2103.13415", "2111.12077", "2111.13679", "2112.00724", "2309.08596", "2303.15951", "2212.07388", "2303.14435"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_2901"} +{"question": "What work proposed Datamodels that fit a model to predict behavior given a subset of training data?", "answer": ["Datamodels: Predicting Predictions from Training Data"], "answer_arxiv_id": ["2202.00622"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_2902"} +{"question": "Can you refer papers that investigated the potential of Transformer on visual tasks in supervised learning?", "answer": ["Training data-efficient image transformers & distillation through attention", "Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions", "Scaling Vision Transformers", "Going deeper with Image Transformers", "Go Wider Instead of Deeper"], "answer_arxiv_id": ["2012.12877", "2101.11986", "2103.14030", "2102.12122", "2106.04560", "2103.17239", "2107.11817"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_2903"} +{"question": "Are there any studies applying KD for lip reading?", "answer": ["ASR is all you need: cross-modal distillation for lip reading"], "answer_arxiv_id": ["1911.12747"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_2904"} +{"question": "What studies have found that Large Language Models (LLMs) can store real-world facts learned during pre-training?", "answer": ["Discovering Knowledge-Critical Subnetworks in Pretrained Language Models", "Quantifying Memorization Across Neural Language Models", "Extracting Training Data from Large Language Models", "Locating and Editing Factual Associations in GPT", "A Primer in BERTology: What We Know About How BERT Works"], "answer_arxiv_id": ["2310.03084", "2202.07646", "2012.07805", "2202.05262", "2002.12327"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_2905"} +{"question": "Which works applied Reinforcement Learning practices in medical image landmark detection?", "answer": ["Partial Policy-based Reinforcement Learning for Anatomical Landmark\n Localization in 3D Medical Images"], "answer_arxiv_id": ["1807.02908"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_2906"} +{"question": "Could you specify studies that show convergence guarantees for normalizing flows?", "answer": ["Gaussianization Flows", "Universal Approximation Property of Neural Ordinary Differential Equations", "Representational aspects of depth and conditioning in normalizing flows"], "answer_arxiv_id": ["2003.01941", "2012.02414", "2010.01155"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_2907"} +{"question": "What paper involves a neural sensor simulator for multi-sensor inputs into a shared implicit field in NeRF?", "answer": ["UniSim: A Neural Closed-Loop Sensor Simulator"], "answer_arxiv_id": ["2308.01898"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_2908"} +{"question": "What works considered the challenging setting of linear MDPs?", "answer": ["Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1907.05388"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_2909"} +{"question": "Could you provide me some works discussing how in-context learning (ICL) is used to learn certain function classes using synthetic data?", "answer": ["What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "What learning algorithm is in-context learning? Investigations with linear models", "Transformers Learn In-Context by Gradient Descent"], "answer_arxiv_id": ["2208.01066", "2211.15661", "2212.07677"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_2910"} +{"question": "What works are about finetuning models on large datasets of human-written instructions?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2109.01652", "2110.08207", "2202.01279", "2203.02155"], "source_meta": {"published_time": "20220824"}, "qid": "AutoScholarQuery_train_2911"} +{"question": "Any researches that concentrate on improving the network initialization?", "answer": ["ReZero is All You Need: Fast Convergence at Large Depth", "GradInit: Learning to Initialize Neural Networks for Stable and Efficient Training"], "answer_arxiv_id": ["2003.04887", "2102.08098"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_2912"} +{"question": "Which work uses the generative pairwise evaluator, JM, to select the response that reflects human-requested feedback?", "answer": ["Training Language Models with Language Feedback at Scale"], "answer_arxiv_id": ["2303.16755"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_2913"} +{"question": "Any works about the usage of Focal Loss for imbalanced datasets in model calibration?", "answer": ["Calibrating Deep Neural Networks using Focal Loss"], "answer_arxiv_id": ["2002.09437"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_2914"} +{"question": "Can you provide me any works that applied soft prompt learning to the image recognition task?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2915"} +{"question": "Can you provide studies that implemented retrieval-augmented methods for code generation?", "answer": ["Retrieval-Based Neural Code Generation", "Retrieval Augmented Code Generation and Summarization", "ReACC: A Retrieval-Augmented Code Completion Framework", "DocPrompting: Generating Code by Retrieving the Docs", "Repository-Level Prompt Generation for Large Language Models of Code", "RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation", "CoCoMIC: Code Completion By Jointly Modeling In-file and Cross-file Context"], "answer_arxiv_id": ["1808.10025", "2108.11601", "2203.07722", "2207.05987", "2206.12839", "2303.12570", "2212.10007"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_2916"} +{"question": "Are there any works that perform image synthesis conditioned on depth maps of a given mesh?", "answer": ["MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion"], "answer_arxiv_id": ["2307.01097"], "source_meta": {"published_time": "20240626"}, "qid": "AutoScholarQuery_train_2917"} +{"question": "What are some studies on controllable material generation?", "answer": ["MaterialGAN: Reflectance Capture using a Generative SVBRDF Model", "Controlling Material Appearance by Examples", "TileGen: Tileable, Controllable Material Generation and Capture"], "answer_arxiv_id": ["2010.00114", "2206.14970", "2206.05649"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_2918"} +{"question": "Are there any works where Image-conditioned diffusion models have been used for tasks such as image enhancement, harmonization and translation?", "answer": ["SRDiff: Single Image Super-Resolution with Diffusion Probabilistic\n Models", "Denoising Diffusion Restoration Models", "Multiscale Structure Guided Diffusion for Image Deblurring", "JPEG Artifact Correction using Denoising Diffusion Restoration Models", "Zero-Shot Image Harmonization with Generative Model Prior", "TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Composition", "Painterly Image Harmonization using Diffusion Model", "Adding Conditional Control to Text-to-Image Diffusion Models", "UNIT-DDPM: UNpaired Image Translation with Denoising Diffusion\n Probabilistic Models", "EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic\n Differential Equations", "Pretraining is All You Need for Image-to-Image Translation", "Diffusion-based Image Translation using Disentangled Style and Content\n Representation", "The Blessing of Randomness: SDE Beats ODE in General Diffusion-based\n Image Editing"], "answer_arxiv_id": ["2104.14951", "2201.11793", "2212.01789", "2209.11888", "2307.08182", "2307.12493", "2308.02228", "2302.05543", "2104.05358", "2207.06635", "2205.12952", "2209.15264", "2311.01410"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_2919"} +{"question": "What papers have used Implicit Neural Spatial Representation for image processing?", "answer": ["Learning Continuous Image Representation with Local Implicit Image Function", "Spatially-Adaptive Pixelwise Networks for Fast Image Translation"], "answer_arxiv_id": ["2012.09161", "2012.02992"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_2920"} +{"question": "Which research papers provide certifiable robustness guarantees for 1-Lipschitz neural networks?", "answer": ["Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation", "Lightweight Lipschitz Margin Training for Certified Defense against Adversarial Examples"], "answer_arxiv_id": ["1705.08475", "1811.08080"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_2921"} +{"question": "What work introduces ControlNet method which uses a zero-initialized ResNet block to piggyback a pre-trained diffusion model?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_2922"} +{"question": "What papers propose to refine the text prompts for better 3D generation?", "answer": ["Re-imagine the Negative Prompt Algorithm: Transform 2D Diffusion into\n 3D, alleviate Janus problem and Beyond"], "answer_arxiv_id": ["2304.04968"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_2923"} +{"question": "Could you provide me some research papers about the implementation of a single Transformer network in VL pre-training models, eliminating the need for additional branches?", "answer": ["VisualBERT: A Simple and Performant Baseline for Vision and Language", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "UNITER: UNiversal Image-TExt Representation Learning"], "answer_arxiv_id": ["1908.03557", "1908.08530", "1909.11740"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_2924"} +{"question": "Could you provide me some works dedicated to speech-to-text translation tasks?", "answer": ["CoVoST: A Diverse Multilingual Speech-To-Text Translation Corpus", "CoVoST 2 and Massively Multilingual Speech-to-Text Translation", "CVSS Corpus and Massively Multilingual Speech-to-Speech Translation"], "answer_arxiv_id": ["2002.01320", "2007.10310", "2201.03713"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_2925"} +{"question": "What piece of research improved the results in uniform stability of SGD and obtained tight guarantees for the stability of SGD?", "answer": ["Stability of SGD: Tightness Analysis and Improved Bounds"], "answer_arxiv_id": ["2102.05274"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_2926"} +{"question": "What paper proposes sub-updates projection to vocabulary space?", "answer": ["Transformer Feed-Forward Layers Build Predictions by Promoting Concepts\n in the Vocabulary Space"], "answer_arxiv_id": ["2203.14680"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_2927"} +{"question": "Which work developed the ConceptNet, a graph of concepts connected by relations for reasoning in language models?", "answer": ["ConceptNet 5.5: An Open Multilingual Graph of General Knowledge"], "answer_arxiv_id": ["1612.03975"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_2928"} +{"question": "What studies labeled lanes attributes of continuity and direction on a massive amount of data?", "answer": ["BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning"], "answer_arxiv_id": ["1805.04687"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_2929"} +{"question": "Could you provide me some works that describe transformer-based models for computer vision tasks?", "answer": ["End-to-End Object Detection with Transformers", "Gold-YOLO: Efficient Object Detector via Gather-and-Distribute Mechanism"], "answer_arxiv_id": ["2005.12872", "2309.11331"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_2930"} +{"question": "What were the studies that extend the method of using improvement methods from bib.bib9 to the Capacitated Vehicle Routing Problem (CVRP) ?", "answer": ["Learning Improvement Heuristics for Solving Routing Problems"], "answer_arxiv_id": ["1912.05784"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_2931"} +{"question": "What research papers are about utilizing prompt generators comprising multiple components, including cross-attention layers, groups of learnable parameters, and linear layers?", "answer": ["When Prompt-based Incremental Learning Does Not Meet Strong Pretraining"], "answer_arxiv_id": ["2308.10445"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_2932"} +{"question": "Could you provide me some research that involved the weighted regression technique in PPO algorithms?", "answer": ["Near-optimal Policy Optimization Algorithms for Learning Adversarial Linear Mixture MDPs", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes"], "answer_arxiv_id": ["2102.08940v2", "2012.08507"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_2933"} +{"question": "Could you mention any research that focuses on improving the inductive bias of MLPs by systematically sparsifying them?", "answer": ["Towards Learning Convolutions from Scratch"], "answer_arxiv_id": ["2007.13657"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_2934"} +{"question": "What studies utilize operations that are insensitive to sampling intervals for processing irregularly sampled data?", "answer": ["Set Functions for Time Series", "Graph-Guided Network for Irregularly Sampled Multivariate Time Series"], "answer_arxiv_id": ["1909.12064", "2110.05357"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_2935"} +{"question": "What supervised and unsupervised methods have evolved regarding deep learning in primitive decomposition?", "answer": ["3D-PRNN: Generating Shape Primitives with Recurrent Neural Networks", "Supervised Fitting of Geometric Primitives to 3D Point Clouds", "Learning Shape Abstractions by Assembling Volumetric Primitives", "Superquadrics Revisited: Learning 3D Shape Parsing beyond Cuboids", "Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networks"], "answer_arxiv_id": ["1708.01648", "1811.08988", "1612.00404", "1904.09970", "2103.10429v1"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_2936"} +{"question": "Which papers use a scorer model to add attribute information to guide an unconditional language model?", "answer": ["Plug and Play Language Models: a Simple Approach to Controlled Text Generation"], "answer_arxiv_id": ["1912.02164"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_2937"} +{"question": "Could you provide some studies that highlighted the robustness of machine-generated text detectors?", "answer": ["CoCo: Coherence-Enhanced Machine-Generated Text Detection Under Data\n Limitation With Contrastive Learning", "Paraphrasing evades detectors of AI-generated text, but retrieval is an\n effective defense", "RADAR: Robust AI-Text Detection via Adversarial Learning", "Assaying on the Robustness of Zero-Shot Machine-Generated Text Detectors"], "answer_arxiv_id": ["2212.10341", "2303.13408", "2307.03838", "2312.12918"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_2938"} +{"question": "Are there any studies that investigate how properties of the data distribution can reduce the tension between group fairness and model performance?", "answer": ["Recovering from Biased Data: Can Fairness Constraints Improve Accuracy?", "Fairness Evaluation in Presence of Biased Noisy Labels", "Robust Optimization for Fairness with Noisy Protected Groups", "Mitigating Bias in Set Selection with Noisy Protected Attributes", "Assessing Fairness in the Presence of Missing Data", "Fairness for Unobserved Characteristics: Insights from Technological Impacts on Queer Communities", "Measurement and Fairness", "Assessing Algorithmic Fairness with Unobserved Protected Class Using Data Combination", "Fairness without Imputation: A Decision Tree Approach for Fair Prediction with Missing Values", "To Split or Not to Split: The Impact of Disparate Treatment in Classification", "Why Is My Classifier Discriminatory?"], "answer_arxiv_id": ["1912.01094", "2003.13808", "2002.09343", "2011.04219v2", "2112.04899", "2102.04257", "1912.05511", "1906.00285", "2109.10431", "2002.04788", "1805.12002"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_2939"} +{"question": "Which work adapted the idea of low-switching to tabular RL and formalized the switching cost as a secondary metric?", "answer": ["Provably Efficient Q-Learning with Low Switching Cost"], "answer_arxiv_id": ["1905.12849"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_2940"} +{"question": "Are there any research works discussing multi-camera Bird’s-Eye-View (BEV) projection for generating 3D occupancy predictions?", "answer": ["Lift, Splat, Shoot: Encoding Images from Arbitrary Camera Rigs by Implicitly Unprojecting to 3D", "BEVDet: High-Performance Multi-Camera 3D Object Detection in Bird-Eye-View", "Cross-view Transformers for real-time Map-view Semantic Segmentation", "BEVFormer: Learning Bird’s-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers", "LaRa: Latents and Rays for Multi-Camera Bird’s-Eye-View Semantic Segmentation", "BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detection"], "answer_arxiv_id": ["2008.05711", "2112.11790", "2205.02833", "2203.17270", "2206.13294", "2206.10092"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_2941"} +{"question": "What works utilized multi-plane images for scene representation?", "answer": ["DeepView: View Synthesis with Learned Gradient Descent", "Layered Neural Rendering for Retiming People in Video", "Local Light Field Fusion: Practical View Synthesis with Prescriptive Sampling Guidelines", "Pushing the Boundaries of View Extrapolation with Multiplane Images", "Stereo Magnification: Learning view synthesis using multiplane images"], "answer_arxiv_id": ["1906.07316v1", "2009.07833", "1905.00889", "1905.00413", "1805.09817"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_2942"} +{"question": "Could you provide me some studies that have used various priors to achieve accurate colorization semantics?", "answer": ["Pixel-level Semantics Guided Image Colorization", "Pixelated Semantic Colorization", "Instance-aware Image Colorization", "Colorful Image Colorization", "BigColor: Colorization using a Generative Color Prior for Natural Images", "Towards Vivid and Diverse Image Colorization with Generative Color Prior"], "answer_arxiv_id": ["1808.01597", "1901.10889", "2005.10825", "1603.08511", "2207.09685", "2108.08826"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_2943"} +{"question": "Which studies argued on how neural fields circumvent the memory-expressivity dilemma?", "answer": ["On the Effectiveness of Weight-Encoded Neural Implicit 3D Shapes"], "answer_arxiv_id": ["2009.09808"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_2944"} +{"question": "Could you mention a study which defined heat equation on Finsler manifolds?", "answer": ["Heat flow on Finsler manifolds"], "answer_arxiv_id": ["0808.1166"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_2945"} +{"question": "What research proposes the method for open vocabulary image classification and semantic segmentation?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2103.00020", "2112.01071"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_2946"} +{"question": "Which studies discovered that the internal states of multilingual LMs can be categorized into language-sensitive and language-agnostic components?", "answer": ["What does it mean to be language-agnostic? Probing multilingual sentence\n encoders for typological properties", "The Geometry of Multilingual Language Model Representations", "Inducing Language-Agnostic Multilingual Representations", "First Align, then Predict: Understanding the Cross-Lingual Ability of\n Multilingual BERT"], "answer_arxiv_id": ["2009.12862", "2205.10964", "2008.09112", "2101.11109"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_2947"} +{"question": "Who was the first to introduce 3D-LLM by adapting LLMs to 3D data?", "answer": ["3D-LLM: Injecting the 3D World into Large Language Models"], "answer_arxiv_id": ["2307.12981"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2948"} +{"question": "Could you provide me some works that encode the target concept using learned embeddings in text-to-image generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "Enhancing Detail Preservation for Customized Text-to-Image Generation: A\n Regularization-Free Approach", "PhotoVerse: Tuning-Free Image Customization with Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2208.01618", "2302.12228", "2305.14720", "2305.13579", "2309.05793"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_2949"} +{"question": "In what papers do discrete-time methods that apply static graph learning methods on each snapshot of a dynamic graph feature?", "answer": ["EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs", "dyngraph2vec: Capturing Network Dynamics using Dynamic Graph Representation Learning", "ROLAND: Graph Learning Framework for Dynamic Graphs"], "answer_arxiv_id": ["1902.10191", "1809.02657", "2208.07239"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_2950"} +{"question": "Has anyone proposed a hierarchical generative network-based approach for crafting targeted transfer-based adversarial examples in a 2D domain?", "answer": ["Boosting Transferability of Targeted Adversarial Examples via\n Hierarchical Generative Networks"], "answer_arxiv_id": ["2107.01809"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_2951"} +{"question": "Which papers explore the PAC-Bayes theory?", "answer": ["PAC-Bayesian Theory Meets Bayesian Inference", "Meta-Learning Reliable Priors in the Function Space", "PACOH: Bayes-Optimal Meta-Learning with PAC-Guarantees"], "answer_arxiv_id": ["1605.08636", "2106.03195", "2002.05551"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_2952"} +{"question": "Which study first conducted MIA on convolution neural networks and used a distributional prior through GAN?", "answer": ["The Secret Revealer: Generative Model-Inversion Attacks Against Deep Neural Networks"], "answer_arxiv_id": ["1911.07135"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2953"} +{"question": "What studies proposed precision and recall variants to separately evaluate fidelity and diversity aspects of generated samples?", "answer": ["Assessing Generative Models via Precision and Recall", "Improved Precision and Recall Metric for Assessing Generative Models"], "answer_arxiv_id": ["1806.00035", "1904.06991"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_2954"} +{"question": "What are some examples of research that solve multiple tasks through joint hidden representation in multi-task learning (MTL)?", "answer": ["Instance-aware Semantic Segmentation via Multi-task Network Cascades", "Learning to Push by Grasping: Using multiple tasks for effective learning", "End-to-End Multi-Task Learning with Attention"], "answer_arxiv_id": ["1512.04412", "1609.09025", "1803.10704"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_2955"} +{"question": "What studies have leveraged game theory in the context of algorithm-curated platforms?", "answer": ["Modeling Content Creator Incentives on Algorithm-Curated Platforms", "Supply-Side Equilibria in Recommender Systems", "From Recommendation Systems to Facility Location Games", "Learning from a Learning User for Optimal Recommendations", "Learning the Optimal Recommendation from Explorative Users"], "answer_arxiv_id": ["2206.13102", "2206.13489", "1809.02931", "2202.01879", "2110.03068"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2956"} +{"question": "What study introduces the original concept of the shortcut learning problem?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_2957"} +{"question": "Which work enabled multi-scale neural rendering of large scenes with a progressive optimization scheme?", "answer": ["BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale\n Scene Rendering"], "answer_arxiv_id": ["2112.05504"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_2958"} +{"question": "Can you mention some works that use TDA-based constructions to infer properties of underlying data manifolds?", "answer": ["Intrinsic persistent homology via density-based metric learning"], "answer_arxiv_id": ["2012.07621"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_2959"} +{"question": "What studies in Natural Language Processing introduced extensive range of data augmentation techniques for diverse applications?", "answer": ["An Empirical Survey of Data Augmentation for Limited Data Learning in\n NLP", "Data Augmentation for Low-Resource Neural Machine Translation", "Data Noising as Smoothing in Neural Network Language Models", "Unsupervised Machine Translation Using Monolingual Corpora Only", "Sequence-Level Mixed Sample Data Augmentation", "SwitchOut: an Efficient Data Augmentation Algorithm for Neural Machine\n Translation"], "answer_arxiv_id": ["2106.07499", "1705.00440", "1703.02573", "1711.00043", "2011.09039", "1808.07512"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_2960"} +{"question": "Any works about leveraging a dataset with structural scene graph supervision?", "answer": ["Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs"], "answer_arxiv_id": ["2305.06343"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2961"} +{"question": "Which works use a pre-trained visual encoder to encode the egocentric images into feature vectors for object goal navigation?", "answer": ["Deep Residual Learning for Image Recognition", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1512.03385", "2103.00020"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_2962"} +{"question": "Can you list the works that have been researching controllable text-to-image diffusion through pre-trained models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "GLIGEN: Open-Set Grounded Text-to-Image Generation", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion Models", "Sketch-Guided Text-to-Image Diffusion Models", "SpaText: Spatio-Textual Representation for Controllable Image Generation"], "answer_arxiv_id": ["2302.05543", "2301.07093", "2302.08453", "2305.16322", "2211.13752", "2211.14305"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2963"} +{"question": "What works have done research on extending normalizing flows to Riemannian manifolds?", "answer": ["Normalizing Flows on Tori and Spheres", "Riemannian Continuous Normalizing Flows", "Smooth Normalizing Flows", "Neural Spline Flows"], "answer_arxiv_id": ["2002.02428", "2006.10605", "2110.00351", "1906.04032"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_2964"} +{"question": "What works have been conducted to improve the efficiency of image matching methods?", "answer": ["Learning to Match Features with Seeded Graph Matching Network", "ClusterGNN: Cluster-based Coarse-to-Fine Graph Neural Network for\n Efficient Feature Matching", "LightGlue: Local Feature Matching at Light Speed"], "answer_arxiv_id": ["2108.08771", "2204.11700", "2306.13643"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_2965"} +{"question": "What research used GAN inversion technique for different purposes in few-shot GDA?", "answer": ["Inverting The Generator Of A Generative Adversarial Network", "Designing an Encoder for StyleGAN Image Manipulation", "Towards Diverse and Faithful One-shot Adaption of Generative Adversarial Networks", "Mind the Gap: Domain Gap Control for Single Shot Domain Adaptation for Generative Adversarial Networks", "Few Shot Generative Model Adaption via Relaxed Spatial Structural Alignment"], "answer_arxiv_id": ["1802.05701", "2102.02766", "2207.08736", "2110.08398", "2203.04121"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_2966"} +{"question": "What papers focused on the parameter optimization for two-layer ReLU neural networks?", "answer": ["Neural Networks are Convex Regularizers: Exact Polynomial-time Convex Optimization Formulations for Two-layer Networks"], "answer_arxiv_id": ["2002.10553"], "source_meta": {"published_time": "20221114"}, "qid": "AutoScholarQuery_train_2967"} +{"question": "Can you cite examples of studies that approached 3D and 4D human parsing by rendering multi-view images and projecting parsing labels onto 3D meshes?", "answer": ["SIZER: A Dataset and Model for Parsing 3D Clothing and Learning Size\n Sensitive 3D Clothing", "Multi-Garment Net: Learning to Dress 3D People from Images"], "answer_arxiv_id": ["2007.11610", "1908.06903"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_2968"} +{"question": "What publications follow the SPOS NAS scheme for Transformer Architecture Search (TAS)?", "answer": ["Single Path One-Shot Neural Architecture Search with Uniform Sampling"], "answer_arxiv_id": ["1904.00420"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_2969"} +{"question": "Which studies establish the correspondence between visual objects and their corresponding sounds by contrastive learning in Audio-Visual Sound Localization?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Big Self-Supervised Models are Strong Semi-Supervised Learners"], "answer_arxiv_id": ["2002.05709", "2006.10029"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_2970"} +{"question": "Which works are about Vision and Language Models (VLMs)?", "answer": ["VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "MomentDiff: Generative Video Moment Retrieval from Random to Real"], "answer_arxiv_id": ["1908.08530", "2201.12086", "2307.02869"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_2971"} +{"question": "In what paper was the score-based generative model proposed?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_2972"} +{"question": "Could you point out some works on simulating humans with language models based on personas?", "answer": ["Social Simulacra: Creating Populated Prototypes for Social Computing Systems", "Generative Agents: Interactive Simulacra of Human Behavior", "Out of One, Many: Using Language Models to Simulate Human Samples"], "answer_arxiv_id": ["2208.04024", "2304.03442", "2209.06899"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_2973"} +{"question": "Which paper established a novel connection between diffusion model and denoising scoring matching?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_2974"} +{"question": "Which works proposed adversarial robustness through the use of project gradient descent or randomized smoothing?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Certified Adversarial Robustness via Randomized Smoothing"], "answer_arxiv_id": ["1706.06083", "1902.02918"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_2975"} +{"question": "Which prior studies show understanding on how to balance policy improvement and constraint?", "answer": ["Hundreds Guide Millions: Adaptive Offline Reinforcement Learning with Expert Guidance", "Boosting Offline Reinforcement Learning with Action Preference Query"], "answer_arxiv_id": ["2309.01448", "2306.03362"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_2976"} +{"question": "What publication defined the system that convolves with truncated Legendre polynomials?", "answer": ["HiPPO: Recurrent Memory with Optimal Polynomial Projections"], "answer_arxiv_id": ["2008.07669"], "source_meta": {"published_time": "20220624"}, "qid": "AutoScholarQuery_train_2977"} +{"question": "Are there any works that propose optimization-free methods for stylized image generation?", "answer": ["Composer: Creative and Controllable Image Synthesis with Composable\n Conditions", "StyleAdapter: A Single-Pass LoRA-Free Model for Stylized Image\n Generation", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.09778", "2309.01770", "2302.08453"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_2978"} +{"question": "Can you name some studies that have proposed various methods to enhance IPS for more stabilizing learning?", "answer": ["Counterfactual Risk Minimization: Learning from Logged Bandit Feedback"], "answer_arxiv_id": ["1502.02362"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_2979"} +{"question": "Could you provide me some works focus on solving empirical risk minimization problems in a (ε,δ)-DP compliant manner?", "answer": ["Differentially private Riemannian optimization"], "answer_arxiv_id": ["2205.09494"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_2980"} +{"question": "Could you provide me some significant studies that worked on latent space to increase the resolution of generated images in the field of diffusion-based image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_2981"} +{"question": "Which studies have developed self-supervised learning losses considering equivariance?", "answer": ["Equivariant Imaging: Learning Beyond the Range Space"], "answer_arxiv_id": ["2103.14756"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_2982"} +{"question": "Which studies proposed ways minimize the worst-case training loss over a set of groups for OOD generalization?", "answer": ["Distributionally Robust Optimization with Probabilistic Group"], "answer_arxiv_id": ["2303.05809"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_2983"} +{"question": "What studies present techniques like Chain of Thought (CoT) and Tree of Thought (ToT) that enhance the reasoning capabilities of LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2201.11903", "2305.10601"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_2984"} +{"question": "What papers incorporate a pre-trained depth-aware image diffusion model to synthesize high-resolution partial textures from multiple viewpoints progressively?", "answer": ["TEXTure: Text-Guided Texturing of 3D Shapes", "Text2Tex: Text-driven Texture Synthesis via Diffusion Models"], "answer_arxiv_id": ["2302.01721", "2303.11396"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_2985"} +{"question": "Which studies have used compiler backend outputs to ground Code-LMs?", "answer": ["SelfEvolve: A Code Evolution Framework via Large Language Models", "Teaching Large Language Models to Self-Debug", "CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing", "RLTF: Reinforcement Learning from Unit Test Feedback"], "answer_arxiv_id": ["2306.02907", "2304.05128v2", "2305.11738v4", "2307.04349"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_2986"} +{"question": "Which studies focused on the frameworks specialized for dynamic neural network training?", "answer": ["DyNet: The Dynamic Neural Network Toolkit", "Deep Learning with Dynamic Computation Graphs"], "answer_arxiv_id": ["1701.03980", "1702.02181"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_2987"} +{"question": "Which paper first discussed the 'harmlessness' requirement for Dataset Ownership Verification?", "answer": ["Untargeted Backdoor Watermark: Towards Harmless and Stealthy Dataset Copyright Protection"], "answer_arxiv_id": ["2210.00875"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_2988"} +{"question": "What papers proposed the augmenting of classical data structures with predictions?", "answer": ["The Case for Learned Index Structures", "A Model for Learned Bloom Filters, and Optimizing by Sandwiching", "Secretary and Online Matching Problems with Machine Learned Advice", "Secretaries with Advice", "A Universal Error Measure for Input Predictions Applied to Online Graph Problems", "Learning-Augmented Algorithms for Online TSP on the Line", "The Primal-Dual method for Learning Augmented Algorithms"], "answer_arxiv_id": ["1712.01208", "1901.00902", "2006.01026", "2011.06726", "2205.12850v2", "2206.00655v1", "2010.11632"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2989"} +{"question": "What research papers have adopted Faster-RCNN to extract image information into object features for VLMs?", "answer": ["VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "UNITER: UNiversal Image-TExt Representation Learning"], "answer_arxiv_id": ["1908.08530", "1908.02265", "1909.11740"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_2990"} +{"question": "Which works have constructed dialogue data for users by promoting crowd-workers to author dialogues based on specific personas?", "answer": ["Personalizing Dialogue Agents: I have a dog, do you have pets too?"], "answer_arxiv_id": ["1801.07243"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_2991"} +{"question": "Could you provide me some studies about transport mapping problems?", "answer": ["The Perception-Distortion Tradeoff"], "answer_arxiv_id": ["1711.06077"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_2992"} +{"question": "Are there any previous works that leverage in-context learning to correct wrongly recognized tokens in hypotheses?", "answer": ["Can Whisper perform speech-based in-context learning?"], "answer_arxiv_id": ["2309.07081"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_2993"} +{"question": "What works aimed to generalize equivariance and focus on continuous groups?", "answer": ["Oriented Response Networks"], "answer_arxiv_id": ["1701.01833"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_2994"} +{"question": "Which works have proposed methods for hypergraph clustering?", "answer": ["Clustering in graphs and hypergraphs with categorical edge labels", "Parameterized Correlation Clustering in Hypergraphs and Bipartite Graphs"], "answer_arxiv_id": ["1910.09943", "2002.09460"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_2995"} +{"question": "What research work is associated with the release of the FACET dataset containing annotations about person-related attributes and objects?", "answer": ["FACET: Fairness in Computer Vision Evaluation Benchmark"], "answer_arxiv_id": ["2309.00035"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_2996"} +{"question": "Could you provide an example of a significant advancement in the field of vision-language models?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2301.12597"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_2997"} +{"question": "What studies on time series processing used recurrent GNNs?", "answer": ["Structured Sequence Modeling with Graph Convolutional Recurrent Networks", "Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting"], "answer_arxiv_id": ["1612.07659", "1707.01926v3"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_2998"} +{"question": "Could you provide me some studies in the clustering domain that have advocated for deep clustering methods?", "answer": ["Unsupervised Deep Embedding for Clustering Analysis", "SCAN: Learning to Classify Images without Labels"], "answer_arxiv_id": ["1511.06335", "2005.12320"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_2999"} +{"question": "What works present methods for prompt generation, scoring or paraphrasing in discrete prompt search?", "answer": ["Making Pre-trained Language Models Better Few-shot Learners", "Commonsense Knowledge Mining from Pretrained Models", "How Can We Know What Language Models Know?", "BARTScore: Evaluating Generated Text as Text Generation"], "answer_arxiv_id": ["2012.15723", "1909.00505", "1911.12543", "2106.11520"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_3000"} +{"question": "Could you name the studies that propose Riemannian optimization-based training models for low-rank training?", "answer": ["Low-rank lottery tickets: finding efficient low-rank neural networks via matrix differential equations"], "answer_arxiv_id": ["2205.13571"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_3001"} +{"question": "What works combined GNN and Transformer to incorporate graph structural information?", "answer": ["Representing Long-Range Context for Graph Neural Networks with Global Attention", "Structure-Aware Transformer for Graph Representation Learning", "Recipe for a General, Powerful, Scalable Graph Transformer"], "answer_arxiv_id": ["2201.08821", "2202.03036", "2205.12454"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_3002"} +{"question": "What studies are related to the integration of ViTs for dense predictions in medical imaging domain?", "answer": ["UNETR: Transformers for 3D Medical Image Segmentation", "TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation", "nnFormer: Interleaved Transformer for Volumetric Segmentation", "TransBTS: Multimodal Brain Tumor Segmentation Using Transformer"], "answer_arxiv_id": ["2103.10504", "2102.04306", "2109.03201v6", "2103.04430"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_3003"} +{"question": "Which work proposes an approach to overcome the conflicting gradients problem in MTL?", "answer": ["Gradient Surgery for Multi-Task Learning"], "answer_arxiv_id": ["2001.06782"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_3004"} +{"question": "What works demonstrate deep linear networks improve the training and performance of deep nonlinear networks?", "answer": ["The Low-Rank Simplicity Bias in Deep Networks", "ExpandNets: Linear Over-parameterization to Train Compact Convolutional Networks", "On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization"], "answer_arxiv_id": ["2103.10427", "1811.10495", "1802.06509"], "source_meta": {"published_time": "20230101"}, "qid": "AutoScholarQuery_train_3005"} +{"question": "Can you indicate methods proposed for aligning DNNs with primate vision, beyond the neural harmonizer?", "answer": ["Harmonizing the object recognition strategies of deep neural networks with humans"], "answer_arxiv_id": ["2211.04533"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3006"} +{"question": "Could you provide me with some researches about learning-based obfuscation for privacy-preserving action recognition?", "answer": ["SPAct: Self-supervised Privacy Preservation for Action Recognition", "Privacy-Preserving Deep Action Recognition: An Adversarial Learning Framework and A New Dataset"], "answer_arxiv_id": ["2203.15205", "1906.05675"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3007"} +{"question": "Which works proposed deep learning based scene text detectors in regression based category?", "answer": ["Detecting Oriented Text in Natural Images by Linking Segments", "TextBoxes: A Fast Text Detector with a Single Deep Neural Network", "EAST: An Efficient and Accurate Scene Text Detector", "Arbitrary-Oriented Scene Text Detection via Rotation Proposals", "MOST: A Multi-Oriented Scene Text Detector with Localization Refinement", "Few Could Be Better Than All: Feature Sampling and Grouping for Scene\n Text Detection"], "answer_arxiv_id": ["1703.06520", "1611.06779", "1704.03155", "1703.01086", "2104.01070", "2203.15221"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_3008"} +{"question": "Can you name some research about using encoder-decoder architectures in self-supervised skeleton-based action recognition?", "answer": ["PREDICT & CLUSTER: Unsupervised Skeleton Based Action Recognition", "Unsupervised Human Action Recognition with Skeletal Graph Laplacian and Self-Supervised Viewpoints Invariance", "Skeleton Cloud Colorization for Unsupervised 3D Action Representation Learning"], "answer_arxiv_id": ["1911.12409", "2204.10312", "2108.01959"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_3009"} +{"question": "What paper introduced Focal Loss that employs a soft weighting scheme to allocate higher weights to more challenging samples?", "answer": ["Focal Loss for Dense Object Detection"], "answer_arxiv_id": ["1708.02002"], "source_meta": {"published_time": "20240709"}, "qid": "AutoScholarQuery_train_3010"} +{"question": "Could you provide me some works that proposed offline RL methods based on estimating conservative q-values?", "answer": ["Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Conservative Q-Learning for Offline Reinforcement Learning", "AlgaeDICE: Policy Gradient from Arbitrary Experience", "Offline Reinforcement Learning with Implicit Q-Learning", "Off-Policy Deep Reinforcement Learning without Exploration", "OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation"], "answer_arxiv_id": ["1906.00949", "2006.04779", "1912.02074", "2110.06169", "1812.02900", "2106.10783"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_3011"} +{"question": "Which research argued that disentangled representations increase fairness of downstream prediction tasks?", "answer": ["On the Fairness of Disentangled Representations"], "answer_arxiv_id": ["1905.13662"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_3012"} +{"question": "Which works use lexical substitutions for embedding watermark messages into the text?", "answer": ["DeepTextMark: A Deep Learning-Driven Text Watermarking Approach for\n Identifying Large Language Model Generated Text", "Tracing Text Provenance via Context-Aware Lexical Substitution", "Robust Multi-bit Natural Language Watermarking through Invariant\n Features", "Watermarking Text Generated by Black-Box Language Models"], "answer_arxiv_id": ["2305.05773", "2112.07873", "2305.01904", "2305.08883v1"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_3013"} +{"question": "Can you provide the reference that showed MF-DP-FTRL subsumes and improves the DP-FTRL algorithm?", "answer": ["Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams"], "answer_arxiv_id": ["2202.08312"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3014"} +{"question": "What papers propose to directly regress poses for end-to-end Structure-from-Motion methods?", "answer": ["SfM-Net: Learning of Structure and Motion from Video", "Unsupervised Learning of Depth and Ego-Motion from Video", "DiffPoseNet: Direct Differentiable Camera Pose Estimation", "RelPose: Predicting Probabilistic Relative Rotation for Single Objects\n in the Wild"], "answer_arxiv_id": ["1704.07804", "1704.07813", "2203.11174", "2208.05963"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_3015"} +{"question": "Which papers discuss about the neurosymbolic perspective that combine aspects of deep learning and symbolic AI?", "answer": ["The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences from Natural Supervision", "Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding", "Neural Module Networks for Reasoning over Text", "Learning by Abstraction: The Neural State Machine"], "answer_arxiv_id": ["1904.12584", "1810.02338", "1912.04971", "1907.03950"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_3016"} +{"question": "Could you provide me some works that focus on synthetic data generated using generative models for various downstream tasks?", "answer": ["Generative Adversarial Networks", "EditGAN: High-Precision Semantic Image Editing", "HandsOff: Labeled Dataset Generation With No Additional Human Annotations", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["2203.00667", "2111.03186", "2212.12645", "1503.03585"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_3017"} +{"question": "Could you cite some research papers about LiDAR-based 3D detection?", "answer": ["PointPillars: Fast Encoders for Object Detection from Point Clouds", "Center-based 3D Object Detection and Tracking", "PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detection"], "answer_arxiv_id": ["1812.05784", "2006.11275", "2102.00463"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_3018"} +{"question": "Which works used masked language modeling in large-scale pre-training?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_3019"} +{"question": "Could you provide me papers that attempt to use powerful generative models such as Gaussians or VAEs for behavior modeling in offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning", "PLAS: Latent Action Space for Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "1906.00949", "1911.11361", "2011.07213"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_3020"} +{"question": "What work developed empathic psychological counseling platforms using LLMs’ APIs?", "answer": ["Human-AI Collaboration Enables More Empathic Conversations in Text-based\n Peer-to-Peer Mental Health Support"], "answer_arxiv_id": ["2203.15144"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_3021"} +{"question": "Any researches about voxel grid based approaches for scene reconstruction?", "answer": ["CoCoNets: Continuous Contrastive 3D Scene Representations", "Scene Representation Transformer: Geometry-Free Novel View Synthesis Through Set-Latent Scene Representations", "Equivariant Neural Rendering"], "answer_arxiv_id": ["2104.03851", "2111.13152", "2006.07630"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_3022"} +{"question": "What research papers have applied selfies for novel view synthesis?", "answer": ["Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "FLAME-in-NeRF : Neural control of Radiance Fields for Free View Face\n Animation", "CoNeRF: Controllable Neural Radiance Fields"], "answer_arxiv_id": ["2011.12948", "2106.13228v2", "2108.04913", "2112.01983"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_3023"} +{"question": "What are the previous studies proposing frameworks like Private-Heavy-UCBVI and Private-Heavy-UCBPO for private RL with bounded rewards?", "answer": ["Differentially Private Regret Minimization in Episodic Markov Decision Processes"], "answer_arxiv_id": ["2112.10599"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3024"} +{"question": "Which researches are on the topic of compositional human-scene interaction synthesis with high-level semantic control?", "answer": ["Compositional Human-Scene Interaction Synthesis with Semantic Control"], "answer_arxiv_id": ["2207.12824"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_3025"} +{"question": "Can you provide a study showing empirical evidence of the effectiveness of the low-frequency attack?", "answer": ["On the Effectiveness of Low Frequency Perturbations"], "answer_arxiv_id": ["1903.00073"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_3026"} +{"question": "Which research papers applied DRO in various fields such as adversarial training, long-tailed learning, and label shift?", "answer": ["Certifying Some Distributional Robustness with Principled Adversarial Training", "Distributional Robustness Loss for Long-tail Learning", "Coping with label shift via distributionally robust optimisation"], "answer_arxiv_id": ["1710.10571v5", "2104.03066", "2010.12230"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_3027"} +{"question": "Any studies addressing memory reduction during inference?", "answer": ["SqueezeNeRF: Further factorized FastNeRF for memory-efficient inference", "MERF: Memory-Efficient Radiance Fields for Real-time View Synthesis in\n Unbounded Scenes"], "answer_arxiv_id": ["2204.02585", "2302.12249"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_3028"} +{"question": "What works propose the use of deep generative models for shape generation tasks?", "answer": ["Generative Adversarial Networks", "Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["2203.00667", "1312.6114"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_3029"} +{"question": "Which work uses multi-scale hash grids to handle large-scale scenes?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_3030"} +{"question": "Which studies proposed solutions for weakly-supervised Human-object Interaction (HOI) detection?", "answer": ["PPR-FCN: Weakly Supervised Visual Relation Detection via Parallel Pairwise R-FCN", "Explanation-based Weakly-supervised Learning of Visual Relations with Graph Networks", "Detecting Human-Object Interaction with Mixed Supervision", "Human-Object Interaction Detection via Weak Supervision"], "answer_arxiv_id": ["1708.01956", "2006.09562", "2011.04971", "2112.00492v1"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_3031"} +{"question": "Which works improved LSVI-UCB to match the minimax lower bound?", "answer": ["Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes"], "answer_arxiv_id": ["2212.06132"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_3032"} +{"question": "What works have focused on the field of mechanistic interpretability?", "answer": ["In-context Learning and Induction Heads", "Progress measures for grokking via mechanistic interpretability"], "answer_arxiv_id": ["2209.11895v1", "2301.05217"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_3033"} +{"question": "Are there any works about the development of the approximate learning methods that are not identical to retrain-from-scratch, but are similar?", "answer": ["Certified Data Removal from Machine Learning Models", "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks"], "answer_arxiv_id": ["1911.03030", "1911.04933"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_3034"} +{"question": "Which studies attempted to incorporate flow-based models into the training process of EBMs?", "answer": ["Flow Contrastive Estimation of Energy-Based Models", "MCMC Should Mix: Learning Energy-Based Model with Neural Transport Latent Space MCMC", "A Tale of Two Flows: Cooperative Learning of Langevin Flow and Normalizing Flow Toward Energy-Based Model"], "answer_arxiv_id": ["1912.00589", "2006.06897", "2205.06924"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_3035"} +{"question": "Which paper shed light on theoretical understanding of hybrid scenario in RL?", "answer": ["Agnostic System Identification for Model-Based Reinforcement Learning", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning"], "answer_arxiv_id": ["1203.1007", "2106.04895"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_3036"} +{"question": "Could you provide me some works that propose methods to regularize the explanation models in learning from explanations?", "answer": ["Interpretations are Useful: Penalizing Explanations to Align Neural Networks with Prior Knowledge"], "answer_arxiv_id": ["1909.13584"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_3037"} +{"question": "What works have used a conditional variational auto-encoder (CVAE) to encode hand-object interaction?", "answer": ["Grasping Field: Learning Implicit Representations for Human Grasps", "A Skeleton-Driven Neural Occupancy Representation for Articulated Hands", "Grasping Field: Learning Implicit Representations for Human Grasps", "Contact2Grasp: 3D Grasp Synthesis via Hand-Object Contact Constraint"], "answer_arxiv_id": ["2008.04451", "2109.11399", "2008.04451", "2210.09245"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_3038"} +{"question": "Which works have attempted image-to-image translation?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "Adding Conditional Control to Text-to-Image Diffusion Models", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "StarGAN: Unified Generative Adversarial Networks for Multi-Domain\n Image-to-Image Translation", "Semantic Image Synthesis with Spatially-Adaptive Normalization", "High-Resolution Image Synthesis and Semantic Manipulation with\n Conditional GANs", "Cross-domain Correspondence Learning for Exemplar-based Image\n Translation", "CoCosNet v2: Full-Resolution Correspondence Learning for Image\n Translation", "Toward Multimodal Image-to-Image Translation", "Pre-Trained Image Processing Transformer", "Taming Transformers for High-Resolution Image Synthesis", "Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["1611.07004", "1703.10593", "2302.05543", "2211.09800", "1711.09020", "1903.07291", "1711.11585", "2004.05571", "2012.02047", "1711.11586", "2012.00364", "2012.09841", "2102.12092"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_3039"} +{"question": "What work combined the contrastive learning objective with negative log-likelihood in sequence-to-sequence models?", "answer": ["Alleviating Exposure Bias via Contrastive Learning for Abstractive Text Summarization"], "answer_arxiv_id": ["2108.11846"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_3040"} +{"question": "Can you tell me about a study that proposed an expander-based graph learning mechanism for message-passing networks?", "answer": ["Expander Graph Propagation"], "answer_arxiv_id": ["2210.02997"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_3041"} +{"question": "What studies have aimed to learn domain-invariant features by minimizing distinct style characteristics?", "answer": ["Domain Generalization via Shuffled Style Assembly for Face Anti-Spoofing", "Instance-Aware Domain Generalization for Face Anti-Spoofing", "Generative Domain Adaptation for Face Anti-Spoofing"], "answer_arxiv_id": ["2203.05340", "2304.05640", "2207.10015"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3042"} +{"question": "Is there any work about a range of methods for editing networks?", "answer": ["Rewriting a Deep Generative Model", "Rewriting Geometric Rules of a GAN", "Locating and Editing Factual Associations in GPT", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2007.15646", "2207.14288", "2202.05262", "2110.11309"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_3043"} +{"question": "What research papers focused on memory-based methods in CL?", "answer": ["Efficient Lifelong Learning with A-GEM", "Learning to Learn without Forgetting By Maximizing Transfer and Minimizing Interference", "Gradient-based Editing of Memory Examples for Online Task-free Continual Learning", "Orthogonal Gradient Descent for Continual Learning", "Gradient Projection Memory for Continual Learning", "TRGP: Trust Region Gradient Projection for Continual Learning"], "answer_arxiv_id": ["1812.00420", "1810.11910", "2006.15294", "1910.07104v1", "2103.09762", "2202.02931"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_3044"} +{"question": "Could you provide me the studies about a memory module in Modular RAG?", "answer": ["Training Data is More Valuable than You Think: A Simple and Effective\n Method by Retrieving from Training Data", "Lift Yourself Up: Retrieval-augmented Text Generation with Self Memory"], "answer_arxiv_id": ["2203.08773", "2305.02437"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_3045"} +{"question": "Can you provide studies about feature contrastive learning for instance information extraction in 3D panoptic segmentation?", "answer": ["Contrastive Lift: 3D Object Instance Segmentation by Slow-Fast\n Contrastive Fusion"], "answer_arxiv_id": ["2306.04633"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_3046"} +{"question": "Could you tell me about some studies on adversarial data collection?", "answer": ["Build it Break it Fix it for Dialogue Safety: Robustness from Adversarial Human Attack", "Adversarial NLI: A New Benchmark for Natural Language Understanding", "Dynabench: Rethinking Benchmarking in NLP", "DynaSent: A Dynamic Benchmark for Sentiment Analysis", "Analyzing Dynamic Adversarial Training Data in the Limit"], "answer_arxiv_id": ["1908.06083", "1910.14599", "2104.14337", "2012.15349", "2110.08514"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_3047"} +{"question": "Could you provide papers that differ mainly in the neighbor aggregation scheme of DGNs than the message passing paradigm?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "Inductive Representation Learning on Large Graphs", "Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering", "Strategies for Pre-training Graph Neural Networks"], "answer_arxiv_id": ["1609.02907", "1710.10903", "1706.02216", "1606.09375", "1905.12265"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_3048"} +{"question": "Could you point me to some works on offline knowledge distillation methods extended for CLIP multi-modal setup?", "answer": ["A Fast Knowledge Distillation Framework for Visual Recognition", "Re-labeling ImageNet: from Single to Multi-Labels, from Global to\n Localized Labels", "Reinforce Data, Multiply Impact: Improved Model Accuracy and Robustness\n with Dataset Reinforcement"], "answer_arxiv_id": ["2112.01528", "2101.05022", "2303.08983"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3049"} +{"question": "Which papers introduced multi-modal large language foundational models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "PaLM-E: An Embodied Multimodal Language Model", "PaLM: Scaling Language Modeling with Pathways", "Visual Instruction Tuning", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "End-to-End Object Detection with Transformers", "Language Is Not All You Need: Aligning Perception with Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2204.14198", "2303.03378", "2204.02311", "2304.08485", "2306.15195", "2301.12597", "2005.12872", "2302.14045", "2304.14178", "2304.10592"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_3050"} +{"question": "What are the studies that found a strong correlation between distribution distance metrics like Wasserstein Distance or Kullback–Leibler divergence and the perceptual quality of images?", "answer": ["The Perception-Distortion Tradeoff"], "answer_arxiv_id": ["1711.06077"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_3051"} +{"question": "What studies have explored the feasibility of using Natural Language Processing methods to guide the manipulation of Excel sheets?", "answer": ["SpreadsheetCoder: Formula Prediction from Semi-structured Context", "FLAME: A Small Language Model for Spreadsheet Formulas"], "answer_arxiv_id": ["2106.15339", "2301.13779"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3052"} +{"question": "What research employs Mathematical Flow Algorithm (MFA) to calculate losses efficiently in the context of CO problems?", "answer": ["Erdos Goes Neural: an Unsupervised Learning Framework for Combinatorial Optimization on Graphs", "DIMES: A Differentiable Meta Solver for Combinatorial Optimization Problems"], "answer_arxiv_id": ["2006.10643", "2210.04123"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_3053"} +{"question": "What paper proposes an adaptive language transformer that improves inference speed by dynamically selecting tokens?", "answer": ["TR-BERT: Dynamic Token Reduction for Accelerating BERT Inference"], "answer_arxiv_id": ["2105.11618"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_3054"} +{"question": "Which works were about enhancing non-autoregressive TTS with more powerful generative models?", "answer": ["Glow-TTS: A Generative Flow for Text-to-Speech via Monotonic Alignment Search", "PortaSpeech: Portable and High-Quality Generative Text-to-Speech"], "answer_arxiv_id": ["2005.11129", "2109.15166"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_3055"} +{"question": "Which works focused on detecting edges at multiple scales by combining features at different layers and in cascades for edge detection?", "answer": ["Holistically-Nested Edge Detection", "Richer Convolutional Features for Edge Detection", "Bi-Directional Cascade Network for Perceptual Edge Detection"], "answer_arxiv_id": ["1504.06375", "1612.02103", "1902.10903"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_3056"} +{"question": "Could you provide me some studies about cross-lingual alignment?", "answer": ["Cross-Lingual Alignment of Contextual Word Embeddings, with Applications\n to Zero-shot Dependency Parsing"], "answer_arxiv_id": ["1902.09492"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_3057"} +{"question": "Which papers explored the application of diffusion models in the field of photo-realistic image generation?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2112.10752"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_3058"} +{"question": "Which studies used network-based learnable eigenfunctions to produce spectral embeddings?", "answer": ["Learning Neural Eigenfunctions for Unsupervised Semantic Segmentation"], "answer_arxiv_id": ["2304.02841"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_3059"} +{"question": "Which paper demonstrated the use of high-quality data to pre-train smaller coding models with results comparable to larger models?", "answer": ["Textbooks Are All You Need"], "answer_arxiv_id": ["2306.11644"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_3060"} +{"question": "Can you recommend some studies in the field of facial recognition using Perturbative availability poison methods?", "answer": ["Fawkes: Protecting Privacy against Unauthorized Deep Learning Models", "LowKey: Leveraging Adversarial Attacks to Protect Social Media Users from Facial Recognition"], "answer_arxiv_id": ["2002.08327", "2101.07922"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_3061"} +{"question": "Could you give some papers about guided assets retrieval?", "answer": ["PSDR-Room: Single Photo to Scene using Differentiable Rendering"], "answer_arxiv_id": ["2307.03244"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_3062"} +{"question": "Which work was the first to derive an accelerated algorithm for decentralized optimization and link its convergence speed to the Laplacian eigengap?", "answer": ["Optimal algorithms for smooth and strongly convex distributed optimization in networks"], "answer_arxiv_id": ["1702.08704"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_3063"} +{"question": "Can you name some papers that discuss the application of knowledge distillation in model compression?", "answer": ["Variational Information Distillation for Knowledge Transfer", "Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1904.05835", "1503.02531"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_3064"} +{"question": "Which paper introduced Insta-SSF, an instance adaptive version of the scale-space flow (SSF) model?", "answer": ["Instance-Adaptive Video Compression: Improving Neural Codecs by Training\n on the Test Set"], "answer_arxiv_id": ["2111.10302"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3065"} +{"question": "Could you name some studies where spatial rewiring is common to access information beyond the 1-hop when updating node features?", "answer": ["MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing", "Diffusion Improves Graph Learning", "Supervised Community Detection with Line Graph Neural Networks", "Path Integral Based Convolution and Pooling for Graph Neural Networks", "Multi-hop Attention Graph Neural Network", "k-hop graph neural networks"], "answer_arxiv_id": ["1905.00067", "1911.05485", "1705.08415", "2006.16811", "2009.14332v5", "1907.06051"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3066"} +{"question": "Which studies are about leveraging feature attentions in knowledge distillation for dense prediction tasks?", "answer": ["Channel-wise Knowledge Distillation for Dense Prediction", "Focal and Global Knowledge Distillation for Detectors"], "answer_arxiv_id": ["2011.13256", "2111.11837"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_3067"} +{"question": "Which papers examine gender stereotypes in language models?", "answer": ["Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word\n Embeddings", "\"Kelly is a Warm Person, Joseph is a Role Model\": Gender Biases in\n LLM-Generated Reference Letters", "Marked Personas: Using Natural Language Prompts to Measure Stereotypes\n in Language Models", "Multi-Dimensional Gender Bias Classification"], "answer_arxiv_id": ["1607.06520", "2310.09219", "2305.18189", "2005.00614"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3068"} +{"question": "Which work established the concept of Neural Radiance Fields (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_3069"} +{"question": "What papers provide a background on reinforcement learning from human feedback?", "answer": ["Fine-Tuning Language Models from Human Preferences", "Learning New Skills after Deployment: Improving open-domain internet-driven dialogue with human feedback"], "answer_arxiv_id": ["1909.08593", "2208.03270"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_3070"} +{"question": "Could you provide me studies about planner-centric metrics for object detection systems?", "answer": ["Learning to Evaluate Perception Models Using Planner-Centric Metrics"], "answer_arxiv_id": ["2004.08745"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_3071"} +{"question": "Which paper introduces CLIP features into hash grid-represented 3D scenes?", "answer": ["LERF: Language Embedded Radiance Fields"], "answer_arxiv_id": ["2303.09553"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_3072"} +{"question": "What work is the source of MVSNet used in Point-NeRF?", "answer": ["MVSNet: Depth Inference for Unstructured Multi-view Stereo"], "answer_arxiv_id": ["1804.02505"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_3073"} +{"question": "What papers talk about the application of transformers in the field of 3D instance processing?", "answer": ["PCT: Point Cloud Transformer", "End-to-End Human Pose and Mesh Reconstruction with Transformers"], "answer_arxiv_id": ["2012.09688", "2012.09760"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_3074"} +{"question": "Can you provide some studies that have applied similar low-switching RL ideas with linear function approximation?", "answer": ["A Provably Efficient Algorithm for Linear Markov Decision Process with Low Switching Cost", "Provably Efficient Reinforcement Learning with Linear Function Approximation under Adaptivity Constraints", "Near-Optimal Deployment Efficiency in Reward-Free Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["2101.00494v1", "2101.02195", "2210.00701"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_3075"} +{"question": "Which research utilized the capabilities of GPT-3 to encourage curiosity-driven questioning among children?", "answer": ["GPT-3-driven pedagogical agents for training children's curious\n question-asking skills"], "answer_arxiv_id": ["2211.14228"], "source_meta": {"published_time": "20240820"}, "qid": "AutoScholarQuery_train_3076"} +{"question": "Which papers examine the development of reward models in the vision domain?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "VILA: Learning Image Aesthetics from User Comments with Vision-Language\n Pretraining", "ImageReward: Learning and Evaluating Human Preferences for Text-to-Image\n Generation", "Human Preference Score: Better Aligning Text-to-Image Models with Human\n Preference", "Human Preference Score v2: A Solid Benchmark for Evaluating Human\n Preferences of Text-to-Image Synthesis", "Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image\n Generation", "Aligning Text-to-Image Models using Human Feedback"], "answer_arxiv_id": ["2103.00020", "2201.12086", "2205.01917", "2303.14302", "2304.05977", "2303.14420", "2306.09341", "2305.01569", "2302.12192"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_3077"} +{"question": "Which papers primarily focus on model-based approaches for 3D face reconstruction?", "answer": ["Face2Face: Real-time Face Capture and Reenactment of RGB Videos"], "answer_arxiv_id": ["2007.14808"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_3078"} +{"question": "Which papers showcased advancements in AI-generated content?", "answer": ["Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Versatile Diffusion: Text, Images and Variations All in One Diffusion Model", "Scalable Diffusion Models with Transformers", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2102.12092", "2205.11487", "2112.10752", "2204.06125", "2211.08332v4", "2212.09748", "2204.02311", "2302.13971"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_3079"} +{"question": "What studies have targeted reducing the limitations of object detection backbones for multi-modal pre-training by applying grid features?", "answer": ["Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal Transformers", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "An Empirical Study of Training End-to-End Vision-and-Language Transformers"], "answer_arxiv_id": ["2004.00849", "1907.11692", "2111.02387"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_3080"} +{"question": "Which studies have managed to estimate contact from a single image using contact-based datasets?", "answer": ["Populating 3D Scenes by Learning Human-Scene Interaction", "Capturing and Inferring Dense Full-Body Human-Scene Contact", "Learning Complex 3D Human Self-Contact"], "answer_arxiv_id": ["2012.11581", "2206.09553", "2012.10366"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_3081"} +{"question": "Could you provide me some studies that obtain small-loss regret for linear-quadratic regulators (LQRs)?", "answer": ["Information Theoretic Regret Bounds for Online Nonlinear Control"], "answer_arxiv_id": ["2006.12466"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3082"} +{"question": "What studies extend the research beyond binary search to active binary classification?", "answer": ["The Geometry of Generalized Binary Search"], "answer_arxiv_id": ["0910.4397"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_3083"} +{"question": "Which papers discusses about understanding model behavior in the aspects of transparency?", "answer": ["Expanding Explainability: Towards Social Transparency in AI systems"], "answer_arxiv_id": ["2101.04719"], "source_meta": {"published_time": "20220928"}, "qid": "AutoScholarQuery_train_3084"} +{"question": "Could you list some studies that focus on simulating acoustic environments by modeling their 3D geometry and material properties?", "answer": ["SoundSpaces: Audio-Visual Navigation in 3D Environments", "SoundSpaces 2.0: A Simulation Platform for Visual-Acoustic Learning", "Scene-Aware Audio for 360° Videos", "Scene-Aware Audio Rendering via Deep Acoustic Analysis", "GWA: A Large High-Quality Acoustic Dataset for Audio Processing"], "answer_arxiv_id": ["1912.11474", "2206.08312", "1805.04792", "1911.06245", "2204.01787"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_3085"} +{"question": "Which work combines synthetic data augmentation with loss-based debiasing methods for mitigating spurious correlation?", "answer": ["From Fake to Real: Pretraining on Balanced Synthetic Images to Prevent\n Bias"], "answer_arxiv_id": ["2308.04553"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3086"} +{"question": "What papers are about generating large amounts of synthetic data for improving APR methods?", "answer": ["LENS: Localization enhanced by NeRF synthesis", "DFNet: Enhance Absolute Pose Regression with Direct Feature Matching"], "answer_arxiv_id": ["2110.06558", "2204.00559"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_3087"} +{"question": "Could you point me to some data-driven approaches for grasp synthesis?", "answer": ["Data-Driven Grasp Synthesis - A Survey"], "answer_arxiv_id": ["1309.2660"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_3088"} +{"question": "Can you cite studies that proposed techniques such as Auto-CoT for diverse and zero-shot prompting for CoT?", "answer": ["Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2210.03493"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_3089"} +{"question": "Could you provide me some studies about compression of the changes of model activations?", "answer": ["Fine-tuning Language Models over Slow Networks using Activation Quantization with Guarantees"], "answer_arxiv_id": ["2206.01299"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_3090"} +{"question": "What techniques are typically used to integrate features from all views in these models to perform tasks like zero-shot 3D semantic segmentation?", "answer": ["OpenScene: 3D Scene Understanding with Open Vocabularies", "ConceptFusion: Open-set Multimodal 3D Mapping", "Semantic Abstraction: Open-World 3D Scene Understanding from 2D\n Vision-Language Models"], "answer_arxiv_id": ["2211.15654", "2302.07241", "2207.11514"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_3091"} +{"question": "Which research papers concentrate on text-to-image generation models in the context of diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2205.11487", "2112.10741"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_3092"} +{"question": "Are there any papers that address the problem of learning equivariant distributions over trajectories in the context of multi-agent dynamics?", "answer": ["Probabilistic Symmetry for Multi-Agent Dynamics"], "answer_arxiv_id": ["2205.01927v3"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_3093"} +{"question": "What works propose dynamically increasing the attack steps for better convergence in adversarial training?", "answer": ["On the Convergence and Robustness of Adversarial Training", "When Adversarial Training Meets Vision Transformers: Recipes from Training to Architecture"], "answer_arxiv_id": ["2112.08304", "2210.07540"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_3094"} +{"question": "Could you show me some papers that provide tail bounds for bounded, sub-Gaussian, or sub-exponential distributions?", "answer": ["Estimation of Spectral Risk Measures", "A Wasserstein distance approach for concentration of empirical risk estimates"], "answer_arxiv_id": ["1912.10398", "1902.10709"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_3095"} +{"question": "Which studies suggest that learning priors using neural networks can be used for sampling?", "answer": ["Variational Inference with Normalizing Flows", "Deep Unsupervised Clustering with Gaussian Mixture Variational\n Autoencoders", "Variational Lossy Autoencoder", "PixelVAE: A Latent Variable Model for Natural Images", "Neural Discrete Representation Learning", "VAE with a VampPrior", "Generating Diverse High-Fidelity Images with VQ-VAE-2", "From Variational to Deterministic Autoencoders", "Taming Transformers for High-Resolution Image Synthesis", "Pythae: Unifying Generative Autoencoders in Python -- A Benchmarking Use\n Case"], "answer_arxiv_id": ["1505.05770", "1611.02648", "1611.02731", "1611.05013", "1711.00937", "1705.07120", "1906.00446", "1903.12436v4", "2012.09841", "2206.08309"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_3096"} +{"question": "Are there any works that improved the Atlas network by splitting a scene into fragments and using a Gated Recurrent Unit (GRU) network to fuse the 3D features of the fragments?", "answer": ["NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video"], "answer_arxiv_id": ["2104.00681"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_3097"} +{"question": "Could you tell me what molecular machine learning libraries exist?", "answer": ["DGL-LifeSci: An Open-Source Toolkit for Deep Learning on Graphs in Life Science", "TorchDrug: A Powerful and Flexible Machine Learning Platform for Drug Discovery"], "answer_arxiv_id": ["2106.14232", "2202.08320"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_3098"} +{"question": "Are there any researches that practiced doubly robust learning to learn propensity scores?", "answer": ["StableDR: Stabilized Doubly Robust Learning for Recommendation on Data Missing Not at Random"], "answer_arxiv_id": ["2205.04701"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_3099"} +{"question": "What works introduced the concept of adding trainable modules to an otherwise frozen network?", "answer": ["Learning multiple visual domains with residual adapters", "Depthwise Convolution is All You Need for Learning Multiple Visual\n Domains", "Piggyback: Adapting a Single Network to Multiple Tasks by Learning to\n Mask Weights", "NetTailor: Tuning the Architecture, Not Just the Weights"], "answer_arxiv_id": ["1705.08045", "1902.00927", "1801.06519", "1907.00274"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_3100"} +{"question": "What were the early works about the concept of implicit bias from optimization algorithm in neural networks?", "answer": ["Margins, Shrinkage, and Boosting", "In Search of the Real Inductive Bias: On the Role of Implicit Regularization in Deep Learning", "On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "The Implicit Bias of Gradient Descent on Separable Data"], "answer_arxiv_id": ["1303.4172", "1412.6614", "1609.04836", "1710.10345"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_3101"} +{"question": "Could you provide me with some research papers that have successfully applied Transformer model in computer vision domain?", "answer": ["End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection", "DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETR", "Sparse R-CNN: End-to-End Object Detection with Learnable Proposals"], "answer_arxiv_id": ["2005.12872", "2010.04159", "2201.12329", "2011.12450"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_3102"} +{"question": "Which research proposed the concept of novel category discovery?", "answer": ["Learning to Discover Novel Visual Categories via Deep Transfer Clustering", "Neighborhood Contrastive Learning for Novel Class Discovery"], "answer_arxiv_id": ["1908.09884", "2106.10731"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_3103"} +{"question": "What are the papers that describe using imaginary transitions to train the actor and/or critic by generating short-horizon trajectories starting at existing state-action pairs?", "answer": ["When to Trust Your Model: Model-Based Policy Optimization", "Model-Augmented Actor-Critic: Backpropagating through Paths", "Sample-Efficient Reinforcement Learning with Stochastic Ensemble Value Expansion"], "answer_arxiv_id": ["1906.08253", "2005.08068", "1807.01675"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_3104"} +{"question": "Any works about designing neural network modules that are equivariant to scaling?", "answer": ["Deep Scale-spaces: Equivariance Over Scale"], "answer_arxiv_id": ["1905.11697"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_3105"} +{"question": "What researches revealed that self-supervised learning is robust to uncurated data?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition", "Divide and Contrast: Self-supervised Learning from Uncurated Data", "Self-supervised Pretraining of Visual Features in the Wild", "Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision"], "answer_arxiv_id": ["1910.09217", "2105.08054", "2103.01988", "2202.08360"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_3106"} +{"question": "Which works have used behavioral tests to assess the compositionality of neural models?", "answer": ["Generalization without Systematicity: On the Compositional Skills of Sequence-to-Sequence Recurrent Networks", "Systematic Generalization: What Is Required and Can It Be Learned?", "Assessing Phrasal Representation and Composition in Transformers"], "answer_arxiv_id": ["1711.00350", "1811.12889", "2010.03763"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_3107"} +{"question": "What studies enable multi-object composition by introducing a layout into the image generation process?", "answer": ["SpaText: Spatio-Textual Representation for Controllable Image Generation", "Collage Diffusion"], "answer_arxiv_id": ["2211.14305", "2303.00262"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_3108"} +{"question": "What study is most closely related to the proposed latent shift method for generating counterfactuals?", "answer": ["Gifsplanation via Latent Shift: A Simple Autoencoder Approach to Counterfactual Generation for Chest X-rays"], "answer_arxiv_id": ["2102.09475"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_3109"} +{"question": "What are some prior works studying time series classification from the distribution level?", "answer": ["AdaRNN: Adaptive Learning and Forecasting of Time Series"], "answer_arxiv_id": ["2108.04443v2"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_3110"} +{"question": "Which works study identifying vulnerable versions of libraries?", "answer": ["ATVHunter: Reliable Version Detection of Third-Party Libraries for\n Vulnerability Identification in Android Applications", "LibDB: An Effective and Efficient Framework for Detecting Third-Party\n Libraries in Binaries"], "answer_arxiv_id": ["2102.08172", "2204.10232"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_3111"} +{"question": "What research papers have looked into introducing more inductive biases to improve the performance of Transformers?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "How Do Vision Transformers Work?"], "answer_arxiv_id": ["2103.14030", "2202.06709"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_3112"} +{"question": "Which work proposed the Transformer architecture?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3113"} +{"question": "Can you name some publications that enhanced control effects, system efficiency, and few-shot training in neural avatars?", "answer": ["RigNeRF: Fully Controllable Neural 3D Portraits", "Semantic-Aware Implicit Neural Audio-Driven Video Portrait Generation", "DFA-NeRF: Personalized Talking Head Generation via Disentangled Face\n Attributes Neural Rendering", "Real-time Neural Radiance Talking Portrait Synthesis via Audio-spatial\n Decomposition", "Efficient Region-Aware Neural Radiance Fields for High-Fidelity Talking\n Portrait Synthesis", "Learning Dynamic Facial Radiance Fields for Few-Shot Talking Head\n Synthesis", "Generalizable One-shot Neural Head Avatar"], "answer_arxiv_id": ["2206.06481", "2201.07786", "2201.00791", "2211.12368", "2307.09323", "2207.11770", "2306.08768"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_3114"} +{"question": "What prior works have applied masking data modeling on visual and audio domains?", "answer": ["wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations", "HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units", "Conformer-Based Self-Supervised Learning for Non-Speech Audio Tasks"], "answer_arxiv_id": ["2006.11477", "2106.07447", "2110.07313"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_3115"} +{"question": "Which works delve into the concept of group convolution networks?", "answer": ["Group Equivariant Convolutional Networks", "On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "A General Theory of Equivariant CNNs on Homogeneous Spaces"], "answer_arxiv_id": ["1602.07576", "1802.03690", "1811.02017"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_3116"} +{"question": "What works propose to constrain the expressivity of the representation of a neural network?", "answer": ["Discrete-Valued Neural Communication"], "answer_arxiv_id": ["2107.02367"], "source_meta": {"published_time": "20220401"}, "qid": "AutoScholarQuery_train_3117"} +{"question": "Which works have introduced methods for multimodal RGB-3D anomaly detection?", "answer": ["Asymmetric Student-Teacher Networks for Industrial Anomaly Detection", "Towards Total Recall in Industrial Anomaly Detection", "Back to the Feature: Classical 3D Features are (Almost) All You Need for\n 3D Anomaly Detection", "Multimodal Industrial Anomaly Detection via Hybrid Fusion"], "answer_arxiv_id": ["2210.07829", "2106.08265", "2203.05550", "2303.00601"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_3118"} +{"question": "Which papers showcase the continuous performance gains by scaling up transformer language models?", "answer": ["PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2204.02311"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3119"} +{"question": "What paper corresponds to the white-box compression technique using Low-Rank Adaptation?", "answer": ["Learning to Filter Context for Retrieval-Augmented Generation"], "answer_arxiv_id": ["2311.08377"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_3120"} +{"question": "What papers have performed the task of fact-checking using multiple datasets?", "answer": ["Fact-Checking Meets Fauxtography: Verifying Claims About Images", "Overview of the CLEF--2021 CheckThat! Lab on Detecting Check-Worthy\n Claims, Previously Fact-Checked Claims, and Fake News", "MM-Claims: A Dataset for Multimodal Claim Detection in Social Media", "MuMiN: A Large-Scale Multilingual Multimodal Fact-Checked Misinformation\n Social Network Dataset", "End-to-End Multimodal Fact-Checking and Explanation Generation: A\n Challenging Dataset and Models"], "answer_arxiv_id": ["1908.11722", "2109.12987", "2205.01989", "2202.11684", "2205.12487"], "source_meta": {"published_time": "20240718"}, "qid": "AutoScholarQuery_train_3121"} +{"question": "Which research contributes a novel benchmark for assessing knowledge localization methods in LLMs?", "answer": ["KLoB: a Benchmark for Assessing Knowledge Locating Methods in Language\n Models"], "answer_arxiv_id": ["2309.16535"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_3122"} +{"question": "Which studies apply the sequential averaging method (SAM) in the context of BLO?", "answer": ["A Generic First-Order Algorithmic Framework for Bi-Level Programming Beyond Lower-Level Singleton", "Improved Bilevel Model: Fast and Optimal Algorithm with Theoretical Guarantee"], "answer_arxiv_id": ["2006.04045", "2009.00690"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_3123"} +{"question": "Which studies incorporated Fourier features in the field of Implicit neural representations (INRs)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains"], "answer_arxiv_id": ["2003.08934", "2006.10739"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_3124"} +{"question": "What studies involve replacing a graph with another where edges have only been added to mitigate bottlenecks?", "answer": ["Understanding over-squashing and bottlenecks on graphs via curvature", "Oversquashing in GNNs through the lens of information contraction and graph expansion"], "answer_arxiv_id": ["2111.14522", "2208.03471"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3125"} +{"question": "Could you provide me some works about instruction design for in-context learning?", "answer": ["Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2109.01652"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_3126"} +{"question": "Which works studied how in-context learning may arise in various contexts having transformer language models?", "answer": ["What learning algorithm is in-context learning? Investigations with linear models", "Data Distributional Properties Drive Emergent In-Context Learning in Transformers", "Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?", "Impact of Pretraining Term Frequencies on Few-Shot Reasoning", "On the Effect of Pretraining Corpora on In-context Learning by a Large-scale Language Model", "Transformers Learn In-Context by Gradient Descent", "An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2211.15661", "2205.05055", "2202.12837", "2202.07206", "2204.13509", "2212.07677", "2111.02080"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3127"} +{"question": "Which research proposed utilizing the hypernetwork to output the parameters of a data generator, drawing inspiration from the generator of GASP?", "answer": ["Generative Models as Distributions of Functions"], "answer_arxiv_id": ["2102.04776"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_3128"} +{"question": "What papers have focused on reducing the numerical precision of the weights and activations for inference?", "answer": ["Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1", "XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks", "Post training 4-bit quantization of convolutional networks for rapid-deployment", "PACT: Parameterized Clipping Activation for Quantized Neural Networks", "LSQ+: Improving low-bit quantization through learnable offsets and better initialization", "Bridging the Accuracy Gap for 2-bit Quantized Neural Networks (QNN)", "Pruning and Quantization for Deep Neural Network Acceleration: A Survey"], "answer_arxiv_id": ["1602.02830", "1603.05279", "1810.05723", "1805.06085", "2004.09576", "1807.06964v1", "2101.09671"], "source_meta": {"published_time": "20211219"}, "qid": "AutoScholarQuery_train_3129"} +{"question": "What works have contributed to model refinement through knowledge distillation, step distillation, and architectural optimization in efficient Text-to-Image Models?", "answer": ["SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two\n Seconds", "Progressive Distillation for Fast Sampling of Diffusion Models", "Latent Consistency Models: Synthesizing High-Resolution Images with\n Few-Step Inference"], "answer_arxiv_id": ["2306.00980", "2202.00512", "2310.04378"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_3130"} +{"question": "What works apply deep state space models to different domains such as time series data, audio, visual data, text, and medical data?", "answer": ["Effectively Modeling Time Series with Simple Discrete State Spaces", "Efficiently Modeling Long Sequences with Structured State Spaces", "Deep Latent State Space Models for Time-Series Generation", "It’s Raw! Audio Generation with State-Space Models", "Mega: Moving Average Equipped Gated Attention", "Long Range Language Modeling via Gated State Spaces", "Hungry Hungry Hippos: Towards Language Modeling with State Space Models"], "answer_arxiv_id": ["2303.09489", "2111.00396", "2212.12749v3", "2202.09729", "2209.10655", "2206.13947", "2212.14052"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_3131"} +{"question": "What specific papers focused on datasets for detection and localization of extreme weather events and natural disasters?", "answer": ["FloodNet: A High Resolution Aerial Imagery Dataset for Post Flood Scene Understanding", "EarthNet2021: A large-scale dataset and challenge for Earth surface forecasting as a guided video prediction task.", "NADBenchmarks - a compilation of Benchmark Datasets for Machine Learning Tasks related to Natural Disasters"], "answer_arxiv_id": ["2012.02951", "2104.10066", "2212.10735"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_3132"} +{"question": "Is there any study that used the Swin Transformer model to process 3D video inputs?", "answer": ["Video Swin Transformer"], "answer_arxiv_id": ["2106.13230"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_3133"} +{"question": "Which papers used a dual fish-eye camera setup to generate synthetic data for pose estimation?", "answer": ["UnrealEgo: A New Dataset for Robust Egocentric 3D Human Motion Capture"], "answer_arxiv_id": ["2208.01633"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_3134"} +{"question": "What papers depict the use of image encoder and reasoning block with attention mechanisms in visual reasoning methods?", "answer": ["GAMR: A Guided Attention Model for (visual) Reasoning", "From Recognition to Cognition: Visual Commonsense Reasoning", "MERLOT: Multimodal Neural Script Knowledge Models", "VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts"], "answer_arxiv_id": ["2206.04928", "1811.10830", "2106.02636", "2111.02358"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_3135"} +{"question": "Which papers conducted adversarial attack on an attributed graph by adding perturbations on the graphic structure or node features?", "answer": ["Adversarial Attacks on Graph Neural Networks via Meta Learning", "Graph Robustness Benchmark: Benchmarking the Adversarial Robustness of Graph Machine Learning"], "answer_arxiv_id": ["1902.08412", "2111.04314"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_3136"} +{"question": "Were there any studies that learned a correction function of conventional PDE solvers to improve accuracy?", "answer": ["Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers"], "answer_arxiv_id": ["2007.00016"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_3137"} +{"question": "Are there any papers focusing on improving model efficiency in video classification?", "answer": ["SlowFast Networks for Video Recognition", "TSM: Temporal Shift Module for Efficient Video Understanding", "X3D: Expanding Architectures for Efficient Video Recognition", "MoViNets: Mobile Video Networks for Efficient Video Recognition", "TDN: Temporal Difference Networks for Efficient Action Recognition"], "answer_arxiv_id": ["1812.03982", "1811.08383", "2004.04730", "2103.11511", "2012.10071"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_3138"} +{"question": "Are there any studies that provide additional evidence complementing the more phenomenological approach to the statistical mechanics of learning for a practical theory of NN performance?", "answer": ["Predicting trends in the quality of state-of-the-art neural networks without access to training or testing data", "Evaluating natural language processing models with generalization metrics that do not need access to any training or testing data"], "answer_arxiv_id": ["2002.06716v2", "2202.02842"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_3139"} +{"question": "Could you provide studies that have focused on the memorization aspect of training data in language models and potential privacy leakage?", "answer": ["Extracting Training Data from Large Language Models", "Quantifying Memorization Across Neural Language Models", "Preventing Verbatim Memorization in Language Models Gives a False Sense\n of Privacy", "Analyzing Leakage of Personally Identifiable Information in Language\n Models"], "answer_arxiv_id": ["2012.07805", "2202.07646", "2210.17546", "2302.00539"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_3140"} +{"question": "What studies discussed improvements in human reconstruction research using Neural Radiance Fields (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2011.13961", "2011.12948", "2106.13228"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_3141"} +{"question": "What works are there that focus on task allocation or behavioral diversity in subgoal assignment of multi-agent setting?", "answer": ["ROMA: Multi-Agent Reinforcement Learning with Emergent Roles", "Reinforced Cross-Modal Matching and Self-Supervised Imitation Learning for Vision-Language Navigation", "MASER: Multi-Agent Reinforcement Learning with Subgoals Generated from Experience Replay Buffer", "Hierarchical Deep Multiagent Reinforcement Learning with Temporal Abstraction", "Hierarchical Cooperative Multi-Agent Reinforcement Learning with Skill Discovery"], "answer_arxiv_id": ["2003.08039", "1811.10092", "2206.10607", "1809.09332", "1912.03558"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_3142"} +{"question": "Could you name the studies that highlight the role of learned knowledge and prior information in Few-shot Classification?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Probabilistic Model-Agnostic Meta-Learning", "Improved Few-Shot Visual Classification", "Prototypical Networks for Few-shot Learning", "Learning to Compare: Relation Network for Few-Shot Learning", "Matching Networks for One Shot Learning"], "answer_arxiv_id": ["1703.03400", "1806.02817", "1912.03432", "1703.05175", "1711.06025", "1606.04080"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3143"} +{"question": "Could you provide works that have combined physics-informed neural networks and neural operators?", "answer": ["Learning the solution operator of parametric partial differential equations with physics-informed DeepOnets", "Physics-Informed Neural Operator for Learning Partial Differential Equations"], "answer_arxiv_id": ["2103.10974", "2111.03794"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_3144"} +{"question": "What are the works related to taxonomy expansion that have shown their methods' effectiveness?", "answer": ["HiExpan: Task-Guided Taxonomy Construction by Hierarchical Tree\n Expansion", "TaxoExpan: Self-supervised Taxonomy Expansion with Position-Enhanced\n Graph Neural Network", "STEAM: Self-Supervised Taxonomy Expansion with Mini-Paths"], "answer_arxiv_id": ["1910.08194", "2001.09522", "2006.10217"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_3145"} +{"question": "Are there any studies about map awareness in motion prediction?", "answer": ["Dynamic Occupancy Grid Mapping with Recurrent Neural Networks"], "answer_arxiv_id": ["2011.08659"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_3146"} +{"question": "What are some recent works that focus on influence estimation?", "answer": ["Explaining Black Box Predictions and Unveiling Data Artifacts through\n Influence Functions", "Studying Large Language Model Generalization with Influence Functions", "Simfluence: Modeling the Influence of Individual Training Examples by\n Simulating Training Runs", "DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and\n Diffusion Models", "If Influence Functions are the Answer, Then What is the Question?"], "answer_arxiv_id": ["2005.06676", "2308.03296v1", "2303.08114", "2310.00902", "2209.05364"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_3147"} +{"question": "Can you provide references demonstrating the data augmentation approach in self-knowledge distillation?", "answer": ["Regularizing Class-wise Predictions via Self-knowledge Distillation", "A Comprehensive Overhaul of Feature Distillation"], "answer_arxiv_id": ["2003.13964", "1904.01866"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3148"} +{"question": "Which work utilizes a score-based diffusion model to generate 3D periodic material structures and design the model to capture physical symmetries in materials?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_3149"} +{"question": "What are the works on graph neural operators?", "answer": ["Learning to Simulate Complex Physics with Graph Networks", "Neural Operator: Graph Kernel Network for Partial Differential Equations"], "answer_arxiv_id": ["2002.09405", "2003.03485"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3150"} +{"question": "What works parameterize the norm and direction of the weight matrices separately and thus reach faster convergence in the process of re-parameterization?", "answer": ["Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks"], "answer_arxiv_id": ["1602.07868"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_3151"} +{"question": "Could you provide me some works about one-stage detectors in object detection task?", "answer": ["Focal Loss for Dense Object Detection", "FCOS: Fully Convolutional One-Stage Object Detection", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["1708.02002", "1904.01355", "2005.12872"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3152"} +{"question": "Any works about extending the approach of using GANs by conditioning on scene meshes and visual signals?", "answer": ["MESH2IR: Neural Acoustic Impulse Response Generator for Complex 3D\n Scenes", "AV-RIR: Audio-Visual Room Impulse Response Estimation"], "answer_arxiv_id": ["2205.09248", "2312.00834"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_3153"} +{"question": "Which papers discussed the use of symbolic systems and formal languages in automatic logical reasoners?", "answer": ["Acquisition of Phrase Correspondences using Natural Deduction Proofs"], "answer_arxiv_id": ["1804.07656"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_3154"} +{"question": "Are there any research papers that integrated a physics-based objective as conditional guidance for physically plausible planning and motion generation?", "answer": ["Diffusion-based Generation, Optimization, and Planning in 3D Scenes"], "answer_arxiv_id": ["2301.06015"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_3155"} +{"question": "Any works that investigated how model scales and retrievers affect in-context learning for semantic parsing?", "answer": ["Evaluating the Impact of Model Scale for Compositional Generalization in Semantic Parsing", "Benchmarking Multimodal Regex Synthesis with Complex Structures", "Diverse Demonstrations Improve In-context Compositional Generalization"], "answer_arxiv_id": ["2205.12253", "2005.00663", "2212.06800"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3156"} +{"question": "Which work proposed adjusting the variance schedule, via a cosine function of time, for the improvement of Diffusion Models?", "answer": ["Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2102.09672"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3157"} +{"question": "Which paper introduced OptiDICE as the first policy optimization algorithm for DICE?", "answer": ["OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation"], "answer_arxiv_id": ["2106.10783"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_3158"} +{"question": "What works propose Iterative Dataset Update (Iterative DU) for NeRF editing?", "answer": ["Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions"], "answer_arxiv_id": ["2303.12789"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_3159"} +{"question": "Which researches have advanced towards human-level text-to-speech using techniques like BERT pre-training and end-to-end training?", "answer": ["Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech", "PnG BERT: Augmented BERT on Phonemes and Graphemes for Neural TTS", "NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level Quality", "Mixed-Phoneme BERT: Improving BERT with Mixed Phoneme and Sup-Phoneme Representations for Text to Speech", "JETS: Jointly Training FastSpeech2 and HiFi-GAN for End to End Text to Speech", "End-to-End Adversarial Text-to-Speech", "Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling"], "answer_arxiv_id": ["2106.06103", "2103.15060", "2205.04421", "2203.17190", "2203.16852", "2006.03575", "2103.14574"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_3160"} +{"question": "Which works propose memorization-augmented methods to enhance language modeling and named entity recognition?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models", "Efficient Nearest Neighbor Language Models"], "answer_arxiv_id": ["1911.00172", "2109.04212"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_3161"} +{"question": "What might be the work that presents an adaptive keyframe and region scheduling policy for VSS?", "answer": ["Dynamic Video Segmentation Network"], "answer_arxiv_id": ["1804.00931"], "source_meta": {"published_time": "20240127"}, "qid": "AutoScholarQuery_train_3162"} +{"question": "What papers have discussed the teacher-forcing strategy in natural language processing?", "answer": ["Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks"], "answer_arxiv_id": ["1506.03099"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_3163"} +{"question": "Which works demonstrated direct usage of pretrained 2D image encoders via supervised fine-tuning?", "answer": ["P2P: Tuning Pre-trained Image Models for Point Cloud Analysis with Point-to-Pixel Prompting"], "answer_arxiv_id": ["2208.02812"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_3164"} +{"question": "What were the initial text-to-image generation models focused on and which studies represented them?", "answer": ["Generative Adversarial Text to Image Synthesis", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked\n Generative Adversarial Networks"], "answer_arxiv_id": ["1605.05396", "1612.03242"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_3165"} +{"question": "Could you provide me some works about the proficiency of language models in synthetic dataset generation?", "answer": ["Textbooks Are All You Need", "Generating Datasets with Pretrained Language Models", "Exploiting Asymmetry for Synthetic Training Data Generation: SynthIE and\n the Case of Information Extraction", "TinyStories: How Small Can Language Models Be and Still Speak Coherent\n English?", "Exploring the Potential of AI-Generated Synthetic Datasets: A Case Study\n on Telematics Data with ChatGPT"], "answer_arxiv_id": ["2306.11644", "2104.07540", "2303.04132", "2305.07759", "2306.13700"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_3166"} +{"question": "Could you provide me some works that inspired this research's interpolation experiments?", "answer": ["Patching open-vocabulary models by interpolating weights"], "answer_arxiv_id": ["2208.05592"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_3167"} +{"question": "Which papers introduce decoupled training using ADMM in the context of improving parallelism in training DNNs?", "answer": ["Training Neural Networks Without Gradients: A Scalable ADMM Approach"], "answer_arxiv_id": ["1605.02026"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3168"} +{"question": "What works identified the mismatch between training and testing degradations as the reason for SR model failure?", "answer": ["Blind Image Super-Resolution: A Survey and Beyond"], "answer_arxiv_id": ["2107.03055"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3169"} +{"question": "Could you provide me some works on directly fine-tuning a GPT-style model on conversation data?", "answer": ["BlenderBot 3: a deployed conversational agent that continually learns to\n responsibly engage", "DialoGPT: Large-Scale Generative Pre-training for Conversational\n Response Generation", "PLATO: Pre-trained Dialogue Generation Model with Discrete Latent\n Variable"], "answer_arxiv_id": ["2208.03188", "1911.00536", "1910.07931"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_3170"} +{"question": "Which works prune unimportant channels through learnable scaling factors added for each structure?", "answer": ["Plug-in, Trainable Gate for Streamlining Arbitrary Neural Networks", "Gate Decorator: Global Filter Pruning Method for Accelerating Deep\n Convolutional Neural Networks", "Operation-Aware Soft Channel Pruning using Differentiable Masks"], "answer_arxiv_id": ["1904.10921", "1909.08174", "2007.03938"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_3171"} +{"question": "Any study introduced trainable visual prompt vectors into the image patch sequence of each Transformer layer and learned them?", "answer": ["Visual Prompt Tuning"], "answer_arxiv_id": ["2203.12119"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_3172"} +{"question": "Could you provide me some studies about self-supervised methods that utilize skeletal representations?", "answer": ["Self-supervised Learning of Interpretable Keypoints from Unlabelled\n Videos", "PersonLab: Person Pose Estimation and Instance Segmentation with a\n Bottom-Up, Part-Based, Geometric Embedding Model", "AutoLink: Self-supervised Learning of Human Skeletons and Object\n Outlines by Linking Keypoints"], "answer_arxiv_id": ["1907.02055", "1803.08225", "2205.10636"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_3173"} +{"question": "Can you provide some studies that propose novel data synthesizing methods?", "answer": ["Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data", "Designing a Practical Degradation Model for Deep Blind Image Super-Resolution"], "answer_arxiv_id": ["2107.10833", "2103.14006"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3174"} +{"question": "What studies empirically explore the negative influence of heterogeneity in federated learning?", "answer": ["Federated Learning with Non-IID Data"], "answer_arxiv_id": ["1806.00582"], "source_meta": {"published_time": "20220110"}, "qid": "AutoScholarQuery_train_3175"} +{"question": "What were the early studies on integrating different knowledge sources or fusing multimodal information for KVQA with KGs?", "answer": ["FVQA: Fact-based Visual Question Answering", "KRISP: Integrating Implicit and Symbolic Knowledge for Open-Domain\n Knowledge-Based VQA"], "answer_arxiv_id": ["1606.05433", "2012.11014"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_3176"} +{"question": "Which studies employ Kalman filters for motion prediction in tracking-by-detection tasks?", "answer": ["Simple Online and Realtime Tracking", "Simple Online and Realtime Tracking with a Deep Association Metric", "ByteTrack: Multi-Object Tracking by Associating Every Detection Box", "Observation-Centric SORT: Rethinking SORT for Robust Multi-Object\n Tracking", "FairMOT: On the Fairness of Detection and Re-Identification in Multiple\n Object Tracking", "MAT: Motion-Aware Multi-Object Tracking"], "answer_arxiv_id": ["1602.00763", "1703.07402", "2110.06864", "2203.14360", "2004.01888", "2009.04794v2"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3177"} +{"question": "What study proposed a loss function to address the problem of entity omission in image generation?", "answer": ["Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2301.13826"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_3178"} +{"question": "What are examples of works that use meshes for 3D data representation?", "answer": ["MeshDiffusion: Score-based Generative 3D Mesh Modeling", "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images"], "answer_arxiv_id": ["2303.08133", "2209.11163"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3179"} +{"question": "Can you provide examples of studies that have used Graph Neural Networks in physical environments?", "answer": ["Learning Domain-Independent Planning Heuristics with Hypergraph Networks", "Structured agents for physical construction"], "answer_arxiv_id": ["1911.13101", "1904.03177"], "source_meta": {"published_time": "20220202"}, "qid": "AutoScholarQuery_train_3180"} +{"question": "Are there any studies proposing lightweight models in this field?", "answer": ["UDepth: Fast Monocular Depth Estimation for Visually-guided Underwater\n Robots"], "answer_arxiv_id": ["2209.12358"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_3181"} +{"question": "Could you provide me some works that manipulated the latent space of GANs to bring about global semantic changes in a supervised manner?", "answer": ["GANalyze: Toward Visual Definitions of Cognitive Image Properties", "Controlling generative models with continuous factors of variations", "InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs", "Interpreting the Latent Space of GANs for Semantic Face Editing"], "answer_arxiv_id": ["1906.10112", "2001.10238", "2005.09635", "1907.10786"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_3182"} +{"question": "What studies investigate that language models can meaningfully encode the concept of color?", "answer": ["Can Language Models Encode Perceptual Structure Without Grounding? A\n Case Study in Color"], "answer_arxiv_id": ["2109.06129"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_3183"} +{"question": "What works contributed to improvements in training speed of Neural Radiance Fields (NeRF)?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "FastNeRF: High-Fidelity Neural Rendering at 200FPS"], "answer_arxiv_id": ["2201.05989", "2112.05131", "2103.10380"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_3184"} +{"question": "Which works used the data augmentation method to address the data scarcity problem in domain-specific NER?", "answer": ["Data Augmentation for Cross-Domain Named Entity Recognition", "MELM: Data Augmentation with Masked Entity Language Modeling for\n Low-Resource NER"], "answer_arxiv_id": ["2109.01758", "2108.13655"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_3185"} +{"question": "What research papers discuss the application of greedy k-center to choose the coreset with good data coverage?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["1708.00489"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_3186"} +{"question": "Which studies applied matrix mechanisms to linear statistical queries?", "answer": ["Optimizing Linear Counting Queries Under Differential Privacy", "Optimizing error of high-dimensional statistical queries under differential privacy", "The Power of Factorization Mechanisms in Local and Central Differential Privacy"], "answer_arxiv_id": ["0912.4742", "1808.03537", "1911.08339"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3187"} +{"question": "Could you give me some examples of studies that employ hierarchical generative models in the vector graphics field?", "answer": ["DeepSVG: A Hierarchical Generative Network for Vector Graphics Animation"], "answer_arxiv_id": ["2007.11301"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_3188"} +{"question": "Which works have contributed to the rapid development of the field of egocentric vision through their large-scale datasets?", "answer": ["The EPIC-KITCHENS Dataset: Collection, Challenges and Baselines", "Rescaling Egocentric Vision", "Ego4D: Around the World in 3,000 Hours of Egocentric Video", "Assembly101: A Large-Scale Multi-View Video Dataset for Understanding Procedural Activities"], "answer_arxiv_id": ["2005.00343", "2006.13256", "2110.07058", "2203.14712v2"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3189"} +{"question": "What are the most recent state-of-the-art HIC methods?", "answer": ["Leveraging Class Hierarchies with Metric-Guided Prototype Learning", "Use All The Labels: A Hierarchical Multi-Label Contrastive Learning Framework"], "answer_arxiv_id": ["2007.03047", "2204.13207"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_3190"} +{"question": "What studies utilized kernel density estimates to estimate calibration in classification settings?", "answer": ["Effect of Confidence and Explanation on Accuracy and Trust Calibration in AI-Assisted Decision Making", "A Consistent and Differentiable Lp Canonical Calibration Error Estimator"], "answer_arxiv_id": ["2001.02114", "2210.07810v1"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_3191"} +{"question": "Could you list the publications that apply exact line search or exploit non-uniform smoothness to achieve linear convergence for PG?", "answer": ["Leveraging Non-uniformity in First-order Non-convex Optimization"], "answer_arxiv_id": ["2105.06072"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_3192"} +{"question": "Are there any studies which used GANs for some specific audio editing tasks?", "answer": ["Audio inpainting with generative adversarial network", "CMGAN: Conformer-Based Metric-GAN for Monaural Speech Enhancement"], "answer_arxiv_id": ["2003.07704", "2209.11112v3"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_3193"} +{"question": "What papers proposed diffusion models for conditional generation?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations"], "answer_arxiv_id": ["2105.05233", "2207.12598", "2108.01073"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3194"} +{"question": "Could you please provide some studies that use Wasserstein distance for measuring proximity in DR?", "answer": ["Robust Wasserstein Profile Inference and Applications to Machine Learning", "Distributionally Robust Stochastic Optimization with Wasserstein Distance", "Wasserstein Distributionally Robust Optimization: Theory and Applications in Machine Learning", "Data-Driven Distributionally Robust Optimization Using the Wasserstein Metric: Performance Guarantees and Tractable Reformulations", "Distributionally Robust Logistic Regression"], "answer_arxiv_id": ["1610.05627", "1604.02199", "1908.08729v1", "1505.05116", "1509.09259"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_3195"} +{"question": "Who leveraged the link between Bayesian perspective of cv and the training of Gaussian process models?", "answer": ["Revisiting Active Sets for Gaussian Process Decoders"], "answer_arxiv_id": ["2209.04636"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3196"} +{"question": "Could you provide me some works proposing diverse gradient-based strong attacks?", "answer": ["Adversarial examples in the physical world", "Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1607.02533", "1608.04644"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_3197"} +{"question": "Which works are based on image matting using deep learning?", "answer": ["Deep Image Matting", "Semantic Image Matting", "Indices Matter: Learning to Index for Deep Image Matting", "Natural Image Matting via Guided Contextual Attention"], "answer_arxiv_id": ["1703.03872", "2104.08201", "1908.00672", "2001.04069"], "source_meta": {"published_time": "20240424"}, "qid": "AutoScholarQuery_train_3198"} +{"question": "What research attempts to formulate more vision tasks in a unified sequence-to-sequence learning paradigm?", "answer": ["Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework"], "answer_arxiv_id": ["2206.08916", "2209.06794", "2202.03052"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_3199"} +{"question": "Which papers are the applications of tensor factorization methods in various deep learning models?", "answer": ["A Tensorized Transformer for Language Modeling", "Tensorizing Neural Networks", "Tensor-Train Recurrent Neural Networks for Video Classification", "Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition", "Multi-Grid Tensorized Fourier Neural Operator for High-Resolution PDEs"], "answer_arxiv_id": ["1906.09777", "1509.06569", "1707.01786", "1412.6553", "2310.00120"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_3200"} +{"question": "What studies attempted to unify reposing, virtual try-on, and text manipulation tasks, and used frozen CLIP embeddings to bridge visual and text domains?", "answer": ["UPGPT: Universal Diffusion Model for Person Image Generation, Editing\n and Pose Transfer"], "answer_arxiv_id": ["2304.08870"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_3201"} +{"question": "What studies utilize the spectral clustering method as a core algorithmic component in their segmentation systems?", "answer": ["Pushing the Boundaries of Boundary Detection using Deep Learning", "Affinity CNN: Learning Pixel-Centric Pairwise Relations for\n Figure/Ground Embedding", "Normalized Cut Loss for Weakly-supervised CNN Segmentation"], "answer_arxiv_id": ["1511.07386", "1512.02767", "1804.01346"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_3202"} +{"question": "Are there any studies that focus on optimal transport (OT) within mixup for interpolating features?", "answer": ["Puzzle Mix: Exploiting Saliency and Local Statistics for Optimal Mixup", "Co-Mixup: Saliency Guided Joint Mixup with Supermodular Diversity"], "answer_arxiv_id": ["2009.06962v2", "2102.03065"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_3203"} +{"question": "Which study proposes a multi-scale autoencoder for 3D point clouds?", "answer": ["Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud\n Pre-training"], "answer_arxiv_id": ["2205.14401"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_3204"} +{"question": "Which studies highlight that models have a hard time discerning between text containing the same words ordered differently?", "answer": ["Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality"], "answer_arxiv_id": ["2204.03162"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_3205"} +{"question": "Could you provide me some research that adapted Visual-Language Model into the object detection framework?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "RegionCLIP: Region-based Language-Image Pretraining"], "answer_arxiv_id": ["2104.13921", "2112.09106"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_3206"} +{"question": "Which paper has derived the first high-probability results for non-convex optimization under Assumption?", "answer": ["High-probability bounds for Non-Convex Stochastic Optimization with Heavy Tails"], "answer_arxiv_id": ["2106.14343"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3207"} +{"question": "Which papers investigated the use of synthetic controls to estimate the effect of a single intervention over time?", "answer": ["Policy Analysis using Synthetic Controls in Continuous-Time"], "answer_arxiv_id": ["2102.01577"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_3208"} +{"question": "Any works about using generative neural networks to refine style representation in image painting?", "answer": ["Neural Painters: A learned differentiable constraint for generating\n brushstroke paintings"], "answer_arxiv_id": ["1904.08410"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_3209"} +{"question": "Which papers predict in 3D in single-view methods by using monocular depth prediction networks?", "answer": ["Deep Ordinal Regression Network for Monocular Depth Estimation", "Unsupervised Monocular Depth Estimation with Left-Right Consistency"], "answer_arxiv_id": ["1806.02446", "1609.03677"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_3210"} +{"question": "Could you provide me some works that use multimodal information from academic presentation videos?", "answer": ["See, Hear, Read: Leveraging Multimodality with Guided Attention for\n Abstractive Text Summarization"], "answer_arxiv_id": ["2105.09601"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_3211"} +{"question": "Could you name the works that proposed to do spectral decomposition of dense DINO features and use of sign of Fiedler eigenvector as criteria for object localization mask?", "answer": ["Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization", "Self-Supervised Transformers for Unsupervised Object Discovery using Normalized Cut"], "answer_arxiv_id": ["2205.07839", "2202.11539"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_3212"} +{"question": "Which articles highlight the importance of evaluating large language models to ensure their effectiveness and reliability?", "answer": ["Evaluating Large Language Models: A Comprehensive Survey", "A Survey on Evaluation of Large Language Models"], "answer_arxiv_id": ["2310.19736", "2307.03109"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_3213"} +{"question": "What work used the concept of global optimism to obtain the optimal sample complexity at the expense of computational efficiency?", "answer": ["Learning Near Optimal Policies with Low Inherent Bellman Error"], "answer_arxiv_id": ["2003.00153"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_3214"} +{"question": "Which contrastive learning (CL) methods are known to optimize the InfoNCE loss?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "1911.05722"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_3215"} +{"question": "What studies provide methods to solve MWPs by generating and executing intermediate expressions?", "answer": ["Unit Dependency Graph and its Application to Arithmetic Word Problem Solving", "MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalisms"], "answer_arxiv_id": ["1612.00969", "1905.13319"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_3216"} +{"question": "Could you provide some references where techniques were developed to train models for outputting images from text prompts?", "answer": ["Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers", "Muse: Text-To-Image Generation via Masked Generative Transformers", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["2206.10789", "2105.13290", "2301.00704", "2204.06125", "2112.10741", "2203.13131"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_3217"} +{"question": "Can you provide references for works that involve tuning hyperparameters on small-scale models and transferring them to large-scale versions?", "answer": ["Tensor Programs V: Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfer"], "answer_arxiv_id": ["2203.03466"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_3218"} +{"question": "Are there any works focused on the CLIP model, a specific type of vision-language model?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_3219"} +{"question": "In the scope of test-time prompt tuning, what is the research that proposed learning prompts at the text side with an entropy minimization objective?", "answer": ["Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models"], "answer_arxiv_id": ["2209.07511"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_3220"} +{"question": "Which studies modeled the binary relationship between samples based on the similarity of the top-k feature dimensions for novel class discovery?", "answer": ["AutoNovel: Automatically Discovering and Learning Novel Visual\n Categories", "Novel Visual Category Discovery with Dual Ranking Statistics and Mutual\n Knowledge Distillation"], "answer_arxiv_id": ["2106.15252", "2107.03358"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_3221"} +{"question": "Are there any papers that demonstrated how our model can enhance alignment through sparse, entailment scores and dense, fine-grained natural language explanations (NLE) feedback?", "answer": ["Training Language Models with Language Feedback at Scale"], "answer_arxiv_id": ["2303.16755"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_3222"} +{"question": "Could you provide me with studies that aimed at distilling a complete 'Chain-of-Thought' from a teacher model?", "answer": ["Symbolic Chain-of-Thought Distillation: Small Models Can Also \"Think\"\n Step-by-Step", "SCOTT: Self-Consistent Chain-of-Thought Distillation", "Distilling Reasoning Capabilities into Smaller Language Models", "Teaching Small Language Models to Reason", "Distilling Step-by-Step! Outperforming Larger Language Models with Less\n Training Data and Smaller Model Sizes"], "answer_arxiv_id": ["2306.14050", "2305.01879", "2212.00193", "2212.08410", "2305.02301"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_3223"} +{"question": "Do any works demonstrate the advancement of DDPM over GANs in computer vision?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2102.09672", "2105.05233"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_3224"} +{"question": "What research showcases the usage of part-based models for fine-grained classification?", "answer": ["DeepVoting: A Robust and Explainable Deep Network for Semantic Part Detection under Partial Occlusion", "Semantic Part Detection via Matching: Learning to Generalize to Novel Viewpoints from Limited Training Data"], "answer_arxiv_id": ["1709.04577", "1811.11823"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_3225"} +{"question": "Which researches are linked to the beginning of model-based RL with low-dimensional, compact state spaces?", "answer": ["When to Trust Your Model: Model-Based Policy Optimization", "Neural Network Dynamics for Model-Based Deep Reinforcement Learning with Model-Free Fine-Tuning"], "answer_arxiv_id": ["1906.08253", "1708.02596"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_3226"} +{"question": "What studies demonstrated the effectiveness of diffusion models in text-to-image generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456", "2010.02502"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3227"} +{"question": "Can you list some studies that constructed datasets for detecting violent actions?", "answer": ["Real-world Anomaly Detection in Surveillance Videos", "RWF-2000: An Open Large Scale Video Database for Violence Detection"], "answer_arxiv_id": ["1801.04264", "1911.05913"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3228"} +{"question": "Could you provide me some works considering adversarial training in the domain of domain adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks", "Adversarial Discriminative Domain Adaptation", "Controllable Invariance through Adversarial Feature Learning", "Learning to Pivot with Adversarial Networks"], "answer_arxiv_id": ["1505.07818", "1702.05464", "1705.11122", "1611.01046"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3229"} +{"question": "What works have explored the data manifold to explain adversarial examples?", "answer": ["On the Geometry of Adversarial Examples", "Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models"], "answer_arxiv_id": ["1811.00525", "1805.06605"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_3230"} +{"question": "Can you specify a work where a similar equivalent substitution of the same order in the local approximation as in this research occurred?", "answer": ["UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2302.04867"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_3231"} +{"question": "Which works focused on increasing design diversity in TO using data-driven approaches?", "answer": ["Deep neural networks for the evaluation and design of photonic devices", "A novel topology design approach using an integrated deep learning network architecture", "A Novel Topology Optimization Approach using Conditional Deep Learning"], "answer_arxiv_id": ["2007.00084", "1808.02334v2", "1901.04859v1"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3232"} +{"question": "Which work proposed the concept of Markov chain variational inference (MCVI)?", "answer": ["Markov Chain Monte Carlo and Variational Inference: Bridging the Gap"], "answer_arxiv_id": ["1410.6460"], "source_meta": {"published_time": "20211125"}, "qid": "AutoScholarQuery_train_3233"} +{"question": "Which work explored the cultural limitations of Text-to-Image models in the South Asian context?", "answer": ["AI's Regimes of Representation: A Community-centered Study of\n Text-to-Image Models in South Asia"], "answer_arxiv_id": ["2305.11844"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_3234"} +{"question": "Are there any studies conducting parameter-efficient transfer learning in the field of computer vision?", "answer": ["Visual Prompt Tuning", "Exploring Visual Prompts for Adapting Large-Scale Models", "AdaptFormer: Adapting Vision Transformers for Scalable Visual\n Recognition", "Convolutional Bypasses Are Better Vision Transformer Adapters", "Visual Prompt Tuning for Test-time Domain Adaptation", "AIM: Adapting Image Models for Efficient Video Action Recognition", "MV-Adapter: Multimodal Video Transfer Learning for Video Text Retrieval", "Semi-Supervised Single-View 3D Reconstruction via Prototype Shape Priors", "Few-shot Single-view 3D Reconstruction with Memory Prior Contrastive\n Network"], "answer_arxiv_id": ["2203.12119", "2203.17274", "2205.13535", "2207.07039", "2210.04831", "2302.03024", "2301.07868", "2209.15383", "2208.00183"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_3235"} +{"question": "What papers proposed methods to reduce catastrophic forgetting in continual learning?", "answer": ["Overcoming Catastrophic Forgetting by Incremental Moment Matching"], "answer_arxiv_id": ["1703.08475"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_3236"} +{"question": "Can you name any works that used Tensor Product Representations in math problem solving?", "answer": ["Enhancing the Transformer With Explicit Relational Encoding for Math Problem Solving"], "answer_arxiv_id": ["1910.06611"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3237"} +{"question": "What works suggest using concatenated unimodal features for integrating information across modalities?", "answer": ["EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action Recognition", "Repetitive Activity Counting by Sight and Sound", "SoundSpaces: Audio-Visual Navigation in 3D Environments"], "answer_arxiv_id": ["1908.08498", "2103.13096", "1912.11474"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_3238"} +{"question": "What are some studies that continue pushing the frontier of scaling for LLMs, going beyond 500B parameters?", "answer": ["Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2112.11446", "2201.11990", "2112.06905", "2204.02311"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_3239"} +{"question": "Are there any studies indicating specific shortcomings of RL-based approaches in a multi-objective setting?", "answer": ["A practical guide to multi-objective reinforcement learning and planning"], "answer_arxiv_id": ["2103.09568"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_3240"} +{"question": "What works study the realizability of so-called importance weight functions in OPE?", "answer": ["Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation", "Minimax Weight and Q-Function Learning for Off-Policy Evaluation"], "answer_arxiv_id": ["1810.12429", "1910.12809"], "source_meta": {"published_time": "20230725"}, "qid": "AutoScholarQuery_train_3241"} +{"question": "What are the standard white-box defense evaluation attacks mentioned in recent years?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1706.06083", "1608.04644"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_3242"} +{"question": "Which papers eliminated the requirement of negative samples in contrastive learning to avoid representation collapse?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning", "Revitalizing CNN Attentions via Transformers in Self-Supervised Visual Representation Learning"], "answer_arxiv_id": ["2006.07733", "2011.10566", "2110.05340"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_3243"} +{"question": "Which work built non-local blocks for capturing long-range dependencies?", "answer": ["Non-local Neural Networks"], "answer_arxiv_id": ["1711.07971"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_3244"} +{"question": "Do any studies point out the 'manifold intrusion' phenomenon in Mixup training?", "answer": ["MixUp as Locally Linear Out-Of-Manifold Regularization"], "answer_arxiv_id": ["1809.02499"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_3245"} +{"question": "Which works extended generalization bounds in DANN to multi-source domains and proposed multisource domain adversarial networks?", "answer": ["Multi-Domain Adversarial Learning"], "answer_arxiv_id": ["1903.09239"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_3246"} +{"question": "Which publication proposes to condition a GAN using an input image to generate variations of that input image?", "answer": ["Instance-Conditioned GAN"], "answer_arxiv_id": ["2109.05070"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_3247"} +{"question": "What studies use text-conditional discrete diffusion models for sound generation?", "answer": ["Diffsound: Discrete Diffusion Model for Text-to-sound Generation"], "answer_arxiv_id": ["2207.09983"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_3248"} +{"question": "Could you provide studies where the PointConv-based discriminator was used in training a function generator?", "answer": ["Generative Models as Distributions of Functions"], "answer_arxiv_id": ["2102.04776"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_3249"} +{"question": "Which papers attempted to parse the question into a logical form for answering questions using a knowledge graph?", "answer": ["Neural-Symbolic Models for Logical Queries on Knowledge Graphs"], "answer_arxiv_id": ["2205.10128"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_3250"} +{"question": "Can you tell me which paper is about the original approach of federated learning termed as FedAvg?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized\n Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_3251"} +{"question": "Can you provide references of works that used masked representation learning in Natural Language Processing?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "ALBERT: A Lite BERT for Self-supervised Learning of Language\n Representations", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["1810.04805", "1909.11942", "1907.11692", "2111.06377"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_3252"} +{"question": "Who introduced the Flat-UCB and BAST algorithm to adapt UCT for the D-chain problem?", "answer": ["Bandit Algorithms for Tree Search"], "answer_arxiv_id": ["1408.2028v1"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_3253"} +{"question": "What papers report the use of CLIP image embeddings in text-to-image synthesis?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10741", "2204.06125"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_3254"} +{"question": "Could you provide me some works that offer significant inspiration for the common-information-based approach?", "answer": ["Decentralized Stochastic Control with Partial History Sharing: A Common Information Approach"], "answer_arxiv_id": ["1209.1695"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_3255"} +{"question": "Could you list some references about recent advancements in optimizers such as low-rank parameterization?", "answer": ["Adafactor: Adaptive Learning Rates with Sublinear Memory Cost", "Shampoo: Preconditioned Stochastic Tensor Optimization"], "answer_arxiv_id": ["1804.04235", "1802.09568"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_3256"} +{"question": "Which studies have demonstrated performance enhancements by generating additional data for training in the field of diffusion models?", "answer": ["Diversity is Definitely Needed: Improving Model-Agnostic Zero-shot\n Classification via Stable Diffusion", "Boosting Human-Object Interaction Detection with Text-to-Image Diffusion\n Model"], "answer_arxiv_id": ["2302.03298", "2305.12252"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_3257"} +{"question": "Which resources can provide general techniques for speeding up model inference?", "answer": ["Compressing Deep Convolutional Networks using Vector Quantization", "HAQ: Hardware-Aware Automated Quantization with Mixed Precision", "Distilling the Knowledge in a Neural Network", "MetaDistiller: Network Self-Boosting via Meta-Learned Top-Down Distillation", "Channel Pruning for Accelerating Very Deep Neural Networks", "Neural Architecture Search: A Survey"], "answer_arxiv_id": ["1412.6115", "1811.08886", "1503.02531", "2008.12094v1", "1707.06168", "1808.05377"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3258"} +{"question": "What research papers conducted self-supervised representation learning on object-centric point clouds?", "answer": ["Deep Closest Point: Learning Representations for Point Cloud\n Registration", "Unsupervised Multi-Task Feature Learning on Point Clouds", "Self-Supervised Deep Learning on Point Clouds by Reconstructing Space", "Info3D: Representation Learning on 3D Objects using Mutual Information\n Maximization and Contrastive Learning"], "answer_arxiv_id": ["1905.03304", "1910.08207", "1901.08396", "2006.02598"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_3259"} +{"question": "Which studies use the cross-view attention mechanism for novel view reconstruction?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering", "VolRecon: Volume Rendering of Signed Ray Distance Functions for\n Generalizable Multi-View Reconstruction", "Explicit Correspondence Matching for Generalizable Neural Radiance\n Fields"], "answer_arxiv_id": ["2102.13090", "2212.08067", "2304.12294"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_3260"} +{"question": "Who pointed out the limitation of the PR metric in sample generation?", "answer": ["Improved Precision and Recall Metric for Assessing Generative Models"], "answer_arxiv_id": ["1904.06991"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_3261"} +{"question": "Which research papers utilize GNN auto-encoders in a non-linear morphable model for face geometry estimation?", "answer": ["Generating 3D faces using Convolutional Mesh Autoencoders"], "answer_arxiv_id": ["1807.10267"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_3262"} +{"question": "Could you provide me some works about enhancing Neural Volumes with mixture of volumetric primitives?", "answer": ["Mixture of Volumetric Primitives for Efficient Neural Rendering"], "answer_arxiv_id": ["2103.01954"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_3263"} +{"question": "Can you tell me about works that utilize neural radiance fields?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields", "VR-NeRF: High-Fidelity Virtualized Walkable Spaces"], "answer_arxiv_id": ["2003.08934", "2304.06706", "2311.02542"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_3264"} +{"question": "Could you tell me about studies that use model-free methods, specifically social pooling methods, to model interactive behaviors?", "answer": ["Social GAN: Socially Acceptable Trajectories with Generative Adversarial\n Networks", "SoPhie: An Attentive GAN for Predicting Paths Compliant to Social and\n Physical Constraints"], "answer_arxiv_id": ["1803.10892", "1806.01482"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_3265"} +{"question": "Are there any studies that have proposed an optimizer continuously using weight averaging to update model weight?", "answer": ["Lookahead Optimizer: k steps forward, 1 step back"], "answer_arxiv_id": ["1907.08610"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_3266"} +{"question": "Could you mention some work that employ a two-stage approach integrating parametric lighting representations and environment map generation?", "answer": ["Editable Indoor Lighting Estimation", "EverLight: Indoor-Outdoor Editable HDR Lighting Estimation"], "answer_arxiv_id": ["2211.03928", "2304.13207"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_3267"} +{"question": "Which references proposed methodologies used for multi-label classification in hierarchical classification tasks?", "answer": ["Improving Pretrained Models for Zero-shot Multi-label Text Classification through Reinforced Label Hierarchy Reasoning", "An Empirical Study on Large-Scale Multi-Label Text Classification Including Few and Zero-Shot Labels"], "answer_arxiv_id": ["2104.01666v1", "2010.01653"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3268"} +{"question": "What are the works that presented continuous-time dynamic GNNs, also known as temporal GNNs?", "answer": ["Temporal Graph Networks for Deep Learning on Dynamic Graphs"], "answer_arxiv_id": ["2006.10637"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_3269"} +{"question": "Are there any works specifically relating to tasks associated with mobile apps?", "answer": ["Spotlight: Mobile UI Understanding using Vision-Language Models with a Focus", "Mapping Natural Language Instructions to Mobile UI Action Sequences"], "answer_arxiv_id": ["2209.14927", "2005.03776"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_3270"} +{"question": "What papers propose architectural modifications to neural networks for selective classification?", "answer": ["SelectiveNet: A Deep Neural Network with an Integrated Reject Option", "Deep Gamblers: Learning to Abstain with Portfolio Theory", "Self-Adaptive Training: beyond Empirical Risk Minimization"], "answer_arxiv_id": ["1901.09192", "1907.00208", "2002.10319"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_3271"} +{"question": "In the field of sample selection, can you list some studies that improved the concept of two models helping each other out?", "answer": ["How does Disagreement Help Generalization against Label Corruption?"], "answer_arxiv_id": ["1901.04215"], "source_meta": {"published_time": "20230318"}, "qid": "AutoScholarQuery_train_3272"} +{"question": "What papers analyzed stochastic composition optimization problems?", "answer": ["Stochastic Compositional Gradient Descent: Algorithms for Minimizing Compositions of Expected-Value Functions", "Sample Complexity of Sample Average Approximation for Conditional Stochastic Optimization"], "answer_arxiv_id": ["1411.3803v1", "1905.11957"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_3273"} +{"question": "What works propose unlikelihood training as an attempt to address the issue of text degeneration?", "answer": ["Neural Text Generation with Unlikelihood Training"], "answer_arxiv_id": ["1908.04319"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_3274"} +{"question": "Could you specify some works that revolve around the single domain generalization problem?", "answer": ["Adversarially Adaptive Normalization for Single Domain Generalization", "Progressive Domain Expansion Network for Single Domain Generalization", "Learning to Diversify for Single Domain Generalization"], "answer_arxiv_id": ["2106.01899", "2103.16050", "2108.11726"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_3275"} +{"question": "Could you name some works that used fixed or object-like masks for learning-based video inpainting in deep learning?", "answer": ["Deep Flow-Guided Video Inpainting", "FuseFormer: Fusing Fine-Grained Information in Transformers for Video Inpainting", "Learning Joint Spatial-Temporal Transformations for Video Inpainting", "Towards An End-to-End Framework for Flow-Guided Video Inpainting", "Flow-Guided Transformer for Video Inpainting"], "answer_arxiv_id": ["1905.02884", "2109.02974", "2007.10247", "2204.02663", "2208.06768"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_3276"} +{"question": "What are some of the key works in Generative Adversarial Networks (GANs) based models for text-to-image generation?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "MirrorGAN: Learning Text-to-image Generation by Redescription", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image\n Synthesis"], "answer_arxiv_id": ["1711.10485", "1903.05854", "1904.01310"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_3277"} +{"question": "Which open-source toolbox in remote PPG sensing offers numerous algorithm implementations for rPPG sensing, but lacks Python support and neural network training and evaluation?", "answer": ["iPhys: An Open Non-Contact Imaging-Based Physiological Measurement Toolbox"], "answer_arxiv_id": ["1901.04366"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_3278"} +{"question": "Which studies propose a flow-based model to generate molecular graphs conditional to a protein target?", "answer": ["Target-aware Molecular Graph Generation"], "answer_arxiv_id": ["2202.04829"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_3279"} +{"question": "Which works have proposed super-resolution and related upsampling methods as an alternative for correcting under-resolved information from coarse-grid simulations in numerical time-stepping?", "answer": ["Super-resolution reconstruction of turbulent flows with machine learning"], "answer_arxiv_id": ["1811.11328"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_3280"} +{"question": "Can you name some papers that studied similar data corruption models but within statistical context?", "answer": ["Learning Discrete Distributions from Untrusted Batches", "On the Sample Complexity of Adversarial Multi-Source PAC Learning"], "answer_arxiv_id": ["1711.08113", "2002.10384"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3281"} +{"question": "Can you name articles that explored the combination of event cameras with other sensors like standard cameras or inertial units?", "answer": ["Ultimate SLAM? Combining Events, Images, and IMU for Robust Visual SLAM\n in HDR and High Speed Scenarios", "Event-aided Direct Sparse Odometry"], "answer_arxiv_id": ["1709.06310", "2204.07640"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_3282"} +{"question": "Are there any works that explored efficient exploration methods relying on the prediction disagreement of a forward dynamics model?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Planning to Explore via Self-Supervised World Models", "Curious Exploration via Structured World Models Yields Zero-Shot Object Manipulation"], "answer_arxiv_id": ["1507.00814", "2005.05960", "2206.11403"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3283"} +{"question": "Can you cite papers studying generative models using implicit fields in 3D data?", "answer": ["Rodin: A Generative Model for Sculpting 3D Digital Avatars Using\n Diffusion", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "HoloFusion: Towards Photo-realistic 3D Generative Modeling", "RenderDiffusion: Image Diffusion for 3D Reconstruction, Inpainting and\n Generation", "DiffRF: Rendering-Guided 3D Radiance Field Diffusion", "Shap-E: Generating Conditional 3D Implicit Functions", "3DShape2VecSet: A 3D Shape Representation for Neural Fields and\n Generative Diffusion Models", "3DGen: Triplane Latent Diffusion for Textured Mesh Generation", "HyperDiffusion: Generating Implicit Neural Fields with Weight-Space\n Diffusion", "Single-Stage Diffusion NeRF: A Unified Approach to 3D Generation and\n Reconstruction", "Pushing the Limits of 3D Shape Generation at Scale"], "answer_arxiv_id": ["2212.06135", "2212.04493", "2308.14244", "2211.09869", "2212.01206", "2305.02463", "2301.11445", "2303.05371", "2303.17015", "2304.06714", "2306.11510"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3284"} +{"question": "What studies argue that large computing resources are essential for text-to-video generation?", "answer": ["Deep Learning Scaling is Predictable, Empirically", "On the Measure of Intelligence"], "answer_arxiv_id": ["1712.00409", "1911.01547"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_3285"} +{"question": "Could you provide me some works that simulate walks in graph neural networks?", "answer": ["RAW-GNN: RAndom Walk Aggregation based Graph Neural Network"], "answer_arxiv_id": ["2206.13953"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_3286"} +{"question": "Could you provide me some works that modeled prompt uncertainty using Gaussian distribution?", "answer": ["Prompt Distribution Learning"], "answer_arxiv_id": ["2205.03340"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_3287"} +{"question": "Which works focused on Domain Generalization using meta-learning methods?", "answer": ["Episodic Training for Domain Generalization", "Compound Domain Generalization via Meta-Knowledge Encoding", "Learning to Learn with Variational Information Bottleneck for Domain Generalization", "Learning to Generalize: Meta-Learning for Domain Generalization", "Learning to Generalize Unseen Domains via Memory-based Multi-Source Meta-Learning for Person Re-Identification"], "answer_arxiv_id": ["1902.00113", "2203.13006", "2007.07645", "1710.03463", "2012.00417"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_3288"} +{"question": "Could you tell me the studies focusing on text-to code-generation using few-shot prompting?", "answer": ["Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2107.03374"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_3289"} +{"question": "What papers have adopted spatial broadcast decoders to predict RGB images and segmentation masks in unsupervised object-centric learning?", "answer": ["Object-Centric Learning with Slot Attention", "Conditional Object-Centric Learning from Video"], "answer_arxiv_id": ["2006.15055", "2111.12594"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_3290"} +{"question": "What papers report training 3D models using the temporal dimension, rather than a single scan, for contrastive pretext tasks?", "answer": ["Spatio-temporal Self-Supervised Representation Learning for 3D Point\n Clouds"], "answer_arxiv_id": ["2109.00179"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_3291"} +{"question": "Which work introduced the video dataset of CLEVRER to investigate the performance of state-of-the-art models on learning complex spatial-temporal-causal structures from interacting objects?", "answer": ["CLEVRER: Collision Events for Video Representation and Reasoning"], "answer_arxiv_id": ["1910.01442"], "source_meta": {"published_time": "20220618"}, "qid": "AutoScholarQuery_train_3292"} +{"question": "Which works indicated that a classifier trained on a source distribution performs worse on a given target distribution due to the distribution shift?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations", "The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization", "Measuring Robustness to Natural Distribution Shifts in Image Classification", "Improving robustness against common corruptions by covariate shift adaptation"], "answer_arxiv_id": ["1903.12261", "2006.16241", "2007.00644v2", "2006.16971v2"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_3293"} +{"question": "Which papers studied adversarial cases in which agents receive the same loss for the same action chosen at the same time step?", "answer": ["Delay and Cooperation in Nonstochastic Bandits", "Individual Regret in Cooperative Nonstochastic Multi-Armed Bandits", "On Regret-optimal Cooperative Nonstochastic Multi-armed Bandits"], "answer_arxiv_id": ["1602.04741", "1907.03346", "2211.17154"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_3294"} +{"question": "Which studies have addressed recognition in low-quality imagery?", "answer": ["Crystal Loss and Quality Pooling for Unconstrained Face Verification and\n Recognition", "An Automatic System for Unconstrained Video-Based Face Recognition", "Uncertainty Modeling of Contextual-Connections between Tracklets for\n Unconstrained Video-based Face Recognition", "Controllable and Guided Face Synthesis for Unconstrained Face\n Recognition", "AdaFace: Quality Adaptive Margin for Face Recognition", "Teaching Where to Look: Attention Similarity Knowledge Distillation for\n Low Resolution Face Recognition", "Latent Fingerprint Recognition: Fusion of Local and Global Embeddings", "Meet-in-the-middle: Multi-scale upsampling and matching for\n cross-resolution face recognition", "Cluster and Aggregate: Face Recognition with Large Probe Set", "Improving Face Recognition from Hard Samples via Distribution\n Distillation Loss", "FAN: Feature Adaptation Network for Surveillance Face Recognition and\n Normalization"], "answer_arxiv_id": ["1804.01159", "1812.04058", "1905.02756", "2207.10180", "2204.00964", "2209.14498", "2304.13800", "2211.15225", "2210.10864", "2002.03662", "1911.11680"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_3295"} +{"question": "What recent works proposed to strike a balance between the expressive power of GNNs and their efficiency by focusing on subgraph GNNs?", "answer": ["Equivariant Subgraph Aggregation Networks", "Reconstruction for Powerful Graph Representations", "Understanding and Extending Subgraph GNNs by Rethinking Their Symmetries", "Boosting the Cycle Counting Power of Graph Neural Networks with I2-GNNs", "Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning", "Ordered Subgraph Aggregation Networks", "Identity-aware Graph Neural Networks", "A Complete Expressiveness Hierarchy for Subgraph GNNs via Subgraph Weisfeiler-Lehman Tests", "Nested Graph Neural Networks", "From Stars to Subgraphs: Uplifting Any GNN with Local Structure Awareness"], "answer_arxiv_id": ["2110.02910", "2110.00577", "2206.11140", "2210.13978", "2009.00142", "2206.11168", "2101.10320", "2302.07090", "2110.13197", "2110.03753"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_train_3296"} +{"question": "What works used reinforcement learning and neural networks to evaluate new features in automated feature generation?", "answer": ["Feature Engineering for Predictive Modeling using Reinforcement Learning"], "answer_arxiv_id": ["1709.07150"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_3297"} +{"question": "Any works related to the development of methods in video action recognition that capture long-term dependencies through scalable self-attention mechanisms?", "answer": ["ViViT: A Video Vision Transformer", "Is Space-Time Attention All You Need for Video Understanding?", "Video Swin Transformer"], "answer_arxiv_id": ["2103.15691", "2102.05095", "2106.13230"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_3298"} +{"question": "Could you provide me with research that employed unsupervised methods based on single images for object-centric learning?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational Inference", "Object-Centric Learning with Slot Attention", "Illiterate DALL-E Learns to Compose"], "answer_arxiv_id": ["1901.11390", "1903.00450", "2006.15055", "2110.11405"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_3299"} +{"question": "What papers implement hard parameter sharing methods in architecture design for multi-task learning?", "answer": ["UberNet : Training a ‘Universal’ Convolutional Neural Network for Low-, Mid-, and High-Level Vision using Diverse Datasets and Limited Memory", "Learning Multiple Tasks with Multilinear Relationship Networks", "Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Generalist Convolution Kernels"], "answer_arxiv_id": ["1609.02132", "1506.02117", "1908.09597"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_3300"} +{"question": "Which researchers further improved the regret bound to O​(T3/4) in the adversarial setting?", "answer": ["Bandit Multi-linear DR-Submodular Maximization and Its Applications on Adversarial Submodular Bandits"], "answer_arxiv_id": ["2305.12402"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_3301"} +{"question": "What research works have focused on prompting a large language model to generate intermediate reasoning steps?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2201.11903", "2205.11916"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_3302"} +{"question": "What work demonstrated converting discrete melody tokens into a continuous latent space for training a diffusion model?", "answer": ["Symbolic Music Generation with Diffusion Models"], "answer_arxiv_id": ["2103.16091"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_3303"} +{"question": "What works specify the late fusion in cooperative perception to be bandwidth-economic and focus only on transferring perception results?", "answer": ["Cooperative Perception for 3D Object Detection in Driving Scenarios\n using Infrastructure Sensors", "DAIR-V2X: A Large-Scale Dataset for Vehicle-Infrastructure Cooperative\n 3D Object Detection"], "answer_arxiv_id": ["1912.12147", "2204.05575"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_3304"} +{"question": "Which research have developed recurrent neural network(RNN)-based models for ZSVOS?", "answer": ["Learning Video Object Segmentation with Visual Memory", "RVOS: End-to-End Recurrent Network for Video Object Segmentation"], "answer_arxiv_id": ["1704.05737", "1903.05612"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_3305"} +{"question": "Which researches use Deep Sets as a full GNN layer or graph pooling architecture?", "answer": ["Graph Matching Networks for Learning the Similarity of Graph Structured Objects", "Graph Matching Networks for Learning the Similarity of Graph Structured Objects"], "answer_arxiv_id": ["1904.12787", "1904.12787"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_3306"} +{"question": "Which references can provide a comprehensive discussion on the development of active SLAM?", "answer": ["A Survey on Active Simultaneous Localization and Mapping: State of the\n Art and New Frontiers"], "answer_arxiv_id": ["2207.00254"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3307"} +{"question": "Which work claims improvement in 'legal skills' in the context of language model evaluation?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_3308"} +{"question": "What papers proposed the use of Transformers as visual encoders or dynamics models in RL?", "answer": ["Masked World Models for Visual Control", "TransDreamer: Reinforcement Learning with Transformer World Models", "Transformers are Sample-Efficient World Models", "Transformer-based World Models Are Happy With 100k Interactions"], "answer_arxiv_id": ["2206.14244", "2202.09481", "2209.00588", "2303.07109"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3309"} +{"question": "What papers extended the results of implicit bias of gradient descent to non-separable data?", "answer": ["Risk and parameter convergence of logistic regression", "Gradient descent follows the regularization path for general losses"], "answer_arxiv_id": ["1803.07300", "2006.11226"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_3310"} +{"question": "Which work has tried to modify FNO to work with irregular structures via learning a coordinate transformation?", "answer": ["Fourier Neural Operator with Learned Deformations for PDEs on General Geometries"], "answer_arxiv_id": ["2207.05209"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_3311"} +{"question": "Which works have been done on weight decay in the context of training dynamics of DNNs in centralized learning?", "answer": ["Decoupled Weight Decay Regularization", "Three Mechanisms of Weight Decay Regularization", "Understanding the Generalization Benefit of Normalization Layers: Sharpness Reduction"], "answer_arxiv_id": ["1711.05101", "1810.12281", "2206.07085"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_3312"} +{"question": "What studies discuss large language models with multi-modalities?", "answer": ["Attention Is All You Need", "Language Models are Few-Shot Learners", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "LaMDA: Language Models for Dialog Applications", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "Vision Transformers are Parameter-Efficient Audio-Visual Learners", "Tackling Data Bias in MUSIC-AVQA: Crafting a Balanced Dataset for\n Unbiased Question-Answering", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "LoRA: Low-Rank Adaptation of Large Language Models", "QLoRA: Efficient Finetuning of Quantized LLMs"], "answer_arxiv_id": ["1706.03762", "2005.14165", "2307.09288", "2201.08239", "2304.10592", "2306.02858", "2212.07983", "2310.06238", "2303.16199", "2304.15010", "2106.09685", "2305.14314"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_3313"} +{"question": "Which papers introduced multimodal LLMs?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation\n Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2301.12597", "2304.10592", "2304.08485", "2303.04671", "2304.14178"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_3314"} +{"question": "Which work proposed the FactorVAE that seeks to isolate the Total Correlation (TC) component by employing a large discriminator neural network?", "answer": ["Disentangling by Factorising"], "answer_arxiv_id": ["1802.05983"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_3315"} +{"question": "What studies have used a Vision Transformer (ViT) for web navigation?", "answer": ["Multimodal Web Navigation with Instruction-Finetuned Foundation Models"], "answer_arxiv_id": ["2305.11854"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_3316"} +{"question": "Which works demonstrated that quantization schemes can efficiently improve memory allocation without performance degradation?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale"], "answer_arxiv_id": ["2208.07339"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3317"} +{"question": "What research has been done in finding anchor examples in classification datasets?", "answer": ["Anchor Points: Benchmarking Models with Much Fewer Examples"], "answer_arxiv_id": ["2309.08638"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_3318"} +{"question": "What studies discuss advanced mixup strategies as a noise-based data augmentation strategy?", "answer": ["Sequence-Level Mixed Sample Data Augmentation"], "answer_arxiv_id": ["2011.09039"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_3319"} +{"question": "Which works discuss the blueprint text plan, a concept used for content selection and organisation of generated text?", "answer": ["Conditional Generation with a Question-Answering Blueprint"], "answer_arxiv_id": ["2207.00397"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_3320"} +{"question": "What papers have shown that when the arms are standard basis vectors, logistic bandits are equivalent to Bernoulli bandits?", "answer": ["Instance-Wise Minimax-Optimal Algorithms for Logistic Bandits"], "answer_arxiv_id": ["2010.12642v2"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_3321"} +{"question": "Could you tell me about the work that developed the geometric aggregation scheme to enhance convolutions in Geom-GCN?", "answer": ["Geom-GCN: Geometric Graph Convolutional Networks"], "answer_arxiv_id": ["2002.05287"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_3322"} +{"question": "What are the works on mechanistic interpretability aiming to reverse engineer neural networks?", "answer": ["Interpreting Neural Networks through the Polytope Lens", "A Tale of Two Circuits: Grokking as Competition of Sparse and Dense Subnetworks"], "answer_arxiv_id": ["2211.12312", "2303.11873"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_3323"} +{"question": "Could you identify the research works that brought domain-specific modules into transformer models for efficiency?", "answer": ["MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Swin Transformer V2: Scaling Up Capacity and Resolution", "CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows"], "answer_arxiv_id": ["2110.02178", "2103.14030", "2111.09883", "2107.00652"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_3324"} +{"question": "Could you identify some research that designs supervised primitive networks to detect and fit primitives within point clouds?", "answer": ["Supervised Fitting of Geometric Primitives to 3D Point Clouds", "ParSeNet: A Parametric Surface Fitting Network for 3D Point Clouds"], "answer_arxiv_id": ["1811.08988", "2003.12181"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_3325"} +{"question": "Which works focused on the problem of designing optimal scoring rules to incentivize an agent to acquire and report costly information in terms of computational problem of information acquisition?", "answer": ["Contracts with Information Acquisition, via Scoring Rules", "Optimization of Scoring Rules", "Binary Scoring Rules that Incentivize Precision"], "answer_arxiv_id": ["2204.01773", "2007.02905v2", "2002.10669"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3326"} +{"question": "What studies used diffusion models in text-to-image generation?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion", "AnyDoor: Zero-shot Object-level Image Customization", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "Cascaded Diffusion Models for High Fidelity Image Generation", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models without Specific Tuning", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2112.10741", "2112.10752", "2208.01618", "2307.09481", "2211.01324", "2108.01073", "2106.15282", "2307.04725", "2006.11239", "2010.02502"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3327"} +{"question": "Could you provide me information about works in point- and mesh-based dynamics models?", "answer": ["Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids", "Flexible Neural Representation for Physics Prediction", "Learning to Simulate Complex Physics with Graph Networks", "Learning Mesh-Based Simulation with Graph Networks"], "answer_arxiv_id": ["1810.01566", "1806.08047", "2002.09405", "2010.03409"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_3328"} +{"question": "What are some studies that applied NeRFs principle to spatial audio?", "answer": ["Learning Neural Acoustic Fields"], "answer_arxiv_id": ["2204.00628"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_3329"} +{"question": "Which work proposed dynamic importance weighting to make traditional IW methods compatible with stochastic optimizers?", "answer": ["Rethinking Importance Weighting for Deep Learning under Distribution Shift"], "answer_arxiv_id": ["2006.04662"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_3330"} +{"question": "Which papers have used F-score in assessing the accuracy of mesh shapes?", "answer": ["Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images", "Local Deep Implicit Functions for 3D Shape", "Fostering Generalization in Single-view 3D Reconstruction by Learning a\n Hierarchy of Local and Global Shape Priors", "What Do Single-view 3D Reconstruction Networks Learn?"], "answer_arxiv_id": ["1804.01654", "1912.06126", "2104.00476", "1905.03678"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_3331"} +{"question": "Which papers studied learning a single neuron by gradient-based algorithms in the context of realisable without bias?", "answer": ["Learning a Single Neuron with Gradient Methods"], "answer_arxiv_id": ["2001.05205"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_3332"} +{"question": "Any works investing in uncertainty estimation in context of distribution shift?", "answer": ["Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift"], "answer_arxiv_id": ["1906.02530"], "source_meta": {"published_time": "20220220"}, "qid": "AutoScholarQuery_train_3333"} +{"question": "What studies are about memory recall based probing in transformers?", "answer": ["Analyzing Transformers in Embedding Space", "Dissecting Recall of Factual Associations in Auto-Regressive Language Models"], "answer_arxiv_id": ["2209.02535", "2304.14767"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3334"} +{"question": "What works have found trajectory balance and subtrajectory balance losses to be more efficient in GFlowNets?", "answer": ["Trajectory balance: Improved credit assignment in GFlowNets", "Learning GFlowNets From Partial Episodes For Improved Convergence And Stability"], "answer_arxiv_id": ["2201.13259", "2209.12782"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_3335"} +{"question": "What works have succeeded in reducing the number of discretization steps while maintaining small approximation error?", "answer": ["Denoising Diffusion Implicit Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "Fast Sampling of Diffusion Models with Exponential Integrator", "GENIE: Higher-Order Denoising Diffusion Solvers"], "answer_arxiv_id": ["2010.02502", "2201.06503", "2204.13902", "2210.05475"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_3336"} +{"question": "What works have looked into random or rule-based policy switching in Intervention-based Reinforcement Learning?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "ReIL: A Framework for Reinforced Intervention-based Imitation Learning"], "answer_arxiv_id": ["1011.0686v3", "2203.15390v1"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_3337"} +{"question": "Which works have implemented quantization as a method of communication compression?", "answer": ["RATQ: A Universal Fixed-Length Quantizer for Stochastic Optimization", "vqSGD: Vector Quantized Stochastic Gradient Descent", "signSGD: Compressed Optimisation for Non-Convex Problems", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "TernGrad: Ternary Gradients to Reduce Communication in Distributed Deep Learning", "Natural Compression for Distributed Deep Learning"], "answer_arxiv_id": ["1908.08200", "1911.07971v4", "1802.04434", "1610.02132", "1705.07878", "1905.10988"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3338"} +{"question": "Which papers discussed the application of beta-VAEs in data compression?", "answer": ["End-to-end Optimized Image Compression", "Joint Autoregressive and Hierarchical Priors for Learned Image Compression", "Improving Inference for Neural Image Compression"], "answer_arxiv_id": ["1611.01704", "1809.02736", "2006.04240"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_3339"} +{"question": "What papers have studied principal component analysis (PCA) in the context of private feature selection?", "answer": ["Beating Randomized Response on Incoherent Matrices"], "answer_arxiv_id": ["1111.0623"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3340"} +{"question": "Which papers have contributed to LiDAR-based 3D detection that aids robots in understanding large-scale scenes?", "answer": ["SSN: Shape Signature Networks for Multi-class Object Detection from\n Point Clouds", "Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR-based\n Perception", "Center-based 3D Object Detection and Tracking", "VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "SqueezeSegV2: Improved Model Structure and Unsupervised Domain\n Adaptation for Road-Object Segmentation from a LiDAR Point Cloud"], "answer_arxiv_id": ["2004.02774", "2109.05441", "2006.11275", "1711.06396", "1812.05784", "1809.08495"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3341"} +{"question": "Can you point me to studies that tested the performance of Chat-GPT in task-oriented dialogues?", "answer": ["Guiding Large Language Models via Directional Stimulus Prompting"], "answer_arxiv_id": ["2302.11520"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_3342"} +{"question": "Can you provide examples of works that have used supervised pre-training in tasks like classification, retrieval, and captioning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "PaLI: A Jointly-Scaled Multilingual Language-Image Model"], "answer_arxiv_id": ["2103.00020", "2202.03052", "2209.06794"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3343"} +{"question": "Which works focus on handling spatial information in Tracking Any Point (TAP)?", "answer": ["CoTracker: It is Better to Track Together"], "answer_arxiv_id": ["2307.07635"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_3344"} +{"question": "Which paper discusses the concept of Label Smoothing for model calibration?", "answer": ["When Does Label Smoothing Help?"], "answer_arxiv_id": ["1906.02629"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_3345"} +{"question": "Which researches improved upon past trust-region approaches using Penalized Proximal Policy Optimization?", "answer": ["Penalized Proximal Policy Optimization for Safe Reinforcement Learning"], "answer_arxiv_id": ["2205.11814"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_3346"} +{"question": "Which studies in class-agnostic object counting use exemplar bounding boxes of a new category for target object counting?", "answer": ["Training-free Object Counting with Prompts", "A Low-Shot Object Counting Network With Iterative Prototype Adaptation", "Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector", "CounTR: Transformer-based Generalised Visual Counting", "Represent, Compare, and Learn: A Similarity-Aware Framework for\n Class-Agnostic Counting", "Class-Agnostic Counting"], "answer_arxiv_id": ["2307.00038", "2211.08217", "1908.01998", "2208.13721", "2203.08354", "1811.00472"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_3347"} +{"question": "What recent recurrent-free models are highlighted for their strong performance?", "answer": ["PastNet: Introducing Physical Inductive Biases for Spatio-temporal Video Prediction", "Implicit Stacked Autoregressive Model for Video Prediction"], "answer_arxiv_id": ["2305.11421", "2303.07849"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3348"} +{"question": "What works found 'skill' neurons highly predictive of the downstream task in soft prompt-tuning of language models?", "answer": ["Finding Skill Neurons in Pre-trained Transformer-based Language Models"], "answer_arxiv_id": ["2211.07349"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_3349"} +{"question": "What papers used uncertainty-based algorithms in multi-view classification for trusted decision-making?", "answer": ["Trusted Multi-View Classification"], "answer_arxiv_id": ["2102.02051"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_3350"} +{"question": "Which papers delved into the implicit regularization of gradient descent-based learning?", "answer": ["Implicit Regularization in ReLU Networks with the Square Loss"], "answer_arxiv_id": ["2012.05156"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_3351"} +{"question": "Which research introduced the idea of using depth maps, reflection maps, and R-PPG signals as diverse supervisory inputs to enhance detection capabilities?", "answer": ["Revisiting Pixel-Wise Supervision for Face Anti-Spoofing", "Structure Destruction and Content Combination for Face Anti-Spoofing"], "answer_arxiv_id": ["2011.12032", "2107.10628"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3352"} +{"question": "What research papers focus on limitations of RMSProp and Adam algorithms?", "answer": ["On the convergence of Adam and Beyond", "Adam Can Converge Without Any Modification On Update Rules"], "answer_arxiv_id": ["1904.09237", "2208.09632v5"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_3353"} +{"question": "Could you provide me some studies about training methods which make unlearning efficient?", "answer": ["SAFE: Machine Unlearning With Shard Graphs", "Training Data Protection with Compositional Diffusion Models", "Machine Unlearning", "Erasing Concepts from Diffusion Models", "Ablating Concepts in Text-to-Image Diffusion Models", "Tangent Transformers for Composition, Privacy and Removal"], "answer_arxiv_id": ["2304.13169", "2308.01937", "1912.03817", "2303.07345", "2303.13516", "2307.08122"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_3354"} +{"question": "What is the original work that introduced the Invariant Risk Minimization (IRM) framework?", "answer": ["Invariant Risk Minimization", "Invariant Risk Minimization Games"], "answer_arxiv_id": ["1907.02893", "2002.04692"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_3355"} +{"question": "What studies focus on aggregating and standardizing classical NLP tasks to evaluate language models?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding", "SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems"], "answer_arxiv_id": ["1804.07461", "1905.00537"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_3356"} +{"question": "Could you provide works that conduct techniques learn representations by contrasting nodes with negative samples?", "answer": ["LINE: Large-scale Information Network Embedding", "PTE: Predictive Text Embedding through Large-scale Heterogeneous Text Networks"], "answer_arxiv_id": ["1503.03578", "1508.00200"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3357"} +{"question": "What papers provide in-depth perspective on Dempster-Shafer thoery as a model for uncertainty estimation?", "answer": ["Trusted Multi-View Classification"], "answer_arxiv_id": ["2102.02051"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_3358"} +{"question": "What research focused on finding optimized methods to elicit the model’s knowledge obtained during pretraining?", "answer": ["Give Me the Facts! A Survey on Factual Knowledge Probing in Pre-trained\n Language Models", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2310.16570", "2107.13586v1"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_3359"} +{"question": "What work utilizes a learnable mechanism that determines necessary tokens during inference?", "answer": ["Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers"], "answer_arxiv_id": ["2305.15805"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_3360"} +{"question": "Could you provide me some works that focused on decoupling gaming and improving effects in both learning and evaluation?", "answer": ["Linear Classifiers that Encourage Constructive Adaptation"], "answer_arxiv_id": ["2011.00355"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_3361"} +{"question": "Which research papers utilize conditional normalizing flow-based models?", "answer": ["SRFlow: Learning the Super-Resolution Space with Normalizing Flow"], "answer_arxiv_id": ["2006.14200"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_3362"} +{"question": "Which works does the researcher refer to when talking about federated learning using a supervised training framework like FedAvg?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_3363"} +{"question": "Are there researches extrapolate full surrounding from a partial scene captured in a limited single view frustum?", "answer": ["DeepLight: Learning Illumination for Unconstrained Mobile Mixed Reality", "Learning to Predict Indoor Illumination from a Single Image", "Deep Parametric Indoor Lighting Estimation", "Sparse Needlets for Lighting Estimation with Spherical Transport Loss", "EMLight: Lighting Estimation via Spherical Distribution Approximation", "GMLight: Lighting Estimation via Geometric Distribution Approximation", "EverLight: Indoor-Outdoor Editable HDR Lighting Estimation"], "answer_arxiv_id": ["1904.01175", "1704.00090", "1910.08812", "2106.13090", "2012.11116", "2102.10244", "2304.13207"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_3364"} +{"question": "Can you cite some studies on online learning in simple auctions?", "answer": ["The Sample Complexity of Revenue Maximization", "Real-Time Optimisation for Online Learning in Auctions", "Online learning in repeated auctions.", "Optimal No-regret Learning in Repeated First-price Auctions", "Learning in Auctions: Regret is Hard, Envy is Easy"], "answer_arxiv_id": ["1502.00963", "2010.10070", "1511.05720", "2003.09795", "1511.01411"], "source_meta": {"published_time": "20220813"}, "qid": "AutoScholarQuery_train_3365"} +{"question": "Which work developed a method for dealing with covariate shift in non-exchangeable data?", "answer": ["Conformal Prediction Under Covariate Shift"], "answer_arxiv_id": ["1904.06019"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_3366"} +{"question": "Could you provide me some studies that used SDF to infer the shape of hand-held objects?", "answer": ["What's in your hands? 3D Reconstruction of Generic Objects in Hands"], "answer_arxiv_id": ["2204.07153"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_3367"} +{"question": "What studies the robustness to delays in continuized and asynchronous algorithms?", "answer": ["Improved Asynchronous Parallel Optimization Analysis for Stochastic Incremental Methods"], "answer_arxiv_id": ["1801.03749"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_3368"} +{"question": "Which papers are focused on open-ended text-guided image editing combining large language models with text-to-image generators?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2211.09800"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_3369"} +{"question": "Which papers used a patch discriminator to model high-frequencies?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks"], "answer_arxiv_id": ["1611.07004", "1703.10593"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3370"} +{"question": "Could you provide me some works that focus on defending against adversarial attacks on RL?", "answer": ["Defense Against Reward Poisoning Attacks in Reinforcement Learning", "Robust Policy Gradient against Strong Data Corruption", "Corruption-robust exploration in episodic reinforcement learning", "Improved Corruption Robust Algorithms for Episodic Reinforcement Learning", "A Model Selection Approach for Corruption Robust Reinforcement Learning"], "answer_arxiv_id": ["2102.05776", "2102.05800v3", "1911.08689", "2102.06875", "2110.03580v1"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_3371"} +{"question": "What papers describe traditional approaches to 3D scene segmentation that focus on point clouds or voxels?", "answer": ["Joint 2D-3D-Semantic Data for Indoor Scene Understanding", "SemanticKITTI: A Dataset for Semantic Scene Understanding of LiDAR Sequences", "nuScenes: A multimodal dataset for autonomous driving", "Matterport3D: Learning from RGB-D Data in Indoor Environments", "ScanRefer: 3D Object Localization in RGB-D Scans using Natural Language", "ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "KITTI-360: A Novel Dataset and Benchmarks for Urban Scene Understanding in 2D and 3D", "PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical Part-level 3D Object Understanding", "Scalability in Perception for Autonomous Driving: Waymo Open Dataset", "SEGCloud: Semantic Segmentation of 3D Point Clouds", "Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs"], "answer_arxiv_id": ["1702.01105", "1904.01416", "1903.11027", "1709.06158", "1912.08830", "1702.04405v2", "2109.13410", "1812.02713", "1912.04838", "1710.07563", "1711.09869"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_3372"} +{"question": "What works developed Best-of-both-worlds (BOBW) algorithms for online decision-making problems?", "answer": ["The best of both worlds: stochastic and adversarial bandits", "An Improved Parametrization and Analysis of the EXP3++ Algorithm for Stochastic and Adversarial Bandits", "Tsallis-INF: An Optimal Algorithm for Stochastic and Adversarial Bandits", "A Second-order Bound with Excess Losses", "More Adaptive Algorithms for Adversarial Bandits", "The best of both worlds: stochastic and adversarial episodic MDPs with unknown transition"], "answer_arxiv_id": ["1202.4473", "1702.06103", "1807.07623", "1402.2044", "1801.03265", "2106.04117"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_3373"} +{"question": "What studies have investigated the connections between Graph Neural Networks and the Weisfeiler-Lehman (WL) test of isomorphism?", "answer": ["Expressiveness and Approximation Properties of Graph Neural Networks", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["2204.04661", "1810.02244", "1810.00826"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_3374"} +{"question": "Which studies have developed dehazing algorithms for DaSID?", "answer": ["Contrastive Learning for Compact Single Image Dehazing", "Vision Transformers for Single Image Dehazing", "Mutual Information-driven Triple Interaction Network for Efficient Image\n Dehazing", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2104.09367", "2204.03883", "2308.06998", "2103.14030"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_3375"} +{"question": "Which research proposes to adjust local model training at the client side for reducing the difference in local models?", "answer": ["Federated Optimization in Heterogeneous Networks", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning"], "answer_arxiv_id": ["1812.06127", "1910.06378"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3376"} +{"question": "What papers have studied control tasks from the perspective of regret?", "answer": ["Online Control with Adversarial Disturbances", "Improper Learning for Non-Stochastic Control", "Naive Exploration is Optimal for Online LQR", "Online Linear Quadratic Control"], "answer_arxiv_id": ["1902.08721", "2001.09254", "2001.09576", "1806.07104"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_3377"} +{"question": "What papers discuss Bayesian neural networks in the context of uncertainty estimation in visual recognition models?", "answer": ["Bayesian Convolutional Neural Networks with Bernoulli Approximate\n Variational Inference", "Efficient and Scalable Bayesian Neural Nets with Rank-1 Factors", "Being Bayesian, Even Just a Bit, Fixes Overconfidence in ReLU Networks"], "answer_arxiv_id": ["1506.02158", "2005.07186", "2002.10118v2"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_3378"} +{"question": "What works study the combinations of pruning and quantization with different degrees of granularity?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding", "Bayesian Bits: Unifying Quantization and Pruning", "OPQ: Compressing Deep Neural Networks with One-shot Pruning-Quantization", "Automatic Neural Network Compression by Sparsity-Quantization Joint Learning: A Constrained Optimization-based Approach"], "answer_arxiv_id": ["1510.00149", "2005.07093", "2205.11141", "1910.05897"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_3379"} +{"question": "Which works mentioned Neural Motion Planning strategies?", "answer": ["Learning Sampling Distributions for Robot Motion Planning", "LEGO: Leveraging Experience in Roadmap Generation for Sampling-Based Planning", "Deeply Informed Neural Sampling for Robot Motion Planning", "Motion Planning Networks", "Motion Planning Networks: Bridging the Gap Between Learning-based and Classical Motion Planners", "Robot Motion Planning in Learned Latent Spaces", "iSDF: Real-Time Neural Signed Distance Fields for Robot Perception", "Regularized Deep Signed Distance Fields for Reactive Motion Generation", "Predicted Composite Signed-Distance Fields for Real-Time Motion Planning in Dynamic Environments", "Vision-Only Robot Navigation in a Neural Radiance World"], "answer_arxiv_id": ["1709.05448", "1907.09574", "1809.10252", "1806.05767", "1907.06013", "1807.10366", "2204.02296", "2203.04739", "2008.00969", "2110.00168"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_3380"} +{"question": "Could you provide me with the studies related to the AraCOCO?", "answer": ["Violet: A Vision-Language Model for Arabic Image Captioning with Gemini\n Decoder"], "answer_arxiv_id": ["2311.08844"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_3381"} +{"question": "Could you provide me some studies about the use of higher-order (k-dimensional) GNNs and hierarchical variants?", "answer": ["Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks"], "answer_arxiv_id": ["1810.02244"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_3382"} +{"question": "Which papers used received words randomly taken from the AWGN channel for training sample selection?", "answer": ["Deep Learning Methods for Improved Decoding of Linear Codes", "Learned Belief-Propagation Decoding with Simple Scaling and SNR Adaptation"], "answer_arxiv_id": ["1706.07043", "1901.08621"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_3383"} +{"question": "Which studies involve text-to-image methods relying on GANs and VQVAEs?", "answer": ["Generative Adversarial Nets", "Analyzing and Improving the Image Quality of StyleGAN", "Neural Discrete Representation Learning", "Taming Transformers for High-Resolution Image Synthesis", "Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers"], "answer_arxiv_id": ["1406.2661", "1912.04958", "1711.00937", "2012.09841", "2102.12092", "2105.13290"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_3384"} +{"question": "What research was conducted on the application of momentum prototypes in weakly supervised learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1911.05722"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_3385"} +{"question": "Which work utilized dense contrast in the field of contrastive learning?", "answer": ["Dense Contrastive Learning for Self-Supervised Visual Pre-Training"], "answer_arxiv_id": ["2011.09157"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_3386"} +{"question": "What works are relevant for training a ConvNet or a ViT from a single image or a single long video?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "A critical analysis of self-supervision, or what we can learn from a\n single image"], "answer_arxiv_id": ["2010.11929", "1904.13132"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_3387"} +{"question": "Can you provide some research on self-supervised learning models based on autoencoder and generative adversarial networks for 3D object recognition?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling", "VConv-DAE: Deep Volumetric Shape Learning Without Object Labels"], "answer_arxiv_id": ["1610.07584", "1604.03755"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_3388"} +{"question": "Which research proposed Riemannian Flow Matching (RFM)?", "answer": ["Riemannian Flow Matching on General Geometries"], "answer_arxiv_id": ["2302.03660"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_3389"} +{"question": "What are some recent sources dealing with Text-To-Image generation task in the context of conditional image synthesis?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2205.11487", "2112.10752"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_3390"} +{"question": "What research discuss about Open-set domain adaptation (OSDA) that assumes the presence of unknown classes in the target domain?", "answer": ["Open Set Domain Adaptation by Backpropagation"], "answer_arxiv_id": ["1804.10427"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_3391"} +{"question": "Which studies show the role of the feed-forward (FF) layer in the construction of an LLM’s prediction?", "answer": ["Transformer Feed-Forward Layers Build Predictions by Promoting Concepts\n in the Vocabulary Space"], "answer_arxiv_id": ["2203.14680"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_3392"} +{"question": "Any studies about the development of MLLMs like Vary and Monkey for document image understanding?", "answer": ["Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models", "Monkey: Image Resolution and Text Label Are Important Things for Large\n Multi-modal Models"], "answer_arxiv_id": ["2312.06109", "2311.06607"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_3393"} +{"question": "What works focus on training networks for scene understanding in a more data-efficient manner using self-supervised and semi-supervised methods?", "answer": ["Learning from 2D: Contrastive Pixel-to-Point Knowledge Transfer for 3D Pretraining", "Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Data", "Segment Any Point Cloud Sequences by Distilling Vision Foundation Models", "Model2Scene: Learning 3D Scene Representation via Contrastive Language-CAD models Pre-training", "NeRF-SOS: Any-View Self-supervised Object Segmentation on Complex Scenes", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Unsupervised Visual Representation Learning by Context Prediction", "Exploring Simple Siamese Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Towards 3D Scene Understanding by Referring Synthetic Models", "Unsupervised Learning of Intrinsic Structural Representation Points", "Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision", "Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels", "ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentation", "Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning", "Semi-Supervised Medical Image Segmentation via Cross Teaching between CNN and Transformer", "A Survey on Label-efficient Deep Image Segmentation: Bridging the Gap between Weak Supervision and Dense Prediction", "Semi-Supervised Semantic Segmentation with Cross-Consistency Training", "Semi-supervised Anatomical Landmark Detection via Shape-regulated Self-training"], "answer_arxiv_id": ["2104.04687", "2203.16258", "2306.09347", "2309.16956", "2209.08776", "1911.05722", "2006.07733", "1505.05192", "2011.10566", "2002.05709", "2006.09882", "2203.10546", "2003.01661", "2106.01226", "2203.03884", "2106.05095", "2110.05474", "2112.04894", "2207.01223", "2003.09005", "2105.13593"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3394"} +{"question": "What are the works that discuss the relative performance of zero-shot learning and supervised learning for Hate Speech Detection (HSD)?", "answer": ["An Investigation of Large Language Models for Real-World Hate Speech\n Detection"], "answer_arxiv_id": ["2401.03346"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_3395"} +{"question": "What papers derived an Stochastic Differential Equation (SDE) for Stochastic Gradient Descent (SGD) approximation?", "answer": ["Stochastic modified equations and adaptive stochastic gradient algorithms", "Stochastic Modified Equations and Dynamics of Stochastic Gradient Algorithms I: Mathematical Foundations"], "answer_arxiv_id": ["1511.06251", "1811.01558"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_3396"} +{"question": "Which works used planning algorithms in NLP tasks to optimize text output for specific objectives?", "answer": ["To Beam Or Not To Beam: That is a Question of Cooperation for Language GANs", "Machine Translation Decoding beyond Beam Search", "PPL-MCTS: Constrained Textual Generation Through Discriminator-Guided MCTS Decoding"], "answer_arxiv_id": ["2106.06363", "2104.05336", "2109.13582"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_3397"} +{"question": "Can you mention some of the studies on adaptation and robustness to various distribution shifts?", "answer": ["Deep Domain Confusion: Maximizing for Domain Invariance", "What is the Effect of Importance Weighting in Deep Learning?", "Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty", "Invariant Risk Minimization", "Do Adversarially Robust ImageNet Models Transfer Better?", "Just Train Twice: Improving Group Robustness without Training Group Information", "A Fine-Grained Analysis on Distribution Shift", "The Evolution of Out-of-Distribution Robustness Throughout Fine-Tuning", "Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization", "Environment Inference for Invariant Learning", "Diversify and Disambiguate: Learning From Underspecified Data", "Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution"], "answer_arxiv_id": ["1412.3474", "1812.03372", "1906.12340", "1907.02893", "2007.08489", "2107.09044", "2110.11328", "2106.15831", "2107.04649", "2010.07249", "2202.03418v3", "2202.10054v1"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_3398"} +{"question": "What research projects focused on learning the deformation from the actual 3D space into another 3D space?", "answer": ["Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Video", "Disentangled3D: Learning a 3D Generative Model with Disentangled Geometry and Appearance from Monocular Images", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Neural Actor: Neural Free-view Synthesis of Human Actors with Pose Control", "Animatable Neural Radiance Fields for Modeling Dynamic Human Bodies"], "answer_arxiv_id": ["2012.12247", "2203.15926", "2011.13961", "2106.02019", "2105.02872"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_train_3399"} +{"question": "What works discuss and apply the Membership Inference Attack in identifying training subsets of a larger data pool?", "answer": ["Membership Inference Attacks against Machine Learning Models", "Are Diffusion Models Vulnerable to Membership Inference Attacks?", "An Efficient Membership Inference Attack for the Diffusion Model by\n Proximal Initialization", "White-box Membership Inference Attacks against Diffusion Models"], "answer_arxiv_id": ["1610.05820", "2302.01316", "2305.18355", "2308.06405"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_3400"} +{"question": "Which autoencoder-based approaches optimize the encoder to adapt to each instance leaving the decoder fixed?", "answer": ["Content Adaptive Optimization for Neural Image Compression", "Improving Inference for Neural Image Compression", "Content Adaptive and Error Propagation Aware Deep Video Compression"], "answer_arxiv_id": ["1906.01223", "2006.04240", "2003.11282"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3401"} +{"question": "Which research papers have applied Learning to Rank for hyperparameter optimization?", "answer": ["Learning to Rank Learning Curves"], "answer_arxiv_id": ["2006.03361"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_3402"} +{"question": "What papers are introducing tiny object detectors that balance between accuracy and speed?", "answer": ["MRDet: A Multi-Head Network for Accurate Oriented Object Detection in\n Aerial Images", "QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small\n Object Detection", "Interactive Multi-Class Tiny-Object Detection", "Small Object Detection via Coarse-to-fine Proposal Generation and\n Imitation Learning"], "answer_arxiv_id": ["2012.13135", "2103.09136", "2203.15266", "2308.09534"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3403"} +{"question": "Which work proposed a framework for transforming offline problems to online methods in an adversarial bandit setting?", "answer": ["Online Learning via Offline Greedy Algorithms: Applications in Market Design and Optimization"], "answer_arxiv_id": ["2102.11050v4"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_3404"} +{"question": "What papers researched the representation theory focusing on the representation power of neural networks?", "answer": ["On the Expressive Power of Deep Neural Networks"], "answer_arxiv_id": ["1606.05336"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_3405"} +{"question": "Could you provide me some studies that proved non-asymptotic convergence rates for policy-based primal-dual algorithms?", "answer": ["A Primal-Dual Approach to Constrained Markov Decision Processes", "Policy Optimization for Constrained MDPs with Provable Fast Global Convergence", "Towards Painless Policy Optimization for Constrained MDPs", "Faster Algorithm and Sharper Analysis for Constrained Markov Decision Process", "Finite-Time Complexity of Online Primal-Dual Natural Actor-Critic Algorithm for Constrained Markov Decision Processes", "Anchor-Changing Regularized Natural Policy Gradient for Multi-Objective Reinforcement Learning"], "answer_arxiv_id": ["2101.10895", "2111.00552", "2204.05176", "2110.10351v1", "2110.11383v2", "2206.05357"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3406"} +{"question": "Can you provide the references where the Transformer neural network is used in the context of neural text generation?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_3407"} +{"question": "Can you mention instances where synthetically generated data have been used to bootstrap performance on real images?", "answer": ["CARLA: An Open Urban Driving Simulator", "Virtual Worlds as Proxy for Multi-Object Tracking Analysis", "6-DoF Pose Estimation of Household Objects for Robotic Manipulation: An\n Accessible Dataset and Benchmark", "BOP Challenge 2020 on 6D Object Localization", "HomebrewedDB: RGB-D Dataset for 6D Pose Estimation of 3D Objects"], "answer_arxiv_id": ["1711.03938", "1605.06457", "2203.05701", "2009.07378", "1904.03167"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_3408"} +{"question": "What works explore methods for probabilistic dynamics forecasting?", "answer": ["Quantifying Uncertainty in Deep Spatiotemporal Forecasting"], "answer_arxiv_id": ["2105.11982"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_3409"} +{"question": "Which work suggested negative impacts of training LLMs for multiple epochs?", "answer": ["Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2203.15556"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_3410"} +{"question": "Can you mention some studies that investigated the convergence properties of gradient-based optimization methods in context with NNGP correspondence?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Global Convergence of Deep Networks with One Wide Layer Followed by Pyramidal Topology", "On Exact Computation with an Infinitely Wide Neural Net", "On the Proof of Global Convergence of Gradient Descent for Deep ReLU Networks with Linear Widths", "A Global Convergence Theory for Deep ReLU Implicit Networks via Over-parameterization", "Gradient Descent Optimizes Infinite-Depth ReLU Implicit Networks with Linear Widths"], "answer_arxiv_id": ["1806.07572", "1811.03804", "1811.03962", "2002.07867", "1904.11955", "2101.09612", "2110.05645", "2205.07463"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_3411"} +{"question": "What is the study that introduces small perturbations into input images to facilitate the separation of softmax score for OOD detection?", "answer": ["Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks"], "answer_arxiv_id": ["1706.02690"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3412"} +{"question": "What studies were the first to propose prompting in NLP tasks?", "answer": ["Language Models are Few-Shot Learners", "Making Pre-trained Language Models Better Few-shot Learners", "How Can We Know What Language Models Know?"], "answer_arxiv_id": ["2005.14165", "2012.15723", "1911.12543"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_3413"} +{"question": "Can you provide studies where LLM's failure modes were analysed by prompts constructed from templates?", "answer": ["Adversarial Examples for Evaluating Reading Comprehension Systems", "Counterfactual Fairness in Text Classification through Robustness", "Capturing Failures of Large Language Models via Human Cognitive Biases"], "answer_arxiv_id": ["1707.07328", "1809.10610", "2202.12299"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_3414"} +{"question": "Which research papers propose versions of IWERM to help control the variance when facing a high variance due to low number of the training data?", "answer": ["Relative Density-Ratio Estimation for Robust Distribution Comparison"], "answer_arxiv_id": ["1106.4729"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3415"} +{"question": "Which studies have shown the applicability of pre-trained diffusion models in image-to-image translation?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Zero-shot Image-to-Image Translation", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2208.01626", "2302.03027", "2211.09800", "2211.12572"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_3416"} +{"question": "Which works introduced disagreement of an ensemble of world models as an estimate of expected information gain?", "answer": ["Self-Supervised Exploration via Disagreement", "Planning to Explore via Self-Supervised World Models", "Curious Exploration via Structured World Models Yields Zero-Shot Object Manipulation"], "answer_arxiv_id": ["1906.04161", "2005.05960", "2206.11403"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_3417"} +{"question": "Could you point out which work proposed the method for learning an invariant representation for covariate shift adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks"], "answer_arxiv_id": ["1505.07818"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3418"} +{"question": "What research propose a relaxed subset selection algorithm?", "answer": ["Reparameterizable Subset Sampling via Continuous Relaxations"], "answer_arxiv_id": ["1901.10517"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_3419"} +{"question": "Which works proposed the TRADES method?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1901.08573"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_3420"} +{"question": "What works provide theoretical guarantees with no boundedness assumptions on the features or labels?", "answer": ["Differentially Private Regression with Unbounded Covariates"], "answer_arxiv_id": ["2202.11199"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_3421"} +{"question": "Could you provide me some studies about multi-scale motion blur removal methods?", "answer": ["Deep Multi-scale Convolutional Neural Network for Dynamic Scene\n Deblurring", "Scale-recurrent Network for Deep Image Deblurring", "Multi-Temporal Recurrent Neural Networks For Progressive Non-Uniform\n Single Image Deblurring With Incremental Temporal Training", "Rethinking Coarse-to-Fine Approach in Single Image Deblurring", "Learning Degradation Representations for Image Deblurring"], "answer_arxiv_id": ["1612.02177", "1802.01770", "1911.07410", "2108.05054", "2208.05244"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_3422"} +{"question": "What are some researches in which Grokking-like phenomena have been identified?", "answer": ["Omnigrok: Grokking Beyond Algorithmic Data", "Grokking of Hierarchical Structure in Vanilla Transformers"], "answer_arxiv_id": ["2210.01117", "2305.18741"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_3423"} +{"question": "Which works show that KD methods can significantly reduce model complexity with barely sacrificing task performance?", "answer": ["Born-Again Neural Networks", "Relational Knowledge Distillation"], "answer_arxiv_id": ["1805.04770", "1904.05068"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_3424"} +{"question": "Which works studied the existence of EF1 allocation for monotone combinatorial functions?", "answer": ["On Approximate Envy-Freeness for Indivisible Chores and Mixed Resources"], "answer_arxiv_id": ["2012.06788"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_3425"} +{"question": "What studies have been made on capturing paired low-resolution and high-resolution images from real-world environments?", "answer": ["Toward Real-World Single Image Super-Resolution: A New Benchmark and A\n New Model", "Component Divide-and-Conquer for Real-World Image Super-Resolution"], "answer_arxiv_id": ["1904.00523", "2008.01928"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3426"} +{"question": "Could you provide some works that propose the use of neural models to assist subsequent algorithms in the hybrid category of Neural Combinatorial Optimization?", "answer": ["NeuroLKH: Combining Deep Learning Model with Lin-Kernighan-Helsgaun Heuristic for Solving the Traveling Salesman Problem", "Deep Policy Dynamic Programming for Vehicle Routing Problems", "Generalize a Small Pre-trained Model to Arbitrarily Large TSP Instances", "Graph Neural Network Guided Local Search for the Traveling Salesperson Problem"], "answer_arxiv_id": ["2110.07983", "2102.11756", "2012.10658", "2110.05291"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_3427"} +{"question": "Which works incorporate the Squeeze-and-Excitation (SE) block into their network architectures?", "answer": ["Searching for MobileNetV3", "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks"], "answer_arxiv_id": ["1905.02244", "1905.11946"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_3428"} +{"question": "Is there any research work focused on semantic scene completion for autonomous vehicles?", "answer": ["SemanticKITTI: A Dataset for Semantic Scene Understanding of LiDAR\n Sequences", "SemanticPOSS: A Point Cloud Dataset with Large Quantity of Dynamic\n Instances", "S3CNet: A Sparse Semantic Scene Completion Network for LiDAR Point\n Clouds", "Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning\n Contextual Shape Priors from Scene Completion", "SCPNet: Semantic Scene Completion on Point Cloud"], "answer_arxiv_id": ["1904.01416", "2002.09147", "2012.09242", "2012.03762", "2303.06884"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_3429"} +{"question": "Which works provided an extensive overview and evaluation of different cause-effect methods in the ANM context?", "answer": ["Distinguishing cause from effect using observational data: methods and benchmarks"], "answer_arxiv_id": ["1412.3773"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_3430"} +{"question": "Which works intensively investigated the spectrum of the Hessian matrix in deep learning?", "answer": ["Eigenvalues of the Hessian in Deep Learning: Singularity and Beyond", "Empirical Analysis of the Hessian of Over-Parametrized Neural Networks", "An Investigation into Neural Net Optimization via Hessian Eigenvalue Density", "A Deeper Look at the Hessian Eigenspectrum of Deep Neural Networks and its Applications to Regularization"], "answer_arxiv_id": ["1611.07476", "1706.04454", "1901.10159", "2012.03801"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_3431"} +{"question": "Which paper discusses the architecture-based approach in Continual Learning methods?", "answer": ["PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning"], "answer_arxiv_id": ["1711.05769"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_3432"} +{"question": "What works argue that for optimal compute, model size should be scaled faster than data size?", "answer": ["Scaling Laws for Neural Language Models"], "answer_arxiv_id": ["2001.08361"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_3433"} +{"question": "Could you provide me some works that linked diffusion models to score matching and energy-based models?", "answer": ["Implicit Generation and Modeling with Energy-Based Models", "Learning Non-Convergent Non-Persistent Short-Run MCMC Toward Energy-Based Model", "Learning the Stein Discrepancy for Training and Evaluating Energy-Based Models without Sampling"], "answer_arxiv_id": ["1903.08689", "1904.09770", "2002.05616"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_3434"} +{"question": "Which paper introduced a model with spiking response model (SRM) neurons and proposed a sample and measure criteria?", "answer": ["Fitting summary statistics of neural data with a differentiable spiking network simulator"], "answer_arxiv_id": ["2106.10064"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_3435"} +{"question": "What research papers discuss the task of temporal video grounding?", "answer": ["Negative Sample Matters: A Renaissance of Metric Learning for Temporal\n Grounding", "Towards Generalisable Video Moment Retrieval: Visual-Dynamic Injection to Image-Text Pre-Training"], "answer_arxiv_id": ["2109.04872", "2303.00040v2"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_3436"} +{"question": "Which works discuss unifying image and text representations within a single multimodal model?", "answer": ["VisualBERT: A Simple and Performant Baseline for Vision and Language", "An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "One Model, Multiple Modalities: A Sparsely Activated Approach for Text,\n Sound, Image, Video and Code", "CM3: A Causal Masked Multimodal Model of the Internet"], "answer_arxiv_id": ["1908.03557", "2010.11929", "2108.10904", "2205.06126", "2201.07520"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_3437"} +{"question": "Which papers present work on semantic segmentation in 3D scene perception?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1711.10275", "1904.08755"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_3438"} +{"question": "Which papers discuss the concept of adversarial examples in the context of machine learning?", "answer": ["Evasion attacks against machine learning at test time", "Intriguing properties of neural networks", "Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1708.06131", "1312.6199", "1412.6572"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_3439"} +{"question": "Which paper presented BART, a denoising autoencoder with an encoder-decoder architecture?", "answer": ["BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension"], "answer_arxiv_id": ["1910.13461"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_3440"} +{"question": "Are there any researches that add a regularization term to prevent the network from overfitting to noisy labels?", "answer": ["Early-Learning Regularization Prevents Memorization of Noisy Labels", "Open-set Label Noise Can Improve Robustness Against Inherent Label Noise"], "answer_arxiv_id": ["2007.00151", "2106.10891"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_3441"} +{"question": "Which researches perform amodal completion for specific object categories such as vehicles?", "answer": ["Visualizing the Invisible: Occluded Vehicle Segmentation and Recovery", "Visiting the Invisible: Layer-by-Layer Completed Scene Decomposition"], "answer_arxiv_id": ["1907.09381", "2104.05367"], "source_meta": {"published_time": "20231224"}, "qid": "AutoScholarQuery_train_3442"} +{"question": "Could you provide me with studies that propose the use of automated attacks to generate jailbreaks?", "answer": ["Universal and Transferable Adversarial Attacks on Aligned Language\n Models", "How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to\n Challenge AI Safety by Humanizing LLMs"], "answer_arxiv_id": ["2307.15043", "2401.06373"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_3443"} +{"question": "What research articles focus on self-distillation, particularly to enhance the multilingual capabilities of language models?", "answer": ["Be Your Own Teacher: Improve the Performance of Convolutional Neural\n Networks via Self Distillation", "Revisiting Self-Distillation"], "answer_arxiv_id": ["1905.08094", "2206.08491v1"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_3444"} +{"question": "Which research works utilized a local quadratic approximation of the loss function, similar to the case in this research?", "answer": ["WoodFisher: Efficient Second-Order Approximation for Neural Network Compression"], "answer_arxiv_id": ["2004.14340"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_3445"} +{"question": "Any prior research papers about semantic image editing methods based on applying noise to the original image?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "DiffEdit: Diffusion-based semantic image editing with mask guidance"], "answer_arxiv_id": ["2108.01073", "2210.11427v1"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_3446"} +{"question": "Which paper proposed a training framework by sharding data and creating multiple models for exact unlearning of certain data partitions?", "answer": ["Machine Unlearning"], "answer_arxiv_id": ["1912.03817v3"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_3447"} +{"question": "Which works are about backdoor detection that leverages reverse engineering techniques?", "answer": ["TABOR: A Highly Accurate Approach to Inspecting and Restoring Trojan Backdoors in AI Systems", "Trigger Hunting with a Topological Prior for Trojan Detection"], "answer_arxiv_id": ["1908.01763", "2110.08335"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_3448"} +{"question": "What studies can be referenced for the implementation of Masked Image Modeling (MIM)?", "answer": ["SimMIM: A Simple Framework for Masked Image Modeling", "Swin Transformer V2: Scaling Up Capacity and Resolution"], "answer_arxiv_id": ["2111.09886", "2111.09883"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_3449"} +{"question": "Which studies revisited the training with stronger teachers and its effect on student’s performance?", "answer": ["Knowledge Distillation from A Stronger Teacher", "On the Efficacy of Knowledge Distillation"], "answer_arxiv_id": ["2205.10536", "1910.01348"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3450"} +{"question": "What studies emphasized the pivotal role of data augmentation in self-supervised learning frameworks?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "Exploring Simple Siamese Representation Learning", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss"], "answer_arxiv_id": ["2104.14294", "2006.07733", "2105.04906", "2103.03230", "2011.10566", "2106.04156"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_3451"} +{"question": "Could you provide me study that employs video sequences to improve monocular 3D object detection?", "answer": ["Kinematic 3D Object Detection in Monocular Video"], "answer_arxiv_id": ["2007.09548"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3452"} +{"question": "Could you provide some references that use confounded POMDP for problem formulation and extend the framework of proximal causal inference to sequential decision making?", "answer": ["Off-Policy Evaluation in Partially Observable Environments", "Proximal Reinforcement Learning: Efficient Off-Policy Evaluation in Partially Observed Markov Decision Processes", "A Spectral Approach to Off-Policy Evaluation for POMDPs", "Off-Policy Evaluation for Episodic Partially Observable Markov Decision Processes under Non-Parametric Models", "A Minimax Learning Approach to Off-Policy Evaluation in Confounded Partially Observable Markov Decision Processes"], "answer_arxiv_id": ["1909.03739", "2110.15332", "2109.10502", "2209.10064", "2111.06784"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_3453"} +{"question": "What research works have considered switching costs in the adversarial bandit learning?", "answer": ["Bandits with Feedback Graphs and Switching Costs"], "answer_arxiv_id": ["1907.12189"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_3454"} +{"question": "What researches have been done on image & video editing using feature manipulation in diffusion models?", "answer": ["Diffusion Models in Vision: A Survey", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models", "Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image\n Diffusion Models", "Improving Sample Quality of Diffusion Models Using Self-Attention\n Guidance", "Directed Diffusion: Direct Control of Object Placement through Attention\n Guidance", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Localizing Object-level Shape Variations with Text-to-Image Diffusion\n Models", "MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing"], "answer_arxiv_id": ["2209.04747", "2105.05233", "2006.11239", "2102.09672", "2301.13826", "2210.00939", "2302.13153", "2211.12572", "2208.01626", "2303.11306", "2304.08465v1"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3455"} +{"question": "Which works address double descent in the context of linear regression by changing the ratio of 'n' to input feature dimension 'd'?", "answer": ["High-dimensional dynamics of generalization error in neural networks", "Benign Overfitting in Linear Regression", "Exact expressions for double descent and implicit regularization via surrogate random design", "Surprises in High-Dimensional Ridgeless Least Squares Interpolation"], "answer_arxiv_id": ["1710.03667", "1906.11300", "1912.04533", "1903.08560"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_3456"} +{"question": "What studies have leveraged an Inception network as a feature extractor to compute a notion of distance or similarity between the generated and the real distribution?", "answer": ["Rethinking the Inception Architecture for Computer Vision"], "answer_arxiv_id": ["1512.00567"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_3457"} +{"question": "Could you show me some works about graph neural networks (GNNs) that have been proposed to achieve higher performance in low homophily settings?", "answer": ["Geom-GCN: Geometric Graph Convolutional Networks", "Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs", "Adaptive Universal Generalized PageRank Graph Neural Network", "Learning How to Propagate Messages in Graph Neural Networks", "Fairness-aware Message Passing for Graph Neural Networks", "Graph Neural Networks for Graphs with Heterophily: A Survey", "Revisiting Heterophily For Graph Neural Networks", "Counterfactual Learning on Graphs: A Survey"], "answer_arxiv_id": ["2002.05287", "2006.11468v2", "2006.07988", "2310.00697", "2306.11132", "2202.07082", "2210.07606", "2304.01391"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_3458"} +{"question": "What researches used multilayer perceptron or neural basis functions with a regression loss to approximate the equality constraints?", "answer": ["Implicit Neural Representations with Periodic Activation Functions", "Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains", "Factor Fields: A Unified Framework for Neural Fields and Beyond", "NeX: Real-time View Synthesis with Neural Basis Expansion"], "answer_arxiv_id": ["2006.09661", "2006.10739", "2302.01226", "2103.05606"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_3459"} +{"question": "What works explain the generalization gap between the training and test errors through the classical Rademacher complexity?", "answer": ["Generalization and Representational Limits of Graph Neural Networks"], "answer_arxiv_id": ["2002.06157"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3460"} +{"question": "Any works about approaches other than fine-tuning and learned embeddings in text-to-image generation?", "answer": ["Subject-driven Text-to-Image Generation via Apprenticeship Learning", "Kosmos-G: Generating Images in Context with Multimodal Large Language\n Models", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2304.00186", "2310.02992", "2307.06949"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3461"} +{"question": "What paper discusses the disadvantages of depending on disagreement coefficient in regret bound and number of queries?", "answer": ["Efficient Active Learning with Abstention"], "answer_arxiv_id": ["2204.00043"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_3462"} +{"question": "Can you name studies where U-Nets are used in the field of low-latency segmentation?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation", "UNet++: A Nested U-Net Architecture for Medical Image Segmentation"], "answer_arxiv_id": ["1505.04597", "1807.10165"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_3463"} +{"question": "Which publications provide guarantees for downstream linear classification tasks in the context of contrastive learning?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning"], "answer_arxiv_id": ["1902.09229v1"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_3464"} +{"question": "Which works introduced the unpaired image-to-image translation setting?", "answer": ["Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks", "Unsupervised Image-to-Image Translation Networks"], "answer_arxiv_id": ["1703.10593", "1703.00848"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_3465"} +{"question": "Which researches suggested assessing individual generated images for the evaluation of generative models?", "answer": ["Improved Precision and Recall Metric for Assessing Generative Models", "Rarity Score : A New Metric to Evaluate the Uncommonness of Synthesized\n Images"], "answer_arxiv_id": ["1904.06991", "2206.08549"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_3466"} +{"question": "What works have combined meta-learning and inverse reinforcement learning?", "answer": ["Learning a Prior over Intent via Meta-Inverse Reinforcement Learning", "Meta-Inverse Reinforcement Learning with Probabilistic Context Variables", "Meta-Adversarial Inverse Reinforcement Learning for Decision-making Tasks", "Multi-task Maximum Entropy Inverse Reinforcement Learning"], "answer_arxiv_id": ["1805.12573", "1909.09314", "2103.12694", "1805.08882v2"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_3467"} +{"question": "Which papers provided guarantees on the probability of satisfaction in finite state MDPs?", "answer": ["Omega-Regular Objectives in Model-Free Reinforcement Learning"], "answer_arxiv_id": ["1810.00950"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_3468"} +{"question": "Which papers employed memory networks in video object tracking and segmentation?", "answer": ["Video Object Segmentation using Space-Time Memory Networks", "Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation"], "answer_arxiv_id": ["1904.00607", "2106.05210"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_3469"} +{"question": "What research introduced additional neurons into the Feed Forward Neural Network (FFN) layer as part of parameter-preserving ME methods?", "answer": ["Transformer-Patcher: One Mistake worth One Neuron", "Calibrating Factual Knowledge in Pretrained Language Models"], "answer_arxiv_id": ["2301.09785", "2210.03329"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_3470"} +{"question": "Could you provide me studies which have attempted to better understand self-training and thresholding theoretically?", "answer": ["Self-training Avoids Using Spurious Features Under Domain Shift", "Understanding Self-Training for Gradual Domain Adaptation"], "answer_arxiv_id": ["2006.10032", "2002.11361v1"], "source_meta": {"published_time": "20220515"}, "qid": "AutoScholarQuery_train_3471"} +{"question": "Can you identify the works that have shown achievements in the RobustBench with larger models like Swin-L and ConvNeXt-L?", "answer": ["RobustBench: a standardized adversarial robustness benchmark", "A Comprehensive Study on Robustness of Image Classification Models:\n Benchmarking and Rethinking", "Revisiting Adversarial Training for ImageNet: Architectures, Training\n and Generalization across Threat Models"], "answer_arxiv_id": ["2010.09670", "2302.14301", "2303.01870"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_3472"} +{"question": "Which paper provides the most common contrastive learning method in computer vision called SimCLR?", "answer": ["Indiscriminate Poisoning Attacks on Unsupervised Contrastive Learning"], "answer_arxiv_id": ["2202.11202"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_3473"} +{"question": "What works provide a more explicit disentanglement by leveraging a canonical volume and deformation in Disentangled3D?", "answer": ["Disentangled3D: Learning a 3D Generative Model with Disentangled\n Geometry and Appearance from Monocular Images"], "answer_arxiv_id": ["2203.15926"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_3474"} +{"question": "What paper recently explored the concept of sensitivity of any trained NeRF to perturbations?", "answer": ["Active Neural Mapping"], "answer_arxiv_id": ["2308.16246"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_3475"} +{"question": "Which studies have successfully imported lower and upper bounds techniques to obtain analogous information-theoretic bounds in the context of classification?", "answer": ["An Equivalence Between Private Classification and Online Prediction", "A Limitation of the PAC-Bayes Framework", "Finite Littlestone Dimension Implies Finite Information Complexity"], "answer_arxiv_id": ["2003.00563", "2006.13508", "2206.13257v1"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_3476"} +{"question": "What papers define a contrastive prediction task using strong data augmentations?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20211202"}, "qid": "AutoScholarQuery_train_3477"} +{"question": "What research works involve large language models with chain-of-thought prompting technique for reasoning ability?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models", "Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2201.11903", "2203.11171", "2205.10625", "2210.03493"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_3478"} +{"question": "Any works that illustrate the inadequacy of the message passing paradigm for detecting structural motifs in graphs?", "answer": ["Theoretically Improving Graph Neural Networks via Anonymous Walk Graph Kernels", "Convolutional Kernel Networks for Graph-Structured Data", "Graph Kernel Neural Networks"], "answer_arxiv_id": ["2104.02995", "2003.05189", "2112.07436"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_3479"} +{"question": "Which studies proposed learning strategies in repeated second-price auctions with budget and return-on-spend (RoS) constraints?", "answer": ["Online Bidding Algorithms for Return-on-Spend Constrained Advertisers"], "answer_arxiv_id": ["2208.13713v3"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_3480"} +{"question": "What is the reference for the MelGAN study which introduced a multi-scale waveform discriminator and feature matching loss?", "answer": ["MelGAN: Generative Adversarial Networks for Conditional Waveform Synthesis"], "answer_arxiv_id": ["1910.06711"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_3481"} +{"question": "What studies have focused on the compositional generalization of standard neural models?", "answer": ["Assessing Phrasal Representation and Composition in Transformers", "COGS: A Compositional Generalization Challenge Based on Semantic Interpretation", "Uncontrolled Lexical Exposure Leads to Overestimation of Compositional Generalization in Pretrained Models", "The Paradox of the Compositionality of Natural Language: A Neural Machine Translation Case Study"], "answer_arxiv_id": ["2010.03763", "2010.05465", "2212.10769", "2108.05885"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_3482"} +{"question": "Could you provide me some works about interactive image editing using reference images?", "answer": ["Paint by Example: Exemplar-based Image Editing with Diffusion Models", "AnyDoor: Zero-shot Object-level Image Customization"], "answer_arxiv_id": ["2211.13227", "2307.09481"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_3483"} +{"question": "Could you provide me some studies about CNNs that can be trained without the need for ground truth data?", "answer": ["Learning Residual Flow as Dynamic Motion from Stereo Videos"], "answer_arxiv_id": ["1909.06999"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_3484"} +{"question": "Can you cite studies that used proxy-based methods to understand the input/output correlation in black-box explanations?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "A Unified Approach to Interpreting Model Predictions", "Interpretable Explanations of Black Boxes by Meaningful Perturbation", "Visualizing Deep Neural Network Decisions: Prediction Difference Analysis"], "answer_arxiv_id": ["1602.04938", "1705.07874", "1704.03296", "1702.04595"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_3485"} +{"question": "Can you name the research that proposes to encode an image with an autoencoder and then use a diffusion model to generate continuous feature maps in the latent space?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_3486"} +{"question": "Could you provide me a study which focused on preserving the ego information of nodes in graph learning?", "answer": ["Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "Simple and Deep Graph Convolutional Networks"], "answer_arxiv_id": ["1810.05997", "2007.02133"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_3487"} +{"question": "Which works employ long-term modeling to improve the performance of 3D object detection models?", "answer": ["Exploring Recurrent Long-term Temporal Fusion for Multi-view 3D Perception", "Time Will Tell: New Outlooks and A Baseline for Temporal Multi-View 3D Object Detection"], "answer_arxiv_id": ["2303.05970", "2210.02443"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_3488"} +{"question": "Could you provide me some studies about stochastic subtokenization?", "answer": ["BPE-Dropout: Simple and Effective Subword Regularization", "Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidates", "Multi-view Subword Regularization"], "answer_arxiv_id": ["1910.13267", "1804.10959", "2103.08490v2"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_3489"} +{"question": "What works use domain weighting strategies in multi-source domain adaptation?", "answer": ["Aligning Domain-specific Distribution and Classifier for Cross-domain Classification from Multiple Sources", "Moment Matching for Multi-Source Domain Adaptation", "Multi-source Distilling Domain Adaptation", "Learning to Combine: Knowledge Aggregation for Multi-Source Domain Adaptation"], "answer_arxiv_id": ["2201.01003", "1812.01754", "1911.11554", "2007.08801"], "source_meta": {"published_time": "20220201"}, "qid": "AutoScholarQuery_train_3490"} +{"question": "Which works investigate the way different fairness policies affect the gap between the qualification rates of different groups?", "answer": ["How Do Fair Decisions Fare in Long-term Qualification?"], "answer_arxiv_id": ["2010.11300"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_3491"} +{"question": "Which papers discuss the use of machine learning combined with image-based rendering in the context of generalizable novel view synthesis?", "answer": ["Free View Synthesis", "Neural Point-Based Graphics", "Point-Based Neural Rendering with Per-View Optimization", "SynSin: End-to-end View Synthesis from a Single Image"], "answer_arxiv_id": ["2008.05511", "1906.08240", "2109.02369", "1912.08804"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_3492"} +{"question": "What are the works where high accuracy models depend on resource-intensive backbones?", "answer": ["Explicit Box Detection Unifies End-to-End Multi-Person Pose Estimation"], "answer_arxiv_id": ["2302.01593"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_3493"} +{"question": "Are there any studies that allowed for the adaption of test data in TTA?", "answer": ["Back to the Source: Diffusion-Driven Test-Time Adaptation", "Energy-Based Test Sample Adaptation for Domain Generalization"], "answer_arxiv_id": ["2207.03442", "2302.11215"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_3494"} +{"question": "Which works proposed offline reinforcement learning algorithms using penalize value function as a method?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble", "Anti-Exploration by Random Network Distillation"], "answer_arxiv_id": ["2006.04779", "2110.01548", "2301.13616"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_3495"} +{"question": "What studies have applied the Segment Anything Model (SAM) to object tracking tasks?", "answer": ["Track Anything: Segment Anything Meets Videos", "Segment and Track Anything"], "answer_arxiv_id": ["2304.11968", "2305.06558"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_3496"} +{"question": "Which works are based on directly training 3D diffusion models using point clouds as 3D representations?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "LION: Latent Point Diffusion Models for 3D Shape Generation", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "3D Shape Generation and Completion through Point-Voxel Diffusion"], "answer_arxiv_id": ["2212.08751", "2210.06978", "2103.01458", "2104.03670"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_3497"} +{"question": "What works have been done on characterizing the capacity of neural networks by assessing their abilities to learn simpler data models?", "answer": ["On the Practical Computational Power of Finite Precision RNNs for Language Recognition", "LSTM Networks Can Perform Dynamic Counting", "Sequential Neural Networks as Automata", "The Limitations of Limited Context for Constituency Parsing", "Self-Attention Networks Can Process Bounded Hierarchical Languages"], "answer_arxiv_id": ["1805.04908", "1906.03648", "1906.01615", "2106.01580", "2105.11115"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_3498"} +{"question": "Which work leverages Low-Rank Adaptation to fine-tune the foundation model for Dialogue State Tracking?", "answer": ["Towards LLM-driven Dialogue State Tracking"], "answer_arxiv_id": ["2310.14970"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_3499"} +{"question": "Could you give me the works about fine-tuning CLIP on egocentric datasets?", "answer": ["Learning Video Representations from Large Language Models", "Egocentric Video-Language Pretraining", "EgoVLPv2: Egocentric Video-Language Pre-training with Fusion in the Backbone"], "answer_arxiv_id": ["2212.04501v1", "2206.01670", "2307.05463v2"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_3500"} +{"question": "Could you provide me some works about neural controlled differential equations?", "answer": ["Neural Controlled Differential Equations for Irregular Time Series", "Neural Controlled Differential Equations for Online Prediction Tasks"], "answer_arxiv_id": ["2005.08926", "2106.11028"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_3501"} +{"question": "Which works use Wikipedia or web pages as an external corpus for open-domain question answering?", "answer": ["Reading Wikipedia to Answer Open-Domain Questions", "Leveraging Passage Retrieval with Generative Models for Open Domain Question Answering", "Internet-augmented language models through few-shot prompting for open-domain question answering"], "answer_arxiv_id": ["1704.00051", "2007.01282", "2203.05115"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_3502"} +{"question": "What works proposed loss correction based on an estimated noise transition matrix or the model’s predictions?", "answer": ["Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe Noise", "Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach", "Unsupervised Label Noise Modeling and Loss Correction", "Beyond Class-Conditional Assumption: A Primary Attempt to Combat Instance-Dependent Label Noise", "Training deep neural networks on noisy labels with bootstrapping", "Joint Optimization Framework for Learning with Noisy Labels", "Error-Bounded Correction of Noisy Labels"], "answer_arxiv_id": ["1802.05300", "1609.03683", "1904.11238", "2012.05458", "1412.6596", "1803.11364", "2011.10077v1"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_3503"} +{"question": "What works used Latent Diffusion Models (LDMs) to enhance training efficiency in T2I generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_3504"} +{"question": "What are the studies that change the input image by cropping or masking it to focus on the foreground object?", "answer": ["ReCLIP: A Strong Zero-Shot Baseline for Referring Expression\n Comprehension", "OvarNet: Towards Open-vocabulary Object Attribute Recognition", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP"], "answer_arxiv_id": ["2204.05991", "2301.09506", "2210.04150"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_3505"} +{"question": "Which studies applied a hierarchical variational autoencoder model for abstractive summarization of reviews?", "answer": ["Unsupervised Opinion Summarization as Copycat-Review Generation"], "answer_arxiv_id": ["1911.02247"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_3506"} +{"question": "Which studies constitute the linkage between object shapes and affordances by a learning-based mapping in 3D space?", "answer": ["3D AffordanceNet: A Benchmark for Visual Object Affordance Understanding", "PartAfford: Part-level Affordance Discovery from 3D Objects", "AdaAfford: Learning to Adapt Manipulation Affordance for 3D Articulated\n Objects via Few-shot Interactions", "Where2Act: From Pixels to Actions for Articulated 3D Objects", "Learning Foresightful Dense Visual Affordance for Deformable Object\n Manipulation", "Background Activation Suppression for Weakly Supervised Object Localization and Semantic Segmentation"], "answer_arxiv_id": ["2103.16397", "2202.13519", "2112.00246", "2101.02692", "2303.11057", "2309.12943v1"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_3507"} +{"question": "Which paper provides an overview of causal feature selection approaches and their evaluation?", "answer": ["Causality-based Feature Selection: Methods and Evaluations"], "answer_arxiv_id": ["1911.07147"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_3508"} +{"question": "What works have been made in the machine learning literature that also utilized the concept of reductions for multi-agent problems?", "answer": ["Bayesian Action Decoder for Deep Multi-Agent Reinforcement Learning", "Improving Policies via Search in Cooperative Partially Observable Games", "Solving Common-Payoff Games with Approximate Policy Iteration", "Scalable Online Planning via Reinforcement Learning Fine-Tuning", "Common Information based Approximate State Representations in Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1811.01458", "1912.02318", "2101.04237", "2109.15316", "2110.12603"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_3509"} +{"question": "Which works adopted the mechanism of 'Chain-of-thought' to enhance the capabilities of LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_3510"} +{"question": "Which study proposes a compositional scene representation to assist in geometry optimization in complex scenes?", "answer": ["Object-Compositional Neural Implicit Surfaces"], "answer_arxiv_id": ["2207.09686"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_3511"} +{"question": "What researches explore generating virtual outliers for model regularization in OOD detection?", "answer": ["VOS: Learning What You Don’t Know by Virtual Outlier Synthesis", "Non-parametric Outlier Synthesis"], "answer_arxiv_id": ["2202.01197", "2303.02966"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_3512"} +{"question": "Could you provide me with some studies that discuss interpretability and feature selection in GNNs?", "answer": ["GNNExplainer: Generating Explanations for Graph Neural Networks", "Discovering and Explaining the Representation Bottleneck of Graph Neural Networks from Multi-order Interactions"], "answer_arxiv_id": ["1903.03894", "2205.07266"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_3513"} +{"question": "What works have discussed CHOMP and its variants in relation to motion optimization?", "answer": ["Functional Gradient Motion Planning in Reproducing Kernel Hilbert Spaces"], "answer_arxiv_id": ["1601.03648"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_3514"} +{"question": "Could you provide me some works about explicit and implicit GR in deep learning?", "answer": ["Implicit Gradient Regularization", "On the Origin of Implicit Regularization in Stochastic Gradient Descent"], "answer_arxiv_id": ["2009.11162", "2101.12176"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_3515"} +{"question": "Which works are recognized as groundbreaking explorations in foundation model development?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3516"} +{"question": "What research introduced PointContrast for conducting point-level contrast on two transformed views of the same point cloud?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud Understanding"], "answer_arxiv_id": ["2007.10985"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_3517"} +{"question": "Could you give me examples of researches that revolve around evaluating video prediction models on standardized datasets?", "answer": ["Unsupervised Learning of Video Representations using LSTMs", "The Cityscapes Dataset for Semantic Urban Scene Understanding", "Unsupervised Learning for Physical Interaction through Video Prediction", "RoboNet: Large-Scale Multi-Robot Learning"], "answer_arxiv_id": ["1502.04681", "1604.01685v2", "1605.07157", "1910.11215"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_3518"} +{"question": "Could you provide me some studies about Guided Set Diffusion Model?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2006.11239", "2011.13456"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_3519"} +{"question": "Could you provide me some works on supervised metric like DCI Disentanglement and MIG?", "answer": ["Isolating Sources of Disentanglement in Variational Autoencoders"], "answer_arxiv_id": ["1802.04942"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_3520"} +{"question": "What works have cited gossip learning as a DL algorithm where each node functions independently?", "answer": ["Gossip Learning with Linear Models on Fully Distributed Data"], "answer_arxiv_id": ["1109.1396v3"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_3521"} +{"question": "Could you give me examples of follow-up works of GraIL?", "answer": ["Communicative Message Passing for Inductive Relation Reasoning", "Subgraph Neighboring Relations Infomax for Inductive Link Prediction on Knowledge Graphs"], "answer_arxiv_id": ["2012.08911", "2208.00850"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_3522"} +{"question": "Could you list some research works which trained Neural Cellular Automata generators to produce images, textures, or 3D structures?", "answer": ["Growing 3D Artefacts and Functional Machines with Neural Cellular Automata"], "answer_arxiv_id": ["2103.08737"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_3523"} +{"question": "Which researchers utilized feature alignment across domains for Domain Generalization?", "answer": ["Domain-Adversarial Training of Neural Networks", "Domain Generalization via Conditional Invariant Representation", "Deep CORAL: Correlation Alignment for Deep Domain Adaptation"], "answer_arxiv_id": ["1505.07818", "1807.08479", "1607.01719"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_3524"} +{"question": "Which works are based on VAE or GAN for semi-supervised conditional image generation?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_3525"} +{"question": "Which datasets are built from reliable sources like case reports, research literature, and medical exams?", "answer": ["CliCR: A Dataset of Clinical Case Reports for Machine Reading Comprehension", "Measuring Massive Multitask Language Understanding"], "answer_arxiv_id": ["1803.09720v1", "2009.03300"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_3526"} +{"question": "Are there papers that applied contrastive objectives to image-text pairs during visual-language pre-training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2107.07651"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_3527"} +{"question": "What works are about the investigation of the timestep scheduling used in the score distillation process?", "answer": ["DreamTime: An Improved Optimization Strategy for Diffusion-Guided 3D\n Generation", "A Variational Perspective on Solving Inverse Problems with Diffusion\n Models"], "answer_arxiv_id": ["2306.12422", "2305.04391"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_3528"} +{"question": "What works in CS used Stochastic Gradient MCMC (SGM) as a prior?", "answer": ["Robust Compressed Sensing MRI with Deep Generative Priors", "Instance-Optimal Compressed Sensing via Posterior Sampling", "SNIPS: Solving Noisy Inverse Problems Stochastically", "Denoising Diffusion Restoration Models", "Improving Diffusion Models for Inverse Problems using Manifold Constraints"], "answer_arxiv_id": ["2108.01368", "2106.11438", "2105.14951", "2201.11793", "2206.00941"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_3529"} +{"question": "Who proposed classifier guidance for repurposing pretrained unconditional diffusion models for conditional image generation?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_3530"} +{"question": "Can you provide examples of studies connecting LLMs with external tools such as a web browser, HuggingFace model hub, chemical software, PowerPoint, and a tool library?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face", "ChemCrow: Augmenting large-language models with chemistry tools", "TaskMatrix.AI: Completing Tasks by Connecting Foundation Models with Millions of APIs", "Toolformer: Language Models Can Teach Themselves to Use Tools", "ART: Automatic multi-step reasoning and tool-use for large language models"], "answer_arxiv_id": ["2112.09332", "2303.17580", "2304.05376v5", "2303.16434", "2302.04761", "2303.09014"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3531"} +{"question": "What papers have explored the dimensional collapse issue in contrastive learning?", "answer": ["Whitening for Self-Supervised Representation Learning", "On Feature Decorrelation in Self-Supervised Learning", "An Investigation into Whitening Loss for Self-supervised Learning"], "answer_arxiv_id": ["2007.06346", "2105.00470", "2210.03586"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_3532"} +{"question": "Which studies were about generating stories using language models?", "answer": ["Automatic Story Generation: Challenges and Attempts", "The Next Chapter: A Study of Large Language Models in Storytelling", "Future Sight: Dynamic Story Generation with Large Pretrained Language\n Models", "MEGATRON-CNTRL: Controllable Story Generation with External Knowledge\n Using Large-Scale Language Models"], "answer_arxiv_id": ["2102.12634", "2301.09790", "2212.09947", "2010.00840"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_3533"} +{"question": "Which studies suggested that models can be merged by interpolating their weights?", "answer": ["Robust fine-tuning of zero-shot models", "Merging Models with Fisher-Weighted Averaging", "Model soups: averaging weights of multiple fine-tuned models improves\n accuracy without increasing inference time"], "answer_arxiv_id": ["2109.01903", "2111.09832", "2203.05482"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_3534"} +{"question": "What research work incorporated Monte-Carlo approach with multiple independent rollouts per iteration?", "answer": ["On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2201.07443"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_3535"} +{"question": "What papers provide more efficient solutions with inexact proximal point methods with a faster convergence rate of 𝒪(ϵ−3)?", "answer": ["Efficient Algorithms for Smooth Minimax Optimization", "An accelerated inexact proximal point method for solving nonconvex-concave min-max problems"], "answer_arxiv_id": ["1907.01543", "1905.13433"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3536"} +{"question": "Which works are focused on improving adversarial training (AT) through various aspects?", "answer": ["Recognizing Object by Components with Human Prior Knowledge Enhances\n Adversarial Robustness of Deep Neural Networks", "Rethinking Softmax Cross-Entropy Loss for Adversarial Robustness", "Boosting Adversarial Training with Hypersphere Embedding", "Adversarial Weight Perturbation Helps Robust Generalization", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["2212.01806", "1905.10626", "2002.08619", "2004.05884", "1901.08573"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_3537"} +{"question": "What is the paper that introduced a 7x7 convolution kernel larger than the previous CNN?", "answer": ["A ConvNet for the 2020s"], "answer_arxiv_id": ["2201.03545"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_3538"} +{"question": "Which works have studied the optimization of shallow ReLU networks using hinge loss?", "answer": ["SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data"], "answer_arxiv_id": ["1710.10174"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_3539"} +{"question": "Which work scaled the batch RL algorithm like REINFORCE with off-policy correction to real-world products?", "answer": ["Top-K Off-Policy Correction for a REINFORCE Recommender System"], "answer_arxiv_id": ["1812.02353"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_3540"} +{"question": "Can you cite a few studies on generator models where the generator model is trained with a discriminator in the context of Generative adversarial network (GAN)?", "answer": ["Generative Adversarial Nets", "Progressive Growing of GANs for Improved Quality, Stability, and Variation", "Spectral Normalization for Generative Adversarial Networks", "Training Generative Adversarial Networks with Limited Data"], "answer_arxiv_id": ["1406.2661", "1710.10196", "1802.05957", "2006.06676"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3541"} +{"question": "Which research papers developed RNN-based methods by introducing exponential decay on observations for sequential learning?", "answer": ["Recurrent Neural Networks for Multivariate Time Series with Missing Values", "The Neural Hawkes Process: A Neurally Self-Modulating Multivariate Point Process"], "answer_arxiv_id": ["1606.01865", "1612.09328"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_3542"} +{"question": "Which work suggests that Head Probing (HP) yields better performance than Fine Tuning (FT) when features are perfect?", "answer": ["To Tune or Not to Tune? Adapting Pretrained Representations to Diverse Tasks"], "answer_arxiv_id": ["1903.05987"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_3543"} +{"question": "Which studies introduced strategies for training models with noisy ground truths?", "answer": ["DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection", "DN-DETR: Accelerate DETR Training by Introducing Query DeNoising"], "answer_arxiv_id": ["2203.03605", "2203.01305"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_3544"} +{"question": "Which papers are about deriving non-asymptotic rates for structured non-convex sampling using Langevin-based schemes?", "answer": ["Non-Convex Learning via Stochastic Gradient Langevin Dynamics: A Nonasymptotic Analysis", "Sharp convergence rates for Langevin dynamics in the nonconvex setting", "Stochastic Runge-Kutta Accelerates Langevin Monte Carlo and Beyond", "Global Convergence of Langevin Dynamics Based Algorithms for Nonconvex Optimization", "Rapid Convergence of the Unadjusted Langevin Algorithm: Isoperimetry Suffices", "Non-asymptotic bounds for sampling algorithms without log-concavity", "Improved Bounds for Discretization of Langevin Diffusions: Near-Optimal Rates without Convexity"], "answer_arxiv_id": ["1702.03849", "1805.01648", "1906.07868", "1707.06618", "1903.08568", "1808.07105", "1907.11331"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_3545"} +{"question": "What references suggest that restricted variants of low-rank approximation are NP-hard?", "answer": ["Regularized Weighted Low Rank Approximation", "Approximation Schemes for Low-Rank Binary Matrix Approximation Problems"], "answer_arxiv_id": ["1911.06958", "1807.07156v1"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_3546"} +{"question": "Which papers aim to improve the rendering quality of NeRF using different types of supervision or by improving the model?", "answer": ["NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw Images", "​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​", "DIVeR: Real-time and Accurate Neural Radiance Fields with Deterministic Integration for Volume Rendering", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2111.13679", "2103.13415", "2111.10427", "2304.06706"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_3547"} +{"question": "What is the work that disentangles attributes and objects with reversed attention in CZSL?", "answer": ["Distilled Reverse Attention Network for Open-world Compositional\n Zero-Shot Learning"], "answer_arxiv_id": ["2303.00404"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_3548"} +{"question": "Which paper is related to solving the problem in the feature space using Fenchel duality in lower bound calculation?", "answer": ["Sparse regression: Scalable algorithms and empirical performance"], "answer_arxiv_id": ["1902.06547"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_3549"} +{"question": "Which papers focus on the Laplace approximations primarily utilized for bnns?", "answer": ["Optimizing Neural Networks with Kronecker-factored Approximate Curvature", "Practical Gauss-Newton Optimisation for Deep Learning", "Adapting the Linearised Laplace Model Evidence for Modern Deep Learning", "Bayesian Deep Learning via Subnetwork Inference", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1503.05671", "1706.03662", "2206.08900", "2010.14689", "1612.01474"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_3550"} +{"question": "What works have integrated diffusion models with PointNet and triplane features?", "answer": ["Diffusion-SDF: Conditional Generative Modeling of Signed Distance\n Functions"], "answer_arxiv_id": ["2211.13757"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_3551"} +{"question": "Are there papers discussing graph feature models that help in link prediction?", "answer": ["DeepPath: A Reinforcement Learning Method for Knowledge Graph Reasoning", "Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning", "Reasoning on Knowledge Graphs with Debate Dynamics", "Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "DRUM: End-To-End Differentiable Rule Mining On Knowledge Graphs"], "answer_arxiv_id": ["1707.06690", "1711.05851", "2001.00461", "1702.08367", "1911.00055"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_3552"} +{"question": "Which papers addressed certified robustness in backdoor attack defense?", "answer": ["PatchGuard: A Provably Robust Defense against Adversarial Patches via\n Small Receptive Fields and Masking", "Minority Reports Defense: Defending Against Adversarial Patches", "Certified Robustness of Nearest Neighbors against Data Poisoning and\n Backdoor Attacks"], "answer_arxiv_id": ["2005.10884", "2004.13799", "2012.03765"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_3553"} +{"question": "Which papers proposed finding a completion of the ground truth matrix with low Frobenius norm error in matrix completion?", "answer": ["Exact Matrix Completion via Convex Optimization", "The Power of Convex Relaxation: Near-Optimal Matrix Completion", "Matrix Completion from a Few Entries", "Recovering Low-Rank Matrices From Few Coefficients In Any Basis", "A Simpler Approach to Matrix Completion", "Restricted strong convexity and weighted matrix completion: Optimal bounds with noise", "Low-rank Matrix Completion using Alternating Minimization", "Noisy Matrix Completion: Understanding Statistical Guarantees for Convex Relaxation via Nonconvex Optimization", "Entrywise Eigenvector Analysis of Random Matrices with Low Expected Rank"], "answer_arxiv_id": ["0805.4471", "0903.1476", "0901.3150", "0910.1879", "0910.0651", "1009.2118v2", "1212.0467", "1902.07698v2", "1709.09565"], "source_meta": {"published_time": "20220825"}, "qid": "AutoScholarQuery_train_3554"} +{"question": "What papers have improved the original AT formulation to speed up the training?", "answer": ["Adversarial Training for Free!", "Fast is better than free: Revisiting adversarial training"], "answer_arxiv_id": ["1904.12843", "2001.03994"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3555"} +{"question": "Which works focused on improving the pioneering policy gradient approach for symbolic regression?", "answer": ["Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients"], "answer_arxiv_id": ["1912.04871"], "source_meta": {"published_time": "20231230"}, "qid": "AutoScholarQuery_train_3556"} +{"question": "Could you provide me studies which propose attacks rely on querying the model with examples nearby or derived from the target?", "answer": ["Revisiting Membership Inference Under Realistic Assumptions", "Understanding Membership Inferences on Well-Generalized Learning Models", "Membership Leakage in Label-Only Exposures", "Label-Only Membership Inference Attacks", "Canary in a Coalmine: Better Membership Inference with Ensembled Adversarial Queries"], "answer_arxiv_id": ["2005.10881", "1802.04889", "2007.15528", "2007.14321", "2210.10750"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_3557"} +{"question": "Which research proposes to recycle soft prompts across models with vocab-to-vocab transformations?", "answer": ["Reducing Retraining by Recycling Parameter-Efficient Prompts"], "answer_arxiv_id": ["2208.05577"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_3558"} +{"question": "Which works utilized supervised learning approaches to approximate convex optimization with DNNs?", "answer": ["Deep learning for efficient frontier calculation in finance", "Learning Convex Optimization Models", "Deep Learning the Efficient Frontier of Convex Vector Optimization Problems", "A stochastic approximation method for approximating the efficient frontier of chance-constrained nonlinear programs"], "answer_arxiv_id": ["2101.02044v4", "2006.04248", "2205.07077", "1812.07066v3"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_3559"} +{"question": "What works have noticed that RLHF can make mistakes in the output more subtle?", "answer": ["Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback"], "answer_arxiv_id": ["2204.05862"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_3560"} +{"question": "Which works made efforts towards enabling image editing capabilities in text-to-image generation?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2208.01626", "2211.09800", "2210.09276"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_3561"} +{"question": "Are there any research papers that proposed alignment based learning rules for networks at large width in the lazy regime?", "answer": ["How to Train Your Wide Neural Network Without Backprop: An Input-Weight Alignment Perspective"], "answer_arxiv_id": ["2106.08453"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_3562"} +{"question": "Which work introduced the use of foundation image-language models in aligning image and text in high-dimensional feature space?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_3563"} +{"question": "Are there any works that proposed refining global pose with different stages of downsampling?", "answer": ["HRegNet: A Hierarchical Network for Large-scale Outdoor LiDAR Point\n Cloud Registration"], "answer_arxiv_id": ["2107.11992"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_3564"} +{"question": "Could you provide some papers that aim to improve the efficiency of conformal prediction?", "answer": ["Classification with Valid and Adaptive Coverage", "A comparison of some conformal quantile regression methods", "\\cdsplit and \\hpdsplit: efficient conformal regions in high dimensions", "Finite-sample Efficient Conformal Prediction", "Learning Optimal Conformal Classifiers"], "answer_arxiv_id": ["2006.02544", "1909.05433", "2007.12778", "2104.13871", "2110.09192"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_3565"} +{"question": "What implementation of the Minecraft Environment was used in this study?", "answer": ["MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledge"], "answer_arxiv_id": ["2206.08853"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_3566"} +{"question": "Which papers have used DiCE for policy evaluation?", "answer": ["Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation", "DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections", "A Deep Reinforcement Learning Approach to Marginalized Importance Sampling with the Successor Representation"], "answer_arxiv_id": ["1810.12429", "1906.04733", "2106.06854"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_3567"} +{"question": "What studies have focused on image-to-music generation in cross-modal music generation?", "answer": ["Vis2Mus: Exploring Multimodal Representation Mapping for Controllable Music Generation", "I hear your true colors: Image Guided Audio Generation"], "answer_arxiv_id": ["2211.05543", "2211.03089"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_3568"} +{"question": "What papers discuss the application of fine-tuning open-source LVLMs, including LLaVA, MiniGPT4, Mplug-Owl, LRV-Instruction, and LLaVAR?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Instruction Tuning with GPT-4", "StableLLaVA: Enhanced Visual Instruction Tuning with Synthesized\n Image-Dialogue Data", "A Survey on Multimodal Large Language Models"], "answer_arxiv_id": ["2304.10592", "2304.14178", "2306.17107v2", "2305.06500", "2304.03277", "2308.10253", "2306.13549"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_3569"} +{"question": "Where can I find information about the application of Label Smoothing in penalizing 100% confident predictions?", "answer": ["When Does Label Smoothing Help?"], "answer_arxiv_id": ["1906.02629"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_3570"} +{"question": "Which studies conclude that existing models lack in-depth image-text fusion mechanisms or are designed for specific tasks rather than multi-modal embedding applications?", "answer": ["UniIR: Training and Benchmarking Universal Multimodal Information\n Retrievers", "Universal Vision-Language Dense Retrieval: Learning A Unified\n Representation Space for Multi-Modal Retrieval", "Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image\n Retrieval", "Composed Image Retrieval using Contrastive Learning and Task-oriented\n CLIP-based Features"], "answer_arxiv_id": ["2311.17136", "2209.00179", "2302.03084", "2308.11485"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_3571"} +{"question": "What work proposes to augment image data by perturbing the amplitude information in the spectral domain?", "answer": ["A Fourier-based Framework for Domain Generalization"], "answer_arxiv_id": ["2105.11120"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_3572"} +{"question": "Which papers focus on reconstructing multiple people jointly from a single image?", "answer": ["Coherent Reconstruction of Multiple Humans from a Single Image", "Monocular, One-stage, Regression of Multiple 3D People"], "answer_arxiv_id": ["2006.08586", "2008.12272"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_3573"} +{"question": "Can you tell me about any existing efforts in open-ended tasks in video question answering?", "answer": ["Just Ask: Learning to Answer Questions from Millions of Narrated Videos", "TGIF-QA: Toward Spatio-Temporal Reasoning in Visual Question Answering", "ActivityNet-QA: A Dataset for Understanding Complex Web Videos via\n Question Answering"], "answer_arxiv_id": ["2012.00451", "1704.04497", "1906.02467"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_3574"} +{"question": "Which works focus on the instability in the performance of GPT-3's few-shot learning tied to the selection of in-context examples?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models", "What Makes Good In-Context Examples for GPT-3?", "A Survey of Deep Learning for Mathematical Reasoning"], "answer_arxiv_id": ["2102.09690", "2101.06804", "2212.10535"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_3575"} +{"question": "What studies have dealt with Macro-level spaces in search spaces?", "answer": ["MobileNets: Efficient Convolutional Neural Networks for Mobile Vision\n Applications", "MobileNetV2: Inverted Residuals and Linear Bottlenecks", "Searching for MobileNetV3", "FBNetV2: Differentiable Neural Architecture Search for Spatial and\n Channel Dimensions", "FBNetV3: Joint Architecture-Recipe Search using Predictor Pretraining", "High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis"], "answer_arxiv_id": ["1704.04861", "1801.04381", "1905.02244", "2004.05565", "2006.02049", "2112.10752", "2307.01952"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_3576"} +{"question": "Can you mention some works that show state-of-the-art performance in the field of image synthesis due to the progress in DMs?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2010.02502"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_3577"} +{"question": "What work illustrated the permutation properties in both forward and backward propagations within the existing Transformer network structure?", "answer": ["Set Transformer: A Framework for Attention-based Permutation-Invariant\n Neural Networks"], "answer_arxiv_id": ["1810.00825"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_3578"} +{"question": "Who introduced Datamodel, a method of re-training-based influence estimation?", "answer": ["Datamodels: Predicting Predictions from Training Data"], "answer_arxiv_id": ["2202.00622"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_3579"} +{"question": "Could you provide some examples of researches that focus on segmentation in 4D perception using RGB-D video?", "answer": ["Efficient Hierarchical Graph-Based Segmentation of RGBD Videos"], "answer_arxiv_id": ["1801.08981"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_3580"} +{"question": "What works have attempted to achieve generalization through feature decorrelation?", "answer": ["Stable Prediction across Unknown Environments", "Stable Learning via Sample Reweighting", "Stable Prediction with Model Misspecification and Agnostic Distribution Shift", "Deep Stable Learning for Out-Of-Distribution Generalization"], "answer_arxiv_id": ["1806.06270v2", "1911.12580", "2001.11713v1", "2104.07876"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_3581"} +{"question": "Which papers concentrate on pseudo-labeling techniques for addressing label scarcity in the target domain?", "answer": ["Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samples"], "answer_arxiv_id": ["2104.13963"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_3582"} +{"question": "Are there any works that generalize the setting of having just a pre-treatment phase and a post-treatment phase to multiple times slots in panel data models?", "answer": ["Matrix Completion Methods for Causal Panel Data Models"], "answer_arxiv_id": ["1710.10251v5"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_3583"} +{"question": "Which studies made a weaker boundedness assumption when LL objective is strongly convex?", "answer": ["Will Bilevel Optimizers Benefit from Loops", "A framework for bilevel optimization that enables stochastic and global variance reduction algorithms"], "answer_arxiv_id": ["2205.14224", "2201.13409"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_3584"} +{"question": "Which works investigate the computational complexity in learning k-parities over d bits?", "answer": ["Failures of Gradient-Based Deep Learning"], "answer_arxiv_id": ["1703.07950v2"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_3585"} +{"question": "Which works talk about the increasing popularity of event-based cameras?", "answer": ["Event-based Vision: A Survey"], "answer_arxiv_id": ["1904.08405"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_3586"} +{"question": "Which work learns visual-language representation from a noisy image-text dataset?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2102.05918"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_3587"} +{"question": "Any papers discussing models trained on various types of audio, and demonstrated universal audio understanding abilities?", "answer": ["SALMONN: Towards Generic Hearing Abilities for Large Language Models", "Qwen-Audio: Advancing Universal Audio Understanding via Unified\n Large-Scale Audio-Language Models"], "answer_arxiv_id": ["2310.13289", "2311.07919"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_3588"} +{"question": "Are there any studies using randomized smoothing to certify white-box ensembles?", "answer": ["On the Certified Robustness for Ensemble Models and Beyond", "Enhancing Certified Robustness via Smoothed Weighted Ensembling"], "answer_arxiv_id": ["2107.10873", "2005.09363"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_3589"} +{"question": "Could you provide me with some of the extensive studies on the various versions of Stochastic Gradient Descent (SGD)?", "answer": ["Stochastic Gradient Descent for Non-smooth Optimization: Convergence Results and Optimal Averaging Schemes", "Making the Last Iterate of SGD Information Theoretically Optimal", "Tight analyses for non-smooth stochastic gradient descent"], "answer_arxiv_id": ["1212.1824", "1904.12443", "1812.05217"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_3590"} +{"question": "Any works that discuss the limitations of a fixed discount factor and present approaches for more flexible discounting?", "answer": ["Rethinking the Discount Factor in Reinforcement Learning: A Decision Theoretic Approach", "Unifying Task Specification in Reinforcement Learning"], "answer_arxiv_id": ["1902.02893", "1609.01995"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3591"} +{"question": "Which studies involve the use of entropy search in multi-objective Bayesian optimization?", "answer": ["Predictive Entropy Search for Multi-objective Bayesian Optimization"], "answer_arxiv_id": ["1511.05467"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_3592"} +{"question": "Can you provide the research papers which explore embedding traffic rules through different logic formulae such as MTL, LTL, and STL in Autonomous Vehicle (AV) planning?", "answer": ["Formalizing Traffic Rules for Machine Interpretability", "Receding Horizon Planning with Rule Hierarchies for Autonomous Vehicles", "Rule-based Optimal Control for Autonomous Driving"], "answer_arxiv_id": ["2007.00330", "2212.03323", "2101.05709"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_3593"} +{"question": "Which works focus on end-to-end motion planning in autonomous driving?", "answer": ["End-to-end Driving via Conditional Imitation Learning", "CIRL: Controllable Imitative Reinforcement Learning for Vision-based\n Self-driving", "ST-P3: End-to-end Vision-based Autonomous Driving via Spatial-Temporal\n Feature Learning", "End-to-end Learning of Driving Models from Large-scale Video Datasets", "Planning-oriented Autonomous Driving", "Multi-Modal Fusion Transformer for End-to-End Autonomous Driving", "Trajectory-guided Control Prediction for End-to-end Autonomous Driving:\n A Simple yet Strong Baseline", "Learning from All Vehicles", "LookOut: Diverse Multi-Future Prediction and Planning for Self-Driving"], "answer_arxiv_id": ["1710.02410", "1807.03776", "2207.07601", "1612.01079", "2212.10156", "2104.09224", "2206.08129", "2203.11934", "2101.06547"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_3594"} +{"question": "What study inspired a large-scale language foundation models for natural language processing?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_3595"} +{"question": "Are there any studies that extended MF-MAB into bandit optimization?", "answer": ["Multi-fidelity Bayesian Optimisation with Continuous Approximations"], "answer_arxiv_id": ["1703.06240"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_3596"} +{"question": "What works introduce the concept of clustering variant of contrastive learning?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2006.09882", "2104.14294"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_3597"} +{"question": "Which works propose to leverage the Python interpreter to augment LLMs ?", "answer": ["PAL: Program-aided Language Models", "Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks", "Lila: A Unified Benchmark for Mathematical Reasoning", "MathPrompter: Mathematical Reasoning using Large Language Models", "ART: Automatic multi-step reasoning and tool-use for large language models", "Multimodal Procedural Planning via Dual Text-Image Prompting"], "answer_arxiv_id": ["2211.10435", "2211.12588", "2210.17517", "2303.05398", "2303.09014", "2305.01795"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_3598"} +{"question": "Can you name any work that fine-tune the LLM to imitate a human-like web-browsing behavior?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "WebGPT: Browser-assisted question-answering with human feedback"], "answer_arxiv_id": ["2302.04761", "2112.09332"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_3599"} +{"question": "Which prior studies have focused on rebalancing approaches for dealing with long-tailed recognition problems?", "answer": ["SMOTE: Synthetic Minority Over-sampling Technique", "Distributed Representations of Words and Phrases and their Compositionality"], "answer_arxiv_id": ["1106.1813", "1310.4546"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_3600"} +{"question": "Which papers talk about the creation of pixel-aligned datasets by altering foreground color in real images?", "answer": ["DoveNet: Deep Image Harmonization via Domain Verification", "Deep Image Harmonization with Learnable Augmentation"], "answer_arxiv_id": ["1911.13239", "2308.00376"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_3601"} +{"question": "Could you provide me some studies about the implementation of nonconvex minimax optimization in federated adversarial training?", "answer": ["Robust Federated Learning: The Case of Affine Distribution Shifts"], "answer_arxiv_id": ["2006.08907"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_3602"} +{"question": "Which papers use masked signal modeling in pre-training for scalable applications?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "SimMIM: A Simple Framework for Masked Image Modeling"], "answer_arxiv_id": ["2111.06377", "1810.04805", "2112.09133", "2111.09886"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_3603"} +{"question": "What works reported the use of Neural Prototype Trees (ProtoTree) in prototype-based networks?", "answer": ["Neural Prototype Trees for Interpretable Fine-grained Image Recognition"], "answer_arxiv_id": ["2012.02046"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_3604"} +{"question": "What papers have highlighted the implicit regularization of GD for several statistical estimation tasks?", "answer": ["Implicit Regularization in Nonconvex Statistical Estimation: Gradient Descent Converges Linearly for Phase Retrieval, Matrix Completion, and Blind Deconvolution", "Nonconvex Matrix Factorization from Rank-One Measurements"], "answer_arxiv_id": ["1711.10467", "1802.06286"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3605"} +{"question": "Could you tell me any studies that have been successful in improving representation learning by using hard negatives?", "answer": ["Smart Mining for Deep Metric Learning", "Sampling Matters in Deep Embedding Learning", "Deep Metric Learning with Hierarchical Triplet Loss"], "answer_arxiv_id": ["1704.01285", "1706.07567", "1810.06951"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_3606"} +{"question": "Which works discuss the concept of adversarial contrastive learning?", "answer": ["Adversarial Self-Supervised Contrastive Learning", "Robust Pre-Training by Adversarial Contrastive Learning", "Contrastive Learning with Adversarial Examples", "When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning?", "Adversarial Contrastive Learning via Asymmetric InfoNCE", "Decoupled Adversarial Contrastive Learning for Self-supervised Adversarial Robustness", "Rethinking the Effect of Data Augmentation in Adversarial Contrastive Learning", "Enhancing Adversarial Contrastive Learning via Adversarial Invariant Regularization"], "answer_arxiv_id": ["2006.07589", "2010.13337", "2010.12050", "2111.01124", "2207.08374", "2207.10899", "2303.01289", "2305.00374"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_3607"} +{"question": "Could you provide me with a work that uses a different technique of forcing the softmax activations to act as sigmoids to approximate GD with general loss functions in transformers?", "answer": ["Looped Transformers as Programmable Computers"], "answer_arxiv_id": ["2301.13196v1"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_3608"} +{"question": "What study proposed the Distribution Matching (DM) model?", "answer": ["Dataset Condensation with Distribution Matching"], "answer_arxiv_id": ["2110.04181"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_3609"} +{"question": "Could you provide me some works about Large Language Models, like Llama and Llama2?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2302.13971", "2307.09288"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_3610"} +{"question": "Which papers have been cited regarding heteroscedastic noise in bandits?", "answer": ["Normal Bandits of Unknown Means and Variances: Asymptotic Optimality, Finite Horizon Regret Bounds, and a Solution to an Open Problem", "Information Directed Sampling and Bandits with Heteroscedastic Noise", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs"], "answer_arxiv_id": ["1504.05823", "1801.09667", "2012.08507", "2205.11507"], "source_meta": {"published_time": "20220228"}, "qid": "AutoScholarQuery_train_3611"} +{"question": "What work leverages a relevance classifier to associate lower level instruction to a higher level task to provide subtask shaping reward?", "answer": ["ELLA: Exploration through Learned Language Abstraction"], "answer_arxiv_id": ["2103.05825"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_3612"} +{"question": "Could you point to a research that introduces sequential Bayesian quadrature?", "answer": ["Optimally-Weighted Herding is Bayesian Quadrature"], "answer_arxiv_id": ["1408.2049v2"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3613"} +{"question": "What studies have proposed advanced backdoor attacks, such as clean-label attacks, invisible-trigger attacks and physical attacks?", "answer": ["Hidden Trigger Backdoor Attacks", "Backdoor Embedding in Convolutional Neural Network Models via Invisible Perturbation", "BppAttack: Stealthy and Efficient Trojan Attacks against Deep Neural Networks via Image Quantization and Contrastive Adversarial Learning", "Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks"], "answer_arxiv_id": ["1910.00033", "1808.10307", "2205.13383", "2007.02343"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3614"} +{"question": "Which research papers have studied stationary RL under the context of Low-rank MDPs?", "answer": ["FLAMBE: Structural Complexity and Representation Learning of Low Rank MDPs", "Representation Learning for Online and Offline RL in Low-rank MDPs", "Improved Sample Complexity for Reward-free Reinforcement Learning under Low-rank MDPs", "Model-free Representation Learning and Exploration in Low-rank MDPs", "Provable Benefit of Multitask Representation Learning in Reinforcement Learning", "Provable Benefits of Representational Transfer in Reinforcement Learning"], "answer_arxiv_id": ["2006.10814", "2110.04652", "2303.10859", "2102.07035", "2206.05900v1", "2205.14571v2"], "source_meta": {"published_time": "20230810"}, "qid": "AutoScholarQuery_train_3615"} +{"question": "Which studies use machine learning techniques to address combinatorial optimization problems?", "answer": ["Accelerate Presolve in Large-Scale Linear Programming via Reinforcement Learning"], "answer_arxiv_id": ["2310.11845"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_3616"} +{"question": "Can you name the work that studied the dynamics of NTK throughout training?", "answer": ["Asymptotics of Wide Networks from Feynman Diagrams", "Dynamics of Deep Neural Networks and Neural Tangent Hierarchy"], "answer_arxiv_id": ["1909.11304", "1909.08156"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_3617"} +{"question": "Which study used the problem of SC as an instance of online linear regression?", "answer": ["Synthetic Control As Online Linear Regression"], "answer_arxiv_id": ["2202.08426"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_3618"} +{"question": "What studies explain about the inability of One-Step Inverse Models to capture long-range dependencies or trivial prediction of actions?", "answer": ["Causal Confusion in Imitation Learning"], "answer_arxiv_id": ["1905.11979"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_3619"} +{"question": "What studies suggest that the transfer performance of instance discrimination methods degrades when tasks of a very different nature are considered?", "answer": ["The iNaturalist Species Classification and Detection Dataset"], "answer_arxiv_id": ["1707.06642"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_3620"} +{"question": "Could you mention any studies where the focus is on differentiating between causal and environment clips in VidQA?", "answer": ["Invariant Grounding for Video Question Answering", "Equivariant and Invariant Grounding for Video Question Answering"], "answer_arxiv_id": ["2206.02349", "2207.12783"], "source_meta": {"published_time": "20240703"}, "qid": "AutoScholarQuery_train_3621"} +{"question": "Which studies have formulated the offline policy evaluation problem as a constrained linear optimization problem?", "answer": ["Reinforcement Learning via Fenchel-Rockafellar Duality", "Off-Policy Evaluation via the Regularized Lagrangian", "Beyond the Return: Off-policy Function Estimation under User-specified Error-measuring Distributions"], "answer_arxiv_id": ["2001.01866", "2007.03438", "2210.15543"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_3622"} +{"question": "Can you name some works that focus on global interpretability methods and consider them as vertical data valuation?", "answer": ["Understanding Global Feature Contributions With Additive Importance Measures"], "answer_arxiv_id": ["2004.00668"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_3623"} +{"question": "What work claimed that a NN can escape the kernel regime by taking one specific large gradient step?", "answer": ["High-dimensional Asymptotics of Feature Learning: How One Gradient Step Improves the Representation"], "answer_arxiv_id": ["2205.01445"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_3624"} +{"question": "Which works provide benchmarks for heterophilic graph learning?", "answer": ["New Benchmarks for Learning on Non-Homophilous Graphs", "Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods"], "answer_arxiv_id": ["2104.01404", "2110.14446"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_3625"} +{"question": "Which works demonstrated the challenges with rewards not being always uniquely identified from expert demonstrations in IRL?", "answer": ["Identifiability in inverse reinforcement learning"], "answer_arxiv_id": ["2106.03498"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_3626"} +{"question": "Which works developed Net2Vec that focused on individual neurons?", "answer": ["Net2Vec: Quantifying and Explaining how Concepts are Encoded by Filters\n in Deep Neural Networks"], "answer_arxiv_id": ["1801.03454"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3627"} +{"question": "Which studies contributed to the field of contrastive learning in the context of computer vision?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments"], "answer_arxiv_id": ["1807.03748", "2002.05709", "1911.05722", "2006.07733", "2011.10566", "2006.09882"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_3628"} +{"question": "Which studies propose a proportional odds model that allows for adjacent categories to have equal weights?", "answer": ["Regularized Ordinal Regression and the ordinalNet R Package"], "answer_arxiv_id": ["1706.05003"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3629"} +{"question": "What research has implemented Fourier Neural Operators in the field of weather forecasting?", "answer": ["FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators"], "answer_arxiv_id": ["2202.11214"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_3630"} +{"question": "Which research demonstrated that the defense of a certain model could be fooled?", "answer": ["Are Odds Really Odd? Bypassing Statistical Detection of Adversarial Examples"], "answer_arxiv_id": ["1907.12138"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_3631"} +{"question": "Could you provide me some studies about adversarial attacks that only perturb a small part of an image or even a single pixel?", "answer": ["Sparse and Imperceivable Adversarial Attacks", "GreedyFool: Distortion-Aware Sparse Adversarial Attack", "One Pixel Attack for Fooling Deep Neural Networks"], "answer_arxiv_id": ["1909.05040", "2010.13773", "1710.08864"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_3632"} +{"question": "Which works introduced sampling layers between flow layers to improve model expressiveness in SNFs?", "answer": ["Stochastic Normalizing Flows", "SurVAE Flows: Surjections to Bridge the Gap between VAEs and Flows"], "answer_arxiv_id": ["2002.06707", "2007.02731"], "source_meta": {"published_time": "20220803"}, "qid": "AutoScholarQuery_train_3633"} +{"question": "Which paper showed that Turing machines that run for T steps can be approximated by encoder-decoder transformers of specific depth and size?", "answer": ["Statistically Meaningful Approximation: a Case Study on Approximating Turing Machines with Transformers"], "answer_arxiv_id": ["2107.13163"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_3634"} +{"question": "Which research introduced the AVSBench-Object and AVSBench-Semantic benchmarks for AVS tasks?", "answer": ["Audio-Visual Segmentation", "Audio-Visual Segmentation with Semantics"], "answer_arxiv_id": ["2207.05042", "2301.13190"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_3635"} +{"question": "Are there any studies that use the hop count prediction for GAD in a self-supervised approach?", "answer": ["Hop-Count Based Self-Supervised Anomaly Detection on Attributed Networks"], "answer_arxiv_id": ["2104.07917"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3636"} +{"question": "What works suggest the application of informative sampling for prognosis in health care or assessing the robustness of predictive models to distribution shifts?", "answer": ["Learning from Clinical Judgments: Semi-Markov-Modulated Marked Hawkes Processes for Risk Prognosis", "DeepJoint: Robust Survival Modelling Under Clinical Presence Shift"], "answer_arxiv_id": ["1705.05267", "2205.13481"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_3637"} +{"question": "Which work introduces the Proximal Policy Gradient Arborescence (PPGA) method, specifically within the field of QD-RL?", "answer": ["Proximal Policy Gradient Arborescence for Quality Diversity Reinforcement Learning"], "answer_arxiv_id": ["2305.13795"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3638"} +{"question": "Which studies explore hardness results for learning algorithms based on label queries?", "answer": ["Hardness of Noise-Free Learning for Two-Hidden-Layer Neural Networks"], "answer_arxiv_id": ["2202.05258"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_3639"} +{"question": "Which papers centered on the robustness aspect of assessing the quality of LLMs?", "answer": ["Measure and Improve Robustness in NLP Models: A Survey", "Robustness Gym: Unifying the NLP Evaluation Landscape"], "answer_arxiv_id": ["2112.08313", "2101.04840"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_3640"} +{"question": "Which works propose methods for simulation agent modelling where sim agent could execute a set of trajectory predictions as its plan?", "answer": ["BARK: Open Behavior Benchmarking in Multi-Agent Environments", "InterSim: Interactive Traffic Simulation via Explicit Relation Modeling"], "answer_arxiv_id": ["2003.02604", "2210.14413"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_3641"} +{"question": "Are there any research works trying to condense prototypes inside the model as a part of explainability?", "answer": ["Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictions", "This Looks Like That: Deep Learning for Interpretable Image Recognition"], "answer_arxiv_id": ["1710.04806", "1806.10574"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_3642"} +{"question": "Which paper extensively explored the topic of CoT prompting in language models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_3643"} +{"question": "In which referenced papers, the computation of a confidence metric is described to find the optimal rejection threshold?", "answer": ["Consistency of plug-in confidence sets for classification in semi-supervised learning"], "answer_arxiv_id": ["1507.07235"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_3644"} +{"question": "What papers propose studies related to person Re-ID in human-centric perception?", "answer": ["Large-Scale Pre-training for Person Re-identification with Noisy Labels", "TransReID: Transformer-based Object Re-Identification", "Self-Supervised Pre-Training for Transformer-Based Person Re-Identification", "PASS: Part-Aware Self-Supervised Pre-Training for Person Re-Identification", "AAformer: Auto-Aligned Transformer for Person Re-Identification", "Self-training with progressive augmentation for unsupervised cross-domain person re-identification", "Implicit Sample Extension for Unsupervised Person Re-Identification"], "answer_arxiv_id": ["2203.16533", "2102.04378", "2111.12084", "2203.03931", "2104.00921", "1907.13315", "2204.06892v1"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_3645"} +{"question": "What research can consider LWR under some specific assumptions in an overparameterized regime?", "answer": ["Overparameterized Nonlinear Learning: Gradient Descent Takes the Shortest Path?", "On the Proof of Global Convergence of Gradient Descent for Deep ReLU Networks with Linear Widths"], "answer_arxiv_id": ["1812.10004", "2101.09612"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_3646"} +{"question": "What works explore weakly-supervised panoptic segmentation?", "answer": ["Weakly- and Semi-Supervised Panoptic Segmentation"], "answer_arxiv_id": ["1808.03575"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_3647"} +{"question": "Any works that employ logic rules to annotate unlabeled samples in weakly supervised learning?", "answer": ["Data Programming: Creating Large Training Sets, Quickly", "Snorkel: Rapid Training Data Creation with Weak Supervision"], "answer_arxiv_id": ["1605.07723", "1711.10160"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_3648"} +{"question": "Which research introduced the idea of using a continuous-time DTW variant known as GDTW?", "answer": ["A General Optimization Framework for Dynamic Time Warping"], "answer_arxiv_id": ["1905.12893"], "source_meta": {"published_time": "20230319"}, "qid": "AutoScholarQuery_train_3649"} +{"question": "Which works achieved optimal contraction rates with respect to the weighted L2 norm using a completely different proof technique?", "answer": ["Bayesian Manifold Regression"], "answer_arxiv_id": ["1305.0617"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_3650"} +{"question": "What works studies the use of self-supervision in building models for visual correspondence?", "answer": ["Learning Correspondence from the Cycle-consistency of Time", "Space-Time Correspondence as a Contrastive Random Walk"], "answer_arxiv_id": ["1903.07593", "2006.14613"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3651"} +{"question": "What works have proposed popular MARL environments with continuous action spaces?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments", "FACMAC: Factored Multi-Agent Centralised Policy Gradients"], "answer_arxiv_id": ["1706.02275", "2003.06709"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_3652"} +{"question": "Which studies proposed certified variants of adversarial training?", "answer": ["Lipschitz-Margin Training: Scalable Certification of Perturbation Invariance for Deep Neural Networks", "Training Certifiably Robust Neural Networks with Efficient Local Lipschitz Bounds", "On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models"], "answer_arxiv_id": ["1802.04034", "2111.01395", "1810.12715"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_3653"} +{"question": "What are some studies about fine-tuning VLMs for improving few-shot classification performance?", "answer": ["SVL-Adapter: Self-Supervised Adapter for Vision-Language Pretrained\n Models", "SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for\n Few-shot Image Classification", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language\n Modeling", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "Enhancing CLIP with GPT-4: Harnessing Visual Descriptions as Prompts"], "answer_arxiv_id": ["2210.03794", "2211.16191", "2111.03930", "2110.04544", "2303.16199", "2307.11661"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_3654"} +{"question": "In which papers is the risk of model reliance on artifacts discussed?", "answer": ["Teach Me to Explain: A Review of Datasets for Explainable Natural\n Language Processing"], "answer_arxiv_id": ["2102.12060"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_3655"} +{"question": "What off-the-shelf inpainting method is utilized to fill in occluded contents?", "answer": ["Resolution-robust Large Mask Inpainting with Fourier Convolutions"], "answer_arxiv_id": ["2109.07161"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_3656"} +{"question": "Which studies have pointed out the texture bias in CNNs?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness"], "answer_arxiv_id": ["1811.12231"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_3657"} +{"question": "Which works fine-tune language models to achieve competitive accuracy on numerous NLP tasks?", "answer": ["Large Language Models Can Be Strong Differentially Private Learners"], "answer_arxiv_id": ["2110.05679"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_3658"} +{"question": "Are there any works that use prompt-based or in-context learning strategies to update the knowledge of Large Language Models (LLMs)?", "answer": ["Memory-assisted prompt editing to improve GPT-3 after deployment", "Can We Edit Factual Knowledge by In-Context Learning?"], "answer_arxiv_id": ["2201.06009", "2305.12740"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_3659"} +{"question": "What work assesses the multi-turn conversation and instruction-following ability of LLMs using MT-Bench?", "answer": ["Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena"], "answer_arxiv_id": ["2306.05685"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_3660"} +{"question": "Which study is developed on DeepIM for template-based methods?", "answer": ["CosyPose: Consistent multi-view multi-object 6D pose estimation"], "answer_arxiv_id": ["2008.08465"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_3661"} +{"question": "What works capitalized on the suodularity of IoU to apply convex Lovasz extensions of suodular functions?", "answer": ["The Lovász Hinge: A Novel Convex Surrogate for Submodular Losses"], "answer_arxiv_id": ["1512.07797"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_3662"} +{"question": "What works discuss existing benchmarks for black-box optimisation?", "answer": ["COCO: A Platform for Comparing Continuous Optimizers in a Black-Box Setting", "IOHprofiler: A Benchmarking and Profiling Tool for Iterative Optimization Heuristics", "IOHexperimenter: Benchmarking Platform for Iterative Optimization Heuristics", "Olympus: a benchmarking framework for noisy optimization and experiment planning"], "answer_arxiv_id": ["1603.08785", "1810.05281", "2111.04077", "2010.04153"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_3663"} +{"question": "Which works provide examples of deep embedding-based multi-view clustering (MVC)?", "answer": ["Deep Embedded Multi-View Clustering via Jointly Learning Latent\n Representations and Graphs", "Deep Embedded Multi-view Clustering with Collaborative Training"], "answer_arxiv_id": ["2205.03803", "2007.13067"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_3664"} +{"question": "Could you provide me with the studies that used classical stereo algorithms for disparity estimation with dual-pixel sensors?", "answer": ["Synthetic Depth-of-Field with a Single-Camera Mobile Phone"], "answer_arxiv_id": ["1806.04171"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_3665"} +{"question": "What are some popular imitation learning methods?", "answer": ["Generative Adversarial Imitation Learning", "Wasserstein Adversarial Imitation Learning", "Primal Wasserstein Imitation Learning"], "answer_arxiv_id": ["1606.03476", "1906.08113", "2006.04678"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_3666"} +{"question": "Which paper discusses a practical alternative parameteriation where neural density estimators are defined on a lower dimensional latent space?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_3667"} +{"question": "What datasets are known for offering unique advantages in training and testing models’ capabilities in the Open-QA community?", "answer": ["TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension"], "answer_arxiv_id": ["1705.03551"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_3668"} +{"question": "What are some papers where image translation is guided by either text or image?", "answer": ["Diffusion-based Image Translation using Disentangled Style and Content\n Representation"], "answer_arxiv_id": ["2209.15264"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_3669"} +{"question": "Which studies proposed variants of FNO?", "answer": ["Multiwavelet-based Operator Learning for Differential Equations", "U-FNO - an enhanced Fourier neural operator-based deep-learning model for multiphase flow", "Factorized Fourier Neural Operators"], "answer_arxiv_id": ["2109.13459", "2109.03697", "2111.13802"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_3670"} +{"question": "Which works propose early fusion architectures for multimodal deep learning?", "answer": ["On the Benefits of Early Fusion in Multimodal Representation Learning"], "answer_arxiv_id": ["2011.07191"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_3671"} +{"question": "What studies argue that invariances enhance generalization in machine learning models?", "answer": ["Unsupervised Learning of Invariant Representations in Hierarchical Architectures", "On Invariance and Selectivity in Representation Learning", "Generalization Error of Invariant Classifiers", "Improved Generalization Bounds of Group Invariant / Equivariant Deep Networks via Quotient Feature Spaces"], "answer_arxiv_id": ["1311.4158v5", "1503.05938", "1610.04574v3", "1910.06552"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_3672"} +{"question": "What studies have incorporated saliency maps to improve classification or object detection performance in neural networks?", "answer": ["Top-Down Saliency Detection Driven by Visual Classification", "Saliency for Fine-grained Object Recognition in Domains with Scarce Training Data", "Hallucinating Saliency Maps for Fine-Grained Image Classification for Limited Data Domains"], "answer_arxiv_id": ["1709.05307", "1808.00262", "2007.12562"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_3673"} +{"question": "Which research used image-text contrastive learning methods in medical object detection?", "answer": ["nnDetection: A Self-configuring Method for Medical Object Detection"], "answer_arxiv_id": ["2106.00817"], "source_meta": {"published_time": "20240203"}, "qid": "AutoScholarQuery_train_3674"} +{"question": "What work proposed a point-wise loss function as a soft monotonicity constraint?", "answer": ["How to Incorporate Monotonicity in Deep Networks While Preserving Flexibility?"], "answer_arxiv_id": ["1909.10662"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_3675"} +{"question": "Are there any papers about early exiting strategy for ViT models?", "answer": ["You Need Multiple Exiting: Dynamic Early Exiting for Accelerating\n Unified Vision Language Model"], "answer_arxiv_id": ["2211.11152"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3676"} +{"question": "Which papers explore disentangled representation learning using generative models?", "answer": ["Isolating Sources of Disentanglement in Variational Autoencoders", "Variational Inference of Disentangled Latent Concepts from Unlabeled Observations", "Deforming Autoencoders: Unsupervised Disentangling of Shape and Appearance", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["1802.04942", "1711.00848", "1806.06503", "1812.04948"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_3677"} +{"question": "What works discuss the concept of text-to-image models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2205.11487", "2112.10741", "2204.06125"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_3678"} +{"question": "Can you name studies that attempt to learn 3D structure of the real world?", "answer": ["3D Neural Scene Representations for Visuomotor Control", "Reinforcement Learning with Neural Radiance Fields"], "answer_arxiv_id": ["2107.04004", "2206.01634"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_3679"} +{"question": "Could you tell me what work introduced the concept of augmentation graph in analysing contrastive learning?", "answer": ["Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss"], "answer_arxiv_id": ["2106.04156"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_3680"} +{"question": "What research papers propose ways to improve the faster computation of Transformer building blocks?", "answer": ["Efficient Transformers: A Survey", "FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness", "Hierarchical Transformers Are More Efficient Language Models"], "answer_arxiv_id": ["2009.06732", "2205.14135", "2110.13711"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_3681"} +{"question": "Which studies are related to the personalization of generative adversarial networks (GANs) through GAN inversion?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["1812.04948", "1912.04958"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3682"} +{"question": "Which works propose methods for low-light enhancement and denoising based on reflectance and illumination maps?", "answer": ["Deep Retinex Decomposition for Low-Light Enhancement"], "answer_arxiv_id": ["1808.04560"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_3683"} +{"question": "What works treated learning 3D geometry prior and reconstructing objects from images as a conditional generation?", "answer": ["GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images", "Locally Attentional SDF Diffusion for Controllable 3D Shape Generation", "HoloDiffusion: Training a 3D Diffusion Model using 2D Images", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "LION: Latent Point Diffusion Models for 3D Shape Generation"], "answer_arxiv_id": ["2209.11163", "2305.04461", "2303.16509", "2212.04493", "2210.06978"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_3684"} +{"question": "Could you provide me studies that switched to bfloat16 when float16 training resulted in loss divergence?", "answer": ["Reproducible scaling laws for contrastive language-image learning"], "answer_arxiv_id": ["2212.07143"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_3685"} +{"question": "Can you list some studies that focused on learning underlying group structure solely from symmetries contained in data?", "answer": ["disentangling images with lie group transformations and sparse coding"], "answer_arxiv_id": ["2012.12071"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_3686"} +{"question": "Which works explain aleatoric uncertainty in the context of fitting a neural network?", "answer": ["A Review of Uncertainty Quantification in Deep Learning: Techniques,\n Applications and Challenges", "What Uncertainties Do We Need in Bayesian Deep Learning for Computer\n Vision?", "Stochastic Segmentation Networks: Modelling Spatially Correlated\n Aleatoric Uncertainty"], "answer_arxiv_id": ["2011.06225", "1703.04977", "2006.06015"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_3687"} +{"question": "Which reference models were augmented with inductive biases that encourage Lagrangian-like magnification?", "answer": ["Learning-based Video Motion Magnification"], "answer_arxiv_id": ["1804.02684v3"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3688"} +{"question": "What studies focus on the issue of compatible representations learning?", "answer": ["Understanding image representations by measuring their equivariance and\n equivalence", "Convergent Learning: Do different neural networks learn the same\n representations?", "Towards Understanding Learning Representations: To What Extent Do\n Different Neural Networks Learn the Same Representation", "Similarity of Neural Network Representations Revisited", "Revisiting Model Stitching to Compare Neural Representations"], "answer_arxiv_id": ["1411.5908", "1511.07543", "1810.11750", "1905.00414", "2106.07682"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_3689"} +{"question": "Which studies demonstrated that an agent’s world model is implicitly a forward model that predict future states?", "answer": ["Recurrent World Models Facilitate Policy Evolution", "Learning to Predict Without Looking Ahead: World Models Without Forward Prediction"], "answer_arxiv_id": ["1809.01999", "1910.13038"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_3690"} +{"question": "What work estimates the global directions by relying on the randomly sampled latent codes in StyleGAN?", "answer": ["GANSpace: Discovering Interpretable GAN Controls"], "answer_arxiv_id": ["2004.02546"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_3691"} +{"question": "Could you cite any works that proposed federated learning with heterogeneous client models to save both computation and communication?", "answer": ["Expanding the Reach of Federated Learning by Reducing Client Resource Requirements", "Federated Learning: Challenges, Methods, and Future Directions", "FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout", "HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients", "Expanding the Reach of Federated Learning by Reducing Client Resource Requirements", "FedRolex: Model-Heterogeneous Federated Learning with Rolling Sub-Model Extraction", "MAC-PO: Multi-Agent Experience Replay via Collective Priority Optimization", "Option-Aware Adversarial Inverse Reinforcement Learning for Robotic Control", "Multi-Agent Reinforcement Learning with Shared Resources for Inventory Management", "A Versatile Multi-Agent Reinforcement Learning Benchmark for Inventory Management", "A Causal Inference Approach to Eliminate the Impacts of Interfering Factors on Traffic Performance Evaluation", "A Multi-agent Reinforcement Learning Approach for Efficient Client Selection in Federated Learning"], "answer_arxiv_id": ["1812.07210", "1908.07873", "2102.13451", "2010.01264", "1812.07210", "2212.01548", "2302.10418v2", "2210.01969", "2212.07684", "2306.07542", "2308.03545v1", "2201.02932"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_3692"} +{"question": "Could you provide me the studies about performing spatial pyramid pooling at several grid scales for Semantic segmentation?", "answer": ["Pyramid Scene Parsing Network"], "answer_arxiv_id": ["1612.01105"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_3693"} +{"question": "What studies target learning a representation of only the common information between views in multi-view contrastive learning?", "answer": ["Contrastive Multiview Coding", "What Makes for Good Views for Contrastive Learning?", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1906.05849", "2005.10243", "2002.05709"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_3694"} +{"question": "What are the papers that discussed the problem of learned manifold not optimized within its latent space representation in most generative models?", "answer": ["Representing Deep Neural Networks Latent Space Geometries with Graphs"], "answer_arxiv_id": ["2011.07343v1"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_3695"} +{"question": "Could you provide me some works on program induction and neuro-symbolic methods?", "answer": ["Neural Programmer-Interpreters", "Making Neural Programming Architectures Generalize via Recursion", "Neural Module Networks", "Neurosymbolic Transformers for Multi-Agent Communication", "Discovering Symbolic Models from Deep Learning with Inductive Biases", "Sequence-to-Sequence Learning with Latent Neural Grammars"], "answer_arxiv_id": ["1511.06279", "1704.06611v1", "1511.02799v4", "2101.03238", "2006.11287", "2109.01135v7"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3696"} +{"question": "Which works show progress in the wide-baseline two-view setting in camera pose estimation?", "answer": ["Wide-Baseline Relative Camera Pose Estimation with Directional Learning", "Extreme Rotation Estimation using Dense Correlation Volumes", "RPNet: an End-to-End Network for Relative Camera Pose Estimation", "Extreme Relative Pose Network under Hybrid Representations", "Extreme Relative Pose Estimation for RGB-D Scans via Scene Completion"], "answer_arxiv_id": ["2106.03336", "2104.13530", "1809.08402", "1912.11695", "1901.00063"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3697"} +{"question": "What works have proposed disklling GNNs onto MLPs in terms of Knowledge Distillation for GNNs?", "answer": ["Graph-less Neural Networks: Teaching Old MLPs New Tricks via Distillation", "Cold Brew: Distilling Graph Node Representations with Incomplete or Missing Neighborhoods"], "answer_arxiv_id": ["2110.08727", "2111.04840"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_3698"} +{"question": "What works are about the popular benchmark datasets like GLUE, SuperGLUE, and SQuAD for evaluating NLP models?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding", "SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems", "SQuAD: 100,000+ Questions for Machine Comprehension of Text"], "answer_arxiv_id": ["1804.07461", "1905.00537", "1606.05250"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_3699"} +{"question": "What studies focused on re-texturizing a given mesh representation of 3D scenes or objects?", "answer": ["RoomDreamer: Text-Driven 3D Indoor Scene Synthesis with Coherent\n Geometry and Texture", "MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion", "Text2Mesh: Text-Driven Neural Stylization for Meshes"], "answer_arxiv_id": ["2305.11337", "2307.01097", "2112.03221"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_3700"} +{"question": "What papers discuss controllable text generation?", "answer": ["Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks", "CTRL: A Conditional Transformer Language Model for Controllable Generation", "Plug and Play Language Models: a Simple Approach to Controlled Text Generation", "GeDi: Generative Discriminator guided Sequence Generation", "DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts"], "answer_arxiv_id": ["2004.10964", "1909.05858", "1912.02164", "2009.06367", "2105.03023"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_3701"} +{"question": "Could you provide me some works on one-stage methods in HOI detection that use transformer-based architectures?", "answer": ["End-to-End Object Detection with Transformers", "MSTR: Multi-Scale Transformer for End-to-End Human-Object Interaction Detection", "Category-Aware Transformer Network for Better Human-Object Interaction Detection", "Mining the Benefits of Two-stage and One-stage HOI Detection", "Efficient Two-Stage Detection of Human–Object Interactions with a Novel Unary–Pairwise Transformer"], "answer_arxiv_id": ["2005.12872", "2203.14709", "2204.04911", "2108.05077", "2112.01838"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_3702"} +{"question": "What papers should be considered for looking at prior works that do not use external proposals and also directly predict tubelets?", "answer": ["Action Tubelet Detector for Spatio-Temporal Action Localization", "Actions as Moving Points", "TACNet: Transition-Aware Context Network for Spatio-Temporal Action\n Detection", "Online Real-time Multiple Spatiotemporal Action Localisation and\n Prediction", "TubeR: Tubelet Transformer for Video Action Detection"], "answer_arxiv_id": ["1705.01861", "2001.04608", "1905.13417", "1611.08563", "2104.00969"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_3703"} +{"question": "Could you provide me the references that proposed to use the attention mechanism to improve target-specific context representations?", "answer": ["Attentive Neural Processes"], "answer_arxiv_id": ["1901.05761"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_3704"} +{"question": "Which research works on a prior derived from a PVL model in the context of expert models?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated\n Mixture-of-Experts Layer", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts", "MoEBERT: from BERT to Mixture-of-Experts via Importance-Guided\n Adaptation"], "answer_arxiv_id": ["1701.06538", "2112.06905", "2204.07675"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_3705"} +{"question": "Which works try to distil a large dataset into a smaller but informative synthetic one in the field of Dataset Distillation?", "answer": ["Dataset Distillation", "Dataset Condensation with Gradient Matching", "Towards Lossless Dataset Distillation via Difficulty-Aligned Trajectory Matching", "DREAM: Efficient Dataset Distillation by Representative Matching", "Dataset Quantization", "DiM: Distilling Dataset into Generative Model", "DARTS: Differentiable Architecture Search", "CAFE: Learning to Condense Dataset by Aligning Features", "Dataset Distillation by Matching Training Trajectories", "Generalizing Dataset Distillation via Deep Generative Prior", "Towards Lossless Dataset Distillation via Difficulty-Aligned Trajectory Matching"], "answer_arxiv_id": ["1811.10959", "2006.05929", "2310.05773", "2302.14416", "2308.10524", "2303.04707", "1806.09055", "2203.01531", "2203.11932", "2305.01649", "2310.05773"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_3706"} +{"question": "Can you provide researches building known linear physical constraints directly into deep learning architectures?", "answer": ["Enforcing Analytic Constraints in Neural-Networks Emulating Physical Systems"], "answer_arxiv_id": ["1909.00912"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_3707"} +{"question": "In what studies did the researcher consider the noisy-label literature to derive the form of the HET classifier?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?"], "answer_arxiv_id": ["1703.04977"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_3708"} +{"question": "Which papers reported on the phenomena of 'grokking'?", "answer": ["Unifying Grokking and Double Descent"], "answer_arxiv_id": ["2303.06173"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_3709"} +{"question": "Which papers introduced advanced versions of Transformer models?", "answer": ["Efficient Transformers: A Survey", "Long Range Arena: A Benchmark for Efficient Transformers"], "answer_arxiv_id": ["2009.06732", "2011.04006"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_3710"} +{"question": "Which studies pertain to Image Generation with Spatial Controls, a form of conditional image synthesis task?", "answer": ["Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "Unsupervised Image-to-Image Translation Networks", "Image-to-Image Translation with Conditional Adversarial Networks", "Video-to-Video Synthesis", "High-Resolution Image Synthesis and Semantic Manipulation with\n Conditional GANs", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "Conditional Image Generation with PixelCNN Decoders", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "GLIGEN: Open-Set Grounded Text-to-Image Generation", "Adding Conditional Control to Text-to-Image Diffusion Models", "LayoutGPT: Compositional Visual Planning and Generation with Large\n Language Models", "Conditional Image Generation and Manipulation for User-Specified Content"], "answer_arxiv_id": ["1703.10593", "1703.00848", "1611.07004", "1808.06601", "1711.11585", "1711.10485", "1606.05328", "2208.01626", "2203.13131", "2301.07093", "2302.05543", "2305.15393", "2005.04909"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_3711"} +{"question": "Which papers have made an attempt to analyze the impacts of pre-training?", "answer": ["Pretrained Transformers Improve Out-of-Distribution Robustness", "Rethinking ImageNet Pre-training", "On Robustness and Transferability of Convolutional Neural Networks", "Rethinking ImageNet Pre-training", "Do Better ImageNet Models Transfer Better?", "Using Pre-Training Can Improve Model Robustness and Uncertainty", "Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuning", "An Empirical Investigation of the Role of Pre-training in Lifelong Learning"], "answer_arxiv_id": ["2004.06100", "1811.08883", "2007.08558", "1811.08883", "1805.08974", "1901.09960", "2003.12862", "2112.09153"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_3712"} +{"question": "Could you provide me some studies related to data-free model extraction attack?", "answer": ["Data-Free Model Extraction"], "answer_arxiv_id": ["2011.14779v2"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_3713"} +{"question": "Any works about training evaluation models for natural language processing to provide stable and effective evaluations at a low cost?", "answer": ["PandaLM: An Automatic Evaluation Benchmark for LLM Instruction Tuning\n Optimization", "Generative Judge for Evaluating Alignment", "Prometheus: Inducing Fine-grained Evaluation Capability in Language\n Models"], "answer_arxiv_id": ["2306.05087", "2310.05470", "2310.08491"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_3714"} +{"question": "What research employed a dataset from two human players collaborating on a RTS game?", "answer": ["Hierarchical Decision Making by Generating and Following Natural Language Instructions"], "answer_arxiv_id": ["1906.00744"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3715"} +{"question": "Which research papers validated the effectiveness of wavelet analysis in image processing?", "answer": ["Wavelet Convolutional Neural Networks for Texture Classification", "Wavelet Knowledge Distillation: Towards Efficient Image-to-Image\n Translation"], "answer_arxiv_id": ["1707.07394", "2203.06321"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_3716"} +{"question": "What works propose to use reusable units stacked hierarchically for multi-grained video content representation?", "answer": ["Hierarchical Conditional Relation Networks for Video Question Answering"], "answer_arxiv_id": ["2002.10698"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_3717"} +{"question": "Which research added conditioning to pre-trained diffusion models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_3718"} +{"question": "Which works demonstrate the connection of 1-Lipschitz neural nets to optimal transport theory?", "answer": ["Achieving robustness in classification using optimal transport with hinge regularization"], "answer_arxiv_id": ["2006.06520"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_3719"} +{"question": "Could you provide me some works that introduced scalable approaches in neural wave functions?", "answer": ["Better, Faster Fermionic Neural Networks", "Deep neural network solution of the electronic Schrödinger equation"], "answer_arxiv_id": ["2011.07125", "1909.08423"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_3720"} +{"question": "Where do the theories justifying the role of probability calibration in label shift come from?", "answer": ["A Unified View of Label Shift Estimation"], "answer_arxiv_id": ["2003.07554"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_3721"} +{"question": "Which studies proposed optimization-based methods to embed subjects into diffusion models?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2212.04488"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_3722"} +{"question": "Which works aim to speed up the sampling of DMs by modifying the underlying stochastic process?", "answer": ["On Fast Sampling of Diffusion Probabilistic Models", "Learning to Efficiently Sample from Diffusion Probabilistic Models"], "answer_arxiv_id": ["2106.00132", "2106.03802"], "source_meta": {"published_time": "20220429"}, "qid": "AutoScholarQuery_train_3723"} +{"question": "Which study has focused on producing stable matchings satisfying the PIIF notion and contrasts with the present research?", "answer": ["On Fairness and Stability in Two-Sided Matchings"], "answer_arxiv_id": ["2111.10885"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_3724"} +{"question": "Can you provide some papers where diffusion models were successful in generating image data?", "answer": ["Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239", "2102.09672"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_3725"} +{"question": "Which research papers have used simulators for data generation in the context of visual-tactile datasets?", "answer": ["ObjectFolder: A Dataset of Objects with Implicit Visual, Auditory, and\n Tactile Representations", "ObjectFolder 2.0: A Multisensory Object Dataset for Sim2Real Transfer"], "answer_arxiv_id": ["2109.07991", "2204.02389"], "source_meta": {"published_time": "20240507"}, "qid": "AutoScholarQuery_train_3726"} +{"question": "What works utilize translations of either MSCOCO or Flickr8k?", "answer": ["Violet: A Vision-Language Model for Arabic Image Captioning with Gemini\n Decoder"], "answer_arxiv_id": ["2311.08844"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_3727"} +{"question": "What works have explored a time-discrete approach, specializing in Runge-Kutta schemes?", "answer": ["The Neural Particle Method - An Updated Lagrangian Physics Informed Neural Network for Computational Fluid Dynamics"], "answer_arxiv_id": ["2003.10208"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_3728"} +{"question": "Could you provide me with studies that proposed to learn 3D-aware representations and generation from ImageNet using transformer-based autoencoders?", "answer": ["VQ3D: Learning a 3D-Aware Generative Model on ImageNet"], "answer_arxiv_id": ["2302.06833"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_3729"} +{"question": "Which were the works that discovered that pretraining with contrastive learning improves robustness to label noise?", "answer": ["Contrastive Learning Improves Model Robustness Under Label Noise"], "answer_arxiv_id": ["2104.08984"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_3730"} +{"question": "Are there any works that use masked language models to couple the original word with top coordinate terms from the pre-trained MLM head?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "SPLADE: Sparse Lexical and Expansion Model for First Stage Ranking"], "answer_arxiv_id": ["1810.04805", "1907.11692", "2107.05720"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_3731"} +{"question": "Which work focuses on the issue of computational efficiency in hybrid RL?", "answer": ["Hybrid RL: Using Both Offline and Online Data Can Make RL Efficient"], "answer_arxiv_id": ["2210.06718"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_3732"} +{"question": "Which works focused on the evaluation quality of multilingual text representation, through the development of cross-lingual and multilingual tasks?", "answer": ["XNLI: Evaluating Cross-lingual Sentence Representations", "MLQA: Evaluating Cross-lingual Extractive Question Answering", "A Comparison of Architectures and Pretraining Methods for Contextualized Multilingual Word Embeddings"], "answer_arxiv_id": ["1809.05053", "1910.07475", "1912.10169"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_3733"} +{"question": "Which papers have focused on reducing PLMs hallucination by guiding the model on correct knowledge usage or improving knowledge selection?", "answer": ["Increasing Faithfulness in Knowledge-Grounded Dialogue with Controllable Features", "Retrieval Augmentation Reduces Hallucination in Conversation"], "answer_arxiv_id": ["2107.06963", "2104.07567"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3734"} +{"question": "What research works have established an O(1/sqrt{N}) optimality gap for the setting of finite-horizon total reward RBs?", "answer": ["An Asymptotically Optimal Index Policy for Finite-Horizon Restless Bandits", "Indexability is Not Enough for Whittle: Improved, Near-Optimal Algorithms for Restless Bandits"], "answer_arxiv_id": ["1707.00205", "2211.00112"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_3735"} +{"question": "What studies utilized the ability of compositional reasoning in various computer vision applications including vision and language?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning", "VisualBERT: A Simple and Performant Baseline for Vision and Language", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "LXMERT: Learning Cross-Modality Encoder Representations from Transformers"], "answer_arxiv_id": ["1909.11740", "1908.03557", "2004.06165", "1908.07490"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_3736"} +{"question": "What research papers have proposed learning rotation equivariant features?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural\n networks for 3D point clouds", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "ZZ-Net: A Universal Rotation Equivariant Architecture for 2D Point\n Clouds", "Quaternion Equivariant Capsule Networks for 3D Point Clouds", "Equivariant Point Cloud Analysis via Learning Orientations for Message\n Passing"], "answer_arxiv_id": ["1802.08219", "2006.10503", "2111.15341", "1912.12098v3", "2203.14486"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_3737"} +{"question": "What works discussed manipulating environment layouts in curriculum learning?", "answer": ["Emergent Tool Use From Multi-Agent Autocurricula", "Teacher algorithms for curriculum learning of Deep RL in continuously parameterized environments"], "answer_arxiv_id": ["1909.07528v2", "1910.07224"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_3738"} +{"question": "Any works about exploring masking strategies in transformer architecture?", "answer": ["GD-MAE: Generative Decoder for MAE Pre-training on LiDAR Point Clouds", "MV-JAR: Masked Voxel Jigsaw and Reconstruction for LiDAR-Based Self-Supervised Pre-Training"], "answer_arxiv_id": ["2212.03010", "2303.13510"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3739"} +{"question": "Can you provide the reference that used loss-capacity curves to analyze the double-descent phenomenon in neural networks?", "answer": ["Deep Double Descent: Where Bigger Models and More Data Hurt"], "answer_arxiv_id": ["1912.02292"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_3740"} +{"question": "Who created a dataset of author profiling from Reddit, with labels of 8 personal attributes for each profile?", "answer": ["Beyond Memorization: Violating Privacy Via Inference with Large Language Models"], "answer_arxiv_id": ["2310.07298v2"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_3741"} +{"question": "What are the studies which present the mathematical model for the evolution of action loss?", "answer": ["Online Learning of Rested and Restless Bandits", "Multi-armed Bandit Problem with Known Trend", "Blocking Bandits", "Recovering Bandits", "Stochastic Bandits with Delay-Dependent Payoffs", "Discrepancy-Based Algorithms for Non-Stationary Rested Bandits"], "answer_arxiv_id": ["1102.3508", "1508.07091", "1907.11975", "1910.14354", "1910.02757", "1710.10657v3"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_3742"} +{"question": "Which studies dealt with learning Human language via token prediction?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_3743"} +{"question": "Which papers explore the transfer learning in the context of PINNs?", "answer": ["One-Shot Transfer Learning of Physics-Informed Neural Networks"], "answer_arxiv_id": ["2110.11286v2"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3744"} +{"question": "What research proposed a reconstruction loss that directly minimizes the blur error of the VAE?", "answer": ["Explicitly Minimizing the Blur Error of Variational Autoencoders"], "answer_arxiv_id": ["2304.05939"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_3745"} +{"question": "Which works tried backtracking methods for white-box explanations?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "Visualizing and Understanding Convolutional Networks", "Striving for Simplicity: The All Convolutional Net"], "answer_arxiv_id": ["1312.6034", "1311.2901", "1412.6806"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_3746"} +{"question": "Any papers conducted about natural adversarial examples indicating that DNNs don't learn real semantic information during training?", "answer": ["Natural Adversarial Examples"], "answer_arxiv_id": ["1907.07174"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_3747"} +{"question": "Can you mention some studies that have applied graph data mining in computer vision tasks such as image retrieval, object detection, and image classification?", "answer": ["Spectral Networks and Locally Connected Networks on Graphs", "Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering", "Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "Deep Visual-Semantic Alignments for Generating Image Descriptions", "SIGMA: Semantic-complete Graph Matching for Domain Adaptive Object Detection", "Image Classification using Graph Neural Network and Multiscale Wavelet Superpixels"], "answer_arxiv_id": ["1312.6203", "1606.09375", "1609.02907", "1710.10903", "1412.2306", "2203.06398", "2201.12633"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_3748"} +{"question": "Could you provide any works about hybrid optimization methods?", "answer": ["Adaptive Gradient Methods with Dynamic Bound of Learning Rate", "Improving Generalization Performance by Switching from Adam to SGD"], "answer_arxiv_id": ["1902.09843", "1712.07628"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_3749"} +{"question": "Which prior works have explored CNNs for novel view synthesis?", "answer": ["Free View Synthesis", "Deferred Neural Rendering: Image Synthesis using Neural Textures"], "answer_arxiv_id": ["2008.05511", "1904.12356"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_3750"} +{"question": "What works discuss the application of DPPs in skill discovery for diversity enhancement?", "answer": ["Determinantal point processes for machine learning"], "answer_arxiv_id": ["1207.6083"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_3751"} +{"question": "Can you name some works that proposed a subspace representation learning as a pretext task in deep clustering?", "answer": ["Deep Subspace Clustering Networks", "Learning a Self-Expressive Network for Subspace Clustering"], "answer_arxiv_id": ["1709.02508", "2110.04318v1"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_3752"} +{"question": "What studies derive a relationship between the network architecture and the largest learning rate that would allow the network to converge to the global minimum when trained with SGD?", "answer": ["Universal Statistics of Fisher Information in Deep Neural Networks: Mean Field Approach"], "answer_arxiv_id": ["1806.01316"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_3753"} +{"question": "Which papers studied the critical ability of continual learning for artificial neural networks to accommodate real-world changes?", "answer": ["A Comprehensive Survey of Continual Learning: Theory, Method and Application", "Incorporating neuro-inspired adaptability for continual learning in artificial intelligence"], "answer_arxiv_id": ["2302.00487", "2308.14991"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_3754"} +{"question": "Could you tell me which paper used contrastive learning in the application of style transfer in diffusion models?", "answer": ["Zero-Shot Contrastive Loss for Text-Guided Diffusion Image Style\n Transfer"], "answer_arxiv_id": ["2303.08622"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_3755"} +{"question": "What paper approaches offline IL through weighted behavior cloning?", "answer": ["Discriminator-Weighted Offline Imitation Learning from Suboptimal Demonstrations"], "answer_arxiv_id": ["2207.10050"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_3756"} +{"question": "Could you provide me some studies that have used contrastive learning to solve repitition issue in language models?", "answer": ["A Contrastive Framework for Neural Text Generation", "A Simple Contrastive Learning Objective for Alleviating Neural Text Degeneration"], "answer_arxiv_id": ["2202.06417", "2205.02517"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_3757"} +{"question": "What research examined and proved the statistical sample complexity of score matching?", "answer": ["Statistical Efficiency of Score Matching: The View from Isoperimetry"], "answer_arxiv_id": ["2210.00726"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_3758"} +{"question": "Which papers discuss projection-based methods for understanding 3D point cloud data?", "answer": ["Multi-view Convolutional Neural Networks for 3D Shape Recognition", "Vehicle Detection from 3D Lidar Using Fully Convolutional Network", "Multi-View 3D Object Detection Network for Autonomous Driving", "PointPillars: Fast Encoders for Object Detection from Point Clouds"], "answer_arxiv_id": ["1505.00880", "1608.07916", "1611.07759", "1812.05784"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_3759"} +{"question": "What works studied the convergence of iterative optimization algorithms to GNEs?", "answer": ["Augmented Lagrangian Methods for the Solution of Generalized Nash Equilibrium Problems", "First-Order Algorithms for Nonlinear Generalized Nash Equilibrium Problems", "Exploitability Minimization in Games and Beyond"], "answer_arxiv_id": ["1807.04474v1", "2204.03132v3", "2210.10207v1"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_3760"} +{"question": "Could you provide me some studies about regret minimization with delayed feedback in Online Optimization and Multi-armed bandit in stochastic and adversarial settings?", "answer": ["Distributed Delayed Stochastic Optimization", "Bandits with Delayed, Aggregated Anonymous Feedback", "Delay and Cooperation in Nonstochastic Bandits", "Nonstochastic Multiarmed Bandits with Unrestricted Delays"], "answer_arxiv_id": ["1104.5525v1", "1709.06853", "1602.04741", "1906.00670"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_3761"} +{"question": "Which papers discuss about Video Captioning?", "answer": ["Video Description: A Survey of Methods, Datasets and Evaluation Metrics", "VATEX: A Large-Scale, High-Quality Multilingual Dataset for\n Video-and-Language Research", "Dense-Captioning Events in Videos", "VALUE: A Multi-Task Benchmark for Video-and-Language Understanding\n Evaluation", "SwinBERT: End-to-End Transformers with Sparse Attention for Video\n Captioning"], "answer_arxiv_id": ["1806.00186", "1904.03493", "1705.00754", "2106.04632", "2111.13196"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_3762"} +{"question": "What research discuss about unrolled dynamical systems leading to chaotic loss landscapes?", "answer": ["Gradients are Not All You Need"], "answer_arxiv_id": ["2111.05803"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_3763"} +{"question": "In what works do the researchers focus on 3D fractured object reassembly?", "answer": ["Neural Shape Mating: Self-Supervised Object Assembly with Adversarial\n Shape Priors"], "answer_arxiv_id": ["2205.14886"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_3764"} +{"question": "Any studies proposed decoupled training as part of their methodology?", "answer": ["BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed Visual Recognition", "Decoupling Representation and Classifier for Long-Tailed Recognition"], "answer_arxiv_id": ["1912.02413", "1910.09217"], "source_meta": {"published_time": "20221230"}, "qid": "AutoScholarQuery_train_3765"} +{"question": "Could you provide the details of the study which used entire kinematic chains of local pose transformations to obtain finer-grained alignment?", "answer": ["gSDF: Geometry-Driven Signed Distance Functions for 3D Hand-Object\n Reconstruction"], "answer_arxiv_id": ["2304.11970"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_3766"} +{"question": "What studies simplify muscular control learning by manually grouping muscles?", "answer": ["Reinforcement Learning of Musculoskeletal Control from Functional Simulations"], "answer_arxiv_id": ["2007.06669"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_3767"} +{"question": "Could you tell me what studies are about generative methods in Graph SSL?", "answer": ["Inductive Representation Learning on Large Graphs", "N-Gram Graph: Simple Unsupervised Representation for Graphs, with Applications to Molecules", "Strategies for Pre-training Graph Neural Networks", "GPT-GNN: Generative Pre-Training of Graph Neural Networks"], "answer_arxiv_id": ["1706.02216", "1806.09206", "1905.12265", "2006.15437"], "source_meta": {"published_time": "20220616"}, "qid": "AutoScholarQuery_train_3768"} +{"question": "Could you provide me some works about the emergence and advancements of GPT-3, ChatGPT and GPT-4?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "GPT-4 Technical Report"], "answer_arxiv_id": ["2201.11903", "2303.08774"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_3769"} +{"question": "Which papers proposed using an energy score and demonstrated its advantages for out-of-distribution uncertainty estimation?", "answer": ["Energy-based Out-of-distribution Detection", "Can multi-label classification networks know what they don’t know?"], "answer_arxiv_id": ["2010.03759", "2109.14162"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_3770"} +{"question": "Which works focus on noise-based strategies involving integration of random continuous noise?", "answer": ["Towards Robust Neural Machine Translation", "NEFTune: Noisy Embeddings Improve Instruction Finetuning", "A Good Sample is Hard to Find: Noise Injection Sampling and\n Self-Training for Neural Language Generation Models"], "answer_arxiv_id": ["1805.06130", "2310.05914v2", "1911.03373"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_3771"} +{"question": "Which studies improve the speed of diffusion models by combining them with GANs and other generative models?", "answer": ["Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed", "Accelerating Diffusion Models via Early Stop of the Diffusion Process", "How Much is Enough? A Study on Diffusion Times in Score-based Generative Models", "D2C: Diffusion-Decoding Models for Few-Shot Conditional Generation", "Diffusion Priors In Variational Autoencoders", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["2101.02388", "2205.12524", "2206.05173", "2106.06819", "2106.15671v1", "2106.05931"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_3772"} +{"question": "What are some regularization-based approaches that address catastrophic forgetting in continual learning?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Continual Learning Through Synaptic Intelligence", "Learning without Forgetting"], "answer_arxiv_id": ["1612.00796", "1703.04200", "1606.09282"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_3773"} +{"question": "What work introduced graph random feature (GRF) mechanism?", "answer": ["Taming graph kernels with random features"], "answer_arxiv_id": ["2305.00156"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_3774"} +{"question": "Could you provide me some works discussing language-based tasks in Continual learning?", "answer": ["Continual Lifelong Learning in Natural Language Processing: A Survey", "LAMOL: LAnguage MOdeling for Lifelong Language Learning"], "answer_arxiv_id": ["2012.09823", "1909.03329"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_3775"} +{"question": "Which papers have discussed the potential of LLMs for content moderation?", "answer": ["Can Large Language Models Transform Computational Social Science?", "Adapting Large Language Models for Content Moderation: Pitfalls in Data\n Engineering and Supervised Fine-tuning"], "answer_arxiv_id": ["2305.03514", "2310.03400"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_3776"} +{"question": "What research papers have focused on disentangling domain-invariant and domain-specific factors for casual invariance?", "answer": ["Deep causal representation learning for unsupervised domain adaptation", "Make the U in UDA Matter: Invariant Consistency Learning for\n Unsupervised Domain Adaptation"], "answer_arxiv_id": ["1910.12417", "2309.12742"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3777"} +{"question": "What works are related to advancements in 3D occupancy prediction such as modality fusion, multi-task learning and end-to-end autonomous driving?", "answer": ["OpenOccupancy: A Large Scale Benchmark for Surrounding Semantic\n Occupancy Perception", "Scene as Occupancy"], "answer_arxiv_id": ["2303.03991", "2306.02851"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_3778"} +{"question": "What works are there about alternative approaches to the Shapley value method by relaxing some of the underlying fair division axioms?", "answer": ["Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning", "The Shapley Value in Machine Learning"], "answer_arxiv_id": ["2110.14049", "2202.05594"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_3779"} +{"question": "What was the first work to address 3D scene semantic completion from camera images?", "answer": ["MonoScene: Monocular 3D Semantic Scene Completion"], "answer_arxiv_id": ["2112.00726"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_3780"} +{"question": "What research proposes using voxel grids, octree, meshes, point clouds, and shape primitives in representing 3D geometry of objects and scenes?", "answer": ["3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction", "Octree Generating Networks: Efficient Convolutional Architectures for High-resolution 3D Outputs", "Neural 3D Mesh Renderer", "A Point Set Generation Network for 3D Object Reconstruction from a Single Image", "3D-PRNN: Generating Shape Primitives with Recurrent Neural Networks"], "answer_arxiv_id": ["1604.00449", "1703.09438", "1711.07566", "1612.00603", "1708.01648"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_3781"} +{"question": "Can you provide me some works that employ graph structures to model complex relationships among multiple activities?", "answer": ["Location-aware Graph Convolutional Networks for Video Question Answering", "Activity Graph Transformer for Temporal Action Localization", "Ego-Topo: Environment Affordances from Egocentric Video"], "answer_arxiv_id": ["2008.09105", "2101.08540", "2001.04583"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_3782"} +{"question": "Which papers have studied Expectation-Maximization assuming Gaussian noise and model-free clustering in MLR?", "answer": ["Statistical guarantees for the EM algorithm: From population to sample-based analysis", "Ten Steps of EM Suffice for Mixtures of Two Gaussians", "Estimating the Coefficients of a Mixture of Two Linear Regressions by Expectation Maximization", "Global Convergence of the EM Algorithm for Mixtures of Two Component Linear Regression", "EM Converges for a Mixture of Many Linear Regressions", "Estimation, Confidence Intervals, and Large-Scale Hypotheses Testing for High-Dimensional Mixed Linear Regression", "On the Minimax Optimality of the EM Algorithm for Learning Two-Component Mixed Linear Regression"], "answer_arxiv_id": ["1408.2156v1", "1609.00368", "1704.08231", "1810.05752", "1905.12106", "2011.03598", "2006.02601v2"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_3783"} +{"question": "Which work achieved a discretization analysis for the probability flow ODE in KL divergence?", "answer": ["Restoration-Degradation Beyond Linear Diffusions: A Non-Asymptotic Analysis For DDIM-Type Samplers"], "answer_arxiv_id": ["2303.03384"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_3784"} +{"question": "What paper showed a generalization bound for parameterized ODEs for manifold learning?", "answer": ["Fitting an immersed submanifold to data via Sussmann’s orbit theorem"], "answer_arxiv_id": ["2204.01119"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_3785"} +{"question": "Are there any papers where non-parametric depth sampling was used for SIDE?", "answer": ["DepthTransfer: Depth Extraction from Video Using Non-parametric Sampling"], "answer_arxiv_id": ["2001.00987"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_3786"} +{"question": "Which works use the adjoint method for differentiating through physics simulators?", "answer": ["A Differentiable Physics Engine for Deep Learning in Robotics", "Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers"], "answer_arxiv_id": ["1611.01652", "2007.00016"], "source_meta": {"published_time": "20220718"}, "qid": "AutoScholarQuery_train_3787"} +{"question": "What papers considered the offline ILfO setting?", "answer": ["Off-Policy Imitation Learning from Observations", "Versatile Offline Imitation from Observations and Examples via Regularized State-Occupancy Matching", "MobILE: Model-Based Imitation Learning From Observation Alone"], "answer_arxiv_id": ["2102.13185", "2202.02433", "2102.10769"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_3788"} +{"question": "What are some papers discussing about the role of different data augmentations in the self-supervised learning process?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction"], "answer_arxiv_id": ["2002.05709", "2103.03230"], "source_meta": {"published_time": "20220216"}, "qid": "AutoScholarQuery_train_3789"} +{"question": "What studies are there about practical algorithms that utilize weighted regression?", "answer": ["Prioritized Experience Replay", "DisCor: Corrective Feedback in Reinforcement Learning via Distribution Correction", "SUNRISE: A Simple Unified Framework for Ensemble Learning in Deep Reinforcement Learning"], "answer_arxiv_id": ["1511.05952", "2003.07305", "2007.04938"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_3790"} +{"question": "What works have examined self-supervised representation learning through a focus on causality?", "answer": ["Representation Learning via Invariant Causal Mechanisms"], "answer_arxiv_id": ["2010.07922"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_3791"} +{"question": "What studies show the use of deep learning methods such as CNN, GNN, and attentions in matching problems?", "answer": ["SuperGlue: Learning Feature Matching with Graph Neural Networks", "LoFTR: Detector-Free Local Feature Matching with Transformers", "Learning Combinatorial Embedding Networks for Deep Graph Matching", "Deep Graph Matching via Blackbox Differentiation of Combinatorial Solvers", "Neural Graph Matching Network: Learning Lawler's Quadratic Assignment Problem with Extension to Hypergraph and Multiple-graph Matching", "Learning Universe Model for Partial Matching Networks over Multiple Graphs"], "answer_arxiv_id": ["1911.11763", "2104.00680", "1904.00597", "2003.11657", "1911.11308", "2210.10374"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3792"} +{"question": "What are some recent works that propose pretraining techniques for satellite imagery and remote sensing?", "answer": ["SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagery", "Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learning"], "answer_arxiv_id": ["2207.08051", "2212.14532"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_3793"} +{"question": "Which works focus on the popular SSL vision encoders, SimSiam and DINO?", "answer": ["Exploring Simple Siamese Representation Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2011.10566", "2104.14294"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_3794"} +{"question": "Could you provide me some studies conducted on decentralized bilevel optimization?", "answer": ["Decentralized Bilevel Optimization", "Decentralized Gossip-Based Stochastic Bilevel Optimization over Communication Networks"], "answer_arxiv_id": ["2206.05670", "2206.10870"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_3795"} +{"question": "Which studies applied diffusion model for image restoration, and introduce intriguing modifications to the model?", "answer": ["Rethinking Real-world Image Deraining via An Unpaired\n Degradation-Conditioned Diffusion Model", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "Residual Denoising Diffusion Models"], "answer_arxiv_id": ["2301.09430", "2212.00490", "2308.13712"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_3796"} +{"question": "What papers discuss the concept of Parameter-efficient fine-tuning (PEFT)?", "answer": ["What Would Elsa Do? Freezing Layers During Transformer Fine-Tuning", "LoRA: Low-Rank Adaptation of Large Language Models", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Compacter: Efficient Low-Rank Hypercomplex Adapter Layers", "Parameter-Efficient Transfer Learning with Diff Pruning", "BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models"], "answer_arxiv_id": ["1911.03090", "2106.09685", "2101.00190", "2106.04647", "2012.07463", "2106.10199"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_3797"} +{"question": "Which works attempted to enhance the performance of MLP by using label propagation?", "answer": ["Combining Label Propagation and Simple Models out-performs Graph Neural Networks"], "answer_arxiv_id": ["2010.13993"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_3798"} +{"question": "Could you provide me research about DLKT models that utilize the adversarial techniques?", "answer": ["Enhancing Knowledge Tracing via Adversarial Training"], "answer_arxiv_id": ["2108.04430"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_3799"} +{"question": "What articles proposed a method for horizontal federated learning where clients share a part of their local data with the server?", "answer": ["Coded Federated Learning", "Coded Computing for Low-Latency Federated Learning over Wireless Edge Networks", "Stochastic Coded Federated Learning with Convergence and Privacy Guarantees", "Stochastic Coded Federated Learning: Theoretical Analysis and Incentive Mechanism Design"], "answer_arxiv_id": ["2002.09574", "2011.06223", "2201.10092", "2211.04132"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_3800"} +{"question": "What papers reported the failure of empirical Bayes in modern larger neural networks?", "answer": ["Weight Uncertainty in Neural Networks"], "answer_arxiv_id": ["1505.05424"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3801"} +{"question": "Which works discuss other noteworthy fairness definitions including causality-based fairness?", "answer": ["Avoiding Discrimination through Causal Reasoning"], "answer_arxiv_id": ["1706.02744"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_3802"} +{"question": "Could you provide me some studies that solve oversquashing by creating positional embeddings for the nodes or edges?", "answer": ["Attention Is All You Need", "Rethinking Graph Transformers with Spectral Attention", "P", "Rewiring with Positional Encodings for Graph Neural Networks"], "answer_arxiv_id": ["1706.03762", "2106.03893", "0704.0320", "2201.12674"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_3803"} +{"question": "Which papers investigated variants of the full data coverage assumption in offline RL?", "answer": ["Batch Value-function Approximation with Only Realizability", "Minimax Weight and Q-Function Learning for Off-Policy Evaluation"], "answer_arxiv_id": ["2008.04990", "1910.12809"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_3804"} +{"question": "Any works that studied the feature-learning process and generalization in two-layer ReLU networks using noisy 2-xor clustered data?", "answer": ["Random Feature Amplification: Feature Learning and Generalization in Neural Networks"], "answer_arxiv_id": ["2202.07626"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_3805"} +{"question": "Are there any studies on the prediction of a molecule’s energy and forces using hand-crafted representations?", "answer": ["Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields"], "answer_arxiv_id": ["1802.09238"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_3806"} +{"question": "What papers focus on the aspect of Differentiable Computing that involves integrating structured computation graphs into neural networks using Tree-LSTMs?", "answer": ["Language to Logical Form with Neural Attention", "Tree-to-tree Neural Networks for Program Translation"], "answer_arxiv_id": ["1601.01280", "1802.03691"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_3807"} +{"question": "What works are involved in treating text as a confounder within causal inference?", "answer": ["Matching with Text Data: An Experimental Evaluation of Methods for Matching Documents and of Measuring Match Quality", "Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimates"], "answer_arxiv_id": ["1801.00644", "2005.00649"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_3808"} +{"question": "Which studies tackled high-performance VSS by leveraging the temporal continuity of input videos?", "answer": ["Efficient Semantic Video Segmentation with Per-frame Inference", "Mining Relations among Cross-Frame Affinities for Video Semantic\n Segmentation", "Learning Local and Global Temporal Contexts for Video Semantic\n Segmentation", "Local Memory Attention for Fast Video Semantic Segmentation"], "answer_arxiv_id": ["2002.11433", "2207.10436", "2204.03330", "2101.01715"], "source_meta": {"published_time": "20240127"}, "qid": "AutoScholarQuery_train_3809"} +{"question": "Could you provide me some works about human-object interactions in egocentric videos?", "answer": ["Forecasting Human-Object Interaction: Joint Prediction of Motor\n Attention and Actions in First Person Video", "HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object\n Interaction"], "answer_arxiv_id": ["1911.10967", "2203.01577"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_3810"} +{"question": "Which works also explored the concept of direct policy optimization with preference information, similar to the research in question?", "answer": ["Beyond Reward: Offline Preference-guided Policy Optimization", "Direct Preference Optimization: Your Language Model is Secretly a Reward Model"], "answer_arxiv_id": ["2305.16217", "2305.18290"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_3811"} +{"question": "Which reference introduced the GLaD method?", "answer": ["Generalizing Dataset Distillation via Deep Generative Prior"], "answer_arxiv_id": ["2305.01649"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_3812"} +{"question": "What works have shown that restricting the type of learning algorithm can make a tractable problem become intractable in Ising models?", "answer": ["Structure learning of antiferromagnetic Ising models", "Computational Implications of Reducing Data to Sufficient Statistics"], "answer_arxiv_id": ["1412.1443", "1409.3821"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_3813"} +{"question": "Can you tell me about some of the research into length generalization in transformers?", "answer": ["Exploring Length Generalization in Large Language Models", "Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation", "Induced Natural Language Rationales and Interleaved Markup Tokens Enable Extrapolation in Large Language Models", "The EOS Decision and Length Extrapolation", "On the Practical Ability of Recurrent Neural Networks to Recognize Hierarchical Languages"], "answer_arxiv_id": ["2207.04901", "2108.12409", "2208.11445", "2010.07174", "2011.03965"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3814"} +{"question": "What studies have tried to leverage the correlations in the data to propose a more efficient semi-supervised object detector from streaming video?", "answer": ["Label-Efficient Online Continual Object Detection in Streaming Video"], "answer_arxiv_id": ["2206.00309"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_3815"} +{"question": "What studies considered the generalization of learning algorithms with surrogates aiming to optimize Hamming loss and subset accuracy?", "answer": ["Multi-label classification: do Hamming loss and subset accuracy really conflict with each other?"], "answer_arxiv_id": ["2011.07805"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_3816"} +{"question": "Could you provide me some studies that address the issues of Iterative Magnitude Pruning in deep neural networks?", "answer": ["To prune, or not to prune: exploring the efficacy of pruning for model compression", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks"], "answer_arxiv_id": ["1710.01878", "1803.03635"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_3817"} +{"question": "Are there any studies that adopt a two-stage paradigm in their intermediate representations?", "answer": ["PolyTransform: Deep Polygon Transformer for Instance Segmentation", "Deep Snake for Real-Time Instance Segmentation"], "answer_arxiv_id": ["1912.02801", "2001.01629"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_3818"} +{"question": "Which research papers presented methods that use soft prompts for continual learning?", "answer": ["PIVOT: Prompting for Video Continual Learning", "DualPrompt: Complementary Prompting for Rehearsal-free Continual\n Learning", "Learning to Prompt for Continual Learning"], "answer_arxiv_id": ["2212.04842", "2204.04799", "2112.08654"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_3819"} +{"question": "Could you provide me some research papers where instruction tuning applied to help VLMs generate satisfactory answers?", "answer": ["InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning"], "answer_arxiv_id": ["2305.06500", "2305.03726", "2304.14178", "2304.10592", "2304.08485"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_3820"} +{"question": "Which studies adopt retrieval-augmented models in the NLP community to improve various NLP tasks?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Improving language models by retrieving from trillions of tokens", "In-Context Retrieval-Augmented Language Models"], "answer_arxiv_id": ["2005.11401", "2112.04426", "2302.00083"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_3821"} +{"question": "Which works discussed the complexity of architectures in understanding the inner workings of LLMs?", "answer": ["A Primer in BERTology: What we know about how BERT works", "Causal Abstraction for Faithful Model Interpretation"], "answer_arxiv_id": ["2002.12327", "2301.04709"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_3822"} +{"question": "Which papers explore the use of generative models in synthesizing images for DFKT?", "answer": ["Data-Free Learning of Student Networks", "Zero-shot Knowledge Transfer via Adversarial Belief Matching"], "answer_arxiv_id": ["1904.01186", "1905.09768"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_3823"} +{"question": "Can you name some research about diffusion probabilistic models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_3824"} +{"question": "Can you list some research work in the field of language driven robot learning?", "answer": ["CLIPort: What and Where Pathways for Robotic Manipulation", "LISA: Learning Interpretable Skill Abstractions from Language", "Learning Language-Conditioned Robot Behavior from Offline Data and Crowd-Sourced Annotation", "Pre-Trained Language Models for Interactive Decision-Making", "RT-1: Robotics Transformer for Real-World Control at Scale", "CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Tasks", "What Matters in Language Conditioned Robotic Imitation Learning over Unstructured Data", "Language as an Abstraction for Hierarchical Deep Reinforcement Learning", "MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledge", "Semantic Exploration from Language Abstractions and Pretrained Representations", "Language-Conditioned Goal Generation: a New Approach to Language Grounding for RL", "CLIPort: What and Where Pathways for Robotic Manipulation", "Language as an Abstraction for Hierarchical Deep Reinforcement Learning", "Correcting Robot Plans with Natural Language Feedback", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "PaLM-E: An Embodied Multimodal Language Model", "TidyBot: Personalized Robot Assistance with Large Language Models", "Text2Motion: From Natural Language Instructions to Feasible Plans", "Can Foundation Models Perform Zero-Shot Task Specification For Robot Manipulation?", "Code as Policies: Language Model Programs for Embodied Control", "VIMA: General Robot Manipulation with Multimodal Prompts"], "answer_arxiv_id": ["2109.12098", "2203.00054", "2109.01115", "2202.01771", "2212.06817", "2112.03227", "2204.06252", "1906.07343", "2206.08853", "2204.05080", "2006.07043", "2109.12098", "1906.07343", "2204.05186", "2204.01691", "2303.03378v1", "2305.05658", "2303.12153", "2204.11134", "2209.07753", "2210.03094"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_3825"} +{"question": "What are some notable research works aimed at finding and dealing with minority group samples to promote model fairness?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "An investigation of why overparameterization exacerbates spurious correlations", "Just Train Twice: Improving Group Robustness without Training Group Information", "Correct-n-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlations", "Spread Spurious Attribute: Improving Worst-group Accuracy with Spurious Attribute Estimation"], "answer_arxiv_id": ["1911.08731", "2005.04345", "2107.09044", "2203.01517", "2204.02070"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_3826"} +{"question": "What works feature adaptation methods to pivot VFMs towards different tasks?", "answer": ["CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language\n Modeling", "SVL-Adapter: Self-Supervised Adapter for Vision-Language Pretrained\n Models", "Extract Free Dense Labels from CLIP", "Unleashing Text-to-Image Diffusion Models for Visual Perception"], "answer_arxiv_id": ["2110.04544", "2111.03930", "2210.03794", "2112.01071", "2303.02153v1"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_3827"} +{"question": "Which studies are related to the unsupervised fine-tuning of Vision&Language models?", "answer": ["Unsupervised Prompt Learning for Vision-Language Models", "Improving Zero-Shot Models with Label Distribution Priors"], "answer_arxiv_id": ["2204.03649", "2212.00784"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3828"} +{"question": "Which studies fabricated a classifier and conditional generative adversarial networks to solve the dataset bias?", "answer": ["Characterizing Bias in Classifiers using Generative Models", "Fair Attribute Classification through Latent Space De-biasing"], "answer_arxiv_id": ["1906.11891", "2012.01469"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_3829"} +{"question": "Any works focusing on single dynamic objects and generalizing to unknown objects using graph optimization?", "answer": ["BundleSDF: Neural 6-DoF Tracking and 3D Reconstruction of Unknown\n Objects"], "answer_arxiv_id": ["2303.14158"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_3830"} +{"question": "Which paper attempts a unifying framework that encompasses many operator learning architectures and codifies them through a particular definition of neural operators?", "answer": ["Neural Operator: Learning Maps Between Function Spaces"], "answer_arxiv_id": ["2108.08481v6"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_3831"} +{"question": "Which works have initially proposed the idea of transforming various language tasks into a generative problem?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_3832"} +{"question": "Could you provide some examples of works that aimed at post-processing cross-attention to improve semantic alignment?", "answer": ["Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image\n Diffusion Models", "Divide & Bind Your Attention for Improved Generative Semantic Nursing"], "answer_arxiv_id": ["2301.13826", "2307.10864"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_3833"} +{"question": "Which research papers have explored changes in neural network architectures to address spectral bias of conventional ReLU MLPs?", "answer": ["On the Spectral Bias of Neural Networks"], "answer_arxiv_id": ["1806.08734"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_3834"} +{"question": "Could you give me examples of research that use the mechanism of injecting keys and values from attention layers for transferring visual features?", "answer": ["MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing", "InFusion: Inject and Attention Fusion for Multi Concept Zero-Shot\n Text-based Video Editing", "Cross-Image Attention for Zero-Shot Appearance Transfer"], "answer_arxiv_id": ["2304.08465v1", "2308.00135", "2311.03335"], "source_meta": {"published_time": "20240105"}, "qid": "AutoScholarQuery_train_3835"} +{"question": "Which research introduced decoupled-guidance attention control as a method for video editing?", "answer": ["Video-P2P: Video Editing with Cross-attention Control"], "answer_arxiv_id": ["2303.04761"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_3836"} +{"question": "What study adopted dynamic filters in compact keypoint heads to increase accuracy and speed?", "answer": ["FCPose: Fully Convolutional Multi-Person Pose Estimation with Dynamic Instance-Aware Convolutions"], "answer_arxiv_id": ["2105.14185v1"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_3837"} +{"question": "Can you mention the study that proposed using the TV prior as a regularization term along with the gradient matching objective?", "answer": ["Inverting Gradients - How easy is it to break privacy in federated learning?"], "answer_arxiv_id": ["2003.14053"], "source_meta": {"published_time": "20220912"}, "qid": "AutoScholarQuery_train_3838"} +{"question": "In what research papers have efforts been made to make ViTs more lightweight and mobile-friendly?", "answer": ["Separable Self-attention for Mobile Vision Transformers", "MobileViTv3: Mobile-Friendly Vision Transformer with Simple and\n Effective Fusion of Local, Global and Input Features", "EfficientFormer: Vision Transformers at MobileNet Speed", "FastViT: A Fast Hybrid Vision Transformer using Structural\n Reparameterization"], "answer_arxiv_id": ["2206.02680", "2209.15159", "2206.01191", "2303.14189"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_3839"} +{"question": "What papers look into the tendency of augmentations to sacrifice performance on some classes in exchange for gains on others or cause models to misrepresent uncertainty?", "answer": ["The Effects of Regularization and Data Augmentation are Class Dependent"], "answer_arxiv_id": ["2204.03632"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_3840"} +{"question": "What papers detail node-embedding-based models in the one-shot category?", "answer": ["Variational Graph Auto-Encoders", "Graphite: Iterative Generative Modeling of Graphs"], "answer_arxiv_id": ["1611.07308", "1803.10459"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_3841"} +{"question": "Which models have been developed for universal audio generation modeling?", "answer": ["AudioGen: Textually Guided Audio Generation", "AudioLDM: Text-to-Audio Generation with Latent Diffusion Models", "Text-to-Audio Generation using Instruction-Tuned LLM and Latent\n Diffusion Model", "Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion\n Models"], "answer_arxiv_id": ["2209.15352", "2301.12503", "2304.13731", "2301.12661"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_3842"} +{"question": "What papers develop human pose estimation with visual methods?", "answer": ["VIBE: Video Inference for Human Body Pose and Shape Estimation", "Bottom-Up Human Pose Estimation Via Disentangled Keypoint Regression", "3D Human Pose Estimation via Intuitive Physics", "Learning Human Mesh Recovery in 3D Scenes"], "answer_arxiv_id": ["1912.05656", "2104.02300", "2303.18246", "2306.03847"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_3843"} +{"question": "Could you provide examples of studies that explored disentanglement for encoding sentence semantics?", "answer": ["Learning Disentangled Representations for Natural Language Definitions"], "answer_arxiv_id": ["2210.02898"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_3844"} +{"question": "Any works about unlearning that propose methods with a two-stage pipeline: ‘neutralization’ / forgetting followed by a ‘retraining’ / restoring?", "answer": ["Federated Unlearning with Knowledge Distillation"], "answer_arxiv_id": ["2201.09441"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_3845"} +{"question": "Which research employ methods using a convolutional projection?", "answer": ["Multiscale Vision Transformers", "MViTv2: Improved Multiscale Vision Transformers for Classification and\n Detection", "Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction\n without Convolutions", "CvT: Introducing Convolutions to Vision Transformers"], "answer_arxiv_id": ["2104.11227", "2112.01526", "2102.12122", "2103.15808"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_3846"} +{"question": "Any works about predicting behaviors by conditioning on desired returns in RCSL?", "answer": ["Reward-Conditioned Policies", "RvS: What is Essential for Offline RL via Supervised Learning?", "Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["1912.13465", "2112.10751", "2106.01345"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_3847"} +{"question": "Which work was focused on OCR-free text-oriented multimodal understanding?", "answer": ["Pix2Struct: Screenshot Parsing as Pretraining for Visual Language\n Understanding"], "answer_arxiv_id": ["2210.03347"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_3848"} +{"question": "What studies proposed stopping underperforming model evaluations to save computational resources?", "answer": ["Non-stochastic Best Arm Identification and Hyperparameter Optimization", "Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization"], "answer_arxiv_id": ["1502.07943v1", "1603.06560"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_3849"} +{"question": "What studies proposed works for faster learning of student models from a pre-trained diffusion teacher?", "answer": ["Knowledge Distillation in Iterative Generative Models for Improved\n Sampling Speed", "Consistency Models", "Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2101.02388", "2303.01469", "2202.00512"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_3850"} +{"question": "Which research works employed multi-modal fusion from concatenation operation in fully supervising the RIS?", "answer": ["Segmentation from Natural Language Expressions", "Recurrent Multimodal Interaction for Referring Image Segmentation"], "answer_arxiv_id": ["1603.06180", "1703.07939"], "source_meta": {"published_time": "20240418"}, "qid": "AutoScholarQuery_train_3851"} +{"question": "Can you provide studies that have focused on policy regret against generic m-memory bounded adversaries?", "answer": ["Online Bandit Learning against an Adaptive Adversary: from Regret to Policy Regret", "Online Learning with Switching Costs and Other Adaptive Adversaries", "Policy Regret in Repeated Games"], "answer_arxiv_id": ["1206.6400", "1302.4387", "1811.04127v2"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_3852"} +{"question": "What studies represent sets of entities as geometric shapes or probabilistic distribution in the context of complex logical query answering?", "answer": ["Embedding Logical Queries on Knowledge Graphs", "Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings", "ConE: Cone Embeddings for Multi-Hop Reasoning over Knowledge Graphs", "Self-Supervised Hyperboloid Representations from Logical Queries over Knowledge Graphs", "Beta Embeddings for Multi-Hop Logical Reasoning in Knowledge Graphs"], "answer_arxiv_id": ["1806.01445", "2002.05969", "2110.13715", "2012.13023", "2010.11465"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_3853"} +{"question": "Could you tell me any examples of the research provided algorithms and empirical heuristics for making IL sample efficient by modeling the experts?", "answer": ["Imitation Learning by Estimating Expertise of Demonstrators", "Learning from Imperfect Demonstrations via Adversarial Confidence Transfer", "Inverse Preference Learning: Preference-based RL without a Reward Function", "Behavior Retrieval: Few-Shot Imitation Learning by Querying Unlabeled Datasets", "Masked Imitation Learning: Discovering Environment-Invariant Modalities in Multimodal Demonstrations"], "answer_arxiv_id": ["2202.01288", "2202.02967", "2305.15363", "2304.08742", "2209.07682"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_3854"} +{"question": "Which study incorporates a backward procedure in PCL to improve the consistency of the Koopman operator?", "answer": ["Forecasting Sequential Data Using Consistent Koopman Autoencoders"], "answer_arxiv_id": ["2003.02236"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3855"} +{"question": "Could you provide me with some works that demonstrate other classical optimization methods for smooth optimization?", "answer": ["Faster Rates for Convex-Concave Games", "Acceleration through Optimistic No-Regret Dynamics", "No-Regret Dynamics in the Fenchel Game: A Unified Framework for Algorithmic Convex Optimization"], "answer_arxiv_id": ["1805.06792v1", "1807.10455", "2111.11309"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_3856"} +{"question": "Which works estimate 3D motion from a pair or a sequence of point clouds?", "answer": ["FlowNet3D: Learning Scene Flow in 3D Point Clouds", "PointPWC-Net: A Coarse-to-Fine Network for Supervised and\n Self-Supervised Scene Flow Estimation on 3D Point Clouds", "HPLFlowNet: Hierarchical Permutohedral Lattice FlowNet for Scene Flow\n Estimation on Large-scale Point Clouds", "CaDeX: Learning Canonical Deformation Coordinate Space for Dynamic\n Surface Representation via Neural Homeomorphism", "Neural Deformation Graphs for Globally-consistent Non-rigid\n Reconstruction"], "answer_arxiv_id": ["1806.01411", "1911.12408", "1906.05332", "2203.16529", "2012.01451"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_3857"} +{"question": "In what works is the learning strategies like classifier free guidance aware distillation used?", "answer": ["On Distillation of Guided Diffusion Models", "SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two\n Seconds"], "answer_arxiv_id": ["2210.03142v3", "2306.00980"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_3858"} +{"question": "What papers discussed architectures such as transformers, graph neural networks, etc., that are claimed to have some degree of compositional generalization?", "answer": ["Attention Is All You Need", "Combinatorial Optimization and Reasoning with Graph Neural Networks"], "answer_arxiv_id": ["1706.03762", "2102.09544"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_3859"} +{"question": "What studies attempted to train CNFs by employing kinetic energy as a regularization term?", "answer": ["How to Train Your Neural ODE: the World of Jacobian and Kinetic Regularization", "OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport"], "answer_arxiv_id": ["2002.02798", "2006.00104"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_3860"} +{"question": "In what papers the researcher extracted vision and language features using convolution networks and recurrent neural networks in referring image segmentation?", "answer": ["Recurrent Multimodal Interaction for Referring Image Segmentation", "Modeling Context Between Objects for Referring Expression Understanding"], "answer_arxiv_id": ["1703.07939", "1608.00525"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_3861"} +{"question": "Which papers introduce localization approaches using an overhead image such as a map, a satellite patch, or a floorplan?", "answer": ["OrienterNet: Visual Localization in 2D Public Maps with Neural Matching", "You Are Here: Geolocation by Embedding Maps and Images", "VIGOR: Cross-View Image Geo-localization beyond One-to-one Retrieval", "Visual Cross-View Metric Localization with Dense Uncertainty Estimates", "LaLaLoc: Latent Layout Localisation in Dynamic, Unvisited Environments", "LASER: LAtent SpacE Rendering for 2D Visual Localization"], "answer_arxiv_id": ["2304.02009", "1911.08797", "2011.12172", "2208.08519", "2104.09169", "2204.00157"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3862"} +{"question": "Which study uses questions as a bridge to connect the appearance and motion features to find the answers?", "answer": ["Bridge to Answer: Structure-aware Graph Interaction Network for Video Question Answering"], "answer_arxiv_id": ["2104.14085"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_3863"} +{"question": "Which works contemplate a setting where the learner has access to some logging policy rather than a fixed set of logged data?", "answer": ["Agnostic System Identification for Model-Based Reinforcement Learning", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning"], "answer_arxiv_id": ["1203.1007", "2106.04895"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_3864"} +{"question": "What are the studies about Large Language Models developed for general generation tasks?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Effective Long-Context Scaling of Foundation Models", "PaLM: Scaling Language Modeling with Pathways", "The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora\n with Web Data, and Web Data Only", "Mistral 7B", "Baichuan 2: Open Large-scale Language Models", "Qwen Technical Report"], "answer_arxiv_id": ["2302.13971", "2309.16039", "2204.02311", "2306.01116", "2310.06825", "2309.10305", "2309.16609v1"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_3865"} +{"question": "What studies have discussed the degree bias inherent in Graph Neural Networks?", "answer": ["Cold Brew: Distilling Graph Node Representations with Incomplete or Missing Neighborhoods"], "answer_arxiv_id": ["2111.04840"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_3866"} +{"question": "Could you give me some examples of studies where contrastive learning has been used extensively to align representations across different modalities for downstream use in multimodal tasks?", "answer": ["UniXcoder: Unified Cross-Modal Pre-training for Code Representation", "CodeBERT: A Pre-Trained Model for Programming and Natural Languages", "Zero-Shot Text-to-Image Generation", "CURL: Contrastive Unsupervised Representations for Reinforcement Learning"], "answer_arxiv_id": ["2203.03850", "2002.08155", "2102.12092", "2004.04136"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_3867"} +{"question": "Could you provide me some studies about adversarial attacks on deep neural networks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Boosting Adversarial Attacks with Momentum", "Towards Evaluating the Robustness of Neural Networks", "Obfuscated Gradients Give a False Sense of Security: Circumventing\n Defenses to Adversarial Examples"], "answer_arxiv_id": ["1412.6572", "1706.06083", "1710.06081", "1608.04644", "1802.00420"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_3868"} +{"question": "What studies have traded interpretability for faster rendering by using light field scene representations?", "answer": ["Light Field Networks: Neural Scene Representations with\n Single-Evaluation Rendering", "Light Field Neural Rendering", "Learning to Render Novel Views from Wide-Baseline Stereo Pairs", "Generalizable Patch-Based Neural Rendering", "Scene Representation Transformer: Geometry-Free Novel View Synthesis\n Through Set-Latent Scene Representations"], "answer_arxiv_id": ["2106.02634", "2112.09687", "2304.08463", "2207.10662", "2111.13152"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_3869"} +{"question": "Which works developed methods that optimize a new word embedding token for each concept in fine-tuning-based TIE?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_3870"} +{"question": "Any recent works bridging offline and online RL?", "answer": ["Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning", "Hybrid RL: Using Both Offline and Online Data Can Make RL Efficient", "Leveraging Offline Data in Online Reinforcement Learning"], "answer_arxiv_id": ["2103.12021v2", "2106.04895", "2210.06718", "2211.04974"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_3871"} +{"question": "What papers have focused on interference in statistical learning and inference with equilibrium effects?", "answer": ["Estimating Average Causal Effects Under General Interference, with Application to a Social Network Experiment", "Exact P-values for Network Interference", "Average Direct and Indirect Causal Effects under Interference", "Random Graph Asymptotics for Treatment Effect Estimation under Network Interference"], "answer_arxiv_id": ["1305.6156", "1506.02084v1", "2104.03802", "2007.13302"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_3872"} +{"question": "Which method uses a rare-token as the pseudo-word and further fine-tunes the entire pre-trained diffusion model for better similarity?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_3873"} +{"question": "What research showed the lack of word order sensitivity in large language models?", "answer": ["Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Little"], "answer_arxiv_id": ["2104.06644"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_3874"} +{"question": "Which papers focus on denoising the features with diffusion models in order to minimize the representation gap between teacher and student?", "answer": ["Knowledge Diffusion for Distillation"], "answer_arxiv_id": ["2305.15712"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_3875"} +{"question": "What papers discuss using limited precedent data for rehearsal-based approaches in language models?", "answer": ["Fine-tuned Language Models are Continual Learners", "CITB: A Benchmark for Continual Instruction Tuning"], "answer_arxiv_id": ["2205.12393", "2310.14510"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_3876"} +{"question": "What is an example of a method that allow real-time rendering of neural radiance field using grid-based neural fields?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_3877"} +{"question": "Any works that offered the first algorithm for training robust tree ensembles on a pool of adversarial examples updated on every iteration of boosting?", "answer": ["Evasion and Hardening of Tree Ensemble Classifiers"], "answer_arxiv_id": ["1509.07892"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_3878"} +{"question": "Which papers propose backdoor attacks on Diffusion Models?", "answer": ["TrojDiff: Trojan Attacks on Diffusion Models with Diverse Targets", "How to Backdoor Diffusion Models?"], "answer_arxiv_id": ["2303.05762", "2212.05400"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_3879"} +{"question": "Which works discussed the robustness of the watermarking scheme used in machine-generated text detection against various attacks?", "answer": ["A Watermark for Large Language Models", "On the Reliability of Watermarks for Large Language Models", "Robust Distortion-free Watermarks for Language Models", "Provable Robust Watermarking for AI-Generated Text", "SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text\n Generation"], "answer_arxiv_id": ["2301.10226", "2306.04634", "2307.15593", "2306.17439", "2310.03991"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_3880"} +{"question": "What are the works related to Neural Machine Translation (NMT)?", "answer": ["Improving Neural Machine Translation Models with Monolingual Data", "Google's Multilingual Neural Machine Translation System: Enabling\n Zero-Shot Translation", "Multilingual Denoising Pre-training for Neural Machine Translation", "Beyond English-Centric Multilingual Machine Translation", "No Language Left Behind: Scaling Human-Centered Machine Translation"], "answer_arxiv_id": ["1511.06709", "1611.04558", "2001.08210", "2010.11125", "2207.04672v3"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_3881"} +{"question": "Which research integrated time-conditioning into Neural Radiance Fields using a set of compact latent codes?", "answer": ["Neural 3D Video Synthesis from Multi-view Video"], "answer_arxiv_id": ["2103.02597"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_3882"} +{"question": "Could you provide me works about LLMs being used as reference-free automated evaluators?", "answer": ["LLM-Mini-CEX: Automatic Evaluation of Large Language Model for\n Diagnostic Conversation", "LLM-Eval: Unified Multi-Dimensional Automatic Evaluation for Open-Domain\n Conversations with Large Language Models", "GPTScore: Evaluate as You Desire"], "answer_arxiv_id": ["2308.07635", "2305.13711", "2302.04166"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_train_3883"} +{"question": "Which papers introduce the method of model stitching?", "answer": ["Understanding image representations by measuring their equivariance and\n equivalence", "Revisiting Model Stitching to Compare Neural Representations"], "answer_arxiv_id": ["1411.5908", "2106.07682"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_3884"} +{"question": "Has any work been done on enhancing the generalizability of super-resolution models using limited degradation data?", "answer": ["Masked Image Training for Generalizable Deep Image Denoising"], "answer_arxiv_id": ["2303.13132"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3885"} +{"question": "Could you provide me some works about maintaining the performance of dense networks at an ultra-high sparse ratio?", "answer": ["Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization"], "answer_arxiv_id": ["1902.05967"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_3886"} +{"question": "Could you mention the papers that study kernels induced by tree ensembles through the Neural Tangent Kernel perspective?", "answer": ["A Neural Tangent Kernel Perspective of Infinite Tree Ensembles"], "answer_arxiv_id": ["2109.04983"], "source_meta": {"published_time": "20220611"}, "qid": "AutoScholarQuery_train_3887"} +{"question": "Could you provide me some works introduced additional modules or training objectives for better segmentation?", "answer": ["Reliability Does Matter: An End-to-End Weakly Supervised Semantic\n Segmentation Approach", "Learning Affinity from Attention: End-to-End Weakly-Supervised Semantic\n Segmentation with Transformers", "Self Correspondence Distillation for End-to-End Weakly-Supervised\n Semantic Segmentation"], "answer_arxiv_id": ["1911.08039", "2203.02664", "2302.13765"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_3888"} +{"question": "Which paper developed the fine-tuning method named Textual Inversion for Stable Diffusion?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_3889"} +{"question": "Which works discussed the use of scene graphs from the perspective of autonomous vehicles?", "answer": ["roadscene2vec: A Tool for Extracting and Embedding Road Scene-Graphs", "Explainable Action Prediction through Self-Supervision on Scene Graphs"], "answer_arxiv_id": ["2109.01183", "2302.03477"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_3890"} +{"question": "Which studies dealt with the representation of surfaces and adaptation to dynamic scenes in NeRF?", "answer": ["UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "Nerfies: Deformable Neural Radiance Fields", "D-NeRF: Neural Radiance Fields for Dynamic Scenes"], "answer_arxiv_id": ["2104.10078", "2106.12052", "2011.12948", "2011.13961"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_3891"} +{"question": "Are there any works where a latent world model is learned on top of VAE features and a latent value function is built?", "answer": ["Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Model"], "answer_arxiv_id": ["1907.00953"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3892"} +{"question": "Can you provide some studies that focused on object grounding benchmarks like YouCook2-BoundingBox?", "answer": ["Look at What I'm Doing: Self-Supervised Spatial Grounding of Narrations\n in Instructional Videos"], "answer_arxiv_id": ["2110.10596"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_3893"} +{"question": "What studies have focused on data augmentation and the performance of using generative models?", "answer": ["Fixing Data Augmentation to Improve Adversarial Robustness"], "answer_arxiv_id": ["2103.01946"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_3894"} +{"question": "Could you reference the studies that focus on optimizing strategies to reduce task conflict, such as loss balancing and gradient balancing in MTL?", "answer": ["Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics", "End-to-End Multi-Task Learning with Attention", "GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks", "Gradient Surgery for Multi-Task Learning", "Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout", "Conflict-Averse Gradient Descent for Multi-task Learning", "RotoGrad: Gradient Homogenization in Multitask Learning", "Multi-Task Learning as a Bargaining Game"], "answer_arxiv_id": ["1705.07115", "1803.10704", "1711.02257", "2001.06782", "2010.06808", "2110.14048", "2103.02631", "2202.01017"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_3895"} +{"question": "What research used recurrent attentive model to spatially fuse image and language features?", "answer": ["Language-Based Image Editing with Recurrent Attentive Models"], "answer_arxiv_id": ["1711.06288"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_3896"} +{"question": "Which papers propose various methods for adaptation without target data?", "answer": ["Generalizing Across Domains via Cross-Gradient Training", "Learning to Generalize: Meta-Learning for Domain Generalization", "Domain Generalization via Model-Agnostic Learning of Semantic Features", "Robustness properties of Facebook’s ResNeXt WSL models", "Fast AutoAugment", "AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "On Feature Normalization and Data Augmentation", "The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization", "Improving Out-of-Distribution Robustness via Selective Augmentation"], "answer_arxiv_id": ["1804.10745", "1710.03463", "1910.13580", "1907.07640", "1905.00397", "1912.02781", "2002.11102", "2006.16241", "2201.00299"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_3897"} +{"question": "Which works apply diffusion model in text generation domains?", "answer": ["Diffusion-LM Improves Controllable Text Generation", "DiffusER: Discrete Diffusion via Edit-based Reconstruction", "DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models"], "answer_arxiv_id": ["2205.14217", "2210.16886v1", "2210.08933"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_3898"} +{"question": "Could you provide me some studies about adaptive algorithms which need information about the problem parameters be provided when the objective function is nonconvex about one variable?", "answer": ["AdaGDA: Faster Adaptive Gradient Descent Ascent Methods for Minimax Optimization", "A Novel Convergence Analysis for Algorithms of the Adam Family"], "answer_arxiv_id": ["2106.16101v6", "2112.03459v1"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_3899"} +{"question": "What studies propose strategies involving fine-tuning and post-processing techniques to enhance the capabilities of pre-trained Large Language Models (LLMs)?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models", "Mistral 7B", "GPT-4 Technical Report", "Llemma: An Open Language Model For Mathematics"], "answer_arxiv_id": ["2307.09288", "2310.06825", "2303.08774", "2310.10631"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_3900"} +{"question": "Which papers made advancements in estimating and tracking the segmentation of the articulated parts of future home-assistant robots?", "answer": ["Deep Part Induction from Articulated Object Pairs", "Act the Part: Learning Interaction Strategies for Articulated Object Part Discovery", "OPD: Single-view 3D Openable Part Detection"], "answer_arxiv_id": ["1809.07417", "2105.01047", "2203.16421"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_3901"} +{"question": "Which paper is about a corpus composed of synthetically generated single-step deductive proofs?", "answer": ["Critical Thinking for Language Models"], "answer_arxiv_id": ["2009.07185"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_3902"} +{"question": "Which papers studied the application of the CLIP model on vision understanding tasks?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "ActionCLIP: A New Paradigm for Video Action Recognition", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting"], "answer_arxiv_id": ["2104.13921", "2109.08472", "2112.01518"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_train_3903"} +{"question": "Any previous research provided theoretical analysis regarding the intriguing alignment properties observed in class of word vectorizers?", "answer": ["A Latent Variable Model Approach to PMI-based Word Embeddings"], "answer_arxiv_id": ["1502.03520v8"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_3904"} +{"question": "What previous work has employed transformers for video retrieval?", "answer": ["VIOLET : End-to-End Video-Language Transformers with Masked Visual-token\n Modeling", "Revealing Single Frame Bias for Video-and-Language Learning", "OmniVL:One Foundation Model for Image-Language and Video-Language Tasks", "VindLU: A Recipe for Effective Video-and-Language Pretraining", "Unmasked Teacher: Towards Training-Efficient Video Foundation Models", "All in One: Exploring Unified Video-Language Pre-training"], "answer_arxiv_id": ["2111.12681", "2206.03428", "2209.07526", "2212.05051", "2303.16058", "2203.07303"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_3905"} +{"question": "What studies utilized stochastic node/edge masking strategies for generating contrastive views?", "answer": ["Graph Contrastive Learning with Augmentations"], "answer_arxiv_id": ["2010.13902"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_3906"} +{"question": "Which works used deformation fields to model the target 3D shape of objects while preserving their geometric details?", "answer": ["Deep Deformable 3D Caricatures with Learned Shape Control", "3DN: 3D Deformation Network", "Deformed Implicit Field: Modeling 3D Shapes with Learned Dense\n Correspondence"], "answer_arxiv_id": ["2207.14593", "1903.03322", "2011.13650"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_3907"} +{"question": "Could you provide me with any works that investigate the use of the 'doubling trick' to make an algorithm parameter-free?", "answer": ["No-Regret Algorithms for Unconstrained Online Convex Optimization"], "answer_arxiv_id": ["1211.2260"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3908"} +{"question": "What research utilizes explicit sparse voxel representation for direct object composition but encounters challenges with storage requirements?", "answer": ["Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2112.05131"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_3909"} +{"question": "In the field of Test-Time Adaptation (TTA), which papers address the utilization of a self-supervised proxy task?", "answer": ["Test-Time Training with Masked Autoencoders", "Test-Time Training with Self-Supervision for Generalization under Distribution Shifts"], "answer_arxiv_id": ["2209.07522", "1909.13231"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_3910"} +{"question": "Where was self-training in speech recognition that uses Language Models (LMs) discussed?", "answer": ["Self-Training for End-to-End Speech Recognition"], "answer_arxiv_id": ["1909.09116"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_3911"} +{"question": "Which references discussed the semi-supervised learning method that assumes unlabeled and labeled data are from the same source?", "answer": ["Semi-Supervised Graph Imbalanced Regression"], "answer_arxiv_id": ["2305.12087"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_3912"} +{"question": "Which papers discuss the dynamic NeRFs that capture scenes by separately encoding each frame?", "answer": ["INV: Towards Streaming Incremental Neural Videos"], "answer_arxiv_id": ["2302.01532"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3913"} +{"question": "Which work proposed the least-to-most prompting as a form of problem decomposition?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language Models"], "answer_arxiv_id": ["2205.10625"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_3914"} +{"question": "What studies demonstrate that policies trained with the basic assumption of Object Oriented RL often achieve better generalization ability?", "answer": ["Object-Category Aware Reinforcement Learning"], "answer_arxiv_id": ["2210.07802"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_3915"} +{"question": "Could you provide me some research about applying Rademacher complexity to two-layer neural networks?", "answer": ["Adversarial Learning Guarantees for Linear Hypotheses and Neural Networks"], "answer_arxiv_id": ["2004.13617"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_3916"} +{"question": "Which studies are there on using stylometric signals to detect AI-based tweet generation?", "answer": ["Stylometric Detection of AI-Generated Text in Twitter Timelines"], "answer_arxiv_id": ["2303.03697"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_3917"} +{"question": "Are there studies that develop memory models for visual object tracking (VOT)?", "answer": ["Learning Dynamic Memory Networks for Object Tracking", "MAVOT: Memory-Augmented Video Object Tracking", "Adaptive Correlation Filters with Long-Term and Short-Term Memory for\n Object Tracking"], "answer_arxiv_id": ["1803.07268", "1711.09414", "1707.02309"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_3918"} +{"question": "Are there any papers about learning general facial representations for facial videos in the same field?", "answer": ["MARLIN: Masked Autoencoder for facial video Representation LearnINg"], "answer_arxiv_id": ["2211.06627"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_3919"} +{"question": "Could you provide me some papers that discuss learning neural implicit functions from 3D point clouds?", "answer": ["Learning Consistency-Aware Unsigned Distance Functions Progressively from Raw Point Clouds", "Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces", "Implicit Geometric Regularization for Learning Shapes", "SAL: Sign Agnostic Learning of Shapes from Raw Data", "Sign-Agnostic Implicit Learning of Surface Self-Similarities for Shape Modeling and Reconstruction from Raw Point Clouds", "SALD: Sign Agnostic Learning with Derivatives", "Latent Partition Implicit with Surface Codes for 3D Representation"], "answer_arxiv_id": ["2210.02757", "2011.13495", "2002.10099", "1911.10414", "2012.07498", "2006.05400", "2207.08631"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_3920"} +{"question": "Which research papers have significantly progressed in reasoning on one-hop relational data?", "answer": ["RotatE: Knowledge Graph Embedding by Relational Rotation in Complex\n Space"], "answer_arxiv_id": ["1902.10197"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_3921"} +{"question": "Which works have used iterative inference in the development of flow or diffusion-based generative models in computer vision?", "answer": ["NICE: Non-linear Independent Components Estimation", "Variational Inference with Normalizing Flows", "Density estimation using Real NVP", "Glow: Generative Flow with Invertible 1x1 Convolutions", "PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "3D Shape Generation and Completion through Point-Voxel Diffusion"], "answer_arxiv_id": ["1410.8516", "1505.05770", "1605.08803", "1807.03039v2", "1906.12320", "1907.05600", "2011.13456", "2006.11239", "2010.02502", "2103.01458", "2104.03670"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_3922"} +{"question": "Which research works have studied the determination of orientation in a super-structure given a super-structure?", "answer": ["Parameterized Complexity Results for Exact Bayesian Network Structure Learning", "High-dimensional learning of linear causal networks via inverse covariance estimation", "Reliable Causal Discovery with Improved Exact Search and Weaker Assumptions"], "answer_arxiv_id": ["1402.0558v1", "1311.3492", "2201.05666"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_3923"} +{"question": "Could you provide me a study about the feasibility of using ChatGPT in psychiatry?", "answer": ["LLM-empowered Chatbots for Psychiatrist and Patient Simulation:\n Application and Evaluation"], "answer_arxiv_id": ["2305.13614"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_3924"} +{"question": "Which papers study differentially private versions of ADMM for centralized settings?", "answer": ["Differentially Private ADMM Algorithms for Machine Learning"], "answer_arxiv_id": ["2011.00164"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_3925"} +{"question": "Could you provide me some studies based on differentiable simulators that assume that gradients of simulation outcomes w. r. t. actions are explicitly given?", "answer": ["Do Differentiable Simulators Give Better Policy Gradients?", "Bundled Gradients through Contact via Randomized Smoothing", "Accelerated Policy Learning with Parallel Differentiable Simulation"], "answer_arxiv_id": ["2202.00817", "2109.05143", "2204.07137"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_3926"} +{"question": "What works have studied the implicit bias to margin maximisation in parameter space for homogenous networks?", "answer": ["Gradient Descent Maximizes the Margin of Homogeneous Neural Networks", "Directional convergence and alignment in deep learning"], "answer_arxiv_id": ["1906.05890", "2006.06657"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_3927"} +{"question": "Which studies indicate that transformer models struggle across various domains like algorithmic and commonsense reasoning?", "answer": ["Neural Algorithmic Reasoning", "Is ChatGPT a General-Purpose Natural Language Processing Task Solver?", "Humans in Humans Out: On GPT Converging Toward Common Sense in both Success and Failure"], "answer_arxiv_id": ["2105.02761", "2302.06476", "2303.17276"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3928"} +{"question": "Where can I find information about the PrivatePGM (PGM) algorithm and its variations?", "answer": ["Graphical-model based estimation and inference for differential privacy", "AIM: An Adaptive and Iterative Mechanism for Differentially Private Synthetic Data"], "answer_arxiv_id": ["1901.09136", "2201.12677"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_3929"} +{"question": "Which paper proposed the first non-autoregressive S2ST model?", "answer": ["TranSpeech: Speech-to-Speech Translation With Bilateral Perturbation"], "answer_arxiv_id": ["2205.12523"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_3930"} +{"question": "Which works describe mitigating robust overfitting by smoothing labels and weights?", "answer": ["Exploring Memorization in Adversarial Training", "Adversarial Weight Perturbation Helps Robust Generalization"], "answer_arxiv_id": ["2106.01606", "2004.05884"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_3931"} +{"question": "Which works focused on object motion generation?", "answer": ["Learning Multi-Object Dynamics with Compositional Neural Radiance Fields", "Flexible Neural Representation for Physics Prediction", "Predicting the Physical Dynamics of Unseen 3D Objects", "Object-Oriented Dynamics Predictor"], "answer_arxiv_id": ["2202.11855", "1806.08047", "2001.06291", "1806.07371"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_3932"} +{"question": "Which papers revealed that post-processing methods could lead to higher performance degradation compared to in-processing methods?", "answer": ["Retiring Adult: New Datasets for Fair Machine Learning"], "answer_arxiv_id": ["2108.04884"], "source_meta": {"published_time": "20220916"}, "qid": "AutoScholarQuery_train_3933"} +{"question": "What are some examples of research that work on contrastive language-image pre-training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3934"} +{"question": "What are the recent works that brought back large kernel convnets with appealing performance?", "answer": ["Patches Are All You Need?", "A ConvNet for the 2020s", "Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs", "More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using\n Sparsity"], "answer_arxiv_id": ["2201.09792", "2201.03545", "2203.06717", "2207.03620"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_3935"} +{"question": "Are there any works about extracting more MDP models with low-rank structures by proposing the bilinear class?", "answer": ["Bilinear Classes: A Structural Framework for Provable Generalization in RL", "Minimax Regret Bounds for Reinforcement Learning", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches", "Provably Efficient Reinforcement Learning with Linear Function Approximation", "Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles", "Provably Efficient Exploration in Policy Optimization", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes"], "answer_arxiv_id": ["2103.10897", "1703.05449", "1811.08540", "1907.05388", "1910.10597", "1912.05830", "2012.08507"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_3936"} +{"question": "What studies revealed that GPT-3 can work as few-shot or zero-shot learners?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_3937"} +{"question": "What research proposed using explanations to enhance few-shot learning in large language models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "STaR: Bootstrapping Reasoning With Reasoning"], "answer_arxiv_id": ["2201.11903", "2203.14465"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_3938"} +{"question": "Which works use the masked modeling paradigm in the field of computer vision?", "answer": ["data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: a Simple Framework for Masked Image Modeling"], "answer_arxiv_id": ["2202.03555", "2111.06377", "2111.09886"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_3939"} +{"question": "Which works proposed learning correlated equilibrium and coarse correlated equilibrium in Markov games?", "answer": ["A Sharp Analysis of Model-based Reinforcement Learning with Self-Play", "Robustness and sample complexity of model-based MARL for general-sum Markov games", "When Can We Learn General-Sum Markov Games with a Large Number of Players Sample-Efficiently?", "V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL", "On Improving Model-Free Algorithms for Decentralized Multi-Agent Reinforcement Learning", "Provably Efficient Reinforcement Learning in Decentralized General-Sum Markov Games"], "answer_arxiv_id": ["2010.01604", "2110.02355", "2110.04184", "2110.14555", "2110.05707v2", "2110.05682v3"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_3940"} +{"question": "Could you provide me papers examining the necessity of sinusoidal encoding in 3D scene representations?", "answer": ["Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains", "PINs: Progressive Implicit Networks for Multi-Scale Neural Representations"], "answer_arxiv_id": ["2006.10739", "2202.04713"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_3941"} +{"question": "What studies found drawbacks in FedDecorr, namely its dependency on large batch size and its tendency to deactivate lots of neuron parameters?", "answer": ["VNE: An Effective Method for Improving Deep Representation by\n Manipulating Eigenvalue Distribution"], "answer_arxiv_id": ["2304.01434"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_3942"} +{"question": "What are the studies that have integrated biomimetic design into machine vision models?", "answer": ["Peripheral Vision Transformer", "Focal Modulation Networks", "Focal Self-attention for Local-Global Interactions in Vision\n Transformers"], "answer_arxiv_id": ["2206.06801", "2203.11926", "2107.00641"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3943"} +{"question": "Can you list works that proposed a mechanism to select intervention targets for efficient causal structure learning?", "answer": ["Learning Neural Causal Models with Active Interventions"], "answer_arxiv_id": ["2109.02429"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_3944"} +{"question": "Could you provide me with the research papers that discuss existing methods that adopt time-dependent discriminators in diffusion models?", "answer": ["Tackling the Generative Learning Trilemma with Denoising Diffusion GANs", "Diffusion-GAN: Training GANs with Diffusion", "Refining Generative Process with Discriminator Guidance in Score-based Diffusion Models"], "answer_arxiv_id": ["2112.07804", "2206.02262", "2211.17091v4"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_3945"} +{"question": "Can you provide papers that explored the representation cost in shallow nonlinear networks or deep networks with very specific structure?", "answer": ["Breaking the Curse of Dimensionality with Convex Neural Networks", "The Role of Linear Layers in Nonlinear Interpolating Networks", "Training invariances and the low-rank phenomenon: beyond linear networks"], "answer_arxiv_id": ["1412.8690", "2202.00856", "2201.11968"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_3946"} +{"question": "What papers demonstrate revealing sensitive text using a large language model in the NLP domain?", "answer": ["The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks", "Extracting Training Data from Large Language Models"], "answer_arxiv_id": ["1802.08232", "2012.07805"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_3947"} +{"question": "Who has proposed the use of auxiliary variables, different from the labels, to solve the OOD problem in a single training domain?", "answer": ["Causally motivated Shortcut Removal Using Auxiliary Labels", "Out-of-distribution Generalization in the Presence of Nuisance-Induced Spurious Correlations"], "answer_arxiv_id": ["2105.06422", "2107.00520"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_3948"} +{"question": "What papers suggested the use of expanded policy for offline-to-online RL?", "answer": ["Policy Expansion for Bridging Offline-to-Online Reinforcement Learning"], "answer_arxiv_id": ["2302.00935"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_3949"} +{"question": "Could you state the research that explores the use of depth maps for data augmentation in the context of semantic segmentation?", "answer": ["Three Ways to Improve Semantic Segmentation with Self-Supervised Depth\n Estimation"], "answer_arxiv_id": ["2012.10782"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_3950"} +{"question": "What research conducted various analyses of different algorithms under the (L0,L1) smoothness condition?", "answer": ["Improved Analysis of Clipping Algorithms for Non-convex Optimization", "Beyond Uniform Smoothness: A Stopped Analysis of Adaptive SGD", "Variance-reduced Clipping for Non-convex Optimization", "Robustness to Unbounded Smoothness of Generalized SignSGD"], "answer_arxiv_id": ["2010.02519", "2302.06570", "2303.00883", "2208.11195"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_3951"} +{"question": "Any works about multi-modality encoder-decoder models adapted for sequence-to-sequence learning?", "answer": ["PaLI: A Jointly-Scaled Multilingual Language-Image Model", "Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks"], "answer_arxiv_id": ["2209.06794", "2206.08916"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_3952"} +{"question": "Could you provide some references about Lagrangian model-based agent method?", "answer": ["Constrained Policy Optimization via Bayesian World Models"], "answer_arxiv_id": ["2201.09802"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_3953"} +{"question": "What works stated that layer freezing technique can achieve almost linear acceleration according to the computation FLOPs reduction?", "answer": ["Layer Freezing & Data Sieving: Missing Pieces of a Generic Framework for Sparse Training"], "answer_arxiv_id": ["2209.11204"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_3954"} +{"question": "Are there any works regarding LVLMs that leveraged high-quality multimodal data to facilitate the training process?", "answer": ["LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2304.15010", "2305.03726", "2304.14178"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_3955"} +{"question": "What works tackle the issue of Out-Of-Distribution (OOD) on graph?", "answer": ["Debiased Graph Neural Networks with Agnostic Label Selection Bias", "Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training Data", "Handling Distribution Shifts on Graphs: An Invariance Perspective", "Discovering Invariant Rationales for Graph Neural Networks", "Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanism", "Graph Invariant Learning with Subgraph Co-mixup for Out-Of-Distribution Generalization", "GOOD: A Graph Out-of-Distribution Benchmark"], "answer_arxiv_id": ["2201.07708", "2108.01099", "2202.02466", "2201.12872", "2201.12987", "2312.10988", "2206.08452"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_3956"} +{"question": "Which works studied non-convex optimization in different settings?", "answer": ["Arxiv"], "answer_arxiv_id": ["2004.12380"], "source_meta": {"published_time": "20230803"}, "qid": "AutoScholarQuery_train_3957"} +{"question": "Could you provide the researcher who regressed correspondences directly without feature matching?", "answer": ["REGTR: End-to-end Point Cloud Correspondences with Transformers"], "answer_arxiv_id": ["2203.14517"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_3958"} +{"question": "Are there any studies that employed overcomplete representations for better detailing and noise robustness in semantic segmentation tasks?", "answer": ["KiU-Net: Towards Accurate Segmentation of Biomedical Images using\n Over-complete Representations"], "answer_arxiv_id": ["2006.04878"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_3959"} +{"question": "What studies are about the pre-training of useful skill representations from offline trajectories?", "answer": ["OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning", "Accelerating Reinforcement Learning with Learned Skill Priors", "Parrot: Data-Driven Behavioral Priors for Reinforcement Learning", "TRAIL: Near-Optimal Imitation Learning with Suboptimal Data"], "answer_arxiv_id": ["2010.13611", "2010.11944", "2011.10024", "2110.14770"], "source_meta": {"published_time": "20220819"}, "qid": "AutoScholarQuery_train_3960"} +{"question": "Could you provide me some works about the template matching methods in instance-level approaches?", "answer": ["Implicit 3D Orientation Learning for 6D Object Detection from RGB Images", "OSOP: A Multi-Stage One Shot Object Pose Estimation Framework", "OVE6D: Object Viewpoint Encoding for Depth-based 6D Object Pose\n Estimation", "Going Further with Point Pair Features", "6D Pose Estimation using an Improved Method based on Point Pair Features", "Templates for 3D Object Pose Estimation Revisited: Generalization to New Objects and Robustness to Occlusions"], "answer_arxiv_id": ["1902.01275", "2203.15533", "2203.01072", "1711.04061", "1802.08516v1", "2203.17234v1"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_3961"} +{"question": "Could you name some studies about pruning method in model compression?", "answer": ["Recent Advances on Neural Network Pruning at Initialization"], "answer_arxiv_id": ["2103.06460"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_3962"} +{"question": "How were large pre-trained models like CLIP and Stable Diffusion applied in human reconstruction according to existing works?", "answer": ["One-shot Implicit Animatable Avatars with Model-based Priors", "TeCH: Text-guided Reconstruction of Lifelike Clothed Humans"], "answer_arxiv_id": ["2212.02469", "2308.08545"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_3963"} +{"question": "Which studies propose to combine machine learning with modern interactive theorem provers?", "answer": ["Learning to Prove Theorems via Interacting with Proof Assistants", "TacticToe: Learning to Prove with Tactics"], "answer_arxiv_id": ["1905.09381", "1804.00596"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_3964"} +{"question": "Could you provide me some researches which explored the use of multi-grid features in INRs?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_3965"} +{"question": "What earlier studies have explored the optimality of AGD optimization vs. averaging the losses?", "answer": ["Non-convex Optimization for Machine Learning", "Improved optimization strategies for deep Multi-Task Networks"], "answer_arxiv_id": ["1712.07897", "2109.11678"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_3966"} +{"question": "Can you name the works concerning neural generative models for group-invariant distributions?", "answer": ["Scalable Normalizing Flows for Permutation Invariant Densities", "E(n) Equivariant Normalizing Flows", "Equivariant Flows: sampling configurations for multi-body systems with symmetric energies", "Graph Normalizing Flows", "Equivariant Hamiltonian Flows"], "answer_arxiv_id": ["2010.03242", "2105.09016", "1910.00753", "1905.13177v1", "1909.13739"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_3967"} +{"question": "What works have explored achieving specific planning or control tasks or policies using methods like model-based, imitation learning and reinforcement learning?", "answer": ["Learning Language-Conditioned Robot Behavior from Offline Data and\n Crowd-Sourced Annotation", "Learning with Latent Language", "Correcting Robot Plans with Natural Language Feedback", "Language Conditioned Imitation Learning over Unstructured Data", "CLIPort: What and Where Pathways for Robotic Manipulation", "Language as an Abstraction for Hierarchical Deep Reinforcement Learning", "PixL2R: Guiding Reinforcement Learning Using Natural Language by Mapping\n Pixels to Rewards", "Mapping Instructions and Visual Observations to Actions with\n Reinforcement Learning"], "answer_arxiv_id": ["2109.01115", "1711.00482", "2204.05186", "2005.07648", "2109.12098", "1906.07343", "2007.15543", "1704.08795"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_3968"} +{"question": "Is there any work that generates human-object interaction using a hierarchical generation pipeline?", "answer": ["Hierarchical Generation of Human-Object Interactions with Diffusion\n Probabilistic Models"], "answer_arxiv_id": ["2310.02242"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_3969"} +{"question": "Which studies proposed a two-stream network for global alignment in cross-modal retrieval?", "answer": ["VSE++: Improving Visual-Semantic Embeddings with Hard Negatives", "COTS: Collaborative Two-Stream Vision-Language Pre-Training Model for\n Cross-Modal Retrieval"], "answer_arxiv_id": ["1707.05612", "2204.07441"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_3970"} +{"question": "Which works are based on uniformly random modifications of graph elements focusing on graph data augmentation methods?", "answer": ["Data Augmentation for Deep Graph Learning: A Survey", "Deep Graph Infomax", "DropEdge: Towards Deep Graph Convolutional Networks on Node Classification", "Graph Random Neural Networks for Semi-Supervised Learning on Graphs", "Graph Contrastive Learning with Augmentations", "Deep Graph Contrastive Representation Learning"], "answer_arxiv_id": ["2202.08235", "1809.10341", "1907.10903", "2005.11079", "2010.13902", "2006.04131"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_3971"} +{"question": "What works address the limitation of diversity and quality of 3D data in 3D diffusion models?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "Shap-E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2212.08751", "2305.02463"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_3972"} +{"question": "Could you provide me studies that used a Manhattan world for robust simultaneous localization and mapping methods?", "answer": ["ManhattanSLAM: Robust Planar Tracking and Mapping Leveraging Mixture of\n Manhattan Frames"], "answer_arxiv_id": ["2103.15068"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_3973"} +{"question": "Which paper found that Vision Transformers are more robust to spurious correlations compared to ConvNets when using larger models and more training data?", "answer": ["Are Vision Transformers Robust to Spurious Correlations?"], "answer_arxiv_id": ["2203.09125"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_3974"} +{"question": "Could you give me some references on empirical analyses of Curriculum Learning?", "answer": ["A Survey on Curriculum Learning", "Curriculum Learning: A Survey"], "answer_arxiv_id": ["2010.13166", "2101.10382"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_3975"} +{"question": "Could you provide some works that used the transformations of neural networks for solving Sudoku puzzles?", "answer": ["Assessing SATNet’s Ability to Solve the Symbol Grounding Problem", "Techniques for Symbol Grounding with SATNet"], "answer_arxiv_id": ["2312.11522", "2106.11072"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_3976"} +{"question": "What works utilize the strategy of encoder-decoder models for training large-scale neural language models on source code?", "answer": ["Unified Pre-training for Program Understanding and Generation", "Competition-Level Code Generation with AlphaCode", "DOBF: A Deobfuscation Pre-Training Objective for Programming Languages", "CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generation"], "answer_arxiv_id": ["2103.06333", "2203.07814v1", "2102.07492", "2109.00859"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_3977"} +{"question": "Could you list down research papers relevant to 'coordination graph' and 'coordinated exploration' methods in multi-agent reinforcement learning?", "answer": ["Deep Coordination Graphs", "Context-Aware Sparse Deep Coordination Graphs", "Self-Organized Polynomial-Time Coordination Graphs", "MAVEN: Multi-Agent Variational Exploration", "Coordinated Exploration via Intrinsic Rewards for Multi-Agent Reinforcement Learning", "Episodic Multi-Agent Reinforcement Learning with Curiosity-Driven Exploration"], "answer_arxiv_id": ["1910.00091", "2106.02886", "2112.03547", "1910.07483", "1905.12127", "2111.11032"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_3978"} +{"question": "Could you provide a study where the model was jointly trained with language-only and multi-modal instruction data to enhance both multi-modal and language-only performance?", "answer": ["mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2304.14178"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_3979"} +{"question": "What works have been conducted on neuron analysis in vision models?", "answer": ["Understanding the Role of Individual Units in a Deep Neural Network", "Compositional Explanations of Neurons"], "answer_arxiv_id": ["2009.05041", "2006.14032"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_3980"} +{"question": "What studies have followed the architecture of HMR for HPS regression?", "answer": ["End-to-end Recovery of Human Shape and Pose"], "answer_arxiv_id": ["1712.06584"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_3981"} +{"question": "Which papers discuss methods for zero-shot point cloud understanding by transferring VLM representation to 3D?", "answer": ["PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "Rethinking Network Design and Local Geometry in Point Cloud: A Simple\n Residual MLP Framework", "ConceptFusion: Open-set Multimodal 3D Mapping", "PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning", "OpenScene: 3D Scene Understanding with Open Vocabularies"], "answer_arxiv_id": ["1706.02413", "2202.07123", "2302.07241", "2211.11682", "2211.15654"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_3982"} +{"question": "What papers discuss methods for tackling the challenge of deciphering correspondences among co-salient objects?", "answer": ["Adaptive Graph Convolutional Network with Attention Graph Clustering for\n Co-saliency Detection", "Gradient-Induced Co-Saliency Detection"], "answer_arxiv_id": ["2003.06167", "2004.13364"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_3983"} +{"question": "Which study designed the SmoothOut framework to improve generalization?", "answer": ["SmoothOut: Smoothing Out Sharp Minima to Improve Generalization in Deep\n Learning"], "answer_arxiv_id": ["1805.07898"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_3984"} +{"question": "Could you name any papers that applied manually written few-shot demonstrations for CoT reasoning in language models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_3985"} +{"question": "Which papers have integrated logical statements with neural networks for interpretable logical reasoning?", "answer": ["VQA-LOL: Visual Question Answering under the Lens of Logic", "Multimodal Logical Inference System for Visual-Textual Entailment"], "answer_arxiv_id": ["2002.08325", "1906.03952"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_3986"} +{"question": "Which research compared trajectories of Bayesian neural networks that had access to different amounts of data to human developmental trajectories?", "answer": ["Emulating Human Developmental Stages with Bayesian Neural Networks"], "answer_arxiv_id": ["1902.07579"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_3987"} +{"question": "What research papers describe the advancements in one-stage object detection algorithms?", "answer": ["FCOS: Fully Convolutional One-Stage Object Detection", "SSD: Single Shot MultiBox Detector", "YOLOX: Exceeding YOLO Series in 2021", "Objects as Points", "CenterNet: Keypoint Triplets for Object Detection"], "answer_arxiv_id": ["1904.01355", "1512.02325", "2107.08430", "1904.07850", "1904.08189"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_3988"} +{"question": "Which research papers have investigated the infinite-depth limit of finite-width architectures, specifically ResNets?", "answer": ["Neural Ordinary Differential Equations", "Scaling Properties of Deep Residual Networks", "Scaling ResNets in the Large-depth Regime", "On the infinite-depth limit of finite-width neural networks"], "answer_arxiv_id": ["1806.07366", "2105.12245", "2206.06929", "2210.00688"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_3989"} +{"question": "Which studies used FCNs, Non-local Module, and Hough Voting for learned correspondence weight in pose estimation?", "answer": ["Deep Global Registration", "PointDSC: Robust Point Cloud Registration using Deep Spatial Consistency", "Deep Hough Voting for Robust Global Registration"], "answer_arxiv_id": ["2004.11540", "2103.05465", "2109.04310"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_3990"} +{"question": "What paper designed a method for LLMs that involves copying reference text tokens to the decoder?", "answer": ["Inference with Reference: Lossless Acceleration of Large Language Models"], "answer_arxiv_id": ["2304.04487"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_3991"} +{"question": "Which studies try to tackle the question of what information to store in a stream of observations in RL with incomplete state information?", "answer": ["Continual Learning Through Synaptic Intelligence", "Overcoming catastrophic forgetting in neural networks", "Progress & Compress: A scalable framework for continual learning"], "answer_arxiv_id": ["1703.04200", "1612.00796", "1805.06370"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_3992"} +{"question": "What works are related to game-theoretic approaches to measure the importance of features in feature attribution?", "answer": ["A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1705.07874"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_3993"} +{"question": "Which paper induced class coherence through a group-sparsity regularizer?", "answer": ["Optimal Transport for Domain Adaptation"], "answer_arxiv_id": ["1507.00504"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_3994"} +{"question": "Could you name the works that propose language models performing gradient descent when learning task in-context?", "answer": ["Transformers Learn In-Context by Gradient Descent"], "answer_arxiv_id": ["2212.07677"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_3995"} +{"question": "What works introduced Generative Adversarial Networks (GANs)?", "answer": ["Generative Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN", "StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets"], "answer_arxiv_id": ["1406.2661", "1812.04948", "1912.04958", "2202.00273"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_3996"} +{"question": "Which work introduces 'The Stack', a large dataset of licensed source code in multiple programming languages?", "answer": ["The Stack: 3 TB of permissively licensed source code"], "answer_arxiv_id": ["2211.15533"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_3997"} +{"question": "Could you provide me some papers of methods for inverse rendering with different constraints?", "answer": ["Deep Reflectance Volumes: Relightable Reconstructions from Multi-View\n Photometric Images", "Deep 3D Capture: Geometry and Reflectance from Sparse Multi-View Images", "Unified Shape and SVBRDF Recovery using Differentiable Monte Carlo\n Rendering"], "answer_arxiv_id": ["2007.09892", "2003.12642", "2103.15208"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_3998"} +{"question": "Which work proposed a training-free method to calculate the posterior PCs in Gaussian denoising?", "answer": ["On the Posterior Distribution in Denoising: Application to Uncertainty Quantification"], "answer_arxiv_id": ["2309.13598"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_3999"} +{"question": "Which studies are about learning the node ordering for autoregressive graph generation?", "answer": ["Order Matters: Probabilistic Modeling of Node Sequence for Graph Generation"], "answer_arxiv_id": ["2106.06189"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_4000"} +{"question": "What are some research papers that have begun to learn disentangled shape representations for potential geometry manipulation in the context of 3D scene editing?", "answer": ["Extracting Triangular 3D Models, Materials, and Lighting From Images", "Editing Conditional Radiance Fields", "CodeNeRF: Disentangled Neural Radiance Fields for Object Categories", "Decomposing 3D Scenes into Objects via Unsupervised Volume Segmentation"], "answer_arxiv_id": ["2111.12503", "2105.06466", "2109.01750", "2104.01148"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_4001"} +{"question": "What is K-Net and where can I find more about it?", "answer": ["K-Net: Towards Unified Image Segmentation"], "answer_arxiv_id": ["2106.14855"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_4002"} +{"question": "What are the papers about the feature propagation method, whereby features in preceding frames are reused to accelerate computation?", "answer": ["Clockwork Convnets for Video Semantic Segmentation", "Temporally Distributed Networks for Fast Video Semantic Segmentation"], "answer_arxiv_id": ["1608.03609", "2004.01800"], "source_meta": {"published_time": "20240127"}, "qid": "AutoScholarQuery_train_4003"} +{"question": "What works exist on image-to-image translation methods that aim to preserve the background while editing the object part only?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation", "Blended Diffusion for Text-driven Editing of Natural Images", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image Translation", "Zero-shot Image-to-Image Translation"], "answer_arxiv_id": ["2108.01073", "2110.02711", "2111.14818", "2210.09276", "2208.01626", "2211.12572", "2302.03027"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4004"} +{"question": "Are there works which improved upon the Assembly101 dataset with a better annotation pipeline?", "answer": ["AssemblyHands: Towards Egocentric Activity Understanding via 3D Hand Pose Estimation"], "answer_arxiv_id": ["2304.12301"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_4005"} +{"question": "What studies propose models that are neither fully autoregressive in nature, nor do they make any Gaussianity assumptions, like MAFs?", "answer": ["Masked Autoregressive Flow for Density Estimation"], "answer_arxiv_id": ["1705.07057"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_4006"} +{"question": "Which research proposed and analyzed the DANE method for distributed optimization under the assumption of strong convexity?", "answer": ["Communication Efficient Distributed Optimization using an Approximate Newton-type Method"], "answer_arxiv_id": ["1312.7853"], "source_meta": {"published_time": "20220906"}, "qid": "AutoScholarQuery_train_4007"} +{"question": "What research attempts to synthesize low-resolution to high-resolution image training pairs as part of the blind image super-resolution process?", "answer": ["Frequency Separation for Real-World Super-Resolution", "Toward Real-World Single Image Super-Resolution: A New Benchmark and A\n New Model", "Component Divide-and-Conquer for Real-World Image Super-Resolution"], "answer_arxiv_id": ["1911.07850", "1904.00523", "2008.01928"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_4008"} +{"question": "What research papers established CLIP-like models for mapping images with corresponding language descriptions?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "FILIP: Fine-grained Interactive Language-Image Pre-Training", "Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm"], "answer_arxiv_id": ["2102.05918", "2111.07783", "2110.05208"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_4009"} +{"question": "Could you tell me about some research papers that focus on managing the variance of IS in long-horizon settings?", "answer": ["Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation", "Towards Optimal Off-Policy Evaluation for Reinforcement Learning with Marginalized Importance Sampling", "DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections"], "answer_arxiv_id": ["1810.12429", "1906.03393", "1906.04733"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_4010"} +{"question": "Could you provide me some studies about graph neural networks?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Distilling Knowledge from Graph Convolutional Networks", "Meta-Aggregator: Learning to Aggregate for 1-bit Graph Neural Networks", "Factorizable Graph Convolutional Networks", "Learning Graph Neural Networks for Image Style Transfer", "SPAGAN: Shortest Path Graph Attention Network"], "answer_arxiv_id": ["1609.02907", "2003.10477", "2109.12872", "2010.05421", "2207.11681", "2101.03464"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_4011"} +{"question": "Can you name the research papers that showed OMWU and Optimistic Gradient Descent Ascent (OGDA) converging to a Nash equilibrium (NE) with a global sublinear and a local linear convergence rate in NFGs?", "answer": ["Linear Last-iterate Convergence in Constrained Saddle-point Optimization"], "answer_arxiv_id": ["2006.09517"], "source_meta": {"published_time": "20220619"}, "qid": "AutoScholarQuery_train_4012"} +{"question": "What study is about the robustness of kernel bandits with misspecification?", "answer": ["High-Dimensional Experimental Design and Kernel Bandits"], "answer_arxiv_id": ["2105.05806"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_4013"} +{"question": "Which works proposed the use of eigenvector distances in maintaining positional information?", "answer": ["Equivariant and Stable Positional Encoding for More Powerful Graph Neural Networks"], "answer_arxiv_id": ["2203.00199"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4014"} +{"question": "What studies discuss the use of polarization for reflection removal?", "answer": ["Polarization Guided Specular Reflection Separation", "Polarized Reflection Removal with Perfect Alignment in the Wild"], "answer_arxiv_id": ["2103.11652", "2003.12789"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_4015"} +{"question": "What research studies have been conducted on using diffusion models for image synthesis?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Probabilistic Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Implicit Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2105.05233", "2006.11239", "2206.00364", "1503.03585", "2010.02502", "2011.13456"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_4016"} +{"question": "What papers made advancements in network architectures in the field of 3D reconstruction?", "answer": ["Mesh R-CNN", "pixelNeRF: Neural Radiance Fields from One or Few Images"], "answer_arxiv_id": ["1906.02739", "2012.02190"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_4017"} +{"question": "What studies are about reducing the number of training iterations with models that initially share parameters?", "answer": ["Speeding up Deep Model Training by Sharing Weights and Then Unsharing", "M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining", "ALBERT: A Lite BERT for Self-supervised Learning of Language Representations", "Universal Transformers"], "answer_arxiv_id": ["2110.03848", "2110.03888", "1909.11942", "1807.03819"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_4018"} +{"question": "Could you give me examples of research that introduced additional distillation loss functions in open-vocabulary object detection?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation", "Learning to Prompt for Open-Vocabulary Object Detection with\n Vision-Language Model", "Aligning Bag of Regions for Open-Vocabulary Object Detection"], "answer_arxiv_id": ["2104.13921", "2203.14940", "2302.13996"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_4019"} +{"question": "Which works applied diffusion models to audio synthesis, protein modeling, and graph modeling?", "answer": ["DiffWave: A Versatile Diffusion Model for Audio Synthesis", "Protein Structure and Sequence Generation with Equivariant Denoising Diffusion Probabilistic Models", "Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations"], "answer_arxiv_id": ["2009.09761v3", "2205.15019", "2202.02514"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_4020"} +{"question": "Can you provide the works that use density estimation to disentangle epistemic and aleatoric uncertainty?", "answer": ["Deep Deterministic Uncertainty: A Simple Baseline"], "answer_arxiv_id": ["2102.11582"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4021"} +{"question": "What was the approach of STGCNN in deep-learning based motion forecasting models?", "answer": ["Social-STGCNN: A Social Spatio-Temporal Graph Convolutional Neural Network for Human Trajectory Prediction"], "answer_arxiv_id": ["2002.11927"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_4022"} +{"question": "Could you provide me with studies about image data rules defined on meta-data?", "answer": ["Webly Supervised Learning of Convolutional Networks"], "answer_arxiv_id": ["1505.01554"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_4023"} +{"question": "Which study includes Multi positives, Anchor-free, and Decoupled Head into the YOLO-series model?", "answer": ["YOLOX: Exceeding YOLO Series in 2021"], "answer_arxiv_id": ["2107.08430"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_4024"} +{"question": "What literature has worked with diffusion processes for score-based flows?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_4025"} +{"question": "Could you provide some relevant work on large language models?", "answer": ["Language Models are Few-Shot Learners", "GPT-4 Technical Report", "Training language models to follow instructions with human feedback", "OPT: Open Pre-trained Transformer Language Models", "LLaMA: Open and Efficient Foundation Language Models", "GLM-130B: An Open Bilingual Pre-trained Model"], "answer_arxiv_id": ["2005.14165", "2303.08774", "2203.02155", "2205.01068", "2302.13971", "2210.02414"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_4026"} +{"question": "What works have used the information bottleneck to extract well generalizing data representations in neural networks?", "answer": ["Opening the black box of Deep Neural Networks via Information"], "answer_arxiv_id": ["1703.00810"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_4027"} +{"question": "What studies present work on finding Nash equilibria of matrix games?", "answer": ["Linear Last-iterate Convergence in Constrained Saddle-point Optimization"], "answer_arxiv_id": ["2006.09517"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_4028"} +{"question": "Can you point me to the studies that used voxels to depict geometry in 3D representation?", "answer": ["High-Resolution Shape Completion Using Deep Neural Networks for Global\n Structure and Local Geometry Inference", "3D-R2N2: A Unified Approach for Single and Multi-view 3D Object\n Reconstruction", "3D Shape Induction from 2D Views of Multiple Objects", "OctNet: Learning Deep 3D Representations at High Resolutions"], "answer_arxiv_id": ["1709.07599", "1604.00449", "1612.05872", "1611.05009"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_4029"} +{"question": "Which papers discussed hard parameter sharing methods in designing multi-task architectures?", "answer": ["UberNet : Training a ‘Universal’ Convolutional Neural Network for Low-, Mid-, and High-Level Vision using Diverse Datasets and Limited Memory", "Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics", "Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Generalist Convolution Kernels", "Many Task Learning with Task Routing", "Multi-Task Learning for Dense Prediction Tasks: A Survey"], "answer_arxiv_id": ["1609.02132", "1705.07115", "1908.09597", "1903.12117", "2004.13379"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_4030"} +{"question": "What research proposes solutions for global explanations of black-box models?", "answer": ["GLocalX - From Local to Global Explanations of Black Box AI Models", "XGNN: Towards Model-Level Explanations of Graph Neural Networks"], "answer_arxiv_id": ["2101.07685", "2006.02587"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_4031"} +{"question": "What studies aim to learn conservative values such as fitted Q-iteration using conservative update, conservative Q-learning (CQL), critic regularization, and subtracting penalties?", "answer": ["Provably Good Batch Reinforcement Learning Without Great Exploration", "Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Fisher Divergence Critic Regularization", "Offline Reinforcement Learning as Anti-Exploration"], "answer_arxiv_id": ["2007.08202", "2006.04779", "2103.08050", "2106.06431"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_4032"} +{"question": "Which papers are focused on modifying the online counterparts of offline RL algorithms by adding regularization?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "MOReL: Model-Based Offline Reinforcement Learning", "Off-Policy Deep Reinforcement Learning without Exploration", "COMBO: Conservative Offline Model-Based Policy Optimization"], "answer_arxiv_id": ["2006.04779", "2005.05951", "1812.02900", "2102.08363"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_4033"} +{"question": "Could you list some studies that observe part models in human recognition?", "answer": ["Articulated Pose Estimation by a Graphical Model with Image Dependent Pairwise Relations", "Actions and Attributes from Wholes and Parts", "Devil in the Details: Towards Accurate Single and Multiple Human Parsing"], "answer_arxiv_id": ["1407.3399", "1412.2604", "1809.05996"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_4034"} +{"question": "What studies expanded the CLIP 2D paradigm for 3D point clouds?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP"], "answer_arxiv_id": ["2112.02413"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_4035"} +{"question": "Which papers have studied marginal contribution-based methods in the context of feature attribution problems?", "answer": ["A Unified Approach to Interpreting Model Predictions", "Explaining by Removing: A Unified Framework for Model Explanation", "WeightedSHAP: analyzing and improving Shapley based feature attributions"], "answer_arxiv_id": ["1705.07874", "2011.14878", "2209.13429"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_4036"} +{"question": "Could you provide some examples of works that suggest point-based rendering via splatting as a potential solution for animatable avatar modeling?", "answer": ["Differentiable Surface Splatting for Point-based Geometry Processing", "Neural Point-Based Graphics", "Pulsar: Efficient Sphere-based Neural Rendering", "Point-Based Neural Rendering with Per-View Optimization", "ADOP: Approximate Differentiable One-Pixel Point Rendering"], "answer_arxiv_id": ["1906.04173v3", "1906.08240", "2004.07484", "2109.02369", "2110.06635"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4037"} +{"question": "Which works utilized open-source C/CPP repositories for function-level vulnerability detection?", "answer": ["Automated Vulnerability Detection in Source Code Using Deep\n Representation Learning"], "answer_arxiv_id": ["1807.04320"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_4038"} +{"question": "What studies did work on extending MetaFormers to focus on the potential of recurrent-free models?", "answer": ["MetaFormer Is Actually What You Need for Vision"], "answer_arxiv_id": ["2111.11418"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_4039"} +{"question": "What research provides information on chain of reasoning explanations in QA systems?", "answer": ["Learning to Explain: Datasets and Models for Identifying Valid Reasoning Chains in Multihop Question-Answering"], "answer_arxiv_id": ["2010.03274"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_4040"} +{"question": "What are the studies about large language models (LLMs) that have demonstrated transferability to various downstream tasks?", "answer": ["Language Models are Few-Shot Learners", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2307.09288", "2204.02311"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_4041"} +{"question": "Which studies have analyzed the application of parallelization techniques like Picard iteration in the context of sampling from log-concave and determinantal distributions?", "answer": ["The Randomized Midpoint Method for Log-Concave Sampling"], "answer_arxiv_id": ["1909.05503"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4042"} +{"question": "What efforts have been made to use diffusion models for Out-Of-Distribution detection?", "answer": ["Multiscale Score Matching for Out-of-Distribution Detection", "Unsupervised Out-of-Distribution Detection with Diffusion Inpainting"], "answer_arxiv_id": ["2010.13132", "2302.10326"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4043"} +{"question": "Which work initially introduced the Structured State Space Sequence (S4) Model?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces"], "answer_arxiv_id": ["2111.00396"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_4044"} +{"question": "Which research paper analyzes KD for vision transformers?", "answer": ["Co-advise: Cross Inductive Bias Distillation"], "answer_arxiv_id": ["2106.12378"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_4045"} +{"question": "Which papers focus on improving the language modeling capability or a particular category of downstream task in semi-parametric language models?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models", "Training Language Models with Memory Augmentation", "Improving Neural Language Models with a Continuous Cache", "Pointer Sentinel Mixture Models", "Episodic Memory in Lifelong Language Learning", "Augmenting Transformers with KNN-Based Composite Memory for Dialog", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["1911.00172", "2205.12674", "1612.04426", "1609.07843", "1906.01076", "2004.12744", "2005.11401"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_4046"} +{"question": "Which papers fix the issue that deep learning models give poor uncertainty estimations and suffer from overconfidence?", "answer": ["Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images", "Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problem", "Fixing Overconfidence in Dynamic Neural Networks", "Periodic Activation Functions Induce Stationarity"], "answer_arxiv_id": ["1412.1897", "1812.05720", "2302.06359v4", "2110.13572v2"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_4047"} +{"question": "Which papers studied RL with function approximation?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches", "Bilinear Classes: A Structural Framework for Provable Generalization in RL", "Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms", "The Statistical Complexity of Interactive Decision Making", "Model-based RL with Optimistic Posterior Sampling: Structural Conditions and Sample Complexity", "Unified Algorithms for RL with Decision-Estimation Coefficients: No-Regret, PAC, and Reward-Free Learning"], "answer_arxiv_id": ["1610.09512v2", "1811.08540", "2103.10897", "2102.00815", "2112.13487v3", "2206.07659", "2209.11745"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_4048"} +{"question": "Which work proposed the universal framework that is suitable for various teacher-student pairs?", "answer": ["UniDistill: A Universal Cross-Modality Knowledge Distillation Framework\n for 3D Object Detection in Bird's-Eye View"], "answer_arxiv_id": ["2303.15083"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_4049"} +{"question": "Could you provide me some examples of recent autonomous driving datasets?", "answer": ["The Cityscapes Dataset for Semantic Urban Scene Understanding", "UrbanLoco: A Full Sensor Suite Dataset for Mapping and Localization in\n Urban Scenes", "nuScenes: A multimodal dataset for autonomous driving", "Scalability in Perception for Autonomous Driving: Waymo Open Dataset", "A*3D Dataset: Towards Autonomous Driving in Challenging Environments", "The H3D Dataset for Full-Surround 3D Multi-Object Detection and Tracking\n in Crowded Urban Scenes"], "answer_arxiv_id": ["1604.01685v2", "1912.09513", "1903.11027", "1912.04838", "1909.07541", "1903.01568"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_4050"} +{"question": "Can you provide some studies that are at the intersection of irl and bc?", "answer": ["Apprenticeship Learning using Inverse Reinforcement Learning and Gradient Methods", "Adversarial Soft Advantage Fitting: Imitation Learning without Policy Optimization", "Augmenting GAIL with BC for sample efficient imitation learning", "Discriminator-Weighted Offline Imitation Learning from Suboptimal Demonstrations"], "answer_arxiv_id": ["1206.5264v1", "2006.13258", "2001.07798", "2207.10050"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4051"} +{"question": "Could you provide me some studies about the Score Distillation Sampling procedure and its applications?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "High-Resolution Image Synthesis with Latent Diffusion Models", "TeCH: Text-guided Reconstruction of Lifelike Clothed Humans", "TADA! Text to Animatable Digital Avatars", "Text-Guided Generation and Editing of Compositional 3D Avatars", "DreamHuman: Animatable 3D Avatars from Text", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models"], "answer_arxiv_id": ["2209.14988", "2112.10752", "2308.08545", "2308.10899", "2309.07125", "2306.09329", "2304.00916"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_4052"} +{"question": "What other research paper applied the GosInE Algorithm in the context unaware scenario and partially context aware scenario of multi-armed bandits?", "answer": ["The Gossiping Insert-Eliminate Algorithm for Multi-Agent Bandits"], "answer_arxiv_id": ["2001.05452v4"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4053"} +{"question": "Which papers expound on the utilization of Variational autoencoders (VAEs) in modeling neural data?", "answer": ["Auto-Encoding Variational Bayes", "Learning identifiable and interpretable latent models of high-dimensional neural activity using pi-VAE", "Variational Autoencoders and Nonlinear ICA: A Unifying Framework", "Drop, Swap, and Generate: A Self-Supervised Approach for Generating Neural Activity"], "answer_arxiv_id": ["1312.6114", "2011.04798", "1907.04809", "2111.02338"], "source_meta": {"published_time": "20230812"}, "qid": "AutoScholarQuery_train_4054"} +{"question": "Can you give examples of research worn on open-vocabulary detection using the COCO dataset?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation", "RegionCLIP: Region-based Language-Image Pretraining", "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection"], "answer_arxiv_id": ["2104.13921", "2112.09106", "2303.05499"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_4055"} +{"question": "What work proposed a teacher-student strategy for transformers in visual tasks?", "answer": ["Training data-efficient image transformers & distillation through attention"], "answer_arxiv_id": ["2012.12877"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_4056"} +{"question": "Which work introduced techniques for augmenting LiDAR datasets?", "answer": ["Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks"], "answer_arxiv_id": ["1609.06666"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_4057"} +{"question": "Can you provide studies that employ genetic programming for symbolic regression?", "answer": ["Learning concise representations for regression by evolving networks of trees"], "answer_arxiv_id": ["1807.00981"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_4058"} +{"question": "What works discuss the effectiveness of CN in applications like style transfer, semantic image synthesis and denoising?", "answer": ["A Learned Representation For Artistic Style", "Exploring the structure of a real-time, arbitrary neural artistic\n stylization network", "Arbitrary Style Transfer in Real-time with Adaptive Instance\n Normalization", "Semantic Image Synthesis with Spatially-Adaptive Normalization", "Transfer Learning from Synthetic to Real-Noise Denoising with Adaptive\n Instance Normalization"], "answer_arxiv_id": ["1610.07629", "1705.06830", "1703.06868", "1903.07291", "2002.11244"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_4059"} +{"question": "Which papers present the use of Gaussian convolution with score matching in generative modelling?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_4060"} +{"question": "What work previously introduced gradient descent optimization for low rank approximation?", "answer": ["Learning-Based Low-Rank Approximations"], "answer_arxiv_id": ["1910.13984"], "source_meta": {"published_time": "20200720"}, "qid": "AutoScholarQuery_train_4061"} +{"question": "In what study showed the worst-case inequality T1/2⋆​(μ)≤2​T⋆​(μ) for any single-parameter exponential families?", "answer": ["Simple Bayesian Algorithms for Best-Arm Identification"], "answer_arxiv_id": ["1602.08448"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_4062"} +{"question": "What papers are about embedding color images within QR codes?", "answer": ["ART-UP: A Novel Method for Generating Scanning-robust Aesthetic QR codes"], "answer_arxiv_id": ["1803.02280"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_4063"} +{"question": "What works are related to model compression methods such as pruning?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding", "To prune, or not to prune: exploring the efficacy of pruning for model compression", "Structured Pruning of Large Language Models", "Structured Pruning Learns Compact and Accurate Models"], "answer_arxiv_id": ["1510.00149", "1710.01878", "1910.04732", "2204.00408"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_4064"} +{"question": "What are the studies that have used discrete neural representation learning in videos, audios, and anomaly detection?", "answer": ["Neural Discrete Representation Learning", "Predicting Video with VQVAE", "VideoGPT: Video Generation using VQ-VAE and Transformers", "vq-wav2vec: Self-Supervised Learning of Discrete Speech Representations", "Jukebox: A Generative Model for Music", "Transformer VQ-VAE for Unsupervised Unit Discovery and Speech Synthesis: ZeroSpeech 2020 Challenge", "AudioLM: a Language Modeling Approach to Audio Generation"], "answer_arxiv_id": ["1711.00937", "2103.01950", "2104.10157", "1910.05453", "2005.00341", "2005.11676", "2209.03143"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_4065"} +{"question": "Which works use supervised learning and lexicon expansion techniques to analyze framing?", "answer": ["A Systematic Media Frame Analysis of 1.5 Million New York Times Articles\n from 2000 to 2017", "Modeling Framing in Immigration Discourse on Social Media", "Framing and Agenda-setting in Russian News: a Computational Analysis of\n Intricate Political Strategies", "Weakly Supervised Learning of Nuanced Frames for Analyzing Polarization\n in News Media"], "answer_arxiv_id": ["2005.01803", "2104.06443", "1808.09386", "2009.09609"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_4066"} +{"question": "Which papers are dedicated to correcting classification bias when new tasks arise?", "answer": ["Large Scale Incremental Learning", "Maintaining Discrimination and Fairness in Class Incremental Learning"], "answer_arxiv_id": ["1905.13260", "1911.07053"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_4067"} +{"question": "In what papers did researchers use agreement margin principle to set the rejection threshold between accepted and rejected examples?", "answer": ["AUC-based Selective Classification"], "answer_arxiv_id": ["2210.10703v2"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_4068"} +{"question": "Could you provide me some works about pre-trained foundation models?", "answer": ["On the Opportunities and Risks of Foundation Models", "A Comprehensive Survey on Pretrained Foundation Models: A History from BERT to ChatGPT"], "answer_arxiv_id": ["2108.07258v3", "2302.09419"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_4069"} +{"question": "What studies established the foundational ideas of diffusion models and score-based generative models in image generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Protein Structure and Sequence Generation with Equivariant Denoising\n Diffusion Probabilistic Models", "Generative Modeling by Estimating Gradients of the Data Distribution", "Improved Techniques for Training Score-Based Generative Models"], "answer_arxiv_id": ["2006.11239", "2205.15019", "1907.05600", "2006.09011"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_4070"} +{"question": "Which works proposed interaction generation in 3D space?", "answer": ["NIFTY: Neural Object Interaction Fields for Guided Human Motion\n Synthesis", "Scene Synthesis from Human Motion", "Object Motion Guided Human Motion Synthesis", "InterDiff: Generating 3D Human-Object Interactions with Physics-Informed\n Diffusion", "Synthesizing Diverse Human Motions in 3D Indoor Scenes"], "answer_arxiv_id": ["2307.07511", "2301.01424", "2309.16237", "2308.16905", "2305.12411"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_4071"} +{"question": "What paper applied discrete diffusion to text generation?", "answer": ["Argmax Flows and Multinomial Diffusion: Learning Categorical Distributions"], "answer_arxiv_id": ["2102.05379"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_4072"} +{"question": "Which studies introduced DETR architecture into open-vocabulary detection?", "answer": ["Simple Open-Vocabulary Object Detection with Vision Transformers", "Scaling Open-Vocabulary Object Detection", "Open-Vocabulary DETR with Conditional Matching"], "answer_arxiv_id": ["2205.06230", "2306.09683", "2203.11876"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_4073"} +{"question": "What papers reported using Transformers for offline reinforcement learning, multi-task behavioural cloning, and algorithm distillation?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "A Generalist Agent", "In-context Reinforcement Learning with Algorithm Distillation"], "answer_arxiv_id": ["2106.01345", "2205.06175", "2210.14215"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_4074"} +{"question": "In what work the Wigner-Eckart theorem was applied to parametrize G-steerable kernel spaces over orbits of a compact G?", "answer": ["A Wigner-Eckart Theorem for Group Equivariant Convolution Kernels"], "answer_arxiv_id": ["2010.10952"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_4075"} +{"question": "What paper proposes to represent the world model as a graph by clustering states into nodes, similar to SEA?", "answer": ["World Model as a Graph: Learning Latent Landmarks for Planning"], "answer_arxiv_id": ["2011.12491"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_4076"} +{"question": "Could you provide me with research that models semantic-level state transitions between two consecutive frames for action reasoning?", "answer": ["Explainable Video Action Reasoning via Prior Knowledge and State\n Transitions"], "answer_arxiv_id": ["1908.10700"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_4077"} +{"question": "Which references present the use of Byte Pair Encoding (BPE) in natural language processing (NLP)?", "answer": ["Neural Machine Translation of Rare Words with Subword Units"], "answer_arxiv_id": ["1508.07909v5"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_4078"} +{"question": "Which paper highlighted the sensitivity of the Improved PR metric to outlier samples?", "answer": ["Reliable Fidelity and Diversity Metrics for Generative Models"], "answer_arxiv_id": ["2002.09797"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_4079"} +{"question": "Which papers are related to the creation of datasets containing more general distribution shifts?", "answer": ["MetaShift: A Dataset of Datasets for Evaluating Contextual Distribution Shifts and Training Conflicts"], "answer_arxiv_id": ["2202.06523"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_4080"} +{"question": "Could you provide me some works that employ 2D text-to-image diffusion models for 3D representations?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2305.16213", "2212.08751"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_4081"} +{"question": "Can you list some research articles that propose alterations to the MPNN framework or introduce extra heuristics?", "answer": ["Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting", "Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks", "Weisfeiler and Lehman Go Cellular: CW Networks", "Equivariant Subgraph Aggregation Networks"], "answer_arxiv_id": ["2006.09252", "2103.03212", "2106.12575", "2110.02910"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_4082"} +{"question": "What independent work connects multicalibration with real-valued boosting to minimize ℓ2 loss?", "answer": ["Multicalibration as Boosting for Regression"], "answer_arxiv_id": ["2301.13767"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_4083"} +{"question": "Which research introduced the technique of using adversarial methods to learn time-invariant representations for continuous domain adaptation?", "answer": ["Continuously Indexed Domain Adaptation"], "answer_arxiv_id": ["2007.01807"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_4084"} +{"question": "What are some complex scene parsing datasets proposed by the community with more classes?", "answer": ["Microsoft COCO: Common Objects in Context"], "answer_arxiv_id": ["1405.0312"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_4085"} +{"question": "What large-scale pre-training datasets use text supervision instead of labelled categories?", "answer": ["HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million\n Narrated Video Clips", "Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval"], "answer_arxiv_id": ["1906.03327", "2104.00650"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_4086"} +{"question": "Which papers suggest maintaining cross-frame consistency in their video synthesis method?", "answer": ["Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation"], "answer_arxiv_id": ["2306.07954"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4087"} +{"question": "Could you tell me the studies which used architectural compression to increase efficiency of each sampling step in DMs?", "answer": ["Cascaded Diffusion Models for High Fidelity Image Generation", "Tackling the Generative Learning Trilemma with Denoising Diffusion GANs", "SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two\n Seconds", "Diffusion Probabilistic Model Made Slim"], "answer_arxiv_id": ["2106.15282", "2112.07804", "2306.00980", "2211.17106"], "source_meta": {"published_time": "20240508"}, "qid": "AutoScholarQuery_train_4088"} +{"question": "What is the paper that introduced the NetHack Learning Environment?", "answer": ["The NetHack Learning Environment"], "answer_arxiv_id": ["2006.13760"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4089"} +{"question": "What studies fall into the category of structure-based methods for continual learning?", "answer": ["Piggyback: Adapting a Single Network to Multiple Tasks by Learning to\n Mask Weights", "Progressive Neural Networks", "DyTox: Transformers for Continual Learning with DYnamic TOken eXpansion", "Expert Gate: Lifelong Learning with a Network of Experts", "Adaptive Aggregation Networks for Class-Incremental Learning"], "answer_arxiv_id": ["1801.06519", "1606.04671", "2111.11326", "1611.06194", "2010.05063"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_4090"} +{"question": "Who proposed text generation as an objective for task-general multimodal models?", "answer": ["Unifying Vision-and-Language Tasks via Text Generation"], "answer_arxiv_id": ["2102.02779"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_4091"} +{"question": "Which researchers proposed to collect real-world HR-LR image pairs for training in blind image super-resolution?", "answer": ["Toward Real-World Single Image Super-Resolution: A New Benchmark and A\n New Model", "Zoom To Learn, Learn To Zoom", "Camera Lens Super-Resolution"], "answer_arxiv_id": ["1904.00523", "1905.05169", "1904.03378"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_4092"} +{"question": "Which works addressed the issue of requiring mitigation with discriminative methods when a customized answer set is provided in inference?", "answer": ["Self-supervised vision-language pretraining for Medical visual question answering"], "answer_arxiv_id": ["2211.13594v1"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_4093"} +{"question": "Which paper developed a method that output hierarchical matrix-Fisher distributions that exploit the SMPL kinematic tree?", "answer": ["Hierarchical Kinematic Probability Distributions for 3D Human Shape and\n Pose Estimation from Images in the Wild"], "answer_arxiv_id": ["2110.00990"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_4094"} +{"question": "What papers laid the foundations of large language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1810.04805", "1910.10683"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_4095"} +{"question": "Could you provide some studies that focused on augmenting data or synthesizing images by using properties of retinal sampling?", "answer": ["On the use of Cortical Magnification and Saccades as Biological Proxies for Data Augmentation"], "answer_arxiv_id": ["2112.07173"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_4096"} +{"question": "Which works discuss the application of transformer token reduction in natural language processing (NLP)?", "answer": ["PoWER-BERT: Accelerating BERT Inference via Progressive Word-vector Elimination", "Length-Adaptive Transformer: Train Once with Length Drop, Use Anytime with Search", "Learned Token Pruning for Transformers", "A Study on Token Pruning for ColBERT"], "answer_arxiv_id": ["2001.08950", "2010.07003", "2107.00910", "2112.06540"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_4097"} +{"question": "Which research noted that training for adversarial examples may also increase robust error?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1706.06083", "1901.08573"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_4098"} +{"question": "Which works attempt to incorporate 3D visual information into large language models?", "answer": ["3D-LLM: Injecting the 3D World into Large Language Models", "PointLLM: Empowering Large Language Models to Understand Point Clouds", "3D-GPT: Procedural 3D Modeling with Large Language Models", "LLM-Grounder: Open-Vocabulary 3D Visual Grounding with Large Language\n Model as an Agent"], "answer_arxiv_id": ["2307.12981", "2308.16911", "2310.12945", "2309.12311"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_4099"} +{"question": "What research papers used formal systems such as string rewriting systems, cellular encoding schemes, or hyperedge replacement graph grammars in NAS?", "answer": ["Path-Level Network Transformation for Efficient Architecture Search"], "answer_arxiv_id": ["1806.02639"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_4100"} +{"question": "What papers presented research about diffusion models used for high fidelity image and video synthesis?", "answer": ["Denoising Diffusion Probabilistic Models", "Variational Diffusion Models", "Video Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2107.00630", "2204.03458"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_4101"} +{"question": "What is the paper that describes MaskFormer?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation"], "answer_arxiv_id": ["2107.06278"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_4102"} +{"question": "Which work showed that Adam with certain hyper-parameters could work on counter-examples?", "answer": ["AdaShift: Decorrelation and Convergence of Adaptive Learning Rate Methods"], "answer_arxiv_id": ["1810.00143"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_4103"} +{"question": "Which works have attempted to incorporate semantic data into NeRF?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene\n Representation", "Decomposing NeRF for Editing via Feature Field Distillation", "Panoptic Lifting for 3D Scene Understanding with Neural Fields", "Panoptic NeRF: 3D-to-2D Label Transfer for Panoptic Urban Scene\n Segmentation", "DM-NeRF: 3D Scene Geometry Decomposition and Manipulation from 2D Images", "Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D\n Image Representations", "LERF: Language Embedded Radiance Fields"], "answer_arxiv_id": ["2103.15875", "2205.15585", "2212.09802", "2203.15224", "2208.07227", "2209.03494", "2303.09553"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_4104"} +{"question": "Which research papers have introduced open-source knowledge graphs for different fields?", "answer": ["OntoProtein: Protein Pretraining With Gene Ontology Embedding"], "answer_arxiv_id": ["2201.11147"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_4105"} +{"question": "Any works about developing ANN-to-SNN conversion for SNN models?", "answer": ["Spiking Deep Networks with LIF Neurons", "Optimal ANN-SNN Conversion for High-accuracy and Ultra-low-latency Spiking Neural Networks", "Training High-Performance Low-Latency Spiking Neural Networks by Differentiation on Spike Representation"], "answer_arxiv_id": ["1510.08829", "2303.04347", "2205.00459"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_4106"} +{"question": "What research proposed the fast gradient sign method (FGSM) to improve the robustness of a DNN model?", "answer": ["Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1412.6572"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_4107"} +{"question": "What studies have proposed methods to estimate the confusion matrix or skill of workers under the D&S model using iterative algorithms?", "answer": ["Budget-Optimal Task Allocation for Reliable Crowdsourcing Systems", "Error Rate Bounds and Iterative Weighted Majority Voting for Crowdsourcing"], "answer_arxiv_id": ["1110.3564", "1411.4086"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_4108"} +{"question": "What are some references about end-to-end approaches in the context of multi-view stereo (MVS)?", "answer": ["Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision", "Multiview Neural Surface Reconstruction by Disentangling Geometry and Appearance", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction", "Improving neural implicit surfaces geometry with patch warping", "Geo-Neus: Geometry-Consistent Neural Implicit Surfaces Learning for Multi-view Reconstruction", "Learning Deformable Tetrahedral Meshes for 3D Reconstruction", "NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in the Wild", "Differentiable Stereopsis: Meshes from multiple views using differentiable rendering", "Extracting Triangular 3D Models, Materials, and Lighting From Images", "Neural 3D Mesh Renderer", "Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning", "Learning to Predict 3D Objects with an Interpolation-based Differentiable Renderer", "Accelerating 3D Deep Learning with PyTorch3D", "Modular Primitives for High-Performance Differentiable Rendering", "PartNeRF: Generating Part-Aware Editable 3D Shapes without 3D Supervision"], "answer_arxiv_id": ["1912.07372", "2003.09852", "2003.08934", "2104.10078", "2106.12052", "2106.10689", "2112.09648", "2205.15848", "2011.01437", "2110.07604", "2110.05472", "2111.12503", "1711.07566", "1904.01786", "1908.01210", "2007.08501", "2011.03277", "2303.09554"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_4109"} +{"question": "Could you tell me the works that discuss latent diffusion models used for efficient text-to-image (T2I) generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Scalable Diffusion Models with Transformers", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis"], "answer_arxiv_id": ["2112.10752", "2212.09748", "2307.01952"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4110"} +{"question": "What research proposes a unifying framework for popular diffusion models?", "answer": ["Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2206.00364"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_4111"} +{"question": "Could you provide some research about the optimization of graph construction in the context of LPA?", "answer": ["Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learning", "Deep Metric Transfer for Label Propagation with Limited Annotated Data"], "answer_arxiv_id": ["1805.10002", "1812.08781"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_4112"} +{"question": "Which studies demonstrate promising rendering results, especially when attaching the methods to a conventional network pipeline?", "answer": ["Free-Viewpoint RGB-D Human Performance Capture and Rendering"], "answer_arxiv_id": ["2112.13889"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4113"} +{"question": "Which work has recognized that TESTES{TES} samples new noise in every truncation window to resolve its bias?", "answer": ["Unbiased Gradient Estimation in Unrolled Computation Graphs with Persistent Evolution Strategies"], "answer_arxiv_id": ["2112.13835"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_4114"} +{"question": "Any research aiming at postponing the interaction until the last layer of the model for neural retrieval models?", "answer": ["COIL: Revisit Exact Lexical Match in Information Retrieval with Contextualized Inverted List", "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT"], "answer_arxiv_id": ["2104.07186", "2004.12832"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_train_4115"} +{"question": "Which works have discussed the impact of covariant-shift in out-of-distribution (OOD) detection?", "answer": ["On the Impact of Spurious Correlation for Out-of-distribution Detection"], "answer_arxiv_id": ["2109.05642"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_4116"} +{"question": "What studies have used contrastive loss in deep learning and specifically the triplet loss?", "answer": ["Deep Speaker: an End-to-End Neural Speaker Embedding System"], "answer_arxiv_id": ["1705.02304"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_4117"} +{"question": "Could you mention the works that use DeepZ relaxation approximated using Cauchy random matrices?", "answer": ["Scaling provable adversarial defenses"], "answer_arxiv_id": ["1805.12514"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_4118"} +{"question": "What papers contextualize the task relationship in a teacher-student network setup?", "answer": ["Continual Learning in the Teacher-Student Setup: Impact of Task Similarity", "Statistical Mechanical Analysis of Catastrophic Forgetting in Continual Learning with Teacher and Student Networks"], "answer_arxiv_id": ["2107.04384", "2105.07385v1"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_4119"} +{"question": "Which papers argue grid cells as a basis for predicting future outcomes?", "answer": ["Prediction and Generalisation Over Directed Actions by Grid Cells"], "answer_arxiv_id": ["2006.03355"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_4120"} +{"question": "Can you provide work that discovered a terminal training stage when the embedding collapses to the geometric means of the classifier for each category?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning training", "Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training", "Neural collapse with unconstrained features"], "answer_arxiv_id": ["2008.08186", "2101.12699", "2011.11619"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_4121"} +{"question": "Which studies highlight that the video-text models struggle in comprehending the semantics of the text focusing on manipulating the verbs, actions, and entities grounded in the video description?", "answer": ["Verbs in Action: Improving verb understanding in video-language models", "Paxion: Patching Action Knowledge in Video-Language Foundation Models"], "answer_arxiv_id": ["2304.06708", "2305.10683"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_4122"} +{"question": "What works are known for providing new approaches for capturing class distributions and aligning image distributions in the area of 2D adversarial attacks?", "answer": ["On Generating Transferable Targeted Perturbations"], "answer_arxiv_id": ["2103.14641"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_4123"} +{"question": "Which works report that prior research on amodal completion, segmentation, and detection is limited to a small set of objects or to synthetic data?", "answer": ["Self-Supervised Scene De-occlusion", "Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers", "Amodal Completion and Size Constancy in Natural Scenes", "SeGAN: Segmenting and Generating the Invisible"], "answer_arxiv_id": ["2004.02788", "2103.12340", "1509.08147", "1703.10239"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_4124"} +{"question": "Could you provide me any works discussing the limitations of annealing as a form of perturbation?", "answer": ["Spread Divergence"], "answer_arxiv_id": ["1811.08968"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_4125"} +{"question": "What papers predefined the bias type and used such prior knowledge to make models robust against that predefined bias type?", "answer": ["Learning Robust Representations by Projecting Superficial Statistics Out", "Learning De-biased Representations with Biased Representations"], "answer_arxiv_id": ["1903.06256", "1910.02806"], "source_meta": {"published_time": "20240430"}, "qid": "AutoScholarQuery_train_4126"} +{"question": "What papers are foundational in exploring multi-task learning with a single model in NLP?", "answer": ["Natural Language Processing (almost) from Scratch"], "answer_arxiv_id": ["1103.0398"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_4127"} +{"question": "Which papers provide fine-grained activity annotations for actions understanding datasets?", "answer": ["The “something something” video database for learning and evaluating visual common sense", "Rescaling Egocentric Vision", "LEMMA: A Multi-view Dataset for LEarning Multi-agent Multi-task Activities", "Home Action Genome: Cooperative Compositional Action Understanding", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["1706.04261", "2006.13256", "2007.15781", "2105.05226", "2110.07058"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_4128"} +{"question": "Could you provide some studies that have developed differentiable point-based and sphere-based rendering?", "answer": ["SynSin: End-to-end View Synthesis from a Single Image", "Pulsar: Efficient Sphere-based Neural Rendering"], "answer_arxiv_id": ["1912.08804", "2004.07484"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4129"} +{"question": "What libraries have been developed to guide the advancement of uncertainty quantification in machine learning?", "answer": ["Fortuna: A Library for Uncertainty Quantification in Deep Learning", "Uncertainty Baselines: Benchmarks for Uncertainty & Robustness in Deep Learning"], "answer_arxiv_id": ["2302.04019", "2106.04015"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_4130"} +{"question": "Could you mention research that utilizes mixup and adversarial training techniques adapted from vision for tabular data?", "answer": ["mixup: Beyond Empirical Risk Minimization", "Explaining and Harnessing Adversarial Examples", "Well-tuned Simple Nets Excel on Tabular Datasets"], "answer_arxiv_id": ["1710.09412", "1412.6572", "2106.11189"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_4131"} +{"question": "Could you provide me with some studies on 'Sharding' that allows for a more efficient model training through the use of additional GPUs?", "answer": ["ZeRO: Memory Optimizations Toward Training Trillion Parameter Models"], "answer_arxiv_id": ["1910.02054"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_4132"} +{"question": "Could you provide me some studies on non-optimistic instance-optimal algorithms for linear bandits?", "answer": ["Asymptotically Optimal Information-Directed Sampling", "Adaptive Exploration in Linear Contextual Bandit"], "answer_arxiv_id": ["2011.05944", "1910.06996"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_4133"} +{"question": "Which are the initial works on the ALFRED dataset that employed end-to-end methods in the context of Visual Language Navigation?", "answer": ["ALFRED A Benchmark for Interpreting Grounded Instructions for Everyday Tasks", "Hierarchical Task Learning from Language Instructions with Unified Transformers and Self-Monitoring"], "answer_arxiv_id": ["1912.01734", "2106.03427"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_4134"} +{"question": "Which research discusses the concept of multiaccuracy in connection with agnostic learning?", "answer": ["Loss Minimization through the Lens of Outcome Indistinguishability"], "answer_arxiv_id": ["2210.08649"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_4135"} +{"question": "Which studies have used 3D models to aid correspondence estimation?", "answer": ["Learning Dense Correspondence via 3D-guided Cycle Consistency", "Canonical Surface Mapping via Geometric Cycle Consistency", "DensePose 3D: Lifting Canonical Surface Maps of Articulated Objects to\n the Third Dimension", "Learning 3D Dense Correspondence via Canonical Point Autoencoder"], "answer_arxiv_id": ["1604.05383", "1907.10043", "2109.00033", "2107.04867"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_4136"} +{"question": "What work introduced a solution to improve the ability to generate personalized images of subjects in specific contexts?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_4137"} +{"question": "Could you provide me some works that used box embeddings to model various relations among tasks?", "answer": ["Box Embeddings: An open-source library for representation learning using\n geometric structures", "Probabilistic Embedding of Knowledge Graphs with Box Lattice Measures", "BoxE: A Box Embedding Model for Knowledge Base Completion", "Query2box: Reasoning over Knowledge Graphs in Vector Space using Box\n Embeddings"], "answer_arxiv_id": ["2109.04997", "1805.06627", "2007.06267v2", "2002.05969"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_4138"} +{"question": "Could you provide me some works that use adapters in parameter-efficient transfer learning?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Parameter-efficient Multi-task Fine-tuning for Transformers via Shared Hypernetworks"], "answer_arxiv_id": ["1902.00751", "2106.04489"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_4139"} +{"question": "What papers have applied the concept of average sensitivity to dynamic programming problems?", "answer": ["Average Sensitivity of Dynamic Programming"], "answer_arxiv_id": ["2111.02657"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_4140"} +{"question": "Where is a presented a framework for modeling a Markov chain with a stationary distribution approximating the data distribution?", "answer": ["Deep Generative Stochastic Networks Trainable by Backprop"], "answer_arxiv_id": ["1306.1091"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4141"} +{"question": "What works have studied backpropagating through an unrolled parameter estimation mapping?", "answer": ["Gradient-based Hyperparameter Optimization through Reversible Learning", "Bilevel Programming for Hyperparameter Optimization and Meta-Learning", "Unrolled Generative Adversarial Networks"], "answer_arxiv_id": ["1502.03492", "1806.04910", "1611.02163"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_4142"} +{"question": "What research papers utilize the mechanism of attention, sampling, edge removing, and meta-edge choosing during neighborhood aggregation in GNN-based GAD methods?", "answer": ["Alleviating the Inconsistency Problem of Applying Graph Neural Network to Fraud Detection"], "answer_arxiv_id": ["2005.00625"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4143"} +{"question": "Which works proposed occlusion-aware adjustments in the field of unsupervised optical flow?", "answer": ["Occlusion Aware Unsupervised Learning of Optical Flow", "UnFlow: Unsupervised Learning of Optical Flow with a Bidirectional\n Census Loss", "UFD-PRiME: Unsupervised Joint Learning of Optical Flow and Stereo Depth\n through Pixel-Level Rigid Motion Estimation"], "answer_arxiv_id": ["1711.05890", "1711.07837", "2310.04712"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_4144"} +{"question": "Which papers are about optimizing objectives like code compilability, readability, or passing test cases in code generation", "answer": ["Execution-based Code Generation using Deep Reinforcement Learning", "Planning with Large Language Models for Code Generation"], "answer_arxiv_id": ["2301.13816", "2303.05510"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_4145"} +{"question": "Which papers had studies about the emergence of transformers employing the attention mechanism in the field of text-to-image generation?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_4146"} +{"question": "Could you indicate some research that accomplished the enhancement of two-stage models by utilising detections of additional objects in the scene?", "answer": ["Unified Graph Structured Models for Video Understanding", "Object Level Visual Reasoning in Videos", "A Structured Model For Action Detection", "Videos as Space-Time Region Graphs"], "answer_arxiv_id": ["2103.15662", "1806.06157", "1812.03544", "1806.01810"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_4147"} +{"question": "What papers examine the impact of LLMs based on Transformers on the language processing field?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_4148"} +{"question": "Which work extends the results of efficient algorithm against the best fixed policy to the setting with general function approximation?", "answer": ["Learning Markov Games with Adversarial Opponents: Efficient Algorithms and Fundamental Limits"], "answer_arxiv_id": ["2203.06803"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_4149"} +{"question": "What papers can be found about pairing human motion with audio?", "answer": ["Dancing to Music", "Audio2Gestures: Generating Diverse Gestures from Speech Audio with\n Conditional Variational Autoencoders", "Generating Holistic 3D Human Motion from Speech"], "answer_arxiv_id": ["1911.02001", "2108.06720", "2212.04420"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_4150"} +{"question": "What work proposed using an anchor-based deep lane detection model in lane detection discipline?", "answer": ["Keep your Eyes on the Lane: Real-time Attention-guided Lane Detection"], "answer_arxiv_id": ["2010.12035v2"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_4151"} +{"question": "Which works did use machine learning to understand the materials and their properties in materials science?", "answer": ["Recent Advances and Applications of Deep Learning Methods in Materials Science"], "answer_arxiv_id": ["2110.14820"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_4152"} +{"question": "Which works studied coreset construction for vanilla (k,z)-Clustering in R?", "answer": ["A Unified Framework for Approximating and Clustering Data", "New Frameworks for Offline and Streaming Coreset Constructions", "A New Coreset Framework for Clustering"], "answer_arxiv_id": ["1106.1379", "1612.00889", "2104.06133"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_4153"} +{"question": "Which works have studied tasks other than segmentation using diffusion models?", "answer": ["DreamTeacher: Pretraining Image Backbones with Deep Generative Models", "Monocular Depth Estimation using Diffusion Models", "DiffusionDet: Diffusion Model for Object Detection"], "answer_arxiv_id": ["2307.07487", "2302.14816", "2211.09788"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_4154"} +{"question": "Could you show an example of a work illustrating the application of KL control problems in a RL context?", "answer": ["Optimal control as a graphical model inference problem"], "answer_arxiv_id": ["0901.0633"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_4155"} +{"question": "Which works have utilized Fourier layers in neural operators?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations"], "answer_arxiv_id": ["2010.08895"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_4156"} +{"question": "Which paper provided an approach to interpret the session embeddings from dual encoders in the domain of information retrieval?", "answer": ["What Are You Token About? Dense Retrieval as Distributions Over the\n Vocabulary"], "answer_arxiv_id": ["2212.10380"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_4157"} +{"question": "What recent work offers broader theoretical guarantees for mixing fairness and utility maximizing but does not take into account general user preferences?", "answer": ["Optimally Interpolating between Ex-Ante Fairness and Welfare"], "answer_arxiv_id": ["2302.03071"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_4158"} +{"question": "What research uses neural 3D points for representing and rendering a continuous radiance volume?", "answer": ["Point-NeRF: Point-based Neural Radiance Fields"], "answer_arxiv_id": ["2201.08845"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_4159"} +{"question": "What papers introduced the attention or gate mechanism in order to improve the representation ability in image segmentation?", "answer": ["PiCANet: Learning Pixel-wise Contextual Attention for Saliency Detection", "Suppress and Balance: A Simple Gated Network for Salient Object\n Detection"], "answer_arxiv_id": ["1708.06433v2", "2007.08074"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4160"} +{"question": "Could you cite some studies treating depth estimation as ordinal regression or classification by discretizing depth?", "answer": ["Deep Ordinal Regression Network for Monocular Depth Estimation", "AdaBins: Depth Estimation using Adaptive Bins"], "answer_arxiv_id": ["1806.02446", "2011.14141"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_4161"} +{"question": "Which studies explored video-to-video translation or video stylization for generating consistent videos?", "answer": ["Deep Video Portraits", "One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing", "VToonify: Controllable High-Resolution Portrait Video Style Transfer"], "answer_arxiv_id": ["1805.11714", "2011.15126", "2209.11224"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_4162"} +{"question": "Which papers have modeled 3D human motion through diffusion using transformers?", "answer": ["Human Motion Diffusion Model", "TransFusion: A Practical and Effective Transformer-based Diffusion Model\n for 3D Human Motion Prediction", "Synthesizing Long-Term Human Motions with Diffusion Models via Coherent\n Sampling", "Single Motion Diffusion", "LongDanceDiff: Long-term Dance Generation with Conditional Diffusion\n Model", "Human Motion Diffusion as a Generative Prior", "Fg-T2M: Fine-Grained Text-Driven Human Motion Generation via Diffusion\n Model", "Can We Use Diffusion Probabilistic Models for 3D Motion Prediction?", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "Human Joint Kinematics Diffusion-Refinement for Stochastic Motion\n Prediction"], "answer_arxiv_id": ["2209.14916", "2307.16106", "2308.01850", "2302.05905", "2308.11945", "2303.01418", "2309.06284", "2302.14503", "2208.15001", "2210.05976"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4163"} +{"question": "What studies attend on nearly on-policy samples?", "answer": ["Experience Replay with Likelihood-free Importance Weights"], "answer_arxiv_id": ["2006.13169"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_4164"} +{"question": "What document proposes graph-based egocentric action annotations in the Ego4D dataset?", "answer": ["Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2110.07058"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_4165"} +{"question": "Which study introduced the method of speculative decoding?", "answer": ["Fast Inference from Transformers via Speculative Decoding"], "answer_arxiv_id": ["2211.17192"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_4166"} +{"question": "Could you provide me some studies about interpolate-style temporal modeling?", "answer": ["Is Space-Time Attention All You Need for Video Understanding?", "TVTSv2: Learning Out-of-the-box Spatiotemporal Visual Representations at\n Scale", "Expanding Language-Image Pretrained Models for General Video Recognition", "ST-Adapter: Parameter-Efficient Image-to-Video Transfer Learning"], "answer_arxiv_id": ["2102.05095", "2305.14173", "2208.02816", "2206.13559"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4167"} +{"question": "What are the foundational studies in optimal transport problem and Wasserstein distance?", "answer": ["A Survey on Optimal Transport for Machine Learning: Theory and Applications"], "answer_arxiv_id": ["2106.01963"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_4168"} +{"question": "Could you provide me a research work that studied mean estimation with compression under local DP?", "answer": ["Lossless Compression of Efficient Private Local Randomizers", "Breaking the Communication-Privacy-Accuracy Trilemma", "Optimal Algorithms for Mean Estimation under Local Differential Privacy", "Minimax Optimal Procedures for Locally Private Estimation"], "answer_arxiv_id": ["2102.12099", "2007.11707", "2205.02466", "1604.02390"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_4169"} +{"question": "Which studies employed the Neumann series approximation with independent mini-batches for solving Stochastic Bilevel Optimization problems?", "answer": ["Approximation Methods for Bilevel Programming"], "answer_arxiv_id": ["1802.02246"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4170"} +{"question": "Are there any research articles discussing the need for adaptation in detectors for varied conditions in self-driving scenarios?", "answer": ["Train in Germany, Test in The USA: Making 3D Object Detectors Generalize"], "answer_arxiv_id": ["2005.08139"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_4171"} +{"question": "Which research papers focus on the few-shot learning approach?", "answer": ["Generalizing from a Few Examples: A Survey on Few-Shot Learning"], "answer_arxiv_id": ["1904.05046"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_4172"} +{"question": "Which studies are relevant to adapting both encoders for multi-modal tasks, such as image recognition and video-text retrieval?", "answer": ["Unified Vision and Language Prompt Learning", "MaPLe: Multi-modal Prompt Learning", "Cross-Modal Adapter for Text-Video Retrieval", "VoP: Text-Video Co-operative Prompt Tuning for Cross-Modal Retrieval"], "answer_arxiv_id": ["2210.07225", "2210.03117", "2211.09623", "2211.12764"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_4173"} +{"question": "Which work refers to the baking of radiance fields into dense voxel grids, thereby speeding convergence compared to the NeRF?", "answer": ["Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2112.05131"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_4174"} +{"question": "What papers discuss semantic segmentation of images and shapes in relation to CAD programs?", "answer": ["3D Shape Segmentation with Projective Convolutional Networks", "MvDeCor: Multi-view Dense Correspondence Learning for Fine-grained 3D\n Segmentation", "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection", "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "Learning Transferable Visual Models From Natural Language Supervision", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "PartSLIP: Low-Shot Part Segmentation for 3D Point Clouds via Pretrained\n Image-Language Models", "Grounded Language-Image Pre-training", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["1612.02808", "2208.08580", "2303.05499", "2401.14159", "2103.00020", "2210.04150", "2212.01558", "2112.03857", "2302.05543"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4175"} +{"question": "What works discuss the use of latent transformer architecture to manipulate image content?", "answer": ["User-Controllable Latent Transformer for StyleGAN Image Layout Editing"], "answer_arxiv_id": ["2208.12408"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_4176"} +{"question": "Can you list some studies that have proposed MPC protocols to ensure privacy in horizontal federated learning?", "answer": ["Turbo-Aggregate: Breaking the Quadratic Aggregation Barrier in Secure Federated Learning", "Communication-Computation Efficient Secure Aggregation for Federated Learning", "LightSecAgg: a Lightweight and Versatile Design for Secure Aggregation in Federated Learning", "Efficient Dropout-resilient Aggregation for Privacy-preserving Machine Learning"], "answer_arxiv_id": ["2002.04156", "2012.05433", "2109.14236v3", "2203.17044"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_4177"} +{"question": "Can you provide me with some studies that developed architectures for algorithm inference by learning programs as compositions of simple instructions?", "answer": ["Neural Programmer-Interpreters", "Neural Arithmetic Expression Calculator"], "answer_arxiv_id": ["1511.06279", "1809.08590"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_4178"} +{"question": "Are there any works that apply keypoint descriptors in biomedical imaging for tasks like anatomy object classification, medical image retrieval, segmentation tasks, etc.?", "answer": ["Multi-scale Cell Instance Segmentation with Keypoint Graph based Bounding Boxes", "Keypoint Transfer for Fast Whole-Body Segmentation", "Estimation of Large Motion in Lung CT by Integrating Regularized Keypoint Correspondences into Dense Deformable Registration", "Adapting the Mean Teacher for keypoint-based lung registration under geometric domain shifts"], "answer_arxiv_id": ["1907.09140", "1806.08723", "1807.00467v1", "2207.00371"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_4179"} +{"question": "What work shows that simple gating and channel attention is sufficient to incorporate global context with better efficiency and performance?", "answer": ["Simple Baselines for Image Restoration"], "answer_arxiv_id": ["2204.04676"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_4180"} +{"question": "Which work introduced the Perceiver architecture that can scale transformers to input sequences of any length?", "answer": ["Perceiver: General Perception with Iterative Attention"], "answer_arxiv_id": ["2103.03206"], "source_meta": {"published_time": "20240718"}, "qid": "AutoScholarQuery_train_4181"} +{"question": "What studies directly match the output or intermediate layers of the teacher network in different KD strategies?", "answer": ["Distilling the Knowledge in a Neural Network", "FitNets: Hints for Thin Deep Nets", "Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer", "Paraphrasing Complex Network: Network Compression via Factor Transfer"], "answer_arxiv_id": ["1503.02531", "1412.6550", "1612.03928", "1802.04977"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_4182"} +{"question": "Which papers served as the inspiration for developing H3 by using linear attention and classical sequence models like RNNs?", "answer": ["Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention"], "answer_arxiv_id": ["2006.16236"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_4183"} +{"question": "What works present benchmark datasets for assessing natural language understanding and reasoning capabilities of LLMs?", "answer": ["RACE: Large-scale ReAding Comprehension Dataset From Examinations", "TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension", "DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs", "GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding", "SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems", "WinoGrande: An Adversarial Winograd Schema Challenge at Scale", "TruthfulQA: Measuring How Models Mimic Human Falsehoods", "Measuring Massive Multitask Chinese Understanding", "Evaluating Large Language Models Trained on Code", "Training Verifiers to Solve Math Word Problems", "C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for Foundation Models", "M3KE: A Massive Multi-Level Multi-Subject Knowledge Evaluation Benchmark for Chinese Large Language Models"], "answer_arxiv_id": ["1704.04683", "1705.03551", "1903.00161", "1804.07461", "1905.00537", "1907.10641", "2109.07958", "2304.12986v2", "2107.03374", "2110.14168", "2305.08322", "2305.10263"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_4184"} +{"question": "What studies propose the implementation of regularized Boltzmann policy in IRL and resemble a Bayes posterior?", "answer": ["Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor"], "answer_arxiv_id": ["1801.01290"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4185"} +{"question": "What works explore the advantages of BEV representation in multi-sensor fusion?", "answer": ["BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird’s-Eye View Representation"], "answer_arxiv_id": ["2205.13542"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_4186"} +{"question": "What works about improving alignment of LLMs by fine-tuning on specialized datasets?", "answer": ["Constitutional AI: Harmlessness from AI Feedback", "Fine-tuning language models to find agreement among humans with diverse\n preferences"], "answer_arxiv_id": ["2212.08073", "2211.15006"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_4187"} +{"question": "What works utilize knowledge distillation technique for model compression of speech models?", "answer": ["DistilHuBERT: Speech Representation Learning by Layer-wise Distillation\n of Hidden-unit BERT", "DPHuBERT: Joint Distillation and Pruning of Self-Supervised Speech\n Models", "Distil-Whisper: Robust Knowledge Distillation via Large-Scale Pseudo\n Labelling"], "answer_arxiv_id": ["2110.01900", "2305.17651", "2311.00430"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_4188"} +{"question": "In which study does intrinsic self-correction based solely on inherent capabilities of LLMs prove to be unreliable?", "answer": ["Large Language Models Cannot Self-Correct Reasoning Yet"], "answer_arxiv_id": ["2310.01798"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_4189"} +{"question": "Could you provide examples of research that used specific structural information of the white-box model in data-free knowledge distillation?", "answer": ["Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion", "Data-Free Learning of Student Networks"], "answer_arxiv_id": ["1912.08795", "1904.01186"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_4190"} +{"question": "What works applied the sprite approach in object-centric learning that learns a dictionary of object appearances?", "answer": ["Unsupervised Layered Image Decomposition into Object Prototypes", "MarioNette: Self-Supervised Sprite Learning"], "answer_arxiv_id": ["2104.14575", "2104.14553"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_4191"} +{"question": "What papers considered the effect of permutation symmetries in parameter space on connectivity of the loss landscape?", "answer": ["Explaining Landscape Connectivity of Low-cost Solutions for Multilayer Nets", "Weight-space symmetry in deep networks gives rise to permutation saddles, connected by equal-loss valleys across the loss landscape", "The Role of Permutation Invariance in Linear Mode Connectivity of Neural Networks", "Git Re-Basin: Merging Models Modulo Permutation Symmetries", "REPAIR: REnormalizing Permuted Activations for Interpolation Repair"], "answer_arxiv_id": ["1906.06247v2", "1907.02911", "2110.06296", "2209.04836", "2211.08403"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_4192"} +{"question": "Which works use future reward sequences as supervisory signals to enhance the performance of visual RL algorithms?", "answer": ["Learning Task-relevant Representations for Generalization via Characteristic Functions of Reward Sequence Distributions", "Generalization in Visual Reinforcement Learning with the Reward Sequence Distribution"], "answer_arxiv_id": ["2205.10218", "2302.09601"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_4193"} +{"question": "Which works have used L2O for domain adaptation and adversarial training tasks?", "answer": ["Automated Synthetic-to-Real Generalization", "Improved Adversarial Training via Learned Optimizer"], "answer_arxiv_id": ["2007.06965", "2004.12227"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_4194"} +{"question": "Which studies review the detection of toxicity, harm and hate speech in the evaluation of language models?", "answer": ["Handling Bias in Toxic Speech Detection: A Survey"], "answer_arxiv_id": ["2202.00126"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_4195"} +{"question": "Which work introduces a lightweight adapter module to predict the network depth map for efficient adaptive inference with sparse convolution?", "answer": ["Deep Adaptive Inference Networks for Single Image Super-Resolution"], "answer_arxiv_id": ["2004.03915"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_4196"} +{"question": "What is the paper that introduced the 3D Gaussian Splatting (3D-GS) model?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_4197"} +{"question": "What studies have focused on generating multi-view images?", "answer": ["MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion"], "answer_arxiv_id": ["2307.01097"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4198"} +{"question": "Could you provide me with studies where each document is associated with an identifier which could be keywords?", "answer": ["TOME: A Two-stage Approach for Model-based Retrieval", "Autoregressive Search Engines: Generating Substrings as Document\n Identifiers", "GLEN: Generative Retrieval via Lexical Index Learning", "Nonparametric Decoding for Generative Retrieval"], "answer_arxiv_id": ["2305.11161", "2204.10628", "2311.03057", "2210.02068"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_4199"} +{"question": "Which papers used distribution correction estimation in conservatism value estimation for offline RL algorithms?", "answer": ["CoinDICE: Off-Policy Confidence Interval Estimation", "Off-Policy Evaluation via the Regularized Lagrangian"], "answer_arxiv_id": ["2010.11652", "2007.03438"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_4200"} +{"question": "Can you provide a study that utilized autoencoder learning techniques to find consistent projections in large-scale MVC tasks?", "answer": ["Graph-Collaborated Auto-Encoder Hashing for Multi-view Binary Clustering"], "answer_arxiv_id": ["2301.02484"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_4201"} +{"question": "Which papers proposed the application of DAE pretraining of point cloud Transformers?", "answer": ["Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling", "Masked Discrimination for Self-Supervised Learning on Point Clouds", "Masked Autoencoders for Point Cloud Self-supervised Learning"], "answer_arxiv_id": ["2111.14819", "2203.11183", "2203.06604"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_4202"} +{"question": "Can you provide works that propose methods for self-supervised learning?", "answer": ["Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles", "Colorful Image Colorization", "Unsupervised Representation Learning by Predicting Image Rotations", "Deep Clustering for Unsupervised Learning of Visual Features", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments"], "answer_arxiv_id": ["1603.09246v3", "1603.08511", "1803.07728", "1807.05520", "2006.09882"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_4203"} +{"question": "Which works studied the approach of modeling Wikipedia edits for text editing?", "answer": ["Learning to Model Editing Processes", "Text Editing by Command", "PEER: A Collaborative Language Model"], "answer_arxiv_id": ["2205.12374", "2010.12826", "2208.11663"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_4204"} +{"question": "What works have been developed to enhance the performance of point-based methods in point cloud completion?", "answer": ["PF-Net: Point Fractal Network for 3D Point Cloud Completion"], "answer_arxiv_id": ["2003.00410"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_4205"} +{"question": "What studies utilize model-based hand pose estimation methods and with which statistical models?", "answer": ["3D Hand Shape and Pose from Images in the Wild", "FrankMocap: Fast Monocular 3D Hand and Body Motion Capture by Regression and Integration", "Fast and Robust Hand Tracking Using Detection-Guided Optimization", "End-to-end Hand Mesh Recovery from a Monocular RGB Image", "Monocular Real-time Hand Shape and Motion Capture using Multi-modal Data", "Embodied Hands: Modeling and Capturing Hands and Bodies Together"], "answer_arxiv_id": ["1902.03451", "2008.08324", "1602.04124", "1902.09305", "2003.09572", "2201.02610"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_4206"} +{"question": "Which research work first proposed that feature collapse can be alleviated with an online predictor and a momentum encoder in non-contrastive learning?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["2006.07733"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_4207"} +{"question": "What studies tackled the variational inference problem using implicit distributions in the field of Implicit Variational Inference?", "answer": ["Variational Inference using Implicit Distributions"], "answer_arxiv_id": ["1702.08235"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_4208"} +{"question": "Which study is related to the authors' general framework and results in terms of approximating functions in arbitrary approximation spaces via an empirical least-squares regression?", "answer": ["Convergence bounds for empirical nonlinear least-squares"], "answer_arxiv_id": ["2001.00639"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_4209"} +{"question": "Could you provide examples of studies utilizing user preference studies for evaluating text-to-3D models?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "DreamBooth3D: Subject-Driven Text-to-3D Generation", "Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models", "CC3D: Layout-Conditioned Generation of Compositional 3D Scenes", "DITTO-NeRF: Diffusion-based Iterative Text To Omni-directional 3D Model", "DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation"], "answer_arxiv_id": ["2211.10440", "2303.13508", "2303.11989", "2303.12074", "2304.02827", "2309.16653"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_4210"} +{"question": "Could you provide me some studies about fully decentralized variants of DP-SGD?", "answer": ["Privacy Amplification by Decentralization", "Muffliato: Peer-to-Peer Privacy Amplification for Decentralized Optimization and Averaging"], "answer_arxiv_id": ["2012.05326", "2206.05091"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_4211"} +{"question": "Which papers compare the effectiveness of different slice sampling methods like GPSS, ESS, and HRUSS?", "answer": ["Elliptical slice sampling", "Geometric Convergence of Elliptical Slice Sampling", "Reversibility of elliptical slice sampling revisited", "Convergence of hybrid slice sampling via spectral gap", "Comparison of hit-and-run, slice sampler and random walk Metropolis", "Parallel MCMC with Generalized Elliptical Slice Sampling"], "answer_arxiv_id": ["1001.0175v2", "2105.03308", "2301.02426", "1409.2709", "1505.00579", "1210.7477"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_4212"} +{"question": "Could you provide the studies that proposed scene synthesis from a floor plan?", "answer": ["ATISS: Autoregressive Transformers for Indoor Scene Synthesis", "SceneFormer: Indoor Scene Generation with Transformers", "Fast and Flexible Indoor Scene Synthesis via Deep Convolutional Generative Models"], "answer_arxiv_id": ["2110.03675v1", "2012.09793", "1811.12463"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_4213"} +{"question": "Which works are based on the iterative differentiation (ITD) methods?", "answer": ["Gradient-based Hyperparameter Optimization through Reversible Learning", "Forward and Reverse Gradient-Based Hyperparameter Optimization", "On First-Order Meta-Learning Algorithms", "Truncated Back-propagation for Bilevel Optimization"], "answer_arxiv_id": ["1502.03492", "1703.01785", "1803.02999", "1810.10667"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_4214"} +{"question": "What previous research papers provided the foundation for the idea of DDIMs?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_4215"} +{"question": "Could you tell me which papers are about written human-to-human TOD datasets?", "answer": ["MultiWOZ - A Large-Scale Multi-Domain Wizard-of-Oz Dataset for Task-Oriented Dialogue Modelling"], "answer_arxiv_id": ["1810.00278"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_4216"} +{"question": "Who has studied 2-dimensional objectives of a certain form in order to understand the emergence of threshold neurons?", "answer": ["Learning threshold neurons via the “edge of stability”"], "answer_arxiv_id": ["2212.07469"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_4217"} +{"question": "What are some papers that discuss the involvement of multiple modalities in the world around us?", "answer": ["Multimodal Machine Learning: A Survey and Taxonomy", "Multimodal Learning with Transformers: A Survey"], "answer_arxiv_id": ["1705.09406", "2206.06488"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_4218"} +{"question": "Which papers discuss BERT and its use of a bidirectional transformer?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_4219"} +{"question": "Are there any works showing how transformers can approximate Turing machines with bounded computation time?", "answer": ["Statistically Meaningful Approximation: a Case Study on Approximating Turing Machines with Transformers"], "answer_arxiv_id": ["2107.13163"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4220"} +{"question": "What is the source of the MOT17 dataset used in Tracking-by-Detection?", "answer": ["MOT16: A Benchmark for Multi-Object Tracking"], "answer_arxiv_id": ["1603.00831"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_4221"} +{"question": "What are the studies contributing to the large-scale multimodal models?", "answer": ["Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "NExT-GPT: Any-to-Any Multimodal LLM", "MiniGPT-5: Interleaved Vision-and-Language Generation via Generative\n Vokens", "MiniGPT-v2: large language model as a unified interface for\n vision-language multi-task learning", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "MultiModal-GPT: A Vision and Language Model for Dialogue with Humans", "X-LLM: Bootstrapping Advanced Large Language Models by Treating\n Multi-Modalities as Foreign Languages", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation\n Models", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face", "MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action", "ViperGPT: Visual Inference via Python Execution for Reasoning", "GPT-4 Technical Report", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "NExT-GPT: Any-to-Any Multimodal LLM", "X-LLM: Bootstrapping Advanced Large Language Models by Treating\n Multi-Modalities as Foreign Languages"], "answer_arxiv_id": ["2306.15195", "2309.05519", "2310.02239", "2310.09478", "2304.15010", "2305.04790", "2305.04160", "2201.12086", "2303.04671", "2303.17580", "2303.11381", "2303.08128", "2303.08774", "2301.12597", "2304.15010", "2306.05424", "2306.02858", "2309.05519", "2305.04160"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_4222"} +{"question": "Any works about the application of DTW in representation learning?", "answer": ["Learning by Aligning Videos in Time", "Temporal Feature Alignment in Contrastive Self-Supervised Learning for Human Activity Recognition"], "answer_arxiv_id": ["2103.17260", "2210.03382"], "source_meta": {"published_time": "20230812"}, "qid": "AutoScholarQuery_train_4223"} +{"question": "What research conducted the unseen panorama reconstruction?", "answer": ["Learning to Look Around: Intelligently Exploring Unseen Environments for Unknown Tasks", "Im2Pano3D: Extrapolating 360 ° Structure and Semantics Beyond the Field of View", "Pathdreamer: A World Model for Indoor Navigation"], "answer_arxiv_id": ["1709.00507", "1712.04569", "2105.08756"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_4224"} +{"question": "Which papers have decoded mathematical operations such as modular addition and the greater-than operation as part of learnt mechanisms?", "answer": ["Progress measures for grokking via mechanistic interpretability", "How does GPT-2 compute greater-than?: Interpreting mathematical\n abilities in a pre-trained language model"], "answer_arxiv_id": ["2301.05217", "2305.00586"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_4225"} +{"question": "Could you provide me some studies about human body reconstruction that incorporated NeRF or its variations?", "answer": ["HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "LISA: Learning Implicit Shape and Appearance of Hands", "PersonNeRF: Personalized Reconstruction from Photo Collections"], "answer_arxiv_id": ["2201.04127", "2204.01695", "2302.08504"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4226"} +{"question": "Could you tell me examples of research that explored hierarchical modeling strategies for VideoQA tasks?", "answer": ["Hierarchical Conditional Relation Networks for Video Question Answering", "Location-aware Graph Convolutional Networks for Video Question Answering", "Hierarchical Object-oriented Spatio-Temporal Reasoning for Video Question Answering", "Video as Conditional Graph Hierarchy for Multi-Granular Question Answering", "Video Graph Transformer for Video Question Answering"], "answer_arxiv_id": ["2002.10698", "2008.09105", "2106.13432", "2112.06197", "2207.05342"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_4227"} +{"question": "Can you provide me some studies on training models on multiple modalities?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Attention Bottlenecks for Multimodal Fusion", "Perceiver IO: A General Architecture for Structured Inputs & Outputs", "Uni-Perceiver: Pre-training Unified Architecture for Generic Perception for Zero-shot and Few-shot Tasks", "CM3: A Causal Masked Multimodal Model of the Internet", "Flamingo: a Visual Language Model for Few-Shot Learning", "Language Is Not All You Need: Aligning Perception with Language Models", "ImageBind: One Embedding Space To Bind Them All"], "answer_arxiv_id": ["2103.00020", "2205.01917", "2107.00135", "2107.14795", "2112.01522", "2201.07520", "2204.14198", "2302.14045", "2305.05665"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_4228"} +{"question": "Are there any researches that focus on proving convergence to CE/CCE and sublinear individual regret within the full-information feedback setting?", "answer": ["Regret Minimization and Convergence to Equilibria in General-sum Markov Games"], "answer_arxiv_id": ["2207.14211"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_4229"} +{"question": "Can you provide studies that discuss the impact of domain shifts on language models, specifically on question answering and text classification?", "answer": ["MultiQA: An Empirical Investigation of Generalization and Transfer in Reading Comprehension", "MRQA 2019 Shared Task: Evaluating Generalization in Reading Comprehension", "Pretrained Transformers Improve Out-of-Distribution Robustness", "Types of Out-of-Distribution Texts and How to Detect Them"], "answer_arxiv_id": ["1905.13453", "1910.09753", "2004.06100", "2109.06827"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_4230"} +{"question": "What works are about the Correctness Ranking Loss method which regularizes the confidence based on the frequency of correct predictions?", "answer": ["Confidence-Aware Learning for Deep Neural Networks"], "answer_arxiv_id": ["2007.01458"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_4231"} +{"question": "What papers discuss the scaling laws and the use of larger corpora in training language models?", "answer": ["Scaling Laws for Neural Language Models", "Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2001.08361", "2203.15556"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_4232"} +{"question": "What papers propose spectral-based convolutional neural networks for learning on graphs?", "answer": ["Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering", "CayleyNets: Graph Convolutional Neural Networks with Complex Rational Spectral Filters", "Semi-Supervised Classification with Graph Convolutional Networks"], "answer_arxiv_id": ["1606.09375", "1705.07664", "1609.02907"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4233"} +{"question": "Can you tell me about research that studied and shown that small adversarial L2-perturbations can be found in random ReLU networks?", "answer": ["A Simple Explanation for the Existence of Adversarial Examples with Small Hamming Distance"], "answer_arxiv_id": ["1901.10861"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_4234"} +{"question": "Which studies suggested to select sub-networks and optimize their parameters in a sub-space to preserve learned knowledge and learn new sessions?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Deconstructing Lottery Tickets: Zeros, Signs, and the Supermask", "Discovering Neural Wirings", "What’s Hidden in a Randomly Weighted Neural Network?", "Meta-ticket: Finding optimal subnetworks for few-shot learning within randomly initialized neural networks"], "answer_arxiv_id": ["1803.03635", "1905.01067", "1906.00586", "1911.13299", "2205.15619"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_4235"} +{"question": "Can you provide studies that define options through a single eigenfunction of the Laplacian?", "answer": ["Discovering Options for Exploration by Minimizing Cover Time"], "answer_arxiv_id": ["1903.00606v2"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_4236"} +{"question": "Which study provided a benchmark for evaluating model accuracy and uncertainty under data distributional shifts?", "answer": ["Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift"], "answer_arxiv_id": ["1906.02530"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_4237"} +{"question": "What papers studied image restoration utilizing deep learning models?", "answer": ["Multi-Stage Progressive Image Restoration"], "answer_arxiv_id": ["2102.02808"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_4238"} +{"question": "What research argue about implicit bias towards rank minimization?", "answer": ["Gradient descent aligns the layers of deep linear networks", "Directional convergence and alignment in deep learning", "Implicit Regularization Towards Rank Minimization in ReLU Networks", "Implicit Regularization in Deep Matrix Factorization", "Implicit Regularization in Deep Learning May Not Be Explainable by Norms", "Towards Resolving the Implicit Bias of Gradient Descent for Matrix Factorization: Greedy Low-Rank Learning"], "answer_arxiv_id": ["1810.02032", "2006.06657", "2201.12760", "1905.13655", "2005.06398", "2012.09839"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_4239"} +{"question": "Which papers discussed the importance of negatives in training discriminative models?", "answer": ["Contrastive Learning with Hard Negative Samples", "Focal Loss for Dense Object Detection"], "answer_arxiv_id": ["2010.04592", "1708.02002"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_4240"} +{"question": "Could you provide me research examples that introduced bisimulation and bisimulation metrics to learn state abstractions in reinforcement learning?", "answer": ["Scalable methods for computing state similarity in deterministic Markov Decision Processes", "Learning Invariant Representations for Reinforcement Learning without Reconstruction"], "answer_arxiv_id": ["1911.09291", "2006.10742"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_4241"} +{"question": "Could you provide me some papers where MCTS-based methods have been applied for complex tasks?", "answer": ["Online and Offline Reinforcement Learning by Planning with a Learned Model"], "answer_arxiv_id": ["2104.06294"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_4242"} +{"question": "What works use customization techniques for subject-driven image editing?", "answer": ["Custom-Edit: Text-Guided Image Editing with Customized Diffusion Models", "Photoswap: Personalized Subject Swapping in Images"], "answer_arxiv_id": ["2305.15779", "2305.18286"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4243"} +{"question": "What works proposed ghost clipping in DP optimizer that can improve both time and space complexity if the feature dimension is small?", "answer": ["Efficient Per-Example Gradient Computations", "Large Language Models Can Be Strong Differentially Private Learners", "Scalable and Efficient Training of Large Convolutional Neural Networks with Differential Privacy"], "answer_arxiv_id": ["1510.01799", "2110.05679", "2205.10683"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_4244"} +{"question": "Can you recommend a comprehensive survey on LLM alignment?", "answer": ["Aligning Large Language Models with Human: A Survey"], "answer_arxiv_id": ["2307.12966"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_4245"} +{"question": "Could you provide me some studies that used gradient-based methods for automatic prompt alterations in LLMs?", "answer": ["Universal and Transferable Adversarial Attacks on Aligned Language\n Models"], "answer_arxiv_id": ["2307.15043"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_4246"} +{"question": "Could you provide me some works about decomposition-based methods that have emerged for dynamic scenes?", "answer": ["HexPlane: A Fast Representation for Dynamic Scenes", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "Tensor4D : Efficient Neural 4D Decomposition for High-fidelity Dynamic\n Reconstruction and Rendering"], "answer_arxiv_id": ["2301.09632", "2301.10241", "2211.11610"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_4247"} +{"question": "Which research proposed a shaping reward based on MI between agents’ transitions in the MARL setting?", "answer": ["Influence-Based Multi-Agent Exploration"], "answer_arxiv_id": ["1910.05512"], "source_meta": {"published_time": "20230319"}, "qid": "AutoScholarQuery_train_4248"} +{"question": "What studies explore statistical efficiency in hybrid RL?", "answer": ["Leveraging Offline Data in Online Reinforcement Learning"], "answer_arxiv_id": ["2211.04974"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_4249"} +{"question": "Could you provide the work that targets on training the U-Net model end-to-end specifically for multimodal diffusion?", "answer": ["Kosmos-G: Generating Images in Context with Multimodal Large Language\n Models"], "answer_arxiv_id": ["2310.02992"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_4250"} +{"question": "What research papers use contrastive learning strategies like SimCLR and CLIP?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2002.05709", "2103.00020"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4251"} +{"question": "Which paper projects the LiDAR points and images into the BEV space to align the LiDAR feature and image feature?", "answer": ["BEVDistill: Cross-Modal BEV Distillation for Multi-View 3D Object Detection"], "answer_arxiv_id": ["2211.09386"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_4252"} +{"question": "Which work proposes an iterative painting scheme to enable text-guided editing of 3D shape textures?", "answer": ["TEXTure: Text-Guided Texturing of 3D Shapes"], "answer_arxiv_id": ["2302.01721"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4253"} +{"question": "Which works demonstrated neural collapse as the global optimum of balanced training using cross entropy (CE) loss?", "answer": ["An Unconstrained Layer-Peeled Perspective on Neural Collapse", "Neural Collapse with Cross-Entropy Loss", "A Geometric Analysis of Neural Collapse with Unconstrained Features", "On the emergence of simplex symmetry in the final and penultimate layers\n of neural network classifiers"], "answer_arxiv_id": ["2110.02796", "2012.08465", "2105.02375", "2012.05420"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_4254"} +{"question": "Which studies deal with unstructured pruning of Vision Transformers?", "answer": ["Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Rigging the Lottery: Making All Tickets Winners"], "answer_arxiv_id": ["2106.04533", "1911.11134"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_4255"} +{"question": "Which papers used graph-based representations of molecules with the development of GNNs in molecular property prediction?", "answer": ["PotentialNet for Molecular Property Prediction", "How Powerful are Graph Neural Networks?", "Molformer: Motif-based Transformer on 3D Heterogeneous Molecular Graphs", "Equivariant Subgraph Aggregation Networks", "Molecular Representation Learning via Heterogeneous Motif Graph Neural Networks", "A graph representation of molecular ensembles for polymer property prediction", "On the Bottleneck of Graph Neural Networks and its Practical Implications"], "answer_arxiv_id": ["1803.04465", "1810.00826", "2110.01191", "2110.02910", "2202.00529", "2205.08619v1", "2006.05205"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_4256"} +{"question": "Which datasets were collected for acoustic research?", "answer": ["SoundCam: A Dataset for Finding Humans Using Room Acoustics"], "answer_arxiv_id": ["2311.03517"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_4257"} +{"question": "Which work developed a method for representing DAGs as sequences of node strings of the node type and adjacency vector of each node, and employs a GRU-based RNN for learning the DAG representation?", "answer": ["Generating Sentences from a Continuous Space"], "answer_arxiv_id": ["1511.06349"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_4258"} +{"question": "Which work discussed the application of DistRL in risk-sensitive RL?", "answer": ["Distributional Reinforcement Learning with Quantile Regression"], "answer_arxiv_id": ["1710.10044"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4259"} +{"question": "Could you provide research that improved the efficiency and scalability of neural radiance fields?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation"], "answer_arxiv_id": ["1901.05103"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_4260"} +{"question": "Which studies used the finite-difference computation for regularization?", "answer": ["The Hessian Penalty: A Weak Prior for Unsupervised Disentanglement", "Scaleable input gradient regularization for adversarial robustness"], "answer_arxiv_id": ["2008.10599", "1905.11468"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_4261"} +{"question": "Which papers propose automatic debiasing methods in the context of dataset biases?", "answer": ["Towards Debiasing NLU Models from Unknown Biases", "Learning from others' mistakes: Avoiding dataset biases without modeling\n them", "Feature-Level Debiased Natural Language Understanding"], "answer_arxiv_id": ["2009.12303", "2012.01300", "2212.05421"], "source_meta": {"published_time": "20240823"}, "qid": "AutoScholarQuery_train_4262"} +{"question": "Which works used tri-plane representations of 3D scenes created by a generator-discriminator framework based on StyleGAN2?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "EpiGRAF: Rethinking training of 3D GANs"], "answer_arxiv_id": ["2112.07945", "2206.10535"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_4263"} +{"question": "Which works studied adaptation as a component of continual learning?", "answer": ["Towards Continual Reinforcement Learning: A Review and Perspectives", "Never Stop Learning: The Effectiveness of Fine-Tuning in Robotic Reinforcement Learning"], "answer_arxiv_id": ["2012.13490", "2004.10190"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_4264"} +{"question": "Can you tell me which research proposes using LiDAR as supervision for street view multi-view reconstruction in NeRF?", "answer": ["StreetSurf: Extending Multi-view Implicit Surface Reconstruction to\n Street Views"], "answer_arxiv_id": ["2306.04988"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_4265"} +{"question": "Could you point out researches that focus on neural point processes (NPPs)?", "answer": ["Deep Reinforcement Learning of Marked Temporal Point Processes", "Self-Attentive Hawkes Process"], "answer_arxiv_id": ["1805.09360", "1907.07561"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_4266"} +{"question": "Which paper discussed a high-resolution 3D human generation within a clean framework?", "answer": ["EVA3D: Compositional 3D Human Generation from 2D Image Collections"], "answer_arxiv_id": ["2210.04888"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4267"} +{"question": "What works demonstrate the sample efficiency of model-based methods?", "answer": ["Model Based Reinforcement Learning for Atari", "When to Trust Your Model: Model-Based Policy Optimization"], "answer_arxiv_id": ["1903.00374", "1906.08253"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_4268"} +{"question": "What studies have proposed methods to improve the efficiency of SAM?", "answer": ["Fast Segment Anything", "Faster Segment Anything: Towards Lightweight SAM for Mobile Applications"], "answer_arxiv_id": ["2306.12156", "2306.14289"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4269"} +{"question": "Which study created a system for time series data selection?", "answer": ["TODS: An Automated Time Series Outlier Detection System"], "answer_arxiv_id": ["2009.09822"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4270"} +{"question": "Can you provide works that established a gradual-variation bound for optimistic FTRL?", "answer": ["A Modular Analysis of Adaptive (Non-)Convex Optimization: Optimism, Composite Objectives, and Variational Bounds"], "answer_arxiv_id": ["1709.02726"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_4271"} +{"question": "What studies develop methods relying on slots for semantic representation in contrastive learning?", "answer": ["Self-Supervised Visual Representation Learning with Semantic Grouping", "Bridging the Gap to Real-World Object-Centric Learning"], "answer_arxiv_id": ["2205.15288", "2209.14860"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_4272"} +{"question": "What are some studies showing large language models' capabilities in performing a diverse range of tasks?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "Language Models are Few-Shot Learners", "GPT-4 Technical Report", "PaLM: Scaling Language Modeling with Pathways", "PaLM 2 Technical Report"], "answer_arxiv_id": ["1910.10683", "2005.14165", "2303.08774", "2204.02311", "2305.10403"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_4273"} +{"question": "Which works combine DNNs with voxel grid representations to minimize discretization artifacts in view synthesis?", "answer": ["A Neural Rendering Framework for Free-Viewpoint Relighting", "Neural Volumes: Learning Dynamic Renderable Volumes from Images", "DeepVoxels: Learning Persistent 3D Feature Embeddings"], "answer_arxiv_id": ["1911.11530", "1906.07751", "1812.01024v2"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_4274"} +{"question": "Which studies have showed that orthogonal gradient descent (OGD) gives a tighter generalization bound than SGD?", "answer": ["Generalisation Guarantees for Continual Learning with Orthogonal Gradient Descent"], "answer_arxiv_id": ["2006.11942"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_4275"} +{"question": "Which studies employed retrieval-augmented data for pre-training of LMs?", "answer": ["Atlas: Few-shot Learning with Retrieval Augmented Language Models", "Improving language models by retrieving from trillions of tokens"], "answer_arxiv_id": ["2208.03299", "2112.04426"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_4276"} +{"question": "What is the work that introduces an adversarial method to dynamically generate the most challenging blending choices for deepfake detection?", "answer": ["Self-supervised Learning of Adversarial Example: Towards Good\n Generalizations for Deepfake Detection"], "answer_arxiv_id": ["2203.12208"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_4277"} +{"question": "Can you provide examples of research papers where knowledge distillation is used to transfer knowledge from larger models to smaller models?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_4278"} +{"question": "Could you name the studies which explored MARL communication under mixed-motive scenarios?", "answer": ["Social Influence as Intrinsic Motivation for Multi-Agent Deep Reinforcement Learning"], "answer_arxiv_id": ["1810.08647"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_4279"} +{"question": "What are the extensions to the standard metrics for the evaluation of image captioning models?", "answer": ["CIDEr-R: Robust Consensus-based Image Description Evaluation", "JaSPICE: Automatic Evaluation Metric Using Predicate-Argument Structures\n for Image Captioning Models"], "answer_arxiv_id": ["2109.13701", "2311.04192"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_4280"} +{"question": "Which study looked into learning curves in a standard or purely supervised-learning environment?", "answer": ["The Shape of Learning Curves: a Review"], "answer_arxiv_id": ["2103.10948"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_4281"} +{"question": "What papers demonstrate that vision-language contrastive learning can generate transferable features to downstream tasks in a multi-modal learning scenario?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2110.05208", "2102.05918"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_4282"} +{"question": "What papers are related to extending motif discovery to hypergraphs?", "answer": ["Hypergraph Motifs: Concepts, Algorithms, and Discoveries", "THyMe+: Temporal Hypergraph Motifs and Fast Algorithms for Exact Counting"], "answer_arxiv_id": ["2003.01853", "2109.08341"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_4283"} +{"question": "What works propose methods to increase accuracy in the context of denoised smoothing?", "answer": ["Denoised Smoothing: A Provable Defense for Pretrained Classifiers"], "answer_arxiv_id": ["2003.01908"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_4284"} +{"question": "Which works have provided solutions for general nonlinear mixing functions in independent component analysis?", "answer": ["Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICA", "Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learning", "Variational Autoencoders and Nonlinear ICA: A Unifying Framework", "ICE-BeeM: Identifiable Conditional Energy-Based Deep Models Based on Nonlinear ICA"], "answer_arxiv_id": ["1605.06336", "1805.08651", "1907.04809", "2002.11537"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_4285"} +{"question": "Which works focus on online time series forecasting that updates their models?", "answer": ["DDG-DA: Data Distribution Generation for Predictable Concept Drift Adaptation", "Generalizing to Evolving Domains with Latent Structure-Aware Sequential Autoencoder", "Learning Fast and Slow for Online Time Series Forecasting"], "answer_arxiv_id": ["2201.04038", "2205.07649", "2202.11672"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_4286"} +{"question": "Could you provide me some studies about the solution of high-dimensional Black-Scholes and Hamilton-Jacobi-Bellman equations?", "answer": ["DGM: A deep learning algorithm for solving partial differential equations", "Solving High-Dimensional Partial Differential Equations Using Deep Learning"], "answer_arxiv_id": ["1708.07469", "1707.02568"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_4287"} +{"question": "What works propose denoising in the latent space to reduce complexity?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4288"} +{"question": "Can you give examples of research that use generative methods in zero-shot generalization to generate fake visual features for training classifiers?", "answer": ["Feature Generating Networks for Zero-Shot Learning", "Dual Adversarial Semantics-Consistent Network for Generalized Zero-Shot Learning", "Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders", "A Meta-Learning Framework for Generalized Zero-Shot Learning", "Zero-Shot Learning from Adversarial Feature Residual to Compact Visual Feature"], "answer_arxiv_id": ["1712.00981", "1907.05570", "1812.01784", "1909.04344", "2008.12962"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_4289"} +{"question": "What are some studies that involved curiosity or similarity-driven exploration?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction", "Unsupervised Control through Non-Parametric Discriminative Rewards"], "answer_arxiv_id": ["1705.05363", "1811.11359"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_4290"} +{"question": "Could you provide me some studies on deep models for permutation invariance?", "answer": ["Deep Sets"], "answer_arxiv_id": ["1703.06114"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_4291"} +{"question": "Which papers use a set of latent codes to represent similar local geometry patterns across objects in 4D dynamics?", "answer": ["3DShape2VecSet: A 3D Shape Representation for Neural Fields and\n Generative Diffusion Models"], "answer_arxiv_id": ["2301.11445"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_4292"} +{"question": "What studies delve into vision and language navigation for enhancing robot navigation?", "answer": ["Vision-and-Language Navigation: Interpreting visually-grounded\n navigation instructions in real environments"], "answer_arxiv_id": ["1711.07280"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_4293"} +{"question": "Is there any work that focuses solely on the deterministic setting?", "answer": ["Accelerated Primal-Dual Gradient Method for Smooth and Convex-Concave Saddle-Point Problems with Bilinear Coupling", "Lifted Primal-Dual Method for Bilinearly Coupled Smooth Minimax Optimization"], "answer_arxiv_id": ["2112.15199", "2201.07427v1"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_4294"} +{"question": "What are some of the research papers on designing an efficient transformer layer for real-time tracking?", "answer": ["Efficient Visual Tracking with Exemplar Transformers"], "answer_arxiv_id": ["2112.09686"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_4295"} +{"question": "In what research is sequence-to-sequence transformer models used for image-text data?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_4296"} +{"question": "What is the work that also considers depth separation similar to this research and showed that GD with a certain three-layer network can learn the ball indicator which cannot be approximated by any two-layer network?", "answer": ["Optimization-Based Separations for Neural Networks"], "answer_arxiv_id": ["2112.02393"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_4297"} +{"question": "Which works developed a method for incorporating skip connections which seem to be ad-hoc, in over-smoothing graph learning?", "answer": ["Representation Learning on Graphs with Jumping Knowledge Networks", "Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "Simple and Deep Graph Convolutional Networks"], "answer_arxiv_id": ["1806.03536v2", "1810.05997", "2007.02133"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_4298"} +{"question": "Which studies have been done on topology-aware feature imputation over missing features?", "answer": ["Graph Convolutional Networks for Graphs Containing Missing Features", "Incomplete Graph Representation and Learning via Partial Graph Neural Networks", "On the Unreasonable Effectiveness of Feature Propagation in Learning on Graphs with Missing Node Features"], "answer_arxiv_id": ["2007.04583", "2003.10130", "2111.12128"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_4299"} +{"question": "What papers focus on the evaluation of language models in summarization tasks?", "answer": ["Teaching Machines to Read and Comprehend"], "answer_arxiv_id": ["1506.03340"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_4300"} +{"question": "What works used learning methods to develop large neighborhood search (LNS) heuristics?", "answer": ["Accelerating Primal Solution Findings for Mixed Integer Programs Based on Solution Prediction", "A General Large Neighborhood Search Framework for Solving Integer Linear Programs"], "answer_arxiv_id": ["1906.09575", "2004.00422"], "source_meta": {"published_time": "20230520"}, "qid": "AutoScholarQuery_train_4301"} +{"question": "In what papers the researcher used Bayesian neural networks as alternative methods for Bayesian optimization?", "answer": ["Scalable Bayesian Optimization Using Deep Neural Networks"], "answer_arxiv_id": ["1502.05700"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_train_4302"} +{"question": "Are there any previous works that proposed using biased task-specific designs?", "answer": ["COGS: A Compositional Generalization Challenge Based on Semantic Interpretation", "Measuring abstract reasoning in neural networks"], "answer_arxiv_id": ["2010.05465", "1807.04225"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_4303"} +{"question": "Can you provide me some works about Multi-armed bandit problems (MAB)?", "answer": ["Regret Analysis of Stochastic and Nonstochastic Multi-armed Bandit Problems", "Introduction to Multi-Armed Bandits"], "answer_arxiv_id": ["1204.5721v2", "1904.07272"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4304"} +{"question": "Which papers discuss the introduction of asynchrony to optimization in deep neural network training?", "answer": ["Local SGD Converges Fast and Communicates Little", "Don’t Use Large Mini-Batches, Use Local SGD", "Deep learning with Elastic Averaging SGD", "SlowMo: Improving Communication-Efficient Distributed SGD with Slow Momentum", "Federated Optimization in Heterogeneous Networks", "Breaking (Global) Barriers in Parallel Stochastic Optimization with Wait-Avoiding Group Averaging", "Collaborative Deep Learning in Fixed Topology Networks", "Stochastic Gradient Push for Distributed Deep Learning"], "answer_arxiv_id": ["1805.09767", "1808.07217", "1412.6651", "1910.00643", "1812.06127", "2005.00124", "1706.07880", "1811.10792v3"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_4305"} +{"question": "Which papers propose alternate formulations of diffusion model to reduce the length of sampling chains?", "answer": ["Denoising Diffusion Implicit Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Learning to Efficiently Sample from Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "On Fast Sampling of Diffusion Probabilistic Models"], "answer_arxiv_id": ["2010.02502", "2206.00364", "2106.03802", "2011.13456", "2106.00132"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_4306"} +{"question": "Which research attempts to model non-rigid deformations caused by a moving person with a thin plate spline-based transformer?", "answer": ["Adversarial T-shirt! Evading Person Detectors in A Physical World"], "answer_arxiv_id": ["1910.11099"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_4307"} +{"question": "In which paper is the slot attention mechanism first proposed for object-centric learning tasks?", "answer": ["Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["2006.15055"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_4308"} +{"question": "Could you provide me some studies about text-to-image diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2205.11487", "2204.06125"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_4309"} +{"question": "Which study shows that representation-based similarity measures are not always reliable for testing functional differences in models?", "answer": ["Grounding Representation Similarity with Statistical Testing"], "answer_arxiv_id": ["2108.01661"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_4310"} +{"question": "Are there any papers indicating that the representation power of the treatment indicator can be significantly diluted when the dimension of the covariates becomes high?", "answer": ["Bayesian Inference of Individualized Treatment Effects using Multi-task Gaussian Processes", "Deep Counterfactual Networks with Propensity-Dropout"], "answer_arxiv_id": ["1704.02801", "1706.05966"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_4311"} +{"question": "Are there any contemporary studies autoring on the problem of grounding using mechanistic interpretability methods?", "answer": ["Characterizing Mechanisms for Factual Recall in Language Models"], "answer_arxiv_id": ["2310.15910"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4312"} +{"question": "Which work applies the tree-sliced variant of GSW in an intrinsic manner?", "answer": ["Tree-Sliced Variants of Wasserstein Distances"], "answer_arxiv_id": ["1902.00342v3"], "source_meta": {"published_time": "20201028"}, "qid": "AutoScholarQuery_train_4313"} +{"question": "Which research introduced prototypical contrastive learning?", "answer": ["Prototypical Contrastive Learning of Unsupervised Representations"], "answer_arxiv_id": ["2005.04966"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_4314"} +{"question": "What references explored the method of style transfer in language models?", "answer": ["Extracting Latent Steering Vectors from Pretrained Language Models"], "answer_arxiv_id": ["2205.05124"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_4315"} +{"question": "What is the study that introduces the Squeeze-and-Excitation Networks?", "answer": ["Squeeze-and-Excitation Networks"], "answer_arxiv_id": ["1709.01507"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_4316"} +{"question": "Which work suggests that IRL methods can in theory exceed longer task horizons by overcoming compounding errors through environment interactions?", "answer": ["Error Bounds of Imitating Policies and Environments"], "answer_arxiv_id": ["2010.11876v1"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4317"} +{"question": "Could you provide me some works that exploited the translation of decision trees into MLP?", "answer": ["Neural Random Forests"], "answer_arxiv_id": ["1604.07143v2"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_4318"} +{"question": "Could you provide me some studies about generative video editing methods that use temporal-aware cross-frame attention techniques?", "answer": ["ControlVideo: Training-free Controllable Text-to-Video Generation", "Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Zero-Shot Video Editing Using Off-The-Shelf Image Diffusion Models", "Pix2Video: Video Editing using Image Diffusion", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "TokenFlow: Consistent Diffusion Features for Consistent Video Editing"], "answer_arxiv_id": ["2305.13077", "2306.07954", "2303.09535", "2303.17599", "2303.12688", "2303.13439", "2307.10373"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_4319"} +{"question": "Which research concerned the realizable setting of online regression with the absolute loss?", "answer": ["Fast Rates for Nonparametric Online Learning: From Realizability to Learning in Games"], "answer_arxiv_id": ["2111.08911"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_4320"} +{"question": "What are some of the studies that set the off-the-shelf 2D body keypoints as a condition to animate the source person image?", "answer": ["Everybody Dance Now", "A Variational U-Net for Conditional Appearance and Shape Generation", "Soft-Gated Warping-GAN for Pose-Guided Person Image Synthesis", "Pose Guided Person Image Generation", "Combining Attention with Flow for Person Image Synthesis", "Deep Image Spatial Transformation for Person Image Generation"], "answer_arxiv_id": ["1808.07371", "1804.04694", "1810.11610", "1705.09368", "2108.01823", "2003.00696"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_4321"} +{"question": "Which works proposed methods for handling covariate mismatch in the context of multi-site causal inference?", "answer": ["Combining multiple observational data sources to estimate causal effects", "Multi-Source Causal Inference Using Control Variates", "Efficient Generalization and Transportation"], "answer_arxiv_id": ["1801.00802", "2103.16689", "2302.00092"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_4322"} +{"question": "Which paper extended the application of Denoising Diffusion Probabilistic Models to noisy non-linear inverse problems?", "answer": ["Parallel Diffusion Models of Operator and Image for Blind Inverse\n Problems"], "answer_arxiv_id": ["2211.10656"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4323"} +{"question": "Which works demonstrated the effectiveness of training on billions of samples in generalizing on a wide range of prompts for text-to-image generation?", "answer": ["Zero-Shot Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2102.12092", "2204.06125"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4324"} +{"question": "Any studies propose to waive the global model by adopting multi-task learning or hyper-network frameworks in federated learning?", "answer": ["Federated Multi-Task Learning", "Personalized Federated Learning using Hypernetworks"], "answer_arxiv_id": ["1705.10467", "2103.04628"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_4325"} +{"question": "Could you provide some works focused on enhancing generation fidelity and optimization stability or explore more application scenarios in the context of 3D generation?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "TextMesh: Generation of Realistic 3D Meshes From Text Prompts", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "Points-to-3D: Bridging the Gap between Sparse Points and\n Shape-Controllable Text-to-3D Generation", "FocalDreamer: Text-driven 3D Editing via Focal-fusion Assembly", "DreamTime: An Improved Optimization Strategy for Diffusion-Guided 3D\n Generation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "DreamEditor: Text-Driven 3D Scene Editing with Neural Fields", "Text-To-4D Dynamic Scene Generation", "DreamBooth3D: Subject-Driven Text-to-3D Generation"], "answer_arxiv_id": ["2211.10440", "2304.12439", "2303.13873", "2307.13908", "2308.10608", "2306.12422", "2211.07600", "2309.16653", "2305.16213", "2306.13455", "2301.11280", "2303.13508"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_4326"} +{"question": "What works have contributed to the study of adapting a given large model for downstream tasks in the field of transfer learning?", "answer": ["Learning multiple visual domains with residual adapters", "Learning Aligned Cross-Modal Representations from Weakly Aligned Data", "End-to-End Object Detection with Transformers", "Cross-domain Few-shot Learning with Task-specific Adapters", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Parameter-efficient Multi-task Fine-tuning for Transformers via Shared Hypernetworks", "Visual Prompt Tuning", "LST: Ladder Side-Tuning for Parameter and Memory Efficient Transfer Learning"], "answer_arxiv_id": ["1705.08045", "1607.07295", "2005.12872", "2107.00358", "2101.00190", "2106.04489", "2203.12119", "2206.06522"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_4327"} +{"question": "Could you provide me a study about building policy with the reverse chain of a conditional diffusion model?", "answer": ["Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning"], "answer_arxiv_id": ["2208.06193"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4328"} +{"question": "Could you give me some examples of researches that explored the use of transformers for image restoration?", "answer": ["SwinIR: Image Restoration Using Swin Transformer", "Restormer: Efficient Transformer for High-Resolution Image Restoration", "BSRT: Improving Burst Super-Resolution with Swin Transformer and Flow-Guided Deformable Alignment"], "answer_arxiv_id": ["2108.10257", "2111.09881", "2204.08332"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_4329"} +{"question": "Which works propose to fuse semantic features from pre-trained ViTs into a neural reconstruction?", "answer": ["Decomposing NeRF for Editing via Feature Field Distillation", "Semantic Attention Flow Fields for Monocular Dynamic Scene Decomposition"], "answer_arxiv_id": ["2205.15585", "2303.01526"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_4330"} +{"question": "What works follow the approach of self-training, where pseudo labels for unlabeled data are generated by a smaller teacher model?", "answer": ["Self-training with Noisy Student improves ImageNet classification", "Self-training Improves Pre-training for Natural Language Understanding"], "answer_arxiv_id": ["1911.04252", "2010.02194"], "source_meta": {"published_time": "20220814"}, "qid": "AutoScholarQuery_train_4331"} +{"question": "What work initially introduced the concept of working with the merit distribution and comparing individuals by the probability that one is more qualified than another?", "answer": ["Fairness in Ranking under Uncertainty"], "answer_arxiv_id": ["2107.06720v2"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_4332"} +{"question": "Who proposed a disentangled conditional diffusion model for MCSR?", "answer": ["DisC-Diff: Disentangled Conditional Diffusion Model for Multi-Contrast\n MRI Super-Resolution"], "answer_arxiv_id": ["2303.13933"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_4333"} +{"question": "Which datasets have been developed for understanding the driving scene in autonomous driving?", "answer": ["GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering", "Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations", "The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale", "Visual Relationship Detection with Language Priors"], "answer_arxiv_id": ["1902.09506", "1602.07332", "1811.00982v2", "1608.00187"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_4334"} +{"question": "Could you provide me some works about designing methods that are robust to misspecification?", "answer": ["A General Framework for Updating Belief Distributions", "Generalized Variational Inference: Three arguments for deriving new Posteriors"], "answer_arxiv_id": ["1306.6430", "1904.02063v4"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4335"} +{"question": "Are there any works that introduced hybrid-based methods in 3D object detection?", "answer": ["PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection", "STD: Sparse-to-Dense 3D Object Detector for Point Cloud", "HVPR: Hybrid Voxel-Point Representation for Single-stage 3D Object\n Detection"], "answer_arxiv_id": ["1912.13192", "1907.10471", "2104.00902"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4336"} +{"question": "Are there any works that emphasized the benefits of input-dependent dynamic quantization?", "answer": ["Dynamic Dual Trainable Bounds for Ultra-low Precision Super-Resolution Networks", "INSTA-BNN: Binary Neural Network with INSTAnce-aware Threshold", "CADyQ: Content-Aware Dynamic Quantization for Image Super-Resolution"], "answer_arxiv_id": ["2203.03844", "2204.07439", "2207.10345v3"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_4337"} +{"question": "What works describe the bandit setting in relation to upper confidence bounds, which originate from Bayesian optimization?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design", "Deep reinforced active learning for multi-class image classification", "Learning Active Learning from Data", "Meta-Learning Transferable Active Learning Policies by Deep Reinforcement Learning"], "answer_arxiv_id": ["0912.3995v4", "2206.13391", "1703.03365", "1806.04798"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4338"} +{"question": "Could you provide me with studies that have analyzed the impact of causal fairness notions on downstream utilities?", "answer": ["Causal Conceptions of Fairness and their Consequences"], "answer_arxiv_id": ["2207.05302"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_4339"} +{"question": "What studies discuss the stability issues in training a signed distance function (SDF) for computer graphics applications?", "answer": ["Spelunking the Deep: Guaranteed Queries on General Neural Implicit Surfaces via Range Analysis"], "answer_arxiv_id": ["2202.02444"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_4340"} +{"question": "Which studies investigated using server-side hypernetwork or learning multiple global models in personalized Federated Learning?", "answer": ["Personalized Federated Learning using Hypernetworks", "Federated Multi-Task Learning under a Mixture of Distributions"], "answer_arxiv_id": ["2103.04628", "2108.10252v4"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_4341"} +{"question": "Which studies investigated the relationship between ImageNet and transfer accuracy for self-supervised networks?", "answer": ["How Well Do Self-Supervised Models Transfer?", "Contrasting Contrastive Self-Supervised Representation Learning Pipelines", "Diverse Imagenet Models Transfer Better"], "answer_arxiv_id": ["2011.13377", "2103.14005", "2204.09134"], "source_meta": {"published_time": "20230111"}, "qid": "AutoScholarQuery_train_4342"} +{"question": "What works introduced state-of-the-art approaches for synthetic time-series generation?", "answer": ["Generating multivariate time series with COmmon Source CoordInated GAN (COSCI-GAN)", "Towards Generating Real-World Time Series Data", "Deep Latent State Space Models for Time-Series Generation"], "answer_arxiv_id": ["2205.13741", "2111.08386", "2212.12749v3"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_4343"} +{"question": "What research papers discuss the technique of Sparsification in the context of improving learning on graphs?", "answer": ["Inductive Representation Learning on Large Graphs"], "answer_arxiv_id": ["1706.02216"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_4344"} +{"question": "What research focused on generating text from abstract meaning representations?", "answer": ["Towards a Decomposable Metric for Explainable Evaluation of Text\n Generation from AMR"], "answer_arxiv_id": ["2008.08896"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_4345"} +{"question": "Could you provide me with some studies about critical points for different types of neural network architectures?", "answer": ["Geometry of Linear Convolutional Networks", "Function Space and Critical Points of Linear Convolutional Networks"], "answer_arxiv_id": ["2108.01538", "2304.05752"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_4346"} +{"question": "Are there any theoretical analysis of how related dynamics affect learning in the supervised setting?", "answer": ["Towards Explaining the Regularization Effect of Initial Large Learning Rate in Training Neural Networks", "The Pitfalls of Simplicity Bias in Neural Networks"], "answer_arxiv_id": ["1907.04595", "2006.07710"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_4347"} +{"question": "Which study proposes a stagewise pre-training strategy that utilizes frozen attention blocks pretrained by image-only data to train the language expert?", "answer": ["VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts"], "answer_arxiv_id": ["2111.02358"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_4348"} +{"question": "What papers have focused on the effects of SGD hyperparameters for the adversarial robustness of the resulting models?", "answer": ["Hessian-based Analysis of Large Batch Training and Robustness to Adversaries", "How do SGD hyperparameters in natural training affect adversarial robustness?"], "answer_arxiv_id": ["1802.08241", "2006.11604"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_4349"} +{"question": "Could you provide me some studies that apply deep generative models to computational biological sequence design?", "answer": ["Generating and designing DNA with deep generative models"], "answer_arxiv_id": ["1712.06148"], "source_meta": {"published_time": "20221015"}, "qid": "AutoScholarQuery_train_4350"} +{"question": "Which works use geometric methods in the field of RIR synthesis?", "answer": ["SoundSpaces: Audio-Visual Navigation in 3D Environments", "SoundSpaces 2.0: A Simulation Platform for Visual-Acoustic Learning"], "answer_arxiv_id": ["1912.11474", "2206.08312"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_4351"} +{"question": "What research has recognized the need for an axiomatisation of causality based on measure-theoretic probability theory?", "answer": ["Subjectivity, Bayesianism, and Causality", "Causal Inference Theory with Information Dependency Models", "Causal Models on Probability Spaces"], "answer_arxiv_id": ["1407.4139", "2108.03099", "1907.01672"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_4352"} +{"question": "Are there any works that prove similar upper bounds for the last-iterate 𝒪​(1/N) rate without relying on the Lipschitzness?", "answer": ["Last-Iterate Convergence of Optimistic Gradient Method for Monotone Variational Inequalities", "Tight Last-Iterate Convergence of the Extragradient and the Optimistic Gradient Descent-Ascent Algorithm for Constrained Monotone Variational Inequalities"], "answer_arxiv_id": ["2205.08446", "2204.09228"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_4353"} +{"question": "What papers about solving ordinary differential equations (ODEs) with fewer steps for fast sampling of DDIM?", "answer": ["Pseudo Numerical Methods for Diffusion Models on Manifolds", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps"], "answer_arxiv_id": ["2202.09778", "2206.00927"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_4354"} +{"question": "What works reported the advances in generative AI using large image datasets like LAION-5B?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_4355"} +{"question": "What research focused on learning single index models where the link function is ReLU?", "answer": ["Learning ReLUs via Gradient Descent"], "answer_arxiv_id": ["1705.04591"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_4356"} +{"question": "What are some papers that draw inspirations from machine learning techniques for preconditioner design?", "answer": ["Combining Differentiable PDE Solvers and Graph Neural Networks for Fluid Flow Prediction", "Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers", "Fourier Neural Operator for Parametric Partial Differential Equations", "Multipole Graph Neural Operator for Parametric Partial Differential Equations"], "answer_arxiv_id": ["2007.04439", "2007.00016", "2010.08895", "2006.09535"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4357"} +{"question": "Which papers introduced logit-level knowledge distillation?", "answer": ["On the Efficacy of Knowledge Distillation", "Snapshot Distillation: Teacher-Student Optimization in One Generation", "Improved Knowledge Distillation via Teacher Assistant", "Rethinking Knowledge Distillation via Cross-Entropy"], "answer_arxiv_id": ["1910.01348", "1812.00123", "1902.03393", "2208.10139"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4358"} +{"question": "Could you provide me some studies where a motion imitator, UHC, has been utilized for egocentric and third-person scene-aware pose estimation?", "answer": ["Dynamics-Regulated Kinematic Policy for Egocentric Pose Estimation", "Embodied Scene-aware Human Pose Estimation"], "answer_arxiv_id": ["2106.05969", "2206.09106"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_4359"} +{"question": "What works improved the performance and usability of the knowledge amalgamation model merging strategy?", "answer": ["Knowledge Amalgamation from Heterogeneous Networks by Common Feature\n Learning", "Customizing Student Networks From Heterogeneous Teachers via Adaptive\n Knowledge Amalgamation", "Data-Free Knowledge Amalgamation via Group-Stack Dual-GAN", "Knowledge Amalgamation for Object Detection with Transformers", "Contrastive Knowledge Amalgamation for Unsupervised Image Classification", "Deep Graph Reprogramming"], "answer_arxiv_id": ["1906.10546", "1908.07121", "2003.09088", "2203.03187", "2307.14781", "2304.14593"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_4360"} +{"question": "What works proposed to add constraints and regularization to the model for obtaining unbiased representations?", "answer": ["Machine Learning Models that Remember Too Much", "Putting Fairness Principles into Practice: Challenges, Metrics, and Improvements", "RUBi: Reducing Unimodal Biases for Visual Question Answering", "EnD: Entangling and Disentangling deep representations for bias correction"], "answer_arxiv_id": ["1709.07886", "1901.04562", "1906.10169", "2103.02023"], "source_meta": {"published_time": "20221110"}, "qid": "AutoScholarQuery_train_4361"} +{"question": "Can you list the studies that discussed the usage of contrastive learning as an effective distillation objective?", "answer": ["Contrastive Representation Distillation"], "answer_arxiv_id": ["1910.10699"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_4362"} +{"question": "What research induced world models from large language models for robust human-like reasoning?", "answer": ["Reasoning with Language Model is Planning with World Model"], "answer_arxiv_id": ["2305.14992"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_4363"} +{"question": "Which papers focus on alleviating 3D annotation efforts by distilling knowledge from 2D networks?", "answer": ["OpenScene: 3D Scene Understanding with Open Vocabularies", "CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP", "2DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Clouds", "Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Data"], "answer_arxiv_id": ["2211.15654", "2301.04926", "2207.04397", "2203.16258"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_4364"} +{"question": "May you list the papers related to the establishment of connections among expressibility, generalizability, and trainability of quantum neural networks?", "answer": ["Barren plateaus preclude learning scramblers", "Theoretical Error Performance Analysis for Variational Quantum Circuit Based Functional Regression"], "answer_arxiv_id": ["2009.14808", "2206.04804v2"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_4365"} +{"question": "Who researched to understand the relationships between tasks for efficient multi-tasking?", "answer": ["Efficiently Identifying Task Groupings for Multi-Task Learning"], "answer_arxiv_id": ["2109.04617"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_4366"} +{"question": "What are some examples of diagnostic benchmarks for compositionality with hard negatives?", "answer": ["Image Retrieval from Contextual Descriptions", "Probing Image-Language Transformers for Verb Understanding", "Does CLIP Bind Concepts? Probing Compositionality in Large Image Models", "VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena"], "answer_arxiv_id": ["2203.15867", "2106.09141", "2212.10537", "2112.07566v2"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4367"} +{"question": "In what papers refer to the black-box nature of neural networks in the context of QA systems?", "answer": ["A Survey of the State of Explainable AI for Natural Language Processing"], "answer_arxiv_id": ["2010.00711"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_4368"} +{"question": "What studies aimed at measuring hallucination in Multimodal LLMs?", "answer": ["Evaluating Object Hallucination in Large Vision-Language Models", "Object Hallucination in Image Captioning"], "answer_arxiv_id": ["2305.10355", "1809.02156"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_4369"} +{"question": "Any works about distilling knowledge from 2D vision-language pre-training to propose unsupervised 3D perception?", "answer": ["Unsupervised 3D Perception with 2D Vision-Language Distillation for\n Autonomous Driving"], "answer_arxiv_id": ["2309.14491"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_4370"} +{"question": "Which work improved the robustness and performance of monocular depth estimation by employing data augmentation to simulate different weather conditions?", "answer": ["Self-supervised Monocular Depth Estimation: Let's Talk About The Weather"], "answer_arxiv_id": ["2307.08357"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_4371"} +{"question": "Could you provide some works that leveraged GANs to address geometry inconsistencies in image compositing?", "answer": ["ST-GAN: Spatial Transformer Generative Adversarial Networks for Image\n Compositing", "Compositional GAN: Learning Image-Conditional Binary Composition"], "answer_arxiv_id": ["1803.01837", "1807.07560"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4372"} +{"question": "Which studies discuss static sparse training in neural networks?", "answer": ["Diversity Networks: Neural Network Compression Using Determinantal Point Processes", "Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science", "Rigging the Lottery: Making All Tickets Winners", "The Unreasonable Effectiveness of Random Pruning: Return of the Most Naive Baseline for Sparse Training"], "answer_arxiv_id": ["1511.05077", "1707.04780v2", "1911.11134", "2202.02643"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4373"} +{"question": "What work introduced the IDC method used in comparative analysis?", "answer": ["Dataset Condensation via Efficient Synthetic-Data Parameterization"], "answer_arxiv_id": ["2205.14959v2"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_4374"} +{"question": "Which works discuss the use of data augmentation or more training data to improve the performance and robustness of deep learning models?", "answer": ["Data Augmentation Can Improve Robustness", "Generalizing to Unseen Domains via Adversarial Data Augmentation", "Unlabeled Data Improves Adversarial Robustness", "Robust Learning Meets Generative Models: Can Proxy Distributions Improve Adversarial Robustness?"], "answer_arxiv_id": ["2111.05328", "1805.12018", "1905.13736", "2104.09425"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_4375"} +{"question": "Which visual synthesis models aim to generate plausible images?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "ImageBART: Bidirectional Context with Multinomial Diffusion for Autoregressive Image Synthesis"], "answer_arxiv_id": ["1812.04948", "2108.08827"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_4376"} +{"question": "Which studies consider the dynamics of responses from human population to algorithmic prediction?", "answer": ["Delayed Impact of Fair Machine Learning", "The Disparate Effects of Strategic Manipulation", "Performative Prediction", "How Do Fair Decisions Fare in Long-term Qualification?", "On the Long-term Impact of Algorithmic Decision Policies: Effort Unfairness and Feature Segregation through Social Learning", "The Disparate Equilibria of Algorithmic Decision Making when Individuals Invest Rationally", "A Short-term Intervention for Long-term Fairness in the Labor Market", "From Fair Decision Making to Social Equality", "Dynamic Modeling and Equilibria in Fair Decision Making", "Unintended Selection: Persistent Qualification Rate Disparities and Interventions"], "answer_arxiv_id": ["1803.04383", "1808.08646", "2002.06673", "2010.11300", "1903.01209", "1910.04123", "1712.00064", "1812.02952", "1911.06837", "2111.01201"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_4377"} +{"question": "Which studies focused on fully cooperative settings in the field of multi-agent reinforcement learning (MARL) communication?", "answer": ["Learning to Communicate with Deep Multi-Agent Reinforcement Learning", "Learning Multiagent Communication with Backpropagation", "Multiagent Bidirectionally-Coordinated Nets: Emergence of Human-level Coordination in Learning to Play StarCraft Combat Games"], "answer_arxiv_id": ["1605.06676", "1605.07736", "1703.10069v4"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_4378"} +{"question": "What studies have designed a side-adapter network to leverage CLIP features for decoupling segmentation and classification?", "answer": ["Side Adapter Network for Open-Vocabulary Semantic Segmentation"], "answer_arxiv_id": ["2302.12242"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4379"} +{"question": "Are there any works that explore visual prompts for task adaptation?", "answer": ["Visual Prompt Tuning", "Diversity-Aware Meta Visual Prompting", "MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2203.12119", "2303.08138", "2210.03117"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_4380"} +{"question": "Which papers examine the memorization of large language models on sensitive information in the training data?", "answer": ["Extracting Training Data from Large Language Models"], "answer_arxiv_id": ["2012.07805"], "source_meta": {"published_time": "20211224"}, "qid": "AutoScholarQuery_train_4381"} +{"question": "What works explored model designs like degradation-adaptive networks and deep unfolding networks on top of multi-degradation settings?", "answer": ["Blind Super-Resolution With Iterative Kernel Correction", "Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Efficient and Degradation-Adaptive Network for Real-World Image\n Super-Resolution", "Unsupervised Degradation Representation Learning for Blind\n Super-Resolution", "Unfolding the Alternating Optimization for Blind Super Resolution", "Unfolded Deep Kernel Estimation for Blind Image Super-resolution", "Deep Unfolding Network for Image Super-Resolution"], "answer_arxiv_id": ["1904.03377", "1511.04587", "2203.14216", "2104.00416", "2010.02631", "2203.05568", "2003.10428"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_4382"} +{"question": "What is the research that introduced HuBERT, where training involved masked prediction with masked continuous audio signals?", "answer": ["HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units"], "answer_arxiv_id": ["2106.07447"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_4383"} +{"question": "What methodologies have been proposed to detect data contamination in LLMs?", "answer": ["Skywork: A More Open Bilingual Foundation Model", "Detecting Pretraining Data from Large Language Models"], "answer_arxiv_id": ["2310.19341", "2310.16789"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_4384"} +{"question": "What research proposed the use of a parametric Bezier curve-based method in lane detection?", "answer": ["Rethinking Efficient Lane Detection via Curve Modeling"], "answer_arxiv_id": ["2203.02431v2"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_4385"} +{"question": "Could you provide me some studies that predicted relative position of image patches or sorted sequential data during vision model pretraining?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles"], "answer_arxiv_id": ["1505.05192", "1603.09246v3"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_4386"} +{"question": "What works developed methods to improve server's aggregation of local updates, quantized at varying levels?", "answer": ["Secure Aggregation with Heterogeneous Quantization in Federated Learning", "Bitwidth Heterogeneous Federated Learning with Progressive Weight\n Dequantization"], "answer_arxiv_id": ["2009.14388", "2202.11453"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_4387"} +{"question": "Could you provide me some research that focuses on leveraging English as a pivot language to efficiently augment multilingual instruction-following capabilities?", "answer": ["Extrapolating Large Language Models to Non-English by Aligning Languages"], "answer_arxiv_id": ["2308.04948"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_4388"} +{"question": "What are the most recent works on model-based robust average-reward MDPs?", "answer": ["Robust Average-Reward Markov Decision Processes"], "answer_arxiv_id": ["2301.00858"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_4389"} +{"question": "Which studies use docstring and function signature as an input form for evaluating program synthesis models?", "answer": ["Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2107.03374"], "source_meta": {"published_time": "20220325"}, "qid": "AutoScholarQuery_train_4390"} +{"question": "Could you provide me a paper where Contrastive denoising, Mix query selection and a look forward twice scheme were introduced for the first time?", "answer": ["DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection"], "answer_arxiv_id": ["2203.03605"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_4391"} +{"question": "Which research work introduces using text prompts as conditioning information for controlling generative model composition?", "answer": ["Zero-Shot Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2102.12092", "2112.10752", "2205.11487"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_4392"} +{"question": "Which papers can be seen within the wider spectrum of effective model scaling?", "answer": ["EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", "MobileNetV2: Inverted Residuals and Linear Bottlenecks"], "answer_arxiv_id": ["1905.11946", "1801.04381"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_4393"} +{"question": "What references contain studies about the application of neural fields for representing 3D shapes?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Occupancy Networks: Learning 3D Reconstruction in Function Space"], "answer_arxiv_id": ["1812.02822", "1901.05103", "1812.03828"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_4394"} +{"question": "Could you provide me some works about efficient transfer features for unsupervised domain adaptation with large-scale networks?", "answer": ["CDTrans: Cross-domain Transformer for Unsupervised Domain Adaptation", "TVT: Transferable Vision Transformer for Unsupervised Domain Adaptation", "Safe Self-Refinement for Transformer-based Domain Adaptation"], "answer_arxiv_id": ["2109.06165", "2108.05988", "2204.07683"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_4395"} +{"question": "Could you provide me studies about tractable dendritic RNNs and Neural Operators as alternatives for NODE?", "answer": ["Tractable Dendritic RNNs for Reconstructing Nonlinear Dynamical Systems", "Neural Operator: Learning Maps Between Function Spaces", "Learning Dissipative Dynamics in Chaotic Systems"], "answer_arxiv_id": ["2207.02542", "2108.08481", "2106.06898"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_4396"} +{"question": "Which works proposed leveraging annealed time-step schedule for efficient training in Text-3D Generation?", "answer": ["DreamTime: An Improved Optimization Strategy for Diffusion-Guided 3D\n Generation", "DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation"], "answer_arxiv_id": ["2306.12422", "2309.16653"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_4397"} +{"question": "What researches have been conducted on multi-task kernel bandits based on composite kernel functions?", "answer": ["Multi-Task Learning for Contextual Bandits"], "answer_arxiv_id": ["1705.08618"], "source_meta": {"published_time": "20211029"}, "qid": "AutoScholarQuery_train_4398"} +{"question": "What research originally proposed fully decoupled discrete control via Q-learning?", "answer": ["Learning to Factor Policies and Action-Value Functions: Factored Action Space Representations for Deep Reinforcement learning"], "answer_arxiv_id": ["1705.07269"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_4399"} +{"question": "Can you name papers that leveraged large language models (LLMs) to enhance the generalization of agents?", "answer": ["CLIP-Nav: Using CLIP for Zero-Shot Vision-and-Language Navigation", "NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large\n Language Models"], "answer_arxiv_id": ["2211.16649", "2305.16986"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4400"} +{"question": "What works proposed co-memory and co-attention mechanisms for capturing question-related spatial and temporal clues?", "answer": ["Motion-Appearance Co-Memory Networks for Video Question Answering", "Heterogeneous Memory Enhanced Multimodal Attention Model for Video Question Answering"], "answer_arxiv_id": ["1803.10906", "1904.04357"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_4401"} +{"question": "Which papers have used clipped versions of the identity function as an alternative to Straight-Through Estimator when training threshold networks?", "answer": ["Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations", "XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks"], "answer_arxiv_id": ["1609.07061", "1603.05279"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_4402"} +{"question": "Which papers are about explicit 3D visual affordances?", "answer": ["Learning Affordance Landscapes for Interaction Exploration in 3D\n Environments", "3D AffordanceNet: A Benchmark for Visual Object Affordance Understanding"], "answer_arxiv_id": ["2008.09241", "2103.16397"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_4403"} +{"question": "Is there any work which proposed a variant of LinCBwK where the selected action must satisfy a single constraint with high probability in all rounds?", "answer": ["Linear Stochastic Bandits Under Safety Constraints"], "answer_arxiv_id": ["1908.05814"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_4404"} +{"question": "Which work hypothesized that emergent abilities may be partially attributed to the metric?", "answer": ["Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models"], "answer_arxiv_id": ["2206.04615"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_4405"} +{"question": "Which paper uses a diffusion model to synthesize sequences of motion codes representing motion as a sequence of flow maps?", "answer": ["LEO: Generative Latent Image Animator for Human Video Synthesis"], "answer_arxiv_id": ["2305.03989"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4406"} +{"question": "Which publication proposed the ODIN method for detecting out-of-distribution inputs using gradient information?", "answer": ["Gradients as a measure of uncertainty in neural networks"], "answer_arxiv_id": ["2008.08030"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_4407"} +{"question": "Which paper popularized the Lookahead algorithm for minimax training?", "answer": ["Taming GANs with Lookahead–Minmax"], "answer_arxiv_id": ["2006.14567"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_4408"} +{"question": "Which papers explored meta-learning methods that learn the update function?", "answer": ["Unbiased Gradient Estimation in Unrolled Computation Graphs with Persistent Evolution Strategies", "Backprop Evolution"], "answer_arxiv_id": ["2112.13835", "1808.02822"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_4409"} +{"question": "Which work found out that byte-level encoding performs worse than subword tokenization for non-Latin scripts?", "answer": ["ByT5: Towards a token-free future with pre-trained byte-to-byte models"], "answer_arxiv_id": ["2105.13626"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4410"} +{"question": "What studies have attempted to improve compositional understanding of contrastive vision-language pre-trained models?", "answer": ["When and why vision-language models behave like bags-of-words, and what\n to do about it?", "Going Beyond Nouns With Vision & Language Models Using Synthetic Data", "Learning to Compose Soft Prompts for Compositional Zero-Shot Learning", "Teaching Structured Vision&Language Concepts to Vision&Language Models", "Coarse-to-Fine Contrastive Learning in Image-Text-Graph Space for\n Improved Vision-Language Compositionality"], "answer_arxiv_id": ["2210.01936", "2303.17590", "2204.03574", "2211.11733", "2305.13812"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_4411"} +{"question": "Could you provide me work that utilizes Distributionally Robust Optimization for Language Models in the group shift setting?", "answer": ["Distributionally Robust Language Modeling"], "answer_arxiv_id": ["1909.02060"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_4412"} +{"question": "Can you give examples of recent works in pose estimation specialized towards human and animal body part?", "answer": ["Jointformer: Single-Frame Lifting Transformer with Error Prediction and\n Refinement for 3D Human Pose Estimation", "A simple yet effective baseline for 3d human pose estimation", "MotionCLIP: Exposing Human Motion Generation to CLIP Space", "Human Motion Diffusion Model"], "answer_arxiv_id": ["2208.03704", "1705.03098", "2203.08063", "2209.14916"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_4413"} +{"question": "Can you name a research work that proposed clip2latent and used a diffusion model instead of a GMM as the translator network?", "answer": ["clip2latent: Text driven sampling of a pre-trained StyleGAN using denoising diffusion and CLIP"], "answer_arxiv_id": ["2210.02347"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_4414"} +{"question": "What research proposes the gated LIF (GLIF) spiking neuron?", "answer": ["GLIF: A Unified Gated Leaky Integrate-and-Fire Neuron for Spiking Neural Networks"], "answer_arxiv_id": ["2210.13768"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_4415"} +{"question": "Could you name some works that rely on using dense depth maps for monocular 3D object detection?", "answer": ["Learning Depth-Guided Convolutions for Monocular 3D Object Detection", "Depth-conditioned Dynamic Message Propagation for Monocular 3D Object\n Detection", "Rethinking Pseudo-LiDAR Representation", "Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object\n Detection for Autonomous Driving", "Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous\n Driving"], "answer_arxiv_id": ["1912.04799", "2103.16470", "2008.04582", "1812.07179", "1906.06310"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_4416"} +{"question": "What studies have tackled the computational cost of VaSSO and SAM?", "answer": ["Towards Efficient and Scalable Sharpness-Aware Minimization", "Efficient Sharpness-aware Minimization for Improved Training of Neural Networks", "An Adaptive Policy to Employ Sharpness-Aware Minimization", "Sharpness-Aware Training for Free"], "answer_arxiv_id": ["2203.02714", "2110.03141", "2304.14647", "2205.14083"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4417"} +{"question": "What studies proposed offline RL for learning a good policy and value initialization, that was then followed by online fine-tuning?", "answer": ["Offline Reinforcement Learning with Implicit Q-Learning", "Mildly Conservative Q-Learning for Offline Reinforcement Learning", "Improving TD3-BC: Relaxed Policy Constraint for Offline Learning and Stable Online Fine-Tuning", "Supported Policy Optimization for Offline Reinforcement Learning", "Offline-to-Online Reinforcement Learning via Balanced Replay and Pessimistic Q-Ensemble"], "answer_arxiv_id": ["2110.06169", "2206.04745v3", "2211.11802", "2202.06239", "2107.00591"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_4418"} +{"question": "What studies propose defense mechanisms against adversarial attacks?", "answer": ["Towards Fast Computation of Certified Robustness for ReLU Networks", "Certified Adversarial Robustness via Randomized Smoothing", "Certified Adversarial Robustness with Additive Noise"], "answer_arxiv_id": ["1804.09699", "1902.02918", "1809.03113"], "source_meta": {"published_time": "20220719"}, "qid": "AutoScholarQuery_train_4419"} +{"question": "What works propose the use of variational autoencoders for data representation learning?", "answer": ["Auto-Encoding Variational Bayes", "Stochastic Backpropagation and Approximate Inference in Deep Generative Models"], "answer_arxiv_id": ["1312.6114", "1401.4082"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_4420"} +{"question": "In what papers do researchers propose high-order Graph Neural Networks based on k-WL for increasing the expressiveness?", "answer": ["Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "Weisfeiler and Leman go sparse: Towards scalable higher-order graph embeddings"], "answer_arxiv_id": ["1810.02244", "1904.01543"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_4421"} +{"question": "What paper conducted a comprehensive empirical evaluation of pre-trained object-oriented models for model and policy learning in multi-object RL?", "answer": ["An Investigation into Pre-Training Object-Centric Representations for Reinforcement Learning"], "answer_arxiv_id": ["2302.04419"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_4422"} +{"question": "Are there any papers discussing the concept of module criticality?", "answer": ["Are All Layers Created Equal?", "The intriguing role of module criticality in the generalization of deep networks"], "answer_arxiv_id": ["1902.01996", "1912.00528"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_4423"} +{"question": "What paper first defined balance in the case of two groups for group-level fairness?", "answer": ["Fair Clustering Through Fairlets"], "answer_arxiv_id": ["1802.05733"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4424"} +{"question": "Which works proposed the use of distill language feature into 3D space for open-vocabulary interactive segmentation?", "answer": ["LERF: Language Embedded Radiance Fields", "Interactive Segmentation of Radiance Fields", "Decomposing NeRF for Editing via Feature Field Distillation", "OpenMask3D: Open-Vocabulary 3D Instance Segmentation"], "answer_arxiv_id": ["2303.09553", "2212.13545", "2205.15585", "2306.13631"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_4425"} +{"question": "Could you provide some papers that developed methods to improve the payload, robustness, and unforgeability of watermark embedding?", "answer": ["Provable Robust Watermarking for AI-Generated Text", "A Semantic Invariant Robust Watermark for Large Language Models", "On the Reliability of Watermarks for Large Language Models", "A Robust Semantics-based Watermark for Large Language Model against\n Paraphrasing", "An Unforgeable Publicly Verifiable Watermark for Large Language Models", "Publicly-Detectable Watermarking for Language Models"], "answer_arxiv_id": ["2306.17439", "2310.06356", "2306.04634", "2311.08721", "2307.16230", "2310.18491"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_4426"} +{"question": "Could you provide information about studies on other notions of fairness such as individual fairness and counterfactual fairness?", "answer": ["Fairness Through Awareness", "Counterfactual Fairness"], "answer_arxiv_id": ["1104.3913", "1703.06856"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_4427"} +{"question": "What works related to synthesizing first-order logic rules from data in inductive logic programming?", "answer": ["The ILASP System for Inductive Learning of Answer Set Programs", "Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "Learn to Explain Efficiently via Neural Logic Inductive Learning"], "answer_arxiv_id": ["2005.00904", "1702.08367", "1910.02481"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_4428"} +{"question": "What research decomposes the image generation into multiple factors and generates images by re-combining them?", "answer": ["Composer: Creative and Controllable Image Synthesis with Composable Conditions"], "answer_arxiv_id": ["2302.09778"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4429"} +{"question": "What research works used vectorized representations for encoding contextual information in trajectory prediction?", "answer": ["ADAPT: Efficient Multi-Agent Trajectory Prediction with Adaptation", "TBP-Former: Learning Temporal Bird's-Eye-View Pyramid for Joint\n Perception and Prediction in Vision-Centric Autonomous Driving", "Holistic Graph-based Motion Prediction", "GOHOME: Graph-Oriented Heatmap Output for future Motion Estimation"], "answer_arxiv_id": ["2307.14187", "2303.09998", "2301.13545", "2109.01827"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_4430"} +{"question": "Which studies combine active learning and semi-supervised learning?", "answer": ["Cost-Effective Active Learning for Deep Image Classification", "Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Variational Adversarial Active Learning"], "answer_arxiv_id": ["1701.03551", "1708.00489", "1904.00370"], "source_meta": {"published_time": "20211227"}, "qid": "AutoScholarQuery_train_4431"} +{"question": "Which research pieces are about employing a Bayesian Neural Network (BNN) and a sparse GP on fingerprint representations of molecules in Bayes Optimization?", "answer": ["Parallel and Distributed Thompson Sampling for Large-scale Accelerated Exploration of Chemical Space", "Scalable Thompson Sampling using Sparse Gaussian Process Models"], "answer_arxiv_id": ["1706.01825", "2006.05356"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_4432"} +{"question": "Could you name some works that contribute to the debate around whether language models can acquire meaning from being trained on form alone?", "answer": ["Meaning without reference in large language models"], "answer_arxiv_id": ["2208.02957"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_4433"} +{"question": "Which studies advanced the conventional Hopfield Networks by introducing continuous queries and states through a new energy function?", "answer": ["Hopfield Networks is All You Need"], "answer_arxiv_id": ["2008.02217"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_4434"} +{"question": "What studies proposed variants of x-formers to enable transformers to attend on longer context?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Linformer: Self-Attention with Linear Complexity", "Longformer: The Long-Document Transformer", "Efficient Content-Based Sparse Attention with Routing Transformers", "Big Bird: Transformers for Longer Sequences"], "answer_arxiv_id": ["1901.02860", "2006.04768", "2004.05150", "2003.05997", "2007.14062"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_4435"} +{"question": "What references can provide standard results on private empirical risk minimization?", "answer": ["Private Empirical Risk Minimization Beyond the Worst Case: The Effect of the Constraint Set Geometry"], "answer_arxiv_id": ["1411.5417"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_4436"} +{"question": "Do you know any work that involved active exploration to reduce model uncertainty?", "answer": ["Chance-Constrained Trajectory Optimization for Safe Exploration and Learning of Nonlinear Systems"], "answer_arxiv_id": ["2005.04374"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_4437"} +{"question": "Which works discuss the universal approximation property of specific neural network architectures?", "answer": ["Discrete Restricted Boltzmann Machines", "Deep Narrow Boltzmann Machines are Universal Approximators", "Universal Approximation of Markov Kernels by Shallow Stochastic Feedforward Networks", "On the number of response regions of deep feedforward networks with piecewise linear activations", "Stochastic Feedforward Neural Networks: Universal Approximation"], "answer_arxiv_id": ["1301.3529", "1411.3784", "1503.07211", "1312.6098", "1910.09763"], "source_meta": {"published_time": "20220818"}, "qid": "AutoScholarQuery_train_4438"} +{"question": "What research works focused on human pose estimation using IMU sensors?", "answer": ["Sparse Inertial Poser: Automatic 3D Human Pose Estimation from Sparse\n IMUs", "Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time", "TransPose: Real-time 3D Human Translation and Pose Estimation with Six\n Inertial Sensors", "Physical Inertial Poser (PIP): Physics-aware Real-time Human Motion\n Tracking from Sparse Inertial Sensors", "Transformer Inertial Poser: Real-time Human Motion Reconstruction from\n Sparse IMUs with Simultaneous Terrain Generation"], "answer_arxiv_id": ["1703.08014", "1810.04703", "2105.04605", "2203.08528", "2203.15720"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_4439"} +{"question": "In what papers did the researchers introduce 'metadata normalization' (MDN), a novel layer normalization module?", "answer": ["Metadata Normalization"], "answer_arxiv_id": ["2104.09052"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_4440"} +{"question": "What studies use Mel-spectrograms of the audio signal as input for the audio encoder?", "answer": ["Active Speakers in Context", "Is Someone Speaking? Exploring Long-term Temporal Features for\n Audio-visual Active Speaker Detection", "How to Design a Three-Stage Architecture for Audio-Visual Active Speaker\n Detection in the Wild", "MAAS: Multi-modal Assignation for Active Speaker Detection"], "answer_arxiv_id": ["2005.09812", "2107.06592", "2106.03932", "2101.03682"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_4441"} +{"question": "What studies argued about the limitation of their architectural designs in scenarios involving multiple subjects?", "answer": ["X&Fuse: Fusing Visual Information in Text-to-Image Generation", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning"], "answer_arxiv_id": ["2303.01000", "2302.13848", "2304.03411"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_4442"} +{"question": "Which papers used attention-based aggregation methods in the context of WSI classification in MIL?", "answer": ["Attention-based Deep Multiple Instance Learning", "Multi-scale Domain-adversarial Multiple-instance CNN for Cancer Subtype Classification with Unannotated Histopathological Images", "Whole Slide Images based Cancer Survival Prediction using Attention Guided Deep Multiple Instance Learning Networks", "Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learning", "DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning for Histopathology Whole Slide Image Classification", "Data Efficient and Weakly Supervised Computational Pathology on Whole Slide Images"], "answer_arxiv_id": ["1802.04712", "2001.01599", "2009.11169v1", "2011.08939", "2203.12081", "2004.09666"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4443"} +{"question": "Could you provide me studies that proposed text-guided strategies in the field of human avatar creation?", "answer": ["AvatarCraft: Transforming Text into Neural Human Avatars with\n Parameterized Shape and Pose Control", "DreamHuman: Animatable 3D Avatars from Text", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "TeCH: Text-guided Reconstruction of Lifelike Clothed Humans", "TADA! Text to Animatable Digital Avatars", "AvatarVerse: High-quality & Stable 3D Avatar Creation from Text and Pose", "DreamWaltz: Make a Scene with Complex 3D Animatable Avatars", "One-shot Implicit Animatable Avatars with Model-based Priors", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars"], "answer_arxiv_id": ["2303.17606", "2306.09329", "2304.00916", "2308.08545", "2308.10899", "2308.03610", "2305.12529", "2212.02469", "2205.08535"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4444"} +{"question": "Which work extended the gradient-based adversarial attack, FGM, to the point cloud domain?", "answer": ["Extending Adversarial Attacks and Defenses to Deep 3D Point Cloud\n Classifiers"], "answer_arxiv_id": ["1901.03006"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_4445"} +{"question": "Could you provide me some studies about single-image inverse rendering?", "answer": ["Invertible Neural BRDF for Object Inverse Rendering", "LIME: Live Intrinsic Material Estimation"], "answer_arxiv_id": ["2008.04030", "1801.01075"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4446"} +{"question": "Which paper first demonstrated the possibility of extending English-only PLMs to other languages by relearning the embedding layer with unsupervised data from the new language?", "answer": ["On the Cross-lingual Transferability of Monolingual Representations"], "answer_arxiv_id": ["1910.11856"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_4447"} +{"question": "Which studies focus on understanding the phenomenon of benign overfitting through implicit regularization mechanisms in overparameterized models?", "answer": ["In Search of the Real Inductive Bias: On the Role of Implicit Regularization in Deep Learning"], "answer_arxiv_id": ["1412.6614"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_4448"} +{"question": "Could you provide me the study that utilizes local dependencies to generate counterfactual samples to enhance sample efficiency?", "answer": ["Counterfactual Data Augmentation using Locally Factored Dynamics"], "answer_arxiv_id": ["2007.02863"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_4449"} +{"question": "Are there any references that suggest Unrolled Sharpness-Aware Minimization (USAM) could escape saddles faster than Stochastic Gradient Descent (SGD)?", "answer": ["Escaping Saddle Points for Effective Generalization on Class-Imbalanced Data"], "answer_arxiv_id": ["2212.13827"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_4450"} +{"question": "Which works discuss the non-normalized nature of Energy-Based Models (EBMs)?", "answer": ["How to Train Your Energy-Based Models"], "answer_arxiv_id": ["2101.03288"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_4451"} +{"question": "Which work explores solutions space using semantic maps?", "answer": ["DeepSEE: Deep Disentangled Semantic Explorative Extreme Super-Resolution"], "answer_arxiv_id": ["2004.04433"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_4452"} +{"question": "Could you provide me some studies on pixel-based methods in object detection?", "answer": ["You Only Look Once: Unified, Real-Time Object Detection", "SSD: Single Shot MultiBox Detector", "Focal Loss for Dense Object Detection", "FCOS: Fully Convolutional One-Stage Object Detection", "Bridging the Gap Between Anchor-based and Anchor-free Detection via\n Adaptive Training Sample Selection"], "answer_arxiv_id": ["1506.02640", "1512.02325", "1708.02002", "1904.01355", "1912.02424"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_4453"} +{"question": "Which work demonstrates the application of LLMs in code generation?", "answer": ["Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2107.03374"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4454"} +{"question": "What work popularized the idea to avoid learning spurious correlation in sub-population shift by exploiting group labels?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_4455"} +{"question": "Which works improve the performance of ICL by retrieving related demonstrations to the test instance?", "answer": ["Learning To Retrieve Prompts for In-Context Learning", "Selective Annotation Makes Language Models Better Few-Shot Learners"], "answer_arxiv_id": ["2112.08633", "2209.01975"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_4456"} +{"question": "Which research papers focused on addressing the issue of exposure bias in algorithm generation?", "answer": ["Relating Neural Text Degeneration to Exposure Bias", "Why Exposure Bias Matters: An Imitation Learning Perspective of Error Accumulation in Language Generation"], "answer_arxiv_id": ["2109.08705", "2204.01171"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_4457"} +{"question": "Which works addressed the challenges of optimizing the forward and backward passes for embedding layers?", "answer": ["TPU v4: An Optically Reconfigurable Supercomputer for Machine Learning with Hardware Support for Embeddings", "Merlin HugeCTR: GPU-accelerated Recommender System Training and Inference"], "answer_arxiv_id": ["2304.01433v3", "2210.08803"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_4458"} +{"question": "Which research papers discuss the use of data augmentation strategies like image manipulation for boosting model generalization?", "answer": ["Image Data Augmentation for Deep Learning: A Survey"], "answer_arxiv_id": ["2204.08610"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_4459"} +{"question": "What studies show that LLMs behave similarly to topic models, with output dependent on a latent topic?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_4460"} +{"question": "Is there any paper measuring the effect of DP-SGD on model accuracy?", "answer": ["Differential Privacy Has Disparate Impact on Model Accuracy"], "answer_arxiv_id": ["1905.12101"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_4461"} +{"question": "Could you tell me which paper introduced the VolSDF method that learns scaling and shrinking coefficients for volume densities?", "answer": ["Volume Rendering of Neural Implicit Surfaces"], "answer_arxiv_id": ["2106.12052"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_4462"} +{"question": "What papers made progress in developing more effective implementations of MCTS-based algorithms?", "answer": ["Online and Offline Reinforcement Learning by Planning with a Learned Model", "Efficient Offline Policy Optimization with a Learned Model"], "answer_arxiv_id": ["2104.06294", "2210.05980"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_4463"} +{"question": "What works have attempted to port existing off-policy value-based online RL methods to the offline setting with various types of additional regularization components?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "EMaQ: Expected-Max Q-Learning Operator for Simple Yet Effective Offline and Online RL", "Behavior Regularized Offline Reinforcement Learning", "Way Off-Policy Batch Deep Reinforcement Learning of Implicit Human Preferences in Dialog", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "AlgaeDICE: Policy Gradient from Arbitrary Experience", "Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Fisher Divergence Critic Regularization"], "answer_arxiv_id": ["1812.02900", "2007.11091", "1911.11361", "1907.00456", "1906.00949", "1912.02074", "2006.04779", "2103.08050"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_4464"} +{"question": "What studies proposed innovative methods to construct curricula that systematically escalate the complexity of subgoals?", "answer": ["Automatic Curriculum Learning through Value Disagreement", "C-Planning: An Automatic Curriculum for Learning Goal-Reaching Tasks"], "answer_arxiv_id": ["2006.09641", "2110.12080"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_4465"} +{"question": "Could you provide me publications where printable adversarial patches were used to increase the effectiveness of physical adversarial attacks against MDE?", "answer": ["Physical Attack on Monocular Depth Estimation with Optimal Adversarial\n Patches", "APARATE: Adaptive Adversarial Patch for CNN-based Monocular Depth Estimation for Autonomous Navigation"], "answer_arxiv_id": ["2207.04718", "2303.01351v3"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_4466"} +{"question": "Which papers discuss the decentralized learning in Multi-Agent RL approaches?", "answer": ["Multiagent Cooperation and Competition with Deep Reinforcement Learning", "Multi-agent Reinforcement Learning in Sequential Social Dilemmas"], "answer_arxiv_id": ["1511.08779", "1702.03037"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_4467"} +{"question": "Which papers dicussed fragment extraction methods for user-generated content video quality assessment?", "answer": ["FAST-VQA: Efficient End-to-end Video Quality Assessment with Fragment\n Sampling"], "answer_arxiv_id": ["2207.02595"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_4468"} +{"question": "Which papers led prior-based matting in image matting?", "answer": ["AlphaGAN: Generative adversarial networks for natural image matting", "Deep Image Matting", "Mask Guided Matting via Progressive Refinement Network", "MatteFormer: Transformer-Based Image Matting via Prior-Tokens"], "answer_arxiv_id": ["1807.10088v1", "1703.03872v3", "2012.06722", "2203.15662"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_4469"} +{"question": "Which papers have applied Transformers to slots from multiple frames in the field of unsupervised object-centric representation learning?", "answer": ["Generative Video Transformer: Can Objects be the Words?"], "answer_arxiv_id": ["2107.09240"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_4470"} +{"question": "Which papers discuss BNNs representing epistemic uncertainty via approximating the posterior distribution over parameters of base neural networks?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?"], "answer_arxiv_id": ["1703.04977"], "source_meta": {"published_time": "20210719"}, "qid": "AutoScholarQuery_train_4471"} +{"question": "Can you list some papers about open-vocabulary semantic segmentation that follow mix supervision?", "answer": ["Zero-Shot Semantic Segmentation", "Language-driven Semantic Segmentation", "A Simple Baseline for Open-Vocabulary Semantic Segmentation with Pre-trained Vision-language Model", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Generalized Decoding for Pixel, Image, and Language", "Side Adapter Network for Open-Vocabulary Semantic Segmentation", "A Simple Framework for Open-Vocabulary Segmentation and Detection"], "answer_arxiv_id": ["1906.00817", "2201.03546", "2112.14757", "2112.12143", "2210.04150", "2212.11270", "2302.12242", "2303.08131"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_4472"} +{"question": "Which studies conducted semantic attacks by implementing transformation to an image, possibly in HSV color space or through rotation and brightness changes?", "answer": ["Semantic Adversarial Examples", "Adversarial Attacks Beyond the Image Space", "Beyond Pixel Norm-Balls: Parametric Adversaries using an Analytically Differentiable Renderer", "ViewFool: Evaluating the Robustness of Visual Recognition to Adversarial Viewpoints"], "answer_arxiv_id": ["1804.00499", "1711.07183", "1808.02651", "2210.03895"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4473"} +{"question": "What are some contemporary efforts in spectral techniques for segmentation challenges in vision?", "answer": ["Mean Shift for Self-Supervised Learning", "Unsupervised Segmentation of Hyperspectral Remote Sensing Images with Superpixels"], "answer_arxiv_id": ["2105.07269", "2204.12296v2"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_4474"} +{"question": "Can you provide some studies on generating pseudo-LiDAR in single-view methods?", "answer": ["Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving", "Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud"], "answer_arxiv_id": ["1812.07179", "1903.09847"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_4475"} +{"question": "What papers applied manual exploration to identify pathologies in ImageNet?", "answer": ["ConvNets and ImageNet Beyond Accuracy: Understanding Mistakes and Uncovering Biases", "When does dough become a bagel? Analyzing the remaining mistakes on ImageNet", "From ImageNet to Image Classification: Contextualizing Progress on Benchmarks"], "answer_arxiv_id": ["1711.11443", "2205.04596", "2005.11295"], "source_meta": {"published_time": "20220629"}, "qid": "AutoScholarQuery_train_4476"} +{"question": "What works proposed learning construction heuristics for VRPs?", "answer": ["Pointer Networks", "Neural Combinatorial Optimization with Reinforcement Learning", "Reinforcement Learning for Solving the Vehicle Routing Problem", "Attention, Learn to Solve Routing Problems!", "POMO: Policy Optimization with Multiple Optima for Reinforcement Learning", "Learning Combinatorial Optimization Algorithms over Graphs", "An Efficient Graph Convolutional Network Technique for the Travelling Salesman Problem", "Combinatorial Optimization by Graph Pointer Networks and Hierarchical Reinforcement Learning", "Matrix Encoding Networks for Neural Combinatorial Optimization", "Multi-Decoder Attention Model with Embedding Glimpse for Solving Vehicle Routing Problems", "Sym-NCO: Leveraging Symmetricity for Neural Combinatorial Optimization"], "answer_arxiv_id": ["1506.03134", "1611.09940", "1802.04240", "1803.08475", "2010.16011", "1704.01665", "1906.01227", "1911.04936", "2106.11113", "2012.10638", "2205.13209"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4477"} +{"question": "What studies showcase the application of implicit generators?", "answer": ["Learning in Implicit Generative Models", "Stein Neural Sampler"], "answer_arxiv_id": ["1610.03483", "1810.03545"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4478"} +{"question": "What work evidences that infinite action space contextual bandits are minimax intractable?", "answer": ["Contextual Bandit Learning with Predictable Rewards"], "answer_arxiv_id": ["1202.1334v2"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_4479"} +{"question": "What papers are about lifting the human-object spatial relation to 3D space?", "answer": ["Detailed 2D-3D Joint Representation for Human-Object Interaction", "Object pop-up: Can we infer 3D objects and their poses from human\n interactions alone?"], "answer_arxiv_id": ["2004.08154", "2306.00777"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_4480"} +{"question": "What papers have proposed methods for text-guided image editing with diffusion models?", "answer": ["Text2LIVE: Text-Driven Layered Image and Video Editing", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation", "Diffusion-based Image Translation using Disentangled Style and Content Representation", "StyleDiffusion: Prompt-Embedding Inversion for Text-Based Editing"], "answer_arxiv_id": ["2204.02491", "2108.02938", "2110.02711", "2209.15264", "2303.15649"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4481"} +{"question": "Are there any works about improving robustness by utilising properties of retinal sampling?", "answer": ["Finding Biological Plausibility for Adversarially Robust Features via Metameric Tasks", "Biologically Inspired Mechanisms for Adversarial Robustness", "Human Eyes Inspired Recurrent Neural Networks are More Robust Against Adversarial Noises"], "answer_arxiv_id": ["2202.00838", "2006.16427", "2206.07282"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_4482"} +{"question": "What studies focus on detecting if a given instance is present in the training data of a machine learning model, through membership inference attacks?", "answer": ["Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "Membership Inference Attacks Against Machine Learning Models", "On the Privacy Risks of Model Explanations", "Understanding Membership Inferences on Well-Generalized Learning Models", "White-box vs Black-box: Bayes Optimal Strategies for Membership Inference", "Reconstructing Training Data from Trained Neural Networks", "Membership Inference Attacks From First Principles", "Extracting Training Data from Diffusion Models", "On the Privacy Risks of Algorithmic Recourse", "Parameters or Privacy: A Provable Tradeoff Between Overparameterization and Membership Inference", "A Blessing of Dimensionality in Membership Inference through Regularization", "Label-Only Membership Inference Attacks", "Enhanced Membership Inference Attacks against Machine Learning Models"], "answer_arxiv_id": ["1709.01604", "1610.05820", "1907.00164", "1802.04889", "1908.11229", "2206.07758", "2112.03570", "2301.13188", "2211.05427", "2202.01243", "2205.14055v2", "2007.14321", "2111.09679"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_4483"} +{"question": "Do you have any references discussing post-processing methods in ensuring algorithmic fairness?", "answer": ["Equality of Opportunity in Supervised Learning", "A Confidence-Based Approach for Balancing Fairness and Accuracy", "On Fairness and Calibration"], "answer_arxiv_id": ["1610.02413", "1601.05764", "1709.02012"], "source_meta": {"published_time": "20220916"}, "qid": "AutoScholarQuery_train_4484"} +{"question": "Are there any research papers on using prompting as method to improve multi-step reasoning part of language models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2205.10625"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_4485"} +{"question": "Could you tell me about the studies that utilized long convolution models?", "answer": ["What Makes Convolutional Models Great on Long Sequence Modeling?", "Simple Hardware-Efficient Long Convolutions for Sequence Modeling", "Toeplitz Neural Network for Sequence Modeling", "Hyena Hierarchy: Towards Larger Convolutional Language Models"], "answer_arxiv_id": ["2210.09298", "2302.06646", "2305.04749", "2302.10866v3"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_4486"} +{"question": "Could you provide me the documents where it was difficult to achieve differential privacy for deep neural networks?", "answer": ["Deep Learning with Differential Privacy"], "answer_arxiv_id": ["1607.00133"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_4487"} +{"question": "Can you specify studies that use student-teacher models to assign pseudo-labels to the unlabelled data in the context of self-training?", "answer": ["Temporal Ensembling for Semi-Supervised Learning", "Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning", "A robust self-learning method for fully unsupervised cross-lingual mappings of word embeddings", "Unsupervised Data Augmentation for Consistency Training", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Zero-Shot Text Classification with Self-Training"], "answer_arxiv_id": ["1610.02242v3", "1703.01780", "1704.03976", "1805.06297", "1904.12848", "2001.07685v2", "2210.17541"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_4488"} +{"question": "Are there studies that improved the initial results regarding the caching problem?", "answer": ["Near-Optimal Bounds for Online Caching with Machine Learned Advice", "Online metric algorithms with untrusted predictions", "Better and Simpler Learning-Augmented Online Caching"], "answer_arxiv_id": ["1910.12172", "2003.02144", "2005.13716"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4489"} +{"question": "Can you name the works that propose local search to find the rejection set efficiently?", "answer": ["Statistical Quantification of Differential Privacy: A Local Approach"], "answer_arxiv_id": ["2108.09528"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4490"} +{"question": "Which research papers utilized unrolling networks in solving sparse-view CT problems?", "answer": ["LEARN: Learned Experts' Assessment-based Reconstruction Network for Sparse-data CT", "FISTA-Net: Learning A Fast Iterative Shrinkage Thresholding Network for\n Inverse Problems in Imaging", "LEARN++: Recurrent Dual-Domain Reconstruction Network for Compressed Sensing CT"], "answer_arxiv_id": ["1707.09636v3", "2008.02683", "2012.06983v1"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_4491"} +{"question": "Are there any studies that focused on datasets capturing drastic changes in long-term indoor environments?", "answer": ["RIO: 3D Object Instance Re-Localization in Changing Indoor Environments", "Rescan: Inductive Instance Segmentation for Indoor RGBD Scans", "Nothing Stands Still: A Spatiotemporal Benchmark on 3D Point Cloud\n Registration Under Large Geometric and Temporal Change"], "answer_arxiv_id": ["1908.06109", "1909.11268", "2311.09346"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_4492"} +{"question": "What studies represent transformer-based video captioning methods?", "answer": ["Memory-Attended Recurrent Network for Video Captioning", "Attention-Based Multimodal Fusion for Video Description", "Video Captioning with Transferred Semantic Attributes", "Hierarchical Boundary-Aware Neural Encoder for Video Captioning", "From Deterministic to Generative: Multi-Modal Stochastic RNNs for Video\n Captioning", "Reconstruction Network for Video Captioning", "MART: Memory-Augmented Recurrent Transformer for Coherent Video\n Paragraph Captioning", "Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense\n Video Captioning", "Sequence to Sequence Learning with Neural Networks", "End-to-end Generative Pretraining for Multimodal Video Captioning"], "answer_arxiv_id": ["1905.03966", "1701.03126", "1611.07675", "1611.09312", "1708.02478", "1803.11438", "2005.05402", "2302.14115", "1409.3215", "2201.08264"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_4493"} +{"question": "Which paper introduced the use of channel splitting to suppress outliers in weights and reduce quantization errors?", "answer": ["Improving Neural Network Quantization without Retraining using Outlier Channel Splitting"], "answer_arxiv_id": ["1901.09504"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_4494"} +{"question": "Any references where diffusion models have been effectively used as generative priors in Image Restoration?", "answer": ["Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "DiffBIR: Towards Blind Image Restoration with Generative Diffusion Prior", "High-Resolution Image Synthesis with Latent Diffusion Models", "Exploiting Diffusion Prior for Real-World Image Super-Resolution"], "answer_arxiv_id": ["2201.11793", "2212.00490", "2308.15070", "2112.10752", "2305.07015"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_4495"} +{"question": "What papers use transfer learning-based methods in semi-supervised learning?", "answer": ["Big Self-Supervised Models are Strong Semi-Supervised Learners"], "answer_arxiv_id": ["2006.10029"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_4496"} +{"question": "Which paper proposes a keyframe-based training strategy to extend NeRF with time-conditioning?", "answer": ["Neural 3D Video Synthesis from Multi-view Video"], "answer_arxiv_id": ["2103.02597"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4497"} +{"question": "What papers are about the Bayesian robustness to uncertainty in the objective of imitation learning?", "answer": ["Bayesian Robust Optimization for Imitation Learning", "Policy Gradient Bayesian Robust Optimization for Imitation Learning"], "answer_arxiv_id": ["2007.12315", "2106.06499"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_4498"} +{"question": "What research addressed deploying a student model at the server and updating it using local gradients computed when minimizing the divergence of soft prediction?", "answer": ["Ensemble Distillation for Robust Model Fusion in Federated Learning"], "answer_arxiv_id": ["2006.07242"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_4499"} +{"question": "Which papers discussed formulating curricula by manipulating goals in curriculum learning?", "answer": ["Intrinsically Motivated Goal Exploration Processes with Automatic Curriculum Learning", "Automatic Goal Generation for Reinforcement Learning Agents"], "answer_arxiv_id": ["1708.02190", "1705.06366"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_4500"} +{"question": "Could you provide me some works on the problem of open-set DA?", "answer": ["Open Set Domain Adaptation by Backpropagation"], "answer_arxiv_id": ["1804.10427"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_4501"} +{"question": "What works proposed likelihood training for SBP approximation based on divergence objectives and FB-SDEs?", "answer": ["Likelihood Training of Schrödinger Bridge using Forward-Backward SDEs Theory"], "answer_arxiv_id": ["2110.11291"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_4502"} +{"question": "Could you provide some studies on region-level image understanding?", "answer": ["AIMS: All-Inclusive Multi-Level Segmentation", "High-Quality Entity Segmentation"], "answer_arxiv_id": ["2305.17768", "2211.05776"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_4503"} +{"question": "Can you list the papers that focus on risk-sensitive RL with the exponential utility criterion?", "answer": ["Risk-Sensitive Reinforcement Learning: Near-Optimal Risk-Sample Tradeoff in Regret", "Exponential Bellman Equation and Improved Regret Bounds for Risk-Sensitive Reinforcement Learning"], "answer_arxiv_id": ["2006.13827", "2111.03947"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_4504"} +{"question": "What work proposed the concept of Knowledge distillation in the context of the multi-step DDIM sampler and the pretrained UNet?", "answer": ["Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed"], "answer_arxiv_id": ["2101.02388"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_4505"} +{"question": "What works have studied max-cut and TSP problems in the context of unsupervised learning for CO?", "answer": ["Experimental performance of graph neural networks on random instances of max-cut", "Graph Neural Network Guided Local Search for the Traveling Salesperson Problem"], "answer_arxiv_id": ["1908.05767", "2110.05291"], "source_meta": {"published_time": "20230108"}, "qid": "AutoScholarQuery_train_4506"} +{"question": "Which work demonstrates the limitations of perplexity metric in language modeling?", "answer": ["MAUVE: Measuring the Gap Between Neural Text and Human Text using\n Divergence Frontiers"], "answer_arxiv_id": ["2102.01454"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_4507"} +{"question": "What works have been done on clustering learning in SSL approaches?", "answer": ["Deep Clustering for Unsupervised Learning of Visual Features", "Unsupervised Feature Learning via Non-Parametric Instance Discrimination"], "answer_arxiv_id": ["1807.05520", "1805.01978"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_4508"} +{"question": "Could you provide me with some papers that demonstrate the effectiveness of sparse MoE on vast language datasets?", "answer": ["Unified Scaling Laws for Routed Language Models", "Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2202.01169", "2203.15556"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_4509"} +{"question": "Any works that have offered a convergence proof of robust policy mirror-descent in the (s,a)-rectangular case?", "answer": ["First-order Policy Optimization for Robust Markov Decision Process"], "answer_arxiv_id": ["2209.10579"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_4510"} +{"question": "Which studies revealed the connections of diffusion models to score-based methods?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Improved Denoising Diffusion Probabilistic Models", "Estimating High Order Gradients of the Data Distribution by Denoising"], "answer_arxiv_id": ["1907.05600", "2102.09672", "2111.04726"], "source_meta": {"published_time": "20220316"}, "qid": "AutoScholarQuery_train_4511"} +{"question": "Which papers have used OT for clustering tasks?", "answer": ["Self-labelling via simultaneous clustering and representation learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "A Unified Objective for Novel Class Discovery"], "answer_arxiv_id": ["1911.05371", "2006.09882", "2108.08536"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_4512"} +{"question": "What works have been cited for the efficient mixed-integer modeling of neural networks?", "answer": ["Evaluating Robustness of Neural Networks with Mixed Integer Programming", "Training for Faster Adversarial Robustness Verification via Inducing ReLU Stability", "Safe Control with Neural Network Dynamic Models", "Strong mixed-integer programming formulations for trained neural networks"], "answer_arxiv_id": ["1711.07356", "1809.03008", "2110.01110", "1811.08359"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_4513"} +{"question": "Can you list some papers that analyze network representations in order to interpret deep learning models?", "answer": ["What you can cram into a single $&!#⁢* vector: Probing sentence embeddings for linguistic properties", "BERT Rediscovers the Classical NLP Pipeline", "Probing for Constituency Structure in Neural Language Models"], "answer_arxiv_id": ["1805.01070", "1905.05950", "2204.06201"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_4514"} +{"question": "Which works have been applied to generating new motions by recombining primitive actions?", "answer": ["MCP: Learning Composable Hierarchical Control with Multiplicative Compositional Policies", "CARL: Controllable Agent with Reinforcement Learning for Quadruped Locomotion"], "answer_arxiv_id": ["1905.09808", "2005.03288"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_4515"} +{"question": "Could you provide me some studies exploring the theory of SGD's dynamics near a manifold of minima with label noise?", "answer": ["Implicit regularization for deep neural networks driven by an Ornstein-Uhlenbeck like process", "Label Noise SGD Provably Prefers Flat Global Minimizers"], "answer_arxiv_id": ["1904.09080", "2106.06530"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_4516"} +{"question": "Which studies discuss approaches that involve backpropagating through the solver in Predict+Optimize framework?", "answer": ["Task-based End-to-end Model Learning in Stochastic Optimization"], "answer_arxiv_id": ["1703.04529"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_4517"} +{"question": "Could you provide me with some works that fall under the representation-based approach in Continual Learning techniques?", "answer": ["Self-Supervised Models are Continual Learners"], "answer_arxiv_id": ["2112.04215"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_4518"} +{"question": "What works reported length generalization problem in pretrained Transformers such as T5 and LaMDA?", "answer": ["Compositional Generalization in Semantic Parsing: Pre-training vs. Specialized Architectures", "Exploring Length Generalization in Large Language Models"], "answer_arxiv_id": ["2007.08970", "2207.04901"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4519"} +{"question": "Could you provide me with some studies on augmenting ICL with fine-grained information?", "answer": ["Cross-Task Generalization via Natural Language Crowdsourcing Instructions", "Multitask Prompted Training Enables Zero-Shot Task Generalization"], "answer_arxiv_id": ["2104.08773", "2110.08207"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_4520"} +{"question": "Which paper provides an evaluation of watermark robustness under unrealistic assumptions about adversary's capabilities?", "answer": ["Large Language Models can be Guided to Evade AI-Generated Text Detection"], "answer_arxiv_id": ["2305.10847"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_4521"} +{"question": "Could you provide information on studies that sampled calibration data from web text or model pre-training data?", "answer": ["SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot", "A Simple and Effective Pruning Approach for Large Language Models", "SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight\n Compression", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large\n Language Models"], "answer_arxiv_id": ["2301.00774", "2306.11695", "2306.03078", "2211.10438"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_4522"} +{"question": "Which studies have proposed non-blind super-resolution (SISR) methods?", "answer": ["Enhanced Deep Residual Networks for Single Image Super-Resolution", "Learning a Single Convolutional Super-Resolution Network for Multiple Degradations", "Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network", "Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "Structure-Preserving Super Resolution with Gradient Guidance", "Basic Binary Convolution Unit for Binarized Image Restoration Network"], "answer_arxiv_id": ["1707.02921", "1712.06116", "1609.04802", "1603.08155", "2003.13081", "2210.00405"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_4523"} +{"question": "Which study proposed using a stop gradient operation to avoid collapse in instance discrimination tasks?", "answer": ["Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2011.10566"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_4524"} +{"question": "What studies contributed to the development of Imagen and DALL.E with diffusion models in text-to-image generation?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2205.11487", "2204.06125", "2006.11239", "2112.10752"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_4525"} +{"question": "What work does enhance VQ tokenizer’s decoder with modulation and proposes multi-channel quantization with a shared codebook?", "answer": ["MoVQ: Modulating Quantized Vectors for High-Fidelity Image Generation"], "answer_arxiv_id": ["2209.09002"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_4526"} +{"question": "What work studied stochastic saddle-point problems with decision-dependent distributions with a focus on performative stable points?", "answer": ["Stochastic Saddle Point Problems with Decision-Dependent Distributions"], "answer_arxiv_id": ["2201.02313"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_4527"} +{"question": "What research developed an energy-constrained diffusion transformer?", "answer": ["DIFFormer: Scalable (Graph) Transformers Induced by Energy Constrained Diffusion"], "answer_arxiv_id": ["2301.09474v4"], "source_meta": {"published_time": "20220611"}, "qid": "AutoScholarQuery_train_4528"} +{"question": "In which paper is A* viewed as a Markov Decision Process with the Q-function equal to the number of steps of A* reaching the solution?", "answer": ["Learning Heuristic Search via Imitation"], "answer_arxiv_id": ["1707.03034"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_4529"} +{"question": "Are there any works that propose training-free strategies for architecture searching?", "answer": ["Neural Architecture Search without Training", "Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective", "Pruning neural networks without any data by iteratively conserving synaptic flow", "Zen-NAS: A Zero-Shot NAS for High-Performance Image Recognition", "MAE-DET: Revisiting Maximum Entropy Principle in Zero-Shot NAS for Efficient Object Detection", "Training-free Transformer Architecture Search", "Neural Architecture Design for GPU-Efficient Networks"], "answer_arxiv_id": ["2006.04647", "2102.11535", "2006.05467", "2102.01063", "2111.13336", "2203.12217", "2006.14090"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_4530"} +{"question": "What studies have drawn inspiration from DAGGER and SEARN for imitation learning for NLP?", "answer": ["Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks", "Parallel Scheduled Sampling", "Scheduled Sampling for Transformers", "Differentiable Scheduled Sampling for Credit Assignment", "Learning to Search Better than Your Teacher", "TextGAIL: Generative Adversarial Imitation Learning for Text Generation", "SeaRnn: Training RNNs with global-local losses"], "answer_arxiv_id": ["1506.03099", "1906.04331", "1906.07651", "1704.06970", "1502.02206", "2004.13796", "1706.04499"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_4531"} +{"question": "What studies have shown interest in contrastive approaches in self-supervised learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Big Self-Supervised Models are Strong Semi-Supervised Learners", "Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "Learning Transferable Visual Models From Natural Language Supervision", "DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2002.05709", "2006.10029", "1911.05722", "2003.04297", "2103.00020", "2304.07193"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_4532"} +{"question": "What studies focused on learning a special text prompt to feature specific objects or persons from the reference images for customized generation?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2212.04488"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_4533"} +{"question": "Could you list works that introduced neural-symbolic methods for query encoding?", "answer": ["Neural-Symbolic Entangled Framework for Complex Query Answering", "GammaE: Gamma Embeddings for Logical Queries on Knowledge Graphs"], "answer_arxiv_id": ["2209.08779", "2210.15578"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4534"} +{"question": "Could you name the study where the hidden layer is trained, and the loss is over a Gaussian data population?", "answer": ["Over-Parameterization Exponentially Slows Down Gradient Descent for Learning a Single Neuron"], "answer_arxiv_id": ["2302.10034"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_4535"} +{"question": "Are there any studies that leveraged large pre-trained vision-language models like CLIP and ALIGN?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_4536"} +{"question": "Could you find studies that have explored the topic of text generation from visions, specially in the image and video captioning task?", "answer": ["Describing Unseen Videos via Multi-Modal Cooperative Dialog Agents", "Saying the Unseen: Video Descriptions via Dialog Agents", "Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering", "Image Captioning with Semantic Attention", "Diverse and Accurate Image Description Using a Variational Auto-Encoder with an Additive Gaussian Encoding Space"], "answer_arxiv_id": ["2008.07935", "2106.14069", "1707.07998", "1603.03925", "1711.07068"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_4537"} +{"question": "Which researches showed impressive results with no negative samples in contrastive learning?", "answer": ["Debiased Contrastive Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["2007.00224", "2006.07733"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4538"} +{"question": "Which works incorporate spatial correlation in the noising process of finite-dimensional diffusion models?", "answer": ["Subspace Diffusion Generative Models", "Wavelet Score-Based Generative Modeling", "Cascaded Diffusion Models for High Fidelity Image Generation", "Image Super-Resolution via Iterative Refinement", "Blurring Diffusion Models", "Generative Modelling With Inverse Heat Dissipation"], "answer_arxiv_id": ["2205.01490", "2208.05003", "2106.15282", "2104.07636", "2209.05557", "2206.13397"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_4539"} +{"question": "Could you list the studies about learning symmetry invariant features and predicting pose distribution for seen object pose estimation?", "answer": ["SurfEmb: Dense and Continuous Correspondence Distributions for Object\n Pose Estimation with Learnt Surface Embeddings", "Implicit-PDF: Non-Parametric Representation of Probability Distributions\n on the Rotation Manifold", "Ki-Pode: Keypoint-based Implicit Pose Distribution Estimation of Rigid\n Objects", "SpyroPose: SE(3) Pyramids for Object Pose Distribution Estimation"], "answer_arxiv_id": ["2111.13489", "2106.05965", "2209.09659", "2303.05308"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_4540"} +{"question": "What papers feature instant-ngp and its use of multi-resolution hashing for efficient encoding?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20220826"}, "qid": "AutoScholarQuery_train_4541"} +{"question": "Which papers discuss the complexity of GDA and GDmax algorithms in deterministic and stochastic settings in the NC-SC setting?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems", "Solving a Class of Non-Convex Min-Max Games Using Iterative First Order Methods"], "answer_arxiv_id": ["1906.00331", "1902.08297"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_4542"} +{"question": "What works extend neural fields to represent dynamic scenes?", "answer": ["Neural Fields in Visual Computing and Beyond"], "answer_arxiv_id": ["2111.11426"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_4543"} +{"question": "Provide some studies that focus on efficient transformers?", "answer": ["Efficient Content-Based Sparse Attention with Routing Transformers", "Image Transformer", "Generating Wikipedia by Summarizing Long Sequences", "Blockwise Self-Attention for Long Document Understanding", "Generating Long Sequences with Sparse Transformers", "Longformer: The Long-Document Transformer", "Reformer: The Efficient Transformer", "Sparse Sinkhorn Attention", "DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter", "Patient Knowledge Distillation for BERT Model Compression", "Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned", "Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks", "Augmenting Self-attention with Persistent Memory", "Challenges in Information-Seeking QA: Unanswerable Questions and Paragraph Retrieval", "Longformer: The Long-Document Transformer", "Linformer: Self-Attention with Linear Complexity", "Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention", "Rethinking Attention with Performers", "Efficient Attention: Attention with Linear Complexities", "Random Feature Attention"], "answer_arxiv_id": ["2003.05997", "1802.05751", "1801.10198", "1911.02972", "1904.10509", "2004.05150", "2001.04451", "2002.11296", "1910.01108", "1908.09355", "1905.09418", "1810.00825", "1907.01470", "2010.11915", "2004.05150", "2006.04768", "2006.16236", "2009.14794", "1812.01243", "2103.02143"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_4544"} +{"question": "Which papers discuss the performance of SGD in terms of 'bias' and 'variance' terms?", "answer": ["Last iterate convergence of SGD for Least-Squares in the Interpolation regime", "Accelerated SGD for Non-Strongly-Convex Least Squares"], "answer_arxiv_id": ["2102.03183", "2203.01744"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_4545"} +{"question": "Can you give examples of research which reports on certified defenses as a way of non-biological defense?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing", "Certified Defense to Image Transformations via Randomized Smoothing", "Center Smoothing: Certified Robustness for Networks with Structured Outputs", "Certified Adversarial Robustness with Additive Noise"], "answer_arxiv_id": ["1902.02918", "2002.12463", "2102.09701", "1809.03113"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_4546"} +{"question": "Which studies have discussed the gap between realizable and non-realizable learning from dependent data?", "answer": ["Least Squares Regression with Markovian Data: Fundamental Limits and Algorithms"], "answer_arxiv_id": ["2006.08916"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_4547"} +{"question": "Which paper explores the idea of in-domain representation learning for remote sensing imagery?", "answer": ["In-domain representation learning for remote sensing"], "answer_arxiv_id": ["1911.06721"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_4548"} +{"question": "What research proposed a causal attention module that generates data partitions and removes confounders progressively?", "answer": ["Causal Attention for Unbiased Visual Recognition"], "answer_arxiv_id": ["2108.08782"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_4549"} +{"question": "Could you mention the research where temporal GNNs were applied?", "answer": ["Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks", "APAN: Asynchronous Propagation Attention Network for Real-time Temporal Graph Embedding"], "answer_arxiv_id": ["1908.01207", "2011.11545"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_4550"} +{"question": "Which works proposed methods to improve in-context learning in aspects like meta-training?", "answer": ["Meta-learning via Language Model In-context Tuning", "MetaICL: Learning to Learn In Context"], "answer_arxiv_id": ["2110.07814", "2110.15943"], "source_meta": {"published_time": "20220905"}, "qid": "AutoScholarQuery_train_4551"} +{"question": "Can you cite some studies about diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2102.09672"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_4552"} +{"question": "Could you provide me some studies that separation rank led to guidelines for architecture design, pretraining schemes and regularizers for countering locality in convolutional neural networks?", "answer": ["Inductive Bias of Deep Convolutional Networks through Pooling Geometry", "Deep Learning and Quantum Entanglement: Fundamental Connections with Implications to Network Design", "The Inductive Bias of In-Context Learning: Rethinking Pretraining Example Design", "Implicit Regularization in Hierarchical Tensor Factorization and Deep Convolutional Neural Networks"], "answer_arxiv_id": ["1605.06743", "1704.01552v2", "2110.04541", "2201.11729"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_4553"} +{"question": "Could you mention some work dealing with the problem of simultaneously sparse and low-rank matrix recovery?", "answer": ["Near-Optimal Estimation of Simultaneously Sparse and Low-Rank Matrices from Nested Linear Measurements", "Jointly Low-Rank and Bisparse Recovery: Questions and Partial Answers"], "answer_arxiv_id": ["1506.08159", "1902.04731"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_4554"} +{"question": "What researches have been done on integrating single-view and multi-view prediction?", "answer": ["Single-View and Multi-View Depth Fusion", "MVS2D: Efficient Multi-view Stereo via Attention-Driven 2D Convolutions", "Multi-View Depth Estimation by Fusing Single-View Depth Probability with\n Multi-View Geometry"], "answer_arxiv_id": ["1611.07245", "2104.13325", "2112.08177"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_4555"} +{"question": "Which works introduced MMD-based calibration metrics?", "answer": ["Calibration tests beyond classification"], "answer_arxiv_id": ["2210.13355"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_4556"} +{"question": "Which works focused on creating early math datasets with basic math problems and equation-based solutions?", "answer": ["Program Induction by Rationale Generation : Learning to Solve and\n Explain Algebraic Word Problems"], "answer_arxiv_id": ["1705.04146"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_4557"} +{"question": "Which papers introduced or expanded on the concept of the Gromov-Wasserstein (GW) distance metric in the context of Optimal Transport (OT)?", "answer": ["Optimal Transport for structured data with application on graphs"], "answer_arxiv_id": ["1805.09114"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_4558"} +{"question": "Which works highlight the use of text aware loss for constraining text image super-resolution?", "answer": ["TextSR: Content-Aware Text Super-Resolution Guided by Recognition", "Scene Text Image Super-Resolution in the Wild"], "answer_arxiv_id": ["1909.07113", "2005.03341"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_4559"} +{"question": "Could you provide me some studies that use reduction of uncertainty under the learned model as intrinsic reward?", "answer": ["VIME: Variational Information Maximizing Exploration", "Self-Supervised Exploration via Disagreement"], "answer_arxiv_id": ["1605.09674", "1906.04161"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_4560"} +{"question": "What studies achieved horizon-free regret bounds for time-homogeneous linear mixture MDPs?", "answer": ["Improved Variance-Aware Confidence Sets for Linear Bandits and Linear Mixture MDP", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs"], "answer_arxiv_id": ["2101.12745", "2205.11507"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_4561"} +{"question": "Which paper incorporated graph neural networks into evolutionary algorithms to tackle large combinatorial spaces?", "answer": ["Neuroevolution-Enhanced Multi-Objective Optimization for Mixed-Precision Quantization"], "answer_arxiv_id": ["2106.07611v2"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_4562"} +{"question": "What works focused on using large-scale datasets of image-text pairs for training Vision-Language Models?", "answer": ["YFCC100M: The New Data in Multimedia Research", "Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize\n Long-Tail Visual Concepts", "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs"], "answer_arxiv_id": ["1503.01817", "2102.08981", "2111.02114"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_4563"} +{"question": "Which works are involved in the evolution of Vision Language (V-L) foundation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "FILIP: Fine-grained Interactive Language-Image Pre-Training", "Florence: A New Foundation Model for Computer Vision"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.07991", "2111.07783", "2111.11432"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_4564"} +{"question": "Who explored the technique of goal relabeling?", "answer": ["Hindsight Experience Replay"], "answer_arxiv_id": ["1707.01495v3"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4565"} +{"question": "Which work proposed the single-agent BatchLinUCB-DG, a system that attempts to save on communication?", "answer": ["Linear Bandits with Limited Adaptivity and Learning Distributional Optimal Design"], "answer_arxiv_id": ["2007.01980"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_4566"} +{"question": "Which works focus on multi-branch networks that select the suitable branches of networks to reduce the computation workload?", "answer": ["Dynamic Routing Networks", "Anytime Inference with Distilled Hierarchical Neural Ensembles", "Multi-Scale Dense Networks for Resource Efficient Image Classification"], "answer_arxiv_id": ["1905.04849", "2003.01474", "1703.09844"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_4567"} +{"question": "Which works have used cuboids for shape abstraction and object part estimation from a single image?", "answer": ["Learning Shape Abstractions by Assembling Volumetric Primitives", "Unsupervised Learning for Cuboid Shape Abstraction via Joint Segmentation from Point Clouds", "Im2Struct: Recovering 3D Shape Structure from a Single RGB Image"], "answer_arxiv_id": ["1612.00404", "2106.03437", "1804.05469"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_4568"} +{"question": "Could you provide me some works discussing OOD detection?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Selective Classification for Deep Neural Networks", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "To Trust Or Not To Trust A Classifier", "Hybrid Models for Open Set Recognition"], "answer_arxiv_id": ["1610.02136", "1705.08500", "1706.02690", "1612.01474", "1805.11783", "2003.12506"], "source_meta": {"published_time": "20220220"}, "qid": "AutoScholarQuery_train_4569"} +{"question": "Are there any studies that focus on improving adversarial robustness through using large generative models for data augmentation or synthetically generating more training samples?", "answer": ["Unlabeled Data Improves Adversarial Robustness", "Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examples", "Fixing Data Augmentation to Improve Adversarial Robustness", "Improving Robustness using Generated Data", "Generating High Fidelity Data from Low-density Regions using Diffusion Models"], "answer_arxiv_id": ["1905.13736", "2010.03593", "2103.01946", "2110.09468", "2203.17260"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_4570"} +{"question": "Which study uses localized and aspect-based feedback to iteratively refine outputs from LLMs?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback"], "answer_arxiv_id": ["2303.17651"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_4571"} +{"question": "What 2D human pose estimators are used for 2D pose predictions?", "answer": ["Cascaded deep monocular 3D human pose estimation with evolutionary\n training data", "Deep High-Resolution Representation Learning for Visual Recognition", "Source-free Domain Adaptive Human Pose Estimation", "GaitSADA: Self-Aligned Domain Adaptation for mmWave Gait Recognition"], "answer_arxiv_id": ["2006.07778", "1908.07919", "2308.03202", "2301.13384"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_4572"} +{"question": "Could you provide me a paper which proposed a new paraphrase robust watermarking method 'XMark' based on 'text redundancy' of text watermark?", "answer": ["WatME: Towards Lossless Watermarking Through Lexical Redundancy"], "answer_arxiv_id": ["2311.09832"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_4573"} +{"question": "Could you provide me some works that consider OOD as an essential topic in offline reinforcement learning?", "answer": ["Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems", "Rethinking Goal-Conditioned Supervised Learning and Its Connection to Offline RL", "Towards Robust Offline-to-Online Reinforcement Learning via Uncertainty and Smoothness", "RORL: Robust Offline Reinforcement Learning via Conservative Smoothing", "What is Essential for Unseen Goal Generalization of Offline Goal-conditioned RL?"], "answer_arxiv_id": ["2005.01643", "2202.04478", "2309.16973", "2206.02829", "2305.18882"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_4574"} +{"question": "Which paper introduces Influence functions and shapley values for quantifying the role of individual samples in data attribution?", "answer": ["Understanding Black-box Predictions via Influence Functions", "Data Shapley: Equitable Valuation of Data for Machine Learning"], "answer_arxiv_id": ["1703.04730", "1904.02868v2"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_4575"} +{"question": "Which works are associated with minimizing various distance metrics to reduce feature discrepancy in Domain Adaptive Object Detection?", "answer": ["Learning Transferable Features with Deep Adaptation Networks", "Deep Transfer Network: Unsupervised Domain Adaptation", "Unsupervised Domain Adaptation with Residual Transfer Networks", "Mind the Class Weight Bias: Weighted Maximum Mean Discrepancy for Unsupervised Domain Adaptation", "Domain-Adversarial Training of Neural Networks", "Deep Transfer Learning with Joint Adaptation Networks", "Spatial Attention Pyramid Network for Unsupervised Domain Adaptation", "Deep Transfer Network: Unsupervised Domain Adaptation"], "answer_arxiv_id": ["1502.02791", "1503.00591", "1602.04433", "1705.00609", "1505.07818", "1605.06636", "2003.12979", "1503.00591"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_4576"} +{"question": "Which datasets have been used for human motion capture in large-scale scenes?", "answer": ["HSC4D: Human-centered 4D Scene Capture in Large-scale Indoor-outdoor\n Space Using Wearable IMUs and LiDAR", "SLOPER4D: A Scene-Aware Dataset for Global 4D Human Pose Estimation in\n Urban Environments", "CIMI4D: A Large Multimodal Climbing Motion Dataset under Human-scene\n Interactions"], "answer_arxiv_id": ["2203.09215", "2303.09095", "2303.17948"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_4577"} +{"question": "What work involves quantile regularization in improving calibration during training by incorporating regularization techniques?", "answer": ["Quantile Regularization: Towards Implicit Calibration of Regression Models"], "answer_arxiv_id": ["2002.12860"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_4578"} +{"question": "What studies introduced supervised methods for image denoising?", "answer": ["Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image\n Denoising", "FFDNet: Toward a Fast and Flexible Solution for CNN based Image\n Denoising", "Toward Convolutional Blind Denoising of Real Photographs", "Real Image Denoising with Feature Attention", "Dual Adversarial Network: Toward Real-world Noise Removal and Noise\n Generation"], "answer_arxiv_id": ["1608.03981", "1710.04026", "1807.04686", "1904.07396v2", "2007.05946"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_4579"} +{"question": "Which works discuss anchor-free detectors in oriented object detection?", "answer": ["FCOS: Fully Convolutional One-Stage Object Detection"], "answer_arxiv_id": ["1904.01355"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_4580"} +{"question": "Which works show that the Molecular Property Neural Networks (MPNNs) can imitate classical graph algorithms like the Bellman-Ford and Prim’s algorithm?", "answer": ["Neural Execution of Graph Algorithms"], "answer_arxiv_id": ["1910.10593"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_4581"} +{"question": "What studies employed input or output filters for detecting and blocking harmful user prompts?", "answer": ["Baseline Defenses for Adversarial Attacks Against Aligned Language\n Models", "LLM Self Defense: By Self Examination, LLMs Know They Are Being Tricked", "SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks"], "answer_arxiv_id": ["2309.00614", "2308.07308", "2310.03684"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_4582"} +{"question": "Who introduced GPTScore and lays the premise that a generative pre-training model is likely to assign a high probability to the generation of high-quality text?", "answer": ["GPTScore: Evaluate as You Desire"], "answer_arxiv_id": ["2302.04166"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_4583"} +{"question": "Which papers have used a relation interaction mechanism and spatio-temporal message passing mechanism through NRI-MPM?", "answer": ["Neural Relational Inference with Efficient Message Passing Mechanisms"], "answer_arxiv_id": ["2101.09486"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_4584"} +{"question": "What are the studies showcasing great success of transformer models in language understanding?", "answer": ["XLNet: Generalized Autoregressive Pretraining for Language Understanding", "ALBERT: A Lite BERT for Self-supervised Learning of Language Representations", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators"], "answer_arxiv_id": ["1906.08237", "1909.11942", "1910.10683", "2003.10555"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_4585"} +{"question": "Which papers proposed fine-tuning techniques for incorporating the layout condition into a pre-trained diffusion model?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "SpaText: Spatio-Textual Representation for Controllable Image Generation", "LayoutDiffuse: Adapting Foundational Diffusion Models for\n Layout-to-Image Generation", "Freestyle Layout-to-Image Synthesis", "SSMG: Spatial-Semantic Map Guided Diffusion Model for Free-form\n Layout-to-Image Generation"], "answer_arxiv_id": ["2302.05543", "2211.14305", "2302.08908", "2303.14412", "2308.10156"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_4586"} +{"question": "Which studies are associated with the reverse-MI approaches in terms of unsupervised skill discovery?", "answer": ["Variational Intrinsic Control", "Diversity is All You Need: Learning Skills without a Reward Function", "Variational Option Discovery Algorithms", "Fast Task Inference with Variational Intrinsic Successor Features"], "answer_arxiv_id": ["1611.07507", "1802.06070", "1807.10299", "1906.05030"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_4587"} +{"question": "What studies used structural causal models to formalize and solve causal problems in machine learning?", "answer": ["On Calibration and Out-of-domain Generalization"], "answer_arxiv_id": ["2102.10395"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_4588"} +{"question": "What works have applied attention mechanisms in text-guided image synthesis?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks"], "answer_arxiv_id": ["1711.10485"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_4589"} +{"question": "Which study first formally proposed the problem of dataset distillation?", "answer": ["Dataset Distillation"], "answer_arxiv_id": ["1811.10959"], "source_meta": {"published_time": "20221119"}, "qid": "AutoScholarQuery_train_4590"} +{"question": "What study improved the inference-time of the forward diffusion process by relaxing the Markov assumption?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4591"} +{"question": "Could you provide me some studies that proposed new segmentation architectures?", "answer": ["Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "Segmenter: Transformer for Semantic Segmentation", "SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers"], "answer_arxiv_id": ["2012.15840", "2107.06278", "2112.01527", "2105.05633", "2105.15203"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_4592"} +{"question": "Are there any works that focus on studying transfer when pre-training with a contrastive objective?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Contrastive learning, multi-view redundancy, and linear models", "Contrastive estimation reveals topic posterior information to linear models"], "answer_arxiv_id": ["1902.09229", "2008.10150", "2003.02234"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_4593"} +{"question": "Could you give me examples of the studies that optimize transmitted and reflected radiance separately to counter limitations in NeRF’s lighting components?", "answer": ["Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance Fields", "NeRFReN: Neural Radiance Fields with Reflections"], "answer_arxiv_id": ["2112.03907", "2111.15234"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_4594"} +{"question": "Which work is about random perturbation of the input query for response selection?", "answer": ["SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks"], "answer_arxiv_id": ["2310.03684"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_4595"} +{"question": "What are the studies that further improved the pseudo-labels with more advanced Test-Time Augmentation loss?", "answer": ["If your data distribution shifts, use self-learning", "Test-Time Adaptation via Conjugate Pseudo-labels"], "answer_arxiv_id": ["2104.12928", "2207.09640"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_4596"} +{"question": "Which works show the application of diffusion model in image generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["1503.03585", "2010.02502", "2105.05233"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_4597"} +{"question": "Which survey was about deep techniques for stereo matching?", "answer": ["A Survey on Deep Learning Techniques for Stereo-based Depth Estimation"], "answer_arxiv_id": ["2006.02535"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_4598"} +{"question": "What papers discuss existence work in hyperspherical uniformity?", "answer": ["Learning towards Minimum Hyperspherical Energy", "Regularizing Neural Networks via Minimizing Hyperspherical Energy", "Learning with Hyperspherical Uniformity"], "answer_arxiv_id": ["1805.09298", "1906.04892", "2103.01649"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_4599"} +{"question": "Which study developed a Multi-Realism model that merges the ELIC codec and PatchGAN for neural image compression?", "answer": ["Multi-Realism Image Compression with a Conditional Generator"], "answer_arxiv_id": ["2212.13824"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_4600"} +{"question": "What works propose deriving the Bayesian inference from a subspace of the parameter space spanned by a few vectors derived from principal component analysis?", "answer": ["Subspace Inference for Bayesian Deep Learning"], "answer_arxiv_id": ["1907.07504"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_4601"} +{"question": "What works have been conducted on utilizing Optimal Transport Theory and Wasserstein Metric as computational tools in generative computer vision models?", "answer": ["Earth Movers in The Big Data Era: A Review of Optimal Transport in Machine Learning", "Optimal Transport for Parameter Identification of Chaotic Dynamics via Invariant Measures"], "answer_arxiv_id": ["2305.05080", "2104.15138"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_4602"} +{"question": "Can you give examples of studies that consider using task-specific finetuning and distillation in their pretraining?", "answer": ["Big Self-Supervised Models are Strong Semi-Supervised Learners"], "answer_arxiv_id": ["2006.10029"], "source_meta": {"published_time": "20220814"}, "qid": "AutoScholarQuery_train_4603"} +{"question": "What studies are associated with static dataset-based approaches using domain-specific questions or tasks for LLM evaluation?", "answer": ["Think you have Solved Question Answering? Try ARC, the AI2 Reasoning\n Challenge", "Measuring Massive Multitask Language Understanding", "C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for\n Foundation Models", "TruthfulQA: Measuring How Models Mimic Human Falsehoods", "Training Verifiers to Solve Math Word Problems"], "answer_arxiv_id": ["1803.05457", "2009.03300", "2305.08322", "2109.07958", "2110.14168"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_4604"} +{"question": "What are the research papers discussing the development and importance of various VideoQA datasets?", "answer": ["ActivityNet-QA: A Dataset for Understanding Complex Web Videos via\n Question Answering", "HERO: Hierarchical Encoder for Video+Language Omni-representation\n Pre-training", "NExT-QA:Next Phase of Question-Answering to Explaining Temporal Actions"], "answer_arxiv_id": ["1906.02467", "2005.00200", "2105.08276"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_4605"} +{"question": "Could you give me examples of research that uses pre-trained RoBERTa and fine-tunes it as a classifier in post-hoc text detection?", "answer": ["Release Strategies and the Social Impacts of Language Models", "Automatic Detection of Generated Text is Easiest when Humans are Fooled", "How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation,\n and Detection", "DPIC: Decoupling Prompt and Intrinsic Characteristics for LLM Generated\n Text Detection"], "answer_arxiv_id": ["1908.09203v2", "1911.00650", "2301.07597", "2305.12519"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_4606"} +{"question": "Could you provide me with an example of work that developed a fully convolutional localization network for dense captioning?", "answer": ["DenseCap: Fully Convolutional Localization Networks for Dense Captioning"], "answer_arxiv_id": ["1511.07571"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4607"} +{"question": "Could you provide me with studies that use sparse MoE for multimodal modeling?", "answer": ["Multimodal Contrastive Learning with LIMoE: the Language-Image Mixture of Experts", "Scaling Vision-Language Models with Sparse Mixture of Experts"], "answer_arxiv_id": ["2206.02770", "2303.07226"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_4608"} +{"question": "Which papers demonstrated the use of various demonstration configuration techniques in the context of QA?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?", "Self-Generated In-Context Learning: Leveraging Auto-regressive Language\n Models as a Demonstration Generator"], "answer_arxiv_id": ["2101.06804", "2206.08082"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4609"} +{"question": "Could you provide me some research about stochastic majorization-minimization (SMM)-type algorithms for constrained and nonconvex problems?", "answer": ["Online matrix factorization for Markovian data and applications to Network Dictionary Learning"], "answer_arxiv_id": ["1911.01931"], "source_meta": {"published_time": "20220329"}, "qid": "AutoScholarQuery_train_4610"} +{"question": "Could you tell me what works have proposed solutions for the problems faced by NeRF when given sparse inputs?", "answer": ["NeRF++: Analyzing and Improving Neural Radiance Fields", "RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs"], "answer_arxiv_id": ["2010.07492", "2112.00724"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_4611"} +{"question": "Could you provide me some studies about achieving precise credit assignment by leveraging path-wise derivatives?", "answer": ["Gradient Estimation Using Stochastic Computation Graphs", "Learning Continuous Control Policies by Stochastic Value Gradients", "Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1506.05254", "1510.09142", "1912.01603", "2010.02193"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_4612"} +{"question": "What studies have shown considerable improvement in large language models when guided through a step-by-step reasoning process?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2201.11903", "2203.11171"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_4613"} +{"question": "What are the references that investigate the effect of increasing pretraining data on the performance of large language models?", "answer": ["Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2203.15556"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_4614"} +{"question": "Could you provide me some studies about interpolating position embeddings?", "answer": ["Extending Context Window of Large Language Models via Positional\n Interpolation"], "answer_arxiv_id": ["2306.15595"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_4615"} +{"question": "Which studies have sought to enhance depth estimation accuracy in monocular 3D object detection by harnessing the synergy between 2D-3D geometric relationships?", "answer": ["Geometry Uncertainty Projection Network for Monocular 3D Object\n Detection", "Geometry-based Distance Decomposition for Monocular 3D Object Detection", "RTM3D: Real-time Monocular 3D Detection from Object Keypoints for\n Autonomous Driving", "Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular\n 3D Object Detection", "Disentangling Monocular 3D Object Detection"], "answer_arxiv_id": ["2107.13774", "2104.03775", "2001.03343", "2205.09373", "1905.12365"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_4616"} +{"question": "Which study demonstrated that OFW attains O(T2/3) regret for strongly convex functions?", "answer": ["Revisiting Projection-free Online Learning: the Strongly Convex Case"], "answer_arxiv_id": ["2010.07572"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4617"} +{"question": "Which studies introduced efficient methods for modeling long sequences with CNNs and transformers?", "answer": ["CKConv: Continuous Kernel Convolution For Sequential Data", "FlexConv: Continuous Kernel Convolutions with Differentiable Kernel Sizes", "Classification of Long Sequential Data using Circular Dilated Convolutional Neural Networks"], "answer_arxiv_id": ["2102.02611", "2110.08059", "2201.02143"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_4618"} +{"question": "What research has been done on discrete diffusion models using binomial noises and multinomial/categorical noises?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Structured Denoising Diffusion Models in Discrete State-Spaces", "Argmax Flows and Multinomial Diffusion: Learning Categorical Distributions"], "answer_arxiv_id": ["1503.03585", "2107.03006", "2102.05379"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_4619"} +{"question": "Which papers documented novel episodic memory structures like associative and generalized memory?", "answer": ["Generalizable Episodic Memory for Deep Reinforcement Learning"], "answer_arxiv_id": ["2103.06469"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_4620"} +{"question": "Which works utilized optical flow for aligning adjacent frames to enhance the quality of compressed videos?", "answer": ["Multi-Frame Quality Enhancement for Compressed Video", "MFQE 2.0: A New Approach for Multi-frame Quality Enhancement on\n Compressed Video"], "answer_arxiv_id": ["1803.04680", "1902.09707"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4621"} +{"question": "Could you provide me some publications studying the role of depth by analyzing the signal propagation and evolution of the NTK in MLPs?", "answer": ["Exponential expressivity in deep neural networks through transient chaos", "Exact Convergence Rates of the Neural Tangent Kernel in the Large Depth Limit"], "answer_arxiv_id": ["1606.05340", "1905.13654"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_4622"} +{"question": "What research incorporate backdoor as a technology to integrate verifiable watermark information in deep learning models?", "answer": ["Untargeted Backdoor Watermark: Towards Harmless and Stealthy Dataset\n Copyright Protection", "Watermarking Vision-Language Pre-trained Models for Multi-modal\n Embedding as a Service", "Are You Copying My Model? Protecting the Copyright of Large Language\n Models for EaaS via Backdoor Watermark"], "answer_arxiv_id": ["2210.00875", "2311.05863", "2305.10036"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_4623"} +{"question": "What works proposed fine-tuning a large model like LLMs?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2302.13971", "2103.00020"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_4624"} +{"question": "Which papers fall under the first line of work that focuses on translating visual information into a textual format for LLMs to process?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Inner Monologue: Embodied Reasoning through Planning with Language\n Models", "EmbodiedGPT: Vision-Language Pre-Training via Embodied Chain of Thought", "NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large\n Language Models", "LM-Nav: Robotic Navigation with Large Pre-Trained Models of Language,\n Vision, and Action", "Code as Policies: Language Model Programs for Embodied Control"], "answer_arxiv_id": ["2204.01691", "2207.05608", "2305.15021", "2305.16986", "2207.04429", "2209.07753"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_4625"} +{"question": "What studies focus on online linear Markov games?", "answer": ["Learning Zero-Sum Simultaneous-Move Markov Games Using Function Approximation and Correlated Equilibrium"], "answer_arxiv_id": ["2002.07066"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_4626"} +{"question": "What studies analyse FTRL with optimism and adaptivity using local-norm?", "answer": ["A Survey of Algorithms and Analysis for Adaptive Online Learning", "Pushing the Efficiency-Regret Pareto Frontier for Online Learning of Portfolios and Quantum States"], "answer_arxiv_id": ["1403.3465", "2202.02765"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_4627"} +{"question": "Which research has questioned the existence of emergent abilities in LLMs?", "answer": ["Are Emergent Abilities of Large Language Models a Mirage?"], "answer_arxiv_id": ["2304.15004"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_4628"} +{"question": "What studies proposed mitigation strategies to quantize in the presence of outliers?", "answer": ["SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models"], "answer_arxiv_id": ["2211.10438"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4629"} +{"question": "Which works used domain-specific feedback such as compilers in their research?", "answer": ["Graph-based, Self-Supervised Program Repair from Diagnostic Feedback"], "answer_arxiv_id": ["2005.10636"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_4630"} +{"question": "Which works proposed training an image description generator to improve the open-set object recognition performance?", "answer": ["Zero-Shot Out-of-Distribution Detection Based on the Pre-trained Model CLIP"], "answer_arxiv_id": ["2109.02748"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_4631"} +{"question": "Could you provide literature that developed provably efficient algorithms for Markov games?", "answer": ["Independent Policy Gradient Methods for Competitive Reinforcement Learning", "Fast Policy Extragradient Methods for Competitive Games with Entropy Regularization", "Provably Efficient Policy Optimization for Two-Player Zero-Sum Markov Games", "Independent Policy Gradient for Large-Scale Markov Potential Games: Sharper Rates, Function Approximation, and Game-Agnostic Convergence", "Faster Last-iterate Convergence of Policy Optimization in Zero-Sum Markov Games"], "answer_arxiv_id": ["2101.04233v1", "2105.15186", "2102.08903v2", "2202.04129", "2210.01050v2"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_4632"} +{"question": "What works discuss about enhancing the self-attention mechanism in Vision Transformer?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows", "Lite Vision Transformer with Enhanced Self-Attention", "Vision Transformer with Deformable Attention", "Learned Queries for Efficient Local Attention"], "answer_arxiv_id": ["2103.14030", "2107.00652", "2112.10809", "2201.00520", "2112.11435"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_4633"} +{"question": "Could you name the works that use Bayesian neural networks for uncertainty quantification?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "DropConnect Is Effective in Modeling Uncertainty of Bayesian Deep Networks"], "answer_arxiv_id": ["1506.02142", "1906.04569"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_4634"} +{"question": "What research works discuss the application of ideas similar to answer set programming?", "answer": ["NeurASP: Embracing Neural Networks into Answer Set Programming"], "answer_arxiv_id": ["2307.07700"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_4635"} +{"question": "Does any research attempt to minimize the stationarity gap of the population function privately?", "answer": ["Differentially Private Stochastic Optimization: New Results in Convex and Non-Convex Settings"], "answer_arxiv_id": ["2107.05585"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_4636"} +{"question": "Are there works related to symmetry detection?", "answer": ["Joint Cuts and Matching of Partitions in One Graph", "Detecting Approximate Reflection Symmetry in a Point Set using Optimization on Manifold", "Reflection and Rotation Symmetry Detection via Equivariant Learning"], "answer_arxiv_id": ["1711.09584", "1706.08801", "2203.16787"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_4637"} +{"question": "Could you provide me some works which used the mixture of experts (MoE) model for specific multi-task applications?", "answer": ["Beyond Distillation: Task-level Mixture-of-Experts for Efficient Inference"], "answer_arxiv_id": ["2110.03742"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_4638"} +{"question": "Which research significantly enhanced the rendering quality by optimizing MLPs to encode 5D radiance fields, described as NeRF?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_4639"} +{"question": "Which work further explored orientation symmetry in an oriented simplicial complex (SC) in simplicial learning?", "answer": ["Convolutional Learning on Simplicial Complexes"], "answer_arxiv_id": ["2301.11163"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_4640"} +{"question": "Which studies discussed the use of fine-grained feedback for evaluating and training better models?", "answer": ["ALICE: Active Learning with Contrastive Natural Language Explanations", "Training Language Models with Language Feedback at Scale"], "answer_arxiv_id": ["2009.10259", "2303.16755"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_4641"} +{"question": "Can you provide studies that investigated the relationship between the number of manipulated data points and the success of backdoor attacks?", "answer": ["Excess Capacity and Backdoor Poisoning", "Backdoor Learning Curves: Explaining Backdoor Poisoning Beyond Influence Functions"], "answer_arxiv_id": ["2109.00685", "2106.07214"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_4642"} +{"question": "Which research suggests that information sharing between the text and vision modalities is skewed in Visio-linguistic compositionality?", "answer": ["Vision-and-Language or Vision-for-Language? On Cross-Modal Influence in Multimodal Transformers"], "answer_arxiv_id": ["2109.04448"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_4643"} +{"question": "What previous works dealt with the learning of audio-visual representations to comprehend the correlation between two distinct modalities from videos?", "answer": ["SoundNet: Learning Sound Representations from Unlabeled Video", "Ambient Sound Provides Supervision for Visual Learning", "Look, Listen and Learn", "Cooperative Learning of Audio and Video Models from Self-Supervised\n Synchronization", "Learning to Localize Sound Source in Visual Scenes", "The Sound of Pixels", "The Sound of Motions", "Music Gesture for Visual Sound Separation", "Robust Audio-Visual Instance Discrimination", "Audio-Visual Instance Discrimination with Cross-Modal Agreement", "Looking to Listen at the Cocktail Party: A Speaker-Independent\n Audio-Visual Model for Speech Separation", "Deep Multimodal Clustering for Unsupervised Audiovisual Learning", "DiffAVA: Personalized Text-to-Audio Generation with Visual Alignment", "Audio-Visual Class-Incremental Learning", "Class-Incremental Grouping Network for Continual Audio-Visual Learning"], "answer_arxiv_id": ["1610.09001", "1608.07017", "1705.08168", "1807.00230", "1803.03849", "1804.03160", "1904.05979", "2004.09476", "2103.15916", "2004.12943", "1804.03619", "1807.03094", "2305.12903", "2308.11073", "2309.05281"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_4644"} +{"question": "What studies have proposed the Straight-Through Estimator (STE) as a heuristic to train threshold networks?", "answer": ["Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation"], "answer_arxiv_id": ["1308.3432"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_4645"} +{"question": "In what papers masked image modeling is used to pretrain vision models?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "iBOT : Image BERT Pre-Training with Online Tokenizer", "Masked Autoencoders Are Scalable Vision Learners", "Masked Feature Prediction for Self-Supervised Visual Pre-Training"], "answer_arxiv_id": ["2106.08254", "2111.07832", "2111.06377", "2112.09133"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_4646"} +{"question": "Which works explored the length generalization ability of LLMs on compositional problems?", "answer": ["Exploring Length Generalization in Large Language Models"], "answer_arxiv_id": ["2207.04901"], "source_meta": {"published_time": "20240705"}, "qid": "AutoScholarQuery_train_4647"} +{"question": "Could you name some works that are the references for the original self-supervised learning tasks used in test-time training?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_4648"} +{"question": "Which paper applied insertion and deletion to Non-Autoregressive (NAR) models to address the lack of flexibility in NAR generation?", "answer": ["Levenshtein Transformer"], "answer_arxiv_id": ["1905.11006"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_4649"} +{"question": "Which research proposes a human-in-the-loop approach for content moderation?", "answer": ["Human-AI Collaboration via Conditional Delegation: A Case Study of\n Content Moderation"], "answer_arxiv_id": ["2204.11788"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_4650"} +{"question": "Could you provide me some studies about KD methods that consider knowledge as relations between such layers?", "answer": ["Heterogeneous Knowledge Distillation using Information Flow Modeling", "Learning Student Networks via Feature Embedding", "Coarse-To-Fine Incremental Few-Shot Learning"], "answer_arxiv_id": ["2005.00727", "1812.06597", "2111.14806"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_4651"} +{"question": "Which works have explored domain shift mitigation through self-training with pseudo labels in the field of Source-Free Domain Adaptation?", "answer": ["Domain Adaptation without Source Data", "ProxyMix: Proxy-based Mixup Training with Label Refinery for Source-Free\n Domain Adaptation", "BMD: A General Class-balanced Multicentric Dynamic Prototype Strategy\n for Source-free Domain Adaptation", "Source Data-absent Unsupervised Domain Adaptation through Hypothesis\n Transfer and Labeling Transfer"], "answer_arxiv_id": ["2007.01524", "2205.14566", "2204.02811", "2012.07297"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_4652"} +{"question": "Which papers have utilized mutual information to calculate generalization bounds?", "answer": ["Information-theoretic analysis of generalization capability of learning algorithms", "Learners that Use Little Information", "Generalization Bounds via Information Density and Conditional Information Density", "Reasoning About Generalization via Conditional Mutual Information"], "answer_arxiv_id": ["1705.07809", "1710.05233", "2005.08044", "2001.09122"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4653"} +{"question": "Could you provide me some studies that demonstrated the scaling law of text embeddings?", "answer": ["SGPT: GPT Sentence Embeddings for Semantic Search", "Large Dual Encoders Are Generalizable Retrievers", "Language Models are Universal Embedders"], "answer_arxiv_id": ["2202.08904", "2112.07899", "2310.08232"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_4654"} +{"question": "Any works on exploring learning subspace DMs and connecting the full space with Langevin dynamics?", "answer": ["Subspace Diffusion Generative Models"], "answer_arxiv_id": ["2205.01490"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_4655"} +{"question": "Which studies present single-loop algorithms for addressing nonconvex-concave problems?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems", "Hybrid Block Successive Approximation for One-Sided Non-Convex Min-Max Problems: Algorithms and Applications"], "answer_arxiv_id": ["1906.00331", "1902.08294"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_4656"} +{"question": "What studies have explored the utilisation of language to adjust robot plans with constraints or specify subgoals in multimodal tasks?", "answer": ["Correcting Robot Plans with Natural Language Feedback", "LILA: Language-Informed Latent Actions", "“No, to the Right” – Online Language Corrections for Robotic Manipulation via Shared Autonomy"], "answer_arxiv_id": ["2204.05186", "2111.03205", "2301.02555"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_4657"} +{"question": "Which works investigated the task robustness of LLMs?", "answer": ["A Comprehensive Capability Analysis of GPT-3 and GPT-3.5 Series Models"], "answer_arxiv_id": ["2303.10420v2"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_4658"} +{"question": "Are there any research papers about frequency domain learning for improving image deblurring?", "answer": ["Efficient Frequency Domain-based Transformers for High-Quality Image\n Deblurring", "Intriguing Findings of Frequency Selection for Image Deblurring"], "answer_arxiv_id": ["2211.12250", "2111.11745"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_4659"} +{"question": "Could you provide me studies that cover optimization-based DG methods related to Group Distributionally robust optimization (DRO)?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_4660"} +{"question": "What studies applied neural architecture search for exploring the quantization space in mixed-precision quantization?", "answer": ["Mixed Precision Quantization of ConvNets via Differentiable Neural\n Architecture Search", "Single Path One-Shot Neural Architecture Search with Uniform Sampling"], "answer_arxiv_id": ["1812.00090", "1904.00420"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_4661"} +{"question": "What works might help to explain the benefits of more complex fine-tuning mechanisms?", "answer": ["SpotTune: Transfer Learning through Adaptive Fine-tuning", "Side-Tuning: A Baseline for Network Adaptation via Additive Side Networks", "Better Fine-Tuning by Reducing Representational Collapse", "Universal Language Model Fine-tuning for Text Classification"], "answer_arxiv_id": ["1811.08737", "1912.13503", "2008.03156", "1801.06146"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_4662"} +{"question": "What papers use backward warping or forward warping in video frame interpolation architectures?", "answer": ["FILM: Frame Interpolation for Large Motion", "Asymmetric Bilateral Motion Estimation for Video Frame Interpolation", "IFRNet: Intermediate Feature Refine Network for Efficient Frame\n Interpolation", "Real-Time Intermediate Flow Estimation for Video Frame Interpolation", "AMT: All-Pairs Multi-Field Transforms for Efficient Frame Interpolation", "Softmax Splatting for Video Frame Interpolation"], "answer_arxiv_id": ["2202.04901", "2108.06815", "2205.14620", "2011.06294", "2304.09790", "2003.05534v1"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_4663"} +{"question": "What research papers present work on improving the expressivity of Graph Neural Networks by simulating higher-order Weisfeiler-Lehman tests?", "answer": ["Graph Matching Networks for Learning the Similarity of Graph Structured Objects", "On the Equivalence between Graph Isomorphism Testing and Function Approximation with GNNs", "Provably Powerful Graph Networks"], "answer_arxiv_id": ["1904.12787", "1905.12560", "1905.11136"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_4664"} +{"question": "What works proposed remedies to loss discontinuity and regression inconsistency in RBox-supervised oriented object detection through modulated losses?", "answer": ["SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated Objects", "Learning Modulated Loss for Rotated Object Detection"], "answer_arxiv_id": ["1811.07126", "1911.08299"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_4665"} +{"question": "What are some offline goal-conditioned reinforcement learning algorithms proposed to address out-of-distribution errors?", "answer": ["Actionable Models: Unsupervised Offline Reinforcement Learning of Robotic Skills", "Latent Plans for Task-Agnostic Offline Reinforcement Learning", "Conservative Q-Learning for Offline Reinforcement Learning", "Rethinking Goal-Conditioned Supervised Learning and Its Connection to Offline RL", "Model-Based Visual Planning with Self-Supervised Functional Distances", "Generalization with Lossy Affordances: Leveraging Broad Offline Data for Learning Visuomotor Tasks", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["2104.07749", "2209.08959", "2006.04779", "2202.04478", "2012.15373", "2210.06601", "2110.06169"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_4666"} +{"question": "Could you provide some studies about the random forest where several decision trees are trained using a randomly selected feature set?", "answer": ["Confidence Intervals for Random Forests: The Jackknife and the Infinitesimal Jackknife", "Generalized Random Forests"], "answer_arxiv_id": ["1311.4555", "1610.01271"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_4667"} +{"question": "Could you provide some studies on convolution-transformer hybrid architectures?", "answer": ["EfficientFormer: Vision Transformers at MobileNet Speed", "Rethinking Vision Transformers for MobileNet Size and Speed", "FastViT: A Fast Hybrid Vision Transformer using Structural\n Reparameterization", "FasterViT: Fast Vision Transformers with Hierarchical Attention", "SwiftFormer: Efficient Additive Attention for Transformer-based\n Real-time Mobile Vision Applications", "MobileViG: Graph-Based Sparse Attention for Mobile Vision Applications"], "answer_arxiv_id": ["2206.01191", "2212.08059", "2303.14189", "2306.06189", "2303.15446", "2307.00395"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4668"} +{"question": "What papers used cross-modality knowledge distillation to enhance detection performance in 3D object detection?", "answer": ["LIGA-Stereo: Learning LiDAR Geometry Aware Representations for\n Stereo-based 3D Detector", "MonoDistill: Learning Spatial Features for Monocular 3D Object Detection", "Cross-Modality Knowledge Distillation Network for Monocular 3D Object\n Detection", "Unifying Voxel-based Representation with Transformer for 3D Object\n Detection", "Boosting 3D Object Detection by Simulating Multimodality on Point Clouds", "BEVDistill: Cross-Modal BEV Distillation for Multi-View 3D Object\n Detection", "UniDistill: A Universal Cross-Modality Knowledge Distillation Framework\n for 3D Object Detection in Bird's-Eye View", "X$^3$KD: Knowledge Distillation Across Modalities, Tasks and Stages for\n Multi-Camera 3D Object Detection", "DistillBEV: Boosting Multi-Camera 3D Object Detection with Cross-Modal\n Knowledge Distillation"], "answer_arxiv_id": ["2108.08258", "2201.10830", "2211.07171", "2206.00630", "2206.14971", "2211.09386", "2303.15083", "2303.02203", "2309.15109"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_4669"} +{"question": "Could you provide the study that carries out a comprehensive evaluation of ChatGPT’s robustness from adversarial and out-of-distribution perspectives?", "answer": ["On the Robustness of ChatGPT: An Adversarial and Out-of-distribution Perspective"], "answer_arxiv_id": ["2302.12095"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_4670"} +{"question": "Which paper introduced the concept of directional smoothness to control the regret in a realizable, online classification problem?", "answer": ["Efficient and Near-Optimal Smoothed Online Learning for Generalized Linear Functions"], "answer_arxiv_id": ["2205.13056"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_4671"} +{"question": "What research works have utilized either the mixup mechanism or generative models to enrich diversity of local datasets in the area of federated learning?", "answer": ["mixup: Beyond Empirical Risk Minimization", "FedMix: Approximation of Mixup under Mean Augmented Federated Learning", "FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Continuous Frequency Space", "Federated Learning in Non-IID Settings Aided by Differentially Private Synthetic Data"], "answer_arxiv_id": ["1710.09412", "2107.00233", "2103.06030", "2206.00686"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_4672"} +{"question": "Which studies recognized the misspecification issues of DM, one of the benchmark estimators in off-policy evaluation?", "answer": ["More Robust Doubly Robust Off-policy Evaluation", "Empirical Study of Off-Policy Policy Evaluation for Reinforcement Learning"], "answer_arxiv_id": ["1802.03493", "1911.06854"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_4673"} +{"question": "Which work is similar to the research in that the LLM’s usage of tools is grounded by feedback?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools"], "answer_arxiv_id": ["2302.04761"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_4674"} +{"question": "What papers discuss the possibilities and benefits of distributed training of LMs to show lower overall perplexity?", "answer": ["Branch-Train-Merge: Embarrassingly Parallel Training of Expert Language Models"], "answer_arxiv_id": ["2208.03306"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_4675"} +{"question": "Could you tell me the studies that have been critical of the effectiveness of existing domain generalization methods?", "answer": ["The Risks of Invariant Risk Minimization", "Out-of-distribution Prediction with Invariant Risk Minimization: The Limitation and An Effective Fix", "In Search of Lost Domain Generalization"], "answer_arxiv_id": ["2010.05761", "2101.07732", "2007.01434"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_4676"} +{"question": "What papers proposed methods for diverse prompt tuning?", "answer": ["PLOT: Prompt Learning with Optimal Transport for Vision-Language Models"], "answer_arxiv_id": ["2210.01253"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_4677"} +{"question": "Which study analyzed clipping under bounded variance?", "answer": ["Stochastic Optimization with Heavy-Tailed Noise via Accelerated Gradient Clipping"], "answer_arxiv_id": ["2005.10785"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_4678"} +{"question": "What works studied reward-free RL under low-rank Markov Decision Processes (MDPs)?", "answer": ["FLAMBE: Structural Complexity and Representation Learning of Low Rank MDPs", "Model-free Representation Learning and Exploration in Low-rank MDPs"], "answer_arxiv_id": ["2006.10814", "2102.07035"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_4679"} +{"question": "Which research work establishes an end-to-end sample complexity bound on learning a robust LQG controller?", "answer": ["Sample Complexity of Linear Quadratic Gaussian (LQG) Control for Output Feedback Systems"], "answer_arxiv_id": ["2011.09929"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_4680"} +{"question": "What are some scholarly works that have explored traditional 2D, 3D CNNs and Transformer-based methods in video recognition?", "answer": ["Temporal Segment Networks: Towards Good Practices for Deep Action\n Recognition", "TSM: Temporal Shift Module for Efficient Video Understanding", "MutualNet: Adaptive ConvNet via Mutual Learning from Different Model\n Configurations", "Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset", "A Closer Look at Spatiotemporal Convolutions for Action Recognition", "Tube Convolutional Neural Network (T-CNN) for Action Detection in Videos", "AIM: Adapting Image Models for Efficient Video Action Recognition", "Is Space-Time Attention All You Need for Video Understanding?", "Video Swin Transformer", "Expanding Language-Image Pretrained Models for General Video Recognition", "Implicit Temporal Modeling with Learnable Alignment for Video\n Recognition", "Multiscale Vision Transformers", "Frozen CLIP Models are Efficient Video Learners"], "answer_arxiv_id": ["1608.00859", "1811.08383", "2105.07085", "1705.07750", "1711.11248", "1703.10664", "2302.03024", "2102.05095", "2106.13230", "2208.02816", "2304.10465", "2104.11227", "2208.03550"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_4681"} +{"question": "Which work proposed the use of norm of the Jacobian to measure oversquashing in GNNs?", "answer": ["Representation Learning on Graphs with Jumping Knowledge Networks"], "answer_arxiv_id": ["1806.03536"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_4682"} +{"question": "What studies use SDF-based methods in creating lifelike human models in human avatar reconstruction?", "answer": ["PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization", "Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view\n Human Reconstruction", "PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution\n 3D Human Digitization", "PaMIR: Parametric Model-Conditioned Implicit Representation for\n Image-based Human Reconstruction", "ICON: Implicit Clothed humans Obtained from Normals", "ECON: Explicit Clothed humans Optimized via Normal integration"], "answer_arxiv_id": ["1905.05172", "2006.08072", "2004.00452", "2007.03858", "2112.09127", "2212.07422"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4683"} +{"question": "Which studies used GANs in text-to-image generation?", "answer": ["Generative Adversarial Text to Image Synthesis", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks", "AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks"], "answer_arxiv_id": ["1605.05396", "1612.03242", "1711.10485"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4684"} +{"question": "Could you tell me the studies which use unified approaches to PEFT methods?", "answer": ["Towards a Unified View of Parameter-Efficient Transfer Learning", "AutoPEFT: Automatic Configuration Search for Parameter-Efficient\n Fine-Tuning", "Neural Architecture Search with Reinforcement Learning", "UniPELT: A Unified Framework for Parameter-Efficient Language Model\n Tuning"], "answer_arxiv_id": ["2110.04366", "2301.12132", "1611.01578", "2110.07577"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_4685"} +{"question": "Which works are about state-of-the-art solvers for linear programming?", "answer": ["Solving Linear Programs in the Current Matrix Multiplication Time"], "answer_arxiv_id": ["1810.07896"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_4686"} +{"question": "Which studies provided efficient methods to handle multi-view setups?", "answer": ["NeuS2: Fast Learning of Neural Implicit Surfaces for Multi-view\n Reconstruction", "MPS-NeRF: Generalizable 3D Human Rendering from Multiview Images", "Representing Volumetric Videos as Dynamic MLP Maps", "Mixed Neural Voxels for Fast Multi-view Video Synthesis"], "answer_arxiv_id": ["2212.05231", "2203.16875", "2304.06717", "2212.00190"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_4687"} +{"question": "What research work supports the concept of weak convergence in general non-log-concave sampling?", "answer": ["Towards a Theory of Non-Log-Concave Sampling: First-Order Stationarity Guarantees for Langevin Monte Carlo"], "answer_arxiv_id": ["2202.05214"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_4688"} +{"question": "What works designed supervisory signals based on the conditional GANs in text-driven image editing?", "answer": ["ManiGAN: Text-Guided Image Manipulation", "Text-Guided Neural Image Inpainting"], "answer_arxiv_id": ["1912.06203", "2004.03212"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_4689"} +{"question": "Can you identify the papers that showed the emergence of the evaluation ability of LLMs?", "answer": ["GPTScore: Evaluate as You Desire"], "answer_arxiv_id": ["2302.04166"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_4690"} +{"question": "Do you know the research that first showed that LLMs’ performance weakens in the middle of the prompt?", "answer": ["Lost in the Middle: How Language Models Use Long Contexts"], "answer_arxiv_id": ["2307.03172"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_4691"} +{"question": "What studies propose the idea of radio-visual SSL in the context of automative radar?", "answer": ["Multimodal contrastive learning for remote sensing tasks", "Self-Supervised Radio-Visual Representation Learning for 6G Sensing", "Look, Radiate, and Learn: Self-Supervised Localisation via Radio-Visual\n Correspondence"], "answer_arxiv_id": ["2209.02329", "2111.02887", "2206.06424"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4692"} +{"question": "Could you mention the work that further extending [bib.bib16] by using a general transformer architecture?", "answer": ["Attention, Learn to Solve Routing Problems!"], "answer_arxiv_id": ["1803.08475"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_4693"} +{"question": "What studies have explored the area of instruction tuning for Large Language Models?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Cross-Task Generalization via Natural Language Crowdsourcing\n Instructions", "Super-NaturalInstructions: Generalization via Declarative Instructions\n on 1600+ NLP Tasks"], "answer_arxiv_id": ["2109.01652", "2104.08773", "2204.07705"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_4694"} +{"question": "Which works in text-to-image generation used autoregressive models?", "answer": ["Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers", "M6: A Chinese Multimodal Pretrainer", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092", "2105.13290", "2103.00823", "2206.10789"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4695"} +{"question": "Can you point me to the papers that applied handcrafted approaches to solve the shape correspondence problem?", "answer": ["Efficient Deformable Shape Correspondence via Kernel Matching", "Continuous and Orientation-preserving Correspondences via Functional Maps"], "answer_arxiv_id": ["1707.08991", "1806.04455v3"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_4696"} +{"question": "What work proposed a novel prompting method for enhancing cultural diversity in LLM responses?", "answer": ["Improving Diversity of Demographic Representation in Large Language\n Models via Collective-Critiques and Self-Voting"], "answer_arxiv_id": ["2310.16523"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_4697"} +{"question": "Which works proposed a multi-stage progressive image recovery network?", "answer": ["Multi-Stage Progressive Image Restoration"], "answer_arxiv_id": ["2102.02808"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_4698"} +{"question": "What are some works that built standardized preprocessing pipelines, patient splits, and task definitions on top of EHR datasets?", "answer": ["Multitask learning and benchmarking with clinical time series data"], "answer_arxiv_id": ["1703.07771"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_4699"} +{"question": "What papers use attribute classifiers or heuristic constraints at decoding time in control of language models?", "answer": ["Gradient-Based Constrained Sampling from Language Models", "Plug and Play Language Models: A Simple Approach to Controlled Text\n Generation", "DExperts: Decoding-Time Controlled Text Generation with Experts and\n Anti-Experts", "FUDGE: Controlled Text Generation With Future Discriminators"], "answer_arxiv_id": ["2205.12558", "1912.02164", "2105.03023", "2104.05218"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_4700"} +{"question": "What research papers proposed the parameter-efficient fine-tuning techniques in the field of NLP?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Towards a Unified View of Parameter-Efficient Transfer Learning", "BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models", "Training Neural Networks with Fixed Sparse Masks"], "answer_arxiv_id": ["1902.00751", "2106.09685", "2104.08691", "2101.00190", "2110.04366", "2106.10199", "2111.09839"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_4701"} +{"question": "Which works have proposed the methods of Group Activity Recognition (GAR) where training requires only the labels of group activity?", "answer": ["Spatio-Temporal Dynamic Inference Network for Group Activity Recognition", "Detector-Free Weakly Supervised Group Activity Recognition"], "answer_arxiv_id": ["2108.11743", "2204.02139"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_4702"} +{"question": "Can you list recent works that used large language models (LLMs) to leverage GPT3 as an implicit knowledge source for reasoning in KVQA?", "answer": ["Language Models are Few-Shot Learners", "An Empirical Study of GPT-3 for Few-Shot Knowledge-Based VQA"], "answer_arxiv_id": ["2005.14165", "2109.05014"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_4703"} +{"question": "What works deal with hierarchical techniques in reinforcement learning?", "answer": ["Hierarchical Reinforcement Learning By Discovering Intrinsic Options", "Data-Efficient Hierarchical Reinforcement Learning", "Sub-policy Adaptation for Hierarchical Reinforcement Learning"], "answer_arxiv_id": ["2101.06521", "1805.08296", "1906.05862"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_4704"} +{"question": "Which papers leveraged the observed structure as regularization for learning representations in Graph-based SSL?", "answer": ["Revisiting Semi-Supervised Learning with Graph Embeddings"], "answer_arxiv_id": ["1603.08861"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_4705"} +{"question": "What studies have leveraged T2I models in zero-shot or few-shot ways to improvise T2V generation?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation", "TokenFlow: Consistent Diffusion Features for Consistent Video Editing", "LAMP: Learn A Motion Pattern for Few-Shot-Based Video Generation"], "answer_arxiv_id": ["2212.11565", "2303.09535", "2306.07954", "2307.10373", "2310.10769"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_4706"} +{"question": "Are there any works that demonstrate the ability to establish semantic correspondence or segmentation by optimizing text tokens and cross-attention maps?", "answer": ["Unsupervised Semantic Correspondence Using Stable Diffusion", "SLiMe: Segment Like Me"], "answer_arxiv_id": ["2305.15581", "2309.03179"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4707"} +{"question": "Could you provide me some work that showed the Decision-Estimation Coefficient (DEC) for B-stable PSRs is bounded?", "answer": ["Unified Algorithms for RL with Decision-Estimation Coefficients: No-Regret, PAC, and Reward-Free Learning", "The Statistical Complexity of Interactive Decision Making"], "answer_arxiv_id": ["2209.11745", "2112.13487v3"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_4708"} +{"question": "Are there any works that explore Multi-Turn Reinforcement Learning with Language Models using LMRL Gym?", "answer": ["LMRL Gym: Benchmarks for Multi-Turn Reinforcement Learning with Language\n Models"], "answer_arxiv_id": ["2311.18232"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_4709"} +{"question": "Could you mention the deep learning-based NiSID methods that adopted data-driven strategies?", "answer": ["NightHazeFormer: Single Nighttime Haze Removal Using Prior Query\n Transformer", "Enhancing Visibility in Nighttime Haze Images Using Guided APSF and\n Gradient Adaptive Convolution"], "answer_arxiv_id": ["2305.09533", "2308.01738"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_4710"} +{"question": "What studies propose methods for applying greedy local learning on model parallelism training or add a local reconstruction loss for preserving information?", "answer": ["Parallel Training of Deep Networks with Local Updates", "Revisiting Locally Supervised Learning: an Alternative to End-to-end Training"], "answer_arxiv_id": ["2012.03837", "2101.10832"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_4711"} +{"question": "Can you name the papers that used occupation fields method for model representation?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "Implicit Functions in Feature Space for 3D Shape Reconstruction and\n Completion"], "answer_arxiv_id": ["1812.03828", "2003.01456"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_4712"} +{"question": "What publications highlight the progress of learning-based approaches in point cloud registration?", "answer": ["3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions", "The Perfect Match: 3D Point Cloud Matching with Smoothed Densities", "PPFNet: Global Context Aware Local Features for Robust 3D Point Matching", "USIP: Unsupervised Stable Interest Point Detection from 3D Point Clouds"], "answer_arxiv_id": ["1603.08182", "1811.06879", "1802.02669", "1904.00229"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4713"} +{"question": "Which research papers discuss learning with differential privacy in the context of empirical risk minimization?", "answer": ["Differentially Private Empirical Risk Minimization"], "answer_arxiv_id": ["0912.0071v5"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_4714"} +{"question": "What work proposes learning visual features via language supervision from scratch?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4715"} +{"question": "Which papers utilize one-stage end-to-end paradigms in Scene Graph Generation?", "answer": ["Structured Sparse R-CNN for Direct Scene Graph Generation", "RelTR: Relation Transformer for Scene Graph Generation"], "answer_arxiv_id": ["2106.10815", "2201.11460"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_4716"} +{"question": "What papers discuss the usage of generative models as a data generator in spite of their limitations such as lack of quality control?", "answer": ["Denoising Diffusion Probabilistic Models", "Generative Adversarial Networks", "Erasing Concepts from Diffusion Models", "DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort", "Fake it till you make it: Learning transferable representations from synthetic ImageNet clones"], "answer_arxiv_id": ["2006.11239", "2203.00667", "2303.07345", "2104.06490", "2212.08420v2"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_4717"} +{"question": "Are there any research papers that adopt tri-plane-based neural fields representation?", "answer": ["One-Shot High-Fidelity Talking-Head Synthesis with Deformable Neural\n Radiance Field", "NOFA: NeRF-based One-shot Facial Avatar Reconstruction", "Generalizable One-shot Neural Head Avatar"], "answer_arxiv_id": ["2304.05097", "2307.03441", "2306.08768"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_4718"} +{"question": "Could you provide me some works that made further improvements of PGD-AT, particularly in terms of initialization?", "answer": ["Prior-Guided Adversarial Initialization for Fast Adversarial Training", "Boosting Fast Adversarial Training with Learnable Adversarial Initialization"], "answer_arxiv_id": ["2207.08859", "2110.05007"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_4719"} +{"question": "What papers are about methods that perform per-video fine-tuning for video editing?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Edit-A-Video: Single Video Editing with Object-Aware Consistency", "Video-P2P: Video Editing with Cross-attention Control"], "answer_arxiv_id": ["2212.11565", "2303.07945", "2303.04761"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_4720"} +{"question": "Which works used gradient-based methods in the context of offline model-based optimization?", "answer": ["Conservative Objective Models for Effective Offline Model-Based Optimization", "RoMA: Robust Model Adaptation for Offline Model-based Optimization", "Offline Model-Based Optimization via Normalized Maximum Likelihood Estimation", "Bidirectional Learning for Offline Infinite-width Model-based Optimization"], "answer_arxiv_id": ["2107.06882", "2110.14188", "2102.07970", "2209.07507"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_4721"} +{"question": "What works used hypothesis-based approaches in relative object pose estimation?", "answer": ["RelPose: Predicting Probabilistic Relative Rotation for Single Objects\n in the Wild", "RelPose++: Recovering 6D Poses from Sparse-view Observations", "3D-Aware Hypothesis & Verification for Generalizable Relative Object\n Pose Estimation"], "answer_arxiv_id": ["2208.05963", "2305.04926", "2310.03534"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_4722"} +{"question": "Can you list studies on multi-view representation learning that are statistic-based?", "answer": ["Multi-View Spectral Clustering via Structured Low-Rank Matrix\n Factorization"], "answer_arxiv_id": ["1709.01212"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_4723"} +{"question": "What papers discussed several other co-teaching schemes used in the literature?", "answer": ["Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels", "DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "UniCon: Combating Label Noise Through Uniform Selection and Contrastive Learning"], "answer_arxiv_id": ["1804.06872", "2002.07394", "2203.14542"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_4724"} +{"question": "Can you name the research papers that use dynamic architecture methods in overcoming catastrophic forgetting?", "answer": ["Progressive Neural Networks", "Progress & Compress: A scalable framework for continual learning", "Lifelong Learning with Dynamically Expandable Networks"], "answer_arxiv_id": ["1606.04671", "1805.06370", "1708.01547"], "source_meta": {"published_time": "20230807"}, "qid": "AutoScholarQuery_train_4725"} +{"question": "Which works have utilized Transformers for visual object tracking?", "answer": ["Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual\n Tracking", "Transformer Tracking", "Transforming Model Prediction for Tracking", "AiATrack: Attention in Attention for Transformer Visual Tracking", "Backbone is All Your Need: A Simplified Architecture for Visual Object\n Tracking", "Joint Feature Learning and Relation Modeling for Tracking: A One-Stream\n Framework", "Revisiting Color-Event based Tracking: A Unified Network, Dataset, and\n Metric"], "answer_arxiv_id": ["2103.11681", "2103.15436", "2203.11192", "2207.09603", "2203.05328", "2203.11991", "2211.11010"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_4726"} +{"question": "What is the study that introduced Snake activations and the improvement of codebook learning by projecting the encodings into a low-dimensional space?", "answer": ["Neural Networks Fail to Learn Periodic Functions and How to Fix It", "BigVGAN: A Universal Neural Vocoder with Large-Scale Training", "Vector-quantized Image Modeling with Improved VQGAN"], "answer_arxiv_id": ["2006.08195", "2206.04658", "2110.04627"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_4727"} +{"question": "What papers employed attention and Transformer to integrate most salient features for prediction in multiple instance learning (MIL)?", "answer": ["Attention-based Deep Multiple Instance Learning", "TransMIL: Transformer based Correlated Multiple Instance Learning for\n Whole Slide Image Classification", "Dual-stream Multiple Instance Learning Network for Whole Slide Image\n Classification with Self-supervised Contrastive Learning", "Diagnose Like a Pathologist: Transformer-Enabled Hierarchical\n Attention-Guided Multiple Instance Learning for Whole Slide Image\n Classification"], "answer_arxiv_id": ["1802.04712", "2106.00908", "2011.08939", "2301.08125"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_4728"} +{"question": "What papers have explored transformer-based model learning to induce causal structures from observational and interventional data?", "answer": ["Learning to Induce Causal Structure"], "answer_arxiv_id": ["2204.04875"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4729"} +{"question": "Any works about utilizing neural network representations for volumetric reconstructions?", "answer": ["DeepVoxels: Learning Persistent 3D Feature Embeddings", "Neural Volumes: Learning Dynamic Renderable Volumes from Images", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["1812.01024", "1906.07751", "1906.01618", "2003.08934"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_4730"} +{"question": "Do any studies explain DNN decision at the level of individual image pixels?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "Interpretable Explanations of Black Boxes by Meaningful Perturbation", "Visualizing Deep Convolutional Neural Networks Using Natural Pre-Images", "Learning visual explanations for DCNN-based image classifiers using an attention mechanism", "TAME: Attention Mechanism Based Feature Fusion for Generating Explanation Maps of Convolutional Neural Networks"], "answer_arxiv_id": ["1312.6034", "1704.03296", "1512.02017v3", "2209.11189", "2301.07407v1"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_4731"} +{"question": "What papers discuss that the phenomenon of neural networks learning similar representations for semantically similar data is particularly pronounced for large and wide models?", "answer": ["Can Neural Nets Learn the Same Model Twice? Investigating Reproducibility and Double Descent from the Decision Boundary Perspective", "You Only Need a Good Embeddings Extractor to Fix Spurious Correlations"], "answer_arxiv_id": ["2203.08124", "2212.06254"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_4732"} +{"question": "Are there any papers on the use of wavelets and tree-structured dilated causal convolutions in convolutional sequence models?", "answer": ["WaveNet: A Generative Model for Raw Audio"], "answer_arxiv_id": ["1609.03499"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_4733"} +{"question": "Which works introduced artificial augmentations like image corruptions and perturbations for testing models' robustness?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["1903.12261"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_4734"} +{"question": "What studies present the development of memory network in the area of natural language processing?", "answer": ["Ask Me Anything: Dynamic Memory Networks for Natural Language Processing", "Dynamic Memory Networks for Visual and Textual Question Answering", "Neural Turing Machines", "End-To-End Memory Networks"], "answer_arxiv_id": ["1506.07285", "1603.01417", "1410.5401", "1503.08895"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4735"} +{"question": "What studies proposed to debias face recognition models through model pruning?", "answer": ["FairGRAPE: Fairness-aware GRAdient Pruning mEthod for Face Attribute Classification"], "answer_arxiv_id": ["2207.10888"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_4736"} +{"question": "Which papers discussed generation of foggy images using physical means like fog or haze machines?", "answer": ["O-HAZE: a dehazing benchmark with real hazy and haze-free outdoor images", "Dense Haze: A benchmark for image dehazing with dense-haze and haze-free\n images", "NH-HAZE: An Image Dehazing Benchmark with Non-Homogeneous Hazy and\n Haze-Free Images", "A Multi-purpose Realistic Haze Benchmark with Quantifiable Haze Levels and Ground Truth"], "answer_arxiv_id": ["1804.05101", "1904.02904", "2005.03560", "2206.06427v3"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_4737"} +{"question": "What research provided 360-degree view multimodal data to advance single-vehicle perception research?", "answer": ["nuScenes: A multimodal dataset for autonomous driving"], "answer_arxiv_id": ["1903.11027"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4738"} +{"question": "Could you mention some examples of research on Visual Dialog, where a dialog about an image is undertaken?", "answer": ["Visual Dialog", "Large-scale Pretraining for Visual Dialog: A Simple State-of-the-Art Baseline"], "answer_arxiv_id": ["1611.08669", "1912.02379"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4739"} +{"question": "Which works incorporated diffusion models for self-supervised pre-training?", "answer": ["Label-Efficient Semantic Segmentation with Diffusion Models"], "answer_arxiv_id": ["2112.03126"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_4740"} +{"question": "Could you provide some studies that focus on federated learning?", "answer": ["Federated Learning: Strategies for Improving Communication Efficiency", "Communication-Efficient Learning of Deep Networks from Decentralized Data", "Federated Machine Learning: Concept and Applications", "Federated Optimization in Heterogeneous Networks"], "answer_arxiv_id": ["1610.05492", "1602.05629", "1902.04885", "1812.06127"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_4741"} +{"question": "In what papers are the latent dynamics models used in visually complex domains, and which are these domains?", "answer": ["dm_control: Software and Tasks for Continuous Control", "ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning", "DeepMind Lab"], "answer_arxiv_id": ["2006.12983", "1605.02097", "1612.03801"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_4742"} +{"question": "Which research works have developed approaches that learn pose-sensitive embeddings for subsequent pose retrieval?", "answer": ["Implicit 3D Orientation Learning for 6D Object Detection from RGB Images"], "answer_arxiv_id": ["1902.01275"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_4743"} +{"question": "Which studies uncover the importance of subnetworks using pruning technique?", "answer": ["The State of Sparsity in Deep Neural Networks", "Deep Compression: Compressing Deep Neural Networks with Pruning, Trained\n Quantization and Huffman Coding", "Pruning Filters for Efficient ConvNets", "Learning Structured Sparsity in Deep Neural Networks"], "answer_arxiv_id": ["1902.09574", "1510.00149", "1608.08710", "1608.03665"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_4744"} +{"question": "Can you specify research which found properties of retinal sampling useful for enhancing performance of neural networks training?", "answer": ["Peripheral Vision Transformer", "FOVEA: Foveated Image Magnification for Autonomous Navigation", "FoveaTer: Foveated Transformer for Image Classification"], "answer_arxiv_id": ["2206.06801", "2108.12102", "2105.14173"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_4745"} +{"question": "Could you find me papers that studied linear mixture MDPs?", "answer": ["Model-Based Reinforcement Learning with Value-Targeted Regression", "Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles", "Provably Efficient Exploration in Policy Optimization", "Improved Variance-Aware Confidence Sets for Linear Bandits and Linear Mixture MDP", "Near-optimal Policy Optimization Algorithms for Learning Adversarial Linear Mixture MDPs", "Provably Efficient Reinforcement Learning for Discounted MDPs with Feature Mapping", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs"], "answer_arxiv_id": ["2006.01107", "1910.10597", "1912.05830", "2101.12745", "2102.08940v2", "2006.13165", "2012.08507", "2205.11507"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_4746"} +{"question": "What works applied Vision Transformers to the field of semantic segmentation?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction\n without Convolutions", "CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image\n Classification", "CvT: Introducing Convolutions to Vision Transformers", "Tokens-to-Token ViT: Training Vision Transformers from Scratch on\n ImageNet", "Transformer in Transformer", "MetaFormer Is Actually What You Need for Vision"], "answer_arxiv_id": ["2010.11929", "2103.14030", "2102.12122", "2103.14899", "2103.15808", "2101.11986", "2103.00112", "2111.11418"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_4747"} +{"question": "Could you provide me some papers about the involvement of transformer in the segmentation domain leading to different novel segmentation architectures?", "answer": ["Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2012.15840", "2107.06278", "2112.01527"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_4748"} +{"question": "Which papers developed approaches that considered irreducible representations to construct roto-translational equivariant neural networks?", "answer": ["Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "Geometric and Physical Quantities improve E(3) Equivariant Message Passing"], "answer_arxiv_id": ["2206.11990", "2006.10503", "2110.02905"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_4749"} +{"question": "What prior studies represent objects as localized object-centric patches?", "answer": ["SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition", "Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects"], "answer_arxiv_id": ["2001.02407", "1603.08575", "1806.01794"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_4750"} +{"question": "What works proposed the strong performance of training objective in sequence modeling and planning tasks?", "answer": ["Imagination is All You Need! Curved Contrastive Learning for Abstract\n Sequence Modeling Utilized on Long Short-Term Dialogue Planning"], "answer_arxiv_id": ["2211.07591"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_4751"} +{"question": "Could you name some research that approaches object-centric learning from the perspective of contrastive and self-supervised learning?", "answer": ["Slot Contrastive Networks: A Contrastive Approach for Representing Objects", "Learning Object-Centric Video Models by Contrasting Sets", "Online Object Representations with Contrastive Learning", "Towards Self-Supervised Learning of Global and Object-Centric Representations"], "answer_arxiv_id": ["2007.09294v1", "2011.10287", "1906.04312", "2203.05997"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_4752"} +{"question": "What papers are about multichannel extensions in E(n)-equivariant GNNs?", "answer": ["Equivariant Graph Mechanics Networks with Constraints"], "answer_arxiv_id": ["2203.06442"], "source_meta": {"published_time": "20220812"}, "qid": "AutoScholarQuery_train_4753"} +{"question": "What are some examples of studies that explore the use of the IMU sensor for human activity recognition?", "answer": ["Attention-Based Sensor Fusion for Human Activity Recognition Using IMU Signals"], "answer_arxiv_id": ["2112.11224"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_4754"} +{"question": "What papers propose solutions to tackle the scalability problem of GNNs by sparsifying the graph structure?", "answer": ["Spectral Sparsification of Graphs", "FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling", "DropEdge: Towards Deep Graph Convolutional Networks on Node Classification", "Graph Coarsening with Neural Networks", "Geometric Knowledge Distillation: Topology Compression for Graph Neural Networks"], "answer_arxiv_id": ["0808.4134v3", "1801.10247", "1907.10903", "2102.01350", "2210.13014"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_4755"} +{"question": "What works focus on designing and analyzing algorithms for MFGs under contraction conditions?", "answer": ["Q-Learning in Regularized Mean-field Games", "Approximately Solving Mean Field Games via Entropy-Regularized Deep Reinforcement Learning", "Oracle-free Reinforcement Learning in Mean-Field Games along a Single Sample Path", "Policy Mirror Ascent for Efficient and Independent Learning in Mean Field Games"], "answer_arxiv_id": ["2003.12151v3", "2102.01585v2", "2208.11639v3", "2212.14449"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_4756"} +{"question": "What papers utilized compactness, diversity, forgetfulness, or the gradient norm for data selection methods?", "answer": ["iCaRL: Incremental Classifier and Representation Learning", "End-to-End Incremental Learning", "Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Gradient based sample selection for online continual learning", "An Empirical Study of Example Forgetting during Deep Neural Network Learning", "Deep Learning on a Data Diet: Finding Important Examples Early in Training"], "answer_arxiv_id": ["1611.07725", "1807.09536", "1708.00489", "1903.08671", "1812.05159", "2107.07075"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_4757"} +{"question": "What studies contributed to the surge of interest in transformer-based LLMs?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["2302.13971", "2211.05100", "2205.01068"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4758"} +{"question": "Could you provide me some papers about human motion generation from audios?", "answer": ["Dancing to Music", "Rhythm is a Dancer: Music-Driven Motion Synthesis with Global Structure", "DanceFormer: Music Conditioned 3D Dance Generation with Parametric\n Motion Transformer", "Rhythmic Gesticulator: Rhythm-Aware Co-Speech Gesture Synthesis with\n Hierarchical Neural Embeddings", "GestureDiffuCLIP: Gesture Diffusion Model with CLIP Latents"], "answer_arxiv_id": ["1911.02001", "2111.12159", "2103.10206", "2210.01448", "2303.14613"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_4759"} +{"question": "In what paper the authors had a profound discussion of dynamic benchmarks that inspired this research?", "answer": ["What Will it Take to Fix Benchmarking in Natural Language Understanding?"], "answer_arxiv_id": ["2104.02145"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_4760"} +{"question": "Could you cite some research that discusses the role of the temperature parameter in controlling the penalty strength on negative samples in InfoNCE?", "answer": ["Understanding the Behaviour of Contrastive Loss", "Simpler, Faster, Stronger: Breaking The log-K Curse On Contrastive Learners With FlatNCE"], "answer_arxiv_id": ["2012.09740", "2107.01152v1"], "source_meta": {"published_time": "20220716"}, "qid": "AutoScholarQuery_train_4761"} +{"question": "What is the standard algorithm for federated learning?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized\n Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_4762"} +{"question": "Could you provide me some studies about multimodal VAEs?", "answer": ["Variational Mixture-of-Experts Autoencoders for Multi-Modal Deep Generative Models", "Generalized Multimodal ELBO", "Mitigating Modality Collapse in Multimodal VAEs via Impartial Optimization"], "answer_arxiv_id": ["1911.03393", "2105.02470", "2206.04496"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_4763"} +{"question": "Can you provide the paper that introduces the Bidirectional-Inference VAE (BIVA) which uses a combination of top-down and bottom-up inference?", "answer": ["BIVA: A Very Deep Hierarchy of Latent Variables for Generative Modeling"], "answer_arxiv_id": ["1902.02102"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_4764"} +{"question": "Which works used these multilingual models for zero-shot cross-lingual dependency parsing?", "answer": ["Cross-Lingual Alignment of Contextual Word Embeddings, with Applications to Zero-shot Dependency Parsing", "Substructure Distribution Projection for Zero-Shot Cross-Lingual Dependency Parsing"], "answer_arxiv_id": ["1902.09492", "2110.08538"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_4765"} +{"question": "Can you tell me about the studies that perform optimization on both the encoder and decoder in relation to an RD loss?", "answer": ["Overfitting for Fun and Profit: Instance-Adaptive Data Compression", "Instance-Adaptive Video Compression: Improving Neural Codecs by Training\n on the Test Set", "Dynamic Low-Rank Instance Adaptation for Universal Neural Image\n Compression"], "answer_arxiv_id": ["2101.08687", "2111.10302", "2308.07733"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4766"} +{"question": "Which works have explored various visual representation for NBV policies in active 3D reconstruction?", "answer": ["Vision-Only Robot Navigation in a Neural Radiance World", "Learning to Explore using Active Neural SLAM", "Multi-Robot Active Mapping via Neural Bipartite Graph Matching", "Asynchronous Collaborative Autoscanning with Mode Switching for\n Multi-Robot Scene Reconstruction"], "answer_arxiv_id": ["2110.00168", "2004.05155", "2203.16319", "2210.04413"], "source_meta": {"published_time": "20240225"}, "qid": "AutoScholarQuery_train_4767"} +{"question": "What research provided the concept of Classification and Regression Diffusion Models (CARD)?", "answer": ["CARD: Classification and Regression Diffusion Models"], "answer_arxiv_id": ["2206.07275"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_4768"} +{"question": "Any papers that extend these convergence guarantees to broad input distributions and finite training data?", "answer": ["Learning a Single Neuron with Gradient Methods", "Learning a Single Neuron with Bias Using Gradient Descent", "The Landscape of Empirical Risk for Non-convex Losses", "Learning a Single Neuron for Non-monotonic Activation Functions"], "answer_arxiv_id": ["2001.05205", "2106.01101", "1607.06534", "2202.08064v1"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_4769"} +{"question": "Which papers developed a Meta Learning-based approach named Editable Training for HyperNetwork-based editors?", "answer": ["Editable Neural Networks"], "answer_arxiv_id": ["2004.00345"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_4770"} +{"question": "Which work showed mathematical reasoning capabilities can be correlated with training data frequency in context of few-shot learning in language models?", "answer": ["Impact of Pretraining Term Frequencies on Few-Shot Reasoning"], "answer_arxiv_id": ["2202.07206"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_4771"} +{"question": "Could you provide me some works that applied the disentanglement to situations with multiple individuals?", "answer": ["Learning to Decompose and Disentangle Representations for Video Prediction"], "answer_arxiv_id": ["1806.04166"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_4772"} +{"question": "Which works discuss 3D gaussians or metaballs?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering", "Approximate Differentiable Rendering with Algebraic Surfaces", "Flexible Techniques for Differentiable Rendering with 3D Gaussians"], "answer_arxiv_id": ["2308.04079", "2207.10606", "2308.14737"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_4773"} +{"question": "Give examples of papers that study knowledge distillation schemes in Class-Incremental Learning.", "answer": ["iCaRL: Incremental Classifier and Representation Learning", "Large Scale Incremental Learning", "Distilling the Knowledge in a Neural Network", "Learning without Forgetting"], "answer_arxiv_id": ["1611.07725", "1905.13260", "1503.02531", "1606.09282"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_4774"} +{"question": "Which work incorporates the use of high-quality images generated by DDPM to enhance adversarial robustness and suggests the use of 'Complementary' as a metric?", "answer": ["Improving Robustness using Generated Data"], "answer_arxiv_id": ["2110.09468"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_4775"} +{"question": "Which works discuss representation learning within the Natural Language Processing (NLP) and computer vision context?", "answer": ["Neural Discrete Representation Learning", "Multi-column Deep Neural Networks for Image Classification", "Unsupervised Representation Learning by Predicting Image Rotations"], "answer_arxiv_id": ["1711.00937", "1202.2745", "1803.07728"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_4776"} +{"question": "Which works discuss adversarial attacks causing the significant performance drop from applying visually imperceptible perturbations to images?", "answer": ["Explaining and Harnessing Adversarial Examples", "Intriguing properties of neural networks", "Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection\n Methods"], "answer_arxiv_id": ["1412.6572", "1312.6199", "1705.07263"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_4777"} +{"question": "What works proposed using cross-attention to compute the similarity between images and texts?", "answer": ["Stacked Cross Attention for Image-Text Matching", "Similarity Reasoning and Filtration for Image-Text Matching"], "answer_arxiv_id": ["1803.08024", "2101.01368"], "source_meta": {"published_time": "20240617"}, "qid": "AutoScholarQuery_train_4778"} +{"question": "Which research cases documented the impact of simplifying a Transformer’s representations on the in-domain performance and systematic generalization?", "answer": ["Interpretability Illusions in the Generalization of Simplified Models"], "answer_arxiv_id": ["2312.03656"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_4779"} +{"question": "What work proposed the expansion of autoregressive models and their application towards scalability?", "answer": ["Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2206.10789"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_4780"} +{"question": "In which work is the EMPATHETICDIALOGUES dataset established for empathetic response generation?", "answer": ["Towards Empathetic Open-domain Conversation Models: a New Benchmark and\n Dataset"], "answer_arxiv_id": ["1811.00207"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_4781"} +{"question": "Which work showed that the sample complexity of adversarial robustness is larger than standard classification tasks in the Gaussian setting?", "answer": ["Adversarially Robust Generalization Requires More Data"], "answer_arxiv_id": ["1804.11285"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_4782"} +{"question": "What works shifted the monolingual pre-training procedure to multilingual scenarios in the field of cross-lingual representation pre-training?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Unsupervised Cross-lingual Representation Learning at Scale"], "answer_arxiv_id": ["1810.04805", "1911.02116"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_4783"} +{"question": "Which work indicates that planning in Dec-POMDPs can be NEXP-hard in finding the team-optimal solution?", "answer": ["The Complexity of Decentralized Control of Markov Decision Processes"], "answer_arxiv_id": ["1301.3836v1"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_4784"} +{"question": "Which work introduced Instant-NGP that encodes features into a multi-resolution hash table?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4785"} +{"question": "Which works discussed the issue of objective inconsistency caused by local data heterogeneity in Federated Learning?", "answer": ["Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning", "Towards Federated Learning on Time-Evolving Heterogeneous Data"], "answer_arxiv_id": ["2007.07481", "1910.06378", "2008.03606", "2112.13246v3"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_4786"} +{"question": "Which work introduced RuleTaker as a benchmark for multistep deductive reasoning?", "answer": ["Transformers as Soft Reasoners over Language", "ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language"], "answer_arxiv_id": ["2002.05867", "2012.13048"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_4787"} +{"question": "What works explored spatio-temporal scene graphs to represent the programs for VidQA?", "answer": ["AGQA: A Benchmark for Compositional Spatio-Temporal Reasoning"], "answer_arxiv_id": ["2103.16002"], "source_meta": {"published_time": "20240703"}, "qid": "AutoScholarQuery_train_4788"} +{"question": "Could you provide me a study that involves proximal policy optimization in their strategy?", "answer": ["Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1707.06347"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_4789"} +{"question": "Could you tell me the papers where classifier-guided diffusion and classifier-free guidance were introduced in conditional generation scenarios?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2105.05233", "2207.12598"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_4790"} +{"question": "What works have investigated various types of distribution shifts, including subpopulation shifts, in supervised learning?", "answer": ["Does Distributionally Robust Supervised Learning Give Robust Classifiers?", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1611.02041", "1911.08731"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_4791"} +{"question": "Which works employed proprietary LLMs chatGPT and Claude2 to assess the quality of instruction data?", "answer": ["AlpaGasus: Training A Better Alpaca with Fewer Data"], "answer_arxiv_id": ["2307.08701"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_4792"} +{"question": "Which works used VAEs for generative modeling of vector graphics?", "answer": ["Auto-Encoding Variational Bayes", "A Learned Representation for Scalable Vector Graphics"], "answer_arxiv_id": ["1312.6114", "1904.02632"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_4793"} +{"question": "What papers contributed to the development of masked language modeling (MLM) that has been used for pretraining large-scale language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "2005.14165", "1907.11692"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_4794"} +{"question": "What research paper presented the game-based approach for BAI with fixed confidence?", "answer": ["Non-Asymptotic Pure Exploration by Solving Games"], "answer_arxiv_id": ["1906.10431v1"], "source_meta": {"published_time": "20230905"}, "qid": "AutoScholarQuery_train_4795"} +{"question": "What works leverage external knowledge from knowledge bases or pretraining paradigms on large-scale datasets for multimodal commonsense benchmarks?", "answer": ["KVL-BERT: Knowledge Enhanced Visual-and-Linguistic BERT for Visual Commonsense Reasoning", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "MERLOT: Multimodal Neural Script Knowledge Models"], "answer_arxiv_id": ["2012.07000", "1908.02265", "2106.02636"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_4796"} +{"question": "Could you list studies for enhancing the training stability and generation fidelity in Text-3D Generation via advanced shape guidance?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation", "EfficientDreamer: High-Fidelity and Robust 3D Creation via\n Orthogonal-view Diffusion Prior", "SweetDreamer: Aligning Geometric Priors in 2D Diffusion for Consistent\n Text-to-3D"], "answer_arxiv_id": ["2308.16512", "2308.13223", "2310.02596"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_4797"} +{"question": "Which study introduced a general approach to deal with general function classes with a finite number of actions?", "answer": ["Beyond UCB: Optimal and Efficient Contextual Bandits with Regression Oracles"], "answer_arxiv_id": ["2002.04926"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_4798"} +{"question": "Which work also discussed various modeling choices and their underlying implicit or explicit normative principles, as well as limitations imposed by their framework?", "answer": ["Fairness in Ranking under Uncertainty"], "answer_arxiv_id": ["2107.06720v2"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_4799"} +{"question": "Could you provide me papers where simpler architectures are used to approximate the decision map in end-to-end learning?", "answer": ["Deep Learning for Portfolio Optimization", "Integrating prediction in mean-variance portfolio optimization", "End-to-End Risk Budgeting Portfolio Optimization with Neural Networks"], "answer_arxiv_id": ["2005.13665", "2102.09287", "2107.04636"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_4800"} +{"question": "What works are related to the analysis of the interpretability of embedding spaces in contrastive learning?", "answer": ["Network Dissection: Quantifying Interpretability of Deep Visual Representations", "Net2Vec: Quantifying and Explaining how Concepts are Encoded by Filters in Deep Neural Networks", "Quantifying Learnability and Describability of Visual Concepts Emerging in Representation Learning"], "answer_arxiv_id": ["1704.05796", "1801.03454", "2010.14551"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_4801"} +{"question": "Any works about combining LLMs and symbolic solvers for solving math problems?", "answer": ["Solving Math Word Problems by Combining Language Models With Symbolic Solvers"], "answer_arxiv_id": ["2304.09102"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_4802"} +{"question": "Could you tell me about the papers that propose QTRAN for value function transformation?", "answer": ["QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement learning"], "answer_arxiv_id": ["1905.05408"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_4803"} +{"question": "Which works focus on the study of parameter identification in auto-regressive models?", "answer": ["Statistical Learning Theory for Control: A Finite Sample Perspective"], "answer_arxiv_id": ["2209.05423v2"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_4804"} +{"question": "What research is working on optimizing policy improvement objectives using just supervised learning?", "answer": ["Trust Region Policy Optimization", "Dynamical Distance Learning for Semi-Supervised and Unsupervised Skill Discovery"], "answer_arxiv_id": ["1502.05477", "1907.08225"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_4805"} +{"question": "Which paper proposed Textual Inversion in the multimodal context?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_4806"} +{"question": "What work used CLIP as a spatial expert?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_4807"} +{"question": "Any works focused on weakly-supervised instance segmentation?", "answer": ["BoxInst: High-Performance Instance Segmentation with Box Annotations", "DiscoBox: Weakly Supervised Instance Segmentation and Semantic Correspondence from Box Supervision", "Box-supervised Instance Segmentation with Level Set Evolution", "SIM: Semantic-aware Instance Mask Generation for Box-Supervised Instance Segmentation", "BoxTeacher: Exploring High-Quality Pseudo Labels for Weakly Supervised Instance Segmentation", "Box2Mask: Box-supervised Instance Segmentation via Level-set Evolution", "Vision Transformers Are Good Mask Auto-Labelers", "BoxSnake: Polygonal Instance Segmentation with Box Supervision"], "answer_arxiv_id": ["2012.02310", "2105.06464", "2207.09055", "2303.08578", "2210.05174", "2212.01579", "2301.03992", "2303.11630"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_4808"} +{"question": "Which works empirically studied the influence of different settings on the positive and negative interference among different directions in MNMT?", "answer": ["Massively Multilingual Neural Machine Translation", "Causes and Cures for Interference in Multilingual Translation"], "answer_arxiv_id": ["1903.00089", "2212.07530"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_4809"} +{"question": "What works discuss the use of Adapter as a method for fine-tuning large models?", "answer": ["Parameter-Efficient Transfer Learning for NLP"], "answer_arxiv_id": ["1902.00751"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_4810"} +{"question": "What works have adapted Stable Diffusion to solve tasks in different domain?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Zero-1-to-3: Zero-shot One Image to 3D Object", "Objaverse-XL: A Universe of 10M+ 3D Objects", "Sin3DM: Learning a Diffusion Model from a Single 3D Textured Shape", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models", "SegDiff: Image Segmentation with Diffusion Probabilistic Models", "Label-Efficient Semantic Segmentation with Diffusion Models"], "answer_arxiv_id": ["2211.09800", "2208.01618", "2208.12242", "2303.11328", "2307.05663", "2305.15399", "2303.04803", "2112.00390", "2112.03126"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_4811"} +{"question": "Can you list some works where the authors independently extract and aggregate features via proposals or queries in the detection head?", "answer": ["TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with\n Transformers", "FUTR3D: A Unified Sensor Fusion Framework for 3D Detection", "Cross Modal Transformer: Towards Fast and Robust 3D Object Detection"], "answer_arxiv_id": ["2203.11496", "2203.10642", "2301.01283"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_4812"} +{"question": "What are some studies to improve the performance of Variational Autoencoders (VAEs) by constructing more expressive tractable variational posteriors?", "answer": ["Hierarchical Variational Models", "Variational Inference using Implicit Distributions", "Semi-Implicit Variational Inference", "Advances in Variational Inference", "Doubly Semi-Implicit Variational Inference", "Unbiased Implicit Variational Inference"], "answer_arxiv_id": ["1511.02386", "1702.08235", "1805.11183", "1711.05597", "1810.02789v2", "1808.02078v3"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_4813"} +{"question": "Are there any research works that model non-conjugate likelihoods with Gaussian approximations, similar to Kalman filtering and smoothing operations?", "answer": ["Fast Variational Learning in State-Space Gaussian Process Models", "Spatio-Temporal Variational Gaussian Processes"], "answer_arxiv_id": ["2007.04731", "2111.01732v1"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_4814"} +{"question": "What previous work introduced the concept of few-shot tuning for a specific class or concept of images?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2212.04488"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_4815"} +{"question": "Could you provide me with some research about the classifier-guidance approach for generating samples from a diffusion model?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2105.05233", "2011.13456"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_4816"} +{"question": "Could you provide studies that delve into the theoretical capabilities, limitations, and internal workings of transformers?", "answer": ["Are Transformers universal approximators of sequence-to-sequence functions?", "On the Turing Completeness of Modern Neural Network Architectures", "Infinite attention: NNGP and NTK for deep attention networks", "Self-Attention Networks Can Process Bounded Hierarchical Languages", "On the Computational Power of Transformers and its Implications in Sequence Modeling", "Unveiling Transformers with LEGO: a synthetic reasoning task", "Transformers Learn Shortcuts to Automata", "Theoretical Limitations of Self-Attention in Neural Sequence Models", "On the Ability and Limitations of Transformers to Recognize Formal Languages", "Approximating How Single Head Attention Learns", "Thinking Like Transformers", "Inductive Biases and Variable Creation in Self-Attention Mechanisms", "In-context Learning and Induction Heads"], "answer_arxiv_id": ["1912.10077", "1901.03429", "2006.10540", "2105.11115", "2006.09286", "2206.04301v3", "2210.10749", "1906.06755", "2009.11264", "2103.07601", "2106.06981", "2110.10090", "2209.11895v1"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_4817"} +{"question": "Which work shows that sparsity is useful in some cases regarding the sparse MAB problem?", "answer": ["Equipping Experts/Bandits with Long-term Memory"], "answer_arxiv_id": ["1905.12950"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_4818"} +{"question": "Which papers studied the discrimination risks associated with missing values?", "answer": ["Fairness without Imputation: A Decision Tree Approach for Fair Prediction with Missing Values", "FairPrep: Promoting Data to a First-Class Citizen in Studies on Fairness-Enhancing Interventions"], "answer_arxiv_id": ["2109.10431", "1911.12587"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4819"} +{"question": "Which publication conditions the model during training with past model predictions?", "answer": ["Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning"], "answer_arxiv_id": ["2208.04202"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_4820"} +{"question": "What works have used YouTube as a source to draw data from?", "answer": ["Title Generation for User Generated Videos", "HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million\n Narrated Video Clips"], "answer_arxiv_id": ["1608.07068", "1906.03327"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_4821"} +{"question": "What works mentioned the concept of the von Neumann winner in contextual dueling bandits?", "answer": ["Contextual Dueling Bandits"], "answer_arxiv_id": ["1502.06362v2"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_4822"} +{"question": "What works present support factorization for disentanglement from a causal perspective?", "answer": ["Desiderata for Representation Learning: A Causal Perspective"], "answer_arxiv_id": ["2109.03795"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_4823"} +{"question": "What studies introduce the use of concept activation vectors (CAVs)?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)"], "answer_arxiv_id": ["1711.11279"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_4824"} +{"question": "Which methods learn gene representations from both single-cell and spatial datasets?", "answer": ["node2vec: Scalable Feature Learning for Networks", "Predicting multicellular function through multi-layer tissue networks"], "answer_arxiv_id": ["1607.00653", "1707.04638"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_4825"} +{"question": "What research designed a differentiable parametric sampler that can be optimized for fast data generation?", "answer": ["Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality"], "answer_arxiv_id": ["2202.05830"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_4826"} +{"question": "What research papers handle the task of pose estimation in animal behavior analysis?", "answer": ["Self-Supervised Keypoint Discovery in Behavioral Videos"], "answer_arxiv_id": ["2112.05121"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_4827"} +{"question": "What references studied offline Markov games using linear function approximation?", "answer": ["Pessimistic Minimax Value Iteration: Provably Efficient Equilibrium Learning from Offline Datasets", "Nearly Minimax Optimal Offline Reinforcement Learning with Linear Function Approximation: Single-Agent MDP and Markov Game"], "answer_arxiv_id": ["2202.07511", "2205.15512"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_4828"} +{"question": "Could you direct me to the work which investigates defensive tactics for LLMs like preprocessing, paraphrasing input prompts, and adversarial training?", "answer": ["Baseline Defenses for Adversarial Attacks Against Aligned Language\n Models"], "answer_arxiv_id": ["2309.00614"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_4829"} +{"question": "What studies compared with the Walk Index Sparsification (WIS) method regarding edge sparsity levels and prediction accuracies?", "answer": ["Graph Sparsification by Effective Resistances", "A Unified Lottery Ticket Hypothesis for Graph Neural Networks"], "answer_arxiv_id": ["0803.0929v4", "2102.06790"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_4830"} +{"question": "What studies have used recurrent networks in developing autoregressive graph generative models?", "answer": ["Learning Deep Generative Models of Graphs", "GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models", "Scalable Deep Generative Modeling for Sparse Graphs"], "answer_arxiv_id": ["1803.03324", "1802.08773", "2006.15502"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_4831"} +{"question": "Which methods have been used to tackle value divergence through an empirical approach?", "answer": ["Addressing Function Approximation Error in Actor-Critic Methods", "Towards Characterizing Divergence in Deep Q-Learning"], "answer_arxiv_id": ["1802.09477", "1903.08894"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_4832"} +{"question": "What papers were the first ones to use model extraction for explaining deep learning models?", "answer": ["Interpreting Blackbox Models via Model Extraction", "Interpretability via Model Extraction"], "answer_arxiv_id": ["1705.08504", "1706.09773"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_4833"} +{"question": "Are there any works that enhanced the encoding ability of the dictionary encoder by focusing more on entities and use phonetically similar phrases as negative examples?", "answer": ["Contextual Speech Recognition with Difficult Negative Training Examples"], "answer_arxiv_id": ["1810.12170"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_4834"} +{"question": "What works proposed methodologies for trajectory planning using diffusion models?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis", "Is Conditional Generative Modeling all you need for Decision-Making?"], "answer_arxiv_id": ["2205.09991", "2211.15657"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_4835"} +{"question": "Which publications enhanced diffusion models for generative novel-view synthesis?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object", "Novel View Synthesis with Diffusion Models", "Generative Novel View Synthesis with 3D-Aware Diffusion Models"], "answer_arxiv_id": ["2303.11328", "2210.04628", "2304.02602"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_4836"} +{"question": "Which studies have effectively utilized diffusion models for discriminative tasks?", "answer": ["Diffusion Models for Zero-Shot Open-Vocabulary Segmentation", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models", "Prompting Diffusion Representations for Cross-Domain Semantic\n Segmentation"], "answer_arxiv_id": ["2306.09316", "2303.04803", "2307.02138"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_4837"} +{"question": "Which works covered the topic of utility in data evaluation, especially in synthetic data?", "answer": ["Machine Learning for Synthetic Data Generation: A Review", "General and specific utility measures for synthetic data", "MLReal: Bridging the gap between training on synthetic data and real data applications in machine learning", "A Study on Improving Realism of Synthetic Data for Machine Learning", "Synthcity: facilitating innovative use cases of synthetic data in different data modalities"], "answer_arxiv_id": ["2302.04062", "1604.06651", "2109.05294", "2304.12463", "2301.07573"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_4838"} +{"question": "Which studies report that GAN inversion techniques often struggle to accurately represent the data that falls outside the GAN distribution?", "answer": ["Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?", "High-Fidelity GAN Inversion for Image Attribute Editing"], "answer_arxiv_id": ["1904.03189", "2109.06590"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4839"} +{"question": "Are there any studies working on incorporating a convergence mechanism for efficient curriculum progresses in high-dimensional goal spaces?", "answer": ["Null-text Inversion for Editing Real Images using Guided Diffusion Models", "Delta Denoising Score"], "answer_arxiv_id": ["2211.09794", "2304.07090"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4840"} +{"question": "Could you provide me some works that use neural networks to synthesize lighting effects in portrait relighting?", "answer": ["Learning Physics-guided Face Relighting under Directional Light", "Single Image Portrait Relighting", "Neural Light Transport for Relighting and View Synthesis", "Neural Video Portrait Relighting in Real-time via Consistency Modeling", "LightPainter: Interactive Portrait Relighting with Freehand Scribble", "Learning to Relight Portrait Images via a Virtual Light Stage and\n Synthetic-to-Real Adaptation"], "answer_arxiv_id": ["1906.03355", "1905.00824", "2008.03806", "2104.00484", "2303.12950", "2209.10510"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_4841"} +{"question": "Which paper used an instance selective whitening loss to selectively remove only feature representations that cause domain shifts?", "answer": ["RobustNet: Improving Domain Generalization in Urban-Scene Segmentation\n via Instance Selective Whitening"], "answer_arxiv_id": ["2103.15597"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_4842"} +{"question": "What works have utilized approximate nearest neighbor (ANN) methods for dense retrieval?", "answer": ["Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs"], "answer_arxiv_id": ["1603.09320v4"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_4843"} +{"question": "Are there any studies that use outputs from other models for conditioning?", "answer": ["Cascaded Diffusion Models for High Fidelity Image Generation", "Image Super-Resolution via Iterative Refinement"], "answer_arxiv_id": ["2106.15282", "2104.07636"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_4844"} +{"question": "Could you mention research papers about datasets that are collected through human-generated data, manual work and heuristics?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "The Pile: An 800GB Dataset of Diverse Text for Language Modeling", "LAION-5B: An open large-scale dataset for training next generation image-text models", "DataComp: In search of the next generation of multimodal datasets"], "answer_arxiv_id": ["1910.10683", "2101.00027", "2210.08402", "2304.14108"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_4845"} +{"question": "Could you provide me with some research about using language models for program synthesis with extension to general-purpose programming languages?", "answer": ["Program Synthesis with Large Language Models", "Measuring Coding Challenge Competence With APPS", "Latent Execution for Neural Program Synthesis", "PyMT5: multi-mode translation of natural language and Python code with transformers", "CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generation"], "answer_arxiv_id": ["2108.07732", "2105.09938", "2107.00101", "2010.03150", "2109.00859"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_4846"} +{"question": "What studies focused on reducing serving costs of retrieval models?", "answer": ["COIL: Revisit Exact Lexical Match in Information Retrieval with Contextualized Inverted List", "CITADEL: Conditional Token Interaction via Dynamic Lexical Routing for Efficient and Effective Multi-Vector Retrieval", "PLAID: An Efficient Engine for Late Interaction Retrieval"], "answer_arxiv_id": ["2104.07186", "2211.10411", "2205.09707"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_4847"} +{"question": "Which research proposed a method of approximating a target ReLU network by pruning an overparametrized ReLU network?", "answer": ["Proving the Lottery Ticket Hypothesis: Pruning is All You Need"], "answer_arxiv_id": ["2002.00585"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_4848"} +{"question": "What works are about learning EBMs through score matching?", "answer": ["Sliced Score Matching: A Scalable Approach to Density and Score Estimation", "Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1905.07088", "1907.05600"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_4849"} +{"question": "What studies proposed loss functions for various ranking metrics?", "answer": ["Ranking via Sinkhorn Propagation", "Optimizing Rank-based Metrics with Blackbox Differentiation"], "answer_arxiv_id": ["1106.1925", "1912.03500"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_4850"} +{"question": "Which works used edge maps in image inpainting tasks?", "answer": ["StructureFlow: Image Inpainting via Structure-aware Appearance Flow"], "answer_arxiv_id": ["1908.03852"], "source_meta": {"published_time": "20200331"}, "qid": "AutoScholarQuery_train_4851"} +{"question": "Can you provide some examples where self-training and bootstrapping approaches are used in sequence generation?", "answer": ["Revisiting Self-Training for Neural Sequence Generation"], "answer_arxiv_id": ["1909.13788"], "source_meta": {"published_time": "20220824"}, "qid": "AutoScholarQuery_train_4852"} +{"question": "What are the works that achieved human-level performance on many NLP tasks using pre-trained language models such as GPT-3, OPT, PaLM?", "answer": ["Language Models are Few-Shot Learners", "OPT: Open Pre-trained Transformer Language Models", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2205.01068", "2204.02311"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_4853"} +{"question": "What flat-running problem-oriented studies introduce VAE variants of discrete-time RNNs?", "answer": ["A Recurrent Latent Variable Model for Sequential Data", "Learning Stochastic Recurrent Networks"], "answer_arxiv_id": ["1506.02216", "1411.7610"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_4854"} +{"question": "Can you provide the example of research that interprets system as the characteristics of a continuity equation?", "answer": ["Sinkformers: Transformers with Doubly Stochastic Attention"], "answer_arxiv_id": ["2110.11773v2"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_4855"} +{"question": "Can you name the works that discuss the use of domain randomization in reinforcement learning for dexterous manipulation?", "answer": ["Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Sim-to-Real Transfer of Robotic Control with Dynamics Randomization"], "answer_arxiv_id": ["1703.06907", "1710.06537"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_4856"} +{"question": "What are the relevant studies related to text-driven editing task using GANs?", "answer": ["CLIP2StyleGAN: Unsupervised Extraction of StyleGAN Edit Directions", "Paint by Word", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "Self-Distilled StyleGAN: Towards Generation from Internet Photos", "TediGAN: Text-Guided Diverse Face Image Generation and Manipulation", "Spatially-Adaptive Multilayer Selection for GAN Inversion and Editing", "Pivotal Tuning for Latent-based Editing of Real Images"], "answer_arxiv_id": ["2112.05219", "2103.10951", "2108.00946", "2202.12211", "2012.03308", "2206.08357", "2106.05744"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4857"} +{"question": "What are the works about structured pruning in Neural Networks?", "answer": ["Structured Pruning of Large Language Models", "LLM-Pruner: On the Structural Pruning of Large Language Models", "Structured Pruning Learns Compact and Accurate Models", "Sheared LLaMA: Accelerating Language Model Pre-training via Structured\n Pruning"], "answer_arxiv_id": ["1910.04732", "2305.11627", "2204.00408", "2310.06694"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_4858"} +{"question": "Which papers discuss the SZ3 method, which finds sparse representations in the space of locally spanning splines?", "answer": ["SZ3: A Modular Framework for Composing Prediction-Based Error-Bounded Lossy Compressors"], "answer_arxiv_id": ["2111.02925"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_4859"} +{"question": "Which works project visual information into semantic space for embedding-based Zero-Shot Learning?", "answer": ["Transductive Unbiased Embedding for Zero-Shot Learning"], "answer_arxiv_id": ["1803.11320"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_4860"} +{"question": "Which research papers focus on DRO under various uncertainty sets?", "answer": ["Variance-based regularization with convex objectives", "Learning Models with Uniform Performance via Distributionally Robust Optimization", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Certifying Some Distributional Robustness with Principled Adversarial Training", "Wasserstein Distributionally Robust Optimization and Variation Regularization", "Distributionally Robust Optimization and Generalization in Kernel Methods", "Kernel Distributionally Robust Optimization"], "answer_arxiv_id": ["1610.02581", "1810.08750", "1911.08731", "1710.10571v5", "1712.06050", "1905.10943", "2006.06981"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_4861"} +{"question": "What research is focused on entity-level conceptualization?", "answer": ["COPEN: Probing Conceptual Knowledge in Pre-trained Language Models"], "answer_arxiv_id": ["2211.04079"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_4862"} +{"question": "Can you provide papers about deep generative model methods in sequence design?", "answer": ["Deep Extrapolation for Attribute-Enhanced Generation", "Conditioning by adaptive sampling for robust design"], "answer_arxiv_id": ["2107.02968", "1901.10060"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_4863"} +{"question": "Can you provide examples of studies that propose ways to improve the performance of fast adversarial training?", "answer": ["Understanding and Improving Fast Adversarial Training", "Revisiting and Advancing Fast Adversarial Training Through the Lens of Bi-Level Optimization"], "answer_arxiv_id": ["2007.02617", "2112.12376"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_4864"} +{"question": "What works utilized Denoising diffusion probabilistic models (DDPMs) for text-to-image synthesis?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["1503.03585", "1907.05600", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_4865"} +{"question": "Could you tell me the work that proposed solving OPE using linear programming when Bellman completeness is not present?", "answer": ["Offline Reinforcement Learning with Realizability and Single-policy Concentrability"], "answer_arxiv_id": ["2202.04634"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_4866"} +{"question": "What are the papers about accelerating self-attention computation by refining attention mechanisms through improved IO management?", "answer": ["FlashAttention: Fast and Memory-Efficient Exact Attention with\n IO-Awareness", "FlashAttention-2: Faster Attention with Better Parallelism and Work\n Partitioning"], "answer_arxiv_id": ["2205.14135", "2307.08691"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_4867"} +{"question": "What research works discussed prompting the language model in a few-shot setup for specific tasks?", "answer": ["PAL: Program-aided Language Models", "Internet-augmented language models through few-shot prompting for open-domain question answering", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2211.10435", "2203.05115", "2210.03629"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_4868"} +{"question": "What studies showcase the ability of diffusion models in creating lifelike images from textual inputs?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Cascaded Diffusion Models for High Fidelity Image Generation", "Palette: Image-to-Image Diffusion Models", "Image Super-Resolution via Iterative Refinement", "Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["2105.05233", "2106.15282", "2111.05826", "2104.07636", "1907.05600"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4869"} +{"question": "What works refer to the extension of diffusion models techniques to the infinite-dimensional space?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2011.13456", "2006.11239"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_4870"} +{"question": "Which research papers first adopt fully convolutional networks for pixel-wise classification during image semantic segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_4871"} +{"question": "Could you name the paper that proposed a graph model for multi-sound source localization?", "answer": ["Mix and Localize: Localizing Sound Sources in Mixtures"], "answer_arxiv_id": ["2211.15058"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_4872"} +{"question": "Which study proposed the Forward Compatible Training (FCT) method?", "answer": ["Forward Compatible Training for Large-Scale Embedding Retrieval Systems"], "answer_arxiv_id": ["2112.02805"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_4873"} +{"question": "Could you provide me some researches about the broader context of distributed learning, where agents work together to achieve a shared goal?", "answer": ["RLlib: Abstractions for Distributed Reinforcement Learning", "Acme: A Research Framework for Distributed Reinforcement Learning", "DeepMTL Pro: Deep Learning Based Multiple Transmitter Localization and Power Estimation"], "answer_arxiv_id": ["1712.09381", "2006.00979", "2112.13181"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_4874"} +{"question": "Which work first determines the word replacement order according to the prediction change after deleting each word, and then replaces words back according to the word importance until adversarial examples are generated?", "answer": ["Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment"], "answer_arxiv_id": ["1907.11932"], "source_meta": {"published_time": "20240202"}, "qid": "AutoScholarQuery_train_4875"} +{"question": "What paper uses PatchNCE for brain MR T1w-T2w registration?", "answer": ["ContraReg: Contrastive Learning of Multi-modality Unsupervised\n Deformable Image Registration"], "answer_arxiv_id": ["2206.13434"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_4876"} +{"question": "What works analyse the hierarchical structure of tasks that guide SGD to learn high degree monomials?", "answer": ["The staircase property: How hierarchical structure can guide deep learning", "The merged-staircase property: a necessary and nearly sufficient condition for SGD learning of sparse functions on two-layer neural networks"], "answer_arxiv_id": ["2108.10573", "2202.08658"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_4877"} +{"question": "What works propose specific methods to bind or connect different modalities?", "answer": ["ImageBind: One Embedding Space To Bind Them All", "Contrastive Multiview Coding"], "answer_arxiv_id": ["2305.05665v2", "1906.05849"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_4878"} +{"question": "Which paper suggested that in-distribution accuracy improvements often directly yield out-of-distribution accuracy improvements?", "answer": ["Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization"], "answer_arxiv_id": ["2107.04649"], "source_meta": {"published_time": "20230111"}, "qid": "AutoScholarQuery_train_4879"} +{"question": "Which papers established the relationship between SBP and stochastic optimal control?", "answer": ["Stochastic bridges of linear systems", "Wasserstein Proximal Algorithms for the Schrödinger Bridge Problem: Density Control with Nonlinear Drift"], "answer_arxiv_id": ["1407.3421", "1912.01244"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_4880"} +{"question": "Could you provide me some researches that focused on learning specific components like feature extraction, matching, and pose or point cloud refinement?", "answer": ["LIFT: Learned Invariant Feature Transform", "Working hard to know your neighbor’s margins: Local descriptor learning loss", "SuperPoint: Self-Supervised Interest Point Detection and Description", "D2-Net: A Trainable CNN for Joint Detection and Description of Local Features", "SOSNet: Second Order Similarity Regularization for Local Descriptor Learning", "R2D2: Repeatable and Reliable Detector and Descriptor", "Beyond Cartesian Representations for Local Descriptors", "DISK: Learning local features with policy gradient", "Learning Feature Descriptors using Camera Pose Supervision", "ASLFeat: Learning Local Features of Accurate Shape and Localization", "Learning to Find Good Correspondences", "Learning Two-View Correspondences and Geometry Using Order-Aware Network", "SuperGlue: Learning Feature Matching with Graph Neural Networks", "LoFTR: Detector-Free Local Feature Matching with Transformers", "COTR: Correspondence Transformer for Matching Across Images", "Progressive Correspondence Pruning by Consensus Learning", "MatchFormer: Interleaving Attention in Transformers for Feature Matching", "LightGlue: Local Feature Matching at Light Speed", "GN-Net: The Gauss-Newton Loss for Multi-Weather Relocalization", "Back to the Feature: Learning Robust Camera Localization from Pixels to Pose", "Pixel-Perfect Structure-from-Motion with Featuremetric Refinement"], "answer_arxiv_id": ["1603.09114", "1705.10872", "1712.07629", "1905.03561", "1904.05019", "1906.06195", "1908.05547", "2006.13566", "2004.13324", "2003.10071", "1711.05971", "1908.04964", "1911.11763", "2104.00680", "2103.14167", "2101.00591", "2203.09645", "2306.13643", "1904.11932", "2103.09213", "2108.08291"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_4881"} +{"question": "Could you tell me about the studies that utilized self-supervised local priors to deal with very sparse inputs in neural networks?", "answer": ["Reconstructing Surfaces for Sparse Point Clouds with On-Surface Priors"], "answer_arxiv_id": ["2204.10603"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_4882"} +{"question": "Which works explore the issues with uninformative priors for Bayesian Neural Networks?", "answer": ["Priors in Bayesian Deep Learning: A Review", "How Good is the Bayes Posterior in Deep Neural Networks Really?", "Bayesian Neural Network Priors Revisited"], "answer_arxiv_id": ["2105.06868", "2002.02405", "2102.06571"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_4883"} +{"question": "Could you provide me some works about scalable approaches for GP inference?", "answer": ["Kernel Interpolation for Scalable Structured Gaussian Processes (KISS-GP)", "A multi-resolution approximation for massive spatial datasets", "Hierarchically Compositional Kernels for Scalable Nonparametric Learning", "Linear-Cost Covariance Functions for Gaussian Random Fields"], "answer_arxiv_id": ["1503.01057", "1507.04789", "1608.00860v2", "1711.05895"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_4884"} +{"question": "What are the references for utterance-level models that differentiate self-other awareness within individual utterances?", "answer": ["Don't Lose Yourself! Empathetic Response Generation via Explicit\n Self-Other Awareness"], "answer_arxiv_id": ["2210.03884"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_4885"} +{"question": "Could you provide me with studies that have conducted explorations on the work mechanism for ICL?", "answer": ["Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "What Makes Good In-Context Examples for GPT-$3$?", "What learning algorithm is in-context learning? Investigations with\n linear models"], "answer_arxiv_id": ["2202.12837", "2101.06804", "2211.15661"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_4886"} +{"question": "Could you provide me some works about representation editing for controllable text generation?", "answer": ["Plug and Play Language Models: A Simple Approach to Controlled Text\n Generation", "Second Thoughts are Best: Learning to Re-Align With Human Values from\n Text Edits"], "answer_arxiv_id": ["1912.02164", "2301.00355"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_4887"} +{"question": "Which papers propose the first models and extensions on graph learning from stationary signals?", "answer": ["​Network​ Topology​ Inference​ from​ Spectral​ Templates", "Joint inference of multiple graphs with hidden variables from stationary graph signals", "Online Topology Inference from Streaming Stationary Graph Signals with Partial Connectivity Information", "Learning Graphs from Smooth and Graph-Stationary Signals with Hidden Variables"], "answer_arxiv_id": ["1608.03008", "2110.03666", "2007.03653v1", "2111.05588"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_4888"} +{"question": "Are there any works that have attempted to aggregate 2D features in 3D space for 3D object detection in indoor scenes?", "answer": ["ImVoxelNet: Image to Voxels Projection for Monocular and Multi-View\n General-Purpose 3D Object Detection", "DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries", "PETR: Position Embedding Transformation for Multi-View 3D Object\n Detection"], "answer_arxiv_id": ["2106.01178", "2110.06922", "2203.05625"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_4889"} +{"question": "Which studies have looked at the data sparsity issues of supervised methods in user modeling?", "answer": ["Towards Universal Sequence Representation Learning for Recommender Systems", "AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks"], "answer_arxiv_id": ["2206.05941", "1810.11921"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_4890"} +{"question": "Can you tell me what works have utilized Gaussian Processes, Neural Processes or Bayesian Neural Networks in the context of Online ED?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design", "Conditional Neural Processes", "Neural Processes", "Convolutional Conditional Neural Processes", "Attentive Neural Processes", "Sequential Neural Processes", "Transformer Neural Processes: Uncertainty-Aware Meta Learning Via Sequence Modeling", "Bayesian Neural Networks: An Introduction and Survey"], "answer_arxiv_id": ["0912.3995v4", "1807.01613", "1807.01622", "1910.13556", "1901.05761", "1906.10264", "2207.04179", "2006.12024"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_4891"} +{"question": "Which papers have applied neural network quantum states for continuous many-body wave function in quantum chemistry?", "answer": ["Deep neural network solution of the electronic Schrödinger equation"], "answer_arxiv_id": ["1909.08423"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_4892"} +{"question": "What paper proposes similar algorithms to the main research in the context of Gaussian mixture estimation?", "answer": ["Learning Gaussian Mixtures Using the Wasserstein-Fisher-Rao Gradient Flow"], "answer_arxiv_id": ["2301.01766"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_4893"} +{"question": "What works have reported on phase transitions in deep learning?", "answer": ["What learning algorithm is in-context learning? Investigations with linear models", "Mechanistic Mode Connectivity"], "answer_arxiv_id": ["2211.15661", "2211.08422"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_4894"} +{"question": "Which studies represent the deep learning-based methods in monocular depth estimation?", "answer": ["Depth Map Prediction from a Single Image using a Multi-Scale Deep\n Network"], "answer_arxiv_id": ["1406.2283"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_4895"} +{"question": "Which studies utilize larger models with higher capacity to support complex multilingual ASR tasks?", "answer": ["SCALING END-TO-END MODELS FOR LARGE-SCALE MULTILINGUAL ASR", "Massively Multilingual ASR: 50 Languages, 1 Model, 1 Billion Parameters"], "answer_arxiv_id": ["2104.14830", "2007.03001"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_4896"} +{"question": "What studies focus on the complexity and rate guarantees in nonconvex and nonsmooth optimization problems?", "answer": ["Mini-batch Stochastic Approximation Methods for Nonconvex Stochastic Composite Optimization", "A globally convergent algorithm for nonconvex optimization based on block coordinate update", "Stochastic model-based minimization of weakly convex functions", "Proximally Guided Stochastic Subgradient Method for Nonsmooth, Nonconvex Problems", "Asynchronous Variance-reduced Block Schemes for Composite Nonconvex Stochastic Optimization: Block-specific Steplengths and Adapted Batch-sizes"], "answer_arxiv_id": ["1308.6594", "1410.1386v2", "1803.06523", "1707.03505", "1808.02543"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_4897"} +{"question": "Are there any papers about the application of score-based generative models in MRI reconstruction?", "answer": ["Score-based diffusion models for accelerated MRI", "Robust Compressed Sensing MRI with Deep Generative Priors", "Solving Inverse Problems in Medical Imaging with Score-Based Generative Models"], "answer_arxiv_id": ["2110.05243", "2108.01368", "2111.08005"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_4898"} +{"question": "What paper presents a partial graph neural network (PaGNN) that uses a partial message-propagation scheme for handling missing node features?", "answer": ["Incomplete Graph Representation and Learning via Partial Graph Neural Networks"], "answer_arxiv_id": ["2003.10130"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_4899"} +{"question": "What are some works that model shapes extrinsically using sparse representations such as octrees?", "answer": ["OctNet: Learning Deep 3D Representations at High Resolutions", "Octree Generating Networks: Efficient Convolutional Architectures for High-resolution 3D Outputs", "O-CNN: Octree-based Convolutional Neural Networks for 3D Shape Analysis"], "answer_arxiv_id": ["1611.05009", "1703.09438", "1712.01537"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_4900"} +{"question": "Which works show computational and oracle-call lower bounds when the contexts and reward functions are adversarial?", "answer": ["The Computational Power of Optimization in Online Learning"], "answer_arxiv_id": ["1504.02089"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_4901"} +{"question": "Which studies proposed supervised data-driven methods for Compressed Sensing (CS)?", "answer": ["Deep Convolutional Neural Network for Inverse Problems in Imaging", "ISTA-Net: Interpretable Optimization-Inspired Deep Network for Image Compressive Sensing", "MoDL: Model Based Deep Learning Architecture for Inverse Problems", "Deep Compressed Sensing", "DR2-Net: Deep Residual Reconstruction Network for Image Compressive Sensing", "Neumann Networks for Linear Inverse Problems in Imaging"], "answer_arxiv_id": ["1611.03679", "1706.07929", "1712.02862", "1905.06723", "1702.05743", "1901.03707"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_4902"} +{"question": "Any works about enhancing point cloud data representation?", "answer": ["Multi-View Transformer for 3D Visual Grounding"], "answer_arxiv_id": ["2204.02174"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_4903"} +{"question": "Could you provide me some studies that have made progress using optimization techniques apart from G-DRO?", "answer": ["Contrastive Adapters for Foundation Model Group Robustness", "Correct-n-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlations", "Focus on the Common Good: Group Distributional Robustness Follows", "Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations"], "answer_arxiv_id": ["2207.07180", "2203.01517", "2110.02619", "2204.02937"], "source_meta": {"published_time": "20221202"}, "qid": "AutoScholarQuery_train_4904"} +{"question": "What studies have carried out masked image modeling by predicting discrete tokens or contextualized representations for masked image patches?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language"], "answer_arxiv_id": ["2106.08254", "2202.03555"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_4905"} +{"question": "Could you list some works that extend chain-of-thought to tree- or graph-based structures for handling more complex problems?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Tree-of-Mixed-Thought: Combining Fast and Slow Thinking for Multi-hop\n Visual Reasoning", "Graph of Thoughts: Solving Elaborate Problems with Large Language Models", "Boosting Logical Reasoning in Large Language Models through a New\n Framework: The Graph of Thought"], "answer_arxiv_id": ["2305.10601", "2308.09658", "2308.09687", "2308.08614"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_4906"} +{"question": "What paper presents a concurrent work with a similar perspective on multi-task learning in diffusion models?", "answer": ["Efficient Diffusion Training via Min-SNR Weighting Strategy"], "answer_arxiv_id": ["2303.09556"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_4907"} +{"question": "Which work explores improving generalization capabilities by alternating diverse sample generation and discriminative style-invariant representation learning?", "answer": ["Learning to Diversify for Single Domain Generalization"], "answer_arxiv_id": ["2108.11726"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_4908"} +{"question": "Which papers provided studies about the implicit bias on the sharpness of gradient descent in some general loss function?", "answer": ["Understanding Gradient Descent on Edge of Stability in Deep Learning", "Understanding the Generalization Benefit of Normalization Layers: Sharpness Reduction"], "answer_arxiv_id": ["2205.09745", "2206.07085"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_4909"} +{"question": "Which study used a different metric for evaluating relative position capabilities of T2I models?", "answer": ["Benchmarking Spatial Relationships in Text-to-Image Generation"], "answer_arxiv_id": ["2212.10015"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_4910"} +{"question": "Can you mention some proposed solutions to network heterophily in graph models?", "answer": ["Geom-GCN: Geometric Graph Convolutional Networks", "Two Sides of the Same Coin: Heterophily and Oversmoothing in Graph Convolutional Neural Networks"], "answer_arxiv_id": ["2002.05287", "2102.06462"], "source_meta": {"published_time": "20211104"}, "qid": "AutoScholarQuery_train_4911"} +{"question": "What works propose more complicated augmentations to boost generalization?", "answer": ["Random Erasing Data Augmentation", "CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features"], "answer_arxiv_id": ["1708.04896", "1905.04899"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_4912"} +{"question": "Which papers have proposed methods for tracking source coverage vectors to adjust the decoding of neural machine translation models?", "answer": ["Modeling Coverage for Neural Machine Translation", "Coverage Embedding Models for Neural Machine Translation", "Neural Machine Translation with Reconstruction", "Modeling Past and Future for Neural Machine Translation"], "answer_arxiv_id": ["1601.04811", "1605.03148", "1611.01874", "1711.09502"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_4913"} +{"question": "Could you tell me about some research papers that provide methods for generating human motion given a single static object?", "answer": ["GOAL: Generating 4D Whole-Body Motion for Hand-Object Grasping", "FLEX: Full-Body Grasping Without Full-Body Grasps", "COUCH: Towards Controllable Human-Chair Interactions", "SAGA: Stochastic Whole-Body Grasping with Contact", "Locomotion-Action-Manipulation: Synthesizing Human-Scene Interactions in\n Complex 3D Environments", "ROAM: Robust and Object-Aware Motion Generation Using Neural Pose\n Descriptors", "NIFTY: Neural Object Interaction Fields for Guided Human Motion\n Synthesis"], "answer_arxiv_id": ["2112.11454", "2211.11903", "2205.00541", "2112.10103", "2301.02667", "2308.12969", "2307.07511"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_4914"} +{"question": "What research developed the trajectory VAE, a variational generative model for learning representations of physical trajectories in space?", "answer": ["Self-Consistent Trajectory Autoencoder: Hierarchical Reinforcement Learning with Trajectory Embeddings"], "answer_arxiv_id": ["1806.02813"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_4915"} +{"question": "Could you provide me studies about having LDM as the foundational model for many controllable or customized image synthesis works?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation"], "answer_arxiv_id": ["2302.05543", "2305.16322", "2208.12242", "2302.13848"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_4916"} +{"question": "Which studies discuss the use of guidance and tool use to improve the performance of large language models in step-by-step problem solving?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "ReAct: Synergizing Reasoning and Acting in Language Models", "Solving math word problems with process- and outcome-based feedback", "Successive Prompting for Decomposing Complex Questions"], "answer_arxiv_id": ["2205.10625", "2210.03629", "2211.14275", "2212.04092"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_4917"} +{"question": "Who proposed a spatial memory architecture and demonstrated that a spatial representation emerges when trained on a localization task?", "answer": ["Multigrid Neural Memory"], "answer_arxiv_id": ["1906.05948"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_4918"} +{"question": "Could you provide research that explored the application of butterfly matrices?", "answer": ["Learning Fast Algorithms for Linear Transforms Using Butterfly Factorizations", "Pixelated Butterfly: Simple and Efficient Sparse Training for Neural Network Models", "Monarch: Expressive Structured Matrices for Efficient and Accurate Training"], "answer_arxiv_id": ["1903.05895", "2112.00029", "2204.00595"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_4919"} +{"question": "Could you list the studies demonstrating that pretrained vision-language models encode valuable information for robotic goal selection and task specification?", "answer": ["Can Foundation Models Perform Zero-Shot Task Specification For Robot Manipulation?", "Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language", "R3M: A Universal Visual Representation for Robot Manipulation", "Deep Residual Learning for Image Recognition", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2204.11134", "2204.00598", "2203.12601", "1512.03385", "2110.07058"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_4920"} +{"question": "What research studied the adaptation of RL agents to new and changing environments using Meta Learning and similar approaches?", "answer": ["Learning to reinforcement learn", "Learning to Adapt in Dynamic, Real-World Environments through Meta-Reinforcement Learning", "Neural Episodic Control", "A Regret Minimization Approach to Iterative Learning Control"], "answer_arxiv_id": ["1611.05763", "1803.11347", "1703.01988", "2102.13478"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_4921"} +{"question": "Could you provide me some references on the use of video as an additional information source in the grouping of low-level features into object entities?", "answer": ["Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects", "SCALOR: Generative World Models with Scalable Object Representations", "Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos"], "answer_arxiv_id": ["1806.01794", "1910.02384", "2205.14065"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_4922"} +{"question": "Please list some works that applied heavy data augmentation to create positive pairs in contrastive learning.", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "1911.05722"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_4923"} +{"question": "Which study estimated the cost of training an 11B-parameter model?", "answer": ["The Cost of Training NLP Models: A Concise Overview"], "answer_arxiv_id": ["2004.08900"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_4924"} +{"question": "Who used second order influence functions and found them to be more predictive than first order influence functions?", "answer": ["On Second-Order Group Influence Functions for Black-Box Predictions"], "answer_arxiv_id": ["1911.00418"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_4925"} +{"question": "What works proposed QTRAN and QPLEX to further achieve full representativeness of IGM?", "answer": ["QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement learning", "QPLEX: Duplex Dueling Multi-Agent Q-Learning"], "answer_arxiv_id": ["1905.05408", "2008.01062"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_4926"} +{"question": "What studies used memory banks of negative samples from recent batches to increase the effective contrastive batch size in vision tasks?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1911.05722"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_4927"} +{"question": "What research made assumptions on the accuracy of the estimation Q(π) requiring an L∞ supremum norm bound?", "answer": ["On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2201.07443"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_4928"} +{"question": "Could you provide me some works on Continual Learning?", "answer": ["Continual Lifelong Learning with Neural Networks: A Review", "Learning without Forgetting", "Continual Learning Through Synaptic Intelligence"], "answer_arxiv_id": ["1802.07569", "1606.09282", "1703.04200"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_4929"} +{"question": "Which study uses a Determinantal Point Process-based retriever to retrieve a set of demonstrations?", "answer": ["Compositional Exemplars for In-context Learning"], "answer_arxiv_id": ["2302.05698"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_4930"} +{"question": "Any studies about the performance of Reinforcement Learning agents with neural networks in large-scale, partially observable games?", "answer": ["Suphx: Mastering Mahjong with Deep Reinforcement Learning"], "answer_arxiv_id": ["2003.13590v2"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_4931"} +{"question": "Any studies that use TDA to define novel losses within ML algorithms?", "answer": ["A Topological Regularizer for Classifiers via Persistent Homology", "A Fast and Robust Method for Global Topological Functional Optimization"], "answer_arxiv_id": ["1806.10714v3", "2009.08496v3"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_4932"} +{"question": "What research has developed closed-loop methods to provide LLMs with dynamic information for planning?", "answer": ["Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents", "Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents", "Text2Motion: From Natural Language Instructions to Feasible Plans"], "answer_arxiv_id": ["2201.07207", "2302.01560", "2303.12153"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_4933"} +{"question": "Which works showed strong generalization for new tasks in multi-task learning?", "answer": ["Minimax Multi-Task Learning and a Generalized Loss-Compositional Paradigm for MTL"], "answer_arxiv_id": ["1209.2784"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_4934"} +{"question": "Which papers discussed about constraining the learnt policy to be similar to the behaviour policy in the model-free offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Offline Reinforcement Learning with Implicit Q-Learning", "Behavior Regularized Offline Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "2110.06169", "1911.11361", "2106.06860"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_4935"} +{"question": "What research paper discusses PromptCache that eliminates the redundant computation of reusable prefix KVs?", "answer": ["Prompt Cache: Modular Attention Reuse for Low-Latency Inference"], "answer_arxiv_id": ["2311.04934"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_4936"} +{"question": "Can you provide examples of studies that adopted specific data augmentation techniques to improve task results?", "answer": ["mixup: Beyond Empirical Risk Minimization", "Differentiable Augmentation for Data-Efficient GAN Training", "Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentation"], "answer_arxiv_id": ["1710.09412", "2006.10738", "2012.07177"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_4937"} +{"question": "Which works used recurrent neural networks or transformer blocks in video-based ReID models?", "answer": ["Person Re-Identification via Recurrent Feature Aggregation"], "answer_arxiv_id": ["1701.06351"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4938"} +{"question": "Which study proposes to normalize the per-user gradient in the federated learning setting?", "answer": ["On the Convergence of Differentially Private Federated Learning on Non-Lipschitz Objectives, and with Normalized Client Updates"], "answer_arxiv_id": ["2106.07094"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_4939"} +{"question": "Could you provide me studies where object-centric models are trained to generate motions for a single character?", "answer": ["Stochastic Scene-Aware Motion Prediction"], "answer_arxiv_id": ["2108.08284"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_4940"} +{"question": "What papers exist on the developments of the neural versions of NHR?", "answer": ["Time-to-Event Prediction with Neural Networks and Cox Regression", "Survival Regression with Proper Scoring Rules and Monotonic Neural Networks"], "answer_arxiv_id": ["1907.00825", "2103.14755v2"], "source_meta": {"published_time": "20230318"}, "qid": "AutoScholarQuery_train_4941"} +{"question": "What works focus on label propagation in graph neural networks?", "answer": ["Unifying Graph Convolutional Neural Networks and Label Propagation"], "answer_arxiv_id": ["2002.06755"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_4942"} +{"question": "Which papers address the Vector-Quantized Image Modeling (VQIM) problem and aim to learn a discrete codebook from scratch?", "answer": ["Neural Discrete Representation Learning", "Taming Transformers for High-Resolution Image Synthesis", "Regularized Vector Quantization for Tokenized Image Synthesis", "Online Clustered Codebook"], "answer_arxiv_id": ["1711.00937", "2012.09841", "2303.06424", "2307.15139"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4943"} +{"question": "What papers propose to predict blending weights based on image features and ray distances?", "answer": ["Stereo Magnification: Learning View Synthesis using Multiplane Images", "Local Light Field Fusion: Practical View Synthesis with Prescriptive\n Sampling Guidelines"], "answer_arxiv_id": ["1805.09817", "1905.00889"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_4944"} +{"question": "Which works are about single-agent kernel bandit problem?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design", "Finite-Time Analysis of Kernelised Contextual Bandits", "Lower Bounds on Regret for Noisy Gaussian Process Bandit Optimization", "Gaussian Process Bandit Optimization with Few Batches"], "answer_arxiv_id": ["0912.3995v4", "1309.6869v1", "1706.00090", "2110.07788v4"], "source_meta": {"published_time": "20211029"}, "qid": "AutoScholarQuery_train_4945"} +{"question": "Could you provide some examples of works that focus on federated visual prompts?", "answer": ["Learning Federated Visual Prompt in Null Space for MRI Reconstruction"], "answer_arxiv_id": ["2303.16181"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_4946"} +{"question": "Which papers reported the limitations of RLHF?", "answer": ["Open Problems and Fundamental Limitations of Reinforcement Learning from\n Human Feedback"], "answer_arxiv_id": ["2307.15217"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_4947"} +{"question": "Could you provide me an example of a study that proposes DIAL to allow gradients to flow cross agents?", "answer": ["Learning to Communicate with Deep Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1605.06676"], "source_meta": {"published_time": "20230319"}, "qid": "AutoScholarQuery_train_4948"} +{"question": "Which works proposed an efficient gradient method for fair PCA?", "answer": ["Efficient Fair Principal Component Analysis"], "answer_arxiv_id": ["1911.04931"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_4949"} +{"question": "What papers proposed the technique of motion mimicking in physics-based animation?", "answer": ["DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills", "SFV: Reinforcement Learning of Physical Skills from Videos"], "answer_arxiv_id": ["1804.02717", "1810.03599"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_4950"} +{"question": "What research efforts are leveraging diffusion models for unified frameworks in image compositing?", "answer": ["Denoising Diffusion Probabilistic Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2006.11239", "1503.03585", "1907.05600", "2112.10752"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_4951"} +{"question": "What are the works on federated quantile computation?", "answer": ["Differentially Private Learning with Adaptive Clipping"], "answer_arxiv_id": ["1905.03871"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_4952"} +{"question": "Which works attempt to optimize prompts in continuous embedding space?", "answer": ["Gradient-based Adversarial Attacks against Text Transformers", "Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with\n Adversarial Examples", "Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt\n Tuning and Discovery"], "answer_arxiv_id": ["2104.13733", "1803.01128", "2302.03668"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_4953"} +{"question": "Can you provide me the works about generative model of human-scene interaction using a conditional variational autoencoder?", "answer": ["Generating 3D People in Scenes without People"], "answer_arxiv_id": ["1912.02923"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_4954"} +{"question": "What study first popularized adversarial examples for neural networks?", "answer": ["Intriguing properties of neural networks"], "answer_arxiv_id": ["1312.6199"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_4955"} +{"question": "Which work proposed Lipschitz-constrained Skill Discovery (LSD) to resolve the limitation of MI-based skill discovery?", "answer": ["Lipschitz-constrained Unsupervised Skill Discovery"], "answer_arxiv_id": ["2202.00914"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_4956"} +{"question": "What is the first work that tried to leverage label propagation in order to exploit relationships among prototypes and query points in FS-PCS?", "answer": ["Few-shot 3D Point Cloud Semantic Segmentation"], "answer_arxiv_id": ["2006.12052"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_4957"} +{"question": "What papers explored 3D reconstruction estimates without ground truth 3D supervision using differentiable renderers?", "answer": ["Neural 3D Mesh Renderer", "Perspective Transformer Nets: Learning Single-View 3D Object Reconstruction without 3D Supervision"], "answer_arxiv_id": ["1711.07566", "1612.00814"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_4958"} +{"question": "Any works about enabling multi-domain transfer by studying learning via selective tuning?", "answer": ["Task Adaptive Parameter Sharing for Multi-Task Learning", "SpotTune: Transfer Learning through Adaptive Fine-tuning"], "answer_arxiv_id": ["2203.16708", "1811.08737"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_4959"} +{"question": "What publications involve SINE, which supports editing a local region of static NeRF from a single view?", "answer": ["SINE: Semantic-driven Image-based NeRF Editing with Prior-guided Editing\n Field"], "answer_arxiv_id": ["2303.13277"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_4960"} +{"question": "Is there any reference providing a benchmark for data valuation?", "answer": ["OpenDataVal: a Unified Benchmark for Data Valuation"], "answer_arxiv_id": ["2306.10577"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_4961"} +{"question": "Can you name studies that used different norm measures to quantify the approximation error in neural network compression?", "answer": ["Exploiting Linear Structure Within Convolutional Networks for Efficient Evaluation", "Speeding up Convolutional Neural Networks with Low Rank Expansions", "Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition", "Tensorizing Neural Networks", "Compression of Deep Convolutional Neural Networks for Fast and Low Power Mobile Applications", "Low-Rank+Sparse Tensor Compression for Neural Networks", "Compressing Neural Networks: Towards Determining the Optimal Layer-wise Decomposition"], "answer_arxiv_id": ["1404.0736", "1405.3866", "1412.6553", "1509.06569", "1511.06530", "2111.01697", "2107.11442"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_4962"} +{"question": "Could you provide me the studies that addressed other problems related to understanding neural network predictions?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)"], "answer_arxiv_id": ["1711.11279"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_4963"} +{"question": "Which research construct a function which can be approximated by a poly-width network with large depth, but not with smaller depth?", "answer": ["Benefits of depth in neural networks"], "answer_arxiv_id": ["1602.04485"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_4964"} +{"question": "Could you name works that proposed kernel-based methods in the context of causal inference?", "answer": ["Kernel Instrumental Variable Regression", "Kernel Methods for Causal Functions: Dose, Heterogeneous, and Incremental Response Curves", "Causal Inference with Treatment Measurement Error: A Nonparametric Instrumental Variable Approach"], "answer_arxiv_id": ["1906.00232", "2010.04855", "2206.09186"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_4965"} +{"question": "Could you provide me some references that use the diffusion approach to learn the Energy-based model (EBM)?", "answer": ["Learning Energy-Based Models by Diffusion Recovery Likelihood"], "answer_arxiv_id": ["2012.08125"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4966"} +{"question": "Are there any references on the meta-learning of auxiliary objectives?", "answer": ["Self-Supervised Generalisation with Meta Auxiliary Learning", "Auxiliary Learning by Implicit Differentiation"], "answer_arxiv_id": ["1901.08933", "2007.02693"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_4967"} +{"question": "Who provided the first set of limitations to information theoretic generalization bounds recently?", "answer": ["Limitations of Information-Theoretic Generalization Bounds for Gradient Descent Methods in Stochastic Convex Optimization"], "answer_arxiv_id": ["2212.13556v3"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_4968"} +{"question": "What research introduced sentence-level perturbations in adversarial attacks on NLP tasks?", "answer": ["Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Models", "Adversarial Example Generation with Syntactically Controlled Paraphrase Networks", "Stress Test Evaluation for Natural Language Inference"], "answer_arxiv_id": ["2111.02840", "1804.06059", "1806.00692"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_4969"} +{"question": "Which papers described the use of random sampling in randomized algorithms to make it more memory-efficient?", "answer": ["Breaking the Linear Iteration Cost Barrier for Some Well-known Conditional Gradient Methods Using MaxIP Data-structures", "RSC: Accelerating Graph Neural Networks Training via Randomized Sparse Computations"], "answer_arxiv_id": ["2111.15139v1", "2210.10737v2"], "source_meta": {"published_time": "20220805"}, "qid": "AutoScholarQuery_train_4970"} +{"question": "Which works explored use of diffusion models in the context of human motion generation?", "answer": ["MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "PhysDiff: Physics-Guided Human Motion Diffusion Model", "Guided Motion Diffusion for Controllable Human Motion Synthesis"], "answer_arxiv_id": ["2208.15001", "2212.02500", "2305.12577"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_4971"} +{"question": "Could you provide me some deep-learning based methods that jointly train an encoder and a decoder to watermark images?", "answer": ["HiDDeN: Hiding Data With Deep Networks", "ReDMark: Framework for Residual Diffusion Watermarking based on Deep Networks"], "answer_arxiv_id": ["1807.09937", "1810.07248"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_4972"} +{"question": "What research have leveraged learnable modules in modular methods?", "answer": ["Active Neural Localization", "No RL, No Simulation: Learning to Navigate without Navigating", "Cognitive Mapping and Planning for Visual Navigation", "Neural Topological SLAM for Visual Navigation", "Semi-parametric Topological Memory for Navigation", "Learning To Explore Using Active Neural SLAM", "EMPNet: Neural Localisation and Mapping Using Embedded Memory Points", "Learning Active Camera for Multi-Object Navigation"], "answer_arxiv_id": ["1801.08214", "2110.09470", "1702.03920v3", "2005.12256", "1803.00653", "2004.05155", "1907.13268", "2210.07505"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_4973"} +{"question": "Could you provide me with some research on continuous prompts?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "GPT Understands, Too", "PTR: Prompt Tuning with Rules for Text Classification", "Knowledgeable Prompt-tuning: Incorporating Knowledge into Prompt Verbalizer for Text Classification"], "answer_arxiv_id": ["2101.00190", "2104.08691", "2103.10385", "2105.11259", "2108.02035"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_4974"} +{"question": "Could you provide me some works that utilized Neural Radiance Fields or NeRF for learning 5D INR and reconstructing radiance fields?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_4975"} +{"question": "What references mentioned point cloud transformers?", "answer": ["Point Transformer", "PCT: Point Cloud Transformer", "Voxel Transformer for 3D Object Detection", "Voxel Set Transformer: A Set-to-Set Approach to 3D Object Detection from Point Clouds", "Embracing Single Stride 3D Object Detector with Sparse Transformer", "Point Transformer V2: Grouped Vector Attention and Partition-based Pooling", "FlatFormer: Flattened Window Attention for Efficient Point Cloud Transformer", "Stratified Transformer for 3D Point Cloud Segmentation", "3DCTN: 3D Convolution-Transformer Network for Point Cloud Classification", "Swin3D: A Pretrained Transformer Backbone for 3D Indoor Scene Understanding", "OctFormer: Octree-based Transformers for 3D Point Clouds"], "answer_arxiv_id": ["2012.09164", "2012.09688", "2109.02497", "2203.10314", "2112.06375", "2210.05666", "2301.08739", "2203.14508", "2203.00828", "2304.06906", "2305.03045"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_4976"} +{"question": "What works provided improvements in diffusion models through various training and sampling techniques?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["1907.05600", "2011.13456", "2006.11239", "2010.02502"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_4977"} +{"question": "Which researchers performed demonstration selection through a kNN-based retriever for example selection in in-context learning?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?"], "answer_arxiv_id": ["2101.06804"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_4978"} +{"question": "Could you give me the reference of studies who have argued that larger batch sizes can lead to improved performance in distributed settings?", "answer": ["Accelerated Methods for Deep Reinforcement Learning"], "answer_arxiv_id": ["1803.02811v2"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_4979"} +{"question": "Which works first explored introducing vanilla NeRF into semantic masks?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene\n Representation"], "answer_arxiv_id": ["2103.15875"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_4980"} +{"question": "What work should we refer to regarding the ESBN and CoRelNet architectures, which both employed the concept of a relational bottleneck in different ways?", "answer": ["Emergent Symbols through Binding in External Memory", "On Neural Architecture Inductive Biases for Relational Tasks"], "answer_arxiv_id": ["2012.14601", "2206.05056"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_4981"} +{"question": "Can you name some studies that focus on feature distillation in distillation-based methods in continual learning?", "answer": ["PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning"], "answer_arxiv_id": ["2004.13513"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_4982"} +{"question": "Who proposed a robust contrastive loss function inspired by symmetric losses for noisy label learning?", "answer": ["Robust Contrastive Learning against Noisy Views"], "answer_arxiv_id": ["2201.04309"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_4983"} +{"question": "Are there any works that discuss using AMP as an inspiration for deriving new spectral methods?", "answer": ["Spectral Clustering of Graphs with the Bethe Hessian", "Constrained Low-rank Matrix Estimation: Phase Transitions, Approximate Message Passing and Applications.", "The spiked matrix model with generative priors", "Fundamental Limits of Weak Recovery with Applications to Phase Retrieval", "Optimal Combination of Linear and Spectral Estimators for Generalized Linear Models", "Construction of optimal spectral methods in phase retrieval", "Estimation in Rotationally Invariant Generalized Linear Models via Approximate Message Passing"], "answer_arxiv_id": ["1406.1880", "1701.00858", "1905.12385", "1708.05932v3", "2008.03326", "2012.04524", "2112.04330"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_4984"} +{"question": "Could you provide me some works about establishing a standard benchmark for Activity Cliff Prediction?", "answer": ["Activity Cliff Prediction: Dataset and Benchmark"], "answer_arxiv_id": ["2302.07541"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_4985"} +{"question": "Could you provide me some research using user instructions to simulate real-world interactions in model training?", "answer": ["WizardLM: Empowering Large Language Models to Follow Complex\n Instructions", "Orca: Progressive Learning from Complex Explanation Traces of GPT-4"], "answer_arxiv_id": ["2304.12244", "2306.02707v1"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_4986"} +{"question": "What research has used a pre-trained LLM to generate adversarial prompts to the victim LLM?", "answer": ["Jailbreaking Black Box Large Language Models in Twenty Queries"], "answer_arxiv_id": ["2310.08419"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_4987"} +{"question": "What work proposes a decentralized training algorithm where each agent runs single-agent offline RL over the individual Q-functions?", "answer": ["Plan Better Amid Conservatism: Offline Multi-Agent Reinforcement Learning with Actor Rectification"], "answer_arxiv_id": ["2111.11188"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_4988"} +{"question": "What is the study which proposed a doubly robust method for the local average treatment effect using continuous instruments?", "answer": ["Robust causal inference with continuous instruments using the local instrumental variable curve"], "answer_arxiv_id": ["1607.02566"], "source_meta": {"published_time": "20220817"}, "qid": "AutoScholarQuery_train_4989"} +{"question": "Can you list the studies that use a detection model as a backdoor defence mechanism?", "answer": ["AEVA: Black-box Backdoor Detection Using Adversarial Extreme Value Analysis", "SCALE-UP: An Efficient Black-box Input-level Backdoor Detection via Analyzing Scaled Prediction Consistency", "Rethinking the Backdoor Attacks’ Triggers: A Frequency Perspective", "Black-box Detection of Backdoor Attacks with Limited Information and Data", "An Adaptive Black-box Defense against Trojan Attacks (TrojDef)", "STRIP: A Defence Against Trojan Attacks on Deep Neural Networks", "Model Agnostic Defence against Backdoor Attacks in Machine Learning", "XGBD: Explanation-Guided Graph Backdoor Detection"], "answer_arxiv_id": ["2110.14880", "2302.03251", "2104.03413", "2103.13127", "2209.01721", "1902.06531", "1908.02203", "2308.04406"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_4990"} +{"question": "Could you provide a study that investigated statistical properties of numerical trajectories by modeling numerical round-off errors as small random perturbations?", "answer": ["Pseudo-Orbits, Stationary Measures and Metastability"], "answer_arxiv_id": ["1211.2952"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_4991"} +{"question": "What studies showed the effect of pixel cooperation on the inference of deep learning models through game-theoretic interactions?", "answer": ["A Game-Theoretic Taxonomy of Visual Concepts in DNNs", "Discovering and Explaining the Representation Bottleneck of DNNs", "A Unified Approach to Interpreting and Boosting Adversarial\n Transferability", "Interpreting and Boosting Dropout from a Game-Theoretic View", "Game-Theoretic Understanding of Misclassification"], "answer_arxiv_id": ["2106.10938", "2111.06236", "2010.04055", "2009.11729", "2210.03349"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_4992"} +{"question": "Could you provide some references about AUC optimization in the interested range?", "answer": ["Large-scale Optimization of Partial AUC in a Range of False Positive Rates"], "answer_arxiv_id": ["2203.01505"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_4993"} +{"question": "Can you name the studies that used implicit methods for 3D human reconstruction?", "answer": ["PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization", "PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution\n 3D Human Digitization", "Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view\n Human Reconstruction", "PaMIR: Parametric Model-Conditioned Implicit Representation for\n Image-based Human Reconstruction", "Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing", "ARCH: Animatable Reconstruction of Clothed Humans", "ARCH++: Animation-Ready Clothed Human Reconstruction Revisited", "ICON: Implicit Clothed humans Obtained from Normals", "ECON: Explicit Clothed humans Optimized via Normal integration", "TeCH: Text-guided Reconstruction of Lifelike Clothed Humans", "DiffuStereo: High Quality Human Reconstruction via Diffusion-based\n Stereo Using Sparse Cameras", "POSEFusion: Pose-guided Selective Fusion for Single-view Human\n Volumetric Capture", "D-IF: Uncertainty-aware Human Digitization via Implicit Distribution\n Field"], "answer_arxiv_id": ["1905.05172", "2004.00452", "2006.08072", "2007.03858", "2204.08906", "2004.04572", "2108.07845", "2112.09127", "2212.07422", "2308.08545", "2207.08000", "2103.15331", "2308.08857"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_4994"} +{"question": "What papers have presented datasets for part detection?", "answer": ["Look into Person: Self-supervised Structure-sensitive Learning and A New Benchmark for Human Parsing", "Detect What You Can: Detecting and Representing Objects using Holistic Models and Body Parts", "PACO: Parts and Attributes of Common Objects"], "answer_arxiv_id": ["1703.05446", "1406.2031", "2301.01795"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_4995"} +{"question": "Which study proposed a tensor-based model that is as powerful as the 3-WL test?", "answer": ["Provably Powerful Graph Networks"], "answer_arxiv_id": ["1905.11136"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_4996"} +{"question": "Could you provide me some studies that used representation-based methods to compare different types of neural networks?", "answer": ["Do Wide and Deep Networks Learn the Same Things? Uncovering How Neural Network Representations Vary with Width and Depth", "Do Vision Transformers See Like Convolutional Neural Networks?", "Similarity Analysis of Contextual Word Representation Models"], "answer_arxiv_id": ["2010.15327", "2108.08810", "2005.01172"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_4997"} +{"question": "Which study proposed a method for learning a consensus representation from multiview graph for attributed graph clustering?", "answer": ["Self-supervised Contrastive Attributed Graph Clustering"], "answer_arxiv_id": ["2110.08264"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_4998"} +{"question": "What papers proposed formulating text generation as a reference-free quality estimation problem aided by contrastive learning?", "answer": ["SimCLS: A Simple Framework for Contrastive Learning of Abstractive Summarization"], "answer_arxiv_id": ["2106.01890"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_4999"} +{"question": "Have there been studies on the use of ensemble of discriminators in GAN-TTS for audio synthesis?", "answer": ["High Fidelity Speech Synthesis with Adversarial Networks"], "answer_arxiv_id": ["1909.11646"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_5000"} +{"question": "What studies have brought diffusion models into NVS?", "answer": ["Novel View Synthesis with Diffusion Models", "RenderDiffusion: Image Diffusion for 3D Reconstruction, Inpainting and\n Generation", "HoloDiffusion: Training a 3D Diffusion Model using 2D Images", "SparseFusion: Distilling View-conditioned Diffusion for 3D\n Reconstruction", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "Light Field Diffusion for Single-View Novel View Synthesis"], "answer_arxiv_id": ["2210.04628", "2211.09869", "2303.16509", "2212.00792", "2306.07881", "2309.11525"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_5001"} +{"question": "Could you provide me some studies that use reflectance as enhanced image in low-light image enhancement?", "answer": ["Toward Fast, Flexible, and Robust Low-Light Image Enhancement", "Retinexformer: One-stage Retinex-based Transformer for Low-light Image\n Enhancement"], "answer_arxiv_id": ["2204.10137", "2303.06705"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_5002"} +{"question": "Do we have any works utilizing Successor Features with GPI for transferring knowledge?", "answer": ["Successor Features for Transfer in Reinforcement Learning"], "answer_arxiv_id": ["1606.05312"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_5003"} +{"question": "What researches apply ML models to directly predict solutions for MILPs?", "answer": ["Solving Mixed Integer Programs Using Neural Networks", "MIP-GNN: A Data-Driven Framework for Guiding Combinatorial Solvers", "A GNN-Guided Predict-and-Search Framework for Mixed-Integer Linear Programming"], "answer_arxiv_id": ["2012.13349", "2205.14210", "2302.05636"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_5004"} +{"question": "Which paper established the asymptotic global convergence of policy gradient under different policy parameterizations?", "answer": ["On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift"], "answer_arxiv_id": ["1908.00261"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_5005"} +{"question": "Which works have proposed Label Distribution Learning (LDL) and Label Enhancement (LE)?", "answer": ["Label Distribution Learning"], "answer_arxiv_id": ["1408.6027"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_5006"} +{"question": "Can you cite researches that searches for LLM's model failure modes by testing manually written prompts?", "answer": ["Beyond Accuracy: Behavioral Testing of NLP Models with CheckList"], "answer_arxiv_id": ["2005.04118"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_5007"} +{"question": "Can you tell me about some works that focus on automated hyper-parameter setting in the field of AutoML?", "answer": ["Fast Bayesian Optimization of Machine Learning Hyperparameters on Large Datasets"], "answer_arxiv_id": ["1605.07079"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_5008"} +{"question": "Can you give me some examples of researches that use dense 2D-3D correspondences for instance-level 6D object pose estimation?", "answer": ["RNNPose: Recurrent 6-DoF Object Pose Refinement with Robust\n Correspondence Field Estimation and Pose Optimization", "Pix2Pose: Pixel-Wise Coordinate Regression of Objects for 6D Pose\n Estimation"], "answer_arxiv_id": ["2203.12870", "1908.07433"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_5009"} +{"question": "What work proposed the use of multi-level hash tables to improve the quality and efficiency of NeRF?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_5010"} +{"question": "Which research papers are associated with generation tasks in the vision foundation model space?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_5011"} +{"question": "What papers discussed the holistic joint generation of expression and gesture?", "answer": ["Learning Speech-driven 3D Conversational Gestures from Video", "Generating Holistic 3D Human Motion from Speech"], "answer_arxiv_id": ["2102.06837", "2212.04420"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_5012"} +{"question": "Could you tell me the research papers that address the challenges of off-policy learning?", "answer": ["Distributed Prioritized Experience Replay"], "answer_arxiv_id": ["1803.00933"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_5013"} +{"question": "Which work presents theoretical considerations on corrupted offline RL and provides suboptimality bounds?", "answer": ["Corruption-Robust Offline Reinforcement Learning"], "answer_arxiv_id": ["2106.06630"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_5014"} +{"question": "Could you provide me with papers that discuss the pitfalls of saliency maps when used post-hoc?", "answer": ["Visualizing and Understanding Convolutional Networks", "Axiomatic Attribution for Deep Networks", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "The (Un)reliability of saliency methods", "Sanity Checks for Saliency Maps"], "answer_arxiv_id": ["1311.2901", "1703.01365", "1610.02391", "1711.00867", "1810.03292v3"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_5015"} +{"question": "Which research optimized normalizing flows through maximum likelihood estimation at the cost of limiting the expressive power of the representation?", "answer": ["Variational Inference with Normalizing Flows", "Density estimation using Real NVP", "Neural Autoregressive Flows", "Neural Spline Flows"], "answer_arxiv_id": ["1505.05770", "1605.08803", "1804.00779", "1906.04032"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5016"} +{"question": "What papers propose learning objectives compatible with stochastic optimization that can be implemented by deep networks in PLL?", "answer": ["Network Cooperation with Progressive Disambiguation for Partial Label Learning", "Progressive Identification of True Labels for Partial-Label Learning"], "answer_arxiv_id": ["2002.11919", "2002.08053"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_5017"} +{"question": "Which research works aim to learn both disentangled and equivariant group representations with the help of topological group structure?", "answer": ["Towards a Definition of Disentangled Representations"], "answer_arxiv_id": ["1812.02230"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_5018"} +{"question": "What works developed reduced variance variants like MIS and doubly robust estimation?", "answer": ["Towards Optimal Off-Policy Evaluation for Reinforcement Learning with Marginalized Importance Sampling", "Doubly Robust Off-policy Value Evaluation for Reinforcement Learning", "Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning"], "answer_arxiv_id": ["1906.03393", "1511.03722", "1604.00923"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_5019"} +{"question": "Which previous works do not remove tokens but apply a mask during training in order to solve the problem of varying number of tokens in images or sentences?", "answer": ["PoWER-BERT: Accelerating BERT Inference via Progressive Word-vector Elimination", "Length-Adaptive Transformer: Train Once with Length Drop, Use Anytime with Search", "Learned Token Pruning for Transformers", "A Study on Token Pruning for ColBERT", "AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "A-ViT: Adaptive Tokens for Efficient Vision Transformer", "SPViT: Enabling Faster Vision Transformers via Soft Token Pruning", "CP-ViT: Cascade Vision Transformer Pruning via Progressive Sparsity Prediction", "DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "A Unified Pruning Framework for Vision Transformers"], "answer_arxiv_id": ["2001.08950", "2010.07003", "2107.00910", "2112.06540", "2111.15668", "2112.07658", "2112.13890v2", "2203.04570", "2106.02034", "2111.15127"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_5020"} +{"question": "Which paper discusses the progression from single-agent systems to advanced multi-agent systems?", "answer": ["The Rise and Potential of Large Language Model Based Agents: A Survey"], "answer_arxiv_id": ["2309.07864v3"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_5021"} +{"question": "What works have been fine-tuned or have reported their scores on Diverse Search Space (DSS)?", "answer": ["MiLeNAS: Efficient Neural Architecture Search via Mixed-Level Reformulation", "PC-DARTS: Partial Channel Connections for Memory-Efficient Architecture Search", "Geometry-Aware Gradient Algorithms for Neural Architecture Search", "DrNAS: Dirichlet Neural Architecture Search", "Robustifying DARTS by Eliminating Information Bypass Leakage via Explicit Sparse Regularization", "Rethinking Architecture Selection in Differentiable NAS", "DARTS-: Robustly Stepping out of Performance Collapse Without Indicators", "Searching for A Robust Neural Architecture in Four GPU Hours", "Stabilizing Differentiable Architecture Search via Perturbation-based Regularization", "Fair DARTS: Eliminating Unfair Advantages in Differentiable Architecture Search", "Understanding and Robustifying Differentiable Architecture Search", "Rapid Neural Architecture Search by Learning to Generate Graphs from Datasets"], "answer_arxiv_id": ["2003.12238", "1907.05737", "2004.07802", "2006.10355", "2306.06858", "2108.04392", "2009.01027", "1910.04465", "2002.05283", "1911.12126", "1909.09656", "2107.00860"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_5022"} +{"question": "What studies explored text-to-video retrieval?", "answer": ["CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "Bridging Video-text Retrieval with Multiple Choice Questions", "TS2-Net: Token Shift and Selection Transformer for Text-Video Retrieval", "X-CLIP: End-to-End Multi-grained Contrastive Learning for Video-Text\n Retrieval"], "answer_arxiv_id": ["2106.11097", "2201.04850", "2207.07852", "2207.07285"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_5023"} +{"question": "What papers showed about the advantages of adding adapter architectures to image foundation models?", "answer": ["Frozen CLIP Models are Efficient Video Learners", "ST-Adapter: Parameter-Efficient Image-to-Video Transfer Learning"], "answer_arxiv_id": ["2208.03550", "2206.13559"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5024"} +{"question": "What papers discuss contrastive learning as a technique for representation learning in reinforcement learning?", "answer": ["Time-Contrastive Networks: Self-Supervised Learning from Video", "Representation Learning with Contrastive Predictive Coding", "Learning Predictive Representations for Deformable Objects Using Contrastive Estimation", "Deep Reinforcement and InfoMax Learning"], "answer_arxiv_id": ["1704.06888", "1807.03748", "2003.05436", "2006.07217"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_5025"} +{"question": "What research allows to decouple the gradient updates from communication?", "answer": ["Optimal Gradient Tracking for Decentralized Optimization"], "answer_arxiv_id": ["2110.05282v4"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_5026"} +{"question": "Could you provide me some papers demonstrating that equivariant methods can increase sample efficiency in reinforcement learning?", "answer": ["MDP Homomorphic Networks: Group Symmetries in Reinforcement Learning", "Group Equivariant Deep Reinforcement Learning"], "answer_arxiv_id": ["2006.16908", "2007.03437"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_5027"} +{"question": "Which works have explored the use of machine learning models to infer variable types?", "answer": ["TypeWriter: Neural Type Prediction with Search-based Validation", "LambdaNet: Probabilistic Type Inference using Graph Neural Networks"], "answer_arxiv_id": ["1912.03768", "2005.02161"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_5028"} +{"question": "Could you provide me some studies about different modalities of Visual Odometry such as visual-inertial odometry (VIO) and stereo VO?", "answer": ["Direct Sparse Visual-Inertial Odometry using Dynamic Marginalization", "Stereo DSO: Large-Scale Direct Sparse Visual Odometry with Stereo Cameras"], "answer_arxiv_id": ["1804.05625", "1708.07878"], "source_meta": {"published_time": "20220808"}, "qid": "AutoScholarQuery_train_5029"} +{"question": "What papers introduced multimodal fusion strategies for scene flow estimation?", "answer": ["Hidden Gems: 4D Radar Scene Flow Learning Using Cross-Modal Supervision", "DeepLiDARFlow: A Deep Learning Architecture For Scene Flow Estimation\n Using Monocular Camera and Sparse LiDAR", "CamLiFlow: Bidirectional Camera-LiDAR Fusion for Joint Optical Flow and\n Scene Flow Estimation", "RPEFlow: Multimodal Fusion of RGB-PointCloud-Event for Joint Optical\n Flow and Scene Flow Estimation"], "answer_arxiv_id": ["2303.00462", "2008.08136", "2111.10502", "2309.15082"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5030"} +{"question": "Which studies aimed to modify the input data in terms of algorithmic fairness?", "answer": ["A Confidence-Based Approach for Balancing Fairness and Accuracy"], "answer_arxiv_id": ["1601.05764"], "source_meta": {"published_time": "20220916"}, "qid": "AutoScholarQuery_train_5031"} +{"question": "Can you identify the prior work that focused on proving the equivalence between GRW and ERM under the zero-one loss with a monotonic relationship?", "answer": ["Does Distributionally Robust Supervised Learning Give Robust Classifiers?"], "answer_arxiv_id": ["1611.02041"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_5032"} +{"question": "What are some works that have explored the integration of contrastive learning techniques into federated learning to prevent local client drift and enhance local training?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1807.03748", "2002.05709", "1911.05722"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_5033"} +{"question": "What papers detail the applicability of NeRF to both rigid and non-rigid dynamic scenes?", "answer": ["Nerfies: Deformable Neural Radiance Fields", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View\n Synthesis of a Dynamic Scene From Monocular Video", "MonoNeRF: Learning a Generalizable Dynamic Radiance Field from Monocular\n Videos", "Dynamic View Synthesis from Dynamic Monocular Video", "Structured Local Radiance Fields for Human Avatar Modeling"], "answer_arxiv_id": ["2011.12948", "2011.13961", "2012.12247", "2212.13056", "2105.06468", "2203.14478"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_5034"} +{"question": "Could you mention some works that targeted open-vocabulary 3D object detection?", "answer": ["End-to-End Object Detection with Transformers", "An End-to-End Transformer Model for 3D Object Detection", "Open-Vocabulary Point-Cloud Object Detection without 3D Annotation", "Learning Transferable Visual Models From Natural Language Supervision", "Detecting Twenty-thousand Classes using Image-level Supervision"], "answer_arxiv_id": ["2005.12872", "2109.08141", "2304.00788v2", "2103.00020", "2201.02605"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_5035"} +{"question": "What studies have successfully applied data programming to medicine and industry?", "answer": ["Cross-Modal Data Programming Enables Rapid Medical Machine Learning", "Overton: A Data System for Monitoring and Improving Machine-Learned Products", "Snorkel DryBell: A Case Study in Deploying Weak Supervision at Industrial Scale"], "answer_arxiv_id": ["1903.11101", "1909.05372", "1812.00417"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_5036"} +{"question": "What works are about the use of CLIP for 3D understanding and manipulation?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields"], "answer_arxiv_id": ["2112.02413", "2112.05139"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_5037"} +{"question": "Can you provide a study where the frequency of correct predictions was interpreted as a sample difficulty measure?", "answer": ["An Empirical Study of Example Forgetting during Deep Neural Network Learning"], "answer_arxiv_id": ["1812.05159"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_5038"} +{"question": "Which works introduced techniques in deep reinforcement learning derived from count-based or prediction-based exploration techniques?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "First return, then explore"], "answer_arxiv_id": ["1606.01868", "2004.12919v6"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_5039"} +{"question": "What work demonstrated that learning the variance schedule improves performance on image density estimation benchmarks?", "answer": ["Variational Diffusion Models"], "answer_arxiv_id": ["2107.00630"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5040"} +{"question": "What works in the Logistic Bandits domain focus on designing methods to handle the heteroskedasticity?", "answer": ["An Experimental Design Approach for Regret Minimization in Logistic Bandits"], "answer_arxiv_id": ["2202.02407"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_5041"} +{"question": "Which studies provided benchmark assessments for resolve referential uncertainty when learning new words in few-shot settings?", "answer": ["MEWL: Few-shot multimodal word learning with referential uncertainty"], "answer_arxiv_id": ["2306.00503"], "source_meta": {"published_time": "20220618"}, "qid": "AutoScholarQuery_train_5042"} +{"question": "Are there any studies that learn representations for 3D scenes?", "answer": ["CoCoNets: Continuous Contrastive 3D Scene Representations"], "answer_arxiv_id": ["2104.03851"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_5043"} +{"question": "Which works have found evidence for unpredictable emergence of several functional linguistic abilities in large language models (LLMs)?", "answer": ["Emergent Abilities of Large Language Models"], "answer_arxiv_id": ["2206.07682"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_5044"} +{"question": "What are the works that apply recurrent neural networks to exploit temporal information?", "answer": ["Frame-Recurrent Video Super-Resolution", "Efficient Video Super-Resolution through Recurrent Latent Space\n Propagation", "Video Super-Resolution with Recurrent Structure-Detail Network", "BasicVSR: The Search for Essential Components in Video Super-Resolution\n and Beyond", "BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation\n and Alignment", "Rethinking Alignment in Video Super-Resolution Transformers"], "answer_arxiv_id": ["1801.04590", "1909.08080", "2008.00455", "2012.02181", "2104.13371", "2207.08494"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_5045"} +{"question": "Are there any studies that provide a theoretical understanding of model compressibility?", "answer": ["Stronger generalization bounds for deep nets via a compression approach", "A Theoretical Understanding of Neural Network Compression from Sparse Linear Approximation", "Data-Dependent Coresets for Compressing Neural Networks with Applications to Generalization Bounds", "Data-Independent Neural Pruning via Coresets", "Good Subnetworks Provably Exist: Pruning via Greedy Forward Selection"], "answer_arxiv_id": ["1802.05296", "2206.05604", "1804.05345", "1907.04018", "2003.01794"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_5046"} +{"question": "Which project applied depth-wise convolutional operations to achieve a better trade-off between accuracy and efficiency for Vision Transformers (ViT)?", "answer": ["CMT: Convolutional Neural Networks Meet Vision Transformers"], "answer_arxiv_id": ["2107.06263"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_5047"} +{"question": "Who developed Binder which addressed mostly answering questions about tables using SQL and SQL-like Python?", "answer": ["Binding Language Models in Symbolic Languages"], "answer_arxiv_id": ["2210.02875"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_5048"} +{"question": "Are there any works that employ Transformer networks for single image super-resolution?", "answer": ["Pre-Trained Image Processing Transformer", "Cross Aggregation Transformer for Image Restoration", "SwinIR: Image Restoration Using Swin Transformer", "Activating More Pixels in Image Super-Resolution Transformer"], "answer_arxiv_id": ["2012.00364", "2211.13654", "2108.10257", "2205.04437"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5049"} +{"question": "Could you provide me some studies that improved regret rates by using a private version of the shrinking dartboard algorithm?", "answer": ["Private Online Prediction from Experts: Separations and Faster Rates"], "answer_arxiv_id": ["2210.13537"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_5050"} +{"question": "Could you provide me some studies related to federated learning methods?", "answer": ["SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Federated Accelerated Stochastic Gradient Descent", "Linear Convergence in Federated Learning: Tackling Client Heterogeneity and Sparse Gradients", "Federated Learning with Buffered Asynchronous Aggregation"], "answer_arxiv_id": ["1910.06378", "2006.08950", "2102.07053", "2106.06639v4"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_5051"} +{"question": "Could you provide me a study that proposed a decomposed cross-modal distillation framework to enhance RGB-based temporal action detection?", "answer": ["Decomposed Cross-modal Distillation for RGB-based Temporal Action\n Detection"], "answer_arxiv_id": ["2303.17285"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_5052"} +{"question": "Which studies used LLMs to improve negotiation scenarios?", "answer": ["Improving Language Model Negotiation with Self-Play and In-Context\n Learning from AI Feedback"], "answer_arxiv_id": ["2305.10142"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_5053"} +{"question": "Could you provide me any work that illustrated the optimal matrix factorization significantly improves the privacy-utility-computation tradeoffs?", "answer": ["Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams"], "answer_arxiv_id": ["2202.08312"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_5054"} +{"question": "Are there any papers investigating the models' sensitivities to non-robust features?", "answer": ["Intriguing properties of neural networks", "Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1312.6199", "1412.6572", "1706.06083"], "source_meta": {"published_time": "20220629"}, "qid": "AutoScholarQuery_train_5055"} +{"question": "Which datasets are based on onboard sensor data collection from moving vehicles equipped with multiple sensors?", "answer": ["nuScenes: A multimodal dataset for autonomous driving", "Scalability in Perception for Autonomous Driving: Waymo Open Dataset"], "answer_arxiv_id": ["1903.11027", "1912.04838"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_5056"} +{"question": "What research papers use GRU4Rec or BERT4Rec to exploit user interaction histories in sequential recommendation?", "answer": ["Session-based Recommendations with Recurrent Neural Networks", "BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer"], "answer_arxiv_id": ["1511.06939", "1904.06690"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_5057"} +{"question": "What works are used to assess the visual reasoning capabilities of tested models?", "answer": ["GuessWhat?! Visual object discovery through multi-modal dialogue"], "answer_arxiv_id": ["1611.08481"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_5058"} +{"question": "What studies proposed solutions for class-imbalanced learning?", "answer": ["Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting", "Class-Balanced Loss Based on Effective Number of Samples"], "answer_arxiv_id": ["1902.07379", "1901.05555"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_5059"} +{"question": "What papers discuss the development of foundational models in the field of multimodal learning?", "answer": ["On the Opportunities and Risks of Foundation Models", "Reproducible scaling laws for contrastive language-image learning", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2108.07258", "2212.07143", "2204.06125", "2205.11487", "2112.10752", "2207.12598"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_5060"} +{"question": "Could you provide me some studies about training V2V models with paired spatial controls and video data?", "answer": ["Structure and Content-Guided Video Synthesis with Diffusion Models", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "Control-A-Video: Controllable Text-to-Video Generation with Diffusion\n Models"], "answer_arxiv_id": ["2302.03011", "2306.02018", "2305.13840"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_5061"} +{"question": "Could you detail some research that proposed using hybrid models for symbolic regression?", "answer": ["Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients", "Data-driven discovery of free-form governing differential equations", "AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularity", "Inferring the Structure of Ordinary Differential Equations"], "answer_arxiv_id": ["1912.04871", "1910.05117", "2006.10782", "2107.07345"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_5062"} +{"question": "Could you provide me some studies about the intrinsic properties of representations produced by the Transformer models?", "answer": ["Attention is not all you need: pure attention loses rank doubly exponentially with depth", "A Contrastive Framework for Neural Text Generation", "How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddings"], "answer_arxiv_id": ["2103.03404", "2202.06417", "1909.00512"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_5063"} +{"question": "Could you cite some works that address the distributional shift in offline RL by making conservative estimates of future rewards?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "COMBO: Conservative Offline Model-Based Policy Optimization", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["2006.04779", "2102.08363", "2106.06860"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_5064"} +{"question": "Which papers focused on task distributions in the context of meta-learning?", "answer": ["Open-Ended Learning Leads to Generally Capable Agents", "Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning", "Leveraging Procedural Generation to Benchmark Reinforcement Learning"], "answer_arxiv_id": ["2107.12808", "1910.10897", "1912.01588"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_5065"} +{"question": "What research attempts to overcome the struggles of conditioning on unseen attribute values via the approach of iteratively quantizing into more fine-grained control codes?", "answer": ["Quark: Controllable Text Generation with Reinforced [Un]learning"], "answer_arxiv_id": ["2205.13636"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_5066"} +{"question": "Can you point to a study that models long-term temporal self attention but only for each individual speaker?", "answer": ["Is Someone Speaking? Exploring Long-term Temporal Features for\n Audio-visual Active Speaker Detection"], "answer_arxiv_id": ["2107.06592"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_5067"} +{"question": "Which papers reviewed previous works that handled reposing, virtual try-on, and text manipulation tasks independently?", "answer": ["Person Image Synthesis via Denoising Diffusion Model", "UPGPT: Universal Diffusion Model for Person Image Generation, Editing\n and Pose Transfer", "Pose with Style: Detail-Preserving Pose-Guided Image Synthesis with\n Conditional StyleGAN", "Exploring Dual-task Correlation for Pose Guided Person Image Generation", "Dressing in Order: Recurrent Person Image Generation for Pose Transfer,\n Virtual Try-on and Outfit Editing", "Self-supervised Correlation Mining Network for Person Image Generation", "Collecting The Puzzle Pieces: Disentangled Self-Driven Human Pose\n Transfer by Permuting Textures", "Towards Scalable Unpaired Virtual Try-On via Patch-Routed\n Spatially-Adaptive GAN", "LaDI-VTON: Latent Diffusion Textual-Inversion Enhanced Virtual Try-On", "TryOnDiffusion: A Tale of Two UNets", "High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled\n Conditions", "GP-VTON: Towards General Purpose Virtual Try-on via Collaborative\n Local-Flow Global-Parsing Learning", "Muse: Text-To-Image Generation via Masked Generative Transformers", "Paint by Example: Exemplar-based Image Editing with Diffusion Models"], "answer_arxiv_id": ["2211.12500", "2304.08870", "2109.06166", "2203.02910", "2104.07021", "2111.13307", "2210.01887", "2111.10544", "2305.13501", "2306.08276", "2206.14180", "2303.13756", "2301.00704", "2211.13227"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_5068"} +{"question": "What papers were referenced in the discussion on susceptibility of machine-generated detection methods to authorship obfuscation attacks?", "answer": ["Adversarial Robustness of Neural-Statistical Features in Detection of\n Generative Transformers", "Paraphrasing evades detectors of AI-generated text, but retrieval is an\n effective defense", "Red Teaming Language Model Detectors with Language Models", "OUTFOX: LLM-Generated Essay Detection Through In-Context Learning with\n Adversarially Generated Examples"], "answer_arxiv_id": ["2203.07983", "2303.13408", "2305.19713", "2307.11729"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_5069"} +{"question": "What are some studies that explored using Word Mover’s Distance for sentence-level matching?", "answer": ["MoverScore: Text Generation Evaluating with Contextualized Embeddings and Earth Mover Distance"], "answer_arxiv_id": ["1909.02622"], "source_meta": {"published_time": "20220801"}, "qid": "AutoScholarQuery_train_5070"} +{"question": "What works discussed the importance of correct ordering of LayerNorm, residual connections and attention layers?", "answer": ["On Layer Normalization in the Transformer Architecture"], "answer_arxiv_id": ["2002.04745"], "source_meta": {"published_time": "20211123"}, "qid": "AutoScholarQuery_train_5071"} +{"question": "Which works developed a method for assigning data-driven weights to conformal e-values from different machine learning models?", "answer": ["Integrative conformal p-values for powerful out-of-distribution testing with labeled outliers"], "answer_arxiv_id": ["2208.11111"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_5072"} +{"question": "What efforts have been made to reduce hallucination for Multimodal Large Language Models?", "answer": ["VIGC: Visual Instruction Generation and Correction", "Woodpecker: Hallucination Correction for Multimodal Large Language\n Models", "Detecting and Preventing Hallucinations in Large Vision Language Models", "Aligning Large Multimodal Models with Factually Augmented RLHF", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2308.12714", "2310.16045", "2308.06394", "2309.14525", "2203.02155"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_5073"} +{"question": "What papers dealt with text-based image-to-image (I2I) generation?", "answer": ["T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Prompt-to-Prompt Image Editing with Cross Attention Control", "SINE: SINgle Image Editing with Text-to-Image Diffusion Models", "Zero-shot Image-to-Image Translation", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "DiffEdit: Diffusion-based semantic image editing with mask guidance"], "answer_arxiv_id": ["2302.08453", "2211.12572", "2210.09276", "2208.01626", "2212.04489", "2302.03027", "2108.01073", "2210.11427"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_5074"} +{"question": "What is the pioneering work in open-vocabulary semantic segmentation that combines a deep visual segmentation model with a generative model of class-dependent features?", "answer": ["Zero-Shot Semantic Segmentation"], "answer_arxiv_id": ["1906.00817"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_5075"} +{"question": "Which studies derived the first line of policy optimization results for RL in linear MDPs without any reachability-style assumption?", "answer": ["PC-PG: Policy Cover Directed Exploration for Provable Policy Gradient Learning", "Cautiously Optimistic Policy Optimization and Exploration with Linear Function Approximation"], "answer_arxiv_id": ["2007.08459", "2103.12923"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_5076"} +{"question": "What studies focused on predicting an intermediate sparse or dense 3D-to-2D correspondence map in 6-DoF pose estimation?", "answer": ["BB8: A Scalable, Accurate, Robust to Partial Occlusion Method for\n Predicting the 3D Poses of Challenging Objects without Using Depth", "PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation", "HybridPose: 6D Object Pose Estimation under Hybrid Representations", "CheckerPose: Progressive Dense Keypoint Localization for Object Pose\n Estimation with Graph Neural Network", "CRT-6D: Fast 6D Object Pose Estimation with Cascaded Refinement\n Transformers", "Single-Stage 6D Object Pose Estimation", "DPODv2: Dense Correspondence-Based 6 DoF Pose Estimation", "SurfEmb: Dense and Continuous Correspondence Distributions for Object\n Pose Estimation with Learnt Surface Embeddings", "ZebraPose: Coarse to Fine Surface Encoding for 6DoF Object Pose\n Estimation", "EPOS: Estimating 6D Pose of Objects with Symmetries"], "answer_arxiv_id": ["1703.10896", "1812.11788", "2001.01869", "2303.16874", "2210.11718", "1911.08324", "2207.02805", "2111.13489", "2203.09418", "2004.00605"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5077"} +{"question": "Any works about improving performance of general contrastive learning approaches by augmenting multiple global and local views simultaneously?", "answer": ["DetCo: Unsupervised Contrastive Learning for Object Detection"], "answer_arxiv_id": ["2102.04803"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_5078"} +{"question": "Are there any studies on structural control over the diffusion process in image generation?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5079"} +{"question": "Could you provide me some works that propose utilization of multiple memory systems?", "answer": ["DualPrompt: Complementary Prompting for Rehearsal-free Continual Learning", "DualNet: Continual Learning, Fast and Slow", "Learning Fast, Learning Slow: A General Continual Learning Method based on Complementary Learning System"], "answer_arxiv_id": ["2204.04799", "2110.00175", "2201.12604"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_5080"} +{"question": "Could you mention some works about prompt engineering?", "answer": ["Prompt Programming for Large Language Models: Beyond the Few-Shot Paradigm", "Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2102.07350", "2104.08786"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_5081"} +{"question": "Any works that employ both a metric head and FiLM layers to adapt the backbone?", "answer": ["TADAM: Task dependent adaptive metric for improved few-shot learning", "Learning a Universal Template for Few-shot Dataset Generalization", "Improved Few-Shot Visual Classification"], "answer_arxiv_id": ["1805.10123", "2105.07029v2", "1912.03432"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_5082"} +{"question": "Which works have focused on distilling conversational responses from LLMs to enhance dialogue agents?", "answer": ["Dialogue Chain-of-Thought Distillation for Commonsense-aware\n Conversational Agents", "SODA: Million-scale Dialogue Distillation with Social Commonsense\n Contextualization"], "answer_arxiv_id": ["2310.09343", "2212.10465"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_5083"} +{"question": "What works have supported the visual object tracking task with autoregressive modeling?", "answer": ["Unified Sequence-to-Sequence Learning for Single- and Multi-Modal Visual\n Object Tracking"], "answer_arxiv_id": ["2304.14394"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_5084"} +{"question": "Can you list the works that experimented with multistep in diffusion integration?", "answer": ["Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2202.09778"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_5085"} +{"question": "Can you provide notable examples of adaptive sorting algorithms?", "answer": ["R"], "answer_arxiv_id": ["1210.6589"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_5086"} +{"question": "Could you tell me about any studies that implemented contrastive learning for molecule pretraining?", "answer": ["Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1807.03748"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_5087"} +{"question": "Can you provide me works about chart-to-summary generation?", "answer": ["Chart-to-Text: Generating Natural Language Descriptions for Charts by\n Adapting the Transformer Model", "Chart-to-Text: A Large-Scale Benchmark for Chart Summarization"], "answer_arxiv_id": ["2010.09142", "2203.06486"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_5088"} +{"question": "What studies have transformations of value function or a return distribution as the main aim?", "answer": ["QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement learning", "QPLEX: Duplex Dueling Multi-Agent Q-Learning"], "answer_arxiv_id": ["1905.05408", "2008.01062"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_5089"} +{"question": "Which work originally presented Generative adversarial networks (GANs)?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_5090"} +{"question": "Any papers about video understanding schemes by decoupling the spatiotemporal patterns?", "answer": ["Complex Sequential Understanding through the Awareness of Spatial and Temporal Concepts", "HOI Analysis: Integrating and Decomposing Human-Object Interaction"], "answer_arxiv_id": ["2006.00212", "2010.16219"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_5091"} +{"question": "What works have used flows as a parametric method for estimating the density of potential outcomes from observational data?", "answer": ["Normalizing Flows for Interventional Density Estimation"], "answer_arxiv_id": ["2209.06203"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_5092"} +{"question": "Could you provide me some studies that utilized pixel-space approaches in object-centric learning?", "answer": ["Multi-Object Representation Learning with Iterative Variational Inference", "Differentiable Mathematical Programming for Object-Centric Representation Learning"], "answer_arxiv_id": ["1903.00450", "2210.02159"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_5093"} +{"question": "What studies center on transferring text-image generation to text-video?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "Pix2Video: Video Editing using Image Diffusion", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "Text2LIVE: Text-Driven Layered Image and Video Editing", "Structure and Content-Guided Video Synthesis with Diffusion Models", "Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "SweetDreamer: Aligning Geometric Priors in 2D Diffusion for Consistent\n Text-to-3D"], "answer_arxiv_id": ["2209.14792", "2211.11018", "2303.12688", "2303.13439", "2204.02491", "2302.03011", "2212.11565", "2310.02596"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5094"} +{"question": "Which research study the typical architecture of instruction-following Vision Language Models (VLMs)?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning", "Improved Baselines with Visual Instruction Tuning"], "answer_arxiv_id": ["2010.11929", "2304.10592", "2304.08485", "2310.03744"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_5095"} +{"question": "Are there any works about using multiple source models for cross-domain knowledge extraction?", "answer": ["Multi-Source Unsupervised Domain Adaptation via Pseudo Target Domain"], "answer_arxiv_id": ["2202.10725"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_5096"} +{"question": "Which works propose to regularize the policy by minimizing some divergence measure?", "answer": ["Behavior Regularized Offline Reinforcement Learning", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction"], "answer_arxiv_id": ["1911.11361", "1906.00949"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_5097"} +{"question": "Which paper trained the retriever according to the feedback from black-box Language Models (LMS) in unsupervised learning of RAG?", "answer": ["REPLUG: Retrieval-Augmented Black-Box Language Models"], "answer_arxiv_id": ["2301.12652"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_5098"} +{"question": "Which works discuss the topics of geometric deep learning?", "answer": ["Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["2104.13478"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5099"} +{"question": "What research is about learning a mask over discriminatory features to reduce reliance on spurious correlations?", "answer": ["MaskTune: Mitigating Spurious Correlations by Forcing to Explore"], "answer_arxiv_id": ["2210.00055"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_5100"} +{"question": "Any works about diagnosing and rectifying vision classifiers using natural language inputs?", "answer": ["Diagnosing and Rectifying Vision Models using Language"], "answer_arxiv_id": ["2302.04269"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_5101"} +{"question": "What work describes the decomposition of a signal into coarse-to-fine LODs through convolution with appropriate kernels?", "answer": ["nerf2nerf: Pairwise Registration of Neural Radiance Fields"], "answer_arxiv_id": ["2211.01600"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_5102"} +{"question": "Are there any works that expand document revision analysis to student essays?", "answer": ["ArgRewrite V.2: an Annotated Argumentative Revisions Corpus"], "answer_arxiv_id": ["2206.01677v1"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_train_5103"} +{"question": "What works learn reward functions using videos from the Something-Something dataset?", "answer": ["Learning Generalizable Robotic Reward Functions from “In-The-Wild” Human Videos"], "answer_arxiv_id": ["2103.16817"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5104"} +{"question": "Could you provide me with some studies about the use of image discretization tokenizers in multimodal generation for LLMs?", "answer": ["Making LLaMA SEE and Draw with SEED Tokenizer"], "answer_arxiv_id": ["2310.01218"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_5105"} +{"question": "Could you provide me with some studies about articulated neural representations of human heads?", "answer": ["Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar Reconstruction", "Learning Compositional Radiance Fields of Dynamic Human Heads", "AD-NeRF: Audio Driven Neural Radiance Fields for Talking Head Synthesis", "I M Avatar: Implicit Morphable Head Avatars from Videos", "CoNeRF: Controllable Neural Radiance Fields", "MoFaNeRF: Morphable Facial Neural Radiance Field", "RigNeRF: Fully Controllable Neural 3D Portraits", "PointAvatar: Deformable Point-based Head Avatars from Videos"], "answer_arxiv_id": ["2012.03065", "2012.09955v1", "2103.11078", "2112.07471", "2112.01983", "2112.02308", "2206.06481", "2212.08377"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5106"} +{"question": "What research extracted features from self-supervised Vision Transformer for various applications?", "answer": ["Deep ViT Features as Dense Visual Descriptors"], "answer_arxiv_id": ["2112.05814"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5107"} +{"question": "Any works about the consistent improvements in diffusion models like efficient samplers, latent-space diffusion, classifier(-free) guidance?", "answer": ["Denoising Diffusion Implicit Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds", "High-Resolution Image Synthesis with Latent Diffusion Models", "Classifier-Free Diffusion Guidance", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2010.02502", "2202.09778", "2112.10752", "2207.12598", "2105.05233"], "source_meta": {"published_time": "20230407"}, "qid": "AutoScholarQuery_train_5108"} +{"question": "Which research introduced disentangled representation learning?", "answer": ["Representation Learning: A Review and New Perspectives"], "answer_arxiv_id": ["1206.5538"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_5109"} +{"question": "Any works about applying autoregressive models in Text-to-image generation?", "answer": ["Muse: Text-To-Image Generation via Masked Generative Transformers", "CogView: Mastering Text-to-Image Generation via Transformers", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2301.00704", "2105.13290", "2206.10789"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_5110"} +{"question": "Which work unified logits distillation and location distillation in object detection?", "answer": ["Localization Distillation for Object Detection"], "answer_arxiv_id": ["2204.05957"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5111"} +{"question": "Could you provide me some research that involves full or partial spectral decomposition?", "answer": ["Spectral Networks and Deep Locally Connected Networks on Graphs", "LanczosNet: Multi-Scale Deep Graph Convolutional Networks"], "answer_arxiv_id": ["1312.6203", "1901.01484"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_5112"} +{"question": "Which studies are about spectral clustering, where clustering is computed with the help of the eigenvectors of the graph Laplacian?", "answer": ["A Tutorial on Spectral Clustering"], "answer_arxiv_id": ["0711.0189"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_5113"} +{"question": "Which work leverage static 2D and 3D masks to constrain the edit area of NeRF?", "answer": ["ED-NeRF: Efficient Text-Guided Editing of 3D Scene with Latent Space\n NeRF", "DreamEditor: Text-Driven 3D Scene Editing with Neural Fields"], "answer_arxiv_id": ["2310.02712", "2306.13455"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_5114"} +{"question": "Could you provide some references that tried to answer the 'benign overfitting' phenomenon?", "answer": ["Just Interpolate: Kernel “Ridgeless” Regression Can Generalize", "Consistency of Interpolation with Laplace Kernels is a High-Dimensional Phenomenon"], "answer_arxiv_id": ["1808.00387", "1812.11167"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_5115"} +{"question": "What papers are about designing specialized watermarking mechanisms for large-scale image generation models?", "answer": ["Tree-Ring Watermarks: Fingerprints for Diffusion Images that are\n Invisible and Robust", "The Stable Signature: Rooting Watermarks in Latent Diffusion Models", "Evading Watermark based Detection of AI-Generated Content", "A Recipe for Watermarking Diffusion Models", "DiffusionShield: A Watermark for Copyright Protection against Generative\n Diffusion Models"], "answer_arxiv_id": ["2305.20030", "2303.15435", "2305.03807", "2303.10137", "2306.04642"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_5116"} +{"question": "Which works discuss counterfactual intervention methodologies?", "answer": ["Understanding Neural Networks through Representation Erasure"], "answer_arxiv_id": ["1612.08220"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_5117"} +{"question": "What studies have evaluated easy-to-hard generalization in NLP using model-based hardness measures?", "answer": ["Dataset Cartography: Mapping and Diagnosing Datasets with Training\n Dynamics", "Complexity-Based Prompting for Multi-Step Reasoning"], "answer_arxiv_id": ["2009.10795", "2210.00720"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_5118"} +{"question": "What papers explain the concept of Gaussian splatting?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis", "4D Gaussian Splatting for Real-Time Dynamic Scene Rendering"], "answer_arxiv_id": ["2308.04079", "2308.09713", "2310.08528"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_5119"} +{"question": "Could you provide me some studies of masked image modeling in the self-supervised learning?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles", "Learning Features by Watching Objects Move", "Unsupervised Representation Learning by Predicting Image Rotations", "Representation Learning with Contrastive Predictive Coding", "Momentum Contrast for Unsupervised Visual Representation Learning", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["1505.05192", "1603.09246v3", "1612.06370", "1803.07728", "1807.03748", "1911.05722", "2111.06377"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_5120"} +{"question": "What research have used nearest neighbor entropy estimation methods to improve performance in visual domains?", "answer": ["Reinforcement Learning with Prototypical Representations", "Behavior From the Void: Unsupervised Active Pre-Training"], "answer_arxiv_id": ["2102.11271v2", "2103.04551"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_5121"} +{"question": "Could you cite some studies about approximate unlearning?", "answer": ["Certified Data Removal from Machine Learning Models", "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks", "Forgetting Outside the Box: Scrubbing Deep Networks of Information Accessible from Input-Output Observations", "Remember What You Want to Forget: Algorithms for Machine Unlearning", "Knowledge Removal in Sampling-based Bayesian Inference", "Certified Graph Unlearning"], "answer_arxiv_id": ["1911.03030", "1911.04933", "2003.02960", "2103.03279v2", "2203.12964", "2206.09140v2"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_5122"} +{"question": "Could you name some papers dealing with offline RL?", "answer": ["CORL: Research-oriented Deep Offline Reinforcement Learning Library"], "answer_arxiv_id": ["2210.07105"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_5123"} +{"question": "Which works involve aligning LLMs with human preferences?", "answer": ["Concrete Problems in AI Safety", "Fine-Tuning Language Models from Human Preferences", "Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasets", "A General Language Assistant as a Laboratory for Alignment", "Unsolved Problems in ML Safety", "Constitutional AI: Harmlessness from AI Feedback", "Improving alignment of dialogue agents via targeted human judgements", "The Capacity for Moral Self-Correction in Large Language Models", "Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned"], "answer_arxiv_id": ["1606.06565", "1909.08593", "2106.10328", "2112.00861", "2109.13916", "2212.08073", "2209.14375", "2302.07459", "2209.07858"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_5124"} +{"question": "Could you provide some references about combining blurring and noising techniques to extend diffusion processes?", "answer": ["Blurring Diffusion Models", "Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise"], "answer_arxiv_id": ["2209.05557", "2208.09392"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_5125"} +{"question": "What research works intended to answer logical queries on knowledge graphs by constructing box embeddings?", "answer": ["Query2box: Reasoning over Knowledge Graphs in Vector Space using Box\n Embeddings"], "answer_arxiv_id": ["2002.05969"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5126"} +{"question": "What studies are about scalable Graph Neural Networks (GNNs)?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Inductive Representation Learning on Large Graphs", "FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling", "Simplifying Graph Convolutional Networks", "GraphSAINT: Graph Sampling Based Inductive Learning Method", "Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks", "DeeperGCN: All You Need to Train Deeper GCNs", "Training Graph Neural Networks with 1000 Layers", "Towards Deeper Graph Neural Networks", "DropEdge: Towards Deep Graph Convolutional Networks on Node Classification", "PairNorm: Tackling Oversmoothing in GNNs", "Scaling Graph Neural Networks with Approximate PageRank", "SIGN: Scalable Inception Graph Neural Networks", "Rethinking and Scaling Up Graph Contrastive Learning: An Extremely Efficient Approach with Group Discrimination", "Sketch-GNN: Scalable Graph Neural Networks with Sublinear Training Complexity", "NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification", "Recipe for a General, Powerful, Scalable Graph Transformer"], "answer_arxiv_id": ["1609.02907", "1706.02216", "1801.10247", "1902.07153", "1907.04931v4", "1905.07953", "2006.07739v1", "2106.07476", "2007.09296", "1907.10903", "1909.12223", "2007.01570", "2004.11198", "2206.01535", "2406.15575v1", "2306.08385", "2205.12454"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_5127"} +{"question": "What publications propose predictive stochastic process models that directly learn mappings from context such as attentive Neural Processes (anps), Convolutional Neural Processes (convnps), Gaussian Neural Processes (gnps)?", "answer": ["Attentive Neural Processes", "Meta-Learning Stationary Stochastic Process Prediction with Convolutional Neural Processes", "The Gaussian Neural Process"], "answer_arxiv_id": ["1901.05761", "2007.01332", "2101.03606"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5128"} +{"question": "What are examples of bigger and more powerful models developed after the original GAN?", "answer": ["Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Progressive Growing of GANs for Improved Quality, Stability, and\n Variation", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["1809.11096", "1710.10196", "1812.04948", "1912.04958"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_5129"} +{"question": "Can you list some papers that have proposed GNN pretraining methods that adopt contrastive learning?", "answer": ["Contrastive Multi-View Representation Learning on Graphs", "GCC: Graph Contrastive Coding for Graph Neural Network Pre-Training", "Deep Graph Contrastive Representation Learning", "Graph Contrastive Learning with Adaptive Augmentation", "Graph Contrastive Learning Automated", "When Does Self-Supervision Help Graph Convolutional Networks?", "Self-supervised Learning on Graphs: Deep Insights and New Direction", "Geometric Graph Representation Learning via Maximizing Rate Reduction", "Self-supervised Training of Graph Convolutional Networks"], "answer_arxiv_id": ["2006.05582", "2006.09963", "2006.04131", "2010.14945", "2106.07594", "2006.09136", "2006.10141v1", "2202.06241", "2006.02380"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5130"} +{"question": "What papers have contributed to the improvement of decision transformers (DTs) and what areas have they improved on?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Generalized Decision Transformer for Offline Hindsight Information Matching", "Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Can Wikipedia Help Offline Reinforcement Learning?", "Multi-Game Decision Transformers", "A Generalist Agent", "Multi-Agent Reinforcement Learning is A Sequence Modeling Problem", "Online Decision Transformer"], "answer_arxiv_id": ["2106.01345", "2111.10364", "2106.02039", "2201.12122", "2205.15241", "2205.06175", "2205.14953", "2202.05607"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_5131"} +{"question": "Which research introduced the concept of Federated learning and proposed the FedAvg algorithm?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized\n Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20220110"}, "qid": "AutoScholarQuery_train_5132"} +{"question": "Which study elaborated a symplectic numerical integrator for manifold constraints and provided a rigorous proof of reversibility?", "answer": ["Multiple projection Markov Chain Monte Carlo algorithms on submanifolds"], "answer_arxiv_id": ["2003.09402"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_5133"} +{"question": "What papers proposed models like FastGCN and SGC to reduce the computational cost of GNNs?", "answer": ["FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling"], "answer_arxiv_id": ["1801.10247"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_5134"} +{"question": "Which study considered the problem where a deep ReLU net approximates an L-infinity-Sobolev function and provides a theorem based on covering number arguments?", "answer": ["Error bounds for approximations with deep ReLU networks"], "answer_arxiv_id": ["1610.01145"], "source_meta": {"published_time": "20200819"}, "qid": "AutoScholarQuery_train_5135"} +{"question": "What are some works that have mentioned or studied the strong in-context learning ability in the context of augmenting LLMs with external tools?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_5136"} +{"question": "Are there any works about the limitations of these API-based applications in capturing fine-grained visual details and understanding complex visual contexts?", "answer": ["Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "MM-ReAct : Prompting ChatGPT for Multimodal Reasoning and Action", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face", "VideoChat : Chat-Centric Video Understanding"], "answer_arxiv_id": ["2303.04671", "2303.11381", "2303.17580", "2305.06355"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_5137"} +{"question": "Which works propose new models for rationalization based on causal interpretations?", "answer": ["Desiderata for Representation Learning: A Causal Perspective"], "answer_arxiv_id": ["2109.03795v2"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_5138"} +{"question": "Are there any works indicating that SDE approximation might fail if the learning rate is too large?", "answer": ["On the Validity of Modeling SGD with Stochastic Differential Equations (SDEs)", "On the Generalization Benefit of Noise in Stochastic Gradient Descent"], "answer_arxiv_id": ["2102.12470", "2006.15081"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_5139"} +{"question": "In what research, does the author further develop the approach of Tracking Any Point (TAP) by globally estimating cost volume decoded to correspondence position and occlusion?", "answer": ["TAP-Vid: A Benchmark for Tracking Any Point in a Video"], "answer_arxiv_id": ["2211.03726"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_5140"} +{"question": "Which works extended the success in open-vocabulary image recognition to object detection?", "answer": ["Contrastive Feature Masking Open-Vocabulary Vision Transformer", "CoDet: Co-Occurrence Guided Region-Word Alignment for Open-Vocabulary\n Object Detection", "DetCLIP: Dictionary-Enriched Visual-Concept Paralleled Pre-training for\n Open-world Detection", "DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via\n Word-Region Alignment", "EdaDet: Open-Vocabulary Object Detection Using Early Dense Alignment", "Betrayed by Captions: Joint Caption Grounding and Generation for Open\n Vocabulary Instance Segmentation", "Multi-Modal Classifiers for Open-Vocabulary Object Detection", "Unified Open-Vocabulary Dense Visual Prediction", "Open-Vocabulary Object Detection via Scene Graph Discovery", "Multi-Modal Classifiers for Open-Vocabulary Object Detection", "CORA: Adapting CLIP for Open-Vocabulary Detection with Region Prompting\n and Anchor Pre-Matching", "Object-Aware Distillation Pyramid for Open-Vocabulary Object Detection", "Aligning Bag of Regions for Open-Vocabulary Object Detection", "Region-Aware Pretraining for Open-Vocabulary Object Detection with\n Vision Transformers", "Learning to Detect and Segment for Open Vocabulary Object Detection", "FACTUAL: A Benchmark for Faithful and Consistent Textual Scene Graph\n Parsing"], "answer_arxiv_id": ["2309.00775", "2310.16667", "2209.09407", "2304.04514", "2309.01151", "2301.00805", "2306.05493", "2307.08238", "2307.03339", "2306.05493", "2303.13076", "2303.05892v1", "2302.13996", "2305.07011", "2212.12130", "2305.17497"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_5141"} +{"question": "What are the works in which self-supervised representations were utilized for pseudo-labels generation in unsupervised object detection/discovery?", "answer": ["TokenCut: Segmenting Objects in Images and Videos with Self-supervised Transformer and Normalized Cut", "Cut and Learn for Unsupervised Object Detection and Instance Segmentation"], "answer_arxiv_id": ["2209.00383", "2301.11320"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_5142"} +{"question": "Which research introduces a hierarchical feature alignment network for improving the accuracy of object segmentation?", "answer": ["Hierarchical Feature Alignment Network for Unsupervised Video Object\n Segmentation"], "answer_arxiv_id": ["2207.08485"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_5143"} +{"question": "What are some studies that implemented Recurrent models for video-based reasoning?", "answer": ["Every Moment Counts: Dense Detailed Labeling of Actions in Complex\n Videos"], "answer_arxiv_id": ["1507.05738"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_5144"} +{"question": "Which studies introduced execution-based evaluation with comprehensive test suites for code generation?", "answer": ["A Syntax-Guided Edit Decoder for Neural Program Repair", "Less Training, More Repairing Please: Revisiting Automated Program\n Repair via Zero-shot Learning"], "answer_arxiv_id": ["2106.08253", "2207.08281"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_5145"} +{"question": "Which papers discussed distributed linear bandits?", "answer": ["Differentially-Private Federated Linear Bandits", "Federated Linear Contextual Bandits", "Distributed Clustering of Linear Bandits in Peer to Peer Networks", "Learning in Distributed Contextual Linear Bandits Without Sharing the Context"], "answer_arxiv_id": ["2010.11425", "2110.14177", "1604.07706", "2206.04180"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_5146"} +{"question": "Which studies estimated the initial computations required for BERT training and found significant improvements?", "answer": ["Large Batch Optimization for Deep Learning: Training BERT in 76 minutes"], "answer_arxiv_id": ["1904.00962"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_5147"} +{"question": "Which papers exemplify data-driven approaches using motion capture datasets for training?", "answer": ["AMASS: Archive of Motion Capture as Surface Shapes"], "answer_arxiv_id": ["1904.03278"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_5148"} +{"question": "What papers applied cascaded diffusion models for 3D synthesis?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "Neural Wavelet-domain Diffusion for 3D Shape Generation"], "answer_arxiv_id": ["2211.10440", "2209.08725"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_5149"} +{"question": "In what paper was DPPs used in object detection and multi-label classification?", "answer": ["Learning Detection with Diverse Proposals"], "answer_arxiv_id": ["1704.03533"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_5150"} +{"question": "What works addressed the significance of initialization in the performance of dimensionality reduction methods?", "answer": ["Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMap, and PaCMAP for Data Visualization"], "answer_arxiv_id": ["2012.04456"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_5151"} +{"question": "Which studies involved in 3D object detection transforms the image features to the BEV space using depth estimation?", "answer": ["Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by\n Implicitly Unprojecting to 3D", "BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object\n Detection", "BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers", "DFA3D: 3D Deformable Attention For 2D-to-3D Feature Lifting"], "answer_arxiv_id": ["2008.05711", "2206.10092", "2203.17270", "2307.12972"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_5152"} +{"question": "Which paper trains a Transformer to interpret regular expressions considering binary outputs?", "answer": ["What Makes Instruction Learning Hard? An Investigation and a New\n Challenge in a Synthetic Environment"], "answer_arxiv_id": ["2204.09148"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_5153"} +{"question": "Which research papers discusses the Non-Markov process in DDIMs?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_5154"} +{"question": "What works are based on the encoder-only architecture for the infilling task?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "1907.11692"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_5155"} +{"question": "Are there any studies addressing the issue of updating large pretrained models while preserving their transferability and the generalizability?", "answer": ["VisualGPT: Data-efficient Adaptation of Pretrained Language Models for\n Image Captioning", "Locating and Editing Factual Associations in GPT", "Editing Models with Task Arithmetic", "Language Models Meet World Models: Embodied Experiences Enhance Language\n Models", "Self-regulating Prompts: Foundational Model Adaptation without Forgetting", "Preventing Zero-Shot Transfer Degradation in Continual Learning of Vision-Language Models"], "answer_arxiv_id": ["2102.10407", "2202.05262", "2212.04089", "2305.10626", "2307.06948v2", "2303.06628v2"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_train_5156"} +{"question": "Which paper introduces a stochastic pruning process to token pruning for ViT?", "answer": ["AdaViT: Adaptive Tokens for Efficient Vision Transformer"], "answer_arxiv_id": ["2112.07658"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_5157"} +{"question": "What is the study that uses Landmarks Attention as a compression scheme to extend the context length of LLaMA-7B?", "answer": ["Landmark Attention: Random-Access Infinite Context Length for Transformers"], "answer_arxiv_id": ["2305.16300v2"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_5158"} +{"question": "Could you provide me some studies about feature attribution methods?", "answer": ["Axiomatic Attribution for Deep Networks", "SmoothGrad: removing noise by adding noise", "Improving performance of deep learning models with axiomatic attribution priors and expected gradients", "RISE: Randomized Input Sampling for Explanation of Black-box Models", "Explaining by Removing: A Unified Framework for Model Explanation"], "answer_arxiv_id": ["1703.01365", "1706.03825", "1906.10670", "1806.07421", "2011.14878"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5159"} +{"question": "What literature provides methods about using the implicit function theorem to find necessary gradients and backpropagate through the solver?", "answer": ["OptNet: Differentiable Optimization as a Layer in Neural Networks", "Differentiable Convex Optimization Layers", "On Differentiating Parameterized Argmin and Argmax Problems with Application to Bi-level Optimization", "Deep Declarative Networks: A New Hope"], "answer_arxiv_id": ["1703.00443", "1910.12430", "1607.05447", "1909.04866"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_5160"} +{"question": "Could you tell me a paper which proposed a scalable method for function space variational inference on deep neural networks?", "answer": ["Tractable Function-Space Variational Inference in Bayesian Neural Networks"], "answer_arxiv_id": ["2312.17199"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_5161"} +{"question": "Are there any research studies that focus on models that represent the process of reading long narratives?", "answer": ["TVShowGuess: Character Comprehension in Stories as Speaker Guessing", "Personality Understanding of Fictional Characters during Book Reading", "RELIC: Retrieving Evidence for Literary Claims", "Plot Retrieval as an Assessment of Abstract Semantic Association"], "answer_arxiv_id": ["2204.07721", "2305.10156", "2203.10053", "2311.01666"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_5162"} +{"question": "Could you provide me some works that employed supervised training with human-annotated sentence and bounding box supervision for spatio-temporal grounding?", "answer": ["Where Does It Exist: Spatio-Temporal Video Grounding for Multi-Form\n Sentences", "Weakly-Supervised Spatio-Temporally Grounding Natural Sentence in Video"], "answer_arxiv_id": ["2001.06891", "1906.02549"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_5163"} +{"question": "Could you provide me some references about embedding differentiable optimization problems in DNNs?", "answer": ["OptNet: Differentiable Optimization as a Layer in Neural Networks", "Differentiable Convex Optimization Layers", "Homogeneous Linear Inequality Constraints for Neural Network Activations"], "answer_arxiv_id": ["1703.00443", "1910.12430", "1902.01785"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_5164"} +{"question": "What work has been done on modeling the separation of objects and background in complex scenes using only images in training?", "answer": ["Unsupervised Discovery of Object Radiance Fields"], "answer_arxiv_id": ["2107.07905"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_5165"} +{"question": "What works used similar objectives as the loss objective used in this study for mutual information estimation and statistical learning?", "answer": ["On Variational Bounds of Mutual Information"], "answer_arxiv_id": ["1905.06922"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_5166"} +{"question": "What studies report on introducing Vision Transformer into the design of lightweight vision backbones?", "answer": ["LightViT: Towards Light-Weight Convolution-Free Vision Transformers", "EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers", "EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications", "MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer", "Mobile-Former: Bridging MobileNet and Transformer"], "answer_arxiv_id": ["2207.05557", "2205.03436", "2206.10589", "2110.02178", "2108.05895"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5167"} +{"question": "Can you list the studies that present the concentration bounds for DRM, CE, and RDEU?", "answer": ["A Wasserstein distance approach for concentration of empirical risk estimates"], "answer_arxiv_id": ["1902.10709"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_5168"} +{"question": "Which papers have contributed to the development of Multimodal Large Language Models (MLLMs) with focus on image modality?", "answer": ["PaLM-E: An Embodied Multimodal Language Model", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "ImageBind: One Embedding Space To Bind Them All", "Meta-Transformer: A Unified Framework for Multimodal Learning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2303.03378", "2303.16199", "2304.15010", "2305.05665v2", "2307.10802", "2304.10592"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_5169"} +{"question": "What papers show the use of large-scale image-text pairwise datasets or arbitrarily interleaved visual and textual data to pre-train VLMs?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize\n Long-Tail Visual Concepts", "Microsoft COCO: Common Objects in Context", "Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with\n Text", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2102.05918", "2102.08981", "1405.0312", "2304.06939", "2204.14198"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_5170"} +{"question": "Could you provide me some works about deduction in the context of modern language models?", "answer": ["Maieutic Prompting: Logically Consistent Reasoning with Recursive Explanations", "Enhancing Self-Consistency and Performance of Pre-Trained Language Models through Natural Language Inference"], "answer_arxiv_id": ["2205.11822", "2211.11875"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_5171"} +{"question": "Which studies are about the use of conditioned diffusion models in text-to-image generation tasks?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2204.06125", "2205.11487"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5172"} +{"question": "What are the studies that used retrieval-augmentation in visual question answering or generative processes?", "answer": ["ReVeaL: Retrieval-Augmented Visual-Language Pre-Training with Multi-Source Multimodal Knowledge Memory", "Retrieval-augmented Image Captioning"], "answer_arxiv_id": ["2212.05221", "2302.08268"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5173"} +{"question": "Which research paper is responsible for the development of the MeshRIR dataset?", "answer": ["MeshRIR: A Dataset of Room Impulse Responses on Meshed Grid Points For\n Evaluating Sound Field Analysis and Synthesis Methods"], "answer_arxiv_id": ["2106.10801"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_5174"} +{"question": "Are there any studies that have used hypergraphs for task selection?", "answer": ["Taskonomy: Disentangling Task Transfer Learning"], "answer_arxiv_id": ["1804.08328"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_5175"} +{"question": "Which studies showcase the construction of a commonsense inferential rule base through crowdsourcing?", "answer": ["ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning"], "answer_arxiv_id": ["1811.00146"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_5176"} +{"question": "Could you provide me some studies that used Langevin dynamics in a diffusion model?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_5177"} +{"question": "Could you provide me some studies about employing diffusion models in tasks such as image editing?", "answer": ["EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations"], "answer_arxiv_id": ["2207.06635"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_5178"} +{"question": "Are there any papers that utilize neural networks to learn shape deformations and reconstruct geometry?", "answer": ["Geometry-Aware Network for Non-Rigid Shape Prediction from a Single View"], "answer_arxiv_id": ["1809.10305"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_5179"} +{"question": "Could you tell me about some papers that provided additional data for specific research tasks in dance generation such as PMSD, PhantomDance, and MMD?", "answer": ["Transflower: probabilistic autoregressive dance generation with\n multimodal attention", "DanceFormer: Music Conditioned 3D Dance Generation with Parametric\n Motion Transformer"], "answer_arxiv_id": ["2106.13871", "2103.10206"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_5180"} +{"question": "What are the examples of studies on implicit 3D representations?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Differentiable Volumetric Rendering: Learning Implicit 3D\n Representations without 3D Supervision"], "answer_arxiv_id": ["2106.10689", "1912.07372"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5181"} +{"question": "Which studies have proposed methods for increasing the distance between the decision boundary and data points?", "answer": ["Large Margin Deep Networks for Classification", "Controlling Neural Level Sets", "MMA Training: Direct Input Space Margin Maximization through Adversarial Training"], "answer_arxiv_id": ["1803.05598", "1905.11911", "1812.02637"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5182"} +{"question": "Could you provide me some studies about online and offline IL?", "answer": ["DITTO: Offline Imitation Learning with World Models", "Strictly Batch Imitation Learning by Energy-based Distribution Matching", "Curriculum Offline Imitating Learning", "Discriminator-Guided Model-Based Offline Imitation Learning"], "answer_arxiv_id": ["2302.03086", "2006.14154", "2111.02056", "2207.00244"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_5183"} +{"question": "Which study presents a theoretical comparison of positive-unlabeled learning against positive-negative learning?", "answer": ["Theoretical Comparisons of Positive-Unlabeled Learning against Positive-Negative Learning"], "answer_arxiv_id": ["1603.03130"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_5184"} +{"question": "Could you provide me some researches that concentrated on enhancing the reward model by eliminating Markovian assumptions?", "answer": ["Non-Markovian Reward Modelling from Trajectory Labels via Interpretable Multiple Instance Learning", "Preference Transformer: Modeling Human Preferences using Transformers for RL"], "answer_arxiv_id": ["2205.15367", "2303.00957"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_5185"} +{"question": "Which papers discuss about model personalization approaches in federated learning?", "answer": ["Federated Multi-Task Learning", "Federated Optimization in Heterogeneous Networks", "Federated Learning with Personalization Layers", "Exploiting Shared Representations for Personalized Federated Learning", "Federated Learning with Partial Model Personalization"], "answer_arxiv_id": ["1705.10467", "1812.06127", "1912.00818v1", "2102.07078", "2204.03809"], "source_meta": {"published_time": "20230911"}, "qid": "AutoScholarQuery_train_5186"} +{"question": "Which studies have discussed the use of Learning to Optimize (L2O) for black-box and Bayesian problems?", "answer": ["Learning to Learn without Gradient Descentby Gradient Descent", "Learning to Optimize in Swarms"], "answer_arxiv_id": ["1611.03824", "1911.03787"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_5187"} +{"question": "Which works are part of the Generic Face Image Quality Assessment (GFIQA) category?", "answer": ["A Survey of Deep Face Restoration: Denoise, Super-Resolution, Deblur,\n Artifact Removal", "Going the Extra Mile in Face Image Quality Assessment: A Novel Database\n and Model", "An Image Quality Assessment Dataset for Portraits", "Going the Extra Mile in Face Image Quality Assessment: A Novel Database\n and Model", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["2211.02831", "2207.04904", "2304.05772", "2207.04904", "1812.04948"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_5188"} +{"question": "Could you provide me some works about using model prediction error or 'Curiosity' as intrinsic rewards?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction"], "answer_arxiv_id": ["1705.05363"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_5189"} +{"question": "Which works establish the AHN and its modern variants?", "answer": ["Long Sequence Hopfield Memory"], "answer_arxiv_id": ["2306.04532"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_5190"} +{"question": "What studies propose using Neural Relational Inference(NRI) for learning and predicting interactions?", "answer": ["Neural Relational Inference for Interacting Systems"], "answer_arxiv_id": ["1802.04687"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5191"} +{"question": "What work suggests the usage of DSVAE with a similar LSTM architecture?", "answer": ["Disentangled Sequential Autoencoder"], "answer_arxiv_id": ["1803.02991"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5192"} +{"question": "Could you give me examples of research where they jointly learned an optimized graph structure and corresponding graph representations?", "answer": ["Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings", "Graph Structure Learning for Robust Graph Neural Networks", "Learning to Drop: Robust Graph Neural Network via Topological Denoising"], "answer_arxiv_id": ["2006.13009", "2005.10203", "2011.07057"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_5193"} +{"question": "Which works studied the optimal strategies for finite horizons in repeated games?", "answer": ["Efficient Stackelberg Strategies for Finitely Repeated Games"], "answer_arxiv_id": ["2207.04192"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_5194"} +{"question": "What are some papers about learning inverse dynamics models, particularly in the field of robotics?", "answer": ["Transfer from Simulation to Real World through Learning Deep Inverse Dynamics Model"], "answer_arxiv_id": ["1610.03518"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_5195"} +{"question": "Which papers study adversarial attacks influencing test-time performance of reinforcement learning agents trained in self-play?", "answer": ["Adversarial Policies: Attacking Deep Reinforcement Learning"], "answer_arxiv_id": ["1905.10615"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_5196"} +{"question": "What are the studies that focus on the statistical trade-off of estimating privacy parameters while dealing with a query with discrete output in a finite space?", "answer": ["Property Testing for Differential Privacy"], "answer_arxiv_id": ["1806.06427"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5197"} +{"question": "Which studies induce word alignment from the contextualized word embeddings of mPLMs pre-trained on non-parallel data?", "answer": ["SimAlign: High Quality Word Alignments without Parallel Training Data\n using Static and Contextualized Embeddings"], "answer_arxiv_id": ["2004.08728"], "source_meta": {"published_time": "20240716"}, "qid": "AutoScholarQuery_train_5198"} +{"question": "Is there any study that proves that MPNNs cannot solve the biconnectivity problem?", "answer": ["Rethinking the Expressive Power of GNNs via Graph Biconnectivity"], "answer_arxiv_id": ["2301.09505"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_5199"} +{"question": "What works are related to acceleration methods?", "answer": ["Acceleration Methods", "Templates for Convex Cone Problems with Applications to Sparse Signal Recovery", "Efficient online algorithms for fast-rate regret bounds under sparsity"], "answer_arxiv_id": ["2101.09545", "1009.2065", "1805.09174"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_5200"} +{"question": "What works utilize a grid or voxel-like data structures to precompute and cache radiance values?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "Baking Neural Radiance Fields for Real-Time View Synthesis", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2112.05131", "2103.14645", "2103.14024", "2201.05989"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_5201"} +{"question": "Can you cite studies that have addressed the task of dense video captioning?", "answer": ["End-to-end Dense Video Captioning as Sequence Generation", "End-to-End Dense Video Captioning with Parallel Decoding", "Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense\n Video Captioning"], "answer_arxiv_id": ["2204.08121", "2108.07781", "2302.14115"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_5202"} +{"question": "Which papers concentrate on the study of CVaR MDP where the objective is to minimize the CVaR of the total cost?", "answer": ["Risk-Sensitive and Robust Decision-Making: a CVaR Optimization Approach"], "answer_arxiv_id": ["1506.02188"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_5203"} +{"question": "What are some regularization-based methods proposed to address catastrophic forgetting in continual learning?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Memory Aware Synapses: Learning what (not) to forget", "Continual Learning with Recursive Gradient Optimization"], "answer_arxiv_id": ["1612.00796", "1711.09601", "2201.12522"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_5204"} +{"question": "What works have been inspired by BERT and included masked-language-modeling, image-text-matching, or mask-region-modeling objectives in the domain of vision and language pre-training?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "VisualBERT: A Simple and Performant Baseline for Vision and Language"], "answer_arxiv_id": ["1908.02265", "1908.07490", "1908.03557"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_5205"} +{"question": "Can you provide some studies that investigated the goal-exploring algorithms?", "answer": ["Discovering and Achieving Goals via World Models", "Skew-Fit: State-Covering Self-Supervised Reinforcement Learning"], "answer_arxiv_id": ["2110.09514", "1903.03698"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_5206"} +{"question": "What works first attempted image generation through a combination of gradient-based optimization and image priors?", "answer": ["CLIPDraw: Exploring Text-to-Drawing Synthesis through Language-Image\n Encoders", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "VQGAN-CLIP: Open Domain Image Generation and Editing with Natural\n Language Guidance"], "answer_arxiv_id": ["2106.14843", "2103.17249", "2204.08583"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_5207"} +{"question": "Which study separates the state space into internal state and external state, and attempts to maximize the mutual information between them in the context of empowerment-based exploration?", "answer": ["Mutual Information State Intrinsic Control"], "answer_arxiv_id": ["2103.08107"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_5208"} +{"question": "What papers have studied the Adam algorithm which is related to clipped SGD?", "answer": ["P", "Adam Can Converge Without Any Modification On Update Rules"], "answer_arxiv_id": ["0704.0320", "2208.09632v5"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_5209"} +{"question": "Any works about how the appointment of dropout can improve the generalization ability of SR networks?", "answer": ["Reflash Dropout in Image Super-Resolution"], "answer_arxiv_id": ["2112.12089"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5210"} +{"question": "Which study established an equivalence between online learnability and differentially private PAC learnability?", "answer": ["An Equivalence Between Private Classification and Online Prediction", "Private PAC learning implies finite Littlestone dimension"], "answer_arxiv_id": ["2003.00563", "1806.00949"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_5211"} +{"question": "How has the research line of robust overfitting been developed?", "answer": ["Overfitting in adversarially robust deep learning", "Label Noise in Adversarial Training: A Novel Perspective to Study Robust Overfitting"], "answer_arxiv_id": ["2002.11569", "2110.03135"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_5212"} +{"question": "Which papers proposed Eulerian methods for motion magnification without explicit motion estimation?", "answer": ["Video Acceleration Magnification"], "answer_arxiv_id": ["1704.04186"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_5213"} +{"question": "Any research applying Tukey depth to private estimation in identifying high-depth points?", "answer": ["Private Center Points and Learning of Halfspaces", "How to Find a Point in the Convex Hull Privately", "Robust and differentially private mean estimation", "Differentially private depth functions and their associated medians"], "answer_arxiv_id": ["1902.10731", "2003.13192", "2102.09159", "2101.02800"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_5214"} +{"question": "What studies assumed disentanglement constraints and invariance in white-box explanations?", "answer": ["Interpretable Convolutional Neural Networks", "Interpretable Compositional Convolutional Neural Networks", "Training Interpretable Convolutional Neural Networks by Differentiating Class-specific Filters", "Concept Whitening for Interpretable Image Recognition", "Self-Interpretable Model with Transformation Equivariant Interpretation"], "answer_arxiv_id": ["1710.00935", "2107.04474", "2007.08194", "2002.01650", "2111.04927"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_5215"} +{"question": "Could you provide me with studies that use prompt engineering to elicit or improve reasoning in LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2305.10601"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_5216"} +{"question": "Could you list the papers that has observed RNNs and LSTMs requiring exponential memory for learning more advanced languages?", "answer": ["Evaluating the Ability of LSTMs to Learn Context-Free Grammars"], "answer_arxiv_id": ["1811.02611"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_5217"} +{"question": "Could you provide me some works about weight editing methods?", "answer": ["Locating and Editing Factual Associations in GPT", "Editing Models with Task Arithmetic", "Editing Implicit Assumptions in Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2202.05262", "2212.04089", "2303.08084"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_5218"} +{"question": "Which work similarly retrieves from premises accessible to the current file in the context of Python, like the proposed ReProver does?", "answer": ["CoCoMIC: Code Completion By Jointly Modeling In-file and Cross-file Context"], "answer_arxiv_id": ["2212.10007"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_5219"} +{"question": "What are the key works that discussed the use and advantages of Latent Diffusion Models (LDMs)?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_5220"} +{"question": "What works ventured to incorporate diffusion models across a broader array of tasks?", "answer": ["Image Super-Resolution via Iterative Refinement", "Motion-Conditioned Diffusion Model for Controllable Video Synthesis", "Gligen: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2104.07636", "2304.14404", "2301.07093"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_5221"} +{"question": "Which works utilized the simple decoder head of Mask2Former in 3D instance segmentation?", "answer": ["Mask3D: Mask Transformer for 3D Semantic Instance Segmentation"], "answer_arxiv_id": ["2210.03105"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_5222"} +{"question": "Can you name a research that stated that when the width of the neural network tends to infinity, it is equivalent to a neural tangent kernel?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_5223"} +{"question": "What works introduced mask classification in semantic segmentation?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2107.06278", "2112.01527"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_5224"} +{"question": "Could you provide me some studies about theoretically guaranteed BO methods?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design"], "answer_arxiv_id": ["0912.3995v4"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_5225"} +{"question": "Could you provide me with works that investigate bridging image and text models?", "answer": ["Linearly Mapping from Image to Text Space", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "Multimodal Few-Shot Learning with Frozen Language Models"], "answer_arxiv_id": ["2209.15162", "2111.07991", "2106.13884"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_5226"} +{"question": "Which study proposes adaptive fusion of ERP and TP predictions, achieving SOTA performance in bi-projection inputs?", "answer": ["HRDFuse: Monocular 360{\\deg}Depth Estimation by Collaboratively Learning\n Holistic-with-Regional Depth Distributions"], "answer_arxiv_id": ["2303.11616"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_5227"} +{"question": "Which papers address the distributional shift issue in offline RL through modifications to policy evaluation?", "answer": ["Bridging the Gap Between Value and Policy Based Reinforcement Learning", "Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning", "Mutual Information Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["1702.08892", "2006.04779", "2110.06169", "2210.07484"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5228"} +{"question": "What papers recently address point counterfactual identification through neural methods?", "answer": ["Deep Structural Causal Models for Tractable Counterfactual Inference", "Counterfactual Generative Networks", "Diffusion Causal Models for Counterfactual Estimation", "Counterfactual Fairness with Disentangled Causal Effect Variational Autoencoder", "Evaluating and Mitigating Bias in Image Classifiers: A Causal Perspective Using Counterfactuals", "Deep Counterfactual Estimation with Categorical Background Variables", "Weakly Supervised Disentangled Generative Causal Representation Learning", "Variational Causal Inference", "Interventional and Counterfactual Inference with Diffusion Models", "Predicting Cellular Responses with Variational Causal Inference and Refined Relational Information"], "answer_arxiv_id": ["2006.06485", "2101.06046", "2202.10166", "2011.11878", "2009.08270", "2210.05811", "2010.02637", "2209.05935", "2302.00860", "2210.00116"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_5229"} +{"question": "Which works proposed diffusion models for the purpose of speeding up inference?", "answer": ["Denoising Diffusion Implicit Models", "Progressive Distillation for Fast Sampling of Diffusion Models", "Score-Based Generative Modeling with Critically-Damped Langevin Diffusion", "Tackling the Generative Learning Trilemma with Denoising Diffusion GANs", "Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality", "High-Resolution Image Synthesis with Latent Diffusion Models", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations"], "answer_arxiv_id": ["2010.02502", "2202.00512", "2112.07068", "2112.07804", "2202.05830", "2112.10752", "2108.01073"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_5230"} +{"question": "Could you provide me some studies about generating prompts to adapt the domain space for effective continual learning?", "answer": ["CODA-Prompt: COntinual Decomposed Attention-based Prompting for\n Rehearsal-Free Continual Learning", "When Prompt-based Incremental Learning Does Not Meet Strong Pretraining"], "answer_arxiv_id": ["2211.13218", "2308.10445"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_5231"} +{"question": "Could you tell me any research on the implicit bias in two-layer leaky ReLU networks trained on linearly separable and symmetric data?", "answer": ["Gradient Descent on Two-layer Nets: Margin Maximization and Simplicity Bias"], "answer_arxiv_id": ["2110.13905"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_5232"} +{"question": "Are there any studies developing a neural diving approach for finding fast MILP solutions?", "answer": ["Solving Mixed Integer Programs Using Neural Networks"], "answer_arxiv_id": ["2012.13349"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_5233"} +{"question": "What works have presented latent ordinary differential equations (ODEs) for temporal data?", "answer": ["Neural Ordinary Differential Equations", "Hamiltonian Generative Networks"], "answer_arxiv_id": ["1806.07366", "1909.13789"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_5234"} +{"question": "Who have proposed different optimization strategies in the context of PINNs?", "answer": ["PSO-PINN: Physics-Informed Neural Networks Trained with Particle Swarm Optimization"], "answer_arxiv_id": ["2202.01943"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_5235"} +{"question": "Which studies incorporate text latent vectors as conditions in text-to-image diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_5236"} +{"question": "Could you provide me some studies on reinforcement learning (RL) and evolution strategies (ES) in neural networks?", "answer": ["Evolution Strategies as a Scalable Alternative to Reinforcement Learning"], "answer_arxiv_id": ["1703.03864v2"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_5237"} +{"question": "Which work integrates information from color and frequency domains to detect manipulated face images and videos?", "answer": ["Two-branch Recurrent Network for Isolating Deepfakes in Videos"], "answer_arxiv_id": ["2008.03412"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_5238"} +{"question": "Which benchmarks have been used to assess the long context processing ability of large language models?", "answer": ["LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding"], "answer_arxiv_id": ["2308.14508v2"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_5239"} +{"question": "What papers are associated with the development of Soft or Posterior Policy iteration?", "answer": ["Taming the Noise in Reinforcement Learning via Soft Updates", "Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor", "Inferring Smooth Control: Monte Carlo Posterior Policy Iteration with Gaussian Processes"], "answer_arxiv_id": ["1512.08562", "1801.01290", "2210.03512"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5240"} +{"question": "Which works focused on the process of eliminating reward engineering in the field of inverse reinforcement learning?", "answer": ["A Survey of Inverse Reinforcement Learning: Challenges, Methods and Progress"], "answer_arxiv_id": ["1806.06877"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_5241"} +{"question": "Any notable research papers demonstrating the use of Large Language Models in generating 3D scene layouts?", "answer": ["LayoutGPT: Compositional Visual Planning and Generation with Large\n Language Models", "Towards Language-guided Interactive 3D Generation: LLMs as Layout Interpreter with Generative Feedback"], "answer_arxiv_id": ["2305.15393", "2305.15808v1"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_5242"} +{"question": "Who first introduced graphical conditioners to the autoregressive flows architecture through input masking?", "answer": ["Graphical Normalizing Flows"], "answer_arxiv_id": ["2006.02548v3"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_5243"} +{"question": "What work used the flow to guide the generation of canonical images?", "answer": ["CoDeF: Content Deformation Fields for Temporally Consistent Video\n Processing"], "answer_arxiv_id": ["2308.07926"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_5244"} +{"question": "Could you provide me some studies that encountered difficulties in training PINNs?", "answer": ["Characterizing possible failure modes in physics-informed neural networks"], "answer_arxiv_id": ["2109.01050"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_5245"} +{"question": "What research proposed a multi-agent adaptation of the MuJoCo environment?", "answer": ["FACMAC: Factored Multi-Agent Centralised Policy Gradients"], "answer_arxiv_id": ["2003.06709"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_5246"} +{"question": "Which work made use of the SC framework to estimate treatment effects in adaptive experimental design while minimizing regret associated with experimentation?", "answer": ["Synthetically Controlled Bandits"], "answer_arxiv_id": ["2202.07079"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_5247"} +{"question": "Which works propose to use canonicalization in an autoencoding setup?", "answer": ["Unsupervised Learning of Group Invariant and Equivariant Representations"], "answer_arxiv_id": ["2202.07559"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_5248"} +{"question": "Can you provide the studies that focus on variable importance methods?", "answer": ["All Models are Wrong, but Many are Useful: Learning a Variable’s Importance by Studying an Entire Class of Prediction Models Simultaneously"], "answer_arxiv_id": ["1801.01489"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_5249"} +{"question": "What works propose dataset and model which achieve high quality semantic segmentation on many labels?", "answer": ["High-Quality Entity Segmentation"], "answer_arxiv_id": ["2211.05776"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_5250"} +{"question": "Could you provide me some citations that deal with arbitrary-scale super resolution methods?", "answer": ["Meta-SR: A Magnification-Arbitrary Network for Super-Resolution", "Learning Continuous Image Representation with Local Implicit Image\n Function", "Local Texture Estimator for Implicit Representation Function", "Cascaded Local Implicit Transformer for Arbitrary-Scale Super-Resolution", "CiaoSR: Continuous Implicit Attention-in-Attention Network for\n Arbitrary-Scale Image Super-Resolution", "Implicit Transformer Network for Screen Content Image Continuous\n Super-Resolution", "UltraSR: Spatial Encoding is a Missing Key for Implicit Image\n Function-based Arbitrary-Scale Super-Resolution", "Local Implicit Normalizing Flow for Arbitrary-Scale Image\n Super-Resolution"], "answer_arxiv_id": ["1903.00875", "2012.09161", "2111.08918", "2303.16513", "2212.04362", "2112.06174", "2103.12716", "2303.05156"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_5251"} +{"question": "Which works consider example-level DP using the subsampled Gaussian mechanism to ensure DP?", "answer": ["Rényi Differential Privacy of the Sampled Gaussian Mechanism"], "answer_arxiv_id": ["1908.10530"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_5252"} +{"question": "Which paper first introduced the Test Time Adaptation (TTA) paradigm?", "answer": ["A Comprehensive Survey on Test-Time Adaptation under Distribution Shifts"], "answer_arxiv_id": ["2303.15361"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_5253"} +{"question": "What researches focused on the problem of generating videos in the task of video prediction?", "answer": ["Video (Language) Modeling: A Baseline For Generative Models of Natural Videos", "Deep Predictive Coding Networks for Video Prediction and Unsupervised Learning"], "answer_arxiv_id": ["1412.6604", "1605.08104"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_5254"} +{"question": "Which works use a single example as the style reference in face stylization?", "answer": ["JoJoGAN: One Shot Face Stylization", "Mind the Gap: Domain Gap Control for Single Shot Domain Adaptation for\n Generative Adversarial Networks", "One-Shot Adaptation of GAN in Just One CLIP", "Towards Diverse and Faithful One-shot Adaption of Generative Adversarial\n Networks"], "answer_arxiv_id": ["2112.11641", "2110.08398", "2203.09301", "2207.08736"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_5255"} +{"question": "Which papers contain information about conditional diffusion models, similar to cGAN and cVAE?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_5256"} +{"question": "What studies have been performed on enhancing and classifying pairwise features via a spatial-temporal transformer for dynamic scene graph generation?", "answer": ["Spatial-Temporal Transformer for Dynamic Scene Graph Generation"], "answer_arxiv_id": ["2107.12309"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_5257"} +{"question": "Which works explore the large scale of recent deep models?", "answer": ["PaLM: Scaling Language Modeling with Pathways", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2204.02311", "2005.14165"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_5258"} +{"question": "What studies have proposed hybrid normalization approaches that utilize running statistics along both minibatch and spatial dimensions?", "answer": ["Differentiable Learning-to-Normalize via Switchable Normalization", "Continual Normalization: Rethinking Batch Normalization for Online Continual Learning"], "answer_arxiv_id": ["1806.10779", "2203.16102"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_5259"} +{"question": "Which papers introduced Generative Adversarial Networks (GANs) for text-based image generation?", "answer": ["Generative Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["2203.00667", "1809.11096", "1812.04948"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5260"} +{"question": "What papers utilized interval bound propagation (IBP) to certify the ensemble?", "answer": ["Enhancing Certifiable Robustness via a Deep Model Ensemble"], "answer_arxiv_id": ["1910.14655v1"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_5261"} +{"question": "Are there any studies that applied LMs for error correction tasks in ASR?", "answer": ["FastCorrect: Fast Error Correction with Edit Alignment for Automatic Speech Recognition", "ASR ERROR CORRECTION and DOMAIN ADAPTATION USING MACHINE TRANSLATION", "Automatic Spelling Correction with Transformer for CTC-based End-to-End Speech Recognition", "Error Correction in ASR using Sequence-to-Sequence Models", "BART based semantic correction for Mandarin automatic speech recognition system", "Mask the Correct Tokens: An Embarrassingly Simple Approach for Error Correction"], "answer_arxiv_id": ["2105.03842", "2003.07692", "1904.10045", "2202.01157", "2104.05507", "2211.13252"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_5262"} +{"question": "Could you give examples of research that deals with test-time domain adaptation?", "answer": ["Test-Time Training with Self-Supervision for Generalization under\n Distribution Shifts", "Tent: Fully Test-time Adaptation by Entropy Minimization", "Do We Really Need to Access the Source Data? Source Hypothesis Transfer\n for Unsupervised Domain Adaptation", "Source Data-absent Unsupervised Domain Adaptation through Hypothesis\n Transfer and Labeling Transfer", "Parameter-free Online Test-time Adaptation", "Test-Time Training with Masked Autoencoders", "Contrastive Test-Time Adaptation", "TTTFlow: Unsupervised Test-Time Training with Normalizing Flow"], "answer_arxiv_id": ["1909.13231", "2006.10726", "2002.08546", "2012.07297", "2201.05718", "2209.07522", "2204.10377", "2210.11389"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_5263"} +{"question": "What are the works on jammers in the Multi-player Multi-Armed Bandit problem?", "answer": ["Learning to Coordinate in a Decentralized Cognitive Radio Network in Presence of Jammers"], "answer_arxiv_id": ["1803.06810"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5264"} +{"question": "Which researchers have leveraged neural networks, deep learning, and reinforcement learning to create more sophisticated and adaptive AI agents in the field of economic research?", "answer": ["Building a Foundation for Data-Driven, Interpretable, and Robust Policy\n Design using the AI Economist"], "answer_arxiv_id": ["2108.02904"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_5265"} +{"question": "Are there any studies using known dynamics models to facilitate exploration using prediction error, surprise, or information gain?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Curiosity-driven Exploration by Self-supervised Prediction", "Exploration by Random Network Distillation", "Surprise-Based Intrinsic Motivation for Deep Reinforcement Learning", "Planning to Explore via Self-Supervised World Models"], "answer_arxiv_id": ["1507.00814", "1705.05363", "1810.12894", "1703.01732", "2005.05960"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_5266"} +{"question": "What works proposed metrics to measure image fidelity in AI-generated images?", "answer": ["Improved Techniques for Training GANs", "GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash\n Equilibrium"], "answer_arxiv_id": ["1606.03498", "1706.08500"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_5267"} +{"question": "Which works studied the definition of functions for one-hidden layer ReLU networks in the univariate case?", "answer": ["How do infinite width bounded norm networks look in function space?", "Convex Geometry and Duality of Over-parameterized Neural Networks"], "answer_arxiv_id": ["1902.05040", "2002.11219"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_5268"} +{"question": "Could you tell me about the studies that worked on weakly-supervised dense video captioning?", "answer": ["Weakly Supervised Dense Event Captioning in Videos", "Watch, Listen and Tell: Multi-modal Weakly Supervised Dense Event\n Captioning"], "answer_arxiv_id": ["1812.03849", "1909.09944"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_5269"} +{"question": "Which paper is related to K-FAC and whose main focus is its speed for large-batch size training?", "answer": ["Optimizing Neural Networks with Kronecker-factored Approximate Curvature"], "answer_arxiv_id": ["1503.05671"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_5270"} +{"question": "Which studies show the superior capacity of Transformer network, with ViT standing out in various downstream tasks?", "answer": ["End-to-End Object Detection with Transformers", "GroupViT: Semantic Segmentation Emerges from Text Supervision"], "answer_arxiv_id": ["2005.12872", "2202.11094"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_5271"} +{"question": "Can you list the studies that began to ground key frames or objects to improve VideoQA?", "answer": ["Invariant Grounding for Video Question Answering", "Equivariant and Invariant Grounding for Video Question Answering", "Discovering Spatio-Temporal Rationales for Video Question Answering", "Self-Chained Image-Language Model for Video Localization and Question\n Answering"], "answer_arxiv_id": ["2206.02349", "2207.12783", "2307.12058", "2305.06988"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_5272"} +{"question": "Who proposed the neural codec, called CDC, which uses a DDPM conditioned on a quantized latent representation?", "answer": ["Lossy Image Compression with Conditional Diffusion Models"], "answer_arxiv_id": ["2209.06950"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_5273"} +{"question": "Which research contributions make use of deep convolutional neural networks in image classification?", "answer": ["Predicting Depth, Surface Normals and Semantic Labels with a Common\n Multi-Scale Convolutional Architecture", "Designing Deep Networks for Surface Normal Estimation", "Marr Revisited: 2D-3D Alignment via Surface Normal Prediction"], "answer_arxiv_id": ["1411.4734", "1411.4958", "1604.01347"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_5274"} +{"question": "Which works applied Multiple Gossip Steps (MGS) into their algorithms to accelerate the training process or achieve fast convergence?", "answer": ["DeEPCA: Decentralized Exact PCA with Linear Convergence Rate", "Communication-Efficient Distributed Optimization in Networks with Gradient Tracking and Variance Reduction"], "answer_arxiv_id": ["2102.03990", "1909.05844v3"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_5275"} +{"question": "What studies demonstrated that attributes are separable in the latent space of generative models?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Interpreting the Latent Space of GANs for Semantic Face Editing"], "answer_arxiv_id": ["1812.04948", "1907.10786"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_5276"} +{"question": "What works propose functional causal models with additional assumptions about the data distribution in the field of causal structural learning?", "answer": ["CAM: Causal additive models, high-dimensional order search and penalized regression"], "answer_arxiv_id": ["1310.1533"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_5277"} +{"question": "In what studies employ density estimators for per-ray basis in 3D GANs?", "answer": ["AdaNeRF: Adaptive Sampling for Real-time Rendering of Neural Radiance\n Fields", "TermiNeRF: Ray Termination Prediction for Efficient Neural Rendering"], "answer_arxiv_id": ["2207.10312", "2111.03643"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_5278"} +{"question": "What research have been done on neural network pruning and sparsification?", "answer": ["Learning both Weights and Connections for Efficient Neural Networks", "Dynamic Network Surgery for Efficient DNNs", "NISP: Pruning Networks using Neuron Importance Score Propagation", "Rethinking the Value of Network Pruning", "Regularized Evolution for Image Classifier Architecture Search", "SNIP: Single-shot Network Pruning based on Connection Sensitivity", "To prune, or not to prune: exploring the efficacy of pruning for model compression", "Exploring Sparsity in Recurrent Neural Networks", "Dynamic Model Pruning with Feedback", "Effective Model Sparsification by Scheduled Grow-and-Prune Methods", "Deep Rewiring: Training very sparse deep networks", "Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization", "Rigging the Lottery: Making All Tickets Winners", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "One ticket to win them all: generalizing lottery ticket initializations across datasets and optimizers"], "answer_arxiv_id": ["1506.02626", "1608.04493", "1711.05908", "1810.05270", "1802.01548", "1810.02340", "1710.01878", "1704.05119", "2006.07253", "2106.09857", "1711.05136", "1902.05967", "1911.11134", "1803.03635", "1906.02773"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_5279"} +{"question": "Which papers improved upon the Neural Radiance Fields (NeRF) technique focusing on the quality of novel view synthesis and the speeds of training and rendering?", "answer": ["Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2111.12077", "2201.05989"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_5280"} +{"question": "Could you list some works about considering smaller function spaces, like smooth functions, to overcome the limitations of the curse of dimensionality?", "answer": ["Deep Network Approximation for Smooth Functions"], "answer_arxiv_id": ["2001.03040"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_5281"} +{"question": "What work first proposed a method to generate images from audio recordings?", "answer": ["Towards Audio to Scene Image Synthesis using Generative Adversarial\n Network"], "answer_arxiv_id": ["1808.04108"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_5282"} +{"question": "Could you provide me some works that have attempted to solve the challenge of exploration in skill learning approaches?", "answer": ["CIC: Contrastive Intrinsic Control for Unsupervised Skill Discovery", "APS: Active Pretraining with Successor Features", "Unsupervised Skill Discovery with Bottleneck Option Learning", "Lipschitz-constrained Unsupervised Skill Discovery"], "answer_arxiv_id": ["2202.00161", "2108.13956", "2106.14305", "2202.00914"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_5283"} +{"question": "What are some examples of research that applied static NeRF by introducing an additional time dimension or latent code?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Forward Flow for Novel View Synthesis of Dynamic Scenes", "Robust Dynamic Radiance Fields", "DynIBaR: Neural Dynamic Image-Based Rendering", "Flow supervision for Deformable NeRF"], "answer_arxiv_id": ["2011.13961", "2309.17390v1", "2301.02239", "2211.11082", "2303.16333"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_5284"} +{"question": "Mention the works where recurrent neural networks (RNNs) such as vanilla RNNs or LSTMs are utilized in reinforcement learning methods?", "answer": ["Recurrent Reinforcement Learning: A Hybrid Approach"], "answer_arxiv_id": ["1509.03044"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_5285"} +{"question": "What are some examples of native Chinese benchmarks used for evaluating LLMs' commonsense reasoning abilities?", "answer": ["LogiQA: A Challenge Dataset for Machine Reading Comprehension with\n Logical Reasoning", "CLUE: A Chinese Language Understanding Evaluation Benchmark", "CMMLU: Measuring massive multitask language understanding in Chinese", "CORECODE: A Common Sense Annotated Dialogue Dataset with Benchmark Tasks\n for Chinese Large Language Models"], "answer_arxiv_id": ["2007.08124", "2004.05986", "2306.09212", "2312.12853"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_5286"} +{"question": "What is the reference that proposes the idea of representing a part-whole hierarchy by weight-sharing columns?", "answer": ["How to represent part-whole hierarchies in a neural network"], "answer_arxiv_id": ["2102.12627"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_5287"} +{"question": "Could you provide me some works focusing on the safe offline RL setting?", "answer": ["Batch Policy Learning under Constraints", "Constraints Penalized Q-learning for Safe Offline Reinforcement Learning", "COptiDICE: Offline Constrained Reinforcement Learning via Stationary Distribution Correction Estimation"], "answer_arxiv_id": ["1903.08738", "2107.09003", "2204.08957"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5288"} +{"question": "Which research work discussed the importance of addressing trustworthiness and distributional shift issues of automated system in healthcare?", "answer": ["Trust Issues: Uncertainty Estimation Does Not Enable Reliable OOD\n Detection On Medical Tabular Data"], "answer_arxiv_id": ["2011.03274"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_5289"} +{"question": "Could you list some works where image priors are applied to image denoising and deraining?", "answer": ["When Image Denoising Meets High-Level Vision Tasks: A Deep Learning\n Approach"], "answer_arxiv_id": ["1706.04284"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5290"} +{"question": "What papers manipulate the sampling procedure of the latent space to embed human preference and controllability into deep generative models?", "answer": ["Controllable and Compositional Generation with Latent-Space Energy-Based Models", "Diffusion-LM Improves Controllable Text Generation", "DiffuseVAE: Efficient, Controllable and High-Fidelity Generation from Low-Dimensional Latents"], "answer_arxiv_id": ["2110.10873", "2205.14217", "2201.00308"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_5291"} +{"question": "Which research papers explored open-vocabulary segmentation frameworks using language data as auxiliary weak supervision?", "answer": ["Segment Everything Everywhere All at Once", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2304.06718", "2210.04150", "2303.04803"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_5292"} +{"question": "Could you list out the works that address finding stationary points of a function?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems", "Solving a Class of Non-Convex Min-Max Games Using Iterative First Order Methods"], "answer_arxiv_id": ["1906.00331", "1902.08297"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_5293"} +{"question": "Could you tell me some studies regarding the scaling of language models?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2005.14165", "2302.13971"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_5294"} +{"question": "Which works discuss the requirement of full coverage in offline RL?", "answer": ["Information-Theoretic Considerations in Batch Reinforcement Learning"], "answer_arxiv_id": ["1905.00360"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_5295"} +{"question": "What is the work that proposed counterfactual Tpps?", "answer": ["Counterfactual Temporal Point Processes"], "answer_arxiv_id": ["2111.07603"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_5296"} +{"question": "What research papers demonstrate that step-by-step reasoning can improve the performance of large language models?", "answer": ["Explain Yourself! Leveraging Language Models for Commonsense Reasoning", "Unsupervised Commonsense Question Answering with Self-Talk", "Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Few-Shot Self-Rationalization with Natural Language Prompts"], "answer_arxiv_id": ["1906.02361", "2004.05483", "2112.00114", "2201.11903", "2111.08284"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_5297"} +{"question": "Which papers present model selection approaches that use random search and grid search?", "answer": ["Grid Search, Random Search, Genetic Algorithm: A Big Comparison for NAS"], "answer_arxiv_id": ["1912.06059"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_5298"} +{"question": "Which works proposed general computational models for transformers?", "answer": ["Statistically Meaningful Approximation: a Case Study on Approximating Turing Machines with Transformers", "What learning algorithm is in-context learning? Investigations with linear models"], "answer_arxiv_id": ["2107.13163", "2211.15661"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_5299"} +{"question": "What works proposed or discussed different models variants of AM for constructive NCO?", "answer": ["POMO: Policy Optimization with Multiple Optima for Reinforcement Learning", "Multi-Decoder Attention Model with Embedding Glimpse for Solving Vehicle Routing Problems", "Matrix Encoding Networks for Neural Combinatorial Optimization", "Learning Collaborative Policies to Solve NP-hard Routing Problems"], "answer_arxiv_id": ["2010.16011", "2012.10638", "2106.11113", "2110.13987"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_5300"} +{"question": "Any works about predicting future agent states using goal predefinition and conditioning?", "answer": ["MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for\n Behavior Prediction", "TNT: Target-driveN Trajectory Prediction", "Map-Adaptive Goal-Based Trajectory Prediction", "Leveraging Future Relationship Reasoning for Vehicle Trajectory\n Prediction", "GANet: Goal Area Network for Motion Forecasting", "Motion Transformer with Global Intention Localization and Local Movement\n Refinement", "ProphNet: Efficient Agent-Centric Motion Forecasting with\n Anchor-Informed Proposals"], "answer_arxiv_id": ["1910.05449", "2008.08294", "2009.04450", "2305.14715", "2209.09723", "2209.13508", "2303.12071"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_5301"} +{"question": "Which work explores predicting the prefix of a model for adaptation using input-output pairs for a task?", "answer": ["HyperTuning: Toward Adapting Large Language Models without Back-propagation"], "answer_arxiv_id": ["2211.12485"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_5302"} +{"question": "Could you provide some references that talk about information maximization as an alternative to distillation methods?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Learning Representations by Predicting Bags of Visual Words", "Exploring Simple Siamese Representation Learning", "Emerging Properties in Self-Supervised Vision Transformers", "BYOL works even without batch statistics", "Understanding Self-Supervised Learning Dynamics without Contrastive Pairs", "Whitening for Self-Supervised Representation Learning", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction"], "answer_arxiv_id": ["2006.07733", "2002.12247", "2011.10566", "2104.14294", "2010.10241", "2102.06810", "2007.06346", "2105.04906", "2103.03230"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_5303"} +{"question": "What study used seq2seq models to study audio-visual event recognition and localization?", "answer": ["dual-modality seq2seq network for Audio-Visual event localization"], "answer_arxiv_id": ["1902.07473"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_5304"} +{"question": "Which papers work on intrinsic rewards using novelty of transitions or diversity of skills?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Diversity is All You Need: Learning Skills without a Reward Function"], "answer_arxiv_id": ["1507.00814", "1802.06070"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_5305"} +{"question": "Which research works discussed spurious correlations within the text?", "answer": ["E-BERT: Efficient-Yet-Effective Entity Embeddings for BERT", "Calibrating Factual Knowledge in Pretrained Language Models"], "answer_arxiv_id": ["1911.03681", "2210.03329"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_5306"} +{"question": "Can you mention the studies that proposed using an adversarial learning method to mitigate unfairness?", "answer": ["Learning Fair Node Representations with Graph Counterfactual Fairness", "Say No to the Discrimination: Learning Fair Graph Neural Networks with Limited Sensitive Attribute Information"], "answer_arxiv_id": ["2201.03662", "2009.01454"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_5307"} +{"question": "Can you name works that use point clouds from SfM or depth sensors as geometry proxies?", "answer": ["Neural Rerendering in the Wild", "Revealing Scenes by Inverting Structure from Motion Reconstructions", "Free-Viewpoint RGB-D Human Performance Capture and Rendering"], "answer_arxiv_id": ["1904.04290", "1904.03303", "2112.13889"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_5308"} +{"question": "What works introduce content semantics adapters for targeted tasks in generative modelling?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_5309"} +{"question": "Which studies investigated the memorization capabilities of generative models?", "answer": ["Quantifying Memorization Across Neural Language Models", "Measuring Forgetting of Memorized Training Examples"], "answer_arxiv_id": ["2202.07646", "2207.00099"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5310"} +{"question": "Which works improve zero-shot transfer by utilizing GPT-3 to generate rich context descriptions?", "answer": ["Visual Classification via Description from Large Language Models"], "answer_arxiv_id": ["2210.07183"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_5311"} +{"question": "What studies used the chaining method for collection of paraphrases in crowdsourcing?", "answer": ["Directed Diversity: Leveraging Language Embedding Distances for Collective Creativity in Crowd Ideation", "Outlier Detection for Improved Data Quality and Diversity in Dialog\n Systems"], "answer_arxiv_id": ["2101.06030v1", "1904.03122"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_5312"} +{"question": "What works have developed oracle-efficient boosting-like algorithms for multicalibration?", "answer": ["Multiaccuracy: Black-Box Post-Processing for Fairness in Classification", "Low-Degree Multicalibration", "Outcome Indistinguishability"], "answer_arxiv_id": ["1805.12317", "2203.01255", "2011.13426"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_5313"} +{"question": "Which works employed cross-task transfer learning using supervision from tasks like Semantic Role Labeling for low-resource IE models?", "answer": ["Unsupervised Label-aware Event Trigger and Argument Classification", "Zero-Shot Transfer Learning for Event Extraction"], "answer_arxiv_id": ["2012.15243", "1707.01066"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_5314"} +{"question": "What works contribute to the sub-field of human action recognition known as human interaction recognition?", "answer": ["Interaction Relational Network for Mutual Action Recognition", "IGFormer: Interaction Graph Transformer for Skeleton-based Human\n Interaction Recognition"], "answer_arxiv_id": ["1910.04963", "2207.12100"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_5315"} +{"question": "What works have been done on non-contrastive methods in self-supervised learning?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "Understanding Self-Supervised Learning Dynamics without Contrastive Pairs", "The Curious Case of Benign Memorization"], "answer_arxiv_id": ["2104.14294", "2102.06810", "2210.14019"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_5316"} +{"question": "Which research papers have focused on the use of non-recurrent neural dynamics models for decision making?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Offline Reinforcement Learning as One Big Sequence Modeling Problem"], "answer_arxiv_id": ["2106.01345", "2106.02039"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_5317"} +{"question": "What are the well-known black-box adversarial attacks according to the literature?", "answer": ["Square Attack: a query-efficient black-box adversarial attack via random search", "Simple Black-box Adversarial Attacks"], "answer_arxiv_id": ["1912.00049", "1905.07121"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_5318"} +{"question": "What are some of the studies that have worked on learning algorithms with uncertainty estimate to improve reliability of learning models?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "A Simple Baseline for Bayesian Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.02142", "1612.01474", "1902.02476"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_5319"} +{"question": "Which research papers made successful predictions on different material properties using message-passing framework for crystal structure?", "answer": ["Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties"], "answer_arxiv_id": ["1710.10324"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_5320"} +{"question": "What studies have developed zero-cost proxy methods to estimate the quality of architectures?", "answer": ["Neural Architecture Search without Training", "Zen-NAS: A Zero-Shot NAS for High-Performance Image Recognition", "Zero-Cost Proxies for Lightweight NAS", "SNIP: Single-shot Network Pruning based on Connection Sensitivity", "Picking Winning Tickets Before Training by Preserving Gradient Flow", "Pruning neural networks without any data by iteratively conserving synaptic flow", "NASI: Label- and Data-agnostic Neural Architecture Search at Initialization", "Unifying and Boosting Gradient-Based Training-Free Neural Architecture Search", "KNAS: Green Neural Architecture Search", "Training-free Transformer Architecture Search", "LiteTransformerSearch: Training-free Neural Architecture Search for Efficient Language Models"], "answer_arxiv_id": ["2006.04647", "2102.01063", "2101.08134", "1810.02340", "2002.07376", "2006.05467", "2109.00817", "2201.09785", "2111.13293", "2203.12217", "2203.02094"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_5321"} +{"question": "Any research papers on memory optimization to speed up the self-attention computation process?", "answer": ["Efficient Memory Management for Large Language Model Serving with\n PagedAttention", "Fast Transformer Decoding: One Write-Head is All You Need", "GQA: Training Generalized Multi-Query Transformer Models from Multi-Head\n Checkpoints"], "answer_arxiv_id": ["2309.06180", "1911.02150", "2305.13245"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_5322"} +{"question": "Any studies discussing the role of PLMs to handle raw input with diverse formats?", "answer": ["Measuring Massive Multitask Language Understanding"], "answer_arxiv_id": ["2009.03300"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_5323"} +{"question": "Could you provide me some studies that fall into the category of heuristic and regularized architectures?", "answer": ["How to Incorporate Monotonicity in Deep Networks While Preserving Flexibility?", "Certified Monotonic Neural Networks", "Counterexample-Guided Learning of Monotonic Neural Networks"], "answer_arxiv_id": ["1909.10662", "2011.10219", "2006.08852"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_5324"} +{"question": "Which works propagated for the development of better divergence measures in EBMs?", "answer": ["Training Deep Energy-Based Models with f-Divergence Minimization", "Pseudo-Spherical Contrastive Divergence", "Improved Contrastive Divergence Training of Energy Based Models"], "answer_arxiv_id": ["2003.03463v2", "2111.00780", "2012.01316v4"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_5325"} +{"question": "Which papers discuss the use of a fine-tuned RoBERTa model for distinguishing between machine-generated and human-written texts?", "answer": ["GPT-Sentinel: Distinguishing Human and ChatGPT Generated Content", "On the Detectability of ChatGPT Content: Benchmarking, Methodology, and\n Evaluation through the Lens of Academic Writing"], "answer_arxiv_id": ["2305.07969", "2306.05524"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_5326"} +{"question": "Could you provide me studies that use diffusion models for learning a distribution over 3D poses?", "answer": ["DiffPose: Toward More Reliable 3D Pose Estimation", "Diffusion-Based 3D Human Pose Estimation with Multi-Hypothesis\n Aggregation"], "answer_arxiv_id": ["2211.16940", "2303.11579"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_5327"} +{"question": "Are there any studies that have explored temporal dynamic routing in the context of Personalized Federated Learning?", "answer": ["Neural Speed Reading via Skim-RNN"], "answer_arxiv_id": ["1711.02085"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_5328"} +{"question": "Are there any works focused on developing learnable semantic map representations in the field of robotics and autonomous driving?", "answer": ["Semantic MapNet: Building Allocentric Semantic Maps and Representations\n from Egocentric Views", "Learning to Explore using Active Neural SLAM"], "answer_arxiv_id": ["2010.01191", "2004.05155"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_5329"} +{"question": "Are there studies on Unlearnable examples (ULEs) that provide textures easy to learn?", "answer": ["Unlearnable Examples: Making Personal Data Unexploitable"], "answer_arxiv_id": ["2101.04898"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_5330"} +{"question": "What works have proposed poisoning attack frameworks trying to violate the predictive parity among subgroups in classification?", "answer": ["Poisoning Attacks on Algorithmic Fairness", "Exacerbating Algorithmic Bias through Fairness Attacks"], "answer_arxiv_id": ["2004.07401", "2012.08723"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_5331"} +{"question": "Could you provide some studies that tackled reprogramming the speech recognition model for time-series data classification tasks?", "answer": ["Voice2Series: Reprogramming Acoustic Models for Time Series Classification"], "answer_arxiv_id": ["2106.09296"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_5332"} +{"question": "What studies elaborate on the components of an LLM-based agent such as planning, memory, reflection, and retrieval?", "answer": ["Generative Agents: Interactive Simulacra of Human Behavior"], "answer_arxiv_id": ["2304.03442"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_5333"} +{"question": "What works are about using Kronecker factored preconditioning to approximate the Fisher-information matrix?", "answer": ["Optimizing Neural Networks with Kronecker-factored Approximate Curvature", "R"], "answer_arxiv_id": ["1503.05671", "1210.6589"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_5334"} +{"question": "Which work describes the per-column regression problem formulation of layout estimation?", "answer": ["HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch\n Data Augmentation"], "answer_arxiv_id": ["1901.03861"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5335"} +{"question": "Which datasets are commonly used for diacritization in Arabic dialects?", "answer": ["Diacritization of Maghrebi Arabic Sub-Dialects"], "answer_arxiv_id": ["1810.06619"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_train_5336"} +{"question": "Which papers does the author refer to for an extensive review of the ANNS structures?", "answer": ["A Comprehensive Survey and Experimental Comparison of Graph-Based Approximate Nearest Neighbor Search"], "answer_arxiv_id": ["2101.12631"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5337"} +{"question": "Which research demonstrated a joint reconstruction of both image and label representation?", "answer": ["Deep Leakage from Gradients"], "answer_arxiv_id": ["1906.08935"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_5338"} +{"question": "What are some of the studies that have been done on the synthesis of datasets using 3D computer simulations?", "answer": ["Virtual Worlds as Proxy for Multi-Object Tracking Analysis", "Virtual KITTI 2"], "answer_arxiv_id": ["1605.06457", "2001.10773"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_5339"} +{"question": "What studies employ disentanglement or identity perturbation to de-identify face identity?", "answer": ["Differentially Private Imaging via Latent Space Manipulation", "CFA-Net: Controllable Face Anonymization Network with Identity\n Representation Manipulation", "IdentityDP: Differential Private Identification Protection for Face\n Images"], "answer_arxiv_id": ["2103.05472", "2105.11137", "2103.01745"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_5340"} +{"question": "What works have utilized CAM for interpreting convolutional neural networks?", "answer": ["This Looks Like That: Deep Learning for Interpretable Image Recognition", "Learning Deep Features for Discriminative Localization", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based\n Localization"], "answer_arxiv_id": ["1806.10574", "1512.04150", "1610.02391"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_5341"} +{"question": "What work extends the methodology of obtaining concentration bounds using the supremum distance?", "answer": ["Functional Sequential Treatment Allocation"], "answer_arxiv_id": ["1812.09408v8"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_5342"} +{"question": "What papers discuss the potential of training a full generative language model for whale communication?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2204.02311"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_5343"} +{"question": "Which works explore understanding the representations induced by GANs for interpretation and control?", "answer": ["GAN Dissection: Visualizing and Understanding Generative Adversarial Networks", "Interpreting the Latent Space of GANs for Semantic Face Editing", "Semantic Hierarchy Emerges in Deep Generative Representations for Scene Synthesis", "Closed-Form Factorization of Latent Semantics in GANs", "GANSpace: Discovering Interpretable GAN Controls", "Unsupervised Discovery of Interpretable Directions in the GAN Latent Space", "Cluster-guided Image Synthesis with Unconditional Models", "WarpedGANSpace: Finding non-linear RBF paths in GAN latent space", "Low-Rank Subspaces in GANs", "Finding Directions in GAN’s Latent Space for Neural Face Reenactment", "StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation", "StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated Images using Conditional Continuous Normalizing Flows"], "answer_arxiv_id": ["1811.10597", "1907.10786", "1911.09267", "2007.06600", "2004.02546", "2002.03754", "2112.12911", "2109.13357", "2106.04488", "2202.00046", "2011.12799", "2008.02401"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_5344"} +{"question": "What research has been conducted on variable misuse detection and identifying “simple stupid bugs” in programs?", "answer": ["Learning to Represent Programs with Graphs"], "answer_arxiv_id": ["1711.00740"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_5345"} +{"question": "What studies proposed the framework of the message-passing architecture for graph neural networks?", "answer": ["Neural Message Passing for Quantum Chemistry"], "answer_arxiv_id": ["1704.01212"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_5346"} +{"question": "Any works about the methods proposed to maintain the benefits of fine-tuning with rehearsal in continuous learning while minimizing the size of the replay buffer?", "answer": ["Playing Atari with Deep Reinforcement Learning", "A continual learning survey: Defying forgetting in classification tasks"], "answer_arxiv_id": ["1312.5602", "1909.08383"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_5347"} +{"question": "Could you provide me with the examples of research that study ways to insert a backdoor that triggers on certain inputs in reinforcement learning?", "answer": ["Poisoning Deep Reinforcement Learning Agents with In-Distribution Triggers"], "answer_arxiv_id": ["2106.07798"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_5348"} +{"question": "Can you tell me about research papers that demonstrated the ability of the MLE-NCM to identify and estimate ℒ2 queries in practice?", "answer": ["The Causal-Neural Connection: Expressiveness, Learnability, and Inference"], "answer_arxiv_id": ["2107.00793"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5349"} +{"question": "What work is closely comparable to the researcher's work which fine-tunes a language model to produce prompts that lead to toxic completions?", "answer": ["Red Teaming Language Models with Language Models"], "answer_arxiv_id": ["2202.03286v1"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_5350"} +{"question": "Which paper first applied spherical harmonics to achieve SE3 equivariance in the context of point clouds?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural\n networks for 3D point clouds"], "answer_arxiv_id": ["1802.08219"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_5351"} +{"question": "Which works showed that CLIP models can perform the reverse task of image-to-text generation?", "answer": ["ZeroCap: Zero-Shot Image-to-Text Generation for Visual-Semantic Arithmetic"], "answer_arxiv_id": ["2111.14447"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5352"} +{"question": "Could you provide studies that utilized data rebalancing in an attempt to mitigate spurious correlations?", "answer": ["Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural\n Language Inference", "HellaSwag: Can a Machine Really Finish Your Sentence?"], "answer_arxiv_id": ["1902.01007", "1905.07830"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_5353"} +{"question": "Which papers discuss mean estimation dealing with bounded data points?", "answer": ["Between Pure and Approximate Differential Privacy", "A Primer on Private Statistics"], "answer_arxiv_id": ["1501.06095", "2005.00010"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_5354"} +{"question": "Could you provide me some papers about the deterministic policy gradient (DPG) algorithm in reinforcement learning?", "answer": ["Continuous control with deep reinforcement learning"], "answer_arxiv_id": ["1509.02971"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_5355"} +{"question": "What research demonstrated that gradient descent converges to the minimum ℓ2-norm solutions in linear regression problems?", "answer": ["Characterizing Implicit Bias in Terms of Optimization Geometry"], "answer_arxiv_id": ["1802.08246"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_5356"} +{"question": "Which papers focus on prompt template design?", "answer": ["Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference"], "answer_arxiv_id": ["2001.07676"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_5357"} +{"question": "Which paper developed the method of contrasting instance-level representations obtained from the same scenario but processed by different model architecture?", "answer": ["Self-Supervised Pretraining of 3D Features on any Point-Cloud"], "answer_arxiv_id": ["2101.02691"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_5358"} +{"question": "Which works in the literature have identified biases in pronoun resolution, especially in cases related to machine translation, use of non-binary pronouns, and gender ambiguity?", "answer": ["Collecting a Large-Scale Gender Bias Dataset for Coreference Resolution and Machine Translation", "Toward Gender-Inclusive Coreference Resolution", "Mind the GAP: A Balanced Corpus of Gendered Ambiguous Pronouns"], "answer_arxiv_id": ["2109.03858", "1910.13913", "1810.05201"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_5359"} +{"question": "What datasets are specifically designed to research in ambiguous environments?", "answer": ["6D Camera Relocalization in Ambiguous Scenes via Continuous Multimodal Inference"], "answer_arxiv_id": ["2004.04807v2"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_5360"} +{"question": "What studies introduced adversarial reprogramming-based prompting for general pre-training using vision-language relationships?", "answer": ["Exploring Visual Prompts for Adapting Large-Scale Models", "Adversarial Reprogramming of Neural Networks"], "answer_arxiv_id": ["2203.17274", "1806.11146"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_5361"} +{"question": "Which works have introduced techniques for image editing with diffusion models, specifically for inpainting tasks?", "answer": ["Denoising Diffusion Implicit Models", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2010.02502", "2112.10741"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_5362"} +{"question": "Which work first proposed a probabilistic box lattice to embed entities in the knowledge graph as n-dimensional rectangles, or box embedding?", "answer": ["Probabilistic Embedding of Knowledge Graphs with Box Lattice Measures"], "answer_arxiv_id": ["1805.06627"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_5363"} +{"question": "Which works discussed the concept of Foundation Models?", "answer": ["On the Opportunities and Risks of Foundation Models", "Large-scale Multi-Modal Pre-trained Models: A Comprehensive Survey"], "answer_arxiv_id": ["2108.07258v3", "2302.10035"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_5364"} +{"question": "Could you provide me works that use BERT for calculating term importance?", "answer": ["Context-Aware Sentence/Passage Term Importance Estimation For First Stage Retrieval"], "answer_arxiv_id": ["1910.10687"], "source_meta": {"published_time": "20230409"}, "qid": "AutoScholarQuery_train_5365"} +{"question": "Which paper focuses on understanding the fundamental trade-offs between group calibration and other fairness criteria?", "answer": ["Inherent Trade-Offs in the Fair Determination of Risk Scores"], "answer_arxiv_id": ["1609.05807"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_5366"} +{"question": "Which studies that have tried applying LLMs for document visual information extraction?", "answer": ["LMDX: Language Model-based Document Information Extraction and\n Localization"], "answer_arxiv_id": ["2309.10952"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_5367"} +{"question": "Are there any research studies that utilize a surrogate model to estimate metrics over the unlabeled test set?", "answer": ["ALT-MAS: A Data-Efficient Framework for Active Testing of Machine\n Learning Algorithms"], "answer_arxiv_id": ["2104.04999"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_5368"} +{"question": "Who worked on using high-level specifications with RL algorithms?", "answer": ["A Composable Specification Language for Reinforcement Learning Tasks", "Compositional Reinforcement Learning from Logical Specifications", "Reinforcement Learning With Temporal Logic Rewards"], "answer_arxiv_id": ["2008.09293", "2106.13906", "1612.03471"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_5369"} +{"question": "Could you provide me some works about Bayesian Optimisation Frameworks specialized in multi-fidelity optimization?", "answer": ["The Parallel Knowledge Gradient Method for Batch Bayesian Optimization", "Emulation of physical processes with Emukit"], "answer_arxiv_id": ["1606.04414", "2110.13293"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_5370"} +{"question": "What studies use the S-learner and Robinson decomposition to tailor specific existing CATE estimators?", "answer": ["Meta-learners for Estimating Heterogeneous Treatment Effects using Machine Learning", "Quasi-Oracle Estimation of Heterogeneous Treatment Effects"], "answer_arxiv_id": ["1706.03461", "1712.04912"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_5371"} +{"question": "Are there any works that use observed intervention targets to identify causal factors in causal representation learning?", "answer": ["CITRIS: Causal Identifiability from Temporal Intervened Sequences", "Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systems"], "answer_arxiv_id": ["2202.03169", "2206.06169"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_5372"} +{"question": "What papers propose extending image-based ControlNet to the video domain with full cross-frame attention?", "answer": ["ControlVideo: Training-free Controllable Text-to-Video Generation"], "answer_arxiv_id": ["2305.13077"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_5373"} +{"question": "Which works focused on the topic of Active Vision that continually acquires new visual observations helpful for object classification, recognition, detection and segmentation?", "answer": ["Multiple Instance Active Learning for Object Detection", "Geometry-Aware Recurrent Neural Networks for Active Visual Recognition", "Revisiting Active Perception", "Reinforced Active Learning for Image Segmentation", "Efficient Active Learning for Image Classification and Segmentation using a Sample Selection and Conditional Generative Adversarial Network", "Embodied Visual Active Learning for Semantic Segmentation", "Look-ahead before you leap: end-to-end active recognition by forecasting the effect of motion"], "answer_arxiv_id": ["2104.02324", "1811.01292v2", "1603.02729v2", "2002.06583", "1806.05473", "2012.09503", "1605.00164"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5374"} +{"question": "What research has been conducted in training models specifically designed for evaluating image quality for text-to-image models?", "answer": ["ImageReward: Learning and Evaluating Human Preferences for Text-to-Image\n Generation", "Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image\n Generation", "Human Preference Score: Better Aligning Text-to-Image Models with Human\n Preference", "Human Preference Score v2: A Solid Benchmark for Evaluating Human\n Preferences of Text-to-Image Synthesis"], "answer_arxiv_id": ["2304.05977", "2305.01569", "2303.14420", "2306.09341"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_5375"} +{"question": "What studies compare sample complexity bounds for standard vs. robust error?", "answer": ["Adversarially Robust Generalization Requires More Data", "Rademacher Complexity for Adversarially Robust Generalization", "Adversarial risk bounds via function transformation"], "answer_arxiv_id": ["1804.11285", "1810.11914", "1810.09519"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_5376"} +{"question": "Which work proposes a quantization scheme called uniform affine or asymmetric quantization?", "answer": ["Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations", "Quantizing deep convolutional networks for efficient inference: A whitepaper", "DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients"], "answer_arxiv_id": ["1609.07061", "1806.08342", "1606.06160"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_5377"} +{"question": "Could you provide me studies that attempted to learn a hypothesis for all groups via importance up-weighting?", "answer": ["What is the Effect of Importance Weighting in Deep Learning?"], "answer_arxiv_id": ["1812.03372"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_5378"} +{"question": "Which research brought language priors into time-series forecasting?", "answer": ["Numeracy for Language Models: Evaluating and Improving their Ability to\n Predict Numbers", "An Empirical Investigation of Contextualized Number Prediction"], "answer_arxiv_id": ["1805.08154", "2011.07961"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_5379"} +{"question": "Which studies have addressed the safety concerns and potential misuses of LLMs?", "answer": ["On the Opportunities and Risks of Foundation Models", "Extracting Training Data from Large Language Models", "Predictability and Surprise in Large Generative Models", "Ethical and social risks of harm from Language Models"], "answer_arxiv_id": ["2108.07258v3", "2012.07805", "2202.07785", "2112.04359"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_5380"} +{"question": "Could you provide me some studies about recently developed binary treatment estimators?", "answer": ["Nonparametric Estimation of Heterogeneous Treatment Effects: From Theory to Learning Algorithms", "Estimating individual treatment effects under unobserved confounding using binary instruments", "Heterogeneous Treatment Effect with Trained Kernels of the Nadaraya-Watson Regression"], "answer_arxiv_id": ["2101.10943v2", "2208.08544", "2207.09139"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_5381"} +{"question": "Any works about automatically estimating the camera poses or finetune them during training?", "answer": ["NeRF-⁣-: Neural Radiance Fields Without Known Camera Parameters", "BARF : Bundle-Adjusting Neural Radiance Fields", "NeROIC: Neural Rendering of Objects from Online Image Collections"], "answer_arxiv_id": ["2102.07064", "2104.06405", "2201.02533v2"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_5382"} +{"question": "What are the works about low rank matrix estimation in the context of supervised learning?", "answer": ["The Power of Convex Relaxation: Near-Optimal Matrix Completion", "Estimation of (near) low-rank matrices with noise and high-dimensional scaling", "Inference and Uncertainty Quantification for Noisy Matrix Completion", "Low-rank Matrix Completion using Alternating Minimization", "Non-convex Optimization for Machine Learning", "A Simpler Approach to Matrix Completion", "Noisy Matrix Completion: Understanding Statistical Guarantees for Convex Relaxation via Nonconvex Optimization", "Nonconvex Optimization Meets Low-Rank Matrix Factorization: An Overview"], "answer_arxiv_id": ["0903.1476", "0912.5100v1", "1906.04159v2", "1212.0467", "1712.07897", "0910.0651", "1902.07698v2", "1809.09573"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_5383"} +{"question": "Which papers focus on models where robots share the same morphology and differ only in kinematics or dynamics parameters?", "answer": ["Hardware Conditioned Policies for Multi-Robot Transfer Learning", "Sim-to-Real Transfer of Robotic Control with Dynamics Randomization", "Bayesian Meta-Learning for Few-Shot Policy Adaptation Across Robotic Platforms", "GenLoco: Generalized Locomotion Controllers for Quadrupedal Robots"], "answer_arxiv_id": ["1811.09864", "1710.06537", "2103.03697", "2209.05309"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_5384"} +{"question": "Which works introduced domain discriminator to extract domain-invariant features?", "answer": ["Domain Adaptive Faster R-CNN for Object Detection in the Wild", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks"], "answer_arxiv_id": ["1803.03243", "1506.01497"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_5385"} +{"question": "Which works are related to aggregation-based methods in stereo matching?", "answer": ["End-to-End Learning of Geometry and Context for Deep Stereo Regression", "Pyramid Stereo Matching Network", "Group-wise Correlation Stereo Network", "Accurate and Efficient Stereo Matching via Attention Concatenation\n Volume", "CGI-Stereo: Accurate and Real-Time Stereo Matching via Context and\n Geometry Interaction", "GA-Net: Guided Aggregation Net for End-to-end Stereo Matching", "AANet: Adaptive Aggregation Network for Efficient Stereo Matching", "PCW-Net: Pyramid Combination and Warping Cost Volume for Stereo Matching"], "answer_arxiv_id": ["1703.04309", "1803.08669", "1903.04025", "2209.12699", "2301.02789", "1904.06587", "2004.09548", "2006.12797"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_5386"} +{"question": "Which works found the performance disparity in radiograph classification?", "answer": ["Deep Learning Predicts Hip Fracture using Confounding Patient and Healthcare Variables"], "answer_arxiv_id": ["1811.03695"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_5387"} +{"question": "What papers applied prompt-learning or a combination of it with Layer Normalization layers fine-tuning?", "answer": ["Visual Prompt Tuning", "Conditional Prompt Learning for Vision-Language Models", "CLIP for All Things Zero-Shot Sketch-Based Image Retrieval, Fine-Grained\n or Not", "What Can Human Sketches Do for Object Detection?"], "answer_arxiv_id": ["2203.12119", "2203.05557", "2303.13440", "2303.15149"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_5388"} +{"question": "Who proposed the text-based anomaly detection method WinCLIP?", "answer": ["WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation"], "answer_arxiv_id": ["2303.14814"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_5389"} +{"question": "What papers are about finetuning models to take images as condition but lose most image details?", "answer": ["VideoComposer: Compositional Video Synthesis with Motion Controllability", "I2VGen-XL: High-Quality Image-to-Video Synthesis via Cascaded Diffusion\n Models"], "answer_arxiv_id": ["2306.02018", "2311.04145"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_5390"} +{"question": "Which papers show that LLMs could answer questions with explicit reasoning steps by being provided with a chain of thoughts?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_5391"} +{"question": "Which studies used words' co-occurrence vectors and PMI for lexical entailment models?", "answer": ["pair2vec: Compositional Word-Pair Embeddings for Cross-Sentence\n Inference"], "answer_arxiv_id": ["1810.08854"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_5392"} +{"question": "Could you provide me some works discussing the expressiveness problem of GNNs?", "answer": ["How Powerful are Graph Neural Networks?", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "Uplifting Message Passing Neural Network with Graph Original Information", "Identity-aware Graph Neural Networks"], "answer_arxiv_id": ["1810.00826", "1810.02244", "2210.05382", "2101.10320"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_5393"} +{"question": "Who proposed the Intra Order invariant (IO), an order-preserving function for post-hoc calibration?", "answer": ["Intra Order-Preserving Functions for Calibration of Multi-Class Neural Networks"], "answer_arxiv_id": ["2003.06820"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_5394"} +{"question": "Which works demonstrated the ability to bridge the gap between visual content and textual descriptions through vision-language pretraining?", "answer": ["VirTex: Learning Visual Representations from Textual Annotations", "Learning Transferable Visual Models From Natural Language Supervision", "VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text\n Understanding", "VT-CLIP: Enhancing Vision-Language Models with Visual-guided Texts"], "answer_arxiv_id": ["2006.06666", "2103.00020", "2109.14084", "2112.02399"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_5395"} +{"question": "What research have used pretrained visual-language models for text-to-3D generation?", "answer": ["CLIP-Forge: Towards Zero-Shot Text-to-Shape Generation", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields"], "answer_arxiv_id": ["2110.02624", "2112.05139"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_5396"} +{"question": "Can you name a study where an algorithm, SPAG, is proposed for a strongly convex setting?", "answer": ["Statistically Preconditioned Accelerated Gradient Method for Distributed Optimization"], "answer_arxiv_id": ["2002.10726"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_5397"} +{"question": "Could you tell me about the studies that pursue performance enhancements by crafting difficult negative samples in visual-linguistic studies?", "answer": ["When and why vision-language models behave like bags-of-words, and what\n to do about it?", "Teaching Structured Vision&Language Concepts to Vision&Language Models"], "answer_arxiv_id": ["2210.01936", "2211.11733"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_5398"} +{"question": "Which paper offered a dataset containing a mix of human and machine-written texts using operations like polishing, completing, adding natural noise, and adapting?", "answer": ["LLM-as-a-Coauthor: Can Mixed Human-Written and Machine-Generated Text Be\n Detected?"], "answer_arxiv_id": ["2401.05952"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_5399"} +{"question": "Could you provide me some papers that applied Natural Language Explanations (NLEs) to various domains?", "answer": ["WT5?! Training Text-to-Text Models to Explain their Predictions", "e-SNLI: Natural Language Inference with Natural Language Explanations", "Generate Natural Language Explanations for Recommendation", "Rationalization: A Neural Machine Translation Approach to Generating\n Natural Language Explanations", "Explaining Chest X-ray Pathologies in Natural Language", "Multimodal Explanations: Justifying Decisions and Pointing to the\n Evidence", "Grounding Visual Explanations", "e-ViL: A Dataset and Benchmark for Natural Language Explanations in\n Vision-Language Tasks", "Knowledge-Grounded Self-Rationalization via Extractive and Natural\n Language Explanations", "Textual Explanations for Self-Driving Vehicles", "Program Induction by Rationale Generation : Learning to Solve and\n Explain Algebraic Word Problems"], "answer_arxiv_id": ["2004.14546", "1812.01193", "2101.03392", "1702.07826", "2207.04343", "1802.08129", "1807.09685", "2105.03761", "2106.13876", "1807.11546", "1705.04146"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_5400"} +{"question": "Can you tell me about the research papers on single-model arbitrary-scale super-resolution since the inception of MetaSR?", "answer": ["Meta-SR: A Magnification-Arbitrary Network for Super-Resolution", "Learning Continuous Image Representation with Local Implicit Image\n Function", "Cascaded Diffusion Models for High Fidelity Image Generation", "Image Super-Resolution via Iterative Refinement", "Pyramidal Denoising Diffusion Probabilistic Models", "Implicit Diffusion Models for Continuous Super-Resolution"], "answer_arxiv_id": ["1903.00875", "2012.09161", "2106.15282", "2104.07636", "2208.01864", "2303.16491"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_5401"} +{"question": "Which works present an analysis of the Fitted Q-Iteration algorithms?", "answer": ["A Theoretical Analysis of Deep Q-Learning"], "answer_arxiv_id": ["1901.00137"], "source_meta": {"published_time": "20221114"}, "qid": "AutoScholarQuery_train_5402"} +{"question": "What work describes a family of distributions for which SSL algorithms do not offer any advantage over SL by using the causality framework?", "answer": ["On Causal and Anticausal Learning"], "answer_arxiv_id": ["1206.6471"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5403"} +{"question": "Which studies showed that out-of-distribution data can lead to performance drop in machine learning models?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and\n Perturbations", "The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution\n Generalization", "Do ImageNet Classifiers Generalize to ImageNet?", "Causal Transportability for Visual Recognition", "Generative Interventions for Causal Learning", "Discrete Representations Strengthen Vision Transformer Robustness"], "answer_arxiv_id": ["1903.12261", "2006.16241", "1902.10811", "2204.12363", "2012.12265", "2111.10493"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_5404"} +{"question": "Who demonstrate the utilization of polarization in rendering?", "answer": ["PANDORA: Polarization-Aided Neural Decomposition Of Radiance"], "answer_arxiv_id": ["2203.13458"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_5405"} +{"question": "Could you tell me which study proposed a method called BAT to adjust the attack strengths and difficulties of each class?", "answer": ["Improving Robust Fairness via Balance Adversarial Training"], "answer_arxiv_id": ["2209.07534"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_5406"} +{"question": "What works mention the selection-based methods of PEFT that involve fine-tuning the last K layers?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Visual Prompt Tuning"], "answer_arxiv_id": ["1902.00751", "2203.12119"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_5407"} +{"question": "What studies explicitly generated the entire proof to improve finetuning and few-shot learning?", "answer": ["Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Explaining Answers with Entailment Trees", "STaR: Bootstrapping Reasoning With Reasoning", "Improved Logical Reasoning of Language Models via Differentiable Symbolic Programming"], "answer_arxiv_id": ["2112.00114", "2104.08661", "2203.14465", "2305.03742"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_5408"} +{"question": "What works focus on unsupervised approaches for detection and correction of artifact generations in GANs by examining local activation and activation frequency?", "answer": ["An Unsupervised Way to Understand Artifact Generating Internal Units in Generative Neural Networks", "Can We Find Neurons that Cause Unrealistic Images in Deep Generative Networks?"], "answer_arxiv_id": ["2112.08814", "2201.06346"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_5409"} +{"question": "Which work leveraged CLIP for learning pixel-level visual embeddings aligned with the text embeddings of CLIP?", "answer": ["Language-driven Semantic Segmentation"], "answer_arxiv_id": ["2201.03546"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_5410"} +{"question": "Which studies discuss performance degradation in machine learning models due to subpopulation shift?", "answer": ["Wilds: A Benchmark of in-the-Wild Distribution Shifts", "A Theory of Label Propagation for Subpopulation Shift"], "answer_arxiv_id": ["2012.07421", "2102.11203"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_5411"} +{"question": "Which research demonstrates how a slight modification of the algorithm in ~\\cite{bib.bib39} could lead to a first-order regret?", "answer": ["Reward-Free Exploration for Reinforcement Learning"], "answer_arxiv_id": ["2002.02794"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_5412"} +{"question": "What works deal with tree or line constraints on the order in which the boxes can be opened in the context of Pandora's Box?", "answer": ["Pandora’s Box Problem with Order Constraints"], "answer_arxiv_id": ["2002.06968"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_5413"} +{"question": "Who presented the conditional context optimization to generate an input-conditional token for each image?", "answer": ["Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2203.05557"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5414"} +{"question": "Are there any works discussing scene and object reconstruction from multiple viewpoints?", "answer": ["Next-Best View Policy for 3D Reconstruction"], "answer_arxiv_id": ["2008.12664"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_5415"} +{"question": "What papers introduce the use of Siamese networks for feature extraction and linear correlation operation for visual object tracking?", "answer": ["Fully-Convolutional Siamese Networks for Object Tracking"], "answer_arxiv_id": ["1606.09549"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_5416"} +{"question": "Which works proposed Bayesian structure learning based on NeuralODEs?", "answer": ["Neural Ordinary Differential Equations", "Neural Granger Causality", "Neural graphical modelling in continuous-time: consistency guarantees and algorithms", "Beyond Predictions in Neural ODEs: Identification and Interventions", "Sparsity in Continuous-Depth Neural Networks"], "answer_arxiv_id": ["1806.07366", "1802.05842", "2105.02522", "2106.12430", "2210.14672"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_5417"} +{"question": "What work proposes to use voxelized smoothed density value (SDV) representation for matching 3D point clouds?", "answer": ["The Perfect Match: 3D Point Cloud Matching with Smoothed Densities"], "answer_arxiv_id": ["1811.06879"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_5418"} +{"question": "What works have proposed the use of finetuning only methods in transforming LLMs into bi-encoders?", "answer": ["SGPT: GPT Sentence Embeddings for Semantic Search", "Fine-Tuning LLaMA for Multi-Stage Text Retrieval", "Language Models are Universal Embedders", "Making Large Language Models A Better Foundation For Dense Retrieval"], "answer_arxiv_id": ["2202.08904", "2310.08319", "2310.08232", "2312.15503"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_5419"} +{"question": "What works proposed the Group Equivariant Convolutions for rotations and flips?", "answer": ["Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1602.07576"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_5420"} +{"question": "Which studies showed that models with low average error can still fail on certain data point groups?", "answer": ["Fairness Without Demographics in Repeated Loss Minimization", "Demographic Dialectal Variation in Social Media: A Case Study of African-American English"], "answer_arxiv_id": ["1806.08010", "1608.08868"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_5421"} +{"question": "What study proposes the use of a single embedding and adding this to original embeddings linearly per the number of moderate-frequency trigger words?", "answer": ["Are You Copying My Model? Protecting the Copyright of Large Language\n Models for EaaS via Backdoor Watermark"], "answer_arxiv_id": ["2305.10036"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_5422"} +{"question": "What research utilized Large Language Models (LLMs) for examining responses in multi-modal environments?", "answer": ["MMDialog: A Large-scale Multi-turn Dialogue Dataset Towards Multi-modal\n Open-domain Conversation"], "answer_arxiv_id": ["2211.05719"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_5423"} +{"question": "Any research about the application of GNNs in large-scale weather predictions?", "answer": ["Forecasting Global Weather with Graph Neural Networks"], "answer_arxiv_id": ["2202.07575"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_5424"} +{"question": "Which research papers are about index-based methods in implicit video representations that take content-independent time vectors and/or spatial coordinates as inputs?", "answer": ["NeRV: Neural Representations for Videos", "E-NeRV: Expedite Neural Video Representation with Disentangled\n Spatial-Temporal Context", "PS-NeRV: Patch-wise Stylized Neural Representations for Videos", "NIRVANA: Neural Implicit Representations of Videos with Adaptive\n Networks and Autoregressive Patch-wise Modeling"], "answer_arxiv_id": ["2110.13903", "2207.08132", "2208.03742", "2212.14593"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_5425"} +{"question": "What research has been conducted on generating image layouts conditioned on an input object label set?", "answer": ["LayoutVAE: Stochastic Scene Layout Generation From a Label Set"], "answer_arxiv_id": ["1907.10719"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5426"} +{"question": "What works are about probing in explanation method context?", "answer": ["What you can cram into a single $&!#⁢* vector: Probing sentence embeddings for linguistic properties", "BERT Rediscovers the Classical NLP Pipeline", "Understanding Learning Dynamics Of Language Models with SVCCA"], "answer_arxiv_id": ["1805.01070", "1905.05950", "1811.00225"], "source_meta": {"published_time": "20220928"}, "qid": "AutoScholarQuery_train_5427"} +{"question": "Are there any methodologies specifically designed for question answering or summarization?", "answer": ["RankQA: Neural Question Answering with Answer Re-Ranking", "SummaReranker: A Multi-Task Mixture-of-Experts Re-ranking Framework for Abstractive Summarization"], "answer_arxiv_id": ["1906.03008", "2203.06569"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_5428"} +{"question": "Which work applies a conditional diffusion model to colorization, inpainting, and restoration?", "answer": ["Palette: Image-to-Image Diffusion Models"], "answer_arxiv_id": ["2111.05826"], "source_meta": {"published_time": "20220316"}, "qid": "AutoScholarQuery_train_5429"} +{"question": "Who developed a system for human-in-the-loop editing in several domains, including Wikipedia and ArXiv?", "answer": ["Read, Revise, Repeat: A System Demonstration for Human-in-the-loop\n Iterative Text Revision"], "answer_arxiv_id": ["2204.03685"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_5430"} +{"question": "Can you name some studies that employed CLIP in 3D generation?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "Zero-Shot Text-Guided Object Generation with Dream Fields", "Understanding Pure CLIP Guidance for Voxel Grid NeRF Models", "CLIP-Mesh: Generating textured meshes from text using pretrained\n image-text models", "CLIP-Forge: Towards Zero-Shot Text-to-Shape Generation", "Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2212.08751", "2112.01455", "2209.15172", "2203.13333", "2110.02624", "2212.14704"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_5431"} +{"question": "Which work demonstrated that self-supervised methods can produce all-purpose visual features?", "answer": ["DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2304.07193"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5432"} +{"question": "Who proposed to use the VOE method to probe the knowledge of neural networks?", "answer": ["Probing Physics Knowledge Using Tools from Developmental Psychology"], "answer_arxiv_id": ["1804.01128v1"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_5433"} +{"question": "Which research paper closed the gap between universal dynamic regret for convex function and the lower bound?", "answer": ["Adaptive Online Learning in Dynamic Environments"], "answer_arxiv_id": ["1810.10815"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5434"} +{"question": "Could you provide me some works about datasets that share the same super-categories, such as weeds and insects, with Species196-L?", "answer": ["DeepWeeds: A Multiclass Weed Species Image Dataset for Deep Learning"], "answer_arxiv_id": ["1810.05726"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_5435"} +{"question": "Could you provide me with studies about models that use a linear layer for basic decoding in MLLMs without specialized vision encoders?", "answer": ["OtterHD: A High-Resolution Multi-modality Model"], "answer_arxiv_id": ["2311.04219"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_5436"} +{"question": "Which works have previously explored the use of NeRFs for path planning?", "answer": ["Vision-Only Robot Navigation in a Neural Radiance World"], "answer_arxiv_id": ["2110.00168"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_5437"} +{"question": "Could you tell me about transformer-based methods on image segmentation that have adopted a multi-stage strategy to iteratively improve predicted segmentation outcomes?", "answer": ["MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "Masked-attention Mask Transformer for Universal Image Segmentation", "K-Net: Towards Unified Image Segmentation", "SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers", "Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers", "DaTaSeg: Taming a Universal Multi-Dataset Multi-Task Segmentation Model"], "answer_arxiv_id": ["2012.00759", "2112.01527", "2106.14855", "2105.15203", "2109.03814", "2306.01736"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_5438"} +{"question": "Which papers discussed the challenge of distribution shift in offline RL?", "answer": ["Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction"], "answer_arxiv_id": ["1906.00949"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_5439"} +{"question": "Which works are part of traditional explainable computer vision methodologies focusing on interpreting features or neurons within deep neural networks?", "answer": ["Learning Deep Features for Discriminative Localization", "A Unified Approach to Interpreting Model Predictions", "Natural Language Descriptions of Deep Visual Features"], "answer_arxiv_id": ["1512.04150", "1705.07874", "2201.11114"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_5440"} +{"question": "Which works used 3D operators, like OpenOccupancy, ResNet3D and FPN3D for autonomous driving perception?", "answer": ["OpenOccupancy: A Large Scale Benchmark for Surrounding Semantic\n Occupancy Perception", "Deep Residual Learning for Image Recognition", "Feature Pyramid Networks for Object Detection", "SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Driving", "Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2303.03991", "1512.03385", "1612.03144", "2303.09551", "2010.04159"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_5441"} +{"question": "Which papers have examined the concept of 'length extrapolation' in Natural Language Processing context?", "answer": ["Train Short, Test Long: Attention with Linear Biases Enables Input\n Length Extrapolation", "KERPLE: Kernelized Relative Positional Embedding for Length\n Extrapolation"], "answer_arxiv_id": ["2108.12409", "2205.09921"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_5442"} +{"question": "What studies proposed improvements to CAM by creating improved training methods?", "answer": ["Self-produced Guidance for Weakly-supervised Object Localization", "Inter-Image Communication for Weakly Supervised Localization", "TS-CAM: Token Semantic Coupled Attention Map for Weakly Supervised\n Object Localization", "LCTR: On Awakening the Local Continuity of Transformer for Weakly\n Supervised Object Localization", "Weakly Supervised Object Localization via Transformer with Implicit\n Spatial Calibration"], "answer_arxiv_id": ["1807.08902", "2008.05096", "2103.14862", "2112.05291", "2207.10447"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5443"} +{"question": "Which papers developed simulation-free Continuous Normalizing Flow training frameworks?", "answer": ["Moser Flow: Divergence-based Generative Modeling on Manifolds", "Matching Normalizing Flows and Probability Paths on Manifolds"], "answer_arxiv_id": ["2108.08052", "2207.04711"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_5444"} +{"question": "What paper connects squaring a circuit to the Born-rule of quantum mechanics?", "answer": ["Tensor-Train Density Estimation"], "answer_arxiv_id": ["2108.00089v2"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5445"} +{"question": "What research papers focused on Language Models' understanding of cross-cultural differences in values and beliefs?", "answer": ["Assessing Cross-Cultural Alignment between ChatGPT and Human Societies:\n An Empirical Study", "Probing Pre-Trained Language Models for Cross-Cultural Differences in\n Values"], "answer_arxiv_id": ["2303.17466", "2203.13722"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_5446"} +{"question": "Which work investigates the fingerprints left by GAN image synthesis models on the images they generate?", "answer": ["Do GANs leave artificial fingerprints?"], "answer_arxiv_id": ["1812.11842"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_5447"} +{"question": "Which works have been conducted on generative models for 3D scene synthesis based on VAEs and GANs?", "answer": ["SG-VAE: Scene Grammar Variational Autoencoder to generate new indoor\n scenes", "Scene Synthesis via Uncertainty-Driven Attribute Synchronization", "Indoor Scene Generation from a Collection of Semantic-Segmented Depth\n Images"], "answer_arxiv_id": ["1912.04554", "2108.13499", "2108.09022"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_5448"} +{"question": "What papers focused on audio caption tasks, mapping audio embeddings to a sequence of prefix vectors for caption generation?", "answer": ["Prefix tuning for automated audio captioning", "SECap: Speech Emotion Captioning with Large Language Model"], "answer_arxiv_id": ["2303.17489", "2312.10381"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_5449"} +{"question": "Which researches concern foundation models in audio recognition?", "answer": ["ImageBind: One Embedding Space To Bind Them All"], "answer_arxiv_id": ["2305.05665"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5450"} +{"question": "Which of the works resorted to optimizing with a KD-aware loss function to obtain a pruned student architecture in KD?", "answer": ["Search for Better Students to Learn Distilled Knowledge"], "answer_arxiv_id": ["2001.11612"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_5451"} +{"question": "Can you provide me findings related to single-view reconstruction in 3D shapes?", "answer": ["Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision"], "answer_arxiv_id": ["1912.07372"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_5452"} +{"question": "What previous works discussed transfer hyperparameter optimization and how to speed it up?", "answer": ["Few-Shot Bayesian Optimization with Deep Kernel Surrogates", "Pre-trained Gaussian processes for Bayesian optimization", "Transferable Neural Processes for Hyperparameter Optimization"], "answer_arxiv_id": ["2101.07667", "2109.08215", "1909.03209"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_5453"} +{"question": "What prior studies empirically consider settings similar to 'training on midpoints'?", "answer": ["Midpoint Regularization: from High Uncertainty Training to Conservative Classification", "Towards Understanding the Data Dependency of Mixup-style Training"], "answer_arxiv_id": ["2106.13913", "2110.07647"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_5454"} +{"question": "Which papers propose a straightforward approach to align visual and textual latent representations in training VLMs?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_5455"} +{"question": "What studies propose novel algorithms for resolving the Mean-Field Game problem?", "answer": ["Deep Generalized Schrödinger Bridge", "Alternating the Population and Control Neural Networks to Solve High-Dimensional Stochastic Mean-Field Games"], "answer_arxiv_id": ["2209.09893", "2002.10113v4"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_5456"} +{"question": "Which papers provide the standard and commonly-used assumptions in reinforcement learning theory?", "answer": ["Reinforcement Learning with General Value Function Approximation: Provably Efficient Approach via Bounded Eluder Dimension", "Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms", "A General Framework for Sample-Efficient Function Approximation in Reinforcement Learning", "The Role of Coverage in Online Reinforcement Learning"], "answer_arxiv_id": ["2005.10804", "2102.00815", "2209.15634", "2210.04157"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_5457"} +{"question": "Are there any studies on AI alignment research that teaches LLMs to follow human instructions?", "answer": ["Training language models to follow instructions with human feedback", "Visual Instruction Tuning"], "answer_arxiv_id": ["2203.02155", "2304.08485"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_5458"} +{"question": "Can you list some studies that explore training of a foundational video generator on limited resources?", "answer": ["VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation", "Swap Attention in Spatiotemporal Diffusions for Text-to-Video Generation", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models"], "answer_arxiv_id": ["2303.08320", "2305.10874", "2307.04725", "2304.08818"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_5459"} +{"question": "Which papers reviewed the adaptation of Ewald summation to encompass more general interactions?", "answer": ["Ewald methods for inverse power-law interactions in tridimensional and quasi-two dimensional systems."], "answer_arxiv_id": ["1009.1255"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_5460"} +{"question": "Which work proposed using a discriminator network with the decoder of VAE acting as generator to improve the visual fidelity of the generated images?", "answer": ["Autoencoding beyond pixels using a learned similarity metric"], "answer_arxiv_id": ["1512.09300v2"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_5461"} +{"question": "Are there studies that have been conducted on 'Early-Exit Neural Networks?'", "answer": ["BranchyNet: Fast Inference via Early Exiting from Deep Neural Networks", "Resolution Adaptive Networks for Efficient Inference", "Dynamic Neural Networks: A Survey", "Multi-Scale Dense Networks for Resource Efficient Image Classification", "Shallow-Deep Networks: Understanding and Mitigating Network Overthinking", "The Right Tool for the Job: Matching Model and Instance Complexities", "Consistent Accelerated Inference via Confident Adaptive Transformers", "A Survey on Dynamic Neural Networks for Natural Language Processing", "Not All Images are Worth 16x16 Words: Dynamic Transformers for Efficient Image Recognition", "Zero Time Waste: Recycling Predictions in Early Exit Neural Networks"], "answer_arxiv_id": ["1709.01686", "2003.07326", "2102.04906", "1703.09844", "1810.07052", "2004.07453", "2104.08803", "2202.07101", "2105.15075", "2106.05409"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_5462"} +{"question": "What works have shown the use of gradient attack in jailbreaking LLMs?", "answer": ["Universal and Transferable Adversarial Attacks on Aligned Language\n Models"], "answer_arxiv_id": ["2307.15043"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_5463"} +{"question": "What studies describe VG-boosting training-schemes involving data augmentation in image or feature space?", "answer": ["MUTANT: A Training Paradigm for Out-of-Distribution Generalization in\n Visual Question Answering", "SwapMix: Diagnosing and Regularizing the Over-Reliance on Visual Context\n in Visual Question Answering"], "answer_arxiv_id": ["2009.08566", "2204.02285"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_5464"} +{"question": "Which works study the out-of-distribution generalization in the context of deep neural networks?", "answer": ["Towards Out-Of-Distribution Generalization: A Survey"], "answer_arxiv_id": ["2108.13624v2"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_5465"} +{"question": "What papers utilized siamese networks applied in the contrastive self-supervised learning methods?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance-level\n Discrimination", "A Simple Framework for Contrastive Learning of Visual Representations", "Big Self-Supervised Models are Strong Semi-Supervised Learners", "Emerging Properties in Self-Supervised Vision Transformers", "Momentum Contrast for Unsupervised Visual Representation Learning", "Exploring Simple Siamese Representation Learning", "Masked Siamese Networks for Label-Efficient Learning", "Siamese Image Modeling for Self-Supervised Vision Representation\n Learning", "Siamese Masked Autoencoders"], "answer_arxiv_id": ["1805.01978", "2002.05709", "2006.10029", "2104.14294", "1911.05722", "2011.10566", "2204.07141", "2206.01204", "2305.14344"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_5466"} +{"question": "What was the first deep learning method for semantic matching?", "answer": ["SCNet: Learning Semantic Correspondence"], "answer_arxiv_id": ["1705.04043"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_5467"} +{"question": "What papers have discussed the use of conditional GANs in generative models for navigating prediction uncertainty?", "answer": ["Conditional Generative Adversarial Nets"], "answer_arxiv_id": ["1411.1784"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_5468"} +{"question": "Which works studied the topic of voice conversion in the area of speech generative models?", "answer": ["The Voice Conversion Challenge 2018: Promoting Development of Parallel and Nonparallel Methods"], "answer_arxiv_id": ["1804.04262"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_5469"} +{"question": "What research proposed prompt-tuning to improve task performance in the continuous space?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2104.08691"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_5470"} +{"question": "Could you provide me a research paper that shows diffusion models implement the optimal transport from the data to the target distribution?", "answer": ["Understanding DDPM Latent Codes Through Optimal Transport"], "answer_arxiv_id": ["2202.07477"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5471"} +{"question": "What research proposes Spectral Attention Networks (SAN) that utilise two attention mechanism?", "answer": ["Rethinking Graph Transformers with Spectral Attention"], "answer_arxiv_id": ["2106.03893"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_5472"} +{"question": "Which works established connections between online learning and multigroup fairness notions?", "answer": ["Online Multivalid Learning: Means, Moments, and Prediction Intervals", "Advancing subgroup fairness via sleeping experts", "Online Learning with an Unknown Fairness Metric"], "answer_arxiv_id": ["2101.01739", "1909.08375", "1802.06936"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_5473"} +{"question": "Could you provide me with some references that belong to the multi-priors fusion category in user-generated content video quality assessment?", "answer": ["A Deep Learning based No-reference Quality Assessment Model for UGC\n Videos", "Blindly Assess Quality of In-the-Wild Videos via Quality-aware\n Pre-training and Motion Perception", "MD-VQA: Multi-Dimensional Quality Assessment for UGC Live Videos", "Ada-DQA: Adaptive Diverse Quality-aware Feature Acquisition for Video\n Quality Assessment"], "answer_arxiv_id": ["2204.14047", "2108.08505", "2303.14933", "2308.00729"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_5474"} +{"question": "What work utilized 360 degree panoramic cameras to capture gaze targets and human faces simultaneously?", "answer": ["Gaze360: Physically Unconstrained Gaze Estimation in the Wild"], "answer_arxiv_id": ["1910.10088"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_5475"} +{"question": "Which papers focus on weighted adversarial training methods that assign larger weights to more vulnerable points closer to the decision boundary?", "answer": ["Are Adversarial Examples Created Equal? A Learnable Weighted Minimax Risk for Robustness under Non-uniform Attacks", "Probabilistic Margins for Instance Reweighting in Adversarial Training", "Geometry-aware Instance-reweighted Adversarial Training"], "answer_arxiv_id": ["2010.12989", "2106.07904", "2010.01736"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5476"} +{"question": "What works focus on a data-driven induction of grammars for molecule generation?", "answer": ["Data-Efficient Graph Grammar Learning for Molecular Generation", "Molecular Hypergraph Grammar with Its Application to Molecular Optimization"], "answer_arxiv_id": ["2203.08031", "1809.02745"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5477"} +{"question": "Which works investigate the benefits of large learning rates to generalization and predict maximum learning rates?", "answer": ["The large learning rate phase of deep learning: the catapult mechanism"], "answer_arxiv_id": ["2003.02218"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_5478"} +{"question": "What research studied prompt-specified video segmentation tasks?", "answer": ["YouTube-VOS: Sequence-to-Sequence Video Object Segmentation", "MOSE: A New Dataset for Video Object Segmentation in Complex Scenes", "BURST: A Benchmark for Unifying Object Recognition, Segmentation and\n Tracking in Video", "Video Object Segmentation in Panoptic Wild Scenes"], "answer_arxiv_id": ["1809.00461", "2302.01872", "2209.12118", "2305.04470"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_5479"} +{"question": "Could you provide me some studies about different methods to fuse information in both early and late fusion?", "answer": ["Multimodal Machine Learning: A Survey and Taxonomy"], "answer_arxiv_id": ["1705.09406"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_5480"} +{"question": "What papers have made attempts to provide upstream data like range-azimuth (RA) maps?", "answer": ["RADIATE: A Radar Dataset for Automotive Perception in Bad Weather"], "answer_arxiv_id": ["2010.09076"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_5481"} +{"question": "Could you provide me some works about quality estimation for MT starting as confidence estimation?", "answer": ["OpenKiwi: An Open Source Framework for Quality Estimation", "TransQuest: Translation Quality Estimation with Cross-lingual\n Transformers", "Unbabel's Participation in the WMT20 Metrics Shared Task"], "answer_arxiv_id": ["1902.08646", "2011.01536", "2010.15535"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_5482"} +{"question": "Which papers used language as human-interpretable medium to explain classification decisions in computer vision?", "answer": ["Generating Visual Explanations", "Grounding Visual Explanations", "Multimodal Explanations: Justifying Decisions and Pointing to the Evidence", "From Recognition to Cognition: Visual Commonsense Reasoning", "Natural Language Descriptions of Deep Visual Features"], "answer_arxiv_id": ["1603.08507", "1807.09685", "1802.08129", "1811.10830", "2201.11114"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_5483"} +{"question": "What papers study the tradeoff between group fairness and accuracy in a centralized setting?", "answer": ["Inherent Tradeoffs in Learning Fair Representations"], "answer_arxiv_id": ["1906.08386"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_5484"} +{"question": "Which works considered ELBOs defined by the normalizing constant estimates from multiple importance sampling?", "answer": ["Importance Weighted Autoencoders"], "answer_arxiv_id": ["1509.00519"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_5485"} +{"question": "Could you give me examples of research that explores generating protective noise for images using adversarial perturbation methods?", "answer": ["DUAW: Data-free Universal Adversarial Watermark against Stable Diffusion\n Customization", "Improving Adversarial Attacks on Latent Diffusion Model", "Towards Prompt-robust Face Privacy Protection via Adversarial Decoupling\n Augmentation Framework", "Unlearnable Examples for Diffusion Models: Protect Data from\n Unauthorized Exploitation"], "answer_arxiv_id": ["2308.09889", "2310.04687", "2305.03980", "2306.01902"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5486"} +{"question": "What papers focus on mathematical programming (MP) based NAS methods?", "answer": ["DeepMAD: Mathematical Architecture Design for Deep Convolutional Neural Network", "LayerNAS: Neural Architecture Search in Polynomial Complexity"], "answer_arxiv_id": ["2303.02165v3", "2304.11517"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_5487"} +{"question": "Could you provide me with studies that use Region of Interest to extract object-level features in LVLMs?", "answer": ["Position-Enhanced Visual Instruction Tuning for Multimodal Large\n Language Models", "GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest", "Mask R-CNN"], "answer_arxiv_id": ["2308.13437", "2307.03601", "1703.06870"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_5488"} +{"question": "Which works utilized high-quality dense view images or accurate annotations like human body keypoints and foreground segmentation for neural rendering techniques?", "answer": ["HUMBI: A Large Multiview Dataset of Human Body Expressions", "Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "Structured Local Radiance Fields for Human Avatar Modeling"], "answer_arxiv_id": ["1812.00281", "2012.15838", "2203.14478"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_5489"} +{"question": "Can you name the papers which encourage fair predictions on unlabeled data through distribution alignment?", "answer": ["AdaMatch: A Unified Approach to Semi-Supervised Learning and Domain Adaptation"], "answer_arxiv_id": ["2106.04732"], "source_meta": {"published_time": "20220515"}, "qid": "AutoScholarQuery_train_5490"} +{"question": "Which studies used loss rebalancing to focus more on the tail categories in long-tail recognition?", "answer": ["Cost Sensitive Learning of Deep Feature Representations from Imbalanced\n Data", "Focal Loss for Dense Object Detection", "Class-Balanced Loss Based on Effective Number of Samples", "Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss", "Intriguing properties of adversarial training at scale", "Balanced Product of Calibrated Experts for Long-Tailed Recognition", "Long-tail learning via logit adjustment", "Large-Scale Long-Tailed Recognition in an Open World", "Contrastive Learning based Hybrid Networks for Long-Tailed Image\n Classification"], "answer_arxiv_id": ["1508.03422", "1708.02002", "1901.05555", "1906.07413", "1906.03787", "2206.05260", "2007.07314", "1904.05160", "2103.14267"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_5491"} +{"question": "What researches have attempted to utilize dynamic encoding or physical mechanisms in scientific machine learning?", "answer": ["DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators", "Scientific Machine Learning through Physics-Informed Neural Networks: Where we are and What’s next", "Characterizing possible failure modes in physics-informed neural networks"], "answer_arxiv_id": ["1910.03193", "2201.05624", "2109.01050"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_5492"} +{"question": "Which research papers proved the existence of double descent curves using forms of linear regression?", "answer": ["Benign Overfitting in Linear Regression", "Benign overfitting in ridge regression", "Surprises in High-Dimensional Ridgeless Least Squares Interpolation", "Harmless interpolation of noisy data in regression"], "answer_arxiv_id": ["1906.11300", "2009.14286", "1903.08560", "1903.09139"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_5493"} +{"question": "What studies focus on active mapping by optimizing NeRFs with next-best-view selection strategies?", "answer": ["Uncertainty Guided Policy for Active Robotic 3D Reconstruction using\n Neural Radiance Fields", "ActiveNeRF: Learning where to See with Uncertainty Estimation"], "answer_arxiv_id": ["2209.08409", "2209.08546"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_5494"} +{"question": "What works approach the pre-trained teacher model as a discriminator and adversarially train the generator?", "answer": ["Contrastive Model Inversion for Data-Free Knowledge Distillation", "Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion", "Data-Free Learning of Student Networks", "Data-Free Knowledge Amalgamation via Group-Stack Dual-GAN", "Generative Low-bitwidth Data Free Quantization"], "answer_arxiv_id": ["2105.08584", "1912.08795", "1904.01186", "2003.09088", "2003.03603"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_5495"} +{"question": "What work also build upon GRAND but instead using additional Beltrami features?", "answer": ["Beltrami Flow and Neural Diffusion on Graphs"], "answer_arxiv_id": ["2110.09443"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_5496"} +{"question": "Which works have shown interest in using Large Language Models (LLMs) for developing autonomous agents?", "answer": ["The Rise and Potential of Large Language Model Based Agents: A Survey", "A Survey on Large Language Model based Autonomous Agents"], "answer_arxiv_id": ["2309.07864v3", "2308.11432"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_5497"} +{"question": "What approach do Neural Module Networks represent in Differentiable Computing?", "answer": ["Neural Module Networks"], "answer_arxiv_id": ["1511.02799v4"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5498"} +{"question": "Which papers utilized GPs with deep kernels for task-specific inference in Bayesian meta-learning?", "answer": ["Bayesian Meta-Learning for the Few-Shot Setting via Deep Kernels", "Bayesian Few-Shot Classification with One-vs-Each Pólya-Gamma Augmented Gaussian Processes"], "answer_arxiv_id": ["1910.05199", "2007.10417v2"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_5499"} +{"question": "Which studies focus on 3D molecule generation?", "answer": ["Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules", "Equivariant Diffusion for Molecule Generation in 3D", "Diffusion-based Molecule Generationwith Informative Prior Bridges", "Geometric Latent Diffusion Models for 3D Molecule Generation"], "answer_arxiv_id": ["1906.00957", "2203.17003", "2209.00865", "2305.01140"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_5500"} +{"question": "Can you list the references that proposed algorithms to efficiently approximate Schrödinger Bridges in the context of machine learning?", "answer": ["Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling", "Solving Schrödinger Bridges via Maximum Likelihood", "Likelihood Training of Schrödinger Bridge using Forward-Backward SDEs Theory"], "answer_arxiv_id": ["2106.01357", "2106.02081", "2110.11291"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_5501"} +{"question": "Which works suggest forming a new update vector by linearly combining task gradients?", "answer": ["Multi-Task Learning as Multi-Objective Optimization", "Gradient Surgery for Multi-Task Learning", "Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout", "Conflict-Averse Gradient Descent for Multi-task Learning", "Multi-Task Learning as a Bargaining Game", "Auto-Lambda: Disentangling Dynamic Task Relationships"], "answer_arxiv_id": ["1810.04650", "2001.06782", "2010.06808", "2110.14048", "2202.01017", "2202.03091"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_5502"} +{"question": "Which works have used hierarchical RL, goal-based RL, and reward shaping in training an agent for video game Minecraft?", "answer": ["MineRL: A Large-Scale Dataset of Minecraft Demonstrations", "MineDojo: Building Open-Ended Embodied Agents with Internet-Scale\n Knowledge", "Open-World Multi-Task Control Through Goal-Aware Representation Learning\n and Adaptive Horizon Prediction", "JueWu-MC: Playing Minecraft with Sample-efficient Hierarchical\n Reinforcement Learning", "CLIP4MC: An RL-Friendly Vision-Language Model for Minecraft", "Zero-Shot Task Generalization with Multi-Task Deep Reinforcement\n Learning"], "answer_arxiv_id": ["1907.13440", "2206.08853", "2301.10034", "2112.04907", "2303.10571", "1706.05064"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_5503"} +{"question": "Which paper proposed to evaluate neural estimators based on variational lower bounds on image data?", "answer": ["Understanding the Limitations of Variational Mutual Information Estimators"], "answer_arxiv_id": ["1910.06222"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_5504"} +{"question": "Which studies have conditioned recent diffusion models with brain activations to reconstruct observed images?", "answer": ["Seeing Beyond the Brain: Conditional Diffusion Model with Sparse Masked Modeling for Vision Decoding", "MindDiffuser: Controlled Image Reconstruction from Human Brain Activity with Semantic and Structural Diffusion", "Second Sight: Using brain-optimized encoding models to align image distributions with human brain activity"], "answer_arxiv_id": ["2211.06956", "2303.14139", "2306.00927"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_5505"} +{"question": "Which work highlighted the significance of implicit learning curricula and the suboptimality of the adversarial used in previous work?", "answer": ["Challenges of Adversarial Image Augmentations"], "answer_arxiv_id": ["2111.12427"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_5506"} +{"question": "Which research proposes Jacobian regularization for DEQ models to improve training stability instead of robustness?", "answer": ["Stabilizing Equilibrium Models by Jacobian Regularization"], "answer_arxiv_id": ["2106.14342"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_5507"} +{"question": "What papers introduce the white-box attack method via utilizing some loss gradient with respect to the model inputs?", "answer": ["Explaining and Harnessing Adversarial Examples", "Adversarial Machine Learning at Scale", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1412.6572", "1611.01236", "1706.06083"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_5508"} +{"question": "Which study developed a method that adaptively changes the weight decay hyper-parameter for each individual parameter?", "answer": ["Adaptive Weight Decay for Deep Neural Networks"], "answer_arxiv_id": ["1907.08931"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5509"} +{"question": "Can you provide references that discovered the double descent effects in state-of-the-art neural networks?", "answer": ["Deep Double Descent: Where Bigger Models and More Data Hurt", "A jamming transition from under- to over-parametrization affects generalization in deep learning", "Scaling description of generalization with number of parameters in deep learning"], "answer_arxiv_id": ["1912.02292", "1810.09665", "1901.01608"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_5510"} +{"question": "Which works applied the use of classifier-free guidance that trades off diversity and quality?", "answer": ["Classifier-Free Diffusion Guidance", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2207.12598", "2105.05233"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_5511"} +{"question": "Can you mention any research that has proposed possible solutions to deal with variations in resolution in Vision Transformers?", "answer": ["Pix2Struct: Screenshot Parsing as Pretraining for Visual Language\n Understanding", "FlexiViT: One Model for All Patch Sizes", "Once-for-All: Train One Network and Specialize it for Efficient\n Deployment", "BigNAS: Scaling Up Neural Architecture Search with Big Single-Stage\n Models", "Super Vision Transformer", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2210.03347", "2212.08013", "1908.09791", "2003.11142", "2205.11397", "2111.06377"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_5512"} +{"question": "Who proposed the diversity-guided searching strategy to select diverse demonstrations?", "answer": ["Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2210.03493"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_5513"} +{"question": "What works address bias in supervision through methods like mutual learning and distribution alignment?", "answer": ["DMT: Dynamic Mutual Training for Semi-Supervised Learning", "Repetitive Reprediction Deep Decipher for Semi-Supervised Learning", "Robust Mutual Learning for Semi-supervised Semantic Segmentation", "Re-distributing Biased Pseudo Labels for Semi-supervised Semantic Segmentation: A Baseline Investigation", "Semi-supervised Semantic Segmentation with Error Localization Network"], "answer_arxiv_id": ["2004.08514", "1908.04345", "2106.00609", "2107.11279", "2204.02078"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5514"} +{"question": "Could you provide me some studies about incremental learning algorithms using knowledge distillation?", "answer": ["Learning without Forgetting"], "answer_arxiv_id": ["1606.09282"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_5515"} +{"question": "What studies have used Tensor Product Representations in synthetic sequence tasks?", "answer": ["Discovering the Compositional Structure of Vector Representations with Role Learning Networks"], "answer_arxiv_id": ["1910.09113"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5516"} +{"question": "Which papers introduce language features to make vision classifiers more robust?", "answer": ["On Guiding Visual Attention with Language Specification", "Diagnosing and Rectifying Vision Models using Language"], "answer_arxiv_id": ["2202.08926", "2302.04269"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5517"} +{"question": "Could you provide me some studies that incorporated deep RL with human preferences?", "answer": ["Deep Reinforcement Learning from Human Preferences"], "answer_arxiv_id": ["1706.03741"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_5518"} +{"question": "Which studies have aimed to improve the adversarial robustness by dealing with high-frequency components?", "answer": ["Bandlimiting Neural Networks Against Adversarial Attacks", "High-frequency Component Helps Explain the Generalization of Convolutional Neural Networks"], "answer_arxiv_id": ["1905.12797", "1905.13545"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_5519"} +{"question": "What works provide uncertainty-based curriculum guidance using absolute reward difference?", "answer": ["Teacher algorithms for curriculum learning of Deep RL in continuously parameterized environments"], "answer_arxiv_id": ["1910.07224"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_5520"} +{"question": "Which publication introduces a two-level variational autoencoder for learning the overall shape structure and detailed part geometries?", "answer": ["SDM-NET: Deep Generative Network for Structured Deformable Mesh"], "answer_arxiv_id": ["1908.04520"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_5521"} +{"question": "Can you suggest some papers where cross-entropy is utilized in node classification or graph classification tasks?", "answer": ["Hyperbolic Graph Convolutional Neural Networks", "Hyperbolic Graph Neural Networks"], "answer_arxiv_id": ["1910.12933", "1910.12892"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_5522"} +{"question": "What is the work that applies an additional distillation loss for Experience Replay (ER) by storing the output logits with the samples?", "answer": ["Dark Experience for General Continual Learning: a Strong, Simple Baseline"], "answer_arxiv_id": ["2004.07211"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_5523"} +{"question": "What paper generalized the principle of selecting samples that have the maximum concensus with other samples in the execution results when handling code generation?", "answer": ["Natural Language to Code Translation with Execution"], "answer_arxiv_id": ["2204.11454"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_5524"} +{"question": "What works focus on synthesizing motions for text-to-video models by using separate training modules?", "answer": ["AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "MotionDirector: Motion Customization of Text-to-Video Diffusion Models"], "answer_arxiv_id": ["2307.04725", "2310.08465"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_5525"} +{"question": "What works have been conducted on sketch for visual understanding, including sketch-based applications in image editing and 3D shape generation?", "answer": ["SketchEdit: Mask-Free Local Image Manipulation with Partial Sketches", "DeFLOCNet: Deep Image Editing via Flexible Low-level Controls", "Free-Form Image Inpainting with Gated Convolution", "SketchyGAN: Towards Diverse and Realistic Sketch to Image Synthesis", "SketchyCOCO: Image Generation from Freehand Scene Sketches", "Sketch Your Own GAN", "Deep Generation of Face Images from Sketches", "Sketch2Mesh: Reconstructing and Editing 3D Shapes from Sketches", "Sketch2Model: View-Aware 3D Modeling from Single Free-Hand Sketches", "Doodle Your 3D: From Abstract Freehand Sketches to Precise 3D Shapes"], "answer_arxiv_id": ["2111.15078", "2103.12723", "1806.03589", "1801.02753", "2003.02683", "2108.02774", "2006.01047", "2104.00482", "2105.06663", "2312.04043"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_5526"} +{"question": "What works have derived MDL-C from the RL-as-inference framework?", "answer": ["Reinforcement Learning and Control as Probabilistic Inference: Tutorial and Review"], "answer_arxiv_id": ["1805.00909"], "source_meta": {"published_time": "20220717"}, "qid": "AutoScholarQuery_train_5527"} +{"question": "Which works analyze methods to interpret graph neural networks applied to geometric data?", "answer": ["GNNExplainer: Generating Explanations for Graph Neural Networks", "Parameterized Explainer for Graph Neural Network", "Interpreting Graph Neural Networks for NLP With Differentiable Edge Masking", "On Explainability of Graph Neural Networks via Subgraph Explorations"], "answer_arxiv_id": ["1903.03894", "2011.04573", "2010.00577", "2102.05152"], "source_meta": {"published_time": "20221030"}, "qid": "AutoScholarQuery_train_5528"} +{"question": "Which studies outline the utilization of implicit surfaces to produce complex non-linear deformations of 3D bodies?", "answer": ["NPMs: Neural Parametric Models for 3D Deformable Shapes", "gDNA: Towards Generative Detailed Neural Avatars", "NPMs: Neural Parametric Models for 3D Deformable Shapes"], "answer_arxiv_id": ["2104.00702", "2201.04123", "2104.00702"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_5529"} +{"question": "Which studies focus on sequence length reduction in transformers as a method of model compression from the width perspective?", "answer": ["How Contextual are Contextualized Word Representations? Comparing the\n Geometry of BERT, ELMo, and GPT-2 Embeddings", "Spying on your neighbors: Fine-grained probing of contextual embeddings\n for information about surrounding words"], "answer_arxiv_id": ["1909.00512", "2005.01810"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_5530"} +{"question": "What studies discuss the use of CLIP adapters for compositional zero-shot learning?", "answer": ["CLIP-Adapter: Better Vision-Language Models with Feature Adapters"], "answer_arxiv_id": ["2110.04544"], "source_meta": {"published_time": "20220407"}, "qid": "AutoScholarQuery_train_5531"} +{"question": "Which works proposed to leverage pretrained image diffusion models in 3D generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_5532"} +{"question": "What work originally established the Stable Diffusion model for text-to-image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_5533"} +{"question": "Could you give me examples of studies of equivariant models using atom embeddings in spaces of much higher dimension?", "answer": ["3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data", "Vector Neurons: A General Framework for SO(3)-Equivariant Networks", "Frame Averaging for Invariant and Equivariant Network Design"], "answer_arxiv_id": ["1807.02547", "2104.12229", "2110.03336"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_5534"} +{"question": "Which studies examined the task of goal relabeling in the area of in-domain representation learning?", "answer": ["Manipulator-Independent Representations for Visual Imitation"], "answer_arxiv_id": ["2103.09016"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_5535"} +{"question": "Which work illustrates a way to achieve a significantly faster speed in practice by reducing the number of accesses to High Bandwidth Memory?", "answer": ["FlashAttention: Fast and Memory-Efficient Exact Attention with\n IO-Awareness"], "answer_arxiv_id": ["2205.14135"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_5536"} +{"question": "Which papers are focused on test-time adaptation/training to overcome performance degradation due to distribution shifts?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "Tent: Fully Test-Time Adaptation by Entropy Minimization"], "answer_arxiv_id": ["1909.13231", "2006.10726"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_5537"} +{"question": "What papers demonstrated the excellent transferability and high performance of self-supervised Vision Transformers?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "An Empirical Study of Training Self-Supervised Vision Transformers", "Emerging Properties in Self-Supervised Vision Transformers", "BEiT: BERT Pre-Training of Image Transformers", "iBOT : Image BERT Pre-Training with Online Tokenizer"], "answer_arxiv_id": ["2111.06377", "2104.02057", "2104.14294", "2106.08254", "2111.07832"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_5538"} +{"question": "What papers introduced adapter-based tuning in PEFT?", "answer": ["Parameter-Efficient Transfer Learning for NLP"], "answer_arxiv_id": ["1902.00751"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_5539"} +{"question": "What papers describe the use of data augmentation in developing solutions for detecting image forgery in deepfake detection?", "answer": ["Exposing DeepFake Videos By Detecting Face Warping Artifacts", "Face X-ray for More General Face Forgery Detection", "Learning Self-Consistency for Deepfake Detection", "Self-supervised Learning of Adversarial Example: Towards Good\n Generalizations for Deepfake Detection", "Detecting Deepfakes with Self-Blended Images"], "answer_arxiv_id": ["1811.00656", "1912.13458", "2012.09311", "2203.12208", "2204.08376"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_5540"} +{"question": "Which works have conducted experimental investigations into other architectural features such as layers and memory?", "answer": ["General-Purpose In-Context Learning by Meta-Learning Transformers"], "answer_arxiv_id": ["2212.04458"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_5541"} +{"question": "What studies have begun to study dense Out-of-distribution (OOD) detection?", "answer": ["The Fishyscapes Benchmark: Measuring Blind Spots in Semantic Segmentation", "Scaling Out-of-Distribution Detection for Real-World Settings", "SegmentMeIfYouCan: A Benchmark for Anomaly Segmentation"], "answer_arxiv_id": ["1904.03215", "1911.11132", "2104.14812"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_5542"} +{"question": "Which research papers present a universal architecture capable of handling semantic, instance, and panoptic segmentation for close-set categories?", "answer": ["Masked-attention Mask Transformer for Universal Image Segmentation", "Mask DINO: Towards A Unified Transformer-based Framework for Object\n Detection and Segmentation"], "answer_arxiv_id": ["2112.01527", "2206.02777"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_5543"} +{"question": "Which works showed that FFN layers in transformers gradually build predictions by promoting concepts that are interpretable in the vocabulary space?", "answer": ["Transformer Feed-Forward Layers Are Key-Value Memories", "Transformer Feed-Forward Layers Build Predictions by Promoting Concepts\n in the Vocabulary Space"], "answer_arxiv_id": ["2012.14913", "2203.14680"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_5544"} +{"question": "What are some of the domain-specific models for legal tasks?", "answer": ["LEGAL-BERT: The Muppets straight out of Law School"], "answer_arxiv_id": ["2010.02559"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_5545"} +{"question": "In what papers the researcher focused on diffusion-based framework for sketch-to-image (S2I) generation?", "answer": ["Pretraining is All You Need for Image-to-Image Translation", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Sketch-Guided Text-to-Image Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2205.12952", "2108.01073", "2211.13752", "2302.05543", "2302.08453"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5546"} +{"question": "Could you tell me about the research that showed all tokens quickly lump together into a single tight cluster without skip connections?", "answer": ["Attention is not all you need: pure attention loses rank doubly exponentially with depth"], "answer_arxiv_id": ["2103.03404"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_5547"} +{"question": "Which studies used diffusion models to transform random samples into data resembling a target?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Variational Diffusion Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2107.00630", "2105.05233"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5548"} +{"question": "Are there any studies that used neural networks for differentially rendering images from 3D scene representation in autonomous driving?", "answer": ["TensoRF: Tensorial Radiance Fields", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction", "NeRF-Det: Learning Geometry-Aware Volumetric Representation for\n Multi-View 3D Object Detection", "Point-NeRF: Point-based Neural Radiance Fields", "UniSim: A Neural Closed-Loop Sensor Simulator"], "answer_arxiv_id": ["2203.09517", "2003.08934", "2104.10078", "2307.14620", "2201.08845", "2308.01898"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_5549"} +{"question": "Which work proposed OTOv1, a method proposed to avoid fine-tuning and perform end-to-end training and compression of the DNN once?", "answer": ["Only Train Once: A One-Shot Neural Network Training And Pruning Framework"], "answer_arxiv_id": ["2107.07467"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_5550"} +{"question": "Could you provide me some studies about utilizing non-robotics datasets in robot learning?", "answer": ["Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2110.07058"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_5551"} +{"question": "Which works applied contrastive learning on convolutional networks and Vision transformers?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "An Empirical Study of Training Self-Supervised Vision Transformers", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["1911.05722", "2003.04297", "2006.07733", "2006.09882", "2104.02057", "2104.14294"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_5552"} +{"question": "Which research papers adjusted the structure of the stacked hourglass network in PSMNet and creatively built the cost volume using group-wise correlation?", "answer": ["Group-wise Correlation Stereo Network"], "answer_arxiv_id": ["1903.04025"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_5553"} +{"question": "Which studies proposed representation learning methods based on VAE with modification?", "answer": ["Q", "Disentangling by Factorising", "Information Dropout: Learning Optimal Representations Through Noisy Computation", "Variational Inference of Disentangled Latent Concepts from Unlabeled Observations", "Learning Hierarchical Features from Generative Models", "Ladder Variational Autoencoders", "InfoVAE: Balancing Learning and Inference in Variational Autoencoders", "Improving Variational Autoencoder with Deep Feature Consistent and Generative Adversarial Training", "Explicit Disentanglement of Appearance and Perspective in Generative Models", "Guided Variational Autoencoder for Disentanglement Learning"], "answer_arxiv_id": ["1611.08152", "1802.05983", "1611.01353", "1711.00848", "1702.08396", "1602.02282", "1706.02262", "1906.01984", "1906.11881", "2004.01255"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_5554"} +{"question": "What studies have made efforts above contrastive methods to improve pre-training quality for specific downstream tasks?", "answer": ["DetCo: Unsupervised Contrastive Learning for Object Detection", "Region Similarity Representation Learning", "CASTing Your Model: Learning to Localize Improves Self-Supervised Representations"], "answer_arxiv_id": ["2102.04803", "2103.12902", "2012.04630"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_5555"} +{"question": "Which papers explore the use of deep Gaussian Processes (deep gps)?", "answer": ["Deep Gaussian Processes"], "answer_arxiv_id": ["1211.0358"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5556"} +{"question": "Could you name the studies that use knowledge distillation or regularization strategies to maintain the representations of old classes in many-shot CIL methods?", "answer": ["Distilling the Knowledge in a Neural Network", "Overcoming catastrophic forgetting in neural networks", "Semantic Drift Compensation for Class-Incremental Learning"], "answer_arxiv_id": ["1503.02531", "1612.00796", "2004.00440"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_5557"} +{"question": "Could you provide me some studies about using inference models to assist in modern program synthesis?", "answer": ["Neural Program Meta-Induction"], "answer_arxiv_id": ["1710.04157"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_5558"} +{"question": "Which research demonstrated the improvement of open-domain multimodal tasks through knowledge augmentation from knowledge bases and GPT-3?", "answer": ["KAT: A Knowledge Augmented Transformer for Vision-and-Language"], "answer_arxiv_id": ["2112.08614"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_5559"} +{"question": "Which works suggest that the robustness of a network is related to its spatial-frequency preferences?", "answer": ["High-frequency Component Helps Explain the Generalization of Convolutional Neural Networks", "A Fourier Perspective on Model Robustness in Computer Vision", "An Extended Study of Human-like Behavior under Adversarial Training"], "answer_arxiv_id": ["1905.13545", "1906.08988", "2303.12669"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_5560"} +{"question": "Which papers uncover the phenomenon wherein SGD exhibits a benign underfitting, where the test loss remains small regardless of a large empirical loss?", "answer": ["Benign Underfitting of Stochastic Gradient Descent"], "answer_arxiv_id": ["2202.13361"], "source_meta": {"published_time": "20220616"}, "qid": "AutoScholarQuery_train_5561"} +{"question": "Which research analysed the impact of decentralization as landscape-dependent noise on the convergence of D-SGD?", "answer": ["Loss Landscape Dependent Self-Adjusting Learning Rates in Decentralized Stochastic Gradient Descent"], "answer_arxiv_id": ["2112.01433"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_5562"} +{"question": "What works dig deeper into RR-based (constrained) minimax optimization algorithms for convex-concave problems?", "answer": ["Zeroth-Order Methods for Convex-Concave Minmax Problems: Applications to Decision-Dependent Risk Minimization"], "answer_arxiv_id": ["2106.09082"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_5563"} +{"question": "Which study proposes a dual-branch network for balance learning in the long-tailed classification challenge?", "answer": ["BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed\n Visual Recognition"], "answer_arxiv_id": ["1912.02413"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_5564"} +{"question": "What work proposed the fine-tuning method named DreamBooth for Stable Diffusion?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5565"} +{"question": "What research paper considered MDP with exogenous inputs?", "answer": ["Hindsight Learning for MDPs with Exogenous Inputs"], "answer_arxiv_id": ["2207.06272"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5566"} +{"question": "Can you list the works that have considered lp extensions, high dimensional variants, or improvements and applications of PSCO?", "answer": ["Differentially Private Algorithms for Graphs Under Continual Observation", "Practical and Private (Deep) Learning Without Sampling or Shuffling", "Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams", "The Price of Differential Privacy under Continual Observation", "Differentially Private Histograms under Continual Observation: Streaming Selection into the Unknown", "Differentially Private Continual Releases of Streaming Frequency Moment Estimations", "Almost Tight Error Bounds on Differentially Private Continual Counting"], "answer_arxiv_id": ["2106.14756", "2103.00039", "2202.08312", "2112.00828", "2103.16787", "2301.05605", "2211.05006"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_5567"} +{"question": "What works select coresets by iteratively matching the (preconditioned) gradient of full training data?", "answer": ["Coresets for Data-efficient Training of Machine Learning Models", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training", "Adaptive Second Order Coresets for Data-efficient Machine Learning"], "answer_arxiv_id": ["1906.01827", "2103.00123", "2207.13887"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_5568"} +{"question": "What is the research that applied MAE in the audio domain?", "answer": ["Masked Autoencoders that Listen"], "answer_arxiv_id": ["2207.06405"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_5569"} +{"question": "Could you provide me some studies that showed the functional form, M2, holds over many orders of magnitude?", "answer": ["Deep Learning Scaling is Predictable, Empirically"], "answer_arxiv_id": ["1712.00409"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_5570"} +{"question": "What papers discuss the importance of integration methods in neural point process literature?", "answer": ["The Neural Hawkes Process: A Neurally Self-Modulating Multivariate Point Process"], "answer_arxiv_id": ["1612.09328"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_5571"} +{"question": "Could you provide me with some references discussing generalization and stable training using generative modeling techniques?", "answer": ["Taming Transformers for High-Resolution Image Synthesis", "VideoGPT: Video Generation using VQ-VAE and Transformers"], "answer_arxiv_id": ["2012.09841", "2104.10157"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_5572"} +{"question": "What studies proposed online learning algorithms to optimize dynamic regret measure under different kinds of feedback information and non-stationarity measures of environments?", "answer": ["Dynamical Models and Tracking Regret in Online Convex Programming", "Non-stationary Stochastic Optimization", "Online Optimization : Competing with Dynamic Comparators", "Tracking the Best Expert in Non-stationary Stochastic Environments", "Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient", "Dynamic Regret of Strongly Adaptive Methods", "A Simple Approach for Non-stationary Linear Bandits", "Adaptive Online Estimation of Piecewise Polynomial Trends", "Improved Analysis for Dynamic Regret of Strongly Convex and Smooth Functions", "Non-stationary Reinforcement Learning without Prior Knowledge: An Optimal Black-box Approach", "Corralling a Larger Band of Bandits: A Case Study on Switching Regret for Linear Bandits"], "answer_arxiv_id": ["1301.1254", "1307.5449", "1501.06225", "1712.00578", "1605.04638", "1701.07570", "2103.05324", "2010.00073", "2006.05876", "2102.05406", "2202.06151"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5573"} +{"question": "Which papers are associated with improving the quality of negatives by preserving a memory queue or generating high-quality negatives in contrastive learning in NLP?", "answer": ["Aligning Cross-lingual Sentence Representations with Dual Momentum Contrast", "Learning to Generalize to More: Continuous Semantic Augmentation for Neural Machine Translation"], "answer_arxiv_id": ["2109.00253", "2204.06812"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_5574"} +{"question": "What studies previously analyzed the case of TD with stochastic gradients and non-iid samples in the on-policy case?", "answer": ["A Finite Time Analysis of Temporal Difference Learning With Linear Function Approximation"], "answer_arxiv_id": ["1806.02450"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_5575"} +{"question": "Which work proposed the lottery ticket hypothesis?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks"], "answer_arxiv_id": ["1803.03635"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_5576"} +{"question": "Could you provide me studies that analyzed the sample complexity of optimal auction design problems?", "answer": ["The Sample Complexity of Revenue Maximization", "The Sample Complexity of Auctions with Side Information", "Sample Complexity of Automated Mechanism Design", "Settling the Sample Complexity of Single-parameter Revenue Maximization"], "answer_arxiv_id": ["1502.00963", "1511.02296v5", "1606.04145v1", "1904.04962"], "source_meta": {"published_time": "20230520"}, "qid": "AutoScholarQuery_train_5577"} +{"question": "Which studies in video grounding enumerate all segments and organize them in a 2D adjacency map to reason about their relations?", "answer": ["Learning 2D Temporal Adjacent Networks for Moment Localization with\n Natural Language", "VLG-Net: Video-Language Graph Matching Network for Video Grounding", "Relation-aware Video Reading Comprehension for Temporal Language\n Grounding"], "answer_arxiv_id": ["1912.03590", "2011.10132", "2110.05717"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_5578"} +{"question": "Which studies have proposed adding offset noise and employing v-prediction to tackle the average brightness issue in diffusion models?", "answer": ["Common Diffusion Noise Schedules and Sample Steps are Flawed", "Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2305.08891", "2202.00512"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_5579"} +{"question": "Are there any works that have optimized functional gradients in the space of probability distributions endowed with a Wasserstein structure?", "answer": ["Approximate inference with Wasserstein gradient flows", "Wasserstein Proximal of GANs"], "answer_arxiv_id": ["1806.04542v1", "2102.06862"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_5580"} +{"question": "What study presented a novel metric named 'RQUGE' for assessing the quality of automatically generated questions?", "answer": ["RQUGE: Reference-Free Metric for Evaluating Question Generation by\n Answering the Question"], "answer_arxiv_id": ["2211.01482"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_5581"} +{"question": "Can you provide me works which utilized the attention mechanism in computer vision?", "answer": ["Squeeze-and-Excitation Networks", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "CamoFormer: Masked Separable Attention for Camouflaged Object Detection", "Referring Camouflaged Object Detection"], "answer_arxiv_id": ["1709.01507", "2103.14030", "2212.06570", "2306.07532"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_5582"} +{"question": "Which paper introduced the concept of augmenting data-space neural ODEs with additional latent variables?", "answer": ["Augmented Neural ODEs"], "answer_arxiv_id": ["1904.01681"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_5583"} +{"question": "Could you provide me some researches about generating explanations with structured reasoning such as entailment trees?", "answer": ["Explaining Answers with Entailment Trees"], "answer_arxiv_id": ["2104.08661"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_5584"} +{"question": "What research papers look into budget-centric training recipes with hard time constraints such as training on a single GPU for one day?", "answer": ["nanoT5: A PyTorch Framework for Pre-training and Fine-tuning T5-style Models with Limited Resources"], "answer_arxiv_id": ["2309.02373"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_5585"} +{"question": "Which papers have studied strategic classification and how algorithmic decisions affect participants' feature changes?", "answer": ["Strategic Classification"], "answer_arxiv_id": ["1506.06980"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_5586"} +{"question": "What studies explored the utilization of language embeddings in the context of generalized zero-shot learning?", "answer": ["A Review of Generalized Zero-Shot Learning Methods"], "answer_arxiv_id": ["2011.08641"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_5587"} +{"question": "What papers have discussed about using pre-trained visual representations for control?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "R3M: A Universal Visual Representation for Robot Manipulation", "Masked Visual Pre-training for Motor Control", "VRL3: A Data-Driven Framework for Visual Deep Reinforcement Learning", "Visual Reinforcement Learning with Self-Supervised 3D Representations", "Pre-Trained Image Encoder for Generalizable Visual Reinforcement Learning", "On the Feasibility of Cross-Task Transfer with Model-Based Reinforcement Learning", "RT-1: Robotics Transformer for Real-World Control at Scale"], "answer_arxiv_id": ["2107.03380", "2203.03580", "2203.12601", "2203.06173", "2202.10324", "2210.07241", "2212.08860", "2210.10763", "2212.06817"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_5588"} +{"question": "Which works build a divide-and-conquer pipeline for multi-modal retrieval?", "answer": ["WebQA: Multihop and Multimodal QA", "Universal Vision-Language Dense Retrieval: Learning A Unified\n Representation Space for Multi-Modal Retrieval"], "answer_arxiv_id": ["2109.00590", "2209.00179"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_5589"} +{"question": "Which studies began to explore weakly supervised RIS by leveraging weak supervisory signals?", "answer": ["Shatter and Gather: Learning Referring Image Segmentation with Text\n Supervision", "Referring Image Segmentation Using Text Supervision"], "answer_arxiv_id": ["2308.15512", "2308.14575"], "source_meta": {"published_time": "20240418"}, "qid": "AutoScholarQuery_train_5590"} +{"question": "What work proposed personalized PCA to identify shared and unique principals in federated learning?", "answer": ["Personalized PCA: Decoupling Shared and Unique Features"], "answer_arxiv_id": ["2207.08041"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5591"} +{"question": "What research papers describe building ODE networks to learn conservation law in the dynamical system?", "answer": ["Neural Ordinary Differential Equations", "Augmented Neural ODEs", "Neural SDE: Stabilizing Neural ODE Networks with Stochastic Noise", "Time-Reversal Symmetric ODE Network"], "answer_arxiv_id": ["1806.07366", "1904.01681", "1906.02355", "2007.11362"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_5592"} +{"question": "Which research papers have proposed methods enabling the generation of high-resolution outputs?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "High-Resolution Image Synthesis with Latent Diffusion Models", "Latent Video Diffusion Models for High-Fidelity Long Video Generation", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "Latent-Shift: Latent Diffusion with Temporal Shift for Efficient\n Text-to-Video Generation", "Simple diffusion: End-to-end diffusion for high resolution images", "On the Importance of Noise Scheduling for Diffusion Models", "f-DM: A Multi-stage Diffusion Model via Progressive Signal\n Transformation", "Relay Diffusion: Unifying diffusion process across resolutions for image\n synthesis", "Matryoshka Diffusion Models"], "answer_arxiv_id": ["2210.02303", "2205.11487", "2211.01324", "2305.10474v3", "2209.14792", "2112.10752", "2211.13221", "2304.08818", "2211.11018", "2304.08477", "2301.11093", "2301.10972", "2210.04955", "2309.03350", "2310.15111"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_5593"} +{"question": "Could you provide me the works that utilized the Jacobian norm as a lower bound on the function norm in machine learning model training?", "answer": ["A Kernel Perspective for Regularizing Deep Neural Networks"], "answer_arxiv_id": ["1810.00363"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_5594"} +{"question": "What are some recent papers about Vision-Language (VL) supervised methods in computational pathology?", "answer": ["Towards a Visual-Language Foundation Model for Computational Pathology", "Visual Language Pretrained Multiple Instance Zero-Shot Transfer for\n Histopathology Images"], "answer_arxiv_id": ["2307.12914", "2306.07831"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_5595"} +{"question": "Are there any previous studies that showed how kernels can't adapt to the low-dimensional structure?", "answer": ["Linearized two-layers neural networks in high dimension"], "answer_arxiv_id": ["1904.12191"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_5596"} +{"question": "Which works introduced the original design for Variational Autoencoders (VAEs)?", "answer": ["Auto-Encoding Variational Bayes", "Stochastic Backpropagation and Approximate Inference in Deep Generative Models"], "answer_arxiv_id": ["1312.6114", "1401.4082v3"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_5597"} +{"question": "What studies have been conducted in designing mechanisms for data sharing focused on collaborative data sharing?", "answer": ["Collaborative Machine Learning with Incentive-Aware Model Rewards"], "answer_arxiv_id": ["2010.12797"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_5598"} +{"question": "Which work proposed a data distillation method based on Matching Training Trajectories?", "answer": ["Dataset Distillation by Matching Training Trajectories"], "answer_arxiv_id": ["2203.11932"], "source_meta": {"published_time": "20221119"}, "qid": "AutoScholarQuery_train_5599"} +{"question": "Could you provide me some literature about the methods to mitigate 'index collapse'?", "answer": ["Neural Discrete Representation Learning", "Robust Training of Vector Quantized Bottleneck Models", "SoundStream: An End-to-End Neural Audio Codec", "Jukebox: A Generative Model for Music", "Theory and Experiments on Vector Quantized Autoencoders", "SQ-VAE: Variational Bayes on Discrete Representation with Self-annealed Stochastic Quantization", "Self-labelling via simultaneous clustering and representation learning"], "answer_arxiv_id": ["1711.00937", "2005.08520", "2107.03312", "2005.00341", "1805.11063", "2205.07547", "1911.05371"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_5600"} +{"question": "Which works proposed parameter regularization-based methods in the field of class-incremental learning?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Task-Free Continual Learning", "Memory Aware Synapses: Learning what (not) to forget", "Continual Learning Through Synaptic Intelligence"], "answer_arxiv_id": ["1612.00796", "1812.03596", "1711.09601", "1703.04200"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_5601"} +{"question": "Which papers presented methods for merging models using Fisher-weighted averaging with the aim of improving performance on a single target task?", "answer": ["Merging Models with Fisher-Weighted Averaging"], "answer_arxiv_id": ["2111.09832"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_5602"} +{"question": "What was the first work to achieve full training in FP8 format?", "answer": ["Training Deep Neural Networks with 8-bit Floating Point Numbers"], "answer_arxiv_id": ["1812.08011"], "source_meta": {"published_time": "20211219"}, "qid": "AutoScholarQuery_train_5603"} +{"question": "What papers discuss the influence of model hyperparameters on performance?", "answer": ["Underspecification Presents Challenges for Credibility in Modern Machine Learning", "Predicting Neural Network Accuracy from Weights", "Classifying the classifier: dissecting the weight space of neural networks"], "answer_arxiv_id": ["2011.03395", "2002.11448", "2002.05688"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_5604"} +{"question": "Which works apply the numerical methods in order to accelerate the sampling process of diffusion models?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5605"} +{"question": "What studies explore the output of Transformer in relation to Gaussian process kernel and NTK formulation?", "answer": ["Infinite attention: NNGP and NTK for deep attention networks"], "answer_arxiv_id": ["2006.10540"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_5606"} +{"question": "What work describes a study that applied a cross-instance swap loss to help encourage multi-view consistency across object instances when training a reconstruction network?", "answer": ["Share With Thy Neighbors: Single-View Reconstruction by Cross-Instance\n Consistency"], "answer_arxiv_id": ["2204.10310"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_5607"} +{"question": "Are there any studies about the application of LLMs in web navigation?", "answer": ["Language Models can Solve Computer Tasks", "BOLAA: Benchmarking and Orchestrating LLM-augmented Autonomous Agents"], "answer_arxiv_id": ["2303.17491", "2308.05960"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_5608"} +{"question": "Which publications focus on deep sequential LVMs for different applications like speech modeling, and video compression to prediction and planning?", "answer": ["A Recurrent Latent Variable Model for Sequential Data", "Disentangled Sequential Autoencoder", "A General Method for Amortizing Variational Filtering", "Learning Latent Dynamics for Planning from Pixels", "Stochastic Variational Video Prediction", "Deep Variational Bayes Filters: Unsupervised Learning of State Space Models from Raw Data", "Clockwork Variational Autoencoders"], "answer_arxiv_id": ["1506.02216", "1803.02991", "1811.05090", "1811.04551", "1710.11252", "1605.06432", "2102.09532"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5609"} +{"question": "What studies have leveraged datasets such as MIMIC-CXR-JPG v2.0.0 and The Cancer Genome Atlas for learning vision-language representations?", "answer": ["MIMIC-CXR-JPG, a large publicly available database of labeled chest radiographs"], "answer_arxiv_id": ["1901.07042"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_5610"} +{"question": "What studies define task difficulty based on the Kolmogorov complexity required to model the relationship between inputs and outputs?", "answer": ["The Information Complexity of Learning Tasks, their Structure and their Distance"], "answer_arxiv_id": ["1904.03292"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_5611"} +{"question": "What works fall under the umbrella term of amortized MCMC technique?", "answer": ["Approximate Inference with Amortised MCMC", "A Contrastive Divergence for Combining Variational Inference and MCMC", "Generative VoxelNet: Learning Energy-Based Models for 3D Shape Synthesis and Analysis", "Learning Energy-Based Model with Variational Auto-Encoder as Amortized Sampler", "A Tale of Two Flows: Cooperative Learning of Langevin Flow and Normalizing Flow Toward Energy-Based Model", "Learning Energy-Based Models by Cooperative Diffusion Recovery Likelihood"], "answer_arxiv_id": ["1702.08343", "1905.04062", "2012.13522v1", "2012.14936", "2205.06924", "2309.05153"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_5612"} +{"question": "What papers are related to the use of tabular language models that encode both tables and text?", "answer": ["TaBert: Pretraining for Joint Understanding of Textual and Tabular Data", "TaPas: Weakly Supervised Table Parsing via Pre-training", "Grappa: Grammar-Augmented Pre-Training for Table Semantic Parsing"], "answer_arxiv_id": ["2005.08314", "2004.02349", "2009.13845"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_5613"} +{"question": "Which studies used DOEs combined with narrow-band coherent laser for generating structured light?", "answer": ["Polka Lines: Learning Structured Illumination and Reconstruction for\n Active Stereo"], "answer_arxiv_id": ["2011.13117"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5614"} +{"question": "Which research derive the regularized policy updates and maximization of the above objective using reparameterized gradients and minibatches from replay buffer?", "answer": ["Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor"], "answer_arxiv_id": ["1801.01290"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5615"} +{"question": "Could you provide me some works that investigate the utilization of multi-turn user feedback for solving a given task?", "answer": ["MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language\n Feedback", "OpenAgents: An Open Platform for Language Agents in the Wild"], "answer_arxiv_id": ["2309.10691", "2310.10634v1"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_5616"} +{"question": "Which research computed the proportion of distinct N-grams generated to characterize the diversity of the generated texts?", "answer": ["A Diversity-Promoting Objective Function for Neural Conversation Models"], "answer_arxiv_id": ["1510.03055"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_5617"} +{"question": "Could you provide me some studies that explored using retrieval augmented generations (RAG) in multimodal few-shot learning?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Improving language models by retrieving from trillions of tokens", "In-Context Retrieval-Augmented Language Models", "Generalization through Memorization: Nearest Neighbor Language Models"], "answer_arxiv_id": ["2005.11401", "2112.04426", "2302.00083", "1911.00172"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_train_5618"} +{"question": "Which studies have approached the self-supervision task through the association with target networks and clustering methods?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments"], "answer_arxiv_id": ["2006.07733", "2006.09882"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_5619"} +{"question": "What papers merged the implicit function with stereo mixture density in order to deliver precise disparity estimations?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Learning Continuous Image Representation with Local Implicit Image\n Function", "SMD-Nets: Stereo Mixture Density Networks"], "answer_arxiv_id": ["2003.08934", "2012.09161", "2104.03866"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_5620"} +{"question": "What work demonstrated the success of 3D generation on the challenging dataset like CO3D?", "answer": ["Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction"], "answer_arxiv_id": ["2109.00512v1"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_5621"} +{"question": "Which works have used retrieval methods for question answering in NLP systems?", "answer": ["GreaseLM: Graph REASoning Enhanced Language Models for Question\n Answering", "QA-GNN: Reasoning with Language Models and Knowledge Graphs for Question\n Answering", "Deep Bidirectional Language-Knowledge Graph Pretraining"], "answer_arxiv_id": ["2201.08860", "2104.06378", "2210.09338"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_5622"} +{"question": "Which papers studied the Lottery ticket hypothesis for pre-trained language models?", "answer": ["The Lottery Ticket Hypothesis for Pre-trained BERT Networks", "When BERT Plays the Lottery, All Tickets Are Winning", "Super Tickets in Pre-Trained Language Models: From Model Compression to Improving Generalization"], "answer_arxiv_id": ["2007.12223", "2005.00561", "2105.12002"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_5623"} +{"question": "Could you provide me some studies about contemporary deep-learning models solving visual analogies in IQ tests?", "answer": ["Solving Bongard Problems with a Visual Language and Pragmatic Reasoning", "RAVEN: A Dataset for Relational and Analogical Visual rEasoNing", "Learning to Make Analogies by Contrasting Abstract Relational Structure"], "answer_arxiv_id": ["1804.04452", "1903.02741", "1902.00120"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_5624"} +{"question": "Any works about approximate unlearning inspired from differential privacy in SCO?", "answer": ["Certified Data Removal from Machine Learning Models", "Descent-to-Delete: Gradient-Based Methods for Machine Unlearning", "Remember What You Want to Forget: Algorithms for Machine Unlearning", "Adaptive Machine Unlearning"], "answer_arxiv_id": ["1911.03030", "2007.02923v1", "2103.03279v2", "2106.04378"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_5625"} +{"question": "What works have developed saliency prediction models to quantitatively study human attention?", "answer": ["Deep Residual Learning for Image Recognition", "Attention Is All You Need", "Very Deep Convolutional Networks for Large-Scale Image Recognition", "DeepFix: A Fully Convolutional Neural Network for predicting Human Eye Fixations", "DeepGaze II: Reading fixations from deep features trained on object recognition", "Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model", "A Dilated Inception Network for Visual Saliency Prediction", "TranSalNet: Towards perceptually relevant visual saliency prediction", "Attention Is All You Need"], "answer_arxiv_id": ["1512.03385", "1706.03762", "1409.1556", "1510.02927", "1610.01563", "1611.09571", "1904.03571", "2110.03593", "1706.03762"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_5626"} +{"question": "What studies apply recurrence for a long sequence in a transformer?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "XLNet: Generalized Autoregressive Pretraining for Language Understanding", "Recurrent Memory Transformer"], "answer_arxiv_id": ["1901.02860", "1906.08237", "2207.06881"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_5627"} +{"question": "Could you provide me some studies that characterized an intriguing property of a specific form of prompting, also known as in-context learning?", "answer": ["What learning algorithm is in-context learning? Investigations with linear models", "Transformers Learn In-Context by Gradient Descent"], "answer_arxiv_id": ["2211.15661", "2212.07677"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5628"} +{"question": "Could you provide me some papers about online algorithms that approximates the task reward value with a neural network?", "answer": ["MetricOpt: Learning to Optimize Black-Box Evaluation Metrics"], "answer_arxiv_id": ["2104.10631"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_5629"} +{"question": "Which work demonstrates the utilization of masked modeling in pre-training in order to learn low-level information from data unsupervisedly?", "answer": ["Revealing the Dark Secrets of Masked Image Modeling"], "answer_arxiv_id": ["2205.13543"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_5630"} +{"question": "Could you cite research around differentiable frameworks for video compression?", "answer": ["Efficient Video Compression via Content-Adaptive Super-Resolution"], "answer_arxiv_id": ["2104.02322"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_5631"} +{"question": "Which papers propose personalization approaches by optimizing text embeddings?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_5632"} +{"question": "Could you provide me some studies about single-entry query complexity for approximate correlated equilibria and Nash equilibria of games with many players?", "answer": ["Query Complexity of Approximate Nash Equilibria", "The Query Complexity of Correlated Equilibria"], "answer_arxiv_id": ["1306.6686", "1305.4874v2"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_5633"} +{"question": "Can you name some recent studies that focused on specific aspects like representing sparse functions, convex-relaxations, and expressive power of self-attention mechanism?", "answer": ["Inductive Biases and Variable Creation in Self-Attention Mechanisms", "Unraveling Attention via Convex Duality: Analysis and Interpretations of Vision Transformers", "Convexifying Transformers: Improving optimization and understanding of transformer networks", "The Quarks of Attention", "Attention is not all you need: pure attention loses rank doubly exponentially with depth"], "answer_arxiv_id": ["2110.10090", "2205.08078", "2211.11052", "2202.08371", "2103.03404"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_5634"} +{"question": "What papers discuss methods that achieve over-optimistic and similar performance for the dynamic link prediction task?", "answer": ["Inductive Representation Learning in Temporal Networks via Causal Anonymous Walks", "Temporal Graph Networks for Deep Learning on Dynamic Graphs", "Do We Really Need Complicated Model Architectures For Temporal Networks?", "Provably expressive temporal graph networks", "Neighborhood-aware Scalable Temporal Network Representation Learning", "Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks"], "answer_arxiv_id": ["2101.05974", "2006.10637", "2302.11636", "2209.15059", "2209.01084", "1908.01207"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_5635"} +{"question": "Which papers focus on the investigations of LLM reasoning capabilities across various domains?", "answer": ["Are Large Language Models Really Good Logical Reasoners? A Comprehensive\n Evaluation and Beyond", "Towards Reasoning in Large Language Models: A Survey"], "answer_arxiv_id": ["2306.09841", "2212.10403"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_5636"} +{"question": "Could you provide me some works where the features of different modalities are used as the inputs to a transformer directly?", "answer": ["Multi-modal Transformer for Video Retrieval"], "answer_arxiv_id": ["2007.10639"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_5637"} +{"question": "Any works about computing paper assignments without short-length cycles to prevent agreements between reviewers?", "answer": ["Combating Collusion Rings is Hard but Possible"], "answer_arxiv_id": ["2112.08444"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_5638"} +{"question": "Which works analyze the use of pseudo-labeling in semi- and self-supervised learning?", "answer": ["SimPLE: Similar Pseudo Label Exploitation for Semi-Supervised\n Classification"], "answer_arxiv_id": ["2103.16725"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_5639"} +{"question": "Could you provide me some works about Zero-Shot Learning that leverage side information from non-visual domains?", "answer": ["Zero-shot Recognition via Semantic Embeddings and Knowledge Graphs", "Zero-Shot Learning Through Cross-Modal Transfer", "Learning Deep Representations of Fine-grained Visual Descriptions"], "answer_arxiv_id": ["1803.08035", "1301.3666", "1605.05395"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_5640"} +{"question": "Which papers examine the issue of overoptimization in Reinforcement Learning?", "answer": ["Reinforcement Learning with a Corrupted Reward Channel", "Uncertainty Estimation for Language Reward Models"], "answer_arxiv_id": ["1705.08417", "2203.07472"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_5641"} +{"question": "Can you point to some studies that have utilized the strategy of transformation selection in the context of data augmentation?", "answer": ["AutoAugment: Learning Augmentation Strategies from Data", "RandAugment: Practical automated data augmentation with a reduced search space"], "answer_arxiv_id": ["1805.09501", "1909.13719"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_5642"} +{"question": "What papers discuss the development of online bidding algorithms for repeated ad auctions?", "answer": ["Online learning in repeated auctions.", "The Best of Many Worlds: Dual Mirror Descent for Online Allocation Problems", "Optimal No-regret Learning in Repeated First-price Auctions", "Learning to Bid Optimally and Efficiently in Adversarial First-price Auctions"], "answer_arxiv_id": ["1511.05720", "2011.10124", "2003.09795", "2007.04568"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_5643"} +{"question": "Any works about distilling teacher model predictions to the student by minimizing cross-entropy?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.14294"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_5644"} +{"question": "Could you provide examples of some studies that propose training a model with early-stopping via a small validation set to surface minority group samples?", "answer": ["Just Train Twice: Improving Group Robustness without Training Group Information", "Correct-n-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlations"], "answer_arxiv_id": ["2107.09044", "2203.01517"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_5645"} +{"question": "What paper used differentiable rendering of RGB-D images for mutual supervision after differentiable registration?", "answer": ["UnsupervisedR&R: Unsupervised Point Cloud Registration via\n Differentiable Rendering"], "answer_arxiv_id": ["2102.11870"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_5646"} +{"question": "Could you provide me some studies about instruction-based image editing?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2211.09800"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_5647"} +{"question": "Could you provide me some studies about LLM-enabled role-playing interactions?", "answer": ["Towards Conversational Diagnostic AI"], "answer_arxiv_id": ["2401.05654"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_5648"} +{"question": "Which works studied replicable algorithms in the context of multi-armed bandits and clustering?", "answer": ["Replicable Bandits", "Replicable Clustering"], "answer_arxiv_id": ["2210.01898v2", "2302.10359v3"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5649"} +{"question": "Which papers proposed learning a dynamics model that only predicts rewards and values over multiple time steps and uses the learned model for planning?", "answer": ["The Predictron: End-To-End Learning and Planning", "Value Prediction Network", "Value Iteration Networks", "Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model", "Mastering Atari Games with Limited Data"], "answer_arxiv_id": ["1612.08810", "1707.03497", "1602.02867", "1911.08265", "2111.00210"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_5650"} +{"question": "Which research reports the possibility of ablating undesired behavior in a model?", "answer": ["Circuit Breaking: Removing Model Behaviors with Targeted Ablation"], "answer_arxiv_id": ["2309.05973"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_5651"} +{"question": "Could you provide me some studies about multi-video summarization?", "answer": ["Diversity-aware Multi-Video Summarization"], "answer_arxiv_id": ["1706.03123"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_5652"} +{"question": "What are some works on label correction to handle noisy annotated data in training sets?", "answer": ["Joint Optimization Framework for Learning with Noisy Labels", "Probabilistic End-to-end Noise Correction for Learning with Noisy Labels", "Learning with Feature-Dependent Label Noise: A Progressive Approach"], "answer_arxiv_id": ["1803.11364", "1903.07788", "2103.07756"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_5653"} +{"question": "What repositories did the COVID images come from?", "answer": ["COVID-19 Image Data Collection", "ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases", "PadChest: A large chest x-ray image dataset with multi-label annotated reports", "BIMCV COVID-19+: a large annotated dataset of RX and CT images from COVID-19 patients"], "answer_arxiv_id": ["2003.11597", "1705.02315", "1901.07441", "2006.01174"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_5654"} +{"question": "Which works propose doubly robust CATE estimators?", "answer": ["Machine Learning Estimation of Heterogeneous Treatment Effects with Instruments"], "answer_arxiv_id": ["1905.10176"], "source_meta": {"published_time": "20220817"}, "qid": "AutoScholarQuery_train_5655"} +{"question": "Are there any studies that use exemplar learning for predicting gene expression from histology images?", "answer": ["Exemplar Guided Deep Neural Network for Spatial Transcriptomics Analysis\n of Gene Expression Prediction"], "answer_arxiv_id": ["2210.16721"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5656"} +{"question": "Which studies have applied diffusion models to adversarial purification for improving model robustness?", "answer": ["Diffusion Models for Adversarial Purification", "Adversarial Purification with Score-based Generative Models"], "answer_arxiv_id": ["2205.07460", "2106.06041"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5657"} +{"question": "Is there a study that uses visual and textual relation graphs to choose prototypes through graph-matching?", "answer": ["Weakly Supervised Learning with Side Information for Noisy Labeled Images"], "answer_arxiv_id": ["2008.11586"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_5658"} +{"question": "Could you provide some studies that developed feature embedding-based methods for unsupervised anomaly detection?", "answer": ["Uninformed Students: Student–Teacher Anomaly Detection with Discriminative Latent Embeddings", "Reconstruction Student with Attention for Student-Teacher Pyramid Matching", "Anomaly Detection via Reverse Distillation from One-Class Embedding", "Asymmetric Student-Teacher Networks for Industrial Anomaly Detection", "Detection and Segmentation of Manufacturing Defects with Convolutional Neural Networks and Transfer Learning", "Patch SVDD: Patch-level SVDD for Anomaly Detection and Segmentation", "PANDA: Adapting Pretrained Features for Anomaly Detection and Segmentation", "CutPaste: Self-Supervised Learning for Anomaly Detection and Localization", "MemSeg: A semi-supervised method for image surface defect detection using differences and commonalities", "A Multi-Scale A Contrario method for Unsupervised Image Anomaly Detection", "Same Same But DifferNet: Semi-Supervised Defect Detection with Normalizing Flows", "Fully Convolutional Cross-Scale-Flows for Image-based Defect Detection", "CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows", "FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flows", "Towards Total Recall in Industrial Anomaly Detection", "Sub-Image Anomaly Detection with Deep Pyramid Correspondences", "CFA: Coupled-hypersphere-based Feature Adaptation for Target-Oriented Anomaly Localization", "Pushing the Limits of Few-shot Anomaly Detection in Industry Vision: Graphcore"], "answer_arxiv_id": ["1911.02357", "2111.15376v2", "2201.10703", "2210.07829", "1808.02518", "2006.16067", "2010.05903v3", "2104.04015", "2205.00908v1", "2110.02407v1", "2008.12577", "2110.02855", "2107.12571", "2111.07677", "2106.08265", "2005.02357", "2206.04325", "2301.12082"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_5659"} +{"question": "What papers propose improvements to diffusion models, such as loss-rescaling techniques or architectural enhancements?", "answer": ["Maximum Likelihood Training of Score-Based Diffusion Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2101.09258", "2105.05233"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_5660"} +{"question": "Which papers have discussed using the Euclidean distance degree to implement efficient solvers in Homotopy Continuation?", "answer": ["HomotopyContinuation.jl: A package for homotopy continuation in Julia"], "answer_arxiv_id": ["1711.10911"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_5661"} +{"question": "Who improved the regret minimization problem of Stochastic Shortest Path to O~​(B∗​S​A​K) when B∗ is known and O~​(B∗3/2​S​A​K) in the parameter-free case?", "answer": ["Near-optimal Regret Bounds for Stochastic Shortest Path"], "answer_arxiv_id": ["2002.09869"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_5662"} +{"question": "Can you name the studies where MVPS was solved for the first time with a SDF representation of the surface?", "answer": ["A Differential Volumetric Approach to Multi-View Photometric Stereo"], "answer_arxiv_id": ["1811.01984"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_5663"} +{"question": "Which papers reported the vulnerabilities of some importance-aware methods to logit scaling attacks or Auto-Attack?", "answer": ["Evaluating the Robustness of Geometry-Aware Instance-Reweighted Adversarial Training", "Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks"], "answer_arxiv_id": ["2103.01914", "2003.01690"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_5664"} +{"question": "What works take reference from the dual process in cognitive science for improving LLMs?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback", "Solving Math Word Problems via Cooperative Reasoning induced Language\n Models"], "answer_arxiv_id": ["2303.17651", "2210.16257"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_5665"} +{"question": "Can you list some studies that develop deep learning based approaches for unsupervised grammar induction?", "answer": ["Compound Probabilistic Context-Free Grammars for Grammar Induction", "Neural Language Modeling by Jointly Learning Syntax and Lexicon", "Ordered Neurons: Integrating Tree Structures into Recurrent Neural Networks", "Tree Transformer: Integrating Tree Structures into Self-Attention", "Visually Grounded Neural Syntax Acquisition", "Visually Grounded Compound PCFGs"], "answer_arxiv_id": ["1906.10225", "1711.02013", "1810.09536", "1909.06639", "1906.02890", "2009.12404"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_5666"} +{"question": "What study proposed the Density/Coverage metric with an aim to fix the overestimation of the manifold around real outliers?", "answer": ["Reliable Fidelity and Diversity Metrics for Generative Models"], "answer_arxiv_id": ["2002.09797"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_5667"} +{"question": "Which works discusss retrieval-based diffusion models?", "answer": ["Semi-Parametric Neural Image Synthesis", "KNN-Diffusion: Image Generation via Large-Scale Retrieval"], "answer_arxiv_id": ["2204.11824", "2204.02849"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_5668"} +{"question": "Could you provide me some works that use dual latent spaces for image editing?", "answer": ["Disentangled Image Generation Through Structured Noise Injection", "Diagonal Attention and Style-based GAN for Content-Style Disentanglement in Image Generation and Translation", "TransEditor: Transformer-Based Dual-Space GAN for Highly Controllable Facial Editing", "CoordGAN: Self-Supervised Dense Correspondences Emerge from GANs"], "answer_arxiv_id": ["2004.12411", "2103.16146", "2203.17266", "2203.16521"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_5669"} +{"question": "Which works are about the efforts on open sourced LLM weights?", "answer": ["OPT: Open Pre-trained Transformer Language Models", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2205.01068", "2302.13971"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_5670"} +{"question": "What studies approach non-identifiability by imposing sparsity in causal representation learning?", "answer": ["Identifiable Deep Generative Models via Sparse Decoding"], "answer_arxiv_id": ["2110.10804"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_5671"} +{"question": "Could you list the papers that measured dataset difficulty by leveraging Arimoto information?", "answer": ["A Theory of Usable Information Under Computational Constraints"], "answer_arxiv_id": ["2002.10689"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_5672"} +{"question": "Can you provide a reference where it has been argued that LLMs face challenges in self-correcting their responses without external feedback?", "answer": ["Large Language Models Cannot Self-Correct Reasoning Yet"], "answer_arxiv_id": ["2310.01798"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_5673"} +{"question": "What are some works leveraged transfer learning for long-tailed recognition?", "answer": ["LPT: Long-tailed Prompt Tuning for Image Classification", "VL-LTR: Learning Class-wise Visual-Linguistic Representation for\n Long-Tailed Visual Recognition"], "answer_arxiv_id": ["2210.01033", "2111.13579"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_5674"} +{"question": "Which papers discuss issues arising in LLM evaluation due to long context and agent abilities?", "answer": ["LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding", "AgentBench: Evaluating LLMs as Agents"], "answer_arxiv_id": ["2308.14508v2", "2308.03688"], "source_meta": {"published_time": "20230913"}, "qid": "AutoScholarQuery_train_5675"} +{"question": "Which paper stipulates that regret bounds depend on pseudo dimension and are thus generally suboptimal for complex F?", "answer": ["Efficient Active Learning with Abstention"], "answer_arxiv_id": ["2204.00043"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_5676"} +{"question": "Can you list the works where researchers demonstrated the benefits of using adversarial pretraining for downstream tasks?", "answer": ["Do Adversarially Robust ImageNet Models Transfer Better?"], "answer_arxiv_id": ["2007.08489"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_5677"} +{"question": "Which studies suggested policy constraint methods to force the trained policy to be close to the behavior policy using KL divergence or direct behavior cloning?", "answer": ["Behavior Regularized Offline Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["1911.11361", "2106.06860"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_5678"} +{"question": "Which works model the alphas of a fixed set of multi-plane images (MPI)?", "answer": ["Stereo Magnification: Learning View Synthesis using Multiplane Images", "Pushing the Boundaries of View Extrapolation with Multiplane Images"], "answer_arxiv_id": ["1805.09817", "1905.00413"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_5679"} +{"question": "What papers proposed alternative fusion mechanisms for audio-visual fusion?", "answer": ["Audiovisual Masked Autoencoders", "Attention Bottlenecks for Multimodal Fusion"], "answer_arxiv_id": ["2212.05922", "2107.00135"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_5680"} +{"question": "What works have been done on learned indexing approach for solving rank problems?", "answer": ["ALEX: An Updatable Adaptive Learned Index"], "answer_arxiv_id": ["1905.08898"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_5681"} +{"question": "What works have focused on relaxing assumptions about data distributions such as exchangeability within conformal prediction?", "answer": ["Distribution-free uncertainty quantification for classification under label shift", "Conformal Prediction Beyond Exchangeability"], "answer_arxiv_id": ["2103.03323", "2202.13415"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_5682"} +{"question": "Which paper contains a parallel work about SSL on pure 3D geometry along the molecule representation learning research line?", "answer": ["Pre-training via Denoising for Molecular Property Prediction"], "answer_arxiv_id": ["2206.00133"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_5683"} +{"question": "What works exist on open-vocabulary motion generation?", "answer": ["BABEL: Bodies, Action and Behavior with English Labels", "Motion-X: A Large-scale 3D Expressive Whole-body Human Motion Dataset", "Human Motion Diffusion Model", "T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete\n Representations", "Diffusion Motion: Generate Text-Guided 3D Human Motion by Diffusion\n Model", "MotionGPT: Finetuned LLMs Are General-Purpose Motion Generators", "MotionGPT: Human Motion as a Foreign Language"], "answer_arxiv_id": ["2106.09696", "2307.00818", "2209.14916", "2301.06052", "2210.12315", "2306.10900", "2306.14795"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_5684"} +{"question": "Can you give examples of studies that used the OTNN for the objective evaluation of the quality of explanations?", "answer": ["Achieving robustness in classification using optimal transport with hinge regularization"], "answer_arxiv_id": ["2006.06520"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_5685"} +{"question": "What studies have looked into the generalization error of representation learning for ordinal data?", "answer": ["Finite Sample Prediction and Recovery Bounds for Ordinal Embedding", "Generalization Error Bound for Hyperbolic Ordinal Embedding"], "answer_arxiv_id": ["1606.07081", "2105.10475"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_5686"} +{"question": "What studies have exploited structure-related features, such as Distance Encoding (DE), that incorporate distance between nodes?", "answer": ["Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning"], "answer_arxiv_id": ["2009.00142"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_5687"} +{"question": "What works proposed deeper networks for single image super-resolution?", "answer": ["Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network"], "answer_arxiv_id": ["1511.04587", "1609.05158"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5688"} +{"question": "What studies estimate the camera pose with inverse NeRF optimization when the neural implicit network is fully trained?", "answer": ["INeRF: Inverting Neural Radiance Fields for Pose Estimation", "NeRF--: Neural Radiance Fields Without Known Camera Parameters"], "answer_arxiv_id": ["2012.05877", "2102.07064"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_5689"} +{"question": "Can you provide examples of studies that have explicitly partitioned the latent state space in the world model into reward-relevant and reward-irrelevant features, like the Task Informed Abstractions (TIA)?", "answer": ["Learning Task Informed Abstractions"], "answer_arxiv_id": ["2106.15612"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_5690"} +{"question": "Who investigates the streaming setting of fair suodular maximization and proposes a 1/2121/2-approximation one pass algorithm?", "answer": ["Fairness in Streaming Submodular Maximization: Algorithms and Hardness"], "answer_arxiv_id": ["2010.07431"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5691"} +{"question": "Which papers discuss Binning-based methods for confidence calibration?", "answer": ["Multi-Class Uncertainty Calibration via Mutual Information Maximization-based Binning"], "answer_arxiv_id": ["2006.13092"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_5692"} +{"question": "Which study proposed an enhancement to random crop augmentation using a simple random shift augmentation specifically in the context of visual off-policy RL?", "answer": ["Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Mastering Visual Continuous Control: Improved Data-Augmented Reinforcement Learning"], "answer_arxiv_id": ["2004.13649", "2107.09645"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_5693"} +{"question": "What studies discuss the utilization of a dense voting strategy to detect keypoints?", "answer": ["PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation", "PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose\n Estimation", "FFB6D: A Full Flow Bidirectional Fusion Network for 6D Pose Estimation", "Vote from the Center: 6 DoF Pose Estimation in RGB-D Images by Radial\n Keypoint Voting"], "answer_arxiv_id": ["1812.11788", "1911.04231", "2103.02242", "2104.02527"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_5694"} +{"question": "Any works that propose diffusion models for motion tracking?", "answer": ["DiffusionTrack: Diffusion Model For Multi-Object Tracking"], "answer_arxiv_id": ["2308.09905"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_5695"} +{"question": "What articles have proposed methods to address domain generalization in the context of image classification for out-of-distribution data?", "answer": ["In Search of Lost Domain Generalization", "A Fine-Grained Analysis on Distribution Shift", "Domain Generalization: A Survey", "Generalizing to Unseen Domains: A Survey on Domain Generalization"], "answer_arxiv_id": ["2007.01434", "2110.11328", "2103.02503", "2103.03097"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_5696"} +{"question": "What papers attempted to solve the Bayesian exploration POMDP in the context of meta-reinforcement learning?", "answer": ["RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning", "Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables"], "answer_arxiv_id": ["1611.02779", "1910.08348", "1903.08254"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_5697"} +{"question": "What papers have explored the role of citations, generated by retrieval models, in building responsible and accountable LLMs?", "answer": ["Citation: A Key to Building Responsible and Accountable Large Language\n Models"], "answer_arxiv_id": ["2307.02185"], "source_meta": {"published_time": "20240225"}, "qid": "AutoScholarQuery_train_5698"} +{"question": "What are some affinity-based methods in the FSS approach?", "answer": ["Simpler is Better: Few-shot Semantic Segmentation with Classifier Weight\n Transformer", "Hypercorrelation Squeeze for Few-Shot Segmentation", "Few-Shot Segmentation via Cycle-Consistent Transformer", "Hierarchical Dense Correlation Distillation for Few-Shot Segmentation"], "answer_arxiv_id": ["2108.03032", "2104.01538", "2106.02320", "2303.14652"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_5699"} +{"question": "Which datasets include multiple-choice and fill-in-the-blank questions for medical QA?", "answer": ["CliCR: A Dataset of Clinical Case Reports for Machine Reading Comprehension", "What Disease does this Patient Have? A Large-scale Open Domain Question Answering Dataset from Medical Exams", "Measuring Massive Multitask Language Understanding", "Large Language Models Encode Clinical Knowledge"], "answer_arxiv_id": ["1803.09720v1", "2009.13081", "2009.03300", "2212.13138"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_5700"} +{"question": "Which works has been used in super-resolution?", "answer": ["Image Super-Resolution via Iterative Refinement"], "answer_arxiv_id": ["2104.07636"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_5701"} +{"question": "Which research paper introduces the use of 3D bounding boxes for coarse detection in instance segmentation?", "answer": ["3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans"], "answer_arxiv_id": ["1812.07003"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_5702"} +{"question": "Which research investigates neuron redundancy in the context of pre-trained language models?", "answer": ["Analyzing Redundancy in Pretrained Transformer Models"], "answer_arxiv_id": ["2004.04010"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5703"} +{"question": "Could you provide me some studies that propose using LLMs for dataset synthesis?", "answer": ["Generating Datasets with Pretrained Language Models", "Self-Instruct: Aligning Language Models with Self-Generated Instructions", "Unnatural Instructions: Tuning Language Models with (Almost) No Human\n Labor", "Self-Alignment with Instruction Backtranslation", "STaR: Bootstrapping Reasoning With Reasoning", "Large Language Models Can Self-Improve"], "answer_arxiv_id": ["2104.07540", "2212.10560", "2212.09689", "2308.06259", "2203.14465", "2210.11610"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_5704"} +{"question": "What studies look at weaker solution concepts that relax stationarity or Markovian properties?", "answer": ["The Complexity of Markov Equilibrium in Stochastic Games", "V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL"], "answer_arxiv_id": ["2204.03991", "2110.14555"], "source_meta": {"published_time": "20220803"}, "qid": "AutoScholarQuery_train_5705"} +{"question": "Can you provide works using pre-trained language models (PLMs) for commonsense reasoning?", "answer": ["TIMEDIAL: Temporal Commonsense Reasoning in Dialog", "Time-Aware Language Models as Temporal Knowledge Bases"], "answer_arxiv_id": ["2106.04571", "2106.15110"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_5706"} +{"question": "Which works represented event data as graphs in order to include more spatio-temporal information?", "answer": ["Graph-based Asynchronous Event Processing for Rapid Object Recognition", "Event-based Motion Segmentation with Spatio-Temporal Graph Cuts", "AEGNN: Asynchronous Event-based Graph Neural Networks"], "answer_arxiv_id": ["2308.14419", "2012.08730", "2203.17149"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5707"} +{"question": "Could you provide me some papers that proposed block-recurrent approaches in RIR-related approaches?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Block-Recurrent Transformers", "Staircase Attention for Recurrent Processing of Sequences", "Temporal Latent Bottleneck: Synthesis of Fast and Slow Processing Mechanisms in Sequence Learning", "Recurrent Memory Transformer"], "answer_arxiv_id": ["1901.02860", "2203.07852", "2106.04279", "2205.14794", "2207.06881"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_5708"} +{"question": "What papers discuss the use of energy distance for direct training of NN based generative models?", "answer": ["Training generative neural networks via Maximum Mean Discrepancy optimization", "Generative Moment Matching Networks", "Demystifying MMD GANs"], "answer_arxiv_id": ["1505.03906", "1502.02761", "1801.01401"], "source_meta": {"published_time": "20221107"}, "qid": "AutoScholarQuery_train_5709"} +{"question": "What are the papers about editing a 3D scene using a NeRF representation?", "answer": ["Editing Conditional Radiance Fields", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields", "NeRF-Editing: Geometry Editing of Neural Radiance Fields", "SNeRF: Stylized Neural Implicit Representations for 3D Scenes", "StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D\n Mutual Learning", "NeRF-Art: Text-Driven Neural Radiance Fields Stylization", "Locally Stylized Neural Radiance Fields", "3DDesigner: Towards Photorealistic 3D Object Generation and Editing with\n Text-guided Diffusion Models", "3D-aware Blending with Generative NeRFs", "Decomposing NeRF for Editing via Feature Field Distillation", "SPIn-NeRF: Multiview Segmentation and Perceptual Inpainting with Neural\n Radiance Fields", "Removing Objects From Neural Radiance Fields"], "answer_arxiv_id": ["2105.06466", "2112.05139", "2205.04978", "2207.02363", "2205.12183", "2212.08070", "2309.10684", "2211.14108", "2302.06608", "2205.15585", "2211.12254", "2212.11966"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_5710"} +{"question": "What studies proposed two-state methods for gaze target detection?", "answer": ["Connecting Gaze, Scene, and Attention: Generalized Attention Estimation via Joint Modeling of Gaze and Scene Saliency"], "answer_arxiv_id": ["1807.10437"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_5711"} +{"question": "Could you provide me some studies that tackled Object Retrieval and Spatiotemporal Detection in Video Understanding?", "answer": ["Query-Dependent Video Representation for Moment Retrieval and Highlight\n Detection", "TubeR: Tubelet Transformer for Video Action Detection"], "answer_arxiv_id": ["2303.13874", "2104.00969"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_5712"} +{"question": "Could you provide me some studies about inductive architectures that integrate ideas from earlier path-based link prediction approaches into modern GNN architectures?", "answer": ["DeepWalk: Online Learning of Social Representations", "node2vec: Scalable Feature Learning for Networks", "Relational Message Passing for Knowledge Graph Completion", "Geodesic Graph Neural Network for Efficient Graph Representation Learning", "Neural Bellman-Ford Networks: A General Graph Neural Network Framework for Link Prediction"], "answer_arxiv_id": ["1403.6652", "1607.00653", "2002.06757", "2210.02636", "2106.06935"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_5713"} +{"question": "Can you tell me about datasets that only focus on sentence-level actions extraction?", "answer": ["An Approach for Process Model Extraction By Multi-Grained Text\n Classification"], "answer_arxiv_id": ["1906.02127"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_5714"} +{"question": "Who introduced greedy rejection coding (GRC) which generalizes the rejection sampling algorithm to arbitrary probability spaces and arbitrary splitting functions?", "answer": ["Faster Relative Entropy Coding with Greedy Rejection Coding"], "answer_arxiv_id": ["2309.15746"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5715"} +{"question": "What studies used evolutionary search algorithms for combinatorial design optimization in robot design?", "answer": ["Scalable Co-Optimization of Morphology and Control in Embodied Machines"], "answer_arxiv_id": ["1706.06133"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5716"} +{"question": "What paper proposed designing the sequences and structures of CDRs simultaneously?", "answer": ["Iterative Refinement Graph Neural Network for Antibody Sequence-Structure Co-design"], "answer_arxiv_id": ["2110.04624"], "source_meta": {"published_time": "20220812"}, "qid": "AutoScholarQuery_train_5717"} +{"question": "Which paper utilized gradients from all layers to train a binary classifier?", "answer": ["Gradients as a measure of uncertainty in neural networks"], "answer_arxiv_id": ["2008.08030"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_5718"} +{"question": "What studies looked into using support constraints as an alternative approach to prevent offline RL algorithms from exploiting out-of-distribution actions?", "answer": ["Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Offline RL With Realistic Datasets: Heteroskedasticity and Support Constraints", "Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble"], "answer_arxiv_id": ["1906.00949", "2211.01052", "2110.01548"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_5719"} +{"question": "What studies showed that a large learning rate leads to a better model?", "answer": ["Towards Explaining the Regularization Effect of Initial Large Learning Rate in Training Neural Networks"], "answer_arxiv_id": ["1907.04595"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_5720"} +{"question": "Which works proposed cdf-based approaches in addressing distributional treatment effects?", "answer": ["Inference on Counterfactual Distributions", "Efficient Estimation of Quantiles in Missing Data Models"], "answer_arxiv_id": ["0904.0951", "1512.08110"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_5721"} +{"question": "Which papers have proposed methods to obtain accurate estimates of and valid inferences for the conditional average treatment effect (CATE)?", "answer": ["Estimating treatment effect heterogeneity in randomized program evaluation", "Recursive Partitioning for Heterogeneous Causal Effects", "Estimation and Inference of Heterogeneous Treatment Effects using Random Forests", "Meta-learners for Estimating Heterogeneous Treatment Effects using Machine Learning", "Quasi-Oracle Estimation of Heterogeneous Treatment Effects"], "answer_arxiv_id": ["1305.5682", "1504.01132", "1510.04342", "1706.03461", "1712.04912"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_5722"} +{"question": "Which papers discuss obtaining lighting representations through joint geometry, material, and lighting optimization with inverse rendering?", "answer": ["Neural Inverse Rendering of an Indoor Scene from a Single Image", "Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying Lighting and SVBRDF from a Single Image", "Object-based Illumination Estimation with Rendering-aware Neural Networks", "Extracting Triangular 3D Models, Materials, and Lighting From Images", "CADSim: Robust and Scalable in-the-wild 3D Reconstruction for Controllable Sensor Simulation", "Shape, Light, and Material Decomposition from Images using Monte Carlo Rendering and Denoising", "Neural Radiance Transfer Fields for Relightable Novel-view Synthesis with Global Illumination", "IntrinsicNeRF: Learning Intrinsic Neural Radiance Fields for Editable Novel View Synthesis", "NeRF for Outdoor Scene Relighting", "Multi-view Inverse Rendering for Large-scale Real-world Indoor Scenes", "Accidental Light Probes", "Neural Fields meet Explicit Geometric Representations for Inverse Rendering of Urban Scenes"], "answer_arxiv_id": ["1901.02453", "1905.02722", "2008.02514", "2111.12503", "2311.01447", "2206.03380", "2207.13607", "2210.00647", "2112.05140", "2211.10206", "2301.05211", "2304.03266"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_5723"} +{"question": "Any works that focused on learning transferable representations via meta-learning to alleviate catastrophic forgetting?", "answer": ["Meta-Learning Representations for Continual Learning", "Learning to Continually Learn"], "answer_arxiv_id": ["1905.12588", "2002.09571"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_5724"} +{"question": "What work presented concepts of Principal Curves?", "answer": ["Nonlinearities and Adaptation of Color Vision from Sequential Principal\n Curves Analysis"], "answer_arxiv_id": ["1602.00236"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_5725"} +{"question": "Which studies analyzed subspace clustering and gave a risk bound?", "answer": ["Testing the manifold hypothesis"], "answer_arxiv_id": ["1310.0425"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_5726"} +{"question": "Can you list some studies that introduced effective strategies for fitting planes and other 3D shapes to 3D data?", "answer": ["Cuboids Revisited: Learning Robust 3D Shape Fitting to Single RGB Images", "MonteBoxFinder: Detecting and Filtering Primitives to Fit a Noisy Point\n Cloud"], "answer_arxiv_id": ["2105.02047", "2207.14268"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_5727"} +{"question": "Which works use a local fitting strategy aimed at learning finer shape details in SDF?", "answer": ["OctField: Hierarchical Implicit Functions for 3D Modeling", "Deep Local Shapes: Learning Local SDF Priors for Detailed 3D\n Reconstruction", "Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D\n Shapes", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2111.01067", "2003.10983", "2101.10994", "2201.05989"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_5728"} +{"question": "Are there any research papers about automatically generating fractals that our study drew inspiration from?", "answer": ["The Chaos Game on a General Iterated Function System"], "answer_arxiv_id": ["1005.0322"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_5729"} +{"question": "Which works are related to structured pruning in the field of network pruning?", "answer": ["Pruning Filters for Efficient ConvNets", "Learning Structured Sparsity in Deep Neural Networks", "Channel Pruning for Accelerating Very Deep Neural Networks", "Soft Filter Pruning for Accelerating Deep Convolutional Neural Networks", "Neural Pruning via Growing Regularization"], "answer_arxiv_id": ["1608.08710", "1608.03665", "1707.06168", "1808.06866", "2012.09243"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_5730"} +{"question": "Which studies utilize feature maps at multiple spatial scales for localization tasks?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation"], "answer_arxiv_id": ["1505.04597"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_5731"} +{"question": "What research papers treated adversarial training as a form of data augmentation?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Subspace Adversarial Training", "Adversarial Examples Improve Image Recognition"], "answer_arxiv_id": ["1706.06083", "2111.12229", "1911.09665"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_5732"} +{"question": "What work was the closest to the evaluation part of this research with a focus on compositional and pragmatic language understanding of VLMs?", "answer": ["Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality"], "answer_arxiv_id": ["2204.03162"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_5733"} +{"question": "Are there studies highlighting the difficulties of SAM in segmenting objects with poor visibility?", "answer": ["SAM Struggles in Concealed Scenes – Empirical Study on “Segment Anything”", "Can SAM Segment Anything? -When SAM Meets Camouflaged Object Detection", "Segment Anything Is Not Always Perfect: An Investigation of SAM on Different Real-world Applications", "Segment Anything Model for Medical Image Analysis: an Experimental Study", "SAM on Medical Images: A Comprehensive Study on Three Prompt Modes", "Can SAM Segment Polyps?", "Segment Anything Model (SAM) Meets Glass: Mirror and Transparent Objects Cannot Be Easily Detected"], "answer_arxiv_id": ["2304.06022", "2304.04709", "2304.05750", "2304.10517", "2305.00035v1", "2304.07583", "2305.00278"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_5734"} +{"question": "Can you list some of the key unnamed contributions to the field of Vision Large Language Models?", "answer": ["LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "InternLM-XComposer: A Vision-Language Large Model for Advanced\n Text-image Comprehension and Composition", "GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest", "NExT-GPT: Any-to-Any Multimodal LLM", "Emu: Generative Pretraining in Multimodality", "Flamingo: a Visual Language Model for Few-Shot Learning", "VideoChat: Chat-Centric Video Understanding", "The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "Monkey: Image Resolution and Text Label Are Important Things for Large\n Multi-modal Models", "mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document\n Understanding"], "answer_arxiv_id": ["2303.16199", "2309.15112", "2307.03601", "2309.05519", "2307.05222", "2204.14198", "2305.06355", "2309.17421", "2209.06794", "2305.03726", "2306.02858", "2311.06607", "2307.02499"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_5735"} +{"question": "What research papers look at the benefits of iteratively learning from signals produced by agents of the previous generation?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback", "Reinforced Self-Training (\\rest) for Language Modeling"], "answer_arxiv_id": ["2303.17651", "2308.08998"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_5736"} +{"question": "What research has been conducted on Speech-to-Speech Translation (A2A)?", "answer": ["Direct speech-to-speech translation with a sequence-to-sequence model", "Translatotron 2: High-quality direct speech-to-speech translation with\n voice preservation"], "answer_arxiv_id": ["1904.06037", "2107.08661"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_5737"} +{"question": "Could you provide me any studies utilizing GMM, GAN, and VAE in human motion priors?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "Generative Adversarial Networks", "End-to-end Recovery of Human Shape and Pose", "HP-GAN: Probabilistic 3D human motion prediction via GAN", "Auto-Encoding Variational Bayes", "HuMoR: 3D Human Motion Model for Robust Pose Estimation", "Learning Motion Priors for 4D Human Body Capture in 3D Scenes", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE"], "answer_arxiv_id": ["1904.05866", "1406.2661", "1712.06584", "1711.09561", "1312.6114", "2105.04668", "2108.10399", "2104.05670"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_5738"} +{"question": "Could you provide me some studies which used the patch discriminator to capture local style statistics for texture synthesis?", "answer": ["Precomputed Real-Time Texture Synthesis with Markovian Generative Adversarial Networks"], "answer_arxiv_id": ["1604.04382"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_5739"} +{"question": "Which works introduced parameter efficiency by training a subset of existing parameters?", "answer": ["BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-models", "Revisiting Batch Normalization For Practical Domain Adaptation"], "answer_arxiv_id": ["2106.10199", "1603.04779"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_5740"} +{"question": "What studies provide methods that have been designed for unsupervised domain adaptation for time series data?", "answer": ["Multi-Source Deep Domain Adaptation with Weak Supervision for Time-Series Sensor Data", "Time Series Domain Adaptation via Sparse Associative Structure Alignment"], "answer_arxiv_id": ["2005.10996", "2012.11797"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_5741"} +{"question": "Can you name some works about tuning-free methods for subject-driven generation techniques?", "answer": ["ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "PhotoMaker: Customizing Realistic Human Photos via Stacked ID Embedding"], "answer_arxiv_id": ["2302.13848", "2304.03411", "2312.04461"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_5742"} +{"question": "Could you reference some works on steganalysis using statistical tests and neural networks?", "answer": ["A Novel Convolutional Neural Network for Image Steganalysis with Shared Normalization"], "answer_arxiv_id": ["1711.07306"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_5743"} +{"question": "What works proposed variations of Adapter and Prefix tuning by updating the weight matrices or varying the placement of Adapters?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "Towards a Unified View of Parameter-Efficient Transfer Learning", "Scaling & Shifting Your Features: A New Baseline for Efficient Model\n Tuning", "Counter-Interference Adapter for Multilingual Machine Translation"], "answer_arxiv_id": ["2106.09685", "2110.04366", "2210.08823", "2104.08154"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_5744"} +{"question": "Which studies discussed resampling methods for long-tailed recognition?", "answer": ["SMOTE: Synthetic Minority Over-sampling Technique", "Deep Over-sampling Framework for Classifying Imbalanced Data"], "answer_arxiv_id": ["1106.1813", "1704.07515"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_5745"} +{"question": "Could you provide me some studies where visual tasks were solved using purely curiosity-based intrinsic reward?", "answer": ["Large-Scale Study of Curiosity-Driven Learning"], "answer_arxiv_id": ["1808.04355"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_5746"} +{"question": "Could you provide me some works that extended the usage of Decision Transformer(DT)?", "answer": ["Online Decision Transformer", "Prompting Decision Transformer for Few-Shot Policy Generalization", "Q-learning Decision Transformer: Leveraging Dynamic Programming for Conditional Sequence Modelling in Offline RL"], "answer_arxiv_id": ["2202.05607", "2206.13499", "2209.03993"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_5747"} +{"question": "Which paper provides a comprehensive overview of dropout in a survey?", "answer": ["Survey of Dropout Methods for Deep Neural Networks"], "answer_arxiv_id": ["1904.13310"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_5748"} +{"question": "What studies transfer the product of the Hessian inverse and a vector to the solution to solve Stochastic Bilevel Optimization problems?", "answer": ["A Fully Single Loop Algorithm for Bilevel Optimization without Hessian Inverse", "A framework for bilevel optimization that enables stochastic and global variance reduction algorithms"], "answer_arxiv_id": ["2112.04660", "2201.13409"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5749"} +{"question": "In which studies did researchers appraoch to elicit the intermediate reasoning steps of LLMs in chains and trees?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2201.11903", "2305.10601"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_5750"} +{"question": "What work first showed that diffusion models outperform GANs in image synthesis?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_5751"} +{"question": "Which works attempted to combine model-based and model-free methods for face animation?", "answer": ["FNeVR: Neural Volume Rendering for Face Animation"], "answer_arxiv_id": ["2209.10340"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_5752"} +{"question": "Could you point to the works that analyze the social impact and potential disparate effects of strategic adaptation?", "answer": ["The Social Cost of Strategic Classification", "The Disparate Effects of Strategic Manipulation", "The Disparate Equilibria of Algorithmic Decision Making when Individuals Invest Rationally"], "answer_arxiv_id": ["1808.08460", "1808.08646", "1910.04123"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_5753"} +{"question": "Could you provide me some works that consider explicitly decomposing the transition kernel?", "answer": ["Spectral Decomposition Representation for Reinforcement Learning"], "answer_arxiv_id": ["2208.09515"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_5754"} +{"question": "Any studies building a dialog with the user by making use of CoIR models?", "answer": ["Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language Feedback", "Dialog-based Interactive Image Retrieval"], "answer_arxiv_id": ["1905.12794", "1805.00145"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_5755"} +{"question": "Which studies attempted dataless distillation in the context of computer vision architectures?", "answer": ["Data-Free Knowledge Distillation for Deep Neural Networks", "Zero-Shot Knowledge Distillation in Deep Networks"], "answer_arxiv_id": ["1710.07535v2", "1905.08114v1"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_5756"} +{"question": "What researches proposed variational techniques for dynamical systems?", "answer": ["Scalable Variational Inference for Dynamical Systems", "Black-box Variational Inference for Stochastic Differential Equations"], "answer_arxiv_id": ["1705.07079v2", "1802.03335v3"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_5757"} +{"question": "Which work addressed the inverse rendering problem for complex indoor scenes?", "answer": ["Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying Lighting and SVBRDF from a Single Image"], "answer_arxiv_id": ["1905.02722"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_5758"} +{"question": "What are the papers which try to improve the expressivity of GNNs by incorporating higher-order neighbourhoods?", "answer": ["Provably Powerful Graph Networks", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks"], "answer_arxiv_id": ["1905.11136", "1810.02244"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_5759"} +{"question": "What work focuses on learning the model change factors and their representation in heterogeneous domains?", "answer": ["AdaRL: What, Where, and How to Adapt in Transfer Reinforcement Learning"], "answer_arxiv_id": ["2107.02729"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_5760"} +{"question": "Could you cite a research paper that investigated deconfounding scores in the context of causal inference?", "answer": ["Deconfounding Scores: Feature Representations for Causal Effect Estimation with Weak Overlap"], "answer_arxiv_id": ["2104.05762"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5761"} +{"question": "Which papers are about developing WSI predictions based on instance scores?", "answer": ["Patch-based Convolutional Neural Network for Whole Slide Tissue Image\n Classification", "Weakly supervised multiple instance learning histopathological tumor\n segmentation", "CAMEL: A Weakly Supervised Learning Framework for Histopathology Image\n Segmentation"], "answer_arxiv_id": ["1504.07947", "2004.05024", "1908.10555"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5762"} +{"question": "Which papers utilize generative models for data augmentation in medicine?", "answer": ["GAN-based Synthetic Medical Image Augmentation for increased CNN Performance in Liver Lesion Classification", "Generate To Adapt: Aligning Domains using Generative Adversarial Networks"], "answer_arxiv_id": ["1803.01229", "1704.01705"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5763"} +{"question": "What works focus on specific issues in prompt-based methods?", "answer": ["Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learning", "Improving and Simplifying Pattern Exploiting Training", "Calibrate Before Use: Improving Few-Shot Performance of Language Models"], "answer_arxiv_id": ["2205.05638", "2103.11955", "2102.09690"], "source_meta": {"published_time": "20221106"}, "qid": "AutoScholarQuery_train_5764"} +{"question": "Can you give examples of studies on unsupervised learning in hyperbolic space?", "answer": ["Hyperbolic Neural Networks++", "Clipped Hyperbolic Classifiers Are Super-Hyperbolic Classifiers", "Hyperbolic Deep Learning in Computer Vision: A Survey"], "answer_arxiv_id": ["2006.08210", "2107.11472", "2305.06611"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_5765"} +{"question": "Which works discussed the potential of pre-trained classifiers in generating data samples within the framework of energy-based models?", "answer": ["Your classifier is secretly an energy based model and you should treat it like one"], "answer_arxiv_id": ["1912.03263"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_5766"} +{"question": "What recent works have explored using LLMs for generating code in general-purpose programming languages?", "answer": ["Program Synthesis with Large Language Models", "A Systematic Evaluation of Large Language Models of Code"], "answer_arxiv_id": ["2108.07732", "2202.13169"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5767"} +{"question": "Which work introduced a framework that formulates the KL divergence between functions as the supremum of marginal KL divergences over finite sets of inputs?", "answer": ["Functional Variational Bayesian Neural Networks"], "answer_arxiv_id": ["1903.05779"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_5768"} +{"question": "Which works contributed to low-resource ASR by exploring continuous pre-training, model adaptation, and data augmentation?", "answer": ["LAE: Language-Aware Encoder for Monolingual and Multilingual ASR", "DARTS-ASR: Differentiable Architecture Search for Multilingual Speech\n Recognition and Adaptation", "Master-ASR: Achieving Multilingual Scalability and Low-Resource\n Adaptation in ASR with Modular Learning", "When Is TTS Augmentation Through a Pivot Language Useful?"], "answer_arxiv_id": ["2206.02093", "2005.07029", "2306.15686", "2207.09889"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_5769"} +{"question": "What research papers focus on regressing 6DoF object poses based on predefined object instance templates?", "answer": ["GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D\n Object Pose Estimation", "SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation"], "answer_arxiv_id": ["2102.12145", "2108.08367"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_5770"} +{"question": "Which papers utilized GAN-based models to predict multiple futures in trajectory predictions?", "answer": ["Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks", "Social-BiGAT: Multimodal Trajectory Forecasting using Bicycle-GAN and Graph Attention Networks"], "answer_arxiv_id": ["1803.10892", "1907.03395"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5771"} +{"question": "What are the studies that propose BRECQ and QDrop methodologies?", "answer": ["Brecq: pushing the limit of post-training quantization by block reconstruction", "QDrop: Randomly Dropping Quantization for Extremely Low-bit Post-Training Quantization"], "answer_arxiv_id": ["2102.05426", "2203.05740"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5772"} +{"question": "Are there studies on matching features that may be visually distinct but remain semantically congruent in cross-spectral matching?", "answer": ["RGB-Multispectral Matching: Dataset, Learning Methodology, Evaluation"], "answer_arxiv_id": ["2206.07047"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_5773"} +{"question": "Can you tell me about the studies where variants of GAN have been proposed?", "answer": ["Conditional Generative Adversarial Nets", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Image-to-Image Translation with Conditional Adversarial Networks", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks", "Data Augmentation Generative Adversarial Networks"], "answer_arxiv_id": ["1411.1784", "1812.04948", "1611.07004", "1703.10593", "1711.04340"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_5774"} +{"question": "Which work focused on studying the linearized learning dynamics and suggested that a competition between the feature signal strength and augmentation strength can lead to dimensional collapse?", "answer": ["Understanding Dimensional Collapse in Contrastive Self-supervised Learning"], "answer_arxiv_id": ["2110.09348"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_5775"} +{"question": "Which works focused on the reconstruction of geometry in 3D clothed humans?", "answer": ["PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution\n 3D Human Digitization", "ICON: Implicit Clothed humans Obtained from Normals", "Structured 3D Features for Reconstructing Controllable Avatars"], "answer_arxiv_id": ["2004.00452", "2112.09127", "2212.06820"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_5776"} +{"question": "What works utilize Stochastic Gradient HMC (SGHMC) in the field of Markov chain Monte Carlo (MCMC)?", "answer": ["Stochastic Gradient Hamiltonian Monte Carlo"], "answer_arxiv_id": ["1402.4102"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5777"} +{"question": "Can you name the works that focus on fine-tuning language models only, without updating retrievers, in retrieval-augmented generation?", "answer": ["Leveraging Passage Retrieval with Generative Models for Open Domain\n Question Answering", "Pre-computed memory or on-the-fly encoding? A hybrid approach to\n retrieval augmentation makes the most of your compute", "GLIMMER: generalized late-interaction memory reranker"], "answer_arxiv_id": ["2007.01282", "2301.10448", "2306.10231"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_5778"} +{"question": "Which works used retrieval-based methods in infusing knowledge into Language Model decoding?", "answer": ["Leveraging Passage Retrieval with Generative Models for Open Domain\n Question Answering", "REALM: Retrieval-Augmented Language Model Pre-Training", "Generalization through Memorization: Nearest Neighbor Language Models", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2007.01282", "2002.08909", "1911.00172", "2005.11401"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_5779"} +{"question": "Are there any studies on improving motion generation results by applying the diffusion formulation to human motion?", "answer": ["MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "FLAME: Free-form Language-based Motion Synthesis & Editing"], "answer_arxiv_id": ["2208.15001", "2209.00349"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_5780"} +{"question": "Can you provide papers that select samples by topological order?", "answer": ["Topological Experience Replay", "DisCor: Corrective Feedback in Reinforcement Learning via Distribution Correction"], "answer_arxiv_id": ["2203.15845", "2003.07305"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_5781"} +{"question": "What works extended the results of finding Nash equilibria to linear and general function approximation setting?", "answer": ["Learning Zero-Sum Simultaneous-Move Markov Games Using Function Approximation and Correlated Equilibrium", "The Power of Exploiter: Provable Multi-Agent RL in Large State Spaces", "Towards General Function Approximation in Zero-Sum Markov Games"], "answer_arxiv_id": ["2002.07066", "2106.03352", "2107.14702"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_5782"} +{"question": "Who proposed methods for pruning vision transformers?", "answer": ["Learned Token Pruning for Transformers"], "answer_arxiv_id": ["2107.00910"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_5783"} +{"question": "What are the papers in which instructions are induced into the model by fine-tuning base models with hundreds of specific tasks?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2109.01652", "2210.11416"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_5784"} +{"question": "Which works are included in the graphics-based simulators for sensor simulation in self-driving?", "answer": ["CARLA: An Open Urban Driving Simulator", "AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles"], "answer_arxiv_id": ["1711.03938", "1705.05065"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_5785"} +{"question": "What research papers utilize deep learning-based methods for image quality analysis?", "answer": ["Blind Image Quality Assessment Using A Deep Bilinear Convolutional\n Neural Network", "MUSIQ: Multi-scale Image Quality Transformer", "Image Quality Assessment: Unifying Structure and Texture Similarity", "TOPIQ: A Top-down Approach from Semantics to Distortions for Image\n Quality Assessment", "FAST-VQA: Efficient End-to-end Video Quality Assessment with Fragment\n Sampling"], "answer_arxiv_id": ["1907.02665", "2108.05997", "2004.07728", "2308.03060", "2207.02595"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_5786"} +{"question": "What are the studies that distill 3D reconstructions from 2D generative models trained on Internet images?", "answer": ["Generative Adversarial Networks", "HoloGAN: Unsupervised learning of 3D representations from natural images", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis", "Efficient Geometry-aware 3D Generative Adversarial Networks", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as\n General Image Priors", "DreamFusion: Text-to-3D using 2D Diffusion", "Zero-1-to-3: Zero-shot One Image to 3D Object", "Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors", "One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization", "DMV3D: Denoising Multi-View Diffusion using 3D Large Reconstruction\n Model", "DreamCraft3D: Hierarchical 3D Generation with Bootstrapped Diffusion\n Prior", "Wonder3D: Single Image to 3D using Cross-Domain Diffusion", "ConsistNet: Enforcing 3D Consistency for Multi-view Images Diffusion", "Consistent123: Improve Consistency for One Image to 3D Object Synthesis", "DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image", "MVDream: Multi-view Diffusion for 3D Generation", "Collaborative Score Distillation for Consistent Visual Synthesis"], "answer_arxiv_id": ["1406.2661", "1904.01326", "2012.00926", "2112.07945", "2006.11239", "2011.13456", "2212.03267", "2209.14988", "2303.11328", "2306.17843", "2306.16928", "2311.09217", "2310.16818", "2310.15008", "2310.10343", "2310.08092", "2309.16653", "2309.03453", "2308.16512", "2307.04787"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_5787"} +{"question": "Which studies utilized box embedding but did not consider the box containment in a global view?", "answer": ["Predicting Visual Overlap of Images Through Interpretable Non-Metric Box\n Embeddings", "Word2Box: Capturing Set-Theoretic Semantics of Words using Box\n Embeddings"], "answer_arxiv_id": ["2008.05785", "2106.14361"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_5788"} +{"question": "Could you provide me some works about modeling equivariant interactions in Cartesian space?", "answer": ["Equivariant message passing for the prediction of tensorial properties and molecular spectra", "E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials", "E(n) Equivariant Graph Neural Networks", "TorchMD-NET: Equivariant Transformers for Neural Network based Molecular Potentials"], "answer_arxiv_id": ["2102.03150", "2101.03164", "2102.09844", "2202.02541"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_5789"} +{"question": "Could you provide a study that discussed the potential of pre-trained foundation models in learning all-purpose features?", "answer": ["On the Opportunities and Risks of Foundation Models"], "answer_arxiv_id": ["2108.07258v3"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5790"} +{"question": "What are some research works that apply the Constrained Markov Decision Process framework for safe RL?", "answer": ["A Review of Safe Reinforcement Learning: Methods, Theory and Applications", "Sauté RL: Almost Surely Safe Reinforcement Learning Using State Augmentation", "SAAC: Safe Reinforcement Learning as an Adversarial Game of Actor-Critics"], "answer_arxiv_id": ["2205.10330", "2202.06558", "2204.09424"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_5791"} +{"question": "What studies have been conducted to improve the generalization ability of neural networks through the use of flat minimizers?", "answer": ["Fantastic Generalization Measures and Where to Find Them", "Relative Flatness and Generalization", "Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data", "Optimal Transport Model Distributional Robustness"], "answer_arxiv_id": ["1912.02178", "2001.00939", "1703.11008", "2306.04178"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5792"} +{"question": "Which studies discuss the inherent issues in defining and learning disentangled representations?", "answer": ["Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations"], "answer_arxiv_id": ["1811.12359"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_5793"} +{"question": "What work found that SSL is more robust to data imbalance?", "answer": ["Self-supervised Learning is More Robust to Dataset Imbalance"], "answer_arxiv_id": ["2110.05025"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_5794"} +{"question": "What are some works about test-time adaptation for robustness to distribution shifting?", "answer": ["Adversarial Attacks are Reversible with Natural Supervision", "MEMO: Test Time Robustness via Adaptation and Augmentation"], "answer_arxiv_id": ["2103.14222", "2110.09506"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_5795"} +{"question": "What works developed datasets for IQA on in-the-wild photographs?", "answer": ["KonIQ-10k: An ecologically valid database for deep learning of blind\n image quality assessment"], "answer_arxiv_id": ["1910.06180"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_5796"} +{"question": "Could you provide me the works following DETR3D which adopt object queries?", "answer": ["DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries", "PETR: Position Embedding Transformation for Multi-View 3D Object Detection", "Polar Parametrization for Vision-based Surround-View 3D Detection"], "answer_arxiv_id": ["2110.06922", "2203.05625", "2206.10965"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_5797"} +{"question": "What are some examples of SSL analysis, particularly on contrastive learning?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Understanding Self-supervised Learning with Dual Deep Networks"], "answer_arxiv_id": ["1902.09229v1", "2005.10242", "2106.04156", "2010.00578"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_5798"} +{"question": "What research has proposed estimators for regularized density ratios with L2-convergence guarantees in the context of policy optimization?", "answer": ["Offline Reinforcement Learning with Realizability and Single-policy Concentrability"], "answer_arxiv_id": ["2202.04634"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_5799"} +{"question": "Which papers discuss masked representation learning in natural language processing and computer vision?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "ERNIE: Enhanced Representation through Knowledge Integration", "Cross-lingual Language Model Pretraining", "Should You Mask 15% in Masked Language Modeling?", "BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "SimMIM: A Simple Framework for Masked Image Modeling", "Context Autoencoder for Self-Supervised Representation Learning", "Object-wise Masked Autoencoders for Fast Pre-training", "Masked Autoencoders As Spatiotemporal Learners", "PeCo: Perceptual Codebook for BERT Pre-training of Vision Transformers"], "answer_arxiv_id": ["1810.04805", "1907.11692", "1904.09223", "1901.07291", "2202.08005", "2106.08254", "2111.06377", "2112.09133", "2111.09886", "2202.03026", "2205.14338", "2205.09113", "2111.12710"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_5800"} +{"question": "Which works observed similar linear trends in sub-type shifts?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations", "Breeds: Benchmarks for Subpopulation Shift"], "answer_arxiv_id": ["1903.12261", "2008.04859"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_5801"} +{"question": "Could you provide me some studies where structured-sparsity optimization problem were formulated and a method to solve it has been proposed?", "answer": ["Only Train Once: A One-Shot Neural Network Training And Pruning\n Framework", "Learning Structured Sparsity in Deep Neural Networks", "Group Sparsity: The Hinge Between Filter Pruning and Decomposition for\n Network Compression", "OTOV2: Automatic, Generic, User-Friendly"], "answer_arxiv_id": ["2107.07467", "1608.03665", "2003.08935", "2303.06862"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_5802"} +{"question": "Could you provide me some studies in which the domain generalization is formulated as a minimax optimization problem?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "On Generalization and Regularization via Wasserstein Distributionally Robust Optimization"], "answer_arxiv_id": ["1911.08731", "2212.05716"], "source_meta": {"published_time": "20210705"}, "qid": "AutoScholarQuery_train_5803"} +{"question": "Which works focus on alternative approaches to entropy-regularized il using the Wasserstein metric, labelling of sparse proxy rewards, feature matching, maximum likelihood, and matching state marginals?", "answer": ["Primal Wasserstein Imitation Learning", "SQIL: Imitation Learning via Reinforcement Learning with Sparse Rewards", "Maximum Entropy Deep Inverse Reinforcement Learning", "Guided Cost Learning: Deep Inverse Optimal Control via Policy Optimization"], "answer_arxiv_id": ["2006.04678", "1905.11108", "1507.04888", "1603.00448"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5804"} +{"question": "Which studies propose methods for uncertainty estimation in deep regression models?", "answer": ["Deep Evidential Regression", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "NOMU: Neural Optimization-based Model Uncertainty", "What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "Accurate Uncertainties for Deep Learning Using Calibrated Regression", "Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift", "Distribution Calibration for Regression", "Graph Posterior Network: Bayesian Predictive Uncertainty for Node Classification", "CRUDE: Calibrating Regression Uncertainty Distributions Empirically"], "answer_arxiv_id": ["1910.02600", "1506.02142", "2102.13640", "1703.04977", "1807.00263", "1906.02530", "1905.06023", "2110.14012", "2005.12496"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_5805"} +{"question": "Could you provide me with the reference that indicates removing positional encoding aids in context expansion?", "answer": ["Transformer Language Models without Positional Encodings Still Learn Positional Information"], "answer_arxiv_id": ["2203.16634"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_5806"} +{"question": "Which papers extended research on the linear predictor under GD/SGD for linearly separable data to linear fully-connected networks and linear Convolutional Neural Networks (CNNs)?", "answer": ["Implicit Bias of Gradient Descent on Linear Convolutional Networks"], "answer_arxiv_id": ["1806.00468"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_5807"} +{"question": "What are some studies about editing image content with text instructions?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "DeltaEdit: Exploring Text-free Training for Text-Driven Image\n Manipulation", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "SINE: SINgle Image Editing with Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2108.01073", "2303.06285", "2211.09800", "2212.04489"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_5808"} +{"question": "Which work trains video diffusion models on paired text-video datasets for video generation?", "answer": ["Structure and Content-Guided Video Synthesis with Diffusion Models"], "answer_arxiv_id": ["2302.03011"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_5809"} +{"question": "Could you provide me some studies about spatial GNNs?", "answer": ["A Comprehensive Survey on Graph Neural Networks", "Neural Message Passing for Quantum Chemistry", "Graph Attention Networks", "Semi-Supervised Classification with Graph Convolutional Networks", "Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering"], "answer_arxiv_id": ["1901.00596", "1704.01212", "1710.10903", "1609.02907", "1606.09375"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_5810"} +{"question": "What learning-based metrics have been used in the field of evaluation of text generation?", "answer": ["BLEURT: Learning Robust Metrics for Text Generation", "LENS: A Learnable Evaluation Metric for Text Simplification", "Not All Errors are Equal: Learning Text Generation Metrics using\n Stratified Error Synthesis", "SESCORE2: Learning Text Generation Evaluation via Synthesizing Realistic\n Mistakes"], "answer_arxiv_id": ["2004.04696", "2212.09739", "2210.05035", "2212.09305"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_5811"} +{"question": "What work proposed using a frozen vision encoder and a large language model for cross-modality alignment?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_5812"} +{"question": "What research has been conducted that uses diffusion models for video editing and generation?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Show-1: Marrying Pixel and Latent Diffusion Models for Text-to-Video\n Generation"], "answer_arxiv_id": ["2212.11565", "2303.09535", "2304.08818", "2309.15818"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_5813"} +{"question": "Which papers explore how Transformers can act as programmable computers?", "answer": ["Looped Transformers as Programmable Computers"], "answer_arxiv_id": ["2301.13196"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5814"} +{"question": "Can you list the papers that discussed using a reference dataset to detect the out-of-context use of images?", "answer": ["Multimedia Semantic Integrity Assessment Using Joint Embedding Of Images\n And Text", "Deep Multimodal Image-Repurposing Detection", "AIRD: Adversarial Learning Framework for Image Repurposing Detection"], "answer_arxiv_id": ["1707.01606", "1808.06686", "1903.00788"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_5815"} +{"question": "What works have documented semantic changes in word embeddings over time?", "answer": ["Diachronic Word Embeddings Reveal Statistical Laws of Semantic Change"], "answer_arxiv_id": ["1605.09096"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_5816"} +{"question": "What works tackled improvement of performance of nODEs by augmenting extra dimensions to the phase space?", "answer": ["Augmented Neural ODEs"], "answer_arxiv_id": ["1904.01681"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_5817"} +{"question": "What work has tried to change the mesh geometry for 3D attacks but failed to perform transferable targeted attacks?", "answer": ["MeshAdv: Adversarial Meshes for Visual Recognition"], "answer_arxiv_id": ["1810.05206"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_5818"} +{"question": "Can you mention the studies that have adopted customized prompts through language models?", "answer": ["What does a platypus look like? Generating customized prompts for\n zero-shot image classification", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2209.03320", "2103.00020"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_5819"} +{"question": "Any works that introduce a dataset related to fractured object reassembly in 3D?", "answer": ["Breaking Bad: A Dataset for Geometric Fracture and Reassembly"], "answer_arxiv_id": ["2210.11463"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_5820"} +{"question": "Which paper discusses the AGDA-RR algorithm and its convergence rate for the two-sided PŁ objective?", "answer": ["Sampling without Replacement Leads to Faster Rates in Finite-Sum Minimax Optimization"], "answer_arxiv_id": ["2206.02953"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_5821"} +{"question": "In what work it is found that the estimation of the 2D bounding box contributes to the prediction of 3D attributes?", "answer": ["Delving into Localization Errors for Monocular 3D Object Detection"], "answer_arxiv_id": ["2103.16237"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_5822"} +{"question": "What research introduced EquiDyn, a family of dynamic topologies in the context of DL algorithms?", "answer": ["Communication-Efficient Topologies for Decentralized Learning with $O(1)$ Consensus Rate"], "answer_arxiv_id": ["2210.07881v2"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_5823"} +{"question": "Can you provide me examples of pFL methods which keep the classifier locally?", "answer": ["Federated Learning with Personalization Layers", "Exploiting Shared Representations for Personalized Federated Learning"], "answer_arxiv_id": ["1912.00818", "2102.07078"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_5824"} +{"question": "Could you provide me some studies concerning saliency methods for attention-based models?", "answer": ["RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism", "Attend and Diagnose: Clinical Time Series Analysis using Attention Models", "RAIM: Recurrent Attentive and Intensive Model of Multimodal Patient Monitoring Data"], "answer_arxiv_id": ["1608.05745", "1711.03905", "1807.08820v1"], "source_meta": {"published_time": "20210729"}, "qid": "AutoScholarQuery_train_5825"} +{"question": "What work focuses on the multi-agent setting where the adversary inserts a backdoor using its behavior in the environment?", "answer": ["BACKDOORL: Backdoor Attack against Competitive Reinforcement Learning"], "answer_arxiv_id": ["2105.00579v3"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_5826"} +{"question": "Which papers relied on Generative Adversarial Networks for text-guided image editing?", "answer": ["Generative Adversarial Networks", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "DeltaEdit: Exploring Text-free Training for Text-Driven Image\n Manipulation", "Predict, Prevent, and Evaluate: Disentangled Text-Driven Image\n Manipulation Empowered by Pre-Trained Vision-Language Model"], "answer_arxiv_id": ["2203.00667", "2103.17249", "2108.00946", "2303.06285", "2111.13333"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_5827"} +{"question": "Could you mention works that stand out in image-to-image regression with regards to childhood obesity?", "answer": ["Image-to-Image Regression with Distribution-Free Uncertainty Quantification and Applications in Imaging"], "answer_arxiv_id": ["2202.05265"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_5828"} +{"question": "What research has looked into data augmentation as a means to explain the CPE phenomenon?", "answer": ["Bayesian Neural Network Priors Revisited"], "answer_arxiv_id": ["2102.06571"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_5829"} +{"question": "What research papers discuss LiDAR-based methods in 3D object detection?", "answer": ["Transformation-Equivariant 3D Object Detection for Autonomous Driving", "From Points to Parts: 3D Object Detection from Point Cloud with\n Part-aware and Part-aggregation Network", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "STD: Sparse-to-Dense 3D Object Detector for Point Cloud", "SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud", "Sparse Fuse Dense: Towards High Quality 3D Detection with Depth\n Completion"], "answer_arxiv_id": ["2211.11962", "1907.03670", "1812.05784", "1907.10471", "2104.09804", "2203.09780"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5830"} +{"question": "Can you provide me some works that propose methods to address shortcut learning where the attribute is known?", "answer": ["Improving Out-of-Distribution Robustness via Selective Augmentation", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "On Feature Learning in the Presence of Spurious Correlations", "Spread Spurious Attribute: Improving Worst-group Accuracy with Spurious Attribute Estimation", "Pulling Up by the Causal Bootstraps: Causal Data Augmentation for Pre-training Debiasing"], "answer_arxiv_id": ["2201.00299", "1911.08731", "2210.11369", "2204.02070", "2108.12510"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_5831"} +{"question": "Which papers focus on learning-based inverse rendering?", "answer": ["Inverse Rendering of Translucent Objects using Physical and Neural\n Renderers", "Neural Inverse Rendering of an Indoor Scene from a Single Image", "Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying\n Lighting and SVBRDF from a Single Image", "Learning Indoor Inverse Rendering with 3D Spatially-Varying Lighting", "IRISformer: Dense Vision Transformers for Single-Image Inverse Rendering\n in Indoor Scenes", "Two-shot Spatially-varying BRDF and Shape Estimation", "Shape and Material Capture at Home"], "answer_arxiv_id": ["2305.08336", "1901.02453", "1905.02722", "2109.06061", "2206.08423", "2004.00403", "2104.06397"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_5832"} +{"question": "Which works have applied zeroth-order optimization to perform various actions, such as adversarial attacks, hyperparameter optimization or transfer learning on black-box models?", "answer": ["Optimal rates for zero-order convex optimization: the power of two function evaluations", "Zeroth-Order Online Alternating Direction Method of Multipliers: Convergence Analysis and Applications", "FLoRA: Single-shot Hyper-parameter Optimization for Federated Learning", "Transfer Learning without Knowing: Reprogramming Black-box Machine Learning Models with Scarce Data and Limited Resources"], "answer_arxiv_id": ["1312.2139", "1710.07804v2", "2112.08524", "2007.08714"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_5833"} +{"question": "Which papers address the detection of video forgery in deepfake detection by focusing on temporal inconsistency?", "answer": ["Lips Don't Lie: A Generalisable and Robust Approach to Face Forgery\n Detection", "Exploring Temporal Coherence for More General Video Face Forgery\n Detection", "AltFreezing for More General Video Face Forgery Detection"], "answer_arxiv_id": ["2012.07657", "2108.06693", "2307.08317"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_5834"} +{"question": "Which papers present conditional behavior cloning methods that can be extended to the online setup?", "answer": ["Online Decision Transformer"], "answer_arxiv_id": ["2202.05607"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_5835"} +{"question": "What work introduced the False Negative Aware Contrastive strategy for single sound source localization?", "answer": ["Learning Audio-Visual Source Localization via False Negative Aware\n Contrastive Learning"], "answer_arxiv_id": ["2303.11302"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_5836"} +{"question": "What works have highlighted the relationship between dataset features and margins?", "answer": ["Hold me tight! Influence of discriminative features on deep network boundaries"], "answer_arxiv_id": ["2002.06349"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_5837"} +{"question": "What papers extended the finetuning process to multi-step RLFT by treating the denoising process as a multi-step Markov Decision Process (MDP)?", "answer": ["Training Diffusion Models with Reinforcement Learning", "DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2305.13301", "2305.16381"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_5838"} +{"question": "Which paper introduced the S4 model as an alternative for capturing long-range dependencies?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces"], "answer_arxiv_id": ["2111.00396"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_5839"} +{"question": "Where can I find the concept of Recall@1 metric?", "answer": ["Efficient Estimation of Word Representations in Vector Space"], "answer_arxiv_id": ["1301.3781"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_5840"} +{"question": "What research has been conducted into membership inference attacks?", "answer": ["Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "Membership Inference Attacks Against Machine Learning Models"], "answer_arxiv_id": ["1709.01604", "1610.05820"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_5841"} +{"question": "What works introduce network compression techniques for parameter reduction by pruning redundancy?", "answer": ["Data-free parameter pruning for Deep Neural Networks", "Data-Driven Sparse Structure Selection for Deep Neural Networks"], "answer_arxiv_id": ["1507.06149", "1707.01213"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_5842"} +{"question": "What researches highlight the importance of relative object pose estimation?", "answer": ["RelPose: Predicting Probabilistic Relative Rotation for Single Objects\n in the Wild", "RelPose++: Recovering 6D Poses from Sparse-view Observations", "3D-Aware Hypothesis & Verification for Generalizable Relative Object\n Pose Estimation", "PoseDiffusion: Solving Pose Estimation via Diffusion-aided Bundle Adjustment"], "answer_arxiv_id": ["2208.05963", "2305.04926", "2310.03534", "2306.15667v4"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_5843"} +{"question": "What are the works about the successful application of diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Diffusion Models Beat GANs on Image Synthesis", "Improved Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Prompt-to-Prompt Image Editing with Cross Attention Control", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "Video Diffusion Models", "Flexible Diffusion Modeling of Long Videos", "Diffusion Models for Video Prediction and Infilling", "LION: Latent Point Diffusion Models for 3D Shape Generation", "3D Shape Generation and Completion through Point-Voxel Diffusion", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "Learning to Generate Realistic LiDAR Point Clouds", "Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed", "Progressive Distillation for Fast Sampling of Diffusion Models", "Score-Based Generative Modeling with Critically-Damped Langevin Diffusion", "Denoising Diffusion Implicit Models", "Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality", "Cascaded Diffusion Models for High Fidelity Image Generation", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["2006.11239", "2011.13456", "1503.03585", "2105.05233", "2102.09672", "2112.10752", "2211.01324", "2112.10741", "2204.06125", "2208.01626", "2108.01073", "2201.09865", "2204.03458", "2205.11495", "2206.07696", "2210.06978", "2104.03670", "2103.01458", "2209.03954", "2101.02388", "2202.00512", "2112.07068", "2010.02502", "2202.05830", "2106.15282", "2106.05931"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_5844"} +{"question": "What studies have shown that combining periodic kernel and noisy expected improvement acquisition function improves the performance of BO?", "answer": ["Constrained Bayesian Optimization with Noisy Experiments"], "answer_arxiv_id": ["1706.07094"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_5845"} +{"question": "Which research papers discussed the improvements made for capturing high-frequency details in initial versions of INRs?", "answer": ["Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains"], "answer_arxiv_id": ["2006.10739"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_5846"} +{"question": "Can you provide some papers revolving around generation methods that utilize point clouds and meshes in 3D representation?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "LION: Latent Point Diffusion Models for 3D Shape Generation", "VPP: Efficient Conditional 3D Generation via Voxel-Point Progressive\n Representation", "MeshDiffusion: Score-based Generative 3D Mesh Modeling"], "answer_arxiv_id": ["2212.08751", "2103.01458", "2210.06978", "2307.16605", "2303.08133"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_5847"} +{"question": "Which researchers carried out a systematic analysis of ChatGPT’s zero-shot abilities on representative NLP tasks?", "answer": ["Is ChatGPT a General-Purpose Natural Language Processing Task Solver?"], "answer_arxiv_id": ["2302.06476"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_5848"} +{"question": "Which works recommend reporting points along the rate-distortion curve in VAEs?", "answer": ["Fixing a Broken ELBO", "Evaluating Lossy Compression Rates of Deep Generative Models"], "answer_arxiv_id": ["1711.00464", "2008.06653"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_5849"} +{"question": "Can you cite work regarding the proposal or examination of gradient perturbation?", "answer": ["Differentially Private Empirical Risk Minimization Revisited: Faster and More General", "Private Stochastic Convex Optimization with Optimal Rates"], "answer_arxiv_id": ["1802.05251v1", "1908.09970"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_5850"} +{"question": "Which studies introduced the general formulation of Multi-armed bandits with feedback graphs?", "answer": ["From Bandits to Experts: On the Value of Side-Observations"], "answer_arxiv_id": ["1106.2436"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_5851"} +{"question": "Which papers employed explicit representations like meshes in scene representation?", "answer": ["Deferred Neural Rendering: Image Synthesis using Neural Textures", "NeuMesh: Learning Disentangled Neural Mesh-based Implicit Field for Geometry and Texture Editing"], "answer_arxiv_id": ["1904.12356", "2207.11911"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_5852"} +{"question": "Could you tell me about some works explaining 3DMM-based methods in talking head synthesis?", "answer": ["Deep Video Portraits", "Audio-driven Talking Face Video Generation with Learning-based\n Personalized Head Pose", "Face2Face: Real-time Face Capture and Reenactment of RGB Videos", "3D Guided Fine-Grained Face Manipulation", "FaceFormer: Speech-Driven 3D Facial Animation with Transformers"], "answer_arxiv_id": ["1805.11714", "2002.10137", "2007.14808", "1902.08900", "2112.05329"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_5853"} +{"question": "Could you give me papers where PointNet is used as a method for point cloud analysis?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space", "PointConv: Deep Convolutional Networks on 3D Point Clouds", "KPConv: Flexible and Deformable Convolution for Point Clouds", "Dynamic Graph CNN for Learning on Point Clouds"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1811.07246", "1904.08889", "1801.07829"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_5854"} +{"question": "What studies have incorporated Stochastic Polyak Step size (SPS) with a line search technique?", "answer": ["Stochastic Polyak Step-size for SGD: An Adaptive Learning Rate for Fast Convergence"], "answer_arxiv_id": ["2002.10542v3"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_5855"} +{"question": "Can you list some studies that control virtual avatars through VR/AR devices?", "answer": ["LoBSTr: Real-time Lower-body Pose Prediction from Sparse Upper-body\n Tracking Signals", "FLAG: Flow-based 3D Avatar Generation from Sparse Observations", "AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion\n Sensing", "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking\n Inputs with Diffusion Model", "HMD-NeMo: Online 3D Avatar Motion Generation From Sparse Observations", "QuestSim: Human Motion Tracking from Sparse Sensors with Simulated\n Avatars", "QuestEnvSim: Environment-Aware Simulated Motion Tracking from Sparse\n Sensors"], "answer_arxiv_id": ["2103.01500", "2203.05789", "2207.13784", "2304.08577", "2308.11261", "2209.09391", "2306.05666"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_5856"} +{"question": "What works discuss quantization as a method to reduce the model size and accelerate the inference process in Graph Neural Networks?", "answer": ["SGQuant: Squeezing the Last Bit on Graph Neural Networks with Specialized Quantization", "Degree-Quant: Quantization-Aware Training for Graph Neural Networks", "Bi-GCN: Binary Graph Convolutional Network", "Binary Graph Neural Networks", "Binarized Graph Neural Network", "Meta-Aggregator: Learning to Aggregate for 1-bit Graph Neural Networks", "VQ-GNN: A Universal Framework to Scale-up11Graph Neural Networks using Vector Quantization"], "answer_arxiv_id": ["2007.05100", "2008.05000", "2010.07565", "2012.15823", "2004.11147", "2109.12872", "2110.14363"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_5857"} +{"question": "Which papers proposed to train a generator network to improve NeRF renderings and an image discriminator network to provide feedback that can be used to improve the reconstruction in a multiview-consistent manner?", "answer": ["GANeRF: Leveraging Discriminators to Optimize Neural Radiance Fields"], "answer_arxiv_id": ["2306.06044"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_5858"} +{"question": "Which study involved using an adversarial approach by optimizing the policy with respect to a worst-case dynamics model?", "answer": ["RAMBO-RL: Robust Adversarial Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2204.12581"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_5859"} +{"question": "What work showed that T5 could answer a large portion of knowledge-intensive questions after fine-tuning on open-book question-answer pairs?", "answer": ["How Much Knowledge Can You Pack Into the Parameters of a Language Model?"], "answer_arxiv_id": ["2002.08910"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_5860"} +{"question": "Which studies provide empirical evidence and provable convergence rate guarantees for entropy-regularized MDPs in reinforcement learning?", "answer": ["A Unified View of Entropy-Regularized Markov Decision Processes", "Reinforcement Learning with Deep Energy-Based Policies", "On the Global Convergence Rates of Softmax Policy Gradient Methods", "Fast Global Convergence of Natural Policy Gradient Methods with Entropy Regularization"], "answer_arxiv_id": ["1705.07798", "1702.08165", "2005.06392", "2007.06558"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_5861"} +{"question": "What works feature a voxel-based method to transform unordered points into regular grids for feature extraction in 3D object detection?", "answer": ["Center-based 3D Object Detection and Tracking"], "answer_arxiv_id": ["2006.11275"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5862"} +{"question": "What works introduced methods to adaptively fuse multi-modal features in the presence of irrelevant backgrounds?", "answer": ["UniCat: Crafting a Stronger Fusion Baseline for Multimodal\n Re-Identification", "TOP-ReID: Multi-spectral Object Re-Identification with Token Permutation"], "answer_arxiv_id": ["2310.18812", "2312.09612"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_5863"} +{"question": "Can you list down the works that used additional guidance input to improve controllability of text-to-image generation?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "High-Fidelity Guided Image Synthesis with Latent Diffusion Models", "SpaText: Spatio-Textual Representation for Controllable Image Generation", "MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation", "Collage Diffusion", "SceneComposer: Any-Level Semantic Image Synthesis", "ReCo: Region-Controlled Text-to-Image Generation", "Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion"], "answer_arxiv_id": ["2108.01073", "2108.02938", "2211.17084", "2211.14305", "2302.08113", "2303.00262", "2211.11742", "2211.15518", "2302.05543", "2302.08453", "2208.12242", "2208.01618"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_5864"} +{"question": "What studies propose methods to simplify GNNs by using MLPs to scale them?", "answer": ["Graph Attention Multi-Layer Perceptron", "Simplifying Graph Convolutional Networks", "SIGN: Scalable Inception Graph Neural Networks", "Scalable and Adaptive Graph Neural Networks with Self-Label-Enhanced Training", "Combining Label Propagation and Simple Models out-performs Graph Neural Networks", "Graph-MLP: Node Classification without Message Passing in Graph"], "answer_arxiv_id": ["2206.04355", "1902.07153", "2004.11198", "2104.09376", "2010.13993", "2106.04051"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5865"} +{"question": "What studies proposed a projection-based pre-training framework for PCQA tasks?", "answer": ["Point Cloud Quality Assessment: Dataset Construction and Learning-based\n No-Reference Metric", "Perceptual Quality Assessment of Colored 3D Point Clouds"], "answer_arxiv_id": ["2012.11895", "2111.05474"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_5866"} +{"question": "What works focused on improving training recipe in Vision Transformers (ViTs) by leveraging distillation?", "answer": ["Training data-efficient image transformers & distillation through\n attention"], "answer_arxiv_id": ["2012.12877"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_5867"} +{"question": "What papers discuss the role of human annotation in the creation of datasets used to benchmark toxicity, harm and hate speech?", "answer": ["Toxicity Detection: Does Context Really Matter?"], "answer_arxiv_id": ["2006.00998"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_5868"} +{"question": "Could you provide me some studies that use off-the-shell tools to generate grasping poses?", "answer": ["Learning joint reconstruction of hands and manipulated objects", "DexGraspNet: A Large-Scale Robotic Dexterous Grasp Dataset for General\n Objects Based on Simulation", "UniDexGrasp: Universal Robotic Dexterous Grasping via Learning Diverse\n Proposal Generation and Goal-Conditioned Policy", "UniDexGrasp++: Improving Dexterous Grasping Policy Learning via\n Geometry-aware Curriculum and Iterative Generalist-Specialist Learning"], "answer_arxiv_id": ["1904.05767", "2210.02697", "2303.00938", "2304.00464"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_5869"} +{"question": "To which study do we owe the discovery of neural networks' tendency to learn shared patterns before resorting to memorization when given real data?", "answer": ["A Closer Look at Memorization in Deep Networks"], "answer_arxiv_id": ["1706.05394"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_5870"} +{"question": "Which studies focused on examining failure modes for generative models?", "answer": ["On GANs and GMMs", "Assessing Generative Models via Precision and Recall"], "answer_arxiv_id": ["1805.12462", "1806.00035"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_5871"} +{"question": "Could you provide me some works that utilize machine learning or deep learning classifiers to identify keyphrases?", "answer": ["Keyphrase Generation with Correlation Constraints", "DivGraphPointer: A Graph Pointer Network for Extracting Diverse\n Keyphrases", "Keyphrase Extraction with Span-based Feature Representations"], "answer_arxiv_id": ["1808.07185", "1905.07689", "2002.05407"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_5872"} +{"question": "In the two-stage framework described, which model is employed as the generative model and what is used to represent a text prompt?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5873"} +{"question": "Could you provide me some works about the reinforcement learning from human feedback (RLFT) of diffusion models?", "answer": ["Aligning Text-to-Image Models using Human Feedback", "Training Diffusion Models with Reinforcement Learning", "DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2302.12192", "2305.13301", "2305.16381"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_5874"} +{"question": "Could you mention some recent research on few-step text-to-image generation?", "answer": ["Latent Consistency Models: Synthesizing High-Resolution Images with\n Few-Step Inference", "InstaFlow: One Step is Enough for High-Quality Diffusion-Based\n Text-to-Image Generation"], "answer_arxiv_id": ["2310.04378", "2309.06380"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_5875"} +{"question": "Which works discuss the mixture of experts (MoE) in the context of natural language processing?", "answer": ["GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding", "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity", "M6-T: Exploring Sparse Expert Models and Beyond", "BASE Layers: Simplifying Training of Large, Sparse Models"], "answer_arxiv_id": ["2006.16668", "2101.03961", "2105.15082", "2103.16716"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_5876"} +{"question": "What work is about nucleus sampling, a decoding algorithm for neural text generation?", "answer": ["The Curious Case of Neural Text Degeneration"], "answer_arxiv_id": ["1904.09751"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_5877"} +{"question": "What works aimed at transferring structured data into natural language text?", "answer": ["Learning from Multiple Sources for Data-to-Text and Text-to-Data"], "answer_arxiv_id": ["2302.11269v1"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_5878"} +{"question": "Which Heterogeneous GNNs assume each edge has a pre-defined type and take such types into consideration during aggregation?", "answer": ["Modeling Relational Data with Graph Convolutional Networks", "Heterogeneous Network Representation Learning: A Unified Framework with Survey and Benchmark"], "answer_arxiv_id": ["1703.06103v4", "2004.00216"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_5879"} +{"question": "Which works present the method UNet for biomedical image segmentation?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation"], "answer_arxiv_id": ["1505.04597"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_5880"} +{"question": "What are some papers about enabling interactions with objects in the field of motion generation?", "answer": ["GRAB: A Dataset of Whole-Body Human Grasping of Objects", "COUCH: Towards Controllable Human-Chair Interactions", "GIMO: Gaze-Informed Human Motion Prediction in Context"], "answer_arxiv_id": ["2008.11200", "2205.00541", "2204.09443"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_5881"} +{"question": "Which studies have used heuristics or fixed grammars to generate driving scenes?", "answer": ["Meta-Sim: Learning to Generate Synthetic Datasets"], "answer_arxiv_id": ["1904.11621"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_5882"} +{"question": "Can you provide works that studied user-level differential privacy?", "answer": ["Learning Differentially Private Recurrent Language Models", "Beyond Inferring Class Representatives: User-Level Privacy Leakage From Federated Learning"], "answer_arxiv_id": ["1710.06963", "1812.00535"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_5883"} +{"question": "Which papers present applications of Graph Neural Networks in molecular property prediction?", "answer": ["Recipe for a General, Powerful, Scalable Graph Transformer"], "answer_arxiv_id": ["2205.12454"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_5884"} +{"question": "Are there any research works about data-driven algorithm design?", "answer": ["A PAC Approach to Application-Specific Algorithm Selection", "How much data is sufficient to learn high-performing algorithms? Generalization guarantees for data-driven algorithm design"], "answer_arxiv_id": ["1511.07147", "1908.02894"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_5885"} +{"question": "Could you provide me some works about disentanglement focused methods that were inspired by Variational Autoencoder?", "answer": ["Disentangling by Factorising", "Isolating Sources of Disentanglement in Variational Autoencoders"], "answer_arxiv_id": ["1802.05983", "1802.04942"], "source_meta": {"published_time": "20200614"}, "qid": "AutoScholarQuery_train_5886"} +{"question": "What works use ALIGN in conjunction with language-based Foundation Models?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2102.05918"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_5887"} +{"question": "Which works develop the concept of grid-based neural fields, notably used for 3D shape and geometry modeling?", "answer": ["3D-R2N2: A Unified Approach for Single and Multi-view 3D Object\n Reconstruction", "Multi-view Supervision for Single-view Reconstruction via Differentiable\n Ray Consistency", "Learning a Multi-View Stereo Machine", "Convolutional Occupancy Networks", "Local Deep Implicit Functions for 3D Shape", "Occupancy Networks: Learning 3D Reconstruction in Function Space"], "answer_arxiv_id": ["1604.00449", "1704.06254", "1708.05375", "2003.04618", "1912.06126", "1812.03828"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5888"} +{"question": "What works proposed semantically preserving methods to produce natural language adversarial examples?", "answer": ["Evaluating the Susceptibility of Pre-Trained Language Models via Handcrafted Adversarial Examples", "Generating Natural Language Attacks in a Hard Label Black Box Setting", "A Geometry-Inspired Attack for Generating Natural Language Adversarial Examples", "Evaluating the Robustness of Neural Language Models to Input Perturbations", "Reevaluating Adversarial Examples in Natural Language", "Generating Natural Language Adversarial Examples on a Large Scale with Generative Models", "PromptAttack: Prompt-based Attack for Language Models via Gradient Search", "On the Transferability of Adversarial Attacks against Neural Text Classifier", "Generating Fluent Adversarial Examples for Natural Languages", "On Robustness of Prompt-based Semantic Parsing with Large Pre-trained Language Model: An Empirical Study on Codex"], "answer_arxiv_id": ["2209.02128", "2012.14956", "2010.01345", "2108.12237v1", "2004.14174", "2003.10388", "2209.01882v1", "2011.08558", "2007.06174", "2301.12868"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_5889"} +{"question": "Which works in the related literature cover activation-based techniques?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "CAMERAS: Enhanced Resolution And Sanity preserving Class Activation Mapping for image saliency", "Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks"], "answer_arxiv_id": ["1610.02391", "2106.10649", "1910.01279"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_5890"} +{"question": "Any papers that explored fairness problem in text classification tasks?", "answer": ["The Authors Matter: Understanding and Mitigating Implicit Bias in Deep Text Classification", "Multi-Dimensional Gender Bias Classification"], "answer_arxiv_id": ["2105.02778", "2005.00614"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_5891"} +{"question": "What studies have made efforts on decomposing the generation process into distinct phases?", "answer": ["Bottom-Up Abstractive Summarization", "Improving the Similarity Measure of Determinantal Point Processes for\n Extractive Multi-Document Summarization", "Summary-Source Proposition-level Alignment: Task, Datasets and\n Supervised Baseline", "Analyzing Sentence Fusion in Abstractive Summarization", "Learning to Fuse Sentences with Transformers for Summarization", "Controlled Text Reduction", "Dont Add, dont Miss: Effective Content Preserving Generation from\n Pre-Selected Text Spans", "Revisiting Sentence Union Generation as a Testbed for Text Consolidation"], "answer_arxiv_id": ["1808.10792", "1906.00072", "2009.00590", "1910.00203", "2010.03726", "2210.13449", "2310.09017", "2305.15605"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_5892"} +{"question": "Which work introduced Adversarial Weight Perturbation (AWP) to enhance model robustness?", "answer": ["Adversarial Weight Perturbation Helps Robust Generalization"], "answer_arxiv_id": ["2004.05884"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_5893"} +{"question": "What studies further improve effectiveness in prompt tuning by enabling optimization in the embedding space?", "answer": ["GPT Understands, Too", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2103.10385", "2104.08691", "2101.00190"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_5894"} +{"question": "Could you tell me research papers that examined prompting strategies through input intervention?", "answer": ["Large Language Models are Zero-Shot Reasoners", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "The Impact of Reasoning Step Length on Large Language Models"], "answer_arxiv_id": ["2205.11916", "2305.10601", "2401.04925"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_5895"} +{"question": "Which research papers in video summarization made use of a multi-modal setup?", "answer": ["Align and Attend: Multimodal Summarization with Dual Contrastive Losses", "VideoXum: Cross-modal Visual and Textural Summarization of Videos"], "answer_arxiv_id": ["2303.07284", "2303.12060"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_5896"} +{"question": "Can you provide me with studies that found deep reinforcement learning algorithms can often overfit to spurious correlations in the observation space?", "answer": ["Observational Overfitting in Reinforcement Learning"], "answer_arxiv_id": ["1912.02975"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_5897"} +{"question": "What studies proposed or used SoftmaxAgg?", "answer": ["DeeperGCN: All You Need to Train Deeper GCNs", "Generalizing Aggregation Functions in GNNs: High-Capacity GNNs via Nonlinear Neighborhood Aggregators"], "answer_arxiv_id": ["2006.07739v1", "2202.09145"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_5898"} +{"question": "Which research focuses on pre-training for mathematical reasoning?", "answer": ["LIME: Learning Inductive Bias for Primitives of Mathematical Reasoning"], "answer_arxiv_id": ["2101.06223"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_5899"} +{"question": "What studies framed the dense video captioning task as set prediction?", "answer": ["End-to-End Dense Video Captioning with Parallel Decoding", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2108.07781", "2005.12872"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_5900"} +{"question": "Which studies used focus groups and workshops to identify specific harm types in model evaluations?", "answer": ["AI’s Regimes of Representation: A Community-centered Study of Text-to-Image Models in South Asia"], "answer_arxiv_id": ["2305.11844"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_5901"} +{"question": "Which work demonstrated that Transformer-based approaches can be used as decoders for panoptic segmentation, despite the slow training issue?", "answer": ["End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2005.12872"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_5902"} +{"question": "What papers detailed a training approach that aligns the features from source and target domains using adversarial losses?", "answer": ["Deep Cocktail Network: Multi-source Unsupervised Domain Adaptation with Category Shift"], "answer_arxiv_id": ["1803.00830"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5903"} +{"question": "Who has worked on breaking down the temporal domain into several subdomains for better long-term temporal integration in physics-informed approaches?", "answer": ["Characterizing possible failure modes in physics-informed neural networks"], "answer_arxiv_id": ["2109.01050"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_5904"} +{"question": "Which works proposed weakly-supervised methods in the field of computational pathology?", "answer": ["Attention-based Deep Multiple Instance Learning", "TransMIL: Transformer based Correlated Multiple Instance Learning for\n Whole Slide Image Classification", "Data Efficient and Weakly Supervised Computational Pathology on Whole\n Slide Images", "DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning\n for Histopathology Whole Slide Image Classification"], "answer_arxiv_id": ["1802.04712", "2106.00908", "2004.09666", "2203.12081"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_5905"} +{"question": "Can you indicate the studies related to debiasing that used different data source or data clean-up methods?", "answer": ["Robot Learning in Homes: Improving Generalization and Reducing Dataset Bias", "Fairness GAN", "FairGAN: Fairness-aware Generative Adversarial Networks"], "answer_arxiv_id": ["1807.07049", "1805.09910", "1805.11202"], "source_meta": {"published_time": "20221110"}, "qid": "AutoScholarQuery_train_5906"} +{"question": "Could you provide me some studies that proposed progressive training techniques?", "answer": ["Progressively Stacking 2.0: A Multi-stage Layerwise Training Method for BERT Training Speedup", "On the Transformer Growth for Progressive BERT Training", "Staged Training for Transformer Language Models", "Automated Progressive Learning for Efficient Training of Vision Transformers", "EfficientTrain: Exploring Generalized Curriculum Learning for Training Visual Backbones"], "answer_arxiv_id": ["2011.13635", "2010.12562", "2203.06211", "2203.14509", "2211.09703"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_5907"} +{"question": "What work developed a diffusion-based approach to create a full-body image?", "answer": ["Person Image Synthesis via Denoising Diffusion Model"], "answer_arxiv_id": ["2211.12500"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_5908"} +{"question": "What papers employ oversampling for imbalanced regression problems, similar to the SMOTE algorithm for classification?", "answer": ["SMOTE: Synthetic Minority Over-sampling Technique"], "answer_arxiv_id": ["1106.1813"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_5909"} +{"question": "Could you provide me some studies about independent and symmetric policy gradient methods for stochastic games?", "answer": ["Gradient play in stochastic games: stationary points, convergence, and sample complexity", "Independent Policy Gradient for Large-Scale Markov Potential Games: Sharper Rates, Function Approximation, and Game-Agnostic Convergence", "Global Convergence of Multi-Agent Policy Gradient in Markov Potential Games"], "answer_arxiv_id": ["2106.00198", "2202.04129", "2106.01969"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_5910"} +{"question": "Which studies focus on lifting 2D images to 3D to leverage capabilities of 2D diffusion models?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "Text-to-3D using Gaussian Splatting", "DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1", "2309.16585", "2309.16653"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_5911"} +{"question": "Can you provide some examples of recent diffusion-based text-to-video models?", "answer": ["LAVIE: High-Quality Video Generation with Cascaded Latent Diffusion\n Models", "VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation", "ModelScope Text-to-Video Technical Report"], "answer_arxiv_id": ["2309.15103", "2303.08320", "2308.06571"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_5912"} +{"question": "What works use uncertainty measures for data selection to improve pseudo-labeling accuracy in SSDA?", "answer": ["Cross-Domain Adaptive Clustering for Semi-Supervised Domain Adaptation", "Semi-supervised Domain Adaptation via Sample-to-Sample Self-Distillation", "Multi-level Consistency Learning for Semi-supervised Domain Adaptation", "ECACL: A Holistic Framework for Semi-Supervised Domain Adaptation", "Deep Co-Training with Task Decomposition for Semi-Supervised Domain\n Adaptation"], "answer_arxiv_id": ["2104.09415", "2111.14353", "2205.04066", "2104.09136", "2007.12684"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_5913"} +{"question": "Could you provide work that endorses an implicit automatic differentiation mechanism under the Jax framework?", "answer": ["Efficient and Modular Implicit Differentiation"], "answer_arxiv_id": ["2105.15183"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_5914"} +{"question": "Which publications introduced the GAN inversion problem?", "answer": ["GAN Inversion: A Survey"], "answer_arxiv_id": ["2101.05278"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_5915"} +{"question": "Which works created the Adaptive Computation Time method for training RNNs on Natural Language Processing tasks?", "answer": ["Adaptive Computation Time for Recurrent Neural Networks"], "answer_arxiv_id": ["1603.08983"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_5916"} +{"question": "Are there any recent works that define diffusion models for molecule generation in 3D?", "answer": ["Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem", "Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2206.04119", "2203.17003"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_5917"} +{"question": "Which papers describe methods in self-supervised learning that use augmentation-based pretext tasks?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Self-Supervised Learning of Pretext-Invariant Representations", "Unsupervised Feature Learning via Non-Parametric Instance-level\n Discrimination", "Exploring Simple Siamese Representation Learning", "Bootstrap your own latent: A new approach to self-supervised Learning", "Self-labelling via simultaneous clustering and representation learning", "Unsupervised Learning of Visual Features by Contrasting Cluster\n Assignments", "Unsupervised Deep Embedding for Clustering Analysis", "Emerging Properties in Self-Supervised Vision Transformers", "DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2002.05709", "1911.05722", "1912.01991", "1805.01978", "2011.10566", "2006.07733", "1911.05371", "2006.09882", "1511.06335", "2104.14294", "2304.07193"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5918"} +{"question": "What studies focused on enabling flexible task output formats in Vision Generalist Models?", "answer": ["VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks"], "answer_arxiv_id": ["2305.11175"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5919"} +{"question": "Could you tell me about the work that designs a co-training scheme to extract semantic embeddings for knowledge transfer from head to tail classes?", "answer": ["Large scale long-tailed product recognition system at Alibaba"], "answer_arxiv_id": ["2102.04652"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_5920"} +{"question": "Any significant references where knowledge is transferred from intermediate features?", "answer": ["FitNets: Hints for Thin Deep Nets", "Masked Autoencoders Enable Efficient Knowledge Distillers"], "answer_arxiv_id": ["1412.6550", "2208.12256"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_5921"} +{"question": "Which papers discuss the challenges of information fusion in the context of retrieval-augmented generation?", "answer": ["Open Domain Multi-document Summarization: A Comprehensive Study of Model Brittleness under Retrieval", "Dense Passage Retrieval for Open-Domain Question Answering", "Challenges in Generalization in Open Domain Question Answering"], "answer_arxiv_id": ["2212.10526v3", "2004.04906", "2109.01156"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_5922"} +{"question": "Which papers contributed to the introduction of gradient guidance including classifier-free guidance to control the generative process?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Generating High Fidelity Data from Low-density Regions using Diffusion Models", "Blended Diffusion for Text-driven Editing of Natural Images", "More Control for Free! Image Synthesis with Semantic Diffusion Guidance", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2203.17260", "2111.14818", "2112.05744", "2112.10741", "2112.10752"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_5923"} +{"question": "What works developed a CoreSet approach for batch selection?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["1708.00489"], "source_meta": {"published_time": "20211227"}, "qid": "AutoScholarQuery_train_5924"} +{"question": "What paper discusses the use of SSLID for language detection?", "answer": ["MADLAD-400: A Multilingual And Document-Level Large Audited Dataset"], "answer_arxiv_id": ["2309.04662"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5925"} +{"question": "Which studies introduced a spatial deformation field, similar to the one used in the PRJ algorithm?", "answer": ["HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields"], "answer_arxiv_id": ["2106.13228v2"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_5926"} +{"question": "Which works track pedestrians and cars in KITTI?", "answer": ["HOTA: A Higher Order Metric for Evaluating Multi-Object Tracking"], "answer_arxiv_id": ["2009.07736"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_5927"} +{"question": "Which study used a sparse voxel grid to interpolate a continuous density field?", "answer": ["Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2112.05131"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_5928"} +{"question": "Which paper did take a linear algebra approach to estimate the dimension of a subspace of adversarial examples?", "answer": ["The Space of Transferable Adversarial Examples"], "answer_arxiv_id": ["1704.03453"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_5929"} +{"question": "What work introduced the Meta-World benchmark in the context of multi-task and meta reinforcement learning?", "answer": ["Meta-World: A Benchmark and Evaluation for Multi-Task and Meta\n Reinforcement Learning"], "answer_arxiv_id": ["1910.10897"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_5930"} +{"question": "What work introduced Diffusion Probabilistic Models (DPMs)?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5931"} +{"question": "Who emphasizes the ability to explore and apply insights from analogous solutions in their research?", "answer": ["Thought Propagation: An Analogical Approach to Complex Reasoning with\n Large Language Models"], "answer_arxiv_id": ["2310.03965"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_5932"} +{"question": "Could you provide me some works that combined semantic or instance segmentation models as object cues to refine optical flow?", "answer": ["Optical Flow with Semantic Segmentation and Localized Layers", "Optical Flow in Mostly Rigid Scenes", "Exploiting Semantic Information and Deep Matching for Optical Flow", "SemARFlow: Injecting Semantics into Unsupervised Optical Flow Estimation\n for Autonomous Driving"], "answer_arxiv_id": ["1603.03911", "1705.01352", "1604.01827", "2303.06209"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_5933"} +{"question": "Which work evaluates semi-supervised learning on datasets that exhibit class imbalance?", "answer": ["A Realistic Evaluation of Semi-Supervised Learning for Fine-Grained\n Classification"], "answer_arxiv_id": ["2104.00679"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_5934"} +{"question": "Which works looked into robustness of neural networks with respect to adversarial examples?", "answer": ["Understanding Zero-Shot Adversarial Robustness for Large-Scale Models", "On the Robustness of Vision Transformers to Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "On Evaluating Adversarial Robustness of Large Vision-Language Models", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["2212.07016", "2104.02610", "1706.06083", "2305.16934", "1901.08573"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_5935"} +{"question": "What studies suggest using point-based shape editing with NeRF?", "answer": ["NeuralEditor: Editing Neural Radiance Fields via Manipulating Point\n Clouds", "Point-NeRF: Point-based Neural Radiance Fields", "Dynamic Point Fields"], "answer_arxiv_id": ["2305.03049", "2201.08845", "2304.02626"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_5936"} +{"question": "Can you name the paper that builds upon the viewpoint transformation in Lift-Splat-Shoot (LSS) and detects 3D objects in BEV features?", "answer": ["BEVDet: High-performance Multi-camera 3D Object Detection in\n Bird-Eye-View"], "answer_arxiv_id": ["2112.11790"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_5937"} +{"question": "Which studies have explored the concept of multi-task segmentation on LiDAR point clouds?", "answer": ["MaskRange: A Mask-classification Model for Range-view based LiDAR\n Segmentation", "LidarMultiNet: Towards a Unified Multi-Task Network for LiDAR Perception", "Position-Guided Point Cloud Panoptic Segmentation Transformer"], "answer_arxiv_id": ["2206.12073", "2209.09385", "2303.13509"], "source_meta": {"published_time": "20240502"}, "qid": "AutoScholarQuery_train_5938"} +{"question": "What works are related to the study of asynchronous proportional response bidding dynamics?", "answer": ["Amortized Analysis of Asynchronous Price Dynamics"], "answer_arxiv_id": ["1806.10952"], "source_meta": {"published_time": "20230709"}, "qid": "AutoScholarQuery_train_5939"} +{"question": "Which studies leveraged hand-crafted dense optical flow to embed motion information in action recognition?", "answer": ["Temporal Segment Networks: Towards Good Practices for Deep Action Recognition", "Learning Spatio-Temporal Representation with Pseudo-3D Residual Networks", "ViViT: A Video Vision Transformer", "G3AN: Disentangling Appearance and Motion for Video Generation", "Conditional Image-to-Video Generation with Latent Flow Diffusion Models", "Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset"], "answer_arxiv_id": ["1608.00859", "1711.10305", "2103.15691", "1912.05523", "2303.13744", "1705.07750"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_5940"} +{"question": "What work showcased improved performance on image classification and object detection by conditioning the position encoding on localized patch token?", "answer": ["Conditional Positional Encodings for Vision Transformers"], "answer_arxiv_id": ["2102.10882"], "source_meta": {"published_time": "20220620"}, "qid": "AutoScholarQuery_train_5941"} +{"question": "Which study first introduced the concept of Dataset Distillation (DD)?", "answer": ["Dataset Distillation"], "answer_arxiv_id": ["1811.10959"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_5942"} +{"question": "What recent efforts focus on learning a robust segmentation model using existing pseudo-labels in WSSS?", "answer": ["Uncertainty Estimation via Response Scaling for Pseudo-mask Noise\n Mitigation in Weakly-supervised Semantic Segmentation", "Adaptive Early-Learning Correction for Segmentation from Noisy\n Annotations", "Out-of-Candidate Rectification for Weakly Supervised Semantic\n Segmentation"], "answer_arxiv_id": ["2112.07431", "2110.03740", "2211.12268"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_5943"} +{"question": "Are there any studies that learn conditional diffusion models tailored for specific image-to-image translation tasks?", "answer": ["Palette: Image-to-Image Diffusion Models", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2111.05826", "2211.09800"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_5944"} +{"question": "What works apply contrastive methods in graph self-supervised learning?", "answer": ["Variational Graph Auto-Encoders", "Inductive Representation Learning on Large Graphs", "Representation Learning on Graphs: Methods and Applications"], "answer_arxiv_id": ["1611.07308", "1706.02216", "1709.05584"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_5945"} +{"question": "Which papers investigated lifting 2D visual features to 3D?", "answer": ["Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D Image Representations", "Decomposing NeRF for Editing via Feature Field Distillation", "LERF: Language Embedded Radiance Fields", "Interactive Segmentation of Radiance Fields"], "answer_arxiv_id": ["2209.03494", "2205.15585", "2303.09553", "2212.13545"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_5946"} +{"question": "What studies use self-supervised learning in speaker recognition to utilize unlabeled data?", "answer": ["wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "an iterative framework for self-supervised deep speaker representation learning", "Self-Supervised Training of Speaker Encoder with Multi-Modal Diverse Positive Pairs", "Self-supervised Speaker Recognition with Loss-gated Learning", "VoxSRC 2021: The Third VoxCeleb Speaker Recognition Challenge"], "answer_arxiv_id": ["2006.11477", "1810.04805", "2006.09882", "2006.07733", "2002.05709", "2010.14751", "2210.15385", "2110.03869", "2201.04583"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_5947"} +{"question": "Can you provide a reference on applying overfit models in the context of level-of-details modeling?", "answer": ["Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes"], "answer_arxiv_id": ["2101.10994"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_5948"} +{"question": "Can you provide references to research that uses reinforcement learning to improve zero-shot transfer and caption style transfer?", "answer": ["Multimodal Knowledge Alignment with Reinforcement Learning"], "answer_arxiv_id": ["2205.12630"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_5949"} +{"question": "What research papers showed advancements in Large Language Models (LLMs) with the increase in model and corpus sizes?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_5950"} +{"question": "Which papers are about deep learning based ISR and focus on model designs?", "answer": ["Enhanced Deep Residual Networks for Single Image Super-Resolution", "Image Super-Resolution Using Very Deep Residual Channel Attention\n Networks", "Residual Dense Network for Image Super-Resolution", "Pre-Trained Image Processing Transformer", "SwinIR: Image Restoration Using Swin Transformer", "Efficient Long-Range Attention Network for Image Super-resolution", "Activating More Pixels in Image Super-Resolution Transformer", "Dual Aggregation Transformer for Image Super-Resolution"], "answer_arxiv_id": ["1707.02921", "1807.02758", "1802.08797", "2012.00364", "2108.10257", "2203.06697", "2205.04437", "2308.03364"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_5951"} +{"question": "What research articles are about enhancing large pre-trained text-to-image diffusion models with task-specific image conditions?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_5952"} +{"question": "Are there any works that look at generalization on symbolic mathematical integration?", "answer": ["Symbolic Brittleness in Sequence Models: on Systematic Generalization in Symbolic Mathematics"], "answer_arxiv_id": ["2109.13986"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_5953"} +{"question": "What works propose the method of increasing the diversity in inputs in black-box attacks?", "answer": ["Improving Transferability of Adversarial Examples with Input Diversity", "Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks", "Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks", "Admix: Enhancing the Transferability of Adversarial Attacks"], "answer_arxiv_id": ["1803.06978", "1904.02884", "1908.06281", "2102.00436"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_5954"} +{"question": "Which paper connects the numerical invertibility of deep residual networks (ResNets) and ODE stability analysis?", "answer": ["Reversible Architectures for Arbitrarily Deep Residual Neural Networks"], "answer_arxiv_id": ["1709.03698"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_5955"} +{"question": "What works have used prior experience without reward annotations for pre-training state representations?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Masked Visual Pre-training for Motor Control", "Real-World Robot Learning with Masked Visual Pre-training", "Representation Matters: Offline Pretraining for Sequential Decision Making", "VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training", "Proto-Value Networks: Scaling Representation Learning with Auxiliary Tasks", "Reinforcement Learning from Passive Data via Latent Intentions"], "answer_arxiv_id": ["2004.04136", "2203.06173", "2210.03109", "2102.05815", "2210.00030", "2304.12567", "2304.04782"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_5956"} +{"question": "What works have contributed to the dramatic reduction of neural rendering train and test times?", "answer": ["MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient\n Neural Field Rendering on Mobile Architectures", "3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2208.00277", "2308.04079"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_5957"} +{"question": "What is the paper that proposed finite-step inference dynamics guided by an energy-based model?", "answer": ["Learning Non-Convergent Non-Persistent Short-Run MCMC Toward Energy-Based Model"], "answer_arxiv_id": ["1904.09770"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_5958"} +{"question": "Which research works use training-based methods for knowledge distillation?", "answer": ["Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed", "Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2101.02388", "2202.00512"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_5959"} +{"question": "In which works did authors consider a multi-agent system with honest and malicious agents in the context of K-armed Multi-Armed Bandits?", "answer": ["Robust Multi-Agent Multi-Armed Bandits", "Robust Multi-Agent Bandits Over Undirected Graphs"], "answer_arxiv_id": ["2007.03812", "2203.00076"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_5960"} +{"question": "Which work applies filters based on n-gram language models?", "answer": ["CCNet: Extracting High Quality Monolingual Datasets from Web Crawl Data"], "answer_arxiv_id": ["1911.00359"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5961"} +{"question": "Could you provide research that tried to defend against adversarial attacks in the input space?", "answer": ["Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models", "Diffusion Models for Adversarial Purification"], "answer_arxiv_id": ["1805.06605", "2205.07460"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_5962"} +{"question": "Which work first introduced the Neural Radiance Fields (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_5963"} +{"question": "Which papers examine architecture and task specific techniques for efficient inference?", "answer": ["MatFormer: Nested Transformer for Elastic Inference"], "answer_arxiv_id": ["2310.07707"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_5964"} +{"question": "What research work classify training into different regimes using training loss?", "answer": ["The Two Regimes of Deep Network Training"], "answer_arxiv_id": ["2002.10376"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_5965"} +{"question": "Which research papers adopted large-scale vision-language instruction tuning data to align LVLMs with human preferences?", "answer": ["PandaGPT: One Model To Instruction-Follow Them All", "InstructionGPT-4: A 200-Instruction Paradigm for Fine-Tuning MiniGPT-4", "MultiModal-GPT: A Vision and Language Model for Dialogue with Humans", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2305.16355", "2308.12067", "2305.04790", "2304.15010", "2305.03726"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_5966"} +{"question": "Can you find any studies dealing with fully or partially recovering clusters in the presence of noise within the 'global algorithm regimes'?", "answer": ["Robust and computationally feasible community detection in the presence of arbitrary outlier nodes", "Community detection in sparse networks via Grothendieck’s inequality", "How Robust are Reconstruction Thresholds for Community Detection?", "Learning Communities in the Presence of Errors", "Clustering Partially Observed Graphs via Convex Optimization", "Tight Error Bounds for Structured Prediction", "Correlation Clustering with Noisy Partial Information"], "answer_arxiv_id": ["1404.6000", "1411.4686", "1511.01473", "1511.03229", "1104.4803", "1409.5834", "1406.5667"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_5967"} +{"question": "Could you provide me with some works about simulation datasets related to typhoon and global atmosphere?", "answer": ["ENS-10: A Dataset For Post-Processing Ensemble Weather Forecasts", "ClimART: A Benchmark Dataset for Emulating Atmospheric Radiative Transfer in Weather and Climate Models", "ExtremeWeather: A large-scale climate dataset for semi-supervised detection, localization, and understanding of extreme weather events"], "answer_arxiv_id": ["2206.14786", "2111.14671", "1612.02095"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_5968"} +{"question": "What works train a transformer-based image encoder with a contrastive objective for semantic segmentation?", "answer": ["Language-driven Semantic Segmentation"], "answer_arxiv_id": ["2201.03546"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_5969"} +{"question": "Could you indicate the datasets used for capturing hand-object interactions?", "answer": ["Learning joint reconstruction of hands and manipulated objects", "DexYCB: A Benchmark for Capturing Hand Grasping of Objects", "H2O: A Benchmark for Visual Human-human Object Handover Analysis", "ARCTIC: A Dataset for Dexterous Bimanual Hand-Object Manipulation", "HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object\n Interaction", "ContactPose: A Dataset of Grasps with Object Contact and Hand Pose"], "answer_arxiv_id": ["1904.05767", "2104.04631", "2104.11466", "2204.13662", "2203.01577", "2007.09545"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_5970"} +{"question": "Which papers present methods of generating answer sentences in the scope of Answer Sentence Selection?", "answer": ["Answer Generation for Retrieval-based Question Answering Systems"], "answer_arxiv_id": ["2106.00955"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_5971"} +{"question": "Which papers introduced the pretrained language models such as BERT and BART?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension"], "answer_arxiv_id": ["1810.04805", "1910.13461"], "source_meta": {"published_time": "20220801"}, "qid": "AutoScholarQuery_train_5972"} +{"question": "What work is about a data augmented MADDPG which reduces the number of environmental interactions?", "answer": ["Experience Augmentation: Boosting and Accelerating Off-Policy Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["2005.09453"], "source_meta": {"published_time": "20220310"}, "qid": "AutoScholarQuery_train_5973"} +{"question": "What work introduced the suite of RL environments AI-Safety-Gridworlds for ensuring safety properties?", "answer": ["AI Safety Gridworlds"], "answer_arxiv_id": ["1711.09883"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_5974"} +{"question": "Which work uses inter-sentence BLEU to access uncertainty in machine translation?", "answer": ["Analyzing Uncertainty in Neural Machine Translation"], "answer_arxiv_id": ["1803.00047"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_5975"} +{"question": "Which works belong to the first category of diffusion-based image editing that involves mixing the latent variables of DPM and the input image?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations"], "answer_arxiv_id": ["2108.02938", "2201.09865", "2108.01073"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_5976"} +{"question": "Which works introduced regularizations into optimization objective for training a fair model on biased data?", "answer": ["Empirical Risk Minimization under Fairness Constraints", "Distributionally Robust Neural Networks for Group Shifts: On the\n Importance of Regularization for Worst-Case Generalization", "Large-Scale Methods for Distributionally Robust Optimization", "Invariant Feature Regularization for Fair Face Recognition", "Enhancing Fairness of Visual Attribute Predictors", "Consistent Instance False Positive Improves Fairness in Face Recognition"], "answer_arxiv_id": ["1802.08626", "1911.08731", "2010.05893", "2310.14652", "2207.05727", "2106.05519"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5977"} +{"question": "Which work focused on mitigating output image flickering by simultaneously generating multi-view images?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2308.16512"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_5978"} +{"question": "Are there any works that tackles image deblurring in unsupervised learning with unpaired data?", "answer": ["Unsupervised Domain-Specific Deblurring via Disentangled Representations", "FCL-GAN: A Lightweight and Real-Time Baseline for Unsupervised Blind\n Image Deblurring", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "DualGAN: Unsupervised Dual Learning for Image-to-Image Translation"], "answer_arxiv_id": ["1903.01594", "2204.07820", "1703.10593", "1704.02510"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_5979"} +{"question": "Could you provide me some works that use learning-based approaches in 3D scene and object reconstruction?", "answer": ["DeepVoxels: Learning Persistent 3D Feature Embeddings", "DISN: Deep Implicit Surface Network for High-quality Single-view 3D Reconstruction", "From Points to Multi-Object 3D Reconstruction"], "answer_arxiv_id": ["1812.01024", "1905.10711", "2012.11575"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_5980"} +{"question": "Which works extend triplets to ranking more than two items with respect to a reference item by using tuple queries?", "answer": ["Active Ordinal Querying for Tuplewise Similarity Learning"], "answer_arxiv_id": ["1910.04115"], "source_meta": {"published_time": "20230908"}, "qid": "AutoScholarQuery_train_5981"} +{"question": "Can you name some studies related to federated minimax optimization?", "answer": ["Agnostic Federated Learning", "Distributionally Robust Federated Averaging", "Federated Minimax Optimization: Improved Convergence Analyses and Algorithms", "FedNest: Federated Bilevel, Minimax, and Compositional Optimization"], "answer_arxiv_id": ["1902.00146", "2102.12660", "2203.04850", "2205.02215v3"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_5982"} +{"question": "What studies have made progress on solving an OICA problem particularly with linear SCMs and non-Gaussian exogenous noises?", "answer": ["Learning Linear Non-Gaussian Causal Models in the Presence of Latent Variables"], "answer_arxiv_id": ["1908.03932"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_5983"} +{"question": "Which paper separates weight decay from the training objective for better generalization?", "answer": ["Decoupled Weight Decay Regularization"], "answer_arxiv_id": ["1711.05101"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_5984"} +{"question": "What studies focus on addressing fairness issues arising from different structural information in GNNs?", "answer": ["Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training Data", "Subgroup Generalization and Fairness of Graph Neural Networks", "EDITS: Modeling and Mitigating Data Bias for Graph Neural Networks"], "answer_arxiv_id": ["2108.01099", "2106.15535", "2108.05233v2"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_5985"} +{"question": "Can you provide the study that extracted a latent variable by a variational autoencoder in the Stable Diffusion model?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_5986"} +{"question": "What are the studies regarding the multiplane image (MPI) representation in the context of novel view synthesis?", "answer": ["Stereo Magnification: Learning View Synthesis using Multiplane Images", "Pushing the Boundaries of View Extrapolation with Multiplane Images", "DeepView: View Synthesis with Learned Gradient Descent", "Local Light Field Fusion: Practical View Synthesis with Prescriptive\n Sampling Guidelines", "NeX: Real-time View Synthesis with Neural Basis Expansion", "Single-View View Synthesis with Multiplane Images", "MINE: Towards Continuous Depth MPI with NeRF for Novel View Synthesis", "Single-View View Synthesis in the Wild with Learned Adaptive Multiplane\n Images"], "answer_arxiv_id": ["1805.09817", "1905.00413", "1906.07316", "1905.00889", "2103.05606", "2004.11364", "2103.14910", "2205.11733"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_5987"} +{"question": "What study explored unrolled computation graphs (UCGs)?", "answer": ["Unbiased Gradient Estimation in Unrolled Computation Graphs with Persistent Evolution Strategies"], "answer_arxiv_id": ["2112.13835"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_5988"} +{"question": "What research has shown significant results in video generation by training a motion module with such temporal layers?", "answer": ["AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning"], "answer_arxiv_id": ["2307.04725"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_5989"} +{"question": "Which studies about unsupervised instance segmentation utilize self-supervised ViT features for segment discovery?", "answer": ["Localizing Objects with Self-Supervised Transformers and no Labels", "TokenCut: Segmenting Objects in Images and Videos with Self-supervised\n Transformer and Normalized Cut"], "answer_arxiv_id": ["2109.14279", "2209.00383"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_5990"} +{"question": "What works utilize Transformer model for natural language modeling?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1810.04805", "2005.14165"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_5991"} +{"question": "Could you provide me with some works that generate data by simulating conversations between model agents?", "answer": ["Enhancing Chat Language Models by Scaling High-quality Instructional Conversations", "Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on\n Self-Chat Data"], "answer_arxiv_id": ["2305.14233v1", "2304.01196"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_5992"} +{"question": "Which papers proposed methods to learn hierarchical feature embeddings for several-minute-long videos?", "answer": ["HierVL: Learning Hierarchical Video-Language Embeddings", "Cross-Modal and Hierarchical Modeling of Video and Text", "HERO: Hierarchical Encoder for Video+Language Omni-representation\n Pre-training"], "answer_arxiv_id": ["2301.02311", "1810.07212", "2005.00200"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_5993"} +{"question": "What works have shown that gradient-based methods can train neural networks of various architectures to achieve small generalization error, provided the networks are sufficiently overparameterized?", "answer": ["Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "Learning and Generalization in Overparameterized Neural Networks, Going Beyond Two Layers", "Benign Overfitting in Linear Regression", "Benign Overfitting in Two-layer Convolutional Neural Networks", "Just Interpolate: Kernel “Ridgeless” Regression Can Generalize", "On the optimization and generalization of overparameterized implicit neural networks"], "answer_arxiv_id": ["1901.08584", "1811.04918", "1906.11300", "2202.06526", "1808.00387", "2209.15562"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_5994"} +{"question": "Can you list down the works that have used sparse MoE for multi-modal and multi-task learning?", "answer": ["Multimodal Contrastive Learning with LIMoE: the Language-Image Mixture of Experts", "Uni-Perceiver-MoE: Learning Sparse Generalist Models with Conditional MoEs", "M3ViT: Mixture-of-Experts Vision Transformer for Efficient Multi-task Learning with Model-Accelerator Co-design"], "answer_arxiv_id": ["2206.02770", "2206.04674", "2210.14793"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_5995"} +{"question": "Which studies combined programs with neural networks for visual reasoning domains?", "answer": ["Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding", "The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences from Natural Supervision"], "answer_arxiv_id": ["1810.02338", "1904.12584"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_5996"} +{"question": "What papers discuss enhancing rendering quality and improving training speeds in NeRF?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Point-NeRF: Point-based Neural Radiance Fields"], "answer_arxiv_id": ["2112.05131", "2201.05989", "2111.11215", "2111.12077", "2201.08845"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_5997"} +{"question": "What works proposed defense strategies based on simple fine-tuning against backdoor attacks?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks", "Revisiting Personalized Federated Learning: Robustness Against Backdoor Attacks", "Enhancing Fine-Tuning Based Backdoor Defense with Sharpness-Aware Minimization"], "answer_arxiv_id": ["1805.12185", "2302.01677", "2304.11823"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_5998"} +{"question": "What works are focused on performing regret analysis in game theory?", "answer": ["Online and Bandit Algorithms for Nonstationary Stochastic Saddle-Point Optimization", "No-Regret Learning in Time-Varying Zero-Sum Games"], "answer_arxiv_id": ["1912.01698", "2201.12736"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_5999"} +{"question": "What studies have focused on preserving features and corresponding decoders in recent advances of dataset distillation?", "answer": ["Synthesizing Informative Training Samples with GAN", "Remember the Past: Distilling Datasets into Addressable Memories for\n Neural Networks", "Dataset Condensation with Latent Space Knowledge Factorization and\n Sharing"], "answer_arxiv_id": ["2204.07513", "2206.02916", "2208.10494"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_6000"} +{"question": "What papers represent the major trend of focusing on the rounding operator itself in data-free quantization research?", "answer": ["SQuant: On-the-Fly Data-Free Quantization via Diagonal Hessian Approximation", "SPIQ: Data-Free Per-Channel Static Input Quantization"], "answer_arxiv_id": ["2202.07471", "2203.14642"], "source_meta": {"published_time": "20220328"}, "qid": "AutoScholarQuery_train_6001"} +{"question": "Which works discussed the sim-to-real problem in the context of robot navigation?", "answer": ["Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Sim2Real Predictivity: Does Evaluation in Simulation Predict Real-World Performance?", "On Embodied Visual Navigation in Real Environments Through Habitat", "Out of the Box: Embodied Navigation in the Real World", "Sim-to-Real Transfer for Vision-and-Language Navigation"], "answer_arxiv_id": ["1703.06907", "1912.06321", "2010.13439v1", "2105.05873", "2011.03807"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_6002"} +{"question": "Who studied the use of English prompts with non-English examples in multilingual settings?", "answer": ["Discrete and Soft Prompting for Multilingual Models", "Language Models are Few-shot Multilingual Learners", "Few-shot Learning with Multilingual Language Models"], "answer_arxiv_id": ["2109.03630", "2109.07684", "2112.10668"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_6003"} +{"question": "What works related over-squashing with the existence of edges with high negative curvature?", "answer": ["Understanding over-squashing and bottlenecks on graphs via curvature"], "answer_arxiv_id": ["2111.14522"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6004"} +{"question": "Could you provide me some works about reinforcement-learning and sequential-decision-making techniques that interpret tokens as actions and utility as a reward?", "answer": ["Deeply AggreVaTeD: Differentiable Imitation Learning for Sequential Prediction", "A Stable and Effective Learning Strategy for Trainable Greedy Decoding", "Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models", "Deep Reinforcement Learning for Sequence-to-Sequence Models", "Machine Translation Decoding beyond Beam Search"], "answer_arxiv_id": ["1703.01030", "1804.07915", "1712.01818", "1805.09461", "2104.05336"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_6005"} +{"question": "Which works provide partitioned datasets used commonly in Federated Learning benchmarks?", "answer": ["LEAF: A Benchmark for Federated Settings", "Flower: A Friendly Federated Learning Research Framework", "FedML: A Research Library and Benchmark for Federated Machine Learning", "FedNLP: Benchmarking Federated Learning Methods for Natural Language Processing Tasks", "FedScale: Benchmarking Model and System Performance of Federated Learning at Scale", "FLBench: A Benchmark Suite for Federated Learning", "The OARF Benchmark Suite: Characterization and Implications for Federated Learning Systems", "FedJAX: Federated learning simulation with JAX"], "answer_arxiv_id": ["1812.01097", "2007.14390v5", "2007.13518", "2104.08815", "2105.11367", "2008.07257", "2006.07856", "2108.02117"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_6006"} +{"question": "Which research papers are about semantic segmentation in autonomous driving?", "answer": ["InverseForm: A Loss Function for Structured Boundary-Aware Segmentation", "Bonnet: An Open-Source Training and Deployment Framework for Semantic\n Segmentation in Robotics using CNNs", "InternImage: Exploring Large-Scale Vision Foundation Models with\n Deformable Convolutions"], "answer_arxiv_id": ["2104.02745", "1802.08960", "2211.05778"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_6007"} +{"question": "Which study showed disentangled representations enables quicker learning for abstract reasoning tasks?", "answer": ["Are Disentangled Representations Helpful for Abstract Visual Reasoning?"], "answer_arxiv_id": ["1905.12506"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_6008"} +{"question": "Which papers discussed the challenges of visual entity recognition tasks, particularly imbalanced training classes and noisy training labels?", "answer": ["Large-Scale Long-Tailed Recognition in an Open World", "WebVision Database: Visual Learning and Understanding from Web Data"], "answer_arxiv_id": ["1904.05160", "1708.02862"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_6009"} +{"question": "Where has been regularization, the main idea behind BPPO, often used?", "answer": ["Unsupervised Domain Adaptation with Dynamics- Aware Rewards in Reinforcement Learning", "Learn Goal-Conditioned Policy with Intrinsic Motivation for Deep Reinforcement Learning", "Wasserstein Adversarial Imitation Learning", "Off-Dynamics Inverse Reinforcement Learning from Hetero-Domain"], "answer_arxiv_id": ["2110.12997", "2104.05043", "1906.08113", "2110.11443"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_6010"} +{"question": "Which papers focused on the Bayesian-based paradigm of uncertainty estimation approaches?", "answer": ["Variational Dropout and the Local Reparameterization Trick", "Variational Dropout Sparsifies Deep Neural Networks", "A Probabilistic U-Net for Segmentation of Ambiguous Images"], "answer_arxiv_id": ["1506.02557", "1701.05369", "1806.05034"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_6011"} +{"question": "What works extended label smoothing to sequence-to-sequence learning?", "answer": ["Semantic Label Smoothing for Sequence to Sequence Problems"], "answer_arxiv_id": ["2010.07447"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_6012"} +{"question": "Could you give me studies that proposed representation learning with the successor measure?", "answer": ["Learning One Representation to Optimize All Rewards"], "answer_arxiv_id": ["2103.07945"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_6013"} +{"question": "What papers discuss next-generation image-text foundation models catalysed by CLIP?", "answer": ["Zero-Shot Text-to-Image Generation", "Flamingo: a Visual Language Model for Few-Shot Learning", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2204.14198", "2112.10752"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_6014"} +{"question": "What are some works in the area of improving reasoning capabilities through requests for explicit reasoning steps?", "answer": ["Show Your Work: Scratchpads for Intermediate Computation with Language Models", "PAL: Program-aided Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2112.00114", "2211.10435", "2201.11903", "2203.11171", "2211.12588", "2305.10601"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6015"} +{"question": "Can you name the works that utilized saliency-based region proposal algorithm in the unsupervised object detection/discovery?", "answer": ["Toward Unsupervised, Multi-Object Discovery in Large-Scale Image Collections", "Localizing Objects with Self-Supervised Transformers and no Labels", "Unsupervised Object Localization: Observing the Background to Discover Objects", "FreeSOLO: Learning to Segment Objects without Annotations"], "answer_arxiv_id": ["2007.02662", "2109.14279", "2212.07834", "2202.12181"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_6016"} +{"question": "What publications have dealt with linear mixture MDPs where the transition probability kernel is a linear mixture of a number of basis kernels?", "answer": ["Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Provably Efficient Reinforcement Learning for Discounted MDPs with Feature Mapping", "Provably Efficient Exploration in Policy Optimization"], "answer_arxiv_id": ["1910.10597", "2006.01107", "2006.01107", "2006.13165", "1912.05830"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_6017"} +{"question": "Could you provide me some studies about tasks involving arithmetic and math that have been developed?", "answer": ["Training Verifiers to Solve Math Word Problems"], "answer_arxiv_id": ["2110.14168"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_6018"} +{"question": "What are some examples of previous studies that applied homotopy continuation method?", "answer": ["An adaptive homotopy method for computing bifurcations of nonlinear parametric systems", "Homotopy-based training of NeuralODEs for accurate dynamics discovery", "Learning to Solve Hard Minimal Problems"], "answer_arxiv_id": ["2002.03460", "2210.01407", "2112.03424v1"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_6019"} +{"question": "Which work proposed a solution to the Fokker-Planck equation based on the Wasserstein gradient flows?", "answer": ["Large-Scale Wasserstein Gradient Flows"], "answer_arxiv_id": ["2106.00736"], "source_meta": {"published_time": "20220214"}, "qid": "AutoScholarQuery_train_6020"} +{"question": "Which works discuss the sensitivity of language models to prompts?", "answer": ["Prompt-based Conservation Learning for Multi-hop Question Answering", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models as Optimizers"], "answer_arxiv_id": ["2209.06923", "2201.11903", "2309.03409"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_6021"} +{"question": "Which works proposed alternative conditions to enforce the conservation law more efficiently in GFlowNets?", "answer": ["Trajectory balance: Improved credit assignment in GFlowNets", "Learning GFlowNets From Partial Episodes For Improved Convergence And Stability", "Better Training of GFlowNets with Local Credit and Incomplete Trajectories"], "answer_arxiv_id": ["2201.13259", "2209.12782", "2302.01687"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6022"} +{"question": "Which papers investigate the effect of negative samples in unsupervised contrastive learning?", "answer": ["Understanding Negative Samples in Instance Discriminative Self-supervised Representation Learning", "Investigating the Role of Negatives in Contrastive Representation Learning", "On the Surrogate Gap between Contrastive and Supervised Losses"], "answer_arxiv_id": ["2102.06866", "2106.09943", "2110.02501"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6023"} +{"question": "What papers developed Diffusion Models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6024"} +{"question": "Which papers propose strategies for early detection and halting of processing for less relevant passages?", "answer": ["Don't Read Too Much into It: Adaptive Computation for Open-Domain\n Question Answering", "Training Adaptive Computation for Open-Domain Question Answering with\n Computational Constraints", "KG-FiD: Infusing Knowledge Graph in Fusion-in-Decoder for Open-Domain\n Question Answering"], "answer_arxiv_id": ["2011.05435", "2107.02102", "2110.04330"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_6025"} +{"question": "Any works that apply the IRM framework to large neural networks?", "answer": ["In Search of Lost Domain Generalization"], "answer_arxiv_id": ["2007.01434"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_6026"} +{"question": "Any works discussing the advantages and limitation of PCA and ICA in modelling some data manifold topologies?", "answer": ["Sparse PCA from Sparse Linear Regression"], "answer_arxiv_id": ["1811.10106"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_6027"} +{"question": "What works showcase the capabilities of LLMs in performing multi-turn interactions with human users?", "answer": ["Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena"], "answer_arxiv_id": ["2306.05685"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_6028"} +{"question": "Which work proposed Centered Kernel Alignment (CKA)?", "answer": ["Similarity of Neural Network Representations Revisited"], "answer_arxiv_id": ["1905.00414"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_6029"} +{"question": "What papers address the issue of performance dip and plateauing observed in offline-to-online RL algorithms?", "answer": ["AWAC: Accelerating Online Reinforcement Learning with Offline Datasets"], "answer_arxiv_id": ["2006.09359"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_6030"} +{"question": "Could you mention the research that employs techniques like word2vec or GloVe to learn vector representations of labels?", "answer": ["Efficient Estimation of Word Representations in Vector Space"], "answer_arxiv_id": ["1301.3781"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_6031"} +{"question": "Which works employed in-context learning for enhancing LLMs in NLP?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Language Models are Few-Shot Learners", "Language Models are General-Purpose Interfaces"], "answer_arxiv_id": ["2204.14198", "2005.14165", "2206.06336"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6032"} +{"question": "Can you provide references that proposed to rectify hallucinations in model outputs?", "answer": ["Analyzing and Mitigating Object Hallucination in Large Vision-Language\n Models", "Woodpecker: Hallucination Correction for Multimodal Large Language\n Models", "Volcano: Mitigating Multimodal Hallucination through Self-Feedback\n Guided Revision"], "answer_arxiv_id": ["2310.00754", "2310.16045", "2311.07362"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_6033"} +{"question": "Which works discuss the role of attention in solving the SVRT tasks?", "answer": ["Understanding the computational demands underlying visual reasoning", "Recurrent Vision Transformer for Solving Visual Reasoning Problems"], "answer_arxiv_id": ["2108.03603", "2111.14576"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_6034"} +{"question": "What works extended the idea with rotation-based iterative Gaussianization?", "answer": ["Iterative Gaussianization: from ICA to Random Rotations"], "answer_arxiv_id": ["1602.00229"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_6035"} +{"question": "Which work is referenced for token generation for using green tokens frequently?", "answer": ["A Watermark for Large Language Models"], "answer_arxiv_id": ["2301.10226"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_6036"} +{"question": "Which paper claims the sample complexity of the MLE in noiseless case to be Ω~​(K2/ε)?", "answer": ["On Statistical Learning of Simplices: Unmixing Problem Revisited"], "answer_arxiv_id": ["1810.07845"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_6037"} +{"question": "What work first studied the kernelized bandit problem and introduced the GP-UCB algorithm?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design"], "answer_arxiv_id": ["0912.3995v4"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_6038"} +{"question": "What works show an interest in using prompts to extract knowledge from large language models?", "answer": ["KnowPrompt: Knowledge-aware Prompt-tuning with Synergistic Optimization for Relation Extraction", "How Many Data Points is a Prompt Worth?", "On Transferability of Prompt Tuning for Natural Language Processing", "Ontology-enhanced Prompt-tuning for Few-shot Learning", "Conditional Prompt Learning for Vision-Language Models", "Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2104.07650", "2103.08493", "2111.06719v2", "2201.11332", "2203.05557", "2205.11916"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_6039"} +{"question": "What research proposed the stochastic Polyak step size (SPS)?", "answer": ["Stochastic Polyak Step-size for SGD: An Adaptive Learning Rate for Fast Convergence"], "answer_arxiv_id": ["2002.10542"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6040"} +{"question": "What studies dealt with the visual acoustic matching problem?", "answer": ["Visual Acoustic Matching", "Image2Reverb: Cross-Modal Reverb Impulse Response Synthesis", "Self-Supervised Visual Acoustic Matching"], "answer_arxiv_id": ["2202.06875", "2103.14201", "2307.15064"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_6041"} +{"question": "Which paper proposed paged attention to reduce memory fragmentation of the KV cache for efficient inference of LLMs?", "answer": ["Efficient Memory Management for Large Language Model Serving with\n PagedAttention"], "answer_arxiv_id": ["2309.06180"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_6042"} +{"question": "What are some examples of works that have used RL techniques to improve NLP applications?", "answer": ["Sequence Level Training with Recurrent Neural Networks", "Text Generation by Learning from Demonstrations", "Improving Dialog Systems for Negotiation with Personality Modeling", "Quark: Controllable Text Generation with Reinforced [Un]learning", "Offline RL for Natural Language Generation with Implicit Language Q Learning", "Is Reinforcement Learning (Not) for Natural Language Processing: Benchmarks, Baselines, and Building Blocks for Natural Language Policy Optimization"], "answer_arxiv_id": ["1511.06732", "2009.07839", "2010.09954", "2205.13636", "2206.11871", "2210.01241"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_6043"} +{"question": "What works represent examples of state-of-the-art Knowledge Graph Embeddings (KGEs)?", "answer": ["Understanding the Performance of Knowledge Graph Embeddings in Drug Discovery", "Relation Prediction as an Auxiliary Training Objective for Improving Multi-Relational Graph Representations"], "answer_arxiv_id": ["2105.10488", "2110.02834"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6044"} +{"question": "What research papers are on 3D human pose estimation from single monocular images?", "answer": ["Context Modeling in 3D Human Pose Estimation: A Unified Perspective", "Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose", "A simple yet effective baseline for 3d human pose estimation", "End-to-end Recovery of Human Shape and Pose", "Learning to Estimate 3D Human Pose and Shape from a Single Color Image", "Integral Human Pose Regression", "SRNet: Improving Generalization in 3D Human Pose Estimation with a Split-and-Recombine Approach"], "answer_arxiv_id": ["2103.15507", "1611.07828", "1705.03098", "1712.06584", "1805.04092", "1711.08229", "2007.09389"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_6045"} +{"question": "Which studies were done on 2D diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Diffusion Models in Vision: A Survey"], "answer_arxiv_id": ["2006.11239", "2112.10752", "2209.04747"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_6046"} +{"question": "What research has been done on vision-language pre-training in multimodal self-supervised representation learning?", "answer": ["MAMO: Masked Multimodal Modeling for Fine-Grained Vision-Language\n Representation Learning", "SMAUG: Sparse Masked Autoencoder for Efficient Video-Language\n Pre-training"], "answer_arxiv_id": ["2210.04183", "2211.11446"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_6047"} +{"question": "What is the foundation work for graph neural networks (GNNs)?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["1609.02907", "1710.10903", "1810.00826"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_6048"} +{"question": "What papers suggest methods for training of energy-based models?", "answer": ["How to Train Your Energy-Based Models"], "answer_arxiv_id": ["2101.03288"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_6049"} +{"question": "Which works did research scaling offline RL in multi-task Atari setting with data from a wide range of games?", "answer": ["An Optimistic Perspective on Offline Reinforcement Learning", "Multi-Game Decision Transformers"], "answer_arxiv_id": ["1907.04543v4", "2205.15241"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_6050"} +{"question": "What papers reported strong results in multimodal applications using weights from large pretrained models and multimodal joint training?", "answer": ["ClipCap: CLIP Prefix for Image Captioning", "CLIP Models are Few-shot Learners: Empirical Studies on VQA and Visual Entailment", "mSLAM: Massively multilingual joint pre-training for speech and text"], "answer_arxiv_id": ["2111.09734", "2203.07190", "2202.01374"], "source_meta": {"published_time": "20220401"}, "qid": "AutoScholarQuery_train_6051"} +{"question": "Which works initiated efforts to emulate complex reasoning with symbolic representations and problem solvers in question-answering?", "answer": ["A Spoken Dialogue System for Spatial Question Answering in a Physical Blocks World"], "answer_arxiv_id": ["1911.02524"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_6052"} +{"question": "What works developed efficient and precise retrieval methods for the refinement of RAG?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models", "EASE: Entity-Aware Contrastive Learning of Sentence Embedding", "Large Language Models with Controllable Working Memory", "Knowledge Graph-Augmented Language Models for Knowledge-Grounded\n Dialogue Generation"], "answer_arxiv_id": ["1911.00172", "2205.04260", "2211.05110", "2305.18846"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_6053"} +{"question": "Which papers utilize model-generated feedback to improve task performance?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback"], "answer_arxiv_id": ["2303.17651"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_6054"} +{"question": "Can you list research that improved contrastive learning by re-weighting the negative pairs or the loss?", "answer": ["Debiased Contrastive Learning", "Contrastive Learning with Hard Negative Samples"], "answer_arxiv_id": ["2007.00224", "2010.04592"], "source_meta": {"published_time": "20230218"}, "qid": "AutoScholarQuery_train_6055"} +{"question": "Which works focus on exploring different baselines in learning with privileged information?", "answer": ["MIML-FCN+: Multi-instance Multi-label Learning via Fully Convolutional Networks with Privileged Information", "Unifying Distillation and Privileged Information", "Deep Learning under Privileged Information Using Heteroscedastic Dropout"], "answer_arxiv_id": ["1702.08681", "1511.03643", "1805.11614"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_6056"} +{"question": "Which works propose variations of policy gradient methods?", "answer": ["Continuous control with deep reinforcement learning", "Trust Region Policy Optimization", "Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1509.02971", "1502.05477", "1707.06347v2"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_6057"} +{"question": "Could you tell me about the research that extended the federated learning ideas of FedPAQ to introduce a local gradient tracking scheme targeting non-i.i.d data?", "answer": ["Federated Learning with Compression: Unified Analysis and Sharp\n Guarantees"], "answer_arxiv_id": ["2007.01154"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_6058"} +{"question": "Could you provide me some papers that explored knowledge distillation in the multimodal domain through human-annotated verbalizations?", "answer": ["Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning"], "answer_arxiv_id": ["2304.08485", "2304.10592", "2305.06500"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_6059"} +{"question": "In what works does the temporal knowledge graph question answering methodology deconstruct the initial question into sub-questions for resolution?", "answer": ["TEQUILA: Temporal Question Answering over Knowledge Bases"], "answer_arxiv_id": ["1908.03650"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_6060"} +{"question": "Could you provide me some research papers involved in prompting fixed pre-trained LLMs with graphical inputs?", "answer": ["Complex Logical Reasoning over Knowledge Graphs using Large Language\n Models", "GPT4Graph: Can Large Language Models Understand Graph Structured Data ?\n An Empirical Evaluation and Benchmarking", "Knowledge-Augmented Language Model Prompting for Zero-Shot Knowledge Graph Question Answering"], "answer_arxiv_id": ["2305.01157", "2305.15066", "2306.04136v1"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_6061"} +{"question": "Can you provide some works that explored in-context few-shot learning in different tasks?", "answer": ["ReAct: Synergizing Reasoning and Acting in Language Models", "Chameleon: Plug-and-Play Compositional Reasoning with Large Language\n Models", "Cross-Task Generalization via Natural Language Crowdsourcing\n Instructions", "Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2210.03629", "2304.09842", "2104.08773", "2210.03493"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_6062"} +{"question": "Which research focus on cycle-based rule learning?", "answer": ["Cycle Representation Learning for Inductive Relation Prediction"], "answer_arxiv_id": ["2110.02510"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_6063"} +{"question": "Which papers propose explicit-based methods for clothed human shape estimation?", "answer": ["Video Based Reconstruction of 3D People Models", "Learning to Reconstruct People in Clothing from a Single RGB Camera", "Detailed Human Shape Estimation from a Single Image by Hierarchical Mesh\n Deformation", "Detailed Human Avatars from Monocular Video", "Tex2Shape: Detailed Full Human Body Geometry From a Single Image", "BCNet: Learning Body and Cloth Shape from A Single Image", "Detailed Avatar Recovery from Single Image"], "answer_arxiv_id": ["1803.04758", "1903.05885", "1904.10506", "1808.01338", "1904.08645", "2004.00214", "2108.02931"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6064"} +{"question": "Could you provide me with some resources that discuss scene design learning from existing 3D scene databases or refining 3D scene with user input iteratively?", "answer": ["SceneSeer: 3D Scene Design with Natural Language", "Text2Scene: Generating Compositional Scenes from Textual Descriptions", "DiffuScene: Denoising Diffusion Models for Generative Indoor Scene\n Synthesis", "RoomDesigner: Encoding Anchor-latents for Style-consistent and\n Shape-compatible Indoor Scene Generation", "SceneFormer: Indoor Scene Generation with Transformers", "LEGO-Net: Learning Regular Rearrangements of Objects in Rooms"], "answer_arxiv_id": ["1703.00050", "1809.01110", "2303.14207", "2310.10027", "2012.09793", "2301.09629"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_6065"} +{"question": "Which papers have investigated when catastrophic forgetting occurs?", "answer": ["How catastrophic can catastrophic forgetting be in linear regression?"], "answer_arxiv_id": ["2205.09588"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6066"} +{"question": "Which studies have proposed alternative architectures for HPS regression?", "answer": ["PARE: Part Attention Regressor for 3D Human Body Estimation", "HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D\n Human Pose and Shape Estimation", "PyMAF-X: Towards Well-aligned Full-body Model Regression from Monocular\n Images", "NIKI: Neural Inverse Kinematics with Invertible Neural Networks for 3D\n Human Pose and Shape Estimation", "Humans in 4D: Reconstructing and Tracking Humans with Transformers", "SPEC: Seeing People in the Wild with an Estimated Camera"], "answer_arxiv_id": ["2104.08527", "2011.14672", "2207.06400", "2305.08590", "2305.20091", "2110.00620"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6067"} +{"question": "What studies extended the HNN's to applications in knowledge graphs?", "answer": ["Multi-relational Poincaré Graph Embeddings", "Low-Dimensional Hyperbolic Knowledge Graph Embeddings"], "answer_arxiv_id": ["1905.09791", "2005.00545"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_6068"} +{"question": "Any works about real-world scenario databases for development and testing of planning module?", "answer": ["nuPlan: A closed-loop ML-based planning benchmark for autonomous vehicles"], "answer_arxiv_id": ["2106.11810"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_6069"} +{"question": "Which studies identified failures of RCSL in stochastic environments?", "answer": ["Planning from Pixels using Inverse Dynamics Models", "Upside-Down Reinforcement Learning Can Diverge in Stochastic Environments With Episodic Resets"], "answer_arxiv_id": ["2012.02419", "2205.06595"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_6070"} +{"question": "Could you provide studies on fine-tuning pre-trained Diffusion Models (DMs) for conditional image manipulation using text prompts?", "answer": ["Imagic: Text-Based Real Image Editing with Diffusion Models", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation"], "answer_arxiv_id": ["2210.09276", "2110.02711"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_6071"} +{"question": "Which paper defined the LegT matrices (𝑨,𝑩)𝑨𝑩(\\{A},\\{B}) to be scaled by a factor of 222 in order to be properly timescale normalized?", "answer": ["HiPPO: Recurrent Memory with Optimal Polynomial Projections"], "answer_arxiv_id": ["2008.07669"], "source_meta": {"published_time": "20220624"}, "qid": "AutoScholarQuery_train_6072"} +{"question": "Which studies showed that reliance on positional information affects LLMs capabilities in arithmetic, multiple-choice question-answering, and text generation evaluation?", "answer": ["Positional Description Matters for Transformers Arithmetic", "Large Language Models Are Not Robust Multiple Choice Selectors", "Large Language Models Sensitivity to The Order of Options in\n Multiple-Choice Questions", "Large Language Models are not Fair Evaluators"], "answer_arxiv_id": ["2311.14737", "2309.03882", "2308.11483", "2305.17926"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_6073"} +{"question": "What works proposed learning one generator that can propose multiple-length subgoals for adaptive planning?", "answer": ["Hierarchical Imitation Learning with Vector Quantized Models"], "answer_arxiv_id": ["2301.12962"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_6074"} +{"question": "What works concentrate on acquiring a collection of local features?", "answer": ["Region Similarity Representation Learning", "Dense Contrastive Learning for Self-Supervised Visual Pre-Training", "VICRegL: Self-Supervised Learning of Local Visual Features", "Self-supervised Feature Learning for 3D Medical Images by Playing a Rubik’s Cube", "Models Genesis", "Transferable Visual Words: Exploiting the Semantics of Anatomical Patterns for Self-supervised Learning"], "answer_arxiv_id": ["2103.12902", "2011.09157", "2210.01571", "1910.02241", "2004.07882", "2102.10680"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_6075"} +{"question": "What papers discuss the use of surrogate gradients as a solution to the non-differentiable nature of activation functions in SNNs?", "answer": ["Convolutional Networks for Fast, Energy-Efficient Neuromorphic Computing", "Spiking Deep Networks with LIF Neurons", "SuperSpike: Supervised learning in multi-layer spiking neural networks", "Training Deep Spiking Neural Networks using Backpropagation"], "answer_arxiv_id": ["1603.08270", "1510.08829", "1705.11146", "1608.08782"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_6076"} +{"question": "Which works touched upon the concepts of diffeomorphism group in the context of disentanglement and nonlinear ICA?", "answer": ["Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style", "Identifiability Results for Multimodal Contrastive Learning"], "answer_arxiv_id": ["2106.04619v4", "2303.09166"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_6077"} +{"question": "Which papers have worked on using large language models as training data generators?", "answer": ["Symbolic Knowledge Distillation: from General Language Models to\n Commonsense Models"], "answer_arxiv_id": ["2110.07178"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_6078"} +{"question": "Could you provide me some works that adapted VLMs for training with pathological images and text?", "answer": ["Towards a Visual-Language Foundation Model for Computational Pathology"], "answer_arxiv_id": ["2307.12914"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_6079"} +{"question": "Which study was done to understand the importance of equations in the Chain-of-Thought prompt in language models?", "answer": ["Complementary Explanations for Effective In-Context Learning"], "answer_arxiv_id": ["2211.13892"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_6080"} +{"question": "Any works about generative AI used for 3D reconstruction?", "answer": ["Multiview Compressive Coding for 3D Reconstruction", "DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Point⋅E: A System for Generating 3D Point Clouds from Complex Prompts", "Shap⋅E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2301.08247", "2209.14988", "2211.10440", "2212.08751", "2305.02463"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_6081"} +{"question": "Which works have researched on compositional action recognition approaches to recognize hand actions?", "answer": ["Something-Else: Compositional Action Recognition with Spatial-Temporal\n Interaction Networks", "Motion Guided Attention Fusion to Recognize Interactions from Videos", "Revisiting spatio-temporal layouts for compositional action recognition", "Object-Region Video Transformers", "Is an Object-Centric Video Representation Beneficial for Transfer?", "Modelling Spatio-Temporal Interactions for Compositional Action\n Recognition"], "answer_arxiv_id": ["1912.09930", "2104.00646", "2111.01936", "2110.06915", "2207.10075", "2305.02673"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_6082"} +{"question": "What literature proposes local methods for addressing the communication bottleneck in distributed algorithms?", "answer": ["Don’t Use Large Mini-Batches, Use Local SGD", "Local SGD Converges Fast and Communicates Little", "Tighter Theory for Local SGD on Identical and Heterogeneous Data", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Is Local SGD Better than Minibatch SGD?", "Minibatch vs Local SGD for Heterogeneous Distributed Learning", "A Unified Theory of Decentralized SGD with Changing Topology and Local Updates", "ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!†"], "answer_arxiv_id": ["1808.07217", "1805.09767", "1909.04746", "1910.06378", "2002.07839", "2006.04735", "2003.10422", "2202.09357"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_6083"} +{"question": "Who first framed GFlowNets as a reinforcement learning (RL) algorithm?", "answer": ["Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation"], "answer_arxiv_id": ["2106.04399"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_6084"} +{"question": "Which works describe the early development of text-to-image generation models which utilized GAN-based models?", "answer": ["Taming Transformers for High-Resolution Image Synthesis", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked\n Generative Adversarial Networks"], "answer_arxiv_id": ["2012.09841", "1612.03242"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_6085"} +{"question": "What studies employ a conditional VAE for predicting a category and bounding box based on previously predicted elements?", "answer": ["Geometry Aligned Variational Transformer for Image-conditioned Layout\n Generation"], "answer_arxiv_id": ["2209.00852"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_6086"} +{"question": "What studies used pruning to identify subnetworks to showcase modular building blocks of model behavior?", "answer": ["Break It Down: Evidence for Structural Compositionality in Neural\n Networks"], "answer_arxiv_id": ["2301.10884"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_6087"} +{"question": "Could you provide me some works that proposed restricting the observation space of skill learning to x-y Cartesian coordinates to increase in traveled distances?", "answer": ["Lipschitz-constrained Unsupervised Skill Discovery", "Mutual Information State Intrinsic Control"], "answer_arxiv_id": ["2202.00914", "2103.08107"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_6088"} +{"question": "What studies employed hierarchical sampling strategies to enhance NeRF’s performance and scalability?", "answer": ["Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Baking Neural Radiance Fields for Real-Time View Synthesis", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "PlenOctrees for Real-time Rendering of Neural Radiance Fields"], "answer_arxiv_id": ["2111.12077", "2103.14645", "2103.13744", "2103.14024"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_6089"} +{"question": "What studies attempted to combine 2D self-supervised object-centric models with neural scene representations to decompose a 3D scene?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational Inference", "SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition"], "answer_arxiv_id": ["1901.11390", "1903.00450", "2001.02407"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_6090"} +{"question": "Can you provide works that utilized VAEs within Latent space Bayesian optimization?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_6091"} +{"question": "Any works about distinguishing generative images from natural images in the context of watermarking?", "answer": ["Responsible Disclosure of Generative Models Using Scalable\n Fingerprinting", "The Stable Signature: Rooting Watermarks in Latent Diffusion Models", "DiffusionShield: A Watermark for Copyright Protection against Generative\n Diffusion Models"], "answer_arxiv_id": ["2012.08726", "2303.15435", "2306.04642"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_6092"} +{"question": "What studies utilized text-to-image diffusion models for image editing tasks?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Prompt-to-Prompt Image Editing with Cross Attention Control", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2111.14818", "2204.06125", "2210.09276", "2208.01626", "2108.01073", "2211.09800"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_6093"} +{"question": "Which papers evaluate the effectiveness of VG-boosting methods in VQA based on the ID/OOD split?", "answer": ["Don't Just Assume; Look and Answer: Overcoming Priors for Visual\n Question Answering", "VisFIS: Visual Feature Importance Supervision with\n Right-for-the-Right-Reason Objectives"], "answer_arxiv_id": ["1712.00377", "2206.11212"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_6094"} +{"question": "Which study proposed modeling human-object relative distance field in data-driven manner and joint post-optimization?", "answer": ["CHORE: Contact, Human and Object REconstruction from a single RGB image", "StackFLOW: Monocular Human-Object Reconstruction by Stacked Normalizing Flow with Offset"], "answer_arxiv_id": ["2204.02445", "2407.20545v1"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_6095"} +{"question": "Which works utilized Optimal Transport (OT) for class-imbalanced learning tasks?", "answer": ["Learning to Re-weight Examples with Optimal Transport for Imbalanced Classification", "SoLar: Sinkhorn Label Refinery for Imbalanced Partial-Label Learning"], "answer_arxiv_id": ["2208.02951", "2209.10365"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_6096"} +{"question": "What prior works found deep network layers hierarchically align with the visual stream?", "answer": ["The Algonauts Project 2021 Challenge: How the Human Brain Makes Sense of a World in Motion", "Teaching Matters: Investigating the Role of Supervision in Vision\n Transformers"], "answer_arxiv_id": ["2104.13714v1", "2212.03862"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_6097"} +{"question": "Which works improved the reader component of the retrieve-then-read model pipeline?", "answer": ["Leveraging Passage Retrieval with Generative Models for Open Domain Question Answering", "UnitedQA: A Hybrid Approach for Open Domain Question Answering", "KG-FiD: Infusing Knowledge Graph in Fusion-in-Decoder for Open-Domain Question Answering"], "answer_arxiv_id": ["2007.01282", "2101.00178", "2110.04330"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_6098"} +{"question": "Which works employed uncertainty estimation to overcome suboptimal conservatism in offline RL?", "answer": ["Confidence-Conditioned Value Functions for Offline Reinforcement Learning", "Offline RL Policies Should be Trained to be Adaptive"], "answer_arxiv_id": ["2212.04607", "2207.02200"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6099"} +{"question": "Can you provide the reference of the used StableDiffusion model in 'bib.bib11' research?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_6100"} +{"question": "What studies discuss creating agents that receive language instructions corresponding to the relevant reward functions in RL?", "answer": ["A Survey of Reinforcement Learning Informed by Natural Language"], "answer_arxiv_id": ["1906.03926"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_6101"} +{"question": "Which work proposed a method for neural density estimation on SO(3) using the exponential map?", "answer": ["Reparameterizing Distributions on Lie Groups"], "answer_arxiv_id": ["1903.02958"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_6102"} +{"question": "In what studies the researchers have worked on autoregressive prediction models?", "answer": ["Delving Deeper into Convolutional Networks for Learning Video Representations", "Decomposing Motion and Content for Natural Video Sequence Prediction", "Stochastic Variational Video Prediction", "Stochastic Video Generation with a Learned Prior"], "answer_arxiv_id": ["1511.06432", "1706.08033", "1710.11252", "1802.07687"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_6103"} +{"question": "Could you provide the study about early stopping variant of ATPGD?", "answer": ["Overfitting in adversarially robust deep learning"], "answer_arxiv_id": ["2002.11569"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6104"} +{"question": "Is there any work related to learning a prior model on synthetic tabular data?", "answer": ["TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second"], "answer_arxiv_id": ["2207.01848"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_6105"} +{"question": "Are there any works inspired by instruction tuning and in-context learning to handle few-shot segmentation?", "answer": ["Instruction Tuning with GPT-4", "Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "What Can Transformers Learn In-Context? A Case Study of Simple Function\n Classes"], "answer_arxiv_id": ["2304.03277", "2202.12837", "2208.01066"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6106"} +{"question": "Which papers discuss the first techniques for self-supervised 3D training, mostly applicable to dense scans of single objects?", "answer": ["Self-supervised Learning of Point Clouds via Orientation Estimation", "Self-Supervised Deep Learning on Point Clouds by Reconstructing Space", "Self-Supervised Pretraining of 3D Features on any Point-Cloud"], "answer_arxiv_id": ["2008.00305", "1901.08396", "2101.02691"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_6107"} +{"question": "Can you tell me about the works that tackled the graph learning problem for non-temporal data in a probabilistic model?", "answer": ["Learning Discrete Structures for Graph Neural Networks"], "answer_arxiv_id": ["1903.11960"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_6108"} +{"question": "What works propose the usage of a singular model for data evaluation in FAL?", "answer": ["Federated Active Learning (F-AL): an Efficient Annotation Strategy for\n Federated Learning"], "answer_arxiv_id": ["2202.00195"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_6109"} +{"question": "What papers have introduced soft-parameter-sharing methods for Multi-Task Learning?", "answer": ["Cross-stitch Networks for Multi-task Learning", "Latent Multi-task Architecture Learning", "MTL-NAS: Task-Agnostic Neural Architecture Search towards General-Purpose Multi-Task Learning", "End-to-End Multi-Task Learning with Attention"], "answer_arxiv_id": ["1604.03539", "1705.08142", "2003.14058", "1803.10704"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6110"} +{"question": "What researches investigated complex trigger patterns such as adversarial perturbation, natural reflection, and sample-wise patterns?", "answer": ["Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks", "Input-Aware Dynamic Backdoor Attack", "Invisible Backdoor Attack with Sample-Specific Triggers"], "answer_arxiv_id": ["2007.02343", "2010.08138", "2012.03816"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_6111"} +{"question": "Which works aim to recognize or localize objects in unsupervised object discovery?", "answer": ["Unsupervised Object Detection with LiDAR Clues", "Self-Supervised Transformers for Unsupervised Object Discovery using\n Normalized Cut"], "answer_arxiv_id": ["2011.12953", "2202.11539"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6112"} +{"question": "What papers made an effort to learn normal patterns by using the One-class models?", "answer": ["Adversarially Learned One-Class Classifier for Novelty Detection"], "answer_arxiv_id": ["1802.09088"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_6113"} +{"question": "What works conduct analysis based on PAC bounds in the study of Contextual MDPs?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "Markov Decision Processes with Continuous Side Information", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches", "Policy Certificates: Towards Accountable Reinforcement Learning", "Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles"], "answer_arxiv_id": ["1610.09512", "1711.05726", "1811.08540", "1811.03056", "1910.10597"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_6114"} +{"question": "What works have attempted to improve predictive performance in Spatio-Temporal Graph Neural Networks (STGNNs) by leveraging an adaptive adjacency matrix?", "answer": ["Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting"], "answer_arxiv_id": ["2007.02842"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_6115"} +{"question": "Can you provide some works where the method of wrapping the input with context is used in the addition-based PEFT?", "answer": ["Prompting Visual-Language Models for Efficient Video Understanding", "Making Pre-trained Language Models Better Few-shot Learners", "Knowledgeable Prompt-tuning: Incorporating Knowledge into Prompt\n Verbalizer for Text Classification", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "OpenPrompt: An Open-source Framework for Prompt-learning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2112.04478", "2012.15723", "2108.02035", "2107.13586v1", "2111.01998", "2101.00190"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_6116"} +{"question": "Can you tell me about the work that proposed to distill 3D models from a pre-trained 2D diffusion model?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6117"} +{"question": "Can you provide me any research papers that employ the iterative proportional fitting technique for solving the DSB problem?", "answer": ["Solving Schrödinger Bridges via Maximum Likelihood", "Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling"], "answer_arxiv_id": ["2106.02081", "2106.01357"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_6118"} +{"question": "What works have utilized reinforcement learning for training controllers in dexterous manipulation?", "answer": ["Deep Dynamics Models for Learning Dexterous Manipulation", "Learning Dexterous In-Hand Manipulation", "Solving Rubik’s Cube with a Robot Hand", "A System for General In-Hand Object Re-Orientation", "State-Only Imitation Learning for Dexterous Manipulation", "Dexterous Manipulation with Deep Reinforcement Learning: Efficient, General, and Low-Cost", "Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations", "Learning Dexterous Manipulation from Suboptimal Experts", "Learning Dexterous Manipulation for a Soft Robotic Hand from Human Demonstrations", "DexMV: Imitation Learning for Dexterous Manipulation from Human Videos"], "answer_arxiv_id": ["1909.11652", "1808.00177", "1910.07113", "2111.03043", "2004.04650", "1810.06045", "1709.10087", "2010.08587", "1603.06348", "2108.05877"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_6119"} +{"question": "What papers studied causal disentanglement and introduced a set of metrics to measure the robustness?", "answer": ["Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustness"], "answer_arxiv_id": ["1811.00007"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_6120"} +{"question": "Which works have developed multimodal web agents using both screenshots and HTML texts?", "answer": ["Multimodal Web Navigation with Instruction-Finetuned Foundation Models"], "answer_arxiv_id": ["2305.11854"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_6121"} +{"question": "Which studies proposed the use of color consistency loss between stereo images in training for self-supervised MDE?", "answer": ["Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescue"], "answer_arxiv_id": ["1603.04992"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6122"} +{"question": "Which works discuss environment-aware text-to-speech?", "answer": ["Environment Aware Text-to-Speech Synthesis"], "answer_arxiv_id": ["2110.03887"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_6123"} +{"question": "What works are related to LiDAR-based egocentric 3D object detection methods?", "answer": ["VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "STD: Sparse-to-Dense 3D Object Detector for Point Cloud", "PointPillars: Fast Encoders for Object Detection from Point Clouds"], "answer_arxiv_id": ["1711.06396", "1907.10471", "1812.05784"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_6124"} +{"question": "Which paper studied mapping natural signals into a compact discrete representation space using VQ-GANs?", "answer": ["Taming Transformers for High-Resolution Image Synthesis", "Vector-quantized Image Modeling with Improved VQGAN"], "answer_arxiv_id": ["2012.09841", "2110.04627"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_6125"} +{"question": "Are there any studies about adaptability to unknown, arbitrary disturbances for model-free reinforcement learning algorithms?", "answer": ["Online Learning in Markov Decision Processes with Adversarially Chosen Transition Probability Distributions"], "answer_arxiv_id": ["1303.3055"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_6126"} +{"question": "Could you provide me some works about scene flow estimation from pairs or sequences of RGB-D frames?", "answer": ["RAFT-3D: Scene Flow using Rigid-Motion Embeddings", "FlowNet3D++: Geometric Losses For Deep Scene Flow Estimation"], "answer_arxiv_id": ["2012.00726", "1912.01438"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_6127"} +{"question": "Any papers about utilising each batch of testing samples to update partial weights in test-time adaptation?", "answer": ["Test-Time Training with Self-Supervision for Generalization under\n Distribution Shifts", "Test-time Unsupervised Domain Adaptation", "Temporal Coherent Test-Time Optimization for Robust Video Classification"], "answer_arxiv_id": ["1909.13231", "2010.01926", "2302.14309"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_6128"} +{"question": "Which research have used multi-task pretraining for video retrieval tasks?", "answer": ["UniVL: A Unified Video and Language Pre-Training Model for Multimodal\n Understanding and Generation"], "answer_arxiv_id": ["2002.06353"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_6129"} +{"question": "What research is there on generating images conditioned on textual input using diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Imagen Video: High Definition Video Generation with Diffusion Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2205.11487", "2210.02303", "2112.10741"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_6130"} +{"question": "Which research papers examined the impact of adversarial training on generalization in the field of natural language processing?", "answer": ["Robust Neural Machine Translation with Doubly Adversarial Inputs", "Improving Neural Language Modeling via Adversarial Training", "SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language\n Models through Principled Regularized Optimization"], "answer_arxiv_id": ["1906.02443", "1906.03805", "1911.03437"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_6131"} +{"question": "What papers used human- and machine-generated natural language feedback for other tasks?", "answer": ["Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback", "PEER: A Collaborative Language Model", "Self-critiquing models for assisting human evaluators", "Constitutional AI: Harmlessness from AI Feedback", "Generating Sequences by Learning to [Self-]Correct"], "answer_arxiv_id": ["2204.05862", "2208.11663", "2206.05802v2", "2212.08073", "2211.00053"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_6132"} +{"question": "Can you point out some papers that focus on step localization in videos?", "answer": ["StepFormer: Self-supervised Step Discovery and Localization in\n Instructional Videos", "Hierarchical Video-Moment Retrieval and Step-Captioning"], "answer_arxiv_id": ["2304.13265", "2303.16406"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6133"} +{"question": "What research provided a description of training trajectories displaying jumps from saddle to saddle until reaching the minimum ℓ1-subscript-ℓ1-norm solution?", "answer": ["Saddle-to-Saddle Dynamics in Diagonal Linear Networks"], "answer_arxiv_id": ["2304.00488"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_6134"} +{"question": "Which paper focused on Markov jump systems where the system dynamics switch between a finite number of linear systems?", "answer": ["Identification and Adaptive Control of Markov Jump Systems: Sample Complexity and Regret Bounds"], "answer_arxiv_id": ["2111.07018"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_6135"} +{"question": "Which research works discussed the use of single-round retrieval augmentation strategy to address the limitation of LLMs?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models", "Improving language models by retrieving from trillions of tokens", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Leveraging Passage Retrieval with Generative Models for Open Domain\n Question Answering", "Retrieval as Attention: End-to-end Learning of Retrieval and Reading\n within a Single Transformer", "REPLUG: Retrieval-Augmented Black-Box Language Models"], "answer_arxiv_id": ["1911.00172", "2112.04426", "2005.11401", "2007.01282", "2212.02027", "2301.12652"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_6136"} +{"question": "What papers have studied distant signals for document newsworthiness?", "answer": ["Modeling \"Newsworthiness\" for Lead-Generation Across Corpora"], "answer_arxiv_id": ["2104.09653"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_6137"} +{"question": "Which works have used temporal-difference error as a priority signal in experience replay?", "answer": ["Prioritized Experience Replay"], "answer_arxiv_id": ["1511.05952"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_6138"} +{"question": "Which works introduced Score Jacobian Chaining that converges towards a similar algorithm as SDS?", "answer": ["Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2212.00774v1"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_6139"} +{"question": "What work conducts separate vertex-and-face sequence approach in the process of mesh generation?", "answer": ["PolyGen: An Autoregressive Generative Model of 3D Meshes"], "answer_arxiv_id": ["2002.10880"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_6140"} +{"question": "Which works introduce LoRA that only tunes new initialized low-rank parameters?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_6141"} +{"question": "Which papers address the problem of tensor network-topology selection (TN-TS)?", "answer": ["Active learning of tree tensor networks using optimal least-squares", "Geometry of tree-based tensor formats in tensor Banach spaces", "Automatic structural optimization of tree tensor networks", "MARS: Masked Automatic Ranks Selection in Tensor Decompositions", "Adaptively Topological Tensor Network for Multi-view Subspace Clustering"], "answer_arxiv_id": ["2104.13436", "2011.08466", "2209.03196", "2006.10859v3", "2305.00716"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_6142"} +{"question": "Could you tell me which work introduced neural abstractions of the classical Production Systems framework to learn rule-based templates?", "answer": ["Neural Production Systems: Learning Rule-Governed Visual Dynamics"], "answer_arxiv_id": ["2103.01937v3"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_6143"} +{"question": "What papers proposed methods for improving input images with occlusions/truncation in the context of Animal 3D Shape Reconstruction?", "answer": ["ARTIC3D: Learning Robust Articulated 3D Shapes from Noisy Web Image\n Collections"], "answer_arxiv_id": ["2306.04619"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_6144"} +{"question": "What studies improve the regret bounds in tabular cases in decentralized policy learning?", "answer": ["Online Reinforcement Learning in Stochastic Games", "Online Learning in Unknown Markov Games"], "answer_arxiv_id": ["1712.00579", "2010.15020"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_6145"} +{"question": "What works utilized provided success examples to reach desired outcome states and how did they approach the issue?", "answer": ["Variational Inverse Control with Events: A General Framework for Data-Driven Reward Definition", "End-to-End Robotic Reinforcement Learning without Reward Engineering", "Replacing Rewards with Examples: Example-Based Policy Search via Recursive Classification", "MURAL: Meta-Learning Uncertainty-Aware Rewards for Outcome-Driven Reinforcement Learning"], "answer_arxiv_id": ["1805.11686", "1904.07854", "2103.12656", "2107.07184"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_6146"} +{"question": "What comprehensive surveys are there on regression-based methods for 3D face reconstruction?", "answer": ["3D Morphable Face Models -- Past, Present and Future"], "answer_arxiv_id": ["1909.01815"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_6147"} +{"question": "Which papers worked on using large language models as zero-shot planners?", "answer": ["Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents", "Inner Monologue: Embodied Reasoning through Planning with Language Models", "Code as Policies: Language Model Programs for Embodied Control", "Text2Motion: From Natural Language Instructions to Feasible Plans", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances"], "answer_arxiv_id": ["2201.07207", "2207.05608", "2209.07753", "2303.12153", "2204.01691"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_6148"} +{"question": "Can you list some works that demonstrated the application of trained ViT for numerous downstream tasks?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "General Facial Representation Learning in a Visual-Linguistic Manner", "Emerging Properties in Self-Supervised Vision Transformers", "Deep ViT Features as Dense Visual Descriptors"], "answer_arxiv_id": ["2103.00020", "2112.03109", "2104.14294", "2112.05814"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_6149"} +{"question": "Could you provide me some studies that proved the training convergence and generalization capacity of networks can be described by some corresponding kernels?", "answer": ["Mean-field theory of two-layers neural networks: dimension-free bounds and kernel limit", "Learning One-hidden-layer ReLU Networks via Gradient Descent", "Deep Convolutional Networks as shallow Gaussian Processes", "Gradient Descent Provably Optimizes Over-parameterized Neural Networks"], "answer_arxiv_id": ["1902.06015", "1806.07808", "1808.05587", "1810.02054"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_6150"} +{"question": "Which works have applied skip connections between distant time steps to improve gradient flow?", "answer": ["A Clockwork RNN", "Dilated Recurrent Neural Networks", "Hierarchical Multiscale Recurrent Neural Networks"], "answer_arxiv_id": ["1402.3511", "1710.02224", "1609.01704"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_6151"} +{"question": "Which works have addressed the problem of template selection in single-step models using a classification neural network?", "answer": ["Retrosynthesis Prediction with Conditional Graph Logic Network"], "answer_arxiv_id": ["2001.01408"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6152"} +{"question": "What works discuss the use of CVaR for action selection in distributional reinforcement learning literature?", "answer": ["Implicit Quantile Networks for Distributional Reinforcement Learning", "Worst Cases Policy Gradients"], "answer_arxiv_id": ["1806.06923", "1911.03618"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_6153"} +{"question": "What works have explored the open vocabulary setting in the field of 3D scene understanding?", "answer": ["CLIP-Fields: Weakly Supervised Semantic Fields for Robotic Memory", "LERF: Language Embedded Radiance Fields", "Visual Language Maps for Robot Navigation", "Open-vocabulary Queryable Scene Representations for Real World Planning", "OpenScene: 3D Scene Understanding with Open Vocabularies", "PLA: Language-Driven Open-Vocabulary 3D Scene Understanding", "Feature-Realistic Neural Fusion for Real-Time, Open Set Scene Understanding", "Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Models", "ConceptFusion: Open-set Multimodal 3D Mapping", "Decomposing NeRF for Editing via Feature Field Distillation"], "answer_arxiv_id": ["2210.05663", "2303.09553", "2210.05714", "2209.09874", "2211.15654", "2211.16312", "2210.03043", "2207.11514", "2302.07241", "2205.15585"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_6154"} +{"question": "Which papers are about zero-shot point cloud understanding using training a 3D encoder?", "answer": ["ULIP: Learning a Unified Representation of Language, Images, and Point\n Clouds for 3D Understanding", "PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning", "OpenScene: 3D Scene Understanding with Open Vocabularies", "CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D\n Recognition", "CLIP2Point: Transfer CLIP to Point Cloud Classification with Image-Depth\n Pre-training", "CLIP$^2$: Contrastive Language-Image-Point Pretraining from Real-World\n Point Cloud Data"], "answer_arxiv_id": ["2212.05171", "2211.11682", "2211.15654", "2303.11313", "2210.01055", "2303.12417"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6155"} +{"question": "What studies have used BEV layouts to augment image data with synthetic single or multi-view images?", "answer": ["BEVControl: Accurately Controlling Street-view Elements with\n Multi-perspective Consistency via BEV Sketch Layout", "Street-View Image Generation from a Bird's-Eye View Layout", "LayoutDiffusion: Controllable Diffusion Model for Layout-to-image\n Generation", "LayoutDiffuse: Adapting Foundational Diffusion Models for\n Layout-to-Image Generation"], "answer_arxiv_id": ["2308.01661", "2301.04634", "2303.17189", "2302.08908"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6156"} +{"question": "Which research papers introduced contrastive learning as part of SSL's shift toward utilizing inductive bias?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "1911.05722"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_6157"} +{"question": "Any manuscripts which discuss about using Reinforcement Learning in medical image segmentation?", "answer": ["Iteratively-Refined Interactive 3D Medical Image Segmentation with\n Multi-Agent Reinforcement Learning", "Searching Learning Strategy with Reinforcement Learning for 3D Medical\n Image Segmentation"], "answer_arxiv_id": ["1911.10334", "2006.05847"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_6158"} +{"question": "Which papers explored ways to extend the amount of information contained in each experiment?", "answer": ["Near-Optimal Multi-Perturbation Experimental Design for Causal Structure Learning"], "answer_arxiv_id": ["2105.14024"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_6159"} +{"question": "Which works discussed the original design of Input-Convex Neural Network (ICNN) in the context of energy models?", "answer": ["Input Convex Neural Networks"], "answer_arxiv_id": ["1609.07152v3"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_6160"} +{"question": "Could you provide me with some papers about Diffusion Models (DMs)?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2105.05233", "2010.02502"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_6161"} +{"question": "Are there any studies that brought insights from recent causality research to model non-stationarity in the model-based method?", "answer": ["Factored Adaptation for Non-stationary Reinforcement Learning"], "answer_arxiv_id": ["2203.16582"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_6162"} +{"question": "What work introduced the Global Mapper method for finding a global direction in StyleGAN using text?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery"], "answer_arxiv_id": ["2103.17249"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_6163"} +{"question": "Could you provide some works that aim to control results based on image style in diffusion models?", "answer": ["BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "Prompt-Free Diffusion: Taking \"Text\" out of Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2305.14720", "2305.16223"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6164"} +{"question": "Which works converted existing NLP datasets into instruction format for instruction tuning?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Scaling Instruction-Finetuned Language Models", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks"], "answer_arxiv_id": ["2109.01652", "2210.11416", "2110.08207", "2204.07705"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_6165"} +{"question": "What works utilized ensemble methods to improve policy performance in Deep RL?", "answer": ["MEPG: A Minimalist Ensemble Policy Gradient Framework for Deep Reinforcement Learning", "SUNRISE: A Simple Unified Framework for Ensemble Learning in Deep Reinforcement Learning", "Probabilistic Mixture-of-Experts for Efficient Deep Reinforcement Learning"], "answer_arxiv_id": ["2109.10552", "2007.04938", "2104.09122"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_6166"} +{"question": "Could you provide me some works about Multimodal Large Language Models (MLLMs) focusing on video modality?", "answer": ["VideoChat: Chat-Centric Video Understanding", "VideoLLM: Modeling Video Sequence with Large Language Models"], "answer_arxiv_id": ["2305.06355", "2305.13292"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_6167"} +{"question": "Which papers describe the process of Visual Document Understanding (VDU)?", "answer": ["V-Doc : Visual questions answers with Documents", "Wukong-Reader: Multi-modal Pre-training for Fine-grained Visual Document\n Understanding", "XYLayoutLM: Towards Layout-Aware Multimodal Networks For Visually-Rich\n Document Understanding"], "answer_arxiv_id": ["2205.13724", "2212.09621", "2203.06947"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_6168"} +{"question": "Could you provide some works that have attempted to use LLMs for evaluation?", "answer": ["G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment"], "answer_arxiv_id": ["2303.16634"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_6169"} +{"question": "Which research papers talk about representation learning and different pre-training strategies that can be applied for downstream tasks?", "answer": ["Do Better ImageNet Models Transfer Better?", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1805.08974", "2002.05709"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_6170"} +{"question": "What works investigate the explicit temporal feature modeling and merge it into the embedding algorithms?", "answer": ["Temporal Knowledge Base Completion: New Algorithms and Evaluation Protocols"], "answer_arxiv_id": ["2005.05035"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_6171"} +{"question": "Could you refer me to the studies that used the DiD method in cases with two time slots (pre-treatment and post-treatment phases)?", "answer": ["Balancing, Regression, Difference-In-Differences and Synthetic Control Methods: A Synthesis", "Two-way fixed effects estimators with heterogeneous treatment effects"], "answer_arxiv_id": ["1610.07748v2", "1803.08807v7"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_6172"} +{"question": "What works have been done on exemplar selection techniques and generative modeling in the context of rehearsal-based methods?", "answer": ["Gradient based sample selection for online continual learning"], "answer_arxiv_id": ["1903.08671"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_6173"} +{"question": "Which works utilized template-based approaches in early video captioning?", "answer": ["Fluency-Guided Cross-Lingual Image Captioning"], "answer_arxiv_id": ["1708.04390"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_6174"} +{"question": "Could you provide me some studies that showed many popular methods for explanation and interpretation are not stable against a perturbation or adversarial attack on the input data and model?", "answer": ["The (Un)reliability of saliency methods", "Interpretation of Neural Networks is Fragile", "Towards Robust Interpretability with Self-Explaining Neural Networks", "Interpretable Deep Learning under Fire", "Explanations can be manipulated and geometry is to blame", "Fooling Neural Network Interpretations via Adversarial Model Manipulation"], "answer_arxiv_id": ["1711.00867", "1710.10547", "1806.07538", "1812.00891", "1906.07983", "1902.02041"], "source_meta": {"published_time": "20230106"}, "qid": "AutoScholarQuery_train_6175"} +{"question": "Could you provide me some studies that use the relationship between different images in the dataset to handle weak supervision in semantic segmentation?", "answer": ["Mining Cross-Image Semantics for Weakly Supervised Semantic Segmentation", "CIAN: Cross-Image Affinity Net for Weakly Supervised Semantic\n Segmentation", "Group-Wise Semantic Mining for Weakly Supervised Semantic Segmentation", "Weakly Supervised Semantic Segmentation by Pixel-to-Prototype Contrast"], "answer_arxiv_id": ["2007.01947", "1811.10842", "2012.05007", "2110.07110"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_6176"} +{"question": "What research works extended Compressed Sensing Generative Models (CSGM) to non-linear observations?", "answer": ["Sample Complexity Bounds for 1-bit Compressive Sensing and Binary Stable Embeddings with Generative Priors", "Non-Iterative Recovery from Nonlinear Observations using Generative Models"], "answer_arxiv_id": ["2002.01697", "2205.15749"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_6177"} +{"question": "Are there any research papers that leverage the duality between graph and matrix for preconditioner design?", "answer": ["Deep Learning of Preconditioners for Conjugate Gradient Solvers in Urban Water Related Problems", "Machine-Learned Preconditioners for Linear Solvers in Geophysical Fluid Flows", "Neural-network preconditioners for solving the Dirac equation in lattice gauge theory"], "answer_arxiv_id": ["1906.06925", "2010.02866", "2208.02728"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6178"} +{"question": "What works showed that generative LMs can generate even the intermediate proofs as well as the final answer?", "answer": ["PRover: Proof Generation for Interpretable Reasoning over Rules", "Explaining Answers with Entailment Trees", "ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language", "FaiRR: Faithful and Robust Deductive Reasoning over Natural Language"], "answer_arxiv_id": ["2010.02830", "2104.08661", "2012.13048", "2203.10261"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_6179"} +{"question": "Are there any studies that have done adversarial training on the whole network but also involved multiple forward-backward passes on the first layer with the aim of speeding up adversarial training?", "answer": ["You Only Propagate Once: Accelerating Adversarial Training via Maximal Principle"], "answer_arxiv_id": ["1905.00877"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6180"} +{"question": "Any studies on the learning point-level affordance for various types of manipulation like articulated, deformable object, language-guided, mobile, and bimanual?", "answer": ["UniDoorManip: Learning Universal Door Manipulation Policy Over\n Large-scale and Diverse Door Manipulation Environments", "Articulated Object Manipulation with Coarse-to-fine Affordance for\n Mitigating the Effect of Point Cloud Noise", "Learning Foresightful Dense Visual Affordance for Deformable Object\n Manipulation", "NaturalVLM: Leveraging Fine-grained Natural Language for\n Affordance-Guided Visual Manipulation"], "answer_arxiv_id": ["2403.02604", "2402.18699", "2303.11057", "2403.08355"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_6181"} +{"question": "What studies can provide insights into the theoretical adjustment for conformal inferences computed after model selection?", "answer": ["Finite-sample Efficient Conformal Prediction"], "answer_arxiv_id": ["2104.13871"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_6182"} +{"question": "What work presents the SmallCap model and discusses retrieval augmentation in image captioning?", "answer": ["SmallCap: Lightweight Image Captioning Prompted with Retrieval\n Augmentation"], "answer_arxiv_id": ["2209.15323"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_6183"} +{"question": "Which studies estimate a dense warp and aim to estimate every matchable pixel pair?", "answer": ["DGC-Net: Dense Geometric Correspondence Network", "GLU-Net: Global-Local Universal Network for Dense Flow and\n Correspondences", "Learning Accurate Dense Correspondences and When to Trust Them", "DKM: Dense Kernelized Feature Matching for Geometry Estimation", "PMatch: Paired Masked Image Modeling for Dense Geometric Matching", "PATS: Patch Area Transportation with Subdivision for Local Feature\n Matching"], "answer_arxiv_id": ["1810.08393", "1912.05524", "2101.01710", "2202.00667", "2303.17342", "2303.07700"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_6184"} +{"question": "Which work introduced the concept of 'lock' and 'unlock' training data by leveraging a class-wise perturbation?", "answer": ["Learnability Lock: Authorized Learnability Control Through Adversarial Invertible Transformations"], "answer_arxiv_id": ["2202.03576"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6185"} +{"question": "What works present usage of data augmentation techniques for consistency regularization?", "answer": ["Unsupervised Data Augmentation for Consistency Training", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["1904.12848", "2001.07685v2"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_6186"} +{"question": "What papers propose return-conditioned agents trained on offline RL datasets?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Online Decision Transformer"], "answer_arxiv_id": ["2106.01345", "2202.05607"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_6187"} +{"question": "What studies present instruction following models that use Reinforcement Learning with AI model feedback (RLAIF)?", "answer": ["Constitutional AI: Harmlessness from AI Feedback"], "answer_arxiv_id": ["2212.08073"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_6188"} +{"question": "What works proposed learning the Energy-based model (EBM) through the noise-contrastive estimation (NCE) method?", "answer": ["A Contrastive Learning Approach for Training Variational Autoencoder Priors", "Adaptive Multi-stage Density Ratio Estimation for Learning Latent Space Energy-based Model"], "answer_arxiv_id": ["2010.02917", "2209.08739"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_6189"} +{"question": "Could you provide me some works about simulated datasets that offer a wide range of environments?", "answer": ["CARLA: An Open Urban Driving Simulator", "TartanAir: A Dataset to Push the Limits of Visual SLAM"], "answer_arxiv_id": ["1711.03938", "2003.14338v2"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_6190"} +{"question": "What sources do discuss the usage of ensemble as a defense method?", "answer": ["Improving Adversarial Robustness via Promoting Ensemble Diversity", "DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensembles", "TRS: Transferability Reduced Ensemble via Promoting Gradient Diversity and Model Smoothness", "Ensemble Adversarial Training: Attacks and Defenses", "Self-ensemble Adversarial Training for Improved Robustness"], "answer_arxiv_id": ["1901.08846", "2009.14720", "2104.00671", "1705.07204", "2203.09678"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_6191"} +{"question": "Could you enumerate works on noise robust loss?", "answer": ["Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels", "Normalized Loss Functions for Deep Learning with Noisy Labels"], "answer_arxiv_id": ["1805.07836", "2006.13554"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_6192"} +{"question": "Which papers discuss the offline RL problem in the tabular setting?", "answer": ["Near-Optimal Provable Uniform Convergence in Offline Policy Evaluation for Reinforcement Learning", "Near-Optimal Offline Reinforcement Learning via Double Variance Reduction", "Towards Instance-Optimal Offline Reinforcement Learning with Pessimism", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning", "Pessimistic Q-Learning for Offline Reinforcement Learning: Towards Optimal Sample Complexity", "Settling the Sample Complexity of Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2007.03760", "2102.01748", "2110.08695v1", "2103.12021v2", "2106.04895", "2202.13890", "2204.05275"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6193"} +{"question": "Which papers contributed in the application of Domain Randomisation in multi-agent domains?", "answer": ["Emergent Tool Use From Multi-Agent Autocurricula", "SMACv2: An Improved Benchmark for Cooperative Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1909.07528", "2212.07489"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_6194"} +{"question": "Can you list the research papers that discuss RecExp, especially about its utility guarantees?", "answer": ["Differentially Private Approximate Quantiles"], "answer_arxiv_id": ["2110.05429"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_6195"} +{"question": "What research has contributed to the progress of interactive segmentation?", "answer": ["Deep Interactive Object Selection", "SimpleClick: Interactive Image Segmentation with Simple Vision Transformers", "FocalClick: Towards Practical Interactive Image Segmentation"], "answer_arxiv_id": ["1603.04042", "2210.11006", "2204.02574"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_6196"} +{"question": "Which studies distilled the volumetric representation of the scene into voxel grids to achieve real-time rendering speeds?", "answer": ["Neural Sparse Voxel Fields", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Baking Neural Radiance Fields for Real-Time View Synthesis"], "answer_arxiv_id": ["2007.11571", "2103.14024", "2103.14645"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_6197"} +{"question": "What are the recent DST models that have shown promising performance?", "answer": ["Dialogue State Tracking with a Language Model using Schema-Driven\n Prompting", "Continual Prompt Tuning for Dialog State Tracking", "Choice Fusion as Knowledge for Zero-Shot Dialogue State Tracking", "Diable: Efficient Dialogue State Tracking as Operations on Tables"], "answer_arxiv_id": ["2109.07506", "2203.06654", "2302.13013", "2305.17020"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_6198"} +{"question": "What research utilized differentiable rendering techniques to aid learning from 2D images?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image\n Synthesis", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature\n Fields", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis", "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images", "Texturify: Generating Textures on 3D Shape Surfaces", "DiffRF: Rendering-Guided 3D Radiance Field Diffusion", "3D generation on ImageNet", "Unsupervised Volumetric Animation", "DisCoScene: Spatially Disentangled Generative Radiance Fields for\n Controllable 3D-aware Scene Synthesis", "3DAvatarGAN: Bridging Domains for Personalized Editable Avatars"], "answer_arxiv_id": ["2112.07945", "2007.02442", "2110.08985", "2011.12100", "2012.00926", "2209.11163", "2204.02411", "2212.01206", "2303.01416", "2301.11326", "2212.11984", "2301.02700"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6199"} +{"question": "Which works proceed with high-order semantic mining within a set of images?", "answer": ["Group-Wise Semantic Mining for Weakly Supervised Semantic Segmentation", "Weakly Supervised Semantic Segmentation by Pixel-to-Prototype Contrast"], "answer_arxiv_id": ["2012.05007", "2110.07110"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_6200"} +{"question": "Which works proposed methods to evaluate images’ physical and geometric realism?", "answer": ["Shadows Don't Lie and Lines Can't Bend! Generative Models don't know\n Projective Geometry...for now"], "answer_arxiv_id": ["2311.17138"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_6201"} +{"question": "What are some works in the field of Class-Incremental Learning that propose solutions to maintain the model’s discriminability ability on old tasks by saving representative instances in old tasks and replaying them in new tasks?", "answer": ["Selective Experience Replay for Lifelong Learning", "Gradient based sample selection for online continual learning", "Experience Replay for Continual Learning"], "answer_arxiv_id": ["1802.10269", "1903.08671", "1811.11682"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_6202"} +{"question": "Are there any research reports that utilize causal inference to address the temporal out-of-distribution issue in sequential data?", "answer": ["Towards Out-of-Distribution Sequential Event Prediction: A Causal Treatment"], "answer_arxiv_id": ["2210.13005"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_6203"} +{"question": "Which paper first derived a generalization error bound for a typical graph embedding setting?", "answer": ["The Importance of Norm Regularization in Linear Graph Embedding: Theoretical Analysis and Empirical Demonstration"], "answer_arxiv_id": ["1802.03560"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_6204"} +{"question": "What papers emphasized the importance of context in visual tasks?", "answer": ["Context Understanding in Computer Vision: A Survey"], "answer_arxiv_id": ["2302.05011"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_6205"} +{"question": "Which papers highlighted the process of text-to-image prompt collection and analysis?", "answer": ["DiffusionDB: A Large-scale Prompt Gallery Dataset for Text-to-Image\n Generative Models", "A Prompt Log Analysis of Text-to-Image Generation Systems", "ImageReward: Learning and Evaluating Human Preferences for Text-to-Image\n Generation", "ArtWhisperer: A Dataset for Characterizing Human-AI Interactions in\n Artistic Creations"], "answer_arxiv_id": ["2210.14896", "2303.04587", "2304.05977", "2306.08141"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_6206"} +{"question": "Which studies propose that the adversarial robustness of a neural network is closely related to its Lipschitz continuity?", "answer": ["Intriguing properties of neural networks", "Parseval Networks: Improving Robustness to Adversarial Examples", "Lipschitz-Margin Training: Scalable Certification of Perturbation Invariance for Deep Neural Networks", "Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation"], "answer_arxiv_id": ["1312.6199", "1704.08847", "1802.04034", "1705.08475"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_6207"} +{"question": "What works develop self-supervised paradigm and vision-language multitask learning in GIQA?", "answer": ["Image Quality Assessment using Contrastive Learning", "Blind Image Quality Assessment via Vision-Language Correspondence: A\n Multitask Learning Perspective"], "answer_arxiv_id": ["2110.13266", "2303.14968"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_6208"} +{"question": "Which studies focused on the problem of aliasing artifacts in neural networks?", "answer": ["Making Convolutional Networks Shift-Invariant Again", "Delving Deeper into Anti-aliasing in ConvNets", "Anti-aliasing Deep Image Classifiers using Novel Depth Adaptive Blurring\n and Activation Function", "WaveCNet: Wavelet Integrated CNNs to Suppress Aliasing Effect for\n Noise-Robust Image Classification", "Impact of Aliasing on Generalization in Deep Convolutional Networks"], "answer_arxiv_id": ["1904.11486", "2008.09604", "2110.00899", "2107.13335", "2108.03489v1"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_6209"} +{"question": "Could you provide the reference that justifies other Bregman divergence losses, paralleling the c-Action Matching method?", "answer": ["Rectified Flow: A Marginal Preserving Approach to Optimal Transport"], "answer_arxiv_id": ["2209.14577v1"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_6210"} +{"question": "Could you provide a paper that proposed marginalising out parameters in the last layer of a neural network?", "answer": ["Last Layer Marginal Likelihood for Invariance Learning"], "answer_arxiv_id": ["2106.07512"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_6211"} +{"question": "Could you provide me some studies about generative modeling aimed at achieving feature disentanglement?", "answer": ["Stochastic Backpropagation and Approximate Inference in Deep Generative Models", "InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets"], "answer_arxiv_id": ["1401.4082", "1606.03657"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_6212"} +{"question": "Which studies have used attention mechanism in their research?", "answer": ["Attention Is All You Need", "Neural Machine Translation by Jointly Learning to Align and Translate", "Show, Attend and Tell: Neural Image Caption Generation with Visual Attention"], "answer_arxiv_id": ["1706.03762", "1409.0473", "1502.03044v3"], "source_meta": {"published_time": "20220129"}, "qid": "AutoScholarQuery_train_6213"} +{"question": "What studies have explored the sensitive nature of generated results to text quality within conditional diffusion generation?", "answer": ["A Picture is Worth a Thousand Words: Principled Recaptioning Improves\n Image Generation"], "answer_arxiv_id": ["2310.16656"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_6214"} +{"question": "Which paper demonstrated that a sparse neural network can match its corresponding dense neural network if its sparse connectivity is designed carefully?", "answer": ["A topological insight into restricted Boltzmann machines"], "answer_arxiv_id": ["1604.05978"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_6215"} +{"question": "Could you provide some works that utilized real-time rendering engines like Unity and Unreal to generate foggy images?", "answer": ["Virtual Worlds as Proxy for Multi-Object Tracking Analysis", "SHIFT: A Synthetic Driving Dataset for Continuous Multi-Task Domain\n Adaptation"], "answer_arxiv_id": ["1605.06457", "2206.08367"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_6216"} +{"question": "Could you name a single-loop algorithm that handles UL constraints for single-objective stochastic BLO problem?", "answer": ["A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic"], "answer_arxiv_id": ["2007.05170"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_6217"} +{"question": "Can you cite papers that leveraged a visual abstractor module to align the two-modalities?", "answer": ["mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2304.14178"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_6218"} +{"question": "Are there any studies that evaluated the ethical and societal impacts of image generation models?", "answer": ["Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scale", "Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models"], "answer_arxiv_id": ["2211.03759", "2211.05105"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_6219"} +{"question": "Is there any work that has encoded relations between words?", "answer": ["SeVeN: Augmenting Word Embeddings with Unsupervised Relation Vectors"], "answer_arxiv_id": ["1808.06068"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_6220"} +{"question": "Which studies performed weight updates across multiple layers to perform simultaneous edits?", "answer": ["Mass-Editing Memory in a Transformer"], "answer_arxiv_id": ["2210.07229"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_6221"} +{"question": "Who investigated applying a personalized language model for downstream tasks in stance classification and demographic inference?", "answer": ["Human Language Modeling"], "answer_arxiv_id": ["2205.05128"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_6222"} +{"question": "Can you list works that tackled representation collapse in non-IID clients?", "answer": ["Heterogeneity for the Win: One-Shot Federated Clustering", "Orchestra: Unsupervised Federated Learning via Globally Consistent\n Clustering", "FedX: Unsupervised Federated Learning with Cross Knowledge Distillation", "L-DAWA: Layer-wise Divergence Aware Weight Aggregation in Federated\n Self-Supervised Visual Representation Learning"], "answer_arxiv_id": ["2103.00697", "2205.11506", "2207.09158", "2307.07393"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_6223"} +{"question": "What works deal with domain-invariant feature learning in domain generalization?", "answer": ["Domain Generalization via Invariant Feature Representation", "Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization"], "answer_arxiv_id": ["1301.2115", "1510.04373v2"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_6224"} +{"question": "What works apply the experimental design approach to distinguish the optimal arm in the context of linear bandits?", "answer": ["Best-Arm Identification in Linear Bandits"], "answer_arxiv_id": ["1409.6110"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_6225"} +{"question": "Can you mention some studies that train vision-language pre-training models for specialist domains like medical, fashion, and remote sensing?", "answer": ["TieNet: Text-Image Embedding Network for Common Thorax Disease Classification and Reporting in Chest X-rays", "Contrastive Learning of Medical Visual Representations from Paired Images and Text", "Multimodal Representation Learning via Maximization of Local Mutual Information", "Making the Most of Text Semantics to Improve Biomedical Vision–Language Processing", "Joint Learning of Localized Representations from Medical Images and Reports", "MedCLIP: Contrastive Learning from Unpaired Medical Images and Text", "FashionBERT: Text and Image Matching with Adaptive Loss for Cross-modal Retrieval", "Kaleido-BERT: Vision-Language Pre-training on Fashion Domain", "FashionViL: Fashion-Focused Vision-and-Language Representation Learning"], "answer_arxiv_id": ["1801.04334", "2010.00747", "2103.04537", "2204.09817", "2112.02889", "2210.10163", "2005.09801v2", "2103.16110", "2207.08150"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_6226"} +{"question": "Could you provide me with studies that exemplify the widespread adoption of transformer models?", "answer": ["Training language models to follow instructions with human feedback", "GPT-4 Technical Report", "OpenAssistant Conversations - Democratizing Large Language Model Alignment"], "answer_arxiv_id": ["2203.02155", "2303.08774", "2304.07327"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6227"} +{"question": "Could you cite some papers that have devised solutions to the problem of conditional moment restrictions using machine learning models?", "answer": ["Deep Generalized Method of Moments for Instrumental Variable Analysis", "Minimax Estimation of Conditional Moment Models"], "answer_arxiv_id": ["1905.12495", "2006.07201v1"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_6228"} +{"question": "Which papers introduced Gaussian Splatting method that used non-isotropic 3D Gaussians with variable scale?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_6229"} +{"question": "What are the works related to tasks involving factual abilities?", "answer": ["LLMs as Factual Reasoners: Insights from Existing Benchmarks and Beyond"], "answer_arxiv_id": ["2305.14540"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_6230"} +{"question": "What references reported the use of multiple action proposal extraction methods for action detection?", "answer": ["R-C3D: Region Convolutional 3D Network for Temporal Activity Detection", "TURN TAP: Temporal Unit Regression Network for Temporal Action Proposals", "DCAN: Improving Temporal Action Detection via Dual Context Aggregation", "Learning Salient Boundary Feature for Anchor-free Temporal Action\n Localization", "An Efficient Spatio-Temporal Pyramid Transformer for Action Detection"], "answer_arxiv_id": ["1703.07814", "1703.06189", "2112.03612", "2103.13137", "2207.10448"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_6231"} +{"question": "Which papers are responsible for providing a comprehensive benchmark of two-stage and one-stage baseline methods?", "answer": ["Panoptic Scene Graph Generation"], "answer_arxiv_id": ["2207.11247"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_6232"} +{"question": "Could you provide me some works employing self-supervised learning schemes for learning with few labeled samples?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Big Self-Supervised Models are Strong Semi-Supervised Learners"], "answer_arxiv_id": ["2002.05709", "2006.10029"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_6233"} +{"question": "What papers are involved in efforts to modify the training data in order to affect the performance of models?", "answer": ["Generative Poisoning Attack Method Against Neural Networks", "Understanding Black-box Predictions via Influence Functions"], "answer_arxiv_id": ["1703.01340", "1703.04730"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_6234"} +{"question": "Which study has proposed a multi-task GP framework that directly outputs the potential outcomes?", "answer": ["Bayesian Inference of Individualized Treatment Effects using Multi-task Gaussian Processes"], "answer_arxiv_id": ["1704.02801"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_6235"} +{"question": "Which work developed a new pooling operator to learn from data in image-text matching?", "answer": ["Learning the Best Pooling Strategy for Visual Semantic Embedding"], "answer_arxiv_id": ["2011.04305"], "source_meta": {"published_time": "20240617"}, "qid": "AutoScholarQuery_train_6236"} +{"question": "Which papers proposed understanding model-based planning algorithms from value equivalence perspective?", "answer": ["The Value Equivalence Principle for Model-Based Reinforcement Learning", "Proper Value Equivalence"], "answer_arxiv_id": ["2011.03506", "2106.10316"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_6237"} +{"question": "Which works first introduced the notions of multi-group fairness?", "answer": ["Multiaccuracy: Black-Box Post-Processing for Fairness in Classification", "Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness"], "answer_arxiv_id": ["1805.12317", "1711.05144"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_6238"} +{"question": "What papers worked on integrating video foundation models and LLMs?", "answer": ["VideoChat: Chat-Centric Video Understanding"], "answer_arxiv_id": ["2305.06355"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_6239"} +{"question": "Any works have proposed GANs for achieving Perceptual Distortion (PD) trade-off in the generation of realistic images?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_6240"} +{"question": "Any works about enhancing consistency in Language Learning Models (LLMs) using special tailored tuning instructions?", "answer": ["MemoChat: Tuning LLMs to Use Memos for Consistent Long-Range Open-Domain\n Conversation"], "answer_arxiv_id": ["2308.08239"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_6241"} +{"question": "What works focused on studying object-centric representations?", "answer": ["Tagger: Deep Unsupervised Perceptual Grouping", "Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition", "Multi-Object Representation Learning with Iterative Variational Inference", "Object-Centric Learning with Slot Attention", "Illiterate DALL-E Learns to Compose", "Bridging the Gap to Real-World Object-Centric Learning"], "answer_arxiv_id": ["1606.06724", "1603.08575", "2001.02407", "1903.00450", "2006.15055", "2110.11405", "2209.14860v2"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_6242"} +{"question": "Which paper primarily used the second approach that presents a deep SCM using normalizing flows and variational inference for efficient and precise training even in high-dimensional cases?", "answer": ["Deep Structural Causal Models for Tractable Counterfactual Inference"], "answer_arxiv_id": ["2006.06485"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_6243"} +{"question": "Which paper uses LLMs in search systems to generate query expansion terms?", "answer": ["Query Expansion by Prompting Large Language Models"], "answer_arxiv_id": ["2305.03653"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_6244"} +{"question": "What work introduces a transformer architecture for addressing AVS task?", "answer": ["AVSegFormer: Audio-Visual Segmentation with Transformer"], "answer_arxiv_id": ["2307.01146"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_6245"} +{"question": "What are some initial studies explored random token dropping?", "answer": ["VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text", "Scaling Language-Image Pre-training via Masking", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2104.11178", "2212.00794", "2111.06377"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_6246"} +{"question": "Which papers introduced Vision-and-Language Navigation (VLN) tasks?", "answer": ["Vision-and-Language Navigation: Interpreting visually-grounded\n navigation instructions in real environments", "REVERIE: Remote Embodied Visual Referring Expression in Real Indoor\n Environments", "Room-Across-Room: Multilingual Vision-and-Language Navigation with Dense\n Spatiotemporal Grounding", "Beyond the Nav-Graph: Vision-and-Language Navigation in Continuous\n Environments"], "answer_arxiv_id": ["1711.07280", "1904.10151", "2010.07954", "2004.02857"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_6247"} +{"question": "Who studied the presence of similar human biases in the large language models (LLMs) like ordering effects in syllogistic tasks?", "answer": ["Language models show human-like content effects on reasoning tasks", "A Systematic Comparison of Syllogistic Reasoning in Humans and Language\n Models", "Using cognitive psychology to understand GPT-3", "Challenging the appearance of machine intelligence: Cognitive bias in LLMs and Best Practices for Adoption", "Cognitive Effects in Large Language Models"], "answer_arxiv_id": ["2207.07051", "2311.00445", "2206.14576", "2304.01358v3", "2308.14337"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_6248"} +{"question": "What is the paper that proposed dropping edges in training to improve training efficiency for GNNs?", "answer": ["DropEdge: Towards Deep Graph Convolutional Networks on Node Classification"], "answer_arxiv_id": ["1907.10903"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_6249"} +{"question": "Which papers discuss methods like ControlNet, T2I-Adapter, and Composer that finetune SD with spatial condition for text-to-image synthesis?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Composer: Creative and Controllable Image Synthesis with Composable\n Conditions"], "answer_arxiv_id": ["2302.05543", "2302.08453", "2302.09778"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_6250"} +{"question": "What papers presented the application of RLHF for instruction-following?", "answer": ["The Wisdom of Hindsight Makes Language Models Better Instruction Followers"], "answer_arxiv_id": ["2302.05206"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_6251"} +{"question": "Could you provide me research that focuses on synthetic cartoon datasets for story visualization?", "answer": ["DeepStory: Video Story QA by Deep Embedded Memory Networks", "Imagine This! Scripts to Compositions to Videos"], "answer_arxiv_id": ["1707.00836v1", "1804.03608"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_6252"} +{"question": "Which works develop multimodal models using more than two modalities for improving the representations of one modality?", "answer": ["Everything at Once -- Multi-modal Fusion Transformer for Video Retrieval", "Self-Supervised MultiModal Versatile Networks", "VATT: Transformers for Multimodal Self-Supervised Learning from Raw\n Video, Audio and Text"], "answer_arxiv_id": ["2112.04446", "2006.16228", "2104.11178"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_6253"} +{"question": "Which research papers have refined and expanded the concept of RPE showcasing its efficiency in natural language processing?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Improve Transformer Models with Better Relative Position Embeddings"], "answer_arxiv_id": ["1901.02860", "2009.13658"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_6254"} +{"question": "What are the recent notable works that focus on finding the groups when group information is not available during training?", "answer": ["Just Train Twice: Improving Group Robustness without Training Group Information", "Environment Inference for Invariant Learning", "Correct-n-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlations", "Identifying Spurious Biases Early in Training through the Lens of Simplicity Bias"], "answer_arxiv_id": ["2107.09044", "2010.07249", "2203.01517", "2305.18761"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_6255"} +{"question": "Which papers assume that the latent variables are conditionally independent given some observed variable?", "answer": ["Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learning", "Variational Autoencoders and Nonlinear ICA: A Unifying Framework"], "answer_arxiv_id": ["1805.08651", "1907.04809"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_6256"} +{"question": "Which study can be seen as a generalization of spectral clustering?", "answer": ["A Tutorial on Spectral Clustering"], "answer_arxiv_id": ["0711.0189"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_6257"} +{"question": "Are there any studies which apply exponential-time spectral algorithm and quasipolynomial-time tensor-based algorithm for density estimation even when centers are arbitrarily closely spaced?", "answer": ["Near-optimal-sample estimators for spherical Gaussian mixtures", "Small Covers for Near-Zero Sets of Polynomials and Learning Latent Variable Models"], "answer_arxiv_id": ["1402.4746", "2012.07774v1"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_6258"} +{"question": "Which research work presents tensor-ring decomposition which assumes a periodic structure on tensor train decomposition?", "answer": ["Tensor Ring Decomposition"], "answer_arxiv_id": ["1606.05535"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_6259"} +{"question": "Which studies have shown significant improvement in terms of text-to-3D quality due to training on large-scale datasets like Objaverse?", "answer": ["Wonder3D: Single Image to 3D using Cross-Domain Diffusion", "Objaverse: A Universe of Annotated 3D Objects", "Objaverse-XL: A Universe of 10M+ 3D Objects", "Instant3D: Fast Text-to-3D with Sparse-View Generation and Large\n Reconstruction Model"], "answer_arxiv_id": ["2310.15008", "2212.08051", "2307.05663", "2311.06214"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_6260"} +{"question": "Which work derived bounds on the spectrum of the kernels of deep CNNs?", "answer": ["On the Spectral Bias of Convolutional Neural Tangent and Gaussian Process Kernels"], "answer_arxiv_id": ["2203.09255"], "source_meta": {"published_time": "20220801"}, "qid": "AutoScholarQuery_train_6261"} +{"question": "What work, being purely online, shows that online guarantees can be obtained in terms of the concentrability coefficient parameter introduced in the offline RL literature?", "answer": ["The Role of Coverage in Online Reinforcement Learning"], "answer_arxiv_id": ["2210.04157"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_6262"} +{"question": "Which works utilize weight averaging for combining the properties of diverse networks in the line of 'model soups'?", "answer": ["Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time", "Diverse Weight Averaging for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2203.05482", "2205.09739"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_6263"} +{"question": "What works introduce the initial YOLOs for single-stage object detection?", "answer": ["You Only Look Once: Unified, Real-Time Object Detection", "YOLO9000: Better, Faster, Stronger", "YOLOv3: An Incremental Improvement"], "answer_arxiv_id": ["1506.02640", "1612.08242", "1804.02767"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_6264"} +{"question": "Which studies provide methods about thinning in the context of selecting quadrature nodes?", "answer": ["Kernel Thinning", "Generalized Kernel Thinning", "Distribution Compression in Near-linear Time"], "answer_arxiv_id": ["2105.05842", "2110.01593", "2111.07941"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6265"} +{"question": "Which papers are foundational in the field of large-scale language models and implementation of attention in Transformers?", "answer": ["Neural Machine Translation by Jointly Learning to Align and Translate", "Attention Is All You Need"], "answer_arxiv_id": ["1409.0473", "1706.03762"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_6266"} +{"question": "What studies proposed models using subquadratic alterations of standard attention?", "answer": ["Non-Local Graph Neural Networks"], "answer_arxiv_id": ["2005.14612"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_6267"} +{"question": "In what papers do they autonomously generate two-channel spoken dialogues?", "answer": ["Generative Spoken Dialogue Language Modeling", "Towards human-like spoken dialogue generation between AI agents from\n written dialogue"], "answer_arxiv_id": ["2203.16502", "2310.01088"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_6268"} +{"question": "Could you provide me some works that estimate the loss for uncertainty assessment in Active Learning?", "answer": ["Learning Loss for Active Learning", "Semi-Supervised Active Learning with Temporal Output Discrepancy"], "answer_arxiv_id": ["1905.03677", "2107.14153"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_6269"} +{"question": "What TTA studies constructed self-supervised auxiliary tasks to learn on test data?", "answer": ["Tent: Fully Test-time Adaptation by Entropy Minimization", "Test-Time Training with Self-Supervision for Generalization under\n Distribution Shifts", "MEMO: Test Time Robustness via Adaptation and Augmentation", "Test-Time Adaptation to Distribution Shift by Confidence Maximization\n and Input Transformation", "Contrastive Test-Time Adaptation", "TeST: Test-time Self-Training under Distribution Shift", "Test-Time Training with Masked Autoencoders"], "answer_arxiv_id": ["2006.10726", "1909.13231", "2110.09506", "2106.14999", "2204.10377", "2209.11459", "2209.07522"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_6270"} +{"question": "In which work is a text-to-video generation model trained on web-scale datasets and expert demonstrations to generate image sequences for planning and inverse modeling?", "answer": ["Replacing Rewards with Examples: Example-Based Policy Search via Recursive Classification"], "answer_arxiv_id": ["2103.12656"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_6271"} +{"question": "Could you provide me with studies that have observed noise interference in retrieved texts to negatively affect performance?", "answer": ["Self-Knowledge Guided Retrieval Augmentation for Large Language Models", "Benchmarking Large Language Models in Retrieval-Augmented Generation", "List-aware Reranking-Truncation Joint Model for Search and\n Retrieval-augmented Generation"], "answer_arxiv_id": ["2310.05002", "2309.01431", "2402.02764"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_6272"} +{"question": "What papers fully determined random feature ridge regression by the limiting spectra of CK or NTK?", "answer": ["A Random Matrix Approach to Neural Networks", "Generalisation error in learning with random features and the hidden manifold model", "A Random Matrix Analysis of Random Fourier Features: Beyond the Gaussian Kernel, a Precise Phase Transition, and the Corresponding Double Descent", "The Neural Tangent Kernel in High Dimensions: Triple Descent and a Multi-Scale Theory of Generalization"], "answer_arxiv_id": ["1702.05419", "2002.09339", "2006.05013", "2008.06786"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_6273"} +{"question": "Could you provide me works about explicit approximations of proximal point method for convex-concave minimax problems?", "answer": ["Min-Max Optimization Made Simple: Approximating the Proximal Point Method via Contraction Maps"], "answer_arxiv_id": ["2301.03931"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_6274"} +{"question": "Which studies have found serious flaws in CoT explanations?", "answer": ["Maieutic Prompting: Logically Consistent Reasoning with Recursive Explanations", "The Unreliability of Explanations in Few-shot Prompting for Textual Reasoning", "ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoning"], "answer_arxiv_id": ["2205.11822", "2205.03401", "2212.07919"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_6275"} +{"question": "What research propose a pioneer dataset for generating explanatory paragraphs in response to open-ended questions?", "answer": ["ELI5: Long Form Question Answering"], "answer_arxiv_id": ["1907.09190"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_6276"} +{"question": "Are there any techniques that employ the marching cubes (MC) algorithm to extract the zero isosurface of the SDF?", "answer": ["OctField: Hierarchical Implicit Functions for 3D Modeling", "Deep Local Shapes: Learning Local SDF Priors for Detailed 3D\n Reconstruction", "Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D\n Shapes", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2111.01067", "2003.10983", "2101.10994", "2201.05989"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_6277"} +{"question": "What are some works that demonstrate achieving polynomial complexity of learning an optimal policy using effective exploration methods?", "answer": ["Minimax Regret Bounds for Reinforcement Learning", "Provably efficient RL with Rich Observations via Latent State Decoding", "Kinematic State Abstraction and Provably Efficient Rich-Observation Reinforcement Learning", "PC-PG: Policy Cover Directed Exploration for Provable Policy Gradient Learning", "Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition"], "answer_arxiv_id": ["1703.05449", "1901.09018", "1911.05815", "2007.08459", "2004.10019"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_6278"} +{"question": "Could you give examples of research comparing human and machine perceptual similarity?", "answer": ["The Unreasonable Effectiveness of Deep Features as a Perceptual Metric", "Eigen-Distortions of Hierarchical Representations", "Towards a Semantic Perceptual Image Metric", "Do better ImageNet classifiers assess perceptual similarity better?"], "answer_arxiv_id": ["1801.03924", "1710.02266", "1808.00447", "2203.04946"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_6279"} +{"question": "What works propose soft prompting as an alternative method to traditional discrete prompts?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Learning How to Ask: Querying LMs with Mixtures of Soft Prompts", "SPoT: Better Frozen Model Adaptation through Soft Prompt Transfer", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2104.08691", "2101.00190", "2104.06599", "2110.07904", "2109.01134"], "source_meta": {"published_time": "20220407"}, "qid": "AutoScholarQuery_train_6280"} +{"question": "Which works proposed traditional methods in 3D representation using meshes?", "answer": ["Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images", "AtlasNet: A Papier-M\\^ach\\'e Approach to Learning 3D Surface Generation", "A Skeleton-bridged Deep Learning Approach for Generating Meshes of\n Complex Topologies from Single RGB Images", "Deep Mesh Reconstruction from Single RGB Images via Topology Modification Networks", "SkeletonNet: A Topology-Preserving Solution for Learning Mesh\n Reconstruction of Object Surfaces from RGB Images"], "answer_arxiv_id": ["1804.01654", "1802.05384", "1903.04704", "1909.00321v1", "2008.05742"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_6281"} +{"question": "Can you list the works that introduced and developed the diffusion model?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20230423"}, "qid": "AutoScholarQuery_train_6282"} +{"question": "Which papers discuss the role of gradient information in generating adversarial examples?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Evaluating the Robustness of Neural Networks", "Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Theoretically Principled Trade-off between Robustness and Accuracy", "Adversarial examples in the physical world", "Adversarial Attacks and Defenses in Images, Graphs and Text: A Review", "Structured Adversarial Attack: Towards General Implementation and Better Interpretability"], "answer_arxiv_id": ["1412.6572", "1608.04644", "2003.01690", "1706.06083", "1901.08573", "1607.02533", "1909.08072", "1808.01664"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_6283"} +{"question": "What studies have been conducted to generate image synthesis from user-scribble based semantic segmentation maps?", "answer": ["Semantic Image Synthesis with Spatially-Adaptive Normalization", "MaskGAN: Towards Diverse and Interactive Facial Image Manipulation", "Taming Transformers for High-Resolution Image Synthesis", "Image-to-Image Translation with Conditional Adversarial Networks", "Learning to Predict Layout-to-image Conditional Convolutions for\n Semantic Image Synthesis", "You Only Need Adversarial Supervision for Semantic Image Synthesis"], "answer_arxiv_id": ["1903.07291", "1907.11922", "2012.09841", "1611.07004", "1910.06809", "2012.04781"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_6284"} +{"question": "Could you provide me some works using Neural Architecture Search strategies?", "answer": ["Large-Scale Evolution of Image Classifiers", "Regularized Evolution for Image Classifier Architecture Search", "Neural Architecture Search with Reinforcement Learning", "Learning Deep Generative Models of Graphs", "Random Search and Reproducibility for Neural Architecture Search", "Exploring the Loss Landscape in Neural Architecture Search", "Neural Architecture Search with Bayesian Optimisation and Optimal Transport", "Interpretable Neural Architecture Search via Bayesian Optimisation with Weisfeiler-Lehman Kernels", "BANANAS: Bayesian Optimization with Neural Architectures for Neural Architecture Search", "Smooth Variational Graph Embeddings for Efficient Neural Architecture Search", "Generative Adversarial Neural Architecture Search", "Learning Where To Look – Generative NAS is Surprisingly Efficient", "Efficient Neural Architecture Search via Parameter Sharing", "DARTS: Differentiable Architecture Search", "ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware", "SNAS: stochastic neural architecture search", "Understanding and Robustifying Differentiable Architecture Search"], "answer_arxiv_id": ["1703.01041", "1802.01548", "1611.01578", "1803.03324", "1902.07638", "2005.02960", "1802.07191", "2006.07556", "1910.11858", "2010.04683v3", "2105.09356", "2203.08734", "1802.03268", "1806.09055", "1812.00332", "1812.09926", "1909.09656"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_6285"} +{"question": "Could you provide me some studies about the application of gradient integration method in multi-task learning?", "answer": ["Gradient Surgery for Multi-Task Learning", "Conflict-Averse Gradient Descent for Multi-task Learning", "Multi-Task Learning as a Bargaining Game"], "answer_arxiv_id": ["2001.06782", "2110.14048", "2202.01017"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_6286"} +{"question": "Which studies focused on designing optimization methods for neural networks that are invariant to scaling symmetries at individual neurons?", "answer": ["Path-SGD: Path-Normalized Optimization in Deep Neural Networks", "Symmetry-invariant optimization in deep networks"], "answer_arxiv_id": ["1506.02617", "1511.01754"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_6287"} +{"question": "What research papers belonged to the loss regularization category, where a regularizer is added to the loss function?", "answer": ["Calibrating Deep Neural Networks using Focal Loss"], "answer_arxiv_id": ["2002.09437"], "source_meta": {"published_time": "20220215"}, "qid": "AutoScholarQuery_train_6288"} +{"question": "Which studies have approached automatic matting by predicting the alpha matte without any user intervention?", "answer": ["Deep Automatic Natural Image Matting", "Semantic Human Matting", "Bridging Composite and Real: Towards End-to-end Deep Image Matting"], "answer_arxiv_id": ["2107.07235", "1809.01354", "2010.16188"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_6289"} +{"question": "Who introduced an alternative aggregation procedure where Model-X Knockoffs are viewed as an e-BH procedure?", "answer": ["Derandomized knockoffs: leveraging e-values for false discovery rate control"], "answer_arxiv_id": ["2205.15461"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_6290"} +{"question": "Who introduced efficient algorithm OLM for the linear bandit problem?", "answer": ["Online Stochastic Linear Optimization under One-bit Feedback"], "answer_arxiv_id": ["1509.07728"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_6291"} +{"question": "What work proposed a method of post-processing SMPLify results by modeling body parts as ellipsoids to overcome self-intersections?", "answer": ["Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a\n Single Image"], "answer_arxiv_id": ["1607.08128"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_6292"} +{"question": "Which works developed the generative langiage models, such as GPT?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_6293"} +{"question": "Could you provide me some studies about the 2-time scale gradient descent ascent?", "answer": ["GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium", "Generative Adversarial Nets", "Unrolled Generative Adversarial Networks"], "answer_arxiv_id": ["1706.08500", "1406.2661", "1611.02163"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_6294"} +{"question": "What studies explored mechanism design in large games along with maintaining privacy?", "answer": ["Mechanism Design in Large Games: Incentives and Privacy", "Efficient Nash Computation in Large Population Games with Bounded Influence", "Robust Mediators in Large Games"], "answer_arxiv_id": ["1207.4084", "1301.0577v1", "1512.02698v2"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_6295"} +{"question": "Can you name the studies that perform premises selection and reasoning in a step-by-step manner?", "answer": ["Entailment Tree Explanations via Iterative Retrieval-Generation Reasoner", "METGEN: A Module-Based Entailment Tree Generation Framework for Answer\n Explanation", "Faithful Question Answering with Monte-Carlo Planning"], "answer_arxiv_id": ["2205.09224", "2205.02593", "2305.02556"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_6296"} +{"question": "What work has been done on calculating confidence scores as a point-wise uncertainty measure?", "answer": ["PointRend: Image Segmentation as Rendering"], "answer_arxiv_id": ["1912.08193"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_6297"} +{"question": "Which foundation models for image and text representation learning are pre-trained by contrastive objectives?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2205.01917", "2204.14198"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_6298"} +{"question": "Which works used Graph Neural Networks (GNNs) in modeling structured dynamical systems?", "answer": ["Learning to Simulate Complex Physics with Graph Networks", "Learning Continuous System Dynamics from Irregularly-Sampled Partial Observations"], "answer_arxiv_id": ["2002.09405", "2011.03880"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_6299"} +{"question": "Could you provide me with the reference of the previous work that aligns with IOMI quantity?", "answer": ["Tightening Mutual Information Based Bounds on Generalization Error"], "answer_arxiv_id": ["1901.04609"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_6300"} +{"question": "Who first introduced the term 'foundation model'?", "answer": ["On the Opportunities and Risks of Foundation Models"], "answer_arxiv_id": ["2108.07258v3"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6301"} +{"question": "Are there any papers that proposed to reduce gradient interference between tasks?", "answer": ["Gradient Surgery for Multi-Task Learning"], "answer_arxiv_id": ["2001.06782"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6302"} +{"question": "Which works further developed practices of diffusion models for image generation applications?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_6303"} +{"question": "Which works discuss about a larger amount of noise injection during the initial phase of learning process?", "answer": ["Parametric Noise Injection: Trainable Randomness to Improve Deep Neural Network Robustness against Adversarial Attack", "Noise Optimization for Artificial Neural Networks"], "answer_arxiv_id": ["1811.09310", "2102.04450"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_6304"} +{"question": "What papers discussed 4D Gaussian Splatting (4DGS)?", "answer": ["4D Gaussian Splatting for Real-Time Dynamic Scene Rendering", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis"], "answer_arxiv_id": ["2310.08528", "2308.09713"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_6305"} +{"question": "Could you point out some research papers that examine text-conditioned visual generation through autoregressive or diffusion models?", "answer": ["Muse: Text-To-Image Generation via Masked Generative Transformers", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["2301.00704", "2206.10789", "2205.11487", "2112.10752", "2211.01324", "2204.06125", "2112.10741", "2111.14822", "2203.13131"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6306"} +{"question": "Which works studied the production of real world fluid datasets through complex velocimetry devices?", "answer": ["Shallow Neural Networks for Fluid Flow Reconstruction with Limited Sensors"], "answer_arxiv_id": ["1902.07358"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_6307"} +{"question": "Which studies aim to accelerate a scene-specific NeRF?", "answer": ["PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Plenoxels: Radiance Fields without Neural Networks", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2103.14024", "2112.05131", "2201.05989"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6308"} +{"question": "In which studies is a persistent memory used?", "answer": ["Memory-assisted prompt editing to improve GPT-3 after deployment"], "answer_arxiv_id": ["2201.06009"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_6309"} +{"question": "Could you provide me some works related to P-FL algorithms that encourage the angle collapse of the local classifier?", "answer": ["Exploiting Shared Representations for Personalized Federated Learning", "FedProto: Federated Prototype Learning across Heterogeneous Clients", "FedBABU: Toward Enhanced Representation for Federated Image Classification"], "answer_arxiv_id": ["2102.07078", "2105.00243", "2106.06042"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_6310"} +{"question": "Which works pretrain 3D autoencoders on object-level point clouds using features from a 2D backbone?", "answer": ["Learning 3D Representations from 2D Pre-trained Models via\n Image-to-Point Masked Autoencoders", "Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image\n Transformers Help 3D Representation Learning?"], "answer_arxiv_id": ["2212.06785", "2212.08320"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_6311"} +{"question": "Which works are about using a NeRF-based parameteric model like HeadNeRF and MofaNeRF?", "answer": ["HeadNeRF: A Real-time NeRF-based Parametric Head Model", "MoFaNeRF: Morphable Facial Neural Radiance Field"], "answer_arxiv_id": ["2112.05637", "2112.02308"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6312"} +{"question": "Which papers focused on the capabilities of large multi-modal models in analyzing and synthesizing images?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Florence: A New Foundation Model for Computer Vision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Flamingo: a Visual Language Model for Few-Shot Learning", "PaLI: A Jointly-Scaled Multilingual Language-Image Model"], "answer_arxiv_id": ["2103.00020", "2111.11432", "2102.05918", "2305.06500", "2204.14198", "2209.06794"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_6313"} +{"question": "What papers aim to learn succinct representations that both factorize the data space efficiently and are robust towards distributional changes?", "answer": ["On the Fairness of Disentangled Representations", "Representation Learning: A Review and New Perspectives", "Are Disentangled Representations Helpful for Abstract Visual Reasoning?"], "answer_arxiv_id": ["1905.13662", "1206.5538", "1905.12506"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_6314"} +{"question": "Which papers have combined the ideas from the Variational Information Bottleneck and Vector Quantization?", "answer": ["ProtoVAE: A Trustworthy Self-Explainable Prototypical Variational Model", "Representation Learning in Deep RL via Discrete Information Bottleneck"], "answer_arxiv_id": ["2210.08151", "2212.13835"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_6315"} +{"question": "Which works combined image augmentation with contrastive learning to achieve better performance in reinforcement learning?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning"], "answer_arxiv_id": ["2004.04136"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_6316"} +{"question": "Were there any papers that pointed out the possibility of obtaining convergence guarantees of Adam optimization algorithm for problem-dependent hyper-parameters?", "answer": ["Adam Can Converge Without Any Modification On Update Rules"], "answer_arxiv_id": ["2208.09632v5"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_6317"} +{"question": "Can you cite a study that used other non-compact metrics to artificially choose a compact subset of graphs from the whole space?", "answer": ["Weisfeiler-Lehman meets Gromov-Wasserstein"], "answer_arxiv_id": ["2202.02495"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6318"} +{"question": "Which datasets are examples of 3D synthetic dataset that consist of category-level objects?", "answer": ["ShapeNet: An Information-Rich 3D Model Repository", "3D ShapeNets: A Deep Representation for Volumetric Shapes"], "answer_arxiv_id": ["1512.03012", "1406.5670"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_6319"} +{"question": "What studies are about instance segmentation?", "answer": ["Mask R-CNN", "Instance-aware Semantic Segmentation via Multi-task Network Cascades", "SOLO: Segmenting Objects by Locations"], "answer_arxiv_id": ["1703.06870", "1512.04412", "1912.04488"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6320"} +{"question": "What papers focus on multimodal contrastive learning through optimizing a contrastive objective for inter-modality pairs?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Self-Supervised MultiModal Versatile Networks", "VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2006.16228", "2104.11178", "2102.05918"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_6321"} +{"question": "Which papers have made observations about the two-stage process in gradient descent?", "answer": ["Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability"], "answer_arxiv_id": ["2103.00065"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_6322"} +{"question": "Which strategies for combating noisy labels are built upon semi-supervised learning methods or self-supervised learning?", "answer": ["DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "SELF: Learning to Filter Noisy Labels with Self-Ensembling", "Selective-Supervised Contrastive Learning with Noisy Labels"], "answer_arxiv_id": ["2002.07394", "1910.01842", "2203.04181"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_6323"} +{"question": "Could you provide me with the research papers which have engineered prompts for obtaining accurate and relevant outputs?", "answer": ["Prompt Programming for Large Language Models: Beyond the Few-Shot Paradigm"], "answer_arxiv_id": ["2102.07350"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_6324"} +{"question": "Which works use context as side information in the field of contextual bandits?", "answer": ["A Contextual-Bandit Approach to Personalized News Article Recommendation", "Thompson Sampling for Contextual Bandits with Linear Payoffs", "Provably Optimal Algorithms for Generalized Linear Contextual Bandits"], "answer_arxiv_id": ["1003.0146", "1209.3352", "1703.00048"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_6325"} +{"question": "Could you provide me with researches based on learnable bases in forecasting models?", "answer": ["Are Transformers Effective for Time Series Forecasting?", "An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling", "DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks", "Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecasting", "Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting", "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting", "FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting"], "answer_arxiv_id": ["2205.13504", "1803.01271", "1704.04110", "1907.00235", "2012.07436", "2106.13008", "2201.12740"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_6326"} +{"question": "Can you list some research works that provided the generalization performance after machine unlearning?", "answer": ["Remember What You Want to Forget: Algorithms for Machine Unlearning", "Algorithms that Approximate Data Removal: New Results and Limitations"], "answer_arxiv_id": ["2103.03279v2", "2209.12269"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_6327"} +{"question": "Which studies have used trichromatic Stokes images for seeing through scattering?", "answer": ["Polarimetric Spatio-Temporal Light Transport Probing"], "answer_arxiv_id": ["2105.11609"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_6328"} +{"question": "Can you provide references that have researched capturing consistent semantics?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Emerging Properties in Self-Supervised Vision Transformers", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Dense Contrastive Learning for Self-Supervised Visual Pre-Training", "GroupViT: Semantic Segmentation Emerges from Text Supervision"], "answer_arxiv_id": ["2103.00020", "2104.14294", "2002.05709", "1911.05722", "2011.09157", "2202.11094"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6329"} +{"question": "Which works are about feed-forward methods for 3D diffusion models?", "answer": ["Shap-E: Generating Conditional 3D Implicit Functions", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization", "Zero-1-to-3: Zero-shot One Image to 3D Object", "SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse\n Views"], "answer_arxiv_id": ["2305.02463", "2212.08751", "2306.16928", "2303.11328", "2206.05737"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_6330"} +{"question": "Which research expressed the nearest point on the surface as a function of the neural signed distance and its gradient?", "answer": ["Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces"], "answer_arxiv_id": ["2011.13495"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_6331"} +{"question": "Are there any works that propose methodologies for time-series specific or tabular data embeddings?", "answer": ["Revisiting Deep Learning Models for Tabular Data"], "answer_arxiv_id": ["2106.11959"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_6332"} +{"question": "Which works introduced group distributionally robust optimization (DRO) method?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_6333"} +{"question": "Which projects are examples of zero-shot visual grounding methods?", "answer": ["ReCLIP: A Strong Zero-Shot Baseline for Referring Expression\n Comprehension", "Adapting CLIP For Phrase Localization Without Further Training", "VGDiffZero: Text-to-image Diffusion Models Can Be Zero-shot Visual\n Grounders"], "answer_arxiv_id": ["2204.05991", "2204.03647", "2309.01141"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6334"} +{"question": "What studies are about deep hashing approaches for encoding design?", "answer": ["A Survey on Deep Hashing Methods", "A Survey on Learning to Hash", "Deep Ordinal Hashing with Spatial Attention"], "answer_arxiv_id": ["2003.03369", "1606.00185", "1805.02459"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_6335"} +{"question": "Could you list some researches on offline low rank matrix completion?", "answer": ["Restricted strong convexity and weighted matrix completion: Optimal bounds with noise", "Noisy Matrix Completion: Understanding Statistical Guarantees for Convex Relaxation via Nonconvex Optimization", "Linear Bandits in High Dimension and Recommendation Systems", "Entrywise Eigenvector Analysis of Random Matrices with Low Expected Rank", "Low-rank Matrix Completion using Alternating Minimization", "Non-convex Optimization for Machine Learning"], "answer_arxiv_id": ["1009.2118v2", "1902.07698v2", "1301.1722", "1709.09565", "1212.0467", "1712.07897"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_6336"} +{"question": "What works have shown that hierarchical classification paradigm can aid in zero-shot learning by attempting to uncover hierarchical relations between classes?", "answer": ["HSVA: Hierarchical Semantic-Visual Adaptation for Zero-Shot Learning"], "answer_arxiv_id": ["2109.15163"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6337"} +{"question": "Could you provide me some works about specialized datasets and benchmarks in later Federated Learning research?", "answer": ["FLAIR: Federated Learning Annotated Image Repository", "FLamby: Datasets and Benchmarks for Cross-Silo Federated Learning in Realistic Healthcare Settings", "Motley: Benchmarking Heterogeneity and Personalization in Federated Learning", "pFL-Bench: A Comprehensive Benchmark for Personalized Federated Learning"], "answer_arxiv_id": ["2207.08869", "2210.04620", "2206.09262", "2206.03655"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_6338"} +{"question": "Which research works focused on the program induction for improving generalization?", "answer": ["Neural Turing Machines", "Reinforcement Learning Neural Turing Machines - Revised", "Neural Programmer-Interpreters", "Neural Logic Machines"], "answer_arxiv_id": ["1410.5401", "1505.00521", "1511.06279", "1904.11694"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_6339"} +{"question": "Could you mention studies that have combined probabilistic inferences with a bivariate Gaussian distribution for multi-modal trajectory generation?", "answer": ["Pedestrian Prediction by Planning using Deep Neural Networks", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for\n Behavior Prediction", "Social-STGCNN: A Social Spatio-Temporal Graph Convolutional Neural\n Network for Human Trajectory Prediction", "Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory\n Prediction", "EvolveGraph: Multi-Agent Trajectory Prediction with Dynamic Relational\n Reasoning", "BiTraP: Bi-directional Pedestrian Trajectory Prediction with Multi-modal\n Goal Estimation", "SGCN:Sparse Graph Convolution Network for Pedestrian Trajectory\n Prediction", "Adaptive Trajectory Prediction via Transferable GNN", "Social-Implicit: Rethinking Trajectory Prediction Evaluation and The\n Effectiveness of Implicit Maximum Likelihood Estimation"], "answer_arxiv_id": ["1706.05904", "1910.05449", "2002.11927", "2005.08514", "2003.13924", "2007.14558", "2104.01528", "2203.05046", "2203.03057"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_6340"} +{"question": "What methods have been utilized thus far to estimate propensity in a heuristic way?", "answer": ["Unbiased Learning for the Causal Effect of Recommendation"], "answer_arxiv_id": ["2008.04563"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_6341"} +{"question": "What research has learned hierarchical representation of images in hyperbolic space?", "answer": ["Hyperbolic Image Embeddings", "Hyperbolic Image Segmentation", "Hyperbolic Vision Transformers: Combining Improvements in Metric\n Learning"], "answer_arxiv_id": ["1904.02239", "2203.05898", "2203.10833"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_6342"} +{"question": "Could you provide me papers that apply pre-trained text-to-image diffusion models for controllable image generation?", "answer": ["T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.08453", "2302.05543"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_6343"} +{"question": "Which works introduced the vision foundation models such as SAM or DINOv2?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2401.14159", "2304.07193"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_6344"} +{"question": "Which studies first described the notion that llms might be performing implicit Bayesian inference?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_6345"} +{"question": "Could you provide some works about employing synthetic data in the field of action recognition?", "answer": ["ElderSim: A Synthetic Data Generation Platform for Human Action Recognition in Eldercare Applications", "Synthetic Humans for Action Recognition from Unseen Viewpoints"], "answer_arxiv_id": ["2010.14742", "1912.04070"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_6346"} +{"question": "What are some of the textual inversion-based methods that have been applied in stylized image generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Inversion-Based Style Transfer with Diffusion Models"], "answer_arxiv_id": ["2208.01618", "2211.13203"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6347"} +{"question": "What papers are about achieving explainability in Computer Vision by mapping relevance maps to a pixel level?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1610.02391", "1703.01365"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_6348"} +{"question": "Which works proposed methods for evaluating generated images using models trained on human preference data?", "answer": ["Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation", "Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesis", "ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation"], "answer_arxiv_id": ["2305.01569", "2306.09341", "2304.05977"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_6349"} +{"question": "Could you provide me the work that uses a diffusion model for generating volumetric radiance fields?", "answer": ["DiffRF: Rendering-Guided 3D Radiance Field Diffusion"], "answer_arxiv_id": ["2212.01206"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_6350"} +{"question": "What studies inspired the discussion on overparameterization being necessary for interpolating training data on neural networks?", "answer": ["A law of robustness for two-layers neural networks", "A Universal Law of Robustness via Isoperimetry"], "answer_arxiv_id": ["2009.14444", "2105.12806"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_6351"} +{"question": "What early studies used generative adversarial networks (GANs) for text-to-image generation?", "answer": ["Generative Adversarial Networks: An Overview", "High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs", "Analyzing and Improving the Image Quality of StyleGAN", "Generative Adversarial Networks"], "answer_arxiv_id": ["1710.07035", "1711.11585", "1912.04958", "2203.00667"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_6352"} +{"question": "What papers specifically focus on the DIAYN method, which maximizes the variational lower bound of mutual information?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function"], "answer_arxiv_id": ["1802.06070"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_6353"} +{"question": "Could you tell me about the papers that introduced the neural compression paradigm and connected it to variational inference?", "answer": ["End-to-end Optimized Image Compression", "Lossy Image Compression with Compressive Autoencoders"], "answer_arxiv_id": ["1611.01704", "1703.00395"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_6354"} +{"question": "Could you provide some references that mentioned the use of graph neural networks on irregular meshes?", "answer": ["Inductive Representation Learning on Large Graphs", "Graph Attention Networks"], "answer_arxiv_id": ["1706.02216", "1710.10903"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_6355"} +{"question": "Which papers introduce coupled learning algorithms for handling exploration in Markov games?", "answer": ["Online Reinforcement Learning in Stochastic Games", "Learning Zero-sum Stochastic Games with Posterior Sampling", "A Self-Play Posterior Sampling Algorithm for Zero-Sum Markov Games"], "answer_arxiv_id": ["1712.00579", "2109.03396", "2210.01907"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_6356"} +{"question": "Are there any studies that focused on Lagrangian Neural Networks (LNNs) and their approach towards parameterizing the inertia matrix and divergence of the potential?", "answer": ["Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning", "Lagrangian Neural Networks"], "answer_arxiv_id": ["1907.04490", "2003.04630"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_6357"} +{"question": "What works introduced a collection of novel but simple programming questions and a Python version of the MathQA dataset?", "answer": ["Program Synthesis with Large Language Models"], "answer_arxiv_id": ["2108.07732"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_6358"} +{"question": "Which studies have used Neural ODEs to tackle irregular time series?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_6359"} +{"question": "Any papers that considered fairness and privacy requirements?", "answer": ["Differentially Private Fair Learning", "On the Privacy Risks of Algorithmic Fairness"], "answer_arxiv_id": ["1812.02696", "2011.03731"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_6360"} +{"question": "What studies indicate that Transformer models are Turing-complete?", "answer": ["Attention Is All You Need", "On the Turing Completeness of Modern Neural Network Architectures", "Universal Transformers", "On the Computational Power of Transformers and its Implications in Sequence Modeling"], "answer_arxiv_id": ["1706.03762", "1901.03429", "1807.03819", "2006.09286"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_6361"} +{"question": "Which papers leveraged 2D GAN inversion for viewpoint and lighting control?", "answer": ["StyleRig: Rigging StyleGAN for 3D Control over Portrait Images", "PIE: Portrait Image Embedding for Semantic Control", "PhotoApp: Photorealistic Appearance Editing of Head Portraits"], "answer_arxiv_id": ["2004.00121", "2009.09485", "2103.07658"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_6362"} +{"question": "What research have discussed the concept of class-balanced re-sampling?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition"], "answer_arxiv_id": ["1910.09217"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_6363"} +{"question": "What papers applied KG prediction tasks, like link prediction, as additional supervision?", "answer": ["Deep Bidirectional Language-Knowledge Graph Pretraining"], "answer_arxiv_id": ["2210.09338"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_6364"} +{"question": "Are there any works which demonstrate empirical progress in learning disentangled representations?", "answer": ["Isolating Sources of Disentanglement in Variational Autoencoders"], "answer_arxiv_id": ["1802.04942"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_6365"} +{"question": "What studies presented the idea of representation disentanglement in domain generalization?", "answer": ["Efficient Domain Generalization via Common-Specific Low-Rank\n Decomposition", "Learning to Balance Specificity and Invariance for In and Out of Domain\n Generalization"], "answer_arxiv_id": ["2003.12815", "2008.12839"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_6366"} +{"question": "Any works that addressed cross-view geo-localization with more extreme view point differences?", "answer": ["VIGOR: Cross-View Image Geo-localization beyond One-to-one Retrieval", "Sample4Geo: Hard Negative Sampling For Cross-View Geo-Localisation", "TransGeo: Transformer Is All You Need for Cross-view Image\n Geo-localization", "Uncertainty-aware Vision-based Metric Cross-view Geolocalization", "Beyond Cross-view Image Retrieval: Highly Accurate Vehicle Localization\n Using Satellite Image", "Boosting 3-DoF Ground-to-Satellite Camera Localization Accuracy via\n Geometry-Guided Cross-View Transformer"], "answer_arxiv_id": ["2011.12172", "2303.11851", "2204.00097", "2211.12145", "2204.04752", "2307.08015"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6367"} +{"question": "Are there any studies that improved training and rendering speed through space discretization?", "answer": ["Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes", "FastNeRF: High-Fidelity Neural Rendering at 200FPS", "Baking Neural Radiance Fields for Real-Time View Synthesis", "TensoRF: Tensorial Radiance Fields", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "PlenOctrees for Real-time Rendering of Neural Radiance Fields"], "answer_arxiv_id": ["2101.10994", "2103.10380", "2103.14645", "2203.09517", "2103.13744", "2103.14024"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_6368"} +{"question": "In which works are patches extracted from the image used in image-based pretext tasks?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles", "Context Encoders: Feature Learning by Inpainting"], "answer_arxiv_id": ["1505.05192", "1603.09246v3", "1604.07379"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6369"} +{"question": "Which work proposes an cross-modal adaptation framework that uses external extra paired data for RGB-to-depth knowledge transfer?", "answer": ["Cross-Modal Knowledge Transfer Without Task-Relevant Source Data"], "answer_arxiv_id": ["2209.04027"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_6370"} +{"question": "What works propose filtering-out or reweighting the teacher’s pseudo-labels based on measures of teacher’s uncertainty?", "answer": ["Fidelity-Weighted Learning", "Training Subset Selection for Weak Supervision", "Few-shot learning of neural networks from scratch by pseudo example optimization"], "answer_arxiv_id": ["1711.02799v2", "2206.02914", "1802.03039"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_6371"} +{"question": "What studies conducted an exploration regarding the impact of the fine-tuning of language models on various downstream tasks' syntactic understanding?", "answer": ["On the Evolution of Syntactic Information Encoded by BERT's\n Contextualized Representations"], "answer_arxiv_id": ["2101.11492"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_6372"} +{"question": "Can you provide some works that have studied controlled text-to-image generation?", "answer": ["Learning to Compose Visual Relations"], "answer_arxiv_id": ["2111.09297"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_6373"} +{"question": "Which study proposed an adaptive weight in online fine-tuning of O2O RL algorithms?", "answer": ["Adaptive Behavior Cloning Regularization for Stable Offline-to-Online Reinforcement Learning"], "answer_arxiv_id": ["2210.13846"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_6374"} +{"question": "Could you tell me some research where additional prior information like surface curvature, 2D image overlap, and scene structure was used for alignment of 3D point clouds?", "answer": ["Q-REG: End-to-End Trainable Point Cloud Registration with Surface\n Curvature", "SGAligner : 3D Scene Alignment with Scene Graphs"], "answer_arxiv_id": ["2309.16023", "2304.14880"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_6375"} +{"question": "Could you provide me some studies that use property filtering during selection in GP methods?", "answer": ["Shape-constrained Symbolic Regression – Improving Extrapolation with Prior Knowledge"], "answer_arxiv_id": ["2103.15624"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_6376"} +{"question": "Which studies suggest methods to tackle the problem of modality laziness and utilize information from subordinate modalities?", "answer": ["What Makes Training Multi-Modal Classification Networks Hard?", "Improving Multimodal Accuracy Through Modality Pre-training and\n Attention", "Modality Competition: What Makes Joint Training of Multi-modal Network\n Fail in Deep Learning? (Provably)", "PMR: Prototypical Modal Rebalance for Multimodal Learning"], "answer_arxiv_id": ["1905.12681", "2011.06102", "2203.12221", "2211.07089"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_6377"} +{"question": "Could you provide me some studies about uncertainty estimation in supervised learning that focus on detecting Out-of-Distribution training samples?", "answer": ["Enhancing The Reliability of Out-of-distribution Image Detection in\n Neural Networks", "Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust\n Deep Learning"], "answer_arxiv_id": ["1706.02690", "1803.04765"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_6378"} +{"question": "Which works contributed to the self-supervised learning of depth estimation?", "answer": ["Unsupervised CNN for Single View Depth Estimation: Geometry to the\n Rescue", "Unsupervised Monocular Depth Estimation with Left-Right Consistency", "Unsupervised Learning of Depth and Ego-Motion from Video", "Digging Into Self-Supervised Monocular Depth Estimation", "How do neural networks see depth in single images?", "Forget About the LiDAR: Self-Supervised Depth Estimators with MED\n Probability Volumes", "On the uncertainty of self-supervised monocular depth estimation", "The Temporal Opportunist: Self-Supervised Multi-Frame Monocular Depth"], "answer_arxiv_id": ["1603.04992", "1609.03677", "1704.07813", "1806.01260", "1905.07005", "2008.03633", "2005.06209", "2104.14540"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_6379"} +{"question": "Which publications describe a disentangled representation as separating the factors of variation?", "answer": ["Disentangling Factors of Variation via Generative Entangling", "Representation Learning: A Review and New Perspectives"], "answer_arxiv_id": ["1210.5474", "1206.5538"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_6380"} +{"question": "Which works applied the attention mechanism on the backbone features to model long-range context dependencies in scene parsing?", "answer": ["OCNet: Object Context Network for Scene Parsing"], "answer_arxiv_id": ["1809.00916"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_6381"} +{"question": "Can you name some studies that have explored applications related to NeRF deformations, such as static scene editing and dynamic scene reconstruction?", "answer": ["NeRF-Editing: Geometry Editing of Neural Radiance Fields", "Virtual Elastic Objects", "NeuPhysics: Editable Neural Geometry and Physics from Monocular Videos", "Dynamic Mesh-Aware Radiance Fields", "Neural Impostor: Editing Neural Radiance Fields with Explicit Shape\n Manipulation"], "answer_arxiv_id": ["2205.04978", "2201.04623", "2210.12352", "2309.04581", "2310.05391"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_6382"} +{"question": "Could you provide me some works that focus on multi-grained semantics grounding by learning the alignment of objects and actions across modalities?", "answer": ["ActBERT: Learning Global-Local Video-Text Representations"], "answer_arxiv_id": ["2011.07231"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_6383"} +{"question": "What papers focus on the impact of up-sampling across the entire image in GAN architectures?", "answer": ["Watch your Up-Convolution: CNN Based Generative Deep Neural Networks are\n Failing to Reproduce Spectral Distributions", "Leveraging Frequency Analysis for Deep Fake Image Recognition"], "answer_arxiv_id": ["2003.01826", "2003.08685"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_6384"} +{"question": "What papers discuss the development and application of Neural Radiance Fields (NeRF) for novel view synthesis with complex scenes?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance\n Fields", "Neural Sparse Voxel Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2103.13415", "2112.03907", "2007.11571", "2201.05989"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_6385"} +{"question": "Which papers fit density estimators on internal model representations to obtain information of interest?", "answer": ["DEUP: Direct Epistemic Uncertainty Prediction"], "answer_arxiv_id": ["2102.08501"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_6386"} +{"question": "Which studies developed ray-based implicit 3D shape representations with MLPs by taking individual rays as input?", "answer": ["Light Field Networks: Neural Scene Representations with Single-Evaluation Rendering", "NeuLF: Efficient Novel View Synthesis with Neural 4D Light Field"], "answer_arxiv_id": ["2106.02634", "2105.07112"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_6387"} +{"question": "Which study proposed mixture invariant training (MixIT) for learning sound separation on noisy data?", "answer": ["Unsupervised Sound Separation Using Mixture Invariant Training"], "answer_arxiv_id": ["2006.12701"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_6388"} +{"question": "Are there any studies that focus on learning from sub-optimal demonstrations by learning a weighting function over demonstrations?", "answer": ["Learning from Imperfect Demonstrations via Adversarial Confidence Transfer", "Imitation Learning by Estimating Expertise of Demonstrators", "Better-than-Demonstrator Imitation Learning via Automatically-Ranked Demonstrations", "Learning from Suboptimal Demonstration via Self-Supervised Reward Regression"], "answer_arxiv_id": ["2202.02967", "2202.01288", "1907.03976", "2010.11723"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_6389"} +{"question": "What works represent the branch of reference-based SR that utilizes patch-matching mechanisms?", "answer": ["Image Super-Resolution by Neural Texture Transfer", "Learning Texture Transformer Network for Image Super-Resolution", "MASA-SR: Matching Acceleration and Spatial Adaptation for\n Reference-Based Image Super-Resolution", "Coarse-to-Fine Embedded PatchMatch and Multi-Scale Dynamic Aggregation\n for Reference-based Super-Resolution"], "answer_arxiv_id": ["1903.00834", "2006.04139", "2106.02299", "2201.04358"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_6390"} +{"question": "Which paper provides an evidence that large generative models are not good at generating synthetic data for tasks with complicated structures?", "answer": ["FlipDA: Effective and Robust Data Augmentation for Few-Shot Learning"], "answer_arxiv_id": ["2108.06332"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_6391"} +{"question": "What work developed a decentralized sparse training technique to lower communication and computation cost in DFL?", "answer": ["DisPFL: Towards Communication-Efficient Personalized Federated Learning via Decentralized Sparse Training"], "answer_arxiv_id": ["2206.00187"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_6392"} +{"question": "What studies have been conducted on computing eigenfunctions for a known spectrum?", "answer": ["Construction of eigenfunctions for scalar-type operators via Laplace averages with connections to the Koopman operator"], "answer_arxiv_id": ["1403.6559"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_6393"} +{"question": "Are there any works that have shown solutions can be little impacted in interpolation regime by such objectives that re-weight losses?", "answer": ["The Implicit Bias of Gradient Descent on Separable Data", "What is the Effect of Importance Weighting in Deep Learning?", "Importance Tempering: Group Robustness for Overparameterized Models"], "answer_arxiv_id": ["1710.10345", "1812.03372", "2209.08745"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6394"} +{"question": "Which studies designed color-constancy, size constancy, and face perception illusions?", "answer": ["Designing Perceptual Puzzles by Differentiating Probabilistic Programs"], "answer_arxiv_id": ["2204.12301"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_6395"} +{"question": "Which studies extend general factoid questions to the domain of scientific literature?", "answer": ["A Dataset of Information-Seeking Questions and Answers Anchored in\n Research Papers"], "answer_arxiv_id": ["2105.03011"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_6396"} +{"question": "Any works about optimizing external knowledge utilization in the generation component of RAG?", "answer": ["Leveraging Passage Retrieval with Generative Models for Open Domain\n Question Answering"], "answer_arxiv_id": ["2007.01282"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_6397"} +{"question": "Can you list the studies that propose end-to-end deep learning approaches for steganography?", "answer": ["HiDDeN: Hiding Data With Deep Networks", "Generating Steganographic Images via Adversarial Training", "SteganoGAN: High Capacity Image Steganography with GANs"], "answer_arxiv_id": ["1807.09937", "1703.00371v3", "1901.03892v2"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_6398"} +{"question": "Any benchmarks offer a varied dataset with 100 hours of video?", "answer": ["Rescaling Egocentric Vision"], "answer_arxiv_id": ["2006.13256"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_6399"} +{"question": "Which paper talks about removal-based explanations, where inputs are withheld from the model?", "answer": ["Explaining by Removing: A Unified Framework for Model Explanation"], "answer_arxiv_id": ["2011.14878"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_6400"} +{"question": "What papers propose self-supervised approaches for sequential disentanglement?", "answer": ["S3VAE: Self-Supervised Sequential VAE for Representation Disentanglement and Data Generation", "Contrastively Disentangled Sequential Variational Autoencoder"], "answer_arxiv_id": ["2005.11437", "2110.12091"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_6401"} +{"question": "Could you provide me some references for the more diverse datasets used in learning 3D models?", "answer": ["Zero-Shot Learning -- A Comprehensive Evaluation of the Good, the Bad\n and the Ugly", "APT-36K: A Large-scale Benchmark for Animal Pose Estimation and Tracking", "Animal Kingdom: A Large and Diverse Dataset for Animal Behavior\n Understanding", "Animal3D: A Comprehensive Dataset of 3D Animal Pose and Shape"], "answer_arxiv_id": ["1707.00600", "2206.05683", "2204.08129", "2308.11737"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_6402"} +{"question": "Which paper proposed to model the Markov chain within the DDPMs with Gaussian transitions?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_6403"} +{"question": "Are there studies that explored the benefits of external feedback in the Markov Decision Process (MDP)?", "answer": ["Control Regularization for Reduced Variance Reinforcement Learning", "Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learning", "Widening the Pipeline in Human-Guided Reinforcement Learning with Explanation and Context-Aware Data Augmentation", "Deep Reinforcement Learning from Human Preferences", "Interactive Learning from Policy-Dependent Human Feedback", "Scaling Laws for Reward Model Overoptimization", "Safe Model-based Reinforcement Learning with Stability Guarantees", "Trustworthy Reinforcement Learning Against Intrinsic Vulnerabilities: Robustness, Safety, and Generalizability", "DARLA: Improving Zero-Shot Transfer in Reinforcement Learning"], "answer_arxiv_id": ["1905.05380", "2108.06266", "2006.14804", "1706.03741", "1701.06049", "2210.10760", "1705.08551", "2209.08025", "1707.08475"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_6404"} +{"question": "Which works introduce one-stage pose estimation methods?", "answer": ["DirectPose: Direct End-to-End Multi-Person Pose Estimation", "Single-Stage Multi-Person Pose Machines", "Objects as Points", "Bottom-Up Human Pose Estimation Via Disentangled Keypoint Regression", "YOLO-Pose: Enhancing YOLO for Multi Person Pose Estimation Using Object Keypoint Similarity Loss"], "answer_arxiv_id": ["1911.07451", "1908.09220", "1904.07850", "2104.02300", "2204.06806v1"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6405"} +{"question": "Could you provide me some works about weight reparameterization to reduce the inference cost of deep learning models?", "answer": ["DiracNets: Training Very Deep Neural Networks Without Skip-Connections", "ACNet: Strengthening the Kernel Skeletons for Powerful CNN via\n Asymmetric Convolution Blocks", "RepVGG: Making VGG-style ConvNets Great Again", "Fourier Series Expansion Based Filter Parametrization for Equivariant\n Convolutions", "Parameter Efficient Training of Deep Convolutional Neural Networks by\n Dynamic Sparse Reparameterization", "ResRep: Lossless CNN Pruning via Decoupling Remembering and Forgetting", "Re-parameterizing Your Optimizers rather than Architectures"], "answer_arxiv_id": ["1706.00388", "1908.03930", "2101.03697", "2107.14519", "1902.05967", "2007.03260", "2205.15242"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_6406"} +{"question": "What research proposed to measure user satisfaction based on the features extracted from an evaluation of interaction quality in a dialogue?", "answer": ["Multi-domain Conversation Quality Evaluation via User Satisfaction\n Estimation"], "answer_arxiv_id": ["1911.08567"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_6407"} +{"question": "What works have observed a power-law relationship between generalization error and training set size?", "answer": ["Recognition in Terra Incognita"], "answer_arxiv_id": ["1807.04975"], "source_meta": {"published_time": "20200624"}, "qid": "AutoScholarQuery_train_6408"} +{"question": "Are there any works that propose model-based fitting of the manifold locations of the events generated by the observation of a line under motion?", "answer": ["A 5-Point Minimal Solver for Event Camera Relative Motion Estimation"], "answer_arxiv_id": ["2309.17054"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_6409"} +{"question": "Could you provide me some studies showing applications of CLIP in segmentation, video understanding, and image generation?", "answer": ["Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models", "VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery"], "answer_arxiv_id": ["2303.04803", "2109.14084", "2103.17249"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_6410"} +{"question": "Which studies mathematically describe the gradient domination structure by the Polyak-Łojasiewicz condition in the settings of LQR and entropy-regularized MDP?", "answer": ["Global Convergence of Policy Gradient Methods for the Linear Quadratic Regulator", "On the Global Convergence Rates of Softmax Policy Gradient Methods", "Fast Global Convergence of Natural Policy Gradient Methods with Entropy Regularization", "Regularized Gradient Descent Ascent for Two-Player Zero-Sum Markov Games"], "answer_arxiv_id": ["1801.05039", "2005.06392", "2007.06558", "2205.13746"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_6411"} +{"question": "Could you provide me any papers that applied range-null space decomposition to existing GAN prior based SR methods to improve performance?", "answer": ["GAN Prior based Null-Space Learning for Consistent Super-Resolution"], "answer_arxiv_id": ["2211.13524"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_6412"} +{"question": "What research papers have proposed detector free methods for geometric matching?", "answer": ["LoFTR: Detector-Free Local Feature Matching with Transformers", "Quadtree Attention for Vision Transformers", "MatchFormer: Interleaving Attention in Transformers for Feature Matching", "DGC-Net: Dense Geometric Correspondence Network", "GLU-Net: Global-Local Universal Network for Dense Flow and Correspondences"], "answer_arxiv_id": ["2104.00680", "2201.02767", "2203.09645", "1810.08393", "1912.05524"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_6413"} +{"question": "Could you mention alternative methods that attain matching performance at the sparsity levels as IMP and feature iterative pruning and retraining?", "answer": ["Comparing Rewinding and Fine-tuning in Neural Network Pruning", "Winning the Lottery with Continuous Sparsification"], "answer_arxiv_id": ["2003.02389", "1912.04427"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_6414"} +{"question": "What studies have proposed to shape the behavior of pre-trained models via task vectors?", "answer": ["Editing Models with Task Arithmetic"], "answer_arxiv_id": ["2212.04089"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_6415"} +{"question": "Could you provide me studies about self-supervised learning where the input is masked and the model is tasked with reconstructing the missing bits?", "answer": ["Masked Siamese Networks for Label-Efficient Learning", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2204.07141", "2111.06377"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_6416"} +{"question": "Any works about 'chunking' approaches for non-causality?", "answer": ["TOWARDS ONLINE END-TO-END TRANSFORMER AUTOMATIC SPEECH RECOGNITION", "Self-Attention Aligner: A Latency-Control End-to-End Model for ASR Using Self-Attention Network and Chunk-Hopping", "DEVELOPING REAL-TIME STREAMING TRANSFORMER TRANSDUCER FOR SPEECH RECOGNITION ON LARGE-SCALE DATASET"], "answer_arxiv_id": ["1910.11871", "1902.06450", "2010.11395"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_6417"} +{"question": "Which research has attempted to train agents that influence humans to behave more optimally?", "answer": ["Learning with Opponent-Learning Awareness", "DiCE: The Infinitely Differentiable Monte Carlo Estimator", "Stable Opponent Shaping in Differentiable Games"], "answer_arxiv_id": ["1709.04326", "1802.05098", "1811.08469"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_6418"} +{"question": "What works propose feature-wise data augmentation in graph representation learning?", "answer": ["mixup: Beyond Empirical Risk Minimization", "Local Augmentation for Graph Neural Networks", "Graph Convolutional Networks for Graphs Containing Missing Features"], "answer_arxiv_id": ["1710.09412", "2109.03856", "2007.04583"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_6419"} +{"question": "Which works propose a co-training strategy and use mutual agreement for annotation filtering?", "answer": ["Temporal Alignment Networks for Long-term Video"], "answer_arxiv_id": ["2204.02968"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_6420"} +{"question": "What sources mention the use of control codes for training models to generate domain-specific texts in controlled text generation?", "answer": ["CTRL: A Conditional Transformer Language Model for Controllable Generation"], "answer_arxiv_id": ["1909.05858"], "source_meta": {"published_time": "20221106"}, "qid": "AutoScholarQuery_train_6421"} +{"question": "Which research papers focused on the multi-task problem by modeling in RKHS of vector-valued functions?", "answer": ["Convex Learning of Multiple Tasks and their Structure"], "answer_arxiv_id": ["1504.03101"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_6422"} +{"question": "Could you provide me some works that discuss the significant impact of the prompt design on the LLM predictions?", "answer": ["Language Models are Few-Shot Learners", "Making Pre-trained Language Models Better Few-shot Learners", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "Scalable Prompt Generation for Semi-supervised Learning with Language\n Models", "Exploiting Cloze Questions for Few Shot Text Classification and Natural\n Language Inference"], "answer_arxiv_id": ["2005.14165", "2012.15723", "2107.13586v1", "2302.09236", "2001.07676"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_6423"} +{"question": "What works designed novel techniques to learn from demonstrations with multiple modes with Transformers?", "answer": ["From Play to Policy: Conditional Behavior Generation from Uncurated Robot Data"], "answer_arxiv_id": ["2210.10047"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_6424"} +{"question": "Which paper showcased the design of a language representation model BERT in a masked language model manner?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_6425"} +{"question": "Could you provide me some studies about using supervised learning techniques to create a training dataset using classical simulations?", "answer": ["Learning Mesh-Based Simulation with Graph Networks", "MultiScale MeshGraphNets", "How Will It Drape Like? Capturing Fabric Mechanics from Depth Images"], "answer_arxiv_id": ["2010.03409", "2210.00612", "2304.06704"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_6426"} +{"question": "What early approaches struggled to achieve good performance in vision-language pre-training due to limited small-scale datasets?", "answer": ["Unicoder-VL: A Universal Encoder for Vision and Language by Cross-modal\n Pre-training", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "SiRi: A Simple Selective Retraining Mechanism for Transformer-based\n Visual Grounding"], "answer_arxiv_id": ["1908.06066", "2004.06165", "1908.02265", "2207.13325"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6427"} +{"question": "What papers introduced unified frameworks for recommendation tasks through fine-tuning on models?", "answer": ["Recommendation as Language Processing (RLP): A Unified Pretrain,\n Personalized Prompt & Predict Paradigm (P5)"], "answer_arxiv_id": ["2203.13366"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_6428"} +{"question": "Which works use a linear or MLP layer as bridge module in multimodal language models?", "answer": ["Visual Instruction Tuning", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Kosmos-2: Grounding Multimodal Large Language Models to the World"], "answer_arxiv_id": ["2304.08485", "2303.16199", "2304.15010", "2306.15195", "2306.14824"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_6429"} +{"question": "What research works represent the use of Transformers for multimodal fusion in classification and generative tasks?", "answer": ["Multimodal Transformer for Unaligned Multimodal Language Sequences", "Multimodal Learning with Transformers: A Survey", "Transformers in Medical Imaging: A Survey"], "answer_arxiv_id": ["1906.00295", "2206.06488", "2201.09873v1"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_6430"} +{"question": "In what studies the concept of visual counterfactuals is applied to computer vision?", "answer": ["Counterfactual Visual Explanations", "Learning the Difference that Makes a Difference with\n Counterfactually-Augmented Data", "Long-Tailed Classification by Keeping the Good and Removing the Bad\n Momentum Causal Effect", "Counterfactual Fairness", "Counterfactual VQA: A Cause-Effect Look at Language Bias"], "answer_arxiv_id": ["1904.07451", "1909.12434", "2009.12991", "1703.06856", "2006.04315v4"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6431"} +{"question": "What datasets are based on data collected from Infrastructure Perception Systems?", "answer": ["A9-Dataset: Multi-Sensor Infrastructure-Based Dataset for Mobility\n Research"], "answer_arxiv_id": ["2204.06527"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_6432"} +{"question": "Are there any papers about using model co-teaching to improve model generalization with noisy labels?", "answer": ["Co-teaching: Robust Training of Deep Neural Networks with Extremely\n Noisy Labels", "How does Disagreement Help Generalization against Label Corruption?"], "answer_arxiv_id": ["1804.06872", "1901.04215"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_6433"} +{"question": "Could you provide me some studies about improvements in the expressive power of GNNs for molecular representation learning (MRL)?", "answer": ["Weisfeiler and Lehman Go Cellular: CW Networks", "Directional Graph Networks", "DeeperGCN: All You Need to Train Deeper GCNs", "Principal Neighbourhood Aggregation for Graph Nets", "Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting"], "answer_arxiv_id": ["2106.12575", "2010.02863", "2006.07739v1", "2004.05718", "2006.09252"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_6434"} +{"question": "Any researches about using pre-trained DPT depths and various adversarial losses to train VQ3D?", "answer": ["Vision Transformers for Dense Prediction"], "answer_arxiv_id": ["2103.13413"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_6435"} +{"question": "What papers discuss about adversarial attacks in computer vision?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Adversarial Attack on Graph Structured Data"], "answer_arxiv_id": ["1412.6572", "1706.06083", "1806.02371"], "source_meta": {"published_time": "20220719"}, "qid": "AutoScholarQuery_train_6436"} +{"question": "Which works applied information geometry to optimization?", "answer": ["A Stochastic Quasi-Newton Method for Large-Scale Nonconvex Optimization with Applications", "Large-Scale Distributed Second-Order Optimization Using Kronecker-Factored Approximate Curvature for Deep Convolutional Neural Networks", "Limitations of the Empirical Fisher Approximation for Natural Gradient Descent", "Lower Bounds on the Generalization Error of Nonlinear Learning Models", "On the Variance of the Fisher Information for Deep Learning"], "answer_arxiv_id": ["1912.04456", "1811.12019", "1905.12558", "2103.14723", "2107.04205v3"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6437"} +{"question": "What research has proposed various reasoning architectures, expanding from naive prompting?", "answer": ["Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Faithful Reasoning Using Large Language Models"], "answer_arxiv_id": ["2205.09712", "2205.10625", "2208.14271"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_6438"} +{"question": "Which works focus on transformer-based models for vision-language tasks?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "VideoBERT: A Joint Model for Video and Language Representation Learning", "VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text"], "answer_arxiv_id": ["1908.02265", "1904.01766", "2104.11178"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_6439"} +{"question": "Which studies emphasize the importance of capturing the model’s uncertainty and explicitly incorporating the uncertainty into planning?", "answer": ["Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models"], "answer_arxiv_id": ["1805.12114"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_6440"} +{"question": "Which studies initially applied simple transformations to labeled samples for data augmentation?", "answer": ["Very Deep Convolutional Networks for Large-Scale Image Recognition", "Deep Residual Learning for Image Recognition", "AutoAugment: Learning Augmentation Policies from Data"], "answer_arxiv_id": ["1409.1556", "1512.03385", "1805.09501"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_6441"} +{"question": "What work describes the RealToxicityPrompts dataset?", "answer": ["RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models"], "answer_arxiv_id": ["2009.11462"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_6442"} +{"question": "Could you tell me about some works that developed superior reasoning architectures like 'Least-to-Most', 'Tree of Thoughts', and 'Graph of Thoughts'?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Graph of Thoughts: Solving Elaborate Problems with Large Language Models"], "answer_arxiv_id": ["2205.10625", "2305.10601", "2308.09687"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_6443"} +{"question": "What papers introduced the concept of Spectral GNNs?", "answer": ["Graph Signal Processing: Overview, Challenges and Applications", "Graph signal processing for machine learning: A review and new perspectives", "Simplifying Graph Convolutional Networks", "Interpreting and Unifying Graph Neural Networks with An Optimization Framework", "Beyond Low-frequency Information in Graph Convolutional Networks", "A New Perspective on the Effects of Spectrum in Graph Neural Networks"], "answer_arxiv_id": ["1712.00468", "2007.16061", "1902.07153", "2101.11859", "2101.00797", "2112.07160"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_6444"} +{"question": "What works have been done on repairing programs using machine learning models?", "answer": ["Graph-based, Self-Supervised Program Repair from Diagnostic Feedback", "SequenceR: Sequence-to-Sequence Learning for End-to-End Program Repair", "Break-It-Fix-It: Unsupervised Learning for Program Repair"], "answer_arxiv_id": ["2005.10636", "1901.01808", "2106.06600"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_6445"} +{"question": "Can you name some works that made the retrieval process iterative?", "answer": ["Active Retrieval Augmented Generation", "Knowledge-Driven CoT: Exploring Faithful Reasoning in LLMs for\n Knowledge-intensive Question Answering", "ReAct: Synergizing Reasoning and Acting in Language Models", "StructGPT: A General Framework for Large Language Model to Reason over\n Structured Data"], "answer_arxiv_id": ["2305.06983", "2308.13259", "2210.03629", "2305.09645"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_6446"} +{"question": "Which papers developed methodologies in object-centric approaches to visual reasoning that combine slot-based object representations with transformer-based architectures?", "answer": ["Attention over learned object embeddings enables complex visual reasoning", "Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos", "SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Models", "Learning to reason over visual objects"], "answer_arxiv_id": ["2012.08508", "2205.14065", "2210.05861", "2303.02260"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_6447"} +{"question": "Which researches employed an object detector to enhance visual understanding in supervised IC?", "answer": ["Bottom-Up and Top-Down Attention for Image Captioning and Visual\n Question Answering", "Meshed-Memory Transformer for Image Captioning", "Attention on Attention for Image Captioning", "Adaptively Aligned Image Captioning via Adaptive Attention Time", "Show, Recall, and Tell: Image Captioning with Recall Mechanism", "Beyond a Pre-Trained Object Detector: Cross-Modal Textual and Visual\n Context for Image Captioning"], "answer_arxiv_id": ["1707.07998", "1912.08226", "1908.06954", "1909.09060", "2001.05876", "2205.04363"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_6448"} +{"question": "Which research introduced an instruction-aware visual feature extraction method for better alignment of two modalities?", "answer": ["InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning"], "answer_arxiv_id": ["2305.06500"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_6449"} +{"question": "Which works have considered Multi-Block Bilevel Optimization and its applications in Machine Learning?", "answer": ["Large-scale Stochastic Optimization of NDCG Surrogates for Deep Learning with Provable Convergence", "Multi-block Min-max Bilevel Optimization with Applications in Multi-task Deep AUC Maximization"], "answer_arxiv_id": ["2202.12183", "2206.00260"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6450"} +{"question": "What papers introduced Multi-modal Large Language Models (MLLMs) like LLaVA, InstructBLIP, and LAMM?", "answer": ["Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "LAMM: Language-Assisted Multi-Modal Instruction-Tuning Dataset,\n Framework, and Benchmark", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Kosmos-2: Grounding Multimodal Large Language Models to the World"], "answer_arxiv_id": ["2304.08485", "2305.06500", "2306.06687", "2304.10592", "2304.14178", "2304.15010", "2306.15195", "2306.14824"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6451"} +{"question": "Which works study autoregressive methods in long video generation?", "answer": ["Video Diffusion Models", "Phenaki: Variable Length Video Generation From Open Domain Textual\n Description", "Flexible Diffusion Modeling of Long Videos"], "answer_arxiv_id": ["2204.03458", "2210.02399", "2205.11495"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_6452"} +{"question": "What works discussed the de-democritization and monopolization of AI due to computing power?", "answer": ["The De-democratization of AI: Deep Learning and the Compute Divide in Artificial Intelligence Research"], "answer_arxiv_id": ["2010.15581v1"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_6453"} +{"question": "What works studied the effect of stochastic gradient noise on generalization by changing the order of learning different patterns?", "answer": ["Stochastic Modified Equations and Dynamics of Stochastic Gradient Algorithms I: Mathematical Foundations"], "answer_arxiv_id": ["1811.01558"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_6454"} +{"question": "Which studies proposed variational optimization of annealing schedules?", "answer": ["Variational Optimization of Annealing Schedules"], "answer_arxiv_id": ["1502.05313"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_6455"} +{"question": "Could you provide me some studies that enhance task-relevant information in augmented graphs with learnable data augmentation methods?", "answer": ["Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "Data Augmentation for Graph Neural Networks", "Measuring and Relieving the Over-smoothing Problem for Graph Neural Networks from the Topological View", "Graph Structure Learning for Robust Graph Neural Networks"], "answer_arxiv_id": ["2106.05819", "2006.06830", "1909.03211", "2005.10203"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_6456"} +{"question": "Are there any studies that conduct a macro analysis to understand the overall dynamics of memorization in large language models?", "answer": ["Memorization Without Overfitting: Analyzing the Training Dynamics of\n Large Language Models"], "answer_arxiv_id": ["2205.10770"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_6457"} +{"question": "Which papers studied privacy verification in the 'white-box setting'?", "answer": ["The Complexity of Verifying Loop-Free Programs as Differentially Private", "The Complexity of Verifying Boolean Programs as Differentially Private"], "answer_arxiv_id": ["1911.03272", "2309.04642"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_6458"} +{"question": "What studies apply in-context learning to vision tasks by converting vision problems to NLP ones?", "answer": ["Pix2seq: A Language Modeling Framework for Object Detection", "A Unified Sequence Interface for Vision Tasks", "UViM: A Unified Modeling Approach for Vision with Learned Guiding Codes", "Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework"], "answer_arxiv_id": ["2109.10852", "2206.07669", "2205.10337", "2206.08916", "2202.03052"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6459"} +{"question": "Can you mention some works that tune prompt embedding vector through SGD?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "GPT Understands, Too", "P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally Across Scales and Tasks"], "answer_arxiv_id": ["2104.08691", "2103.10385", "2110.07602"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_6460"} +{"question": "Which papers propose generating more accurate mixed labels with saliency information or attention maps for Transformer architectures?", "answer": ["All Tokens Matter: Token Labeling for Training Better Vision Transformers", "TransMix: Attend to Mix for Vision Transformers", "TokenMixup: Efficient Attention-guided Token-level Data Augmentation for Transformers", "SMMix: Self-Motivated Image Mixing for Vision Transformers"], "answer_arxiv_id": ["2104.10858", "2111.09833", "2210.07562", "2212.12977"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_6461"} +{"question": "Which papers in the field of machine vision have studied fine-grained correspondence from video frames, specifically in forms of optical flow and motion estimation?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks", "Accurate Optical Flow via Direct Cost Volume Processing", "PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume", "RAFT: Recurrent All-Pairs Field Transforms for Optical Flow"], "answer_arxiv_id": ["1504.06852", "1612.01925", "1704.07325", "1709.02371", "2003.12039"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_6462"} +{"question": "Which study conducted the theoretical analysis of the extrapolation behavior of multi-layer perceptrons?", "answer": ["How Neural Networks Extrapolate: From Feedforward to Graph Neural Networks"], "answer_arxiv_id": ["2009.11848"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_6463"} +{"question": "Which studies contributed to the advancement of diffusion models in text-to-image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2112.10752", "2307.01952", "2102.12092", "2205.11487"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_6464"} +{"question": "Are there studies that use deformation parameterized by both skeletons and an embedded graph of a pre-scanned template?", "answer": ["Real-time Deep Dynamic Characters", "HDHumans: A Hybrid Approach for High-fidelity Digital Humans"], "answer_arxiv_id": ["2105.01794", "2210.12003"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_6465"} +{"question": "What work inspired the design of a relevance attention mechanism specifically for visual object tracking?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification"], "answer_arxiv_id": ["2106.02034"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_6466"} +{"question": "What study adopted MLPs for learning standard signed distance function (SDF) and a volumetric deformation function?", "answer": ["Deformed Implicit Field: Modeling 3D Shapes with Learned Dense\n Correspondence"], "answer_arxiv_id": ["2011.13650"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_6467"} +{"question": "What research can provide insight on the evaluation of various model's performance in driving trajectories?", "answer": ["Real-time Dynamic Object Detection for Autonomous Driving using Prior 3D-Maps", "End-to-End Urban Driving by Imitating a Reinforcement Learning Coach", "Learning from All Vehicles"], "answer_arxiv_id": ["1809.11036", "2108.08265", "2203.11934"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_6468"} +{"question": "Could you provide me some studies which use the neural radiance field for constructing interactive 3D environments from large-scale visual captures?", "answer": ["Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual Fly-Throughs", "BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale Scene Rendering"], "answer_arxiv_id": ["2112.10703", "2112.05504"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_6469"} +{"question": "Which studies learn individual filtering for each channel in line with a channel-independent structure?", "answer": ["How Powerful are Spectral Graph Neural Networks", "A New Perspective on the Effects of Spectrum in Graph Neural Networks"], "answer_arxiv_id": ["2205.11172", "2112.07160"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_6470"} +{"question": "Which paper describes OSRT that is built on top of Slot Attention?", "answer": ["Object Scene Representation Transformer"], "answer_arxiv_id": ["2206.06922"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_6471"} +{"question": "Which papers explored the resolution of this indeterminacy of nonlinear ICA's underspecification through labeling a small number of data points?", "answer": ["Semi-Supervised StyleGAN for Disentanglement Learning"], "answer_arxiv_id": ["2003.03461"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_6472"} +{"question": "What paper studied no-regret learning in repeated first-prices auctions with budgets?", "answer": ["No-regret Learning in Repeated First-Price Auctions with Budget Constraints"], "answer_arxiv_id": ["2205.14572"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_6473"} +{"question": "Which studies established the sublinear convergence results for the linear MDP?", "answer": ["Cautiously Optimistic Policy Optimization and Exploration with Linear Function Approximation", "Actor-critic is implicitly biased towards high entropy optimal policies", "Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["2103.12923", "2110.11280", "1907.05388"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_6474"} +{"question": "What researches relied on large pretrained vision-language models for text promptable segmentation tasks?", "answer": ["CRIS: CLIP-Driven Referring Image Segmentation", "Image Segmentation Using Text and Image Prompts", "Zero-shot Referring Image Segmentation with Global-Local Context Features", "CLIP-Driven Universal Model for Organ Segmentation and Tumor Detection", "Segment Anything"], "answer_arxiv_id": ["2111.15174", "2112.10003", "2303.17811", "2301.00785", "2401.14159"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_6475"} +{"question": "Could you provide me some papers about implicit surfaces?", "answer": ["Multiview Neural Surface Reconstruction by Disentangling Geometry and\n Appearance", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction"], "answer_arxiv_id": ["2003.09852", "2106.10689"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_6476"} +{"question": "Could you provide me studies that done on the development of algorithmic approaches to learn accurate models given a training set with noisy labels?", "answer": ["mixup: Beyond Empirical Risk Minimization", "Training deep neural networks on noisy labels with bootstrapping", "Beyond Synthetic Noise: Deep Learning on Controlled Noisy Labels", "Learning to Reweight Examples for Robust Deep Learning", "Identifying and Correcting Label Bias in Machine Learning", "Normalized Loss Functions for Deep Learning with Noisy Labels", "Robust Loss Functions under Label Noise for Deep Neural Networks", "Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe Noise", "Combating Noisy Labels by Agreement: A Joint Training Method with Co-Regularization", "DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "Instance-Dependent Label-Noise Learning with Manifold-Regularized Transition Matrix Estimation"], "answer_arxiv_id": ["1710.09412", "1412.6596", "1911.09781", "1803.09050", "1901.04966", "2006.13554", "1712.09482", "1802.05300", "2003.02752", "2002.07394", "2206.02791"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_6477"} +{"question": "Can you provide some studies about the popular PETL technique, prompt tuning?", "answer": ["GPT Understands, Too", "The Power of Scale for Parameter-Efficient Prompt Tuning", "P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally\n Across Scales and Tasks"], "answer_arxiv_id": ["2103.10385", "2104.08691", "2110.07602"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_6478"} +{"question": "Which papers proposed a method that shares the computational cost of the multi-branch network for fast segmentation?", "answer": ["Fast-SCNN: Fast Semantic Segmentation Network"], "answer_arxiv_id": ["1902.04502"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_6479"} +{"question": "Are there any works about a hybrid CNN-GNN architecture to design an efficient computer vision backbone?", "answer": ["MobileViG: Graph-Based Sparse Attention for Mobile Vision Applications"], "answer_arxiv_id": ["2307.00395"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_6480"} +{"question": "Which research treated LLMs as search agents to accomplish a range of search tasks?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback", "WebCPM: Interactive Web Search for Chinese Long-form Question Answering", "WebGLM: Towards An Efficient Web-Enhanced Question Answering System with\n Human Preferences"], "answer_arxiv_id": ["2112.09332", "2305.06849", "2306.07906"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_6481"} +{"question": "Which paper discusses a dataset that focuses on explanation and reasoning similar to WikiWhy in Explainable QA?", "answer": ["Explain Yourself! Leveraging Language Models for Commonsense Reasoning", "Learning to Explain: Datasets and Models for Identifying Valid Reasoning Chains in Multihop Question-Answering", "Explaining Answers with Entailment Trees"], "answer_arxiv_id": ["1906.02361", "2010.03274", "2104.08661"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_6482"} +{"question": "Could you reference some works about data pruning in ANNS?", "answer": ["Accelerating Large-Scale Inference with Anisotropic Vector Quantization", "Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs"], "answer_arxiv_id": ["1908.10396", "1603.09320v4"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6483"} +{"question": "Which research papers discuss sparse linear bandits by using sparsity-inducing methods often Lasso?", "answer": ["A Simple Unified Framework for High Dimensional Bandit Problems", "Doubly-Robust Lasso Bandit", "Sparsity-Agnostic Lasso Bandit"], "answer_arxiv_id": ["2102.09626", "1907.11362", "2007.08477"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_6484"} +{"question": "What works proposed enhancing model generalizability in Domain Generalized Semantic Segmentation (DGSS)?", "answer": ["HGFormer: Hierarchical Grouping Transformer for Domain Generalized\n Semantic Segmentation", "Domain Generalization via Balancing Training Difficulty and Model\n Capability", "Order-preserving Consistency Regularization for Domain Adaptation and\n Generalization", "Semantic-Aware Domain Generalized Segmentation"], "answer_arxiv_id": ["2305.13031", "2309.00844", "2309.13258", "2204.00822"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6485"} +{"question": "What are the key research papers that used attention mechanisms to model mutual influences among agents?", "answer": ["Social Attention: Modeling Attention in Human Crowds", "The Trajectron: Probabilistic Multi-Agent Trajectory Modeling With\n Dynamic Spatiotemporal Graphs", "Soft + Hardwired Attention: An LSTM Framework for Human Trajectory\n Prediction and Abnormal Event Detection", "Trajectron++: Dynamically-Feasible Trajectory Forecasting With\n Heterogeneous Data"], "answer_arxiv_id": ["1710.04689", "1810.05993", "1702.05552", "2001.03093"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_6486"} +{"question": "Which studies have focused on the concept of uniform stability in the context of adversarial training?", "answer": ["Stability Analysis and Generalization Bounds of Adversarial Training", "What is a Good Metric to Study Generalization of Minimax Learners?"], "answer_arxiv_id": ["2210.00960", "2206.04502"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_6487"} +{"question": "Which papers gave proposals for various event representation types?", "answer": ["Event-based Vision meets Deep Learning on Steering Prediction for\n Self-driving Cars", "EV-FlowNet: Self-Supervised Optical Flow Estimation for Event-based\n Cameras", "End-to-End Learning of Representations for Asynchronous Event-Based Data", "A Differentiable Recurrent Surface for Asynchronous Event-Based Data", "Distance Surface for Event-Based Optical Flow", "A Voxel Graph CNN for Object Classification with Event Cameras", "Chasing Day and Night: Towards Robust and Efficient All-Day Object\n Detection Guided by an Event Camera"], "answer_arxiv_id": ["1804.01310", "1802.06898", "1904.08245", "2001.03455v2", "2003.12680", "2106.00216", "2309.09297"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_6488"} +{"question": "Which researches mirrored the Q-former design developed by BLIP-2 to handle lengthy visual sequences?", "answer": ["Language Is Not All You Need: Aligning Perception with Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2302.14045", "2304.14178", "2304.10592"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6489"} +{"question": "What works provide embodied AI datasets with interaction opportunities that can struggle with poor transferability to real-world scenarios?", "answer": ["Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments\n for Embodied AI", "Habitat-Matterport 3D Semantics Dataset", "Habitat Synthetic Scenes Dataset (HSSD-200): An Analysis of 3D Scene\n Scale and Realism Tradeoffs for ObjectGoal Navigation"], "answer_arxiv_id": ["2109.08238", "2210.05633", "2306.11290"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_6490"} +{"question": "Could you provide me the studies that discussed the DP ERM and SCO in a centralized setting?", "answer": ["Differentially Private Empirical Risk Minimization Revisited: Faster and More General", "Private Stochastic Convex Optimization with Optimal Rates", "Private Stochastic Convex Optimization: Optimal Rates in Linear Time", "Output Perturbation for Differentially Private Convex Optimization with Improved Population Loss Bounds, Runtimes and Applications to Private Adversarial Training"], "answer_arxiv_id": ["1802.05251v1", "1908.09970", "2005.04763", "2102.04704"], "source_meta": {"published_time": "20210617"}, "qid": "AutoScholarQuery_train_6491"} +{"question": "Which studies have proposed expectation maximization-based methods focusing on communication compression?", "answer": ["An Expectation-Maximization Perspective on Federated Learning"], "answer_arxiv_id": ["2111.10192"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_6492"} +{"question": "What studies propose a control variates-based method in order to reduce the clients’ distribution drift brought by the discrepancies?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_6493"} +{"question": "Can you provide some papers where mitigation techniques for instability in solving constraint reinforcement learning have been discussed?", "answer": ["Constrained Variational Policy Optimization for Safe Reinforcement Learning", "Constrained Policy Optimization", "Projection-Based Constrained Policy Optimization", "Efficient Off-Policy Safe Reinforcement Learning Using Trust Region Conditional Value at Risk", "ReLOAD: Reinforcement Learning with Optimistic Ascent-Descent for Last-Iterate Convergence in Constrained MDPs", "Responsive Safety in Reinforcement Learning by PID Lagrangian Methods"], "answer_arxiv_id": ["2201.11927", "1705.10528", "2010.03152", "2312.00342", "2302.01275", "2007.03964"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_6494"} +{"question": "Which work introduced depthwise convolution to sharpen the features prior to fusing the decode features in a UNet-like architecture?", "answer": ["Sharp U-Net: Depthwise Convolutional Network for Biomedical Image Segmentation"], "answer_arxiv_id": ["2107.12461"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_6495"} +{"question": "Which research papers focus on learning general facial representations with contrastive learning and mask image modeling?", "answer": ["Pre-training strategies and datasets for facial representation learning", "General Facial Representation Learning in a Visual-Linguistic Manner", "Pose-disentangled Contrastive Learning for Self-supervised Facial\n Representation"], "answer_arxiv_id": ["2103.16554", "2112.03109", "2211.13490"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_6496"} +{"question": "Which paper details an exploration of the concept of Grokking, wherein test accuracy increased sharply after achieving perfect train accuracy?", "answer": ["Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets"], "answer_arxiv_id": ["2201.02177"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_6497"} +{"question": "Could you provide me some references where the Vision Transformers (ViTs) were adopted in the contrastive learning domain?", "answer": ["An Empirical Study of Training Self-Supervised Vision Transformers", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.02057", "2104.14294"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_6498"} +{"question": "Can you provide research studies that have used a robotic platform for collecting sounds and learning action-sound synergy?", "answer": ["Swoosh! Rattle! Thump! -- Actions that Sound"], "answer_arxiv_id": ["2007.01851"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_6499"} +{"question": "Which works focus on understanding if neural networks encode and use concepts?", "answer": ["On Interpretability of Deep Learning based Skin Lesion Classifiers using Concept Activation Vectors", "Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)", "Acquisition of Chess Knowledge in AlphaZero"], "answer_arxiv_id": ["2005.02000", "1711.11279", "2111.09259"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_6500"} +{"question": "Which research papers implemented the L-BFGS algorithm to carry out an optimization process in a reconstruction attack?", "answer": ["Deep Leakage from Gradients"], "answer_arxiv_id": ["1906.08935"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_6501"} +{"question": "In what papers the authors focus on anomaly detection through the reconstruction of pre-trained image features?", "answer": ["A Unified Model for Multi-class Anomaly Detection", "DSR -- A dual subspace re-projection network for surface anomaly\n detection", "Anomaly Detection via Reverse Distillation from One-Class Embedding"], "answer_arxiv_id": ["2206.03687", "2208.01521", "2201.10703"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_6502"} +{"question": "Which work obtained rates for strictly realizable square loss similarly to this study?", "answer": ["Learning with little mixing"], "answer_arxiv_id": ["2206.08269"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_6503"} +{"question": "Which papers make note of the challenges faced by amortized methods when performing uncertainty estimation in deep learning?", "answer": ["Auto-Encoding Variational Bayes", "Stochastic Backpropagation and Approximate Inference in Deep Generative Models", "Do Deep Generative Models Know What They Don’t Know?", "Reliable training and estimation of variance networks"], "answer_arxiv_id": ["1312.6114", "1401.4082v3", "1810.09136", "1906.03260"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_6504"} +{"question": "Which works proposed the ideas to update machine learning models without retraining in consideration of data deletion requests?", "answer": ["DeltaGrad: Rapid retraining of machine learning models", "Making AI Forget You: Data Deletion in Machine Learning", "Approximate Data Deletion from Machine Learning Models", "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks", "Forgetting Outside the Box: Scrubbing Deep Networks of Information Accessible from Input-Output Observations"], "answer_arxiv_id": ["2006.14755", "1907.05012", "2002.10077", "1911.04933", "2003.02960"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_6505"} +{"question": "Which studies worked on optimizing neural signed distance functions by training neural radiance fields?", "answer": ["UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Neuralangelo: High-Fidelity Neural Surface Reconstruction", "Improving neural implicit surfaces geometry with patch warping", "NeuMesh: Learning Disentangled Neural Mesh-based Implicit Field for\n Geometry and Texture Editing"], "answer_arxiv_id": ["2104.10078", "2106.12052", "2106.10689", "2306.03092", "2112.09648", "2207.11911"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_6506"} +{"question": "What works are related to the task of human parsing in semantic segmentation?", "answer": ["Deep Learning Technique for Human Parsing: A Survey and Outlook"], "answer_arxiv_id": ["2301.00394"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_6507"} +{"question": "What early studies involve GAN-based text-to-image models?", "answer": ["Generative Adversarial Text to Image Synthesis", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked\n Generative Adversarial Networks", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "DF-GAN: A Simple and Effective Baseline for Text-to-Image Synthesis"], "answer_arxiv_id": ["1605.05396", "1612.03242", "1711.10485", "2008.05865"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_6508"} +{"question": "Could you provide me the references highlighting the proposed human demonstrations for a robot arm?", "answer": ["What Matters in Learning from Offline Human Demonstrations for Robot Manipulation"], "answer_arxiv_id": ["2108.03298"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_6509"} +{"question": "What studies indicate that inference in transformer models is memory bound?", "answer": ["Fast Transformer Decoding: One Write-Head is All You Need", "Data Movement Is All You Need: A Case Study on Optimizing Transformers", "Efficiently Scaling Transformer Inference"], "answer_arxiv_id": ["1911.02150", "2007.00072v3", "2211.05102v1"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6510"} +{"question": "What works use CLIP to train on the data of region-text pairs?", "answer": ["RegionCLIP: Region-based Language-Image Pretraining"], "answer_arxiv_id": ["2112.09106"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_6511"} +{"question": "What research papers have proposed to simplify the MIM framework by directly utilizing the pixel RGB values as reconstruction targets?", "answer": ["SimMIM: a Simple Framework for Masked Image Modeling"], "answer_arxiv_id": ["2111.09886"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_6512"} +{"question": "Could you provide me references about voxel-based approaches to facilitate 3D convolution operations in deep learning?", "answer": ["Semantic Scene Completion from a Single Depth Image", "O-CNN: Octree-based Convolutional Neural Networks for 3D Shape Analysis", "3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1611.08974", "1712.01537", "1711.10275", "1904.08755"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_6513"} +{"question": "What works explore Butterfly-based methods of approximated matrix multiplication in the context of neural networks?", "answer": ["Pixelated Butterfly: Simple and Efficient Sparse Training for Neural Network Models", "Monarch: Expressive Structured Matrices for Efficient and Accurate Training"], "answer_arxiv_id": ["2112.00029", "2204.00595"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_6514"} +{"question": "What works incorporate 3D awareness into 2D diffusion for better 3D consistency in text-to-3D generation?", "answer": ["Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generation"], "answer_arxiv_id": ["2303.07937"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6515"} +{"question": "Can you name some benchmarks used for evaluating LLM-based code synthesis?", "answer": ["Evaluating Large Language Models Trained on Code", "Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task", "CodeGeeX: A Pre-Trained Model for Code Generation with Multilingual Evaluations on HumanEval-X", "Competition-Level Code Generation with AlphaCode", "SWE-bench: Can Language Models Resolve Real-World GitHub Issues?"], "answer_arxiv_id": ["2107.03374", "1809.08887", "2303.17568", "2203.07814", "2310.06770"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_6516"} +{"question": "What papers are related to the task of speech or music inpainting based on the corresponding text or music score?", "answer": ["Audio Inpainting: Revisited and Reweighted", "SpeechPainter: Text-conditioned Speech Inpainting", "Vision-Infused Deep Audio Inpainting"], "answer_arxiv_id": ["2001.02480", "2202.07273", "1910.10997"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_6517"} +{"question": "Are there research papers that focused on applying targeted learning for estimating the average treatment effects?", "answer": ["Adapting Neural Networks for the Estimation of Treatment Effects"], "answer_arxiv_id": ["1906.02120"], "source_meta": {"published_time": "20220319"}, "qid": "AutoScholarQuery_train_6518"} +{"question": "What studies demonstrate models learning to segment with relatively few labels?", "answer": ["Label-Efficient Semantic Segmentation with Diffusion Models", "Self-Supervised Learning of Object Parts for Semantic Segmentation"], "answer_arxiv_id": ["2112.03126", "2204.13101"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_6519"} +{"question": "Could you provide the study that takes into account the latency in channel pruning?", "answer": ["Structural Pruning via Latency-Saliency Knapsack"], "answer_arxiv_id": ["2210.06659"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_6520"} +{"question": "Which papers suggest that the diffusion/SDE models outperform GANs in image synthesis in both quality and diversity?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_6521"} +{"question": "What studies have used Masked Image Modeling (MIM) for geospatial pretraining?", "answer": ["SatMAE: Pre-training Transformers for Temporal and Multi-Spectral\n Satellite Imagery"], "answer_arxiv_id": ["2207.08051"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_6522"} +{"question": "Which research paper discusses the SDC method that estimates prototypes of each learned class to use in a nearest class mean classifier?", "answer": ["Semantic Drift Compensation for Class-Incremental Learning"], "answer_arxiv_id": ["2004.00440"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_6523"} +{"question": "Which tools for data extraction and interacting with proof assistants are available for Coq?", "answer": ["GamePad: A Learning Environment for Theorem Proving", "Learning to Prove Theorems via Interacting with Proof Assistants"], "answer_arxiv_id": ["1806.00608", "1905.09381"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_6524"} +{"question": "Can you list some research papers that have an object-centric method for video anomaly detection?", "answer": ["SSMTL++: Revisiting Self-Supervised Multi-Task Learning for Video\n Anomaly Detection", "Anomaly Detection in Video via Self-Supervised and Multi-Task Learning", "A Background-Agnostic Framework with Adversarial Training for Abnormal\n Event Detection in Video", "Object-centric Auto-encoders and Dummy Anomalies for Abnormal Event\n Detection in Video", "Attribute-based Representations for Accurate and Interpretable Video\n Anomaly Detection", "Video Anomaly Detection by Solving Decoupled Spatio-Temporal Jigsaw\n Puzzles", "Multimodal Motion Conditioned Diffusion Model for Skeleton-based Video\n Anomaly Detection"], "answer_arxiv_id": ["2207.08003", "2011.07491", "2008.12328", "1812.04960", "2212.00789", "2207.10172", "2307.07205"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_6525"} +{"question": "Could you provide me information about studies associated with spatio-temporal features in Action Quality Assessment?", "answer": ["Learning To Score Olympic Events"], "answer_arxiv_id": ["1611.05125"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_6526"} +{"question": "What studies are about teaching LMs to use calculators?", "answer": ["Training Verifiers to Solve Math Word Problems", "LaMDA: Language Models for Dialog Applications"], "answer_arxiv_id": ["2110.14168", "2201.08239"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_6527"} +{"question": "Who were the first to investigate approaches to reference-free NLG evaluation using LLMs?", "answer": ["Exploring the Use of Large Language Models for Reference-Free Text\n Quality Evaluation: An Empirical Study"], "answer_arxiv_id": ["2304.00723"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_6528"} +{"question": "What research creates proxy masks based on the information provided by segmentation maps or other conditions?", "answer": ["Editing in Style: Uncovering the Local Semantics of GANs"], "answer_arxiv_id": ["2004.14367"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_6529"} +{"question": "What works provide the convergence rates for multiple student neurons with a single teacher neuron?", "answer": ["Over-Parameterization Exponentially Slows Down Gradient Descent for Learning a Single Neuron"], "answer_arxiv_id": ["2302.10034"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_6530"} +{"question": "Could you provide me with some studies that developed applications based on the multi-agent collaborative paradigm?", "answer": ["ProAgent: Building Proactive Cooperative Agents with Large Language\n Models"], "answer_arxiv_id": ["2308.11339"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_6531"} +{"question": "What works represent the second approach in blind image super-resolution, that is, zero-shot learning?", "answer": ["\"Zero-Shot\" Super-Resolution using Deep Internal Learning"], "answer_arxiv_id": ["1712.06087"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_6532"} +{"question": "Which papers have contributed to deep active learning for example-label-based labels?", "answer": ["Batch Active Learning at Scale"], "answer_arxiv_id": ["2107.14263"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_6533"} +{"question": "What works have utilized a generalized uncertainty-weighted regression to improve the robustness of algorithms?", "answer": ["Reinforcement Learning with General Value Function Approximation: Provably Efficient Approach via Bounded Eluder Dimension", "Online Sub-Sampling for Reinforcement Learning with General Function Approximation"], "answer_arxiv_id": ["2005.10804", "2106.07203v2"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_6534"} +{"question": "What research adjusted challenges associated with unbounded scenes in large-scale 3D reconstruction?", "answer": ["Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields"], "answer_arxiv_id": ["2111.12077"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_6535"} +{"question": "Which papers have used string-based representations such as SMILES and InChI for representation learning on molecules?", "answer": ["Large-Scale Chemical Language Representations Capture Molecular Structure and Properties"], "answer_arxiv_id": ["2106.09553v3"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_6536"} +{"question": "What works have proposed applying different types of identifiers in various search scenarios?", "answer": ["Autoregressive Entity Retrieval", "Transformer Memory as a Differentiable Search Index", "Autoregressive Search Engines: Generating Substrings as Document\n Identifiers", "Learning to Rank in Generative Retrieval", "IRGen: Generative Modeling for Image Retrieval"], "answer_arxiv_id": ["2010.00904", "2202.06991", "2204.10628", "2306.15222", "2303.10126"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_6537"} +{"question": "Which works used regularization to make the policy favor the actions from the dataset in offline reinforcement learning?", "answer": ["Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning", "Off-Policy Deep Reinforcement Learning without Exploration"], "answer_arxiv_id": ["1906.00949", "1911.11361", "1812.02900"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_6538"} +{"question": "Could you provide me with some works that prove self-attention realizes low-complexity circuits?", "answer": ["Theoretical Limitations of Self-Attention in Neural Sequence Models", "Inductive Biases and Variable Creation in Self-Attention Mechanisms"], "answer_arxiv_id": ["1906.06755", "2110.10090"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_6539"} +{"question": "Can you provide some papers which talks about Local Differential Privacy?", "answer": ["Local Differential Privacy for Regret Minimization in Reinforcement Learning", "Locally Differentially Private Reinforcement Learning for Linear Mixture Markov Decision Processes", "Differentially Private Exploration in Reinforcement Learning with Linear Representation"], "answer_arxiv_id": ["2010.07778", "2110.10133", "2112.01585"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_6540"} +{"question": "Which paper studied and proved the complexity result of the single-loop algorithm for the nonconvex-concave minimax problem?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems"], "answer_arxiv_id": ["1906.00331"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_6541"} +{"question": "What are some papers that focus on perceiving humans with varied poses and activities in diverse daily-life scenarios?", "answer": ["Human-centric Scene Understanding for 3D Large-scale Scenarios"], "answer_arxiv_id": ["2307.14392"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6542"} +{"question": "Which paper introduced a scalable Frank-Wolfe-based optimizer to counter the scalability issues in MGDA?", "answer": ["Multi-Task Learning as Multi-Objective Optimization"], "answer_arxiv_id": ["1810.04650"], "source_meta": {"published_time": "20230827"}, "qid": "AutoScholarQuery_train_6543"} +{"question": "Which paper first introduced the time-invariant log-barrier?", "answer": ["Sparsity, variance and curvature in multi-armed bandits"], "answer_arxiv_id": ["1711.01037"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_6544"} +{"question": "Which research seeks to answer a similar set of questions on performing system identification in order to learn a good controller, but restricted to the setting of linear dynamics?", "answer": ["Task-Optimal Exploration in Linear Dynamical Systems"], "answer_arxiv_id": ["2102.05214"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_6545"} +{"question": "Which work proposed to approximate unnoticeability by preserving the degree distribution?", "answer": ["Adversarial Attacks on Neural Networks for Graph Data"], "answer_arxiv_id": ["1805.07984"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_6546"} +{"question": "What papers proposed to use neural networks to improve the accuracy and efficiency of numerical solvers?", "answer": ["Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers", "Learning data driven discretizations for partial differential equations", "Using Machine Learning to Augment Coarse-Grid Computational Fluid Dynamics Simulations"], "answer_arxiv_id": ["2007.00016", "1808.04930", "2010.00072"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_6547"} +{"question": "Which studies explore creating diverse and realistic pairwise data by adapting the physical collection means for real-world image SR?", "answer": ["Camera Lens Super-Resolution", "Toward Real-World Single Image Super-Resolution: A New Benchmark and A\n New Model", "Component Divide-and-Conquer for Real-World Image Super-Resolution"], "answer_arxiv_id": ["1904.03378", "1904.00523", "2008.01928"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_6548"} +{"question": "What researches are focused on generating 3D human motion from textual descriptions using diffusion models?", "answer": ["MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "Human Motion Diffusion Model", "FLAME: Free-form Language-based Motion Synthesis & Editing"], "answer_arxiv_id": ["2208.15001", "2209.14916", "2209.00349"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_6549"} +{"question": "What work was the ViT-g text encoder network and ViT-G image encoder used in the experiment cited from?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_6550"} +{"question": "In what research was kernel mean embeddings for distributional representation suggested for the first time?", "answer": ["Counterfactual Mean Embeddings"], "answer_arxiv_id": ["1805.08845"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_6551"} +{"question": "Could you provide some research papers that discuss the adversarial transferability in black-box attacks?", "answer": ["Backpropagating Linearly Improves Transferability of Adversarial Examples", "Enhancing Adversarial Example Transferability with an Intermediate Level Attack", "An Intermediate-level Attack Framework on The Basis of Linear Regression", "Improving Transferability of Adversarial Examples with Input Diversity", "Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks", "Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks", "Admix: Enhancing the Transferability of Adversarial Attacks"], "answer_arxiv_id": ["2012.03528", "1907.10823", "2203.10723", "1803.06978", "1904.02884", "1908.06281", "2102.00436"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_6552"} +{"question": "What works classify object-centric learning methods into categories such as pixel-space approaches, glimpse approaches and sprite approaches?", "answer": ["ClevrTex: A Texture-Rich Benchmark for Unsupervised Multi-Object Segmentation"], "answer_arxiv_id": ["2111.10265"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_6553"} +{"question": "What work introduced the idea of enhancing the generalization of the soft prompt to a wider range of unseen classes and datasets?", "answer": ["Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2203.05557"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_6554"} +{"question": "Which research studies explored animal shape estimation using statistical body models?", "answer": ["3D Menagerie: Modeling the 3D Shape and Pose of Animals", "Three-D Safari: Learning to Estimate Zebra Pose, Shape, and Texture from Images “In the Wild”"], "answer_arxiv_id": ["1611.07700", "1908.07201"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6555"} +{"question": "What research has been done to prune ViT architectures or reduce image-text tokens for faster inference of vision-language models?", "answer": ["TinyCLIP: CLIP Distillation via Affinity Mimicking and Weight\n Inheritance", "PuMer: Pruning and Merging Tokens for Efficient Vision Language Models"], "answer_arxiv_id": ["2309.12314", "2305.17530"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6556"} +{"question": "Which works are key examples of the NeRF framework for novel view synthesis?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_6557"} +{"question": "What works have tried to enforce hard constraints in PDE-solving by enforcing boundary conditions?", "answer": ["Physics-informed neural networks with hard constraints for inverse design"], "answer_arxiv_id": ["2102.04626v1"], "source_meta": {"published_time": "20220718"}, "qid": "AutoScholarQuery_train_6558"} +{"question": "What paper indicates that the degenerate parabolic Generalized Porous Medium Equation (GPME) has presented challenges for classical averaged-based finite volume methods?", "answer": ["Numerical Artifacts in the Generalized Porous Medium Equation: Why Harmonic Averaging Itself Is Not to Blame"], "answer_arxiv_id": ["1709.02581"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_6559"} +{"question": "Which works obtained superpolynomial lower bounds for Correlational SQ (CSQ) algorithms for learning depth-222 networks with Gaussian inputs?", "answer": ["Superpolynomial Lower Bounds for Learning One-Layer Neural Networks using Gradient Descent", "Algorithms and SQ Lower Bounds for PAC Learning One-Hidden-Layer ReLU Networks"], "answer_arxiv_id": ["2006.12011", "2006.12476"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6560"} +{"question": "What papers discuss about training an environment generator that maximises the student’s regret in PAIRED?", "answer": ["Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Design", "Environment Generation for Zero-Shot Compositional Reinforcement Learning"], "answer_arxiv_id": ["2012.02096", "2201.08896"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_6561"} +{"question": "Could you give some examples of works that demonstrated the advantage of Group Sparse Training (GST) in DRL?", "answer": ["GST: Group-Sparse Training for Accelerating Deep Reinforcement Learning"], "answer_arxiv_id": ["2101.09650"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_6562"} +{"question": "What studies focus on tensor decomposition?", "answer": ["Tensor Decompositions for Learning Latent Variable Models", "Tensor Decomposition for Signal Processing and Machine Learning", "A Tensorized Transformer for Language Modeling"], "answer_arxiv_id": ["1210.7559", "1607.01668", "1906.09777"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_6563"} +{"question": "What studies have conducted mathematical research on the grounds behind neural collapse (NC)?", "answer": ["On the emergence of simplex symmetry in the final and penultimate layers of neural network classifiers", "On the Optimization Landscape of Neural Collapse under MSE Loss: Global Optimality with Unconstrained Features", "Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training", "A Geometric Analysis of Neural Collapse with Unconstrained Features", "An unconstrained layer-peeled perspective on neural collapse", "Neural collapse with unconstrained features", "Neural Collapse Under MSE Loss: Proximity to and Dynamics on the Central Path", "Generalizing and Decoupling Neural Collapse via Hyperspherical Uniformity Gap", "Perturbation Analysis of Neural Collapse", "Neural Collapse with Normalized Features: A Geometric Analysis over the Riemannian Manifold", "Separation and Concentration in Deep Networks"], "answer_arxiv_id": ["2012.05420", "2203.01238", "2101.12699", "2105.02375", "2110.02796", "2011.11619", "2106.02073", "2303.06484", "2210.16658", "2209.09211", "2012.10424"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_6564"} +{"question": "What papers used contrastive learning methods to learn invariant representations?", "answer": ["Unsupervised Learning of Visual Representations using Videos", "Time-Contrastive Networks: Self-Supervised Learning from Video", "Unsupervised Feature Learning via Non-Parametric Instance Discrimination", "Temporal Cycle-Consistency Learning", "Learning deep representations by mutual information estimation and maximization", "Momentum Contrast for Unsupervised Visual Representation Learning", "Self-Supervised Learning of Pretext-Invariant Representations", "A Simple Framework for Contrastive Learning of Visual Representations", "Decoupled Contrastive Learning"], "answer_arxiv_id": ["1505.00687", "1704.06888", "1805.01978", "1904.07846", "1808.06670", "1911.05722", "1912.01991", "2002.05709", "2110.06848"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_6565"} +{"question": "Which research papers employed simulated annealing-based methods for solving the chip placement task?", "answer": ["Placement in Integrated Circuits using Cyclic Reinforcement Learning and Simulated Annealing"], "answer_arxiv_id": ["2011.07577"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_6566"} +{"question": "What papers deal with the challenge of the mismatch between the logging policy and the learning policy in off-policy learning?", "answer": ["Estimating Position Bias without Intrusive Interventions", "Top-K Off-Policy Correction for a REINFORCE Recommender System"], "answer_arxiv_id": ["1812.05161", "1812.02353"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_6567"} +{"question": "What research are about adjusting layer-level scales in dynamic networks?", "answer": ["Learning Dynamic Routing for Semantic Segmentation", "Resolution Adaptive Networks for Efficient Inference"], "answer_arxiv_id": ["2003.10401", "2003.07326"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_6568"} +{"question": "Which papers refer to the development of Bayesian Neural Networks (BNNs)?", "answer": ["A Complete Recipe for Stochastic Gradient MCMC", "Learnable Uncertainty under Laplace Approximations", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.04696", "2010.02720", "1506.02142"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_6569"} +{"question": "Can you list down the studies that applied optimization-based methods in Model Inversion?", "answer": ["Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion", "Re-thinking Model Inversion Attacks Against Deep Neural Networks"], "answer_arxiv_id": ["1912.08795", "2304.01669"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_6570"} +{"question": "What studies have applied diffusion models to mel-based text-to-speech?", "answer": ["Diff-TTS: A Denoising Diffusion Model for Text-to-Speech", "Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech", "ProDiff: Progressive Fast Diffusion Model for High-Quality Text-to-Speech", "DiffGAN-TTS: High-Fidelity and Efficient Text-to-Speech with Denoising Diffusion GANs", "Diffsound: Discrete Diffusion Model for Text-to-sound Generation"], "answer_arxiv_id": ["2104.01409", "2105.06337", "2207.06389", "2201.11972", "2207.09983"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_6571"} +{"question": "What research works focus on image-to-image translation using diffusion models?", "answer": ["Palette: Image-to-Image Diffusion Models", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Zero-shot Image-to-Image Translation"], "answer_arxiv_id": ["2111.05826", "2108.01073", "2302.03027"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_6572"} +{"question": "What studies have showed that choosing the action with largest lower confidence bound leads to better performance?", "answer": ["On the Optimality of Batch Policy Optimization Algorithms", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism"], "answer_arxiv_id": ["2104.02293", "2103.12021v2"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_6573"} +{"question": "Which works focused on efficient methods to adapt pretrained models to downstream tasks?", "answer": ["Delta Tuning: A Comprehensive Study of Parameter Efficient Methods for Pre-trained Language Models", "Scaling Down to Scale Up: A Guide to Parameter-Efficient Fine-Tuning"], "answer_arxiv_id": ["2203.06904v2", "2303.15647"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_6574"} +{"question": "Can you name some works that have explored pruning tokens without requiring fine-tuning?", "answer": ["Adaptive Token Sampling For Efficient Vision Transformers", "Token Merging: Your ViT But Faster"], "answer_arxiv_id": ["2111.15667", "2210.09461"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_6575"} +{"question": "Which papers analyze the main factors that contribute to the success of in-context learning?", "answer": ["Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?"], "answer_arxiv_id": ["2202.12837"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_6576"} +{"question": "What papers focus on the topic of Mirror Langevin in constrained sampling?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Flow Matching for Generative Modeling"], "answer_arxiv_id": ["2011.13456", "2210.02747"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_6577"} +{"question": "Which papers explored the use of heatmaps or CNN features from adjacent frames to improve the performance of multi-person human pose estimation in video sequences?", "answer": ["Learning Temporal Pose Estimation from Sparsely-Labeled Videos", "Deep Dual Consecutive Network for Human Pose Estimation", "Temporal Feature Alignment and Mutual Information Maximization for\n Video-Based Human Pose Estimation"], "answer_arxiv_id": ["1906.04016", "2103.07254", "2203.15227"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_6578"} +{"question": "Which works used a cascaded multi-scale manner to prune the disparity search space?", "answer": ["CFNet: Cascade and Fused Cost Volume for Robust Stereo Matching", "Deep Stereo using Adaptive Thin Volume Representation with Uncertainty\n Awareness"], "answer_arxiv_id": ["2104.04314", "1911.12012"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_6579"} +{"question": "What studies discuss the limitations of simply concatenating the context features to the state features in reinforcement learning?", "answer": ["On the Modularity of Hypernetworks"], "answer_arxiv_id": ["2002.10006"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_6580"} +{"question": "What works propose simulation-free methods for training diffusion ODEs?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Maximum Likelihood Training for Score-Based Diffusion ODEs by High-Order Denoising Score Matching", "Flow Matching for Generative Modeling", "Building Normalizing Flows with Stochastic Interpolants", "Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow"], "answer_arxiv_id": ["2011.13456", "2206.08265", "2210.02747", "2209.15571", "2209.03003"], "source_meta": {"published_time": "20230506"}, "qid": "AutoScholarQuery_train_6581"} +{"question": "What works centered around the proposition of various low-level pretext tasks like self-reconstruction in designing self-supervised learning techniques tailored for point cloud understanding?", "answer": ["Learning Representations and Generative Models for 3D Point Clouds", "PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local\n Descriptors"], "answer_arxiv_id": ["1707.02392", "1808.10322"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_6582"} +{"question": "What research work leverages 2D Gaussian Filter to generate viewports with corresponding locations?", "answer": ["Blind Omnidirectional Image Quality Assessment with Viewport Oriented Graph Convolutional Networks", "Predicting Head Movement in Panoramic Video: A Deep Reinforcement Learning Approach"], "answer_arxiv_id": ["2002.09140", "1710.10755"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_6583"} +{"question": "Could you provide me some studies about synthesizing input-output pairs for program synthesis and code generation?", "answer": ["Synthetic Datasets for Neural Program Synthesis", "Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["1912.12345", "2203.07814"], "source_meta": {"published_time": "20220729"}, "qid": "AutoScholarQuery_train_6584"} +{"question": "Which work argues that benchmark leaderboards fail to adequately capture model utility through a microeconomics lens?", "answer": ["Utility is in the Eye of the User: A Critique of NLP Leaderboards"], "answer_arxiv_id": ["2009.13888"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_6585"} +{"question": "Which work does the paper refer to as an inspiration for their proposed MORBiT?", "answer": ["A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic"], "answer_arxiv_id": ["2007.05170"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_6586"} +{"question": "Which works discussed allowing the coefficient of message to be negative as a method to deal with heterophily?", "answer": ["Adaptive Universal Generalized PageRank Graph Neural Network", "Beyond Low-frequency Information in Graph Convolutional Networks", "Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?", "Two Sides of the Same Coin: Heterophily and Oversmoothing in Graph Convolutional Neural Networks"], "answer_arxiv_id": ["2006.07988", "2101.00797", "2109.05641", "2102.06462"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_6587"} +{"question": "What studies used conjugate gradient (CG) to approximate the product in Bi-Level Optimization (BLO)?", "answer": ["Hyperparameter optimization with approximate gradient", "Meta-Learning with Implicit Gradients"], "answer_arxiv_id": ["1602.02355", "1909.04630"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6588"} +{"question": "Could you provide some studies about the convergence of FedAvg and related algorithms in FL strategies?", "answer": ["Local SGD Converges Fast and Communicates Little", "Cooperative SGD: A Unified Framework for the Design and Analysis of Communication-Efficient SGD Algorithms", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Federated Optimization in Heterogeneous Networks", "Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning"], "answer_arxiv_id": ["1805.09767", "1808.07576", "1910.06378", "1812.06127", "1807.06629"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6589"} +{"question": "Which studies have proposed computationally efficient solutions for gradients to reversible SDE solvers?", "answer": ["Scalable Gradients for Stochastic Differential Equations"], "answer_arxiv_id": ["2001.01328"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_6590"} +{"question": "Can you provide research where rank one and random sensing matrices were studied?", "answer": ["ROP: Matrix recovery via rank-one projections", "Exact and Stable Covariance Estimation from Quadratic Sampling via Convex Programming", "Low rank matrix recovery from rank one measurements", "Optimal convex lifted sparse phase retrieval and PCA with an atomic matrix norm regularizer", "Alternating minimization for generalized rank one matrix sensing: Sharp predictions from a random initialization"], "answer_arxiv_id": ["1310.5791", "1310.0807", "1410.6913", "2111.04652v2", "2207.09660v1"], "source_meta": {"published_time": "20230908"}, "qid": "AutoScholarQuery_train_6591"} +{"question": "Does any research use StyleGAN2 for image editing to generate anomalous images?", "answer": ["Few-Shot Defect Image Generation via Defect-Aware Feature Manipulation"], "answer_arxiv_id": ["2303.02389"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_6592"} +{"question": "Could you provide a study which proved convergence of the Adam optimization algorithm for general non-convex functions, assuming gradients are bounded?", "answer": ["Convergence guarantees for RMSProp and ADAM in non-convex optimization and an empirical comparison to Nesterov acceleration"], "answer_arxiv_id": ["1807.06766"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_6593"} +{"question": "Which research works have considered the penalty-dependent bound in the context of BOBW algorithms?", "answer": ["Nearly Optimal Best-of-Both-Worlds Algorithms for Online Learning with Feedback Graphs", "Best-of-Both-Worlds Algorithms for Partial Monitoring"], "answer_arxiv_id": ["2206.00873", "2207.14550"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_6594"} +{"question": "What study made a step forward in both scale and diversity that included 100K street-level images worldwide?", "answer": ["The Mapillary Traffic Sign Dataset for Detection and Classification on a Global Scale"], "answer_arxiv_id": ["1909.04422"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_6595"} +{"question": "Could you provide me some studies that applied tensor-train decomposition to efficiently capture high-order temporal dependencies in RNNs?", "answer": ["Long-Term Forecasting using Higher-Order Tensor RNNs", "On the Memory Mechanism of Tensor-Power Recurrent Models"], "answer_arxiv_id": ["1711.00073", "2103.01521v1"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_6596"} +{"question": "Which studies use transformer networks in various computer vision tasks?", "answer": ["ViViT: A Video Vision Transformer", "End-to-End Object Detection with Transformers", "An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Segmenter: Transformer for Semantic Segmentation", "VideoBERT: A Joint Model for Video and Language Representation Learning"], "answer_arxiv_id": ["2103.15691", "2005.12872", "2010.11929", "2105.05633", "1904.01766"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_6597"} +{"question": "What research papers discuss the phenomenon of shortcut learning during pretraining?", "answer": ["Unmasking Clever Hans Predictors and Assessing What Machines Really Learn", "Invariant Risk Minimization", "Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference", "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness", "The Origins and Prevalence of Texture Bias in Convolutional Neural Networks", "Recognition in Terra Incognita", "Noise or Signal: The Role of Image Backgrounds in Object Recognition"], "answer_arxiv_id": ["1902.10178", "1907.02893", "1902.01007", "1811.12231", "1911.09071", "1807.04975", "2006.09994"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6598"} +{"question": "Which paper explores the concept of enhancing larger models through weak supervision and training them on labels generated by weaker models?", "answer": ["Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak\n Supervision"], "answer_arxiv_id": ["2312.09390"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_6599"} +{"question": "In what paper does the researcher use LRT to correct labels progressively and provides a theoretical proof for convergence to the Bayes optimal classifier?", "answer": ["Learning with feature-dependent label noise: a progressive approach"], "answer_arxiv_id": ["2103.07756"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_6600"} +{"question": "Which papers studied the impact of scaling up model capacity and training tokens in enhancing the performance of computer vision architectures and neural language models?", "answer": ["Deep Residual Learning for Image Recognition", "Identity Mappings in Deep Residual Networks", "NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detection", "CoAtNet: Marrying Convolution and Attention for All Data Sizes", "Language Models are Few-Shot Learners", "Scaling Laws for Neural Language Models", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism", "Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["1512.03385", "1603.05027", "1904.07392", "2106.04803", "2005.14165", "2001.08361", "1910.10683", "1909.08053", "2203.15556"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_6601"} +{"question": "Which works discuss about test-time normalization technique in Test-time adaptation (TTA)?", "answer": ["Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shift", "Improving robustness against common corruptions by covariate shift adaptation"], "answer_arxiv_id": ["2006.10963", "2006.16971v2"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_6602"} +{"question": "Could you provide some references about using quantization as a gradient compression method for communication efficiency in FL?", "answer": ["QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "TernGrad: Ternary Gradients to Reduce Communication in Distributed Deep Learning", "Distributed Mean Estimation with Limited Communication", "signSGD: Compressed Optimisation for Non-Convex Problems", "Optimizing the Communication–Accuracy Trade-off in Federated Learning with Rate–Distortion Theory"], "answer_arxiv_id": ["1610.02132", "1705.07878", "1611.00429", "1802.04434", "2201.02664"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_6603"} +{"question": "Which work is associated with the Perspective API used in automated content moderation?", "answer": ["A New Generation of Perspective API: Efficient Multilingual Character-level Transformers"], "answer_arxiv_id": ["2202.11176"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_6604"} +{"question": "What is the work that discusses about the relevance of IW invariant in the context of RuLSIF?", "answer": ["Relative Density-Ratio Estimation for Robust Distribution Comparison"], "answer_arxiv_id": ["1106.4729"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_6605"} +{"question": "Which research work proposed solving the dual problem to entropic OT in LSOT?", "answer": ["Large-Scale Optimal Transport and Mapping Estimation"], "answer_arxiv_id": ["1711.02283"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_6606"} +{"question": "What papers define CEs in the context of understanding black box ML models?", "answer": ["Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR"], "answer_arxiv_id": ["1711.00399"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_6607"} +{"question": "Could you provide me some studies aimed to reduce the computational complexity of self-attention or improve its ability of modeling visual dependencies?", "answer": ["UFO-ViT: High Performance Linear Vision Transformer without Softmax", "DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "Rethinking Attention with Performers"], "answer_arxiv_id": ["2109.14382", "2106.02034", "2009.14794"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_6608"} +{"question": "What studies contributed to the customization of text-to-image diffusion models for specific objects or styles?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion", "Break-A-Scene: Extracting Multiple Concepts from a Single Image"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2212.04488", "2305.16311"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6609"} +{"question": "Could you provide me with some studies about Vision-Language Models that show impressive performance on various vision-and-language tasks?", "answer": ["CPT: Colorful Prompt Tuning for Pre-trained Vision-Language Models", "Turning a CLIP Model into a Scene Text Detector"], "answer_arxiv_id": ["2109.11797", "2302.14338"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6610"} +{"question": "Could you provide me some recent research about text-to-image generation that employed auto-regressive transformers as generators?", "answer": ["Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["2102.12092", "2105.13290", "2203.13131"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_6611"} +{"question": "What research efforts have been made to simplify, understand, and improve the Structured State Space Sequence (S4) Model?", "answer": ["Diagonal State Spaces are as Effective as Structured State Spaces", "On the Parameterization and Initialization of Diagonal State Space Models", "Long Range Language Modeling via Gated State Spaces", "How to Train Your HiPPO: State Space Models with Generalized Orthogonal Basis Projections", "Simplified State Space Layers for Sequence Modeling", "Simplifying and Understanding State Space Models with Diagonal Linear RNNs", "Resurrecting Recurrent Neural Networks for Long Sequences"], "answer_arxiv_id": ["2203.14343", "2206.11893", "2206.13947", "2206.12037", "2208.04933", "2212.00768", "2303.06349"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_6612"} +{"question": "Could you provide the reference that introduced Conditional Normalizing Flows models?", "answer": ["Normalizing Flows: An Introduction and Review of Current Methods"], "answer_arxiv_id": ["1908.09257"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_6613"} +{"question": "What studies have successfully extracted interpretable symbolic computations from trained models?", "answer": ["What Does BERT Look At? An Analysis of BERT’s Attention", "Visualizing Attention in Transformer-Based Language Representation Models", "BERT Rediscovers the Classical NLP Pipeline", "Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small"], "answer_arxiv_id": ["1906.04341", "1904.02679", "1905.05950", "2211.00593"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_6614"} +{"question": "What papers discuss the interpretation of machine learning models, specifically language models?", "answer": ["A Survey of the State of Explainable AI for Natural Language Processing", "Analysis Methods in Neural Language Processing: A Survey", "A Primer in BERTology: What We Know About How BERT Works"], "answer_arxiv_id": ["2010.00711", "1812.08951", "2002.12327"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_6615"} +{"question": "What works proposed an elimination-based algorithm with optimism in model-free RL?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable"], "answer_arxiv_id": ["1610.09512v2"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_6616"} +{"question": "Which works maximise mutual information between the representations of augmented and non-augmented images?", "answer": ["Generalization in Reinforcement Learning by Soft Data Augmentation"], "answer_arxiv_id": ["2011.13389"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_6617"} +{"question": "Which works leveraged object motion across video frames to learn separate representations for movable foreground and static background in 2D?", "answer": ["Layered Neural Atlases for Consistent Video Editing", "Deformable Sprites for Unsupervised Video Decomposition", "Discovering Objects that Can Move"], "answer_arxiv_id": ["2109.11418", "2204.07151", "2203.10159"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_6618"} +{"question": "Can you provide the study that is recognized as the first end-to-end framework for building vectorized HD map?", "answer": ["VectorMapNet: End-to-end Vectorized HD Map Learning"], "answer_arxiv_id": ["2206.08920"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_6619"} +{"question": "What researches have been conducted on synthesizing scenes given a single image and camera motion?", "answer": ["Pathdreamer: A World Model for Indoor Navigation", "PixelSynth: Generating a 3D-Consistent Experience from a Single Image", "SynSin: End-to-end View Synthesis from a Single Image", "Look Outside the Room: Synthesizing A Consistent Long-Term 3D Scene Video from A Single Image", "Consistent View Synthesis with Pose-Guided Diffusion Models", "Painting 3D Nature in 2D: View Synthesis of Natural Scenes from a Single Semantic Mask", "Worldsheet: Wrapping the World in a 3D Sheet for View Synthesis from a Single Image", "Infinite Nature:Perpetual View Generation of Natural Scenes from a Single Image", "InfiniteNature-Zero: Learning Perpetual View Generation of Natural Scenes from Single Images"], "answer_arxiv_id": ["2105.08756", "2108.05892", "1912.08804", "2203.09457", "2303.17598", "2302.07224", "2012.09854", "2012.09855", "2207.11148"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_6620"} +{"question": "Which works are confined to simplified simulation environments for automating web navigation?", "answer": ["Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration", "WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agents"], "answer_arxiv_id": ["1802.08802", "2207.01206"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_6621"} +{"question": "Can you give me an example of studies that incorporates semantics into appearance and geometry for Neural Radiance Fields?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene Representation"], "answer_arxiv_id": ["2103.15875"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_6622"} +{"question": "Which work is related to decision-making over lesson sequences based on skills in personalized learning?", "answer": ["Latent Skill Embedding for Personalized Lesson Sequence Recommendation"], "answer_arxiv_id": ["1602.07029"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_6623"} +{"question": "Which paper provides an upper bound on the generalization error of representation learning algorithms?", "answer": ["How Does Information Bottleneck Help Deep Learning?"], "answer_arxiv_id": ["2305.18887"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_6624"} +{"question": "What research applied geometry-based methods that focus on removing redundant examples?", "answer": ["Super-Samples from Kernel Herding", "Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["1203.3472", "1708.00489"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_6625"} +{"question": "Could you provide me works about designing neural modules?", "answer": ["Neural Module Networks for Reasoning over Text"], "answer_arxiv_id": ["1912.04971"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_6626"} +{"question": "What works suggested the use of a projection determintal point process (DPP) for quadrature with nodes ?", "answer": ["Kernel quadrature with DPPs", "Monte Carlo with Determinantal Point Processes", "An analysis of Ermakov-Zolotukhin quadrature using kernels"], "answer_arxiv_id": ["1906.07832", "1605.00361", "2309.01200"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6627"} +{"question": "Could you provide some studies about reducing harmfulness using RLHF?", "answer": ["Constitutional AI: Harmlessness from AI Feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback", "Quark: Controllable Text Generation with Reinforced [Un]learning", "The Capacity for Moral Self-Correction in Large Language Models"], "answer_arxiv_id": ["2212.08073", "2204.05862", "2205.13636", "2302.07459"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_6628"} +{"question": "Which work introduced the use of a pre-trained deep steganography model to embed fingerprints into the training set?", "answer": ["Artificial Fingerprinting for Generative Models: Rooting Deepfake\n Attribution in Training Data"], "answer_arxiv_id": ["2007.08457"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6629"} +{"question": "Could you provide me some works on pseudo-label self-training approaches?", "answer": ["Rethinking Pre-training and Self-training"], "answer_arxiv_id": ["2006.06882"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_6630"} +{"question": "What studies have used fairlets in their approach towards fair clustering?", "answer": ["Fair Coresets and Streaming Algorithms for Fair k-Means Clustering", "Scalable Fair Clustering"], "answer_arxiv_id": ["1812.10854", "1902.03519"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_6631"} +{"question": "Which study automatically generates a mask by comparing different text prompts to guide the areas of the image that need editing?", "answer": ["DiffEdit: Diffusion-based semantic image editing with mask guidance"], "answer_arxiv_id": ["2210.11427"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_6632"} +{"question": "What works set the prediction targets as keypoints in single-view methods?", "answer": ["Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular 3D Object Detection", "Objects are Different: Flexible Monocular 3D Object Detection"], "answer_arxiv_id": ["2205.09373", "2104.02323"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_6633"} +{"question": "Can you name some studies that implemented weight regularization in continual learning?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Memory Aware Synapses: Learning what (not) to forget", "AFEC: Active Forgetting of Negative Transfer in Continual Learning"], "answer_arxiv_id": ["1612.00796", "1711.09601", "2110.12187"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_6634"} +{"question": "Which studies aim to extend the capabilities of pre-trained Language Models (LMs) to unseen languages?", "answer": ["Parsing with Multilingual BERT, a Small Corpus, and a Small Treebank", "UNKs Everywhere: Adapting Multilingual Language Models to New Scripts", "Adapting BigScience Multilingual Model to Unseen Languages", "Adapting Pre-trained Language Models to African Languages via\n Multilingual Adaptive Fine-Tuning", "Glot500: Scaling Multilingual Corpora and Language Models to 500\n Languages"], "answer_arxiv_id": ["2009.14124", "2012.15562", "2204.04873", "2204.06487", "2305.12182"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_6635"} +{"question": "Which works on Monocular 3D object detection do not utilize additional data?", "answer": ["Mono3D++: Monocular 3D Vehicle Detection with Two-Scale 3D Hypotheses\n and Task Priors", "M3D-RPN: Monocular 3D Region Proposal Network for Object Detection", "MonoPair: Monocular 3D Object Detection Using Pairwise Spatial\n Relationships", "Monocular 3D Object Detection: An Extrinsic Parameter Free Approach"], "answer_arxiv_id": ["1901.03446", "1907.06038", "2003.00504", "2106.15796"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_6636"} +{"question": "Can you outline the studies that target the generation of a coherently connected sequence of diverse scenes?", "answer": ["Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models", "Long-Term Photometric Consistent Novel View Synthesis with Diffusion\n Models", "Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision", "ZeroNVS: Zero-Shot 360-Degree View Synthesis from a Single Image"], "answer_arxiv_id": ["2303.11989", "2304.10700", "2306.11719", "2310.17994"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_6637"} +{"question": "What papers utilized variants of straight through estimation for DNN quantization?", "answer": ["Training and Inference with Integers in Deep Neural Networks", "Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference", "Relaxed Quantization for Discretized Neural Networks", "A Quantization-Friendly Separable Convolution for MobileNets", "Soft Weight-Sharing for Neural Network Compression", "DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients"], "answer_arxiv_id": ["1802.04680", "1712.05877", "1810.01875", "1803.08607", "1702.04008", "1606.06160"], "source_meta": {"published_time": "20220328"}, "qid": "AutoScholarQuery_train_6638"} +{"question": "Which works carried out an extensive performance analysis of natural policy gradient for average-reward MDPs?", "answer": ["On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift"], "answer_arxiv_id": ["1908.00261"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_6639"} +{"question": "Are there any papers that discuss instance-level explanation methods, particularly saliency maps?", "answer": ["Auditing Visualizations: Transparency Methods Struggle to Detect Anomalous Behavior"], "answer_arxiv_id": ["2206.13498"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_6640"} +{"question": "Which studies introduced projection-based methods for 3D point cloud shape analysis?", "answer": ["Volumetric and Multi-View CNNs for Object Classification on 3D Data", "Multi-view Convolutional Neural Networks for 3D Shape Recognition"], "answer_arxiv_id": ["1604.03265", "1505.00880"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_6641"} +{"question": "Which studies explore using human written programs to precisely specify tasks?", "answer": ["ProTo: Program-Guided Transformer for Program-Guided Tasks"], "answer_arxiv_id": ["2110.00804"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_6642"} +{"question": "Could you provide me the works that explore pixel discrimination and object discrimination to ameliorate image representation learning?", "answer": ["Dense Contrastive Learning for Self-Supervised Visual Pre-Training", "Propagate Yourself: Exploring Pixel-Level Consistency for Unsupervised\n Visual Representation Learning", "Self-EMD: Self-Supervised Object Detection without ImageNet", "Efficient Visual Pretraining with Contrastive Detection", "Unsupervised Object-Level Representation Learning from Scene Images", "Aligning Pretraining for Detection via Object-Level Contrastive Learning", "Self-Supervised Visual Representation Learning with Semantic Grouping", "Object discovery and representation networks"], "answer_arxiv_id": ["2011.09157", "2011.10043", "2011.13677", "2103.10957", "2106.11952", "2106.02637", "2205.15288", "2203.08777"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_6643"} +{"question": "What works discuss different dimensions of data quality in machine learning including accuracy, completeness, consistency, timeliness, and accessibility?", "answer": ["A Survey of Data Quality Measurement and Monitoring Tools"], "answer_arxiv_id": ["1907.08138"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_6644"} +{"question": "Which research has used prediction error to a randomly initialized function as exploration bonus?", "answer": ["Exploration by Random Network Distillation"], "answer_arxiv_id": ["1810.12894"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_6645"} +{"question": "Which study proposed to utilize maximum softmax probability as an initial solution to separate ID and OOD data in OOD detection?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks"], "answer_arxiv_id": ["1610.02136"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_6646"} +{"question": "What studies in theoretical multimodal learning have proposed algorithms based on total correlation or utilized partial information decomposition to quantify relationships between modalities?", "answer": ["TCGM: An Information-Theoretic Framework for Semi-Supervised Multi-Modality Learning"], "answer_arxiv_id": ["2007.06793"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_6647"} +{"question": "What works proposed and used efficient methods for transfer learning like feature selection and adding affine parameters?", "answer": ["Head2Toe: Utilizing Intermediate Representations for Better Transfer Learning", "Scaling & Shifting Your Features: A New Baseline for Efficient Model Tuning", "Parameter-Efficient Transfer Learning for NLP"], "answer_arxiv_id": ["2201.03529", "2210.08823", "1902.00751"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6648"} +{"question": "Which works are focused on robust forecast aggregation used in the study of information aggregation problems?", "answer": ["Robust Forecast Aggregation", "Are You Smarter Than a Random Expert? The Robust Aggregation of Substitutable Signals", "Robust Merging of Information"], "answer_arxiv_id": ["1710.02838v3", "2111.03153", "2106.00088v1"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_6649"} +{"question": "Which works introduced linear transformers providing equivalent recurrent and closed-form formulations?", "answer": ["Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention", "Linear Transformers Are Secretly Fast Weight Programmers", "RoFormer: Enhanced Transformer with Rotary Position Embedding"], "answer_arxiv_id": ["2006.16236", "2102.11174", "2104.09864"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_6650"} +{"question": "What works proposed normal based methods for filtering, such as Moving Robust Principal Component Analysis and Graph Laplacian Regularization?", "answer": ["Feature Graph Learning for 3D Point Cloud Denoising"], "answer_arxiv_id": ["1907.09138"], "source_meta": {"published_time": "20240514"}, "qid": "AutoScholarQuery_train_6651"} +{"question": "Which studies adopted a feed-forward network for arbitrary stylization in real-time in the field of neural style transfer?", "answer": ["Arbitrary Style Transfer with Style-Attentional Networks", "Arbitrary Style Transfer via Multi-Adaptation Network", "Arbitrary Style Transfer in Real-time with Adaptive Instance\n Normalization", "Arbitrary Style Transfer with Deep Feature Reshuffle", "Universal Style Transfer via Feature Transforms", "AdaAttN: Revisit Attention Mechanism in Arbitrary Neural Style Transfer", "Avatar-Net: Multi-scale Zero-shot Style Transfer by Feature Decoration"], "answer_arxiv_id": ["1812.02342", "2005.13219", "1703.06868", "1805.04103", "1705.08086", "2108.03647", "1805.03857"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_6652"} +{"question": "What study uses object-centric abstractions as ‘files’ to store factorized declarative knowledge and the dynamics of the environment?", "answer": ["Object Files and Schemata: Factorizing Declarative and Procedural Knowledge in Dynamical Systems"], "answer_arxiv_id": ["2006.16225"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_6653"} +{"question": "What work considers the trade-off between robustness and accuracy of a neural network via a regularized loss?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1901.08573"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_6654"} +{"question": "What papers conduct research on invariant learning for out-of-distribution generalization on Euclidean data?", "answer": ["Invariant Risk Minimization", "Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization", "Environment Inference for Invariant Learning"], "answer_arxiv_id": ["1907.02893", "2106.06607", "2010.07249"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_6655"} +{"question": "Could you provide me some works that have used normalizing flows in human pose-related tasks?", "answer": ["Normalizing Flows: An Introduction and Review of Current Methods", "Probabilistic Monocular 3D Human Pose Estimation with Normalizing Flows", "Normalizing Flows for Human Pose Anomaly Detection"], "answer_arxiv_id": ["1908.09257", "2107.13788", "2211.10946"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6656"} +{"question": "What studies focus on step-by-step reasoning in text-related tasks?", "answer": ["Explaining Answers with Entailment Trees", "Entailment Tree Explanations via Iterative Retrieval-Generation Reasoner", "Natural Language Deduction through Search over Statement Compositions", "PRover: Proof Generation for Interpretable Reasoning over Rules", "Measuring Systematic Generalization in Neural Proof Generation with Transformers", "multiPRover: Generating Multiple Proofs for Improved Interpretability in Rule Reasoning", "ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language"], "answer_arxiv_id": ["2104.08661", "2205.09224", "2201.06028", "2010.02830", "2009.14786", "2106.01354", "2012.13048"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_6657"} +{"question": "Any works about Active Learning (AL) methods that select informative instances using a scoring function?", "answer": ["Deep Bayesian Active Learning with Image Data", "Learning Loss for Active Learning"], "answer_arxiv_id": ["1703.02910", "1905.03677"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_6658"} +{"question": "What researches adopt data generators to synthesize OOD data for model training?", "answer": ["Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samples", "Out-of-distribution Detection in Classifiers via Generation", "Building robust classifiers through generation of confident out of distribution examples", "G2D: Generate to Detect Anomaly"], "answer_arxiv_id": ["1711.09325", "1910.04241", "1812.00239", "2006.11629"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_6659"} +{"question": "Which works exemplify the adaptation of pretrained features for deep anomaly detection methods?", "answer": ["Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty", "PANDA: Adapting Pretrained Features for Anomaly Detection and Segmentation", "Mean-Shifted Contrastive Loss for Anomaly Detection", "Learning Deep Features for One-Class Classification"], "answer_arxiv_id": ["1906.12340", "2010.05903v3", "2106.03844", "1801.05365"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_6660"} +{"question": "What works explored the concept of higher-order pooling in convolutional networks?", "answer": ["Global Second-order Pooling Convolutional Networks"], "answer_arxiv_id": ["1811.12006"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_6661"} +{"question": "Could you provide me examples of studies that focused on automatically generating high-quality data to enhance the instruction-following capability of LLMs?", "answer": ["Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena", "WizardLM: Empowering Large Language Models to Follow Complex\n Instructions"], "answer_arxiv_id": ["2306.05685", "2304.12244"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_6662"} +{"question": "What papers provide a basis for supervised approaches in machine-generated text detection?", "answer": ["M4: Multi-generator, Multi-domain, and Multi-lingual Black-Box\n Machine-Generated Text Detection", "TURINGBENCH: A Benchmark Environment for Turing Test in the Age of\n Neural Text Generation", "Defending Against Neural Fake News", "Neural Deepfake Detection with Factual Structure of Text", "CoCo: Coherence-Enhanced Machine-Generated Text Detection Under Data\n Limitation With Contrastive Learning"], "answer_arxiv_id": ["2305.14902", "2109.13296", "1905.12616", "2010.07475", "2212.10341"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_6663"} +{"question": "Which papers focused on incorporating structural information in the stages of protein sequence modeling?", "answer": ["Pre-training of Deep Bidirectional Protein Sequence Representations with Structural Information"], "answer_arxiv_id": ["1912.05625"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_6664"} +{"question": "What research has been done on applying object-centric representations to visual question answering and visual reasoning?", "answer": ["Object-Centric Representation Learning for Video Question Answering"], "answer_arxiv_id": ["2104.05166"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_6665"} +{"question": "Any recent works about breakthrough maximum flow algorithm that extends to solve optimal transport?", "answer": ["Maximum Flow and Minimum-Cost Flow in Almost-Linear Time"], "answer_arxiv_id": ["2203.00671"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_6666"} +{"question": "Can you name some works employing point-based methods, extracting features directly from raw point clouds in LiDAR-based 3D object detection?", "answer": ["PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud", "STD: Sparse-to-Dense 3D Object Detector for Point Cloud"], "answer_arxiv_id": ["1812.04244", "1907.10471"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_6667"} +{"question": "Could you provide me some works about utilizing abstract representation rehearsal by storing and replaying representations from intermediate layers in DNNs?", "answer": ["Latent Replay for Real-Time Continual Learning", "Memory-Efficient Incremental Learning Through Feature Adaptation", "Online Learned Continual Compression with Adaptive Quantization Modules"], "answer_arxiv_id": ["1912.01100", "2004.00713v2", "1911.08019"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_6668"} +{"question": "What research work discusses replacing summary statistics with statistical distances in ABC methods?", "answer": ["A Comparison of Likelihood-Free Methods With and Without Summary Statistics"], "answer_arxiv_id": ["2103.02407"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6669"} +{"question": "Which papers proposed the concept of 'learning to learn' in meta-learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "One-shot Learning with Memory-Augmented Neural Networks", "Meta-learning of Sequential Strategies"], "answer_arxiv_id": ["1703.03400", "1605.06065", "1905.03030"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_6670"} +{"question": "Any studies demonstrating that long-term temporal fusion can lead to inadequate detection of dynamic objects?", "answer": ["BEVStereo: Enhancing Depth Estimation in Multi-view 3D Object Detection with Dynamic Temporal Stereo", "Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D Object Detection"], "answer_arxiv_id": ["2209.10248", "2303.11926"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_6671"} +{"question": "What papers are concerning the use of an unsupervised learning objective in prompt engineering?", "answer": ["An Information-theoretic Approach to Prompt Engineering Without Ground Truth Labels"], "answer_arxiv_id": ["2203.11364"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_6672"} +{"question": "Which paper(s) have found that the contrastive and non-contrastive methods end up leading to very similar representations?", "answer": ["On the duality between contrastive and non-contrastive self-supervised learning"], "answer_arxiv_id": ["2206.02574"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_6673"} +{"question": "Can you give examples of studies on the application of LLMs to improve access to legal information?", "answer": ["Chatlaw: A Multi-Agent Collaborative Legal Assistant with Knowledge\n Graph Enhanced Mixture-of-Experts Large Language Model", "Explaining Legal Concepts with Augmented Large Language Models (GPT-4)"], "answer_arxiv_id": ["2306.16092", "2306.09525"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_6674"} +{"question": "Which works proposed normalization-based approaches for TTA?", "answer": ["Improving robustness against common corruptions by covariate shift adaptation", "The Norm Must Go On: Dynamic Unsupervised Domain Adaptation by\n Normalization"], "answer_arxiv_id": ["2006.16971v2", "2112.00463"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_6675"} +{"question": "Which paper considered corruptions on the Bellman operator in their study of MDPs with general function approximation?", "answer": ["A Model Selection Approach for Corruption Robust Reinforcement Learning"], "answer_arxiv_id": ["2110.03580v1"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_6676"} +{"question": "Which studies implemented the method of choosing an item with the highest rank by a verifier for solving math word problems?", "answer": ["Training Verifiers to Solve Math Word Problems"], "answer_arxiv_id": ["2110.14168"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_6677"} +{"question": "Are there any works on DNNs learning mechanisms that focus on the Fourier analysis?", "answer": ["On the Spectral Bias of Neural Networks"], "answer_arxiv_id": ["1806.08734"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_6678"} +{"question": "What work introduces the concept of Dataset Condensation (DC)?", "answer": ["Dataset Condensation with Gradient Matching"], "answer_arxiv_id": ["2006.05929"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_6679"} +{"question": "What are the papers where the application of DDPM in the super-resolution field is discussed?", "answer": ["Come-Closer-Diffuse-Faster: Accelerating Conditional Diffusion Models for Inverse Problems through Stochastic Contraction", "Image Super-Resolution via Iterative Refinement"], "answer_arxiv_id": ["2112.05146", "2104.07636"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_6680"} +{"question": "Could you mention the studies that utilize an extra reward model to assess data quality as part of their method?", "answer": ["MoDS: Model-oriented Data Selection for Instruction Tuning", "Data Diversity Matters for Robust Instruction Tuning"], "answer_arxiv_id": ["2311.15653", "2311.14736"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_6681"} +{"question": "Could you name the studies that attempted to overcome the restrictions in ZSD settings by hallucinating novel classes?", "answer": ["Don’t Even Look Once: Synthesizing Features for Zero-Shot Detection", "GTNet: Generative Transfer Network for Zero-Shot Object Detection"], "answer_arxiv_id": ["1911.07933", "2001.06812"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_6682"} +{"question": "Which studies have adopted the Chain-of-thought prompting paradigm?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Complexity-Based Prompting for Multi-Step Reasoning", "Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2201.11903", "2203.11171", "2210.00720", "2210.03493"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_6683"} +{"question": "What works have addressed the task of annotation-free object segmentation?", "answer": ["Multi-Object Representation Learning with Iterative Variational Inference", "MONet: Unsupervised Scene Decomposition and Representation", "Object-Centric Learning with Slot Attention", "GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinement"], "answer_arxiv_id": ["1903.00450", "1901.11390", "2006.15055", "2104.09958"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_6684"} +{"question": "Could you provide me some studies about the intersection of multi-objective optimization and risk-control?", "answer": ["Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control", "SMAC3: A Versatile Bayesian Optimization Package for Hyperparameter Optimization"], "answer_arxiv_id": ["2110.01052", "2109.09831"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_6685"} +{"question": "Could you provide me with studies that have used NeRFs for LIDAR simulations?", "answer": ["Neural LiDAR Fields for Novel View Synthesis"], "answer_arxiv_id": ["2305.01643v2"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_6686"} +{"question": "Which works use masking technique in bidirectional language model?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "SpanBERT: Improving Pre-training by Representing and Predicting Spans"], "answer_arxiv_id": ["1810.04805", "1910.10683", "1907.10529"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_6687"} +{"question": "Could you tell me about research work that discussed collective prediction as a dynamic consensus-finding procedure?", "answer": ["Test-time collective prediction"], "answer_arxiv_id": ["2106.12012"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_6688"} +{"question": "Could you provide me some works that applied transfer-learning-based methods in few-shot learning?", "answer": ["A Closer Look at Few-shot Classification", "A Baseline for Few-Shot Image Classification", "SimpleShot: Revisiting Nearest-Neighbor Classification for Few-Shot Learning", "Associative Alignment for Few-shot Image Classification", "On the Importance of Distractors for Few-Shot Classification", "POODLE: Improving Few-shot Learning via Penalizing Out-of-Distribution Samples"], "answer_arxiv_id": ["1904.04232", "1909.02729", "1911.04623", "1912.05094", "2109.09883", "2206.04679"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_6689"} +{"question": "Can you mention some publications that have connected non-PDE-based integral kernels and the attention mechanism?", "answer": ["LieTransformer: Equivariant Self-Attention for Lie Groups", "Adaptive Fourier Neural Operators: Efficient Token Mixers for Transformers"], "answer_arxiv_id": ["2012.10885", "2111.13587"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_6690"} +{"question": "What studies focused on finetuning the pre-trained models for In-context learning (ICL)?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization", "Finetuned Language Models Are Zero-Shot Learners", "MetaICL: Learning to Learn In Context", "Meta-learning via Language Model In-context Tuning"], "answer_arxiv_id": ["2110.08207", "2109.01652", "2110.15943", "2110.07814"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_6691"} +{"question": "What paper developed the idea of reducing redundancy in the input context to compress the KV cache?", "answer": ["Compressing Context to Enhance Inference Efficiency of Large Language\n Models"], "answer_arxiv_id": ["2310.06201"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_6692"} +{"question": "What are some of the works that extended Bayesian optimization (BO) to the batch setting in recent years?", "answer": ["On Batch Bayesian Optimization", "Parallel Gaussian Process Optimization with Upper Confidence Bound and Pure Exploration", "Batch Bayesian Optimization via Local Penalization", "Bayesian Optimization under Stochastic Delayed Feedback", "The Parallel Knowledge Gradient Method for Batch Bayesian Optimization"], "answer_arxiv_id": ["1911.01032", "1304.5350", "1505.08052v4", "2206.09341", "1606.04414"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_6693"} +{"question": "What papers provide information on hierarchical RL algorithms enabling few-shot generalization to unseen tasks?", "answer": ["Abstract Value Iteration for Hierarchical Reinforcement Learning"], "answer_arxiv_id": ["2010.15638v2"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6694"} +{"question": "Are there any papers that introduced data poisoning attack methods that generate malicious samples to sabotage TTA?", "answer": ["Uncovering Adversarial Risks of Test-Time Adaptation", "Test-Time Poisoning Attacks Against Test-Time Adaptation Models"], "answer_arxiv_id": ["2301.12576", "2308.08505"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_6695"} +{"question": "Which studies showcase LLM-based agents abilities to emulate human social dynamics?", "answer": ["Generative Agents: Interactive Simulacra of Human Behavior", "CharacterChat: Learning towards Conversational AI with Personalized\n Social Support", "Out of One, Many: Using Language Models to Simulate Human Samples", "Apollo's Oracle: Retrieval-Augmented Reasoning in Multi-Agent Debates"], "answer_arxiv_id": ["2304.03442", "2308.10278", "2209.06899", "2312.04854"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_6696"} +{"question": "Which papers have proposed approximate machine unlearning methods?", "answer": ["Certified Data Removal from Machine Learning Models", "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks", "SSSE: Efficiently Erasing Samples from Trained Machine Learning Models", "Remember What You Want to Forget: Algorithms for Machine Unlearning", "Mixed-Privacy Forgetting in Deep Networks", "Certifiable Machine Unlearning for Linear Models", "Algorithms that Approximate Data Removal: New Results and Limitations", "Deep Unlearning via Randomized Conditionally Independent Hessians", "DeltaGrad: Rapid retraining of machine learning models", "Variational Bayesian Unlearning", "Descent-to-Delete: Gradient-Based Methods for Machine Unlearning"], "answer_arxiv_id": ["1911.03030", "1911.04933", "2107.03860", "2103.03279v2", "2012.13431", "2106.15093v3", "2209.12269", "2204.07655", "2006.14755", "2010.12883", "2007.02923v1"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_6697"} +{"question": "Which studies proposed transformers for video-based 3D HPE?", "answer": ["3D Human Pose Estimation with Spatial and Temporal Transformers", "MHFormer: Multi-Hypothesis Transformer for 3D Human Pose Estimation", "MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose\n Estimation in Video", "MotionBERT: A Unified Perspective on Learning Human Motion\n Representations"], "answer_arxiv_id": ["2103.10455", "2111.12707", "2203.00859", "2210.06551"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_6698"} +{"question": "Which papers proposed the use of higher-order solvers for diffusion integration?", "answer": ["Gotta Go Fast When Generating Data with Score-Based Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "GENIE: Higher-Order Denoising Diffusion Solvers"], "answer_arxiv_id": ["2105.14080", "2206.00364v2", "2210.05475"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_6699"} +{"question": "What research papers have been published on Rehearsal-based methods in Continual Learning?", "answer": ["Experience Replay for Continual Learning", "DyTox: Transformers for Continual Learning with DYnamic TOken eXpansion", "POP: Prompt Of Prompts for Continual Learning", "Rainbow Memory: Continual Learning with a Memory of Diverse Samples", "Gradient based sample selection for online continual learning", "Selective Experience Replay for Lifelong Learning", "BiRT: Bio-inspired Replay in Vision Transformers for Continual Learning"], "answer_arxiv_id": ["1811.11682", "2111.11326", "2306.08200", "2103.17230", "1903.08671", "1802.10269", "2305.04769"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_6700"} +{"question": "Could you provide me some work on the Binder model?", "answer": ["Binding Language Models in Symbolic Languages"], "answer_arxiv_id": ["2210.02875"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_6701"} +{"question": "Can you name some research that investigated how to adapt pre-trained representations for downstream tasks?", "answer": ["Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time", "Patching open-vocabulary modelsby interpolating weights", "Surgical Fine-Tuning Improves Adaptation to Distribution Shifts", "Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations", "Finetune like you pretrain: Improved finetuning of zero-shot vision models", "CLIP Itself is a Strong Fine-tuner: Achieving 85.7% and 88.0% Top-1 Accuracy with ViT-B and ViT-L on ImageNet"], "answer_arxiv_id": ["2202.10054v1", "2203.05482", "2208.05592", "2210.11466", "2204.02937", "2212.00638", "2212.06138"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_6702"} +{"question": "Which work introduced a technique using Total Amount of Noise (TAN) and a scaling law for DP-SGD?", "answer": ["TAN Without a Burn: Scaling Laws of DP-SGD"], "answer_arxiv_id": ["2210.03403"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_6703"} +{"question": "Could you provide some studies about memory models in the field of multi-object tracking (MOT)?", "answer": ["MeMOT: Multi-Object Tracking with Memory"], "answer_arxiv_id": ["2203.16761"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_6704"} +{"question": "Can you provide studies that explain the success of the arithmetic mean for model merging from the perspective of loss landscapes and linear mode connectivity?", "answer": ["Linear Mode Connectivity and the Lottery Ticket Hypothesis", "Essentially No Barriers in Neural Network Energy Landscape", "Git Re-Basin: Merging Models Modulo Permutation Symmetries"], "answer_arxiv_id": ["1912.05671", "1803.00885", "2209.04836"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_6705"} +{"question": "What studies focused on utilizing set predictors in conformal risk control for image regressions?", "answer": ["Image-to-Image Regression with Distribution-Free Uncertainty Quantification and Applications in Imaging"], "answer_arxiv_id": ["2202.05265"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_6706"} +{"question": "Which paper introduces an architecture that initially outlines a plan and then applies environmental feedback to iteratively refine it?", "answer": ["Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents"], "answer_arxiv_id": ["2302.01560v2"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_6707"} +{"question": "Which works focused on understanding the feature learning mechanism and how it leads to sample complexity improvements?", "answer": ["Regularization Matters: Generalization and Optimization of Neural Nets v.s. their Induced Kernel", "On the Power of Over-parametrization in Neural Networks with Quadratic Activation", "On the Power and Limitations of Random Features for Understanding Neural Networks", "What Can ResNet Learn Efficiently, Going Beyond Kernels?", "When Do Neural Networks Outperform Kernel Methods?", "Learning Parities with Neural Networks", "Kernel and Rich Regimes in Overparametrized Models", "Quantifying the Benefit of Using Differentiable Learning over Tangent Kernels"], "answer_arxiv_id": ["1810.05369", "1803.01206", "1904.00687", "1905.10337", "2006.13409", "2002.07400", "2002.09277", "2103.01210v1"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_6708"} +{"question": "Could you provide me with several examples of research on the application of curve-based techniques in image vectorization?", "answer": ["Inverse Diffusion Curves using Shape Optimization"], "answer_arxiv_id": ["1610.02769"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_6709"} +{"question": "What papers have studied the optimization geometry of general objective function with Burer-Monteiro type factorization?", "answer": ["Global Optimality in Low-rank Matrix Optimization", "Symmetry, Saddle Points, and Global Optimization Landscape of Nonconvex Matrix Factorization", "The Global Optimization Geometry of Low-Rank Matrix Optimization"], "answer_arxiv_id": ["1702.07945", "1612.09296", "1703.01256"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_6710"} +{"question": "Which studies focus on accelerating rendering speed in NeRF framework?", "answer": ["PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Baking Neural Radiance Fields for Real-Time View Synthesis", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "TensoRF: Tensorial Radiance Fields", "MERF: Memory-Efficient Radiance Fields for Real-time View Synthesis in\n Unbounded Scenes", "MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient\n Neural Field Rendering on Mobile Architectures"], "answer_arxiv_id": ["2103.14024", "2103.14645", "2111.11215", "2203.09517", "2302.12249", "2208.00277"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_6711"} +{"question": "Which papers extended the variance-weighted regression technique to achieve minimax optimality?", "answer": ["Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes"], "answer_arxiv_id": ["2206.11489", "2212.06132"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_6712"} +{"question": "What works are about applying differential volume rendering in learning 3D shape from 2D images?", "answer": ["Differentiable Volumetric Rendering: Learning Implicit 3D\n Representations without 3D Supervision", "HumanNeRF: Efficiently Generated Human Radiance Field from Sparse Inputs", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video"], "answer_arxiv_id": ["1912.07372", "2112.02789", "2201.04127"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_6713"} +{"question": "What research has been done on using predictions to improve metrics such as time, space, and communication complexity?", "answer": ["Competitive caching with machine learned advice", "Learning Predictions for Algorithms with Predictions"], "answer_arxiv_id": ["1802.05399", "2202.09312"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_6714"} +{"question": "What research work has introduced pixel-level reconstruction to tackle the modal-missing problem?", "answer": ["Dynamic Enhancement Network for Partial Multi-modality Person\n Re-identification"], "answer_arxiv_id": ["2305.15762"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_6715"} +{"question": "What paper pioneered the approach of training various natural language processing tasks in a unified text-to-text format?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer"], "answer_arxiv_id": ["1910.10683"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_6716"} +{"question": "Which paper first introduced the application of diffusion models to continuous text space?", "answer": ["Diffusion-LM Improves Controllable Text Generation"], "answer_arxiv_id": ["2205.14217"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_6717"} +{"question": "Which works relate model extraction attacks to knowledge distillation?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6718"} +{"question": "What studies revised the novel class discovery approach and proposed the Generalized Novel Class Discovery setting?", "answer": ["Generalized Category Discovery"], "answer_arxiv_id": ["2201.02609"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_6719"} +{"question": "Any works presented optimization algorithms which balance the scale of bias and variance for DR in OPE?", "answer": ["Doubly Robust Bias Reduction in Infinite Horizon Off-Policy Estimation"], "answer_arxiv_id": ["1910.07186"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_6720"} +{"question": "Which paper introduced improvements to both the testing algorithm and the oracle in graph clustering?", "answer": ["Testing Graph Clusterability: Algorithms and Lower Bounds", "Spectral Clustering Oracles in Sublinear Time"], "answer_arxiv_id": ["1808.04807", "2101.05549"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_6721"} +{"question": "Which work generalizes information theory to multiple variables?", "answer": ["Nonnegative Decomposition of Multivariate Information"], "answer_arxiv_id": ["1004.2515"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_6722"} +{"question": "Which papers introduced the concept of robust top-k accuracy?", "answer": ["Relaxing Local Robustness"], "answer_arxiv_id": ["2106.06624"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_6723"} +{"question": "Which works propose models for obtaining generic representations with minimal dependence on human labels in self-supervised learning?", "answer": ["Unsupervised Learning of Visual Representations using Videos", "Unsupervised Visual Representation Learning by Context Prediction", "Context Encoders: Feature Learning by Inpainting", "Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles", "Colorful Image Colorization", "Unsupervised Representation Learning by Predicting Image Rotations", "Unsupervised Feature Learning via Non-Parametric Instance Discrimination"], "answer_arxiv_id": ["1505.00687", "1505.05192", "1604.07379", "1603.09246v3", "1603.08511", "1803.07728", "1805.01978"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_6724"} +{"question": "Are there any papers that discuss interpretability methods in reinforcement learning?", "answer": ["Explainable Reinforcement Learning: A Survey"], "answer_arxiv_id": ["2005.06247"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_6725"} +{"question": "What research considers learning symmetry from training data only?", "answer": ["Learning Invariances in Neural Networks", "Residual Pathway Priors for Soft Equivariance Constraints", "Generative Adversarial Symmetry Discovery", "Relaxing Equivariance Constraints with Non-stationary Continuous Filters"], "answer_arxiv_id": ["2010.11882", "2112.01388", "2302.00236", "2204.07178"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_6726"} +{"question": "Which works revealed substantial vulnerabilities in Federated Learning indicating how it can infer certain properties of the clients’ data?", "answer": ["Exploiting Unintended Feature Leakage in Collaborative Learning∗"], "answer_arxiv_id": ["1805.04049"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_6727"} +{"question": "Could you provide me some works that used transformers to estimate treatment effects over time?", "answer": ["Causal Transformer for Estimating Counterfactual Outcomes"], "answer_arxiv_id": ["2204.07258"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6728"} +{"question": "Which study probed representations to look into the memory abilities of agents?", "answer": ["Shaping Belief States with Generative Environment Models for RL"], "answer_arxiv_id": ["1906.09237"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_6729"} +{"question": "What works have adopted Instance Discrimination for aligning and distinguishing representations in contrastive representation learning?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance Discrimination"], "answer_arxiv_id": ["1805.01978"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_6730"} +{"question": "Could you indicate the studies that proposed to learn shared vision-language representations?", "answer": ["Learning Visual Features from Large Weakly Supervised Data", "VirTex: Learning Visual Representations from Textual Annotations", "Learning Transferable Visual Models From Natural Language Supervision", "Exploring the Limits of Weakly Supervised Pretraining"], "answer_arxiv_id": ["1511.02251", "2006.06666", "2103.00020", "1805.00932"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_6731"} +{"question": "Which works introduced Markov convex game (MCG)?", "answer": ["Shapley Q-value: A Local Reward Approach to Solve Global Reward Games", "SHAQ: Incorporating Shapley Value Theory into Multi-Agent Q-Learning"], "answer_arxiv_id": ["1907.05707", "2105.15013"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_6732"} +{"question": "Which works discuss the autoregressive model?", "answer": ["SimulLR: Simultaneous Lip Reading Transducer with Attention-Guided Adaptive Memory", "SimulSLT: End-to-End Simultaneous Sign Language Translation"], "answer_arxiv_id": ["2108.13630", "2112.04228"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_6733"} +{"question": "What are the recent advances in self-supervised pretraining?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Big Self-Supervised Models are Strong Semi-Supervised Learners", "An Empirical Study of Training Self-Supervised Vision Transformers", "ReSSL: Relational Self-Supervised Learning with Weak Augmentation", "Similarity Contrastive Estimation for Self-Supervised Soft Contrastive Learning"], "answer_arxiv_id": ["2006.07733", "2006.09882", "2006.10029", "2104.02057", "2107.09282", "2111.14585"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_6734"} +{"question": "What studies propose deep equilibrium model with proximal gradient descent for inverse problems in imaging?", "answer": ["Deep Equilibrium Architectures for Inverse Problems in Imaging"], "answer_arxiv_id": ["2102.07944"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_6735"} +{"question": "What works use variational autoencoders, autoregressive hidden Markov models, and UMAP to characterize the latent structure of behavior in behavioral representation learning?", "answer": ["Auto-Encoding Variational Bayes", "UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction"], "answer_arxiv_id": ["1312.6114", "1802.03426"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_6736"} +{"question": "Could you name studies that used more sophisticated models such as GMMs or PCAs for modeling robust pose prior?", "answer": ["Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a\n Single Image"], "answer_arxiv_id": ["1607.08128"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6737"} +{"question": "Which work proposed the Offset equivariant networks that become equivariant to an additive bias to the RGB input channels?", "answer": ["Offset equivariant networks and their applications"], "answer_arxiv_id": ["2207.00292"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_6738"} +{"question": "What work proposes foundational models that can operate across diverse modalities and domains?", "answer": ["On the Opportunities and Risks of Foundation Models"], "answer_arxiv_id": ["2108.07258v3"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_6739"} +{"question": "Does any paper point out how Segment Anything Model (SAM) contributes to promptable segmentation and its influence in many downstream tasks?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "Segment Anything Meets Point Tracking", "Tracking Anything with Decoupled Video Segmentation", "Segment Everything Everywhere All at Once"], "answer_arxiv_id": ["2401.14159", "2307.01197", "2309.03903", "2304.06718"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_6740"} +{"question": "Which researches have adopted a Transformer-based architecture for predicting interactions between humans and objects?", "answer": ["Reformulating HOI Detection as Adaptive Set Prediction", "HOTR: End-to-End Human-Object Interaction Detection with Transformers", "QPIC: Query-Based Pairwise Human-Object Interaction Detection with Image-Wide Contextual Information", "End-to-End Human Object Interaction Detection with HOI Transformer", "Mining the Benefits of Two-stage and One-stage HOI Detection", "MSTR: Multi-Scale Transformer for End-to-End Human-Object Interaction Detection", "Efficient Two-Stage Detection of Human–Object Interactions with a Novel Unary–Pairwise Transformer", "Human-Object Interaction Detection via Disentangled Transformer", "GEN-VLKT: Simplify Association and Enhance Interaction Understanding for HOI Detection", "Interactiveness Field in Human-Object Interactions", "RLIP: Relational Language-Image Pre-training for Human-Object Interaction Detection", "Towards Hard-Positive Query Mining for DETR-based Human-Object Interaction Detection"], "answer_arxiv_id": ["2103.05983", "2104.13682", "2103.05399", "2103.04503", "2108.05077", "2203.14709", "2112.01838", "2204.09290", "2203.13954", "2204.07718v1", "2209.01814", "2207.05293"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_6741"} +{"question": "In what studies the equivariant architectures have been extended to more general data types under arbitrary finite group and matrix group symmetries?", "answer": ["Equivariance Through Parameter-Sharing", "A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups"], "answer_arxiv_id": ["1702.08389", "2104.09459"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_6742"} +{"question": "What works proposed to adapt foundation models by prompting, instead of full fine-tuning?", "answer": ["On the Opportunities and Risks of Foundation Models"], "answer_arxiv_id": ["2108.07258v3"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_6743"} +{"question": "What studies represent recent advances in using Deep Learning for OCC?", "answer": ["Classification-Based Anomaly Detection for General Data", "Deep Anomaly Detection Using Geometric Transformations", "DROCC: Deep Robust One-Class Classification", "Adversarially Learned Anomaly Detection", "Adversarially Learned One-Class Classifier for Novelty Detection"], "answer_arxiv_id": ["2005.02359", "1805.10917", "2002.12718", "1812.02288", "1802.09088"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_6744"} +{"question": "What works have started to build self-supervised 3D representation learning on scene-centric data?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts", "Self-Supervised Pretraining of 3D Features on any Point-Cloud", "Spatio-temporal Self-Supervised Representation Learning for 3D Point\n Clouds", "Masked Scene Contrast: A Scalable Framework for Unsupervised 3D\n Representation Learning", "UniPAD: A Universal Pre-training Paradigm for Autonomous Driving", "PonderV2: Pave the Way for 3D Foundation Model with A Universal\n Pre-training Paradigm"], "answer_arxiv_id": ["2007.10985", "2012.09165", "2101.02691", "2109.00179", "2303.14191", "2310.08370", "2310.08586"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_6745"} +{"question": "What are the works that studied the effect of width on generalization in the feature-learning regime?", "answer": ["The Onset of Variance-Limited Behavior for Networks in the Lazy and Rich Regimes"], "answer_arxiv_id": ["2212.12147"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_6746"} +{"question": "What papers study MOBO problem in the setting with preferences over the different objectives?", "answer": ["Multi-objective Bayesian optimisation with preferences over objectives", "Preference Exploration for Efficient Bayesian Optimization with Multiple Outcomes"], "answer_arxiv_id": ["1902.04228", "2203.11382v1"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_6747"} +{"question": "Could you name the work that uses the mean of attention matrix column values of transformers to determine the importance of each token for pruning?", "answer": ["Learned Token Pruning for Transformers"], "answer_arxiv_id": ["2107.00910"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_6748"} +{"question": "Are there any studies that utilize Large Language Models (LLMs) in sequence modelling and transformers?", "answer": ["Attention Is All You Need", "Neural Machine Translation by Jointly Learning to Align and Translate", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1706.03762", "1409.0473", "1810.04805"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_6749"} +{"question": "Which work proposed to transform the problem into a narrative cloze task, creating synthetic datasets for implicit SRL?", "answer": ["Implicit Argument Prediction with Event Knowledge"], "answer_arxiv_id": ["1802.07226"], "source_meta": {"published_time": "20240808"}, "qid": "AutoScholarQuery_train_6750"} +{"question": "Which researches focus on image-level out-of-distribution detection?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "Deep Anomaly Detection with Outlier Exposure", "Energy-based Out-of-distribution Detection", "ReAct: Out-of-distribution Detection With Rectified Activations"], "answer_arxiv_id": ["1610.02136", "1706.02690", "1807.03888", "1812.04606", "2010.03759", "2111.12797"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_6751"} +{"question": "Which works discuss the time and space complexity challenges faced by autoregressive models for large-scale NPC problems?", "answer": ["Attention, Learn to Solve Routing Problems!", "Attention Is All You Need"], "answer_arxiv_id": ["1803.08475", "1706.03762"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_6752"} +{"question": "Could you provide me some works that extend trust region learning to a cooperative MARL setting?", "answer": ["Trust Region Policy Optimisation in Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["2109.11251"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_6753"} +{"question": "What studies use ResNet as the backbone for computing face images in face recognition?", "answer": ["Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1512.03385"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_6754"} +{"question": "Which works view each hyperedge as a multi-set of nodes and each node as a multi-set of hyperedges?", "answer": ["HNHN: Hypergraph Networks with Hyperedge Neurons", "UniGNN: a Unified Framework for Graph and Hypergraph Neural Networks", "You are AllSet: A Multiset Function Framework for Hypergraph Neural Networks", "Hypergraph Convolution and Hypergraph Attention", "HyperSAGE: Generalizing Inductive Representation Learning on Hypergraphs", "Edge Representation Learning with Hypergraphs"], "answer_arxiv_id": ["2006.12278", "2105.00956", "2106.13264v4", "1901.08150", "2010.04558", "2106.15845"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_6755"} +{"question": "Can you name some works that use neural differentiable rendering techniques in 3D diffusion models?", "answer": ["DiffRF: Rendering-Guided 3D Radiance Field Diffusion", "Texture Generation on 3D Meshes with Point-UV Diffusion"], "answer_arxiv_id": ["2212.01206", "2308.10490"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_6756"} +{"question": "What studies have been undertaken to improve the standard Chain-of-thought (CoT) in terms of self-consistency?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_6757"} +{"question": "Which studies propose multimodal foundation models that leverage LLMs?", "answer": ["LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "DetGPT: Detect What You Need via Reasoning", "Visual Instruction Tuning", "Kosmos-2: Grounding Multimodal Large Language Models to the World"], "answer_arxiv_id": ["2304.15010", "2301.12597", "2305.14167", "2304.08485", "2306.14824"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_6758"} +{"question": "What work initially proposed to generate Key Points (KPs) for each argument?", "answer": ["Team Enigma at ArgMining-EMNLP 2021: Leveraging Pre-trained Language\n Models for Key Point Matching"], "answer_arxiv_id": ["2110.12370"], "source_meta": {"published_time": "20240719"}, "qid": "AutoScholarQuery_train_6759"} +{"question": "What paper introduced the standard invariant regularization to enforce invariance of data augmentations in causal graphs of SCL?", "answer": ["Representation Learning via Invariant Causal Mechanisms"], "answer_arxiv_id": ["2010.07922"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_6760"} +{"question": "What works have contributed to studies on two-layer mean field networks trained online with Gaussian data?", "answer": ["Phase diagram of Stochastic Gradient Descent in high-dimensional two-layer neural networks"], "answer_arxiv_id": ["2202.00293"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_6761"} +{"question": "Can you mention some references that describe voxel-based methods for 3D perception?", "answer": ["Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detection", "M3DETR: Multi-representation, Multi-scale, Mutual-relation 3D Object Detection with Transformers", "PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection", "PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detection", "From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network", "CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Clouds", "Center-based 3D Object Detection and Tracking"], "answer_arxiv_id": ["2012.15712", "2104.11896", "1912.13192", "2102.00463", "1907.03670", "2210.04264", "2006.11275"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_6762"} +{"question": "What works focus on training networks to learn occupancy functions or distance functions in the context of implicit neural functions?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "Convolutional Occupancy Networks", "DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation", "Implicit Neural Representations with Periodic Activation Functions", "SAL: Sign Agnostic Learning of Shapes from Raw Data"], "answer_arxiv_id": ["1812.03828", "2003.04618", "1901.05103", "2006.09661", "1911.10414"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_6763"} +{"question": "Is there any work that proposed a version of Network Dissection which doesn't require dense annotations?", "answer": ["Understanding the Role of Individual Units in a Deep Neural Network"], "answer_arxiv_id": ["2009.05041"], "source_meta": {"published_time": "20220423"}, "qid": "AutoScholarQuery_train_6764"} +{"question": "How were pyramid structures in Transofmers first introduced ?", "answer": ["Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction\n without Convolutions"], "answer_arxiv_id": ["2102.12122"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_6765"} +{"question": "Could you provide me names of the research works that leverage reconstruction error from the diffusion model for diffusion-generated image detection?", "answer": ["DIRE for Diffusion-Generated Image Detection", "Exposing the Fake: Effective Diffusion-Generated Images Detection"], "answer_arxiv_id": ["2303.09295", "2307.06272"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_6766"} +{"question": "Which works propose the use of non-convex surrogate losses in the adversarial setting?", "answer": ["Calibrated Surrogate Losses for Adversarially Robust Classification", "Calibration and Consistency of Adversarial Surrogate Losses", "A Finer Calibration Analysis for Adversarial Robustness"], "answer_arxiv_id": ["2005.13748", "2104.09658", "2105.01550v2"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_6767"} +{"question": "Which works focused on learning optimal decision trees?", "answer": ["Learning Optimal and Fair Decision Trees for Non-Discriminative Decision-Making", "A Scalable MIP-based Method for Learning Optimal Multivariate Decision Trees", "Optimal Sparse Decision Trees", "Generalized and Scalable Optimal Sparse Decision Trees", "MurTree: Optimal Decision Trees via Dynamic Programming and Search"], "answer_arxiv_id": ["1903.10598", "2011.03375", "1904.12847", "2006.08690", "2007.12652"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_6768"} +{"question": "What research works used fine-tuning strategy in the arithmetic field?", "answer": ["WizardMath: Empowering Mathematical Reasoning for Large Language Models\n via Reinforced Evol-Instruct"], "answer_arxiv_id": ["2308.09583"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_6769"} +{"question": "Which works in knowledge distillation algorithms use relation-based methods?", "answer": ["Relational Knowledge Distillation", "Correlation Congruence for Knowledge Distillation", "Knowledge Distillation for Object Detection via Rank Mimicking and\n Prediction-guided Feature Imitation", "Similarity-Preserving Knowledge Distillation", "Knowledge Distillation from A Stronger Teacher"], "answer_arxiv_id": ["1904.05068", "1904.01802", "2112.04840", "1907.09682", "2205.10536"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_6770"} +{"question": "Any works that discuss applying generative models to UQ?", "answer": ["Uncertainty Quantification with Generative Models", "Uncertainty-Aware Deep Classifiers using Generative Models", "GFlowOut: Dropout with Generative Flow Networks"], "answer_arxiv_id": ["1910.10046", "2006.04183", "2210.12928"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_6771"} +{"question": "Can you name the studies that incorporate MoE layers into Transformers?", "answer": ["GShard: Scaling Giant Models with Conditional Computation and Automatic\n Sharding"], "answer_arxiv_id": ["2006.16668"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_6772"} +{"question": "Could you provide me a study that compares the performance of GPT4-32k and junior lawyers in locating legal issues in contracts?", "answer": ["Better Call GPT, Comparing Large Language Models Against Lawyers"], "answer_arxiv_id": ["2401.16212"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_6773"} +{"question": "Who proposed LAVA as a scalable solution that evaluates data influence on the model performance?", "answer": ["LAVA: Data Valuation without Pre-Specified Learning Algorithms"], "answer_arxiv_id": ["2305.00054"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_6774"} +{"question": "What works used dynamic regret in online convex optimization to evaluate the performance of a learner?", "answer": ["Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient", "Online Optimization : Competing with Dynamic Comparators", "Non-stationary Stochastic Optimization", "Tracking Moving Agents via Inexact Online Gradient Descent Algorithm"], "answer_arxiv_id": ["1605.04638", "1501.06225", "1307.5449", "1710.05133v2"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_6775"} +{"question": "Could you provide me some works that adapt the transformer model scaling laws for related settings?", "answer": ["Data Scaling Laws in NMT: The Effect of Noise and Architecture", "Unified Scaling Laws for Routed Language Models", "Explaining Neural Scaling Laws"], "answer_arxiv_id": ["2202.01994", "2202.01169", "2102.06701"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_6776"} +{"question": "Which research paper creates a modular network that includes modules to attend to subjects, locations, and relationships?", "answer": ["MAttNet: Modular Attention Network for Referring Expression Comprehension"], "answer_arxiv_id": ["1801.08186"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_6777"} +{"question": "Which research papers have introduced a token pruning strategy for ViTs?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "SPViT: Enabling Faster Vision Transformers via Soft Token Pruning", "AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "Adaptive Token Sampling For Efficient Vision Transformers", "Not All Patches are What You Need: Expediting Vision Transformers via\n Token Reorganizations", "Zero-TPrune: Zero-Shot Token Pruning through Leveraging of the Attention\n Graph in Pre-Trained Transformers", "AdaViT: Adaptive Tokens for Efficient Vision Transformer"], "answer_arxiv_id": ["2106.02034v2", "2112.13890", "2111.15668", "2111.15667", "2202.07800", "2305.17328", "2112.07658"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_6778"} +{"question": "Who used the DV formula with stability assumptions to derive some PAC-Bayes bounds?", "answer": ["A PAC-Bayesian Analysis of Randomized Learning with Application to Stochastic Gradient Descent"], "answer_arxiv_id": ["1709.06617"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_6779"} +{"question": "Which works focus on linear MDPs in the field of single-agent RL with linear function approximation?", "answer": ["Sample-Optimal Parametric Q-Learning Using Linearly Additive Features", "Provably Efficient Reinforcement Learning with Linear Function Approximation", "Learning Near Optimal Policies with Low Inherent Bellman Error", "A Unifying View of Optimism in Episodic Reinforcement Learning", "Logarithmic Regret for Reinforcement Learning with Linear Function Approximation", "Provably Efficient Reinforcement Learning with Linear Function Approximation under Adaptivity Constraints", "Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes", "Pessimism in the Face of Confounders: Provably Efficient Offline Reinforcement Learning in Partially Observable Markov Decision Processes"], "answer_arxiv_id": ["1902.04779", "1907.05388", "2003.00153", "2007.01891", "2011.11566", "2101.02195", "2206.11489", "2212.06132", "2205.13589"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_6780"} +{"question": "What are some papers which have contributed to self-supervised pre-training benefiting downstream tasks?", "answer": ["Transferability in Deep Learning: A Survey"], "answer_arxiv_id": ["2201.05867"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_6781"} +{"question": "Which research work use specified slicing functions in the importance weighting for more accurate estimation?", "answer": ["Mandoline: Model Evaluation under Distribution Shift"], "answer_arxiv_id": ["2107.00643"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_6782"} +{"question": "Which works provide a review or discussion on the connection between DDMs and previous iterative denoising methods?", "answer": ["Elucidating the Design Space of Diffusion-Based Generative Models", "Image Denoising: The Deep Learning Revolution and Beyond -- A Survey\n Paper --"], "answer_arxiv_id": ["2206.00364v2", "2301.03362"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_6783"} +{"question": "In what studies did researchers propose a bisimulation metric?", "answer": ["Metrics for Finite Markov Decision Processes"], "answer_arxiv_id": ["1207.4114v1"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_6784"} +{"question": "According to which studies are IPCW Brier score and integrated Brier score used in survival analysis?", "answer": ["Time-to-Event Prediction with Neural Networks and Cox Regression", "Effective Ways to Build and Evaluate Individual Survival Distributions", "Inverse-Weighted Survival Games"], "answer_arxiv_id": ["1907.00825", "1811.11347", "2111.08175"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_6785"} +{"question": "Which works propose combining Large Language Models with a visual-language model and a pre-trained language-conditioned robot policy to perform tasks?", "answer": ["Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language"], "answer_arxiv_id": ["2204.00598"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_6786"} +{"question": "Could you name studies that explored multi-task self-supervised learning in the field of computer vision?", "answer": ["12-in-1: Multi-Task Vision and Language Representation Learning", "Multi-task Self-Supervised Visual Learning", "Cross-Domain Self-supervised Multi-task Feature Learning using Synthetic Imagery", "BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning", "M3P: Learning Universal Representations via Multitask Multilingual Multimodal Pre-training"], "answer_arxiv_id": ["1912.02315", "1708.07860", "1711.09082", "1805.04687", "2006.02635"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_6787"} +{"question": "Which works propose interpolation-based Mixup methods for graph augmentation?", "answer": ["Graph Transplant: Node Saliency-Guided Graph Mixup with Local Structure Preservation", "Model-Agnostic Augmentation for Accurate Graph Classification", "G-Mixup: Graph Data Augmentation for Graph Classification"], "answer_arxiv_id": ["2111.05639", "2202.10107", "2202.07179v2"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_6788"} +{"question": "Could you provide the work that proposes the choice of g as D2ϕ when M is a convex body?", "answer": ["Sampling with Riemannian Hamiltonian Monte Carlo in a Constrained Space"], "answer_arxiv_id": ["2202.01908"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_6789"} +{"question": "Which papers deal with LLM-assisted generations and how they can help in improving text prompt expressiveness?", "answer": ["Investigating Prompt Engineering in Diffusion Models", "A Taxonomy of Prompt Modifiers for Text-To-Image Generation", "Design Guidelines for Prompt Engineering Text-to-Image Generative Models", "Zero-shot Generation of Coherent Storybook from Plain Text Story using Diffusion Models", "LoGoPrompt: Synthetic Text Images Can Be Good Visual Prompts for Vision-Language Models"], "answer_arxiv_id": ["2211.15462", "2204.13988", "2109.06977", "2302.03900", "2309.01155"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_6790"} +{"question": "What works are about the problem of severe discriminator overfitting in GANs training?", "answer": ["FakeCLR: Exploring Contrastive Learning for Solving Latent Discontinuity in Data-Efficient GANs", "A Comprehensive Survey on Data-Efficient GANs in Image Generation"], "answer_arxiv_id": ["2207.08630", "2204.08329"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6791"} +{"question": "What papers demonstrated the effectiveness of using pre-trained large language models with prompting for text classification tasks?", "answer": ["Making Pre-trained Language Models Better Few-shot Learners"], "answer_arxiv_id": ["2012.15723"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_6792"} +{"question": "Any papers around the extension of the forward diffusion process to general signal degradation?", "answer": ["Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise", "Progressive Deblurring of Diffusion Models for Coarse-to-Fine Image\n Synthesis", "Blurring Diffusion Models"], "answer_arxiv_id": ["2208.09392", "2207.11192", "2209.05557"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6793"} +{"question": "Any works where reverse generation from the answer to the given conditions is applied in machine translation and reinforcement learning?", "answer": ["Understanding Back-Translation at Scale", "Learning to Navigate in a VUCA Environment: Hierarchical Multi-expert\n Approach", "A Closed-Loop Perception, Decision-Making and Reasoning Mechanism for\n Human-Like Navigation"], "answer_arxiv_id": ["1808.09381", "2111.08364", "2207.11901"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_6794"} +{"question": "Which papers focused on reprogramming a black-box model for downstream classification tasks with limited data?", "answer": ["Transfer Learning without Knowing: Reprogramming Black-box Machine Learning Models with Scarce Data and Limited Resources"], "answer_arxiv_id": ["2007.08714"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_6795"} +{"question": "What is the work that computes the R-D function by solving a version of the EOT problem in a finite and known alphabet setting?", "answer": ["A Communication Optimal Transport Approach to the Computation of Rate Distortion Functions"], "answer_arxiv_id": ["2212.10098v1"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_6796"} +{"question": "What works leverage the advancements of DETR in object detection for action localization?", "answer": ["STMixer: A One-Stage Sparse Action Detector", "Efficient Video Action Detection with Token Dropout and Context\n Refinement"], "answer_arxiv_id": ["2303.15879", "2304.08451"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_6797"} +{"question": "What studies extract 2D skeleton from dance videos using 2D pose estimation?", "answer": ["Dancing to Music"], "answer_arxiv_id": ["1911.02001"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_6798"} +{"question": "Can you name studies that noted the occurrence of oversmoothing in GNNs with as few as 2 to 4 layers?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "A Comprehensive Survey on Graph Neural Networks"], "answer_arxiv_id": ["1609.02907", "1901.00596"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_6799"} +{"question": "Could you provide me some works that studied smooth nonconvex-concave min-max optimization?", "answer": ["Solving a Class of Non-Convex Min-Max Games Using Iterative First Order Methods", "An accelerated inexact proximal point method for solving nonconvex-concave min-max problems", "On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems", "Efficient Search of First-Order Nash Equilibria in Nonconvex-Concave Smooth Min-Max Problems"], "answer_arxiv_id": ["1902.08297", "1905.13433", "1906.00331", "2002.07919"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_6800"} +{"question": "Could you list some researches on the application of Diffusion Probabilistic Models in generating images, text, and molecules?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Diffusion-LM Improves Controllable Text Generation", "Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2105.05233", "2205.14217", "2203.17003"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_6801"} +{"question": "Which papers focused on generating extra triplet data using generative models in the context of Composed Image Retrieval?", "answer": ["CompoDiff: Versatile Composed Image Retrieval With Latent Diffusion", "Zero-shot Composed Text-Image Retrieval"], "answer_arxiv_id": ["2303.11916", "2306.07272"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_6802"} +{"question": "What works proposed various data structures for approximate membership queries with better memory efficiency than the original Bloom filter?", "answer": ["Xor Filters: Faster and Smaller Than Bloom and Cuckoo Filters", "Ribbon filter: practically smaller than Bloom and Xor1"], "answer_arxiv_id": ["1912.08258", "2103.02515"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_6803"} +{"question": "What papers present and study the technique of supervised fine-tuning (SFT) used in LLM alignment?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Cross-Task Generalization via Natural Language Crowdsourcing\n Instructions"], "answer_arxiv_id": ["2109.01652", "2104.08773"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_6804"} +{"question": "Which studies indicate that PnP methods based on least squares (LS) gradient steps require many iterations?", "answer": ["The Little Engine that Could: Regularization by Denoising (RED)"], "answer_arxiv_id": ["1611.02862"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_6805"} +{"question": "Could you name some studies that employed different 3D representations for learning-based reconstruction methods?", "answer": ["3D ShapeNets: A Deep Representation for Volumetric Shapes", "Learning a Predictable and Generative Vector Representation for Objects", "3D-R2N2: A Unified Approach for Single and Multi-view 3D Object\n Reconstruction", "Pix2Vox: Context-aware 3D Reconstruction from Single and Multi-view\n Images", "A Point Set Generation Network for 3D Object Reconstruction from a\n Single Image", "Dense 3D Point Cloud Reconstruction Using a Deep Pyramid Network", "Category-Specific Object Reconstruction from a Single Image", "Learning Category-Specific Mesh Reconstruction from Image Collections", "CASA: Category-agnostic Skeletal Animal Reconstruction", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization", "Implicit Functions in Feature Space for 3D Shape Reconstruction and\n Completion"], "answer_arxiv_id": ["1406.5670", "1603.08637", "1604.00449", "1901.11153", "1612.00603", "1901.08906", "1411.6069", "1803.07549", "2211.03568", "1812.03828", "1905.05172", "2003.01456"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6806"} +{"question": "Which works propose unsupervised methods based on white-box features for detecting machine-generated text?", "answer": ["Release Strategies and the Social Impacts of Language Models", "Automatic Detection of Generated Text is Easiest when Humans are Fooled", "DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability\n Curvature", "DetectLLM: Leveraging Log Rank Information for Zero-Shot Detection of\n Machine-Generated Text", "MGTBench: Benchmarking Machine-Generated Text Detection", "Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated\n Text"], "answer_arxiv_id": ["1908.09203v2", "1911.00650", "2301.11305", "2306.05540", "2303.14822", "2401.12070"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_6807"} +{"question": "What papers proposed to use a signed distance function to recover a scene's surface in 3D reconstruction?", "answer": ["Multiview Neural Surface Reconstruction by Disentangling Geometry and\n Appearance", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "HF-NeuS: Improved Surface Reconstruction Using High-Frequency Details", "Improving neural implicit surfaces geometry with patch warping", "Humans as Light Bulbs: 3D Human Reconstruction from Thermal Reflection", "MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface\n Reconstruction", "NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in\n the Wild"], "answer_arxiv_id": ["2003.09852", "2106.10689", "2106.12052", "2206.07850", "2112.09648", "2305.01652", "2206.00665", "2110.07604"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_6808"} +{"question": "Could you provide me some papers that focus on reinforcement learning in LQR systems?", "answer": ["Finite-time Analysis of Approximate Policy Iteration for the Linear Quadratic Regulator", "Least-Squares Temporal Difference Learning for the Linear Quadratic Regulator", "On the Sample Complexity of the Linear Quadratic Regulator", "The Gap Between Model-Based and Model-Free Methods on the Linear Quadratic Regulator: An Asymptotic Viewpoint", "Regret Bounds for Robust Adaptive Control of the Linear Quadratic Regulator", "Global Convergence of Policy Gradient Methods for the Linear Quadratic Regulator", "Continuous Deep Q-Learning with Model-based Acceleration"], "answer_arxiv_id": ["1905.12842", "1712.08642", "1710.01688", "1812.03565", "1805.09388", "1801.05039", "1603.00748"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_6809"} +{"question": "Could you cite some research works demonstrating that the top eigenfunctions of the kernel align with the target function learned by the NN?", "answer": ["Neural Spectrum Alignment: Empirical Study", "Neural Anisotropy Directions", "What can linearized neural networks actually say about generalization?"], "answer_arxiv_id": ["1910.08720", "2006.09717", "2106.06770"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_6810"} +{"question": "Any recent works that demonstrate the connection between robust statistical estimator and differential privacy?", "answer": ["Robustness Implies Privacy in Statistical Estimation"], "answer_arxiv_id": ["2212.05015v3"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_6811"} +{"question": "What papers introduced fine-tuning free methods for addressing length extrapolation in LLMs?", "answer": ["YaRN: Efficient Context Window Extension of Large Language Models"], "answer_arxiv_id": ["2309.00071"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_6812"} +{"question": "In which papers was output perturbation and objective perturbation approaches analyzed for convex optimization problems?", "answer": ["Differentially Private Empirical Risk Minimization"], "answer_arxiv_id": ["0912.0071v5"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_6813"} +{"question": "What works have studied hard-to-process living languages and compared non-contextual baselines with pre-trained LM-based methods?", "answer": ["When Being Unseen from mBERT is just the Beginning: Handling New\n Languages With Multilingual Language Models"], "answer_arxiv_id": ["2010.12858"], "source_meta": {"published_time": "20240808"}, "qid": "AutoScholarQuery_train_6814"} +{"question": "Which papers discuss the inductive biases of Stochastic Gradient Descent (SGD) and structured architectures in deep learning?", "answer": ["In Search of the Real Inductive Bias: On the Role of Implicit\n Regularization in Deep Learning", "Inductive Bias of Deep Convolutional Networks through Pooling Geometry", "What Algorithms can Transformers Learn? A Study in Length Generalization"], "answer_arxiv_id": ["1412.6614", "1605.06743", "2310.16028"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_6815"} +{"question": "What research works examine the failure of LLMs in reasoning about real world dynamics without grounding?", "answer": ["Experience Grounds Language"], "answer_arxiv_id": ["2004.10151"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_6816"} +{"question": "Could you provide me with studies related to memory skills of autonomous agents?", "answer": ["Generative Agents: Interactive Simulacra of Human Behavior", "Cognitive Architectures for Language Agents"], "answer_arxiv_id": ["2304.03442", "2309.02427"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_6817"} +{"question": "Can you provide a reference where dropout during the inference phase is enabled to obtain predictive uncertainty?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.02142"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_6818"} +{"question": "Could you tell me what papers extended best arm identification in CoPE to best m arm identification?", "answer": ["Collaborative Top Distribution Identifications with Limited Interaction"], "answer_arxiv_id": ["2004.09454v2"], "source_meta": {"published_time": "20211029"}, "qid": "AutoScholarQuery_train_6819"} +{"question": "What studies have focused on renovating diffusion models by focusing on architectures, scheduling, weighting, and fast sampling?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Perception Prioritized Training of Diffusion Models", "Denoising Diffusion Implicit Models", "Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality"], "answer_arxiv_id": ["2102.09672", "2206.00364", "2204.00227", "2010.02502", "2202.05830"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_6820"} +{"question": "What are some research papers about vision-language co-training approaches?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks"], "answer_arxiv_id": ["2103.00020", "2208.10442"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_6821"} +{"question": "Which works introduced techniques like multi-resolution grids or explored Instant-NGP and TensoRF for efficient mapping?", "answer": ["NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural\n Real-Time SLAM", "ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of\n Signed Distance Fields"], "answer_arxiv_id": ["2112.12130", "2304.14377", "2211.11704"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6822"} +{"question": "Could you provide me some works that explored temporal VAE framework and transformer architecture in motion generation?", "answer": ["Action2video: Generating Videos of Human 3D Actions", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE"], "answer_arxiv_id": ["2111.06925", "2104.05670"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_6823"} +{"question": "What studies dealt with dynamic human modeling from sparse multi-view cameras or a monocular camera using NeRF?", "answer": ["Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "HumanNeRF: Efficiently Generated Human Radiance Field from Sparse Inputs", "Tensor4D : Efficient Neural 4D Decomposition for High-fidelity Dynamic\n Reconstruction and Rendering", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "NeuMan: Neural Human Radiance Field from a Single Video", "Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via\n Self-supervised Scene Decomposition"], "answer_arxiv_id": ["2012.15838", "2112.02789", "2211.11610", "2201.04127", "2203.12575", "2302.11566"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6824"} +{"question": "Which works document the failure modes of large language models, including propagating biases and stereotypes?", "answer": ["The Woman Worked as a Babysitter: On Biases in Language Generation", "StereoSet: Measuring stereotypical bias in pretrained language models", "Investigating African-American Vernacular English in Transformer-Based Text Generation", "Persistent Anti-Muslim Bias in Large Language Models", "Debiased Large Language Models Still Associate Muslims with Uniquely Violent Acts"], "answer_arxiv_id": ["1909.01326", "2004.09456", "2010.02510", "2101.05783", "2208.04417v2"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_6825"} +{"question": "What studies proposed histology-based survival models utilizing Multiple Instance Learning (MIL)?", "answer": ["Whole Slide Images based Cancer Survival Prediction using Attention\n Guided Deep Multiple Instance Learning Networks"], "answer_arxiv_id": ["2009.11169"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_6826"} +{"question": "What studies addressed the problem of ETF structure collapsing in imbalanced datasets?", "answer": ["Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training"], "answer_arxiv_id": ["2101.12699"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_6827"} +{"question": "Which paper first introduced the Denoising Diffusion Probabilistic Model?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_6828"} +{"question": "Which studies alternatively address the problem formulation of recovery of the cluster tree of the probability density from which data are sampled?", "answer": ["Stability of Density-Based Clustering", "Beyond Hartigan Consistency: Merge Distortion Metric for Hierarchical Clustering", "Statistical Inference for Cluster Trees"], "answer_arxiv_id": ["1011.2771", "1506.06422", "1605.06416"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_6829"} +{"question": "What work applies the parallel Kalman smoother in SVAE?", "answer": ["Temporal Parallelization of Bayesian Smoothers"], "answer_arxiv_id": ["1905.13002"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_6830"} +{"question": "Could you provide me with some works that use SBP for generative modeling?", "answer": ["Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling", "Deep Generative Learning via Schrödinger Bridge"], "answer_arxiv_id": ["2106.01357", "2106.10410"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_6831"} +{"question": "Which papers proposed multiscale and hierarchical processing in the field of computer vision?", "answer": ["Multiscale Vision Transformers", "Multi-Scale Vision Longformer: A New Vision Transformer for High-Resolution Image Encoding", "Multi-level Wavelet-CNN for Image Restoration"], "answer_arxiv_id": ["2104.11227", "2103.15358", "1805.07071"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_6832"} +{"question": "What research papers discuss augmenting PTLMs using knowledge graphs?", "answer": ["Human Parity on CommonsenseQA: Augmenting Self-Attention with External Attention"], "answer_arxiv_id": ["2112.03254"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_6833"} +{"question": "What works describe the use of 3D convolutions in the process of lifting image-based models to videos?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models"], "answer_arxiv_id": ["2304.08818"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_6834"} +{"question": "What research has been done on Bayesian autoencoders?", "answer": ["Model Selection for Bayesian Autoencoders"], "answer_arxiv_id": ["2106.06245v1"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_6835"} +{"question": "What papers proposed methods for unsupervised manifold alignment using generative adversarial networks and the maximum mean discrepancy?", "answer": ["MAGAN: Aligning Biological Manifolds"], "answer_arxiv_id": ["1803.00385"], "source_meta": {"published_time": "20220706"}, "qid": "AutoScholarQuery_train_6836"} +{"question": "Which studies discuss the use of Transformers to parameterize action probability distributions in syntactic language modelling?", "answer": ["Transformer Grammars: Augmenting Transformer Language Models with\n Syntactic Inductive Biases at Scale", "Pushdown Layers: Encoding Recursive Structure in Transformer Language\n Models"], "answer_arxiv_id": ["2203.00633", "2310.19089"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_6837"} +{"question": "Which work first formalized and empirically demonstrated the phenomenon of gradient descent on the Edge of Stability (EoS)?", "answer": ["Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability"], "answer_arxiv_id": ["2103.00065"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_6838"} +{"question": "What research investigated the continual learning strategy and mask image modeling technique in self-supervised learning for remote sensing tasks?", "answer": ["Towards Geospatial Foundation Models via Continual Pretraining"], "answer_arxiv_id": ["2302.04476"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_6839"} +{"question": "What are some works that use the traditional algorithm based methods for superpixel decomposition?", "answer": ["SEEDS: Superpixels Extracted via Energy-Driven Sampling"], "answer_arxiv_id": ["1309.3848"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_6840"} +{"question": "Which work pioneered the concept of Generative adversarial network (GAN) which is instrumental in image generation?", "answer": ["Generative Adversarial Nets"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20230829"}, "qid": "AutoScholarQuery_train_6841"} +{"question": "What are the papers related to Message passing neural PDE solver?", "answer": ["Message Passing Neural PDE Solvers"], "answer_arxiv_id": ["2202.03376"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_6842"} +{"question": "Which papers focused on designing long-term information propagation modules in video segmentation?", "answer": ["XMem: Long-Term Video Object Segmentation with an Atkinson-Shiffrin\n Memory Model", "Decoupling Features in Hierarchical Propagation for Video Object\n Segmentation", "Language as Queries for Referring Video Object Segmentation", "Grounding Referring Expressions in Images by Variational Context", "Spectrum-guided Multi-granularity Referring Video Object Segmentation", "Putting the Object Back into Video Object Segmentation"], "answer_arxiv_id": ["2207.07115", "2210.09782", "2201.00487", "1712.01892", "2307.13537", "2310.12982"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_6843"} +{"question": "Any works proving that the neural network-based approaches for minimizing KL-divergence over the Wasserstein space?", "answer": ["Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks"], "answer_arxiv_id": ["2106.00774"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_6844"} +{"question": "Which works have utilized generative adversarial networks in the development of 3D models?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6845"} +{"question": "Can you show me studies addressing ambiguity in coreference resolution?", "answer": ["Mind the GAP: A Balanced Corpus of Gendered Ambiguous Pronouns"], "answer_arxiv_id": ["1810.05201"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_6846"} +{"question": "Which research studied and tested GPT-3 using SocialIQA and ToMi?", "answer": ["Neural Theory-of-Mind? On the Limits of Social Intelligence in Large LMs"], "answer_arxiv_id": ["2210.13312"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_6847"} +{"question": "Which work implements RGBD fusion for real-time human rendering?", "answer": ["LookinGood: Enhancing Performance Capture with Real-time Neural\n Re-Rendering"], "answer_arxiv_id": ["1811.05029"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6848"} +{"question": "What work referenced in this section utilizes reinforcement learning in symbolic regression?", "answer": ["Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients"], "answer_arxiv_id": ["1912.04871"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_6849"} +{"question": "What are the examples of benchmarks that aggregate pre-existing tasks into comprehensive ones?", "answer": ["L-Eval: Instituting Standardized Evaluation for Long Context Language Models", "LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding"], "answer_arxiv_id": ["2307.11088v3", "2308.14508v2"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_6850"} +{"question": "Can you point out some research that used node-specific dynamics in a neural network?", "answer": ["Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting"], "answer_arxiv_id": ["2007.02842"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_6851"} +{"question": "What studies are geared towards high-quality instance segmentation by making full use of high-resolution masks?", "answer": ["DCT-Mask: Discrete Cosine Transform Mask Representation for Instance Segmentation", "SOLQ: Segmenting Objects by Learning Queries"], "answer_arxiv_id": ["2011.09876", "2106.02351"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6852"} +{"question": "Could you tell me about the recent advances in 3D diffusion models that improved 3D generation?", "answer": ["Denoising Diffusion Probabilistic Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "DreamFusion: Text-to-3D using 2D Diffusion", "Shap-E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2006.11239", "2112.10741", "2205.11487", "2112.10752", "2204.06125", "2209.14988", "2305.02463"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_6853"} +{"question": "Could you provide me some studies where the researchers performed image-to-image transformations for data augmentation using text-to-image diffusion models?", "answer": ["Effective Data Augmentation With Diffusion Models"], "answer_arxiv_id": ["2302.07944"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_6854"} +{"question": "Which research implemented LoRA to model the residuals of parameters with low-rank approximation?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_6855"} +{"question": "Which studies focus on learning of conservation laws or invariant properties within a physical system?", "answer": ["Data-driven discovery of coordinates and governing equations"], "answer_arxiv_id": ["1904.02107"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6856"} +{"question": "Could you provide me some works about capturing people dressed in real-world clothing?", "answer": ["Learning to Dress 3D People in Generative Clothing", "Detailed, accurate, human shape estimation from clothed 3D scan\n sequences", "X-Avatar: Expressive Human Avatars", "4DHumanOutfit: a multi-subject 4D dataset of human motion sequences in\n varying outfits exhibiting large displacements", "Learning Locally Editable Virtual Humans", "High-fidelity 3D Human Digitization from Single 2K Resolution Images", "SIZER: A Dataset and Model for Parsing 3D Clothing and Learning Size\n Sensitive 3D Clothing", "Deep Fashion3D: A Dataset and Benchmark for 3D Garment Reconstruction\n from Single Images", "Function4D: Real-time Human Volumetric Capture from Very Sparse Consumer\n RGBD Sensors", "HumanRF: High-Fidelity Neural Radiance Fields for Humans in Motion", "DeepCloth: Neural Garment Representation for Shape and Style Editing"], "answer_arxiv_id": ["1907.13615", "1703.04454", "2303.04805", "2306.07399", "2305.00121v1", "2303.15108", "2007.11610", "2003.12753", "2105.01859", "2305.06356", "2011.14619"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_6857"} +{"question": "Could you provide some studies on DETR-based methods for temporal action detection?", "answer": ["Relaxed Transformer Decoders for Direct Action Proposal Generation", "End-to-end Temporal Action Detection with Transformer", "End-to-end Temporal Action Detection with Transformer", "ReAct: Temporal Action Detection with Relational Queries"], "answer_arxiv_id": ["2102.01894", "2106.10271", "2106.10271", "2207.07097"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6858"} +{"question": "Which work covered StyleCLIP as a method for exploring GAN-based latent space?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery"], "answer_arxiv_id": ["2103.17249"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_6859"} +{"question": "Could you provide me some studies about building kernels using irreducible group representations?", "answer": ["Steerable CNNs", "General E(2) - Equivariant Steerable CNNs"], "answer_arxiv_id": ["1612.08498", "1911.08251"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_6860"} +{"question": "Which papers propose optimizations-based methods for customized text-to-image generation?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion", "Generate Anything Anywhere in Any Scene", "SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2212.04488", "2306.17154", "2303.11305", "2106.09685"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_6861"} +{"question": "Which work used the graph signal viewpoint and showed that the algorithmic alignment of GNNs with graph isomorphism algorithms can be utilized to prove universal approximation theorems for MPNNs?", "answer": ["On the Equivalence between Graph Isomorphism Testing and Function Approximation with GNNs"], "answer_arxiv_id": ["1905.12560"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6862"} +{"question": "Which papers proposed solutions to the hypothesis collapse problem, observed in the Winner-Takes-All training scheme?", "answer": ["Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypotheses", "Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Prediction"], "answer_arxiv_id": ["1612.00197", "1906.03631"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_6863"} +{"question": "What study indicated that sparse parities are learnable on a two layers network under certain conditions?", "answer": ["Learning Parities with Neural Networks"], "answer_arxiv_id": ["2002.07400"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6864"} +{"question": "What are some examples of research on low-rank matrix recovery?", "answer": ["Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization", "Exact Matrix Completion via Convex Optimization", "Matrix Completion with Noise", "Robust PCA via Outlier Pursuit", "Robust Principal Component Analysis?", "Low-rank matrix recovery via iteratively reweighted least squares minimization", "Guarantees of Riemannian Optimization for Low Rank Matrix Recovery", "Accelerating Ill-Conditioned Low-Rank Matrix Estimation via Scaled Gradient Descent"], "answer_arxiv_id": ["0706.4138", "0805.4471", "0903.3131", "1010.4237", "0912.3599", "1010.2471", "1511.01562", "2005.08898"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_6865"} +{"question": "Could you mention some works that developed training algorithms such as DAgger or scheduled sampling to mitigate the mismatch problem?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks"], "answer_arxiv_id": ["1011.0686v3", "1506.03099"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_6866"} +{"question": "Which works demonstrate the in-context learning ability of large language models?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6867"} +{"question": "What works proposed a cascading fashion for predicting part-aware panoptic segmentation?", "answer": ["Visual Recognition by Request"], "answer_arxiv_id": ["2207.14227"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_6868"} +{"question": "Any research discussing the data scarcity problems in tasks related to graphic design?", "answer": ["Retrieve-Then-Adapt: Example-based Automatic Generation for\n Proportion-related Infographics"], "answer_arxiv_id": ["2008.01177"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_6869"} +{"question": "What is a study that leveraged retrieved web texts as unlabeled questions?", "answer": ["Z-ICL: Zero-Shot In-Context Learning with Pseudo-Demonstrations"], "answer_arxiv_id": ["2212.09865"], "source_meta": {"published_time": "20240712"}, "qid": "AutoScholarQuery_train_6870"} +{"question": "Which works introduce various tokenization approaches such as Discrete VAE’s, VQVAE, and VQGAN?", "answer": ["Discrete Variational Autoencoders", "Neural Discrete Representation Learning", "Taming Transformers for High-Resolution Image Synthesis"], "answer_arxiv_id": ["1609.02200", "1711.00937", "2012.09841"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_6871"} +{"question": "Can you cite some studies that have carried out work on domain-generalization through the use of weight averaging on models?", "answer": ["SWAD: Domain Generalization by Seeking Flat Minima", "Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization", "Diverse Weight Averaging for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2102.08604", "2110.10832", "2205.09739"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_6872"} +{"question": "Which papers introduced generative masked modeling for text-to-image generation and editing?", "answer": ["Muse: Text-To-Image Generation via Masked Generative Transformers"], "answer_arxiv_id": ["2301.00704"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_6873"} +{"question": "What is the first large-scale Chinese NLU benchmark?", "answer": ["CLUE: A Chinese Language Understanding Evaluation Benchmark"], "answer_arxiv_id": ["2004.05986"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_6874"} +{"question": "What work investigated the performance of various CNN architectures on datasets, analyzing the correlation with ImageNet accuracy?", "answer": ["Is it enough to optimize CNN architectures on ImageNet?"], "answer_arxiv_id": ["2103.09108"], "source_meta": {"published_time": "20230111"}, "qid": "AutoScholarQuery_train_6875"} +{"question": "What works have investigated 3D language pretraining using advanced techniques, such as mask modeling and contrastive learning on paired object-caption data?", "answer": ["UniT3D: A Unified Transformer for 3D Dense Captioning and Visual\n Grounding"], "answer_arxiv_id": ["2212.00836"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_6876"} +{"question": "What paper developed BatchBALD, an incremental selection approach based on BALD score?", "answer": ["BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning"], "answer_arxiv_id": ["1906.08158"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_6877"} +{"question": "What recent works deal with missing data in the context of treatment effect estimation?", "answer": ["Doubly robust treatment effect estimation with missing attributes"], "answer_arxiv_id": ["1910.10624"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_6878"} +{"question": "Can you list the works that adopted the 'layer peeled model' for studying neural collapse?", "answer": ["Neural collapse with unconstrained features", "Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training"], "answer_arxiv_id": ["2011.11619", "2101.12699"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_6879"} +{"question": "Which paper was the first to show an exact contraction for TD with a frozen target network and general K?", "answer": ["Target-Based Temporal-Difference Learning"], "answer_arxiv_id": ["1904.10945"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_6880"} +{"question": "Which work incorporates an iterative IK module into the image-to-mesh recovery network to better align the 3D keypoints with a parametric body representation?", "answer": ["HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D\n Human Pose and Shape Estimation"], "answer_arxiv_id": ["2011.14672"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_6881"} +{"question": "Could you provide me some works that provided a Bayesian perspective of cv?", "answer": ["Understanding predictive information criteria for Bayesian models"], "answer_arxiv_id": ["1307.5928v1"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_6882"} +{"question": "Which papers have attempted to incorporate linguistically-informed labels, specifically syntactic labels, into neural networks?", "answer": ["Linguistic Input Features Improve Neural Machine Translation", "Linguistically-Informed Self-Attention for Semantic Role Labeling", "Do Syntax Trees Help Pre-trained Transformers Extract Information?", "Structural Guidance for Transformer Language Models", "Transformer Grammars: Augmenting Transformer Language Models with\n Syntactic Inductive Biases at Scale"], "answer_arxiv_id": ["1606.02892", "1804.08199", "2008.09084", "2108.00104", "2203.00633"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_6883"} +{"question": "What research papers modify VGGish to extract per-frame audio embeddings for their audio encoder?", "answer": ["CNN Architectures for Large-Scale Audio Classification"], "answer_arxiv_id": ["1609.09430"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_6884"} +{"question": "Which research introduced joint unsupervised 2D detection and 3D instance segmentation from sequential point clouds and images?", "answer": ["4D Unsupervised Object Discovery"], "answer_arxiv_id": ["2210.04801"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_6885"} +{"question": "Any works suggested the idea of graph-based modeling for physical systems?", "answer": ["Discovering Symbolic Models from Deep Learning with Inductive Biases", "Hamiltonian Neural Networks"], "answer_arxiv_id": ["2006.11287", "1906.01563"], "source_meta": {"published_time": "20220922"}, "qid": "AutoScholarQuery_train_6886"} +{"question": "What papers leverage the use of ensemble methods in overcoming the exploration-exploitation trade-off in deep RL?", "answer": ["Deep Exploration via Bootstrapped DQN", "UCB Exploration via Q-Ensembles", "Ensemble Sampling", "Model-Ensemble Trust-Region Policy Optimization", "Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models", "SUNRISE: A Simple Unified Framework for Ensemble Learning in Deep Reinforcement Learning"], "answer_arxiv_id": ["1602.04621", "1706.01502v3", "1705.07347", "1802.10592", "1805.12114", "2007.04938"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_6887"} +{"question": "What works propose RGBD object-centric datasets?", "answer": ["Category-Level 6D Object Pose Estimation in the Wild: A Semi-Supervised\n Learning Approach and A New Dataset", "A Large Dataset of Object Scans"], "answer_arxiv_id": ["2206.15436", "1602.02481"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_6888"} +{"question": "Can you provide me some papers that have explored the topic of making language models truthful?", "answer": ["Truthful AI Developing and governing AI that does not lie", "TruthfulQA: Measuring How Models Mimic Human Falsehoods"], "answer_arxiv_id": ["2110.06674", "2109.07958"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_6889"} +{"question": "Which papers are about linear MDPs?", "answer": ["Sample-Optimal Parametric Q-Learning Using Linearly Additive Features", "Provably Efficient Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes", "A Theoretical Analysis of Optimistic Proximal Policy Optimization in Linear Markov Decision Processes"], "answer_arxiv_id": ["1902.04779", "1907.05388", "2206.11489", "2212.06132", "2305.08841"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_6890"} +{"question": "Which studies utilize human-object interaction videos to pre-train feature representations for robotic policies?", "answer": ["R3M: A Universal Visual Representation for Robot Manipulation", "Real-World Robot Learning with Masked Visual Pre-training", "Masked Visual Pre-training for Motor Control"], "answer_arxiv_id": ["2203.12601", "2210.03109", "2203.06173"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6891"} +{"question": "Any works that tested the continuous noise across all tokens during the instruction tuning?", "answer": ["NEFTune: Noisy Embeddings Improve Instruction Finetuning"], "answer_arxiv_id": ["2310.05914v2"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_6892"} +{"question": "Which works have focused on explaining predictions in supervised learning tasks through feature attributions?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "A Unified Approach to Interpreting Model Predictions", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1602.04938", "1705.07874", "1703.01365"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_6893"} +{"question": "What research utilized Structural Equation Models (SEMs) in the form of continuous optimization methods for DAG learning?", "answer": ["CAM: Causal additive models, high-dimensional order search and penalized regression", "DAGs with NO TEARS: Continuous Optimization for Structure Learning"], "answer_arxiv_id": ["1310.1533", "1803.01422"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_6894"} +{"question": "Which works focus on mitigating the catastrophic forgetting phenomenon by modifying and growing the model architecture in Continual Learning?", "answer": ["Progressive Neural Networks", "Expert Gate: Lifelong Learning with a Network of Experts", "Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgetting", "Incremental Learning Through Deep Adaptation"], "answer_arxiv_id": ["1606.04671", "1611.06194", "1904.00310", "1705.04228"], "source_meta": {"published_time": "20230326"}, "qid": "AutoScholarQuery_train_6895"} +{"question": "Could you provide research where the BakedSDF method was used?", "answer": ["BakedSDF: Meshing Neural SDFs for Real-Time View Synthesis"], "answer_arxiv_id": ["2302.14859"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_6896"} +{"question": "What research addressed the problem of overestimated supports approximation by Precision and Recall method?", "answer": ["Improved Precision and Recall Metric for Assessing Generative Models"], "answer_arxiv_id": ["1904.06991"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_6897"} +{"question": "What papers learn Network embedding?", "answer": ["A Simple and Powerful Framework for Stable Dynamic Network Embedding"], "answer_arxiv_id": ["2311.09251"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_6898"} +{"question": "Which research studies extended the use of object-level mapping into a more versatile per object depth fusion using an RGB-D sensor and mask proposals?", "answer": ["Fusion++: Volumetric Object-Level SLAM"], "answer_arxiv_id": ["1808.08378"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_6899"} +{"question": "Could you provide me some studies about two-stage methods in referring video object segmentation?", "answer": ["Actor and Action Modular Network for Text-based Video Segmentation", "Rethinking Cross-modal Interaction from a Top-down Perspective for\n Referring Video Object Segmentation", "ClawCraneNet: Leveraging Object-level Relation for Text-based Video\n Segmentation"], "answer_arxiv_id": ["2011.00786", "2106.01061", "2103.10702"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_6900"} +{"question": "Can you name papers that observed the double descent phenomenon before it was popularized?", "answer": ["High-Dimensional Asymptotics of Prediction: Ridge Regression and Classification", "A Brief Prehistory of Double Descent"], "answer_arxiv_id": ["1507.03003", "2004.04328"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_6901"} +{"question": "Could you provide a reference where wavelet frequency decomposition is introduced into multi-view stereo for achieving generalizable NeRF?", "answer": ["WaveNeRF: Wavelet-based Generalizable Neural Radiance Fields"], "answer_arxiv_id": ["2308.04826"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6902"} +{"question": "Which research works are related to the use of LLMs for code generation tasks?", "answer": ["Evaluating Large Language Models Trained on Code", "Competition-Level Code Generation with AlphaCode", "Mapping Language to Code in Programmatic Context"], "answer_arxiv_id": ["2107.03374", "2203.07814", "1808.09588"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_6903"} +{"question": "Which papers approach CSG tree reconstruction by learning fixed-order assemblies?", "answer": ["CAPRI-Net: Learning Compact CAD Shapes with Adaptive Primitive Assembly", "CSG-Stump: A Learning Friendly CSG-Like Representation for Interpretable Shape Parsing", "ExtrudeNet: Unsupervised Inverse Sketch-and-Extrude for Shape Parsing"], "answer_arxiv_id": ["2104.05652", "2108.11305", "2209.15632"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_6904"} +{"question": "Can you name the papers that proposed an energy-based method for estimating distributions over relative rotations?", "answer": ["RelPose: Predicting Probabilistic Relative Rotation for Single Objects\n in the Wild"], "answer_arxiv_id": ["2208.05963"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_6905"} +{"question": "What papers propose more efficient denoising processes for accelerating the diffusion model?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["2112.10752", "2106.05931"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_6906"} +{"question": "Which work proposes PAWS, the semi-supervised extension of the siamese network approach?", "answer": ["Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samples"], "answer_arxiv_id": ["2104.13963"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_6907"} +{"question": "Are there any works providing a rigorous discussion on SE(3)-equivariant representation function?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "e3nn: Euclidean Neural Networks"], "answer_arxiv_id": ["1802.08219", "2006.10503", "2207.09453"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_6908"} +{"question": "What are some papers about utilizing multiple datasets for training?", "answer": ["BigDetection: A Large-scale Benchmark for Improved Object Detector Pre-training", "Detection Hub: Unifying Object Detection Datasets via Query Adaptation on Language Embedding", "Simple Multi-dataset Detection", "OmDet: Language-Aware Object Detection with Large-scale Vision-Language Multi-dataset Pre-training", "InternVid: A Large-scale Video-Text Dataset for Multimodal Understanding and Generation", "AIMS: All-Inclusive Multi-Level Segmentation"], "answer_arxiv_id": ["2203.13249", "2206.03484", "2102.13086", "2209.05946", "2307.06942", "2305.17768"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_6909"} +{"question": "What study introduced disentangled attention?", "answer": ["DeBERTa: Decoding-enhanced BERT with Disentangled Attention"], "answer_arxiv_id": ["2006.03654"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_6910"} +{"question": "Can you provide some studies that proposed special fine-tuning strategies for language-image pre-trained models?", "answer": ["Robust fine-tuning of zero-shot models", "Finetune like you pretrain: Improved finetuning of zero-shot vision models"], "answer_arxiv_id": ["2109.01903", "2212.00638"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_6911"} +{"question": "Which works explored the application of neural networks towards solving the shape correspondence problem?", "answer": ["Very Deep Convolutional Networks for Large-Scale Image Recognition", "Rethinking the Inception Architecture for Computer Vision", "Deep Residual Learning for Image Recognition", "Reconstruction of 3D Porous Media From 2D Slices"], "answer_arxiv_id": ["1409.1556", "1512.00567", "1512.03385", "1901.10233"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_6912"} +{"question": "Which studies use lower-fidelity controlled datasets such as CLEVR, Biased Cars, and ShapeNet for representation learning?", "answer": ["CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning", "ShapeNet: An Information-Rich 3D Model Repository"], "answer_arxiv_id": ["1612.06890", "1512.03012"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_6913"} +{"question": "What papers have discussed the successful imitation of human driving behaviour for learning control policies in autonomous vehicles?", "answer": ["Urban Driving with Conditional Imitation Learning"], "answer_arxiv_id": ["1912.00177"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_6914"} +{"question": "Is there research that establishes general sufficient conditions for the universal approximation property of continuous-depth residual networks?", "answer": ["Deep Learning via Dynamical Systems: An Approximation Perspective"], "answer_arxiv_id": ["1912.10382"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_6915"} +{"question": "Which papers use higher-order voxelized lighting representations and 3D neural networks for lighting estimation?", "answer": ["Lighthouse: Predicting Lighting Volumes for Spatially-Coherent\n Illumination", "Learning Indoor Inverse Rendering with 3D Spatially-Varying Lighting", "Learning-based Inverse Rendering of Complex Indoor Scenes with\n Differentiable Monte Carlo Raytracing"], "answer_arxiv_id": ["2003.08367", "2109.06061", "2211.03017"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_6916"} +{"question": "Who developed A/B testing algorithms for randomized online experiments?", "answer": ["A framework for Multi-A(rmed)/B(andit) testing with online FDR control", "Time series experiments and causal estimands: exact randomization tests and trading", "Cluster-Adaptive Network A/B Testing: From Randomization to Estimation", "An Online Sequential Test for Qualitative Treatment Effects", "Anytime-Valid Confidence Sequences in an Enterprise A/B Testing Platform", "Experimentation Platforms Meet Reinforcement Learning: Bayesian Sequential Decision-Making for Continuous Monitoring"], "answer_arxiv_id": ["1706.05378v2", "1706.07840v2", "2008.08648", "2111.03908", "2302.10108", "2304.00420"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_6917"} +{"question": "Which work first presented CAM in the field of weakly supervised object localization?", "answer": ["Learning Deep Features for Discriminative Localization"], "answer_arxiv_id": ["1512.04150"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_6918"} +{"question": "What works rely on intermediate program execution states for pruning the search space in execution-guided code generation?", "answer": ["Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervision", "Robust Text-to-SQL Generation with Execution-Guided Decoding", "Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["1611.00020", "1807.03100", "2203.07814"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_6919"} +{"question": "Did any studies analyze the static Ethereum transaction network following studies on social networks, citation networks and the Internet?", "answer": ["On the Ethereum Blockchain Structure: a Complex Networks Theory Perspective"], "answer_arxiv_id": ["1908.11808v1"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_6920"} +{"question": "What are the notable works on detecting photometric image manipulation?", "answer": ["On the Detection of Digital Face Manipulation", "Proactive Image Manipulation Detection", "Hierarchical Fine-Grained Image Forgery Detection and Localization"], "answer_arxiv_id": ["1910.01717", "2203.15880", "2303.17111"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_6921"} +{"question": "What research studied the pruning unit in different granularity?", "answer": ["Structured Pruning Learns Compact and Accurate Models"], "answer_arxiv_id": ["2204.00408"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_6922"} +{"question": "Which papers focus on the impact of OOD data on the performance of machine learning models?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations", "The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization", "Do ImageNet Classifiers Generalize to ImageNet?", "Causal Transportability for Visual Recognition", "Generative Interventions for Causal Learning", "Discrete Representations Strengthen Vision Transformer Robustness"], "answer_arxiv_id": ["1903.12261", "2006.16241", "1902.10811", "2204.12363", "2012.12265", "2111.10493v2"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_6923"} +{"question": "Could you provide me a study about joint solutions for sound source localization, separation, and dereverberation by using audio-visual and geometric features?", "answer": ["Novel-View Acoustic Synthesis from 3D Reconstructed Rooms"], "answer_arxiv_id": ["2310.15130"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_6924"} +{"question": "Which research focuses on achieving detailed image-to-3D results by fusing and balancing guidance from pre-trained 2D diffusion model and viewpoint transformation model?", "answer": ["Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors"], "answer_arxiv_id": ["2306.17843"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_6925"} +{"question": "What research works study learning from audiovisual correspondences in the field of action recognition?", "answer": ["Cooperative Learning of Audio and Video Models from Self-Supervised\n Synchronization", "EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action\n Recognition"], "answer_arxiv_id": ["1807.00230", "1908.08498"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_6926"} +{"question": "What studies proposed techniques to alleviate the biases caused by using adaptive optimizer or different hyper-parameters on local clients in federated learning?", "answer": ["Local Adaptivity in Federated Learning: Convergence and Consistency", "Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization"], "answer_arxiv_id": ["2106.02305", "2007.07481"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_6927"} +{"question": "Any study proposed the Bregeman divergence-based DRE framework and its implementation with the deep neural network?", "answer": ["Non-Negative Bregman Divergence Minimization for Deep Direct Density Ratio Estimation"], "answer_arxiv_id": ["2006.06979"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_6928"} +{"question": "What research proves that pessimism can effectively alleviate overestimation and achieve good performance in Offline RL?", "answer": ["The Importance of Pessimism in Fixed-Dataset Policy Optimization", "Is Pessimism Provably Efficient for Offline RL?", "Provably Good Batch Reinforcement Learning Without Great Exploration", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Bellman-consistent Pessimism for Offline Reinforcement Learning", "Provable Benefits of Actor-Critic Methods for Offline Reinforcement Learning", "Adversarially Trained Actor Critic for Offline Reinforcement Learning"], "answer_arxiv_id": ["2009.06799", "2012.15085", "2007.08202", "2103.12021v2", "2106.06926", "2108.08812", "2202.02446"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_6929"} +{"question": "Could you provide me with details of research discussing L1 and L2 norm bounds in OCO?", "answer": ["No-Regret Algorithms for Unconstrained Online Convex Optimization"], "answer_arxiv_id": ["1211.2260"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6930"} +{"question": "Can you provide some works that utilized self-supervised learning techniques to alleviate constraints imposed by human annotations?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2104.14294", "2304.07193"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6931"} +{"question": "Which works study RFE with safety constraints?", "answer": ["A Simple Reward-free Approach to Constrained Reinforcement Learning", "Safe Exploration Incurs Nearly No Additional Sample Complexity for Reward-free RL"], "answer_arxiv_id": ["2107.05216", "2206.14057"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_6932"} +{"question": "Which works demonstrated the use of pre-trained ResNet as effective visual backbone for simulated dexterous manipulation RL tasks?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["2107.03380", "1512.03385"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_6933"} +{"question": "Could you provide me some works about computed feature correlation in the 2D image space for point tracking?", "answer": ["TAPIR: Tracking Any Point with per-frame Initialization and temporal\n Refinement", "CoTracker: It is Better to Track Together"], "answer_arxiv_id": ["2306.08637", "2307.07635"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_6934"} +{"question": "Could you provide me some studies that have investigated the use of ensemble learning methods to exploit model-specific information in the domain generalization?", "answer": ["Best sources forward: domain generalization through source-specific nets", "Batch Normalization Embeddings for Deep Domain Generalization", "Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization", "Domain Generalization using Pretrained Models without Fine-tuning"], "answer_arxiv_id": ["1806.05810", "2011.12672", "2110.10832", "2203.04600"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_6935"} +{"question": "Which papers mentioned using the OT cost as the loss function to update the generator in generative models?", "answer": ["Improved Training of Wasserstein GANs", "Learning Generative Models with Sinkhorn Divergences"], "answer_arxiv_id": ["1704.00028", "1706.00292v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_6936"} +{"question": "Can you mention some studies about the application of LMDP in multi-task RL?", "answer": ["Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning"], "answer_arxiv_id": ["1910.10897"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_6937"} +{"question": "Which papers are associated with finetuning weights of pretrained generative models in text-driven image editing?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_6938"} +{"question": "Which papers revealed that node representations within each connected component of the graph converge to the same values as the number of GNN layers goes to infinity?", "answer": ["Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning", "Graph Neural Networks Exponentially Lose Expressive Power for Node Classification"], "answer_arxiv_id": ["1801.07606", "1905.10947"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_6939"} +{"question": "Which works are currently in progress regarding the spectral analysis of Koopman operators?", "answer": ["Ergodic theory, Dynamic Mode Decomposition and Computation of Spectral Properties of the Koopman operator", "Delay-coordinate maps and the spectra of Koopman operators", "An Operator Theoretic Approach for Analyzing Sequence Neural Networks"], "answer_arxiv_id": ["1611.06664", "1706.08544", "2102.07824"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_6940"} +{"question": "Which works historically applied self-supervised learning via generative models?", "answer": ["Adversarial Feature Learning"], "answer_arxiv_id": ["1605.09782"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_6941"} +{"question": "What research motivates applications of SGD to neural networks training?", "answer": ["Deep Learning", "Deep Learning"], "answer_arxiv_id": ["1901.10233", "1901.10233"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_6942"} +{"question": "Any work provided single-infrastructure datasets available for public for roadside perception?", "answer": ["BoxCars: Improving Fine-Grained Recognition of Vehicles using 3-D\n Bounding Boxes in Traffic Surveillance", "BAAI-VANJEE Roadside Dataset: Towards the Connected Automated Vehicle\n Highway technologies in Challenging Environments of China", "Rope3D: TheRoadside Perception Dataset for Autonomous Driving and\n Monocular 3D Object Detection Task", "A9-Dataset: Multi-Sensor Infrastructure-Based Dataset for Mobility\n Research"], "answer_arxiv_id": ["1703.00686", "2105.14370", "2203.13608", "2204.06527"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_6943"} +{"question": "What studies have used ranking losses to structure GNN embedding spaces?", "answer": ["Contrastive Neural Architecture Search with Neural Architecture\n Comparators", "Representation Learning for Frequent Subgraph Mining"], "answer_arxiv_id": ["2103.05471", "2402.14367"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_6944"} +{"question": "Which works show that the performances of teachers in Vision-Language Models (VLMs) are largely unaware of fine-grained region-word alignment?", "answer": ["RegionCLIP: Region-based Language-Image Pretraining", "Open Vocabulary Object Detection with Proposal Mining and Prediction Equalization"], "answer_arxiv_id": ["2112.09106", "2206.11134"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_6945"} +{"question": "What works have been done to encode spatial information in protein structures by utilizing 3D CNNs?", "answer": ["Deep convolutional networks for quality assessment of protein folds"], "answer_arxiv_id": ["1801.06252"], "source_meta": {"published_time": "20220311"}, "qid": "AutoScholarQuery_train_6946"} +{"question": "What studies have demonstrated the use of web APIs with large language models?", "answer": ["ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs", "Gorilla: Large Language Model Connected with Massive APIs"], "answer_arxiv_id": ["2307.16789", "2305.15334"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6947"} +{"question": "Any works about the collection of a video dataset with textual description annotations?", "answer": ["Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval"], "answer_arxiv_id": ["2104.00650"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_train_6948"} +{"question": "Which studies in the field of portrait segmentation have used convolutional networks or MLPs specifically for segmentation refinement?", "answer": ["PointRend: Image Segmentation as Rendering", "Mask R-CNN", "Mask Transfiner for High-Quality Instance Segmentation"], "answer_arxiv_id": ["1912.08193", "1703.06870", "2111.13673"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_6949"} +{"question": "Who have conducted research on multi-agent frameworks for communication?", "answer": ["Generative Agents: Interactive Simulacra of Human Behavior", "CAMEL: Communicative Agents for \"Mind\" Exploration of Large Language Model Society", "ChatDev: Communicative Agents for Software Development", "ChatDB: Augmenting LLMs with Databases as Their Symbolic Memory"], "answer_arxiv_id": ["2304.03442", "2303.17760v2", "2307.07924", "2306.03901"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_6950"} +{"question": "Could you provide studies that talk about the popular intensive search-based alternatives like beam search for high-likelihood generations?", "answer": ["A Simple, Fast Diverse Decoding Algorithm for Neural Generation"], "answer_arxiv_id": ["1611.08562"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_6951"} +{"question": "Which works proposed fine-tuning Stable Diffusion models using T5 and high-quality paired data?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Emu: Enhancing Image Generation Models Using Photogenic Needles in a\n Haystack"], "answer_arxiv_id": ["2205.11487", "2309.15807"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_6952"} +{"question": "What research implemented consistency models to enable a single-step student to do further steps in distillation?", "answer": ["Consistency Models", "Latent Consistency Models: Synthesizing High-Resolution Images with\n Few-Step Inference"], "answer_arxiv_id": ["2303.01469", "2310.04378"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_6953"} +{"question": "What works provided practical tips for performing knowledge distillation?", "answer": ["Knowledge distillation: A good teacher is patient and consistent"], "answer_arxiv_id": ["2106.05237"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_6954"} +{"question": "What are some papers about the application of Optimal Transport (OT) in deep learning?", "answer": ["GOT: An Optimal Transport framework for Graph comparison", "Graph Optimal Transport for Cross-Domain Alignment", "Domain Generalization via Optimal Transport with Metric Similarity Learning", "Multi-source Domain Adaptation via Weighted Joint Distributions Optimal Transport", "Hierarchical Optimal Transport for Multimodal Distribution Alignment", "Multimodal Learning using Optimal Transport for Sarcasm and Humor Detection"], "answer_arxiv_id": ["1906.02085", "2006.14744", "2007.10573", "2006.12938", "1906.11768", "2110.10949"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_6955"} +{"question": "What papers derived predictions from multiple analyses that would be considered as part of a beam in the field of human language processing?", "answer": ["Finding Syntax in Human Encephalography with Beam Search"], "answer_arxiv_id": ["1806.04127"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_6956"} +{"question": "May you list some works that have used diffusion models for 2D image synthesis applications?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Image Super-Resolution via Iterative Refinement", "Cascaded Diffusion Models for High Fidelity Image Generation", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Blended Diffusion for Text-driven Editing of Natural Images", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "Video Diffusion Models", "Diffusion Probabilistic Modeling for Video Generation"], "answer_arxiv_id": ["2108.01073", "2104.07636", "2106.15282", "2108.01073", "2111.14818", "2112.10741", "2112.10752", "2110.02711", "2204.03458", "2203.09481"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_6957"} +{"question": "Could you mention some works about generative efforts in autonomous driving using BEV layouts?", "answer": ["BEVControl: Accurately Controlling Street-view Elements with\n Multi-perspective Consistency via BEV Sketch Layout", "Street-View Image Generation from a Bird's-Eye View Layout", "LayoutDiffusion: Controllable Diffusion Model for Layout-to-image\n Generation", "LayoutDiffuse: Adapting Foundational Diffusion Models for\n Layout-to-Image Generation"], "answer_arxiv_id": ["2308.01661", "2301.04634", "2303.17189", "2302.08908"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_6958"} +{"question": "Could you provide me some research papers that achieved top performance on the COCO 2017 dataset?", "answer": ["Hybrid Task Cascade for Instance Segmentation", "Dynamic Head: Unifying Object Detection Heads with Attentions"], "answer_arxiv_id": ["1901.07518", "2106.08322"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_6959"} +{"question": "Which work improved on the continual release algorithm initially proposed by bib.bibx60?", "answer": ["Frequency Estimation Under Multiparty Differential Privacy: One-shot and Streaming"], "answer_arxiv_id": ["2104.01808"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_6960"} +{"question": "What are some of the studies that explored the problem of ordinal embedding?", "answer": ["Landmark Ordinal Embedding", "Kernel functions based on triplet comparisons", "Preference Completion: Large-scale Collaborative Ranking from Pairwise Comparisons"], "answer_arxiv_id": ["1910.12379", "1607.08456", "1507.04457"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_6961"} +{"question": "What works innovatively tackled the problem of image inpainting?", "answer": ["Generative Adversarial Networks", "Free-Form Image Inpainting with Gated Convolution", "Contextual Residual Aggregation for Ultra High-Resolution Image\n Inpainting", "Recurrent Feature Reasoning for Image Inpainting", "Large Scale Image Completion via Co-Modulated Generative Adversarial\n Networks", "Resolution-robust Large Mask Inpainting with Fourier Convolutions", "Image Completion with Heterogeneously Filtered Spectral Hints", "CM-GAN: Image Inpainting with Cascaded Modulation GAN and Object-Aware Training", "Image Inpainting for Irregular Holes Using Partial Convolutions", "HD-Painter: High-Resolution and Prompt-Faithful Text-Guided Image\n Inpainting with Diffusion Models"], "answer_arxiv_id": ["2203.00667", "1806.03589", "2005.09704", "2008.03737", "2103.10428", "2109.07161", "2211.03700", "2203.11947v3", "1804.07723", "2312.14091"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_6962"} +{"question": "Which studies apply the new promising techniques, Tree of Thoughts and Self-Consistency, in the realm of self-evaluation in Language Models?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2305.10601", "2203.11171"], "source_meta": {"published_time": "20240525"}, "qid": "AutoScholarQuery_train_6963"} +{"question": "Which research papers focus on instruction tuning for multiple modalities or special domains like science and medicine?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Inner Monologue: Embodied Reasoning through Planning with Language Models", "MultiInstruct: Improving Multi-Modal Zero-Shot Learning via Instruction Tuning", "Solving Quantitative Reasoning Problems with Language Models", "Large Language Models Encode Clinical Knowledge"], "answer_arxiv_id": ["2204.01691", "2207.05608", "2212.10773", "2206.14858", "2212.13138"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6964"} +{"question": "Could you name the research papers that have developed algorithms for Distributionally Robust Optimization or DRO?", "answer": ["Learning Not to Learn: Training Deep Neural Networks with Biased Data", "Distributionally Robust Neural Networks for Group Shifts: On the\n Importance of Regularization for Worst-Case Generalization", "Large-Scale Methods for Distributionally Robust Optimization", "Environment Inference for Invariant Learning", "Learning Debiased Representation via Disentangled Feature Augmentation", "Correct-N-Contrast: A Contrastive Approach for Improving Robustness to\n Spurious Correlations"], "answer_arxiv_id": ["1812.10352", "1911.08731", "2010.05893", "2010.07249", "2107.01372", "2203.01517"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_6965"} +{"question": "Which papers are relevant in regard to the utilization of transformers in return-conditioned imitation learning?", "answer": ["Training Agents using Upside-Down Reinforcement Learning", "Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["1912.02877", "2106.01345"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_6966"} +{"question": "Which work introduced a method to restrict the search space of adversarial attacks to the low-frequency domain?", "answer": ["Low Frequency Adversarial Perturbation"], "answer_arxiv_id": ["1809.08758"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_6967"} +{"question": "What studies investigate automatic metrics for natural language generation?", "answer": ["A Study of Automatic Metrics for the Evaluation of Natural Language Explanations"], "answer_arxiv_id": ["2103.08545"], "source_meta": {"published_time": "20221215"}, "qid": "AutoScholarQuery_train_6968"} +{"question": "What papers highlighted the remarkable performance of LLMs on natural language processing (NLP) tasks?", "answer": ["Language Models are Few-Shot Learners", "Emergent Abilities of Large Language Models", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2005.14165", "2206.07682", "2203.02155"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_6969"} +{"question": "What papers introduced the Vision Transformer (ViT) and associated methods?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "CCNet: Criss-Cross Attention for Semantic Segmentation", "Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation", "CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows", "CvT: Introducing Convolutions to Vision Transformers", "CMT: Convolutional Neural Networks Meet Vision Transformers", "ViTAE: Vision Transformer Advanced by Exploring Intrinsic Inductive Bias"], "answer_arxiv_id": ["2010.11929", "2103.14030", "1811.11721", "2003.07853", "2107.00652", "2103.15808", "2107.06263", "2106.03348"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_6970"} +{"question": "Which research has proposed appreciating procedure planning as a distribution-fitting problem?", "answer": ["P3IV: Probabilistic Procedure Planning from Instructional Videos with\n Weak Supervision", "PDPP:Projected Diffusion for Procedure Planning in Instructional Videos", "Event-Guided Procedure Planning from Instructional Videos with Text\n Supervision"], "answer_arxiv_id": ["2205.02300", "2303.14676", "2308.08885"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6971"} +{"question": "Which works represent 3D scenes as an implicit MLP-based function and use volume rendering technology?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "NeRF++: Analyzing and Improving Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "NeRF--: Neural Radiance Fields Without Known Camera Parameters", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Compressible-composable NeRF via Rank-residual Decomposition", "Cross-Ray Neural Radiance Fields for Novel-view Synthesis from\n Unconstrained Image Collections"], "answer_arxiv_id": ["2003.08934", "2103.13415", "2010.07492", "2201.05989", "2102.07064", "2011.13961", "2205.14870", "2307.08093"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_6972"} +{"question": "What studies implement contrastive learning for 3D pre-training?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D\n Point Cloud Understanding"], "answer_arxiv_id": ["2007.10985", "2203.00680"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_6973"} +{"question": "Could you list the works that propose to leverage established large vision-language models for human-scene interaction synthesis?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Learning Transferable Visual Models From Natural Language Supervision", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2112.10752", "2205.11487", "2103.00020", "2301.12597"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_6974"} +{"question": "What studies related to showing generalization bounds for deep neural networks?", "answer": ["Spectrally-normalized margin bounds for neural networks", "A PAC-Bayesian Approach to Spectrally-Normalized Margin Bounds for Neural Networks"], "answer_arxiv_id": ["1706.08498", "1707.09564"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_6975"} +{"question": "What works discuss unsupervised skill discovery in the context of training a skill-conditioned policy?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function", "Dynamics-Aware Unsupervised Discovery of Skills"], "answer_arxiv_id": ["1802.06070", "1907.01657"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_6976"} +{"question": "What works have designed 'good' non-conformity scores in theory to achieve properties beyond validity in Conversational Intent Slot Prediction?", "answer": ["Classification with Valid and Adaptive Coverage", "Conformalized Quantile Regression"], "answer_arxiv_id": ["2006.02544", "1905.03222"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_6977"} +{"question": "Which works introduce frequency analysis into the synthetic image detection framework?", "answer": ["Thinking in Frequency: Face Forgery Detection by Mining Frequency-aware\n Clues"], "answer_arxiv_id": ["2007.09355"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_6978"} +{"question": "Which works have applied their uncertainty estimates to unseen datasets?", "answer": ["Learning Sample Difficulty from Pre-trained Models for Reliable Prediction", "Massively Scaling Heteroscedastic Classifiers", "Uncertainty in Contrastive Learning: On the Predictability of Downstream Performance", "Probabilistic Embeddings Revisited"], "answer_arxiv_id": ["2304.10127", "2301.12860", "2207.09336", "2202.06768"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_6979"} +{"question": "What study highlighted a connection between spectral learning and supervised learning, and based on that, proposed a two-stage regression learning algorithm?", "answer": ["Supervised Learning for Dynamical System Learning"], "answer_arxiv_id": ["1505.05310"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_6980"} +{"question": "Which paper developed PAC-Bayesian bounds for the analysis of adversarial generative models?", "answer": ["PAC-Bayesian Generalization Bounds for Adversarial Generative Models"], "answer_arxiv_id": ["2302.08942"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_6981"} +{"question": "What papers show that gradient-based methods share the same computational lower bound as the statistical query (SQ) framework when the gradient precision is not sufficient?", "answer": ["Failures of Gradient-Based Deep Learning"], "answer_arxiv_id": ["1703.07950v2"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_6982"} +{"question": "What studies have adapted diffusion models for image synthesis?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Variational Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2107.00630"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_6983"} +{"question": "Which works are about employing grammar induction algorithms for action analysis?", "answer": ["Predicting Human Activities Using Stochastic Grammar", "Generalized Earley Parser: Bridging Symbolic Grammars and Sequence Data for Future Prediction"], "answer_arxiv_id": ["1708.00945v1", "1806.03497v1"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_6984"} +{"question": "What papers argue that agents can generalize better thanks to the partial compositionality and recursivity of language?", "answer": ["Grounded Language Learning in a Simulated 3D World", "Environmental drivers of systematicity and generalization in a situated agent", "Language as a Cognitive Tool to Imagine Goals in Curiosity-Driven Exploration"], "answer_arxiv_id": ["1706.06551", "1910.00571v4", "2002.09253"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_6985"} +{"question": "What research proposes the weight factorization method for multilingual ASR?", "answer": ["Adaptive multilingual speech recognition with pretrained models"], "answer_arxiv_id": ["2205.12304"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_6986"} +{"question": "Any works aimed at speeding up training rather than inference by finding N:M masks?", "answer": ["Accelerated Sparse Neural Training: A Provable and Efficient Method to Find N:M Transposable Masks"], "answer_arxiv_id": ["2102.08124"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_6987"} +{"question": "What works have shown that Transformer-based language models can't generalize to out-of-distribution inputs for non-regular languages?", "answer": ["Neural Networks and the Chomsky Hierarchy"], "answer_arxiv_id": ["2207.02098"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_6988"} +{"question": "Could you provide me some studies delving into the fast solver of SDE or ODE to create efficient sampling of diffusion models?", "answer": ["DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps", "Fast Sampling of Diffusion Models with Exponential Integrator", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in\n Diffusion Probabilistic Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2206.00927", "2204.13902", "2201.06503", "2202.09778"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_6989"} +{"question": "What works propose soft prompts to adapt representations for specific tasks?", "answer": ["Visual Prompt Tuning", "Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Read-only Prompt Optimization for Vision-Language Few-shot Learning", "Prompt Pre-Training with Twenty-Thousand Classes for Open-Vocabulary\n Visual Recognition"], "answer_arxiv_id": ["2203.12119", "2109.01134", "2203.05557", "2308.14960", "2304.04704"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_6990"} +{"question": "Which works discussed the dissemination of disinformation texts through generative LLMs?", "answer": ["Defending Against Neural Fake News", "TweepFake: about Detecting Deepfake Tweets", "Generating Sentiment-Preserving Fake Online Reviews Using Neural\n Language Models and Their Human- and Machine-based Detection"], "answer_arxiv_id": ["1905.12616", "2008.00036", "1907.09177"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_6991"} +{"question": "Which papers elaborated the use of recurrent approaches in generating sequences of 3D human motion?", "answer": ["Recurrent Network Models for Human Dynamics", "Structured Prediction Helps 3D Human Motion Modelling", "Action-Agnostic Human Pose Forecasting", "A Neural Temporal Model for Human Motion Prediction", "Structural-RNN: Deep Learning on Spatio-Temporal Graphs", "On human motion prediction using recurrent neural networks"], "answer_arxiv_id": ["1508.00271", "1910.09070", "1810.09676", "1809.03036", "1511.05298", "1705.02445"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_6992"} +{"question": "What papers have contributed teachings on semantic scene completion?", "answer": ["3D Semantic Scene Completion: a Survey", "BSP-Net: Generating Compact Meshes via Binary Space Partitioning", "Anisotropic Convolutional Networks for 3D Semantic Scene Completion", "SCFusion: Real-time Incremental Scene Reconstruction with Semantic Completion", "LMSCNet: Lightweight Multiscale 3D Semantic Completion", "Semantic Scene Completion using Local Deep Implicit Functions on LiDAR Data", "Semantic Scene Completion via Integrating Instances and Scene in-the-Loop"], "answer_arxiv_id": ["2103.07466", "1911.06971", "2004.02122v1", "2010.13662", "2008.10559", "2011.09141v3", "2104.03640"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_6993"} +{"question": "What works have shown that unlearnable datasets are vulnerable to adversarial training?", "answer": ["Unlearnable Examples: Making Personal Data Unexploitable", "Better Safe Than Sorry: Preventing Delusive Adversaries with Adversarial Training"], "answer_arxiv_id": ["2101.04898", "2102.04716"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6994"} +{"question": "What papers showcase the use of extending CQL within the model-based regime for regularization?", "answer": ["COMBO: Conservative Offline Model-Based Policy Optimization"], "answer_arxiv_id": ["2102.08363"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_6995"} +{"question": "Can you list works that considered domain-adapted anomaly detection methods?", "answer": ["Meta-Learning with Fewer Tasks through Task Interpolation", "Few-shot Scene-adaptive Anomaly Detection"], "answer_arxiv_id": ["2106.02695", "2007.07843"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_6996"} +{"question": "Which collaboration observed that for some neural networks trained by full-batch gradient descent, the loss is not monotonically decreasing?", "answer": ["A Walk with SGD"], "answer_arxiv_id": ["1802.08770"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_6997"} +{"question": "What paper uses a lightweight adapter module to align visual tokens and text tokens in projection-based methods?", "answer": ["LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention"], "answer_arxiv_id": ["2303.16199"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_6998"} +{"question": "Are there any studies about ways to quantify fairness of ML models when demographic information is missing?", "answer": ["Assessing Algorithmic Fairness with Unobserved Protected Class Using Data Combination", "Assessing Fairness in the Presence of Missing Data"], "answer_arxiv_id": ["1906.00285", "2112.04899"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_6999"} +{"question": "Which papers provide comprehensive surveys of different ANN approaches?", "answer": ["Survey of Nearest Neighbor Techniques", "Hashing for Similarity Search: A Survey", "A Comprehensive Survey and Experimental Comparison of Graph-Based Approximate Nearest Neighbor Search"], "answer_arxiv_id": ["1007.0085v1", "1408.2927", "2101.12631"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_7000"} +{"question": "Could you provide me some studies where AT was combined with data augmentation techniques?", "answer": ["Unlabeled Data Improves Adversarial Robustness", "Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuning", "Data Augmentation Can Improve Robustness"], "answer_arxiv_id": ["1905.13736", "2003.12862", "2111.05328"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_7001"} +{"question": "Which works employ knowledge distillation as a means of model compression?", "answer": ["Knowledge Distillation: A Survey"], "answer_arxiv_id": ["2006.05525"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_7002"} +{"question": "What work suggests that a single environment map is insufficient for compositing multiple, large, or moving virtual objects into the captured scene?", "answer": ["Lighthouse: Predicting Lighting Volumes for Spatially-Coherent\n Illumination"], "answer_arxiv_id": ["2003.08367"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_7003"} +{"question": "Which works are the initial proposals in the Generative Pretrained Transformer (GPT) series?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_7004"} +{"question": "What research introduced a framework for generating looser counterfactuals allowing larger edits of original examples?", "answer": ["NeuroCounterfactuals: Beyond Minimal-Edit Counterfactuals for Richer Data Augmentation"], "answer_arxiv_id": ["2210.12365"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_7005"} +{"question": "Which work uses LAMA to evaluate whether models can correctly predict masked object entities in a cloze-style prompt?", "answer": ["Language Models as Knowledge Bases?"], "answer_arxiv_id": ["1909.01066"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_7006"} +{"question": "What paper is most relevant to the study of phrase retrieval problem for question-answering tasks?", "answer": ["Learning Dense Representations of Phrases at Scale"], "answer_arxiv_id": ["2012.12624"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_7007"} +{"question": "What works are there that explores differentially private algorithms for training deep generative models?", "answer": ["Differentially Private Generative Adversarial Network", "DP-CGAN : Differentially Private Synthetic Data and Label Generation", "P3GM: Private High-Dimensional Data Release via Privacy Preserving Phased Generative Model"], "answer_arxiv_id": ["1802.06739", "2001.09700", "2006.12101v4"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_7008"} +{"question": "Any works about exploring the inclusion of regressing semantic labels into the process of Novel View Synthesis in decompositional neural rendering?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene\n Representation"], "answer_arxiv_id": ["2103.15875"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_7009"} +{"question": "What works use primal-dual approaches to solve the Lagrangian problem of constrained policy optimization in Safe RL by CMDP?", "answer": ["Achieving Zero Constraint Violation for Constrained Reinforcement Learning via Primal-Dual Approach", "Risk-Constrained Reinforcement Learning with Percentile Risk Criteria", "Provably Efficient Safe Exploration via Primal-Dual Policy Optimization", "Responsive Safety in Reinforcement Learning by PID Lagrangian Methods", "First Order Constrained Optimization in Policy Space", "CRPO: A New Approach for Safe Reinforcement Learning with Convergence Guarantee", "Penalized Proximal Policy Optimization for Safe Reinforcement Learning"], "answer_arxiv_id": ["2109.06332", "1512.01629", "2003.00534", "2007.03964", "2002.06506", "2011.05869", "2205.11814"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_7010"} +{"question": "What are the recent advancements in reinforcement learning environments since 2021?", "answer": ["Open-Ended Learning Leads to Generally Capable Agents", "Human-Timescale Adaptation in an Open-Ended Task Space"], "answer_arxiv_id": ["2107.12808", "2301.07608"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_7011"} +{"question": "Which research papers are about combining Tactron2 with the Transformer for enhancing training efficiency?", "answer": ["Neural Speech Synthesis with Transformer Network"], "answer_arxiv_id": ["1809.08895"], "source_meta": {"published_time": "20240717"}, "qid": "AutoScholarQuery_train_7012"} +{"question": "Which papers propose solutions for tackling document retrieval using dense retrievers?", "answer": ["Answering Complex Open-Domain Questions with Multi-Hop Dense Retrieval", "Pre-training via Paraphrasing", "Optimizing Dense Retrieval Model Training with Hard Negatives", "More Robust Dense Retrieval with Contrastive Dual Learning", "Few-Shot Conversational Dense Retrieval"], "answer_arxiv_id": ["2009.12756", "2006.15020", "2104.08051", "2107.07773", "2105.04166"], "source_meta": {"published_time": "20220901"}, "qid": "AutoScholarQuery_train_7013"} +{"question": "Which papers apply discriminative reranking approaches in summarization tasks?", "answer": ["Multi-Document Summarization via Discriminative Summary Reranking"], "answer_arxiv_id": ["1507.02062"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_7014"} +{"question": "Which papers mention the use of SimCLR method in solving the contrastive loss function?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_7015"} +{"question": "What studies blend flows and kernels for VFI?", "answer": ["Depth-Aware Video Frame Interpolation", "MEMC-Net: Motion Estimation and Motion Compensation Driven Neural\n Network for Video Interpolation and Enhancement", "ST-MFNet: A Spatio-Temporal Multi-Flow Network for Frame Interpolation", "H-VFI: Hierarchical Frame Interpolation for Videos with Large Motions"], "answer_arxiv_id": ["1904.00830", "1810.08768", "2111.15483", "2211.11309"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_7016"} +{"question": "What works are involved in applying transformers to various computer vision tasks?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "OutfitTransformer: Learning Outfit Representations for Fashion\n Recommendation", "Learning Canonical View Representation for 3D Shape Recognition with\n Arbitrary Views"], "answer_arxiv_id": ["2010.11929", "2204.04812", "2108.07084"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_7017"} +{"question": "Which research paper found superior accuracy by fine-tuning the ViT backbone with a lower learning rate?", "answer": ["AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition"], "answer_arxiv_id": ["2205.13535"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_7018"} +{"question": "Which works highlighted the use of unsupervised or self-supervised learning to generate additional tasks?", "answer": ["STraTA: Self-Training with Task Augmentation for Better Few-shot Learning", "Cross-Domain Few-Shot Classification via Adversarial Task Augmentation"], "answer_arxiv_id": ["2109.06270", "2104.14385"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_7019"} +{"question": "What are some prior works on Language-Conditioned Control in physics-based animation and robot manipulation?", "answer": ["Language-Conditioned Imitation Learning for Robot Manipulation Tasks", "Learning Object Placements For Relational Instructions by Hallucinating Scene Representations", "Composing Pick-and-Place Tasks By Grounding Language"], "answer_arxiv_id": ["2010.12083", "2001.08481", "2102.08094"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_7020"} +{"question": "Which works employed PointNet families for feature extraction from raw point clouds?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space"], "answer_arxiv_id": ["1612.00593", "1706.02413"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7021"} +{"question": "Can you list the papers that have proposed methods to manage data heterogeneity and imbalance in federated learning?", "answer": ["FedShuffle: Recipes for Better Use of Local Work in Federated Learning", "Model-Contrastive Federated Learning"], "answer_arxiv_id": ["2204.13169", "2103.16257"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_7022"} +{"question": "What papers discuss the use of machine learning for program synthesis?", "answer": ["The Three Pillars of Machine Programming", "A Survey of Machine Learning for Big Code and Naturalness"], "answer_arxiv_id": ["1803.07244", "1709.06182"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_7023"} +{"question": "Could you provide me some studies that reinterpreted the original HCA formulation from different perspectives?", "answer": ["An Information-Theoretic Perspective on Credit Assignment in Reinforcement Learning", "Hindsight Network Credit Assignment: Efficient Credit Assignment in Networks of Discrete Stochastic Units"], "answer_arxiv_id": ["2103.06224", "2110.07700"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_7024"} +{"question": "What works studied online convex optimization in presence of switching costs?", "answer": ["Beyond Online Balanced Descent: An Optimal Algorithm for Smoothed Online Optimization"], "answer_arxiv_id": ["1905.12776"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_7025"} +{"question": "What works studied the characteristics of CT-RNNs to understand their applicability and limitations in learning sequential data and flows?", "answer": ["Augmented Neural ODEs", "Neural Spline Flows", "Neural Jump Stochastic Differential Equations", "On the Verification of Neural ODEs with Stochastic Guarantees", "On Robustness of Neural Ordinary Differential Equations", "Learning to Control PDEs with Differentiable Physics", "SNODE: Spectral Discretization of Neural ODEs for System Identification", "Neural Controlled Differential Equations for Irregular Time Series", "Sparse Flows: Pruning Continuous-depth Models"], "answer_arxiv_id": ["1904.01681", "1906.04032", "1905.10403", "2012.08863", "1910.05513", "2001.07457", "1906.07038", "2005.08926", "2106.12718"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_7026"} +{"question": "Which studies have found errors in widely used benchmarks such as CoNLL 2003 for Named Entity Recognition?", "answer": ["CrossWeigh: Training Named Entity Tagger from Imperfect Annotations", "CleanCoNLL: A Nearly Noise-Free Named Entity Recognition Dataset"], "answer_arxiv_id": ["1909.01441", "2310.16225"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_7027"} +{"question": "What studies have proposed wavelet-based representations to enhance NeRF’s ability to capture fine texture details?", "answer": ["Masked Wavelet Representation for Compact Neural Radiance Fields", "WaveNeRF: Wavelet-based Generalizable Neural Radiance Fields", "WIRE: Wavelet Implicit Neural Representations"], "answer_arxiv_id": ["2212.09069", "2308.04826", "2301.05187"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_7028"} +{"question": "Any works indicating the difficulty of the problem for general norms in matrix approximation?", "answer": ["Low-Rank Matrix Approximation in the Infinity Norm"], "answer_arxiv_id": ["1706.00078"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_7029"} +{"question": "What papers focus on active recognition as a way to improve agents' performance?", "answer": ["A Dataset for Developing and Benchmarking Active Vision", "Move to See Better: Self-Improving Embodied Object Detection", "Geometry-Aware Recurrent Neural Networks for Active Visual Recognition"], "answer_arxiv_id": ["1702.08272", "2012.00057", "1811.01292"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_7030"} +{"question": "What studies discuss multi-modal video adaptation?", "answer": ["Multi-Modal Domain Adaptation for Fine-Grained Action Recognition", "Learning Cross-Modal Contrastive Features for Video Domain Adaptation", "Overcoming Label Noise for Source-free Unsupervised Video Domain Adaptation", "CycDA: Unsupervised Cycle Domain Adaptation to Learn from Image to Video", "Audio-Adaptive Activity Recognition Across Video Domains"], "answer_arxiv_id": ["2001.09691", "2108.11974", "2311.18572", "2203.16244", "2203.14240"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_7031"} +{"question": "Can you provide examples of studies on the universal approximation and error bounds of Fourier Neural Operators?", "answer": ["On universal approximation and error bounds for Fourier Neural Operators"], "answer_arxiv_id": ["2107.07562"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_7032"} +{"question": "Can you name studies that aim to disentangle semantic, domain, and noise variables, and use semantic variables that are better aligned with target graphs for prediction?", "answer": ["Graph Domain Adaptation: A Generative View"], "answer_arxiv_id": ["2106.07482"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_7033"} +{"question": "Did any paper explore the robustness of LLM-based dialogue evaluators using perturbation strategies?", "answer": ["A Comprehensive Analysis of the Effectiveness of Large Language Models\n as Automatic Dialogue Evaluators"], "answer_arxiv_id": ["2312.15407"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_7034"} +{"question": "In what work is the Blended Diffusion method, which relies on a user-provided background mask, discussed for use in image-to-image translation?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images"], "answer_arxiv_id": ["2111.14818"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7035"} +{"question": "Any works on neural network-based video compression algorithms?", "answer": ["Learning for Video Compression with Hierarchical Quality and Recurrent\n Enhancement", "DVC: An End-to-end Deep Video Compression Framework", "Neural network-based arithmetic coding of intra prediction modes in HEVC", "ELF-VC: Efficient Learned Flexible-Rate Video Coding", "Deep Contextual Video Compression", "Video Compression through Image Interpolation"], "answer_arxiv_id": ["2003.01966", "1812.00101", "1709.05737", "2104.14335", "2109.15047", "1804.06919"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_7036"} +{"question": "Could you provide some published works on benchmarks such as HAB, ManipulaTHOR and ThreeDWorld Transport Challenge for mobile manipulation tasks?", "answer": ["Habitat 2.0: Training Home Assistants to Rearrange their Habitat", "ManipulaTHOR: A Framework for Visual Object Manipulation", "The ThreeDWorld Transport Challenge: A Visually Guided Task-and-Motion Planning Benchmark for Physically Realistic Embodied AI"], "answer_arxiv_id": ["2106.14405", "2104.11213", "2103.14025"], "source_meta": {"published_time": "20220906"}, "qid": "AutoScholarQuery_train_7037"} +{"question": "Which works proposed Intermediate CTC and Self-conditioned CTC?", "answer": ["Intermediate Loss Regularization for CTC-based Speech Recognition", "Relaxing the Conditional Independence Assumption of CTC-based ASR by\n Conditioning on Intermediate Predictions"], "answer_arxiv_id": ["2102.03216", "2104.02724"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_7038"} +{"question": "Which paper was the first to adopt ensemble Q-learning for DRL?", "answer": ["Averaged-DQN: Variance Reduction and Stabilization for Deep Reinforcement Learning"], "answer_arxiv_id": ["1611.01929"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_train_7039"} +{"question": "Is there any study using transformers to process graphs and outperform classical GNN models?", "answer": ["GraphiT: Encoding Graph Structure in Transformers"], "answer_arxiv_id": ["2106.05667"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_7040"} +{"question": "Which papers presented the concept of distillation techniques in speeding up the inference process?", "answer": ["On Distillation of Guided Diffusion Models", "Progressive Distillation for Fast Sampling of Diffusion Models", "Knowledge Distillation in Iterative Generative Models for Improved\n Sampling Speed"], "answer_arxiv_id": ["2210.03142v3", "2202.00512", "2101.02388"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_7041"} +{"question": "Are there any studies on developing multi-modal LLMs in the speech modality?", "answer": ["SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal\n Conversational Abilities", "LLaSM: Large Language and Speech Model"], "answer_arxiv_id": ["2305.11000", "2308.15930"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_7042"} +{"question": "In what papers was prompt learning first proposed for natural language processing?", "answer": ["Factual Probing Is [MASK]: Learning vs. Learning to Recall", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2104.05240", "2101.00190", "2104.08691", "2107.13586v1"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_7043"} +{"question": "Which studies focused on curiosity-based intrinsic motivation via model error?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Curiosity-driven Exploration by Self-supervised Prediction", "BYOL-Explore: Exploration by Bootstrapped Prediction"], "answer_arxiv_id": ["1507.00814", "1705.05363", "2206.08332"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_7044"} +{"question": "Could you provide me some works that apply the Shapley values in feature-based explanations?", "answer": ["A Unified Approach to Interpreting Model Predictions", "Axiomatic Attribution for Deep Networks", "Consistent Individualized Feature Attribution for Tree Ensembles"], "answer_arxiv_id": ["1705.07874", "1703.01365", "1802.03888"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_7045"} +{"question": "Which study presented a neural ODE model and provided a generalization bound in a transfer learning context?", "answer": ["LEADS: Learning Dynamical Systems that Generalize Across Environments"], "answer_arxiv_id": ["2106.04546"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_7046"} +{"question": "Which studies are related to selective classification?", "answer": ["The Relationship Between Agnostic Selective Classification Active Learning and the Disagreement Coefficient"], "answer_arxiv_id": ["1703.06536"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_7047"} +{"question": "Which studies have emerged in the field of 2D-AD?", "answer": ["A Unified Model for Multi-class Anomaly Detection", "Explicit Boundary Guided Semi-Push-Pull Contrastive Learning for Supervised Anomaly Detection"], "answer_arxiv_id": ["2206.03687", "2207.01463"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_7048"} +{"question": "In the context of RL, what papers are about offline pretraining?", "answer": ["Improving Sample Efficiency in Model-Free Reinforcement Learning from Images", "Mask-based Latent Reconstruction for Reinforcement Learning", "Masked Contrastive Representation Learning for Reinforcement Learning", "Masked Visual Pre-training for Motor Control", "Data-Efficient Reinforcement Learning with Self-Predictive Representations", "Pretraining Representations for Data-Efficient Reinforcement Learning", "Reinforcement Learning with Action-Free Pre-Training from Videos", "On the Importance of Feature Decorrelation for Unsupervised Representation Learning in Reinforcement Learning", "Accelerating Reinforcement Learning with Learned Skill Priors", "OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning", "Parrot: Data-Driven Behavioral Priors for Reinforcement Learning", "Accelerating Reinforcement Learning with Learned Skill Priors", "Hierarchical Few-Shot Imitation with Skill Transition Models", "TRAIL: Near-Optimal Imitation Learning with Suboptimal Data", "CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "DRIBO: Robust Deep Reinforcement Learning via Multi-View Information Bottleneck", "Representation Learning with Contrastive Predictive Coding", "Decoupling Representation Learning from Reinforcement Learning"], "answer_arxiv_id": ["1910.01741", "2201.12096", "2010.07470", "2203.06173", "2007.05929", "2106.04799", "2203.13880", "2306.05637", "2010.11944", "2010.13611", "2011.10024", "2010.11944", "2107.08981", "2110.14770", "2004.04136", "2102.13268", "1807.03748", "2009.08319"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_7049"} +{"question": "Which research papers look into the use of hypervolume improvement in multi-objective Bayesian optimization?", "answer": ["Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization"], "answer_arxiv_id": ["2006.05078"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_7050"} +{"question": "What papers have studied the application of conditional SI in problems?", "answer": ["Selective Sequential Model Selection", "Selective inference with a randomized response", "Selective Inference for Group-Sparse Linear Models", "Post-Selection Inference for Changepoint Detection Algorithms with Application to Copy Number Variation Data", "Computing Valid p-value for Optimal Changepoint by Selective Inference using Dynamic Programming", "Valid and Exact Statistical Inference for Multi-dimensional Multiple Change-Points by Selective Inference", "Valid Inference Corrected for Outlier Removal", "Conditional Selective Inference for Robust Regression and Outlier Detection using Piecewise-Linear Homotopy Continuation", "Computing Valid p-values for Image Segmentation by Selective Inference", "More Powerful Conditional Selective Inference for Generalized Lasso by Parametric Programming", "Exact Post-Selection Inference for Sequential Regression Procedures", "More Powerful and General Selective Inference for Stepwise Feature Selection using the Homotopy Continuation Approach", "Fast and More Powerful Selective Inference for Sparse High-order Interaction Model"], "answer_arxiv_id": ["1512.02565", "1507.06739", "1607.08211v1", "1812.03644", "2002.09132v2", "2110.08989", "1711.10635", "2104.10840", "1906.00629v2", "2105.04920", "1401.3889", "2012.13545v2", "2106.04929v1"], "source_meta": {"published_time": "20230106"}, "qid": "AutoScholarQuery_train_7051"} +{"question": "Who performed extensive studies on the generalisation performance of SSL algorithms on medical data?", "answer": ["Robust and Efficient Medical Imaging with Self-Supervision"], "answer_arxiv_id": ["2205.09723"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_7052"} +{"question": "What paper proposed using progressive distillation in diffusion models and found improved performance with original image prediction?", "answer": ["Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2202.00512"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7053"} +{"question": "Could you provide some works that exploit the non-auto-regressive nature of CTC?", "answer": ["Mask CTC: Non-Autoregressive End-to-End ASR with CTC and Mask Predict", "Glancing Transformer for Non-Autoregressive Neural Machine Translation"], "answer_arxiv_id": ["2005.08700", "2008.07905"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_7054"} +{"question": "Which papers propose heuristic metrics for comparing the quality of two sets of images?", "answer": ["Training Generative Adversarial Networks with Limited Data", "Improved Techniques for Training GANs", "Fourier Spectrum Discrepancies in Deep Network Generated Images", "Spatial Frequency Bias in Convolutional Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN", "Generalization and Equilibrium in Generative Adversarial Nets (GANs)", "Classification Accuracy Score for Conditional Generative Models"], "answer_arxiv_id": ["2006.06676", "1606.03498", "1911.06465", "2010.01473", "1912.04958", "1703.00573", "1905.10887"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_7055"} +{"question": "Which works established out-of-distribution benchmarks for measuring model performance on images with common corruptions?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Surface Variations"], "answer_arxiv_id": ["1807.01697"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_7056"} +{"question": "Which works consider distances between the tokens in the input sequence as a part of the relative position encoding?", "answer": ["Self-Attention with Relative Position Representations"], "answer_arxiv_id": ["1803.02155"], "source_meta": {"published_time": "20210222"}, "qid": "AutoScholarQuery_train_7057"} +{"question": "What research employs optical flow information for video deblurring?", "answer": ["Generalized Video Deblurring for Dynamic Scenes", "Cascaded Deep Video Deblurring Using Temporal Sharpness Prior"], "answer_arxiv_id": ["1507.02438", "2004.02501"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_7058"} +{"question": "What studies applied episodic memory buffers to mimic hippocampal episodic control and rapidly assimilate recent experience?", "answer": ["Model-Free Episodic Control", "Neural Episodic Control"], "answer_arxiv_id": ["1606.04460", "1703.01988"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_7059"} +{"question": "Which are the works referenced about Generative Flow Networks?", "answer": ["Biological Sequence Design with GFlowNets", "Generative Flow Networks for Discrete Probabilistic Modeling", "Bayesian Structure Learning with Generative Flow Networks", "GFlowCausal: Generative Flow Networks for Causal Discovery", "CFlowNets: Continuous Control with Generative Flow Networks"], "answer_arxiv_id": ["2203.04115", "2202.01361", "2202.13903", "2210.08185", "2303.02430"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_7060"} +{"question": "Which studies incorporated appearance embeddings to handle appearance variations in the scene?", "answer": ["NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections"], "answer_arxiv_id": ["2008.02268"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_7061"} +{"question": "What papers propose different ways to represent 3D features in the field of unsupervised pre-training?", "answer": ["CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D\n Point Cloud Understanding", "PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts", "Masked Autoencoders for Point Cloud Self-supervised Learning", "Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point\n Modeling", "Self-Supervised Pretraining of 3D Features on any Point-Cloud", "Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud\n Pre-training"], "answer_arxiv_id": ["2203.00680", "2007.10985", "2012.09165", "2203.06604", "2111.14819", "2101.02691", "2205.14401"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_7062"} +{"question": "Could you cite some studies that proposed averaging the parameters of multiple fine-tuned models in the multi-task setting?", "answer": ["Editing Models with Task Arithmetic", "Branch-Train-Merge: Embarrassingly Parallel Training of Expert Language Models", "Patching open-vocabulary modelsby interpolating weights", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time"], "answer_arxiv_id": ["2212.04089", "2208.03306", "2208.05592", "2203.05482"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_7063"} +{"question": "What work compared the features learnt by one-hidden layer ReLU networks for the square loss and the cross-entropy loss with univariate data?", "answer": ["Regression as Classification: Influence of Task Formulation on Neural Network Features"], "answer_arxiv_id": ["2211.05641v2"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_7064"} +{"question": "Which research papers have focused on introducing patch-level contrastive losses for visual-language pretraining?", "answer": ["FILIP: Fine-grained Interactive Language-Image Pre-Training", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting"], "answer_arxiv_id": ["2111.07783", "2112.01518"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7065"} +{"question": "What works advocate for NLI-based approaches in addressing error-propagation in QA?", "answer": ["SummaC: Re-Visiting NLI-based Models for Inconsistency Detection in\n Summarization", "X-PARADE: Cross-Lingual Textual Entailment and Information Divergence\n across Paragraphs"], "answer_arxiv_id": ["2111.09525", "2309.08873"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_7066"} +{"question": "Could you provide me some works about identifying/learning linear dynamical systems from measurements?", "answer": ["Learning Linear Dynamical Systems with Semi-Parametric Least Squares", "Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification", "Near optimal finite time identification of arbitrary linear dynamical systems", "Finite Time Identification in Unstable Linear Systems", "Linear System Identification via Atomic Norm Regularization", "Gradient Descent Learns Linear Dynamical Systems", "Spectral Filtering for General Linear Dynamical Systems", "Learning Linear Dynamical Systems via Spectral Filtering"], "answer_arxiv_id": ["1902.00768", "1802.08334", "1812.01251", "1710.01852", "1204.0590", "1609.05191", "1802.03981", "1711.00946"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_7067"} +{"question": "Can you name some papers that focus on the prediction model’s uncertainty-based exploration?", "answer": ["Exploration by Random Network Distillation", "Self-Supervised Exploration via Disagreement"], "answer_arxiv_id": ["1810.12894", "1906.04161"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7068"} +{"question": "What research works attempted to automatically produce transferable adversarial suffixes?", "answer": ["Universal and Transferable Adversarial Attacks on Aligned Language\n Models"], "answer_arxiv_id": ["2307.15043"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_7069"} +{"question": "Can you point me to studies that have addressed the issues of semi-supervised learning techniques, particularly using generative models and approximate Bayesian inference?", "answer": ["Semi-supervised Learning with Deep Generative Models"], "answer_arxiv_id": ["1406.5298"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_7070"} +{"question": "What related work exists that defines generated sample in a fixed resolution?", "answer": ["Diffusion Probabilistic Fields"], "answer_arxiv_id": ["2303.00165"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_7071"} +{"question": "Which studies have used code for logical planning?", "answer": ["Interactive Natural Language Processing", "LeTI: Learning to Generate from Textual Interactions", "Chameleon: Plug-and-Play Compositional Reasoning with Large Language\n Models"], "answer_arxiv_id": ["2305.13246", "2305.10314v2", "2304.09842"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_7072"} +{"question": "Which studies provided information-theoretic generalization bounds for Stochastic Gradient Langevine Dynamics?", "answer": ["Generalization Error Bounds for Noisy, Iterative Algorithms", "On Random Subset Generalization Error Bounds and the Stochastic Gradient Langevin Dynamics Algorithm", "Information-Theoretic Generalization Bounds for SGLD via Data-Dependent Estimates", "Sharpened Generalization Bounds based on Conditional Mutual Information and an Application to Noisy, Iterative Algorithms"], "answer_arxiv_id": ["1801.04295v1", "2010.10994", "1911.02151v3", "2004.12983v2"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_7073"} +{"question": "What papers discuss the efficiency of Vision Transformers and ConvNext?", "answer": ["Training data-efficient image transformers & distillation through attention", "How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers"], "answer_arxiv_id": ["2012.12877", "2106.10270"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_7074"} +{"question": "What works proposed methods for generating descriptions using large language models?", "answer": ["Visual Classification via Description from Large Language Models", "What does a platypus look like? Generating customized prompts for\n zero-shot image classification"], "answer_arxiv_id": ["2210.07183", "2209.03320"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_7075"} +{"question": "What is the closest work to the researcher's, which adjusts heuristic value by a policy estimated from the neural network?", "answer": ["Single-Agent Policy Tree Search With Guarantees", "Policy-Guided Heuristic Search with Guarantees"], "answer_arxiv_id": ["1811.10928", "2103.11505"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7076"} +{"question": "Which works involved the use of Weight-sharing strategies like LargeKernel3D and Link to handle overfitting in 3D representation learning?", "answer": ["LinK: Linear Kernel for LiDAR-based 3D Perception"], "answer_arxiv_id": ["2303.16094"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_7077"} +{"question": "Which work took Fully Convolutional Networks (FCN) as the dominant approach for semantic segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_7078"} +{"question": "What papers are about the learning mixtures of Generative Adversarial Networks?", "answer": ["MGAN: Training Generative Adversarial Nets with Multiple Generators", "AdaGAN: Boosting Generative Models"], "answer_arxiv_id": ["1708.02556", "1701.02386"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_7079"} +{"question": "Have there been any works about mitigating the limitations by applying the method of data augmentation?", "answer": ["Few-shot Object Grounding and Mapping for Natural Language Robot Instruction Following", "Learning from Unlabeled 3D Environments for Vision-and-Language Navigation"], "answer_arxiv_id": ["2011.07384", "2208.11781"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_7080"} +{"question": "Any papers about enhancing reasoning performance by optimizing demonstrations selection within a prompt?", "answer": ["Automatic Chain of Thought Prompting in Large Language Models", "Complexity-Based Prompting for Multi-Step Reasoning", "Synthetic Prompting: Generating Chain-of-Thought Demonstrations for\n Large Language Models", "Boosted Prompt Ensembles for Large Language Models"], "answer_arxiv_id": ["2210.03493", "2210.00720", "2302.00618", "2304.05970"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_7081"} +{"question": "What research demonstrates the power of compositional generalization in Diffusion models?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2112.10741"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_7082"} +{"question": "Could you provide me some works on algorithms achieving the minimax optimal regret in adversarial regime?", "answer": ["No Internal Regret via Neighborhood Watch"], "answer_arxiv_id": ["1108.6088"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_7083"} +{"question": "What papers detail a huge performance gap due to the design of different prompts in language tasks?", "answer": ["Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2112.00114", "2201.11903"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_7084"} +{"question": "Can you name some studies that worked on surrogate models in hyperparameter optimization?", "answer": ["Practical Bayesian Optimization of Machine Learning Algorithms", "BANANAS: Bayesian Optimization with Neural Architectures for Neural Architecture Search"], "answer_arxiv_id": ["1206.2944", "1910.11858"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_7085"} +{"question": "Which work proposed a Bayesian inference framework explaining how ICL works with formatting differences?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_7086"} +{"question": "What paper proposes a defensive strategy of utilizing safety filters on input prompt sub-strings for LLMs?", "answer": ["Certifying LLM Safety against Adversarial Prompting"], "answer_arxiv_id": ["2309.02705"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_7087"} +{"question": "What works are about distributional robust optimization?", "answer": ["Certifying Some Distributional Robustness with Principled Adversarial Training", "Generalizing to Unseen Domains via Adversarial Data Augmentation", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Improved OOD Generalization via Adversarial Training and Pre-training", "Large-Scale Methods for Distributionally Robust Optimization"], "answer_arxiv_id": ["1710.10571v5", "1805.12018", "1911.08731", "2105.11144", "2010.05893"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_7088"} +{"question": "Could you provide me some works about explicit NeRF modeling?", "answer": ["Neural Sparse Voxel Fields", "PlenOctrees for Real-time Rendering of Neural Radiance Fields"], "answer_arxiv_id": ["2007.11571", "2103.14024"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_7089"} +{"question": "What are some recent studies that enabled the compression of large neural networks into smaller networks without using any real data?", "answer": ["Zero-shot Knowledge Transfer via Adversarial Belief Matching", "Contrastive Model Inversion for Data-Free Knowledge Distillation", "Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion", "Data-Free Learning of Student Networks", "Dream Distillation: A Data-Independent Model Compression Framework", "The Knowledge Within: Methods for Data-Free Model Compression", "DENSE: Data-Free One-Shot Federated Learning"], "answer_arxiv_id": ["1905.09768", "2105.08584", "1912.08795", "1904.01186", "1905.07072", "1912.01274", "2112.12371"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_7090"} +{"question": "Which work describes the implementation of atrous convolutions in DeepLabv3+ to enhance feature extraction?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation"], "answer_arxiv_id": ["1505.04597"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7091"} +{"question": "What research implemented the region subdivision method?", "answer": ["Complexity of Linear Regions in Deep Networks", "SplineCam: Exact Visualization and Characterization of Deep Network Geometry and Decision Boundaries"], "answer_arxiv_id": ["1901.09021", "2302.12828"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_7092"} +{"question": "Which papers focus on the computation of approximately Maximum Makespan (MMS) allocations for goods?", "answer": ["An Improved Approximation Algorithm for Maximin Shares"], "answer_arxiv_id": ["1903.00029v3"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_7093"} +{"question": "What are some works that employ sparsity assumptions on the nonlinear generating function?", "answer": ["Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICA"], "answer_arxiv_id": ["2107.10098"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_7094"} +{"question": "Could you point the works that apply parameter isolation methods in continual learning?", "answer": ["Learn to Grow: A Continual Structure Learning Framework for Overcoming\n Catastrophic Forgetting", "DER: Dynamically Expandable Representation for Class Incremental\n Learning", "Continual Learning of a Mixed Sequence of Similar and Dissimilar Tasks"], "answer_arxiv_id": ["1904.00310", "2103.16788", "2112.10017"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_7095"} +{"question": "Which works have applied uncertainty scores provided by Deep Neural Networks (DNNs) to re-weighting training strategy?", "answer": ["Long-Tailed Recognition via Weight Balancing", "Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss", "Decoupling Representation and Classifier for Long-Tailed Recognition", "BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed\n Visual Recognition", "Long-Tailed Classification by Keeping the Good and Removing the Bad\n Momentum Causal Effect"], "answer_arxiv_id": ["2203.14197", "1906.07413", "1910.09217", "1912.02413", "2009.12991"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_7096"} +{"question": "Could you provide a work that adopted masked image modeling to perform in-context learning with supervised datasets?", "answer": ["Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning"], "answer_arxiv_id": ["2212.02499"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_7097"} +{"question": "Which studies introduced regularizations such as the spatial gradient constraint based on the Eikonal equation to improve the convergence of neural networks?", "answer": ["Implicit Geometric Regularization for Learning Shapes"], "answer_arxiv_id": ["2002.10099"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_7098"} +{"question": "Can you name the works that mask the low logits tokens in the generated contents?", "answer": ["Active Retrieval Augmented Generation"], "answer_arxiv_id": ["2305.06983"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_7099"} +{"question": "What work is conducted to address the drift of classifier weights using language information in the field of FSCIL?", "answer": ["Few-Shot Class-Incremental Learning by Sampling Multi-Phase Tasks"], "answer_arxiv_id": ["2203.17030"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_7100"} +{"question": "Which works have tried to improve average-based aggregations in Federated Learning?", "answer": ["Bayesian Nonparametric Federated Learning of Neural Networks", "Federated learning with matched averaging"], "answer_arxiv_id": ["1905.12022", "2002.06440"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_7101"} +{"question": "Which works dealt with VEA as a classification task and predicted the emotion within an image?", "answer": ["Learning Multi-level Deep Representations for Image Emotion\n Classification", "Stimuli-Aware Visual Emotion Analysis", "SOLVER: Scene-Object Interrelated Visual Emotion Reasoning Network"], "answer_arxiv_id": ["1611.07145", "2109.01812", "2110.12334"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_7102"} +{"question": "What papers have introduced extra scale interaction modules to improve feature extraction in referring image detection and segmentation?", "answer": ["Shatter and Gather: Learning Referring Image Segmentation with Text\n Supervision", "Cross-Modal Self-Attention Network for Referring Image Segmentation"], "answer_arxiv_id": ["2308.15512", "1904.04745"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_7103"} +{"question": "Which researches use 3D data as supervision for generating 3D humans?", "answer": ["Learning to Dress 3D People in Generative Clothing", "gDNA: Towards Generative Detailed Neural Avatars", "SMPLicit: Topology-aware Generative Model for Clothed People", "NPMs: Neural Parametric Models for 3D Deformable Shapes"], "answer_arxiv_id": ["1907.13615", "2201.04123", "2103.06871", "2104.00702"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_7104"} +{"question": "Could you provide me with some works that propose OCR-free methods for VDU?", "answer": ["PaLI-3 Vision Language Models: Smaller, Faster, Stronger", "Monkey: Image Resolution and Text Label Are Important Things for Large\n Multi-modal Models"], "answer_arxiv_id": ["2310.09199", "2311.06607"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_7105"} +{"question": "Are there any research papers that highlighted the connection between the lifted neural network and two-phase CHL?", "answer": ["Bilevel Programs Meet Deep Learning: A Unifying View on Inference Learning Methods"], "answer_arxiv_id": ["2105.07231"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_7106"} +{"question": "Could you provide me some works about two-stage detectors in object detection?", "answer": ["Rich feature hierarchies for accurate object detection and semantic\n segmentation", "Fast R-CNN", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks", "Feature Pyramid Networks for Object Detection", "Mask R-CNN", "Region Proposal by Guided Anchoring", "Hybrid Task Cascade for Instance Segmentation"], "answer_arxiv_id": ["1311.2524", "1504.08083", "1506.01497", "1612.03144", "1703.06870", "1901.03278", "1901.07518"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_7107"} +{"question": "Could you provide me with references of the works that introduced dihedral angles to the distance and bond angle information in their models?", "answer": ["ComENet: Towards Complete and Efficient Message Passing for 3D Molecular Graphs"], "answer_arxiv_id": ["2206.08515"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_7108"} +{"question": "What research can be considered as a visual counterpart to the chain-of-thought technique used in large language models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_7109"} +{"question": "What paper established that neural network verification is generally NP-complete?", "answer": ["Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks"], "answer_arxiv_id": ["1702.01135"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_7110"} +{"question": "Which papers talked about techniques for text-to-image synthesis using normalizing flow-based priors?", "answer": ["Latent Normalizing Flows for Many-to-Many Cross-Domain Mappings"], "answer_arxiv_id": ["2002.06661"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_7111"} +{"question": "Could you provide me some works related to the distributed dual averaging method?", "answer": ["Dual Averaging for Distributed Optimization: Convergence Analysis and Network Scaling"], "answer_arxiv_id": ["1005.2012v3"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_7112"} +{"question": "Can you list some studies which focuse on multi-modal instruction-following models?", "answer": ["GPT-4 Technical Report", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models"], "answer_arxiv_id": ["2303.08774", "2304.15010", "2303.16199", "2304.08485", "2304.10592"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_7113"} +{"question": "Which studies used advanced deep generative learning for designing or discovering novel molecules?", "answer": ["Auto-Encoding Variational Bayes", "Generative Adversarial Nets", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1312.6114", "1406.2661", "1907.05600", "2006.11239"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_7114"} +{"question": "Can you list some works where meta-learning methods in dataset distillation are introduced?", "answer": ["Flexible Dataset Distillation: Learn Labels Instead of Images", "Dataset Meta-Learning from Kernel Ridge-Regression", "Dataset Distillation with Infinitely Wide Convolutional Networks", "Dataset Distillation using Neural Feature Regression", "Efficient Dataset Distillation using Random Feature Approximation", "Remember the Past: Distilling Datasets into Addressable Memories for Neural Networks", "Meta Knowledge Condensation for Federated Learning"], "answer_arxiv_id": ["2006.08572", "2011.00050", "2107.13034", "2206.00719", "2210.12067", "2206.02916", "2209.14851"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7115"} +{"question": "What work studies contrastive learning process but only with single-layer networks under toy problems?", "answer": ["Toward Understanding the Feature Learning Process of Self-supervised Contrastive Learning"], "answer_arxiv_id": ["2105.15134"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_7116"} +{"question": "Any studies about applying machine learning to the odometry problem in radar systems?", "answer": ["milliEgo: Single-chip mmWave Radar Aided Egomotion Estimation via Deep\n Sensor Fusion", "Milli-RIO: Ego-Motion Estimation with Low-Cost Millimetre-Wave Radar"], "answer_arxiv_id": ["2006.02266", "1909.05774"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_7117"} +{"question": "Could you provide me some works that have explored the use of selfies for face recognition?", "answer": ["WSD: Wild Selfie Dataset for Face Recognition in Selfie Images", "Fun Selfie Filters in Face Recognition: Impact Assessment and Removal"], "answer_arxiv_id": ["2302.07245", "2202.06022"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_7118"} +{"question": "Which papers propose learning multiple tasks concurrently via RL?", "answer": ["Policy Distillation", "Learning Shared Representations in Multi-task Reinforcement Learning", "EPOpt: Learning Robust Neural Network Policies Using Model Ensembles", "Modular Multitask Reinforcement Learning with Policy Sketches", "Sharing Knowledge in Multi-Task Deep Reinforcement Learning", "Gradient Surgery for Multi-Task Learning", "Multitask Soft Option Learning", "Multi-Task Reinforcement Learning with Context-based Representations"], "answer_arxiv_id": ["1511.06295", "1603.02041v1", "1610.01283", "1611.01796", "2401.09561", "2001.06782", "1904.01033", "2102.06177"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_7119"} +{"question": "In what papers have Large Language Models been harnessed to guide embodied actions?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2204.01691", "2210.03629"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_7120"} +{"question": "Could you provide me some studies on the use of neural networks in point processes?", "answer": ["Modeling The Intensity Function Of Point Process Via Recurrent Neural Networks", "Neural Spatio-Temporal Point Processes"], "answer_arxiv_id": ["1705.08982", "2011.04583"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_7121"} +{"question": "What papers propose the usage of rehearsal-based methods for addressing catastrophic forgetting in DNNs?", "answer": ["Replay in Deep Learning: Current Approaches and Missing Biological Elements"], "answer_arxiv_id": ["2104.04132v2"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_7122"} +{"question": "Which works discuss the use of block coordinate descent (BCD) in solving the SMF training problem?", "answer": ["Coordinate Descent Algorithms", "Supervised Dictionary Learning", "Supervised PCA: A Multiobjective Approach"], "answer_arxiv_id": ["1502.04759v1", "0809.3083v1", "2011.05309"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_7123"} +{"question": "Which articles studied the generation of non-factual information by language models?", "answer": ["On Faithfulness and Factuality in Abstractive Summarization", "Entity-Based Knowledge Conflicts in Question Answering", "Evaluating Factuality in Text Simplification"], "answer_arxiv_id": ["2005.00661", "2109.05052", "2204.07562"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_7124"} +{"question": "What are some works on offline tabular zero-sum Markov games?", "answer": ["When is Offline Two-Player Zero-Sum Markov Game Solvable?"], "answer_arxiv_id": ["2201.03522"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_7125"} +{"question": "Which work is about Llama 2-Chat?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2307.09288"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_7126"} +{"question": "What studies have been done on models in the first category of large multimodal models (LMM) that has been trained from scratch or using smaller language models for text processing?", "answer": ["Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "CoCa: Contrastive Captioners are Image-Text Foundation Models"], "answer_arxiv_id": ["2206.08916", "2202.03052", "2205.01917"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_7127"} +{"question": "Could you give me examples of research that utilized the learnable temporal decay mechanism for input and hidden state of GRU?", "answer": ["BRITS: Bidirectional Recurrent Imputation for Time Series"], "answer_arxiv_id": ["1805.10572"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_7128"} +{"question": "Could you provide me some studies about continuous-depth GNNs?", "answer": ["Dissecting the Diffusion Process in Linear Graph Convolutional Networks", "LT-OCF: Learnable-Time ODE-based Collaborative Filtering", "Climate Modeling with Neural Diffusion Equations"], "answer_arxiv_id": ["2102.10739", "2108.06208", "2111.06011"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_7129"} +{"question": "What are the recent works that extend mixup methods to more than two elements or regression tasks?", "answer": ["Co-Mixup: Saliency Guided Joint Mixup with Supermodular Diversity", "SuperMix: Supervising the Mixing Data Augmentation", "C-Mixup: Improving Generalization in Regression"], "answer_arxiv_id": ["2102.03065", "2003.05034", "2210.05775"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_7130"} +{"question": "Can you provide some resources on Automated Feature Transformation (AFT)?", "answer": ["Techniques for Automated Machine Learning"], "answer_arxiv_id": ["1907.08908"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_7131"} +{"question": "Can you provide some studies that utilized self-supervised learning in long-tailed learning?", "answer": ["Contrastive Learning based Hybrid Networks for Long-Tailed Image Classification", "Self Supervision to Distillation for Long-Tailed Visual Recognition"], "answer_arxiv_id": ["2103.14267", "2109.04075"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_7132"} +{"question": "Which works have derived efficient estimators for policy evaluation and subsequent optimization for dosages?", "answer": ["Policy Evaluation and Optimization with Continuous Treatments"], "answer_arxiv_id": ["1802.06037"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_7133"} +{"question": "What papers have proposed to fine-tune generative models to manipulate images based on text descriptions?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "L-CAD: Language-based Colorization with Any-level Descriptions using\n Diffusion Priors", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2302.05543", "2305.15217", "2211.09800"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_7134"} +{"question": "What studies have focused on the behavior of neural networks when both the width and depth are very large?", "answer": ["Exponential expressivity in deep neural networks through transient chaos", "The Future is Log-Gaussian: ResNets and Their Infinite-Depth-and-Width Limit at Initialization", "The Neural Covariance SDE: Shaped Infinite Depth-and-Width Networks at Initialization"], "answer_arxiv_id": ["1606.05340", "2106.04013", "2206.02768"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_7135"} +{"question": "Which works have advanced the understanding of two-level smooth SCO?", "answer": ["Stochastic Compositional Gradient Descent: Algorithms for Minimizing Compositions of Expected-Value Functions", "A Stochastic Composite Gradient Method with Incremental Variance Reduction", "A Single Time-Scale Stochastic Approximation Method for Nested Stochastic Optimization", "An Online Method for A Class of Distributionally Robust Optimization with Non-Convex Objectives", "Solving Stochastic Compositional Optimization is Nearly as Easy as Solving Stochastic Optimization"], "answer_arxiv_id": ["1411.3803v1", "1906.10186", "1812.01094", "2006.10138", "2008.10847"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_7136"} +{"question": "In which work a method for learning to reconstruct CSG trees with arbitrary assembly orders is presented?", "answer": ["UCSG-Net - Unsupervised Discovering of Constructive Solid Geometry Tree"], "answer_arxiv_id": ["2006.09102"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7137"} +{"question": "Any studies use alternative benchmarks such as dynamic regret and approximate regret for adversarial contexts and reward functions?", "answer": ["Efficient Contextual Bandits in Non-stationary Worlds", "A New Algorithm for Non-stationary Contextual Bandits: Efficient, Optimal, and Parameter-free", "Adversarial Online Learning with Changing Action Sets: Efficient Algorithms with Approximate Regret Bounds"], "answer_arxiv_id": ["1708.01799v4", "1902.00980", "2003.03490v2"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_7138"} +{"question": "Which project utilized a second-order system to overcome the issue of oversmoothing in deep graph neural networks?", "answer": ["Graph-Coupled Oscillator Networks"], "answer_arxiv_id": ["2202.02296"], "source_meta": {"published_time": "20220611"}, "qid": "AutoScholarQuery_train_7139"} +{"question": "What paper describes a method that adaptively changes the learning rate for each layer?", "answer": ["Large Batch Training of Convolutional Networks"], "answer_arxiv_id": ["1708.03888"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_7140"} +{"question": "Could you provide me some works that have made progress in multi-concept customization?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion", "SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "Cones: Concept Neurons in Diffusion Models for Customized Generation"], "answer_arxiv_id": ["2212.04488", "2303.11305", "2303.05125"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7141"} +{"question": "Which papers have contributed methods for visual grounding that utilize bounding boxes during supervised learning?", "answer": ["TransVG: End-to-End Visual Grounding with Transformers", "Improving Visual Grounding by Encouraging Consistent Gradient-based\n Explanations", "MDETR -- Modulated Detection for End-to-End Multi-Modal Understanding"], "answer_arxiv_id": ["2104.08541", "2206.15462", "2104.12763"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_7142"} +{"question": "What research paper proposes masked attention to improve upon previous methods?", "answer": ["Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2112.01527"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_7143"} +{"question": "Which work discusses the Memorizing Transformer performing a nearest-neighbor search over previous keys?", "answer": ["Memorizing Transformers"], "answer_arxiv_id": ["2203.08913"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7144"} +{"question": "Which benchmark introduced a workflow that helps a human annotate keypoint tracks in a sequence of video frames?", "answer": ["TAP-Vid: A Benchmark for Tracking Any Point in a Video"], "answer_arxiv_id": ["2211.03726"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_7145"} +{"question": "Which studies refer to automatic prompt generation or using RL for discrete prompt optimization?", "answer": ["Making Pre-trained Language Models Better Few-shot Learners", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts", "TEMPERA: Test-Time Prompting via Reinforcement Learning", "RLPrompt: Optimizing Discrete Text Prompts with Reinforcement Learning"], "answer_arxiv_id": ["2012.15723", "2010.15980", "2211.11890", "2205.12548"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_7146"} +{"question": "Which works tackled the problem of customization by producing a representation of the subject to be used for controlled generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_7147"} +{"question": "Which research discuss about exploring the 𝒲+limit-from𝒲 space?", "answer": ["StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation"], "answer_arxiv_id": ["2011.12799"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_7148"} +{"question": "What studies explored human pose estimation using Event Camera?", "answer": ["Lifting Monocular Events to 3D Human Poses", "EventCap: Monocular 3D Capture of High-Speed Human Motions using an\n Event Camera", "EventHPE: Event-based 3D Human Pose and Shape Estimation", "Efficient Human Pose Estimation via 3D Event Point Cloud"], "answer_arxiv_id": ["2104.10609", "1908.11505", "2108.06819", "2206.04511"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_7149"} +{"question": "Could you provide me some recent studies about diffusion-based models in text-to-image generation?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction\n Tuning", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2205.11487", "2112.10752", "2309.02591", "2204.06125"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_7150"} +{"question": "What papers discussed the issue of deep neural networks overfitting noisy labels?", "answer": ["A Closer Look at Memorization in Deep Networks"], "answer_arxiv_id": ["1706.05394"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_7151"} +{"question": "Could you provide me some studies that fine-tuned a pre-trained diffusion model on a synthetic dataset to generate multiple images of the same object?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2308.16512"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_7152"} +{"question": "Which papers demonstrate the effectiveness of diffusion models in optimizing the diversity-fidelity trade-off in guided image generation?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2105.05233", "2207.12598"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_7153"} +{"question": "Are there any studies where GFlowNets is jointly trained with an energy or reward function?", "answer": ["Generative Flow Networks for Discrete Probabilistic Modeling"], "answer_arxiv_id": ["2202.01361"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_7154"} +{"question": "What studies are about architecture-based methods in CL?", "answer": ["Progressive Neural Networks", "PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning"], "answer_arxiv_id": ["1606.04671", "1711.05769"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_7155"} +{"question": "Can you list some research that designed GNN models to utilize the graph structure?", "answer": ["Graph Attention Networks", "Diffusion Improves Graph Learning"], "answer_arxiv_id": ["1710.10903", "1911.05485"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_7156"} +{"question": "Which works have used a part-based image partitioning approach for object ReID?", "answer": ["Beyond Part Models: Person Retrieval with Refined Part Pooling (and a\n Strong Convolutional Baseline)", "Learning Discriminative Features with Multiple Granularities for Person\n Re-Identification"], "answer_arxiv_id": ["1711.09349", "1804.01438"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_7157"} +{"question": "What are the studies that demonstrate the effectiveness of pretraining representation in Offline RL settings?", "answer": ["Provable Representation Learning for Imitation Learning via Bi-level Optimization", "Pretraining Representations for Data-Efficient Reinforcement Learning", "Provable Representation Learning for Imitation with Contrastive Fourier Features"], "answer_arxiv_id": ["2002.10544", "2106.04799", "2105.12272"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_7158"} +{"question": "Which early works proposed direct transformation of random vectors to video clips for 2D video generation?", "answer": ["Generating Videos with Scene Dynamics", "Temporal Generative Adversarial Nets with Singular Value Clipping"], "answer_arxiv_id": ["1609.02612", "1611.06624"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_7159"} +{"question": "Which papers introduced Large Language Models like GPT-3, T0, Flan-T5, Galactica and LLaMa?", "answer": ["Language Models are Few-Shot Learners", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "Scaling Instruction-Finetuned Language Models", "Galactica: A Large Language Model for Science", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2005.14165", "2110.08207", "2210.11416", "2211.09085", "2302.13971"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_7160"} +{"question": "What are some works that have approached scene completion by optimizing a signed distance field with only LiDAR measurements?", "answer": ["LODE: Locally Conditioned Eikonal Implicit Scene Completion from Sparse\n LiDAR"], "answer_arxiv_id": ["2302.14052"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_7161"} +{"question": "Could you provide me some works that conducted research on domain-agnostic approaches like contrastive learning and masked image modeling?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Unsupervised Feature Learning via Non-Parametric Instance-level\n Discrimination", "Contrastive Multiview Coding", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "BEiT: BERT Pre-Training of Image Transformers", "SimMIM: A Simple Framework for Masked Image Modeling", "Masked Autoencoders Are Scalable Vision Learners", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "Masked Siamese Networks for Label-Efficient Learning", "iBOT: Image BERT Pre-Training with Online Tokenizer", "data2vec: A General Framework for Self-supervised Learning in Speech,\n Vision and Language", "EVA: Exploring the Limits of Masked Visual Representation Learning at\n Scale"], "answer_arxiv_id": ["1807.03748", "1805.01978", "1906.05849", "1911.05722", "2002.05709", "2106.08254", "2111.09886", "2111.06377", "2112.09133", "2204.07141", "2111.07832", "2202.03555", "2211.07636"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_7162"} +{"question": "Which studies demonstrated that a novel query formulation could enhance the performance of DETR?", "answer": ["DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETR", "DN-DETR: Accelerate DETR Training by Introducing Query DeNoising"], "answer_arxiv_id": ["2201.12329", "2203.01305"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_7163"} +{"question": "What works were done on robust MDPs with uncertainty set?", "answer": ["Policy Gradient Method For Robust Reinforcement Learning"], "answer_arxiv_id": ["2205.07344"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_7164"} +{"question": "Which works investigated gradients of Self-Supervised Learning?", "answer": ["Exploring the Equivalence of Siamese Self-Supervised Learning via A\n Unified Gradient Framework"], "answer_arxiv_id": ["2112.05141"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_7165"} +{"question": "Which works demonstrate benefits from structured or 'importance' based strategies?", "answer": ["Auto-Scaling Vision Transformers without Training", "Token Merging: Your ViT But Faster", "A-ViT: Adaptive Tokens for Efficient Vision Transformer"], "answer_arxiv_id": ["2202.11921", "2210.09461", "2112.07658"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_7166"} +{"question": "What paper proposed an image-and-pose conditioned diffusion method for still fashion image animation?", "answer": ["DreamPose: Fashion Image-to-Video Synthesis via Stable Diffusion"], "answer_arxiv_id": ["2304.06025"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_7167"} +{"question": "Which studies utilize statistical methods to remove redundant features in automated feature generation?", "answer": ["SAFE: Scalable Automatic Feature Engineering Framework for Industrial Tasks"], "answer_arxiv_id": ["2003.02556"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_7168"} +{"question": "Can you name the studies that use VQ to perform a first-pass pruning over the dataset and then a multi-codebook quantization to compute more accurate distance estimates?", "answer": ["Accelerating Large-Scale Inference with Anisotropic Vector Quantization", "Billion-scale similarity search with GPUs"], "answer_arxiv_id": ["1908.10396", "1702.08734"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_7169"} +{"question": "What research explores adaptive algorithms for the convex-concave regime based on EG and AdaGrad stepsize?", "answer": ["A Universal Algorithm for Variational Inequalities Adaptive to Smoothness and Noise"], "answer_arxiv_id": ["1902.01637"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_7170"} +{"question": "What studies propose to integrate MLM and MIM and conduct VL pretraining in a joint manner?", "answer": ["MLIM: Vision-and-Language Model Pre-training with Masked Language and\n Image Modeling", "Masked Vision and Language Modeling for Multi-modal Representation\n Learning"], "answer_arxiv_id": ["2109.12178", "2208.02131"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_7171"} +{"question": "Which paper proposed the dual averaging algorithm for distributed convex optimization?", "answer": ["Dual Averaging for Distributed Optimization: Convergence Analysis and Network Scaling"], "answer_arxiv_id": ["1005.2012v3"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_7172"} +{"question": "Which works demonstrated that efficient RNNs like QRNNs and SRUs can leverage parallel scans?", "answer": ["Parallelizing Linear Recurrent Neural Nets Over Sequence Length"], "answer_arxiv_id": ["1709.04057"], "source_meta": {"published_time": "20220809"}, "qid": "AutoScholarQuery_train_7173"} +{"question": "What paper introduced Flow Matching to Riemannian manifolds?", "answer": ["Flow Matching on General Geometries"], "answer_arxiv_id": ["2302.03660"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_7174"} +{"question": "Could you specify any works about generalization in RL that use information-theoretic approaches, non-stationarity reduction, curriculum learning, planning, forward-backward representations, or diverse policies?", "answer": ["Reinforcement Learning Generalization with Surprise Minimization", "Deep Reinforcement and InfoMax Learning", "The Primacy Bias in Deep Reinforcement Learning", "Prioritized Level Replay", "Open-Ended Learning Leads to Generally Capable Agents", "Replay-Guided Adversarial Environment Design", "Evolving Curricula with Regret-Based Environment Design", "Procedural generalization by planning with self-supervised world models", "Learning One Representation to Optimize All Rewards", "One Solution is Not All You Need: Few-Shot Extrapolation via Structured MaxEnt RL"], "answer_arxiv_id": ["2004.12399", "2006.07217", "2205.07802", "2010.03934", "2107.12808", "2110.02439", "2203.01302", "2111.01587", "2103.07945", "2010.14484"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_7175"} +{"question": "Which works focus on improving the stability of reinforcement learning algorithms, particularly with the introduction of adaptive normalization of value targets?", "answer": ["Learning values across many orders of magnitude"], "answer_arxiv_id": ["1602.07714"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_7176"} +{"question": "What studies present zero-shot point cloud understanding in a training-free manner?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP", "PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning", "ConceptFusion: Open-set Multimodal 3D Mapping"], "answer_arxiv_id": ["2112.02413", "2211.11682", "2302.07241"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_7177"} +{"question": "Could you provide me some works that developed neural-net based architectures to learn from audio-visual correspondences?", "answer": ["Learning to Localize Sound Source in Visual Scenes", "Deep Multimodal Clustering for Unsupervised Audiovisual Learning", "Self-Supervised Learning of Audio-Visual Objects from Video", "Multiple Sound Sources Localization from Coarse to Fine", "Localizing Visual Sounds the Hard Way", "LEARNING SOUND LOCALIZATION BETTER FROM SEMANTICALLY SIMILAR SAMPLES", "Localizing Visual Sounds the Easy Way", "A Closer Look at Weakly-Supervised Audio-Visual Source Localization", "Audio-Visual Grouping Network for Sound Localization from Mixtures", "AV-SAM: Segment Anything Model Meets Audio-Visual Localization and Segmentation"], "answer_arxiv_id": ["1803.03849v1", "1807.03094", "2008.04237", "2007.06355", "2104.02691", "2202.03007", "2203.09324", "2209.09634", "2303.17056", "2305.01836"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7178"} +{"question": "Which works offer information about learning from uncurated datasets for safety alignment strategies?", "answer": ["Mistral 7B", "Zephyr: Direct Distillation of LM Alignment"], "answer_arxiv_id": ["2310.06825", "2310.16944"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_7179"} +{"question": "What studies are about the utilization of randomized value function methods?", "answer": ["Deep Exploration via Bootstrapped DQN", "Randomized Prior Functions for Deep Reinforcement Learning", "Deep Exploration via Randomized Value Functions", "Generalization and Exploration via Randomized Value Functions", "Efficient Exploration through Bayesian Deep Q-Networks", "The Uncertainty Bellman Equation and Exploration", "Successor Uncertainties: Exploration and Uncertainty in Temporal Difference Learning", "Temporal Difference Uncertainties as a Signal for Exploration"], "answer_arxiv_id": ["1602.04621", "1806.03335v2", "1703.07608", "1402.0635", "1802.04412", "1709.05380", "1810.06530", "2010.02255"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_7180"} +{"question": "Which works used loss re-weighting as a learning strategy in VRD benchmarks?", "answer": ["PCPL: Predicate-Correlation Perception Learning for Unbiased Scene Graph\n Generation", "Bridging Knowledge Graphs to Generate Scene Graphs", "Iterative Scene Graph Generation"], "answer_arxiv_id": ["2009.00893", "2001.02314", "2207.13440"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_7181"} +{"question": "Which research proposed the continual semantic segmentation method RECALL which uses a pre-trained model as an encoder?", "answer": ["RECALL: Replay-based Continual Learning in Semantic Segmentation"], "answer_arxiv_id": ["2108.03673"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_7182"} +{"question": "Can you give me some studies that discuss the techniques used for discovering options in the options framework?", "answer": ["A Laplacian Framework for Option Discovery in Reinforcement Learning", "Diversity is All You Need: Learning Skills without a Reward Function", "Generalizing Skills with Semi-Supervised Reinforcement Learning"], "answer_arxiv_id": ["1703.00956", "1802.06070", "1612.00429v2"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_7183"} +{"question": "Which research article includes a discussion about convex MDPs?", "answer": ["Reward is Enough for Convex MDPs"], "answer_arxiv_id": ["2106.00661"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_7184"} +{"question": "Can you give references for various tasks that use Implicit Neural Representations (INRs) for efficient querying of continuous locations?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "Learning Shape Templates with Structured Implicit Functions", "Local Deep Implicit Functions for 3D Shape", "Learning Implicit Fields for Generative Shape Modeling", "LION: Latent Point Diffusion Models for 3D Shape Generation", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "Local Implicit Grid Representations for 3D Scenes", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Implicit Neural Representations with Periodic Activation Functions", "Generalised Implicit Neural Representations", "Generative Models as Distributions of Functions", "From data to functa: Your data point is a function and you can treat it like one"], "answer_arxiv_id": ["1812.03828", "1904.06447", "1912.06126", "1812.02822", "2210.06978", "1906.01618", "2003.08981", "2003.08934", "2006.09661", "2205.15674", "2102.04776", "2201.12204"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_7185"} +{"question": "What studies have developed the theory or application of soft prompt tuning methods?", "answer": ["Visual Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2203.12119", "2101.00190", "2109.01134"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_7186"} +{"question": "Which methods learn to generate raster fonts from a large set of reference glyphs or a few exemplar images?", "answer": ["Few-shot Compositional Font Generation with Dual Memory", "Multiple Heads are Better than One: Few-shot Font Generation with\n Multiple Localized Experts", "Few-Shot Font Generation by Learning Fine-Grained Local Styles", "Look Closer to Supervise Better: One-Shot Font Generation via\n Component-Based Discriminator", "Multi-Content GAN for Few-Shot Font Style Transfer", "Artistic Glyph Image Synthesis via One-Stage Few-Shot Learning"], "answer_arxiv_id": ["2005.10510", "2104.00887", "2205.09965", "2205.00146", "1712.00516", "1910.04987"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_7187"} +{"question": "What papers present methods that enforce consistency by distilling what diffusion models have learned to 3D representations?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_7188"} +{"question": "Which works propose solutions for nonconvex-concave saddle point problems through projected gradient based algorithms?", "answer": ["Near-Optimal Algorithms for Minimax Optimization", "Alternating proximal-gradient steps for (stochastic) nonconvex-concave minimax problems", "Efficient Search of First-Order Nash Equilibria in Nonconvex-Concave Smooth Min-Max Problems", "SAPD+ : An Accelerated Stochastic Method for Nonconvex-Concave Minimax Problems", "Accelerated Primal-dual Scheme for a Class of Stochastic Nonconvex-concave Saddle Point Problems"], "answer_arxiv_id": ["2002.02417", "2007.13605", "2002.07919", "2205.15084", "2303.00211"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_7189"} +{"question": "What is the very recent work that is closely related to our study and also adds a matrix inversion-based normalization step after each propagation step?", "answer": ["From Local to Global: Spectral-Inspired Graph Neural Networks"], "answer_arxiv_id": ["2209.12054"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_7190"} +{"question": "Which works discuss the variation in the linear attenuation coefficients (LACs) of human body tissues and metals with the change in X-ray’s energy level?", "answer": ["Convolutional Neural Network Based Metal Artifact Reduction in X-ray Computed Tomography", "DuDoNet: Dual Domain Network for CT Metal Artifact Reduction", "ADN: Artifact Disentanglement Network for Unsupervised Metal Artifact Reduction"], "answer_arxiv_id": ["1709.01581v2", "1907.00273", "1908.01104"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_7191"} +{"question": "What studies considered more complicated models for Neural Collapse but did not incorporate the role of the input distribution?", "answer": ["Revealing the Structure of Deep Neural Networks via Convex Duality", "Extended Unconstrained Features Model for Exploring Deep Neural Collapse", "On the emergence of simplex symmetry in the final and penultimate layers of neural network classifiers"], "answer_arxiv_id": ["2002.09773", "2202.08087", "2012.05420"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_7192"} +{"question": "What work design an alternative neural architecture known as the neural Turing machine?", "answer": ["Neural Turing Machines"], "answer_arxiv_id": ["1410.5401"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7193"} +{"question": "Who proposed DP-SGD with per-example gradient clipping?", "answer": ["Deep Learning with Differential Privacy"], "answer_arxiv_id": ["1607.00133"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7194"} +{"question": "Which papers suggest the ways to detect overfitting in generative models like extracting an exact likelihood or a lower bound to it?", "answer": ["On Memorization in Probabilistic Deep Generative Models", "On the Quantitative Analysis of Decoder-Based Generative Models"], "answer_arxiv_id": ["2106.03216", "1611.04273"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_7195"} +{"question": "What research handles tasks like stylization and customization by training models jointly on images and videos in video editing?", "answer": ["Structure and Content-Guided Video Synthesis with Diffusion Models"], "answer_arxiv_id": ["2302.03011"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_7196"} +{"question": "Could you cite some works about combining neural functions with explicit geometry through neural point clouds?", "answer": ["Point-NeRF: Point-based Neural Radiance Fields", "Neural Point Light Fields", "SPIDR: SDF-based Neural Point Fields for Illumination and Deformation", "PointAvatar: Deformable Point-based Head Avatars from Videos"], "answer_arxiv_id": ["2201.08845", "2112.01473", "2210.08398", "2212.08377"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7197"} +{"question": "What studies have been conducted on text-driven video generation using GAN and VAE models?", "answer": ["To Create What You Tell: Generating Videos from Captions", "Video Generation From Text", "Sync-DRAW: Automatic Video Generation using Deep Recurrent Attentive\n Architectures", "Attentive Semantic Video Generation using Captions"], "answer_arxiv_id": ["1804.08264", "1710.00421", "1611.10314", "1708.05980"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_7198"} +{"question": "Could you provide me some works on recursive prompting techniques?", "answer": ["Recitation-Augmented Language Models", "Measuring and Narrowing the Compositionality Gap in Language Models", "Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models"], "answer_arxiv_id": ["2210.01296", "2210.03350", "2205.09712", "2205.10625"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_7199"} +{"question": "In what studies did the authors express concerns regarding academic independence and integrity related to the proliferation of industry presence in AI research?", "answer": ["The Grey Hoodie Project: Big Tobacco, Big Tech, and the threat on\n academic integrity"], "answer_arxiv_id": ["2009.13676"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_7200"} +{"question": "Which studies explored the impact of duplicate data on the model performance?", "answer": ["The Adverse Effects of Code Duplication in Machine Learning Models of\n Code", "Deduplicating Training Data Makes Language Models Better", "Pythia: A Suite for Analyzing Large Language Models Across Training and\n Scaling", "To Repeat or Not To Repeat: Insights from Scaling LLM under Token-Crisis"], "answer_arxiv_id": ["1812.06469", "2107.06499", "2304.01373", "2305.13230"], "source_meta": {"published_time": "20240709"}, "qid": "AutoScholarQuery_train_7201"} +{"question": "What literature derived a single gradient step to obtain the perturbation in the white-box setting of adversarial examples?", "answer": ["Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1412.6572"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_7202"} +{"question": "Which papers focus on the use of bilevel optimization in modern machine learning problems like meta learning and hyperparameter optimization?", "answer": ["Prototypical Networks for Few-shot Learning", "Meta-learning with differentiable closed-form solvers", "Meta-Learning with Implicit Gradients", "Hyperparameter optimization with approximate gradient", "Bilevel Programming for Hyperparameter Optimization and Meta-Learning"], "answer_arxiv_id": ["1703.05175", "1805.08136", "1909.04630", "1602.02355", "1806.04910"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_7203"} +{"question": "Could you name the works that adopted DPMs to solve problems in the chemistry and biology domain?", "answer": ["GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation", "Equivariant Diffusion for Molecule Generation in 3D", "Diffusion-based Molecule Generationwith Informative Prior Bridges", "Molecular Geometry Pretraining with SE(3)-Invariant Denoising Distance Matching", "Protein structure generation via folding diffusion", "DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking", "Protein Structure and Sequence Generation with Equivariant Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2203.02923", "2203.17003", "2209.00865", "2206.13602", "2209.15611", "2210.01776", "2205.15019"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_7204"} +{"question": "Which works discuss the measure of hyperbolicity for each graph and propose studying graphs with small hyperbolicity using hyperbolic geometry?", "answer": ["Hyperbolic Graph Convolutional Neural Networks", "Hyperbolic Attention Networks", "Lorentzian Graph Convolutional Networks"], "answer_arxiv_id": ["1910.12933", "1805.09786", "2104.07477"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_7205"} +{"question": "Any studies converted 3D to 4D dynamics using a learnable deformation embedding between each frame and its canonical shape?", "answer": ["CaDeX: Learning Canonical Deformation Coordinate Space for Dynamic\n Surface Representation via Neural Homeomorphism"], "answer_arxiv_id": ["2203.16529"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_7206"} +{"question": "What work measure the degree of memorization required of individual examples?", "answer": ["A Closer Look at Memorization in Deep Networks", "Gradient Descent with Early Stopping is Provably Robust to Label Noise for Overparameterized Neural Networks", "Does Learning Require Memorization? A Short Tale about a Long Tail", "What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation"], "answer_arxiv_id": ["1706.05394", "1903.11680", "1906.05271", "2008.03703"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_7207"} +{"question": "Could you mention some research papers that proposed to train VLMs with sequential multi-modal data?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2106.13884", "2204.14198", "2305.03726"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_7208"} +{"question": "Could you provide me some studies about utilizing accumulated attention scores in transformer models?", "answer": ["SpAtten: Efficient Sparse Attention Architecture with Cascade Token and Head Pruning"], "answer_arxiv_id": ["2012.09852"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_7209"} +{"question": "What studies have demonstrated competitive performance with supervised baselines through contrastive learning techniques in visual recognition?", "answer": ["Learning deep representations by mutual information estimation and maximization", "Representation Learning with Contrastive Predictive Coding", "Unsupervised Feature Learning via Non-Parametric Instance Discrimination", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Big Self-Supervised Models are Strong Semi-Supervised Learners", "An Empirical Study of Training Self-Supervised Vision Transformers"], "answer_arxiv_id": ["1808.06670", "1807.03748", "1805.01978", "2002.05709", "1911.05722", "2006.07733", "2006.10029", "2104.02057"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_7210"} +{"question": "What works propose benchmark datasets following the direction of human exams like MMLU, AGIEval, C-Eval, GAOKAO, and IgakuQA for evaluating LLMs?", "answer": ["Measuring Massive Multitask Language Understanding", "AGIEval: A Human-Centric Benchmark for Evaluating Foundation Models", "C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for Foundation Models", "Evaluating the Performance of Large Language Models on GAOKAO Benchmark", "Evaluating GPT-4 and ChatGPT on Japanese Medical Licensing Examinations"], "answer_arxiv_id": ["2009.03300", "2304.06364", "2305.08322", "2305.12474", "2303.18027"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_7211"} +{"question": "Can you provide some works on the subject of gender bias in machine translation?", "answer": ["On Measuring Gender Bias in Translation of Gender-neutral Pronouns", "Evaluating Gender Bias in Machine Translation", "Gender Bias in Machine Translation"], "answer_arxiv_id": ["1905.11684", "1906.00591", "2104.06001"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_7212"} +{"question": "What papers discuss human activity analysis in the context of egocentric videos?", "answer": ["The Evolution of First Person Vision Methods: A Survey"], "answer_arxiv_id": ["1409.1484"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_7213"} +{"question": "What paper proposed a fixed method for SiMT through a wait-k policy?", "answer": ["STACL: Simultaneous Translation with Implicit Anticipation and Controllable Latency using Prefix-to-Prefix Framework"], "answer_arxiv_id": ["1810.08398"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_7214"} +{"question": "Which studies are associated with the emergence of neural implicit reconstruction?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Learning Implicit Fields for Generative Shape Modeling", "Occupancy Networks: Learning 3D Reconstruction in Function Space"], "answer_arxiv_id": ["1901.05103", "1812.02822", "1812.03828"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_7215"} +{"question": "Which works have established user-generated content (UGC) databases for video quality assessment?", "answer": ["UGC-VQA: Benchmarking Blind Video Quality Assessment for User Generated\n Content", "Large-Scale Study of Perceptual Video Quality", "Patch-VQ: 'Patching Up' the Video Quality Problem", "MD-VQA: Multi-Dimensional Quality Assessment for UGC Live Videos", "Towards Explainable In-the-Wild Video Quality Assessment: A Database and\n a Language-Prompted Approach"], "answer_arxiv_id": ["2005.14354", "1803.01761", "2011.13544", "2303.14933", "2305.12726"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_7216"} +{"question": "What studies made use of relative camera viewpoints as conditions for diffusion modeling?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2303.11328v1"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_7217"} +{"question": "What papers propose self-supervised semantic correspondences to directly optimize the 3D shape from image or video collections?", "answer": ["LASSIE: Learning Articulated Shapes from Sparse Image Ensemble via 3D\n Part Discovery", "Hi-LASSIE: High-Fidelity Articulated Shape and Skeleton Discovery from\n Sparse Image Ensemble", "ARTIC3D: Learning Robust Articulated 3D Shapes from Noisy Web Image\n Collections"], "answer_arxiv_id": ["2207.03434", "2212.11042", "2306.04619"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_7218"} +{"question": "Could you give some examples of papers that explore self-supervised learning to decrease the demand for annotated data?", "answer": ["Geography-Aware Self-Supervised Learning", "Seasonal Contrast: Unsupervised Pre-Training from Uncurated Remote\n Sensing Data"], "answer_arxiv_id": ["2011.09980", "2103.16607"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_7219"} +{"question": "What studies have introduced cost metrics for evaluating the efficiency of deep learning models?", "answer": ["How Do Adam and Training Strategies Help BNNs Optimization?", "ReActNet: Towards Precise Binary Neural Network with Generalized\n Activation Functions", "Bi-Real Net: Enhancing the Performance of 1-bit CNNs With Improved\n Representational Capability and Advanced Training Algorithm", "PokeBNN: A Binary Pursuit of Lightweight Accuracy"], "answer_arxiv_id": ["2106.11309", "2003.03488", "1808.00278", "2112.00133"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_7220"} +{"question": "Which papers are about constructing the 3D feature/cost volume and utilizing the voxel feature for decoding density and color?", "answer": ["MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo", "GeoNeRF: Generalizing NeRF with Geometry Priors", "NeRFusion: Fusing Radiance Fields for Large-Scale Scene Reconstruction", "SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse Views"], "answer_arxiv_id": ["2103.15595", "2111.13539", "2203.11283", "2206.05737"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_7221"} +{"question": "What are the existing GAN-based approaches for the virtual try-on problem?", "answer": ["VITON-HD: High-Resolution Virtual Try-On via Misalignment-Aware\n Normalization", "High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled\n Conditions", "GP-VTON: Towards General Purpose Virtual Try-on via Collaborative\n Local-Flow Global-Parsing Learning", "Parser-Free Virtual Try-on via Distilling Appearance Flows"], "answer_arxiv_id": ["2103.16874", "2206.14180", "2303.13756", "2103.04559"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_7222"} +{"question": "Which papers discussed image retrieval methods for same-view localization tasks?", "answer": ["Fine-tuning CNN Image Retrieval with No Human Annotation", "SuperGlue: Learning Feature Matching with Graph Neural Networks", "NetVLAD: CNN architecture for weakly supervised place recognition"], "answer_arxiv_id": ["1711.02512", "1911.11763", "1511.07247"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_7223"} +{"question": "Can you provide references that have made significant contributions to the theoretical basis of Submodular Function Minimization (SFM)?", "answer": ["A Faster Cutting Plane Method and its Implications for Combinatorial and Convex Optimization", "Subquadratic Submodular Function Minimization", "Near-optimal Approximate Discrete and Continuous Submodular Function Minimization", "Geometric Rescaling Algorithms for Submodular Function Minimization", "Minimizing Convex Functions with Rational Minimizers", "New Query Lower Bounds for Submodular Function Minimization", "Improved Lower Bounds for Submodular Function Minimization"], "answer_arxiv_id": ["1508.04874", "1610.09800", "1909.00171", "1707.05065v4", "2007.01445v5", "1911.06889", "2207.04342"], "source_meta": {"published_time": "20230908"}, "qid": "AutoScholarQuery_train_7224"} +{"question": "Any works about comparing the expressivity of spectral MPNNs on spaces of graphons?", "answer": ["On the Universality of Graph Neural Networks on Large Random Graphs"], "answer_arxiv_id": ["2105.13099"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7225"} +{"question": "Which research papers attempted to encode inductive biases for generalization in image-based RL tasks?", "answer": ["Contrastive Behavioral Similarity Embeddings for Generalization in Reinforcement Learning", "CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Pretraining Representations for Data-Efficient Reinforcement Learning", "Deep Reinforcement and InfoMax Learning"], "answer_arxiv_id": ["2101.05265", "2004.04136", "2106.04799", "2006.07217"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_7226"} +{"question": "Any studies about using NeRF-specific proxies for uncertainty quantification?", "answer": ["ActiveRMAP: Radiance Field for Active Mapping And Planning", "ProbNeRF: Uncertainty-Aware Inference of 3D Shapes from 2D Images"], "answer_arxiv_id": ["2211.12656", "2210.17415"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_7227"} +{"question": "What paper introduced the concept of Atrous spatial pyramid pooling (ASPP) in Semantic Segmentation?", "answer": ["Rethinking Atrous Convolution for Semantic Image Segmentation"], "answer_arxiv_id": ["1706.05587"], "source_meta": {"published_time": "20240416"}, "qid": "AutoScholarQuery_train_7228"} +{"question": "Could you provide me some studies that focus on the discovery of a diverse range of policies in RL?", "answer": ["Diversify and Disambiguate: Learning From Underspecified Data"], "answer_arxiv_id": ["2202.03418"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_7229"} +{"question": "What papers introduced KL characteristic time in the context of efficient algorithm design?", "answer": ["Optimal Best Arm Identification with Fixed Confidence"], "answer_arxiv_id": ["1602.04589v2"], "source_meta": {"published_time": "20230905"}, "qid": "AutoScholarQuery_train_7230"} +{"question": "Which papers demonstrate that alternating gradient descent ascent (AGDA) is often more stable than its simultaneous counterpart?", "answer": ["Negative Momentum for Improved Game Dynamics", "Finite Regret and Cycles with Fixed Step-Size via Alternating Gradient Descent-Ascent"], "answer_arxiv_id": ["1807.04740v5", "1907.04392"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_7231"} +{"question": "What papers reported on refined reasoning errors in language models, particularly self-improvement and self-refinement techniques?", "answer": ["Large Language Models Can Self-Improve", "Self-Refine: Iterative Refinement with Self-Feedback"], "answer_arxiv_id": ["2210.11610", "2303.17651"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_7232"} +{"question": "What papers utilized counterfactual generation methods for 'distributional counterfactuals'?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Learning Model-Agnostic Counterfactual Explanations for Tabular Data"], "answer_arxiv_id": ["1812.04948", "1910.09398"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_7233"} +{"question": "What are the studies that first explored strategic classification in a distributional and online model?", "answer": ["Strategic Classification", "Strategic Classification from Revealed Preferences"], "answer_arxiv_id": ["1506.06980", "1710.07887"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7234"} +{"question": "What papers discussed the utilization of these systems in a transfer learning configuration?", "answer": ["GPL: Generative Pseudo Labeling for Unsupervised Domain Adaptation of\n Dense Retrieval"], "answer_arxiv_id": ["2112.07577"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_7235"} +{"question": "Which papers are relevant to the progress made in 2D image segmentation?", "answer": ["Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers", "SegNeXt: Rethinking Convolutional Attention Design for Semantic\n Segmentation", "Encoder-Decoder with Atrous Separable Convolution for Semantic Image\n Segmentation", "Pyramid Scene Parsing Network", "Conditional Convolutions for Instance Segmentation", "SOLO: Segmenting Objects by Locations", "Mask R-CNN", "YOLACT++: Better Real-time Instance Segmentation", "YOLACT: Real-time Instance Segmentation", "SOLOv2: Dynamic and Fast Instance Segmentation", "Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with\n Transformers", "OneFormer: One Transformer to Rule Universal Image Segmentation", "Attention-guided Unified Network for Panoptic Segmentation", "Panoptic Feature Pyramid Networks", "UPSNet: A Unified Panoptic Segmentation Network", "Panoptic Segmentation with a Joint Semantic and Instance Segmentation\n Network"], "answer_arxiv_id": ["2012.15840", "2209.08575", "1802.02611", "1612.01105", "2003.05664", "1912.04488", "1703.06870", "1912.06218", "1904.02689", "2003.10152", "2109.03814", "2211.06220", "1812.03904", "1901.02446", "1901.03784", "1809.02110"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_7236"} +{"question": "Can you name research works that investigate quantized weight exchange in the context of decentralized distributed learning?", "answer": ["Decentralized Stochastic Optimization and Gossip Algorithms with Compressed Communication", "Asynchronous Decentralized SGD with Quantized and Local Updates", "Moniqua: Modulo Quantized Communication in Decentralized SGD"], "answer_arxiv_id": ["1902.00340", "1910.12308v4", "2002.11787"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_7237"} +{"question": "Could you provide some GAN-based and Transformer-based methods proposed for image fusion?", "answer": ["TransMEF: A Transformer-Based Multi-Exposure Image Fusion Framework\n using Self-Supervised Multi-Task Learning"], "answer_arxiv_id": ["2112.01030"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_7238"} +{"question": "Which works apply policy optimization methods in games and other applications?", "answer": ["Deep Learning for Reward Design to Improve Monte Carlo Tree Search in ATARI Games", "ELF OpenGo: An Analysis and Open Reimplementation of AlphaZero"], "answer_arxiv_id": ["1604.07095", "1902.04522"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_7239"} +{"question": "Could you provide me the studies that propose a forward blurring process as an alternative to the additive Gaussian noising process?", "answer": ["Generative Modelling With Inverse Heat Dissipation"], "answer_arxiv_id": ["2206.13397"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_7240"} +{"question": "What papers have presented a dual form for Distributional Robustness?", "answer": ["Quantifying Distributional Model Risk via Optimal Transport", "Certifying Some Distributional Robustness with Principled Adversarial Training"], "answer_arxiv_id": ["1604.01446", "1710.10571v5"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_7241"} +{"question": "Could you provide me with studies that utilized pre-trained visual encoder with LLM?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2301.12597"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_7242"} +{"question": "Could you provide me some studies that proposed other methods for score matching?", "answer": ["Sliced Score Matching: A Scalable Approach to Density and Score Estimation", "Denoising Likelihood Score Matching for Conditional Score-based Data Generation"], "answer_arxiv_id": ["1905.07088", "2203.14206"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_7243"} +{"question": "Are there any recent works that focus on developing empirically efficient algorithms to explore the Pareto Front ?", "answer": ["Efficient Continuous Pareto Exploration in Multi-Task Learning", "Controllable Pareto Multi-Task Learning", "Scalable Pareto Front Approximation for Deep Multi-Objective Learning", "Pareto Navigation Gradient Descent: a First-Order Algorithm for Optimization in Pareto Set"], "answer_arxiv_id": ["2006.16434", "2010.06313", "2103.13392", "2110.08713"], "source_meta": {"published_time": "20230827"}, "qid": "AutoScholarQuery_train_7244"} +{"question": "What studies address reasoning about human social interactions from a first-person perspective in egocentric vision?", "answer": ["In the Eye of Transformer: Global-Local Correlation for Egocentric Gaze\n Estimation", "Future Person Localization in First-Person Videos", "4D Human Body Capture from Egocentric Video via 3D Scene Grounding"], "answer_arxiv_id": ["2208.04464", "1711.11217", "2011.13341"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_7245"} +{"question": "What research papers present methods to optimize the geometry and texture of NeRF based on text descriptions or exemplar images?", "answer": ["CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields", "SINE: Semantic-driven Image-based NeRF Editing with Prior-guided Editing\n Field"], "answer_arxiv_id": ["2112.05139", "2303.13277"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_7246"} +{"question": "What works made significant contributions to the application of asynchronous distributed methods in real-world?", "answer": ["Building High-level Features Using Large Scale Unsupervised Learning"], "answer_arxiv_id": ["1112.6209"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_7247"} +{"question": "Which paper introduced the methodology NRC, which has the intuition that a test data and its nearest neighbors share the same label under domain shift?", "answer": ["Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation"], "answer_arxiv_id": ["2110.04202"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_7248"} +{"question": "Which works have focused on developing efficient algorithms for data valuation using Monte Carlo methods or surrogate utility functions etc.?", "answer": ["Towards Efficient Data Valuation Based on the Shapley Value", "Measuring the Effect of Training Data on Deep Learning Predictions via Randomized Experiments", "Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms", "LAVA: Data Valuation without Pre-Specified Learning Algorithms", "Incentivizing Collaboration in Machine Learning via Synthetic Data Rewards", "Data Valuation using Reinforcement Learning"], "answer_arxiv_id": ["1902.10275", "2206.10013", "1908.08619", "2305.00054", "2112.09327", "1909.11671"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_7249"} +{"question": "Could you provide me examples of works which used redundancy definitions in multi-view learning?", "answer": ["Contrastive learning, multi-view redundancy, and linear models", "What Makes for Good Views for Contrastive Learning?"], "answer_arxiv_id": ["2008.10150", "2005.10243"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_7250"} +{"question": "Which works have incorporated data augmentation techniques to training GANs?", "answer": ["Consistency Regularization for Generative Adversarial Networks", "Training Generative Adversarial Networks with Limited Data", "Differentiable Augmentation for Data-Efficient GAN Training"], "answer_arxiv_id": ["1910.12027", "2006.06676", "2006.10738"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_7251"} +{"question": "Which papers have explored the use of autoregressive models for zero-shot text to image generation?", "answer": ["Zero-Shot Text-to-Image Generation", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092", "2206.10789"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_7252"} +{"question": "Could you provide me some studies about designing agents that interact with humans?", "answer": ["On the Utility of Learning about Humans for Human-AI Coordination"], "answer_arxiv_id": ["1910.05789"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_7253"} +{"question": "Which papers describe handling DG in reinforcement learning via environment generation?", "answer": ["Prioritized Level Replay"], "answer_arxiv_id": ["2010.03934"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7254"} +{"question": "Could you provide me some works about using language in reinforcement learning tasks, especially in text-based games?", "answer": ["Language Understanding for Text-based Games using Deep Reinforcement Learning", "Counting to Explore and Generalize in Text-based Games", "The NetHack Learning Environment"], "answer_arxiv_id": ["1506.08941", "1806.11525", "2006.13760"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_7255"} +{"question": "Which works demonstrated the vulnerability of deep neural networks?", "answer": ["Adversarial examples in the physical world", "Intriguing properties of neural networks"], "answer_arxiv_id": ["1607.02533", "1312.6199"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_7256"} +{"question": "Which papers introduced the mixture-of-LoRA (MoLoRA) architecture to improve the performance of LoRA?", "answer": ["Pushing Mixture of Experts to the Limit: Extremely Parameter Efficient\n MoE for Instruction Tuning", "LoRAMoE: Alleviate World Knowledge Forgetting in Large Language Models\n via MoE-Style Plugin"], "answer_arxiv_id": ["2309.05444", "2312.09979"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_7257"} +{"question": "Can you cite studies about generative adversarial learning in the context of removing motion-induced blurring?", "answer": ["DeblurGAN: Blind Motion Deblurring Using Conditional Adversarial\n Networks", "DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better", "Deblurring by Realistic Blurring"], "answer_arxiv_id": ["1711.07064", "1908.03826", "2004.01860"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_7258"} +{"question": "Which work applied a gradient-based proxy to RoBERTa?", "answer": ["KNAS: Green Neural Architecture Search"], "answer_arxiv_id": ["2111.13293"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_7259"} +{"question": "Could you provide me some works studying network motifs in the context of improving network representation learning?", "answer": ["MotifNet: a motif-based Graph Convolutional Network for directed graphs", "MODEL: Motif-based Deep Feature Learning for Link Prediction"], "answer_arxiv_id": ["1802.01572", "2008.03637"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7260"} +{"question": "What research works are related to Annealed Stein Variational Gradient Descent?", "answer": ["Annealed Stein Variational Gradient Descent"], "answer_arxiv_id": ["2101.09815"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_7261"} +{"question": "What works are about the approach that leverages an unconditional pre-trained diffusion model in diffusion-based SR?", "answer": ["Come-Closer-Diffuse-Faster: Accelerating Conditional Diffusion Models\n for Inverse Problems through Stochastic Contraction", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "DifFace: Blind Face Restoration with Diffused Error Contraction"], "answer_arxiv_id": ["2112.05146", "2108.02938", "2212.06512"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_7262"} +{"question": "Which works have exploited GAN, VAE, and RNN to learn graph distributions?", "answer": ["Generative Adversarial Nets", "Auto-Encoding Variational Bayes", "Recurrent Neural Network Regularization", "A Systematic Survey on Deep Generative Models for Graph Generation"], "answer_arxiv_id": ["1406.2661", "1312.6114", "1409.2329", "2007.06686"], "source_meta": {"published_time": "20220710"}, "qid": "AutoScholarQuery_train_7263"} +{"question": "What research creates soft labels for images using all the human annotations from the data collection step?", "answer": ["Human uncertainty makes classification more robust"], "answer_arxiv_id": ["1908.07086"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_7264"} +{"question": "Which works introduced methods for inferring continuous time models of dynamical systems by using the weak form of the differential equations?", "answer": ["A Unified Approach for Sparse Dynamical System Inference from Temporal Measurements"], "answer_arxiv_id": ["1710.00718"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_7265"} +{"question": "Which research constructed a pseudo-multi-view perspective for weakly supervised 3D object detection?", "answer": ["Weakly Supervised Monocular 3D Object Detection using Multi-View\n Projection and Direction Consistency"], "answer_arxiv_id": ["2303.08686"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_7266"} +{"question": "Could you provide me with the work that introduced a unique data fusion strategy using CutMix for object detection efficacy?", "answer": ["Few-shot Adaptive Object Detection with Cross-Domain CutMix"], "answer_arxiv_id": ["2208.14586"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_7267"} +{"question": "Which studies provide details about generating general purpose datasets for multilingual machine translation and language modeling?", "answer": ["mT5: A Massively Multilingual Pre-trained Text-to-Text Transformer", "Unsupervised Cross-lingual Representation Learning at Scale", "Few-shot Learning with Multilingual Language Models", "Towards a Cleaner Document-Oriented Multilingual Crawled Corpus", "Glot500: Scaling Multilingual Corpora and Language Models to 500 Languages", "Building Machine Translation Systems for the Next Thousand Languages", "No Language Left Behind: Scaling Human-Centered Machine Translation"], "answer_arxiv_id": ["2010.11934", "1911.02116", "2112.10668", "2201.06642", "2305.12182", "2205.03983", "2207.04672"], "source_meta": {"published_time": "20230909"}, "qid": "AutoScholarQuery_train_7268"} +{"question": "Which publications discuss about neural representation of 3D assets (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "TensoRF: Tensorial Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2003.08934", "2203.09517", "2201.05989"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_7269"} +{"question": "Which works have shown that the use of an Equiangular Tight Frame(ETF) in training neural networks can improve performance?", "answer": ["A Geometric Analysis of Neural Collapse with Unconstrained Features", "Imbalance Trouble: Revisiting Neural-Collapse Geometry", "Balanced Contrastive Learning for Long-Tailed Visual Recognition", "Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class Incremental Learning"], "answer_arxiv_id": ["2105.02375", "2208.05512", "2207.09052", "2302.03004"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7270"} +{"question": "Which papers focus on learning skills in a multi-task setting?", "answer": ["Learning and Transfer of Modulated Locomotor Controllers", "Learning by Playing – Solving Sparse Reward Tasks from Scratch"], "answer_arxiv_id": ["1610.05182", "1802.10567"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_7271"} +{"question": "Are there any works that utilize keypoint information to expand the number of depth prediction branches?", "answer": ["Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular\n 3D Object Detection"], "answer_arxiv_id": ["2205.09373"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_7272"} +{"question": "What papers are about developing scene synthesis methods from human motions?", "answer": ["MIME: Human-Aware 3D Scene Generation", "Scene Synthesis from Human Motion", "Pose2Room: Understanding 3D Scenes from Human Activities"], "answer_arxiv_id": ["2212.04360", "2301.01424", "2112.03030"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_7273"} +{"question": "What study introduced FeatureNMS that encodes features for predictions?", "answer": ["FeatureNMS: Non-Maximum Suppression by Learning Feature Embeddings"], "answer_arxiv_id": ["2002.07662"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_7274"} +{"question": "Which research works show examples of LLMs being elaborately prompted to generate plans when provided with task instructions?", "answer": ["Language Models as Zero-Shot Planners: Extracting Actionable Knowledge\n for Embodied Agents", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Reflexion: Language Agents with Verbal Reinforcement Learning", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2201.07207", "2204.01691", "2303.11366", "2210.03629"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_7275"} +{"question": "What are the studies that use 2D diffusion models for text-to-3D?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7276"} +{"question": "Which works discuss representation learning in the context of supervised learning?", "answer": ["A Model of Inductive Bias Learning", "The Benefit of Multitask Representation Learning", "Few-Shot Learning via Learning the Representation, Provably", "Provable Meta-Learning of Linear Representations"], "answer_arxiv_id": ["1106.0245v1", "1505.06279", "2002.09434", "2002.11684"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_7277"} +{"question": "What research used Semantic ID representation of items for recommendation systems?", "answer": ["Neural Machine Translation of Rare Words with Subword Units", "Recommendation as Language Processing (RLP): A Unified Pretrain, Personalized Prompt & Predict Paradigm (P5)"], "answer_arxiv_id": ["1508.07909v5", "2203.13366"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_7278"} +{"question": "Could you provide me some works involving transformer-based self-supervised methods for unsupervised correspondence discovery?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "DINOv2: Learning Robust Visual Features without Supervision", "iBOT: Image BERT Pre-Training with Online Tokenizer"], "answer_arxiv_id": ["2104.14294", "2304.07193", "2111.07832"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_7279"} +{"question": "Which work proposes utilizing single-sample data augmentation while maintaining the privacy guarantee?", "answer": ["Unlocking High-Accuracy Differentially Private Image Classification through Scale"], "answer_arxiv_id": ["2204.13650"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7280"} +{"question": "What papers are about the use of L2-normalized embedding for stable learning in Contrastive Learning?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Learning Transferable Visual Models From Natural Language Supervision", "Spherical Latent Spaces for Stable Variational Autoencoders"], "answer_arxiv_id": ["1807.03748", "1911.05722", "2002.05709", "2103.00020", "1808.10805"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_7281"} +{"question": "Could you provide me some studies about learning mathematical tasks by neural network?", "answer": ["Integration of Neural Network-Based Symbolic Regression in Deep Learning for Scientific Discovery", "Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets", "Hidden Progress in Deep Learning: SGD Learns Parities Near the Computational Limit", "The Quantization Model of Neural Scaling"], "answer_arxiv_id": ["1912.04825", "2201.02177", "2207.08799", "2303.13506v3"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_7282"} +{"question": "What are some works that have adopted the causal representation learning perspective?", "answer": ["Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICA", "Partial Disentanglement via Mechanism Sparsity", "Learning Temporally Causal Latent Processes from General Temporal Data", "Weakly supervised causal representation learning"], "answer_arxiv_id": ["2107.10098", "2207.07732", "2110.05428", "2203.16437"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_7283"} +{"question": "Which works have studied widely used estimators under undiscounted-state-sampling setting and their impacts on policy optimization?", "answer": ["Is the Policy Gradient a Gradient?"], "answer_arxiv_id": ["1906.07073"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_7284"} +{"question": "Which study has accomplished IAD for multiple classes by a unified framework for unified IAD?", "answer": ["A Unified Model for Multi-class Anomaly Detection"], "answer_arxiv_id": ["2206.03687"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_7285"} +{"question": "Which studies involve employing powerful pretrained T2I models to deal with the real-ISR problem?", "answer": ["Exploiting Diffusion Prior for Real-World Image Super-Resolution", "Pixel-Aware Stable Diffusion for Realistic Image Super-resolution and\n Personalized Stylization", "DiffBIR: Towards Blind Image Restoration with Generative Diffusion Prior"], "answer_arxiv_id": ["2305.07015", "2308.14469", "2308.15070"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7286"} +{"question": "Which works explored semi-supervised learning in the context of visual recognition?", "answer": ["Generalized Product Quantization Network for Semi-supervised Image\n Retrieval", "Self-Training Boosted Multi-Faceted Matching Network for Composed Image\n Retrieval", "Decoupled Deep Neural Network for Semi-supervised Semantic Segmentation", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "MixMatch: A Holistic Approach to Semi-Supervised Learning", "Temporal Ensembling for Semi-Supervised Learning", "Virtual Adversarial Training: A Regularization Method for Supervised and\n Semi-Supervised Learning", "Mean teachers are better role models: Weight-averaged consistency\n targets improve semi-supervised deep learning results"], "answer_arxiv_id": ["2002.11281", "2305.09979", "1506.04924", "2001.07685v2", "1905.02249", "1610.02242v3", "1704.03976", "1703.01780"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_7287"} +{"question": "Which studies used MI as a metric in summarization?", "answer": ["Unsupervised Extractive Summarization using Pointwise Mutual Information", "Mutual Information Alleviates Hallucinations in Abstractive\n Summarization"], "answer_arxiv_id": ["2102.06272", "2210.13210"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_7288"} +{"question": "Could you provide me some works about representing 3D data in implicit neural functions?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation", "Neural Unsigned Distance Fields for Implicit Function Learning", "RangeUDF: Semantic Surface Reconstruction from 3D Point Clouds", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​", "GRF: Learning a General Radiance Field for 3D Representation and Rendering"], "answer_arxiv_id": ["1812.02822", "1812.03828", "1901.05103", "2010.13938", "2204.09138", "2003.08934", "2103.13415", "2010.04595"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_7289"} +{"question": "Which research works discuss that there is an assumed underlying ultrametric space whose geometry specifies the unknown tree in theory of hierarchical clustering?", "answer": ["A cost function for similarity-based hierarchical clustering", "Hierarchical Clustering via Spreading Metrics", "Approximate Hierarchical Clustering via Sparsest Cut and Spreading Metrics", "Hierarchical Clustering: Objective Functions and Algorithms", "Maximizing Agreements for Ranking, Clustering and Hierarchical Clustering via MAX-CUT"], "answer_arxiv_id": ["1510.05043", "1610.09269", "1609.09548", "1704.02147", "2102.11548"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_7290"} +{"question": "Which studies have explored changing the transformer architecture to reduce its space or time requirements?", "answer": ["Efficient Transformers: A Survey"], "answer_arxiv_id": ["2009.06732"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_7291"} +{"question": "Which paper introduces an image quality adaptive loss function that reduces the influence of low-quality or unidentifiable samples?", "answer": ["AdaFace: Quality Adaptive Margin for Face Recognition"], "answer_arxiv_id": ["2204.00964"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_7292"} +{"question": "Which works designed multiple-choice questions to measure the knowledge and reasoning ability of LLMs?", "answer": ["Measuring Massive Multitask Language Understanding", "Beyond the Imitation Game: Quantifying and extrapolating the\n capabilities of language models"], "answer_arxiv_id": ["2009.03300", "2206.04615"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_7293"} +{"question": "Could you provide me some research about denoising and deblurring?", "answer": ["Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image\n Denoising", "FFDNet: Toward a Fast and Flexible Solution for CNN based Image\n Denoising", "Masked Image Training for Generalizable Deep Image Denoising", "Hierarchical Integration Diffusion Model for Realistic Image Deblurring", "Deep Multi-scale Convolutional Neural Network for Dynamic Scene\n Deblurring", "Scale-recurrent Network for Deep Image Deblurring"], "answer_arxiv_id": ["1608.03981", "1710.04026", "2303.13132", "2305.12966", "1612.02177", "1802.01770"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_7294"} +{"question": "Which works made use of parameterized mixture models for identifying and correcting noisy labels?", "answer": ["Unsupervised Label Noise Modeling and Loss Correction"], "answer_arxiv_id": ["1904.11238"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_7295"} +{"question": "Which references can provide information about Sparse Coding?", "answer": ["A survey of sparse representation: algorithms and applications"], "answer_arxiv_id": ["1602.07017"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7296"} +{"question": "Which authors have developed understanding of neural networks' connectivity patterns on its trainability?", "answer": ["Deep Architecture Connectivity Matters for Its Convergence: A Fine-Grained Analysis"], "answer_arxiv_id": ["2205.05662"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_7297"} +{"question": "What works describe mechanisms to extract 3D models from images through optimization of SDS distillation loss?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_7298"} +{"question": "Are there any existing works related to the design of contrastive loss?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Big Self-Supervised Models are Strong Semi-Supervised Learners"], "answer_arxiv_id": ["1807.03748", "2002.05709", "2006.10029"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_7299"} +{"question": "What studies have applied autoregressive transformer decoders to object-centric learning?", "answer": ["Illiterate DALL-E Learns to Compose", "Simple Unsupervised Object-Centric Learning for Complex and Naturalistic\n Videos"], "answer_arxiv_id": ["2110.11405", "2205.14065"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_7300"} +{"question": "Which works have followed a retrieve-then-generate architecture in supervised knowledge base question answering (KBQA)?", "answer": ["Sequence-to-Sequence Knowledge Graph Completion and Question Answering", "Subgraph Retrieval Enhanced Model for Multi-hop Knowledge Base Question\n Answering", "Knowledge Base Question Answering by Case-based Reasoning over Subgraphs", "RnG-KBQA: Generation Augmented Iterative Ranking for Knowledge Base\n Question Answering", "Case-based Reasoning for Natural Language Queries over Knowledge Bases", "Don't Generate, Discriminate: A Proposal for Grounding Language Models\n to Real-World Environments"], "answer_arxiv_id": ["2203.10321", "2202.13296", "2202.10610", "2109.08678", "2104.08762", "2212.09736"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_7301"} +{"question": "What research proposed a CNN-based VAE architecture with a Gaussian Process prior for multivariate time series data?", "answer": ["GP-VAE: Deep Probabilistic Time Series Imputation"], "answer_arxiv_id": ["1907.04155"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_7302"} +{"question": "Who proposed a new method of achieving conditional text generation where controlled information is also involved in the diffusion process?", "answer": ["DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models"], "answer_arxiv_id": ["2210.08933"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_7303"} +{"question": "What research explored the potential of Diffusion Models in the context of structured reconstruction?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2006.11239", "2011.13456"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_7304"} +{"question": "Which work employs unsupervised anomaly detectors to assign training labels?", "answer": ["Deep Semi-Supervised Anomaly Detection", "Learning Representations of Ultrahigh-dimensional Data for Random Distance-based Outlier Detection"], "answer_arxiv_id": ["1906.02694", "1806.04808"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_7305"} +{"question": "What studies mentioned the performance drop when GAN image detection methods are applied to diffusion-generated images?", "answer": ["On the detection of synthetic images generated by diffusion models", "Towards the Detection of Diffusion Model Deepfakes"], "answer_arxiv_id": ["2211.00680", "2210.14571"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_7306"} +{"question": "Which studies show progress in Text-to-Image models using pretrained models?", "answer": ["Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion Models", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models", "Prompt-Free Diffusion: Taking \"Text\" out of Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2305.16322", "2308.06721", "2305.16223"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_7307"} +{"question": "Which research proposes a specular albedo prior model for 3DMMs?", "answer": ["A Morphable Face Albedo Model"], "answer_arxiv_id": ["2004.02711"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_7308"} +{"question": "Which studies are focused on preventing catastrophic forgetting by storing parameters of models trained on prior tasks in continual reinforcement learning?", "answer": ["Progressive Neural Networks", "Superposition of many models into one", "Supermasks in Superposition"], "answer_arxiv_id": ["1606.04671", "1902.05522", "2006.14769"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_7309"} +{"question": "Any research about the asymptotic fixed point of the velocity field in relation with the classic Fokker-Planck equation(FPE)?", "answer": ["Self-Consistency of the Fokker-Planck Equation"], "answer_arxiv_id": ["2206.00860"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_7310"} +{"question": "Which studies are involved in retraining-based approaches in sample-based explanation approaches?", "answer": ["Datamodels: Predicting Predictions from Training Data", "Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning", "Counterfactual Memorization in Neural Language Models", "Characterizing Structural Regularities of Labeled Data in Overparameterized Models", "Scalability vs. Utility: Do We Have to Sacrifice One for the Other in Data Importance Quantification?", "Data Valuation using Reinforcement Learning", "A Distributional Framework for Data Valuation", "Data Shapley: Equitable Valuation of Data for Machine Learning", "Towards Efficient Data Valuation Based on the Shapley Value"], "answer_arxiv_id": ["2202.00622", "2110.14049", "2112.12938", "2002.03206", "1911.07128", "1909.11671", "2002.12334", "1904.02868v2", "1902.10275"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_7311"} +{"question": "Which works focus on applying Machine Learning to improve BnB by learning variable selection for branching?", "answer": ["Exact Combinatorial Optimization with Graph Convolutional Neural Networks", "Hybrid Models for Learning to Branch", "Parameterizing Branch-and-Bound Search Trees to Learn Branching Policies"], "answer_arxiv_id": ["1906.01629", "2006.15212", "2002.05120"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_7312"} +{"question": "In what research was Frechet Inception Distance used as a metric in the evaluation of a generative model?", "answer": ["GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash\n Equilibrium"], "answer_arxiv_id": ["1706.08500"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_7313"} +{"question": "Which works discussed the drawbacks of adversarial training in terms of large computations and test accuracy trade-off?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1706.06083", "1901.08573"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_7314"} +{"question": "Which papers establish the relationship between contrastive learning and graph neural networks?", "answer": ["A Message Passing Perspective on Learning Dynamics of Contrastive Learning"], "answer_arxiv_id": ["2303.04435"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_7315"} +{"question": "Which papers discuss novel object captioning based on unpaired image-sentence sources or novel object detectors?", "answer": ["Deep Compositional Captioning: Describing Novel Object Categories\n without Paired Training Data", "Captioning Images with Diverse Objects", "Neural Baby Talk", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "VinVL: Revisiting Visual Representations in Vision-Language Models"], "answer_arxiv_id": ["1511.05284", "1606.07770", "1803.09845", "2004.06165", "2101.00529"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7316"} +{"question": "What studies apply channel pruning to ViT pruning that includes structured and unstructured approaches?", "answer": ["Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Vision Transformer Slimming: Multi-Dimension Searching in Continuous\n Optimization Space", "Unified Visual Transformer Compression"], "answer_arxiv_id": ["2106.04533", "2201.00814", "2203.08243"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_7317"} +{"question": "Which papers discussed query-based attacks within the black-box attacks scope?", "answer": ["Practical Black-Box Attacks against Machine Learning", "Black-box Adversarial Attacks with Limited Queries and Information", "One Pixel Attack for Fooling Deep Neural Networks"], "answer_arxiv_id": ["1602.02697", "1804.08598", "1710.08864"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_7318"} +{"question": "Could you provide me the works that propose a practical algorithm that minimizes the Maximum Mean Discrepancy (MMD) without a sample complexity analysis?", "answer": ["Distributional Reinforcement Learning for Multi-Dimensional Reward Functions"], "answer_arxiv_id": ["2110.13578"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_7319"} +{"question": "What are the research papers that investigated the possibility to combine motion prediction with other modules?", "answer": ["MP3: A Unified Model to Map, Perceive, Predict and Plan", "Deep Interactive Motion Prediction and Planning: Playing Games with Motion Prediction Models", "FIERY: Future Instance Prediction in Bird’s-Eye View from Surround Monocular Cameras", "PredictionNet: Real-Time Joint Probabilistic Traffic Prediction for Planning, Control, and Simulation", "Model-Based Imitation Learning for Urban Driving", "Planning-oriented Autonomous Driving", "End-to-End Urban Driving by Imitating a Reinforcement Learning Coach"], "answer_arxiv_id": ["2101.06806", "2204.02392v1", "2104.10490", "2109.11094", "2210.07729", "2212.10156", "2108.08265"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_7320"} +{"question": "Which studies replaced the convolutional network with other model architectures in the field of cellular automata?", "answer": ["Attention-based Neural Cellular Automata", "E(n)-Equivariant Graph Neural Cellular Automata"], "answer_arxiv_id": ["2211.01233", "2301.10497"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_7321"} +{"question": "Which papers use voxels and meshes to assist the 3D-aware generation?", "answer": ["HoloGAN: Unsupervised Learning of 3D Representations From Natural Images", "Escaping Plato’s Cave: 3D Shape From Adversarial Rendering", "Do 2D GANs Know 3D Shape? Unsupervised 3D shape reconstruction from 2D Image GANs"], "answer_arxiv_id": ["1904.01326", "1811.11606", "2011.00844"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_7322"} +{"question": "Any works report that contrastive visual learning could enhance contrastive sentence embeddings?", "answer": ["Non-Linguistic Supervision for Contrastive Learning of Sentence Embeddings"], "answer_arxiv_id": ["2209.09433"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_7323"} +{"question": "Could you name studies that applied Information Bottleneck principle in graph learning tasks?", "answer": ["Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "InfoGCL: Information-Aware Graph Contrastive Learning", "Graph Information Bottleneck for Subgraph Recognition", "Improving Subgraph Recognition with Variational Graph Information Bottleneck", "Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanism", "Graph Information Bottleneck", "Graph Structure Learning with Variational Information Bottleneck"], "answer_arxiv_id": ["2106.05819", "2110.15438", "2010.05563", "2112.09899", "2201.12987", "2010.12811", "2112.08903"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_7324"} +{"question": "Which paper employs an energy function to model the reverse transition kernel?", "answer": ["Learning Energy-Based Models by Diffusion Recovery Likelihood"], "answer_arxiv_id": ["2012.08125"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7325"} +{"question": "What work initiated the theoretical study of reproducibility in convex minimization problems?", "answer": ["Reproducibility in Optimization: Theoretical Framework and Limits"], "answer_arxiv_id": ["2202.04598"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_7326"} +{"question": "What are the existing data valuation methods?", "answer": ["Understanding Black-box Predictions via Influence Functions", "Towards Efficient Data Valuation Based on the Shapley Value", "Data Shapley: Equitable Valuation of Data for Machine Learning", "A Note on “Towards Efficient Data Valuation Based on the Shapley Value”", "Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning", "Data Valuation using Reinforcement Learning"], "answer_arxiv_id": ["1703.04730", "1902.10275", "1904.02868v2", "2302.11431", "2110.14049", "1909.11671"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_7327"} +{"question": "Which papers explored the different behavior of graph Laplacians in manifold’s boundary and its interior?", "answer": ["Diffusion Maps for Embedded Manifolds with Boundary with Applications to PDEs", "When locally linear embedding hits boundary"], "answer_arxiv_id": ["1912.01391", "1811.04423"], "source_meta": {"published_time": "20210728"}, "qid": "AutoScholarQuery_train_7328"} +{"question": "Which studies propose to condition a T2I model on face embeddings?", "answer": ["Face0: Instantaneously Conditioning a Text-to-Image Model on a Face"], "answer_arxiv_id": ["2306.06638"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_7329"} +{"question": "Which works proposed different fine-tuning methods for transfer learning in deep learning?", "answer": ["Big Transfer (BiT): General Visual Representation Learning", "Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning", "Revisiting Batch Normalization For Practical Domain Adaptation"], "answer_arxiv_id": ["1912.11370", "2102.03983", "1603.04779"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_7330"} +{"question": "Which works gave an evidence about the usefulness of polarization in aiding multi-view stereo depth?", "answer": ["Depth from a polarisation + RGB stereo pair"], "answer_arxiv_id": ["1903.12061"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_7331"} +{"question": "Which papers have characterized the optimal regret bound in one dimension?", "answer": ["Tight Regret Bounds for Bayesian Optimization in One Dimension"], "answer_arxiv_id": ["1805.11792"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_7332"} +{"question": "Which work proposed LoRA, a method to transfer VLP models without additional calculation overhead?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_7333"} +{"question": "Can you provide works that represent the long history in the evaluation of the faithfulness of explanations?", "answer": ["Towards Faithful Model Explanation in NLP: A Survey"], "answer_arxiv_id": ["2209.11326"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_7334"} +{"question": "What studies use diffusion models for motion generation in the context of body motion reconstruction?", "answer": ["Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking\n Inputs with Diffusion Model", "BoDiffusion: Diffusing Sparse Observations for Full-Body Human Motion\n Synthesis"], "answer_arxiv_id": ["2304.08577", "2304.11118"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_7335"} +{"question": "Which studies successfully used feed-forward layers of neural networks with fixed point iterations in computing inferences?", "answer": ["Implicitly Defined Layers in Neural Networks", "Nonsmooth Implicit Differentiation for Machine Learning and Optimization"], "answer_arxiv_id": ["2003.01822", "2106.04350"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_7336"} +{"question": "Which papers have considered convergence theorems for deep linear networks with the square cost?", "answer": ["Gradient descent with identity initialization efficiently learns positive definite linear transformations by deep residual networks", "A Convergence Analysis of Gradient Descent for Deep Linear Neural Networks"], "answer_arxiv_id": ["1802.06093", "1810.02281"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_7337"} +{"question": "What works contributed to the improvement of SISR performance through the introduction of various CNN-based architectures?", "answer": ["Residual Dense Network for Image Restoration", "Enhanced Deep Residual Networks for Single Image Super-Resolution", "ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks", "Deep Back-Projection Networks For Super-Resolution", "Image Super-Resolution Using Very Deep Residual Channel Attention\n Networks", "Residual Dense Network for Image Super-Resolution", "Photo-Realistic Single Image Super-Resolution Using a Generative\n Adversarial Network", "Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Fast and Accurate Image Super-Resolution with Deep Laplacian Pyramid\n Networks"], "answer_arxiv_id": ["1812.10477", "1707.02921", "1809.00219", "1803.02735", "1807.02758", "1802.08797", "1609.04802", "1511.04587", "1710.01992"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_7338"} +{"question": "Which studies have used self-training/bootstrapping approaches in machine translation?", "answer": ["Improving Neural Machine Translation Models with Monolingual Data"], "answer_arxiv_id": ["1511.06709v4"], "source_meta": {"published_time": "20220824"}, "qid": "AutoScholarQuery_train_7339"} +{"question": "Which papers proposed diffusion-based concept customization strategies for image generation?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2212.04488", "2106.09685"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7340"} +{"question": "Any studies about coreference relations in the context of Event Relation Extraction?", "answer": ["End-to-End Neural Event Coreference Resolution"], "answer_arxiv_id": ["2009.08153"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_7341"} +{"question": "Could you provide some works that use Transformers with spatio-temporal attention for VOS methods?", "answer": ["Associating Objects with Transformers for Video Object Segmentation", "HODOR: High-level Object Descriptors for Object Re-segmentation in Video Learned from Static Images"], "answer_arxiv_id": ["2106.02638", "2112.09131v2"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_7342"} +{"question": "Can you mention some works which solve permutations through a series of cycles?", "answer": ["Q-Match: Iterative Shape Matching via Quantum Annealing"], "answer_arxiv_id": ["2105.02878"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_7343"} +{"question": "What papers propose the generation of arbitrary-length human motions through autoregressive methods?", "answer": ["Context-aware Human Motion Prediction", "Long-term Human Motion Prediction with Scene Context", "Stochastic Scene-Aware Motion Prediction", "GOAL: Generating 4D Whole-Body Motion for Hand-Object Grasping", "Contact-aware Human Motion Forecasting", "T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete\n Representations"], "answer_arxiv_id": ["1904.03419", "2007.03672", "2108.08284", "2112.11454", "2210.03954", "2301.06052"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_7344"} +{"question": "Which research fine-tuned the reconstruction of human avatar with DreamBooth?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7345"} +{"question": "Which works apply the success of CLIP to diverse downstream tasks?", "answer": ["DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "ActionCLIP: A New Paradigm for Video Action Recognition", "PointCLIP: Point Cloud Understanding by CLIP"], "answer_arxiv_id": ["2112.01518", "2109.08472", "2112.02413"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_7346"} +{"question": "What research improved upon CommNet using an exponential kernel-based attention?", "answer": ["VAIN: Attentional Multi-agent Predictive Modeling"], "answer_arxiv_id": ["1706.06122"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_7347"} +{"question": "Which paper used techniques like Natural Language Inference (NLI) and dataset benchmarking for improving dialogue consistency?", "answer": ["Generating Persona Consistent Dialogues by Exploiting Natural Language\n Inference", "I like fish, especially dolphins: Addressing Contradictions in Dialogue Modeling", "Don't be Contradicted with Anything! CI-ToD: Towards Benchmarking\n Consistency for Task-oriented Dialogue System"], "answer_arxiv_id": ["1911.05889", "2012.13391v2", "2109.11292"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_7348"} +{"question": "Any works use adversarial training to improve realism in the context of lighting estimation?", "answer": ["Neural Light Field Estimation for Street Scenes with Differentiable Virtual Object Insertion"], "answer_arxiv_id": ["2208.09480v1"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_7349"} +{"question": "What papers discuss how privacy incentives can be directly proportional to a party's privacy budget?", "answer": ["Trading Data For Learning: Incentive Mechanism For On-Device Federated Learning"], "answer_arxiv_id": ["2009.05604"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_7350"} +{"question": "What research is the state-of-the-art Likelihood Ratio Attack (LiRA),which our attacks are based on, discussed in?", "answer": ["Membership Inference Attacks From First Principles"], "answer_arxiv_id": ["2112.03570"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_7351"} +{"question": "Which papers on Vision and Language Models (VLM) demonstrated remarkable potential in addressing vision and language tasks?", "answer": ["Learning multiple visual domains with residual adapters", "Efficient parametrization of multi-domain deep neural networks", "Parameter-efficient Multi-task Fine-tuning for Transformers via Shared\n Hypernetworks", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["1705.08045", "1803.10082", "2106.04489", "2101.00190"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_7352"} +{"question": "Which research papers have made strides in improved training paradigms in the field of Neural Combinatorial Optimization?", "answer": ["Neural Combinatorial Optimization with Reinforcement Learning", "POMO: Policy Optimization with Multiple Optima for Reinforcement Learning", "Sym-NCO: Leveraging Symmetricity for Neural Combinatorial Optimization", "ASP: Learn a Universal Neural Solver!", "Learning Generalizable Models for Vehicle Routing Problems via Knowledge Distillation", "DIMES: A Differentiable Meta Solver for Combinatorial Optimization Problems", "Efficient Active Search for Combinatorial Optimization Problems", "Learning to Solve Routing Problems via Distributionally Robust Optimization"], "answer_arxiv_id": ["1611.09940", "2010.16011", "2205.13209", "2303.00466", "2210.07686", "2210.04123", "2106.05126", "2202.07241"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_7353"} +{"question": "Which studies were developed to handle domain skew in federated learning?", "answer": ["FedBN: Federated Learning on Non-IID Features via Local Batch\n Normalization", "Generalizable Heterogeneous Federated Cross-Correlation and Instance\n Similarity Learning"], "answer_arxiv_id": ["2102.07623", "2309.16286"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_7354"} +{"question": "Which work suggested a recurrent generator architecture for video restoration and trained the model using losses that conserve joint statistics between consecutive frames?", "answer": ["Perceptual Video Super Resolution with Enhanced Temporal Consistency"], "answer_arxiv_id": ["1807.07930"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_7355"} +{"question": "What studies proposed meta-RL algorithms for HiP-MDP modelling?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Offline Meta-Reinforcement Learning with Advantage Weighting"], "answer_arxiv_id": ["1703.03400", "2008.06043"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7356"} +{"question": "Could you list some works propose dynamically expandable network for Continual Learning?", "answer": ["Lifelong Learning with Dynamically Expandable Networks", "Compacting, Picking and Growing for Unforgetting Continual Learning", "Efficient Continual Learning with Modular Networks and Task-Driven Priors"], "answer_arxiv_id": ["1708.01547", "1910.06562", "2012.12631"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_7357"} +{"question": "Which studies developed methods based on the lifting paradigm in the area of Camera-based 3D Perception?", "answer": ["BEVDet: High-performance Multi-camera 3D Object Detection in\n Bird-Eye-View", "BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object\n Detection", "BEVStereo: Enhancing Depth Estimation in Multi-view 3D Object Detection\n with Dynamic Temporal Stereo", "Time Will Tell: New Outlooks and A Baseline for Temporal Multi-View 3D\n Object Detection"], "answer_arxiv_id": ["2112.11790", "2206.10092", "2209.10248", "2210.02443"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_7358"} +{"question": "What works show that if a player is paired with an optimistic best-response opponent, the first player's strategy can converge to the minimax policy?", "answer": ["Online Reinforcement Learning in Stochastic Games", "A Self-Play Posterior Sampling Algorithm for Zero-Sum Markov Games"], "answer_arxiv_id": ["1712.00579", "2210.01907"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_7359"} +{"question": "What works have shown enhancements in generalization using video of a human performing the desired task?", "answer": ["One-Shot Imitation from Observing Humans via Domain-Adaptive Meta-Learning", "Learning One-Shot Imitation from Humans without Humans"], "answer_arxiv_id": ["1802.01557", "1911.01103"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_7360"} +{"question": "What works use the ensemble model in model-based reinforcement learning?", "answer": ["When to Trust Your Model: Model-Based Policy Optimization", "Model-Augmented Actor-Critic: Backpropagating through Paths", "MOPO: Model-based Offline Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["1906.08253", "2005.08068", "2005.13239", "2005.05951"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7361"} +{"question": "What works explored the area of neural architecture search-based methods in domain generalization research?", "answer": ["NAS-OoD: Neural Architecture Search for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2109.02038"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_7362"} +{"question": "What research works provided small-loss bounds in multi-arm bandits?", "answer": ["Learning in Games: Robustness of Fast Convergence"], "answer_arxiv_id": ["1606.06244"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7363"} +{"question": "What are the studies that utilized efficient feature grids based on hashing for accelerating rendering?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Compact Neural Graphics Primitives with Learned Hash Probing"], "answer_arxiv_id": ["2201.05989", "2312.17241"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_7364"} +{"question": "Which works use transformer architectures for model editing?", "answer": ["Modifying Memories in Transformer Models", "Language Anisotropic Cross-Lingual Model Editing"], "answer_arxiv_id": ["2012.00363", "2205.12677"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_7365"} +{"question": "Which research papers detail the online VIS methods that employ the tracking-by-detection approach?", "answer": ["Video Instance Segmentation", "Mask R-CNN", "SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation", "Crossover Learning for Fast Online Video Instance Segmentation", "SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation", "MinVIS: A Minimal Video Instance Segmentation Framework without Video-based Training"], "answer_arxiv_id": ["1905.04804", "1703.06870", "2007.14772", "2104.05970", "2103.10284", "2208.02245"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_7366"} +{"question": "What are some studies that deal with the effectiveness of random pruning masks in different situations?", "answer": ["Sanity-Checking Pruning Methods: Random Tickets can Win the Jackpot"], "answer_arxiv_id": ["2009.11094"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_7367"} +{"question": "Which paper first established the concept of Vision Transformer (ViT)?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_7368"} +{"question": "Which works are associated with diffusion probabilistic models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_7369"} +{"question": "Which papers highlighted the use of unsupervised segmentation methods in image analysis?", "answer": ["SLIC: Self-Supervised Learning with Iterative Clustering for Human\n Action Videos"], "answer_arxiv_id": ["2206.12534"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_7370"} +{"question": "What study introduces a learned canonical shape space as a means of category-level pose estimation?", "answer": ["Learning Canonical Shape Space for Category-Level 6D Object Pose and\n Size Estimation"], "answer_arxiv_id": ["2001.09322"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_7371"} +{"question": "What works focus on optimizing local learning algorithms by leveraging objective regularization for improving the global model learning of FL?", "answer": ["Federated Optimization in Heterogeneous Networks", "Federated Learning Based on Dynamic Regularization", "Model-Contrastive Federated Learning"], "answer_arxiv_id": ["1812.06127", "2111.04263", "2103.16257"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_7372"} +{"question": "What papers discuss and deal with the objective gap between lexicon-recovering language model pre-training and document-compressing dense-vector fine-tuning?", "answer": ["Latent Retrieval for Weakly Supervised Open Domain Question Answering", "Unsupervised Corpus Aware Language Model Pre-training for Dense Passage Retrieval", "Hyperlink-induced Pre-training for Passage Retrieval in Open-domain Question Answering", "Condenser: a Pre-training Architecture for Dense Retrieval", "Unsupervised Corpus Aware Language Model Pre-training for Dense Passage Retrieval", "SimLM: Pre-training with Representation Bottleneck for Dense Passage Retrieval"], "answer_arxiv_id": ["1906.00300", "2108.05540", "2203.06942", "2104.08253", "2108.05540", "2207.02578"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_7373"} +{"question": "Which research proposes a generative teacher?", "answer": ["Iterative Teaching by Data Hallucination"], "answer_arxiv_id": ["2210.17467v2"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_7374"} +{"question": "What works employed sparse representation techniques in HSI reconstruction?", "answer": ["Compressive Hyperspectral Imaging with Side Information"], "answer_arxiv_id": ["1502.06260"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_7375"} +{"question": "What works propose methods to construct parameterized skills in meta-RL setting?", "answer": ["Learning Parameterized Skills", "Meta Learning Shared Hierarchies", "Skill-based Meta-Reinforcement Learning", "Continuous Meta-Learning without Tasks"], "answer_arxiv_id": ["1206.6398", "1710.09767", "2204.11828", "1912.08866"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_7376"} +{"question": "Which works proposed variants of the primal-dual method in the context of randomized block-coordinate settings?", "answer": ["Randomized First-Order Methods for Saddle Point Optimization", "Adaptive Stochastic Primal-Dual Coordinate Descent for Separable Saddle Point Problems", "Stochastic Primal-Dual Coordinate Method for Regularized Empirical Risk Minimization"], "answer_arxiv_id": ["1409.8625v4", "1506.04093", "1409.3257v2"], "source_meta": {"published_time": "20220910"}, "qid": "AutoScholarQuery_train_7377"} +{"question": "What papers presented different versions like attention rollout and attention flow to analyze attention across multiple layers of Transformers?", "answer": ["Quantifying Attention Flow in Transformers"], "answer_arxiv_id": ["2005.00928"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_7378"} +{"question": "Which papers provide methods based on input-convex neural networks for OT computation?", "answer": ["2-Wasserstein Approximation via Restricted Convex Potentials with Application to Improved Training for GANs", "Optimal transport mapping via input convex neural networks", "Wasserstein-2 Generative Networks", "Continuous Wasserstein-2 Barycenter Estimation without Minimax Optimization"], "answer_arxiv_id": ["1902.07197", "1908.10962", "1909.13082v4", "2102.01752"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_7379"} +{"question": "What works apply regularization techniques to either the proxy itself or the design under consideration in Offline Model-based Optimization?", "answer": ["RoMA: Robust Model Adaptation for Offline Model-based Optimization", "Conservative Objective Models for Effective Offline Model-Based Optimization", "Offline Model-Based Optimization via Normalized Maximum Likelihood Estimation", "Bidirectional Learning for Offline Infinite-width Model-based Optimization", "Bidirectional Learning for Offline Model-based Biological Sequence Design"], "answer_arxiv_id": ["2110.14188", "2107.06882", "2102.07970", "2209.07507", "2301.02931"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_7380"} +{"question": "Could you provide me some works that analyze optimization of DNNs in the infinite-width regime for overparameterized nonlinear networks?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Mean-field theory of two-layers neural networks: dimension-free bounds and kernel limit", "On the Global Convergence of Gradient Descent for Over-parameterized Models using Optimal Transport"], "answer_arxiv_id": ["1806.07572", "1811.03804", "1811.03962", "1902.06015", "1805.09545"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_7381"} +{"question": "Can you list the works that included the inductive bias directly into the network layers and designed local versions of higher-order GNNs?", "answer": ["Weisfeiler and Leman go sparse: Towards scalable higher-order graph embeddings", "SpeqNets: Sparsity-aware Permutation-equivariant Graph Networks", "A Practical, Progressively-Expressive GNN"], "answer_arxiv_id": ["1904.01543", "2203.13913", "2210.09521"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_7382"} +{"question": "Could you provide me some works on box-based methods that decompose panoptic segmentation?", "answer": ["Panoptic Feature Pyramid Networks", "UPSNet: A Unified Panoptic Segmentation Network", "DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution", "Mask R-CNN", "Fully Convolutional Networks for Semantic Segmentation", "DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs", "Attention-guided Unified Network for Panoptic Segmentation", "Seamless Scene Segmentation", "An End-to-End Network for Panoptic Segmentation", "DeeperLab: Single-Shot Image Parser", "Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation", "Unifying Training and Inference for Panoptic Segmentation"], "answer_arxiv_id": ["1901.02446", "1901.03784", "2006.02334", "1703.06870", "1411.4038", "1606.00915", "1812.03904", "1905.01220", "1903.05027", "1902.05093", "1911.10194", "2001.04982"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_7383"} +{"question": "Which research improved on quality by making the softness parameter spatially-varying in implicit surface representation?", "answer": ["Adaptive Shells for Efficient Neural Radiance Field Rendering"], "answer_arxiv_id": ["2311.10091"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_7384"} +{"question": "Which works in the field of video summarization fall under the category of unsupervised approaches?", "answer": ["Discriminative Feature Learning for Unsupervised Video Summarization"], "answer_arxiv_id": ["1811.09791"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_7385"} +{"question": "Which papers are providing benign overfitting results similar to the settings of this study?", "answer": ["Benign Overfitting without Linearity: Neural Network Classifiers Trained by Gradient Descent for Noisy Linear Data", "Benign Overfitting in Linear Classifiers and Leaky ReLU Networks from KKT Conditions for Margin Maximization", "Benign Overfitting in Two-layer Convolutional Neural Networks", "From Tempered to Benign Overfitting in ReLU Neural Networks"], "answer_arxiv_id": ["2202.05928", "2303.01462", "2202.06526", "2305.15141"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_7386"} +{"question": "What are some examples of Transformer-based models in text-to-video generation?", "answer": ["GODIVA: Generating Open-DomaIn Videos from nAtural Descriptions", "NÜWA: Visual Synthesis Pre-training for Neural visUal World creAtion", "CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformers", "Phenaki: Variable Length Video Generation From Open Domain Textual Description"], "answer_arxiv_id": ["2104.14806", "2111.12417", "2205.15868", "2210.02399v1"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_7387"} +{"question": "What papers discuss methods for improving localization quality in weakly-supervised object detection?", "answer": ["Multiple Instance Detection Network with Online Instance Classifier Refinement"], "answer_arxiv_id": ["1704.00138"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_7388"} +{"question": "What works generates audio based on videos?", "answer": ["Taming Visually Guided Sound Generation", "Diff-Foley: Synchronized Video-to-Audio Synthesis with Latent Diffusion\n Models", "Conditional Generation of Audio from Video via Foley Analogies", "Physics-Driven Diffusion Models for Impact Sound Synthesis from Videos"], "answer_arxiv_id": ["2110.08791v1", "2306.17203", "2304.08490", "2303.16897"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_7389"} +{"question": "Which research argues that self-correction performance gains may rely on some high-quality external feedback?", "answer": ["Large Language Models Cannot Self-Correct Reasoning Yet"], "answer_arxiv_id": ["2310.01798"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_7390"} +{"question": "Can you list some works that used Bi-Level Optimization (BLO) in hyper-parameter optimization?", "answer": ["Self-Tuning Networks: Bilevel Optimization of Hyperparameters using Structured Best-Response Functions"], "answer_arxiv_id": ["1903.03088"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_7391"} +{"question": "Can you cite some studies that proposed algorithms to handle general deviations in the misspecified linear bandit (MLB) problem?", "answer": ["Learning with Good Feature Representations in Bandits and in RL with a Generative Model"], "answer_arxiv_id": ["1911.07676"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_7392"} +{"question": "Which papers discuss advanced systems such as ReAct, ReWOO, SwiftSage, DyLAN, and DP-LLM that are used to further develop LLM Agents?", "answer": ["ReAct: Synergizing Reasoning and Acting in Language Models", "ReWOO: Decoupling Reasoning from Observations for Efficient Augmented\n Language Models", "SwiftSage: A Generative Agent with Fast and Slow Thinking for Complex\n Interactive Tasks", "Dynamic LLM-Agent Network: An LLM-agent Collaboration Framework with\n Agent Team Optimization", "Dynamic Planning with a LLM"], "answer_arxiv_id": ["2210.03629", "2305.18323", "2305.17390", "2310.02170", "2308.06391"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_7393"} +{"question": "What works mention that maximizing diversity of outcomes can lead to a greater diversity of outcomes than undirected exploration?", "answer": ["GEP-PG: Decoupling Exploration and Exploitation in Deep Reinforcement Learning Algorithms"], "answer_arxiv_id": ["1802.05054"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_7394"} +{"question": "Could you provide me some studies about semi-supervised node classification problem where node features are coupled with relational information?", "answer": ["Community Detection in Networks with Node Attributes", "Inductive Representation Learning on Large Graphs", "Stochastic Blockmodels meet Graph Neural Networks", "Node Feature Extraction by Self-Supervised Multi-scale Neighborhood Prediction"], "answer_arxiv_id": ["1401.7267", "1706.02216", "1905.05738", "2111.00064"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_train_7395"} +{"question": "Are there any studies that created datasets from instruction videos?", "answer": ["HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million\n Narrated Video Clips", "Hierarchical Video-Moment Retrieval and Step-Captioning"], "answer_arxiv_id": ["1906.03327", "2303.16406"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_7396"} +{"question": "What work tackled the suboptimal performance on new classes by conditioning prompts directly on image instances?", "answer": ["Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2203.05557"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_7397"} +{"question": "Which papers are central to the task of fine-grained action understanding?", "answer": ["Fine-grained Activity Recognition in Baseball Videos", "FineAction: A Fine-Grained Video Dataset for Temporal Action\n Localization", "Fine-grained Temporal Contrastive Learning for Weakly-supervised\n Temporal Action Localization", "Fine-grained Video Categorization with Redundancy Reduction Attention", "Multi-Modal Domain Adaptation for Fine-Grained Action Recognition", "Video Pose Distillation for Few-Shot, Fine-Grained Sports Action\n Recognition", "Temporal Query Networks for Fine-grained Video Understanding", "ANetQA: A Large-scale Benchmark for Fine-grained Compositional Reasoning\n over Untrimmed Videos", "Fine-grained Video-Text Retrieval with Hierarchical Graph Reasoning", "How Do You Do It? Fine-Grained Action Understanding with Pseudo-Adverbs", "FineGym: A Hierarchical Video Dataset for Fine-grained Action\n Understanding", "SportsCap: Monocular 3D Human Motion Capture and Fine-grained\n Understanding in Challenging Sports Videos", "MultiSports: A Multi-Person Video Dataset of Spatio-Temporally Localized\n Sports Actions", "Temporal Query Networks for Fine-grained Video Understanding", "Weakly-Supervised Temporal Action Detection for Fine-Grained Videos with\n Hierarchical Atomic Actions"], "answer_arxiv_id": ["1804.03247", "2105.11107", "2203.16800", "1810.11189", "2001.09691", "2109.01305", "2104.09496", "2305.02519", "2003.00392", "2203.12344", "2004.06704", "2104.11452", "2105.07404", "2104.09496", "2207.11805"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_7398"} +{"question": "What studies have constructed the pseudo source domain through a generative model for source-free domain adaptation?", "answer": ["VDM-DA: Virtual Domain Modeling for Source Data-free Domain Adaptation"], "answer_arxiv_id": ["2103.14357"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7399"} +{"question": "What works utilized pre-deterministic methods to preprocess graph structure?", "answer": ["Adversarial Attack on Graph Structured Data", "Adversarial Attacks on Neural Networks for Graph Data"], "answer_arxiv_id": ["1806.02371", "1805.07984"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_7400"} +{"question": "Are there any studies that have used the extraction and alignment of optical flows, spatial maps, and nn-fields from the source video to improve consistency in video editing results?", "answer": ["Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation", "TokenFlow: Consistent Diffusion Features for Consistent Video Editing"], "answer_arxiv_id": ["2306.07954", "2307.10373"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_7401"} +{"question": "What are the studies developing Banach norm bounds in OCO?", "answer": ["Online Learning: Sufficient Statistics and the Burkholder Method", "Black-Box Reductions for Parameter-free Online Learning in Banach Spaces"], "answer_arxiv_id": ["1803.07617", "1802.06293"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_7402"} +{"question": "What researches are about the study of centralized Stackelberg games?", "answer": ["Can Reinforcement Learning Find Stackelberg-Nash Equilibria in General-Sum Markov Games with Myopic Followers?", "Sample-Efficient Learning of Stackelberg Equilibria in General-Sum Games", "Oracles & Followers: Stackelberg Equilibria in Deep Multi-Agent Reinforcement Learning", "Learning Correlated Stackelberg Equilibrium in General-Sum Multi-Leader-Single-Follower Games"], "answer_arxiv_id": ["2112.13521", "2102.11494", "2210.11942", "2210.12470v1"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7403"} +{"question": "Which works consider reward-free RL and proposed algorithms such as RF-UCRL and RF-Express?", "answer": ["Reward-Free Exploration for Reinforcement Learning", "Adaptive Reward-Free Exploration", "Fast active learning for pure exploration in reinforcement learning"], "answer_arxiv_id": ["2002.02794", "2006.06294", "2007.13442"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_7404"} +{"question": "Could you provide studies where DreamField was used?", "answer": ["Zero-Shot Text-Guided Object Generation with Dream Fields"], "answer_arxiv_id": ["2112.01455"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_7405"} +{"question": "What papers propose rasterizing the scene context of target agents into a bird-eye-view image for trajectory prediction?", "answer": ["Multimodal Trajectory Predictions for Autonomous Driving using Deep\n Convolutional Networks", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for\n Behavior Prediction"], "answer_arxiv_id": ["1809.10732", "1910.05449"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_7406"} +{"question": "What research papers have relied on continued or adaptive pre-training to extend the capabilities of LMs?", "answer": ["InfoXLM: An Information-Theoretic Framework for Cross-Lingual Language\n Model Pre-Training", "Adapting Pre-trained Language Models to African Languages via\n Multilingual Adaptive Fine-Tuning", "Glot500: Scaling Multilingual Corpora and Language Models to 500\n Languages"], "answer_arxiv_id": ["2007.07834", "2204.06487", "2305.12182"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_7407"} +{"question": "Could you provide some research references that have contributed to improving the affordability of large pre-trained Transformers models?", "answer": ["Movement Pruning: Adaptive Sparsity by Fine-Tuning", "The Lottery Ticket Hypothesis for Pre-trained BERT Networks", "Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Prune Once for All: Sparse Pre-Trained Language Models", "The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models", "Rethinking Network Pruning— under the Pre-train and Fine-tune Paradigm", "Block Pruning For Faster Transformers", "PLATON: Pruning Large Transformer Models with Upper Confidence Bound of Weight Importance", "M-FAC: Efficient Matrix-Free Approximations of Second-Order Information"], "answer_arxiv_id": ["2005.07683", "2007.12223", "2106.04533", "2111.05754", "2203.07259", "2104.08682", "2109.04838", "2206.12562v1", "2107.03356"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7408"} +{"question": "What work proposes an adaptive search engine-assisted learning method that can self-assess whether the LLM requires retrieval augmentation?", "answer": ["The Web Can Be Your Oyster for Improving Large Language Models"], "answer_arxiv_id": ["2305.10998"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_7409"} +{"question": "What works have been done on uncertainty estimation in segmentation?", "answer": ["Is segmentation uncertainty useful?"], "answer_arxiv_id": ["2103.16265"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_7410"} +{"question": "Provide me some works about the approximate implicit differentiation (AID) based approach in bilevel optimization?", "answer": ["Hyperparameter optimization with approximate gradient", "On Differentiating Parameterized Argmin and Argmax Problems with Application to Bi-level Optimization", "Approximation Methods for Bilevel Programming", "On the Iteration Complexity of Hypergradient Computation", "Optimizing Millions of Hyperparameters by Implicit Differentiation", "Bilevel Optimization: Convergence Analysis and Enhanced Design"], "answer_arxiv_id": ["1602.02355", "1607.05447", "1802.02246", "2006.16218", "1911.02590", "2010.07962"], "source_meta": {"published_time": "20230807"}, "qid": "AutoScholarQuery_train_7411"} +{"question": "Who made a comparison between the computation of interactions and dropout regularization?", "answer": ["Interpreting and Boosting Dropout from a Game-Theoretic View"], "answer_arxiv_id": ["2009.11729"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_7412"} +{"question": "Could you provide me some works about generative spoken language modeling using the discrete speech units?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation", "Generative Spoken Language Modeling from Raw Audio", "Text-Free Prosody-Aware Generative Spoken Language Modeling", "Generative Spoken Dialogue Language Modeling", "Textually Pretrained Speech Language Models"], "answer_arxiv_id": ["2209.03143", "2102.01192", "2109.03264", "2203.16502", "2305.13009"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_7413"} +{"question": "Could you mention a study that addresses how to eliminate artifacts on the Stefan problem within finite volume methods?", "answer": ["Numerical Artifacts in the Discontinuous Generalized Porous Medium Equation: How to Avoid Spurious Temporal Oscillations"], "answer_arxiv_id": ["1712.00132"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_7414"} +{"question": "What works are there on transformer-based models in visual question answering?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "Deep Modular Co-Attention Networks for Visual Question Answering"], "answer_arxiv_id": ["1908.02265", "1908.07490", "1906.10770"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_7415"} +{"question": "Which studies focused on the learning of signed distance fields (SDFs)?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Deep Local Shapes: Learning Local SDF Priors for Detailed 3D\n Reconstruction", "Implicit Neural Representations with Periodic Activation Functions", "Geometry-Consistent Neural Shape Representation with Implicit\n Displacement Fields"], "answer_arxiv_id": ["1901.05103", "2003.10983", "2006.09661", "2106.05187"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_7416"} +{"question": "What research has been conducted into the implementation of watermarks for language model detection?", "answer": ["Paraphrasing evades detectors of AI-generated text, but retrieval is an\n effective defense", "Protecting Language Generation Models via Invisible Watermarking", "A Watermark for Large Language Models"], "answer_arxiv_id": ["2303.13408", "2302.03162", "2301.10226"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_7417"} +{"question": "What research introduced the dual-domain method DuDoNet for sparse-view CT reconstruction?", "answer": ["DuDoNet: Dual Domain Network for CT Metal Artifact Reduction"], "answer_arxiv_id": ["1907.00273"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_7418"} +{"question": "What works propose solving exploration problems in reset-free RL using multi-task learning?", "answer": ["Reset-Free Reinforcement Learning via Multi-Task Learning: Learning Dexterous Manipulation Behaviors without Human Intervention", "Learning to Walk in the Real World with Minimal Human Effort"], "answer_arxiv_id": ["2104.11203", "2002.08550"], "source_meta": {"published_time": "20230106"}, "qid": "AutoScholarQuery_train_7419"} +{"question": "Any works about the application of expert models in computer vision?", "answer": ["BotBuster: Multi-platform Bot Detection Using A Mixture of Experts"], "answer_arxiv_id": ["2207.13658"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_7420"} +{"question": "Are there any studies that have leveraged motion as a cue for self-supervised object disentanglement?", "answer": ["Conditional Object-Centric Learning from Video"], "answer_arxiv_id": ["2111.12594"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_7421"} +{"question": "What are some papers that explore architectures and techniques for better representation of images?", "answer": ["Two-Stream Convolutional Networks for Action Recognition in Videos", "Aggregated Residual Transformations for Deep Neural Networks", "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications", "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "A ConvNet for the 2020s", "Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistles"], "answer_arxiv_id": ["1406.2199", "1611.05431", "1704.04861", "1905.11946", "2010.11929", "2103.14030", "2201.03545", "2306.00989v1"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7422"} +{"question": "What are some early attempts that used a CNN-based encoder and RNN/LSTM-based decoder for supervised IC?", "answer": ["Show and Tell: A Neural Image Caption Generator", "Long-term Recurrent Convolutional Networks for Visual Recognition and\n Description", "Show, Attend and Tell: Neural Image Caption Generation with Visual\n Attention", "An Empirical Study of Language CNN for Image Captioning"], "answer_arxiv_id": ["1411.4555", "1411.4389", "1502.03044", "1612.07086"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_7423"} +{"question": "Can you provide literature that discussed minimum ℓ2-norm solutions for models like kernel ridgeless regression, classification, and the random feature model?", "answer": ["Just Interpolate: Kernel “Ridgeless” Regression Can Generalize", "Finite-sample Analysis of Interpolating Linear Classifiers in the Overparameterized Regime", "Interpolating Classifiers Make Few Mistakes", "Classification vs regression in overparameterized regimes: Does the loss function matter?"], "answer_arxiv_id": ["1808.00387", "2004.12019", "2101.11815", "2005.08054"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_7424"} +{"question": "Could you mention the research where local relation layer was designed to model the context of local pixel pairs for image classification?", "answer": ["Local Relation Networks for Image Recognition"], "answer_arxiv_id": ["1904.11491"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_7425"} +{"question": "What research covers Pre-existing techniques like Domain Randomization (DR) and minimax adversarial curriculum learning?", "answer": ["CAD2RL: Real Single-Image Flight Without a Single Real Image", "Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Supervision via Competition: Robot Adversaries for Learning Tasks"], "answer_arxiv_id": ["1611.04201", "1703.06907", "1610.01685"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_7426"} +{"question": "Who proposed the intrinsic optimization objective in mesa-optimization for language models?", "answer": ["Risks from Learned Optimization in Advanced Machine Learning Systems"], "answer_arxiv_id": ["1906.01820"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_7427"} +{"question": "Which papers proposed methods for probabilistic simulation of dynamical systems?", "answer": ["ODE2VAE: Deep generative second order ODEs with Bayesian neural networks", "Leveraging Stochastic Predictions of Bayesian Neural Networks for Fluid Simulations", "Learning interacting dynamical systems with latent Gaussian process ODEs", "Neural Stochastic PDEs: Resolution-Invariant Learning of Continuous Spatiotemporal Dynamics"], "answer_arxiv_id": ["1905.10994", "2205.01222v1", "2205.11894", "2110.10249"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7428"} +{"question": "Are there any studies that investigated the cause of training instability of Vision Transformers?", "answer": ["Going deeper with Image Transformers", "Vision Transformers with Patch Diversification"], "answer_arxiv_id": ["2103.17239", "2104.12753"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_7429"} +{"question": "Are there any studies that optimized kernels for faster on-device inference?", "answer": ["Speed Is All You Need: On-Device Acceleration of Large Diffusion Models\n via GPU-Aware Optimizations"], "answer_arxiv_id": ["2304.11267"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_7430"} +{"question": "What studies address developing equivariant neural networks for local gauge transformations?", "answer": ["Gauge Equivariant Convolutional Networks and the Icosahedral CNN"], "answer_arxiv_id": ["1902.04615"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_7431"} +{"question": "Could you provide me some methods that have been proposed to automatically generate counterfactuals as a form of explanation?", "answer": ["Linguistically-Informed Transformations (LIT): A Method for\n Automatically Generating Contrast Sets", "Generating Plausible Counterfactual Explanations for Deep Transformers\n in Financial Text Classification", "Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and\n Improving Models", "CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generation", "Tailor: Generating and Perturbing Text with Semantic Controls", "Contrastive Explanations for Model Interpretability"], "answer_arxiv_id": ["2010.08580", "2010.12512", "2101.00288", "2210.04873", "2107.07150", "2103.01378"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_7432"} +{"question": "What works use Local SGD for distributed optimization in training large-scale deep learning systems?", "answer": ["Local SGD Converges Fast and Communicates Little", "Local SGD with Periodic Averaging: Tighter Analysis and Adaptive Synchronization", "On the Convergence of Local Descent Methods in Federated Learning"], "answer_arxiv_id": ["1805.09767", "1910.13598", "1910.14425v2"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_7433"} +{"question": "What papers have developed two-stage-method models for Scene Graph Generation (SGG)?", "answer": ["Learning to Compose Dynamic Tree Structures for Visual Contexts", "Scene Graph Generation by Iterative Message Passing", "Neural Motifs: Scene Graph Parsing with Global Context", "Energy-Based Learning for Scene Graph Generation", "Learning to Generate Scene Graph from Natural Language Supervision"], "answer_arxiv_id": ["1812.01880", "1701.02426", "1711.06640", "2103.02221", "2109.02227"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_7434"} +{"question": "What papers proposed the development of various techniques to make CNNs more lightweight and mobile-friendly?", "answer": ["MobileNets: Efficient Convolutional Neural Networks for Mobile Vision\n Applications", "MobileNetV2: Inverted Residuals and Linear Bottlenecks", "ShuffleNet: An Extremely Efficient Convolutional Neural Network for\n Mobile Devices", "ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture\n Design", "MixConv: Mixed Depthwise Convolutional Kernels", "MnasNet: Platform-Aware Neural Architecture Search for Mobile", "RepVGG: Making VGG-style ConvNets Great Again"], "answer_arxiv_id": ["1704.04861", "1801.04381", "1707.01083", "1807.11164", "1907.09595", "1807.11626", "2101.03697"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_7435"} +{"question": "What studies propose datasets for various aspects of video-language understanding such as spatiotemporal reasoning?", "answer": ["TGIF-QA: Toward Spatio-Temporal Reasoning in Visual Question Answering"], "answer_arxiv_id": ["1704.04497"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_7436"} +{"question": "Could you provide me some works that proposed accelerated training methods for DEQs?", "answer": ["On Training Implicit Models", "JFB: Jacobian-Free Backpropagation for Implicit Networks"], "answer_arxiv_id": ["2111.05177", "2103.12803"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_7437"} +{"question": "What research papers suggested using multiple embeddings per sample data in image-text matching?", "answer": ["Improving Cross-Modal Retrieval with Set of Diverse Embeddings"], "answer_arxiv_id": ["2211.16761"], "source_meta": {"published_time": "20240617"}, "qid": "AutoScholarQuery_train_7438"} +{"question": "What Knowledge-driven BVQA models extract handcrafted features from both the spatial and temporal domains to evaluate video quality?", "answer": ["UGC-VQA: Benchmarking Blind Video Quality Assessment for User Generated\n Content"], "answer_arxiv_id": ["2005.14354"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_7439"} +{"question": "Which works have developed efficient algorithms to learn adversarially noisy LTFs?", "answer": ["The Power of Localization for Efficiently Learning Linear Separators with Noise", "A PTAS for Agnostically Learning Halfspaces"], "answer_arxiv_id": ["1307.8371", "1410.7050"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_7440"} +{"question": "What are some studies that use online methods in subject-driven generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Key-Locked Rank One Editing for Text-to-Image Personalization"], "answer_arxiv_id": ["2208.01618", "2212.04488", "2208.12242", "2305.01644"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_7441"} +{"question": "Which papers used transformers like SST and DSVT as replacements for sparse CNNs in 3D representation learning?", "answer": ["Embracing Single Stride 3D Object Detector with Sparse Transformer", "DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets"], "answer_arxiv_id": ["2112.06375", "2301.06051"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_7442"} +{"question": "What research used paraphraser as the attacker to rewrite AI content, demonstrating effective attacks on many detectors?", "answer": ["Can AI-Generated Text be Reliably Detected?", "Paraphrasing evades detectors of AI-generated text, but retrieval is an\n effective defense"], "answer_arxiv_id": ["2303.11156", "2303.13408"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_7443"} +{"question": "What studies generalized group-equivariance to rotation and dilation?", "answer": ["Polar Transformer Networks"], "answer_arxiv_id": ["1709.01889"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_7444"} +{"question": "Could you provide me some studies about generating CEs by accounting for causal relations among input features?", "answer": ["Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers"], "answer_arxiv_id": ["1912.03277"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_7445"} +{"question": "Could you provide me with examples of studies that utilized LLMs to evaluate the proficiency of generative models in adhering to human instructions?", "answer": ["Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena", "M$^3$IT: A Large-Scale Dataset towards Multi-Modal Multilingual\n Instruction Tuning"], "answer_arxiv_id": ["2306.05685", "2306.04387"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7446"} +{"question": "What research papers exist that propose metric-based approaches of Membership Inference Attacks (MIAs) infering the presence of a sample by monitoring model behavior in terms of correctness?", "answer": ["Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "Label-Only Membership Inference Attacks", "Quantifying Membership Inference Vulnerability via Generalization Gap and Other Model Metrics", "White-box vs Black-box: Bayes Optimal Strategies for Membership Inference"], "answer_arxiv_id": ["1709.01604", "2007.14321", "2009.05669v1", "1908.11229"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_7447"} +{"question": "Which research developed methods for detecting 3D changes by employing geometric transformation consistency?", "answer": ["Objects Can Move: 3D Change Detection by Geometric Transformation\n Constistency"], "answer_arxiv_id": ["2208.09870"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_7448"} +{"question": "Who first demonstrated that the private labels can be estimated before solving the optimization problem to retrieve the client’s private data, reducing its complexity and improving the attack results?", "answer": ["iDLG: Improved Deep Leakage from Gradients"], "answer_arxiv_id": ["2001.02610"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_7449"} +{"question": "Which traditional post-hoc explanation methods have been adapted for Transformers in recent works?", "answer": ["Transformer Interpretability Beyond Attention Visualization", "XAI for Transformers: Better Explanations through Conservative\n Propagation"], "answer_arxiv_id": ["2012.09838", "2202.07304"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_7450"} +{"question": "What studies have been conducted on the challenges posed by applying model compression techniques to Large Language Models (LLMs)?", "answer": ["SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot"], "answer_arxiv_id": ["2301.00774"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_7451"} +{"question": "What studies proposed human pose estimation with inertial methods?", "answer": ["Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time", "Physical Inertial Poser (PIP): Physics-aware Real-time Human Motion\n Tracking from Sparse Inertial Sensors"], "answer_arxiv_id": ["1810.04703", "2203.08528"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_7452"} +{"question": "Which works illustrate the advancements in diffusion model technology?", "answer": ["Denoising Diffusion Probabilistic Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["2006.11239", "1503.03585"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7453"} +{"question": "Could you provide me some works that discuss the information maximization methods?", "answer": ["VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "Whitening for Self-Supervised Representation Learning", "Neural Manifold Clustering and Embedding"], "answer_arxiv_id": ["2105.04906", "2103.03230", "2007.06346", "2201.10000"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_7454"} +{"question": "Could you provide me with papers where the integration of inferential rules and LLMs is explored?", "answer": ["Logic-Driven Context Extension and Data Augmentation for Logical\n Reasoning of Text", "LINC: A Neurosymbolic Approach for Logical Reasoning by Combining\n Language Models with First-Order Logic Provers"], "answer_arxiv_id": ["2105.03659", "2310.15164"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_7455"} +{"question": "Can you name the paper that provides a benchmark showing no significant difference among different unsupervised anomaly detection methods?", "answer": ["ADBench: Anomaly Detection Benchmark"], "answer_arxiv_id": ["2206.09426"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_7456"} +{"question": "Which papers have shown promising results in reinforcement learning from human feedback in classical RL tasks?", "answer": ["Deep Reinforcement Learning from Human Preferences", "Reward learning from human preferences and demonstrations in Atari"], "answer_arxiv_id": ["1706.03741", "1811.06521"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_7457"} +{"question": "Which work introduced a diffusion on shapes with the learnable time parameter in geometric learning?", "answer": ["DiffusionNet: Discretization Agnostic Learning on Surfaces"], "answer_arxiv_id": ["2012.00888v3"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_7458"} +{"question": "Which paper proposed the first double-loop BSA approach to solve single-objective stochastic BLO problem?", "answer": ["Approximation Methods for Bilevel Programming"], "answer_arxiv_id": ["1802.02246"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_7459"} +{"question": "What paper discusses the unique nature of point clouds in 3D visual grounding?", "answer": ["Deep Learning for 3D Point Clouds: A Survey"], "answer_arxiv_id": ["1912.12033"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_7460"} +{"question": "Which works optimize a collection of learnable prompt tokens for few-shot adaptation, like the CoOp?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_7461"} +{"question": "What studies propose overparameterized neural nets can be optimized to global minima close to initialization by assuming sufficient width of several layers?", "answer": ["Gradient Descent Provably Optimizes Over-parameterized Neural Networks", "Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Stochastic Gradient Descent Optimizes Over-parameterized Deep ReLU Networks", "An Improved Analysis of Training Over-parameterized Deep Neural Networks", "Towards moderate overparameterization: global convergence guarantees for training shallow neural networks"], "answer_arxiv_id": ["1810.02054", "1811.03804", "1811.03962", "1811.08888", "1906.04688", "1902.04674"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_7462"} +{"question": "Could you provide me studies that explored the role of pessimism in offline RL?", "answer": ["Is Pessimism Provably Efficient for Offline RL?", "Adversarially Trained Actor Critic for Offline Reinforcement Learning"], "answer_arxiv_id": ["2012.15085", "2202.02446"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_7463"} +{"question": "Which initial works used Low Resolution (LR) images as the condition for the diffusion processes in image super-resolution techniques?", "answer": ["SRDiff: Single Image Super-Resolution with Diffusion Probabilistic\n Models", "Image Super-Resolution via Iterative Refinement"], "answer_arxiv_id": ["2104.14951", "2104.07636"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_7464"} +{"question": "Which papers discuss sample selection-based methods for identifying noisy samples using a small-loss criterion?", "answer": ["Learning to Learn from Noisy Labeled Data"], "answer_arxiv_id": ["1812.05214"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_7465"} +{"question": "Could you provide me some literature that proposed model decoupling in Federated Learning approaches?", "answer": ["Federated Learning with Personalization Layers", "Exploiting Shared Representations for Personalized Federated Learning", "FedDAR: Federated Domain-Aware Representation Learning", "FedBABU: Towards Enhanced Representation for Federated Image Classification", "Think Locally, Act Globally: Federated Learning with Local and Global Representations", "An Efficient Framework for Clustered Federated Learning"], "answer_arxiv_id": ["1912.00818", "2102.07078", "2209.04007", "2106.06042v3", "2001.01523", "2006.04088"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_7466"} +{"question": "What researches specialize in interleaved vision-language generation?", "answer": ["Generating Images with Multimodal Language Models", "Emu: Generative Pretraining in Multimodality", "DreamLLM: Synergistic Multimodal Comprehension and Creation"], "answer_arxiv_id": ["2305.17216", "2307.05222", "2309.11499"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_7467"} +{"question": "Which papers introduce the Score Distillation Sampling (SDS) method in text-to-3D generation using 2D diffusion?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_7468"} +{"question": "Which studies proposed blind SR methods?", "answer": ["Blind Super-Resolution With Iterative Kernel Correction", "Unfolding the Alternating Optimization for Blind Super Resolution"], "answer_arxiv_id": ["1904.03377", "2010.02631"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7469"} +{"question": "Can you mention the studies focused on adaptive gradient approach (AdaGrad) and its versions in nonconvex optimization?", "answer": ["Adaptive Bound Optimization for Online Convex Optimization", "Less Regret via Online Conditioning", "AdaGrad stepsizes: Sharp convergence over nonconvex landscapes", "On the Convergence of Stochastic Gradient Descent with Adaptive Stepsizes", "High Probability Bounds for a Class of Nonconvex Algorithms with AdaGrad Stepsize", "A High Probability Analysis of Adaptive SGD with Momentum"], "answer_arxiv_id": ["1002.4908", "1002.4862", "1806.01811", "1805.08114", "2204.02833", "2007.14294"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_7470"} +{"question": "What are some papers that studied the use of the multi-head self-attention mechanism and Transformer architectures like Set Transformer and UPDeT in MARL?", "answer": ["Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks", "Attention Is All You Need", "UPDeT: Universal Multi-agent Reinforcement Learning via Policy Decoupling with Transformers"], "answer_arxiv_id": ["1810.00825", "1706.03762", "2101.08001"], "source_meta": {"published_time": "20220310"}, "qid": "AutoScholarQuery_train_7471"} +{"question": "What paper shows that the equivalence of realizable and agnostic learnability extends across a wide variety of settings?", "answer": ["Realizable Learning is All You Need"], "answer_arxiv_id": ["2111.04746"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_7472"} +{"question": "Any works about the fairness of collaborative filtering in recommendation?", "answer": ["Beyond Parity: Fairness Objectives for Collaborative Filtering", "Fairness in Recommendation Ranking through Pairwise Comparisons", "An Intersectional Definition of Fairness", "Fairness in Rankings and Recommendations: An Overview"], "answer_arxiv_id": ["1705.08804", "1903.00780", "1807.08362", "2104.05994"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_7473"} +{"question": "Can you list studies that proposed methods to estimate unrestricted geometry?", "answer": ["Unrestricted Facial Geometry Reconstruction Using Image-to-Image\n Translation", "Learning Detailed Face Reconstruction from a Single Image"], "answer_arxiv_id": ["1703.10131", "1611.05053"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_7474"} +{"question": "What papers does not parametrize the exploration noise by the current state in unstructured exploration?", "answer": ["Continuous control with deep reinforcement learning"], "answer_arxiv_id": ["1509.02971"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_7475"} +{"question": "What studies have shown that the diversity of last layer features is positively correlated with the transferability of neural networks?", "answer": ["Rethinking supervised pre-training for better downstream transferring", "Discriminability-Transferability Trade-Off: An Information-Theoretic Perspective"], "answer_arxiv_id": ["2110.06014", "2203.03871"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7476"} +{"question": "Which papers developed hateful memes dataset?", "answer": ["Hate Speech in Pixels: Detection of Offensive Memes towards Automatic\n Moderation", "The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memes", "Exploring Hate Speech Detection in Multimodal Publications", "Detecting Harmful Memes and Their Targets"], "answer_arxiv_id": ["1910.02334", "2005.04790", "1910.03814", "2110.00413"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_7477"} +{"question": "Which paper analyzes non-linear CCA, an asymmetric variant of Kernel PCA with K+subscript𝐾K_{+}?", "answer": ["Predicting What You Already Know Helps: Provable Self-Supervised Learning"], "answer_arxiv_id": ["2008.01064"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_7478"} +{"question": "What works propose heuristics for minimizing the size of the ambiguity sets?", "answer": ["Beyond Confidence Regions: Tight Bayesian Ambiguity Sets for Robust MDPs", "Optimizing Percentile Criterion Using Robust MDPs"], "answer_arxiv_id": ["1902.07605", "1910.10786v3"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_7479"} +{"question": "Can you provide references for works where high-capacity models are used to predict future state information in high-dimensional data?", "answer": ["Probabilistic Recurrent State-Space Models", "Learning and Querying Fast Generative Models for Reinforcement Learning", "Learning Latent Dynamics for Planning from Pixels"], "answer_arxiv_id": ["1801.10395", "1802.03006", "1811.04551"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_7480"} +{"question": "Which works does the researcher use to show that transformers can approximate gradient descent on a broad range of loss functions?", "answer": ["Breaking the Curse of Dimensionality with Convex Neural Networks"], "answer_arxiv_id": ["1412.8690"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_7481"} +{"question": "Which studies demonstrate the benefits of RNNs for visual classification?", "answer": ["Recurrent computations for visual pattern completion"], "answer_arxiv_id": ["1706.02240v2"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_7482"} +{"question": "Can you provide some studies that used gradient-based methods like FGSM or PGD for adversarial attacks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1412.6572", "1706.06083"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7483"} +{"question": "Could you provide me some works that introduced adversarial priors, leverage multi-view cues or learn denoising models to enhance robustness in monocular human motion reconstruction?", "answer": ["Learning 3D Human Dynamics from Video", "VIBE: Video Inference for Human Body Pose and Shape Estimation", "Human Mesh Recovery from Multiple Shots", "Cyclic Test-Time Adaptation on Monocular Video for 3D Human Mesh\n Reconstruction"], "answer_arxiv_id": ["1812.01601", "1912.05656", "2012.09843", "2308.06554"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_7484"} +{"question": "Can you name some works about diffusion models in text-to-image generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Zero-Shot Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["1503.03585", "2006.11239", "1907.05600", "2010.02502", "2105.05233", "2112.10741", "2102.12092", "2204.06125", "2205.11487", "2206.10789", "2112.10752", "2301.07093"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_7485"} +{"question": "Are there any papers about Transformers in the context of operator learning?", "answer": ["Choose a Transformer: Fourier or Galerkin", "Learning Operators with Coupled Attention", "Transformer for Partial Differential Equations’ Operator Learning", "Continuous Spatiotemporal Transformers"], "answer_arxiv_id": ["2105.14995", "2201.01032", "2205.13671", "2301.13338v2"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_7486"} +{"question": "Could you provide me some works that investigated the use of external solvers to enhance reasoning in large language models?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks", "Faithful Chain-of-Thought Reasoning", "Faithful Reasoning Using Large Language Models", "Natural Language Deduction through Search over Statement Compositions", "ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language"], "answer_arxiv_id": ["2302.04761", "2211.12588", "2301.13379", "2208.14271", "2201.06028", "2012.13048"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7487"} +{"question": "Any research about pairwise sentence relations (e.g. NLI) in the context of large text units?", "answer": ["Composition of Sentence Embeddings:Lessons from Statistical Relational\n Learning"], "answer_arxiv_id": ["1904.02464"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_7488"} +{"question": "What is the first work to pre-train world models with datasets from different domains?", "answer": ["Reinforcement Learning with Action-Free Pre-Training from Videos"], "answer_arxiv_id": ["2203.13880"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7489"} +{"question": "Could you name some studies that use group-wise robust optimization to address the sub-population shift problem?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_7490"} +{"question": "Which papers discuss using learned models of transitions between states in conjunction with SFs for zero-shot methods?", "answer": ["Model-based Reinforcement Learning: A Survey."], "answer_arxiv_id": ["2006.16712"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_7491"} +{"question": "Which works discuss the use of gaussian splats in accelerating rendering?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_7492"} +{"question": "Which papers have extended 3D Gaussian Sphere (GS) to dynamic scenarios?", "answer": ["4D Gaussian Splatting for Real-Time Dynamic Scene Rendering", "Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene\n Reconstruction", "Real-time Photorealistic Dynamic Scene Representation and Rendering with\n 4D Gaussian Splatting"], "answer_arxiv_id": ["2310.08528", "2309.13101", "2310.10642"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_7493"} +{"question": "Could you provide me some studies understanding the maximally oriented partially directed acyclic graphs (MPDAGs)?", "answer": ["Interpreting and using CPDAGs with background knowledge", "Identifying causal effects in maximally oriented partially directed acyclic graphs", "Minimal Enumeration of All Possible Total Effects in a Markov Equivalence Class"], "answer_arxiv_id": ["1707.02171", "1910.02997", "2010.08611"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_7494"} +{"question": "Could you name the studies that focus on approximating or replacing the attention mechanism entirely in long-range transformers?", "answer": ["Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention", "Rethinking Attention with Performers", "FNet: Mixing Tokens with Fourier Transforms"], "answer_arxiv_id": ["2006.16236", "2009.14794", "2105.03824"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_7495"} +{"question": "Could you provide me some studies about parameter-isolation-based continual learning methods?", "answer": ["PathNet: Evolution Channels Gradient Descent in Super Neural Networks", "Progressive Neural Networks", "Overcoming catastrophic forgetting with hard attention to the task", "Incremental Learning Through Deep Adaptation", "Expert Gate: Lifelong Learning with a Network of Experts"], "answer_arxiv_id": ["1701.08734", "1606.04671", "1801.01423", "1705.04228", "1611.06194"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_7496"} +{"question": "In which researches does the agent samples a goal from previously visited states in goals to explore from methods?", "answer": ["PBCS: Efficient Exploration and Exploitation Using a Synergy between Reinforcement Learning and Motion Planning"], "answer_arxiv_id": ["2004.11667"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_7497"} +{"question": "Which works on OOD detection are based on softmax probability or logits?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "Scaling Out-of-Distribution Detection for Real-World Settings", "Energy-based Out-of-distribution Detection"], "answer_arxiv_id": ["1610.02136", "1706.02690", "1911.11132", "2010.03759"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_7498"} +{"question": "What papers propose solutions for learning on imbalanced data in the field of self-supervised learning?", "answer": ["Rethinking the Value of Labels for Improving Class-Imbalanced Learning", "Self-supervised Learning is More Robust to Dataset Imbalance", "Is Self-Supervised Learning More Robust Than Supervised Learning?", "Beyond Supervised vs. Unsupervised: Representative Benchmarking and Analysis of Image Representation Learning", "Divide and Contrast: Self-supervised Learning from Uncurated Data", "Self-Damaging Contrastive Learning", "Contrastive Learning with Boosted Memorization"], "answer_arxiv_id": ["2006.07529", "2110.05025", "2206.05259", "2206.08347", "2105.08054", "2106.02990", "2205.12693"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_7499"} +{"question": "Which studies focused on enhancing the training efficiency of DETR?", "answer": ["Conditional DETR for Fast Training Convergence", "Anchor DETR: Query Design for Transformer-Based Object Detection", "DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETR", "QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small Object Detection", "DN-DETR: Accelerate DETR Training by Introducing Query DeNoising"], "answer_arxiv_id": ["2108.06152", "2109.07107", "2201.12329", "2103.09136", "2203.01305"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_7500"} +{"question": "Could you provide me some studies that used different algorithms like reinforcement learning, evolution, gradient descent, or Bayesian Optimization for search in NAS?", "answer": ["Neural Architecture Search with Reinforcement Learning", "Large-Scale Evolution of Image Classifiers", "DARTS: Differentiable Architecture Search", "DrNAS: Dirichlet Neural Architecture Search", "Searching for A Robust Neural Architecture in Four GPU Hours", "Neural Architecture Search with Bayesian Optimisation and Optimal Transport", "BANANAS: Bayesian Optimization with Neural Architectures for Neural Architecture Search", "Interpretable Neural Architecture Search via Bayesian Optimisation with Weisfeiler-Lehman Kernels"], "answer_arxiv_id": ["1611.01578", "1703.01041", "1806.09055", "2006.10355", "1910.04465", "1802.07191", "1910.11858", "2006.07556"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_7501"} +{"question": "What papers have explored interpretability using task reasoning through textual descriptions or vision-language similarity in Vision-Language methods?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Towards a Visual-Language Foundation Model for Computational Pathology", "LLaVA-Med: Training a Large Language-and-Vision Assistant for\n Biomedicine in One Day", "VL-InterpreT: An Interactive Visualization Tool for Interpreting\n Vision-Language Transformers", "Visual Language Pretrained Multiple Instance Zero-Shot Transfer for\n Histopathology Images", "MDNet: A Semantically and Visually Interpretable Medical Image Diagnosis\n Network"], "answer_arxiv_id": ["2103.00020", "2307.12914", "2306.00890", "2203.17247", "2306.07831", "1707.02485"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_7502"} +{"question": "Identify some studies where DPMs have been utilized for solving inverse problems.", "answer": ["Diffusion Posterior Sampling for General Noisy Inverse Problems", "Denoising Diffusion Restoration Models"], "answer_arxiv_id": ["2209.14687", "2201.11793"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_7503"} +{"question": "Which work points out that the Hamiltonian neural network is equivalent to a second order Node?", "answer": ["Deconstructing the Inductive Biases of Hamiltonian Neural Networks"], "answer_arxiv_id": ["2202.04836"], "source_meta": {"published_time": "20220922"}, "qid": "AutoScholarQuery_train_7504"} +{"question": "What studies analyze the computational complexity theory perspective of autoregressive models' limitations?", "answer": ["Limitations of Autoregressive Models and Their Alternatives"], "answer_arxiv_id": ["2010.11939"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7505"} +{"question": "What is the seminal paper that explored how neural networks can memorize completely random labels?", "answer": ["Understanding deep learning requires rethinking generalization"], "answer_arxiv_id": ["1611.03530v2"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_7506"} +{"question": "Which papers utilized probabilistic extensions of IoU in the field of information retrieval?", "answer": ["Maximally Consistent Sampling and the Jaccard Index of Probability Distributions"], "answer_arxiv_id": ["1809.04052"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_7507"} +{"question": "Which studies proposed sampling-based motion planning?", "answer": ["Sampling-based Algorithms for Optimal Motion Planning", "Fast Marching Tree: a Fast Marching Sampling-Based Method for Optimal Motion Planning in Many Dimensions", "Batch Informed Trees (BIT*): Sampling-based Optimal Planning via the Heuristically Guided Search of Implicit Random Geometric Graphs"], "answer_arxiv_id": ["1105.1186", "1306.3532", "1405.5848"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_7508"} +{"question": "Are there any studies that utilized Influence function in statistical machine learning?", "answer": ["Understanding Black-box Predictions via Influence Functions", "Coresets via Bilevel Optimization for Continual Learning and Streaming", "Understanding Black-box Predictions via Influence Functions"], "answer_arxiv_id": ["1703.04730", "2006.03875", "1703.04730"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_7509"} +{"question": "Could you provide me some studies that explore early versions of adaptive training?", "answer": ["Enabling Machine Learning-Ready HPC Ensembles with Merlin", "A Survey of Deep Active Learning"], "answer_arxiv_id": ["1912.02892", "2009.00236"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_7510"} +{"question": "Can you name studies that applied a similar idea to residual flows by masking the weight matrices accordingly?", "answer": ["Graphical Residual Flows"], "answer_arxiv_id": ["2204.11846"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_7511"} +{"question": "Could you list some studies that looked into rehearsal-based methods in CL?", "answer": ["Co2L: Contrastive Continual Learning", "Dark Experience for General Continual Learning: a Strong, Simple Baseline"], "answer_arxiv_id": ["2106.14413", "2004.07211"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_7512"} +{"question": "Which techniques utilized coordinate classification for pose estimation?", "answer": ["SimCC: a Simple Coordinate Classification Perspective for Human Pose\n Estimation", "RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose"], "answer_arxiv_id": ["2107.03332", "2303.07399"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_7513"} +{"question": "What works suggest using edited NeRF for shading-driven adjustments, such as relighting and texturing?", "answer": ["NeRV: Neural Reflectance and Visibility Fields for Relighting and View\n Synthesis", "Editing Conditional Radiance Fields", "NeRF for Outdoor Scene Relighting", "PaletteNeRF: Palette-based Color Editing for NeRFs", "IntrinsicNeRF: Learning Intrinsic Neural Radiance Fields for Editable\n Novel View Synthesis"], "answer_arxiv_id": ["2012.03927", "2105.06466", "2112.05140", "2212.12871", "2210.00647"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_7514"} +{"question": "Could you provide me some studies on diffusion models in generative modeling?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Imagen Video: High Definition Video Generation with Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2112.10752", "2205.11487", "2210.02303"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7515"} +{"question": "Which works use the Transformer in image matching?", "answer": ["SuperGlue: Learning Feature Matching with Graph Neural Networks", "The Animation Transformer: Visual Correspondence via Segment Matching"], "answer_arxiv_id": ["1911.11763", "2109.02614"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_7516"} +{"question": "Can you tell me which studies employed diffusion models for text-to-image tasks?", "answer": ["DALLE-2 is Seeing Double: Flaws in Word-to-Concept Mapping in Text2Image\n Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2210.10606", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_7517"} +{"question": "Which papers discuss dynamic neural networks and their advantages?", "answer": ["Dynamic Neural Networks: A Survey", "Computation-efficient Deep Learning for Computer Vision: A Survey", "Hundreds Guide Millions: Adaptive Offline Reinforcement Learning with Expert Guidance", "Zero-shot Generative Model Adaptation via Image-specific Prompt Learning", "Value-Consistent Representation Learning for Data-Efficient Reinforcement Learning", "Boosting Offline Reinforcement Learning with Action Preference Query", "Efficient Knowledge Distillation from Model Checkpoints", "Adaptive Rotated Convolution for Rotated Object Detection"], "answer_arxiv_id": ["2102.04906", "2308.13998", "2309.01448", "2304.03119", "2206.12542", "2306.03362", "2210.06458", "2303.07820"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_7518"} +{"question": "Which papers focused on investigating adversarial attacks against visual question answering?", "answer": ["Improving Question Answering Model Robustness with Synthetic Adversarial Data Generation", "TASA: Deceiving Question Answering Models by Twin Answer Sentences Attack", "On the Efficacy of Adversarial Data Collection for Question Answering: Results from a Large-Scale Randomized Study", "Adversarial VQA: A New Benchmark for Evaluating the Robustness of VQA Models", "Human-Adversarial Visual Question Answering", "Trick Me If You Can: Human-in-the-loop Generation of Adversarial Examples for Question Answering", "Fooling Vision and Language Models Despite Localization and Attention Mechanism", "Towards Adversarial Attack on Vision-Language Pre-training Models"], "answer_arxiv_id": ["2104.08678", "2210.15221", "2106.00872", "2106.00245v2", "2106.02280", "1809.02701", "1709.08693v2", "2206.09391"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_7519"} +{"question": "Could you provide me a paper which introduced the strategy of converting images into visual tokens using the BERT model in CV tasks?", "answer": ["BEiT: BERT Pre-Training of Image Transformers"], "answer_arxiv_id": ["2106.08254"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_7520"} +{"question": "Could you provide me a study about adversarial training of DNNs for more accurate prediction of neural responses?", "answer": ["Joint rotational invariance and adversarial training of a dual-stream Transformer yields state of the art Brain-Score for Area V4"], "answer_arxiv_id": ["2203.06649"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7521"} +{"question": "Could you provide me some research about medical domain for VLM multi-tasks learning?", "answer": ["Multi-modal Understanding and Generation for Medical Images and Text via Vision-Language Pre-Training"], "answer_arxiv_id": ["2105.11333"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_7522"} +{"question": "Which study proposes a method that employs a related but different teacher-student objective?", "answer": ["Can Bad Teaching Induce Forgetting? Unlearning in Deep Networks Using an Incompetent Teacher"], "answer_arxiv_id": ["2205.08096"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_7523"} +{"question": "Which research proposed to match the features generated by the surrogate model for generating synthetic datasets?", "answer": ["Dataset Condensation with Distribution Matching"], "answer_arxiv_id": ["2110.04181"], "source_meta": {"published_time": "20221119"}, "qid": "AutoScholarQuery_train_7524"} +{"question": "Which research introduced Network Dissection, a method for understanding Deep Neural Networks (DNNs)?", "answer": ["Network Dissection: Quantifying Interpretability of Deep Visual Representations"], "answer_arxiv_id": ["1704.05796"], "source_meta": {"published_time": "20220423"}, "qid": "AutoScholarQuery_train_7525"} +{"question": "Are there any studies that utilize additional object signals like flow and depth in order to enable slot attention in complex scenes?", "answer": ["Conditional Object-Centric Learning from Video", "SAVi++: Towards End-to-End Object-Centric Learning from Real-World\n Videos"], "answer_arxiv_id": ["2111.12594", "2206.07764"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_7526"} +{"question": "Which works perform pixel-level matching with an equivariance loss?", "answer": ["Unsupervised Learning of Landmarks by Descriptor Vector Exchange", "Unsupervised Discovery of Object Landmarks as Structural Representations"], "answer_arxiv_id": ["1908.06427", "1804.04412"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_7527"} +{"question": "Which papers talk about mutual regularization in the context of knowledge exchange between heterogeneous agents?", "answer": ["Robust Domain Randomised Reinforcement Learning through Peer-to-Peer Distillation", "Non-local Policy Optimization via Diversity-regularized Collaborative Exploration"], "answer_arxiv_id": ["2012.04839", "2006.07781"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_7528"} +{"question": "Could you provide me some works that discuss the application of sphere pixelation in satellite missions to measure the cosmic microwave background in astrophysics?", "answer": ["DeepSphere: Efficient spherical Convolutional Neural Network with\n HEALPix sampling for cosmological applications"], "answer_arxiv_id": ["1810.12186"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_7529"} +{"question": "Could you provide me with a study that maintained the VAE architecture in LayoutVAE but replaced the encoder and decoder with Transformers?", "answer": ["Variational Transformer Networks for Layout Generation"], "answer_arxiv_id": ["2104.02416v1"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7530"} +{"question": "Could you provide me some research on vision-language pre-training for complex vision-language tasks, such as image captioning and visual question answering?", "answer": ["CoCa: Contrastive Captioners are Image-Text Foundation Models", "Scaling Up Vision-Language Pre-training for Image Captioning", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2205.01917", "2111.12233", "2201.12086", "2204.14198"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7531"} +{"question": "Which studies applied ensemble as a teacher model to long-tailed recognition?", "answer": ["Class-Balanced Distillation for Long-Tailed Visual Recognition", "Long-tailed Recognition by Routing Diverse Distribution-Aware Experts"], "answer_arxiv_id": ["2104.05279", "2010.01809"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_7532"} +{"question": "What papers cover the use of IL regularization in offline RL approaches like TD3+BC?", "answer": ["A Minimalist Approach to Offline Reinforcement Learning", "Conservative Q-Learning for Offline Reinforcement Learning"], "answer_arxiv_id": ["2106.06860", "2006.04779"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7533"} +{"question": "What works discuss Large Language Models (LLMs)?", "answer": ["Training language models to follow instructions with human feedback", "PaLM: Scaling Language Modeling with Pathways", "Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2203.02155", "2204.02311", "2210.11416"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_7534"} +{"question": "Can you mention any follow-up works that studied specialized architectures for incorporating structured DOM information?", "answer": ["Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration", "Learning to Navigate the Web", "DOM-Q-NET: Grounded RL on Structured Language"], "answer_arxiv_id": ["1802.08802", "1812.09195", "1902.07257"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_7535"} +{"question": "What works have built differentiable representations of 3D scenes using implicit neural networks?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "Learning Implicit Fields for Generative Shape Modeling", "SAL: Sign Agnostic Learning of Shapes from Raw Data", "Implicit Geometric Regularization for Learning Shapes", "Local Implicit Grid Representations for 3D Scenes", "Deep Local Shapes: Learning Local SDF Priors for Detailed 3D Reconstruction", "Convolutional Occupancy Networks", "Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes", "Implicit Neural Representations with Periodic Activation Functions", "Neural Lumigraph Rendering", "SDFDiff: Differentiable Rendering of Signed Distance Fields for 3D Shape Optimization", "Multiview Neural Surface Reconstruction by Disentangling Geometry and Appearance", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "DeepVoxels: Learning Persistent 3D Feature Embeddings", "PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization", "Learning to Infer Implicit Surfaces without 3D Supervision", "Neural Volumes: Learning Dynamic Renderable Volumes from Images", "Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision", "DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing"], "answer_arxiv_id": ["1901.05103", "1812.03828", "1812.02822", "1911.10414", "2002.10099", "2003.08981", "2003.10983", "2003.04618", "2101.10994", "2006.09661", "2103.11571", "1912.07109", "2003.09852", "1906.01618", "1812.01024v2", "1905.05172", "1911.00767", "1906.07751", "1912.07372", "1911.13225"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7536"} +{"question": "What studies developed a model that does not need to estimate the manifold and can utilize standard algorithms for posterior computation in GPs?", "answer": ["Bayesian Manifold Regression"], "answer_arxiv_id": ["1305.0617"], "source_meta": {"published_time": "20230116"}, "qid": "AutoScholarQuery_train_7537"} +{"question": "In what papers did the researcher improve expressiveness using multi-vector representations in the context of representation based models?", "answer": ["Sparse, Dense, and Attentional Representations for Text Retrieval"], "answer_arxiv_id": ["2005.00181"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_train_7538"} +{"question": "Which papers have made significant advances in understanding the nature of Neural Network Hessian maps?", "answer": ["Eigenvalues of the Hessian in Deep Learning: Singularity and Beyond", "An Investigation into Neural Net Optimization via Hessian Eigenvalue Density", "Empirical Analysis of the Hessian of Over-Parametrized Neural Networks", "Traces of Class/Cross-Class Structure Pervade Deep Learning Spectra"], "answer_arxiv_id": ["1611.07476", "1901.10159", "1706.04454", "2008.11865"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_7539"} +{"question": "What works discard irrelevant documents when cooperating LLMs with the retrieved documents?", "answer": ["When Not to Trust Language Models: Investigating Effectiveness of\n Parametric and Non-Parametric Memories"], "answer_arxiv_id": ["2212.10511"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_7540"} +{"question": "What is the reference for the miniF2F benchmark that this work relies on?", "answer": ["miniF2F: a cross-system benchmark for formal Olympiad-level mathematics"], "answer_arxiv_id": ["2109.00110"], "source_meta": {"published_time": "20220203"}, "qid": "AutoScholarQuery_train_7541"} +{"question": "Could you provide me some papers that contributed to the field of deep representation learning?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", "Mask R-CNN", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Exploring Simple Siamese Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: a Simple Framework for Masked Image Modeling", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1506.01497", "1703.06870", "1911.05722", "2002.05709", "2011.10566", "2006.07733", "2111.06377", "2111.09886", "1810.04805", "2005.14165", "1910.10683"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_7542"} +{"question": "Could you provide me some research papers that utilized the graph-based representation for audio-video diarization in egocentric videos?", "answer": ["Intel Labs at Ego4D Challenge 2022: A Better Baseline for Audio-Visual\n Diarization", "STHG: Spatial-Temporal Heterogeneous Graph Learning for Advanced\n Audio-Visual Diarization"], "answer_arxiv_id": ["2210.07764", "2306.10608"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_7543"} +{"question": "Could you provide me some studies about autoencoder-based methods in text-conditioned models for image synthesis?", "answer": ["Generating Diverse High-Fidelity Images with VQ-VAE-2", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["1906.00446", "2111.06377"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_7544"} +{"question": "What studies leverage the sparsity of spatial features to achieve different output activations for the input feature maps in dynamic networks?", "answer": ["Channel Gating Neural Networks", "Precision Gating: Improving Neural Network Efficiency with Dynamic\n Dual-Precision Activations"], "answer_arxiv_id": ["1805.12549", "2002.07136"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_7545"} +{"question": "Can you mention a research that utilized heuristic-based filtering mechanisms?", "answer": ["mT5: A massively multilingual pre-trained text-to-text transformer"], "answer_arxiv_id": ["2010.11934"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_7546"} +{"question": "What state-of-the-art models adopt the classifier-free guidance technique?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2203.13131", "2206.10789", "2205.11487"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_7547"} +{"question": "Could you provide me some works that combined the nonmonotone method with various algorithms for different purposes, including training shallow neural networks, speech recognition, and solving sparse optimization problems?", "answer": ["The Search direction Correction makes first-order methods faster"], "answer_arxiv_id": ["1905.06507"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_7548"} +{"question": "Could you provide me the studies that jointly learn pixel-level dense prediction tasks including depth estimation?", "answer": ["PAD-Net: Multi-Tasks Guided Prediction-and-Distillation Network for\n Simultaneous Depth Estimation and Scene Parsing"], "answer_arxiv_id": ["1805.04409"], "source_meta": {"published_time": "20240412"}, "qid": "AutoScholarQuery_train_7549"} +{"question": "What work applies multi-task prompt training which is similar to instruction tuning?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization"], "answer_arxiv_id": ["2110.08207"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_7550"} +{"question": "Can you indicate the studies that focus on token pruning methods in Transformers?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "Adaptive Token Sampling For Efficient Vision Transformers", "SPViT: Enabling Faster Vision Transformers via Soft Token Pruning", "AdaViT: Adaptive Tokens for Efficient Vision Transformer"], "answer_arxiv_id": ["2106.02034v2", "2111.15668", "2111.15667", "2112.13890", "2112.07658"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_7551"} +{"question": "Which studies proposed prompt learning to improve the generalization to unseen classes in large-scale vision-language models?", "answer": ["Conditional Prompt Learning for Vision-Language Models", "Visual-Language Prompt Tuning with Knowledge-guided Context Optimization", "Align Your Prompts: Test-Time Prompting with Distribution Alignment for\n Zero-Shot Generalization"], "answer_arxiv_id": ["2203.05557", "2303.13283", "2311.01459"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_7552"} +{"question": "Which paper originally proposed knowledge distillation as a technique to reduce the cost of training and deploying deep learning models?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_7553"} +{"question": "Could you list some works which proposed efficient variants of Transformer?", "answer": ["Generating Long Sequences with Sparse Transformers", "Reformer: The Efficient Transformer", "Big Bird: Transformers for Longer Sequences", "Random Feature Attention", "cosFormer : Rethinking Softmax in Attention", "Stable, Fast and Accurate: Kernelized Attention with Relative Positional Encoding"], "answer_arxiv_id": ["1904.10509", "2001.04451", "2007.14062", "2103.02143", "2202.08791", "2106.12566"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_7554"} +{"question": "Any works directly distilling labels from Mask2former’s predicted probabilities?", "answer": ["Panoptic Lifting for 3D Scene Understanding with Neural Fields"], "answer_arxiv_id": ["2212.09802"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_7555"} +{"question": "Which works focus on online HD map construction directly from on-board sensor data?", "answer": ["HDMapNet: An Online HD Map Construction and Evaluation Framework", "LiDAR2Map: In Defense of LiDAR-Based Semantic Map Construction Using\n Online Camera Distillation", "VectorMapNet: End-to-end Vectorized HD Map Learning", "MapTR: Structured Modeling and Learning for Online Vectorized HD Map\n Construction"], "answer_arxiv_id": ["2107.06307", "2304.11379", "2206.08920", "2208.14437"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_7556"} +{"question": "Which studies have proposed separate networks to identify frame-wise visual relationships?", "answer": ["Query Twice: Dual Mixture Attention Meta Learning for Video\n Summarization"], "answer_arxiv_id": ["2008.08360"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_7557"} +{"question": "What research introduced the KernelGAN method?", "answer": ["Blind Super-Resolution Kernel Estimation using an Internal-GAN"], "answer_arxiv_id": ["1909.06581"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_7558"} +{"question": "Could you give me examples of research that favored implicit functions like SDFs in direct 3D generation?", "answer": ["Get3DHuman: Lifting StyleGAN-Human into a 3D Generative Model using\n Pixel-aligned Reconstruction Priors", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "EVA3D: Compositional 3D Human Generation from 2D Image Collections", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "DreamHuman: Animatable 3D Avatars from Text"], "answer_arxiv_id": ["2302.01162", "2304.00916", "2210.04888", "2205.08535", "2306.09329"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_7559"} +{"question": "What papers have focused on analyzing incentives-aware collaboration by examining the equilibrium and stable formation of coalitions in FL?", "answer": ["One for One, or All for All: Equilibria and Optimality of Collaboration in Federated Learning", "Model-sharing Games: Analyzing Federated Learning Under Voluntary Participation"], "answer_arxiv_id": ["2103.03228", "2010.00753"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_7560"} +{"question": "What method uses a rotation prediction task for test-time training?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts"], "answer_arxiv_id": ["1909.13231"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_7561"} +{"question": "Could you give me research that uses diffusion models for 2D open-vocabulary segmentation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7562"} +{"question": "Could you provide me some studies on developing a single visual foundation model to serve as a general feature extractor in visuo-motor control?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "R3M: A Universal Visual Representation for Robot Manipulation", "Masked Visual Pre-training for Motor Control", "Visual Reinforcement Learning with Self-Supervised 3D Representations", "On Pre-Training for Visuo-Motor Control: Revisiting a Learning-from-Scratch Baseline", "Where are we in the search for an Artificial Visual Cortex for Embodied Intelligence?", "GNFactor: Multi-Task Real Robot Learning with Generalizable Neural Feature Fields"], "answer_arxiv_id": ["2107.03380", "2203.03580", "2203.12601", "2203.06173", "2210.07241", "2212.05749", "2303.18240", "2308.16891"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_7563"} +{"question": "Which research papers have used genetic programming for interpretable RL policies?", "answer": ["Interpretable Policies for Reinforcement Learning by Genetic Programming"], "answer_arxiv_id": ["1712.04170"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_7564"} +{"question": "What works found that using VQ for inter-component communication within neural networks can enhance model generalization ability?", "answer": ["Discrete-Valued Neural Communication", "Discrete Key-Value Bottleneck", "Adaptive Discrete Communication Bottlenecks with Dynamic Vector Quantization"], "answer_arxiv_id": ["2107.02367", "2207.11240", "2202.01334"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_7565"} +{"question": "Any research papers addressing the issue of representation collapse in supervised federated learning?", "answer": ["Towards Understanding and Mitigating Dimensional Collapse in\n Heterogeneous Federated Learning"], "answer_arxiv_id": ["2210.00226"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_7566"} +{"question": "Which works used QA as automatic evaluation metric for summarization?", "answer": ["Question Answering as an Automatic Evaluation Metric for News Article\n Summarization", "Towards Question-Answering as an Automatic Metric for Evaluating the\n Content Quality of a Summary"], "answer_arxiv_id": ["1906.00318", "2010.00490"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_7567"} +{"question": "Which papers apply the overfitting procedure to domain of reconstruction?", "answer": ["Deep Geometric Prior for Surface Reconstruction"], "answer_arxiv_id": ["1811.10943"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7568"} +{"question": "What are some studies that utilized soft-predictions as the knowledge between server and clients in federated learning?", "answer": ["FedMD: Heterogenous Federated Learning via Model Distillation", "Ensemble Distillation for Robust Model Fusion in Federated Learning", "Parameterized Knowledge Transfer for Personalized Federated Learning"], "answer_arxiv_id": ["1910.03581", "2006.07242", "2111.02862"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_7569"} +{"question": "Which studies utilize a large-scale language model to enhance the text-spelling knowledge in the field of image generation?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers"], "answer_arxiv_id": ["2205.11487", "2211.01324"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_7570"} +{"question": "What works have investigated the use of transformer-based methods in surgical instrument segmentation?", "answer": ["TraSeTR: Track-to-Segment Transformer with Contrastive Query for Instance-level Instrument Segmentation in Robotic Surgery", "MATIS: Masked-Attention Transformers for Surgical Instrument Segmentation"], "answer_arxiv_id": ["2202.08453", "2303.09514"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7571"} +{"question": "Which researches offer alternative ways of representing OBBs?", "answer": ["Gliding vertex on the horizontal bounding box for multi-oriented object\n detection", "Oriented Objects as pairs of Middle Lines", "Oriented Object Detection in Aerial Images with Box Boundary-Aware\n Vectors", "PolarDet: A Fast, More Precise Detector for Rotated Target in Aerial\n Images"], "answer_arxiv_id": ["1911.09358", "1912.10694", "2008.07043", "2010.08720"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_7572"} +{"question": "What study describes the use of a zero-shot prompt for rectifying errors in generated SQL queries?", "answer": ["DIN-SQL: Decomposed In-Context Learning of Text-to-SQL with\n Self-Correction"], "answer_arxiv_id": ["2304.11015"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_7573"} +{"question": "What are some works that studied private fine-tuning?", "answer": ["Differentially Private Language Models Benefit from Public Pre-training", "One size does not fit all: Investigating strategies for differentially-private learning across NLP tasks", "Benchmarking Differential Privacy and Federated Learning for BERT Models", "Large Scale Private Learning via Low-rank Reparametrization"], "answer_arxiv_id": ["2009.05886", "2112.08159", "2106.13973", "2106.09352"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_7574"} +{"question": "Could you provide me with research that proposed algorithms to reduce the communication cost?", "answer": ["Local SGD Converges Fast and Communicates Little", "On the Linear Speedup Analysis of Communication Efficient Momentum SGD for Distributed Non-Convex Optimization", "Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Sparsified SGD with Memory", "Qsparse-local-SGD: Distributed SGD with Quantization, Sparsification, and Local Computations", "Gradient Sparsification for Communication-Efficient Distributed Optimization", "signSGD: Compressed Optimisation for Non-Convex Problems"], "answer_arxiv_id": ["1805.09767", "1905.03817", "1807.06629", "1910.06378", "1809.07599", "1906.02367", "1710.09854v1", "1802.04434"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_7575"} +{"question": "Which paper generalized the DDIM and explained the advantages of a deterministic sampling scheme for fast sampling?", "answer": ["gDDIM: Generalized denoising diffusion implicit models"], "answer_arxiv_id": ["2206.05564"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_7576"} +{"question": "What papers have worked on making multimodal web document datasets publicly accessible?", "answer": ["Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with Text"], "answer_arxiv_id": ["2304.06939"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_7577"} +{"question": "Which paper used persistent homology to estimate the performance gap between training and testing error without utilizing a test dataset?", "answer": ["Computing the Testing Error without a Testing Set"], "answer_arxiv_id": ["2005.00450"], "source_meta": {"published_time": "20210503"}, "qid": "AutoScholarQuery_train_7578"} +{"question": "What are the studies that employed a coarse-to-fine hierarchical sampling strategy where the final samples are obtained by importance sampling of a coarse proposal distribution concerning NeRF?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Baking Neural Radiance Fields for Real-Time View Synthesis"], "answer_arxiv_id": ["2003.08934", "2103.13415", "2111.12077", "2103.14645"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_7579"} +{"question": "Which works explored the moral knowledge encoded in Language Models (LM)?", "answer": ["Does Moral Code Have a Moral Code? Probing Delphi's Moral Philosophy", "Large Pre-trained Language Models Contain Human-like Biases of What is\n Right and Wrong to Do", "Speaking Multiple Languages Affects the Moral Bias of Language Models", "Align on the Fly: Adapting Chatbot Behavior to Established Norms"], "answer_arxiv_id": ["2205.12771", "2103.11790", "2211.07733", "2312.15907"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7580"} +{"question": "Could you provide me some works that incorporated color invariants directly into deep neural network architectures?", "answer": ["Zero-Shot Day-Night Domain Adaptation with a Physics Prior"], "answer_arxiv_id": ["2108.05137"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7581"} +{"question": "Which research works focus on fine-tuning pre-trained models for improving their performance?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts", "Training language models to follow instructions with human feedback", "Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2109.01652", "2202.01279", "2203.02155", "2210.11416"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_7582"} +{"question": "What are some works that focus on surface rendering for neural surface reconstruction?", "answer": ["Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision", "Multiview Neural Surface Reconstruction by Disentangling Geometry and Appearance", "DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing", "Neural Lumigraph Rendering"], "answer_arxiv_id": ["1912.07372", "2003.09852", "1911.13225", "2103.11571"], "source_meta": {"published_time": "20220826"}, "qid": "AutoScholarQuery_train_7583"} +{"question": "Could you name some works that discussed the risk of trustworthiness in automated systems under out-of-distribution examples or distributional shift?", "answer": ["Can You Trust Your Model's Uncertainty? Evaluating Predictive\n Uncertainty Under Dataset Shift", "Underspecification Presents Challenges for Credibility in Modern Machine\n Learning"], "answer_arxiv_id": ["1906.02530", "2011.03395"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_7584"} +{"question": "Where can I find research that discusses editing methods inspired by the attention-based structure of stable diffusion?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Energy-Based Cross Attention for Bayesian Context Update in\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2211.12572", "2208.01626", "2306.09869"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_7585"} +{"question": "What is an example of a token pruning work for CV tasks?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification"], "answer_arxiv_id": ["2106.02034v2"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_7586"} +{"question": "Could you name some contributions to modeling developments contributed by Open Catalyst Project (OCP)?", "answer": ["GemNet: Universal Directional Graph Neural Networks for Molecules", "Directional Message Passing for Molecular Graphs"], "answer_arxiv_id": ["2106.08903v10", "2003.03123"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_7587"} +{"question": "Can you provide me with some researches about equipping LMs with the ability to use web browsers?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback"], "answer_arxiv_id": ["2112.09332"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_7588"} +{"question": "Could you provide studies that utilized non-attention neural modules for efficiency?", "answer": ["Conformer: Convolution-augmented Transformer for Speech Recognition", "When Attention Meets Fast Recurrence: Training Language Models with Reduced Compute"], "answer_arxiv_id": ["2005.08100", "2102.12459"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_7589"} +{"question": "Which works have been done on decomposing dynamic scenes into objects?", "answer": ["NeuralDiff: Segmenting 3D objects that move in egocentric videos", "Semantic Attention Flow Fields for Monocular Dynamic Scene Decomposition", "D2NeRF: Self-Supervised Decoupling of Dynamic and Static Objects from a Monocular Video"], "answer_arxiv_id": ["2110.09936", "2303.01526", "2205.15838"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_7590"} +{"question": "Are there any works that combine semantically similar tokens to reduce redundancies?", "answer": ["Token Merging: Your ViT But Faster"], "answer_arxiv_id": ["2210.09461"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_7591"} +{"question": "Could you provide some researches that improved NeRF in terms of rendering quality and capability?", "answer": ["Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields"], "answer_arxiv_id": ["2111.12077"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_7592"} +{"question": "What work proposes algorithms for permutation spaces, which occur in problems such as compiler optimization?", "answer": ["Bayesian Optimization over Permutation Spaces"], "answer_arxiv_id": ["2112.01049"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_7593"} +{"question": "Which papers propose an algorithm to determine identifiability of counterfactual queries from interventional data?", "answer": ["What Counterfactuals Can Be Tested", "Nested Counterfactual Identification from Arbitrary Surrogate Experiments"], "answer_arxiv_id": ["1206.5294", "2107.03190"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_7594"} +{"question": "Which references discussed the concept of learning the dynamics of the task as the shared structure in decision-making?", "answer": ["One-Shot Learning of Manipulation Skills with Online Dynamics Adaptation and Neural Network Priors", "Learning to Adapt in Dynamic, Real-World Environments through Meta-Reinforcement Learning", "A Model-based Approach for Sample-efficient Multi-task Reinforcement Learning"], "answer_arxiv_id": ["1509.06841", "1803.11347", "1907.04964"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_7595"} +{"question": "What research works focused on more sparse configurations, particularly six-IMU setting for inertial motion capture systems?", "answer": ["Sparse Inertial Poser: Automatic 3D Human Pose Estimation from Sparse\n IMUs", "Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time", "TransPose: Real-time 3D Human Translation and Pose Estimation with Six\n Inertial Sensors", "Physical Inertial Poser (PIP): Physics-aware Real-time Human Motion\n Tracking from Sparse Inertial Sensors"], "answer_arxiv_id": ["1703.08014", "1810.04703", "2105.04605", "2203.08528"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_7596"} +{"question": "Which datasets are specifically designed for action classification in the mmWave radar frameworks?", "answer": ["CubeLearn: End-to-end Learning for Human Motion Recognition from Raw mmWave Radar Signals"], "answer_arxiv_id": ["2111.03976"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_7597"} +{"question": "What are the works that have studied the min ℓ2-norm solutions for linear regression by using simple features like Gaussian or Fourier features?", "answer": ["To Understand Deep Learning We Need to Understand Kernel Learning", "Two models of double descent for weak features", "Benign Overfitting in Linear Regression", "Surprises in High-Dimensional Ridgeless Least Squares Interpolation", "Harmless interpolation of noisy data in regression"], "answer_arxiv_id": ["1802.01396", "1903.07571", "1906.11300", "1903.08560", "1903.09139"], "source_meta": {"published_time": "20230409"}, "qid": "AutoScholarQuery_train_7598"} +{"question": "Which papers proposed a method for predicting a hairstyle using only a single image?", "answer": ["NeuralHDHair: Automatic High-fidelity Hair Modeling from a Single Image\n Using Implicit Neural Representations", "HairStep: Transfer Synthetic to Real Using Strand and Depth Maps for\n Single-View 3D Hair Modeling"], "answer_arxiv_id": ["2205.04175", "2303.02700"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_7599"} +{"question": "Are there any works that showed connection between P&R and any DeGroot’s divergences?", "answer": ["On the Theoretical Equivalence of Several Trade-Off Curves Assessing Statistical Proximity"], "answer_arxiv_id": ["2006.11809"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7600"} +{"question": "What works are about the models acquiring 'true' representations of synthetic environments?", "answer": ["Emergent World Representations: Exploring a Sequence Model Trained on a\n Synthetic Task", "Emergence of Abstract State Representations in Embodied Sequence\n Modeling", "Evidence of Meaning in Language Models Trained on Programs"], "answer_arxiv_id": ["2210.13382", "2311.02171", "2305.11169"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_7601"} +{"question": "What research studies proposed strategies for generating training data for retrieval model?", "answer": ["REPLUG: Retrieval-Augmented Black-Box Language Models"], "answer_arxiv_id": ["2301.12652"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_7602"} +{"question": "What are the studies that have focused on improving factuality by abstaining from answering questions?", "answer": ["Language Models (Mostly) Know What They Know", "Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation\n in Natural Language Generation", "Adaptation with Self-Evaluation to Improve Selective Prediction in LLMs", "R-Tuning: Instructing Large Language Models to Say `I Don't Know'"], "answer_arxiv_id": ["2207.05221", "2302.09664", "2310.11689", "2311.09677"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_7603"} +{"question": "Could you provide me some works about the limitations and expressivity of basic graph neural networks?", "answer": ["How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["1810.00826"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_7604"} +{"question": "What works are related to adversarial watermarking in image processing?", "answer": ["Adversarial Example Does Good: Preventing Painting Imitation from\n Diffusion Models via Adversarial Examples", "Anti-DreamBooth: Protecting users from personalized text-to-image\n synthesis", "DUAW: Data-free Universal Adversarial Watermark against Stable Diffusion\n Customization", "Generative Watermarking Against Unauthorized Subject-Driven Image\n Synthesis", "Catch You Everything Everywhere: Guarding Textual Inversion via Concept\n Watermarking"], "answer_arxiv_id": ["2302.04578", "2303.15433", "2308.09889", "2306.07754", "2309.05940"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_7605"} +{"question": "What works estimate non-rigid deformations and textures on explicit mesh templates to fit the RGB input?", "answer": ["Dressing Avatars: Deep Photorealistic Appearance for Physically Simulated Clothing", "Drivable Avatar Clothing: Faithful Full-Body Telepresence with Dynamic Clothing Driven by Sparse RGB-D Input"], "answer_arxiv_id": ["2206.15470", "2310.05917"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_7606"} +{"question": "Could you mention the works which discussed combining 𝚒𝙻𝚀𝚁 with learned dynamics models utilising Jacobian linearization matrices?", "answer": ["Learning-based Model Predictive Control for Safe Exploration"], "answer_arxiv_id": ["1803.08287"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_7607"} +{"question": "What works propose generating programs hierarchically from abstractions to improve the scalability?", "answer": ["Neural Sketch Learning for Conditional Program Generation", "Learning to Infer Program Sketches", "Symbolic Learning to Optimize: Towards Interpretability and Scalability", "Symbolic Visual Reinforcement Learning: A Scalable Framework with Object-Level Abstraction and Differentiable Expression Search", "Generative Code Modeling with Graphs", "Hierarchical Neural Program Synthesis"], "answer_arxiv_id": ["1703.05698", "1902.06349", "2203.06578", "2212.14849", "1805.08490", "2303.06018"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_7608"} +{"question": "What works focused on text-based image retrieval?", "answer": ["Linking Image and Text with 2-Way Nets", "Deep Visual-Semantic Alignments for Generating Image Descriptions", "Conditional Image-Text Embedding Networks", "CAMP: Cross-Modal Adaptive Message Passing for Text-Image Retrieval", "VSE++: Improving Visual-Semantic Embeddings with Hard Negatives", "Probabilistic Embeddings for Cross-Modal Retrieval", "Learning Transferable Visual Models From Natural Language Supervision", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image\n Retrieval", "Zero-Shot Composed Image Retrieval with Textual Inversion", "FashionViL: Fashion-Focused Vision-and-Language Representation Learning", "Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language\n Feedback", "Composing Text and Image for Image Retrieval - An Empirical Odyssey"], "answer_arxiv_id": ["1608.07973", "1412.2306", "1711.08389", "1909.05506", "1707.05612", "2101.05068", "2103.00020", "2004.06165", "2102.05918", "2208.01618", "2302.03084", "2303.15247", "2207.08150", "1905.12794", "1812.07119"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_7609"} +{"question": "Could you provide me some works suggesting that inflexible high-performance kernels and limited programming abstractions are hindering innovative machine learning research?", "answer": ["The Hardware Lottery"], "answer_arxiv_id": ["2009.06489"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_7610"} +{"question": "What are the studies that improved single-change captioning by maximizing cross-view contrastive alignment between two images?", "answer": ["Self-supervised Cross-view Representation Reconstruction for Change\n Captioning"], "answer_arxiv_id": ["2309.16283"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_train_7611"} +{"question": "Any works studying the estimation of causal quantities in time series experiments?", "answer": ["Time series experiments and causal estimands: exact randomization tests and trading"], "answer_arxiv_id": ["1706.07840v2"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_7612"} +{"question": "What studies looked into RL with linear function approximation including linear and linear mixture MDPs?", "answer": ["Sample-Optimal Parametric Q-Learning Using Linearly Additive Features", "Provably Efficient Reinforcement Learning with Linear Function Approximation", "Is a Good Representation Sufficient for Sample Efficient Reinforcement Learning?", "Frequentist Regret Bounds for Randomized Least-Squares Value Iteration", "Reinforcement Learning with General Value Function Approximation: Provably Efficient Approach via Bounded Eluder Dimension", "Optimism in Reinforcement Learning with Generalized Linear Function Approximation", "Logarithmic Regret for Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Improved Variance-Aware Confidence Sets for Linear Bandits and Linear Mixture MDP", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Logarithmic Regret for Reinforcement Learning with Linear Function Approximation", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs", "Nearly Minimax Optimal Regret for Learning Infinite-horizon Average-reward MDPs with Linear Function Approximation", "Learning Stochastic Shortest Path with Linear Function Approximation"], "answer_arxiv_id": ["1902.04779", "1907.05388", "1910.03016", "1911.00567v7", "2005.10804", "1912.04136", "2011.11566", "2206.11489", "2212.06132", "2006.01107", "2101.12745", "2012.08507", "2011.11566", "2205.11507", "2102.07301v2", "2110.12727"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_7613"} +{"question": "Which works established methods for minimizing top-k error in multi-class loss functions?", "answer": ["Top-k Multiclass SVM", "Loss Functions for Top-k Error: Analysis and Insights", "Analysis and Optimization of Loss Functions for Multiclass, Top-k, and Multilabel Classification", "On the Consistency of Top-k Surrogate Losses"], "answer_arxiv_id": ["1511.06683", "1512.00486", "1612.03663", "1901.11141v2"], "source_meta": {"published_time": "20211230"}, "qid": "AutoScholarQuery_train_7614"} +{"question": "What papers discuss the divergence in the local training process?", "answer": ["Convergence and Accuracy Trade-Offs in Federated Learning and Meta-Learning", "From Local SGD to Local Fixed-Point Methods for Federated Learning", "Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization", "Local Adaptivity in Federated Learning: Convergence and Consistency"], "answer_arxiv_id": ["2103.05032v1", "2004.01442", "2007.07481", "2106.02305"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_7615"} +{"question": "Which papers use compact transformer architectures for building efficient vision transformers?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Rethinking Spatial Dimensions of Vision Transformers", "Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction\n without Convolutions", "Multiscale Vision Transformers"], "answer_arxiv_id": ["2103.14030", "2103.16302", "2102.12122", "2104.11227"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_7616"} +{"question": "What paper developed the Gibbs-With-Gradients (GWG) method, aiming to pick promising pixel(s) as the proposal?", "answer": ["Oops I Took A Gradient: Scalable Sampling for Discrete Distributions"], "answer_arxiv_id": ["2102.04509"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_7617"} +{"question": "What works have been done to evaluate the complex reasoning abilities of Large Language Models?", "answer": ["Towards Reasoning in Large Language Models: A Survey"], "answer_arxiv_id": ["2212.10403"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_7618"} +{"question": "Which works offer clarity on energy score as a method for modeling Out-of-Distribution (OOD) uncertainty?", "answer": ["Energy-based Out-of-distribution Detection", "POEM: Out-of-Distribution Detection with Posterior Sampling", "ATOM: Robustifying Out-of-distribution Detection Using Outlier Mining", "Towards neural networks that provably know when they don’t know", "Deep Anomaly Detection with Outlier Exposure"], "answer_arxiv_id": ["2010.03759", "2206.13687", "2006.15207", "1909.12180", "1812.04606"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_7619"} +{"question": "Which papers discuss the challenge of Q overestimation introduced by the distribution shift between the behavior policy and the learned policy?", "answer": ["Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems"], "answer_arxiv_id": ["2005.01643"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_7620"} +{"question": "Could you provide me some works that utilized large foundation model to improve audio-visual segmentation?", "answer": ["AV-SAM: Segment Anything Model Meets Audio-Visual Localization and\n Segmentation", "Prompting Segmentation with Sound Is Generalizable Audio-Visual Source\n Localizer", "BAVS: Bootstrapping Audio-Visual Segmentation by Integrating Foundation\n Knowledge"], "answer_arxiv_id": ["2305.01836", "2309.07929", "2308.10175"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_7621"} +{"question": "Could you provide me some research about multi-graph convolutions?", "answer": ["Multi-GCN: Graph Convolutional Networks for Multi-View Networks, with Applications to Global Poverty"], "answer_arxiv_id": ["1901.11213"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_7622"} +{"question": "What paper demonstrated the use of feature grids or tokens, which is a shift from using preprocess regions?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks"], "answer_arxiv_id": ["1506.01497"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_7623"} +{"question": "Could you provide the paper that used GPT-3’s sequence likelihood to estimate model performance?", "answer": ["GPTScore: Evaluate as You Desire"], "answer_arxiv_id": ["2302.04166"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_7624"} +{"question": "Which studies presented synthetic data application in scenarios with privacy constraints?", "answer": ["How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Models", "Beyond Privacy: Navigating the Opportunities and Challenges of Synthetic Data", "K-SHAP: Policy Clustering Algorithm for Anonymous Multi-Agent State-Action Pairs", "Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs"], "answer_arxiv_id": ["2102.08921", "2304.03722v1", "2302.11996", "1706.02633"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_7625"} +{"question": "What studies have looked into enhancing the robustness properties of ensembles?", "answer": ["Improving Adversarial Robustness via Promoting Ensemble Diversity", "Improving Adversarial Robustness of Ensembles with Diversity Training", "Boosting Randomized Smoothing with Variance Reduced Classifiers", "On the Certified Robustness for Ensemble Models and Beyond", "Mixture of Robust Experts (MoRE):A Robust Denoising Method towards multiple perturbations", "On the Adversarial Robustness of Mixture of Experts"], "answer_arxiv_id": ["1901.08846", "1901.09981", "2106.06946", "2107.10873", "2104.10586", "2210.10253"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_7626"} +{"question": "What papers used k-nearest neighbors for generating pseudo labels in this domain?", "answer": ["Unsupervised Person Re-identification via Multi-label Classification", "Invariance Matters: Exemplar Memory for Domain Adaptive Person\n Re-identification", "Unsupervised Person Re-identification by Soft Multilabel Learning"], "answer_arxiv_id": ["2004.09228", "1904.01990", "1903.06325"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_7627"} +{"question": "What are some research works that have studied distribution shifts and their effects on learning models?", "answer": ["Bridging Theory and Algorithm for Domain Adaptation", "On Localized Discrepancy for Domain Adaptation", "On the Value of Target Data in Transfer Learning"], "answer_arxiv_id": ["1904.05801", "2008.06242", "2002.04747"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_7628"} +{"question": "Which works criticize the use of balanced attack accuracy as the metric for assessing the efficacy of a MIA?", "answer": ["Membership Inference Attacks From First Principles"], "answer_arxiv_id": ["2112.03570"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_7629"} +{"question": "Who further improved the FGSM method by proposing iterated FGSM and PGD for adversarial training?", "answer": ["Adversarial examples in the physical world", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1607.02533", "1706.06083"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_7630"} +{"question": "On what papers can I find information about model compression using distillation technique?", "answer": ["Distilling the Knowledge in a Neural Network", "On the Efficacy of Knowledge Distillation", "Distilling Task-Specific Knowledge from BERT into Simple Neural Networks", "Training data-efficient image transformers & distillation through\n attention"], "answer_arxiv_id": ["1503.02531", "1910.01348", "1903.12136", "2012.12877"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_7631"} +{"question": "Which works propose diffusion models to continuous-time diffusion using forward and backward SDEs?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_7632"} +{"question": "What studies extended the Segment Anything Model to predict semantic categories?", "answer": ["Segment Everything Everywhere All at Once", "Hierarchical Open-vocabulary Universal Image Segmentation", "Semantic-SAM: Segment and Recognize Anything at Any Granularity"], "answer_arxiv_id": ["2304.06718", "2307.00764", "2307.04767"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_7633"} +{"question": "Which datasets provide entailment and contradiction pairs for metaphorical sentences and were used in recent studies?", "answer": ["FLUTE: Figurative Language Understanding through Textual Explanations"], "answer_arxiv_id": ["2205.12404"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_7634"} +{"question": "Could you provide me some studies about inference-efficient ensembling methods by sharing parameters?", "answer": ["BatchEnsemble: An alternative approach to Efficient Ensemble and Lifelong Learning", "Efficient and Scalable Bayesian Neural Nets with Rank-1 Factors"], "answer_arxiv_id": ["2002.06715", "2005.07186"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_7635"} +{"question": "Give examples of studies that researched hierarchical clustering in different models of computation.", "answer": ["Learning Hierarchical Structure of Clusterable Graphs"], "answer_arxiv_id": ["2207.02581"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_7636"} +{"question": "Which work proposed a benchmark for composed video retrieval (CoVR) and proposed the CoVR-BLIP framework?", "answer": ["CoVR: Learning Composed Video Retrieval from Web Video Captions"], "answer_arxiv_id": ["2308.14746"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_7637"} +{"question": "Could you provide references that employed detection and reconstruction tasks simultaneously between 2D and 3D graphs?", "answer": ["Pre-training Molecular Graph Representation with 3D Geometry"], "answer_arxiv_id": ["2110.07728"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_7638"} +{"question": "Are there works that implement an upper confidence bound acquisition function using a reparameterization trick?", "answer": ["Efficient Model-Based Reinforcement Learning through Optimistic Policy Search and Planning", "Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design"], "answer_arxiv_id": ["2006.08684", "0912.3995v4"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_7639"} +{"question": "Which studies primarily focus on enhancing model performance in task decomposition?", "answer": ["Decomposed Prompting: A Modular Approach for Solving Complex Tasks", "Successive Prompting for Decomposing Complex Questions", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2210.02406", "2212.04092", "2210.03629"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_7640"} +{"question": "Who demonstrated that cross-lingual ICL can be elicited with proper alignment between the source and target language examples?", "answer": ["Multilingual LLMs are Better Cross-lingual In-context Learners with\n Alignment"], "answer_arxiv_id": ["2305.05940"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_7641"} +{"question": "Which studies proposed leveraging execution information to enhance code generation?", "answer": ["Natural Language to Code Translation with Execution", "Competition-Level Code Generation with AlphaCode", "CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning"], "answer_arxiv_id": ["2204.11454", "2203.07814", "2207.01780"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_7642"} +{"question": "What are the papers about using an alternative noising process such as non-Markovian, a second-order Langevin dynamics, and non-linear diffusion processes?", "answer": ["Denoising Diffusion Implicit Models", "Score-Based Generative Modeling with Critically-Damped Langevin Diffusion", "Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling", "Likelihood Training of Schrödinger Bridge using Forward-Backward SDEs Theory"], "answer_arxiv_id": ["2010.02502", "2112.07068", "2106.01357", "2110.11291"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_7643"} +{"question": "What work proposed the idea of using transformer encoders and masked language modeling task to pre-train the model for NLP tasks?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_7644"} +{"question": "Could you provide me with a work about SGD that, even when a constant fraction of the training labels are corrupted by an adversary, produces neural networks that have a generalization error close to the label noise rate?", "answer": ["Provable Generalization of SGD-trained Neural Networks of Any Width in the Presence of Adversarial Label Noise"], "answer_arxiv_id": ["2101.01152"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_7645"} +{"question": "What works optimize entropy regularized optimal transport loss between observed samples and samples simulated from a current model in the trajectory inference?", "answer": ["Scaling Algorithms for Unbalanced Optimal Transport Problems", "Optimal Transport Tools (OTT): A JAX Toolbox for all things Wasserstein"], "answer_arxiv_id": ["1607.05816", "2201.12324"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_7646"} +{"question": "Can you list research studies that explored diffusion models in a variety of domains like image, video, and audio?", "answer": ["Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Flexible Diffusion Modeling of Long Videos", "Video Diffusion Models", "DiffWave: A Versatile Diffusion Model for Audio Synthesis"], "answer_arxiv_id": ["2006.11239", "2105.05233", "2205.11495", "2204.03458", "2009.09761v3"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_7647"} +{"question": "What works are about deep learning-based 3D feature descriptors?", "answer": ["3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions", "The Perfect Match: 3D Point Cloud Matching with Smoothed Densities", "RPM-Net: Robust Point Matching using Learned Features"], "answer_arxiv_id": ["1603.08182", "1811.06879", "2003.13479"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_7648"} +{"question": "In which research papers has active learning been shown to achieve exponentially smaller label complexity than passive learning?", "answer": ["Active and passive learning of linear separators under log-concave distributions", "Convergence Rates of Active Learning for Maximum Likelihood Estimation", "Active Learning for Cost-Sensitive Classification", "Improved Algorithms for Agnostic Pool-based Active Classification"], "answer_arxiv_id": ["1211.1082", "1506.02348", "1703.01014", "2105.06499"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_7649"} +{"question": "What studies have narrated the successful curation of cleaner datasets from videos using automatic speech recognition models?", "answer": ["MERLOT: Multimodal Neural Script Knowledge Models", "Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos"], "answer_arxiv_id": ["2106.02636", "2206.11795"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_7650"} +{"question": "Could you provide me some researches about the applying speed-up algorithms or product-quantization for dense-vector retrieval methods?", "answer": ["Learning Discrete Representations via Constrained Clustering for Effective and Efficient Dense Retrieval", "Distill-VQ: Learning Retrieval Oriented Vector Quantization By Distilling Knowledge from Dense Embeddings"], "answer_arxiv_id": ["2110.05789", "2204.00185"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_7651"} +{"question": "Which study found that a modality gap exists between the image embeddings and text embeddings sub-space?", "answer": ["Mind the Gap: Understanding the Modality Gap in Multi-modal Contrastive Representation Learning"], "answer_arxiv_id": ["2203.02053"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7652"} +{"question": "Could you provide me some works that used probabilistic grammar-based methods for code completion?", "answer": ["Mining Idioms from Source Code"], "answer_arxiv_id": ["1404.0417"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_7653"} +{"question": "Which work is comparatively discussed in connection with the effective resistance estimation problem?", "answer": ["A New Approach to Estimating Effective Resistances and Counting Spanning Trees in Expander Graphs"], "answer_arxiv_id": ["2211.01468"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_7654"} +{"question": "Which works demonstrate that nonparametric embeddings help reduce instances of hallucination in generation of accurate and factual content?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Improving language models by retrieving from trillions of tokens"], "answer_arxiv_id": ["2005.11401", "2112.04426"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_7655"} +{"question": "What research propose to predict dense semantic labels in addition to RGB colours?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene Representation", "NeSF: Neural Semantic Fields for Generalizable Semantic Segmentation of 3D Scenes"], "answer_arxiv_id": ["2103.15875", "2111.13260"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_7656"} +{"question": "Who proposed techniques for data generation under learning-to-denoise framework?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1907.05600", "2006.11239", "1503.03585"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_7657"} +{"question": "What paper introduces neural implicit scene representation with volume rendering for 3D scene reconstruction from monocular videos?", "answer": ["MonoNeuralFusion: Online Monocular Neural 3D Reconstruction with Geometric Priors"], "answer_arxiv_id": ["2209.15153"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_7658"} +{"question": "Which papers have previously studied the non-contextual setting of bandits with concave rewards?", "answer": ["Bandits with concave rewards and convex knapsacks", "Multi-objective Bandits: Optimizing the Generalized Gini Index"], "answer_arxiv_id": ["1402.5758", "1706.04933"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_7659"} +{"question": "What works aim to address multi-objective reinforcement learning with concave aggregation functions?", "answer": ["Learning Fair Policies in Multiobjective (Deep) Reinforcement Learning with Average and Discounted Rewards", "Socially Fair Reinforcement Learning", "Concave Utility Reinforcement Learning: the Mean-Field Game Viewpoint"], "answer_arxiv_id": ["2008.07773", "2208.12584", "2106.03787"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_7660"} +{"question": "Could you provide me some works using voxel-based neural radiance field (NeRF) with occupancy prediction?", "answer": ["UniOcc: Unifying Vision-Centric 3D Occupancy Prediction with Geometric\n and Semantic Rendering"], "answer_arxiv_id": ["2306.09117"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_7661"} +{"question": "Which works attempted to resolve 3D hand-mesh reconstruction by regressing MANO coefficients?", "answer": ["End-to-end Hand Mesh Recovery from a Monocular RGB Image", "Monocular Real-time Hand Shape and Motion Capture using Multi-modal Data", "BiHand: Recovering Hand Mesh with Multi-stage Bisected Hourglass\n Networks", "Learning joint reconstruction of hands and manipulated objects", "Hand Image Understanding via Deep Multi-Task Learning", "3D Hand Shape and Pose from Images in the Wild", "Pushing the Envelope for RGB-based Dense 3D Hand Pose Estimation via\n Neural Rendering", "FreiHAND: A Dataset for Markerless Capture of Hand Pose and Shape from\n Single RGB Images", "Model-based 3D Hand Reconstruction via Self-Supervised Learning", "Reconstructing Hand-Object Interactions in the Wild", "Hand-Object Contact Consistency Reasoning for Human Grasps Generation", "Semi-Supervised 3D Hand-Object Poses Estimation with Interactions in\n Time", "Collaborative Learning for Hand and Object Reconstruction with\n Attention-guided Graph Convolution"], "answer_arxiv_id": ["1902.09305", "2003.09572", "2008.05079", "1904.05767", "2107.11646", "1902.03451", "1904.04196", "1909.04349", "2103.11703", "2012.09856", "2104.03304", "2106.05266", "2204.13062"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_7662"} +{"question": "What papers investigated fairness under the covariate shift?", "answer": ["Robust Fairness under Covariate Shift"], "answer_arxiv_id": ["2010.05166"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_7663"} +{"question": "Which studies give the query complexity for the unconstrained case with curvature?", "answer": ["Linear Lower Bounds and Conditioning of Differentiable Games", "On Lower Iteration Complexity Bounds for the Convex Concave Saddle Point Problems", "Near-Optimal Algorithms for Minimax Optimization"], "answer_arxiv_id": ["1906.07300", "1912.07481", "2002.02417"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_7664"} +{"question": "Which studies utilized joint-embedding architectures for self-supervised learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2002.05709", "2103.03230", "2105.04906", "2104.14294"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_7665"} +{"question": "Which studies propose algorithms for contextual bandits with linear functions?", "answer": ["A Contextual-Bandit Approach to Personalized News Article Recommendation"], "answer_arxiv_id": ["1003.0146"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_7666"} +{"question": "What research works aid performance by optimizing the examples chosen for the prompt?", "answer": ["Active Prompting with Chain-of-Thought for Large Language Models", "What Makes Good In-Context Examples for GPT-3?"], "answer_arxiv_id": ["2302.12246", "2101.06804"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_7667"} +{"question": "Which study improved prototypes with a momentum update policy for smooth label adjustment in weakly supervised learning?", "answer": ["MoPro: Webly Supervised Learning with Momentum Prototypes"], "answer_arxiv_id": ["2009.07995"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_7668"} +{"question": "What works study extrapolation TKG reasoning methods?", "answer": ["Recurrent Event Network: Autoregressive Structure Inference over\n Temporal Knowledge Graphs", "Temporal Knowledge Graph Reasoning Based on Evolutional Representation\n Learning", "Temporal Knowledge Graph Forecasting with Neural ODE", "Learning from History: Modeling Temporal Knowledge Graphs with\n Sequential Copy-Generation Networks", "TimeTraveler: Reinforcement Learning for Temporal Knowledge Graph\n Forecasting", "Complex Evolutional Pattern Learning for Temporal Knowledge Graph\n Reasoning"], "answer_arxiv_id": ["1904.05530", "2104.10353", "2101.05151", "2012.08492", "2109.04101", "2203.07782"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_7669"} +{"question": "In which papers the researchers propose the use of physically inspired but computationally expensive deformation models to address the assumption of piece-wise rigid motion?", "answer": ["Virtual Elastic Objects"], "answer_arxiv_id": ["2201.04623"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7670"} +{"question": "Which studies introduce the concept of Generalized Category Discovery?", "answer": ["Generalized Category Discovery", "XCon: Learning with Experts for Fine-grained Category Discovery", "OpenCon: Open-world Contrastive Learning"], "answer_arxiv_id": ["2201.02609", "2208.01898", "2208.02764"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7671"} +{"question": "Could you provide me a paper that discussed the convergence of Damped PPM and showed their method converges with only one-sided dominance?", "answer": ["The Landscape of the Proximal Point Method for Nonconvex-Nonconcave Minimax Optimization"], "answer_arxiv_id": ["2006.08667"], "source_meta": {"published_time": "20221226"}, "qid": "AutoScholarQuery_train_7672"} +{"question": "What work can I refer to for understanding the role of Markov chains in image generation?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_7673"} +{"question": "Is there any work that studies text-only interactions in the context of Language-conditioned RL?", "answer": ["ALFWorld: Aligning Text and Embodied Environments for Interactive Learning"], "answer_arxiv_id": ["2010.03768"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_7674"} +{"question": "Which works dealt with linearly constrained neural networks and how are they different from the proposed research?", "answer": ["Linearly Constrained Neural Networks"], "answer_arxiv_id": ["2002.01600"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7675"} +{"question": "Could you provide some studies introducing GNNs?", "answer": ["Learning from Protein Structure with Geometric Vector Perceptrons", "DiffMD: A Geometric Diffusion Model for Molecular Dynamics Simulations"], "answer_arxiv_id": ["2009.01411", "2204.08672"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7676"} +{"question": "Could you give me some examples of works that utilized diffusion probabilistic models for scene synthesis?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Improved Techniques for Training Score-Based Generative Models", "Denoising Diffusion Probabilistic Models", "Diffusion Probabilistic Models for 3D Point Cloud Generation"], "answer_arxiv_id": ["1503.03585", "2006.09011", "2006.11239", "2103.01458"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_7677"} +{"question": "Which works suggest that representations extracted by pre-trained models are near to linear separable?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1911.05722", "2002.05709", "2006.07733", "2103.00020"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_7678"} +{"question": "Could you provide me a study that proposed a stratified class-specific attention-based transformer for construction of fine-grained relationships between support and query features?", "answer": ["Few-Shot 3D Point Cloud Semantic Segmentation via Stratified\n Class-Specific Attention Based Transformer Network"], "answer_arxiv_id": ["2303.15654"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_7679"} +{"question": "What research work has been done to solve degradation of neural network performance due to common corruptions such as Gaussian noise or blur?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["1903.12261"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_7680"} +{"question": "What work proposed Vision Transformer (ViT)?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_7681"} +{"question": "Could you example some works utilizing Transformer architectures for the segmentation task?", "answer": ["Segmenter: Transformer for Semantic Segmentation", "SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2105.05633", "2105.15203", "2012.15840", "2107.06278", "2112.01527"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_7682"} +{"question": "Which paper identified the equivalence between diffusion and score-based generative models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7683"} +{"question": "Which research proposes a method to compose dynamic storyboards with changing camera views in a virtual environment?", "answer": ["Dynamic Storyboard Generation in an Engine-based Virtual Environment for\n Video Production"], "answer_arxiv_id": ["2301.12688"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_7684"} +{"question": "Could you mention some studies that fall into the category of domain adaptation in dark domain learning?", "answer": ["Cross-Domain Car Detection Using Unsupervised Image-to-Image\n Translation: From Day to Night", "Guided Curriculum Model Adaptation and Uncertainty-Aware Evaluation for\n Semantic Nighttime Image Segmentation", "DANNet: A One-Stage Domain Adaptation Network for Unsupervised Nighttime\n Semantic Segmentation", "Multitask AET with Orthogonal Tangent Regularity for Dark Object\n Detection", "Cross-Domain Correlation Distillation for Unsupervised Domain Adaptation\n in Nighttime Semantic Segmentation", "HLA-Face: Joint High-Low Adaptation for Low Light Face Detection", "Self-Aligned Concave Curve: Illumination Enhancement for Unsupervised\n Adaptation", "Unsupervised Monocular Depth Estimation for Night-time Images using\n Adversarial Domain Feature Adaptation", "HLA-Face: Joint High-Low Adaptation for Low Light Face Detection", "Unsupervised Domain Adaptation for Nighttime Aerial Tracking", "2PCNet: Two-Phase Consistency Training for Day-to-Night Unsupervised\n Domain Adaptive Object Detection", "NightLab: A Dual-level Architecture with Hardness Detection for\n Segmentation at Night", "Guided Curriculum Model Adaptation and Uncertainty-Aware Evaluation for\n Semantic Nighttime Image Segmentation", "Dark Model Adaptation: Semantic Image Segmentation from Daytime to\n Nighttime"], "answer_arxiv_id": ["1907.08719", "1901.05946", "2104.10834", "2205.03346", "2205.00858", "2104.01984", "2210.03792", "2010.01402", "2104.01984", "2203.10541", "2303.13853", "2204.05538", "1901.05946", "1810.02575"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_7685"} +{"question": "What works have made use of Large Language Models for empathetic response generation?", "answer": ["Facilitating Multi-turn Emotional Support Conversation with Positive\n Emotion Elicitation: A Reinforcement Learning Approach"], "answer_arxiv_id": ["2307.07994"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_7686"} +{"question": "Which studies have introduced the use of auxiliary classification in the discriminator for Conditional Generative Adversarial Networks?", "answer": ["Conditional Image Synthesis With Auxiliary Classifier GANs"], "answer_arxiv_id": ["1610.09585"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_7687"} +{"question": "Which works weigh instances based on its context/relevance compared with the overall data distribution?", "answer": ["Instance-Conditional Timescales of Decay for Non-Stationary Learning"], "answer_arxiv_id": ["2212.05908"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_7688"} +{"question": "What research proposes decoupling the training of policy and value function to improve sample efficiency?", "answer": ["Phasic Policy Gradient"], "answer_arxiv_id": ["2009.04416"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_7689"} +{"question": "Which paper initially proposed the Activation Addition approach?", "answer": ["Activation Addition: Steering Language Models Without Optimization"], "answer_arxiv_id": ["2308.10248"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_7690"} +{"question": "Which works can be cited as early research in motion prediction?", "answer": ["Learning Human Motion Models for Long-term Predictions", "A Spatio-temporal Transformer for 3D Human Motion Prediction", "Long-term Human Motion Prediction with Scene Context", "On human motion prediction using recurrent neural networks", "Spatiotemporal Co-attention Recurrent Neural Networks for Human-Skeleton\n Motion Prediction"], "answer_arxiv_id": ["1704.02827", "2004.08692", "2007.03672", "1705.02445", "1909.13245"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_7691"} +{"question": "What study developed the Activity-coupled Cartesian Direction of Arrival (ACCDOA) vector?", "answer": ["ACCDOA: Activity-Coupled Cartesian Direction of Arrival Representation for Sound Event Localization and Detection"], "answer_arxiv_id": ["2010.15306"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7692"} +{"question": "What studies have utilized diffusion models for generating panoramic images?", "answer": ["DiffCollage: Parallel Generation of Large Content with Diffusion Models", "PanoGen: Text-Conditioned Panoramic Environment Generation for\n Vision-and-Language Navigation", "360-Degree Panorama Generation from Few Unregistered NFoV Images", "Customizing 360-Degree Panoramas through Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2303.17076", "2305.19195", "2308.14686", "2310.18840"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_7693"} +{"question": "What dataset gathered both human and ChatGPT responses for questions generated by humans on public datasets?", "answer": ["How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation,\n and Detection"], "answer_arxiv_id": ["2301.07597"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_7694"} +{"question": "Which resource discusses the expressiveness limitations of random-walk-based GNNs?", "answer": ["Walk Message Passing Neural Networks and Second-Order Graph Neural Networks"], "answer_arxiv_id": ["2006.09499"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_7695"} +{"question": "What works have established NODE as a universal approximator to ODEs?", "answer": ["Neural Ordinary Differential Equations", "Universal Approximation Property of Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366", "2012.02414"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_7696"} +{"question": "Which study is about instruction-tuning of models and datasets?", "answer": ["QLoRA: Efficient Finetuning of Quantized LLMs"], "answer_arxiv_id": ["2305.14314"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_7697"} +{"question": "Could you give me examples of research papers that employed fine-tuning in generative models with limited data?", "answer": ["Transferring GANs: generating images from limited data", "Freeze the Discriminator: a Simple Baseline for Fine-Tuning GANs", "MineGAN: effective knowledge transfer from GANs to target domains with few images", "Few-shot Image Generation with Elastic Weight Consolidation", "Few-shot Image Generation via Cross-domain Correspondence"], "answer_arxiv_id": ["1805.01677", "2002.10964", "1912.05270", "2012.02780", "2104.06820"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_7698"} +{"question": "Which study demonstrates that the ensemble of peer networks can boost performance and facilitate faster convergence in non-continual scenarios?", "answer": ["Large scale distributed neural network training through online\n distillation"], "answer_arxiv_id": ["1804.03235"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_7699"} +{"question": "Could you provide me some papers explored the Admissible Bellman Characterization (ABC) class to generalize Bellman Eluder (BE)?", "answer": ["A General Framework for Sample-Efficient Function Approximation in Reinforcement Learning"], "answer_arxiv_id": ["2209.15634"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7700"} +{"question": "Which pieces of research have suggested whitening or transforming the feature correlations to improve corruption robustness?", "answer": ["Decorrelated Batch Normalization", "Unsupervised Domain Adaptation using Feature-Whitening and Consensus Loss", "RobustNet: Improving Domain Generalization in Urban-Scene Segmentation via Instance Selective Whitening"], "answer_arxiv_id": ["1804.08450", "1903.03215", "2103.15597"], "source_meta": {"published_time": "20220624"}, "qid": "AutoScholarQuery_train_7701"} +{"question": "What research works demonstrate the similarity-based demonstration selection?", "answer": ["Selective Annotation Makes Language Models Better Few-Shot Learners", "Learning To Retrieve Prompts for In-Context Learning"], "answer_arxiv_id": ["2209.01975", "2112.08633"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7702"} +{"question": "Which papers proposed curricula based on the difficulty level of the curriculum?", "answer": ["Automatic Goal Generation for Reinforcement Learning Agents", "Intrinsic Motivation and Automatic Curricula via Asymmetric Self-Play"], "answer_arxiv_id": ["1705.06366", "1703.05407"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7703"} +{"question": "What papers have studied experimental designs with spatial/network spillover effects?", "answer": ["Graph Cluster Randomization: Network Exposure to Multiple Universes", "Randomization Inference for Peer Effects", "Designs for estimating the treatment effect in Networks with Interference", "A/B Testing in Dense Large-Scale Networks: Design and Inference", "Experimental Design under Network Interference", "Cluster-Adaptive Network A/B Testing: From Randomization to Estimation", "Rate-Optimal Cluster-Randomized Designs for Spatial Interference"], "answer_arxiv_id": ["1305.6979", "1807.01635", "1705.08524", "1901.10505", "2003.08421v4", "2008.08648", "2111.04219"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_7704"} +{"question": "Can you provide me with several papers on image generation using a pre-trained GAN?", "answer": ["Conditional Image Generation and Manipulation for User-Specified Content", "Efficient Neural Architecture for Text-to-Image Synthesis", "Network-to-Network Translation with Conditional Invertible Neural Networks", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "TediGAN: Text-Guided Diverse Face Image Generation and Manipulation"], "answer_arxiv_id": ["2005.04909", "2004.11437", "2005.13580", "2103.17249", "2012.03308"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_7705"} +{"question": "Could you provide me some datasets derived from simulators for cooperative perception?", "answer": ["CARLA: An Open Urban Driving Simulator", "OpenCDA:An Open Cooperative Driving Automation Framework Integrated with\n Co-Simulation"], "answer_arxiv_id": ["1711.03938", "2107.06260"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_7706"} +{"question": "What work built up on COMBO by establishing a closed-form expression for the diffusion kernel and proposing kernels tailored for mixed spaces?", "answer": ["Bayesian Optimization over Hybrid Spaces"], "answer_arxiv_id": ["2106.04682"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_7707"} +{"question": "Could you provide me some works that achieve strong empirical performance using fully-quadratic estimator?", "answer": ["Learning Bellman Complete Representations for Offline Policy Evaluation"], "answer_arxiv_id": ["2207.05837"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_7708"} +{"question": "Could you provide some works that exploit invariance across environments to infer causal structure in asymmetry-based methods?", "answer": ["Learning Causal Structures Using Regression Invariance", "Invariant Models for Causal Transfer Learning", "Causal Structure Learning"], "answer_arxiv_id": ["1705.09644", "1507.05333", "1706.09141"], "source_meta": {"published_time": "20220411"}, "qid": "AutoScholarQuery_train_7709"} +{"question": "What research contains text prompts specially composed for T2V generation?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data"], "answer_arxiv_id": ["2209.14792"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_7710"} +{"question": "Which papers are about the 2D reposing algorithms for humans to provide additional pose controllability?", "answer": ["Synthesizing Images of Humans in Unseen Poses", "Soft-Gated Warping-GAN for Pose-Guided Person Image Synthesis", "Fashion Editing with Adversarial Parsing Learning", "Coordinate-based Texture Inpainting for Pose-Guided Image Generation", "Reposing Humans by Warping 3D Features", "A Variational U-Net for Conditional Appearance and Shape Generation", "Unsupervised Person Image Synthesis in Arbitrary Poses", "Towards Purely Unsupervised Disentanglement of Appearance and Shape for\n Person Images Generation", "Learning Realistic Human Reposing using Cyclic Self-Supervision with 3D\n Shape, Pose, and Appearance Consistency", "Disentangled Person Image Generation", "Unsupervised Person Image Synthesis in Arbitrary Poses", "Learning Realistic Human Reposing using Cyclic Self-Supervision with 3D\n Shape, Pose, and Appearance Consistency", "Unsupervised Person Image Generation with Semantic Parsing\n Transformation", "Style and Pose Control for Image Synthesis of Humans from a Single\n Monocular View", "Pose with Style: Detail-Preserving Pose-Guided Image Synthesis with\n Conditional StyleGAN"], "answer_arxiv_id": ["1804.07739", "1810.11610", "1906.00884", "1811.11459", "2006.04898v1", "1804.04694", "1809.10280", "2007.13098", "2110.05458", "1712.02621", "1809.10280", "2110.05458", "1904.03379", "2102.11263", "2109.06166"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_7711"} +{"question": "Can you list some research that formats feedback as numerical scores?", "answer": ["Dialogue Learning with Human Teaching and Feedback in End-to-End\n Trainable Task-Oriented Dialogue Systems", "High Quality Rather than High Model Probability: Minimum Bayes Risk\n Decoding with Neural Metrics"], "answer_arxiv_id": ["1804.06512", "2111.09388"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_7712"} +{"question": "What works highlighted the use of Stable Diffusion for generating images from text prompts?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_7713"} +{"question": "What is the state-of-the-art report on diffusion models the author referred to?", "answer": ["State of the Art on Diffusion Models for Visual Computing"], "answer_arxiv_id": ["2310.07204"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_7714"} +{"question": "Could you provide me some studies that show the quality improvement of text summarization through fine-tuning pre-trained language models?", "answer": ["Text Summarization with Pretrained Encoders", "HIBERT: Document Level Pre-training of Hierarchical Bidirectional\n Transformers for Document Summarization"], "answer_arxiv_id": ["1908.08345", "1905.06566"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_7715"} +{"question": "Which work proposed a keypoint-based pooling mechanism for recognition tasks?", "answer": ["Unified Keypoint-based Action Recognition Framework via Structured\n Keypoint Pooling"], "answer_arxiv_id": ["2303.15270"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_7716"} +{"question": "Which work applied the Gumbel Softmax reparameterization trick to VAE to make gradients tractable?", "answer": ["Categorical Reparameterization with Gumbel-Softmax"], "answer_arxiv_id": ["1611.01144"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_7717"} +{"question": "Can you cite studies that employed the convex Gaussian MinMax theorem in their works?", "answer": ["Theoretical Insights Into Multiclass Classification: A High-dimensional Asymptotic View"], "answer_arxiv_id": ["2011.07729"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_7718"} +{"question": "Are there any studies that evaluate the performance of Neural Language Models (NLMs) in the Blocks World and other similar domains?", "answer": ["Neural Logic Machines"], "answer_arxiv_id": ["1904.11694"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_7719"} +{"question": "What is the study where the HifiGAN recipe was extended by introducing a periodic inductive bias and replacing the MSD with the MRSD using the Snake activation function?", "answer": ["BigVGAN: A Universal Neural Vocoder with Large-Scale Training"], "answer_arxiv_id": ["2206.04658"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_7720"} +{"question": "Could you name the studies which allow for content-preserving edits by overfitting the diffusion model to the input image?", "answer": ["UniTune: Text-Driven Image Editing by Fine Tuning a Diffusion Model on a\n Single Image"], "answer_arxiv_id": ["2210.09477"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_7721"} +{"question": "Which papers discuss advancements that justify the capabilities of LLMs to digest visual inputs?", "answer": ["Visual Instruction Tuning", "The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)"], "answer_arxiv_id": ["2304.08485", "2309.17421"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_7722"} +{"question": "What works utilize late interaction strategies to better leverage the visual and audio features for Audio-Visual Event Localization?", "answer": ["Audio-Visual Event Localization in Unconstrained Videos"], "answer_arxiv_id": ["1803.08842"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_7723"} +{"question": "What papers have proposed novel bellman operators as strategies in robust RL?", "answer": ["Efficient Adversarial Training without Attacking: Worst-Case-Aware Robust Reinforcement Learning", "Online Robust Reinforcement Learning with Model Uncertainty", "Policy Gradient Method For Robust Reinforcement Learning"], "answer_arxiv_id": ["2210.05927v1", "2109.14523", "2205.07344"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_7724"} +{"question": "Which works are associated with generating 3D human motions from text?", "answer": ["TEMOS: Generating diverse human motions from textual descriptions", "Human Motion Diffusion Model", "FLAME: Free-form Language-based Motion Synthesis & Editing", "MotionCLIP: Exposing Human Motion Generation to CLIP Space", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars"], "answer_arxiv_id": ["2204.14109", "2209.14916", "2209.00349", "2203.08063", "2205.08535"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7725"} +{"question": "Any work proposed a computational model to identify which tasks could be learned by a Transformer?", "answer": ["Thinking Like Transformers"], "answer_arxiv_id": ["2106.06981"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_7726"} +{"question": "Which studies propose methods for collaborative perception in multi-agent systems?", "answer": ["When2com: Multi-Agent Perception via Communication Graph Grouping", "Who2com: Collaborative Perception via Learnable Handshake Communication", "V2X-ViT: Vehicle-to-Everything Cooperative Perception with Vision Transformer", "Where2comm: Communication-Efficient Collaborative Perception via Spatial Confidence Maps", "HM-ViT: Hetero-modal Vehicle-to-Vehicle Cooperative perception with vision transformer", "UMC: A Unified Bandwidth-efficient and Multi-resolution based Collaborative Perception Framework", "Spatio-Temporal Domain Awareness for Multi-Agent Collaborative Perception"], "answer_arxiv_id": ["2006.00176", "2003.09575", "2203.10638", "2209.12836", "2304.10628", "2303.12400", "2307.13929"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_7727"} +{"question": "Could you provide me the work about a GAN-based controller that generates actor-driven camera movements taking into account spatial, emotional, and aesthetic factors?", "answer": ["Automatic Camera Trajectory Control with Enhanced Immersion for Virtual\n Cinematography"], "answer_arxiv_id": ["2303.17041"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_7728"} +{"question": "Can you provide me some works that use the technique of comparing the training dataset with a cleaned reference dataset for data quality assurance?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models", "The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset"], "answer_arxiv_id": ["2005.14165", "2302.13971", "2303.03915"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_7729"} +{"question": "What research papers focus on equipping open-source Large Language Models with agent planning abilities?", "answer": ["FireAct: Toward Language Agent Fine-tuning", "AgentTuning: Enabling Generalized Agent Abilities for LLMs"], "answer_arxiv_id": ["2310.05915", "2310.12823v2"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_7730"} +{"question": "What paper describes the concept of training data extraction?", "answer": ["Extracting Training Data from Large Language Models"], "answer_arxiv_id": ["2012.07805"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_7731"} +{"question": "Which papers worked on improving uncertainty estimations through post-hoc calibration on validation sets?", "answer": ["On Calibration of Modern Neural Networks", "Posterior calibration and exploratory analysis for natural language processing models"], "answer_arxiv_id": ["1706.04599", "1508.05154"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_7732"} +{"question": "What works are about coreset selection methods for faster model training?", "answer": ["Coverage-centric Coreset Selection for High Pruning Rates"], "answer_arxiv_id": ["2210.15809"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_7733"} +{"question": "What studies explored the creativity of Language Learning Models (LLMs)?", "answer": ["Unleashing the Creative Mind: Language Model As Hierarchical Policy For\n Improved Exploration on Challenging Problem Solving", "Inspire creativity with ORIBA: Transform Artists' Original Characters\n into Chatbots through Large Language Model"], "answer_arxiv_id": ["2311.00694", "2306.09776"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_7734"} +{"question": "Which studies have approached skill discovery by defining intrinsic rewards based on successors features rather than mutual information?", "answer": ["Discovering a set of policies for the worst case reward"], "answer_arxiv_id": ["2102.04323"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_7735"} +{"question": "Which papers discuss the theoretical properties of a large class of models relation to local variants of relational Weisfeiler-Leman algorithms?", "answer": ["How Powerful are Graph Neural Networks?", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "Weisfeiler and Leman go sparse: Towards scalable higher-order graph embeddings"], "answer_arxiv_id": ["1810.00826", "1810.02244", "1904.01543"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_7736"} +{"question": "Could you provide me some works that proposed other non-similarity-based approaches to program equilibrium?", "answer": ["Robust Cooperation in the Prisoner’s Dilemma: Program Equilibrium via Provability Logic"], "answer_arxiv_id": ["1401.5577"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_7737"} +{"question": "Which research proposed a method for combining pose estimation and NeRFs?", "answer": ["iNeRF: Inverting Neural Radiance Fields for Pose Estimation"], "answer_arxiv_id": ["2012.05877"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_7738"} +{"question": "Which research principles were used in the regularization-based approach for Continual Learning techniques?", "answer": ["Overcoming catastrophic forgetting in neural networks"], "answer_arxiv_id": ["1612.00796"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_7739"} +{"question": "What studies utilize large visual language models (VLMs) in describing differences in small groups of images?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_7740"} +{"question": "What are some studies that use contrastive learning with data augmentation for image-based RL?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Data-Efficient Reinforcement Learning with Self-Predictive Representations", "Decoupling Representation Learning from Reinforcement Learning", "Behavior From the Void: Unsupervised Active Pre-Training", "Learning Visual Robotic Control Efficiently with Contrastive Pre-training and Data Augmentation"], "answer_arxiv_id": ["2004.04136", "2007.05929", "2009.08319", "2103.04551", "2012.07975"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7741"} +{"question": "What paper suggests the autonomous learning capacity of LLMs to formulate plans could be limited?", "answer": ["On the Planning Abilities of Large Language Models : A Critical Investigation"], "answer_arxiv_id": ["2305.15771v2"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_7742"} +{"question": "What was the first MARL algorithm applied to the offline setting?", "answer": ["Believe What You See: Implicit Constraint Approach for Offline Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["2106.03400"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_7743"} +{"question": "What work has derived accelerated rates in the strongly convex case using robust distance estimation techniques?", "answer": ["From low probability to high confidence in stochastic convex optimization"], "answer_arxiv_id": ["1907.13307"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_7744"} +{"question": "Which research works have studied diffusion models for discrete data by applying continuous diffusion models with Gaussian noise?", "answer": ["DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models", "Diffusion-LM Improves Controllable Text Generation", "Continuous diffusion for categorical data"], "answer_arxiv_id": ["2210.08933", "2205.14217", "2211.15089"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_7745"} +{"question": "What research provided an overview of inference-time data protection in images?", "answer": ["Towards a Visual Privacy Advisor: Understanding and Predicting Privacy Risks in Images"], "answer_arxiv_id": ["1703.10660"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_7746"} +{"question": "What research provides the inspiration for PACO dataset?", "answer": ["PACO: Parts and Attributes of Common Objects"], "answer_arxiv_id": ["2301.01795"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_7747"} +{"question": "What works present how SDM can be treated as an instance of Single Positive Multi-Label (SPML) learning?", "answer": ["Multi-Label Learning from Single Positive Labels", "Spatial Consistency Loss for Training Multi-Label Classifiers from Single-Label Annotations", "Acknowledging the Unknown for Multi-label Learning with Single Positive Labels"], "answer_arxiv_id": ["2106.09708", "2203.06127", "2203.16219"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_7748"} +{"question": "What studies used clustered federated learning technique for personalization?", "answer": ["Clustered Federated Learning: Model-Agnostic Distributed Multi-Task Optimization under Privacy Constraints", "Three Approaches for Personalization with Applications to Federated Learning", "An Efficient Framework for Clustered Federated Learning"], "answer_arxiv_id": ["1910.01991", "2002.10619", "2006.04088"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_7749"} +{"question": "Which works used synthetic data for egocentric pose estimation?", "answer": ["UnrealEgo: A New Dataset for Robust Egocentric 3D Human Motion Capture", "SelfPose: 3D Egocentric Pose Estimation from a Headset Mounted Camera"], "answer_arxiv_id": ["2208.01633", "2011.01519"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_7750"} +{"question": "What studies applied dynamics prediction to dynamics planning?", "answer": ["PHYRE: A New Benchmark for Physical Reasoning", "Learning Long-term Visual Dynamics with Region Proposal Interaction Networks", "On the Learning Mechanisms in Physical Reasoning"], "answer_arxiv_id": ["1908.05656", "2008.02265", "2210.02075"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_7751"} +{"question": "What research showed that the accuracy of LLMs improve when using a pure multiple choice question style vs a cloze question answering style?", "answer": ["Leveraging Large Language Models for Multiple Choice Question Answering"], "answer_arxiv_id": ["2210.12353"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_7752"} +{"question": "Do we have any studies that derived PAC-Bayes bounds for adaptative sliced Wasserstein distances?", "answer": ["Shedding a PAC-Bayesian Light on Adaptive Sliced-Wasserstein Distances"], "answer_arxiv_id": ["2206.03230"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_7753"} +{"question": "Could you provide a paper that focuses on generating synthetic data by utilizing various procedural noise models?", "answer": ["Learning to See by Looking at Noise"], "answer_arxiv_id": ["2106.05963"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_7754"} +{"question": "Which works present existing solutions for panoptic segmentation of 3D point clouds?", "answer": ["SceneGraphFusion: Incremental 3D Scene Graph Prediction from RGB-D\n Sequences", "PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff\n and Things"], "answer_arxiv_id": ["2103.14898", "1903.01177"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_7755"} +{"question": "Could you mention some works that utilize query-based methods in multimodal language models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "X-LLM: Bootstrapping Advanced Large Language Models by Treating\n Multi-Modalities as Foreign Languages", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2304.10592", "2305.06500", "2305.04160", "2306.02858", "2304.14178"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_7756"} +{"question": "In which study the contrastive learner used in the experiment is mentioned?", "answer": ["LiT: Zero-Shot Transfer with Locked-image text Tuning"], "answer_arxiv_id": ["2111.07991"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_7757"} +{"question": "Which research papers developed motion planners and controllers for manipulating articulated objects?", "answer": ["Articulated Object Interaction in Unknown Scenes with Whole-Body Mobile Manipulation", "DexArt: Benchmarking Generalizable Dexterous Manipulation with Articulated Objects"], "answer_arxiv_id": ["2103.10534", "2305.05706"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_7758"} +{"question": "What work proposed the parallelized method for FullES, an algorithm used in evolution strategies?", "answer": ["Evolution Strategies as a Scalable Alternative to Reinforcement Learning"], "answer_arxiv_id": ["1703.03864v2"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_7759"} +{"question": "What studies can be directly used for visual imitation?", "answer": ["SQIL: Imitation Learning via Reinforcement Learning with Sparse Rewards", "IQ-Learn: Inverse soft-Q Learning for Imitation"], "answer_arxiv_id": ["1905.11108", "2106.12142"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_7760"} +{"question": "What works focus on using data augmentation as a solution for Domain Generalization (DG)?", "answer": ["Learning to Generate Novel Domains for Domain Generalization", "Domain Generalization with MixStyle", "Improving Generalization in Reinforcement Learning with Mixture Regularization"], "answer_arxiv_id": ["2007.03304", "2104.02008", "2010.10814"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7761"} +{"question": "What studies have used text and image conditional diffusion models to tackle material generation?", "answer": ["ControlMat: A Controlled Generative Approach to Material Capture", "MatFuse: Controllable Material Generation with Diffusion Models"], "answer_arxiv_id": ["2309.01700", "2308.11408"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_7762"} +{"question": "What papers study invariant risk minimization games in FL?", "answer": ["FL Games: A Federated Learning Framework for Distribution Shifts"], "answer_arxiv_id": ["2211.00184"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7763"} +{"question": "Who laid the foundation for the link between gradient descent and mirror descent, which was used for analyzing the effects of initialization scale on obtained solutions in diagonal linear networks?", "answer": ["Exponentiated Gradient Meets Gradient Descent"], "answer_arxiv_id": ["1902.01903v1"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_7764"} +{"question": "Can you list some works that use multiple normalization layers in a parallel manner?", "answer": ["Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net", "Adversarial Examples Improve Image Recognition", "AugMax: Adversarial Composition of Random Augmentations for Robust Training", "Pure Noise to the Rescue of Insufficient Data: Improving Imbalanced Classification by Training on Random Noise Images"], "answer_arxiv_id": ["1807.09441", "1911.09665", "2110.13771", "2112.08810"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_7765"} +{"question": "Any research demonstrating that in-context learning is more effective with diverse and relevant context examples?", "answer": ["Complementary Explanations for Effective In-Context Learning"], "answer_arxiv_id": ["2211.13892"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_7766"} +{"question": "Which works utilize shift mechanism for temporal modeling in streaming data analysis?", "answer": ["TSM: Temporal Shift Module for Efficient Video Understanding"], "answer_arxiv_id": ["1811.08383"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_7767"} +{"question": "Which datasets were used in the studies of egocentric vision?", "answer": ["Rescaling Egocentric Vision", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2006.13256", "2110.07058"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_7768"} +{"question": "What works follow the approach of compressing the length of the input context?", "answer": ["Perceiver: General Perception with Iterative Attention"], "answer_arxiv_id": ["2103.03206"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7769"} +{"question": "Which work introduces the procedural generation framework ProcTHOR, and which work develops a similar concept, Phone2Proc?", "answer": ["ProcTHOR: Large-Scale Embodied AI Using Procedural Generation", "Phone2Proc: Bringing Robust Robots Into Our Chaotic World"], "answer_arxiv_id": ["2206.06994", "2212.04819"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_7770"} +{"question": "Which studies propose treating the offline RL problem as a sequence modeling problem?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Multi-Game Decision Transformers", "A Generalist Agent", "RvS: What is Essential for Offline RL via Supervised Learning?"], "answer_arxiv_id": ["2106.01345", "2106.02039", "2205.15241", "2205.06175", "2112.10751"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_7771"} +{"question": "Which studies proposed lookahead strategies and future environmental representations for Vision-and-Language Navigation?", "answer": ["Improving Vision-and-Language Navigation by Generating Future-View Image\n Semantics", "DREAMWALKER: Mental Planning for Continuous Vision-Language Navigation", "ULN: Towards Underspecified Vision-and-Language Navigation", "Active Visual Information Gathering for Vision-Language Navigation", "Tactical Rewind: Self-Correction via Backtracking in Vision-and-Language\n Navigation"], "answer_arxiv_id": ["2304.04907", "2308.07498", "2210.10020", "2007.08037", "1903.02547"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_7772"} +{"question": "Which studies have called for finer-grain skin tone measurements within dermatology and for fairness in computer vision?", "answer": ["Designing Disaggregated Evaluations of AI Systems: Choices, Considerations, and Tradeoffs"], "answer_arxiv_id": ["2103.06076"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_7773"} +{"question": "What researches have been conducted on specific representation for deepfake detection, such as forgery region location, capsule network, disentanglement learning, and image reconstruction?", "answer": ["Multi-task Learning For Detecting and Segmenting Manipulated Facial Images and Videos", "CAPSULE-FORENSICS: USING CAPSULE NETWORKS TO DETECT FORGED IMAGES AND VIDEOS", "UCF: Uncovering Common Features for Generalizable Deepfake Detection", "Exploring Disentangled Content Information for Face Forgery Detection"], "answer_arxiv_id": ["1906.06876", "1810.11215", "2304.13949", "2207.09202"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_7774"} +{"question": "Are there any studies that demonstrate that MPNNs can execute the complex Ford-Fulkerson algorithm?", "answer": ["Neural Bipartite Matching"], "answer_arxiv_id": ["2005.11304"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_7775"} +{"question": "Which research provided theoretical justification for calibrating the predictive model in label shift adaptation?", "answer": ["A Unified View of Label Shift Estimation"], "answer_arxiv_id": ["2003.07554"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7776"} +{"question": "What works developed latent models in the context of offline RL and OPE?", "answer": ["Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models", "SOLAR: Deep Structured Representations for Model-Based Reinforcement Learning", "Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Model", "Model-Based Reinforcement Learning via Latent-Space Collocation", "MOPO: Model-based Offline Policy Optimization", "COMBO: Conservative Offline Model-Based Policy Optimization", "Offline Reinforcement Learning from Images with Latent Space Models", "Variational Latent Branching Model for Off-Policy Evaluation"], "answer_arxiv_id": ["1811.04551", "1912.01603", "2010.02193", "1808.09105", "1907.00953", "2106.13229", "2005.13239", "2102.08363", "2012.11547", "2301.12056"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_7777"} +{"question": "What research introduced contextual sparsity by sparsifying MLP and attention blocks in LLMs?", "answer": ["DejaVu: Conditional Regenerative Learning to Enhance Dense Prediction"], "answer_arxiv_id": ["2303.01573"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_7778"} +{"question": "What work(s) proposed the Mixture of Normal-Inverse Gamma (MoNIG) algorithm for multimodal regression?", "answer": ["Trustworthy Multimodal Regression with Mixture of Normal-inverse Gamma Distributions"], "answer_arxiv_id": ["2111.08456"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_7779"} +{"question": "Which papers represent original work on diffusion models in generative applications?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "1907.05600", "2006.11239"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_7780"} +{"question": "Are there any research work discussing real-world datasets with well-aligned CAD annotations?", "answer": ["Aria Digital Twin: A New Benchmark Dataset for Egocentric 3D Machine\n Perception", "SCoDA: Domain Adaptive Shape Completion for Real Scans"], "answer_arxiv_id": ["2306.06362", "2304.10179"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_7781"} +{"question": "What studies focus on quantization, which provides inference but introduces challenges for training?", "answer": ["Training Quantized Nets: A Deeper Understanding"], "answer_arxiv_id": ["1706.02379"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_7782"} +{"question": "Which studies explored the transferability of neural networks to unknown tasks?", "answer": ["A Survey on Deep Transfer Learning", "A Comprehensive Survey on Transfer Learning"], "answer_arxiv_id": ["1808.01974", "1911.02685"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7783"} +{"question": "What benchmarks have been used to demonstrate the versatility of MLLMs in visual perception and comprehension?", "answer": ["SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension", "LVLM-eHub: A Comprehensive Evaluation Benchmark for Large\n Vision-Language Models", "MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language\n Models"], "answer_arxiv_id": ["2307.16125", "2306.09265", "2306.13394"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_7784"} +{"question": "Which papers proposed self-supervised learning on nonhomophilous graphs?", "answer": ["Decoupled Self-supervised Learning for Non-Homophilous Graphs"], "answer_arxiv_id": ["2206.03601v3"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_7785"} +{"question": "Could you provide me a reference that discusses Variational Autoencoders (VAEs)?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7786"} +{"question": "Who has conducted research on the development of robust evaluation frameworks and benchmarks for LLMs?", "answer": ["Who is ChatGPT? Benchmarking LLMs' Psychological Portrayal Using\n PsychoBench", "ConceptPsy:A Benchmark Suite with Conceptual Comprehensiveness in\n Psychology", "PsyCoT: Psychological Questionnaire as Powerful Chain-of-Thought for\n Personality Detection"], "answer_arxiv_id": ["2310.01386", "2311.09861", "2310.20256"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_7787"} +{"question": "Any works that focused on bridging frame-based and event-based vision?", "answer": ["Events-to-Video: Bringing Modern Computer Vision to Event Cameras", "High Speed and High Dynamic Range Video with an Event Camera", "Reducing the Sim-to-Real Gap for Event Cameras", "The Event-Camera Dataset and Simulator: Event-based Data for Pose\n Estimation, Visual Odometry, and SLAM", "v2e: From Video Frames to Realistic DVS Events", "EventGAN: Leveraging Large Scale Image Datasets for Event Cameras", "EvDistill: Asynchronous Events to End-task Learning via Bidirectional\n Reconstruction-guided Cross-modal Knowledge Distillation", "ESS: Learning Event-based Semantic Segmentation from Still Images", "Bridging the Gap between Events and Frames through Unsupervised Domain\n Adaptation", "Learning to Exploit Multiple Vision Modalities by Using Grafted Networks"], "answer_arxiv_id": ["1904.08298", "1906.07165", "2003.09078", "1610.08336", "2006.07722", "1912.01584", "2111.12341", "2203.10016", "2109.02618", "2003.10959"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_7788"} +{"question": "What works have been done in the field of self-supervised pretraining approaches in computer vision?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.14294"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_7789"} +{"question": "In what papers did researchers discuss the capacity loss of agents in new prediction problems?", "answer": ["Understanding and Preventing Capacity Loss in Reinforcement Learning"], "answer_arxiv_id": ["2204.09560"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_7790"} +{"question": "What are the notable studies related to object-object affordances?", "answer": ["O2O-Afford: Annotation-Free Large-Scale Object-Object Affordance\n Learning"], "answer_arxiv_id": ["2106.15087"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_7791"} +{"question": "What prior studies observed performance improvement of scaling block output instead of skip connections in ResNet and Transformer?", "answer": ["ReZero is All You Need: Fast Convergence at Large Depth", "Fixup Initialization: Residual Learning Without Normalization", "Batch Normalization Biases Residual Blocks Towards the Identity Function in Deep Networks", "On Layer Normalization in the Transformer Architecture", "Deep Transformers without Shortcuts: Modifying Self-attention for Faithful Signal Propagation", "Stable ResNet", "How to Start Training: The Effect of Initialization and Architecture"], "answer_arxiv_id": ["2003.04887", "1901.09321", "2002.10444", "2002.04745", "2302.10322", "2010.12859v2", "1803.01719"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_7792"} +{"question": "Who identified the issue of paths connecting disconnected regions yielding high-confidence predictions in a narrow network?", "answer": ["Neural Networks Should Be Wide Enough to Learn Disconnected Decision Regions"], "answer_arxiv_id": ["1803.00094"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7793"} +{"question": "What research studied the sample complexity of classification in the domain of label differential privacy (DP)?", "answer": ["Private Learning and Sanitization: Pure vs. Approximate Differential Privacy"], "answer_arxiv_id": ["1407.2674v1"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_7794"} +{"question": "What work was done using knowledge distillation to generate datasets of similar quality to crowd-sourced data?", "answer": ["Symbolic Knowledge Distillation: from General Language Models to\n Commonsense Models"], "answer_arxiv_id": ["2110.07178"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_7795"} +{"question": "Is there any work that does the scoring differently, e.g., adopts numerical score reward for training while pivoting on preference ranking data collection?", "answer": ["Aligning Large Multimodal Models with Factually Augmented RLHF"], "answer_arxiv_id": ["2309.14525"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_7796"} +{"question": "Which work has related analysis that captures the interplay between the metric and the sampling distribution in the context of k-nearest neighbour risk decay rates?", "answer": ["Rates of Convergence for Nearest Neighbor Classification"], "answer_arxiv_id": ["1407.0067"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_7797"} +{"question": "Could you provide me some works that utilized human perception annotations in evaluating machine learning models?", "answer": ["RoBERTa: A Robustly Optimized BERT Pretraining Approach", "ImageNet Large Scale Visual Recognition Challenge"], "answer_arxiv_id": ["1907.11692", "1409.0575"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_7798"} +{"question": "Which works have integrated 2D Vision-Language Models with 3D point cloud processing marking recent progress in open vocabulary scene understanding?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP", "PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning", "CLIP2Point: Transfer CLIP to Point Cloud Classification with Image-Depth\n Pre-training", "ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding"], "answer_arxiv_id": ["2112.02413", "2211.11682", "2210.01055", "2305.08275"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_7799"} +{"question": "Can you list any research that involved custom decoders or inference algorithms for syntactic parsing?", "answer": ["Constituency Parsing with a Self-Attentive Encoder", "Bottom-Up Constituency Parsing and Nested Named Entity Recognition with Pointer Networks", "In-Order Transition-based Constituent Parsing"], "answer_arxiv_id": ["1805.01052", "2110.05419", "1707.05000"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_7800"} +{"question": "Could you provide me some studies about reducing the quadratic time complexity for computing attention through efficient attention?", "answer": ["Reformer: The Efficient Transformer", "Rethinking Attention with Performers"], "answer_arxiv_id": ["2001.04451", "2009.14794"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_7801"} +{"question": "Which work addresses the sample complexity and regret bounds of LMDPs?", "answer": ["RL for Latent MDPs: Regret Guarantees and a Lower Bound"], "answer_arxiv_id": ["2102.04939"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_7802"} +{"question": "Which papers propose data-driven 3D pose transfer methods based on a parametric human model?", "answer": ["The Power of Points for Modeling Humans in Clothing", "MetaAvatar: Learning Animatable Clothed Human Models from Few Depth\n Images", "Unsupervised Shape and Pose Disentanglement for 3D Meshes"], "answer_arxiv_id": ["2109.01137", "2106.11944", "2007.11341"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_7803"} +{"question": "Which research considers equivariant learning of stochastic fields?", "answer": ["Equivariant Learning of Stochastic Fields: Gaussian Processes and Steerable Conditional Neural Processes"], "answer_arxiv_id": ["2011.12916"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7804"} +{"question": "Which papers explored the connection of GFlowNets with the variational inference literature and other generative models?", "answer": ["GFlowNets and variational inference", "A Variational Perspective on Generative Flow Networks", "Unifying Generative Models with GFlowNets and Beyond"], "answer_arxiv_id": ["2210.00580", "2210.07992", "2209.02606"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7805"} +{"question": "Any works about methods that consider using a part of the test input image as a reference?", "answer": ["Learning the Degradation Distribution for Blind Image Super-Resolution"], "answer_arxiv_id": ["2203.04962"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7806"} +{"question": "What is the initial paper that established the Gaussian Splatting approach used in this research?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_7807"} +{"question": "Which research papers have studied the augmentation of the loss function in the lifted neural network framework?", "answer": ["Lifted Neural Networks"], "answer_arxiv_id": ["1805.01532"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_7808"} +{"question": "What research was done regarding global-level methods for image-text matching?", "answer": ["VSE++: Improving Visual-Semantic Embeddings with Hard Negatives", "Visual Semantic Reasoning for Image-Text Matching", "Learning the Best Pooling Strategy for Visual Semantic Embedding"], "answer_arxiv_id": ["1707.05612", "1909.02701", "2011.04305"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_7809"} +{"question": "Any works about combining multi-rate input sampling and hierarchical interpolation with MLPs for univariate forecasting?", "answer": ["N-HiTS: Neural Hierarchical Interpolation for Time Series Forecasting"], "answer_arxiv_id": ["2201.12886"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_7810"} +{"question": "What papers contain information on the efficacy of integrating structural and attribute information in attribute graph clustering?", "answer": ["Structural Deep Clustering Network"], "answer_arxiv_id": ["2002.01633"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_7811"} +{"question": "Could you provide me some studies about group fairness constraint application on spectral clustering?", "answer": ["Guarantees for Spectral Clustering with Fairness Constraints", "Scalable Spectral Clustering with Group Fairness Constraints"], "answer_arxiv_id": ["1901.08668", "2210.16435v3"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_7812"} +{"question": "Which works describe Human Label Variation (HLV) in Natural Language Processing (NLP)?", "answer": ["The 'Problem' of Human Label Variation: On Ground Truth in Data,\n Modeling and Evaluation"], "answer_arxiv_id": ["2211.02570"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_7813"} +{"question": "Could you provide me some research works that use human feedback in the form of pairwise comparisons or rankings to learn preference reward functions?", "answer": ["Asking Easy Questions: A User-Friendly Approach to Active Reward Learning", "Learning Multimodal Rewards from Rankings", "Learning Reward Functions from Diverse Sources of Human Feedback: Optimally Integrating Demonstrations and Preferences", "Extrapolating Beyond Suboptimal Demonstrations via Inverse Reinforcement Learning from Observations", "PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-training", "Few-Shot Preference Learning for Human-in-the-Loop RL"], "answer_arxiv_id": ["1910.04365", "2109.12750", "2006.14091", "1904.06387", "2106.05091", "2212.03363"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_7814"} +{"question": "Could you provide me some works that modify neuron outputs for directly editing model behavior?", "answer": ["Knowledge Neurons in Pretrained Transformers", "Kformer: Knowledge Injection in Transformer Feed-Forward Layers"], "answer_arxiv_id": ["2104.08696", "2201.05742"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_7815"} +{"question": "What works have formulated open-vocabulary object detection as image-text matching?", "answer": ["RegionCLIP: Region-based Language-Image Pretraining", "Detecting Twenty-thousand Classes using Image-level Supervision", "F-VLM: Open-Vocabulary Object Detection upon Frozen Vision and Language\n Models", "Learning to Prompt for Open-Vocabulary Object Detection with\n Vision-Language Model", "Aligning Bag of Regions for Open-Vocabulary Object Detection"], "answer_arxiv_id": ["2112.09106", "2201.02605", "2209.15639", "2203.14940", "2302.13996"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_7816"} +{"question": "What is the paper claiming to have broken the defense of defensive distillation?", "answer": ["Defensive Distillation is Not Robust to Adversarial Examples"], "answer_arxiv_id": ["1607.04311"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_7817"} +{"question": "What studies utilize counterfactual approaches in designing post hoc explanations?", "answer": ["Explanation by Progressive Exaggeration"], "answer_arxiv_id": ["1911.00483"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_7818"} +{"question": "What works have explored constrained fine-tuning or meta-learning in updating LLM parameters?", "answer": ["Recall and Learn: Fine-tuning Deep Pretrained Language Models with Less\n Forgetting", "Plug-and-Play Adaptation for Continuously-updated QA", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2004.12651", "2204.12785", "2110.11309"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_7819"} +{"question": "What papers propose to make the forward diffusion nonlinear and trainable?", "answer": ["Diffusion Normalizing Flow", "Solving Schrödinger Bridges via Maximum Likelihood", "Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling", "Deep Generative Learning via Schrödinger Bridge", "Likelihood Training of Schrödinger Bridge using Forward-Backward SDEs Theory"], "answer_arxiv_id": ["2110.07579v1", "2106.02081", "2106.01357", "2106.10410", "2110.11291"], "source_meta": {"published_time": "20220429"}, "qid": "AutoScholarQuery_train_7820"} +{"question": "What are the existing studies on few-shot learning?", "answer": ["Generalizing from a Few Examples: A Survey on Few-Shot Learning", "Learning to Compare: Relation Network for Few-Shot Learning", "Prototypical Networks for Few-shot Learning", "Few-Shot Learning with Graph Neural Networks", "NÜWA: Visual Synthesis Pre-training for Neural visUal World creAtion", "Few-Shot Bayesian Optimization with Deep Kernel Surrogates", "AutoTransfer: AutoML with Knowledge Transfer - An Application to Graph Neural Networks"], "answer_arxiv_id": ["1904.05046", "1711.06025", "1703.05175", "1711.04043", "2111.12417", "2101.07667", "2303.07669"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7821"} +{"question": "Which work discussed the complexity and performance characterization issues of modern neural networks?", "answer": ["The Principles of Deep Learning Theory"], "answer_arxiv_id": ["2106.10165"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_7822"} +{"question": "Can you name recent papers that built a general theory of online learning in the statistical learning setting?", "answer": ["Online Learning with Predictable Sequences", "Majorizing Measures, Sequential Complexities, and Online Learning"], "answer_arxiv_id": ["1208.3728", "2102.01729"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_7823"} +{"question": "Could you provide me some works that confirmed the presence of extreme outliers at scale?", "answer": ["GLM-130B: An Open Bilingual Pre-trained Model", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models"], "answer_arxiv_id": ["2210.02414", "2211.10438"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7824"} +{"question": "Can you list the papers discussing conservative value function updates in model-free offline RL?", "answer": ["Adversarially Trained Actor Critic for Offline Reinforcement Learning", "Offline Reinforcement Learning with Fisher Divergence Critic Regularization", "Conservative Q-Learning for Offline Reinforcement Learning", "Bellman-consistent Pessimism for Offline Reinforcement Learning"], "answer_arxiv_id": ["2202.02446", "2103.08050", "2006.04779", "2106.06926"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_7825"} +{"question": "Which works introduced the CLIP guidance in text-to-image synthesis?", "answer": ["VQGAN-CLIP: Open Domain Image Generation and Editing with Natural Language Guidance", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "Towards Counterfactual Image Manipulation via CLIP"], "answer_arxiv_id": ["2204.08583", "2103.17249", "2207.02812"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7826"} +{"question": "What paper first proposed TimeGrad for multivariate time series forecasting?", "answer": ["Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting"], "answer_arxiv_id": ["2101.12072"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_7827"} +{"question": "Any works about the use of dropout for information bottleneck?", "answer": ["Variational Dropout and the Local Reparameterization Trick"], "answer_arxiv_id": ["1506.02557"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_7828"} +{"question": "Can you provide some works about neural network-based 3D surface reconstruction?", "answer": ["Volume Rendering of Neural Implicit Surfaces", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse\n Views", "HF-NeuS: Improved Surface Reconstruction Using High-Frequency Details", "Geo-Neus: Geometry-Consistent Neural Implicit Surfaces Learning for\n Multi-view Reconstruction", "Improving neural implicit surfaces geometry with patch warping"], "answer_arxiv_id": ["2106.12052", "2106.10689", "2206.05737", "2206.07850", "2205.15848", "2112.09648"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_7829"} +{"question": "Could you provide me some works about retrieval-augmented language models and their alignment with a language model?", "answer": ["RA-DIT: Retrieval-Augmented Dual Instruction Tuning", "REPLUG: Retrieval-Augmented Black-Box Language Models"], "answer_arxiv_id": ["2310.01352", "2301.12652"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_7830"} +{"question": "What works focus on improving the quality of text-driven video editing with pre-trained image diffusion models?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Video-P2P: Video Editing with Cross-attention Control", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models"], "answer_arxiv_id": ["2212.11565", "2303.04761", "2208.01626", "2211.09794"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_7831"} +{"question": "What papers utilized manual masks to constrain the editing regions in image editing?", "answer": ["SceneComposer: Any-Level Semantic Image Synthesis"], "answer_arxiv_id": ["2211.11742"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7832"} +{"question": "Which papers provide formal guarantees on system identification in different classes of nonlinear systems, but only consider noiseless systems or significantly easier to excite systems?", "answer": ["Learning nonlinear dynamical systems from a single trajectory", "Stochastic Gradient Descent Learns State Equations with Nonlinear Activations", "Non-asymptotic and Accurate Learning of Nonlinear Dynamical Systems"], "answer_arxiv_id": ["2004.14681", "1809.03019v1", "2002.08538"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7833"} +{"question": "Which paper illustrates the cases demonstrating the non-closedness and best approximation property of function spaces of neural networks?", "answer": ["Topological properties of the set of functions generated by neural networks of fixed size", "Nonclosedness of Sets of Neural Networks in Sobolev Spaces"], "answer_arxiv_id": ["1806.08459", "2007.11730"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_7834"} +{"question": "What previous works propose tailored spatio-temporal attention as an extension to the T2I model in text-guided video editing?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation"], "answer_arxiv_id": ["2212.11565"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_7835"} +{"question": "What are the papers that studied the relation between the bilevel problem and its penalty reformulation?", "answer": ["BOME! Bilevel Optimization Made Easy: A Simple First-Order Approach"], "answer_arxiv_id": ["2209.08709"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_7836"} +{"question": "What examples of research are there on the lifting of knowledge from 2D images to 3D representations for 3D scene synthesis?", "answer": ["LDM3D: Latent Diffusion Model for 3D", "Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models"], "answer_arxiv_id": ["2305.10853", "2303.11989"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_7837"} +{"question": "What works does personalization by optimizing low-rank approximations of weight residuals?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_7838"} +{"question": "Which papers proposed enforcing equivariance constraints and auxiliary losses for Unsupervised Landmark Detection (ULD)?", "answer": ["Unsupervised Learning of Landmarks by Descriptor Vector Exchange", "Unsupervised learning of object frames by dense equivariant image\n labelling", "Unsupervised learning of object landmarks by factorized spatial\n embeddings", "On Equivariant and Invariant Learning of Object Landmark Representations"], "answer_arxiv_id": ["1908.06427", "1706.02932", "1705.02193", "2006.14787"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_7839"} +{"question": "What works proposed learning a conditional sequence model where the control code encodes a discrete or scalar value specifying the target attribute?", "answer": ["CTRL: A Conditional Transformer Language Model for Controllable Generation", "ProGen: Language Modeling for Protein Generation"], "answer_arxiv_id": ["1909.05858", "2004.03497"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_7840"} +{"question": "Which studies have contributed towards strengthening deep gradient leakage attacks?", "answer": ["iDLG: Improved Deep Leakage from Gradients"], "answer_arxiv_id": ["2001.02610"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_7841"} +{"question": "What are the studies that showed convergence of adversarial training error?", "answer": ["Over-parameterized Adversarial Training: An Analysis Overcoming the Curse of Dimensionality"], "answer_arxiv_id": ["2002.06668"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_7842"} +{"question": "Which work specifically studied universal dynamic regret without a bounded domain?", "answer": ["Parameter-free Mirror Descent"], "answer_arxiv_id": ["2203.00444"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_7843"} +{"question": "Which work approximated the gradient statistics with the outer product of 'row' and 'column'?", "answer": ["Adafactor: Adaptive Learning Rates with Sublinear Memory Cost"], "answer_arxiv_id": ["1804.04235"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_7844"} +{"question": "What papers employed neural architecture search in the field of vision?", "answer": ["Neural Architecture Search with Reinforcement Learning", "DARTS: Differentiable Architecture Search", "Neural Architecture Optimization"], "answer_arxiv_id": ["1611.01578", "1806.09055", "1808.07233"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_7845"} +{"question": "Which studies model preconditioners with neural networks?", "answer": ["Deep Learning of Preconditioners for Conjugate Gradient Solvers in Urban Water Related Problems", "Machine-Learned Preconditioners for Linear Solvers in Geophysical Fluid Flows", "Neural-network preconditioners for solving the Dirac equation in lattice gauge theory"], "answer_arxiv_id": ["1906.06925", "2010.02866", "2208.02728"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7846"} +{"question": "Is there any research on the training stability of RLFT methods in text-to-image diffusion models?", "answer": ["Secrets of RLHF in Large Language Models Part I: PPO", "Open Problems and Fundamental Limitations of Reinforcement Learning from\n Human Feedback"], "answer_arxiv_id": ["2307.04964", "2307.15217"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_7847"} +{"question": "Can you list some research papers that propose the detector-free approach for feature matching?", "answer": ["LoFTR: Detector-Free Local Feature Matching with Transformers", "A case for using rotation invariant features in state of the art feature\n matchers", "QuadTree Attention for Vision Transformers", "ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformer"], "answer_arxiv_id": ["2104.00680", "2204.10144", "2201.02767", "2208.14201"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_7848"} +{"question": "What works suggested that BERT embeddings could encode grammatical and morphological information?", "answer": ["Grammatical information in BERT sentence embeddings as two-dimensional\n arrays"], "answer_arxiv_id": ["2312.09890"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_7849"} +{"question": "What are some works centered on accelerating the training of neural fields by using improved encoding schemes?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Fourier Features Let Networks Learn High Frequency Functions in Low\n Dimensional Domains", "Implicit Neural Representations with Periodic Activation Functions", "Neural Sparse Voxel Fields", "Plenoxels: Radiance Fields without Neural Networks", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "ACORN: Adaptive Coordinate Networks for Neural Scene Representation", "Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D\n Shapes", "PlenOctrees for Real-time Rendering of Neural Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2006.10739", "2006.09661", "2007.11571", "2112.05131", "2201.05989", "2105.02788", "2101.10994", "2103.14024"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_7850"} +{"question": "What is an example of a study that collected an artifact regions dataset for image synthesis tasks?", "answer": ["Perceptual Artifacts Localization for Image Synthesis Tasks"], "answer_arxiv_id": ["2310.05590"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_7851"} +{"question": "What are the works that utilized knowledge distillation for efficient and accurate inference in tracking?", "answer": ["Distilled Siamese Networks for Visual Tracking", "Unsupervised Cross-Modal Distillation for Thermal Infrared Tracking", "Real-Time Correlation Tracking via Joint Model Compression and Transfer", "Distilling Channels for Efficient Deep Tracking"], "answer_arxiv_id": ["1907.10586", "2108.00187", "1907.09831", "2409.11785v1"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_7852"} +{"question": "What are some of the works that proposed parameter-efficient fine-tuning methods?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Parameter-Efficient Transfer Learning for NLP", "AdapterFusion: Non-Destructive Task Composition for Transfer Learning", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2101.00190", "1902.00751", "2005.00247", "2106.09685"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_7853"} +{"question": "Could you provide me some researches that use a noninvertible mapping in manifold learning?", "answer": ["Flows for simultaneous manifold learning and density estimation", "Rectangular Flows for Manifold Learning"], "answer_arxiv_id": ["2003.13913", "2106.01413"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_7854"} +{"question": "Which research was shown to achieve minimax-optimal sample complexity in reward-free exploration?", "answer": ["Minimax-Optimal Reward-Agnostic Exploration in Reinforcement Learning"], "answer_arxiv_id": ["2304.07278"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_7855"} +{"question": "Any works that introduce an online clustering algorithm for the representation of dictionary keys?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments"], "answer_arxiv_id": ["2006.09882"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_7856"} +{"question": "What is the fundamental building block of the SSSD model?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces"], "answer_arxiv_id": ["2111.00396"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_7857"} +{"question": "What papers present crowd-sourcing methods to generate visual content for dataset building?", "answer": ["Hollywood in Homes: Crowdsourcing Data Collection for Activity Understanding", "Scaling Egocentric Vision: The EPIC-KITCHENS Dataset", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["1604.01753", "1804.02748", "2110.07058"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_7858"} +{"question": "Which study adapted the findings related to sample complexity of off-policy evaluation to an infinite-horizon setting?", "answer": ["A Variant of the Wang-Foster-Kakade Lower Bound for the Discounted Setting"], "answer_arxiv_id": ["2011.01075"], "source_meta": {"published_time": "20230725"}, "qid": "AutoScholarQuery_train_7859"} +{"question": "What papers developed methods for AI-generated text detection treating the problem as a binary classification task?", "answer": ["Automatic Detection of Machine Generated Text: A Critical Survey", "DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability\n Curvature", "Real or Fake? Learning to Discriminate Machine from Human Generated Text"], "answer_arxiv_id": ["2011.01314", "2301.11305", "1906.03351"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_7860"} +{"question": "Could you provide works that have used feature statistics fitting of the target dataset for unsupervised accuracy estimation?", "answer": ["What can we Learn by Predicting Accuracy?"], "answer_arxiv_id": ["2208.01358"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_7861"} +{"question": "What research papers introduce the use of the signed distance function into neural rendering?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces"], "answer_arxiv_id": ["2106.10689", "2106.12052"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_7862"} +{"question": "Which studies have achieved impressive synthesis quality across different domains with cascaded diffusion models?", "answer": ["Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2106.15282"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_7863"} +{"question": "Which works incorporated attention mechanisms and transformer architectures into tabular neural architectures?", "answer": ["SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training", "Revisiting Deep Learning Models for Tabular Data", "TabNet: Attentive Interpretable Tabular Learning", "TabTransformer: Tabular Data Modeling Using Contextual Embeddings", "AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks", "Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning"], "answer_arxiv_id": ["2106.01342", "2106.11959", "1908.07442", "2012.06678", "1810.11921", "2106.02584v2"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_7864"} +{"question": "Which papers discussed about regularizing the learning policy to stay close to the behavior policy in offline RL algorithms setup?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Behavior Regularized Offline Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning", "Keep Doing What Worked: Behavior Modelling Priors for Offline Reinforcement Learning", "Critic Regularized Regression", "Boosting Offline Reinforcement Learning via Data Rebalancing"], "answer_arxiv_id": ["1812.02900", "1911.11361", "2106.06860", "2002.08396", "2006.15134", "2210.09241"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_7865"} +{"question": "Could you provide me some studies that discussed eliminating the necessity of explicit negative samples in self-supervised learning?", "answer": ["Deep Clustering for Unsupervised Learning of Visual Features", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Emerging Properties in Self-Supervised Vision Transformers", "Self-labelling via simultaneous clustering and representation learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Emerging Properties in Self-Supervised Vision Transformers", "Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning"], "answer_arxiv_id": ["1807.05520", "2006.09882", "2104.14294", "1911.05371", "2006.07733", "2104.14294", "1703.01780", "2006.07733", "2011.10566", "2103.03230", "2105.04906"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_7866"} +{"question": "Are there any studies about auditory or video LLMs?", "answer": ["SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal\n Conversational Abilities", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding"], "answer_arxiv_id": ["2305.11000", "2306.02858"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_7867"} +{"question": "Any works regarding adapting VLLMs for embodied applications?", "answer": ["PaLM-E: An Embodied Multimodal Language Model", "EmbodiedGPT: Vision-Language Pre-Training via Embodied Chain of Thought"], "answer_arxiv_id": ["2303.03378", "2305.15021"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_7868"} +{"question": "What works have used Graph Neural Networks (GNNs) to model physical entities to predict the dynamics of various systems?", "answer": ["Interaction Networks for Learning about Objects, Relations and Physics", "Semi-Supervised Classification with Graph Convolutional Networks", "Inductive Representation Learning on Large Graphs", "Relational inductive biases, deep learning, and graph networks"], "answer_arxiv_id": ["1612.00222", "1609.02907", "1706.02216", "1806.01261"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_7869"} +{"question": "Which previous study showcases the use of unified task representations combined with task-specific queries in multi-task learning?", "answer": ["OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework", "Perceiver IO: A General Architecture for Structured Inputs & Outputs"], "answer_arxiv_id": ["2202.03052", "2107.14795"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_7870"} +{"question": "What studies conducted research on visual correspondences, related to structure-from-motion or 3D reconstruction?", "answer": ["A Survey of Structure from Motion"], "answer_arxiv_id": ["1701.08493"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7871"} +{"question": "Could you provide me with some literature about the need for an additional inversion step for editing existing images in diffusion models?", "answer": ["Denoising Diffusion Implicit Models", "Null-text Inversion for Editing Real Images using Guided Diffusion Models"], "answer_arxiv_id": ["2010.02502", "2211.09794"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_7872"} +{"question": "Which previous works explored the use of expert priors and local search for hyperparameter optimization?", "answer": ["Frugal Optimization for Cost-related Hyperparameters"], "answer_arxiv_id": ["2005.01571"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_7873"} +{"question": "What papers discuss the recent progress in neural fields?", "answer": ["Neural Fields in Visual Computing and Beyond", "Implicit Neural Representations with Periodic Activation Functions", "Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains"], "answer_arxiv_id": ["2111.11426", "2006.09661", "2006.10739"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_7874"} +{"question": "Which papers adopted diffusion models for motion generation following the success of denoising diffusion models in vision generation tasks?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Imagen Video: High Definition Video Generation with Diffusion Models", "Human Motion Diffusion Model", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "FLAME: Free-form Language-based Motion Synthesis & Editing"], "answer_arxiv_id": ["2112.10741", "2205.11487", "2209.14792", "2210.02303", "2209.14916", "2208.15001", "2209.00349"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_7875"} +{"question": "Which works have demonstrated the generalization capabilities of LLMs?", "answer": ["Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2203.15556"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7876"} +{"question": "What studies have discussed the use of pre-trained backbones in GANs?", "answer": ["The Unreasonable Effectiveness of Deep Features as a Perceptual Metric", "Projected GANs Converge Faster", "Enhancing Photorealism Enhancement", "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", "Ensembling Off-the-shelf Models for GAN Training"], "answer_arxiv_id": ["1801.03924", "2111.01007", "2105.04619", "1905.11946", "2112.09130"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_7877"} +{"question": "What works provide insights on robust UCB algorithms for heavy-tailed distributions?", "answer": ["Bandits with heavy tail", "No-Regret Reinforcement Learning with Heavy-Tailed Rewards"], "answer_arxiv_id": ["1209.1727", "2102.12769v1"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7878"} +{"question": "Could you provide me with some research papers about the adaptability of diffusion models in managing various forms of controls and conditions?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models", "GLIGEN: Open-Set Grounded Text-to-Image Generation", "Compositional Visual Generation with Composable Diffusion Models", "Composer: Creative and Controllable Image Synthesis with Composable\n Conditions", "DiffGuard: Semantic Mismatch-Guided Out-of-Distribution Detection using\n Pre-trained Diffusion Models", "Follow Your Pose: Pose-Guided Text-to-Video Generation using Pose-Free\n Videos"], "answer_arxiv_id": ["2112.10752", "2302.05543", "2301.07093", "2206.01714", "2302.09778", "2308.07687", "2304.01186"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_7879"} +{"question": "Which papers exist that utilize intermediate representation in the field of natural language processing", "answer": ["UM4: Unified Multilingual Multiple Teacher-Student Model for\n Zero-Resource Neural Machine Translation", "m3P: Towards Multimodal Multilingual Translation with Multimodal Prompt", "Low-Resource Response Generation with Template Prior", "Machine-Created Universal Language for Cross-lingual Transfer"], "answer_arxiv_id": ["2207.04900", "2403.17556", "1909.11968", "2305.13071"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_7880"} +{"question": "Which works discussed the method of Wasserstein Autoencoder (WAE) in relation to minimizing the primal form of the Wasserstein metric?", "answer": ["Wasserstein Auto-Encoders"], "answer_arxiv_id": ["1711.01558"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_7881"} +{"question": "Which works analyze the training dynamics of single-head attention in Transformers?", "answer": ["Approximating How Single Head Attention Learns"], "answer_arxiv_id": ["2103.07601"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7882"} +{"question": "Could you provide me some works on certifiably robust black-box ensembles?", "answer": ["Scaling provable adversarial defenses", "Boosting Barely Robust Learners: A New Perspective on Adversarial Robustness"], "answer_arxiv_id": ["1805.12514", "2202.05920"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_7883"} +{"question": "Which studies have introduced methods of training classifiers in the latent space of neural networks?", "answer": ["Distilling Model Failures as Directions in Latent Space", "Explaining Deep Convolutional Neural Networks via Latent Visual-Semantic Filter Attention"], "answer_arxiv_id": ["2206.14754", "2204.04601"], "source_meta": {"published_time": "20230408"}, "qid": "AutoScholarQuery_train_7884"} +{"question": "Are there any works focusing on over-fitting prevention to resist noise in training data?", "answer": ["Dimensionality-Driven Learning with Noisy Labels", "Early-Learning Regularization Prevents Memorization of Noisy Labels", "Augmentation Strategies for Learning with Noisy Labels", "Contrast to Divide: Self-Supervised Pre-Training for Learning with Noisy\n Labels", "Dynamic Loss For Robust Learning", "Compressing Features for Learning with Noisy Labels"], "answer_arxiv_id": ["1806.02612", "2007.00151", "2103.02130", "2103.13646", "2211.12506", "2206.13140"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_7885"} +{"question": "What researches proposed algorithms to extract more semantically compact representations or to extend contrastive methods to multi-modality data structures?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Contrastive Clustering", "Task-Customized Self-Supervised Pre-training with Scalable Dynamic Routing", "Learning Transferable Visual Models From Natural Language Supervision", "FILIP: Fine-grained Interactive Language-Image Pre-Training"], "answer_arxiv_id": ["2006.09882", "2009.09687", "2205.13267", "2103.00020", "2111.07783"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_7886"} +{"question": "Could you provide references for research that developed sign language datasets from online video sharing platforms?", "answer": ["Open-Domain Sign Language Translation Learned from Online Video"], "answer_arxiv_id": ["2205.12870"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_7887"} +{"question": "Which papers have looked into neural architecture search (NAS) in the realm of AutoML?", "answer": ["Efficient Multi-objective Neural Architecture Search via Lamarckian Evolution"], "answer_arxiv_id": ["1804.09081"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_7888"} +{"question": "What papers utilize Mask R-CNN as an object detector in proposal-based methods for captioning specific objects in images?", "answer": ["Mask R-CNN"], "answer_arxiv_id": ["1703.06870"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_7889"} +{"question": "Could you list some works related to inverse models predicting sequences of actions?", "answer": ["Planning from Pixels using Inverse Dynamics Models"], "answer_arxiv_id": ["2012.02419"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_7890"} +{"question": "Which papers in SSDA use pseudo-labeling for label information propagation?", "answer": ["Semi-supervised Domain Adaptation via Sample-to-Sample Self-Distillation"], "answer_arxiv_id": ["2111.14353"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_7891"} +{"question": "What studies observed that accuracy-on-the-line may not hold in some circumstances like corruption shifts or in-the-wild shifts?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations", "Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization"], "answer_arxiv_id": ["1903.12261", "2107.04649"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_7892"} +{"question": "What works use a hybrid approach that combines a learned feature grid with a much smaller MLP than the original NeRF?", "answer": ["TensoRF: Tensorial Radiance Fields", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2203.09517", "2301.10241", "2201.05989"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_7893"} +{"question": "Which works are considered pioneers in cross-modal image-level contrastive learning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_7894"} +{"question": "What paper explains that even complex Atari games like Montezuma’s Revenge are made almost fully observable by displaying the player’s score and inventory?", "answer": ["Exploration by Random Network Distillation"], "answer_arxiv_id": ["1810.12894"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_7895"} +{"question": "What papers give an overview of data poisoning and backdoor attacks in adversarial machine learning?", "answer": ["An Overview of Backdoor Attacks Against Deep Neural Networks and Possible Defences"], "answer_arxiv_id": ["2111.08429"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_7896"} +{"question": "What studies have dealt with surface noise in NeRF through the utilization of implicit surface representations like Occupancy?", "answer": ["Differentiable Volumetric Rendering: Learning Implicit 3D\n Representations without 3D Supervision", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction"], "answer_arxiv_id": ["1912.07372", "2104.10078"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_7897"} +{"question": "Which papers adopted methods involving the relaxation of set counting as norm functions for interpolating IoU values?", "answer": ["Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations", "Tversky loss function for image segmentation using 3D fully convolutional deep networks", "A Novel Focal Tversky loss function with Improved Attention U-Net for lesion segmentation"], "answer_arxiv_id": ["1707.03237", "1706.05721", "1810.07842"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_7898"} +{"question": "Can you list some works that studied partial monitoring in the stochastic regime?", "answer": ["An Adaptive Algorithm for Finite Stochastic Partial Monitoring", "Regret Lower Bound and Optimal Algorithm in Finite Stochastic Partial Monitoring", "Analysis and Design of Thompson Sampling for Stochastic Partial Monitoring"], "answer_arxiv_id": ["1206.6487", "1509.09011", "2006.09668"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_7899"} +{"question": "What works introduced the idea of measuring domain-shift sensitivity by comparing gradients?", "answer": ["Improving Test-Time Adaptation via Shift-agnostic Weight Regularization and Nearest Source Prototypes"], "answer_arxiv_id": ["2207.11707"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_7900"} +{"question": "Could you provide the work that proposed the Algorithm Distillation (AD) approach, alternating between online exploration and offline training?", "answer": ["In-context Reinforcement Learning with Algorithm Distillation"], "answer_arxiv_id": ["2210.14215"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_7901"} +{"question": "Which studies introduced more challenging benchmarks like ROxford and RParis?", "answer": ["Revisiting Oxford and Paris: Large-Scale Image Retrieval Benchmarking"], "answer_arxiv_id": ["1803.11285"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_7902"} +{"question": "Which papers focus on qualitative visualizations of real-world multimodal datasets and models?", "answer": ["DIME: Fine-grained Interpretations of Multimodal Models via Disentangled Local Explanations", "M2Lens: Visualizing and Explaining Multimodal Models for Sentiment Analysis"], "answer_arxiv_id": ["2203.02013", "2107.08264"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_7903"} +{"question": "What work provides a geometrical characterization of this manifold?", "answer": ["Geometry of the Loss Landscape in Overparameterized Neural Networks: Symmetries and Invariances"], "answer_arxiv_id": ["2105.12221v2"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_7904"} +{"question": "Is there a study arguing that idealized SSL-trained representations simultaneously cluster data into multiple equivalence classes?", "answer": ["Improving Self-Supervised Learning by Characterizing Idealized Representations"], "answer_arxiv_id": ["2209.06235"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_7905"} +{"question": "What work pre-trains a world model from large-scale human videos and transfers it to robotic manipulation tasks?", "answer": ["Structured World Models from Human Videos"], "answer_arxiv_id": ["2308.10901"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_7906"} +{"question": "What paper considers a repeated Stackelberg game where both the leader and agent learn their optimal actions?", "answer": ["Sample-Efficient Learning of Stackelberg Equilibria in General-Sum Games"], "answer_arxiv_id": ["2102.11494"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7907"} +{"question": "Could you name some works that focus on the Labels-at-Client scenario where each client contains both labeled and unlabeled data?", "answer": ["Federated Semi-Supervised Learning with Inter-Client Consistency & Disjoint Learning"], "answer_arxiv_id": ["2006.12097"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_7908"} +{"question": "Could you provide me a study where adding auxiliary learning tasks around the center found to improve the generalization performance?", "answer": ["Learning Auxiliary Monocular Contexts Helps Monocular 3D Object\n Detection"], "answer_arxiv_id": ["2112.04628"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_7909"} +{"question": "Which papers include studies of visuo-tactile datasets which are commonly used with vision-based sensors?", "answer": ["ObjectFolder: A Dataset of Objects with Implicit Visual, Auditory, and\n Tactile Representations", "ObjectFolder 2.0: A Multisensory Object Dataset for Sim2Real Transfer", "The ObjectFolder Benchmark: Multisensory Learning with Neural and Real\n Objects", "Connecting Look and Feel: Associating the visual and tactile properties\n of physical materials", "MidasTouch: Monte-Carlo inference over distributions across sliding\n touch", "Touch and Go: Learning from Human-Collected Vision and Touch", "Real-time Soft Body 3D Proprioception via Deep Vision-based Sensing", "Visual-Tactile Sensing for In-Hand Object Reconstruction"], "answer_arxiv_id": ["2109.07991", "2204.02389", "2306.00956", "1704.03822", "2210.14210", "2211.12498", "1904.03820", "2303.14498"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_7910"} +{"question": "Which research papers propose novel methods for data valuation?", "answer": ["Estimating Training Data Influence by Tracing Gradient Descent", "LAVA: Data Valuation without Pre-Specified Learning Algorithms"], "answer_arxiv_id": ["2002.08484", "2305.00054"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_7911"} +{"question": "Which studies focus on benign overfitting and consistency in binary classification context?", "answer": ["Classification vs regression in overparameterized regimes: Does the loss function matter?", "The Implicit Bias of Benign Overfitting", "Finite-sample Analysis of Interpolating Linear Classifiers in the Overparameterized Regime", "Benign Overfitting without Linearity: Neural Network Classifiers Trained by Gradient Descent for Noisy Linear Data"], "answer_arxiv_id": ["2005.08054", "2201.11489", "2004.12019", "2202.05928"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_7912"} +{"question": "Which works have examined the capabilities of common machine learning architectures like RNNs, GRUs, SCNs, LSTMs, and Transformers in learning formal languages?", "answer": ["Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation", "Attention Is All You Need"], "answer_arxiv_id": ["1406.1078", "1706.03762"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_7913"} +{"question": "Could you provide me some studies that proposed multi-scale improvements for ambiguous medical image segmentation?", "answer": ["PHiSeg: Capturing Uncertainty in Medical Image Segmentation", "A Hierarchical Probabilistic U-Net for Modeling Multi-Scale Ambiguities"], "answer_arxiv_id": ["1906.04045", "1905.13077"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_7914"} +{"question": "Could you provide me some works about difference and change captioning in images?", "answer": ["Otter: A Multi-Modal Model with In-Context Instruction Tuning", "MIMIC-IT: Multi-Modal In-Context Instruction Tuning", "Flamingo: a Visual Language Model for Few-Shot Learning", "Image Difference Captioning with Pre-training and Contrastive Learning", "Robust Change Captioning", "Changes to Captions: An Attentive Network for Remote Sensing Change\n Captioning"], "answer_arxiv_id": ["2305.03726", "2306.05425", "2204.14198", "2202.04298", "1901.02527", "2304.01091"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_7915"} +{"question": "Could you provide me some works that researched supervising NeRF to generate 3D objects using only 2D diffusion?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_7916"} +{"question": "Could you provide me research papers on strategies that predict context information online through history analysis or selective request of context data from an expert?", "answer": ["Analysis of Thompson Sampling for Partially Observable Contextual Multi-Armed Bandits", "Robust Bandit Learning with Imperfect Context", "Robust Contextual Linear Bandits"], "answer_arxiv_id": ["2110.12175", "2102.05018", "2210.14483v1"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_7917"} +{"question": "Which works are representative of the instance-based approach in self-supervised Vision Transformers?", "answer": ["An Empirical Study of Training Self-Supervised Vision Transformers", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.02057", "2104.14294"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_7918"} +{"question": "What studies considered actual examples of heterophilic graphs such as transaction, ecological food, and molecular networks?", "answer": ["A General Offline Reinforcement Learning Framework for Interactive Recommendation", "Breaking the Limit of Graph Neural Networks by Improving the Assortativity of Graphs with Local Mixing Patterns", "Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs"], "answer_arxiv_id": ["2310.00678", "2106.06586", "2006.11468v2"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_7919"} +{"question": "Which studies proposed methods to enhance the efficiency of cross-validation?", "answer": ["Approximate Cross-Validation for Structured Models", "Approximate Cross-validation: Guarantees for Model Assessment and Selection"], "answer_arxiv_id": ["2006.12669", "2003.00617v2"], "source_meta": {"published_time": "20210503"}, "qid": "AutoScholarQuery_train_7920"} +{"question": "Any works about developing supervised (imitation) learning to train a policy?", "answer": ["Pointer Networks", "Attention, Learn to Solve Routing Problems!", "POMO: Policy Optimization with Multiple Optima for Reinforcement Learning"], "answer_arxiv_id": ["1506.03134", "1803.08475", "2010.16011"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_7921"} +{"question": "What papers introduce stratified techniques for training deep models using the semantics of fuzzy logic?", "answer": ["Logic Tensor Networks for Semantic Image Interpretation", "Analyzing Differentiable Fuzzy Implications", "Deep Adaptive Semantic Logic (DASL): Compiling Declarative Knowledge into Deep Neural Networks"], "answer_arxiv_id": ["1705.08968", "2006.03472", "2003.07344"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7922"} +{"question": "Which work first introduces the vision-language landscape through a contrastive learning framework?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_7923"} +{"question": "Which studies investigate reinforcement learning with CVaR-based constraints?", "answer": ["Algorithms for CVaR Optimization in MDPs", "Risk-Constrained Reinforcement Learning with Percentile Risk Criteria"], "answer_arxiv_id": ["1406.3339", "1512.01629"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_7924"} +{"question": "Which studies constructed datasets of AI-generated images used for specialized purposes such as detecting generated art images?", "answer": ["Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models", "Benchmarking Deepart Detection"], "answer_arxiv_id": ["2212.03860", "2302.14475"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_7925"} +{"question": "Which research work pioneered the creation of the Wild6D dataset?", "answer": ["Category-Level 6D Object Pose Estimation in the Wild: A Semi-Supervised\n Learning Approach and A New Dataset", "TEASER: Fast and Certifiable Point Cloud Registration"], "answer_arxiv_id": ["2206.15436", "2001.07715"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_7926"} +{"question": "Which research papers developed models for visual instruction tuning using pre-trained LLMs?", "answer": ["Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Improved Baselines with Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485", "2304.10592", "2304.14178", "2305.03726", "2305.06500", "2310.03744"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_7927"} +{"question": "Can you show me research papers that focus on using frequency information in computer vision?", "answer": ["Drop an Octave: Reducing Spatial Redundancy in Convolutional Neural\n Networks with Octave Convolution", "Learning in the Frequency Domain", "Frequency Separation for Real-World Super-Resolution", "Fast Vision Transformers with HiLo Attention"], "answer_arxiv_id": ["1904.05049", "2002.12416", "1911.07850", "2205.13213"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_7928"} +{"question": "What works proposed the approach of model repairing as a white-box defense method to remove hidden backdoors in attacked DNNs?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks", "Adversarial Neuron Pruning Purifies Backdoored Deep Models"], "answer_arxiv_id": ["1805.12185", "2110.14430"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_7929"} +{"question": "Which works explored the impact of node degrees on the accuracy discrepancy in Graph Neural Networks?", "answer": ["ResNorm: Tackling Long-tailed Degree Distribution Issue in Graph Neural Networks via Normalization", "LTE4G: Long-Tail Experts for Graph Neural Networks"], "answer_arxiv_id": ["2206.08181v2", "2208.10205"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7930"} +{"question": "What are some papers around the topic of prompt and instruction learning in LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Dynamic Prompt Learning via Policy Gradient for Semi-structured Mathematical Reasoning", "Decomposed Prompting: A Modular Approach for Solving Complex Tasks", "LLaMA: Open and Efficient Foundation Language Models", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention", "Instruction Tuning with GPT-4", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model"], "answer_arxiv_id": ["2201.11903", "2209.14610", "2210.02406", "2302.13971", "2303.16199", "2304.03277", "2304.15010"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_7931"} +{"question": "Could you cite some works about NeRF applications for handling complex and large-scale environments like urban outdoor scenes?", "answer": ["Urban Radiance Fields", "Block-NeRF: Scalable Large Scene Neural View Synthesis", "S-NeRF: Neural Radiance Fields for Street Views", "MARS: An Instance-aware, Modular and Realistic Simulator for Autonomous\n Driving", "Urban Radiance Field Representation with Deformable Neural Mesh\n Primitives", "UrbanGIRAFFE: Representing Urban Scenes as Compositional Generative\n Neural Feature Fields", "READ: Large-Scale Neural Scene Rendering for Autonomous Driving", "SUDS: Scalable Urban Dynamic Scenes", "EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via\n Self-Supervision"], "answer_arxiv_id": ["2111.14643", "2202.05263", "2303.00749", "2307.15058", "2307.10776", "2303.14167", "2205.05509", "2303.14536", "2311.02077"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_7932"} +{"question": "What papers discuss the concept of constrained policy optimization (CPO) in the context of reinforcement learning?", "answer": ["Constrained Policy Optimization"], "answer_arxiv_id": ["1705.10528"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_7933"} +{"question": "Which works have proposed Unlearnable Contrastive Learning (UCL) under unsupervised setting?", "answer": ["Indiscriminate Poisoning Attacks on Unsupervised Contrastive Learning"], "answer_arxiv_id": ["2202.11202"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_7934"} +{"question": "Which work focused on detecting of explicit language?", "answer": ["Automated Hate Speech Detection and the Problem of Offensive Language"], "answer_arxiv_id": ["1703.04009"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_7935"} +{"question": "Can you cite studies that use myopic temperature scaling in natural language generation?", "answer": ["Language Models are Few-Shot Learners", "On the Opportunities and Risks of Foundation Models"], "answer_arxiv_id": ["2005.14165", "2108.07258v3"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_7936"} +{"question": "Which works have attempted to improve the quality of language models by grounding the generation on a set of retrieved materials?", "answer": ["A Survey on Retrieval-Augmented Text Generation", "REALM: Retrieval-Augmented Language Model Pre-Training", "A Retrieve-and-Edit Framework for Predicting Structured Outputs", "Retrieve and Refine: Improved Sequence Generation Models For Dialogue", "Skeleton-to-Response: Dialogue Generation Guided by Retrieval Memory", "Generalization through Memorization: Nearest Neighbor Language Models", "Response Generation by Context-aware Prototype Editing", "REALM: Retrieval-Augmented Language Model Pre-Training", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Improving language models by retrieving from trillions of tokens", "Inference with Reference: Lossless Acceleration of Large Language Models"], "answer_arxiv_id": ["2202.01110", "2002.08909", "1812.01194", "1808.04776", "1809.05296", "1911.00172", "1806.07042", "2002.08909", "2005.11401", "2112.04426", "2304.04487"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_7937"} +{"question": "Which papers have developed modifications in GAN architecture to stabilize the training process?", "answer": ["Improving GAN Equilibrium by Raising Spatial Awareness", "GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium", "Progressive Growing of GANs for Improved Quality, Stability, and Variation", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Training Generative Adversarial Networks with Limited Data", "Analyzing and Improving the Image Quality of StyleGAN", "Alias-Free Generative Adversarial Networks"], "answer_arxiv_id": ["2112.00718", "1706.08500", "1710.10196", "1812.04948", "2006.06676", "1912.04958", "2106.12423"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_7938"} +{"question": "What studies propose other models for frame transition encodings in 3D GNNs?", "answer": ["E(n) Equivariant Graph Neural Networks"], "answer_arxiv_id": ["2102.09844"], "source_meta": {"published_time": "20230407"}, "qid": "AutoScholarQuery_train_7939"} +{"question": "What paper proposed Differentially Private Stochastic Gradient Descent (DP-SGD)?", "answer": ["Deep Learning with Differential Privacy"], "answer_arxiv_id": ["1607.00133"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7940"} +{"question": "What research exists on Class Incremental Learning in federated learning?", "answer": ["Better Generative Replay for Continual Federated Learning", "Federated Reconnaissance: Efficient, Distributed, Class-Incremental Learning"], "answer_arxiv_id": ["2302.13001", "2109.00150"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_7941"} +{"question": "Which papers try to evaluate explanations by correlation analysis with human judgements?", "answer": ["Towards Explainable Evaluation Metrics for Natural Language Generation", "NaturalProver: Grounded Mathematical Proof Generation with Language Models"], "answer_arxiv_id": ["2203.11131", "2205.12910"], "source_meta": {"published_time": "20221215"}, "qid": "AutoScholarQuery_train_7942"} +{"question": "Could you provide me some studies about the self-supervised segmentation framework using object discovery networks?", "answer": ["Object discovery and representation networks"], "answer_arxiv_id": ["2203.08777"], "source_meta": {"published_time": "20220919"}, "qid": "AutoScholarQuery_train_7943"} +{"question": "What papers are about the application of contrastive learning on imbalanced datasets and different feature spaces?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Partner-Assisted Learning for Few-Shot Image Classification"], "answer_arxiv_id": ["2006.09882", "2006.07733", "2109.07607"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_7944"} +{"question": "What sources proposed to divide mismatched pairs from training data for more practical retrieval?", "answer": ["BiCro: Noisy Correspondence Rectification for Multi-modality Data via\n Bi-directional Cross-modal Similarity Consistency"], "answer_arxiv_id": ["2303.12419"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_7945"} +{"question": "What studies utilize codewords with erroneous trapping sets for training sample selection?", "answer": ["Decoding Short LDPC Codes via BP-RNN Diversity and Reliability-Based Post-Processing", "FAID Diversity via Neural Networks"], "answer_arxiv_id": ["2206.12150v3", "2105.04118v1"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_7946"} +{"question": "Which studies have applied prompt tuning to image or video domains for open-vocabulary visual recognition?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2107.13586v1"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_7947"} +{"question": "Any papers about applying knowledge distillation in the natural language processing domain?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_7948"} +{"question": "Could you tell me about studies where ResNet models trained with unsupervised objectives surpassed supervised objectives for both visual navigation and control tasks?", "answer": ["The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2203.03580", "1911.05722"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_7949"} +{"question": "What works learn semantic to visual mapping in Generative Zero-Shot Learning?", "answer": ["Latent Embedding Feedback and Discriminative Features for Zero-Shot\n Classification", "FREE: Feature Refinement for Generalized Zero-Shot Learning", "Contrastive Embedding for Generalized Zero-Shot Learning", "Evolving Semantic Prototype Improves Generative Zero-Shot Learning"], "answer_arxiv_id": ["2003.07833", "2107.13807", "2103.16173", "2306.06931"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_7950"} +{"question": "Could you list the works that aim to learn network depth, filter sizes and other layer types from training data?", "answer": ["Depth Uncertainty in Neural Networks", "DNArch: Learning Convolutional Neural Architectures by Backpropagation"], "answer_arxiv_id": ["2006.08437", "2302.05400"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_7951"} +{"question": "Could you provide me some studies that are similar to DCLS in terms of building a scale equivariant neural network?", "answer": ["Deep Scale-spaces: Equivariance Over Scale", "Scale-Equivariant Steerable Networks", "DISCO: accurate Discrete Scale Convolutions", "B-Spline CNNs on Lie Groups", "Scaling-Translation-Equivariant Networks with Decomposed Convolutional Filters"], "answer_arxiv_id": ["1905.11697", "1910.11093", "2106.02733", "1909.12057", "1909.11193"], "source_meta": {"published_time": "20211207"}, "qid": "AutoScholarQuery_train_7952"} +{"question": "What studies does the works draw from in its efforts to directly learn optimal policies?", "answer": ["The turnpike property in finite-dimensional nonlinear optimal control"], "answer_arxiv_id": ["1402.3263"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_7953"} +{"question": "Could you list down the studies related to unsupervised CSG tree reconstruction?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation"], "answer_arxiv_id": ["1812.02822", "1812.03828", "1901.05103"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_7954"} +{"question": "Which paper proposed the Recurrent Independent Mechanisms (RIMs)?", "answer": ["Recurrent Independent Mechanisms"], "answer_arxiv_id": ["1909.10893"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_7955"} +{"question": "Which works provide examples of diffusion generative models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Zero-shot Generation of Coherent Storybook from Plain Text Story using Diffusion Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2105.05233", "2302.03900", "2010.02502"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_7956"} +{"question": "Which works deal with the problem of 'learning shortcuts' where models overly rely on spurious features?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_7957"} +{"question": "Could you name any studies that leveraged advancements in large-scale text-to-image diffusion models?", "answer": ["Exploiting Diffusion Prior for Real-World Image Super-Resolution", "DiffBIR: Towards Blind Image Restoration with Generative Diffusion Prior", "Pixel-Aware Stable Diffusion for Realistic Image Super-resolution and\n Personalized Stylization"], "answer_arxiv_id": ["2305.07015", "2308.15070", "2308.14469"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_7958"} +{"question": "Which works contain high levels of variation (HLV) in NLI?", "answer": ["What Can We Learn from Collective Human Opinions on Natural Language\n Inference Data?", "Investigating Reasons for Disagreement in Natural Language Inference", "Ecologically Valid Explanations for Label Variation in NLI"], "answer_arxiv_id": ["2010.03532", "2209.03392", "2310.13850"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_7959"} +{"question": "What studies have been conducted on adding new instances to scenes to generate diverse training samples for images?", "answer": ["Modeling Visual Context is Key to Augmenting Object Detection Datasets", "Learning to Generate Synthetic Data via Compositing", "Cut, Paste and Learn: Surprisingly Easy Synthesis for Instance Detection", "Simple Copy-Paste is a Strong Data Augmentation Method for Instance\n Segmentation"], "answer_arxiv_id": ["1807.07428", "1904.05475", "1708.01642", "2012.07177"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_7960"} +{"question": "What work invoked a foreshadowing of the clustering mechanism to reduce the quadratic complexity of self-attention?", "answer": ["Fast Transformers with Clustered Attention"], "answer_arxiv_id": ["2007.04825"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_7961"} +{"question": "Which work formulated that traditional message-passing neural networks (MPNN) are not more expressive than 1-WL?", "answer": ["How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["1810.00826"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_7962"} +{"question": "Which studies discuss different approaches in the context of StyleGAN editing, such as additive perturbations and affine transformation on latents?", "answer": ["Enjoy Your Editing: Controllable GANs for Image Editing via Latent Space Navigation", "Closed-Form Factorization of Latent Semantics in GANs", "WarpedGANSpace: Finding non-linear RBF paths in GAN latent space", "StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation", "StyleFusion: A Generative Model for Disentangling Spatial Segments"], "answer_arxiv_id": ["2102.01187", "2007.06600v4", "2109.13357", "2011.12799", "2107.07437"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_7963"} +{"question": "Could you provide me the study that used the improvement method from bib.bib20 for inducing changes in the image?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2204.06125"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_7964"} +{"question": "Are there any works about using behavior cloning as a pretraining algorithm?", "answer": ["Provable Representation Learning for Imitation Learning via Bi-level Optimization", "OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning", "Behavior Prior Representation learning for Offline Reinforcement Learning"], "answer_arxiv_id": ["2002.10544", "2010.13611", "2211.00863"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_7965"} +{"question": "Could you provide the references where clips have been grouped by rough spatial location as topological graphs or activity threads?", "answer": ["Ego-Topo: Environment Affordances from Egocentric Video", "UnweaveNet: Unweaving Activity Stories"], "answer_arxiv_id": ["2001.04583", "2112.10194"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_7966"} +{"question": "What works practice using a conditional GAN to estimate diffuse, specular, and normal maps for relightable avatars?", "answer": ["AvatarMe: Realistically Renderable 3D Facial Reconstruction\n \"in-the-wild\"", "AvatarMe++: Facial Shape and BRDF Inference with Photorealistic\n Rendering-Aware GANs", "FitMe: Deep Photorealistic 3D Morphable Model Avatars", "Relightify: Relightable 3D Faces from a Single Image via Diffusion\n Models"], "answer_arxiv_id": ["2003.13845", "2112.05957", "2305.09641", "2305.06077"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_7967"} +{"question": "Which works have used the GAN framework for the synthetic data generation task?", "answer": ["Generating Multi-label Discrete Patient Records using Generative Adversarial Networks", "Data Synthesis based on Generative Adversarial Networks", "Modeling Tabular Data using Conditional GAN", "CTAB-GAN: Effective Table Data Synthesizing"], "answer_arxiv_id": ["1703.06490", "1806.03384v5", "1907.00503", "2102.08369"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_7968"} +{"question": "Which research works proposed regularization-based methods for Catastrophic forgetting in Continual Learning?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Continual Learning Through Synaptic Intelligence", "Overcoming Catastrophic Forgetting by Incremental Moment Matching", "Continual Learning Through Synaptic Intelligence", "Memory Aware Synapses: Learning what (not) to forget", "Task-Free Continual Learning", "Gradient Projection Memory for Continual Learning", "Orthogonal Gradient Descent for Continual Learning"], "answer_arxiv_id": ["1612.00796", "1703.04200", "1703.08475", "1703.04200", "1711.09601", "1812.03596", "2103.09762", "1910.07104"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_7969"} +{"question": "What papers have focused on the training or fine-tuning of large-scale text-to-image models?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2211.09800", "2210.09276", "2206.10789", "2108.01073", "2204.06125", "2205.11487", "2112.10741"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_7970"} +{"question": "Which works adapted discrete LVMs to deep learning-scale data for robust classification?", "answer": ["Mining self-similarity: Label super-resolution with epitomic representations"], "answer_arxiv_id": ["2004.11498"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_7971"} +{"question": "What publications focused on using Transformer models for long-term time series forecasting?", "answer": ["Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecasting", "Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting", "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting", "FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting"], "answer_arxiv_id": ["1907.00235", "2012.07436", "2106.13008", "2201.12740"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_7972"} +{"question": "Which papers proposed the use of value decomposition in the modelling of Q(s, a) by aggregating Qi with the sum method?", "answer": ["Value-Decomposition Networks For Cooperative Multi-Agent Learning"], "answer_arxiv_id": ["1706.05296"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_7973"} +{"question": "What research introduced the use of language instructions for various NLP tasks and examples for large language models?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_7974"} +{"question": "Could you provide me with some works discussing the necessity for proposal distributions for scalability in the context of implicit policies?", "answer": ["Q-Learning in enormous action spaces via amortized approximate maximization"], "answer_arxiv_id": ["2001.08116"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_7975"} +{"question": "What works pose a promising approach for modeling random communication on graphs?", "answer": ["A principled framework for the design and analysis of token algorithms"], "answer_arxiv_id": ["2205.15015"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_7976"} +{"question": "What papers used a contrastive learning approach in the semantic segmentation problem?", "answer": ["Contrastive Learning for Label Efficient Semantic Segmentation", "Exploring Cross-Image Pixel Contrast for Semantic Segmentation"], "answer_arxiv_id": ["2012.06985", "2101.11939"], "source_meta": {"published_time": "20200331"}, "qid": "AutoScholarQuery_train_7977"} +{"question": "Which papers proposed methods for sparsifying graphs while attempting to maintain distances between vertices?", "answer": ["Structure-Preserving Sparsification Methods for Social Networks"], "answer_arxiv_id": ["1601.00286v1"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_7978"} +{"question": "Which papers acknowledge that modern CNNs are not shift-equivariant due to the usage of pooling layers?", "answer": ["Why do deep convolutional networks generalize so poorly to small image\n transformations?", "Making Convolutional Networks Shift-Invariant Again"], "answer_arxiv_id": ["1805.12177", "1904.11486"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7979"} +{"question": "Which papers propose algorithmic improvements to help stabilize the training process of transformers", "answer": ["On Layer Normalization in the Transformer Architecture", "Understanding the Difficulty of Training Transformers", "Robust Optimization for Multilingual Translation with Imbalanced Data"], "answer_arxiv_id": ["2002.04745", "2004.08249", "2104.07639"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_7980"} +{"question": "Which research introduced the method to adapt future models by focusing on detecting AI-generated images?", "answer": ["Online Detection of AI-Generated Images"], "answer_arxiv_id": ["2310.15150"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_7981"} +{"question": "What papers contributed in the development of deep ensembles for uncertainty estimation?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Training independent subnetworks for robust prediction"], "answer_arxiv_id": ["1612.01474", "2010.06610"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_7982"} +{"question": "Which research implemented attention graphs for spatiotemporal social interactions?", "answer": ["TRiPOD: Human Trajectory and Pose Dynamics Forecasting in the Wild"], "answer_arxiv_id": ["2104.04029"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_7983"} +{"question": "What works have investigated various loss formations in depth estimation?", "answer": ["Unsupervised Monocular Depth Estimation with Left-Right Consistency", "SfM-Net: Learning of Structure and Motion from Video", "GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera\n Pose", "Competitive Collaboration: Joint Unsupervised Learning of Depth, Camera\n Motion, Optical Flow and Motion Segmentation", "Digging Into Self-Supervised Monocular Depth Estimation", "Every Pixel Counts ++: Joint Learning of Geometry and Motion with 3D\n Holistic Understanding", "Self-supervised Object Motion and Depth Estimation from Video", "3D Packing for Self-Supervised Monocular Depth Estimation", "RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching"], "answer_arxiv_id": ["1609.03677", "1704.07804", "1803.02276", "1805.09806", "1806.01260", "1810.06125", "1912.04250", "1905.02693", "2109.07547"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_7984"} +{"question": "Could you provide some recent approaches for covariate shift detection?", "answer": ["Learning Deep Kernels for Non-Parametric Two-Sample Tests"], "answer_arxiv_id": ["2002.09116"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_7985"} +{"question": "What papers discuss learning a prior distribution on a similar task?", "answer": ["PACOH: Bayes-Optimal Meta-Learning with PAC-Guarantees", "Probabilistic Model-Agnostic Meta-Learning", "Bayesian Neural Network Priors Revisited", "Deterministic Variational Inference for Robust Bayesian Neural Networks", "Scalable Marginal Likelihood Estimation for Model Selection in Deep Learning", "Specifying Weight Priors in Bayesian Deep Neural Networks with Empirical Bayes", "Pre-Train Your Loss: Easy Bayesian Transfer Learning with Informative Priors", "Plex: Towards Reliability Using Pretrained Large Model Extensions"], "answer_arxiv_id": ["2002.05551", "1806.02817", "2102.06571", "1810.03958", "2104.04975", "1906.05323", "2205.10279", "2207.07411"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_7986"} +{"question": "Which research papers used paraphrasing techniques to find adversarial examples for natural language processing tasks and for training a robust language model via adversarial training?", "answer": ["Discrete Adversarial Attacks and Submodular Optimization with Applications to Text Classification", "Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with Adversarial Examples"], "answer_arxiv_id": ["1812.00151", "1803.01128"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_7987"} +{"question": "What works propose the development of the human pose and shape (HPS) using images and videos?", "answer": ["End-to-end Recovery of Human Shape and Pose", "Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loop", "VIBE: Video Inference for Human Body Pose and Shape Estimation", "PARE: Part Attention Regressor for 3D Human Body Estimation", "CLIFF: Carrying Location Information in Full Frames into Human Pose and Shape Estimation", "Zolly: Zoom Focal Length Correctly for Perspective-Distorted Human Mesh Reconstruction", "Learning Dense UV Completion for Human Mesh Recovery"], "answer_arxiv_id": ["1712.06584", "1909.12828", "1912.05656", "2104.08527", "2208.00571", "2303.13796", "2307.11074"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_7988"} +{"question": "Which study reported differences in performance between methods of 'bib.bib126' and 'bib.bib102'?", "answer": ["Balancing Stability and Plasticity through Advanced Null Space in Continual Learning"], "answer_arxiv_id": ["2207.12061"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_7989"} +{"question": "What studies attempted to address the issue of heterogeneity in Federated Learning?", "answer": ["SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning", "Model-Contrastive Federated Learning"], "answer_arxiv_id": ["1910.06378", "2008.03606", "2103.16257"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_7990"} +{"question": "What are some of the high-performing GBDT variants used in tabular datasets?", "answer": ["XGBoost: A Scalable Tree Boosting System", "CatBoost: unbiased boosting with categorical features"], "answer_arxiv_id": ["1603.02754", "1706.09516"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_7991"} +{"question": "What works discuss the unique anti-aliasing challenges presented by 3D Gaussian splatting?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_7992"} +{"question": "Can you list the studies that try to modify training for KGEs to introduce a penalty for triples that do not satisfy given logical constraints?", "answer": ["Type-Constrained Representation Learning in Knowledge Graphs", "Improving Knowledge Graph Embedding Using Simple Constraints"], "answer_arxiv_id": ["1508.02593", "1805.02408"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_7993"} +{"question": "Can you list the papers that work on extracting symmetry learned by neural network?", "answer": ["LieGG: Studying Learned Lie Group Generators"], "answer_arxiv_id": ["2210.04345"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_7994"} +{"question": "Could you provide me with some cross-encoder architecture based methodologies for KGC?", "answer": ["KG-BERT: BERT for Knowledge Graph Completion"], "answer_arxiv_id": ["1909.03193"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_7995"} +{"question": "Are there any studies providing general analysis on Adam-type optimizers?", "answer": ["On the Convergence of A Class of Adam-Type Algorithms for Non-Convex Optimization", "On the Convergence of Adaptive Gradient Methods for Nonconvex Optimization", "MixML: A Unified Analysis of Weakly Consistent Parallel Learning", "Recent Theoretical Advances in Non-Convex Optimization", "A Sufficient Condition for Convergences of Adam and RMSProp"], "answer_arxiv_id": ["1808.02941", "1808.05671", "2005.06706", "2012.06188", "1811.09358"], "source_meta": {"published_time": "20220212"}, "qid": "AutoScholarQuery_train_7996"} +{"question": "Which papers propose non-parametric methods for domain adaptation by referencing a datastore of similar instances?", "answer": ["Nearest Neighbor Machine Translation", "Adaptive Nearest Neighbor Machine Translation", "Learning Kernel-Smoothed Machine Translation with Retrieved Examples"], "answer_arxiv_id": ["2010.00710", "2105.13022", "2109.09991"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_7997"} +{"question": "Which works map 2D image inputs to elements of various groups allowing for disentanglement and equivariance constraints?", "answer": ["Learning Disentangled Representations and Group Structure of Dynamical Environments", "Equivariant Representation Learning via Class-Pose Decomposition"], "answer_arxiv_id": ["2002.06991", "2207.03116v3"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_7998"} +{"question": "What studies use the CLIP model in the text-guided image synthesis, specifically for object generation?", "answer": ["FuseDream: Training-Free Text-to-Image Generation with Improved CLIP+GAN Space Optimization", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields"], "answer_arxiv_id": ["2112.01573", "2112.05139"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_7999"} +{"question": "What studies propose the use of category descriptions from language models for improved zero-shot image classification?", "answer": ["What does a platypus look like? Generating customized prompts for\n zero-shot image classification", "Visual Classification via Description from Large Language Models"], "answer_arxiv_id": ["2209.03320", "2210.07183"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8000"} +{"question": "Which work explores the idea of tracking an object first and then utilizing the cross-modality features?", "answer": ["Cross-Modality Time-Variant Relation Learning for Generating Dynamic\n Scene Graphs"], "answer_arxiv_id": ["2305.08522"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_8001"} +{"question": "What papers show that CLIP-based CZSL methods continue the vein of prompt tuning?", "answer": ["Learning to Compose Soft Prompts for Compositional Zero-Shot Learning", "Prompting Large Pre-trained Vision-Language Models For Compositional\n Concept Learning", "Decomposed Soft Prompt Guided Fusion Enhancing for Compositional\n Zero-Shot Learning"], "answer_arxiv_id": ["2204.03574", "2211.05077", "2211.10681"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_8002"} +{"question": "Which research discuss the use of retrieval augmented generation in the field of Open-Domain QA?", "answer": ["Dense Passage Retrieval for Open-Domain Question Answering", "Interleaving Retrieval with Chain-of-Thought Reasoning for\n Knowledge-Intensive Multi-Step Questions"], "answer_arxiv_id": ["2004.04906", "2212.10509"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_8003"} +{"question": "Could you provide me some studies that scaled up the graph contrastive learning?", "answer": ["Rethinking and Scaling Up Graph Contrastive Learning: An Extremely Efficient Approach with Group Discrimination"], "answer_arxiv_id": ["2206.01535"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_8004"} +{"question": "What research has been done in 3D generation focused on learning the distribution of 3D shapes and textures from category-specific datasets?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "3D Shape Generation and Completion through Point-Voxel Diffusion", "Escaping Plato's Cave: 3D Shape From Adversarial Rendering", "HoloGAN: Unsupervised learning of 3D representations from natural images"], "answer_arxiv_id": ["1610.07584", "1906.12320", "2104.03670", "1811.11606", "1904.01326"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_8005"} +{"question": "What studies show the introduction of Transformer based models have improved self-supervised pretraining?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: A Simple Framework for Masked Image Modeling"], "answer_arxiv_id": ["2106.08254", "2111.06377", "2111.09886"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_8006"} +{"question": "Which study is the source of metadata employed in this research?", "answer": ["Graph Learning Indexer: A Contributor-Friendly and Metadata-Rich Platform for Graph Learning Benchmarks"], "answer_arxiv_id": ["2212.04537"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_8007"} +{"question": "Could you name works that demonstrated zero-shot learning success in multilingual question answering and image classification tasks?", "answer": ["Large-Scale Zero-Shot Image Classification from Rich and Diverse Textual Descriptions", "K-Lite: Learning Transferable Visual Models with External Knowledge"], "answer_arxiv_id": ["2103.09669", "2204.09222"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_8008"} +{"question": "What works use the framework of VQVAE and autoregressive transformers for video generation?", "answer": ["Neural Discrete Representation Learning", "Attention Is All You Need", "Language Models are Few-Shot Learners", "GODIVA: Generating Open-DomaIn Videos from nAtural Descriptions", "Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive Transformer"], "answer_arxiv_id": ["1711.00937", "1706.03762", "2005.14165", "2104.14806", "2204.03638"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_8009"} +{"question": "What papers discuss having the variational posterior depend on the prior parameters to address the amortization gap issue?", "answer": ["A General Method for Amortizing Variational Filtering", "Composing graphical models with neural networks for structured representations and fast inference", "Variational Message Passing with Structured Inference Networks", "VAE with a VampPrior"], "answer_arxiv_id": ["1811.05090", "1603.06277", "1803.05589", "1705.07120v5"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8010"} +{"question": "Could you name some studies that aim to tackle the well-known problem of projection domain shift in Visual-Semantic Domain Shift and Projection Domain Shift?", "answer": ["Dual Progressive Prototype Network for Generalized Zero-Shot Learning", "Transductive Multi-view Zero-Shot Learning", "Transductive Zero-Shot Learning with Visual Structure Constraint"], "answer_arxiv_id": ["2111.02073", "1501.04560", "1901.01570"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_8011"} +{"question": "What are the studies addressing the limitations of neglecting inter-person occlusions and individual positions in the full frame in the existing HPS methods?", "answer": ["Monocular, One-stage, Regression of Multiple 3D People", "Putting People in their Place: Monocular Regression of 3D People in\n Depth", "TRACE: 5D Temporal Regression of Avatars with Dynamic Cameras in 3D\n Environments"], "answer_arxiv_id": ["2008.12272", "2112.08274", "2306.02850"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_8012"} +{"question": "Which work proposed learning the teacher model with the supervision of the student model resulting in a student-friendly teacher?", "answer": ["Learning Student-Friendly Teacher Networks for Knowledge Distillation"], "answer_arxiv_id": ["2102.07650"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8013"} +{"question": "Could you provide me some works about using artificial neural networks as image-computable encoders or predictors of the visual pathway?", "answer": ["Brain-Like Object Recognition with High-Performing Shallow Recurrent ANNs"], "answer_arxiv_id": ["1909.06161"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_8014"} +{"question": "What papers contains successful applications of Modern Hopfield Networks?", "answer": ["Modern Hopfield Networks and Attention for Immune Repertoire Classification", "CLOOB: Modern Hopfield Networks with InfoLOOB Outperform CLIP", "History Compression via Language Models in Reinforcement Learning", "Hopular: Modern Hopfield Networks for Tabular Data", "Txt2Img-MHN: Remote Sensing Image Generation from Text Using Modern Hopfield Networks"], "answer_arxiv_id": ["2007.13505", "2110.11316", "2205.12258", "2206.00664", "2208.04441"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_8015"} +{"question": "What studies have been conducted on counterfactual reasoning in machine learning?", "answer": ["Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models", "Decisions, Counterfactual Explanations and Strategic Behavior", "Counterfactual Explanations in Sequential Decision Making Under Uncertainty", "Meaningfully Debugging Model Mistakes using Conceptual Counterfactual Explanations", "Counterfactual Temporal Point Processes"], "answer_arxiv_id": ["1905.05824", "2002.04333", "2107.02776", "2106.12723", "2111.07603"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_8016"} +{"question": "What research introduces Variational Score Distillation (VSD) for better diversity and quality in text-guided 3D content generation?", "answer": ["ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation"], "answer_arxiv_id": ["2305.16213"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_8017"} +{"question": "What studies used confidence thresholding in the context of self-training?", "answer": ["SimPLE: Similar Pseudo Label Exploitation for Semi-Supervised Classification", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["2103.16725", "2001.07685"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_8018"} +{"question": "Could you provide me the study about the creation of high-quality, diverse multi-modal datasets from GPT4 and GPT-4V?", "answer": ["The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)"], "answer_arxiv_id": ["2309.17421"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_8019"} +{"question": "Which work introduced the first non-asymptotic convergence analysis for non-convex Stochastic Bilevel Optimization with a strongly convex lower level problem?", "answer": ["Approximation Methods for Bilevel Programming"], "answer_arxiv_id": ["1802.02246"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8020"} +{"question": "Which papers analyzed theoretical communication, statistical, and privacy tradeoffs in federated learning?", "answer": ["Breaking the Communication-Privacy-Accuracy Trilemma"], "answer_arxiv_id": ["2007.11707"], "source_meta": {"published_time": "20230722"}, "qid": "AutoScholarQuery_train_8021"} +{"question": "What studies have been conducted on few-shot learning approaches for semantic segmentation?", "answer": ["One-Shot Learning for Semantic Segmentation", "PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment", "MSANet: Multi-Similarity and Attention Guidance for Boosting Few-Shot Segmentation"], "answer_arxiv_id": ["1709.03410", "1908.06391", "2206.09667"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_8022"} +{"question": "Can you provide references on problem-solving simulations based on competitive games?", "answer": ["Avalon's Game of Thoughts: Battle Against Deception through Recursive\n Contemplation", "Language Agents with Reinforcement Learning for Strategic Play in the\n Werewolf Game", "Put Your Money Where Your Mouth Is: Evaluating Strategic Planning and\n Execution of LLM Agents in an Auction Arena"], "answer_arxiv_id": ["2310.01320", "2310.18940", "2310.05746"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_8023"} +{"question": "Are there works which discuss that without auxiliary rewards, the learned behavior alignment rewards may be interpreted as intrinsic rewards?", "answer": ["On Learning Intrinsic Rewards for Policy Gradient Methods", "What Can Learned Intrinsic Rewards Capture?"], "answer_arxiv_id": ["1804.06459", "1912.05500"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_8024"} +{"question": "Which papers involve classifier-free guidance in text-to-image diffusion models?", "answer": ["Classifier-Free Diffusion Guidance", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers"], "answer_arxiv_id": ["2207.12598", "2112.10752", "2204.06125", "2211.01324"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_8025"} +{"question": "Could you tell me what papers extend masked language modeling to video domains for learning good representations for action recognition?", "answer": ["VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training", "Masked Autoencoders As Spatiotemporal Learners"], "answer_arxiv_id": ["2203.12602", "2205.09113"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_8026"} +{"question": "Which works incorporate a framework for deep node clustering on temporal graphs?", "answer": ["Deep Temporal Graph Clustering"], "answer_arxiv_id": ["2305.10738"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_8027"} +{"question": "Which research papers adopted neuron analysis to study sentiment analysis, knowledge storing, and task solving in language models?", "answer": ["Learning to Generate Reviews and Discovering Sentiment", "Knowledge Neurons in Pretrained Transformers", "Finding Skill Neurons in Pre-trained Transformer-based Language Models"], "answer_arxiv_id": ["1704.01444", "2104.08696", "2211.07349"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_8028"} +{"question": "Can you list the works that utilized token merging and pruning for model compression?", "answer": ["Token Merging: Your ViT But Faster"], "answer_arxiv_id": ["2210.09461"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_8029"} +{"question": "Could you provide me some works where critiques used for reflection took the form of scores or natural language?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback", "Reflexion: Language Agents with Verbal Reinforcement Learning", "DERA: Enhancing Large Language Model Completions with Dialog-Enabled\n Resolving Agents", "How FaR Are Large Language Models From Agents with Theory-of-Mind?"], "answer_arxiv_id": ["2303.17651", "2303.11366", "2303.17071", "2310.03051"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_8030"} +{"question": "Which papers proposed algorithms for linear bandit problems where the action set is fixed?", "answer": ["Linearly Parameterized Bandits"], "answer_arxiv_id": ["0812.3465"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_8031"} +{"question": "Which papers discuss methods using attention maps for image editing and token identification in the context of Diffusion Models?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Localizing Object-level Shape Variations with Text-to-Image Diffusion\n Models", "MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing", "Diffusion Self-Guidance for Controllable Image Generation"], "answer_arxiv_id": ["2208.01626", "2303.11306", "2304.08465v1", "2306.00986"], "source_meta": {"published_time": "20240508"}, "qid": "AutoScholarQuery_train_8032"} +{"question": "Which papers mostly focus on the interpretation of deep learning methods applied in the classification tasks?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "Striving for Simplicity: The All Convolutional Net", "Learning Important Features Through Propagating Activation Differences", "Axiomatic Attribution for Deep Networks", "Interpreting Deep Visual Representations via Network Dissection", "A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1312.6034", "1412.6806", "1704.02685", "1703.01365", "1711.05611", "1705.07874"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_8033"} +{"question": "What are the works about adaptive quantization in machine learning?", "answer": ["Adaptive Gradient Quantization for Data-Parallel SGD", "Adaptive Communication Strategies to Achieve the Best Error-Runtime Trade-off in Local-Update SGD"], "answer_arxiv_id": ["2010.12460", "1810.08313v2"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_8034"} +{"question": "Could you provide me some studies that designed new computationally efficient algorithms with specific sample complexities?", "answer": ["Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes", "Reward-Free RL is No Harder Than Reward-Aware RL in Linear Markov Decision Processes"], "answer_arxiv_id": ["2212.06132", "2201.11206"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_8035"} +{"question": "Can you list some of the research studies on developing representation learning methods like state extractor from vision-based features, bi-simulation, successor feature, spectral representation from transition operator decomposition, and contrastive learning?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Reinforcement Learning with Augmented Data", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Metrics for Finite Markov Decision Processes", "DeepMDP: Learning Continuous Latent Space Models for Representation Learning", "Learning Invariant Representations for Reinforcement Learning without Reconstruction", "Successor Features for Transfer in Reinforcement Learning", "Deep Successor Reinforcement Learning", "The Laplacian in RL: Learning Representations with Efficient Approximations", "State Aggregation Learning from Markov Transition Data", "Representation Learning with Contrastive Predictive Coding", "Provable Representation Learning for Imitation with Contrastive Fourier Features"], "answer_arxiv_id": ["2004.04136", "2004.14990", "2004.13649", "1207.4114v1", "1906.02736", "2006.10742", "1606.05312", "1606.02396", "1810.04586", "1811.02619", "1807.03748", "2105.12272"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_8036"} +{"question": "Could you provide me with some studies that used only the shape information for ligand generation?", "answer": ["Zero-Shot 3D Drug Design by Sketching and Generating", "Equivariant Shape-Conditioned Generation of 3D Molecules for Ligand-Based Drug Design"], "answer_arxiv_id": ["2209.13865", "2210.04893"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_8037"} +{"question": "Could you tell me about the research work that uses VQ-GAN trained over diverse data for more expressive generation capabilities?", "answer": ["VQGAN-CLIP: Open Domain Image Generation and Editing with Natural Language Guidance"], "answer_arxiv_id": ["2204.08583"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_8038"} +{"question": "In what paper small language models were improved by being trained on their refined outputs generated using constrained decoding and filtered with a simple supervised critic model?", "answer": ["I2D2: Inductive Knowledge Distillation with NeuroLogic and\n Self-Imitation"], "answer_arxiv_id": ["2212.09246"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_8039"} +{"question": "What papers cover the use of feedback and refinement pairs to learn supervised refiners?", "answer": ["PEER: A Collaborative Language Model", "Read, Revise, Repeat: A System Demonstration for Human-in-the-loop Iterative Text Revision", "Graph-based, Self-Supervised Program Repair from Diagnostic Feedback", "Think about it! Improving defeasible reasoning by first modeling the question scenario"], "answer_arxiv_id": ["2208.11663", "2204.03685", "2005.10636", "2110.12349"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_8040"} +{"question": "Which studies focus on improving language models' performance by careful design of natural language task specifications?", "answer": ["Ask Me Anything: A simple strategy for prompting language models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2210.02441", "2201.11903"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8041"} +{"question": "Which study involved training a sequence-to-sequence model to improve the efficiency of a prompt?", "answer": ["Black-Box Prompt Optimization: Aligning Large Language Models without\n Model Training"], "answer_arxiv_id": ["2311.04155"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8042"} +{"question": "What research proposes GCond to reduce a large-scale graph for node classification using the gradient matching scheme in DC?", "answer": ["Graph Condensation for Graph Neural Networks"], "answer_arxiv_id": ["2110.07580"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_8043"} +{"question": "What are the works that have explored the concept of Bayes consistency in multiclass classification?", "answer": ["Weston-Watkins Hinge Loss and Ordered Partitions", "Convex Calibration Dimension for Multiclass Loss Matrices", "An Embedding Framework for Consistent Polyhedral Surrogates", "Surrogate Regret Bounds for Polyhedral Losses", "An Embedding Framework for the Design and Analysis of Consistent Polyhedral Surrogates", "On Classification-Calibration of Gamma-Phi Losses"], "answer_arxiv_id": ["2006.07346", "1408.2764", "1907.07330", "2110.14031", "2206.14707", "2302.07321"], "source_meta": {"published_time": "20230414"}, "qid": "AutoScholarQuery_train_8044"} +{"question": "Where can I find the information about the prompt-based method that employs energy self-normalization in continual learning?", "answer": ["Isolation and Impartial Aggregation: A Paradigm of Incremental Learning\n without Interference"], "answer_arxiv_id": ["2211.15969"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_8045"} +{"question": "What studies have applied the Optimal Transport (OT) framework in various machine learning tasks?", "answer": ["Scalable Optimal Transport Methods in Machine Learning: A Contemporary\n Survey", "Joint Distribution Optimal Transportation for Domain Adaptation", "DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised\n Domain Adaptation", "Improved Training of Wasserstein GANs", "Improving GANs Using Optimal Transport", "Graph Optimal Transport for Cross-Domain Alignment", "GOT: An Optimal Transport framework for Graph comparison", "ZegOT: Zero-shot Segmentation Through Optimal Transport of Text Prompts"], "answer_arxiv_id": ["2305.05080", "1705.08848", "1803.10081", "1704.00028", "1803.05573", "2006.14744", "1906.02085", "2301.12171"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8046"} +{"question": "Could you provide me studies about object detection, a type of sparse prediction task in visual perception?", "answer": ["End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2005.12872"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_8047"} +{"question": "Which works propose methods for reducing maximization bias with double learning or cross-validation estimators?", "answer": ["Estimating the Maximum Expected Value: An Analysis of (Nested) Cross Validation and the Maximum Sample Average", "Deep Reinforcement Learning with Double Q-learning"], "answer_arxiv_id": ["1302.7175", "1509.06461"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_8048"} +{"question": "What works propose regularization-based methods for OOD detection?", "answer": ["Discriminative out-of-distribution detection for semantic segmentation", "Deep Anomaly Detection with Outlier Exposure", "Predictive Uncertainty Estimation via Prior Networks", "Energy-based Out-of-distribution Detection", "VOS: Learning What You Don’t Know by Virtual Outlier Synthesis", "POEM: Out-of-Distribution Detection with Posterior Sampling"], "answer_arxiv_id": ["1808.07703", "1812.04606", "1802.10501", "2010.03759", "2202.01197", "2206.13687"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_8049"} +{"question": "What studies developed flow-based and diffusion-based generative models to leverage E(n)-Equivariant GNN?", "answer": ["E(n) Equivariant Normalizing Flows", "Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2105.09016", "2203.17003"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_8050"} +{"question": "Which papers exemplify Spatial-Attention Models in object-centric learning?", "answer": ["Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects"], "answer_arxiv_id": ["1603.08575", "1806.01794"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_8051"} +{"question": "Are there any studies that proposed alternate efficient generalized linear bandit algorithm following the line of Thompson sampling scheme?", "answer": ["An Efficient Algorithm For Generalized Linear Bandit: Online Stochastic Gradient Descent and Thompson Sampling"], "answer_arxiv_id": ["2006.04012"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_8052"} +{"question": "What works have explored the intersection of LLMs and planning?", "answer": ["LLM+P: Empowering Large Language Models with Optimal Planning Proficiency"], "answer_arxiv_id": ["2304.11477"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_8053"} +{"question": "What studies presented methods to address the structural limitation of the CT-DE framework?", "answer": ["QPLEX: Duplex Dueling Multi-Agent Q-Learning", "Weighted QMIX: Expanding Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning", "QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement learning"], "answer_arxiv_id": ["2008.01062", "2006.10800", "1905.05408"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_8054"} +{"question": "Any works about the development of YOLO series detectors?", "answer": ["You Only Look Once: Unified, Real-Time Object Detection", "YOLO9000: Better, Faster, Stronger", "YOLOv3: An Incremental Improvement", "YOLOv4: Optimal Speed and Accuracy of Object Detection", "RTMDet: An Empirical Study of Designing Real-Time Object Detectors", "YOLOX: Exceeding YOLO Series in 2021"], "answer_arxiv_id": ["1506.02640", "1612.08242", "1804.02767", "2004.10934", "2212.07784", "2107.08430"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_8055"} +{"question": "Which research papers discuss the emergence of foundation models in AI?", "answer": ["EVA: Exploring the Limits of Masked Visual Representation Learning at Scale", "Flamingo: a Visual Language Model for Few-Shot Learning", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "LLaMA: Open and Efficient Foundation Language Models", "ImageBind: One Embedding Space To Bind Them All", "The effectiveness of MAE pre-pretraining for billion-scale pretraining"], "answer_arxiv_id": ["2211.07636", "2204.14198", "2204.06125", "2302.13971", "2305.05665", "2303.13496"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_8056"} +{"question": "Could you provide me some works on the AI choreographer, generating the motion from music signals?", "answer": ["Learning to Generate Diverse Dance Motions with Transformer", "AI Choreographer: Music Conditioned 3D Dance Generation with AIST++", "Bailando: 3D Dance Generation by Actor-Critic GPT with Choreographic\n Memory"], "answer_arxiv_id": ["2008.08171", "2101.08779", "2203.13055"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_8057"} +{"question": "Which paper described In Context Learning (ICL) in synthetic settings and showed how transformers can function as complex classifiers through ICL?", "answer": ["What Can Transformers Learn In-Context? A Case Study of Simple Function Classes"], "answer_arxiv_id": ["2208.01066"], "source_meta": {"published_time": "20230117"}, "qid": "AutoScholarQuery_train_8058"} +{"question": "Which works proposed a method that uses a pseudo-label strategy in Source-Free Domain Adaptation (SFDA)?", "answer": ["Do We Really Need to Access the Source Data? Source Hypothesis Transfer\n for Unsupervised Domain Adaptation", "Source Data-absent Unsupervised Domain Adaptation through Hypothesis\n Transfer and Labeling Transfer"], "answer_arxiv_id": ["2002.08546", "2012.07297"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_8059"} +{"question": "What work entails about Physics Informed Neural Networks (PINN) that parameterizes the solution as a neural network?", "answer": ["The Deep Ritz method: A deep learning-based numerical algorithm for solving variational problems", "Unsupervised Deep Learning Algorithm for PDE-based Forward and Inverse Problems", "EikoNet: Solving the Eikonal equation with Deep Neural Networks"], "answer_arxiv_id": ["1710.00211", "1904.05417", "2004.00361"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_8060"} +{"question": "Which dataset is the only large-scale, publicly available, spatially-varying material dataset?", "answer": ["Single-Image SVBRDF Capture with a Rendering-Aware Deep Network"], "answer_arxiv_id": ["1810.09718"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_8061"} +{"question": "What papers address the computational challenges of MCMC and VI on large datasets in terms of time and space complexity by using Bayesian Coreset?", "answer": ["Sparse Variational Inference: Bayesian Coresets from Scratch", "Bayesian Coreset Construction via Greedy Iterative Geodesic Ascent", "Automated Scalable Bayesian Inference via Hilbert Coresets", "Stochastic Gradient Hamiltonian Monte Carlo", "Control Variates for Stochastic Gradient MCMC", "Speeding up MCMC by Efficient Data Subsampling"], "answer_arxiv_id": ["1906.03329", "1802.01737", "1710.05053", "1402.4102", "1706.05439", "1404.4178"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_8062"} +{"question": "Which studies use the diffusion model for the text-to-image generative task?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2204.06125"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_8063"} +{"question": "Which papers introduce open-vocabulary semantic segmentation approaches?", "answer": ["Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "Language-driven Semantic Segmentation", "RegionCLIP: Region-based Language-Image Pretraining", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP"], "answer_arxiv_id": ["2112.12143", "2201.03546", "2112.09106", "2210.04150"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_8064"} +{"question": "Could you provide the work that incorporated vision encoders as sensor modalities to a language model?", "answer": ["PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2303.03378"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_8065"} +{"question": "What are the works that support human-object contact estimation in images?", "answer": ["Detecting Human-Object Contact in Images"], "answer_arxiv_id": ["2303.03373"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_8066"} +{"question": "Which paper discusses using a VQ-VAE transformer-based method in the field of text-to-image generation?", "answer": ["Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_8067"} +{"question": "What papers have observed the disparate impact of privacy constraints on demographic subgroups in the field of differentially private fair learning?", "answer": ["Differential Privacy Has Disparate Impact on Model Accuracy", "Decision Making with Differential Privacy under a Fairness Lens"], "answer_arxiv_id": ["1905.12101", "2105.07513"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_8068"} +{"question": "What papers discussed training diffusion models on billions of image-text pairs?", "answer": ["LAION-5B: An open large-scale dataset for training next generation\n image-text models"], "answer_arxiv_id": ["2210.08402"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8069"} +{"question": "Which research introduced an approach to make self-contacts natural and generate pseudo ground-truth data?", "answer": ["On Self-Contact and Human Pose"], "answer_arxiv_id": ["2104.03176"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_8070"} +{"question": "In which studies was the convex notion of group CVaR introduced?", "answer": ["Fairness risk measures"], "answer_arxiv_id": ["1901.08665"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_8071"} +{"question": "Can you provide some examples of one-shot NAS methods?", "answer": ["DSNAS: Direct Neural Architecture Search without Parameter Retraining", "SNAS: stochastic neural architecture search", "ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware"], "answer_arxiv_id": ["2002.09128", "1812.09926", "1812.00332"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_8072"} +{"question": "What are some examples of diffusion models trained with large collections of images?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis"], "answer_arxiv_id": ["2112.10752", "2205.11487", "2204.06125", "2307.01952"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_8073"} +{"question": "What is the remedy proposed by studies for the situation where contrastive learning incorrectly assumes that for a given sample, every other sample in the dataset is dissimilar?", "answer": ["Debiased Contrastive Learning"], "answer_arxiv_id": ["2007.00224"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_8074"} +{"question": "Are there any works that evaluate LLM's ability in using multiple tools for solving challenging tasks?", "answer": ["Gorilla: Large Language Model Connected with Massive APIs", "On the Tool Manipulation Capability of Open-source Large Language Models"], "answer_arxiv_id": ["2305.15334", "2305.16504"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_8075"} +{"question": "Could you provide me some papers about soft prompting technique?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2104.08691", "2101.00190"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_8076"} +{"question": "Could you provide me examples of work that learn task-specific token-level guidance for LMs?", "answer": ["Toward Diverse Text Generation with Inverse Reinforcement Learning", "Long Text Generation via Adversarial Training with Leaked Information", "Unsupervised Text Style Transfer using Language Models as Discriminators"], "answer_arxiv_id": ["1804.11258v3", "1709.08624", "1805.11749"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_8077"} +{"question": "Which studies investigated robust mean teacher framework in the context of Test-time adaptation?", "answer": ["Robust Mean Teacher for Continual and Gradual Test-Time Adaptation"], "answer_arxiv_id": ["2211.13081"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_8078"} +{"question": "Can you name some studies about supervised learning approaches that frame the task as a kernel mean embedding classification problem? ", "answer": ["The Randomized Causation Coefficient", "Towards a Learning Theory of Cause-Effect Inference"], "answer_arxiv_id": ["1409.4366", "1502.02398"], "source_meta": {"published_time": "20220411"}, "qid": "AutoScholarQuery_train_8079"} +{"question": "Could you provide some research examples showing the application of diffusion models in creating content guidance from various sources?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions", "High-Resolution Image Synthesis with Latent Diffusion Models", "ImageBind: One Embedding Space To Bind Them All", "Zero-Shot Text-to-Image Generation", "Image Super-Resolution via Iterative Refinement", "Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2211.09800", "2112.10752", "2305.05665v2", "2102.12092", "2104.07636", "2210.09276"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_8080"} +{"question": "Are there any papers where scene improvisation was achieved through non-linear optimization?", "answer": ["Human-centric Indoor Scene Synthesis Using Stochastic Grammar"], "answer_arxiv_id": ["1808.08473"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_8081"} +{"question": "Where can I find research about MoE models based on the GPT family of models?", "answer": ["Mixtral of Experts"], "answer_arxiv_id": ["2401.04088"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_8082"} +{"question": "Which papers introduce techniques related to layout generation?", "answer": ["LayoutGAN: Generating Graphic Layouts with Wireframe Discriminators", "LayoutVAE: Stochastic Scene Layout Generation From a Label Set", "Variational Transformer Networks for Layout Generation", "LayoutTransformer: Layout Generation and Completion with Self-attention"], "answer_arxiv_id": ["1901.06767", "1907.10719", "2104.02416v1", "2006.14615"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_8083"} +{"question": "Could you provide a study about simplifying the exact matching in previous KL divergence loss during knowledge distillation?", "answer": ["Knowledge Distillation from A Stronger Teacher"], "answer_arxiv_id": ["2205.10536"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8084"} +{"question": "Which datasets provide large-extent coverage but with low-resolution annotations?", "answer": ["BigEarthNet-MM: A Large Scale Multi-Modal Multi-Label Benchmark Archive for Remote Sensing Image Classification and Retrieval", "SEN12MS – A Curated Dataset of Georeferenced Multi-Spectral Sentinel-1/2 Imagery for Deep Learning and Data Fusion"], "answer_arxiv_id": ["2105.07921", "1906.07789"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_8085"} +{"question": "Which libraries excel in areas of hyperparameter sampling, input warping and parallel optimization in Bayesian Optimization?", "answer": ["Practical Bayesian Optimization of Machine Learning Algorithms"], "answer_arxiv_id": ["1206.2944"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_8086"} +{"question": "Which studies proposed solutions employing compression and similarity for minimizing distributed problems?", "answer": ["Permutation Compressors for Provably Faster Distributed Nonconvex Optimization", "Compression and Data Similarity: Combination of Two Techniques for Communication-Efficient Solving of Distributed Variational Inequalities"], "answer_arxiv_id": ["2110.03300", "2206.09446v2"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_8087"} +{"question": "Could you provide me some studies developing an operator network from control variables to solutions of PDEs or objective functions?", "answer": ["Amortized Synthesis of Constrained Configurations Using a Differentiable Surrogate", "Learning Composable Energy Surrogates for PDE Order Reduction"], "answer_arxiv_id": ["2106.09019", "2005.06549"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_8088"} +{"question": "What research papers discuss optimization of LMs with reinforcement learning techniques?", "answer": ["Fine-Tuning Language Models from Human Preferences", "Recursively Summarizing Books with Human Feedback", "Quark: Controllable Text Generation with Reinforced [Un]learning", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["1909.08593", "2109.10862", "2205.13636", "2203.02155"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_8089"} +{"question": "Could you mention some works about expandable networks for class-incremental learning?", "answer": ["DER: Dynamically Expandable Representation for Class Incremental\n Learning", "FOSTER: Feature Boosting and Compression for Class-Incremental Learning", "DyTox: Transformers for Continual Learning with DYnamic TOken eXpansion", "Dynamic Residual Classifier for Class Incremental Learning", "Dense Network Expansion for Class Incremental Learning", "Resolving Task Confusion in Dynamic Expansion Architectures for Class\n Incremental Learning"], "answer_arxiv_id": ["2103.16788", "2204.04662", "2111.11326", "2308.13305", "2303.12696", "2212.14284"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_8090"} +{"question": "What are some of the studies that explored criteria for hardness, anticurriculum, pacing functions, and mixing rates?", "answer": ["Curriculum Learning: A Survey"], "answer_arxiv_id": ["2101.10382"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_8091"} +{"question": "What works used two-part encoder to avoid homological obstruction?", "answer": ["Spherical Latent Spaces for Stable Variational Autoencoders", "Spherical Text Embedding"], "answer_arxiv_id": ["1808.10805", "1911.01196"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_8092"} +{"question": "Could you tell me which research set the threshold for flat clipping as privately estimated quantile of gradient norms?", "answer": ["Differentially Private Learning with Adaptive Clipping"], "answer_arxiv_id": ["1905.03871"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_8093"} +{"question": "Are there any research papers that derives confidence bands allowing the distribution to change over time in a data-dependent manner, remaining time-uniform and applicable to off-policy problems in contextual bandits?", "answer": ["Sequential estimation of quantiles with applications to A/B testing and best-arm identification", "Anytime-valid off-policy inference for contextual bandits"], "answer_arxiv_id": ["1906.09712", "2210.10768"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_8094"} +{"question": "Which studies propose dense object descriptors for robotic manipulation?", "answer": ["Dense Object Nets: Learning Dense Visual Object Descriptors By and For\n Robotic Manipulation"], "answer_arxiv_id": ["1806.08756"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_8095"} +{"question": "What research provides insights on whether or not it should be possible to grant copyrights to AI-authored works?", "answer": ["Extracting Training Data from Diffusion Models"], "answer_arxiv_id": ["2301.13188"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_8096"} +{"question": "Which research has tried to train a neural network model for procedural graph extraction?", "answer": ["PET: An Annotated Dataset for Process Extraction from Natural Language\n Text"], "answer_arxiv_id": ["2203.04860"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_8097"} +{"question": "What paper describes MDM, a method from the category of diffusion models used for human motion generation?", "answer": ["Human Motion Diffusion Model"], "answer_arxiv_id": ["2209.14916"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_8098"} +{"question": "Which studies explore Arbitrary-Scale Super-Resolution with a single network?", "answer": ["Meta-SR: A Magnification-Arbitrary Network for Super-Resolution", "Learning Continuous Image Representation with Local Implicit Image\n Function", "Implicit Transformer Network for Screen Content Image Continuous\n Super-Resolution", "Local Texture Estimator for Implicit Representation Function"], "answer_arxiv_id": ["1903.00875", "2012.09161", "2112.06174", "2111.08918"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_8099"} +{"question": "In what papers track multiple objects using multiple object tracking (MOT) models?", "answer": ["MOT16: A Benchmark for Multi-Object Tracking"], "answer_arxiv_id": ["1603.00831"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_8100"} +{"question": "Could you provide me some works on the inference for dynamic neural networks?", "answer": ["Cortex: A Compiler for Recursive Deep Learning Models", "Just-in-Time Dynamic-Batching"], "answer_arxiv_id": ["2011.01383v2", "1904.07421"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_8101"} +{"question": "What papers include a two-stage reranking system using sequence-to-sequence generation models?", "answer": ["BRIO: Bringing Order to Abstractive Summarization"], "answer_arxiv_id": ["2203.16804"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_8102"} +{"question": "What research papers evaluated LLM-generated summaries of medical evidence?", "answer": ["Summarizing, Simplifying, and Synthesizing Medical Evidence Using GPT-3\n (with Varying Success)"], "answer_arxiv_id": ["2305.06299"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_8103"} +{"question": "What studies utilized instruction-following tasks as a testbed for language grounding?", "answer": ["Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments", "Room-Across-Room: Multilingual Vision-and-Language Navigation with Dense Spatiotemporal Grounding", "ALFRED A Benchmark for Interpreting Grounded Instructions for Everyday Tasks"], "answer_arxiv_id": ["1711.07280", "2010.07954", "1912.01734"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_8104"} +{"question": "What studies did research on metric learning methods in the Few-Shot Class-Incremental Learning paradigm?", "answer": ["MgSvF: Multi-Grained Slow vs. Fast Framework for Few-Shot\n Class-Incremental Learning", "Semantic-aware Knowledge Distillation for Few-Shot Class-Incremental\n Learning", "Forward Compatible Few-Shot Class-Incremental Learning"], "answer_arxiv_id": ["2006.15524", "2103.04059", "2203.06953"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_8105"} +{"question": "Are there any works demonstrating that some algorithms may converge to stationary points which are not saddle points?", "answer": ["The Limit Points of (Optimistic) Gradient Descent in Min-Max Optimization"], "answer_arxiv_id": ["1807.03907"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_8106"} +{"question": "In what works have Transformers been extended to solve vision-related tasks such as object detection?", "answer": ["End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2005.12872"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8107"} +{"question": "What works in offline RL are most similar to this work and what they focus on?", "answer": ["Offline Reinforcement Learning with Implicit Q-Learning", "Conservative Q-Learning for Offline Reinforcement Learning", "Behavior Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["2110.06169", "2006.04779", "1911.11361"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_8108"} +{"question": "Which works have proposed neural networks to learn Lyapunov functions?", "answer": ["Learning Stable Deep Dynamics Models", "Learning Dynamics Models with Stable Invariant Sets"], "answer_arxiv_id": ["2001.06116", "2006.08935"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_8109"} +{"question": "What studies involved using multi-view volumetric capture systems to collect datasets of humans in real-world clothing?", "answer": ["Panoptic Studio: A Massively Multiview System for Social Interaction\n Capture"], "answer_arxiv_id": ["1612.03153"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_8110"} +{"question": "Could you provide me some works about GAN-based unpaired image-to-image translations that use two-side way translation?", "answer": ["Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks", "DualGAN: Unsupervised Dual Learning for Image-to-Image Translation", "Learning to Discover Cross-Domain Relations with Generative Adversarial Networks", "SCAN: Learning to Classify Images without Labels", "U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation"], "answer_arxiv_id": ["1703.10593", "1704.02510", "1703.05192", "2005.12320", "1907.10830"], "source_meta": {"published_time": "20230804"}, "qid": "AutoScholarQuery_train_8111"} +{"question": "Which work revealed the overconfidence issue of neural networks in out-of-distribution data?", "answer": ["Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images"], "answer_arxiv_id": ["1412.1897"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_8112"} +{"question": "In what studies has the functioning of S4 models in a largely offline RL setting been investigated?", "answer": ["Decision S4: Efficient Sequence-Based RL via State Spaces Layers"], "answer_arxiv_id": ["2306.05167"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_8113"} +{"question": "Which works demonstrate impressive performance in zero-shot retrieval and classification with contrastive pre-training models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_8114"} +{"question": "Which papers introduced integration of SAM with video object segmentation approaches for interactive tracking of objects?", "answer": ["Tracking Anything with Decoupled Video Segmentation", "Track Anything: Segment Anything Meets Videos", "Segment and Track Anything"], "answer_arxiv_id": ["2309.03903", "2304.11968", "2305.06558"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_8115"} +{"question": "In what studies is a Class-Incremental Learning (CIL) problem addressed?", "answer": ["iCaRL: Incremental Classifier and Representation Learning"], "answer_arxiv_id": ["1611.07725"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_8116"} +{"question": "Which papers discuss the guidance of training language models with human-preference annotations towards the “Helpful, Honest, and Harmless” (3H) objectives?", "answer": ["A General Language Assistant as a Laboratory for Alignment", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback"], "answer_arxiv_id": ["2112.00861", "2204.05862"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_8117"} +{"question": "Could you provide me some works that used ensemble method for debiasing?", "answer": ["Learning Debiased Classifier with Biased Committee", "Evading the Simplicity Bias: Training a Diverse Set of Models Discovers Solutions with Superior OOD Generalization"], "answer_arxiv_id": ["2206.10843", "2105.05612"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_8118"} +{"question": "Which work introduced a backward accumulation technique which is more robust to occlusions than the forward strategies?", "answer": ["AccFlow: Backward Accumulation for Long-Range Optical Flow"], "answer_arxiv_id": ["2308.13133"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_8119"} +{"question": "Which works are about constructing different hierarchies for hierarchical graph generation applied to molecule generation?", "answer": ["Hierarchical Generation of Molecular Graphs using Structural Motifs", "Hierarchical Graph-to-Graph Translation for Molecules", "Multiscale Planar Graph Generation", "HDMapGen: A Hierarchical Graph Generative Model of High Definition Maps", "De Novo Molecular Generation via Connection-aware Motif Mining", "MolGrow: A Graph Normalizing Flow for Hierarchical Molecular Generation"], "answer_arxiv_id": ["2002.03230", "1907.11223", "1802.09617", "2106.14880", "2302.01129", "2106.05856"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_train_8120"} +{"question": "What papers used fragment-based methods or pre-trained models for generating more realistic molecules in the field of structure-based drug design?", "answer": ["Zero-Shot 3D Drug Design by Sketching and Generating"], "answer_arxiv_id": ["2209.13865"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_8121"} +{"question": "Who introduced Denoising Diffusion Probabilistic Models (DDPMs) as a powerful class of generative models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_8122"} +{"question": "Who proposed that visual quality depends on explanations lying on the data manifold?", "answer": ["The Manifold Hypothesis for Gradient-Based Explanations"], "answer_arxiv_id": ["2206.07387"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8123"} +{"question": "Could you provide me some studies on spectral-based methods for shape correspondence?", "answer": ["Self-Supervised Learning for Multimodal Non-Rigid 3D Shape Matching"], "answer_arxiv_id": ["2303.10971"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_8124"} +{"question": "Which works discuss the use of diffusion models in text-to-motion generative systems?", "answer": ["Human Motion Diffusion Model", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model"], "answer_arxiv_id": ["2209.14916", "2208.15001"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_8125"} +{"question": "Which studies are essentially focused on reducing the inference time in DNNs quantization?", "answer": ["Post training 4-bit quantization of convolutional networks for rapid-deployment", "ZeroQ: A Novel Zero Shot Quantization Framework", "Low-bit Quantization of Neural Networks for Efficient Inference", "Post-Training Piecewise Linear Quantization for Deep Neural Networks", "Confounding Tradeoffs for Neural Network Quantization", "Improving Neural Network Quantization without Retraining using Outlier Channel Splitting", "Data-Free Quantization Through Weight Equalization and Bias Correction", "SQuant: On-the-Fly Data-Free Quantization via Diagonal Hessian Approximation"], "answer_arxiv_id": ["1810.05723", "2001.00281", "1902.06822", "2002.00104", "2102.06366", "1901.09504", "1906.04721", "2202.07471"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_8126"} +{"question": "Which work introduced the concept of contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_8127"} +{"question": "Which works initially proposed prompting to modify the input text string for adaption to new tasks without labeled data?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2107.13586v1"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_8128"} +{"question": "Which studies propose ways to handle the reading process in an efficient manner by manipulating the context length or decoder's attention?", "answer": ["FiD-Light: Efficient and Effective Retrieval-Augmented Text Generation", "FiDO: Fusion-in-Decoder optimized for stronger performance and faster\n inference"], "answer_arxiv_id": ["2209.14290", "2212.08153"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_8129"} +{"question": "Could you provide works that focused on visual place recognition (VPR)?", "answer": ["Visual Place Recognition: A Tutorial", "NetVLAD: CNN architecture for weakly supervised place recognition", "MixVPR: Feature Mixing for Visual Place Recognition", "SeqNet: Learning Descriptors for Sequence-based Hierarchical Place\n Recognition", "Collaborative Visual Place Recognition", "PointNetVLAD: Deep Point Cloud Based Retrieval for Large-Scale Place\n Recognition"], "answer_arxiv_id": ["2303.03281", "1511.07247", "2303.02190", "2102.11603", "2310.05541", "1804.03492"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_8130"} +{"question": "Are there studies that have researched the cultural bias present in large language models?", "answer": ["Having Beer after Prayer? Measuring Cultural Bias in Large Language\n Models", "Towards Measuring the Representation of Subjective Global Opinions in\n Language Models"], "answer_arxiv_id": ["2305.14456", "2306.16388"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_8131"} +{"question": "What papers discussed re-training the classifier from scratch or normalizing the classifier weights with class-balanced sampling?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition"], "answer_arxiv_id": ["1910.09217"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_8132"} +{"question": "What papers are about learning from states and rewards rolled out from the expert policies?", "answer": ["Recent Advances in Imitation Learning from Observation", "State-Only Imitation Learning for Dexterous Manipulation", "Imitation Learning from Observations by Minimizing Inverse Dynamics Disagreement", "Transformers for One-Shot Visual Imitation"], "answer_arxiv_id": ["1905.13566", "2004.04650", "1910.04417", "2011.05970"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_8133"} +{"question": "What works devise a transformer-based framework that uses language to guide the extraction of discriminative visual features when encoding images and sentences?", "answer": ["Visual Grounding with Transformers"], "answer_arxiv_id": ["2105.04281"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_8134"} +{"question": "Could you provide me some studies which applied the ensemble learning method for domain generalization?", "answer": ["Best sources forward: domain generalization through source-specific nets", "DART: Diversify-Aggregate-Repeat Training Improves Generalization of\n Neural Networks", "DoFE: Domain-oriented Feature Embedding for Generalizable Fundus Image\n Segmentation on Unseen Datasets", "Domain Generalization with Domain-Specific Aggregation Modules"], "answer_arxiv_id": ["1806.05810v1", "2302.14685", "2010.06208", "1809.10966"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_8135"} +{"question": "Which research efforts utilized diffusion models for Out of Distribution (OOD) detection?", "answer": ["Denoising diffusion models for out-of-distribution detection", "Unsupervised Out-of-Distribution Detection with Diffusion Inpainting"], "answer_arxiv_id": ["2211.07740", "2302.10326"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_8136"} +{"question": "Which studies propose vision-based gaze collection systems that define gaze as direction vectors originating from facial centers towards specific gaze targets?", "answer": ["MPIIGaze: Real-World Dataset and Deep Appearance-Based Gaze Estimation", "ETH-XGaze: A Large Scale Dataset for Gaze Estimation under Extreme Head\n Pose and Gaze Variation"], "answer_arxiv_id": ["1711.09017", "2007.15837"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_8137"} +{"question": "Which work proposed a communication-efficient distributed GNN training technique (LLCG)?", "answer": ["Learn Locally, Correct Globally: A Distributed Algorithm for Training Graph Neural Networks"], "answer_arxiv_id": ["2111.08202"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_8138"} +{"question": "Which state-of-the-art text-to-image models have been used for generating high-quality, photorealistic images?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_8139"} +{"question": "Could you provide me some works that proposed to explicitly model 3D independent flow field for dynamical objects?", "answer": ["Depth Prediction Without the Sensors: Leveraging Structure for Unsupervised Learning from Monocular Videos", "Depth from Videos in the Wild: Unsupervised Monocular Depth Learning from Unknown Cameras", "Learning Monocular Depth in Dynamic Scenes viaInstance-Aware Projection Consistency", "Attentive and Contrastive Learning for Joint Depth and Motion Field Estimation", "Instance-aware multi-object self-supervision for monocular depth prediction", "Dyna-DM: Dynamic Object-aware Self-supervised Monocular Depth Maps"], "answer_arxiv_id": ["1811.06152", "1904.04998", "2102.02629", "2110.06853", "2203.00809", "2206.03799"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_8140"} +{"question": "Which study used MLPs to model both the occupancy value and color of the human body from one or several images?", "answer": ["PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization"], "answer_arxiv_id": ["1905.05172"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_8141"} +{"question": "Which studies extended binary PU classification to the multi-label setting and modeled MLPUL as cost-sensitive learning?", "answer": ["A Unified Positive-Unlabeled Learning Framework for Document-Level\n Relation Extraction with Different Levels of Labeling"], "answer_arxiv_id": ["2210.08709"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_8142"} +{"question": "Which research adopt the approach that leverages temporal knowledge graph embeddings for semantic similarity assessments in answer determination?", "answer": ["Question Answering Over Temporal Knowledge Graphs"], "answer_arxiv_id": ["2106.01515"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_8143"} +{"question": "Which study suggests fine-tuning LLMs using pairs consisting of errors and their respective corrections?", "answer": ["Learning From Mistakes Makes LLM Better Reasoner"], "answer_arxiv_id": ["2310.20689"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_8144"} +{"question": "What works employ reinforcement learning for NAS and HPO in GL?", "answer": ["Auto-GNN: Neural Architecture Search of Graph Neural Networks", "Policy-GNN: Aggregation Optimization for Graph Neural Networks"], "answer_arxiv_id": ["1909.03184", "2006.15097"], "source_meta": {"published_time": "20220618"}, "qid": "AutoScholarQuery_train_8145"} +{"question": "Could you provide me with studies focused on the application of text to drive perpetual view generation methods?", "answer": ["SceneScape: Text-Driven Consistent Scene Generation"], "answer_arxiv_id": ["2302.01133"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_8146"} +{"question": "Which works focus on mathematical reasoning in natural language?", "answer": ["Measuring Mathematical Problem Solving With the MATH Dataset", "Solving Quantitative Reasoning Problems with Language Models", "NaturalProofs: Mathematical Theorem Proving in Natural Language", "NaturalProver: Grounded Mathematical Proof Generation with Language Models", "Lila: A Unified Benchmark for Mathematical Reasoning", "A Survey of Deep Learning for Mathematical Reasoning", "A Survey in Mathematical Language Processing"], "answer_arxiv_id": ["2103.03874", "2206.14858", "2104.01112", "2205.12910", "2210.17517", "2212.10535", "2205.15231"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_8147"} +{"question": "Any works about leveraging Large Language Models to improve reasoning over Knowledge Bases?", "answer": ["Don't Generate, Discriminate: A Proposal for Grounding Language Models\n to Real-World Environments", "StructGPT: A General Framework for Large Language Model to Reason over\n Structured Data", "Think-on-Graph: Deep and Responsible Reasoning of Large Language Model\n on Knowledge Graph"], "answer_arxiv_id": ["2212.09736", "2305.09645", "2307.07697"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_8148"} +{"question": "Could you indicate some works where contrastive learning has been applied on images?", "answer": ["Improved Baselines with Momentum Contrastive Learning", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2003.04297", "1911.05722"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_8149"} +{"question": "Could you provide me some studies about query rewriting using a specially fine-tuned model?", "answer": ["Query2doc: Query Expansion with Large Language Models", "Query Rewriting for Retrieval-Augmented Large Language Models"], "answer_arxiv_id": ["2303.07678", "2305.14283"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_8150"} +{"question": "Could you provide me any work that discusses the phenomenon of posterior collapse?", "answer": ["Fixing a Broken ELBO", "Don’t Blame the ELBO! A Linear VAE Perspective on Posterior Collapse"], "answer_arxiv_id": ["1711.00464", "1911.02469"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8151"} +{"question": "Which work presented the Denoising Diffusion Implicit Model (DDIM) and demonstrated its equivalence to the probability flow?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_8152"} +{"question": "What studies have focused on training neural networks for landmark-based 2D/3D registration?", "answer": ["Automatic Annotation of Hip Anatomy in Fluoroscopy for Robust and\n Efficient 2D/3D Registration", "Deep Iterative 2D/3D Registration"], "answer_arxiv_id": ["1911.07042", "2107.10004"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8153"} +{"question": "Which studies proposed leveraging pretrained models in one domain and aligning the embeddings of another domain?", "answer": ["LiT: Zero-Shot Transfer with Locked-image text Tuning"], "answer_arxiv_id": ["2111.07991"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_8154"} +{"question": "Could you list some studies that attempted to solve complex tasks by initially generating a plan and then executing the relevant APIs?", "answer": ["HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face", "Chameleon: Plug-and-Play Compositional Reasoning with Large Language Models"], "answer_arxiv_id": ["2303.17580", "2304.09842"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_8155"} +{"question": "Which studies suffer from noisy annotations or long-tailed predicate distribution in the field of Visual Relation Detection (VRD)?", "answer": ["The Devil is in the Labels: Noisy Label Correction for Robust Scene Graph Generation", "Unbiased Scene Graph Generation from Biased Training"], "answer_arxiv_id": ["2206.03014", "2002.11949"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_8156"} +{"question": "Which works use pre-trained generative adversarial networks in the field of generative prior for restoration?", "answer": ["Generative Adversarial Nets", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis"], "answer_arxiv_id": ["1406.2661", "1812.04948", "1809.11096"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_8157"} +{"question": "Which studies propose sample efficient no-regret algorithms for MDPs with linear features?", "answer": ["Reinforcement Learning in Feature Space: Matrix Bandit, Kernels, and Regret Bound", "Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1905.10389", "1907.05388"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_8158"} +{"question": "What research papers have been dedicated to the theoretical study of data augmentation in sample-contrastive learning?", "answer": ["Towards the Generalization of Contrastive Self-Supervised Learning", "Toward Understanding the Feature Learning Process of Self-supervised Contrastive Learning"], "answer_arxiv_id": ["2111.00743", "2105.15134"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_8159"} +{"question": "What research provides insight into visual concept learning through the use of generalizable properties and neuro-symbolic programs or embeddings?", "answer": ["FALCON: Fast Visual Concept Learning by Integrating Images, Linguistic\n descriptions, and Conceptual Relations", "Unsupervised Compositional Concepts Discovery with Text-to-Image\n Generative Models"], "answer_arxiv_id": ["2203.16639", "2306.05357"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_8160"} +{"question": "What are the studies that have used butterfly matrices in kernel methods and deep learning methods?", "answer": ["Quadrature-based features for kernel approximation", "Butterfly Transform: An Efficient FFT Based Neural Architecture Design", "Deformable Butterfly: A Highly Structured and Sparse Linear Transform", "Sparse Linear Networks with a Fixed Butterfly Structure: Theory and Practice"], "answer_arxiv_id": ["1802.03832", "1906.02256", "2203.13556", "2007.08864"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_8161"} +{"question": "Which papers discuss recent advancements in the research topic of multimodal learning?", "answer": ["Multimodal Machine Learning: A Survey and Taxonomy", "Vision+X: A Survey on Multimodal Learning in the Light of Data"], "answer_arxiv_id": ["1705.09406", "2210.02884"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_8162"} +{"question": "Which studies have examined the effects of utilizing Group Normalization instead of Batch Normalization in federated learning?", "answer": ["The Non-IID Data Quagmire of Decentralized Machine Learning", "Federated Visual Classification with Real-World Data Distribution"], "answer_arxiv_id": ["1910.00189v2", "2003.08082"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_8163"} +{"question": "Which reference introduced the Vimeo90K dataset, a widely-used dataset for training and evaluating VFI methods?", "answer": ["Video Enhancement with Task-Oriented Flow"], "answer_arxiv_id": ["1711.09078"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_8164"} +{"question": "Which works study the method of domain invariant learning for domain generalization?", "answer": ["Domain Generalization via Invariant Feature Representation", "Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization", "Unified Deep Supervised Domain Adaptation and Generalization", "Learning to Optimize Domain Specific Normalization for Domain Generalization", "A Bit More Bayesian: Domain-Invariant Learning with Uncertainty", "Domain Generalization using Causal Matching", "Domain Invariant Representation Learning with Domain Density Transformations", "On Learning Domain-Invariant Representations for Transfer Learning with Multiple Sources", "Gradient Matching for Domain Generalization"], "answer_arxiv_id": ["1301.2115", "1510.04373v2", "1709.10190", "1907.04275", "2105.04030", "2006.07500", "2102.05082", "2111.13822", "2104.09937"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_8165"} +{"question": "What studies proposed proposal-free methods for object captioning tasks?", "answer": ["MDETR - Modulated Detection for End-to-End Multi-Modal Understanding", "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision"], "answer_arxiv_id": ["2104.12763", "2102.03334"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_8166"} +{"question": "Which works presented the introduction of INR in the field of novel view synthesis for static scenes?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Few-View Object Reconstruction with Unknown Categories and Camera Poses", "TensoRF: Tensorial Radiance Fields", "NoPe-NeRF: Optimising Neural Radiance Field with No Pose Prior"], "answer_arxiv_id": ["2103.13415", "2201.05989", "2212.04492", "2203.09517", "2212.07388"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_8167"} +{"question": "Could you name some studies that implemented embedding-based methods in unsupervised anomaly detectors?", "answer": ["Towards Total Recall in Industrial Anomaly Detection", "Fully Convolutional Cross-Scale-Flows for Image-based Defect Detection", "Anomaly Detection via Reverse Distillation from One-Class Embedding", "SimpleNet: A Simple Network for Image Anomaly Detection and Localization"], "answer_arxiv_id": ["2106.08265", "2110.02855", "2201.10703", "2303.15140"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_8168"} +{"question": "Which works propose knowledge editing methods that predict updates to the weights of the base model by knowledge locating or meta-learning?", "answer": ["Fast Model Editing at Scale", "Mass-Editing Memory in a Transformer"], "answer_arxiv_id": ["2110.11309", "2210.07229"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_8169"} +{"question": "Which paper discusses the use of the CLIPScore metric for reference-free evaluation of image-text alignment in automated T2I?", "answer": ["CLIPScore: A Reference-free Evaluation Metric for Image Captioning"], "answer_arxiv_id": ["2104.08718"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_8170"} +{"question": "Can you provide some works that used diffusion models for generating motions conditioned on textural input and spatial data?", "answer": ["Human Motion Diffusion Model", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model"], "answer_arxiv_id": ["2209.14916", "2208.15001"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_8171"} +{"question": "What research analyzed the convergence of an online gradient estimator under hysteresis?", "answer": ["Convergence of Online Adaptive and Recurrent Optimization Algorithms"], "answer_arxiv_id": ["2005.05645"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_8172"} +{"question": "Which studies proposed extensions focusing on the design of forward corruption kernels for the discrete diffusion model?", "answer": ["Argmax Flows and Multinomial Diffusion: Learning Categorical Distributions", "Structured Denoising Diffusion Models in Discrete State-Spaces", "Autoregressive Diffusion Models", "Beyond In-Place Corruption: Insertion and Deletion In Denoising Probabilistic Models"], "answer_arxiv_id": ["2102.05379", "2107.03006", "2110.02037", "2107.07675v1"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_8173"} +{"question": "Could you provide me some references about the studies on strategic adaptation by users under a classifier?", "answer": ["Strategic Classification"], "answer_arxiv_id": ["1506.06980"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_8174"} +{"question": "Is there any research that might investigate when to stop measuring in active sensing?", "answer": ["Balancing Suspense and Surprise: Timely Decision Making with Endogenous Information Acquisition"], "answer_arxiv_id": ["1610.07505"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_8175"} +{"question": "What research has been conducted on RL-tailored fusion schemes with respect to data augmentation?", "answer": ["Learning to Compose Domain-Specific Transformations for Data Augmentation", "RandAugment: Practical automated data augmentation with a reduced search space", "AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "AugMax: Adversarial Composition of Random Augmentations for Robust Training", "UniformAugment: A Search-free Probabilistic Data Augmentation Approach", "TrivialAugment: Tuning-free Yet State-of-the-Art Data Augmentation"], "answer_arxiv_id": ["1709.01643", "1909.13719", "1912.02781", "2110.13771", "2003.14348", "2103.10158"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8176"} +{"question": "Could you provide me some works that explored deep reinforcement learning (RL) based search algorithms?", "answer": ["Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients"], "answer_arxiv_id": ["1912.04871"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_8177"} +{"question": "Could you list some models that have explored multimodal in-context learning and the management of interleaved text and image examples?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "OpenFlamingo: An Open-Source Framework for Training Large Autoregressive\n Vision-Language Models", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "M$^3$IT: A Large-Scale Dataset towards Multi-Modal Multilingual\n Instruction Tuning", "MetaVL: Transferring In-Context Learning Ability From Language Models to\n Vision-Language Models", "Sparkles: Unlocking Chats Across Multiple Images for Multimodal\n Instruction-Following Models", "MMICL: Empowering Vision-language Model with Multi-Modal In-Context\n Learning"], "answer_arxiv_id": ["2204.14198", "2308.01390", "2305.03726", "2306.04387", "2306.01311", "2308.16463", "2309.07915"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_8178"} +{"question": "Which studies present taxonomies of Vision-Language models?", "answer": ["ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision"], "answer_arxiv_id": ["2102.03334"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_8179"} +{"question": "What papers apply similar techniques to linear MDPs?", "answer": ["Online Learning in MDPs with Linear Function Approximation and Bandit Feedback", "Policy Optimization in Adversarial MDPs: Improved Exploration via Dilated Bonuses", "Refined Regret for Adversarial MDPs with Linear Function Approximation", "Improved Regret for Efficient Online Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["2007.01612", "2107.08346", "2301.12942", "2301.13087"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_8180"} +{"question": "Which works presented federated learning frameworks focusing on reducing communication bandwidth by quantizing local updates?", "answer": ["Bitwidth Heterogeneous Federated Learning with Progressive Weight\n Dequantization", "DAdaQuant: Doubly-adaptive quantization for communication-efficient\n Federated Learning", "FedPAQ: A Communication-Efficient Federated Learning Method with\n Periodic Averaging and Quantization"], "answer_arxiv_id": ["2202.11453", "2111.00465", "1909.13014"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_8181"} +{"question": "Which papers are about learning-based approaches for providing more generalizable magnification?", "answer": ["Learning-based Video Motion Magnification", "Unsupervised Behaviour Analysis and Magnification (uBAM) using Deep\n Learning"], "answer_arxiv_id": ["1804.02684", "2012.09237"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8182"} +{"question": "Any works about real-world image denoising?", "answer": ["Natural Image Noise Dataset", "NBNet: Noise Basis Learning for Image Denoising with Subspace Projection", "Transfer Learning from Synthetic to Real-Noise Denoising with Adaptive\n Instance Normalization", "Invertible Denoising Network: A Light Solution for Real Noise Removal", "Variational Denoising Network: Toward Blind Noise Modeling and Removal", "Dual Adversarial Network: Toward Real-world Noise Removal and Noise\n Generation", "CVF-SID: Cyclic multi-Variate Function for Self-Supervised Image\n Denoising by Disentangling Noise from Image", "AP-BSN: Self-Supervised Denoising for Real-World Images via Asymmetric\n PD and Blind-Spot Network"], "answer_arxiv_id": ["1906.00270", "2012.15028", "2002.11244", "2104.10546", "1908.11314", "2007.05946", "2203.13009", "2203.11799"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_8183"} +{"question": "Which works validated the innovation of disentangled attention by outperforming previous methods in NLU benchmarks?", "answer": ["SuperGLUE: A Stickier Benchmark for General-Purpose Language\n Understanding Systems"], "answer_arxiv_id": ["1905.00537"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_8184"} +{"question": "What papers propose methods for multi-view refinement using flow or dense features in Structure from Motion (SfM)?", "answer": ["Multi-View Optimization of Local Feature Geometry", "Pixel-Perfect Structure-from-Motion with Featuremetric Refinement"], "answer_arxiv_id": ["2003.08348", "2108.08291"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_8185"} +{"question": "Are there any works about the application of active learning algorithms to deep learning?", "answer": ["Adversarial Active Learning for Deep Networks: a Margin Based Approach", "Gone Fishing: Neural Active Learning with Fisher Embeddings"], "answer_arxiv_id": ["1802.09841", "2106.09675"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_8186"} +{"question": "Which studies presented algorithms for multi-agent RL with convergence guarantees such as federated version of TD and Q-learning, and policy gradient with fault tolerance?", "answer": ["Fully Decentralized Multi-Agent Reinforcement Learning with Networked Agents", "Multi-Agent Reinforcement Learning via Double Averaging Primal-Dual Optimization", "Federated Reinforcement Learning: Linear Speedup Under Markovian Sampling", "Fault-Tolerant Federated Reinforcement Learning with Theoretical Guarantee"], "answer_arxiv_id": ["1802.08757v2", "1806.00877", "2206.10185", "2110.14074"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_8187"} +{"question": "What works discussed the limitations of SMILES representation in molecule generation and proposed graph-based methods instead?", "answer": ["Junction Tree Variational Autoencoder for Molecular Graph Generation", "Constrained Graph Variational Autoencoders for Molecule Design", "GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation", "Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation", "Optimization of Molecules via Deep Reinforcement Learning"], "answer_arxiv_id": ["1802.04364", "1805.09076", "2001.09382", "1806.02473", "1810.08678"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_8188"} +{"question": "What study proposed a reconciling method for error accumulation problem, known as 'pushforward trick'?", "answer": ["Message Passing Neural PDE Solvers"], "answer_arxiv_id": ["2202.03376"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_8189"} +{"question": "Which papers make extensions for conventional convolution and different network architectures to manage free-form inpainting mask?", "answer": ["Image Inpainting for Irregular Holes Using Partial Convolutions", "Free-Form Image Inpainting with Gated Convolution", "Resolution-robust Large Mask Inpainting with Fourier Convolutions"], "answer_arxiv_id": ["1804.07723", "1806.03589", "2109.07161"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_8190"} +{"question": "What papers discussed the architectures of ResNet and EfficientNet?", "answer": ["Deep Residual Learning for Image Recognition", "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks"], "answer_arxiv_id": ["1512.03385", "1905.11946"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_8191"} +{"question": "Could you give me examples of work in which/Federated Learning was applied in machine translation?", "answer": ["Communication-Efficient Federated Learning for Neural Machine Translation", "Training Mixed-Domain Translation Models via Federated Learning"], "answer_arxiv_id": ["2112.06135", "2205.01557"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_8192"} +{"question": "What research include the use of internal language models in language-aware scene text recognition?", "answer": ["Focusing Attention: Towards Accurate Text Recognition in Natural Images"], "answer_arxiv_id": ["1709.02054"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_8193"} +{"question": "Which work used multiple views of a dynamic scene to learn object-centric features?", "answer": ["Object-Centric Representation Learning with Generative Spatial-Temporal Factorization"], "answer_arxiv_id": ["2111.05393"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_8194"} +{"question": "Which studies focused on pose estimation for humans?", "answer": ["CrowdPose: Efficient Crowded Scenes Pose Estimation and A New Benchmark"], "answer_arxiv_id": ["1812.00324"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_8195"} +{"question": "Could you name the study which utilized negative pairs repulsing sample in a batch to avoid solution collapses?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_8196"} +{"question": "Which papers adopt a direct regression approach for category-agnostic sparse view camera pose estimation?", "answer": ["Relative Camera Pose Estimation Using Convolutional Neural Networks", "The 8-Point Algorithm as an Inductive Bias for Relative Pose Prediction\n by ViTs"], "answer_arxiv_id": ["1702.01381", "2208.08988"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_8197"} +{"question": "Could you provide me some works where deformable attention is used?", "answer": ["DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object\n Detection", "Masked-attention Mask Transformer for Universal Image Segmentation", "Mask DINO: Towards A Unified Transformer-based Framework for Object\n Detection and Segmentation", "BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers", "BEVFormer v2: Adapting Modern Image Backbones to Bird's-Eye-View\n Recognition via Perspective Supervision", "Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with\n Transformers"], "answer_arxiv_id": ["2203.03605", "2112.01527", "2206.02777", "2203.17270", "2211.10439", "2109.03814"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_8198"} +{"question": "Who have incorporated synaptic plasticity into ANNs in different settings?", "answer": ["Learning Associative Inference Using Fast Weight Memory", "Meta-Learning through Hebbian Plasticity in Random Networks", "Differentiable plasticity: training plastic neural networks with backpropagation", "Using Fast Weights to Attend to the Recent Past", "Short-Term Plasticity Neurons Learning to Learn and Forget", "Linear Transformers Are Secretly Fast Weight Programmers", "Going Beyond Linear Transformers with Recurrent Fast Weight Programmers", "Meta-Learning Deep Energy-Based Memory Models", "Metalearned Neural Memory", "Biological learning in key-value memory networks"], "answer_arxiv_id": ["2011.07831", "2007.02686", "1804.02464", "1610.06258", "2206.14048", "2102.11174", "2106.06295", "1910.02720", "1907.09720", "2110.13976"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_8199"} +{"question": "What are some of the first studies on learning in bilateral trade?", "answer": ["A Regret Analysis of Bilateral Trade"], "answer_arxiv_id": ["2102.08754v1"], "source_meta": {"published_time": "20220813"}, "qid": "AutoScholarQuery_train_8200"} +{"question": "Which works have used VQ-VAE in discrete diffusion models for generating high quality images?", "answer": ["ImageBART: Bidirectional Context with Multinomial Diffusion for Autoregressive Image Synthesis", "Global Context with Discrete Diffusion in Vector Quantised Modelling for Image Generation", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Improved Vector Quantized Diffusion Models"], "answer_arxiv_id": ["2108.08827", "2112.01799", "2111.14822", "2205.16007"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_8201"} +{"question": "Who proposed non-Markovian diffusion processes for sampling boosting?", "answer": ["Denoising Diffusion Implicit Models", "On Fast Sampling of Diffusion Probabilistic Models"], "answer_arxiv_id": ["2010.02502", "2106.00132"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_8202"} +{"question": "Can you mention any studies that focus on open-source LLMs?", "answer": ["GLM-130B: An Open Bilingual Pre-trained Model", "Mistral 7B", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2210.02414", "2310.06825", "2307.09288"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_8203"} +{"question": "Which papers are about using forward KL-divergence to derive policy objectives or for regularization?", "answer": ["Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning", "Trust Region Policy Optimization", "Maximum a Posteriori Policy Optimisation"], "answer_arxiv_id": ["1910.00177", "1502.05477", "1806.06920"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_8204"} +{"question": "What papers adopted the entire layer as the minimal unit in the pruning unit of structural pruning?", "answer": ["Reducing Transformer Depth on Demand with Structured Dropout"], "answer_arxiv_id": ["1909.11556"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_8205"} +{"question": "Which works have explored open-vocabulary segmentation through leveraging large pretrained vision-language models?", "answer": ["Language-driven Semantic Segmentation", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Decoupling Zero-Shot Semantic Segmentation", "GroupViT: Semantic Segmentation Emerges from Text Supervision", "Extract Free Dense Labels from CLIP", "Side Adapter Network for Open-Vocabulary Semantic Segmentation", "Generalized Decoding for Pixel, Image, and Language", "Open-vocabulary Semantic Segmentation with Frozen Vision-Language Models", "ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation", "DaTaSeg: Taming a Universal Multi-Dataset Multi-Task Segmentation Model"], "answer_arxiv_id": ["2201.03546", "2112.12143", "2210.04150", "2112.07910", "2202.11094", "2112.01071", "2302.12242", "2212.11270", "2210.15138", "2212.03588", "2306.01736"], "source_meta": {"published_time": "20230804"}, "qid": "AutoScholarQuery_train_8206"} +{"question": "What study pioneered large-scale NLI annotation by collecting multiple annotations per instance?", "answer": ["What Can We Learn from Collective Human Opinions on Natural Language\n Inference Data?"], "answer_arxiv_id": ["2010.03532"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_8207"} +{"question": "Any works about offline VIS methods and how they predict masks and instance trajectories?", "answer": ["STEm-Seg: Spatio-temporal Embeddings for Instance Segmentation in Videos", "Classifying, Segmenting, and Tracking Object Instances in Video with Mask Propagation", "Video Instance Segmentation with a Propose-Reduce Paradigm", "End-to-End Video Instance Segmentation with Transformers", "End-to-End Object Detection with Transformers", "Mask2Former for Video Instance Segmentation", "VITA: Video Instance Segmentation via Object Token Association"], "answer_arxiv_id": ["2003.08429", "1912.04573", "2103.13746", "2011.14503", "2005.12872", "2112.10764", "2206.04403v2"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8208"} +{"question": "What works employed attention maps for modeling temporal consistency in text-driven video editing?", "answer": ["FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Pix2Video: Video Editing using Image Diffusion"], "answer_arxiv_id": ["2303.09535", "2303.12688"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_8209"} +{"question": "Which work is based on the tri-perspective view method for predicting 3D occupancy in the area of 3D Occupancy Prediction?", "answer": ["Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction"], "answer_arxiv_id": ["2302.07817"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_8210"} +{"question": "What research introduced sparsely-activated models and achieved state-of-the-art performance?", "answer": ["A Review of Sparse Expert Models in Deep Learning"], "answer_arxiv_id": ["2209.01667"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_8211"} +{"question": "Which paper summarized the 'optimism in the face of uncertainty' principle in RL?", "answer": ["A Unifying View of Optimism in Episodic Reinforcement Learning"], "answer_arxiv_id": ["2007.01891"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_8212"} +{"question": "Could you give me some examples of research that focused on utilizing co-occurrent patterns to boost the decoder performance?", "answer": ["GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond", "CCNet: Criss-Cross Attention for Semantic Segmentation", "ISNet: Integrate Image-Level and Semantic-Level Context for Semantic Segmentation", "Non-local Neural Networks", "Disentangled Non-Local Neural Networks", "Context Prior for Scene Segmentation"], "answer_arxiv_id": ["1904.11492", "1811.11721", "2108.12382", "1711.07971", "2006.06668", "2004.01547"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_8213"} +{"question": "Which works have considered non-parametric models or Gaussian processes to improve the performance of variational autoencoders?", "answer": ["Stick-Breaking Variational Autoencoders", "Gaussian Process Prior Variational Autoencoders"], "answer_arxiv_id": ["1605.06197", "1810.11738"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_8214"} +{"question": "Is there any research about contrastive learning applied to boost the performance of MLP?", "answer": ["Graph-MLP: Node Classification without Message Passing in Graph"], "answer_arxiv_id": ["2106.04051"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_8215"} +{"question": "Which works proposed the idea of adapting Generative Flow Networks to continuous tasks?", "answer": ["Trajectory balance: Improved credit assignment in GFlowNets"], "answer_arxiv_id": ["2201.13259"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_8216"} +{"question": "What are some early works that utilised differentiable renderers to render learned implicit functions into images?", "answer": ["Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing", "SDFDiff: Differentiable Rendering of Signed Distance Fields for 3D Shape Optimization", "Autolabeling 3D Objects with Differentiable Rendering of SDF Shape Priors", "Learning to Infer Implicit Surfaces without 3D Supervision", "Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision", "SDF-SRN: Learning Signed Distance 3D Object Reconstruction from Static Images", "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video"], "answer_arxiv_id": ["1906.01618", "1911.13225", "1912.07109", "1911.11288", "1911.00767", "1912.07372", "2010.10505", "2104.00681"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_8217"} +{"question": "What works enhanced diffusion models with pre-trained models and used them to solve multi-modal tasks?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Classifier-Free Diffusion Guidance", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2103.00020", "2112.10741", "2207.12598", "2204.06125"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_8218"} +{"question": "Which works first proposed a neural network architecture for finite point sets?", "answer": ["Deep Sets"], "answer_arxiv_id": ["1703.06114"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_8219"} +{"question": "Which studies have used DensePose images as conditions to reposition input images?", "answer": ["DreamPose: Fashion Image-to-Video Synthesis via Stable Diffusion"], "answer_arxiv_id": ["2304.06025"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_8220"} +{"question": "Can you provide some references of double robust methods applied in conditionally randomized trials settings?", "answer": ["Demystifying Double Robustness: A Comparison of Alternative Strategies for Estimating a Population Mean from Incomplete Data", "Approximate Residual Balancing: De-Biased Inference of Average Treatment Effects in High Dimensions"], "answer_arxiv_id": ["0804.2958", "1604.07125"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_8221"} +{"question": "Which works utilized human exam questions for automated evaluations?", "answer": ["Measuring Massive Multitask Language Understanding", "Evaluating the Performance of Large Language Models on GAOKAO Benchmark"], "answer_arxiv_id": ["2009.03300", "2305.12474"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_8222"} +{"question": "Any works about monotonic functions for normalizing flows?", "answer": ["Block Neural Autoregressive Flow", "Neural Autoregressive Flows", "Q", "Unconstrained Monotonic Neural Networks"], "answer_arxiv_id": ["1904.04676", "1804.00779", "1611.08152", "1908.05164"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_8223"} +{"question": "Where can I find the details of the large-scale NetHack Learning Dataset?", "answer": ["Dungeons and Data: A Large-Scale NetHack Dataset"], "answer_arxiv_id": ["2211.00539"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_8224"} +{"question": "Which papers have developed results that connect φ-divergence DRO to variance regularization?", "answer": ["Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach"], "answer_arxiv_id": ["1610.03425"], "source_meta": {"published_time": "20220304"}, "qid": "AutoScholarQuery_train_8225"} +{"question": "What paper is about a cross-model knowledge distillation strategy for 3D object detection?", "answer": ["Weakly Supervised 3D Object Detection from Point Clouds"], "answer_arxiv_id": ["2007.13970"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_8226"} +{"question": "Which papers discuss the use of multiply robust estimators to relax model specifications across heterogeneous sites?", "answer": ["Oracle, Multiple Robust and Multipurpose Calibration in a Missing Response Problem"], "answer_arxiv_id": ["1410.3958"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_8227"} +{"question": "Any works about the development of video recognition architectures in the field of egocentric computer vision?", "answer": ["SlowFast Networks for Video Recognition", "ViViT: A Video Vision Transformer", "Keeping Your Eye on the Ball: Trajectory Attention in Video Transformers", "MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognition"], "answer_arxiv_id": ["1812.03982", "2103.15691", "2106.05392", "2201.08383"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_8228"} +{"question": "Which works tried to solve the Domain generalization task in a centralized setting?", "answer": ["Domain Generalization: A Survey", "Domain Invariant Representation Learning with Domain Density\n Transformations", "Unsupervised Domain Adaptive Learning via Synthetic Data for Person\n Re-identification", "Domain Generalization via Conditional Invariant Representation"], "answer_arxiv_id": ["2103.02503", "2102.05082", "2109.05542", "1807.08479"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_8229"} +{"question": "Could you provide me some studies that adapted shift module within convolution blocks for video generation tasks?", "answer": ["Latent-Shift: Latent Diffusion with Temporal Shift for Efficient\n Text-to-Video Generation", "Temporal Shift GAN for Large Scale Video Generation"], "answer_arxiv_id": ["2304.08477", "2004.01823"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_8230"} +{"question": "Could you provide me some studies about the surrogate task, membership inference attack?", "answer": ["Membership Inference Attacks Against Machine Learning Models"], "answer_arxiv_id": ["1610.05820"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_8231"} +{"question": "Could you provide studies on diffusion-based conditional image generation using semantic masks and other inputs?", "answer": ["Denoising Diffusion Probabilistic Models", "Sketch-Guided Text-to-Image Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2211.13752", "2112.10752", "2302.05543"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_8232"} +{"question": "Which papers consider training data memorization from a privacy perspective, especially focusing on if LMs inadvertently reveal private text?", "answer": ["The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks", "Extracting Training Data from Large Language Models", "Deduplicating Training Data Makes Language Models Better"], "answer_arxiv_id": ["1802.08232", "2012.07805", "2107.06499"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_8233"} +{"question": "Could you provide me with some studies about the challenge of aligning LLMs to human values?", "answer": ["Artificial Intelligence, Values and Alignment"], "answer_arxiv_id": ["2001.09768v2"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_8234"} +{"question": "Which references provide comprehensive overview on differentiable rendering, neural rendering, and point-based rendering?", "answer": ["Differentiable Rendering: A Survey", "Advances in Neural Rendering", "Neural Fields in Visual Computing and Beyond"], "answer_arxiv_id": ["2006.12057", "2111.05849", "2111.11426"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_8235"} +{"question": "Which research papers have contributed to dataset-agnostic approaches in biological applications?", "answer": ["CytoImageNet: A large-scale pretraining dataset for bioimage transfer learning"], "answer_arxiv_id": ["2111.11646"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_8236"} +{"question": "What are the works that use depth modality in tracking?", "answer": ["RGBD1K: A Large-scale Dataset and Benchmark for RGB-D Object Tracking", "Depth-only Object Tracking", "Learning Dual-Fused Modality-Aware Representations for RGBD Tracking"], "answer_arxiv_id": ["2208.09787", "2110.11679", "2211.03055"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_8237"} +{"question": "What are some example papers on multi-task learning approaches in computer vision?", "answer": ["Taskonomy: Disentangling Task Transfer Learning", "UberNet: Training a `Universal' Convolutional Neural Network for Low-,\n Mid-, and High-Level Vision using Diverse Datasets and Limited Memory", "Multi-task Self-Supervised Visual Learning", "Multi-Task Learning as Multi-Objective Optimization", "Perceiver IO: A General Architecture for Structured Inputs & Outputs"], "answer_arxiv_id": ["1804.08328", "1609.02132", "1708.07860", "1810.04650", "2107.14795"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_8238"} +{"question": "Could you provide me studies that proposed methods to infer the learning rules governing weight updates from post-learning neural activity or spike train recordings?", "answer": ["A framework for studying synaptic plasticity with neural spike train data"], "answer_arxiv_id": ["1411.4077"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_8239"} +{"question": "What studies worked on policy-constraint methods as a solution to distribution shift in offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Provably Good Batch Reinforcement Learning Without Great Exploration", "Behavior Regularized Offline Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning", "Keep Doing What Worked: Behavioral Modelling Priors for Offline Reinforcement Learning", "Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning", "Off-Policy Policy Gradient with State Distribution Correction", "AlgaeDICE: Policy Gradient from Arbitrary Experience"], "answer_arxiv_id": ["1812.02900", "1906.00949", "2007.08202", "1911.11361", "2106.06860", "2002.08396", "1910.00177", "1904.08473", "1912.02074"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_8240"} +{"question": "What studies used piecewise linear regression for improving flexibility of linear models?", "answer": ["Piecewise Linear Regression via a Difference of Convex Functions", "Efficient Algorithms for Multidimensional Segmented Regression"], "answer_arxiv_id": ["2007.02422", "2003.11086v1"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_8241"} +{"question": "Which works discuss top-k sampling as a decoding algorithm in neural text generation?", "answer": ["Hierarchical Neural Story Generation"], "answer_arxiv_id": ["1805.04833"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_8242"} +{"question": "Who are the researchers that explored the linear model’s evidence for model selection?", "answer": ["Scalable Marginal Likelihood Estimation for Model Selection in Deep Learning", "Adapting the Linearised Laplace Model Evidence for Modern Deep Learning"], "answer_arxiv_id": ["2104.04975", "2206.08900"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_8243"} +{"question": "What papers are about single image 3D reconstruction?", "answer": ["Depth Map Prediction from a Single Image using a Multi-Scale Deep Network", "Single-Image Depth Perception in the Wild", "Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer", "Vision Transformers for Dense Prediction"], "answer_arxiv_id": ["1406.2283", "1604.03901", "1907.01341", "2103.13413"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_8244"} +{"question": "Can you tell me about any research that has used self-supervised instance discrimination task for video analysis?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance Discrimination"], "answer_arxiv_id": ["1805.01978"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_8245"} +{"question": "Which study discusses improving exploration by reducing policy horizons in text-based game Zork ?", "answer": ["How to Avoid Being Eaten by a Grue: Structured Exploration Strategies for Textual Worlds"], "answer_arxiv_id": ["2006.07409"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_8246"} +{"question": "Which studies used self-supervised learning (SSL) for the maximization of mutual information between paired views?", "answer": ["Learning Robust Representations via Multi-View Information Bottleneck", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2002.07017", "2103.00020"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_8247"} +{"question": "Could you provide examples of works that have used adversarial training methods in FRL studies?", "answer": ["Censoring Representations with an Adversary", "Controllable Invariance through Adversarial Feature Learning", "Learning Adversarially Fair and Transferable Representations", "Learning Controllable Fair Representations", "Learning Fair Representations via an Adversarial Framework", "Mitigating Information Leakage in Image Representations: A Maximum Entropy Approach", "Generating Fair Universal Representations using Adversarial Models", "Learning fair representation with a parametric integral probability metric"], "answer_arxiv_id": ["1511.05897", "1705.11122", "1802.06309", "1812.04218", "1904.13341", "1904.05514", "1910.00411", "2202.02943"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_8248"} +{"question": "Which studies applied test-time optimization for generating semantic labels?", "answer": ["Semantic Segmentation with Generative Models: Semi-Supervised Learning\n and Strong Out-of-Domain Generalization"], "answer_arxiv_id": ["2104.05833"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_8249"} +{"question": "Could you tell me which work presented that decision calibration can be achieved when the set of possible actions is finite?", "answer": ["Calibrating Predictions to Decisions: A Novel Approach to Multi-Class Calibration"], "answer_arxiv_id": ["2107.05719"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_8250"} +{"question": "Could you provide me some studies about visual question answering under the 3D setting?", "answer": ["RUBi: Reducing Unimodal Biases for Visual Question Answering", "SQA3D: Situated Question Answering in 3D Scenes", "3D Concept Learning and Reasoning from Multi-View Images", "3D Question Answering", "Embodied Question Answering"], "answer_arxiv_id": ["1906.10169", "2210.07474", "2303.11327", "2112.08359", "1711.11543"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_8251"} +{"question": "Could you give me the reference that proposed entropy regularized PG with stochastic gradient in the contextual MDP setting?", "answer": ["Beyond Exact Gradients: Convergence of Stochastic Soft-Max Policy Gradient Methods with Entropy Regularization"], "answer_arxiv_id": ["2110.10117v3"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_8252"} +{"question": "What works demonstrated that benchmarks are not robust to minor perturbations?", "answer": ["Large Language Models Sensitivity to The Order of Options in\n Multiple-Choice Questions", "Large Language Models Are Not Robust Multiple Choice Selectors"], "answer_arxiv_id": ["2308.11483", "2309.03882"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_8253"} +{"question": "Which papers can you cite that have extended the use of Masked Auto-Encoder (MAE) across various domains?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "GraphMAE: Self-Supervised Masked Graph Autoencoders", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Masked Autoencoders that Listen", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers", "data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language", "Efficient Self-supervised Learning with Contextualized Target Representations for Vision, Speech and Language", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "Architecture-Agnostic Masked Image Modeling – From ViT back to CNN", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks", "Taming Transformers for High-Resolution Image Synthesis", "MAGE: MAsked Generative Encoder to Unify Representation Learning and Image Synthesis"], "answer_arxiv_id": ["2111.06377", "2205.10803", "1810.04805", "2207.06405", "1810.04805", "2111.06377", "2106.08254", "2202.03555", "2212.07525", "2112.09133", "2205.13943", "2208.10442", "2012.09841", "2211.09117"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_8254"} +{"question": "Could you provide some references where researchers encouraged permutation equivariance through data augmentation?", "answer": ["Learning to Learn with Generative Models of Neural Network Checkpoints", "VeLO: Training Versatile Learned Optimizers by Scaling Up"], "answer_arxiv_id": ["2209.12892", "2211.09760"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_8255"} +{"question": "Which works introduced Dirichlet prior over multinomial likelihood for evidential classification?", "answer": ["Evidential Deep Learning for Open Set Action Recognition", "Uncertainty Aware Semi-Supervised Learning on Graph Data"], "answer_arxiv_id": ["2107.10161", "2010.12783"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_8256"} +{"question": "What studies considered data heterogeneity in the context of Federated Learning?", "answer": ["Federated Learning with Non-IID Data", "Personalized Federated Learning: A Meta-Learning Approach", "FedBN: Federated Learning on Non-IID Features via Local Batch Normalization", "Federated Learning on Non-IID Data Silos: An Experimental Study"], "answer_arxiv_id": ["1806.00582", "2002.07948", "2102.07623", "2102.02079"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_8257"} +{"question": "Any studies performing rendering at low-resolution and using a 2D post-processing CNN in 3D-aware GANs?", "answer": ["GIRAFFE HD: A High-Resolution 3D-aware Generative Model", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature\n Fields", "Efficient Geometry-aware 3D Generative Adversarial Networks", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image\n Synthesis", "StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation"], "answer_arxiv_id": ["2203.14954", "2011.12100", "2112.07945", "2110.08985", "2112.11427"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_8258"} +{"question": "In what papers do researchers extract bounding box proposals and classify them?", "answer": ["3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans", "Learning Object Bounding Boxes for 3D Instance Segmentation on Point\n Clouds"], "answer_arxiv_id": ["1812.07003", "1906.01140"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_8259"} +{"question": "What papers are major studies in the domain of Group-Distributionally Robust Optimization?", "answer": ["Does Distributionally Robust Supervised Learning Give Robust Classifiers?", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1611.02041", "1911.08731"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_8260"} +{"question": "Could you provide me some works on minimax optimization in the convex-concave case?", "answer": ["Lower complexity bounds of first-order methods for convex-concave bilinear saddle-point problems"], "answer_arxiv_id": ["1808.02901"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_8261"} +{"question": "Which works discuss the importance of measuring meaningful distances between neural networks?", "answer": ["Measuring and regularizing networks in function space", "On the distance between two neural networks and the stability of learning", "Amortized Proximal Optimization"], "answer_arxiv_id": ["1805.08289", "2002.03432", "2203.00089"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_8262"} +{"question": "Which studies discuss the application of Transformers in natural language processing?", "answer": ["Attention Is All You Need", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context"], "answer_arxiv_id": ["1706.03762", "1810.04805", "1901.02860"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8263"} +{"question": "Who conducted research that considers the Gaussian Processes State Space Models (gpssm) in partially observable unstable settings?", "answer": ["Structured Variational Inference in Partially Observable Unstable Gaussian Process State Space Models"], "answer_arxiv_id": ["1907.07035"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_8264"} +{"question": "Which papers argued that disentanglement is not necessary for style edit?", "answer": ["Controllable Unsupervised Text Attribute Transfer via Editing Entangled Latent Representation"], "answer_arxiv_id": ["1905.12926"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_8265"} +{"question": "Are there any studies that proposed the dynamic composition of experts in MoEs to tackle the systematic generalization problem?", "answer": ["Dynamic Inference with Neural Interpreters"], "answer_arxiv_id": ["2110.06399"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_8266"} +{"question": "Which research attempts to edit factual knowledge in LLMs by making targeted modifications of the model’s neurons?", "answer": ["Editing Factual Knowledge in Language Models", "Fast Model Editing at Scale", "Memory-Based Model Editing at Scale"], "answer_arxiv_id": ["2104.08164", "2110.11309", "2206.06520"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_8267"} +{"question": "What research works utilize sparse 3D convolutional backbones for Lidar-based 3D object detection?", "answer": ["4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection"], "answer_arxiv_id": ["1904.08755", "1711.06396"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_8268"} +{"question": "Could you provide me a study where VQGAN is used to enhance the reconstruction quality with adversarial and perceptual objectives?", "answer": ["Taming Transformers for High-Resolution Image Synthesis"], "answer_arxiv_id": ["2012.09841"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_8269"} +{"question": "What work proposed learning a verifier to predict the correctness of the program based on the NL, program and execution results?", "answer": ["LEVER: Learning to Verify Language-to-Code Generation with Execution"], "answer_arxiv_id": ["2302.08468"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_8270"} +{"question": "Are there any works that employed GANs for learning universal representations from different viewpoints?", "answer": ["Third-Person Imitation Learning"], "answer_arxiv_id": ["1703.01703"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_8271"} +{"question": "Which research artworks integrated capsule networks for image colorization tasks?", "answer": ["Image Colorization By Capsule Networks"], "answer_arxiv_id": ["1908.08307"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_8272"} +{"question": "Could you provide me some works that expanded the boundary of novel view synthesis for dynamic scenes?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2011.12948", "2106.13228v2"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_8273"} +{"question": "Which paper first introduced the slot-attention module for image feature distillation?", "answer": ["Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["2006.15055"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_8274"} +{"question": "What works form the basis of the notion of feature disentanglement in representation learning?", "answer": ["Towards a Definition of Disentangled Representations", "Emergence of Invariance and Disentanglement in Deep Representations"], "answer_arxiv_id": ["1812.02230", "1706.01350"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_8275"} +{"question": "What are some studies in the field of semantic segmentation in 2D images?", "answer": ["Unsupervised Hierarchical Semantic Segmentation with Multiview\n Cosegmentation and Clustering Transformers", "PiCIE: Unsupervised Semantic Segmentation using Invariance and\n Equivariance in Clustering", "Unsupervised Semantic Segmentation by Contrasting Object Mask Proposals"], "answer_arxiv_id": ["2204.11432", "2103.17070", "2102.06191"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_8276"} +{"question": "Could you provide some studies that propose the use of external knowledge bases for image classification or object detection in vision-language models?", "answer": ["K-Lite: Learning Transferable Visual Models with External Knowledge", "Visual Classification via Description from Large Language Models", "Language in a Bottle: Language Model Guided Concept Bottlenecks for Interpretable Image Classification"], "answer_arxiv_id": ["2204.09222", "2210.07183", "2211.11158"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_8277"} +{"question": "What are the papers suggesting methods that add task-specific parameters for Continual Learning?", "answer": ["Lifelong Learning with Dynamically Expandable Networks", "Supermasks in Superposition", "Episodic Memory in Lifelong Language Learning"], "answer_arxiv_id": ["1708.01547", "2006.14769", "1906.01076"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_8278"} +{"question": "Which papers discuss the 8-bit quantization for large language models?", "answer": ["SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models", "LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "DQ-BART: Efficient Sequence-to-Sequence Model via Joint Distillation and Quantization", "GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers", "I-BERT: Integer-only BERT Quantization"], "answer_arxiv_id": ["2211.10438", "2208.07339", "2203.11239", "2210.17323", "2101.01321"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_8279"} +{"question": "What works have studied deterministic decentralized methods for nonconvex problems with a nonsmooth regularizer?", "answer": ["NEXT: In-Network Nonconvex Optimization", "Decentralized Frank-Wolfe Algorithm for Convex and Non-convex Problems", "On Nonconvex Decentralized Gradient Descent", "On Distributed Non-convex Optimization: Projected Subgradient Method For Weakly Convex Problems in Networks", "Distributed Nonconvex Constrained Optimization over Time-Varying Digraphs"], "answer_arxiv_id": ["1602.00591", "1612.01216", "1608.05766", "2004.13233", "1809.01106v1"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_8280"} +{"question": "Which research paper presents recent approaches to pose and shape estimation from images?", "answer": ["I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human\n Pose and Mesh Estimation from a Single RGB Image", "Beyond Static Features for Temporally Consistent 3D Human Pose and Shape\n from a Video", "Capturing Humans in Motion: Temporal-Attentive 3D Human Pose and Shape\n Estimation from Monocular Video", "On Self-Contact and Human Pose", "PARE: Part Attention Regressor for 3D Human Body Estimation", "Humans in 4D: Reconstructing and Tracking Humans with Transformers"], "answer_arxiv_id": ["2008.03713", "2011.08627", "2203.08534", "2104.03176", "2104.08527", "2305.20091"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_8281"} +{"question": "What study proposed an intra- and inter-modal contrastive learning framework among augmented point clouds?", "answer": ["CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understanding"], "answer_arxiv_id": ["2203.00680"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_8282"} +{"question": "Which works have adopted the methodology proposed for entity substitution in producing counterfactual datasets?", "answer": ["DisentQA: Disentangling Parametric and Contextual Knowledge with\n Counterfactual Question Answering", "Context-faithful Prompting for Large Language Models", "Entity-Based Knowledge Conflicts in Question Answering", "Large Language Models with Controllable Working Memory"], "answer_arxiv_id": ["2211.05655", "2303.11315", "2109.05052", "2211.05110"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_8283"} +{"question": "Are there any researches that propose subgraph sampling technique for efficient mini-batch training for large-scale graphs?", "answer": ["Inductive Representation Learning on Large Graphs", "Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks", "GraphSAINT: Graph Sampling Based Inductive Learning Method"], "answer_arxiv_id": ["1706.02216", "1905.07953", "1907.04931v4"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_8284"} +{"question": "Which papers talk about multiple mechanisms that contribute to the implicit regularization in Stochastic Gradient Descent (SGD)?", "answer": ["Neurashed: A Phenomenological Model for Imitating Deep Learning Training", "The Local Elasticity of Neural Networks", "On the Implicit Bias in Deep-Learning Algorithms"], "answer_arxiv_id": ["2112.09741", "1910.06943", "2208.12591"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8285"} +{"question": "Which work examines backpropagation through quantized weights at a scale beyond 1 billion parameters, apart from our work?", "answer": ["Stable and low-precision training for large-scale vision-language models"], "answer_arxiv_id": ["2304.13013"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_8286"} +{"question": "What researches can you give me about unsupervised speech enhancement approaches?", "answer": ["Unsupervised Noise Adaptive Speech Enhancement by Discriminator-Constrained Optimal Transport", "Self-Supervised Learning for Speech Enhancement Through Synthesis", "MetricGAN-U: Unsupervised speech enhancement/ dereverberation based only on noisy/ reverberated speech"], "answer_arxiv_id": ["2111.06316", "2211.02542", "2110.05866v1"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_8287"} +{"question": "Which papers represent complex functions and generate high-dimensional data using Implicit Neural Representation (INR)?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections", "Block-NeRF: Scalable Large Scene Neural View Synthesis", "Coordinates Are NOT Lonely - Codebook Prior Helps Implicit Neural 3D Representations"], "answer_arxiv_id": ["2011.13961", "2008.02268", "2202.05263", "2210.11170"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_8288"} +{"question": "What research presented the TRADES method in the field of adversarial training?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1901.08573"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_8289"} +{"question": "Who used meta-learning methods to compress time slices of climate data into latent vectors?", "answer": ["COIN++: Neural Compression Across Modalities", "Meta-Learning Sparse Compression Networks"], "answer_arxiv_id": ["2201.12904", "2205.08957"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_8290"} +{"question": "What works explored methods of using large language models to generate instructional data for the enhancement of instruction tuning?", "answer": ["Self-Instruct: Aligning Language Models with Self-Generated Instructions", "WizardLM: Empowering Large Language Models to Follow Complex\n Instructions", "AnnoLLM: Making Large Language Models to Be Better Crowdsourced\n Annotators", "Reflection-Tuning: Data Recycling Improves LLM Instruction-Tuning"], "answer_arxiv_id": ["2212.10560", "2304.12244", "2303.16854", "2310.11716"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_8291"} +{"question": "What is the study that initiates the investigation of candidate selection in statistical discrimination scenario?", "answer": ["On Fair Selection in the Presence of Implicit Variance"], "answer_arxiv_id": ["2006.13699"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_8292"} +{"question": "Could you provide some works that used Transformers-based methods for capturing spatial and temporal correlations in human motion prediction?", "answer": ["A Spatio-temporal Transformer for 3D Human Motion Prediction"], "answer_arxiv_id": ["2004.08692"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_8293"} +{"question": "In which work has ORB-SLAM been used to find short clips where the camera pose was stable?", "answer": ["Generative Hybrid Representations for Activity Forecasting with No-Regret Learning"], "answer_arxiv_id": ["1904.06250"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_8294"} +{"question": "Could you provide me some works that introduce a benchmark for evaluating LMMs' perception and cognition?", "answer": ["MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language\n Models"], "answer_arxiv_id": ["2306.13394"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_8295"} +{"question": "What papers propose direct methods for VO that utilize intensity to optimize camera pose and 3D scene point positions?", "answer": ["Direct Sparse Odometry", "Stereo DSO: Large-Scale Direct Sparse Visual Odometry with Stereo\n Cameras"], "answer_arxiv_id": ["1607.02565", "1708.07878"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_8296"} +{"question": "Which papers are related to the usage of distance-based models for video anomaly detection?", "answer": ["Object-centric Auto-encoders and Dummy Anomalies for Abnormal Event\n Detection in Video", "Detecting abnormal events in video using Narrowed Normality Clusters", "Street Scene: A new dataset and evaluation protocol for video anomaly\n detection", "Learning a distance function with a Siamese network to localize\n anomalies in videos", "Plug-and-Play CNN for Crowd Motion Analysis: An Application in Abnormal\n Event Detection", "Deep-Anomaly: Fully Convolutional Neural Network for Fast Anomaly\n Detection in Crowded Scenes"], "answer_arxiv_id": ["1812.04960", "1801.05030", "1902.05872", "2001.09189", "1610.00307", "1609.00866"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_8297"} +{"question": "What papers introduced variance reduction to value-based methods in generative setting RL?", "answer": ["Near-Optimal Time and Sample Complexities for Solving Discounted Markov Decision Process with a Generative Model", "Variance Reduced Value Iteration and Faster Algorithms for Solving Markov Decision Processes"], "answer_arxiv_id": ["1806.01492", "1710.09988"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_8298"} +{"question": "Who used the NaturalQA dataset to construct a new benchmark testing the impact of key information position in long contexts on the processing capability of large language models?", "answer": ["Lost in the Middle: How Language Models Use Long Contexts"], "answer_arxiv_id": ["2307.03172"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_8299"} +{"question": "Which papers have made remarkable performances in code generation using decoder-only models?", "answer": ["Evaluating Large Language Models Trained on Code", "CodeFill: Multi-token Code Completion by Jointly Learning from Structure\n and Naming Sequences", "CodeGen: An Open Large Language Model for Code with Multi-Turn Program\n Synthesis"], "answer_arxiv_id": ["2107.03374", "2202.06689", "2203.13474"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_8300"} +{"question": "Which work selects examples based on the uncertainty of a small proxy model for active learning?", "answer": ["Selection via Proxy: Efficient Data Selection for Deep Learning"], "answer_arxiv_id": ["1906.11829"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_8301"} +{"question": "Which works incorporate continuous embeddings to represent the image and calculate either a loss for the next token prediction or the next visual embedding regression?", "answer": ["Generative Multimodal Models are In-Context Learners"], "answer_arxiv_id": ["2312.13286"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_8302"} +{"question": "What studies used local or global descriptors for cross-view positional estimation?", "answer": ["VIGOR: Cross-View Image Geo-localization beyond One-to-one Retrieval", "Visual Cross-View Metric Localization with Dense Uncertainty Estimates", "Beyond Cross-view Image Retrieval: Highly Accurate Vehicle Localization Using Satellite Image", "Satellite Image Based Cross-view Localization for Autonomous Vehicle", "SliceMatch: Geometry-guided Aggregation for Cross-View Pose Estimation"], "answer_arxiv_id": ["2011.12172", "2208.08519", "2204.04752", "2207.13506", "2211.14651"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_8303"} +{"question": "Are there any studies on generating driving videos for supporting video-based BEV perception methods?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models"], "answer_arxiv_id": ["2304.08818"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_8304"} +{"question": "Which studies facilitate improving sampling by designing efficient training-free sampling algorithms?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations", "Fast Sampling of Diffusion Models with Exponential Integrator", "gDDIM: Generalized denoising diffusion implicit models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps"], "answer_arxiv_id": ["2011.13456", "2204.13902", "2206.05564", "2206.00927"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_8305"} +{"question": "Which works are about 2D Generative Adversarial Networks (GANs) used for image generation?", "answer": ["Generative Adversarial Networks", "Progressive Growing of GANs for Improved Quality, Stability, and\n Variation", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN", "Alias-Free Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661", "1710.10196", "1812.04948", "1912.04958", "2106.12423"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_8306"} +{"question": "Could you provide me some studies about transformer-based architectures in the context of UP-DETR and DETReg?", "answer": ["UP-DETR: Unsupervised Pre-training for Object Detection with Transformers", "DETReg: Unsupervised Pretraining with Region Priors for Object Detection"], "answer_arxiv_id": ["2011.09094", "2106.04550"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_8307"} +{"question": "Which work suggested the use of a gate unit for the pruning of different ranks?", "answer": ["Sparse Low-rank Adaptation of Pre-trained Language Models"], "answer_arxiv_id": ["2311.11696v1"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8308"} +{"question": "What work has recently unified the PGD-AT, TRADES, and MART methods under the W-DRO framework?", "answer": ["A Unified Wasserstein Distributional Robustness Framework for Adversarial Training"], "answer_arxiv_id": ["2202.13437"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_8309"} +{"question": "Which works integrate 3D representations into the network for better pose accuracy and consistency in views?", "answer": ["ConsistNet: Enforcing 3D Consistency for Multi-view Images Diffusion", "iNVS: Repurposing Diffusion Inpainters for Novel View Synthesis"], "answer_arxiv_id": ["2310.10343", "2310.16167"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8310"} +{"question": "What researches successfully established the dynamics of scenes to a differentiable high-quality physical simulation?", "answer": ["3D Human Pose Estimation via Intuitive Physics", "PhysDiff: Physics-Guided Human Motion Diffusion Model", "InterDiff: Generating 3D Human-Object Interactions with Physics-Informed\n Diffusion"], "answer_arxiv_id": ["2303.18246", "2212.02500", "2308.16905"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_8311"} +{"question": "Could you name some works where random features were used to approximate input-output maps in Banach spaces and to solve partial differential equations?", "answer": ["Local Extreme Learning Machines and Domain Decomposition for Solving Linear and Nonlinear Partial Differential Equations", "Bridging Traditional and Machine Learning-based Algorithms for Solving PDEs: The Random Feature Method"], "answer_arxiv_id": ["2012.02895", "2207.13380"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_8312"} +{"question": "Which paper generalizes the concept in SortNet?", "answer": ["Rethinking Lipschitz Neural Networks and Certified Robustness: A Boolean Function Perspective"], "answer_arxiv_id": ["2210.01787"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_8313"} +{"question": "Could you provide me some studies about modeling challenging problems in contact mechanics and turbulent flows using classical time integration schemes?", "answer": ["The Neural Particle Method - An Updated Lagrangian Physics Informed Neural Network for Computational Fluid Dynamics"], "answer_arxiv_id": ["2003.10208"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_8314"} +{"question": "Can you list the works where the successive application of propagation operators was studied in the context of utilizing node features?", "answer": ["Unifying Graph Convolutional Neural Networks and Label Propagation"], "answer_arxiv_id": ["2002.06755"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_8315"} +{"question": "Which studies address Hypothesis Transfer Learning methods?", "answer": ["Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation"], "answer_arxiv_id": ["2002.08546"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_8316"} +{"question": "In the field of text-to-image generation, which researches have explored text-conditional diffusion models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2205.11487", "2112.10741"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_8317"} +{"question": "Which paper proposed a memory model in n-transformers called kNN-LM?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models"], "answer_arxiv_id": ["1911.00172"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8318"} +{"question": "Which paper extended the convergence result for the affine variance noise model for the SGD?", "answer": ["Optimization Methods for Large-Scale Machine Learning"], "answer_arxiv_id": ["1606.04838"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_8319"} +{"question": "Are there any papers that detail the use of a meta-learning classifier-based uncertainty metrics?", "answer": ["MURAL: Meta-Learning Uncertainty-Aware Rewards for Outcome-Driven Reinforcement Learning", "Outcome-directed Reinforcement Learning by Uncertainty & Temporal Distance-Aware Curriculum Goal Generation"], "answer_arxiv_id": ["2107.07184", "2301.11741"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_8320"} +{"question": "Could you tell me which work leverages explicit image priors from strong pretrained 2D human image generators and 3D geometry priors?", "answer": ["HumanGen: Generating Human Radiance Fields with Explicit Priors"], "answer_arxiv_id": ["2212.05321"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8321"} +{"question": "Could you provide me the research that stresses the role of choices in decision tree learning but operates in a stylized theoretical setting?", "answer": ["Properly learning decision trees in almost polynomial time"], "answer_arxiv_id": ["2109.00637v2"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_8322"} +{"question": "What work investigated the deductive reasoning ability of LMs on a corpus which is composed of a specific type of multistep inference?", "answer": ["Measuring Systematic Generalization in Neural Proof Generation with Transformers"], "answer_arxiv_id": ["2009.14786"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_8323"} +{"question": "Can you provide references that don't fall into the categories of exemplar-based and model-based methods in Continual Incremental Learning (CIL)?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Learning without Forgetting", "Gradient-based Editing of Memory Examples for Online Task-free Continual Learning", "Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning"], "answer_arxiv_id": ["1612.00796", "1606.09282", "2006.15294", "2106.09701"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_8324"} +{"question": "Which paper has revolutionized the novel-view synthesis technology with its impressive rendering quality?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_8325"} +{"question": "Could you provide me some studies about camera pose estimation using ground images and 3D points?", "answer": ["Satellite Image Based Cross-view Localization for Autonomous Vehicle"], "answer_arxiv_id": ["2207.13506"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_8326"} +{"question": "Which papers are about techniques that utilized high-level features extracted from pre-trained deep neural networks to represent image style?", "answer": ["Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "Neural Neighbor Style Transfer", "Arbitrary Style Transfer in Real-time with Adaptive Instance\n Normalization", "Universal Style Transfer via Feature Transforms", "Two-Stage Peer-Regularized Feature Recombination for Arbitrary Image\n Style Transfer"], "answer_arxiv_id": ["1603.08155", "2203.13215", "1703.06868", "1705.08086", "1906.02913"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_8327"} +{"question": "Which documents report application of Federated Learning in autonomous driving and IoTs?", "answer": ["Federated Transfer Reinforcement Learning for Autonomous Driving", "Adaptive Federated Learning in Resource Constrained Edge Computing Systems", "Federated Learning in Mobile Edge Networks: A Comprehensive Survey"], "answer_arxiv_id": ["1910.06001v1", "1804.05271", "1909.11875"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_8328"} +{"question": "What papers primarily focus on parameter-efficient fine-tuning (PEFT) techniques?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Parameter-Efficient Transfer Learning for NLP", "Visual Prompt Tuning", "SAM Fails to Segment Anything? -- SAM-Adapter: Adapting SAM in\n Underperformed Scenes: Camouflage, Shadow, Medical Image Segmentation, and\n More", "Conditional Prompt Learning for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language\n Modeling"], "answer_arxiv_id": ["2106.09685", "2101.00190", "1902.00751", "2203.12119", "2304.09148", "2203.05557", "2111.03930"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_8329"} +{"question": "What works extended already trained test-to-image latent diffusion models like Stable Diffusion?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation"], "answer_arxiv_id": ["2304.08818", "2307.04725", "2303.08320"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_8330"} +{"question": "Which studies encountered difficulty in dividing molecule graphs in raw space due to the complex and entangled molecular structure?", "answer": ["Graph Rationalization with Environment-based Augmentations"], "answer_arxiv_id": ["2206.02886"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_8331"} +{"question": "Who developed extensions to the marginal sensitivity model for continuous treatments?", "answer": ["Partial Identification of Dose Responses with Hidden Confounders"], "answer_arxiv_id": ["2204.11206"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_8332"} +{"question": "Which works tried to reduce the amount of required data for cross-modal knowledge transfer?", "answer": ["Knowledge as Priors: Cross-Modal Knowledge Generalization for Datasets\n without Superior Knowledge"], "answer_arxiv_id": ["2004.00176"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_8333"} +{"question": "What are the works on mechanism design for double auctions?", "answer": ["Concurrent Auctions Across The Supply Chain"], "answer_arxiv_id": ["1107.0028v1"], "source_meta": {"published_time": "20220813"}, "qid": "AutoScholarQuery_train_8334"} +{"question": "What research works have shown promising results in the field of Offline BBO?", "answer": ["Model Inversion Networks for Model-Based Optimization", "Conservative Objective Models for Effective Offline Model-Based Optimization", "Conditioning by adaptive sampling for robust design", "Offline Model-Based Optimization via Normalized Maximum Likelihood Estimation", "RoMA: Robust Model Adaptation for Offline Model-based Optimization"], "answer_arxiv_id": ["1912.13464", "2107.06882", "1901.10060", "2102.07970", "2110.14188"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_8335"} +{"question": "What are the relevant studies on the application of visual prompts for vision tasks and foundation models?", "answer": ["CPT: Colorful Prompt Tuning for Pre-trained Vision-Language Models", "Visual Prompt Tuning", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Doubly Right Object Recognition: A Why Prompt for Visual Rationales", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2109.11797", "2203.12119", "2010.11929", "2212.06202v2", "2103.00020"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_8336"} +{"question": "What are some studies about multi-view semantic fusion methods?", "answer": ["SemanticFusion: Dense 3D Semantic Mapping with Convolutional Neural Networks", "Meaningful Maps With Object-Oriented Semantic Mapping", "Multi-View Deep Learning for Consistent Semantic Mapping with RGB-D Cameras"], "answer_arxiv_id": ["1609.05130", "1609.07849", "1703.08866"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_8337"} +{"question": "What works employ various networks to learn box and mask proposals for category-agnostic proposal learning?", "answer": ["Class-agnostic Object Detection", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks", "A Simple Baseline for Open-Vocabulary Semantic Segmentation with\n Pre-trained Vision-language Model", "CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation\n via Global and Local Refinement"], "answer_arxiv_id": ["2011.14204", "1506.01497", "2112.14757", "2005.02551"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_8338"} +{"question": "Can you provide me some studies about video-language models with frozen modules including BLIP-2, Video-LLaMA, Video-ChatGPT, and VideoChat?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models", "VideoChat: Chat-Centric Video Understanding"], "answer_arxiv_id": ["2301.12597", "2201.12086", "2306.02858", "2306.05424", "2305.06355"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_8339"} +{"question": "Can you give examples of studies about 3D-aware image synthesize?", "answer": ["GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis", "Efficient Geometry-aware 3D Generative Adversarial Networks", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis", "StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation"], "answer_arxiv_id": ["2007.02442", "2011.12100", "2012.00926", "2112.07945", "2110.08985", "2112.11427"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_8340"} +{"question": "Which paper proposed a pretraining scheme tailored for QA tasks by designing a recurring span selection objective?", "answer": ["Few-Shot Question Answering by Pretraining Span Selection"], "answer_arxiv_id": ["2101.00438"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_8341"} +{"question": "Are there any projects that integrate LLMs with external tools?", "answer": ["Augmented Language Models: a Survey"], "answer_arxiv_id": ["2302.07842"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8342"} +{"question": "What works incorporated contrastive learning in various tasks, including vision tasks?", "answer": ["FaceNet: A Unified Embedding for Face Recognition and Clustering", "Deep Metric Learning via Lifted Structured Feature Embedding", "Smart Mining for Deep Metric Learning"], "answer_arxiv_id": ["1503.03832", "1511.06452", "1704.01285"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_8343"} +{"question": "Are there any research papers on the mixture of experts in the context of dialogue generation?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer"], "answer_arxiv_id": ["1701.06538"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_8344"} +{"question": "What papers discuss the limitations of Yang-Res’s approach in the application of makeup estimation methods?", "answer": ["Makeup like a superstar: Deep Localized Makeup Transfer Network", "PSGAN: Pose and Expression Robust Spatial-Aware GAN for Customizable\n Makeup Transfer", "LADN: Local Adversarial Disentangling Network for Facial Makeup and\n De-Makeup", "Lipstick ain't enough: Beyond Color Matching for In-the-Wild Makeup\n Transfer", "SOGAN: 3D-Aware Shadow and Occlusion Robust GAN for Makeup Transfer", "EleGANt: Exquisite and Locally Editable GAN for Makeup Transfer", "BeautyREC: Robust, Efficient, and Content-preserving Makeup Transfer"], "answer_arxiv_id": ["1604.07102", "1909.06956", "1904.11272", "2104.01867", "2104.10567", "2207.09840", "2212.05855"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_8345"} +{"question": "Could you give me an example of a work that modeled each denoising step using a multimodal conditional GAN?", "answer": ["Tackling the Generative Learning Trilemma with Denoising Diffusion GANs"], "answer_arxiv_id": ["2112.07804"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8346"} +{"question": "What works consider uncertainty in the number of changes in multi-armed bandit literature?", "answer": ["A New Algorithm for Non-stationary Contextual Bandits: Efficient, Optimal, and Parameter-free"], "answer_arxiv_id": ["1902.00980"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_8347"} +{"question": "What studies have been conducted on mutual learning approaches in the context of self-distillation?", "answer": ["Deep Mutual Learning", "R-Drop: Regularized Dropout for Neural Networks"], "answer_arxiv_id": ["1706.00384", "2106.14448"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_8348"} +{"question": "What works proposed pioneering methods for 3D occupancy prediction using occupancy grids?", "answer": ["Two Stream 3D Semantic Scene Completion", "LMSCNet: Lightweight Multiscale 3D Semantic Completion", "SCFusion: Real-time Incremental Scene Reconstruction with Semantic\n Completion", "Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning\n Contextual Shape Priors from Scene Completion"], "answer_arxiv_id": ["1804.03550", "2008.10559", "2010.13662", "2012.03762"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_8349"} +{"question": "In what work bounded similarity matching (BSM) was proposed for uncorrelated source separation?", "answer": ["Blind Bounded Source Separation Using Neural Networks with Local Learning Rules"], "answer_arxiv_id": ["2004.05479"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_8350"} +{"question": "Which work introduced the concept of 'grokking,' where models gain generalization capabilities when training significantly beyond overfitting?", "answer": ["Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets"], "answer_arxiv_id": ["2201.02177"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_8351"} +{"question": "Which works have explored the use of calculators in large language models?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "ToolkenGPT: Augmenting Frozen Language Models with Massive Tools via\n Tool Embeddings"], "answer_arxiv_id": ["2302.04761", "2305.11554"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8352"} +{"question": "What are the papers that discussed the concept of approximate symmetry in literature?", "answer": ["Learning Invariances using the Marginal Likelihood", "Incorporating Symmetry into Deep Dynamics Models for Improved Generalization"], "answer_arxiv_id": ["1808.05563", "2002.03061"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_8353"} +{"question": "What works have utilized video-based methods for human pose and shape estimation?", "answer": ["Learning 3D Human Dynamics from Video", "VIBE: Video Inference for Human Body Pose and Shape Estimation", "Beyond Static Features for Temporally Consistent 3D Human Pose and Shape from a Video", "Encoder-decoder with Multi-level Attention for 3D Human Shape and Pose Estimation"], "answer_arxiv_id": ["1812.01601", "1912.05656", "2011.08627", "2109.02303"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_8354"} +{"question": "Could you provide me with the reference discussing that the variants of the method either fail to achieve spectral accuracy or require an explicit Mercer decomposition?", "answer": ["Positively Weighted Kernel Quadrature via Subsampling"], "answer_arxiv_id": ["2107.09597"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_8355"} +{"question": "What studies proposed to discriminate the region-wise message passing with graph attention networks?", "answer": ["GMAN: A Graph Multi-Attention Network for Traffic Prediction"], "answer_arxiv_id": ["1911.08415"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_8356"} +{"question": "Could you provide me some studies that look into representation learning for reinforcement learning?", "answer": ["On the Power of Multitask Representation Learning in Linear MDP", "Provable General Function Class Representation Learning in Multitask Bandits and MDPs", "Joint Representation Training in Sequential Tasks with Shared Structure", "Provably Efficient Multi-Task Reinforcement Learning with Model Transfer", "Provable Benefit of Multitask Representation Learning in Reinforcement Learning", "Provable Benefits of Representational Transfer in Reinforcement Learning"], "answer_arxiv_id": ["2106.08053", "2205.15701", "2206.12441", "2107.08622", "2206.05900v1", "2205.14571v2"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_8357"} +{"question": "Who proposed the Transformer Framework for fMRI that utilizes a Transformer model for temporal feature extraction?", "answer": ["Self-Supervised Transformers for fMRI representation"], "answer_arxiv_id": ["2112.05761"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_8358"} +{"question": "What studies explore the use of Gaussian Processes in the context of time-index models?", "answer": ["Time series forecasting with Gaussian Processes needs priors"], "answer_arxiv_id": ["2009.08102"], "source_meta": {"published_time": "20220713"}, "qid": "AutoScholarQuery_train_8359"} +{"question": "Which studies explore diverse prompting techniques to enhance the quality of prompts in image generation for diffusion models?", "answer": ["DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic\n Segmentation Using Diffusion Models", "ImaginaryNet: Learning Object Detectors without Real Images and\n Annotations", "Is synthetic data from generative models ready for image recognition?"], "answer_arxiv_id": ["2303.11681", "2210.06886", "2210.07574"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_8360"} +{"question": "What studies proposed memory-augmented networks capable of performing a linear number of steps per token to solve superlinear tasks?", "answer": ["Improving the Neural GPU Architecture for Algorithm Learning", "Neural GPUs Learn Algorithms", "Extensions and Limitations of the Neural GPU"], "answer_arxiv_id": ["1702.08727", "1511.08228", "1611.00736"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_8361"} +{"question": "Which studies attempted to update a few pre-trained network parameters during fine-tuning in early PETL?", "answer": ["BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models"], "answer_arxiv_id": ["2106.10199"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_8362"} +{"question": "Any literature that discusses performing diffusion in a low dimensional latent space or in a down-sampled pixel space?", "answer": ["Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Improved Vector Quantized Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Unleashing Transformers: Parallel Token Prediction with Discrete\n Absorbing Diffusion for Fast High-Resolution Image Generation from\n Vector-Quantized Codes", "ImageBART: Bidirectional Context with Multinomial Diffusion for\n Autoregressive Image Synthesis", "Global Context with Discrete Diffusion in Vector Quantised Modelling for\n Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2111.14822", "2205.16007", "2112.10752", "2111.12701", "2108.08827", "2112.01799", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_8363"} +{"question": "Which research introduces a re-parameterization method to trade-off between old and new knowledge in DFCIL?", "answer": ["Self-Sustaining Representation Expansion for Non-Exemplar\n Class-Incremental Learning"], "answer_arxiv_id": ["2203.06359"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_8364"} +{"question": "Could you list the works that aimed to capture geometry information using deep features in piece matching?", "answer": ["Neural Shape Mating: Self-Supervised Object Assembly with Adversarial Shape Priors"], "answer_arxiv_id": ["2205.14886"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_8365"} +{"question": "What studies demonstrated that a semantic parse of a natural language can be generated by asking a large language model to continue a prompt that includes the sentence?", "answer": ["MTOP: A Comprehensive Multilingual Task-Oriented Semantic Parsing Benchmark", "The Power of Prompt Tuning for Low-Resource Semantic Parsing"], "answer_arxiv_id": ["2008.09335", "2110.08525"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_8366"} +{"question": "What works built on the initial prompting methods of transformer networks, showing improvements in performance?", "answer": ["Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigence", "PromptFusion: Decoupling Stability and Plasticity for Continual Learning"], "answer_arxiv_id": ["1801.10112", "2303.07223"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_8367"} +{"question": "Which papers provide a theoretical investigation of data augmentations as label-preserving group actions and discuss an inherent invariance-variance trade-off?", "answer": ["A Group-Theoretic Framework for Data Augmentation"], "answer_arxiv_id": ["1907.10905"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_8368"} +{"question": "What works facilitated the unsupervised training of FMNet by incorporating isometry losses in both spatial and spectral domains?", "answer": ["Unsupervised Deep Learning for Structured Shape Matching"], "answer_arxiv_id": ["1812.03794"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_8369"} +{"question": "Which study created six new datasets for link prediction on continuous-time dynamic graphs?", "answer": ["Towards Better Evaluation for Dynamic Link Prediction"], "answer_arxiv_id": ["2207.10128"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_8370"} +{"question": "Which papers studies visual ambiguities caused by optical illusions?", "answer": ["HallusionBench: An Advanced Diagnostic Suite for Entangled Language\n Hallucination and Visual Illusion in Large Vision-Language Models", "Holistic Analysis of Hallucination in GPT-4V(ision): Bias and\n Interference Challenges", "A Challenger to GPT-4V? Early Explorations of Gemini in Visual Expertise"], "answer_arxiv_id": ["2310.14566", "2311.03287", "2312.12436"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_8371"} +{"question": "Could you provide me some studies about hierarchical interpolation and multi-rate data sampling specially for the LSTF task?", "answer": ["N-HiTS: Neural Hierarchical Interpolation for Time Series Forecasting"], "answer_arxiv_id": ["2201.12886"], "source_meta": {"published_time": "20220713"}, "qid": "AutoScholarQuery_train_8372"} +{"question": "What study demonstrated the slow performance of the REC algorithm?", "answer": ["Adaptive Greedy Rejection Sampling"], "answer_arxiv_id": ["2304.10407"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_8373"} +{"question": "What are some works that approached the offline RL problem from a pessimistic value iteration perspective under the single-policy concentrability assumption?", "answer": ["Is Pessimism Provably Efficient for Offline RL?", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Settling the Sample Complexity of Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2012.15085", "2103.12021v2", "2204.05275"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_8374"} +{"question": "Could you provide me some studies that introduced synthetic datasets for fluid dynamics?", "answer": ["Graph neural networks for laminar flow prediction around random two-dimensional shapes", "Learning Mesh-Based Simulation with Graph Networks", "Predicting physics in mesh-reduced space with temporal attention", "Learned Coarse Models for Efficient Turbulence Simulation"], "answer_arxiv_id": ["2107.11529", "2010.03409", "2201.09113", "2112.15275"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_8375"} +{"question": "Any works about the use of DINO-extracted features in instance segmentation?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.14294"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8376"} +{"question": "Are there any researches that proposed richer modalities for multi-modal large language models?", "answer": ["PandaGPT: One Model To Instruction-Follow Them All"], "answer_arxiv_id": ["2305.16355"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_8377"} +{"question": "Could you provide me some studies that work on state space entropy maximization?", "answer": ["Provably Efficient Maximum Entropy Exploration", "Efficient Exploration via State Marginal Matching"], "answer_arxiv_id": ["1812.02690", "1906.05274"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_8378"} +{"question": "Which works proved that a certain number of random variables were sufficient for the existence of a solution to the random Subset-Sum problem?", "answer": ["Revisiting the Random Subset Sum Problem"], "answer_arxiv_id": ["2204.13929"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_8379"} +{"question": "Could you provide me some papers that proposed solutions for boundary discontinuity in angle regression?", "answer": ["Dense Label Encoding for Boundary Discontinuity Free Rotation Detection", "Phase-Shifting Coder: Predicting Accurate Orientation in Oriented Object\n Detection"], "answer_arxiv_id": ["2011.09670", "2211.06368"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_8380"} +{"question": "Could you give an example of research that uses text-to-image diffusion and image-based guidance to generate 3D shape neural fields?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_8381"} +{"question": "What papers has developed accelerations for NeRF rendering?", "answer": ["Efficient Region-Aware Neural Radiance Fields for High-Fidelity Talking\n Portrait Synthesis"], "answer_arxiv_id": ["2307.09323"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_8382"} +{"question": "What research work suggests modifications to the architecture that introduce more separability and decorrelation similar to Forward Gradient?", "answer": ["Local Learning with Neuron Groups"], "answer_arxiv_id": ["2301.07635"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_8383"} +{"question": "What papers justify the action hierarchy design by examining activities in sport and kitchen contexts?", "answer": ["FineGym: A Hierarchical Video Dataset for Fine-grained Action Understanding", "FineDiving: A Fine-grained Dataset for Procedure-aware Action Quality Assessment", "Rescaling Egocentric Vision"], "answer_arxiv_id": ["2004.06704", "2204.03646", "2006.13256"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_8384"} +{"question": "Which works are related to sentiment detection in financial NLP tasks?", "answer": ["Good Debt or Bad Debt: Detecting Semantic Orientations in Economic Texts"], "answer_arxiv_id": ["1307.5336"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_8385"} +{"question": "Which research paper features ZebraPose, a method that uses coarse-to-fine surface encoding to represent correspondences?", "answer": ["ZebraPose: Coarse to Fine Surface Encoding for 6DoF Object Pose\n Estimation"], "answer_arxiv_id": ["2203.09418"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_8386"} +{"question": "What works proposed the idea of language-image contrastive learning (CLIP) for pretraining?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_8387"} +{"question": "What papers support all possible growth dimensions in progressive training?", "answer": ["bert2BERT: Towards Reusable Pretrained Language Models", "Learning to Grow Pretrained Models for Efficient Transformer Training"], "answer_arxiv_id": ["2110.07143", "2303.00980"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_8388"} +{"question": "Which research papers have focused on the problem of pure exploration in quantum bandits?", "answer": ["Quantum Bandits", "Quantum Exploration Algorithms for Multi-Armed Bandits"], "answer_arxiv_id": ["2002.06395", "2007.07049"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_8389"} +{"question": "Could you indicate the studies that made efforts towards combining NeRF models with explicit human models?", "answer": ["Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "Neural Actor: Neural Free-view Synthesis of Human Actors with Pose\n Control", "PINA: Learning a Personalized Implicit Neural Avatar from a Single RGB-D\n Video Sequence", "Capturing and Animation of Body and Clothing from Monocular Video", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video"], "answer_arxiv_id": ["2012.15838", "2106.02019", "2203.01754", "2210.01868", "2201.04127"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_8390"} +{"question": "Can you name some works that introduce new architectures to directly incorporate uncertainty estimates?", "answer": ["Predictive Uncertainty Estimation via Prior Networks", "Posterior Network: Uncertainty Estimation without OOD Samples via Density-Based Pseudo-Counts", "On Feature Collapse and Deep Kernel Learning for Single Forward Pass Uncertainty"], "answer_arxiv_id": ["1802.10501", "2006.09239", "2102.11409"], "source_meta": {"published_time": "20210719"}, "qid": "AutoScholarQuery_train_8391"} +{"question": "Which works are utilizing uncertainty estimation in model-based algorithms for addressing the distribution shift in offline RL?", "answer": ["MOPO: Model-based Offline Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2005.13239", "2005.05951"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_8392"} +{"question": "What papers introduced methods to enhance the training efficiency in Vision Language Models?", "answer": ["LiT: Zero-Shot Transfer with Locked-image text Tuning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2111.07991", "2301.12597"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_8393"} +{"question": "Which works introduced large kernel in the self-attention method?", "answer": ["Visual Attention Network"], "answer_arxiv_id": ["2202.09741"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_8394"} +{"question": "Which works utilize post-hoc methods due to their easy to use without modifying the training procedure and objective?", "answer": ["A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "Energy-based Out-of-distribution Detection", "On the Importance of Gradients for Detecting Distributional Shifts in the Wild", "ReAct: Out-of-distribution Detection With Rectified Activations", "RankFeat: Rank-1 Feature Removal for Out-of-distribution Detection"], "answer_arxiv_id": ["1807.03888", "1706.02690", "2010.03759", "2110.00218", "2111.12797", "2209.08590"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_8395"} +{"question": "Which studies propose a diffusion model formulation that enable high quality image synthesis?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_8396"} +{"question": "Could you list some studies that are based on single-view view synthesis for 3-D scene reconstruction?", "answer": ["3D Ken Burns Effect from a Single Image", "Single-View View Synthesis with Multiplane Images", "GRF: Learning a General Radiance Field for 3D Representation and\n Rendering", "pixelNeRF: Neural Radiance Fields from One or Few Images", "MINE: Towards Continuous Depth MPI with NeRF for Novel View Synthesis", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis", "SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single\n Image", "Generative Novel View Synthesis with 3D-Aware Diffusion Models"], "answer_arxiv_id": ["1909.05483", "2004.11364", "2010.04595", "2012.02190", "2103.14910", "2012.00926", "2204.00928", "2304.02602"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8397"} +{"question": "What research proposed the Regret Minimization Experience Replay (ReMERN)?", "answer": ["Regret Minimization Experience Replay in Off-Policy Reinforcement Learning"], "answer_arxiv_id": ["2105.07253v3"], "source_meta": {"published_time": "20220822"}, "qid": "AutoScholarQuery_train_8398"} +{"question": "Which works involve pretraining state representations for capturing dynamical information of the environment?", "answer": ["Provable Representation Learning for Imitation with Contrastive Fourier Features"], "answer_arxiv_id": ["2105.12272"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_8399"} +{"question": "Are there any works related to solving the orthogonal Procrustes problem between sets of representations?", "answer": ["Generalized Shape Metrics on Neural Representations"], "answer_arxiv_id": ["2110.14739"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_8400"} +{"question": "Can you name the studies that used schema for pronoun resolution as AI challenges?", "answer": ["Gender Bias in Coreference Resolution"], "answer_arxiv_id": ["1804.09301"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_8401"} +{"question": "What study introduced the C&W attack framework in point cloud attacks?", "answer": ["Generating 3D Adversarial Point Clouds"], "answer_arxiv_id": ["1809.07016"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_8402"} +{"question": "Which works proposed Neural Ordinary Differential Equations (NODE) to model dynamics?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_8403"} +{"question": "What studies talk about the emergent mechanism known as in-context learning in LLMs?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_8404"} +{"question": "Which study makes the appearance embedding a global representation across different views?", "answer": ["Hallucinated Neural Radiance Fields in the Wild"], "answer_arxiv_id": ["2111.15246"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8405"} +{"question": "Which work showed that a class of continuous-time RNNs can be written as input-driven ODEs and provided a generalization bound?", "answer": ["Framing RNN as a kernel method: A neural ODE approach"], "answer_arxiv_id": ["2106.01202"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_8406"} +{"question": "Which works focused on generating human motion with audio signals?", "answer": ["AI Choreographer: Music Conditioned 3D Dance Generation with AIST++", "EDGE: Editable Dance Generation From Music"], "answer_arxiv_id": ["2101.08779", "2211.10658"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_8407"} +{"question": "Any studies on performing offline high-fidelity reconstruction by combining neural networks with numerical solvers?", "answer": ["A Physics-informed Diffusion Model for High-fidelity Flow Field Reconstruction", "Super-resolution reconstruction of turbulent flows with machine learning", "Data-Driven Synthesis of Smoke Flows with CNN-based Feature Descriptors"], "answer_arxiv_id": ["2211.14680", "1811.11328", "1705.01425"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8408"} +{"question": "Can you name the paper that gives a comprehensive overview of the OOD generalization problem?", "answer": ["Towards Out-Of-Distribution Generalization: A Survey"], "answer_arxiv_id": ["2108.13624"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_8409"} +{"question": "What research papers proposed variants of Mixup for data augmentation?", "answer": ["CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features", "TransMix: Attend to Mix for Vision Transformers", "Manifold Mixup: Better Representations by Interpolating Hidden States", "Evolving Image Compositions for Feature Representation Learning", "Puzzle Mix: Exploiting Saliency and Local Statistics for Optimal Mixup", "Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification", "SaliencyMix: A Saliency Guided Data Augmentation Strategy for Better Regularization"], "answer_arxiv_id": ["1905.04899", "2111.09833", "1806.05236", "2106.09011", "2009.06962v2", "2003.13048", "2006.01791"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_8410"} +{"question": "Could you provide an example of work that introduces scaling invariance in DGL via cosine similarity?", "answer": ["Inverting Gradients - How easy is it to break privacy in federated learning?"], "answer_arxiv_id": ["2003.14053"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_8411"} +{"question": "Are there any research papers using KD for medical image segmentation?", "answer": ["Knowledge distillation from multi-modal to mono-modal segmentation networks", "Towards Cross-modality Medical Image Segmentation with Online Mutual Knowledge Distillation"], "answer_arxiv_id": ["2106.09564", "2010.01532"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_8412"} +{"question": "What studies or surveys provide insights into the design of message-passing graph neural networks for molecules?", "answer": ["Geometrically Equivariant Graph Neural Networks: A Survey"], "answer_arxiv_id": ["2202.07230"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_8413"} +{"question": "What research efforts have been made to minimize the trade-off and directly use adversarial examples as data augmentation?", "answer": ["Understanding and Mitigating the Tradeoff Between Robustness and Accuracy", "Revisiting adapters with adversarial training", "Adversarial Examples Improve Image Recognition"], "answer_arxiv_id": ["2002.10716", "2210.04886", "1911.09665"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_8414"} +{"question": "Can you provide some articles about path patching and causal scrubbing techniques?", "answer": ["Towards Automated Circuit Discovery for Mechanistic Interpretability", "Localizing Model Behavior with Path Patching"], "answer_arxiv_id": ["2304.14997", "2304.05969"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8415"} +{"question": "Which papers discussed the application of Deep Set and GNN to Multi-Agent Reinforcement Learning (MARL)?", "answer": ["From Few to More: Large-scale Dynamic Multiagent Curriculum Learning", "Permutation Invariant Policy Optimization for Mean-Field Multi-Agent Reinforcement Learning: A Principled Approach", "Graph Convolutional Reinforcement Learning", "PIC: Permutation Invariant Critic for Multi-Agent Deep Reinforcement Learning"], "answer_arxiv_id": ["1909.02790", "2105.08268", "1810.09202", "1911.00025"], "source_meta": {"published_time": "20220310"}, "qid": "AutoScholarQuery_train_8416"} +{"question": "What papers have aimed to address monocular flow estimation in fluid settings?", "answer": ["Global Transport for Fluid Reconstruction with Learned Self-Supervision"], "answer_arxiv_id": ["2104.06031"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_8417"} +{"question": "What work proposes a diffusion model conditioned on rasterized FLAME meshes for avatar animation?", "answer": ["DiffusionRig: Learning Personalized Priors for Facial Appearance Editing"], "answer_arxiv_id": ["2304.06711"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8418"} +{"question": "Which papers delve into improving optimization strategies, specifically through designing loss function to better address class imbalance?", "answer": ["Focal Loss for Dense Object Detection"], "answer_arxiv_id": ["1708.02002"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_8419"} +{"question": "Could you provide me some papers dealing with alternative approaches to replace manual resets?", "answer": ["Leave no Trace: Learning to Reset for Safe and Autonomous Reinforcement Learning", "Automating Reinforcement Learning with Example-based Resets"], "answer_arxiv_id": ["1711.06782", "2204.02041"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_8420"} +{"question": "Could you provide me study examples on developing query-based black-box adversarial attacks?", "answer": ["Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models", "ZOO: Zeroth Order Optimization Based Black-box Attacks to Deep Neural Networks without Training Substitute Models", "Triangle Attack: A Query-efficient Decision-based Adversarial Attack"], "answer_arxiv_id": ["1712.04248", "1708.03999", "2112.06569"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_8421"} +{"question": "Which works tackle subject identity generalization in human reconstruction research?", "answer": ["LOLNeRF: Learn from One Look", "pixelNeRF: Neural Radiance Fields from One or Few Images", "IBRNet: Learning Multi-View Image-Based Rendering", "PVA: Pixel-aligned Volumetric Avatars", "Neural Human Performer: Learning Generalizable Radiance Fields for Human Performance Rendering", "Geometry-Guided Progressive NeRF for Generalizable and Efficient Neural Human Rendering", "HumanNeRF: Efficiently Generated Human Radiance Field from Sparse Inputs", "Generalizable Neural Performer: Learning Robust Radiance Fields for Human Novel View Synthesis"], "answer_arxiv_id": ["2111.09996", "2012.02190", "2102.13090", "2101.02697v1", "2109.07448", "2112.04312", "2112.02789", "2204.11798"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_8422"} +{"question": "Which works suggested adopting a unified perspective with Poisson Flow Generative Models as an approach for diffusion process?", "answer": ["PFGM++: Unlocking the Potential of Physics-Inspired Generative Models"], "answer_arxiv_id": ["2302.04265"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_8423"} +{"question": "Which studies used RL-LNS and bipartite graph representations of ILPs to learn the destroy heuristics represented by GCNs?", "answer": ["Learning Large Neighborhood Search Policy for Integer Programming", "Learning a Large Neighborhood Search Algorithm for Mixed Integer Programs"], "answer_arxiv_id": ["2111.03466", "2107.10201"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_8424"} +{"question": "Any works about transforming-based models for temporal action segmentation?", "answer": ["ASFormer: Transformer for Action Segmentation", "Unified Fully and Timestamp Supervised Temporal Action Segmentation via Sequence to Sequence Translation"], "answer_arxiv_id": ["2110.08568", "2209.00638"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8425"} +{"question": "Can you specify the studies investigating the boundary of the Frequency Principle?", "answer": ["Machine Learning from a Continuous Viewpoint I"], "answer_arxiv_id": ["1912.12777"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_8426"} +{"question": "Which prior work examined the intersection of the right to explanation and the right to be forgotten?", "answer": ["On the Trade-Off between Actionable Explanations and the Right to be Forgotten"], "answer_arxiv_id": ["2208.14137"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_8427"} +{"question": "Could you mention some works that try to avoid instance discrimination tasks by proposing alternative methods?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "FaceMAE: Privacy-Preserving Face Recognition via Masked Autoencoders", "How to Boost Face Recognition with StyleGAN?"], "answer_arxiv_id": ["2104.14294", "2205.11090", "2210.10090"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_8428"} +{"question": "In what papers is a neural SDF representation introduced for surface reconstruction?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction"], "answer_arxiv_id": ["2106.10689"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_8429"} +{"question": "Can you list the studies that used diffusion models by conditioning on the given image?", "answer": ["Multiscale Structure Guided Diffusion for Image Deblurring", "Palette: Image-to-Image Diffusion Models", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2212.01789", "2111.05826", "2201.09865"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8430"} +{"question": "Could you tell me about the datasets containing human motion data along with text descriptions?", "answer": ["The KIT Motion-Language Dataset", "Motion-X: A Large-scale 3D Expressive Whole-body Human Motion Dataset"], "answer_arxiv_id": ["1607.03827", "2307.00818"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_8431"} +{"question": "What work has further investigated the notion of PAC stabilizability, used for ε-fractional core stability, in the context of HGs with underlying interaction networks?", "answer": ["Forming Probably Stable Communities with Limited Interactions"], "answer_arxiv_id": ["1811.04616"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_8432"} +{"question": "Could you provide me some works that considered priors over the optimum in the single-fidelity setting?", "answer": ["Bayesian Optimization with a Prior for the Optimum"], "answer_arxiv_id": ["2006.14608"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_8433"} +{"question": "Could you tell me about the datasets equipped with first order logic (FOL) annotations?", "answer": ["FOLIO: Natural Language Reasoning with First-Order Logic"], "answer_arxiv_id": ["2209.00840"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_8434"} +{"question": "Which papers works on the problem of choosing task weights in ATL?", "answer": ["In Defense of the Unitary Scalarization for Deep Multi-Task Learning"], "answer_arxiv_id": ["2201.04122"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_8435"} +{"question": "Are there any research works that use the Myers–Briggs Type Indicator (MBTI) for LLM evaluations?", "answer": ["Can ChatGPT Assess Human Personalities? A General Evaluation Framework", "Do LLMs Possess a Personality? Making the MBTI Test an Amazing\n Evaluation for Large Language Models"], "answer_arxiv_id": ["2303.01248", "2307.16180"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_8436"} +{"question": "Any works about exploring adaptive mixing strategies in multilingual pretraining?", "answer": ["Balancing Average and Worst-case Accuracy in Multitask Learning"], "answer_arxiv_id": ["2110.05838"], "source_meta": {"published_time": "20230418"}, "qid": "AutoScholarQuery_train_8437"} +{"question": "What research use different conditions like image, video, and 3D sketch for guiding generation in Controllable Generation?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "ControlVideo: Training-free Controllable Text-to-Video Generation", "Control-A-Video: Controllable Text-to-Video Generation with Diffusion\n Models", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures"], "answer_arxiv_id": ["2208.12242", "2302.13848", "2305.13077", "2305.13840", "2211.07600"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8438"} +{"question": "What documents present the extraction of temporal relations among events?", "answer": ["Joint Reasoning for Temporal and Causal Relations", "Joint Constrained Learning for Event-Event Relation Extraction"], "answer_arxiv_id": ["1906.04941", "2010.06727"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_8439"} +{"question": "Could you provide me some works that used monocular mocap methods using optimization?", "answer": ["Towards Accurate Markerless Human Shape and Pose Estimation over Time", "Unite the People: Closing the Loop Between 3D and 2D Human\n Representations", "Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a\n Single Image", "Convolutional Mesh Regression for Single-Image Human Shape\n Reconstruction"], "answer_arxiv_id": ["1707.07548", "1701.02468", "1607.08128", "1905.03244"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8440"} +{"question": "What works have been conducted on certified and provable defense techniques for FL?", "answer": ["SparseFed: Mitigating Model Poisoning Attacks in Federated Learning with Sparsification", "Provably Secure Federated Learning against Malicious Clients", "CRFL: Certifiably Robust Federated Learning against Backdoor Attacks"], "answer_arxiv_id": ["2112.06274v1", "2102.01854", "2106.08283"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_8441"} +{"question": "Can you mention studies tackling issues in multiple IL settings, like offline LfO, cross-domain LfO, and cross-domain offline IL?", "answer": ["Offline Learning from Demonstrations and Unlabeled Experience", "State Alignment-based Imitation Learning", "Cross-domain Imitation from Observations"], "answer_arxiv_id": ["2011.13885", "1911.10947", "2105.10037"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_8442"} +{"question": "Could you give examples of research papers that employ zero-shot learning?", "answer": ["Zero-Shot Learning -- A Comprehensive Evaluation of the Good, the Bad\n and the Ugly"], "answer_arxiv_id": ["1707.00600"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_8443"} +{"question": "What studies used prompts to transfer visual models trained with natural language supervision to downstream tasks in zero-shot setting?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_8444"} +{"question": "Could you provide me some works about recent models trained with coding instructions?", "answer": ["CodeT5+: Open Code Large Language Models for Code Understanding and\n Generation", "WizardCoder: Empowering Code Large Language Models with Evol-Instruct", "PanGu-Coder2: Boosting Large Language Models for Code with Ranking\n Feedback"], "answer_arxiv_id": ["2305.07922", "2306.08568", "2307.14936"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_8445"} +{"question": "What studies have refined these theoretical analyses by considering linear computation steps and logarithmic precision?", "answer": ["A Survey of Neural Networks and Formal Languages", "On the Ability and Limitations of Transformers to Recognize Formal Languages", "Theoretical Limitations of Self-Attention in Neural Sequence Models", "Formal Language Recognition by Hard Attention Transformers: Perspectives from Circuit Complexity", "On the Computational Power of RNNs", "Sequential Neural Networks as Automata", "A Formal Hierarchy of RNN Architectures", "On the Practical Computational Power of Finite Precision RNNs for Language Recognition"], "answer_arxiv_id": ["2006.01338v1", "2009.11264", "1906.06755", "2204.06618", "1906.06349", "1906.01615", "2004.08500", "1805.04908"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_8446"} +{"question": "Could you provide me some works where the theoretical aspects of self-distillation are analyzed?", "answer": ["Self-Distillation Amplifies Regularization in Hilbert Space"], "answer_arxiv_id": ["2002.05715v3"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_8447"} +{"question": "Could you provide me some studies about music style transfer?", "answer": ["Play as You Like: Timbre-enhanced Multi-modal Music Style Transfer", "MelGAN-VC: Voice Conversion and Audio Style Transfer on arbitrarily long samples using Spectrograms"], "answer_arxiv_id": ["1811.12214", "1910.03713"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_8448"} +{"question": "What works are related to the proposition of latent dynamics models for high-dimensional inputs?", "answer": ["PlaNet of the Bayesians: Reconsidering and Improving Deep Planning Network by Incorporating Bayesian Inference"], "answer_arxiv_id": ["2003.00370"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_8449"} +{"question": "Which papers discuss the challenge of ensuring fine-grained physical plausibility in physical skills learning?", "answer": ["PhysDiff: Physics-Guided Human Motion Diffusion Model", "Holistic++ Scene Understanding: Single-view 3D Holistic Scene Parsing\n and Human Pose Estimation with Human-Object Interaction and Physical\n Commonsense", "Synthesizing Long-Term 3D Human Motion and Interaction in 3D Scenes", "Scene-aware Generative Network for Human Motion Synthesis", "MIME: Human-Aware 3D Scene Generation", "Generating 3D People in Scenes without People", "Compositional Human-Scene Interaction Synthesis with Semantic Control", "Locomotion-Action-Manipulation: Synthesizing Human-Scene Interactions in\n Complex 3D Environments", "EDGE: Editable Dance Generation From Music", "3D Human Pose Estimation via Intuitive Physics", "Stochastic Scene-Aware Motion Prediction", "Full-Body Articulated Human-Object Interaction", "Human-centric Indoor Scene Synthesis Using Stochastic Grammar", "ProBio: A Protocol-guided Multimodal Dataset for Molecular Biology Lab", "Move as You Say, Interact as You Can: Language-guided Human Motion\n Generation with Scene Affordance", "COAP: Compositional Articulated Occupancy of People", "Recovering 3D Human Mesh from Monocular Images: A Survey", "Humanoid Self-Collision Avoidance Using Whole-Body Control with Control\n Barrier Functions", "Physics-based Human Motion Estimation and Synthesis from Videos", "Differentiable Dynamics for Articulated 3d Human Motion Reconstruction", "Motron: Multimodal Probabilistic Human Motion Forecasting", "Isaac Gym: High Performance GPU-Based Physics Simulation For Robot\n Learning", "AMP: Adversarial Motion Priors for Stylized Physics-Based Character\n Control", "PADL: Language-Directed Physics-Based Character Control", "CALM: Conditional Adversarial Latent Models for Directable Virtual\n Characters", "ASE: Large-Scale Reusable Adversarial Skill Embeddings for Physically\n Simulated Characters"], "answer_arxiv_id": ["2212.02500", "1909.01507", "2012.05522", "2105.14804", "2212.04360", "1912.02923", "2207.12824", "2301.02667", "2211.10658", "2303.18246", "2108.08284", "2212.10621", "1808.08473", "2311.00556v1", "2403.18036", "2204.06184", "2203.01923", "2207.00692", "2109.09913", "2205.12256", "2203.04132", "2108.10470", "2104.02180", "2301.13868", "2305.02195", "2205.01906"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_8450"} +{"question": "What study integrated NeRF for edge mapping from multi-view images in the context of learning-based 3D line/curve reconstruction?", "answer": ["NEF: Neural Edge Fields for 3D Parametric Curve Reconstruction from\n Multi-view Images"], "answer_arxiv_id": ["2303.07653"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_8451"} +{"question": "What studies leverage non-stationary data for causal discovery?", "answer": ["Causal Discovery and Forecasting in Nonstationary Environments with State-Space Models"], "answer_arxiv_id": ["1905.10857"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_8452"} +{"question": "Which works proposed a theory of neuron activation subspace match and algorithms to compute such matches?", "answer": ["Towards Understanding Learning Representations: To What Extent Do Different Neural Networks Learn the Same Representation"], "answer_arxiv_id": ["1810.11750"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_8453"} +{"question": "Can you showcase any research papers where the method of learning a mask over weights based on Fisher information is utilized as a Parameter Efficient FineTuning (PEFT) method?", "answer": ["Training Neural Networks with Fixed Sparse Masks"], "answer_arxiv_id": ["2111.09839"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_8454"} +{"question": "Which works focus on conditional sequence modelling in offline RL - learning a policy based on a particular metric for future trajectories?", "answer": ["Training Agents using Upside-Down Reinforcement Learning", "Decision Transformer: Reinforcement Learning via Sequence Modeling", "End-to-end Driving via Conditional Imitation Learning", "Learning to Reach Goals via Iterated Supervised Learning", "Learning Latent Plans from Play", "Generalized Decision Transformer for Offline Hindsight Information Matching"], "answer_arxiv_id": ["1912.02877", "2106.01345", "1710.02410", "1912.06088", "1903.01973", "2111.10364"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_8455"} +{"question": "Which research works have considered differentiable function approximation (DFA) for the off-policy evaluation (OPE) task?", "answer": ["Off-Policy Fitted Q-Evaluation with Differentiable Function Approximators: Z-Estimation and Inference Theory"], "answer_arxiv_id": ["2202.04970"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_8456"} +{"question": "What research deals with extracting policy from the dataset through weighted regression to reduce extrapolation error?", "answer": ["Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning", "AWAC: Accelerating Online Reinforcement Learning with Offline Datasets", "Critic Regularized Regression"], "answer_arxiv_id": ["1910.00177", "2006.09359", "2006.15134"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_8457"} +{"question": "Could you provide me some studies about improving video synthesis by finetuning pre-trained text-to-image diffusion models on video data?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "LAMP: Learn A Motion Pattern for Few-Shot-Based Video Generation"], "answer_arxiv_id": ["2304.08818", "2209.14792", "2310.10769"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_8458"} +{"question": "What methods use representation interpolation for data augmentation in computer vision research?", "answer": ["mixup: Beyond Empirical Risk Minimization"], "answer_arxiv_id": ["1710.09412"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_8459"} +{"question": "Which papers have discussed the importance of model diversity in ensembles for their generalization performance?", "answer": ["Ensemble deep learning: A review"], "answer_arxiv_id": ["2104.02395"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_8460"} +{"question": "What paper includes realizing the ray sampling in a log scale and uses the depth priors to improve ray sampling?", "answer": ["DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields using Depth Oracle Networks"], "answer_arxiv_id": ["2103.03231"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_8461"} +{"question": "What papers proposed applying GCN and GRU for addressing spatial correlations and temporal dependencies in multivariate time series forecasting?", "answer": ["Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting"], "answer_arxiv_id": ["2007.02842"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_8462"} +{"question": "What studies refine biased pseudo-labels by iteratively solving a convex optimization problem and blending semantic pseudo-labels and linear pseudo-labels?", "answer": ["Distribution Aligning Refinery of Pseudo-label for Imbalanced\n Semi-supervised Learning", "DASO: Distribution-Aware Semantics-Oriented Pseudo-label for Imbalanced\n Semi-Supervised Learning"], "answer_arxiv_id": ["2007.08844", "2106.05682"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_8463"} +{"question": "Which works developed datasets for hand-drawn animation production?", "answer": ["Deep Animation Video Interpolation in the Wild", "AnimeRun: 2D Animation Visual Correspondence from Open Source 3D Movies"], "answer_arxiv_id": ["2104.02495", "2211.05709"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_8464"} +{"question": "Which papers implemented a projection-based baseline that performs on par with 3D approaches in a point cloud classification task?", "answer": ["Deep Projective 3D Semantic Segmentation", "Revisiting Point Cloud Shape Classification with a Simple and Effective\n Baseline"], "answer_arxiv_id": ["1705.03428", "2106.05304"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_8465"} +{"question": "Which works utilized temperature scaling to rescale the neural network’s logits?", "answer": ["On Calibration of Modern Neural Networks"], "answer_arxiv_id": ["1706.04599"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_8466"} +{"question": "Could you provide me some papers about proposed losses for training retrieval models?", "answer": ["Large-Margin Softmax Loss for Convolutional Neural Networks", "SphereFace: Deep Hypersphere Embedding for Face Recognition", "CosFace: Large Margin Cosine Loss for Deep Face Recognition", "ArcFace: Additive Angular Margin Loss for Deep Face Recognition", "Deep Metric Learning via Lifted Structured Feature Embedding", "Representation Learning with Contrastive Predictive Coding", "Multi-Similarity Loss with General Pair Weighting for Deep Metric\n Learning", "Supervised Contrastive Learning", "Circle Loss: A Unified Perspective of Pair Similarity Optimization"], "answer_arxiv_id": ["1612.02295", "1704.08063", "1801.09414", "1801.07698", "1511.06452", "1807.03748", "1904.06627", "2004.11362", "2002.10857"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_8467"} +{"question": "Can you list the studies on dynamic sparsity in neural networks?", "answer": ["Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science", "Sparse evolutionary Deep Learning with over one million artificial neurons on commodity hardware", "Do We Actually Need Dense Over-Parameterization? In-Time Over-Parameterization in Sparse Training", "Rigging the Lottery: Making All Tickets Winners", "Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization", "Sparse Networks from Scratch: Faster Training without Losing Performance", "Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Aggregated Residual Transformations for Deep Neural Networks", "A ConvNet for the 2020s"], "answer_arxiv_id": ["1707.04780v2", "1901.09181", "2102.02887", "1911.11134", "1902.05967", "1907.04840", "2106.04533", "1510.00149", "1803.03635", "1611.05431", "2201.03545"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_8468"} +{"question": "Which papers studied learning variances in the reverse process?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models"], "answer_arxiv_id": ["2102.09672", "2206.07309"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_8469"} +{"question": "Which studies leveraged VAE and GAN for early 3D shape generation methods?", "answer": ["3D Shape Induction from 2D Views of Multiple Objects", "Escaping Plato's Cave: 3D Shape From Adversarial Rendering", "Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "Learning Representations and Generative Models for 3D Point Clouds", "PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation"], "answer_arxiv_id": ["1612.05872", "1811.11606", "1610.07584", "1707.02392", "1906.12320", "1901.05103"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_8470"} +{"question": "What research utilized vision-based traffic detection systems in aircraft collision avoidance?", "answer": ["AirTrack: Onboard Deep Learning Framework for Long-Range Aircraft Detection and Tracking"], "answer_arxiv_id": ["2209.12849"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_8471"} +{"question": "Which papers have explored automated approaches to extract the conservation laws from data?", "answer": ["Machine Learning Topological Invariants with Neural Networks", "Discovering conservation laws from trajectories via machine learning", "AI Poincaré 2.0: Machine Learning Conservation Laws from Differential Equations", "AI Feynman: a Physics-Inspired Method for Symbolic Regression"], "answer_arxiv_id": ["1708.09401", "2102.04008", "2203.12610", "1905.11481"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_8472"} +{"question": "Are there any studies that show failures due to different word orderings in vision-language-models?", "answer": ["Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality"], "answer_arxiv_id": ["2204.03162"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_8473"} +{"question": "Could you provide the studies that proposed to automatically annotate data with subtitles to enlarge the dataset scale?", "answer": ["HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million\n Narrated Video Clips", "MERLOT: Multimodal Neural Script Knowledge Models", "Advancing High-Resolution Video-Language Representation with Large-Scale\n Video Transcriptions"], "answer_arxiv_id": ["1906.03327", "2106.02636", "2111.10337"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_8474"} +{"question": "Which studies use CNNs to predict a discrete set of future trajectories for the ego agent in trajectory forecasting?", "answer": ["Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction", "CoverNet: Multimodal Behavior Prediction using Trajectory Sets"], "answer_arxiv_id": ["1809.10732", "1910.05449", "1911.10298"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_8475"} +{"question": "Were there any works on dialogue state tracking that focused on modeling dependencies among slot values?", "answer": ["Scaling Multi-Domain Dialogue State Tracking via Query Reformulation", "Transferable Multi-Domain State Generator for Task-Oriented Dialogue\n Systems", "Leveraging Slot Descriptions for Zero-Shot Cross-Domain Dialogue State\n Tracking", "Zero-Shot Dialogue State Tracking via Cross-Task Transfer", "Zero-shot Generalization in Dialog State Tracking through Generative\n Question Answering", "Dialogue State Tracking with a Language Model using Schema-Driven\n Prompting", "Knowledge-grounded Dialog State Tracking", "Dialogue Summaries as Dialogue States (DS2), Template-Guided\n Summarization for Few-shot Dialogue State Tracking"], "answer_arxiv_id": ["1903.05164", "1905.08743", "2105.04222", "2109.04655", "2101.08333", "2109.07506", "2210.06656", "2203.01552"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_8476"} +{"question": "Could you provide the references about ensemble Kalman filters (EnKF) used as an approximate filtering technique?", "answer": ["The Ensemble Kalman Filter: A Signal Processing Perspective"], "answer_arxiv_id": ["1702.08061"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8477"} +{"question": "What research adapted these methods for training neural ODEs more quickly than adjoint methods?", "answer": ["Weak Form Generalized Hamiltonian Learning"], "answer_arxiv_id": ["2104.05096"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_8478"} +{"question": "What are some of the works where the attention mechanism was used, like in Attentive Neural Processes, to enhance Neural Processes?", "answer": ["Attentive Neural Processes", "Transformer Neural Processes: Uncertainty-Aware Meta Learning Via Sequence Modeling"], "answer_arxiv_id": ["1901.05761", "2207.04179"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_8479"} +{"question": "What work proposed a method to interpret T2I pipelines by analyzing the influence of input words on generated images?", "answer": ["What the DAAM: Interpreting Stable Diffusion Using Cross Attention"], "answer_arxiv_id": ["2210.04885"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_8480"} +{"question": "Which papers use interactive supervision to mitigate the limitations of training language-conditioned reinforcement learning models?", "answer": ["Guiding Policies with Language via Meta-Learning", "Interactive Learning from Activity Description"], "answer_arxiv_id": ["1811.07882", "2102.07024"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_8481"} +{"question": "Which paper proposes to reorder the demonstration based on the entropy of the predicted labels?", "answer": ["Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2104.08786"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_8482"} +{"question": "Which study introduced the use of Network Dissection?", "answer": ["Network Dissection: Quantifying Interpretability of Deep Visual\n Representations"], "answer_arxiv_id": ["1704.05796"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_8483"} +{"question": "What papers describe exploration methods based on Thompson sampling in model-based reinforcement learning?", "answer": ["Model-Based Bayesian Exploration", "Online Learning in Kernelized Markov Decision Processes"], "answer_arxiv_id": ["1301.6690v1", "1805.08052"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_8484"} +{"question": "What work describes the process of dynamically dividing the vocabulary space into a green and red list, where the size of the green list is a fraction of the total vocabulary size?", "answer": ["A Watermark for Large Language Models"], "answer_arxiv_id": ["2301.10226"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_8485"} +{"question": "Could you provide me a study that proposed the integration of a large T5 language model to enhance semantic representation?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2205.11487"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_8486"} +{"question": "Could you provide me some references about novel class discovery in an unsupervised manner?", "answer": ["Learning to Discover Novel Visual Categories via Deep Transfer\n Clustering", "Automatically Discovering and Learning New Visual Categories with\n Ranking Statistics", "AutoNovel: Automatically Discovering and Learning Novel Visual\n Categories"], "answer_arxiv_id": ["1908.09884", "2002.05714", "2106.15252"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_8487"} +{"question": "Can you name the studies that introduce recalibration approach for improving the reflectance of data’s distribution?", "answer": ["Accurate Uncertainties for Deep Learning Using Calibrated Regression", "Conformal calibrators"], "answer_arxiv_id": ["1807.00263", "1902.06579"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_8488"} +{"question": "Are there any research studies which propose a two-layer representation for more robust scene capture?", "answer": ["DoubleFusion: Real-time Capture of Human Performances with Inner Body\n Shapes from a Single Depth Sensor"], "answer_arxiv_id": ["1804.06023"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_8489"} +{"question": "Which works employed generative methods for optimizing biological sequences?", "answer": ["Conditioning by adaptive sampling for robust design", "Model Inversion Networks for Model-Based Optimization", "Autofocused oracles for model-based design", "Biological Sequence Design with GFlowNets", "Consistent Training via Energy-Based GFlowNets for Modeling Discrete Joint Distributions"], "answer_arxiv_id": ["1901.10060", "1912.13464", "2006.08052", "2203.04115", "2211.00568"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_8490"} +{"question": "Could you reference studies that added an extra regularization term to the denoising score loss for satisfying properties of the diffusion process?", "answer": ["Consistent Diffusion Models: Mitigating Sampling Drift by Learning to be\n Consistent"], "answer_arxiv_id": ["2302.09057"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8491"} +{"question": "Any works about tasks involving coding skills?", "answer": ["Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2107.03374"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_8492"} +{"question": "What studies have focused on the performance of large language models in specific domains or tasks?", "answer": ["Measuring Massive Multitask Language Understanding", "Beyond the Imitation Game: Quantifying and extrapolating the\n capabilities of language models"], "answer_arxiv_id": ["2009.03300", "2206.04615"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_8493"} +{"question": "Which studies showed that global convergence of gradient descent is possible if the network is sufficiently wide and stays close to its random initialization?", "answer": ["A Convergence Theory for Deep Learning via Over-Parameterization", "Gradient Descent Provably Optimizes Over-parameterized Neural Networks", "On Exact Computation with an Infinitely Wide Neural Net", "Theoretical insights into the optimization landscape of over-parameterized shallow neural networks", "Algorithm-Dependent Generalization Bounds for Overparameterized Deep Residual Networks"], "answer_arxiv_id": ["1811.03962", "1810.02054", "1904.11955", "1707.04926", "1910.02934"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_8494"} +{"question": "Are there any works reviewing the expressive power of Graph Neural Networks?", "answer": ["A Survey on The Expressive Power of Graph Neural Networks", "Theory of Graph Neural Networks: Representation and Learning"], "answer_arxiv_id": ["2003.04078", "2204.07697"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_8495"} +{"question": "Which papers use a dynamic strategy to reduce tokens in a model?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "AdaViT: Adaptive Tokens for Efficient Vision Transformer", "You Need Multiple Exiting: Dynamic Early Exiting for Accelerating\n Unified Vision Language Model"], "answer_arxiv_id": ["2106.02034v2", "2112.07658", "2211.11152"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_8496"} +{"question": "Could you provide me some studies about list-decodable linear regression?", "answer": ["List-Decodable Linear Regression", "List Decodable Learning via Sum of Squares"], "answer_arxiv_id": ["1905.05679", "1905.04660v1"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_8497"} +{"question": "Are there any researches using pre-trained surrogate to minimize classification errors as in Error-Minimizing (EM) poisons?", "answer": ["Better Safe Than Sorry: Preventing Delusive Adversaries with Adversarial Training"], "answer_arxiv_id": ["2102.04716"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_8498"} +{"question": "Which studies discussed Inpainting and Text-guided image editing tasks in conditional image synthesis?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "DiffEdit: Diffusion-based semantic image editing with mask guidance"], "answer_arxiv_id": ["2111.14818", "2201.09865", "2211.09800", "2210.11427"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_8499"} +{"question": "Which work estimated the lighting from shiny objects?", "answer": ["Accidental Light Probes"], "answer_arxiv_id": ["2301.05211"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8500"} +{"question": "What research has focused on open vocabulary object recognition in the image domain?", "answer": ["Open-Vocabulary Object Detection Using Captions", "Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "Detecting Twenty-thousand Classes using Image-level Supervision", "Learning to Compose Soft Prompts for Compositional Zero-Shot Learning"], "answer_arxiv_id": ["2011.10678", "2109.01134", "2203.05557", "2104.13921", "2201.02605", "2204.03574"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_8501"} +{"question": "Which papers have applied KD to vision tasks?", "answer": ["Similarity-Preserving Knowledge Distillation", "Correlation Congruence for Knowledge Distillation", "Knowledge Adaptation for Efficient Semantic Segmentation"], "answer_arxiv_id": ["1907.09682", "1904.01802", "1903.04688v1"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_8502"} +{"question": "What work utilized neighboring views to infer per-view depth and normal maps?", "answer": ["MVPSNet: Fast Generalizable Multi-view Photometric Stereo"], "answer_arxiv_id": ["2305.11167"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_8503"} +{"question": "Which works have been conducted on multilingual neural machine translation?", "answer": ["Multi-Way, Multilingual Neural Machine Translation with a Shared Attention Mechanism", "Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation"], "answer_arxiv_id": ["1601.01073", "1611.04558"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8504"} +{"question": "Which works focused on the value underestimation methods in offline GCRL?", "answer": ["Actionable Models: Unsupervised Offline Reinforcement Learning of Robotic Skills"], "answer_arxiv_id": ["2104.07749"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8505"} +{"question": "Which work proposed a method for molecule SSL pretraining on multiple modalities by optimizing the mutual information between topological and conformational modalities using contrastive and generative objectives?", "answer": ["Pre-training Molecular Graph Representation with 3D Geometry"], "answer_arxiv_id": ["2110.07728"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_8506"} +{"question": "What research has been done on the theoretical and empirical aspects of active learning?", "answer": ["Batch Active Learning at Scale"], "answer_arxiv_id": ["2107.14263"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_8507"} +{"question": "What studies originally proposed the field of Few-Shot Segmentation (FSS)?", "answer": ["One-Shot Learning for Semantic Segmentation", "PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment", "Prior Guided Feature Enrichment Network for Few-Shot Segmentation", "Few-Shot Segmentation via Cycle-Consistent Transformer", "Self-Support Few-Shot Semantic Segmentation", "Hierarchical Dense Correlation Distillation for Few-Shot Segmentation"], "answer_arxiv_id": ["1709.03410", "1908.06391", "2008.01449", "2106.02320", "2207.11549", "2303.14652"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_8508"} +{"question": "Could you mention some studies that examine the relatedness between tasks through various methods such as fully fine-tuning, task vectors, example-based graphs, representation-level similarities, and human prior knowledge?", "answer": ["Domain Adaptation: Learning Bounds and Algorithms", "Characterizing and Avoiding Negative Transfer", "Taskonomy: Disentangling Task Transfer Learning", "Task2Vec: Task Embedding for Meta-Learning", "Representation Similarity Analysis for Efficient Task taxonomy & Transfer Learning", "DEPARA: Deep Attribution Graph for Deep Knowledge Transferability", "Duality Diagram Similarity: a generic framework for initialization selection in task transfer learning", "gulp: a prediction-based metric between representations", "Deep Cross Residual Learning for Multitask Visual Recognition", "HyperFace: A Deep Multi-task Learning Framework for Face Detection, Landmark Localization, Pose Estimation, and Gender Recognition"], "answer_arxiv_id": ["0902.3430", "1811.09751", "1804.08328", "1902.03545", "1904.11740", "2003.07496", "2008.02107", "2210.06545", "1604.01335", "1603.01249"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_8509"} +{"question": "What papers contributes to specification-based methods where they sparsely select part of the foundation model parameters for adjustment and freeze other parameters?", "answer": ["What Would Elsa Do? Freezing Layers During Transformer Fine-Tuning", "BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models", "Parameter-Efficient Transfer Learning with Diff Pruning", "Masking as an Efficient Alternative to Finetuning for Pretrained\n Language Models"], "answer_arxiv_id": ["1911.03090", "2106.10199", "2012.07463", "2004.12406"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_8510"} +{"question": "What papers discuss the emergent abilities of LLMs like in-context learning and chain-of-thoughts reasoning?", "answer": ["Language Models are Few-Shot Learners", "Emergent Abilities of Large Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2005.14165", "2206.07682", "2201.11903"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_8511"} +{"question": "Which works have investigated influence functions in the field of computer vision?", "answer": ["Understanding Black-box Predictions via Influence Functions"], "answer_arxiv_id": ["1703.04730"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_8512"} +{"question": "What prior research has been conducted on voxel-based 3D shape generation?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "3D-R2N2: A Unified Approach for Single and Multi-view 3D Object\n Reconstruction", "Unsupervised Learning of 3D Structure from Images", "Generative and Discriminative Voxel Modeling with Convolutional Neural\n Networks", "Octree Generating Networks: Efficient Convolutional Architectures for\n High-resolution 3D Outputs", "VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids", "Generative Adversarial Networks"], "answer_arxiv_id": ["1610.07584", "1604.00449", "1607.00662", "1608.04236", "1703.09438", "2206.07695", "2203.00667"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_8513"} +{"question": "Which papers have demonstrated the effectiveness of self-conditioning or temporal conditioning?", "answer": ["Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning"], "answer_arxiv_id": ["2208.04202"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_8514"} +{"question": "What work illustrates the relation of self-predictive representation to spectral decomposition?", "answer": ["Understanding Self-Predictive Learning for Reinforcement Learning"], "answer_arxiv_id": ["2212.03319"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_8515"} +{"question": "Which studies focused on generalist models aiming to handle diverse tasks using shared architecture?", "answer": ["Uni-Perceiver: Pre-training Unified Architecture for Generic Perception for Zero-shot and Few-shot Tasks", "Unified-IO: A unified model for vision, language, and multi-modal tasks", "Universal Instance Perception as Object Discovery and Retrieval"], "answer_arxiv_id": ["2112.01522", "2206.08916", "2303.06674"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_8516"} +{"question": "Which studies represented the progress of data augmentations for perspective images?", "answer": ["mixup: Beyond Empirical Risk Minimization", "CutMix: Regularization Strategy to Train Strong Classifiers with\n Localizable Features", "SuperMix: Supervising the Mixing Data Augmentation", "AutoAugment: Learning Augmentation Policies from Data", "Fast AutoAugment", "RandAugment: Practical automated data augmentation with a reduced search\n space", "Random Erasing Data Augmentation"], "answer_arxiv_id": ["1710.09412", "1905.04899", "2003.05034", "1805.09501", "1905.00397", "1909.13719", "1708.04896"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8517"} +{"question": "Which papers discuss the combination of epistemic and aleatoric uncertainties in deep learning settings?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?"], "answer_arxiv_id": ["1703.04977"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_8518"} +{"question": "What research deals with instance constraints by focusing on robustness, individual fairness, and must-link constraints in clustering?", "answer": ["Robust Optimal Classification Trees Against Adversarial Examples", "Learning Optimal and Fair Decision Trees for Non-Discriminative Decision-Making"], "answer_arxiv_id": ["2109.03857", "1903.10598"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_8519"} +{"question": "In context of graphs, what strategies have been proposed for learning contrastive representations?", "answer": ["Graph Contrastive Learning with Augmentations", "Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "Bringing Your Own View: Graph Contrastive Learning without Prefabricated Data Augmentations"], "answer_arxiv_id": ["2010.13902", "2106.05819", "2201.01702"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_8520"} +{"question": "Could you provide me some works that are focused on adapting to novel tasks from previously learned tasks or skills?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-Reinforcement Learning of Structured Exploration Strategies", "Learning Quickly to Plan Quickly Using Modular Meta-Learning", "Neural Task Graphs: Generalizing to Unseen Tasks from a Single Video Demonstration"], "answer_arxiv_id": ["1703.03400", "1802.07245", "1809.07878", "1807.03480"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_8521"} +{"question": "Which works utilized enhanced cone sampling strategies for Neural Radiance Fields?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural\n Radiance Fields"], "answer_arxiv_id": ["2103.13415", "2111.12077", "2307.11335"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_8522"} +{"question": "What studies used adaptive encoders for diffuser in diffusion models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8523"} +{"question": "Could you provide me some research exploring the potential of ChatGPT to complete the Vietnam National High School Graduation Exam?", "answer": ["Can ChatGPT pass the Vietnamese National High School Graduation\n Examination?"], "answer_arxiv_id": ["2306.09170"], "source_meta": {"published_time": "20240820"}, "qid": "AutoScholarQuery_train_8524"} +{"question": "What papers suggest that increasing the number of stochastic layers in hierarchical VAEs (HVAEs) improves performance?", "answer": ["NVAE: A Deep Hierarchical Variational Autoencoder", "Very Deep VAEs Generalize Autoregressive Models and Can Outperform Them on Images"], "answer_arxiv_id": ["2007.03898", "2011.10650"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_8525"} +{"question": "What works improved NeRF's rendering quality by introducing 3D conical frustum?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields"], "answer_arxiv_id": ["2103.13415", "2111.12077"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_8526"} +{"question": "Which papers discuss the use of transformer-based textual models such as BERT, RoBERTa and XLNET?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "XLNet: Generalized Autoregressive Pretraining for Language Understanding"], "answer_arxiv_id": ["1810.04805", "1907.11692", "1906.08237"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_8527"} +{"question": "What research papers describe the application of diffusion probabilistic models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2105.05233", "2006.11239", "2102.09672"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_8528"} +{"question": "Could you name the study that uses a slow kmeans-based approach to token merging?", "answer": ["Token Pooling in Vision Transformers"], "answer_arxiv_id": ["2110.03860"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_8529"} +{"question": "Could you provide me research papers about DNNs explained through subnetworks?", "answer": ["A Review of Modularization Techniques in Artificial Neural Networks", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks"], "answer_arxiv_id": ["1904.12770", "1803.03635"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_8530"} +{"question": "What studies have employed manual interaction in the optimization of new scenes?", "answer": ["SceneSeer: 3D Scene Design with Natural Language", "SceneSuggest: Context-driven 3D Scene Design"], "answer_arxiv_id": ["1703.00050", "1703.00061"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_8531"} +{"question": "Which papers discuss implementations that consider scaling vectors?", "answer": ["Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than\n In-Context Learning", "VeRA: Vector-based Random Matrix Adaptation"], "answer_arxiv_id": ["2205.05638", "2310.11454"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_8532"} +{"question": "Which papers discuss recent advancements in diffusion model-based image editing?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2108.01073"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_8533"} +{"question": "Which papers focus on policy optimization-based methods for solving adversarial linear mixture MDPs?", "answer": ["Provably Efficient Exploration in Policy Optimization", "Near-optimal Policy Optimization Algorithms for Learning Adversarial Linear Mixture MDPs"], "answer_arxiv_id": ["1912.05830", "2102.08940v2"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_8534"} +{"question": "Could you provide me some studies where the similarity-based objective of VLMs has been repurposed to specialized domains?", "answer": ["Learning to Prompt for Vision-Language Models", "Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation", "Teaching CLIP to Count to Ten"], "answer_arxiv_id": ["2109.01134", "2104.13921", "2302.12066"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_8535"} +{"question": "What studies showed that for degree-2 polynomial threshold functions there are polynomial-time algorithms that can assert the robustness of the model or identify an adversarial case?", "answer": ["On Robustness to Adversarial Examples and Polynomial Optimization"], "answer_arxiv_id": ["1911.04681"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_8536"} +{"question": "Which work introduced a dataset and a model for estimating gaze direction on a screen in real-time on mobile devices?", "answer": ["Eye Tracking for Everyone"], "answer_arxiv_id": ["1606.05814"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_8537"} +{"question": "Which works regard end-to-end trained models for building instruction-following agents?", "answer": ["Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments", "Towards Learning a Generic Agent for Vision-and-Language Navigation via Pre-training", "Habitat 2.0: Training Home Assistants to Rearrange their Habitat", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["1711.07280", "2002.10638", "2106.14405", "2211.09800"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_8538"} +{"question": "What studies have been conducted on backdoor detection based on prediction bias?", "answer": ["Rethinking the Trigger of Backdoor Attack"], "answer_arxiv_id": ["2004.04692"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_8539"} +{"question": "Can you name the work that was the inspiration for average and maximum similarity-based approaches?", "answer": ["ColBERT: Efficient and Effective Passage Search via Contextualized Late\n Interaction over BERT", "ColBERTv2: Effective and Efficient Retrieval via Lightweight Late\n Interaction"], "answer_arxiv_id": ["2004.12832", "2112.01488"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_8540"} +{"question": "What papers are about encoder-based models in text-to-image generation?", "answer": ["Taming Encoder for Zero Fine-tuning Image Customization with\n Text-to-Image Diffusion Models", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2304.02642", "2302.13848", "2304.03411", "2305.14720", "2302.12228"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_8541"} +{"question": "Can you provide examples of recent text-to-image models that attempted to mitigate the prior knowledge gap?", "answer": ["Muse: Text-To-Image Generation via Masked Generative Transformers", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "Improving Sample Quality of Diffusion Models Using Self-Attention Guidance", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2301.00704", "2203.13131", "2210.00939", "2204.06125", "2102.12092", "2205.11487"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_8542"} +{"question": "What works proposed the Gumbel-Top-k trick, a technique used in Gumbel AlphaZero for action sampling?", "answer": ["Stochastic Beams and Where to Find Them: The Gumbel-Top-k Trick for Sampling Sequences Without Replacement"], "answer_arxiv_id": ["1903.06059"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_8543"} +{"question": "Any studies that explore neighborhood similarity as a method for learning with noisy labels?", "answer": ["Clusterability as an Alternative to Anchor Points When Learning with Noisy Labels", "A Second-Order Approach to Learning with Instance-Dependent Label Noise"], "answer_arxiv_id": ["2102.05291", "2012.11854"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8544"} +{"question": "Could you provide me some readings about conversational query rewriting in conversational search?", "answer": ["Few-Shot Generative Conversational Query Rewriting", "CONQRR: Conversational Query Rewriting for Retrieval with Reinforcement\n Learning", "Query Resolution for Conversational Search with Limited Supervision", "Conversational Question Reformulation via Sequence-to-Sequence\n Architectures and Pretrained Language Models", "Conversational Query Rewriting with Self-supervised Learning", "Question Rewriting for Conversational Question Answering", "A Comparison of Question Rewriting Methods for Conversational Passage\n Retrieval", "Large Language Models Know Your Contextual Search Intent: A Prompting\n Framework for Conversational Search", "ConvGQR: Generative Query Reformulation for Conversational Search"], "answer_arxiv_id": ["2006.05009", "2112.08558", "2005.11723", "2004.01909", "2102.04708", "2004.14652", "2101.07382", "2303.06573", "2305.15645"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_8545"} +{"question": "What papers propose utilizing meta RL, domain randomization, and system identification for domain adaptation in reinforcement learning?", "answer": ["Policy Transfer with Strategy Optimization", "Learning to Adapt in Dynamic, Real-World Environments through Meta-Reinforcement Learning", "Fast Adaptation to New Environments via Policy-Dynamics Value Functions", "EPOpt: Learning Robust Neural Network Policies Using Model Ensembles", "Sim-to-Real Transfer of Robotic Control with Dynamics Randomization", "Active Domain Randomization", "Environment Probing Interaction Policies", "Preparing for the Unknown: Learning a Universal Policy with Online System Identification", "Auto-Tuned Sim-to-Real Transfer", "Robust Policy Learning over Multiple Uncertainty Sets"], "answer_arxiv_id": ["1810.05751", "1803.11347", "2007.02879", "1610.01283", "1710.06537", "1904.04762", "1907.11740", "1702.02453", "2104.07662", "2202.07013"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_8546"} +{"question": "What papers have given attention to learning in scheduling?", "answer": ["Scheduling jobs with stochastic holding costs"], "answer_arxiv_id": ["2105.13655"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_8547"} +{"question": "Are there any works that discuss variants of CL such as self-paced CL and implicit curriculum?", "answer": ["An Empirical Study of Example Forgetting during Deep Neural Network Learning"], "answer_arxiv_id": ["1812.05159"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_8548"} +{"question": "Are there any works that use end-to-end learning schemes for designing optimal coded mask patterns?", "answer": ["Data Driven Coded Aperture Design for Depth Recovery"], "answer_arxiv_id": ["1705.10021"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_8549"} +{"question": "Are there any datasets that have been proposed for human-centric detection in crowded scenarios?", "answer": ["STCrowd: A Multimodal Dataset for Pedestrian Perception in Crowded\n Scenes"], "answer_arxiv_id": ["2204.01026"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_8550"} +{"question": "What are the works that analyze non-smooth settings?", "answer": ["Optimal rates for zero-order convex optimization: the power of two function evaluations"], "answer_arxiv_id": ["1312.2139"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8551"} +{"question": "Which paper talks about RSTPReid and the baseline algorithm, DSSL?", "answer": ["DSSL: Deep Surroundings-person Separation Learning for Text-based Person\n Retrieval"], "answer_arxiv_id": ["2109.05534"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_8552"} +{"question": "Which papers proposed datasets of long input samples used to evaluate models?", "answer": ["ZeroSCROLLS: A Zero-Shot Benchmark for Long Text Understanding", "SCROLLS: Standardized CompaRison Over Long Language Sequences", "L-Eval: Instituting Standardized Evaluation for Long Context Language Models", "L-Eval: Instituting Standardized Evaluation for Long Context Language Models"], "answer_arxiv_id": ["2305.14196", "2201.03533", "2307.11088v3", "2307.11088v3"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_8553"} +{"question": "What papers have shown that CNNs can learn locally invariant features with respect to arbitrary transformation groups?", "answer": ["Understanding Deep Convolutional Networks"], "answer_arxiv_id": ["1601.04920"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_8554"} +{"question": "What works suggest a dropout-based technique for calibrating deep neural networks?", "answer": ["Concrete Dropout"], "answer_arxiv_id": ["1705.07832"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_8555"} +{"question": "Could you provide me some works that propose loss re-weighting method for long-tailed image classification?", "answer": ["Class-Balanced Loss Based on Effective Number of Samples", "Long-Tail Learning via Logit Adjustment", "Long-tailed Visual Recognition via Gaussian Clouded Logit Adjustment", "Rethinking Class-Balanced Methods for Long-Tailed Visual Recognition from a Domain Adaptation Perspective", "Equalization Loss for Long-Tailed Object Recognition"], "answer_arxiv_id": ["1901.05555", "2007.07314", "2305.11733", "2003.10780", "2003.05176"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_8556"} +{"question": "What are the studies that introduced a cascaded diffusion structure for image generation and editing?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2112.10741"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8557"} +{"question": "Which papers showed that neural collapse is induced even in imbalanced datasets using a fixed ETF classifier?", "answer": ["Inducing Neural Collapse in Imbalanced Learning: Do We Really Need a\n Learnable Classifier at the End of Deep Neural Network?"], "answer_arxiv_id": ["2203.09081"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_8558"} +{"question": "What papers introduced the HT-SR theory as a tool to measure the quality of publicly-available pre-trained neural networks?", "answer": ["Implicit Self-Regularization in Deep Neural Networks: Evidence from Random Matrix Theory and Implications for Learning"], "answer_arxiv_id": ["1810.01075"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_8559"} +{"question": "What research is about panoptic segmentation?", "answer": ["Panoptic Segmentation", "Panoptic Segmentation"], "answer_arxiv_id": ["1801.00868", "1801.00868"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_8560"} +{"question": "Which studies revealed that LLMs often fail to correctly perform multi-hop reasoning?", "answer": ["Measuring and Narrowing the Compositionality Gap in Language Models", "Faith and Fate: Limits of Transformers on Compositionality"], "answer_arxiv_id": ["2210.03350", "2305.18654"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_8561"} +{"question": "What are some previous stratified frameworks, and what are their limitations?", "answer": ["Neural-Symbolic Integration: A Compositional Perspective", "DeepProbLog: Neural Probabilistic Logic Programming", "NeurASP: Embracing Neural Networks into Answer Set Programming"], "answer_arxiv_id": ["2010.11926", "1805.10872", "2307.07700"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_8562"} +{"question": "What are the seminal works on neural scene representations that focus on volume structures?", "answer": ["DeepVoxels: Learning Persistent 3D Feature Embeddings", "Neural Volumes: Learning Dynamic Renderable Volumes from Images"], "answer_arxiv_id": ["1812.01024v2", "1906.07751"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_8563"} +{"question": "What works in the literature attempted to solve the 'terpret problem'?", "answer": ["TerpreT: A Probabilistic Programming Language for Program Induction"], "answer_arxiv_id": ["1608.04428"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_8564"} +{"question": "What are the studies about Open-Vocabulary Detection (OVD)?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation", "MDETR -- Modulated Detection for End-to-End Multi-Modal Understanding", "Grounded Language-Image Pre-training", "Open-Vocabulary Object Detection Using Captions", "RegionCLIP: Region-based Language-Image Pretraining", "Detecting Twenty-thousand Classes using Image-level Supervision"], "answer_arxiv_id": ["2104.13921", "2104.12763", "2112.03857", "2011.10678", "2112.09106", "2201.02605"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_8565"} +{"question": "Which works extended the unified sequence-to-sequence based transformer framework to the field of computer vision?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "A Generalist Agent", "PaLI: A Jointly-Scaled Multilingual Language-Image Model"], "answer_arxiv_id": ["2204.14198", "2202.03052", "2205.06175", "2209.06794"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8566"} +{"question": "What work discussed language model's sensitivity to the order of few-shot demonstrations and introduced an Entropy-based metric?", "answer": ["Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2104.08786"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_8567"} +{"question": "What studies have extended the POIS, an off-policy optimization algorithm that relies on Monte Carlo simulation?", "answer": ["Policy Optimization via Importance Sampling"], "answer_arxiv_id": ["1809.06098"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_8568"} +{"question": "Which research papers are about offline methods in subject-driven generation?", "answer": ["ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models", "Subject-driven Text-to-Image Generation via Apprenticeship Learning"], "answer_arxiv_id": ["2302.13848", "2307.06949", "2304.00186"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_8569"} +{"question": "Could you provide me some works where noisy samples are treated as unlabeled data for semi-supervised learning?", "answer": ["DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "UniCon: Combating Label Noise Through Uniform Selection and Contrastive Learning"], "answer_arxiv_id": ["2002.07394", "2203.14542"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_8570"} +{"question": "Could you provide me some sources that offer principled measures of calibration?", "answer": ["Low-Degree Multicalibration", "A Unifying Theory of Distance from Calibration"], "answer_arxiv_id": ["2203.01255", "2211.16886"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8571"} +{"question": "Could you tell me which study improved the performance of hard prompt-based fine-tuning on single sentence tasks through conventional continued pre-training?", "answer": ["AdaPrompt: Adaptive Model Training for Prompt-based NLP"], "answer_arxiv_id": ["2202.04824"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_8572"} +{"question": "Could you provide me some papers about achieving coverage guarantees across groups aimed at producing fairer outcomes in machine learning systems?", "answer": ["Achieving Equalized Odds by Resampling Sensitive Attributes"], "answer_arxiv_id": ["2006.04292"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_8573"} +{"question": "Which researches utilize Graph Neural Networks for link prediction in knowledge graphs?", "answer": ["Modeling Relational Data with Graph Convolutional Networks", "Composition-based Multi-Relational Graph Convolutional Networks", "Inductive Relation Prediction by Subgraph Reasoning"], "answer_arxiv_id": ["1703.06103v4", "1911.03082", "1911.06962"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_8574"} +{"question": "Which studies discuss the use of prompt engineering or training a costly viewpoint-aware model to alleviate the problems of 2D observations lifting into 3D?", "answer": ["Re-imagine the Negative Prompt Algorithm: Transform 2D Diffusion into\n 3D, alleviate Janus problem and Beyond", "Zero-1-to-3: Zero-shot One Image to 3D Object", "MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2304.04968", "2303.11328", "2308.16512"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8575"} +{"question": "Is there any piece of research observing that LLM-based evaluators have more biases in non-Latin languages?", "answer": ["Are Large Language Model-based Evaluators the Solution to Scaling Up\n Multilingual Evaluation?"], "answer_arxiv_id": ["2309.07462"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_8576"} +{"question": "Is there any work that discusses heavy-tailed rewards where the variance could be non-existent in Reinforcement Learning (RL)?", "answer": ["No-Regret Reinforcement Learning with Heavy-Tailed Rewards"], "answer_arxiv_id": ["2102.12769v1"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_8577"} +{"question": "Who developed NR-PCQA metrics using hand-crafted features?", "answer": ["No-Reference Quality Assessment for 3D Colored Point Cloud and Mesh\n Models", "Blind Quality Assessment of 3D Dense Point Clouds with Structure Guided\n Resampling"], "answer_arxiv_id": ["2107.02041", "2208.14603"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_8578"} +{"question": "Could you provide me with research papers focusing on continuous prompt methods?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Multimodal Few-Shot Learning with Frozen Language Models", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2101.00190", "2106.13884", "2109.01134"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_8579"} +{"question": "Are there any model-based approaches that use a dynamics model of other agent's policies change?", "answer": ["Influencing Long-Term Behavior in Multiagent Reinforcement Learning", "Influencing Towards Stable Multi-Agent Interactions"], "answer_arxiv_id": ["2203.03535", "2110.08229"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_8580"} +{"question": "Which research introduces a querying transformer in Vision-Language models?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models"], "answer_arxiv_id": ["2301.12597"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_8581"} +{"question": "What works have proven the convergence for SGD type of bilevel methods via the AID approach?", "answer": ["Approximation Methods for Bilevel Programming", "A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic", "Bilevel Optimization: Convergence Analysis and Enhanced Design", "Amortized Implicit Differentiation for Stochastic Bilevel Optimization"], "answer_arxiv_id": ["1802.02246", "2007.05170v4", "2010.07962", "2111.14580"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_8582"} +{"question": "What works used LLMs with specific templates to replace object names in language-based detection tasks?", "answer": ["DesCo: Learning Object Recognition with Rich Language Descriptions"], "answer_arxiv_id": ["2306.14060"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_8583"} +{"question": "Could you give me examples of studies that explored BEV segmentation and transformed the camera plane into BEV via Inverse Perspective Mapping (IPM)?", "answer": ["Orthographic Feature Transform for Monocular 3D Object Detection", "Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by\n Implicitly Unprojecting to 3D", "FIERY: Future Instance Prediction in Bird's-Eye View from Surround\n Monocular Cameras"], "answer_arxiv_id": ["1811.08188", "2008.05711", "2104.10490"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_8584"} +{"question": "Are there any works that applied adversarial training strategy with graph contrastive learning for graph augmentation?", "answer": ["Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "Graph Contrastive Learning Automated", "Adversarial Graph Contrastive Learning with Information Regularization"], "answer_arxiv_id": ["2106.05819", "2106.07594", "2202.06491"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_8585"} +{"question": "Which paper presented an approach to learning the mapping function from LR images to HR images just using three convolutional layers?", "answer": ["Image Super-Resolution Using Deep Convolutional Networks"], "answer_arxiv_id": ["1501.00092"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_8586"} +{"question": "Could you provide me some studies about deduplication techniques used in dataset preprocessing?", "answer": ["SemDeDup: Data-efficient learning at web-scale through semantic\n deduplication"], "answer_arxiv_id": ["2303.09540"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_8587"} +{"question": "Which paper suggests incorporating few-shot demonstrations in the prompt to improve self-correction of LLMs?", "answer": ["Teaching Large Language Models to Self-Debug"], "answer_arxiv_id": ["2304.05128v2"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_8588"} +{"question": "Which work first derive lower bounds for classical fixed-confidence BAI setting without privacy?", "answer": ["On the Complexity of Best-Arm Identification in Multi-Armed Bandit Models"], "answer_arxiv_id": ["1407.4443"], "source_meta": {"published_time": "20230905"}, "qid": "AutoScholarQuery_train_8589"} +{"question": "Which studies perturb the actions in continuous action space using independent random noise?", "answer": ["Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor", "Continuous control with deep reinforcement learning"], "answer_arxiv_id": ["1801.01290", "1509.02971"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_8590"} +{"question": "What works showed that there is also 'good heterophily' that are friendly to GNNs?", "answer": ["Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?", "Is Homophily a Necessity for Graph Neural Networks?"], "answer_arxiv_id": ["2109.05641", "2106.06134"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_8591"} +{"question": "In which paper is the concept of 'Chain-of-hindsight' explored, transforming all (binary or multi-scale) feedback into a sentence that consists of chain of all feedback?", "answer": ["Chain of Hindsight aligns Language Models with Feedback"], "answer_arxiv_id": ["2302.02676"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_8592"} +{"question": "What papers are related to the application of learning-augmented algorithms to binary search trees?", "answer": ["Learning-Augmented B-Trees"], "answer_arxiv_id": ["2211.09251"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_8593"} +{"question": "Are there studies which used episodic memory to search the optimal hyperparameter for policy gradient methods?", "answer": ["Episodic Policy Gradient Training"], "answer_arxiv_id": ["2112.01853"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_8594"} +{"question": "Could you provide me the work that studies the converges of Adam under the (L0,L1) smoothness condition?", "answer": ["Provable Adaptivity in Adam"], "answer_arxiv_id": ["2208.09900"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_8595"} +{"question": "What is the study that creates a globally balanced distribution by using local augmentation?", "answer": ["Astraea: Self-balancing Federated Learning for Improving Classification Accuracy of Mobile Deep Learning Applications"], "answer_arxiv_id": ["1907.01132"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_8596"} +{"question": "Could you provide me some studies about enhanced versions of the Stable Diffusion model?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "Composer: Creative and Controllable Image Synthesis with Composable Conditions", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.09778", "2302.08453"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_8597"} +{"question": "Which papers discuss the state-of-the-art sample quality and likelihood achieved by diffusion models or score-based generative models?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Elucidating the Design Space of Diffusion-Based Generative Models", "Variational Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2206.00364", "2107.00630"], "source_meta": {"published_time": "20230506"}, "qid": "AutoScholarQuery_train_8598"} +{"question": "What works focused on roles of Transformer as pure data-driven model in absence of known governing PDE?", "answer": ["Earthformer: Exploring Space-Time Transformers for Earth System Forecasting"], "answer_arxiv_id": ["2207.05833"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8599"} +{"question": "Which works have used evolutionary algorithms such as NSGA-II in various multi-objective optimization problems?", "answer": ["pymoo: Multi-objective Optimization in Python"], "answer_arxiv_id": ["2002.04504"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_8600"} +{"question": "What studies discuss the usage of instance-wise adversarial perturbation in contrastive learning?", "answer": ["Adversarial Self-Supervised Contrastive Learning", "Robust Pre-Training by Adversarial Contrastive Learning"], "answer_arxiv_id": ["2006.07589", "2010.13337"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_8601"} +{"question": "What studies made use of implicit representations when dealing with 3D recovery?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "DISN: Deep Implicit Surface Network for High-quality Single-view 3D\n Reconstruction"], "answer_arxiv_id": ["1812.03828", "1905.10711"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_8602"} +{"question": "What papers have introduced a deep learning-based approach for high-dimensional parabolic PDEs?", "answer": ["Solving High-Dimensional Partial Differential Equations Using Deep Learning"], "answer_arxiv_id": ["1707.02568"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_8603"} +{"question": "Which studies use online refinement modules to obtain final pseudo-labels in one-stage weakly supervised semantic segmentation?", "answer": ["Single-Stage Semantic Segmentation from Image Labels"], "answer_arxiv_id": ["2005.08104"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_8604"} +{"question": "What research proposes making the FC layer sparse for a more interpretable final layer?", "answer": ["Leveraging Sparse Linear Layers for Debuggable Deep Networks"], "answer_arxiv_id": ["2105.04857"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_8605"} +{"question": "Which papers discuss using a knowledge attribution approach in parameter-modifying ME methods?", "answer": ["Knowledge Neurons in Pretrained Transformers"], "answer_arxiv_id": ["2104.08696"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_8606"} +{"question": "What studies have explored the effectiveness of large-scale self-supervised speech language models in enhancing text-to-speech quality?", "answer": ["VQTTS: High-Fidelity Text-to-Speech Synthesis with Self-Supervised VQ Acoustic Feature", "WavThruVec: Latent speech representation as intermediate features for neural speech synthesis", "M2-CTTS: End-to-End Multi-scale Multi-modal Conversational Text-to-Speech Synthesis"], "answer_arxiv_id": ["2204.00768", "2203.16930", "2305.02269"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8607"} +{"question": "Could you inform me of studies researching backdoor attack via inserting specific words or sentences?", "answer": ["Weight Poisoning Attacks on Pre-trained Models", "A backdoor attack against LSTM-based text classification systems"], "answer_arxiv_id": ["2004.06660", "1905.12457v2"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_8608"} +{"question": "What works have extended the image-based noise prediction model, DDPM?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2102.09672", "2010.02502", "2105.05233"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8609"} +{"question": "What works have developed model selection methods specifically for graph learning models?", "answer": ["Auto-GNN: Neural Architecture Search of Graph Neural Networks", "Policy-GNN: Aggregation Optimization for Graph Neural Networks", "Automated Graph Learning via Population Based Self-Tuning GCN", "JITuNE: Just-In-Time Hyperparameter Tuning for Network Embedding Algorithms"], "answer_arxiv_id": ["1909.03184", "2006.15097", "2107.04713", "2101.06427"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_8610"} +{"question": "Which studies focused on incorporating convolutional neural networks (CNNs) and generative adversarial networks (GANs) in video inpainting?", "answer": ["Free-Form Image Inpainting with Gated Convolution", "Video Inpainting by Jointly Learning Temporal Structure and Spatial\n Details", "Context Encoders: Feature Learning by Inpainting", "Image Inpainting for Irregular Holes Using Partial Convolutions", "Recurrent Feature Reasoning for Image Inpainting"], "answer_arxiv_id": ["1806.03589", "1806.08482", "1604.07379", "1804.07723", "2008.03737"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_8611"} +{"question": "Could you provide me studies about integrating INR and diffusion models?", "answer": ["Implicit Diffusion Models for Continuous Super-Resolution"], "answer_arxiv_id": ["2303.16491"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_8612"} +{"question": "Can you cite studies that made theoretical advances on the topic of offline robust RL?", "answer": ["Towards Theoretical Understandings of Robust Markov Decision Processes: Sample Complexity and Asymptotics", "Robust Reinforcement Learning using Offline Data", "Sample Complexity of Robust Reinforcement Learning with a Generative Model", "Distributionally Robust Model-Based Offline Reinforcement Learning with Near-Optimal Sample Complexity", "Distributionally Robust Offline Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["2105.03863", "2208.05129", "2112.01506v3", "2208.05767", "2209.06620"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_8613"} +{"question": "What works have used consistency modeling between stereo images and their warped images for self-supervised training in multiview stereo depth estimation?", "answer": ["Unsupervised Monocular Depth Estimation with Left-Right Consistency", "Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescue", "Unsupervised Learning of Depth and Ego-Motion from Video"], "answer_arxiv_id": ["1609.03677", "1603.04992", "1704.07813"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_8614"} +{"question": "What work introduced auxiliary tasks in RL?", "answer": ["Loss is its own Reward: Self-Supervision for Reinforcement Learning"], "answer_arxiv_id": ["1612.07307"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_8615"} +{"question": "Can you provide some research that applies parameter-efficient transfer-learning to dialogue state tracking?", "answer": ["Continual Learning in Task-Oriented Dialogue Systems", "Continual Prompt Tuning for Dialog State Tracking", "Divide, Conquer, and Combine: Mixture of Semantic-Independent Experts\n for Zero-Shot Dialogue State Tracking", "Prompter: Zero-shot Adaptive Prefixes for Dialogue State Tracking Domain\n Adaptation", "Towards LLM-driven Dialogue State Tracking"], "answer_arxiv_id": ["2012.15504", "2203.06654", "2306.00434", "2306.04724", "2310.14970"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_8616"} +{"question": "Which study found that the task performance of in-context learning can be highly sensitive to how the in-context prompt is written?", "answer": ["Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2104.08786"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_8617"} +{"question": "Which paper applied variational method in neural networks?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_8618"} +{"question": "Which research papers note LLM agents' ability to respond like humans and perform self-oriented planning?", "answer": ["A Survey on Large Language Model based Autonomous Agents", "The Rise and Potential of Large Language Model Based Agents: A Survey"], "answer_arxiv_id": ["2308.11432", "2309.07864v3"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_8619"} +{"question": "What research papers have been released on the subject of video captioning tasks?", "answer": ["End-to-End Dense Video Captioning with Masked Transformer", "Towards Automatic Learning of Procedures from Web Instructional Videos", "SwinBERT: End-to-End Transformers with Sparse Attention for Video\n Captioning", "Multi-modal Dense Video Captioning", "End-to-End Dense Video Captioning with Parallel Decoding", "Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense\n Video Captioning"], "answer_arxiv_id": ["1804.00819", "1703.09788", "2111.13196", "2003.07758", "2108.07781", "2302.14115"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_8620"} +{"question": "What are some works that have applied deep active learning to image classification?", "answer": ["Deep Bayesian Active Learning with Image Data"], "answer_arxiv_id": ["1703.02910"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_8621"} +{"question": "What research has been conducted on cross-block orchestration as a critical component of state-of-the-art visual recognition algorithms?", "answer": ["Cross-X Learning for Fine-Grained Visual Categorization", "Co-Scale Conv-Attentional Image Transformers", "SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers"], "answer_arxiv_id": ["1909.04412", "2104.06399", "2105.15203", "2012.15840"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_8622"} +{"question": "What papers have received attention for work in domain adaptation?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness", "AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "Tent: Fully Test-Time Adaptation by Entropy Minimization", "Source Data-absent Unsupervised Domain Adaptation through Hypothesis Transfer and Labeling Transfer"], "answer_arxiv_id": ["1811.12231", "1912.02781", "2006.10726", "2012.07297"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_8623"} +{"question": "What data sets are used for TV/movie screenplays summary?", "answer": ["SummScreen: A Dataset for Abstractive Screenplay Summarization"], "answer_arxiv_id": ["2104.07091"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_8624"} +{"question": "Can you list the papers that use contrastive learning for learning representations for input to a meta RL policy?", "answer": ["FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization", "Momentum Contrast for Unsupervised Visual Representation Learning", "Provably Improved Context-Based Offline Meta-RL with Attention and Contrastive Learning", "Multi-task Batch Reinforcement Learning with Metric Learning", "Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables", "Improving Context-Based Meta-Reinforcement Learning with Self-Supervised Trajectory Contrastive Learning", "Towards Effective Context for Meta-Reinforcement Learning: an Approach based on Contrastive Learning"], "answer_arxiv_id": ["2010.01112", "1911.05722", "2102.10774", "1909.11373", "1903.08254", "2103.06386", "2009.13891"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_8625"} +{"question": "What works talked about integrating non-rigid deformations based on the body prior of estimated SMPL parameters in order to learn highly articulated humans from unstructured 2D images?", "answer": ["AvatarGen: A 3D Generative Model for Animatable Human Avatars", "EVA3D: Compositional 3D Human Generation from 2D Image Collections", "StyleGAN-Human: A Data-Centric Odyssey of Human Generation", "HumanGen: Generating Human Radiance Fields with Explicit Priors", "AG3D: Learning to Generate 3D Avatars from 2D Image Collections", "3DHumanGAN: 3D-Aware Human Image Generation with 3D Pose Mapping"], "answer_arxiv_id": ["2211.14589", "2210.04888", "2204.11823", "2212.05321", "2305.02312", "2212.07378"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_8626"} +{"question": "What works improved the multi-view consistency across different novel views?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image"], "answer_arxiv_id": ["2308.16512", "2309.03453"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_8627"} +{"question": "What studies have used GPUs for various convex optimizations?", "answer": ["Exploiting GPU/SIMD Architectures for Solving Linear-Quadratic MPC Problems", "GPU Acceleration of ADMM for Large-Scale Quadratic Programming"], "answer_arxiv_id": ["2209.13049v1", "1912.04263"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_8628"} +{"question": "Could you provide me some references which prove Transformers can implement gradient descent?", "answer": ["What learning algorithm is in-context learning? Investigations with linear models", "Transformers Learn In-Context by Gradient Descent", "Uncovering mesa-optimization algorithms in Transformers"], "answer_arxiv_id": ["2211.15661", "2212.07677", "2309.05858"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8629"} +{"question": "What works propose diffusion-based image editing methods given image condition?", "answer": ["Paint by Example: Exemplar-based Image Editing with Diffusion Models"], "answer_arxiv_id": ["2211.13227"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_8630"} +{"question": "What papers explored single-view reconstruction using voxels?", "answer": ["An Effective Loss Function for Generating 3D Models from Single 2D Image without Rendering"], "answer_arxiv_id": ["2103.03390"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_8631"} +{"question": "What are some studies that used pruning or sparsification for model compression?", "answer": ["Pruning Convolutional Neural Networks for Resource Efficient Inference", "Rethinking the Value of Network Pruning", "Filter Pruning via Geometric Median for Deep Convolutional Neural\n Networks Acceleration", "Sparsity in Deep Learning: Pruning and growth for efficient inference\n and training in neural networks", "Deja Vu: Contextual Sparsity for Efficient LLMs at Inference Time", "Rethinking the Role of Scale for In-Context Learning: An\n Interpretability-based Case Study at 66 Billion Scale"], "answer_arxiv_id": ["1611.06440", "1810.05270", "1811.00250", "2102.00554", "2310.17157", "2212.09095"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_8632"} +{"question": "Who proposed AutoPrompt in the context of better prompt generation?", "answer": ["AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts"], "answer_arxiv_id": ["2010.15980"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_8633"} +{"question": "What examples of research are available on machine-assisted annotation methods within the field of computer vision?", "answer": ["ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation", "Interactive Full Image Segmentation by Considering All Regions Jointly", "Deep Extreme Cut: From Extreme Points to Object Segmentation", "Guide Me: Interacting with Deep Networks", "Connecting Vision and Language with Localized Narratives"], "answer_arxiv_id": ["1604.05144v1", "1812.01888", "1711.09081v2", "1803.11544", "1912.03098v4"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_8634"} +{"question": "What previous studies have focused on introducing ED-specific benchmarks?", "answer": ["EHRSHOT: An EHR Benchmark for Few-Shot Evaluation of Foundation Models"], "answer_arxiv_id": ["2307.02028"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_8635"} +{"question": "Which studies have addressed the impact of text perturbations on AI text detection?", "answer": ["A Robust Semantics-based Watermark for Large Language Model against\n Paraphrasing", "A Ship of Theseus: Curious Cases of Paraphrasing in LLM-Generated Texts", "Large Language Models can be Guided to Evade AI-Generated Text Detection", "Mutation-Based Adversarial Attacks on Neural Text Detectors", "Evade ChatGPT Detectors via A Single Space"], "answer_arxiv_id": ["2311.08721", "2311.08374", "2305.10847", "2302.05794", "2307.02599"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_8636"} +{"question": "What papers proposed to augment the ST training data using techniques like spec-augmentation or mixing at frame, word, and sentence levels?", "answer": ["On Using SpecAugment for End-to-End Speech Translation", "\"M\"³ST: MIX AT THREE LEVELS FOR SPEECH TRANSLATION"], "answer_arxiv_id": ["1911.08876v1", "2212.03657"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_8637"} +{"question": "What models focus on efficient long-term temporal modeling for long-duration video input?", "answer": ["Learning Latent Super-Events to Detect Multiple Activities in Videos", "Temporal Gaussian Mixture Layer for Videos", "Coarse-Fine Networks for Temporal Activity Detection in Videos", "MS-TCT: Multi-Scale Temporal ConvTransformer for Action Detection"], "answer_arxiv_id": ["1712.01938", "1803.06316", "2103.01302", "2112.03902"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_8638"} +{"question": "Which works improved the error bounds via designing an efficient algorithm that performs gradient descent on a convex surrogate for the loss in ReLU regression problem?", "answer": ["Algorithms and SQ Lower Bounds for PAC Learning One-Hidden-Layer ReLU Networks"], "answer_arxiv_id": ["2006.12476"], "source_meta": {"published_time": "20220804"}, "qid": "AutoScholarQuery_train_8639"} +{"question": "Can you provide an example of a paper in which an alternative parameterisation specifying an inverse lengthscale, a ratio of noise variance to kernelscale was used?", "answer": ["Hierarchical Nearest-Neighbor Gaussian Process Models for Large Geostatistical Datasets"], "answer_arxiv_id": ["1406.7343"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_8640"} +{"question": "Which studies have used self-supervised or weakly-supervised learning on large-scale static datasets such as YFCC-100M, Instagram-1B, or LAION-5B?", "answer": ["YFCC100M: The New Data in Multimedia Research", "Exploring the Limits of Weakly Supervised Pretraining", "LAION-5B: An open large-scale dataset for training next generation image-text models"], "answer_arxiv_id": ["1503.01817", "1805.00932", "2210.08402"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_8641"} +{"question": "What work has proposed weight equalization as a technique to reduce quantization error?", "answer": ["Data-Free Quantization Through Weight Equalization and Bias Correction"], "answer_arxiv_id": ["1906.04721"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_8642"} +{"question": "Which papers established the core concept of geometric deep learning?", "answer": ["Geometric deep learning: going beyond Euclidean data", "Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["1611.08097", "2104.13478"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_8643"} +{"question": "Which works have utilized SAM for diverse tasks like 3D understanding and video processing?", "answer": ["Segment Anything Model (SAM) Enhanced Pseudo Labels for Weakly\n Supervised Semantic Segmentation", "Scalable Mask Annotation for Video Text Spotting", "A Dive into SAM Prior in Image Restoration", "Matte Anything: Interactive Natural Image Matting with Segment Anything\n Models", "UVOSAM: A Mask-free Paradigm for Unsupervised Video Object Segmentation\n via Segment Anything Model", "SAM3D: Segment Anything in 3D Scenes", "Zero-Shot Co-salient Object Detection Framework"], "answer_arxiv_id": ["2305.05803", "2305.01443", "2305.13620", "2306.04121", "2305.12659", "2306.03908", "2309.05499"], "source_meta": {"published_time": "20240501"}, "qid": "AutoScholarQuery_train_8644"} +{"question": "Where can Bayesian networks be found in use in applications such as gene regulatory networks, medical decision making and spam filtering?", "answer": ["Medical idioms for clinical Bayesian network development"], "answer_arxiv_id": ["2007.00364v2"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_8645"} +{"question": "Which works explored the use of image encoder outputs as soft prompts for the LLM (Language Loss Minimization)?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models"], "answer_arxiv_id": ["2106.13884"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_8646"} +{"question": "Could you provide me studies about concept learning from images?", "answer": ["ReVersion: Diffusion-Based Relation Inversion from Images"], "answer_arxiv_id": ["2303.13495"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8647"} +{"question": "Which works applied Riemannian geometry to the latent spaces of deep generative models?", "answer": ["Latent Space Oddity: on the Curvature of Deep Generative Models", "The Riemannian Geometry of Deep Generative Models", "Metrics for Deep Generative Models", "Geometrically Enriched Latent Spaces", "On Explicit Curvature Regularization in Deep Generative Models"], "answer_arxiv_id": ["1710.11379", "1711.08014", "1711.01204v2", "2008.00565", "2309.10237"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_8648"} +{"question": "What studies deal with VLMs trained either only on images or only on text?", "answer": ["Text-Only Training for Image Captioning using Noise-Injected CLIP"], "answer_arxiv_id": ["2211.00575"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_8649"} +{"question": "Any works that proposed new metrics like Geometric Evaluation of Data Representations (GCA) which uses geometric and topological properties?", "answer": ["GeomCA: Geometric Evaluation of Data Representations"], "answer_arxiv_id": ["2105.12486"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8650"} +{"question": "In what work, LLMs are used to reformulate queries in search systems?", "answer": ["Generative Query Reformulation for Effective Adhoc Search"], "answer_arxiv_id": ["2308.00415"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_8651"} +{"question": "Which works successfully extended 3D representations to 4D, capturing object dynamics effectively?", "answer": ["NPMs: Neural Parametric Models for 3D Deformable Shapes", "Neural Deformation Graphs for Globally-consistent Non-rigid\n Reconstruction", "Learning Compositional Representation for 4D Captures with Neural ODE", "Learning Parallel Dense Correspondence from Spatio-Temporal Descriptors\n for Efficient and Robust 4D Reconstruction", "CaDeX: Learning Canonical Deformation Coordinate Space for Dynamic\n Surface Representation via Neural Homeomorphism", "A Point Set Generation Network for 3D Object Reconstruction from a\n Single Image", "Neural Shape Deformation Priors", "NAP: Neural 3D Articulation Prior"], "answer_arxiv_id": ["2104.00702", "2012.01451", "2103.08271", "2103.16341", "2203.16529", "1612.00603", "2210.05616", "2305.16315"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_8652"} +{"question": "What paper learns an extra [rln]-token in addition to [obj]-tokens?", "answer": ["Relationformer: A Unified Framework for Image-to-Graph Generation"], "answer_arxiv_id": ["2203.10202"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_8653"} +{"question": "Which works proposed discriminative reranking approach in the context of NMT?", "answer": ["Masked Language Model Scoring", "Residual Energy-Based Models for Text Generation"], "answer_arxiv_id": ["1910.14659", "2004.11714"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_8654"} +{"question": "What are the works utilizing Detection Transformers for pose estimation?", "answer": ["Pose Recognition with Cascade Transformers", "TFPose: Direct Human Pose Estimation with Transformers", "End-to-End Trainable Multi-Instance Pose Estimation with Transformers", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2104.06976", "2103.15320", "2103.12115", "2005.12872"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_8655"} +{"question": "What works attempt to identify the sparsity patterns of the network before training?", "answer": ["Pruning neural networks without any data by iteratively conserving synaptic flow", "Picking Winning Tickets Before Training by Preserving Gradient Flow", "SNIP: Single-shot Network Pruning based on Connection Sensitivity", "Learning Fast Algorithms for Linear Transforms Using Butterfly Factorizations", "Monarch: Expressive Structured Matrices for Efficient and Accurate Training", "Deformable Butterfly: A Highly Structured and Sparse Linear Transform", "Kaleidoscope: An Efficient, Learnable Representation For All Structured Linear Maps", "Pixelated Butterfly: Simple and Efficient Sparse Training for Neural Network Models"], "answer_arxiv_id": ["2006.05467", "2002.07376", "1810.02340", "1903.05895", "2204.00595", "2203.13556", "2012.14966", "2112.00029"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_8656"} +{"question": "Which papers have contributed to the growth of vision-language pre-training (VLP) with the collection large-scale visual and linguistic pairs from the internet?", "answer": ["LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "UNITER: UNiversal Image-TExt Representation Learning", "Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal Transformers", "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision", "UNIMO: Towards Unified-Modal Understanding and Generation via Cross-Modal Contrastive Learning", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "VideoBERT: A Joint Model for Video and Language Representation Learning", "UniVL: A Unified Video and Language Pre-Training Model for Multimodal Understanding and Generation", "Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "Lavender: Unifying Video-Language Understanding as Masked Language Modeling"], "answer_arxiv_id": ["1908.07490", "1909.11740", "2004.00849", "2102.03334", "2012.15409", "2108.10904", "1904.01766", "2002.06353", "2104.00650", "2201.12086", "2206.07160"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_8657"} +{"question": "Could you provide me some studies about generative 3D-aware image synthesis through Generative Adversarial Nets (GANs)?", "answer": ["Generative Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Conditional Generative Adversarial Nets", "Learning Efficient GANs for Image Translation via Differentiable Masks\n and co-Attention Distillation", "Auto-Embedding Generative Adversarial Networks for High Resolution Image\n Synthesis"], "answer_arxiv_id": ["1406.2661", "1812.04948", "1411.1784", "2011.08382", "1903.11250"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_8658"} +{"question": "Could you provide me some research works about benefitting from the controllability in the textual embedding space for diffusion models?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "P+: Extended Textual Conditioning in Text-to-Image Generation"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2303.09522"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_8659"} +{"question": "What papers focused on solving zero-sum Markov games under general function approximation?", "answer": ["The Power of Exploiter: Provable Multi-Agent RL in Large State Spaces"], "answer_arxiv_id": ["2106.03352"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_8660"} +{"question": "What works attempt to discover 'steering' vectors or tokens in control of language models?", "answer": ["Extracting Latent Steering Vectors from Pretrained Language Models", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2205.05124", "2101.00190"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_8661"} +{"question": "What are some studies that implement various visual tasks in a training-free zero-shot manner following the new paradigm of Multimodal Large Language Models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2103.00020", "2112.03857"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_8662"} +{"question": "What are the studies that worked on generative transformers?", "answer": ["Taming Transformers for High-Resolution Image Synthesis", "MaskGIT: Masked Generative Image Transformer", "Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["2012.09841", "2202.04200", "2102.12092"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_8663"} +{"question": "Which studies have established that concept-based explanations are meaningful for users in classification tasks?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)", "Concept Bottleneck Models", "Towards Automatic Concept-based Explanations"], "answer_arxiv_id": ["1711.11279", "2007.04612", "1902.03129"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_8664"} +{"question": "Are there any research works that have proposed dynamic least-to-most prompting in improving CoT?", "answer": ["Compositional Semantic Parsing with Large Language Models"], "answer_arxiv_id": ["2209.15003"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_8665"} +{"question": "Which papers discuss API-based applications for solving vision-centric tasks with large language models?", "answer": ["Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "MM-ReAct : Prompting ChatGPT for Multimodal Reasoning and Action", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face", "VideoChat : Chat-Centric Video Understanding"], "answer_arxiv_id": ["2303.04671", "2303.11381", "2303.17580", "2305.06355"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_8666"} +{"question": "Which research studied the linear bandits with adversarial corruptions?", "answer": ["Stochastic Linear Optimization with Adversarial Corruption", "Achieving Near Instance-Optimality and Minimax-Optimality in Stochastic and Adversarial Linear Bandits Simultaneously"], "answer_arxiv_id": ["1909.02109", "2102.05858"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_8667"} +{"question": "Which papers have studied reducing the bias and variance of gradient estimation for policy gradient methods to stabilize the RL process?", "answer": ["High-Dimensional Continuous Control Using Generalized Advantage Estimation"], "answer_arxiv_id": ["1506.02438"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_8668"} +{"question": "What researchers have combined regularization-based methods, memory-based methods, and expansion-based methods for continual learning?", "answer": ["DER: Dynamically Expandable Representation for Class Incremental Learning", "General Incremental Learning with Domain-aware Categorical Representations"], "answer_arxiv_id": ["2103.16788", "2204.04078"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_8669"} +{"question": "What research developed a balanced contrastive loss to ensure optimal regular simplex configuration in long-tailed classification?", "answer": ["Balanced Contrastive Learning for Long-Tailed Visual Recognition"], "answer_arxiv_id": ["2207.09052"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_8670"} +{"question": "Which studies explore the problem of labeling a large dataset and deciding which points to re-classify when provided a new model?", "answer": ["Backward-Compatible Prediction Updates: A Probabilistic Approach"], "answer_arxiv_id": ["2107.01057"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_8671"} +{"question": "What works contribute to the textual inversion in text-to-image generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_8672"} +{"question": "What paper improves the state-of-the-art result when specializing to the tabular case according to the researcher's analysis?", "answer": ["Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism"], "answer_arxiv_id": ["2103.12021v2"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_8673"} +{"question": "Could you provide me with the paper that proposes a robust alignment check function to filter harmful queries?", "answer": ["Defending Against Alignment-Breaking Attacks via Robustly Aligned LLM"], "answer_arxiv_id": ["2309.14348"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_8674"} +{"question": "Which studies propose methods that work towards fullys exploating the temporal coherence of a video?", "answer": ["See More, Know More: Unsupervised Video Object Segmentation with\n Co-Attention Siamese Networks", "Zero-Shot Video Object Segmentation via Attentive Graph Neural Networks", "Anchor Diffusion for Unsupervised Video Object Segmentation", "F2Net: Learning to Focus on the Foreground for Unsupervised Video Object\n Segmentation", "Iteratively Selecting an Easy Reference Frame Makes Unsupervised Video\n Object Segmentation Easier"], "answer_arxiv_id": ["2001.06810", "2001.06807", "1910.10895", "2012.02534", "2112.12402"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_8675"} +{"question": "Could you provide me some studies that proposed to increase the speed of NeRF?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "MERF: Memory-Efficient Radiance Fields for Real-time View Synthesis in\n Unbounded Scenes", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "TensoRF: Tensorial Radiance Fields", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "BakedSDF: Meshing Neural SDFs for Real-Time View Synthesis", "Baking Neural Radiance Fields for Real-Time View Synthesis", "MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient\n Neural Field Rendering on Mobile Architectures", "Delicate Textured Mesh Recovery from NeRF via Adaptive Surface\n Refinement", "R2L: Distilling Neural Radiance Field to Neural Light Field for\n Efficient Novel View Synthesis", "Real-Time Neural Light Field on Mobile Devices"], "answer_arxiv_id": ["2112.05131", "2302.12249", "2201.05989", "2203.09517", "2103.13744", "2103.14024", "2111.11215", "2302.14859", "2103.14645", "2208.00277", "2303.02091", "2203.17261", "2212.08057"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8676"} +{"question": "Any works that provide a comprehensive review of the recent advancements in VL pre-training?", "answer": ["Vision-Language Pre-training: Basics, Recent Advances, and Future Trends"], "answer_arxiv_id": ["2210.09263"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_8677"} +{"question": "Could you provide me some works that applied diffusion models for style transfer?", "answer": ["Inversion-Based Style Transfer with Diffusion Models"], "answer_arxiv_id": ["2211.13203"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_8678"} +{"question": "Which research work uses pure terms for the second order update?", "answer": ["The Numerics of GANs"], "answer_arxiv_id": ["1705.10461"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_8679"} +{"question": "In which work does iBOT heavily dependent on for its vanilla DINO loss?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.14294"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_8680"} +{"question": "Are there any works specialising in task-specific prompt engineering for image classification, object detection, and visual question answering?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language\n Models", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework"], "answer_arxiv_id": ["2103.00020", "2209.07511", "2202.03052"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8681"} +{"question": "Which research paper proposed A coding, a REC algorithm, which generalizes PFR by introducing a partitioning scheme?", "answer": ["Fast Relative Entropy Coding with A* coding"], "answer_arxiv_id": ["2201.12857"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_8682"} +{"question": "Which works use the discrete tokens derived from a VQ-VAE model as the target in S2ST?", "answer": ["SPEECH-TO-SPEECH TRANSLATION BETWEEN UNTRANSCRIBED UNKNOWN LANGUAGES", "UWSpeech: Speech to Speech Translation for Unwritten Languages"], "answer_arxiv_id": ["1910.00795", "2006.07926"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_8683"} +{"question": "Which works showcased significant capabilities in universal generation or recognition tasks using Multimodal Large Language Models?", "answer": ["GPT-4 Technical Report", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Instruction Tuning with GPT-4", "Visual Instruction Tuning"], "answer_arxiv_id": ["2303.08774", "2201.12086", "2301.12597", "2304.03277", "2304.08485"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_8684"} +{"question": "Which studies have tried to deal with the learning rate in MAML-based methods?", "answer": ["Meta-Learning with Adaptive Hyperparameters", "Meta-learning with negative learning rates"], "answer_arxiv_id": ["2011.00209", "2102.00940"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_8685"} +{"question": "Which work critiqued benchmark reporting metrics for unfairly favouring those with more resources to run experiments?", "answer": ["Show Your Work: Improved Reporting of Experimental Results"], "answer_arxiv_id": ["1909.03004"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_8686"} +{"question": "Could you give me examples of recent papers developing the GPT family of models?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2005.14165", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_8687"} +{"question": "Influence functions are used in what studies for modern machine learning models?", "answer": ["Understanding Black-box Predictions via Influence Functions", "Fair Infinitesimal Jackknife: Mitigating the Influence of Biased Training Data Points Without Refitting", "Achieving Fairness at No Utility Cost via Data Reweighing with Influence"], "answer_arxiv_id": ["1703.04730", "2212.06803", "2202.00787"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_8688"} +{"question": "Which papers showcase the use of LLMs in the field of mathematical problem-solving?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "MathPrompter: Mathematical Reasoning using Large Language Models"], "answer_arxiv_id": ["2201.11903", "2303.05398"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_8689"} +{"question": "Which works explore Membership Inference Attack for classification models?", "answer": ["Membership Inference Attacks Against Machine Learning Models"], "answer_arxiv_id": ["1610.05820"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_8690"} +{"question": "Which works contribute to the modeling of the solution of a PDE using kernel methods?", "answer": ["Solving and Learning Nonlinear PDEs with Gaussian Processes", "Learning Partial Differential Equations in Reproducing Kernel Hilbert Spaces"], "answer_arxiv_id": ["2103.12959", "2108.11580"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_8691"} +{"question": "Which study introduced the concept of neural ODE?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_8692"} +{"question": "What research papers focus on the study of online reward poisoning attacks in single agent RL?", "answer": ["Vulnerability of Deep Reinforcement Learning to Policy Induction Attacks", "Adaptive Reward-Poisoning Attacks against Reinforcement Learning", "Understanding the Limits of Poisoning Attacks in Episodic Reinforcement Learning"], "answer_arxiv_id": ["1701.04143", "2003.12613", "2208.13663"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_8693"} +{"question": "Which papers focus on model-free reinforcement learning algorithms?", "answer": ["Proximal Policy Optimization Algorithms", "Playing Atari with Deep Reinforcement Learning", "Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor", "Continuous control with deep reinforcement learning"], "answer_arxiv_id": ["1707.06347v2", "1312.5602", "1801.01290", "1509.02971"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_8694"} +{"question": "What paper fine-tuned 2D diffusion model with multi-view dataset to mitigate a problem identified in Score Distillation?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2303.11328"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_8695"} +{"question": "Could you provide me some studies about using GFlowNets in the field of biological sequences design?", "answer": ["Biological Sequence Design with GFlowNets"], "answer_arxiv_id": ["2203.04115"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_8696"} +{"question": "Which papers have studied the general non-realizable case in the private online optimization literature?", "answer": ["Differentially Private Online Learning", "The Price of Differential Privacy for Online Learning", "Practical and Private (Deep) Learning Without Sampling or Shuffling"], "answer_arxiv_id": ["1109.0105", "1701.07953", "2103.00039"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_8697"} +{"question": "What studies have introduced the fixed memory to store real samples in rehearsal-based continual learning (CL) methods?", "answer": ["Gradient Episodic Memory for Continual Learning", "Selective Experience Replay for Lifelong Learning", "Experience Replay for Continual Learning", "Episodic Memory in Lifelong Language Learning"], "answer_arxiv_id": ["1706.08840", "1802.10269", "1811.11682", "1906.01076"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_8698"} +{"question": "Which work extended CoOp to solve weak generalizability on unseen category?", "answer": ["Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2203.05557"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_8699"} +{"question": "Could you provide me some works about non-autoregressive models in ASR?", "answer": ["Listen and Fill in the Missing Letters: Non-Autoregressive Transformer\n for Speech Recognition", "Mask CTC: Non-Autoregressive End-to-End ASR with CTC and Mask Predict", "Pushing the Limits of Non-Autoregressive Speech Recognition", "Align-Refine: Non-Autoregressive Speech Recognition via Iterative\n Realignment", "Intermediate Loss Regularization for CTC-based Speech Recognition", "Relaxing the Conditional Independence Assumption of CTC-based ASR by\n Conditioning on Intermediate Predictions"], "answer_arxiv_id": ["1911.04908", "2005.08700", "2104.03416", "2010.14233", "2102.03216", "2104.02724"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_8700"} +{"question": "Which research proposed the J-Invariance theory that achieves self-supervised denoising?", "answer": ["Noise2Self: Blind Denoising by Self-Supervision"], "answer_arxiv_id": ["1901.11365"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_8701"} +{"question": "Which paper first proposed implicitly differentiating through KKT conditions in the area of differentiable optimization?", "answer": ["OptNet: Differentiable Optimization as a Layer in Neural Networks"], "answer_arxiv_id": ["1703.00443"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_8702"} +{"question": "Which papers propose methods for precision enhancement through hard truncation in generative model training?", "answer": ["StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets"], "answer_arxiv_id": ["2202.00273"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8703"} +{"question": "Could you mention research papers that focus on metric learning in the context of Few-Shot Learning?", "answer": ["Prototypical Networks for Few-shot Learning", "Matching Networks for One Shot Learning", "Negative Margin Matters: Understanding Margin in Few-shot Classification", "Prototype Rectification for Few-Shot Learning"], "answer_arxiv_id": ["1703.05175", "1606.04080", "2003.12060", "1911.10713"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_8704"} +{"question": "In the context of machine learning, which studies have achieved differential privacy through modifying either the training algorithm?", "answer": ["Differentially Private Empirical Risk Minimization", "Deep Learning with Differential Privacy"], "answer_arxiv_id": ["0912.0071v5", "1607.00133"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_8705"} +{"question": "Which work showed the BCR bound for a multi-objective extension of GP-UCB?", "answer": ["A Flexible Framework for Multi-Objective Bayesian Optimization using Random Scalarizations"], "answer_arxiv_id": ["1805.12168"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_8706"} +{"question": "Could you provide me with the works where the visual encoder from one study was used and a linear layer was used for projecting visual features into another study's feature space?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2304.10592", "2301.12597"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_8707"} +{"question": "Which paper suggested that adjacent frames in video data have semantically similar information?", "answer": ["Unsupervised Learning of Spatiotemporally Coherent Metrics"], "answer_arxiv_id": ["1412.6056"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_8708"} +{"question": "Which works proposed methods for instruction optimization?", "answer": ["Large Language Models Are Human-Level Prompt Engineers", "Automatic Prompt Optimization with \"Gradient Descent\" and Beam Search"], "answer_arxiv_id": ["2211.01910", "2305.03495"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_8709"} +{"question": "What research focuses on enhancing the UNet model by extending to 3D volumes?", "answer": ["V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation", "Non-local U-Nets for Biomedical Image Segmentation"], "answer_arxiv_id": ["1606.04797", "1812.04103"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_8710"} +{"question": "Which works described the phenomenon leading poor performance in offline RL due to distributional shift?", "answer": ["Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems", "Boosting Offline Reinforcement Learning via Data Rebalancing"], "answer_arxiv_id": ["2005.01643", "2210.09241"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_8711"} +{"question": "Can you provide some research papers about multi-task offline RL algorithms that use data filtering from datasets collected from multiple tasks?", "answer": ["Conservative Data Sharing for Multi-Task Offline Reinforcement Learning", "MT-Opt: Continuous Multi-Task Robotic Reinforcement Learning at Scale"], "answer_arxiv_id": ["2109.08128", "2104.08212"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_8712"} +{"question": "Which studies are about image-to-image translation?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "Toward Multimodal Image-to-Image Translation", "A Variational U-Net for Conditional Appearance and Shape Generation"], "answer_arxiv_id": ["1611.07004", "1711.11586", "1804.04694"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_8713"} +{"question": "What works have been adapting language models for complex interactive reasoning tasks in ScienceWorld?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2204.01691", "2210.03629"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8714"} +{"question": "Can you name some studies where attention-based techniques have been used in image dehazing?", "answer": ["GridDehazeNet: Attention-Based Multi-Scale Network for Image Dehazing", "FFA-Net: Feature Fusion Attention Network for Single Image Dehazing"], "answer_arxiv_id": ["1908.03245", "1911.07559"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_8715"} +{"question": "Which studies showed the promise of pre-trained models in visual navigation?", "answer": ["Simple but Effective: CLIP Embeddings for Embodied AI", "Offline Visual Representation Learning for Embodied Navigation", "OVRL-V2: A simple state-of-art baseline for ImageNav and ObjectNav"], "answer_arxiv_id": ["2111.09888", "2204.13226", "2303.07798"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_8716"} +{"question": "What studies successfully applied Retrieval Augmented Generation (RAG) to large language models (LLMs)?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models", "REPLUG: Retrieval-Augmented Black-Box Language Models", "In-Context Retrieval-Augmented Language Models", "SILO Language Models: Isolating Legal Risk In a Nonparametric Datastore"], "answer_arxiv_id": ["1911.00172", "2301.12652", "2302.00083", "2308.04430"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_8717"} +{"question": "What research mentions the use of two parameters that capture different spectral statistics as fingerprints?", "answer": ["Fourier Spectrum Discrepancies in Deep Network Generated Images"], "answer_arxiv_id": ["1911.06465"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_8718"} +{"question": "What studies employed SGD-based approximation in Bayesian deep learning?", "answer": ["Stochastic Gradient Descent as Approximate Bayesian Inference", "A Simple Baseline for Bayesian Uncertainty in Deep Learning", "Fast Adaptation with Linearized Neural Networks", "Bayesian Deep Learning and a Probabilistic Perspective of Generalization"], "answer_arxiv_id": ["1704.04289", "1902.02476", "2103.01439", "2002.08791v4"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_8719"} +{"question": "What work introduced OneFormer as the pioneer 2D image segmentation approach?", "answer": ["OneFormer: One Transformer to Rule Universal Image Segmentation"], "answer_arxiv_id": ["2211.06220"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_8720"} +{"question": "Are there any studies that used SCVIS or UMAP before applying K-Means clustering in SDMs?", "answer": ["No Subclass Left Behind: Fine-Grained Robustness in Coarse-Grained Classification Problems"], "answer_arxiv_id": ["2011.12945"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8721"} +{"question": "Which work used voxel grids as world representations in the field of Neural Radiance Fields?", "answer": ["Neural Sparse Voxel Fields"], "answer_arxiv_id": ["2007.11571"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_8722"} +{"question": "Could you provide some studies where positive pairs in contrastive approaches are constructed by co-occurrence?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Contrastive Multiview Coding", "Learning Representations by Maximizing Mutual Information Across Views"], "answer_arxiv_id": ["1807.03748", "1906.05849", "1906.00910"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_8723"} +{"question": "Which studies made initial explorations into the problem of generating gesture videos directly?", "answer": ["Audio-driven Neural Gesture Reenactment with Video Motion Graphs", "Audio-Driven Co-Speech Gesture Video Generation"], "answer_arxiv_id": ["2207.11524", "2212.02350"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_8724"} +{"question": "Can you provide references about researches that proposed supervised contrastive learning?", "answer": ["Supervised Contrastive Learning", "Self-supervised Co-training for Video Representation Learning"], "answer_arxiv_id": ["2004.11362", "2010.09709"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_8725"} +{"question": "Which research paper proposes a 2-D relative position encoding for image classification?", "answer": ["Attention Augmented Convolutional Networks"], "answer_arxiv_id": ["1904.09925"], "source_meta": {"published_time": "20210222"}, "qid": "AutoScholarQuery_train_8726"} +{"question": "What papers are about the application of diffusion models in audio and video data?", "answer": ["Diffsound: Discrete Diffusion Model for Text-to-sound Generation", "Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models", "AudioLDM: Text-to-Audio Generation with Latent Diffusion Models", "Text-to-Audio Generation using Instruction-Tuned LLM and Latent Diffusion Model", "Video Diffusion Models", "Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2207.09983", "2301.12661", "2301.12503", "2304.13731", "2204.03458", "2304.08818"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_8727"} +{"question": "What datasets are focused on action sounds in kitchen environments?", "answer": ["Epic-Sounds: A Large-scale Dataset of Actions That Sound"], "answer_arxiv_id": ["2302.00646"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_8728"} +{"question": "What researches study benign overfitting in regression problems?", "answer": ["Benign Overfitting in Linear Regression"], "answer_arxiv_id": ["1906.11300"], "source_meta": {"published_time": "20220616"}, "qid": "AutoScholarQuery_train_8729"} +{"question": "Could you provide me some studies about the non-Euclidean embedding method in knowledge graph embedding?", "answer": ["Multi-relational Poincaré Graph Embeddings", "Low-Dimensional Hyperbolic Knowledge Graph Embeddings", "Modeling Heterogeneous Hierarchies with Relation-specific Hyperbolic Cones", "Ultrahyperbolic Knowledge Graph Embeddings"], "answer_arxiv_id": ["1905.09791", "2005.00545", "2110.14923", "2206.00449"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_8730"} +{"question": "Which research introduced a fine-tuning approach for mitigating spurious correlations in pre-trained multi-modal frameworks?", "answer": ["Mitigating Spurious Correlations in Multi-modal Models during\n Fine-tuning"], "answer_arxiv_id": ["2304.03916"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8731"} +{"question": "Which works extend the GAMMA framework?", "answer": ["Synthesizing Diverse Human Motions in 3D Indoor Scenes"], "answer_arxiv_id": ["2305.12411"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_8732"} +{"question": "Are there any studies reporting the limitations of LSTMs and Transformers in learning certain types of languages?", "answer": ["How Can Self-Attention Networks Recognize Dyck-n Languages?"], "answer_arxiv_id": ["2010.04303"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_8733"} +{"question": "Can you provide research that outputs point estimates while tackling inverse problems using optimization?", "answer": ["Diffusion models as plug-and-play priors"], "answer_arxiv_id": ["2206.09012"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8734"} +{"question": "What research proposed the framework for human action recognition (HAR)?", "answer": ["Action Transformer: A Self-Attention Model for Short-Time Pose-Based\n Human Action Recognition", "Action Recognition via Pose-Based Graph Convolutional Networks with\n Intermediate Dense Supervision", "SpatioTemporal Focus for Skeleton-based Action Recognition", "Temporal Segment Networks: Towards Good Practices for Deep Action\n Recognition", "Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset", "TSM: Temporal Shift Module for Efficient Video Understanding", "SlowFast Networks for Video Recognition", "Temporal Relational Reasoning in Videos", "Temporal Interlacing Network", "Learning Spatiotemporal Features with 3D Convolutional Networks", "Is Space-Time Attention All You Need for Video Understanding?", "Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action\n Recognition"], "answer_arxiv_id": ["2107.00606", "1911.12509", "2203.16767", "1608.00859", "1705.07750", "1811.08383", "1812.03982", "1711.08496", "2001.06499", "1412.0767", "2102.05095", "1801.07455"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_8735"} +{"question": "Which studies are notable approaches in the field of learned generative models?", "answer": ["Auto-Encoding Variational Bayes", "Neural Discrete Representation Learning", "Generative Adversarial Networks", "Variational Inference with Normalizing Flows", "NICE: Non-linear Independent Components Estimation", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Score-Based Generative Modeling through Stochastic Differential Equations", "Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["1312.6114", "1711.00937", "2203.00667", "1505.05770", "1410.8516", "1503.03585", "2011.13456", "2106.15282"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_8736"} +{"question": "Can you provide any work that focuses on utilising prompt ensembling to combine multiple prompts?", "answer": ["ECO: Ensembling Context Optimization for Vision-Language Models"], "answer_arxiv_id": ["2307.14063"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8737"} +{"question": "Which work established minimax optimal universal dynamic regret for exp-concave functions?", "answer": ["Optimal Dynamic Regret in Exp-Concave Online Learning"], "answer_arxiv_id": ["2104.11824"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_8738"} +{"question": "Which paper addressed an apply-accept interaction protocol in two-sided markets?", "answer": ["Optimizing Rankings for Recommendation in Matching Markets"], "answer_arxiv_id": ["2106.01941"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_8739"} +{"question": "Which papers have studied algorithms with flavors of online optimization for multicalibration?", "answer": ["Online Multivalid Learning: Means, Moments, and Prediction Intervals"], "answer_arxiv_id": ["2101.01739"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_8740"} +{"question": "What studies developed view-volume networks?", "answer": ["View-volume Network for Semantic Scene Completion from a Single Depth\n Image", "ForkNet: Multi-branch Volumetric Semantic Completion from a Single Depth\n Image", "LMSCNet: Lightweight Multiscale 3D Semantic Completion", "Anisotropic Convolutional Networks for 3D Semantic Scene Completion"], "answer_arxiv_id": ["1806.05361", "1909.01106", "2008.10559", "2004.02122"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_8741"} +{"question": "What research works proposed matching the trajectory distribution in the context of exploiting teacher policy?", "answer": ["Generative Adversarial Imitation Learning", "On Value Discrepancy of Imitation Learning"], "answer_arxiv_id": ["1606.03476", "1911.07027"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_8742"} +{"question": "Any works about the approximation propagation strategy for scalable GNN?", "answer": ["Scalable Graph Neural Networks via Bidirectional Propagation", "GRAND+: Scalable Graph Random Neural Networks", "Scaling Graph Neural Networks with Approximate PageRank"], "answer_arxiv_id": ["2010.15421", "2203.06389", "2007.01570"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_8743"} +{"question": "Which works discussed affordance for the classic grasping task in robotics?", "answer": ["Learning Dexterous Grasping with Object-Centric Visual Affordances", "Learning Task-Oriented Grasping from Human Activity Datasets", "Robotic Pick-and-Place of Novel Objects in Clutter with Multi-Affordance Grasping and Cross-Domain Image Matching"], "answer_arxiv_id": ["2009.01439", "1910.11669", "1710.01330"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_8744"} +{"question": "What works have been done to study federated learning settings where free-riders may send in fabricated gradients?", "answer": ["Free-rider Attacks on Model Aggregation in Federated Learning", "Free-riders in Federated Learning: Attacks and Defenses"], "answer_arxiv_id": ["2006.11901v5", "1911.12560"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_8745"} +{"question": "Could you tell me some works that utilized context encoders or gradient descents for few-shot RL?", "answer": ["Robust Task Representations for Offline Meta-Reinforcement Learning via Contrastive Learning", "VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning", "SMART: Self-supervised Multi-task pretrAining with contRol Transformers", "Multi-Game Decision Transformers"], "answer_arxiv_id": ["2206.10442", "1910.08348", "2301.09816v1", "2205.15241"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_8746"} +{"question": "What are some works that solve network verification through linear programming?", "answer": ["Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Neural Network Robustness Verification", "General Cutting Planes for Bound-Propagation-Based Neural Network Verification"], "answer_arxiv_id": ["2103.06624", "2208.05740"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_8747"} +{"question": "Can you name studies that considered diminishing regularization to provide an unbiased solution in NPG?", "answer": ["Policy Mirror Descent for Reinforcement Learning: Linear Convergence, New Sampling Complexity, and Generalized Problem Classes"], "answer_arxiv_id": ["2102.00135"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_8748"} +{"question": "What papers proposed the concept of Composable Diffusion?", "answer": ["Compositional Visual Generation with Composable Diffusion Models"], "answer_arxiv_id": ["2206.01714"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_8749"} +{"question": "What research is there on using graphs derived from egocentric videos to enhance long-term video understanding and egocentric action anticipation?", "answer": ["EGO-TOPO: Environment Affordances from Egocentric Video"], "answer_arxiv_id": ["2001.04583"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_8750"} +{"question": "Could provide me studies that have leveraged the code generation capabilities of LLMs to design multi-agent collaborative systems?", "answer": ["CAMEL: Communicative Agents for \"Mind\" Exploration of Large Language Model Society"], "answer_arxiv_id": ["2303.17760v2"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8751"} +{"question": "Which researchers explored non-uniform resizing in vision tasks?", "answer": ["Learning to Zoom: a Saliency-Based Sampling Layer for Neural Networks", "Looking for the Devil in the Details: Learning Trilinear Attention Sampling Network for Fine-grained Image Recognition", "FOVEA: Foveated Image Magnification for Autonomous Navigation", "SALISA: Saliency-based Input Sampling for Efficient Video Object Detection", "Learning to Zoom and Unzoom"], "answer_arxiv_id": ["1809.03355", "1903.06150", "2108.12102", "2204.02397", "2303.15390"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_8752"} +{"question": "Which papers discuss network distillation?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_8753"} +{"question": "Can you provide the works which proposed the PrivUnit mechanism and its variant PrivUnitG for privatizing unit L2 vectors under a communication constraint?", "answer": ["Protection Against Reconstruction and Its Applications in Private Federated Learning", "Optimal Algorithms for Mean Estimation under Local Differential Privacy"], "answer_arxiv_id": ["1812.00984", "2205.02466"], "source_meta": {"published_time": "20221108"}, "qid": "AutoScholarQuery_train_8754"} +{"question": "Which concurrent works have used retrieval to improve diffusion models?", "answer": ["Memory-Driven Text-to-Image Generation", "KNN-Diffusion: Image Generation via Large-Scale Retrieval"], "answer_arxiv_id": ["2208.07022", "2204.02849"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_8755"} +{"question": "What papers propose representing human motions with discrete latents?", "answer": ["PoseGPT: Quantization-based 3D Human Motion Generation and Forecasting", "MotionGPT: Human Motion as a Foreign Language", "T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete\n Representations"], "answer_arxiv_id": ["2210.10542", "2306.14795", "2301.06052"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_8756"} +{"question": "Could you provide me some methods proposed for large language models?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["1902.00751", "2106.09685", "2104.08691"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_8757"} +{"question": "Any works that demonstrate generative models pre-trained on code can produce Python snippets to tackle competitive programming challenges?", "answer": ["Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2107.03374"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_8758"} +{"question": "What were the advancements made in reinforcement learning by integrating differentiable simulators into policy optimization?", "answer": ["Accelerated Policy Learning with Parallel Differentiable Simulation"], "answer_arxiv_id": ["2204.07137"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_8759"} +{"question": "What studies perform inversion in the image space of diffusion models?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Implicit Models", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models"], "answer_arxiv_id": ["2108.02938", "2105.05233", "2010.02502", "2211.09794"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_8760"} +{"question": "What papers have extended the annotation scope to whole documents to construct more practical datasets in event extraction", "answer": ["Multi-Sentence Argument Linking", "Document-Level Event Argument Extraction by Conditional Generation"], "answer_arxiv_id": ["1911.03766", "2104.05919"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_8761"} +{"question": "Which works investigated the non-asymptotic convergence for mixture of past policies?", "answer": ["A Primal-Dual Approach to Constrained Markov Decision Processes", "Faster Algorithm and Sharper Analysis for Constrained Markov Decision Process"], "answer_arxiv_id": ["2101.10895", "2110.10351v1"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_8762"} +{"question": "Are there any works about the application of LLMs in event-related information extraction tasks?", "answer": ["Exploring the Feasibility of ChatGPT for Event Extraction", "Large Language Model Is Not a Good Few-shot Information Extractor, but a\n Good Reranker for Hard Samples!"], "answer_arxiv_id": ["2303.03836", "2303.08559"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_8763"} +{"question": "Any research papers that studied transformers models in the perspective of logical and ethical reasoning?", "answer": ["Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models"], "answer_arxiv_id": ["2206.04615"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_8764"} +{"question": "Which works utilized local attention approach in Streaming ASR?", "answer": ["Monotonic Chunkwise Attention"], "answer_arxiv_id": ["1712.05382"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_8765"} +{"question": "Which papers discuss deterministic methods for next-frame prediction in video generation?", "answer": ["Unsupervised Learning for Physical Interaction through Video Prediction"], "answer_arxiv_id": ["1605.07157"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_8766"} +{"question": "Which benchmarks were used to test the performance of the efficient Transformer variants?", "answer": ["Long Range Arena: A Benchmark for Efficient Transformers", "Scrolls: Standardized CompaRison Over Long Language Sequences"], "answer_arxiv_id": ["2011.04006", "2201.03533"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_8767"} +{"question": "What works introduced a Bayesian strategy for debiasing scene graphs in images?", "answer": ["Probabilistic Debiasing of Scene Graphs"], "answer_arxiv_id": ["2211.06444"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_8768"} +{"question": "What work marked the surge in popularity of Text-to-Image (T2I) models in 2021?", "answer": ["Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_8769"} +{"question": "Which series of works integrated pretrained ViT models to an encoder-decoder language framework?", "answer": ["PaLI: A Jointly-Scaled Multilingual Language-Image Model", "PaLI-X: On Scaling up a Multilingual Vision and Language Model", "PaLI-3 Vision Language Models: Smaller, Faster, Stronger"], "answer_arxiv_id": ["2209.06794", "2305.18565", "2310.09199"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_8770"} +{"question": "Could you tell me a research that provides a comprehensive overview of MLLMs, including an evaluation of their performance and capabilities?", "answer": ["A Survey on Multimodal Large Language Models"], "answer_arxiv_id": ["2306.13549"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_8771"} +{"question": "Which works ignore the structure information in aleatoric uncertainty estimation for semantic segmentation and may suffer from inconsistent estimation?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks"], "answer_arxiv_id": ["1703.04977", "1807.07356"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_8772"} +{"question": "What works are representative of two-stage convolution-based object detectors?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", "Mask R-CNN"], "answer_arxiv_id": ["1506.01497", "1703.06870"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_8773"} +{"question": "Which works introduced the use of transformers in graph tasks through structure encoding techniques?", "answer": ["Graph-Bert: Only Attention is Needed for Learning Graph Representations", "A Generalization of Transformer Networks to Graphs", "Rethinking Graph Transformers with Spectral Attention", "Pure Transformers are Powerful Graph Learners", "Representing Long-Range Context for Graph Neural Networks with Global Attention", "GRPE: Relative Positional Encoding for Graph Transformer", "Molecule Attention Transformer", "Structure-Aware Transformer for Graph Representation Learning", "GraphiT: Encoding Graph Structure in Transformers", "From block-Toeplitz matrices to differential equations on graphs: towards a general theory for scalable masked Transformers"], "answer_arxiv_id": ["2001.05140", "2012.09699", "2106.03893", "2207.02505", "2201.08821", "2201.12787", "2002.08264v1", "2202.03036", "2106.05667", "2107.07999"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_8774"} +{"question": "Which work supports the idea that non-sparse latent features are usually encoded in the subspace orthogonal to the vocabulary representation?", "answer": ["Toy Models of Superposition"], "answer_arxiv_id": ["2209.10652v1"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_8775"} +{"question": "What is the study that explored the factors contributing to object hallucinations in VLMs?", "answer": ["Analyzing and Mitigating Object Hallucination in Large Vision-Language\n Models"], "answer_arxiv_id": ["2310.00754"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_8776"} +{"question": "What works regard its top-1 prediction as true labels in Partial Label Learning?", "answer": ["Progressive Identification of True Labels for Partial-Label Learning", "Structured Prediction with Partial Labelling through the Infimum Loss"], "answer_arxiv_id": ["2002.08053", "2003.00920"], "source_meta": {"published_time": "20230917"}, "qid": "AutoScholarQuery_train_8777"} +{"question": "Any studies extended NeRF models to predict and render feature fields?", "answer": ["Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D\n Image Representations", "Decomposing NeRF for Editing via Feature Field Distillation", "CLIP-Fields: Weakly Supervised Semantic Fields for Robotic Memory"], "answer_arxiv_id": ["2209.03494", "2205.15585", "2210.05663"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_8778"} +{"question": "What papers are about program synthesis via language models?", "answer": ["Evaluating Large Language Models Trained on Code", "Code Llama: Open Foundation Models for Code", "Sparks of Artificial General Intelligence: Early experiments with GPT-4"], "answer_arxiv_id": ["2107.03374", "2308.12950", "2303.12712v5"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_8779"} +{"question": "Could you provide me some works discussing about the differences between UPop and other methods?", "answer": ["Playing Lottery Tickets with Vision and Language"], "answer_arxiv_id": ["2104.11832"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_8780"} +{"question": "Could you provide me the studies to preserve the energy of the system in Continuous Dynamic Models?", "answer": ["PDE-GCN: Novel Architectures for Graph Neural Networks Motivated by Partial Differential Equations", "Graph-Coupled Oscillator Networks"], "answer_arxiv_id": ["2108.01938", "2202.02296"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_8781"} +{"question": "Which papers proposed reducing GPU memory consumption of intermediate tensors through methods such as lightweight operations, distributed optimization scheduling, and mixed precision training?", "answer": ["MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications", "Aggregated Residual Transformations for Deep Neural Networks", "MorphMLP: An Efficient MLP-Like Backbone for Spatial-Temporal Representation Learning", "ZeRO: Memory Optimizations Toward Training Trillion Parameter Models", "Mixed Precision Training"], "answer_arxiv_id": ["1704.04861", "1611.05431", "2111.12527", "1910.02054", "1710.03740"], "source_meta": {"published_time": "20220228"}, "qid": "AutoScholarQuery_train_8782"} +{"question": "Which papers discuss the safety and reliability problems of MLLMs, such as value alignment and hallucination issues?", "answer": ["Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' Alignment", "Evaluating Object Hallucination in Large Vision-Language Models", "Detecting and Preventing Hallucinations in Large Vision Language Models"], "answer_arxiv_id": ["2308.05374v2", "2305.10355", "2308.06394"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_8783"} +{"question": "Can you provide me researches that used style transfer to create paraphrases of a certain linguistic style?", "answer": ["Reformulating Unsupervised Style Transfer as Paraphrase Generation"], "answer_arxiv_id": ["2010.05700"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_8784"} +{"question": "Which works have utilized a multi-role dialogue to collaborate or debate with each other for a more comprehensive response?", "answer": ["Improving Language Model Negotiation with Self-Play and In-Context\n Learning from AI Feedback", "Encouraging Divergent Thinking in Large Language Models through\n Multi-Agent Debate", "PEER: A Collaborative Language Model", "Self-collaboration Code Generation via ChatGPT", "Generative Agents: Interactive Simulacra of Human Behavior", "Dynamic LLM-Agent Network: An LLM-agent Collaboration Framework with\n Agent Team Optimization"], "answer_arxiv_id": ["2305.10142", "2305.19118", "2208.11663", "2304.07590", "2304.03442", "2310.02170"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_8785"} +{"question": "Can you find any studies that apply a SARSA-style TD-update algorithm to avoid querying values for out-of-distribution actions?", "answer": ["Offline RL Without Off-Policy Evaluation", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["2106.08909", "2110.06169"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_8786"} +{"question": "Who came up with model-agnostic frameworks like MAML that ADKF-IFT compares to?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["1703.03400"], "source_meta": {"published_time": "20220505"}, "qid": "AutoScholarQuery_train_8787"} +{"question": "Can you provide some references that extended the framework of OCO with finite memory to incorporate aspects like non-stationarity and switching costs?", "answer": ["Non-stationary Online Learning with Memory and Non-stochastic Control", "Online Optimization with Memory and Competitive Control"], "answer_arxiv_id": ["2102.03758", "2002.05318"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_8788"} +{"question": "What works proposed metric learning approaches related to contrastive learning?", "answer": ["FaceNet: A Unified Embedding for Face Recognition and Clustering"], "answer_arxiv_id": ["1503.03832"], "source_meta": {"published_time": "20221110"}, "qid": "AutoScholarQuery_train_8789"} +{"question": "What work is based on independence and goodness-of-fit testing for the aggregated tests?", "answer": ["Adaptive test of independence based on HSIC measures", "KSD Aggregated Goodness-of-fit Test"], "answer_arxiv_id": ["1902.06441v5", "2202.00824"], "source_meta": {"published_time": "20211028"}, "qid": "AutoScholarQuery_train_8790"} +{"question": "What papers showed advancements towards training on real-world scene-centric point clouds in 3D representation learning?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts", "Masked Scene Contrast: A Scalable Framework for Unsupervised 3D\n Representation Learning", "Self-supervised Pre-training with Masked Shape Prediction for 3D Scene\n Understanding"], "answer_arxiv_id": ["2007.10985", "2012.09165", "2303.14191", "2305.05026"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_8791"} +{"question": "What are the studies that proposed 3D asset generation methods requiring simple text prompts?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "CLIP-Actor: Text-Driven Recommendation and Stylization for Animating\n Human Meshes", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "AvatarCraft: Transforming Text into Neural Human Avatars with\n Parameterized Shape and Pose Control", "DreamWaltz: Make a Scene with Complex 3D Animatable Avatars", "Zero-Shot Text-Guided Object Generation with Dream Fields", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "Text2Tex: Text-driven Texture Synthesis via Diffusion Models", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "TANGO: Text-driven Photorealistic and Robust 3D Stylization via Lighting\n Decomposition"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2206.04382", "2205.08535", "2303.17606", "2305.12529", "2112.01455", "2112.03221", "2303.11396", "2303.13873", "2210.11277"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_8792"} +{"question": "Can you mention a study that proposes a surrogate task to learn motion representation from sharp videos in an unsupervised manner?", "answer": ["Bringing Alive Blurred Moments"], "answer_arxiv_id": ["1804.02913"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_8793"} +{"question": "Could you give me examples of research that uses reinforcement learning and evolutionary algorithms for selecting important structures?", "answer": ["AMC: AutoML for Model Compression and Acceleration on Mobile Devices", "MetaPruning: Meta Learning for Automatic Neural Network Channel Pruning", "DHP: Differentiable Meta Pruning via HyperNetworks"], "answer_arxiv_id": ["1802.03494", "1903.10258", "2003.13683"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_8794"} +{"question": "Could you provide me research where multiple commodity cameras are used to simplify fluid flow acquisition?", "answer": ["ScalarFlow: A Large-Scale Volumetric Data Set of Real-world Scalar Transport Flows for Computer Animation and Machine Learning"], "answer_arxiv_id": ["2011.10284"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_8795"} +{"question": "Can you name works that have used variational approach focusing on differential equations in linear combination form or closed-form first-order ODEs?", "answer": ["Weak SINDy for Partial Differential Equations", "Weak SINDy: Galerkin-Based Data-Driven Model Selection", "Using Noisy or Incomplete Data to Discover Models of Spatiotemporal Dynamics"], "answer_arxiv_id": ["2007.02848", "2005.04339", "1911.03365"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_8796"} +{"question": "Could you provide me some studies that utilized 3D convolution to enhance point features?", "answer": ["FKAConv: Feature-Kernel Alignment for Point Cloud Convolution", "PointConv: Deep Convolutional Networks on 3D Point Clouds", "KPConv: Flexible and Deformable Convolution for Point Clouds", "A Closer Look at Local Aggregation Operators in Point Cloud Analysis", "PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies"], "answer_arxiv_id": ["2004.04462", "1811.07246", "1904.08889", "2007.01294", "2206.04670"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8797"} +{"question": "What works propose the concept of OOD detection and energy score for pseudo-labeling in imbalanced SSL?", "answer": ["Energy-based Out-of-distribution Detection"], "answer_arxiv_id": ["2010.03759"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_8798"} +{"question": "What research did not show the superiority of randomized communication in distributed optimization?", "answer": ["Communication-Efficient Topologies for Decentralized Learning with $O(1)$ Consensus Rate"], "answer_arxiv_id": ["2210.07881v2"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_8799"} +{"question": "What research works propose the notion of Bayesian differential privacy?", "answer": ["Bayesian Differential Privacy for Machine Learning"], "answer_arxiv_id": ["1901.09697"], "source_meta": {"published_time": "20210620"}, "qid": "AutoScholarQuery_train_8800"} +{"question": "Which paper offers OffWorld gym, a platform for navigation tasks with a wheeled robot?", "answer": ["OffWorld Gym: open-access physical robotics environment for real-world reinforcement learning benchmark and research"], "answer_arxiv_id": ["1910.08639"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_8801"} +{"question": "What references are about neural-ODE models for sequential prediction that predict the process derivative?", "answer": ["Neural Ordinary Differential Equations", "Neural SDE: Stabilizing Neural ODE Networks with Stochastic Noise", "Q"], "answer_arxiv_id": ["1806.07366", "1906.02355", "1611.08152"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_8802"} +{"question": "What studies focus on achieving alignment autonomously?", "answer": ["Principle-Driven Self-Alignment of Language Models from Scratch with\n Minimal Human Supervision", "Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignment", "Human-Instruction-Free LLM Self-Alignment with Limited Samples", "Self-Rewarding Language Models", "Self-Play Fine-Tuning Converts Weak Language Models to Strong Language\n Models", "Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak\n Supervision"], "answer_arxiv_id": ["2305.03047", "2401.12474v1", "2401.06785v1", "2401.10020", "2401.01335", "2312.09390"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_8803"} +{"question": "Which learning-based metrics were specifically designed for image captioning?", "answer": ["PR-MCS: Perturbation Robust Metric for MultiLingual Image Captioning", "UMIC: An Unreferenced Metric for Image Captioning via Contrastive\n Learning", "Quality Estimation for Image Captions Based on Large-scale Human\n Evaluations"], "answer_arxiv_id": ["2303.08389", "2106.14019", "1909.03396"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_8804"} +{"question": "Which studies discuss fine-grained designed choices in GNNs such as data augmentation, layer type, and graph pooling?", "answer": ["Graph Contrastive Learning Automated", "Rethinking Graph Neural Architecture Search from Message-passing", "DiffMG: Differentiable Meta Graph Search for Heterogeneous Graph Neural Networks", "Search to aggregate neighborhood for graph neural network", "Pooling Architecture Search for Graph Classification"], "answer_arxiv_id": ["2106.07594", "2103.14282", "2010.03250", "2104.06608", "2108.10587"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_8805"} +{"question": "Which papers discussed data pruning methods to accelerate model training?", "answer": ["Deep Learning on a Data Diet: Finding Important Examples Early in Training", "An Empirical Study of Example Forgetting during Deep Neural Network Learning", "Identifying Mislabeled Data using the Area Under the Margin Ranking"], "answer_arxiv_id": ["2107.07075", "1812.05159", "2001.10528"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_8806"} +{"question": "Which papers have found that neural networks in reinforcement learning often result in multiple failure modes?", "answer": ["Deep Reinforcement Learning and the Deadly Triad"], "answer_arxiv_id": ["1812.02648"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_8807"} +{"question": "Which paper suggested a concatenation scheme for vectors in a database lineage tracking?", "answer": ["Efficient Approximate Search for Sets of Vectors"], "answer_arxiv_id": ["2107.06817"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_8808"} +{"question": "What works presented the utilization of latent diffusion models like Stable Diffusion for various tasks?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_8809"} +{"question": "What study claims that, especially Codex davinci v2, models consistently outperform the text models in reasoning-intensive scenarios?", "answer": ["Holistic Evaluation of Language Models"], "answer_arxiv_id": ["2211.09110"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_8810"} +{"question": "Could you list some studies that apply attention mechanism to refine the visual features extracted from CNN backbone for ZSL?", "answer": ["Semantic-Guided Multi-Attention Localization for Zero-Shot Learning", "Attribute Prototype Network for Zero-Shot Learning", "Goal-Oriented Gaze Estimation for Zero-Shot Learning", "TransZero++: Cross Attribute-Guided Transformer for Zero-Shot Learning"], "answer_arxiv_id": ["1903.00502", "2008.08290", "2103.03433", "2112.08643"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_8811"} +{"question": "Which datasets contain the greatest number of frames collected in the mmWave radar frameworks?", "answer": ["mRI: Multi-modal 3D Human Pose Estimation Dataset using mmWave, RGB-D, and Inertial Sensors"], "answer_arxiv_id": ["2210.08394"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_8812"} +{"question": "Which papers discuss advancements made in pool-based active learning for deep neural networks?", "answer": ["A Survey of Deep Active Learning"], "answer_arxiv_id": ["2009.00236"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_8813"} +{"question": "What research has been done on reducibility of tasks to small subnetworks?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Piggyback: Adapting a Single Network to Multiple Tasks by Learning to\n Mask Weights", "Movement Pruning: Adaptive Sparsity by Fine-Tuning", "Disentangling Representations of Text by Masking Transformers", "Masking as an Efficient Alternative to Finetuning for Pretrained\n Language Models", "Are Neural Nets Modular? Inspecting Functional Modularity Through\n Differentiable Weight Masks", "Parameter-Efficient Transfer Learning with Diff Pruning", "Identifying and Adapting Transformer-Components Responsible for Gender\n Bias in an English Language Model"], "answer_arxiv_id": ["1803.03635", "1801.06519", "2005.07683", "2104.07155", "2004.12406", "2010.02066", "2012.07463", "2310.12611"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_8814"} +{"question": "Could you provide me some papers using low-rank priors for HSI reconstruction?", "answer": ["Rank Minimization for Snapshot Compressive Imaging"], "answer_arxiv_id": ["1807.07837"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_8815"} +{"question": "Which studies have discussed the use of Gumbel random variables in the context of the STGS estimator?", "answer": ["A^∗ Sampling", "Categorical Reparameterization with Gumbel-Softmax"], "answer_arxiv_id": ["1411.0030", "1611.01144"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_8816"} +{"question": "What papers are about watermark methods that maintain the semantic information of texts?", "answer": ["A Semantic Invariant Robust Watermark for Large Language Models", "SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text\n Generation"], "answer_arxiv_id": ["2310.06356", "2310.03991"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_8817"} +{"question": "Could you provide me some works that use opt-out values to address the non-correspondence issue in two-player zero-sum games?", "answer": ["DeepStack: Expert-Level Artificial Intelligence in Heads-Up No-Limit Poker"], "answer_arxiv_id": ["1701.01724"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_8818"} +{"question": "Which traditional domain adaptation methods jointly optimize on labeled source data and unlabeled target data?", "answer": ["Deep Domain Confusion: Maximizing for Domain Invariance", "Unsupervised Domain Adaptation by Backpropagation", "Learning Transferable Features with Deep Adaptation Networks", "Deep Transfer Learning with Joint Adaptation Networks", "Adversarial Discriminative Domain Adaptation", "Conditional Adversarial Domain Adaptation", "Alignment Attention by Matching Key and Query Distributions", "Learning with Different Amounts of Annotation: From Zero to Many Labels"], "answer_arxiv_id": ["1412.3474", "1409.7495", "1502.02791", "1605.06636", "1702.05464", "1705.10667", "2110.12567", "2109.04408"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_8819"} +{"question": "Which works demonstrated that a synthesized small dataset from a large target dataset was effective in various scenarios?", "answer": ["Data Distillation: A Survey"], "answer_arxiv_id": ["2301.04272"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_8820"} +{"question": "What papers have attempted to use calibration loss to balance the prediction results between seen and unseen classes in embedding-based Zero-Shot Learning?", "answer": ["TransZero: Attribute-guided Transformer for Zero-Shot Learning", "TransZero++: Cross Attribute-Guided Transformer for Zero-Shot Learning"], "answer_arxiv_id": ["2112.01683", "2112.08643"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_8821"} +{"question": "What studies use the coordinated models approach to building instruction-following agents?", "answer": ["Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "Generalized Decoding for Pixel, Image, and Language", "MM-ReAct : Prompting ChatGPT for Multimodal Reasoning and Action", "Visual Programming: Compositional visual reasoning without training", "ViperGPT: Visual Inference via Python Execution for Reasoning"], "answer_arxiv_id": ["2303.04671", "2212.11270", "2303.11381", "2211.11559", "2303.08128v1"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_8822"} +{"question": "Can you mention some studies that used masked signal modeling as a self-supervised task in computer vision?", "answer": ["ConvMAE: Masked Convolution Meets Masked Autoencoders", "Bootstrapped Masked Autoencoders for Vision BERT Pretraining", "BEiT: BERT Pre-Training of Image Transformers", "SimMIM: A Simple Framework for Masked Image Modeling", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "Masked Autoencoders Are Scalable Vision Learners", "Masked Autoencoders As Spatiotemporal Learners"], "answer_arxiv_id": ["2205.03892", "2207.07116", "2106.08254", "2111.09886", "2112.09133", "2111.06377", "2205.09113"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8823"} +{"question": "What papers discussed the use of self-distillation in masked image modeling methods?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Emerging Properties in Self-Supervised Vision Transformers", "An Empirical Study of Training Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2006.07733", "2104.14294", "2104.02057"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_8824"} +{"question": "Which papers have discussed dynamic sparse training?", "answer": ["Deep Rewiring: Training very sparse deep networks", "Scalable Training of Artificial Neural Networks with Adaptive Sparse\n Connectivity inspired by Network Science", "Sparse Networks from Scratch: Faster Training without Losing Performance", "Sparse evolutionary Deep Learning with over one million artificial\n neurons on commodity hardware", "Rigging the Lottery: Making All Tickets Winners", "Parameter Efficient Training of Deep Convolutional Neural Networks by\n Dynamic Sparse Reparameterization", "Top-KAST: Top-K Always Sparse Training", "Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Selfish Sparse RNN Training"], "answer_arxiv_id": ["1711.05136", "1707.04780", "1907.04840", "1901.09181", "1911.11134", "1902.05967", "2106.03517", "2106.04533", "2101.09048"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_8825"} +{"question": "Which work utilizes ConvGRU to construct decoders in a network and decodes correlation and context information for frame-based optical flow estimation?", "answer": ["RAFT: Recurrent All-Pairs Field Transforms for Optical Flow"], "answer_arxiv_id": ["2003.12039"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_8826"} +{"question": "Which works voiced concerns about data contamination in LLMs?", "answer": ["Pretraining on the Test Set Is All You Need", "Don't Make Your LLM an Evaluation Benchmark Cheater", "Proving Test Set Contamination in Black Box Language Models"], "answer_arxiv_id": ["2309.08632", "2311.01964", "2310.17623"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_8827"} +{"question": "What works used the REINFORCE-based method with a proper baseline for reducing variances in constructive DRL methods for TSP?", "answer": ["POMO: Policy Optimization with Multiple Optima for Reinforcement Learning", "Sym-NCO: Leveraging Symmetricity for Neural Combinatorial Optimization"], "answer_arxiv_id": ["2010.16011", "2205.13209"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_8828"} +{"question": "What studies utilize LLM for paraphrasing as a strategy for replacing a subset of words?", "answer": ["Paraphrasing evades detectors of AI-generated text, but retrieval is an\n effective defense", "Can AI-Generated Text be Reliably Detected?"], "answer_arxiv_id": ["2303.13408", "2303.11156"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_8829"} +{"question": "Could you provide me some studies about using fine-tuning in the healthcare and finance sectors?", "answer": ["GeneGPT: Augmenting Large Language Models with Domain Tools for Improved\n Access to Biomedical Information", "BloombergGPT: A Large Language Model for Finance"], "answer_arxiv_id": ["2304.09667", "2303.17564"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_8830"} +{"question": "What are the efforts to use the real-world 2D image prior from text-to-2D generative models to reconstruct 3D shape from a single image?", "answer": ["RealFusion: 360{\\deg} Reconstruction of Any Object from a Single Image", "NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as\n General Image Priors", "Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion\n Prior"], "answer_arxiv_id": ["2302.10663", "2212.03267", "2303.14184"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_8831"} +{"question": "What papers contributed to the understanding of the number of regions in a neural network?", "answer": ["On the Expressive Power of Deep Neural Networks", "On the Number of Linear Regions of Deep Neural Networks"], "answer_arxiv_id": ["1606.05336", "1402.1869"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_8832"} +{"question": "Can you name a study that employs set-abstraction to obtain query points and infer AABBs?", "answer": ["GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in\n Point Cloud"], "answer_arxiv_id": ["1812.03320"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_8833"} +{"question": "What papers present ensemble methods for embracing efficient exploration?", "answer": ["Deep Exploration via Bootstrapped DQN", "Randomized Ensembled Double Q-Learning: Learning Fast Without a Model", "SUNRISE: A Simple Unified Framework for Ensemble Learning in Deep Reinforcement Learning"], "answer_arxiv_id": ["1602.04621", "2101.05982", "2007.04938"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_8834"} +{"question": "Could you provide me some works that apply self-training methods in UDA for object detection?", "answer": ["Unbiased Mean Teacher for Cross-domain Object Detection", "Contrastive Mean Teacher for Domain Adaptive Object Detectors"], "answer_arxiv_id": ["2003.00707", "2305.03034"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_8835"} +{"question": "Which papers are about the development of algorithms for causal inference in non-linear additive models?", "answer": ["Gradient-Based Neural DAG Learning", "Score matching enables causal discovery of nonlinear additive noise models", "Causal Discovery with Score Matching on Additive Models with Arbitrary Noise", "Diffusion Models for Causal Discovery via Topological Ordering", "Scalable Causal Discovery with Score Matching"], "answer_arxiv_id": ["1906.02226", "2203.04413", "2304.03265", "2210.06201", "2304.03382v1"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_8836"} +{"question": "What is the original paper proposing the DDIM scheme for diffusion-based image editing?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_8837"} +{"question": "Which studies provide datasets that are captured from a specially designed dome or studio?", "answer": ["Hand Keypoint Detection in Single Images using Multiview Bootstrapping", "First-Person Hand Action Benchmark with RGB-D Videos and 3D Hand Pose Annotations", "Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "HUMBI: A Large Multiview Dataset of Human Body Expressions", "HUMBI: A Large Multiview Dataset of Human Body Expressions and Benchmark Challenge", "InterHand2.6M: A Dataset and Baseline for 3D Interacting Hand Pose Estimation from a Single RGB Image"], "answer_arxiv_id": ["1704.07809", "1704.02463", "1904.05866", "1812.00281", "2110.00119", "2008.09309"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_8838"} +{"question": "Which works proposed the extension of static KGs to TKGs?", "answer": ["Learning Sequence Encoders for Temporal Knowledge Graph Completion", "Temporal Knowledge Graph Embedding Model based on Additive Time Series\n Decomposition", "Tensor Decompositions for temporal knowledge base completion", "Temporal Knowledge Base Completion: New Algorithms and Evaluation\n Protocols"], "answer_arxiv_id": ["1809.03202", "1911.07893", "2004.04926", "2005.05035"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_8839"} +{"question": "Any works that add adversarial patches to physical-world scenes to mislead machine learning classification models?", "answer": ["Adversarial Patch", "Robust Physical-World Attacks on Deep Learning Visual Classification", "Synthesizing Robust Adversarial Examples", "Adversarial examples in the physical world"], "answer_arxiv_id": ["1712.09665", "1707.08945", "1707.07397", "1607.02533"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8840"} +{"question": "What references demonstrated that increasing depth in linear diagonal networks drives the network to sparser solutions?", "answer": ["Implicit Bias of Gradient Descent on Linear Convolutional Networks"], "answer_arxiv_id": ["1806.00468"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_8841"} +{"question": "What studies mentioned the issue of slow training and rendering speed in Neural radiance fields (NeRF)?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8842"} +{"question": "Which work uses contrastive loss to model pairwise similarities among samples, generates pseudo labels from the cross-entropy loss, and calibrates the prediction distribution of two branches?", "answer": ["Semi-supervised Contrastive Learning with Similarity Co-calibration"], "answer_arxiv_id": ["2105.07387"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_8843"} +{"question": "What papers studied a federated multi-armed bandit problem?", "answer": ["Federated Bandit: A Gossiping Approach", "Federated Multi-armed Bandits with Personalization"], "answer_arxiv_id": ["2010.12763", "2102.13101v1"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_8844"} +{"question": "What research has explored the role of prompt-engineering in relation to large language models?", "answer": ["Legal Prompting: Teaching a Language Model to Think Like a Lawyer"], "answer_arxiv_id": ["2212.01326"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_8845"} +{"question": "Which studies have researched on online learning equilibria of two-player zero-sum Markov games?", "answer": ["Online Reinforcement Learning in Stochastic Games", "Provable Self-Play Algorithms for Competitive Reinforcement Learning", "Near-Optimal Reinforcement Learning with Self-Play", "A Sharp Analysis of Model-based Reinforcement Learning with Self-Play", "Gap-Dependent Bounds for Two-Player Markov Games"], "answer_arxiv_id": ["1712.00579", "2002.04017", "2006.12007", "2010.01604", "2107.00685"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_8846"} +{"question": "Could you suggest a reference for VQA-v2 dataset on which Modular Co-Attention Network (MCAN) model was evaluated?", "answer": ["Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering"], "answer_arxiv_id": ["1612.00837"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_8847"} +{"question": "Which research covered Semantic Scene Completion using RGB images?", "answer": ["MonoScene: Monocular 3D Semantic Scene Completion", "OccDepth: A Depth-Aware Method for 3D Semantic Scene Completion", "VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene\n Completion"], "answer_arxiv_id": ["2112.00726", "2302.13540", "2302.12251"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8848"} +{"question": "Are there any papers that explored the creation of noise robust methods focusing on loss functions?", "answer": ["Generalized Cross Entropy Loss for Training Deep Neural Networks with\n Noisy Labels", "Symmetric Cross Entropy for Robust Learning with Noisy Labels"], "answer_arxiv_id": ["1805.07836", "1908.06112"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_8849"} +{"question": "Are there any works that compile the queries into the graphs and then solve queries with graph neural networks?", "answer": ["Message Passing Query Embedding", "Mask and Reason: Pre-Training Knowledge Graph Transformers for Complex Logical Queries"], "answer_arxiv_id": ["2002.02406", "2208.07638"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_8850"} +{"question": "Which works used diffusion-based generate models to achieve zero-shot editing?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "Blended Diffusion for Text-driven Editing of Natural Images", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "Prompt-to-Prompt Image Editing with Cross Attention Control"], "answer_arxiv_id": ["2108.02938", "2111.14818", "2201.09865", "2208.01626"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_8851"} +{"question": "Could you provide me some references that propose to improve the processes of shape prior deformation and correspondence matching in 6D object pose estimation based on the SPD?", "answer": ["Category-Level 6D Object Pose and Size Estimation using Self-Supervised\n Deep Prior Deformation Networks", "Category-Level 6D Object Pose Estimation via Cascaded Relation and\n Recurrent Reconstruction Networks", "RBP-Pose: Residual Bounding Box Projection for Category-Level Pose\n Estimation"], "answer_arxiv_id": ["2207.05444", "2108.08755", "2208.00237"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_8852"} +{"question": "What works extended the notion of regular convolutional layers embedding equivariance to translation to other groups by convolving over groups?", "answer": ["Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1602.07576"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_8853"} +{"question": "What studies examined the sample complexity of POMDPs by bounding the Bellman rank?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "Bilinear Classes: A Structural Framework for Provable Generalization in RL", "Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms"], "answer_arxiv_id": ["1610.09512v2", "2103.10897", "2102.00815"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_8854"} +{"question": "Could you provide examples of work on dynamic Vision Transformers (ViTs) aiming to eliminate non-essential tokens adaptively?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "A-ViT: Adaptive Tokens for Efficient Vision Transformer", "Token Merging: Your ViT But Faster", "MIA-Former: Efficient and Robust Vision Transformers via Multi-grained Input Adaptation"], "answer_arxiv_id": ["2106.02034", "2112.07658", "2210.09461", "2112.11542"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_8855"} +{"question": "What research papers can be cited as resources for methods estimating 2D/3D room layout from dense 3D point clouds or a single panoramic RGB image?", "answer": ["SceneCAD: Predicting Object Alignments and Layouts in RGB-D Scans", "PQ-Transformer: Jointly Parsing 3D Objects and Layouts from Point Clouds", "From Semi-supervised to Omni-supervised Room Layout Estimation Using Point Clouds", "DuLa-Net: A Dual-Projection Network for Estimating Room Layouts from a Single RGB Panorama", "HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch Data Augmentation", "Pano2CAD: Room Layout From A Single Panorama Image", "LayoutNet: Reconstructing the 3D Room Layout from a Single RGB Image"], "answer_arxiv_id": ["2003.12622", "2109.05566", "2301.13865", "1811.11977", "1901.03861", "1609.09270", "1803.08999"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_8856"} +{"question": "What works have identified unimportant weights via a certain importance criterion and then utilized a penalty term to produce sparsity in the field of network pruning?", "answer": ["Neural Pruning via Growing Regularization"], "answer_arxiv_id": ["2012.09243"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_8857"} +{"question": "Could you provide some studies where positional or structural encodings capturing spectral information were used?", "answer": ["Benchmarking Graph Neural Networks"], "answer_arxiv_id": ["2003.00982"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8858"} +{"question": "What studies explore training Language and Vision-Integrated Language Models (LVLMs) to optimize goals such as becoming helpful using reinforcement learning?", "answer": ["Can Neural Machine Translation be Improved with User Feedback?", "A General Language Assistant as a Laboratory for Alignment"], "answer_arxiv_id": ["1804.05958", "2112.00861"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_8859"} +{"question": "What studies focus on strategic learning in Stackelberg games?", "answer": ["Strategic Classification", "Strategic Classification from Revealed Preferences"], "answer_arxiv_id": ["1506.06980", "1710.07887"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_8860"} +{"question": "Could you provide me with some references where reconstruction learning is used in detecting image forgery?", "answer": ["Representative Forgery Mining for Fake Face Detection"], "answer_arxiv_id": ["2104.06609"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_8861"} +{"question": "Could you tell me who carried out the comprehensive analysis of the multi-modality problem in training data?", "answer": ["A Study of Syntactic Multi-Modality in Non-Autoregressive Machine Translation", "On the Learning of Non-Autoregressive Transformers"], "answer_arxiv_id": ["2207.04206", "2206.05975"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_8862"} +{"question": "Could you provide some works that train a diffusion model on a single image to produce its variations?", "answer": ["SinDiffusion: Learning a Diffusion Model from a Single Natural Image", "SinFusion: Training Diffusion Models on a Single Image or Video", "SinDDM: A Single Image Denoising Diffusion Model", "SinGAN: Learning a Generative Model from a Single Natural Image", "Drop the GAN: In Defense of Patches Nearest Neighbors as Single Image\n Generative Models"], "answer_arxiv_id": ["2211.12445", "2211.11743", "2211.16582", "1905.01164", "2103.15545"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_8863"} +{"question": "Which works focused on fitting a parametric 3D shape model to animal images using annotated 2D keypoints and segmentation masks?", "answer": ["3D Menagerie: Modeling the 3D shape and pose of animals"], "answer_arxiv_id": ["1611.07700"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_8864"} +{"question": "Which work has been followed for calibration of the canary losses?", "answer": ["Truth Serum: Poisoning Machine Learning Models to Reveal Their Secrets"], "answer_arxiv_id": ["2204.00032"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_8865"} +{"question": "Which works originally proposed e-values as quantitative measures of statistical evidence?", "answer": ["E-values: Calibration, combination, and applications"], "answer_arxiv_id": ["1912.06116"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_8866"} +{"question": "What works cover Markov decision process (MDP) in the context of an offline Reinforcement Learning (RL) environment?", "answer": ["D4RL: Datasets for Deep Data-Driven Reinforcement Learning"], "answer_arxiv_id": ["2004.07219"], "source_meta": {"published_time": "20220812"}, "qid": "AutoScholarQuery_train_8867"} +{"question": "What researches integrate self or weakly supervised pretraining for semantic search in classical models?", "answer": ["Latent Retrieval for Weakly Supervised Open Domain Question Answering", "Unsupervised Corpus Aware Language Model Pre-training for Dense Passage\n Retrieval", "Unsupervised Dense Information Retrieval with Contrastive Learning", "Text and Code Embeddings by Contrastive Pre-Training", "Text Embeddings by Weakly-Supervised Contrastive Pre-training", "Towards General Text Embeddings with Multi-stage Contrastive Learning", "C-Pack: Packaged Resources To Advance General Chinese Embedding"], "answer_arxiv_id": ["1906.00300", "2108.05540", "2112.09118", "2201.10005", "2212.03533", "2308.03281v1", "2309.07597"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_8868"} +{"question": "What study focus on single-task poisoning for small LMs?", "answer": ["Concealed Data Poisoning Attacks on NLP Models"], "answer_arxiv_id": ["2010.12563"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_8869"} +{"question": "What research provided an alternative to our equivariant flow and equivariant CNFs through the proposal of equivariant residual flows?", "answer": ["Equivariant Finite Normalizing Flows"], "answer_arxiv_id": ["2110.08649"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_8870"} +{"question": "Any works about employing clustering or random grouping to form pseudo-bags or labels for multiple instance learning (MIL)?", "answer": ["Multi-level Multiple Instance Learning with Transformer for Whole Slide\n Image Classification", "DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning\n for Histopathology Whole Slide Image Classification", "DGMIL: Distribution Guided Multiple Instance Learning for Whole Slide\n Image Classification", "ProtoDiv: Prototype-guided Division of Consistent Pseudo-bags for\n Whole-slide Image Classification"], "answer_arxiv_id": ["2306.05029", "2203.12081", "2206.08861", "2304.06652"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_8871"} +{"question": "Can you provide studies that explored in-context configurations in Natural Language Processing?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "How Can We Know What Language Models Know?", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?", "What Makes Good In-Context Examples for GPT-3?", "Making Pre-trained Language Models Better Few-shot Learners", "Learning To Retrieve Prompts for In-Context Learning", "Selective Annotation Makes Language Models Better Few-Shot Learners", "Reordering Examples Helps during Priming-based Few-Shot Learning", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2107.13586v1", "1911.12543", "2104.08691", "2202.12837", "2101.06804", "2012.15723", "2112.08633", "2209.01975", "2106.01751", "2205.10625", "2104.08786"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_8872"} +{"question": "Which papers proposed solutions for text-to-video (T2V) generation?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data"], "answer_arxiv_id": ["2210.02303", "2209.14792"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_8873"} +{"question": "Which papers have conducted research on label refinement in AVVP?", "answer": ["Joint-Modal Label Denoising for Weakly-Supervised Audio-Visual Video Parsing"], "answer_arxiv_id": ["2204.11573"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8874"} +{"question": "What work proposed the use of spectral bandpass filters with structured light or a light field camera as an alternative to CASSI for hyperspectral 3D imaging?", "answer": ["Single-shot Hyperspectral-Depth Imaging with Learned Diffractive Optics", "Hyperspectral Light Field Stereo Matching"], "answer_arxiv_id": ["2009.00463", "1709.00835"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8875"} +{"question": "Could you provide me some studies about new VQIM techniques developed from the perspective of codebook update, quantization, or regularization?", "answer": ["Regularized Vector Quantization for Tokenized Image Synthesis", "SQ-VAE: Variational Bayes on Discrete Representation with Self-annealed\n Stochastic Quantization", "Online Clustered Codebook"], "answer_arxiv_id": ["2303.06424", "2205.07547", "2307.15139"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_8876"} +{"question": "Who developed a unified analysis for the convergence of OGDA and EG methods in nonconvex-strongly-concave and nonconvex-concave settings?", "answer": ["Tight Analysis of Extra-gradient and Optimistic Gradient Methods For Nonconvex Minimax Problems"], "answer_arxiv_id": ["2210.09382v1"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_8877"} +{"question": "Any studies showed that the matrix mechanism achieves state-of-the-art results in Differential Privacy Machine Learning?", "answer": ["Multi-Epoch Matrix Factorization Mechanisms for Private Machine Learning"], "answer_arxiv_id": ["2211.06530"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_8878"} +{"question": "What is the work that demonstrates variational autoencoders' ability to learn data manifolds with complex topologies?", "answer": ["Auto-Encoding Variational Bayes", "Topological Autoencoders"], "answer_arxiv_id": ["1312.6114", "1906.00722"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_8879"} +{"question": "Which paper proposed a learnable poisoning sample selection strategy in backdoor attacks?", "answer": ["Boosting Backdoor Attack with A Learnable Poisoning Sample Selection Strategy"], "answer_arxiv_id": ["2307.07328"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_8880"} +{"question": "What studies are about prompt-patching which focuses on identifying mistakes in the predictions generated from a particular prompt?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2107.13586v1"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_8881"} +{"question": "What studies have enhanced outdoor distant low-overlap registration in fulloy convolutional methods?", "answer": ["APR: Online Distant Point Cloud Registration Through Aggregated Point\n Cloud Reconstruction", "Density-invariant Features for Distant Point Cloud Registration"], "answer_arxiv_id": ["2305.02893", "2307.09788"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_8882"} +{"question": "Could you provide me some research that proposed solutions for the missing modality problem without needing data imputation?", "answer": ["Deep Partial Multi-View Learning", "SMIL: Multimodal Learning with Severely Missing Modality", "Are Multimodal Transformers Robust to Missing Modality?"], "answer_arxiv_id": ["2011.06170", "2103.05677", "2204.05454"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_8883"} +{"question": "Can you name some studies that focus on measuring LLMs’ understanding and mastery of world knowledge?", "answer": ["CREAK: A Dataset for Commonsense Reasoning over Entity Knowledge", "When Not to Trust Language Models: Investigating Effectiveness of\n Parametric and Non-Parametric Memories", "The KITMUS Test: Evaluating Knowledge Integration from Multiple Sources in Natural Language Understanding Systems", "KoLA: Carefully Benchmarking World Knowledge of Large Language Models"], "answer_arxiv_id": ["2109.01653", "2212.10511", "2212.08192v2", "2306.09296v3"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_8884"} +{"question": "What works introduced or developed the high-quality AIST++ database?", "answer": ["AI Choreographer: Music Conditioned 3D Dance Generation with AIST++"], "answer_arxiv_id": ["2101.08779"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_8885"} +{"question": "Could you tell me the reference that created the RF100 (Roboflow-100) dataset?", "answer": ["Roboflow 100: A Rich, Multi-Domain Object Detection Benchmark"], "answer_arxiv_id": ["2211.13523"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_8886"} +{"question": "Which studies have extended the lottery ticket hypothesis to different types of neural networks?", "answer": ["GANs Can Play Lottery Tickets Too", "Data-Efficient GAN Training Beyond (Just) Augmentations: A Lottery Ticket Perspective", "Winning Lottery Tickets in Deep Generative Models", "Successfully Applying the Stabilized Lottery Ticket Hypothesis to the Transformer Architecture", "When BERT Plays the Lottery, All Tickets Are Winning", "The Lottery Ticket Hypothesis for Pre-trained BERT Networks", "On Lottery Tickets and Minimal Task Representations in Deep Reinforcement Learning", "A Unified Lottery Ticket Hypothesis for Graph Neural Networks"], "answer_arxiv_id": ["2106.00134", "2103.00397", "2010.02350", "2005.03454", "2005.00561", "2007.12223", "2105.01648", "2102.06790"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_8887"} +{"question": "What research enables separate editing of video layers based on textual descriptions?", "answer": ["Text2LIVE: Text-Driven Layered Image and Video Editing"], "answer_arxiv_id": ["2204.02491"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_8888"} +{"question": "What works provided assessments on the impact of imputation algorithms on fairness risks?", "answer": ["FairPrep: Promoting Data to a First-Class Citizen in Studies on Fairness-Enhancing Interventions", "Impact Of Missing Data Imputation On The Fairness And Accuracy Of Graph Node Classifiers", "Fairness without Imputation: A Decision Tree Approach for Fair Prediction with Missing Values", "Fairness in Missing Data Imputation"], "answer_arxiv_id": ["1911.12587", "2211.00783", "2109.10431", "2110.12002"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8889"} +{"question": "Which studies contributed to improving self-supervision in MDE with the introduction of new loss terms?", "answer": ["Unsupervised Monocular Depth Estimation with Left-Right Consistency", "Unsupervised Scale-consistent Depth and Ego-motion Learning from Monocular Video", "Learning Depth from Monocular Videos using Direct Methods", "GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose", "Single Image Depth Prediction with Wavelet Decomposition", "D3VO: Deep Depth, Deep Pose and Deep Uncertainty for Monocular Visual Odometry"], "answer_arxiv_id": ["1609.03677", "1908.10553", "1712.00175v1", "1803.02276", "2106.02022", "2003.01060"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_8890"} +{"question": "What studies suggest a method for determining the optimal binning scheme and hyperparameters?", "answer": ["Mitigating Bias in Calibration Error Estimation", "Distribution-free calibration guarantees for histogram binning without sample splitting"], "answer_arxiv_id": ["2012.08668", "2105.04656"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_8891"} +{"question": "Which researches tried to use language models for code generation?", "answer": ["Competition-Level Code Generation with AlphaCode", "Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2203.07814", "2107.03374"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_8892"} +{"question": "What research papers describe datasets that focus on providing sequential images with camera and object pose annotations?", "answer": ["I Like to Move It: 6D Pose Estimation as an Action Decision Process", "A Framework for Evaluating 6-DOF Object Trackers", "DeepIM: Deep Iterative Matching for 6D Pose Estimation"], "answer_arxiv_id": ["2009.12678", "1803.10075", "1804.00175"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_8893"} +{"question": "Which papers elaborate on methods learning neural network policies that can be quickly fine-tuned to new tasks at test time?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Recasting Gradient-Based Meta-Learning as Hierarchical Bayes", "Gotta Learn Fast: A New Benchmark for Generalization in RL", "ProMP: Proximal Meta-Policy Search", "Model-Based Reinforcement Learning via Meta-Policy Optimization"], "answer_arxiv_id": ["1703.03400", "1801.08930", "1804.03720", "1810.06784", "1809.05214"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_8894"} +{"question": "Which works discuss denoising diffusion models?", "answer": ["Q"], "answer_arxiv_id": ["1611.08152"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_8895"} +{"question": "Are there any studies about processing diverse forms of input, not only text but also visual and other sensory data, in the context of LLMs?", "answer": ["Visual Instruction Tuning", "Ferret: Refer and Ground Anything Anywhere at Any Granularity", "Chameleon: Plug-and-Play Compositional Reasoning with Large Language\n Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2304.08485", "2310.07704", "2304.09842", "2304.14178", "2304.15010", "2305.03726"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_8896"} +{"question": "What studies proposed reparameterization-based methods that transform the optimization process of trainable parameters into a low-dimensional subspace?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "QLoRA: Efficient Finetuning of Quantized LLMs"], "answer_arxiv_id": ["2106.09685", "2305.14314"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_8897"} +{"question": "Are there any studies that have conducted research in the infinite-width limit?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "On Lazy Training in Differentiable Programming", "Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Data", "On the Global Convergence of Gradient Descent for Over-parameterized Models using Optimal Transport", "A Mean Field View of the Landscape of Two-Layer Neural Networks", "A Rigorous Framework for the Mean Field Limit of Multilayer Neural Networks", "Modeling from Features: a Mean-field Framework for Over-parameterized Deep Neural Networks"], "answer_arxiv_id": ["1806.07572", "1812.07956", "1811.03804", "1811.03962", "1901.08584", "1808.01204", "1805.09545", "1804.06561", "2001.11443v3", "2007.01452"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8898"} +{"question": "What works consider matching-based methods that compute unit (dis)similarity based on estimated propensity scores or use adaptive similarity measures?", "answer": ["Estimation and Inference of Heterogeneous Treatment Effects using Random Forests"], "answer_arxiv_id": ["1510.04342"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_8899"} +{"question": "Which study flags the issue of possible aliasing errors with specific operator learning architectures?", "answer": ["Spectral Neural Operators"], "answer_arxiv_id": ["2205.10573v2"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_8900"} +{"question": "What are some examples of end-to-end trainable multimodal Large Language Models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models", "Language Is Not All You Need: Aligning Perception with Language Models", "PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2302.14045", "2303.03378v1"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_8901"} +{"question": "Which works have explored editing, personalization and inversion to token space in diffusion models?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "Imagic: Text-Based Real Image Editing with Diffusion Models", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Null-text Inversion for Editing Real Images using Guided Diffusion Models"], "answer_arxiv_id": ["2208.01618", "2208.12242", "2210.09276", "2211.09800", "2208.01626", "2211.09794"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_8902"} +{"question": "What works focused on OV-Det by knowledge distillation from VLM and using handcrafted prompt or focus on prompt representation learning for object regions?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "Open-Vocabulary One-Stage Detection with Hierarchical Visual-Language Knowledge Distillation", "Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model", "PromptDet: Towards Open-vocabulary Detection using Uncurated Images"], "answer_arxiv_id": ["2104.13921", "2203.10593", "2203.14940", "2203.16513"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_8903"} +{"question": "Which studies focus on aligning a language model with another multilingual language model for multilingual tasks?", "answer": ["LLM Augmented LLMs: Expanding Capabilities through Composition", "LangBridge: Multilingual Reasoning Without Multilingual Supervision"], "answer_arxiv_id": ["2401.02412", "2401.10695"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_8904"} +{"question": "Could you provide me some studies that propose supervised methods in machine-generated text detection?", "answer": ["How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation,\n and Detection", "RADAR: Robust AI-Text Detection via Adversarial Learning", "Fine-tuning Large Language Models for Multigenerator, Multidomain, and\n Multilingual Machine-Generated Text Detection"], "answer_arxiv_id": ["2301.07597", "2307.03838", "2401.12326"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_8905"} +{"question": "What studies focus on feature-based explanations for explainability of neural models?", "answer": ["A Unified Approach to Interpreting Model Predictions", "Learning Important Features Through Propagating Activation Differences", "Learning from the Best: Rationalizing Prediction by Adversarial\n Information Calibration"], "answer_arxiv_id": ["1705.07874", "1704.02685", "2012.08884"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_8906"} +{"question": "Could you provide me some works that have explored generative models to generate new data for model training?", "answer": ["Synthetic Data from Diffusion Models Improves ImageNet Classification", "Random Multi-Channel Image Synthesis for Multiplexed Immunofluorescence Imaging", "CompoDiff: Versatile Composed Image Retrieval With Latent Diffusion", "Active Generative Adversarial Network for Image Classification", "Unsupervised Cardiac Segmentation Utilizing Synthesized Images from Anatomical Labels", "DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic Segmentation Using Diffusion Models"], "answer_arxiv_id": ["2304.08466", "2109.09004", "2303.11916", "1906.07133", "2301.06043", "2303.11681"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_8907"} +{"question": "Which work established the ShapeNet dataset for predicting and representing 3D shapes?", "answer": ["ShapeNet: An Information-Rich 3D Model Repository"], "answer_arxiv_id": ["1512.03012"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_8908"} +{"question": "What work uses a diffusion model to generate data for a downstream unsupervised domain adaptation (UDA) architecture?", "answer": ["One-shot Unsupervised Domain Adaptation with Personalized Diffusion\n Models"], "answer_arxiv_id": ["2303.18080"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_8909"} +{"question": "Which papers discussed 'unawareness' as a concept of fairness in machine learning?", "answer": ["Loss Balancing for Fair Supervised Learning", "Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatment", "Fair Sequential Selection Using Supervised Learning Models", "A Reductions Approach to Fair Classification", "Fairness and Accuracy under Domain Generalization"], "answer_arxiv_id": ["2311.03714", "1610.08452", "2110.13986", "1803.02453", "2301.13323"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_8910"} +{"question": "Any researches about improving generalization by conditioning agents on a goal image?", "answer": ["Goal-conditioned Imitation Learning", "Learning to Reach Goals via Iterated Supervised Learning"], "answer_arxiv_id": ["1906.05838", "1912.06088"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_8911"} +{"question": "What research papers focus on handover by considering grasping and dynamic motion planning?", "answer": ["Flexible Handover with Real-Time Robust Dynamic Grasp Trajectory\n Generation", "Reactive Human-to-Robot Handovers of Arbitrary Objects", "AnyGrasp: Robust and Efficient Grasp Perception in Spatial and Temporal\n Domains"], "answer_arxiv_id": ["2308.15622", "2011.08961", "2212.08333"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_8912"} +{"question": "What methods have been used to optimistically estimate the Conditional Value at Risk (CVaR) value of a policy’s return?", "answer": ["Being Optimistic to Be Conservative: Quickly Learning a CVaR Policy"], "answer_arxiv_id": ["1911.01546"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_8913"} +{"question": "What works have used mutual-information-based decoding techniques to steer the generation of Language Models?", "answer": ["Mutual Information Alleviates Hallucinations in Abstractive\n Summarization"], "answer_arxiv_id": ["2210.13210"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_8914"} +{"question": "Which works have developed initial results to characterize the optimization and generalization dynamics of attention?", "answer": ["Vision Transformers provably learn spatial structure", "A Theoretical Understanding of Shallow Vision Transformers: Learning, Generalization, and Sample Complexity"], "answer_arxiv_id": ["2210.09221", "2302.06015v3"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_8915"} +{"question": "In which papers did researchers study neuro-symbolic approaches that have attracted increasing interest in the AI and NLP communities?", "answer": ["Categorical Reparameterization with Gumbel-Softmax", "Rationalizing Neural Predictions", "Jumper: Learning When to Make Classification Decisions in Reading", "Object-oriented Neural Programming (OONP) for Document Understanding", "Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervision", "Coupling Distributed and Symbolic Execution for Natural Language Queries", "DeepPath: A Reinforcement Learning Method for Knowledge Graph Reasoning", "An Imitation Learning Approach to Unsupervised Parsing"], "answer_arxiv_id": ["1611.01144", "1606.04155", "1807.02314", "1709.08853v6", "1611.00020", "1612.02741", "1707.06690", "1906.02276"], "source_meta": {"published_time": "20210918"}, "qid": "AutoScholarQuery_train_8916"} +{"question": "Which papers contributed to the generation of consistent multi-view images given camera poses and text prompts by using the pre-trained perspective-image Diffusion Models?", "answer": ["Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models"], "answer_arxiv_id": ["2303.11989"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_8917"} +{"question": "Which studies reflect the two main determinants of asymptotic performance in spectral algorithms?", "answer": ["Optimal rates for the regularized learning algorithms under general source condition", "Optimal Rates For Regularization Of Statistical Inverse Learning Problems", "Optimal Rates for Spectral Algorithms with Least-Squares Regression over Hilbert Spaces"], "answer_arxiv_id": ["1611.01900", "1604.04054", "1801.06720"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_8918"} +{"question": "What research was about Life-Long Learning?", "answer": ["Continual Lifelong Learning with Neural Networks: A Review"], "answer_arxiv_id": ["1802.07569"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_8919"} +{"question": "Could you provide me studies about molecular generation using SMILES representations and techniques from natural language processing (NLP)?", "answer": ["Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules", "LIMO: Latent Inceptionism for Targeted Molecule Generation", "Objective-Reinforced Generative Adversarial Networks (ORGAN) for Sequence Generation Models", "BOSS: Bayesian Optimization over String Spaces"], "answer_arxiv_id": ["1610.02415", "2206.09010", "1705.10843", "2010.00979"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_8920"} +{"question": "What studies implement a learnable neural network model to automate the selection of augmentation?", "answer": ["Graph Contrastive Learning Automated"], "answer_arxiv_id": ["2106.07594"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_8921"} +{"question": "What studies are about implementation of pessimism in reinforcement learning problems?", "answer": ["Settling the Sample Complexity of Model-Based Offline Reinforcement Learning", "Is Pessimism Provably Efficient for Offline RL?", "Pessimistic Q-Learning for Offline Reinforcement Learning: Towards Optimal Sample Complexity"], "answer_arxiv_id": ["2204.05275", "2012.15085", "2202.13890"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_8922"} +{"question": "Which works proposed the Constrained beam search (CBS) algorithm?", "answer": ["Guided Open Vocabulary Image Captioning with Constrained Beam Search"], "answer_arxiv_id": ["1612.00576"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_8923"} +{"question": "Which papers explore one-stage methods for temporal action detection?", "answer": ["ActionFormer: Localizing Moments of Actions with Transformers", "TriDet: Temporal Action Detection with Relative Boundary Modeling", "Revisiting Anchor Mechanisms for Temporal Action Localization", "BasicTAD: an Astounding RGB-Only Baseline for Temporal Action Detection", "Action Sensitivity Learning for Temporal Action Localization"], "answer_arxiv_id": ["2202.07925", "2303.07347v2", "2008.09837", "2205.02717", "2305.15701"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_8924"} +{"question": "What papers address the novel view synthesis scenarios with more complex camera trajectories?", "answer": ["GAUDI: A Neural Architect for Immersive 3D Scene Generation", "MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion", "NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion\n Models", "Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models", "SceneScape: Text-Driven Consistent Scene Generation", "Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision", "Consistent View Synthesis with Pose-Guided Diffusion Models", "Long-Term Photometric Consistent Novel View Synthesis with Diffusion\n Models"], "answer_arxiv_id": ["2207.13751", "2307.01097", "2304.09787", "2303.11989", "2302.01133", "2306.11719", "2303.17598", "2304.10700"], "source_meta": {"published_time": "20240626"}, "qid": "AutoScholarQuery_train_8925"} +{"question": "What research papers have proposed low-rank methods for reducing spatial and temporal complexities in neural networks?", "answer": ["Tensor Networks Meet Neural Networks: A Survey and Future Perspectives", "TedNet: A Pytorch Toolkit for Tensor Decomposition Networks", "DQ-LoRe: Dual Queries with Low Rank Approximation Re-ranking for In-Context Learning", "LoRA: Low-Rank Adaptation of Large Language Models", "FedPETuning: When Federated Learning Meets the Parameter-Efficient Tuning Methods of Pre-trained Language Models", "FedPara: Low-rank Hadamard Product for Communication-Efficient Federated Learning", "Compressing Recurrent Neural Networks with Tensor Ring for Action Recognition", "Heuristic Rank Selection with Progressively Searching Tensor Ring Network", "T-Net: Parametrizing Fully Convolutional Nets with a Single High-Order Tensor", "Block-term Tensor Neural Networks", "A Tensorized Transformer for Language Modeling", "Towards Efficient Tensor Decomposition-Based DNN Model Compression with Optimization Framework"], "answer_arxiv_id": ["2302.09019", "2104.05018", "2310.02954", "2106.09685", "2212.10025", "2108.06098", "1811.07503", "2009.10580", "1904.02698", "2010.04963", "1906.09777", "2107.12422"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_8926"} +{"question": "Which works discuss the usage of machine learning and computer vision techniques in animal behavior analysis?", "answer": ["Task Programming: Learning Data Efficient Behavior Representations"], "answer_arxiv_id": ["2011.13917"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_8927"} +{"question": "Which works have studied the application of VLMs in open-vocabulary object detection?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation", "Simple Open-Vocabulary Object Detection with Vision Transformers"], "answer_arxiv_id": ["2104.13921", "2205.06230"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_8928"} +{"question": "Which method restricts the multiparameter persistence module to certain families of one-parameter lines and leverages available vectorizations for one-parameter barcodes?", "answer": ["A Kernel for Multi-Parameter Persistent Homology", "Multiparameter Persistence Landscapes"], "answer_arxiv_id": ["1809.10231", "1812.09935"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_8929"} +{"question": "What works are related to consistency regularization methods in SSL, focusing on minimizing the distance among different perturbed outputs?", "answer": ["Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data", "Contrastive Regularization for Semi-Supervised Learning"], "answer_arxiv_id": ["1703.01780", "2010.03622", "2201.06247"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_8930"} +{"question": "Which work employes a cGAN to predict RIRs from mesh room structures quickly?", "answer": ["MESH2IR: Neural Acoustic Impulse Response Generator for Complex 3D\n Scenes"], "answer_arxiv_id": ["2205.09248"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_8931"} +{"question": "Can you name a study that proposed the transfer of distilled knowledge from a ranker to a retriever?", "answer": ["Metric-guided Distillation: Distilling Knowledge from the Metric to\n Ranker and Retriever for Generative Commonsense Reasoning"], "answer_arxiv_id": ["2210.11708"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_8932"} +{"question": "In which papers the researchers study on TF-IDF and BM25 based traditional document retrieval techniques?", "answer": ["Reading Wikipedia to Answer Open-Domain Questions"], "answer_arxiv_id": ["1704.00051"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_8933"} +{"question": "Which work demonstrates the close performance of the ANN2SNN method to ANNs on the ImageNet dataset?", "answer": ["ImageNet Large Scale Visual Recognition Challenge"], "answer_arxiv_id": ["1409.0575"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_8934"} +{"question": "What are some studies that focus on social simulations?", "answer": ["Generative Agents: Interactive Simulacra of Human Behavior", "Welfare Diplomacy: Benchmarking Language Model Cooperation", "SOTOPIA: Interactive Evaluation for Social Intelligence in Language\n Agents"], "answer_arxiv_id": ["2304.03442", "2310.08901", "2310.11667"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_8935"} +{"question": "Which studies adjusted the denoising trajectory of DPMs using classifier guidance?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_8936"} +{"question": "What works employed meta-learning to train more robust models in the field of DGSS?", "answer": ["Pin the Memory: Learning to Generalize Semantic Segmentation"], "answer_arxiv_id": ["2204.03609"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8937"} +{"question": "Could you mention some one-stage object detectors?", "answer": ["SSD: Single Shot MultiBox Detector", "YOLOv3: An Incremental Improvement", "YOLOv4: Optimal Speed and Accuracy of Object Detection"], "answer_arxiv_id": ["1512.02325", "1804.02767", "2004.10934"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_8938"} +{"question": "What are some works that indicate that nonlinear ICA and disentangled representation learning are provably underspecified?", "answer": ["Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations"], "answer_arxiv_id": ["1811.12359"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_8939"} +{"question": "Could you inform me about the work that proposed a prediction model similar to ours?", "answer": ["Learning Augmented Energy Minimization via Speed Scaling"], "answer_arxiv_id": ["2010.11629"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_8940"} +{"question": "What works introduced point-voxel correlation fields to capture short-range and long-range movements?", "answer": ["PV-RAFT: Point-Voxel Correlation Fields for Scene Flow Estimation of Point Clouds"], "answer_arxiv_id": ["2012.00987"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_8941"} +{"question": "What work proposed the use of Normalized Object Coordinate Space (NOCS) for category-level pose estimation?", "answer": ["Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimation"], "answer_arxiv_id": ["1901.02970"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_8942"} +{"question": "Could you provide me some studies about utilizing sparsity-invariant convolution that keeps track of validation masks at each layer?", "answer": ["Sparsity Invariant CNNs"], "answer_arxiv_id": ["1708.06500"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_8943"} +{"question": "What works learn UDF from multi-view images in the context of new view synthesis?", "answer": ["NeUDF: Leaning Neural Unsigned Distance Fields with Volume Rendering", "NeuralUDF: Learning Unsigned Distance Fields for Multi-view\n Reconstruction of Surfaces with Arbitrary Topologies"], "answer_arxiv_id": ["2304.10080", "2211.14173"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_8944"} +{"question": "Which work discovered that the after kernels of neural networks trained with larger learning rates generalize better and stay more stable?", "answer": ["Properties of the After Kernel"], "answer_arxiv_id": ["2105.10585"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_8945"} +{"question": "What studies have been conducted on the synthetic setup of Dyck focusing on its solution space in contrast to natural languages?", "answer": ["Self-Attention Networks Can Process Bounded Hierarchical Languages", "Do Transformers Parse while Predicting the Masked Word?"], "answer_arxiv_id": ["2105.11115", "2303.08117"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_8946"} +{"question": "Which papers are about successful unified vision-language models (VLMs)?", "answer": ["Unifying Vision-and-Language Tasks via Text Generation", "Multimodal Few-Shot Learning with Frozen Language Models", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2102.02779", "2106.13884", "2202.03052", "2204.14198"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_8947"} +{"question": "Which studies proposed a differential private version of sliced Wasserstein distance?", "answer": ["Differentially Private Sliced Wasserstein Distance"], "answer_arxiv_id": ["2107.01848"], "source_meta": {"published_time": "20220927"}, "qid": "AutoScholarQuery_train_8948"} +{"question": "Which research demonstrated the performance degradation issue under detected boxes?", "answer": ["Towards Robust and Expressive Whole-body Human Pose and Shape Estimation"], "answer_arxiv_id": ["2312.08730"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_8949"} +{"question": "Which works explored the convergence properties of time-varying graphs in distributed optimization?", "answer": ["Decentralized Stochastic Optimization and Gossip Algorithms with Compressed Communication", "Distributed optimization over time-varying directed graphs", "Network Topology and Communication-Computation Tradeoffs in Decentralized Optimization"], "answer_arxiv_id": ["1902.00340", "1303.2289", "1709.08765"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_8950"} +{"question": "Which papers discuess the develeopment of prompt tuning methods for Vision foundation models?", "answer": ["Visual-Language Prompt Tuning with Knowledge-guided Context Optimization", "Prompt, Generate, then Cache: Cascade of Foundation Models makes Strong\n Few-shot Learners", "Exploring Visual Prompts for Adapting Large-Scale Models"], "answer_arxiv_id": ["2303.13283", "2303.02151", "2203.17274"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_8951"} +{"question": "Which studies discuss using a signed distance function for fast rendering?", "answer": ["BakedSDF: Meshing Neural SDFs for Real-Time View Synthesis", "MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction", "Geo-Neus: Geometry-Consistent Neural Implicit Surfaces Learning for Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces"], "answer_arxiv_id": ["2302.14859", "2208.00277", "2106.10689", "2104.10078", "2205.15848", "2106.12052"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_8952"} +{"question": "What works have empirically examined the relational understanding of DALL-E 2, particularly its ability to capture spatial relations?", "answer": ["Testing Relational Understanding in Text-Guided Image Generation"], "answer_arxiv_id": ["2208.00005"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_8953"} +{"question": "Could you provide me some papers that proposed practical tricks to improve reproducibility in machine learning?", "answer": ["Anti-Distillation: Improving Reproducibility of Deep Networks", "The Numerics of GANs", "Large Scale GAN Training for High Fidelity Natural Image Synthesis"], "answer_arxiv_id": ["2010.09923", "1705.10461", "1809.11096"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_8954"} +{"question": "What research has been done in generating Image Captioning?", "answer": ["Show and Tell: A Neural Image Caption Generator", "Probabilistic Embeddings for Cross-Modal Retrieval", "SmallCap: Lightweight Image Captioning Prompted with Retrieval Augmentation", "Retrieval-Augmented Transformer for Image Captioning"], "answer_arxiv_id": ["1411.4555", "2101.05068", "2209.15323", "2207.13162"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_8955"} +{"question": "Which works have shown the benefits of transfer learning to low-resource language-pairs in the NMT?", "answer": ["Transfer Learning for Low-Resource Neural Machine Translation", "Trivial Transfer Learning for Low-Resource Neural Machine Translation"], "answer_arxiv_id": ["1604.02201", "1809.00357"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_8956"} +{"question": "What studies have focused on attention modules for enhancing point features in point cloud semantic segmentation?", "answer": ["RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds", "PCT: Point Cloud Transformer", "Point Transformer", "Point Transformer V2: Grouped Vector Attention and Partition-based Pooling", "Stratified Transformer for 3D Point Cloud Segmentation", "Collect-and-Distribute Transformer for 3D Point Cloud Analysis"], "answer_arxiv_id": ["1911.11236", "2012.09688", "2012.09164", "2210.05666", "2203.14508", "2306.01257"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8957"} +{"question": "Are there any works that used Transformer-based dialogue modeling methods for response retrieval tasks?", "answer": ["The Ubuntu Dialogue Corpus: A Large Dataset for Research in Unstructured Multi-Turn Dialogue Systems"], "answer_arxiv_id": ["1506.08909"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_8958"} +{"question": "What work proposed the generalized rank invariant landscape and how did it process the invariant?", "answer": ["GRIL: A 2-parameter Persistence Based Vectorization for Machine Learning"], "answer_arxiv_id": ["2304.04970"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_8959"} +{"question": "What papers studied making a unified loss by using a linear combination of the main and auxiliary losses?", "answer": ["S4L: Self-Supervised Semi-Supervised Learning", "Entire Space Multi-Task Modeling via Post-Click Behavior Decomposition for Conversion Rate Prediction"], "answer_arxiv_id": ["1905.03670", "1910.07099"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_8960"} +{"question": "In which studies do researchers learn latent 3D representations to help synthesize views across different scenes?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "IBRNet: Learning Multi-View Image-Based Rendering", "Unsupervised Learning of 3D Object Categories from Videos in the Wild", "Neural Rays for Occlusion-aware Image-based Rendering"], "answer_arxiv_id": ["2012.02190", "2103.15595", "2102.13090", "2103.16552", "2107.13421"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_8961"} +{"question": "What papers discuss the use of the CPS framework in semi-supervised semantic segmentation?", "answer": ["Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision", "Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic\n Segmentation", "Conflict-Based Cross-View Consistency for Semi-Supervised Semantic\n Segmentation"], "answer_arxiv_id": ["2106.01226", "2208.09910", "2303.01276"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_8962"} +{"question": "Any papers where the document identifier used is a number?", "answer": ["Transformer Memory as a Differentiable Search Index", "A Neural Corpus Indexer for Document Retrieval", "Bridging the Gap Between Indexing and Retrieval for Differentiable\n Search Index with Query Generation", "Ultron: An Ultimate Retriever on Corpus with a Model-based Indexer"], "answer_arxiv_id": ["2202.06991", "2206.02743", "2206.10128", "2208.09257"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_8963"} +{"question": "Which works introduced non-generative methods for explaining Graph Neural Networks using gradients?", "answer": ["Explainability Techniques for Graph Convolutional Networks"], "answer_arxiv_id": ["1905.13686"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_8964"} +{"question": "Can you give some examples of the application of PAC-Bayes theory in supervised and reinforcement learning studies?", "answer": ["Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data", "A PAC-Bayesian Approach to Spectrally-Normalized Margin Bounds for Neural Networks", "Exploring Generalization in Deep Learning", "Spectrally-normalized margin bounds for neural networks", "Stronger generalization bounds for deep nets via a compression approach", "PAC-Bayes with Backprop", "Tighter Risk Certificates for Neural Networks", "Fantastic Generalization Measures and Where to Find Them", "PAC-Bayes Compression Bounds So Tight That They Can Explain Generalization", "PAC-Bayesian Policy Evaluation for Reinforcement Learning", "PAC-Bayes Control: Learning Policies that Provably Generalize to Novel Environments", "Probably Approximately Correct Vision-Based Planning using Motion Primitives", "Generalization Guarantees for Imitation Learning"], "answer_arxiv_id": ["1703.11008", "1707.09564", "1706.08947", "1706.08498", "1802.05296", "1908.07380", "2007.12911", "1912.02178", "2211.13609", "1202.3717v1", "1806.04225", "2002.12852", "2008.01913"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_8965"} +{"question": "What is the common approach for distributional Offline Policy Evaluation (OPE)?", "answer": ["Distributional Reinforcement Learning with Quantile Regression"], "answer_arxiv_id": ["1710.10044"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_8966"} +{"question": "Are there any studies about using gradient clipping to handle heavy-tailed noise in stochastic gradients?", "answer": ["Why are Adaptive Methods Good for Attention Models?"], "answer_arxiv_id": ["1912.03194"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_8967"} +{"question": "Which study proposes a linear lighting model for face relighting that eliminates the need of teacher-student distillation?", "answer": ["Towards Practical Capture of High-Fidelity Relightable Avatars"], "answer_arxiv_id": ["2309.04247"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_8968"} +{"question": "Which studies have proposed other DP fair learning algorithms after the initiation work?", "answer": ["Fair Learning with Private Demographic Data", "Differentially Private and Fair Deep Learning: A Lagrangian Dual Approach", "Differentially Private Empirical Risk Minimization under the Fairness Lens", "SF-PATE: Scalable, Fair, and Private Aggregation of Teacher Ensembles"], "answer_arxiv_id": ["2002.11651", "2009.12562", "2106.02674", "2204.05157"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_8969"} +{"question": "Which paper specifically suggests that large batch sizes are essential for effective training in SimCLR?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_8970"} +{"question": "What studies demonstrated the effectiveness of token-based keypoint embedding?", "answer": ["TokenPose: Learning Keypoint Tokens for Human Pose Estimation", "Poseur: Direct Human Pose Regression with Transformers"], "answer_arxiv_id": ["2104.03516", "2201.07412"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_8971"} +{"question": "Which work outlined the concept of sliding window attention for large language models?", "answer": ["Longformer: The Long-Document Transformer"], "answer_arxiv_id": ["2004.05150"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_8972"} +{"question": "What papers highlight the presence of Human Label Variation?", "answer": ["The 'Problem' of Human Label Variation: On Ground Truth in Data,\n Modeling and Evaluation"], "answer_arxiv_id": ["2211.02570"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_8973"} +{"question": "In what studies have runtime improvements been made using predictions?", "answer": ["Faster Matchings via Learned Duals", "Faster Fundamental Graph Algorithms via Learned Predictions", "Discrete-Convex-Analysis-Based Framework for Warm-Starting Algorithms with Predictions", "Generalized Sorting with Predictions"], "answer_arxiv_id": ["2107.09770", "2204.12055", "2205.09961", "2011.00172"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_8974"} +{"question": "What papers revealed that strong adaptivity cannot be achieved in the unconstrained setting?", "answer": ["Parameter-free Mirror Descent"], "answer_arxiv_id": ["2203.00444"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_8975"} +{"question": "What papers model the problem through variational inference and KL divergence optimization for Neural Radiance Fields?", "answer": ["Stochastic Neural Radiance Fields: Quantifying Uncertainty in Implicit\n 3D Representations", "Conditional-Flow NeRF: Accurate 3D Modelling with Reliable Uncertainty\n Quantification"], "answer_arxiv_id": ["2109.02123", "2203.10192"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_8976"} +{"question": "Which papers highlight words or sentences as rationales for the generated summaries?", "answer": ["A Unified Model for Extractive and Abstractive Summarization using Inconsistency Loss", "Bottom-Up Abstractive Summarization", "Text Summarization with Pretrained Encoders", "SEAL: Segment-wise Extractive-Abstractive Long-form Text Summarization", "EASE: Extractive-Abstractive Summarization with Explanations"], "answer_arxiv_id": ["1805.06266", "1808.10792", "1908.08345", "2006.10213", "2105.06982"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_8977"} +{"question": "What is the study that takes Double Envy-Freeness Up To c Matches into account which ensures that individuals are satisfied?", "answer": ["Two-Sided Matching Meets Fair Division"], "answer_arxiv_id": ["2107.07404v1"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_8978"} +{"question": "What work introduced a multi-scale spatial distillation scheme for continual semantic segmentation?", "answer": ["PLOP: Learning without Forgetting for Continual Semantic Segmentation"], "answer_arxiv_id": ["2011.11390"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8979"} +{"question": "Which works analyze the substructure counting power of Recursive Neighborhood Pooling?", "answer": ["Counting Substructures with Higher-Order Graph Neural Networks: Possibility and Impossibility Results"], "answer_arxiv_id": ["2012.03174"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_train_8980"} +{"question": "Which research papers delved into few-shot segmentation by adopting a prototype-guided approach?", "answer": ["Self-Support Few-Shot Semantic Segmentation", "Learning What Not to Segment: A New Perspective on Few-Shot Segmentation", "Mining Latent Classes for Few-shot Segmentation", "Feature Weighting and Boosting for Few-Shot Segmentation", "Adaptive Prototype Learning and Allocation for Few-Shot Segmentation", "Prototypical Kernel Learning and Open-set Foreground Perception for\n Generalized Few-shot Semantic Segmentation"], "answer_arxiv_id": ["2207.11549", "2203.07615", "2103.15402", "1909.13140", "2104.01893", "2308.04952"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_8981"} +{"question": "Which projects have developed single-modality solutions for 3D detection and occupancy prediction, such as LiDAR-based or camera-only approaches?", "answer": ["VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud", "Deep Hough Voting for 3D Object Detection in Point Clouds", "Group-Free 3D Object Detection via Transformers", "FCAF3D: Fully Convolutional Anchor-Free 3D Object Detection", "Disentangling Monocular 3D Object Detection", "FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection", "SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint\n Estimation", "ImVoxelNet: Image to Voxels Projection for Monocular and Multi-View\n General-Purpose 3D Object Detection", "Probabilistic and Geometric Depth: Detecting Objects in Perspective", "Monocular 3D Object Detection with Depth from Motion"], "answer_arxiv_id": ["1711.06396", "1812.05784", "1812.04244", "1904.09664", "2104.00678", "2112.00322", "1905.12365", "2104.10956", "2002.10111", "2106.01178", "2107.14160", "2207.12988"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_8982"} +{"question": "Which research paper introduced correlation on images by leveraging a wavelet basis?", "answer": ["Wavelet Score-Based Generative Modeling"], "answer_arxiv_id": ["2208.05003"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_8983"} +{"question": "What papers are about creating complex query embeddings using geometric objects?", "answer": ["Embedding Logical Queries on Knowledge Graphs", "Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings", "ConE: Cone Embeddings for Multi-Hop Reasoning over Knowledge Graphs", "Self-Supervised Hyperboloid Representations from Logical Queries over Knowledge Graphs"], "answer_arxiv_id": ["1806.01445", "2002.05969", "2110.13715", "2012.13023"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_8984"} +{"question": "What research established a baseline for semantic segmentation with a hierarchical MiT encoder and a lightweight All-MLP decoder?", "answer": ["SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers"], "answer_arxiv_id": ["2105.15203"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_8985"} +{"question": "Which paper explores experimentally and theoretically the interpolations between forward and reverse KL objectives?", "answer": ["A Variational Perspective on Generative Flow Networks"], "answer_arxiv_id": ["2210.07992"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_8986"} +{"question": "Which work have studied stochastic gradient methods with independently and identically distributed (i. i. d. ) noise?", "answer": ["Better Mini-Batch Algorithms via Accelerated Gradient Methods", "A Smoothing Stochastic Gradient Method for Composite Optimization", "Fast and Faster Convergence of SGD for Over-Parameterized Models and an Accelerated Perceptron", "Stochastic first-order methods: non-asymptotic and computer-aided analyses via potential functions", "A Universally Optimal Multistage Accelerated Stochastic Gradient Method", "Near-Optimal High Probability Complexity Bounds for Non-Smooth Stochastic Optimization with Heavy-Tailed Noise", "An Even More Optimal Stochastic Optimization Algorithm: Minibatching and Interpolation Learning", "Adam: A Method for Stochastic Optimization"], "answer_arxiv_id": ["1106.4574", "1008.5204", "1810.07288", "1902.00947v6", "1901.08022", "2106.05958", "2106.02720", "1412.6980v9"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_8987"} +{"question": "What studies documented the reliance of methodologies on beam search decoding and top-1 hypothesis selection for inference?", "answer": ["Multi-task Language Modeling for Improving Speech Recognition of Rare\n Words"], "answer_arxiv_id": ["2011.11715"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_8988"} +{"question": "What works focused on learning coordinate systems that approximately linearize a system’s dynamics?", "answer": ["Forecasting Sequential Data Using Consistent Koopman Autoencoders"], "answer_arxiv_id": ["2003.02236"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_8989"} +{"question": "Which projects generated data for multiple languages by processing common crawl dumps?", "answer": ["Towards a Cleaner Document-Oriented Multilingual Crawled Corpus", "Unsupervised Cross-lingual Representation Learning at Scale", "mT5: A massively multilingual pre-trained text-to-text transformer"], "answer_arxiv_id": ["2201.06642", "1911.02116", "2010.11934"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_8990"} +{"question": "Which researches focus on utilizing roadside sensor data for 3D object detection?", "answer": ["Rope3D: The Roadside Perception Dataset for Autonomous Driving and Monocular 3D Object Detection Task", "BEVHeight: A Robust Framework for Vision-based Roadside 3D Object Detection", "A9-Dataset: Multi-Sensor Infrastructure-Based Dataset for Mobility Research"], "answer_arxiv_id": ["2203.13608", "2303.08498", "2204.06527"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_8991"} +{"question": "Which research papers analyze calibration for empirical risk minimization and highlight the role played by regularization?", "answer": ["On double-descent in uncertainty quantification in overparametrized models", "Theoretical characterization of uncertainty in high-dimensional linear classification"], "answer_arxiv_id": ["2210.12760v4", "2202.03295"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_8992"} +{"question": "What works studied the policy optimization in linear MDP?", "answer": ["Near-optimal Offline Reinforcement Learning with Linear Representation: Leveraging Variance Information with Pessimism"], "answer_arxiv_id": ["2203.05804v1"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_8993"} +{"question": "What are the current methods focused on improving the performance of tasks or datasets in the context of adapting the vision-language model?", "answer": ["Conditional Prompt Learning for Vision-Language Models", "Learning to Prompt for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters"], "answer_arxiv_id": ["2203.05557", "2109.01134", "2111.03930", "2110.04544"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_8994"} +{"question": "Which papers discussed re-sampling as a method for class-imbalanced learning?", "answer": ["SMOTE: Synthetic Minority Over-sampling Technique", "A systematic study of the class imbalance problem in convolutional neural networks", "What is the Effect of Importance Weighting in Deep Learning?"], "answer_arxiv_id": ["1106.1813", "1710.05381", "1812.03372"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_8995"} +{"question": "What research works focused on finding the most salient reasoning pathway for simple synthetic tasks or toy models?", "answer": ["Progress measures for grokking via mechanistic interpretability", "In-context Learning and Induction Heads", "Interpretability in the Wild: a Circuit for Indirect Object\n Identification in GPT-2 small", "Towards Automated Circuit Discovery for Mechanistic Interpretability", "Towards a Mechanistic Interpretation of Multi-Step Reasoning\n Capabilities of Language Models", "Does Circuit Analysis Interpretability Scale? Evidence from Multiple\n Choice Capabilities in Chinchilla", "The Hydra Effect: Emergent Self-repair in Language Model Computations"], "answer_arxiv_id": ["2301.05217", "2209.11895v1", "2211.00593", "2304.14997", "2310.14491", "2307.09458", "2307.15771"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_8996"} +{"question": "Are there recent works where new GAN and autoregressive models for T2I are developed?", "answer": ["Scaling up GANs for Text-to-Image Synthesis", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2303.05511", "2206.10789"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_8997"} +{"question": "Can you provide me with studies discussing loss surfaces with extreme local sensitivity in dynamical systems under the parameter of interest?", "answer": ["Gradients are Not All You Need"], "answer_arxiv_id": ["2111.05803"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_8998"} +{"question": "Could you provide me the work that shows the performance of in-context learning can vary according to the choice of prompt format, training examples, and prompt order?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models"], "answer_arxiv_id": ["2102.09690"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_8999"} +{"question": "Can you mention any work that applied VAEs in an action-conditioned manner?", "answer": ["Action-Conditioned 3D Human Motion Synthesis with Transformer VAE"], "answer_arxiv_id": ["2104.05670"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_9000"} +{"question": "What studies employed t-SVD in t-product layers in DNNs?", "answer": ["Stable Tensor Neural Networks for Rapid Deep Learning"], "answer_arxiv_id": ["1811.06569v1"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_9001"} +{"question": "Are there any works that discuss the application of optimization-based approaches for motion generation in 3D scenes?", "answer": ["Synthesizing Long-Term 3D Human Motion and Interaction in 3D Scenes"], "answer_arxiv_id": ["2012.05522"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_9002"} +{"question": "Can you name the study that showed promising results by connecting vision encoders and LLMs with a small intermediate model?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2201.12086", "2305.14720", "2304.08485", "2304.10592"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_9003"} +{"question": "Can you name some works about adversarial distribution alignment in cross-domain object detection?", "answer": ["Adaptive Object Detection with Dual Multi-Label Prediction"], "answer_arxiv_id": ["2003.12943"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_9004"} +{"question": "Can you provide some references to works on Transfer Learning(TL) that focus on transferring knowledge from source domains?", "answer": ["A Comprehensive Survey on Transfer Learning"], "answer_arxiv_id": ["1911.02685"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_9005"} +{"question": "What papers discussed the use of BYOL in learning representations of sequential data?", "answer": ["Bootstrap Latent-Predictive Representations for Multitask Reinforcement Learning", "Broaden Your Views for Self-Supervised Video Learning", "BYOL for Audio: Self-Supervised Learning for General-Purpose Audio Representation"], "answer_arxiv_id": ["2004.14646", "2103.16559", "2103.06695"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_9006"} +{"question": "Can you cite studies that use mixed integer linear programming in verifying networks?", "answer": ["Evaluating Robustness of Neural Networks with Mixed Integer Programming"], "answer_arxiv_id": ["1711.07356"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_9007"} +{"question": "Which studies improved the quality of DALL-E's text-to-image (T2I) generation model?", "answer": ["Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "CogView2: Faster and Better Text-to-Image Generation via Hierarchical\n Transformers", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["2206.10789", "2204.14217", "2203.13131"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_9008"} +{"question": "Which papers discuss the Transformer-XL technique of memorization?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context"], "answer_arxiv_id": ["1901.02860"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_9009"} +{"question": "Could you provide me with some research that explores progressively growing models during training?", "answer": ["Shallow-to-Deep Training for Neural Machine Translation", "Staged Training for Transformer Language Models"], "answer_arxiv_id": ["2010.03737", "2203.06211"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_9010"} +{"question": "Which papers talk about the extension of CLIP for high-efficiency model training and cycle consistency?", "answer": ["Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models"], "answer_arxiv_id": ["2107.07651", "2201.12086", "2301.12597"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_9011"} +{"question": "What works are related to the application of image editing methods to strong generative models?", "answer": ["InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "Photorealistic Style Transfer via Wavelet Transforms"], "answer_arxiv_id": ["2005.09635v2", "2103.17249", "1903.09760"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_9012"} +{"question": "What papers discuss a coarse-to-fine pipeline in learning-based methods for visual localization in 2D images?", "answer": ["From Coarse to Fine: Robust Hierarchical Localization at Large Scale"], "answer_arxiv_id": ["1812.03506"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9013"} +{"question": "Can you provide any studies concerned with the discovery of entire subspaces as concepts?", "answer": ["Multi-dimensional concept discovery (MCD): A unifying framework with completeness guarantees"], "answer_arxiv_id": ["2301.11911"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_9014"} +{"question": "What works have address semantic segmentation for parts in part-level image segmentation?", "answer": ["FLOAT: Factorized Learning of Object Attributes for Improved\n Multi-object Multi-part Scene Parsing", "GMNet: Graph Matching Network for Large Scale Part Semantic Segmentation\n in the Wild", "Going Denser with Open-Vocabulary Part Segmentation", "OV-PARTS: Towards Open-Vocabulary Part Segmentation", "Learning Part Segmentation through Unsupervised Domain Adaptation from\n Synthetic Vehicles", "Deep Hierarchical Semantic Segmentation", "Graphonomy: Universal Human Parsing via Graph Transfer Learning", "Instance-level Human Parsing via Part Grouping Network", "Self-Correction for Human Parsing", "Holistic, Instance-Level Human Parsing", "Look into Person: Joint Body Parsing & Pose Estimation Network and A New\n Benchmark", "Devil in the Details: Towards Accurate Single and Multiple Human Parsing", "Hierarchical Human Parsing with Typed Part-Relation Reasoning", "Understanding Humans in Crowded Scenes: Deep Nested Adversarial Learning\n and A New Benchmark for Multi-Human Parsing"], "answer_arxiv_id": ["2203.16168", "2007.09073", "2305.11173", "2310.05107", "2103.14098", "2203.14335", "1904.04536", "1808.00157", "1910.09777", "1709.03612", "1804.01984", "1809.05996", "2003.04845", "1804.03287"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_9015"} +{"question": "Which papers developed methods for model calibration including isotonic regression, temperature scaling and Bayesian binning?", "answer": ["On Calibration of Modern Neural Networks"], "answer_arxiv_id": ["1706.04599"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_9016"} +{"question": "Are there any papers that proposed grammar-based approaches to improve compositional generalization?", "answer": ["Sequence-to-Sequence Learning with Latent Neural Grammars"], "answer_arxiv_id": ["2109.01135v7"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_9017"} +{"question": "Which works are about generative models based on MLPs for learning continous image distributions?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "Generative Models as Distributions of Functions"], "answer_arxiv_id": ["1812.02822", "2102.04776"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_9018"} +{"question": "What works provide open-source corpora for low-resource languages in specific regions?", "answer": ["IndoNLG: Benchmark and Resources for Evaluating Indonesian Natural\n Language Generation", "Towards Leaving No Indic Language Behind: Building Monolingual Corpora,\n Benchmark and Models for Indic Languages"], "answer_arxiv_id": ["2104.08200", "2212.05409"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_9019"} +{"question": "Which works discuss the unique challenges of satellite imagery in relation to supervised learning?", "answer": ["Foreground-Aware Relation Network for Geospatial Object Segmentation in\n High Spatial Resolution Remote Sensing Imagery"], "answer_arxiv_id": ["2011.09766"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_9020"} +{"question": "Which works have examined fairness guarantees in the setting of ranking?", "answer": ["Fairness Through Regularization for Learning to Rank"], "answer_arxiv_id": ["2102.05996"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_9021"} +{"question": "Which paper involves the use of an autoencoding model for compressing the image space in inpainting approaches?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_9022"} +{"question": "Could you provide me papers where they model a variety of dynamic loss functions?", "answer": ["Learning to Teach with Dynamic Loss Functions", "Addressing the Loss-Metric Mismatch with Adaptive Loss Alignment", "Meta-Learning with Task-Adaptive Loss Function for Few-Shot Learning"], "answer_arxiv_id": ["1810.12081", "1905.05895", "2110.03909"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_9023"} +{"question": "What work has been discussed that lacks access to raw RGB-D data?", "answer": ["Habitat-Matterport 3D Semantics Dataset"], "answer_arxiv_id": ["2210.05633"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_9024"} +{"question": "What work articulates how large language models are affected by changes in task instructions and context?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models"], "answer_arxiv_id": ["2102.09690"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_train_9025"} +{"question": "What studies investigate using LLMs for tool usage?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "Tool Learning with Foundation Models", "Voyager: An Open-Ended Embodied Agent with Large Language Models"], "answer_arxiv_id": ["2302.04761", "2304.08354", "2305.16291"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9026"} +{"question": "Which study performed Bayesian inference directly over the Transformer output by fitting a GP over the last layer output?", "answer": ["Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness"], "answer_arxiv_id": ["2006.10108"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_9027"} +{"question": "Which papers employ PointNet for 3D point cloud supervision?", "answer": ["A Point Set Generation Network for 3D Object Reconstruction from a\n Single Image"], "answer_arxiv_id": ["1612.00603"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_9028"} +{"question": "Which papers discuss about the high computational cost related to the eNTK computations?", "answer": ["Towards NNGP-guided Neural Architecture Search", "A Framework and Benchmark for Deep Batch Active Learning for Regression", "Deep learning versus kernel learning: an empirical study of loss landscape geometry and the time evolution of the Neural Tangent Kernel"], "answer_arxiv_id": ["2011.06006", "2203.09410v4", "2010.15110"], "source_meta": {"published_time": "20220625"}, "qid": "AutoScholarQuery_train_9029"} +{"question": "What resources offer a deep dive into heterogeneous parallel linear MDP and Markov games in linear multi-agent MDPs?", "answer": ["Provably Efficient Cooperative Multi-Agent Reinforcement Learning with Function Approximation"], "answer_arxiv_id": ["2103.04972"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_9030"} +{"question": "What papers are about large multi-modal models?", "answer": ["The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2309.17421", "2304.10592"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_9031"} +{"question": "Could you provide me the work where the experimental setup, ViT-g text encoder network and ViT-G image encoder were used?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9032"} +{"question": "Could you provide me the reference where researchers identified the regions that compromise safety alignment of a Language Model?", "answer": ["Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank\n Modifications"], "answer_arxiv_id": ["2402.05162"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_9033"} +{"question": "Can you mention some studies that use the concept of 'symmetry' in their works?", "answer": ["Towards a Definition of Disentangled Representations", "Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["1812.02230", "2104.13478"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_9034"} +{"question": "Can you reference some studies that have applied deep learning to create puns?", "answer": ["Pun-GAN: Generative Adversarial Network for Pun Generation"], "answer_arxiv_id": ["1910.10950v1"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_9035"} +{"question": "Which works have demonstrated that robust GNN architectures can be formed via graph propagation layers that mirror the unfolded descent iterations of a graph-regularized energy function?", "answer": ["Elastic Graph Neural Networks", "A Unified View on Graph Neural Networks as Graph Signal Denoising", "Graph Neural Networks Inspired by Classical Iterative Algorithms", "Revisiting Graph Convolutional Network on Semi-Supervised Node Classification from an Optimization Perspective", "Interpreting and Unifying Graph Neural Networks with An Optimization Framework", "Descent Steps of a Relation-Aware Energy Produce Heterogeneous Graph Neural Networks"], "answer_arxiv_id": ["2107.06996", "2010.01777", "2103.06064v4", "2009.11469", "2101.11859", "2206.11081"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_9036"} +{"question": "Which works used the copy mechanism for grammatical error correction by copying correct text from the input to the target?", "answer": ["Improving Grammatical Error Correction via Pre-Training a Copy-Augmented\n Architecture with Unlabeled Data"], "answer_arxiv_id": ["1903.00138"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_9037"} +{"question": "What papers are about the training dynamics near a manifold of minima and study the effect of noise on sharpness?", "answer": ["Implicit regularization for deep neural networks driven by an Ornstein-Uhlenbeck like process", "Label Noise SGD Provably Prefers Flat Global Minimizers", "What Happens after SGD Reaches Zero Loss? –A Mathematical Framework"], "answer_arxiv_id": ["1904.09080", "2106.06530", "2110.06914"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_9038"} +{"question": "Could you provide me a work which proposed mitigating hallucinations with a negative task vector?", "answer": ["Elastic Weight Removal for Faithful and Abstractive Dialogue Generation"], "answer_arxiv_id": ["2303.17574"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_9039"} +{"question": "Which papers discuss the challenges for LLMs to understand user intent and responses accurately?", "answer": ["HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "TaskMatrix.AI: Completing Tasks by Connecting Foundation Models with Millions of APIs", "MM-ReAct : Prompting ChatGPT for Multimodal Reasoning and Action"], "answer_arxiv_id": ["2303.17580", "2303.04671", "2303.16434v1", "2303.11381"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_9040"} +{"question": "Which study discusses the EDICT method that enables mathematically exact DDIM-inversion of real images?", "answer": ["EDICT: Exact Diffusion Inversion via Coupled Transformations"], "answer_arxiv_id": ["2211.12446"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_9041"} +{"question": "Could you give me examples of research that significantly improved the synthesis capability on high-resolution datasets?", "answer": ["Progressive Growing of GANs for Improved Quality, Stability, and Variation", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Self-Attention Generative Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["1710.10196", "1809.11096", "1805.08318", "1812.04948", "1912.04958"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_9042"} +{"question": "What works apply the Lipschitz constraint in adversarial training and learning implicit neural functions?", "answer": ["Learning Smooth Neural Functions via Lipschitz Regularization", "Adversarial Lipschitz Regularization"], "answer_arxiv_id": ["2202.08345", "1907.05681"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_9043"} +{"question": "Which works propose oracle efficient algorithms for general policy classes?", "answer": ["BISTRO: An Efficient Relaxation-Based Method for Contextual Bandits", "Improved Regret Bounds for Oracle-Based Adversarial Contextual Bandits"], "answer_arxiv_id": ["1602.02196", "1606.00313v1"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_9044"} +{"question": "What are the available studies regarding the use of masked token prediction in self-supervised learning?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "SimMIM: a Simple Framework for Masked Image Modeling", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2106.08254", "2111.09886", "2111.06377"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_9045"} +{"question": "Are there any studies that proposed to reduce the computational cost of model training related with the graph structure in Transformers?", "answer": ["Gophormer: Ego-Graph Transformer for Node Classification"], "answer_arxiv_id": ["2110.13094"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_9046"} +{"question": "Which work combines the visual and linguistic branches of CLIP to learn hierarchical prompts?", "answer": ["MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2210.03117"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_9047"} +{"question": "Which papers worked on training a meta-learner on how to update the parameters of a downstream learner in the context of meta-learning?", "answer": ["Learning to Optimize"], "answer_arxiv_id": ["1606.01885"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_9048"} +{"question": "In which paper does the researcher use first-order Euler’s method to reduce the inference timesteps in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_9049"} +{"question": "What works have been done on ensembles of differentiable learners in the context of tabular data?", "answer": ["Neural Oblivious Decision Ensembles for Deep Learning on Tabular Data", "The Tree Ensemble Layer: Differentiability meets Conditional Computation", "Gradient Boosting Neural Networks: GrowNet"], "answer_arxiv_id": ["1909.06312", "2002.07772", "2002.07971"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_9050"} +{"question": "Which work utilized generic non-semantic queries and semantic queries for segmentation tasks?", "answer": ["Generalized Decoding for Pixel, Image, and Language"], "answer_arxiv_id": ["2212.11270"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_9051"} +{"question": "Which research replaced the Features Line of Sight Projection of MonoScene with TPVFormer to enhance the performance of surround-view occupancy prediction?", "answer": ["Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction"], "answer_arxiv_id": ["2302.07817"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_9052"} +{"question": "What studies have been done on Diffusion Models or score-based generative models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Improved Denoising Diffusion Probabilistic Models", "Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential Equations", "Diffusion Models Beat GANs on Image Synthesis", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "MVDiffusion: Enabling Holistic Multi-view Image Generation with Correspondence-Aware Diffusion", "Improved Denoising Diffusion Probabilistic Models", "Variational Diffusion Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["1503.03585", "2102.09672", "1907.05600", "2011.13456", "2105.05233", "2205.11487", "2204.06125", "2112.10752", "2307.01097", "2102.09672", "2107.00630", "2206.00364", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_9053"} +{"question": "Could you provide me some works that focused on behavior regularization in modern offline reinforcement learning for dealing with partial data coverage?", "answer": ["Addressing Function Approximation Error in Actor-Critic Methods", "Safe Policy Improvement with Baseline Bootstrapping", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Keep Doing What Worked: Behavioral Modelling Priors for Offline Reinforcement Learning", "Behavior Regularized Offline Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["1802.09477", "1712.06924", "1906.00949", "2002.08396", "1911.11361", "2106.06860", "2110.06169"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_9054"} +{"question": "Are there any studies that further enrich the signal propagation with personalised PageRank and graph diffusion?", "answer": ["Diffusion Improves Graph Learning"], "answer_arxiv_id": ["1911.05485"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_9055"} +{"question": "Which studies extend the line of research beyond prediction intervals in the domain of conformal prediction?", "answer": ["A Tutorial on Conformal Prediction", "Distribution-Free Predictive Inference For Regression", "Conformalized Quantile Regression"], "answer_arxiv_id": ["0706.3188", "1604.04173", "1905.03222"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_9056"} +{"question": "Which papers have examined the theory of Graph Neural Networks, including their generalization and computational complexity?", "answer": ["Generalization and Representational Limits of Graph Neural Networks", "What Can Neural Networks Reason About?", "Inductive Representation Learning on Large Graphs", "FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling", "Layer-Dependent Importance Sampling for Training Deep and Large Graph Convolutional Networks", "Constant Time Graph Neural Networks"], "answer_arxiv_id": ["2002.06157", "1905.13211", "1706.02216", "1801.10247", "1911.07323", "1901.07868"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_9057"} +{"question": "Which studies have introduced latent diffusion models that conduct iterative denoising processes for text-to-image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_9058"} +{"question": "What papers involved in designing detailed error schemes for LLM to output fine-grained error annotations?", "answer": ["GEMBA-MQM: Detecting Translation Quality Error Spans with GPT-4"], "answer_arxiv_id": ["2310.13988"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_9059"} +{"question": "Which studies apply techniques such as data augmentation, self-supervised learning, and the integration of task-specific information to address the data scarcity issue in visual representation learning for robotics?", "answer": ["Reinforcement Learning with Augmented Data", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement\n Learning from Pixels", "Scaling Robot Learning with Semantically Imagined Experience", "RoboAgent: Generalization and Efficiency in Robot Manipulation via\n Semantic Augmentations and Action Chunking", "CURL: Contrastive Unsupervised Representations for Reinforcement\n Learning", "The Surprising Effectiveness of Representation Learning for Visual\n Imitation", "Robot Learning with Sensorimotor Pre-training", "Learning Invariant Representations for Reinforcement Learning without\n Reconstruction"], "answer_arxiv_id": ["2004.14990", "2004.13649", "2302.11550", "2309.01918", "2004.04136", "2112.01511", "2306.10007", "2006.10742"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_9060"} +{"question": "Which work first used a triplet loss to improve model performance in cross-modal retrieval?", "answer": ["VSE++: Improving Visual-Semantic Embeddings with Hard Negatives"], "answer_arxiv_id": ["1707.05612"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_9061"} +{"question": "Which papers proposed sophisticated pseudo label filtering strategies in semi-supervised learning?", "answer": ["FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo\n Labeling", "FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning", "SoftMatch: Addressing the Quantity-Quality Trade-off in Semi-supervised\n Learning"], "answer_arxiv_id": ["2110.08263", "2205.07246", "2301.10921"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_9062"} +{"question": "What are some of the papers which focus on sub-quadratic approximations of attention and employ sparsification?", "answer": ["Longformer: The Long-Document Transformer", "Reformer: The Efficient Transformer", "Luna: Linear Unified Nested Attention", "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts"], "answer_arxiv_id": ["2004.05150", "2001.04451", "2106.01540", "2101.03961", "2112.06905"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_9063"} +{"question": "What are the studies that offer complementary work on mechanism design?", "answer": ["Optimal Data Acquisition for Statistical Estimation", "Optimal and Quantized Mechanism Design for Fresh Data Acquisition"], "answer_arxiv_id": ["1711.01295", "2006.15751"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_9064"} +{"question": "What papers proposed improving label noise problem using robust clustering techniques?", "answer": ["Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive\n Object Re-ID", "AD-Cluster: Augmented Discriminative Clustering for Domain Adaptive\n Person Re-identification"], "answer_arxiv_id": ["2006.02713", "2004.08787"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_9065"} +{"question": "What studies have shown that SSL representations encode richer visual details about input images compared to supervised models?", "answer": ["High Fidelity Visualization of What Your Self-Supervised Representation Knows About"], "answer_arxiv_id": ["2112.09164"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_9066"} +{"question": "What are the research papers that introduced benchmarks such as ZeroSCROLLS, L-Eval and LongBench to evaluate long text modelling capability of LLMs?", "answer": ["ZeroSCROLLS: A Zero-Shot Benchmark for Long Text Understanding", "L-Eval: Instituting Standardized Evaluation for Long Context Language Models", "LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding"], "answer_arxiv_id": ["2305.14196", "2307.11088v3", "2308.14508v2"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_9067"} +{"question": "What are some of the papers that signify human-object interaction as triplets in 2D domain?", "answer": ["HOTR: End-to-End Human-Object Interaction Detection with Transformers", "End-to-End Human Object Interaction Detection with HOI Transformer", "Learning to Detect Human-Object Interactions", "Detecting and Recognizing Human-Object Interactions", "Neural-Logic Human-Object Interaction Detection"], "answer_arxiv_id": ["2104.13682", "2103.04503", "1702.05448", "1704.07333", "2311.09817"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_9068"} +{"question": "Could you provide me some studies conducted on bi-level optimization methods in relation to dataset distillation?", "answer": ["Dataset Distillation with Infinitely Wide Convolutional Networks", "Dataset Meta-Learning from Kernel Ridge-Regression", "Dataset Distillation using Neural Feature Regression", "Efficient Dataset Distillation Using Random Feature Approximation", "Dataset Distillation with Convexified Implicit Gradients", "Remember the Past: Distilling Datasets into Addressable Memories for\n Neural Networks"], "answer_arxiv_id": ["2107.13034", "2011.00050", "2206.00719", "2210.12067", "2302.06755v2", "2206.02916"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9069"} +{"question": "Could you indicate some research where motion is controlled with high-level guidance from audio and natural language?", "answer": ["AI Choreographer: Music Conditioned 3D Dance Generation with AIST++", "Rhythm is a Dancer: Music-Driven Motion Synthesis with Global Structure", "Language2Pose: Natural Language Grounded Pose Forecasting", "TEMOS: Generating diverse human motions from textual descriptions"], "answer_arxiv_id": ["2101.08779", "2111.12159", "1907.01108", "2204.14109"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_9070"} +{"question": "What papers proposed generative networks that are equivariant under rotation or translation?", "answer": ["Explicitly disentangling image content from translation and rotation with spatial-VAE", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Image Generators with Conditionally-Independent Pixel Synthesis", "Alias-Free Generative Adversarial Networks", "Vector Neurons: A General Framework for SO(3)-Equivariant Networks", "Learning Continuous Image Representation with Local Implicit Image Function", "Unsupervised Object Representation Learning using Translation and Rotation Group Equivariant VAE"], "answer_arxiv_id": ["1909.11663", "2003.08934", "2011.13775", "2106.12423", "2104.12229", "2012.09161", "2210.12918"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_9071"} +{"question": "Which research showed the importance of the relative smoothness of the representations and target functions in the ODE limit of TD dynamics?", "answer": ["A Geometric Perspective on Optimal Representations for Reinforcement Learning"], "answer_arxiv_id": ["1901.11530"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_9072"} +{"question": "Could you provide me some studies that have introduced eased and refined dictionaries with multiple choices?", "answer": ["mc-BEiT: Multi-choice Discretization for Image BERT Pre-training"], "answer_arxiv_id": ["2203.15371"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_9073"} +{"question": "What are some examples of recent studies that used regularization in learning within games in various different settings?", "answer": ["From Poincaré Recurrence to Convergence in Imperfect Information Games: Finding Equilibrium via Regularization", "Exploration-Exploitation in Multi-Agent Competition: Convergence with Bounded Rationality"], "answer_arxiv_id": ["2002.08456", "2106.12928"], "source_meta": {"published_time": "20220619"}, "qid": "AutoScholarQuery_train_9074"} +{"question": "Could you tell me about the works which used cross-attention through query-based methods, viewing segmentation as a set prediction problem?", "answer": ["MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "Masked-attention Mask Transformer for Universal Image Segmentation", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "SeMask: Semantically Masked Transformers for Semantic Segmentation", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2012.00759", "2112.01527", "2107.06278", "2112.12782", "2005.12872"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_9075"} +{"question": "Which work established SMO for VQE optimization using NFT?", "answer": ["Sequential minimal optimization for quantum-classical hybrid algorithms"], "answer_arxiv_id": ["1903.12166"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_9076"} +{"question": "Which papers proposed differentially private quantiles?", "answer": ["Differentially Private Quantiles"], "answer_arxiv_id": ["2102.08244"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_9077"} +{"question": "What studies have used pseudo-labeling for localization in language-based object detectors?", "answer": ["Exploiting Unlabeled Data with Vision and Language Models for Object\n Detection", "Scaling Open-Vocabulary Object Detection"], "answer_arxiv_id": ["2207.08954", "2306.09683"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_9078"} +{"question": "Can you name the studies on federated evaluation of classifiers?", "answer": ["Federated Calibration and Evaluation of Binary Classifiers"], "answer_arxiv_id": ["2210.12526"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_9079"} +{"question": "What papers introduced memory to Transformers through recurrence?", "answer": ["Memformer: A Memory-Augmented Transformer for Sequence Modeling"], "answer_arxiv_id": ["2010.06891"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_9080"} +{"question": "What studies improved performance at inference time in translating natural language into executable forms?", "answer": ["Synchromesh: Reliable code generation from pre-trained language models", "Natural Language to Code Translation with Execution"], "answer_arxiv_id": ["2201.11227", "2204.11454"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_9081"} +{"question": "What are the recent surveys on non-rigid shape matching?", "answer": ["Deep Learning for 3D Point Clouds: A Survey", "Geometric deep learning: going beyond Euclidean data"], "answer_arxiv_id": ["1912.12033", "1611.08097"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_9082"} +{"question": "Which studies explored curiosity-based intrinsic motivation using diversity?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function"], "answer_arxiv_id": ["1802.06070"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_9083"} +{"question": "What work uses Expectation-Maximization to train an optical flow model in a semi-supervised setting?", "answer": ["RealFlow: EM-based Realistic Optical Flow Dataset Generation from Videos"], "answer_arxiv_id": ["2207.11075"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_9084"} +{"question": "In which papers can I find models that achieve large kernel sizes by training large and small kernels in parallel or using a training method with vertical and horizontal rectangular kernels?", "answer": ["More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using\n Sparsity"], "answer_arxiv_id": ["2207.03620"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_9085"} +{"question": "Could you provide me some studies that extend the idea of maximum entropy RL by using learnable state-independent prior policy?", "answer": ["Mutual-Information Regularization in Markov Decision Processes and Actor-Critic Learning"], "answer_arxiv_id": ["1909.05950"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_9086"} +{"question": "Which work referred to the poor performance of the State of The Art online adaptation model FSNet in the ECL dataset?", "answer": ["Learning Fast and Slow for Online Time Series Forecasting"], "answer_arxiv_id": ["2202.11672"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_9087"} +{"question": "What papers are focused on merging point cloud data with visual features extracted from image for scene analysis?", "answer": ["Towards Open Vocabulary Learning: A Survey"], "answer_arxiv_id": ["2306.15880"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9088"} +{"question": "What papers are about using LLMs to generate sentence pairs?", "answer": ["Generating Datasets with Pretrained Language Models"], "answer_arxiv_id": ["2104.07540"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_9089"} +{"question": "Which papers discuss the effectiveness of image-language models in representation learning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "FILIP: Fine-grained Interactive Language-Image Pre-Training", "FLAVA: A Foundational Language And Vision Alignment Model"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2205.01917", "2107.07651", "2111.07783", "2112.04482"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_9090"} +{"question": "Could you name some studies that propose post hoc methods in OOD detection?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "Energy-based Out-of-distribution Detection", "On the Importance of Gradients for Detecting Distributional Shifts in the Wild", "ReAct: Out-of-distribution Detection With Rectified Activations", "Out-of-Distribution Detection with Deep Nearest Neighbors"], "answer_arxiv_id": ["1610.02136", "1706.02690", "1807.03888", "2010.03759", "2110.00218", "2111.12797", "2204.06507"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9091"} +{"question": "What papers propose the use of hierarchical features in Vision Transformers?", "answer": ["Shunted Self-Attention via Multi-Scale Token Aggregation", "CMT: Convolutional Neural Networks Meet Vision Transformers", "MViTv2: Improved Multiscale Vision Transformers for Classification and Detection", "CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows", "Global Context Vision Transformers", "RegionViT: Regional-to-Local Attention for Vision Transformers", "ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases", "MPViT : Multi-Path Vision Transformer for Dense Prediction"], "answer_arxiv_id": ["2111.15193", "2107.06263", "2112.01526", "2107.00652", "2206.09959", "2106.02689", "2103.10697", "2112.11010"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_9092"} +{"question": "Can you list the works that explored the integration of various foundation models?", "answer": ["Pre-Trained Language Models for Interactive Decision-Making", "Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language"], "answer_arxiv_id": ["2202.01771", "2204.00598"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_9093"} +{"question": "What studies improve the segmentation performance by increasing the number of prototypes?", "answer": ["Adaptive Prototype Learning and Allocation for Few-Shot Segmentation"], "answer_arxiv_id": ["2104.01893"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_9094"} +{"question": "What works explored the use of differentiable interpreters to learn subfunctions from program sketches and datasets?", "answer": ["Programming with a Differentiable Forth Interpreter", "Neural Programmer-Interpreters"], "answer_arxiv_id": ["1605.06640", "1511.06279"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9095"} +{"question": "Could you name some studies related to audio-visual retrieval?", "answer": ["Image2song: Song Retrieval via Bridging Image Content and Lyric Words", "Cross-modal Embeddings for Video and Audio Retrieval"], "answer_arxiv_id": ["1708.05851", "1801.02200"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_9096"} +{"question": "Which studies aim to prevent representation collapse in Contrastive SSL without negative samples by applying cross-correlation loss or training strategies?", "answer": ["Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2103.03230", "2105.04906", "2006.07733", "2011.10566"], "source_meta": {"published_time": "20211101"}, "qid": "AutoScholarQuery_train_9097"} +{"question": "Could you provide me studies that characterized temperature as noise scale in the optimization process?", "answer": ["Rethinking generalization requires revisiting old ideas: statistical mechanics approaches and complex learning behavior", "An Empirical Model of Large-Batch Training", "Taxonomizing local versus global structure in neural network loss landscapes"], "answer_arxiv_id": ["1710.09553", "1812.06162", "2107.11228"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_9098"} +{"question": "Which papers introduced the episodic memory benchmark's VQL tasks?", "answer": ["Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2110.07058"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9099"} +{"question": "Which work used active tracking methods and record sounds from a speaker with special ultrasonic speakers?", "answer": ["PoseKernelLifter: Metric Lifting of 3D Human Pose using Sound"], "answer_arxiv_id": ["2112.00216"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_9100"} +{"question": "Who demonstrated that a large pre-trained language model as decoder can improve captioning performance when training data is limited?", "answer": ["Encoder-Agnostic Adaptation for Conditional Language Generation", "VisualGPT: Data-efficient Adaptation of Pretrained Language Models for Image Captioning"], "answer_arxiv_id": ["1908.06938", "2102.10407"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9101"} +{"question": "Which works claim that treating offline RL as a sequence modeling problem has high stability and scalability compared to traditional value-based approaches?", "answer": ["Multi-Game Decision Transformers", "A Generalist Agent"], "answer_arxiv_id": ["2205.15241", "2205.06175"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9102"} +{"question": "What are some main studies in Document Question Answering (DocQA)?", "answer": ["DocVQA: A Dataset for VQA on Document Images", "A Dataset of Information-Seeking Questions and Answers Anchored in\n Research Papers"], "answer_arxiv_id": ["2007.00398", "2105.03011"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9103"} +{"question": "What works faced issues with finding optimal splat or kernel radius in splat-based rasterization methods for point cloud rendering?", "answer": ["Differentiable Rendering: A Survey"], "answer_arxiv_id": ["2006.12057"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_9104"} +{"question": "Which model is a generative model that exploits the intuition behind nonequilibrium thermodynamics?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_9105"} +{"question": "Can you point me to some resources for fine-grained alignment?", "answer": ["Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations", "DenseCap: Fully Convolutional Localization Networks for Dense Captioning", "Broaden the Vision: Geo-Diverse Visual Commonsense Reasoning"], "answer_arxiv_id": ["1602.07332", "1511.07571", "2109.06860"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9106"} +{"question": "What are some of the methods in generating 3D data using voxel representation?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "Escaping Plato's Cave: 3D Shape From Adversarial Rendering"], "answer_arxiv_id": ["1610.07584", "1811.11606"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_9107"} +{"question": "What are the studies that have worked on the referring segmentation task focused on segmenting the target object based on a given explicit text description?", "answer": ["Modeling Context Between Objects for Referring Expression Understanding"], "answer_arxiv_id": ["1608.00525"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_9108"} +{"question": "What studies propose model-based meta learning?", "answer": ["Meta Networks", "Recasting Gradient-Based Meta-Learning as Hierarchical Bayes"], "answer_arxiv_id": ["1703.00837", "1801.08930"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_9109"} +{"question": "Could you provide me a study which suggests combining Hearst patterns with Poincaré embeddings for refining existing taxonomy construction approaches?", "answer": ["Every child should have parents: a taxonomy refinement algorithm based\n on hyperbolic term embeddings"], "answer_arxiv_id": ["1906.02002"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_9110"} +{"question": "Could you show me the papers that propose a cubed sphere representation?", "answer": ["Improving data-driven global weather prediction using deep convolutional neural networks on a cubed sphere", "Global extreme heat forecasting using neural weather models"], "answer_arxiv_id": ["2003.11927", "2205.10972"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_9111"} +{"question": "Are there any studies that report predictable improvements on standard internet benchmarks?", "answer": ["Scaling Laws for Neural Language Models", "GPT-4 Technical Report"], "answer_arxiv_id": ["2001.08361", "2303.08774"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_9112"} +{"question": "What research have been conducted on diffusion models for text-to-image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Cascaded Diffusion Models for High Fidelity Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2106.15282", "2205.11487"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_9113"} +{"question": "Who discusses designing efficient compression techniques, used in the silos approach to Federated Learning?", "answer": ["The Convergence of Sparsified Gradient Methods", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Bidirectional compression in heterogeneous settings for distributed or federated learning with partial participation: tight convergence guarantees", "MARINA: Faster Non-Convex Distributed Learning with Compression", "EF21: A New, Simpler, Theoretically Better, and Practically Faster Error Feedback"], "answer_arxiv_id": ["1809.10505", "1910.06378", "2006.14591", "2102.07845", "2106.05203"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_9114"} +{"question": "What research applied the MIM pipeline and designed different reconstruction targets?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "Context Autoencoder for Self-Supervised Representation Learning", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "PeCo: Perceptual Codebook for BERT Pre-training of Vision Transformers", "Masked Frequency Modeling for Self-Supervised Visual Pre-Training", "The Devil is in the Frequency: Geminated Gestalt Autoencoder for Self-Supervised Visual Pre-Training", "Understanding Masked Image Modeling via Learning Occlusion Invariant Feature"], "answer_arxiv_id": ["2106.08254", "2202.03026", "2112.09133", "2111.12710", "2206.07706", "2204.08227", "2208.04164"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_9115"} +{"question": "Which research proposes that scaling up the kernel size of existing ConvNets results in improvements on downstream tasks?", "answer": ["Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs"], "answer_arxiv_id": ["2203.06717"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9116"} +{"question": "Are there any works discussing the similarity between the pretrained and finetuned models in transfer learning?", "answer": ["Explicit Inductive Bias for Transfer Learning with Convolutional Networks"], "answer_arxiv_id": ["1802.01483"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_9117"} +{"question": "What are some examples of work using deep learning for NR-PCQA?", "answer": ["Point Cloud Quality Assessment: Dataset Construction and Learning-based\n No-Reference Metric", "A No-reference Quality Assessment Metric for Point Cloud Based on\n Captured Video Sequences", "PCQA-GRAPHPOINT: Efficients Deep-Based Graph Metric For Point Cloud\n Quality Assessment", "MM-PCQA: Multi-Modal Learning for No-reference Point Cloud Quality\n Assessment"], "answer_arxiv_id": ["2012.11895", "2206.05054", "2211.02459", "2209.00244"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_9118"} +{"question": "What papers are there on Black-box methods in the context of few-shot learning?", "answer": ["Conditional Neural Processes", "A Simple Neural Attentive Meta-Learner", "Fast and Flexible Multi-Task Classification Using Conditional Neural Adaptive Processes"], "answer_arxiv_id": ["1807.01613", "1707.03141v3", "1906.07697"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_9119"} +{"question": "Which studies focus on intrinsic motivation-based exploration algorithms in reinforcement learning?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Unifying Count-Based Exploration and Intrinsic Motivation", "#Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning", "Curiosity-driven Exploration by Self-supervised Prediction", "Large-Scale Study of Curiosity-Driven Learning", "Exploration by Random Network Distillation", "Never Give Up: Learning Directed Exploration Strategies"], "answer_arxiv_id": ["1507.00814", "1606.01868", "1611.04717v3", "1705.05363", "1808.04355", "1810.12894", "2002.06038"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_9120"} +{"question": "Which works use keypoints as a form of supervision while optimizing the 3D shape from an image or video collection?", "answer": ["Birds of a Feather: Capturing Avian Shape Models from Images"], "answer_arxiv_id": ["2105.09396"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_9121"} +{"question": "In what studies do the researchers leverage the chamfer distance loss for self-supervised scene flow learning?", "answer": ["FlowStep3D: Model Unrolling for Self-Supervised Scene Flow Estimation", "Scene Flow from Point Clouds with or without Learning", "PointPWC-Net: A Coarse-to-Fine Network for Supervised and\n Self-Supervised Scene Flow Estimation on 3D Point Clouds", "Neural Scene Flow Prior"], "answer_arxiv_id": ["2011.10147", "2011.00320", "1911.12408", "2111.01253"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_9122"} +{"question": "What papers introduce other destruction processes as an alternative to Gaussian diffusion?", "answer": ["Generative Modelling With Inverse Heat Dissipation", "Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise"], "answer_arxiv_id": ["2206.13397", "2208.09392"], "source_meta": {"published_time": "20220912"}, "qid": "AutoScholarQuery_train_9123"} +{"question": "Which papers can be found about the blind OIQA", "answer": ["Deep Neural Networks for No-Reference and Full-Reference Image Quality Assessment"], "answer_arxiv_id": ["1612.01697"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_9124"} +{"question": "Can you provide me some works that have incorporated 3D geometries into the RL process in molecular generation?", "answer": ["Reinforcement Learning for Molecular Design Guided by Quantum Mechanics", "Symmetry-Aware Actor-Critic for 3D Molecular Design", "Reinforced Genetic Algorithm for Structure-based Drug Design"], "answer_arxiv_id": ["2002.07717", "2011.12747", "2211.16508"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_9125"} +{"question": "What studies proposed methods to ease the problem in multi-modal models by improving the uni-modal optimization in the jointly trained multi-modal framework?", "answer": ["What Makes Training Multi-Modal Classification Networks Hard?", "Balanced Multimodal Learning via On-the-fly Gradient Modulation", "Characterizing and overcoming the greedy nature of learning in\n multi-modal deep neural networks"], "answer_arxiv_id": ["1905.12681", "2203.15332", "2202.05306"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_9126"} +{"question": "Could you name some papers where binary-tree mechanism has been extended to work for sums of real values?", "answer": ["Private Continual Release of Real-Valued Data Streams"], "answer_arxiv_id": ["1811.03197v1"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_9127"} +{"question": "Which studies are about prompting-based learning designed to help pre-trained models generalize to downstream tasks?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["1902.00751", "2101.00190", "2106.09685"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_9128"} +{"question": "Which papers have utilized natural language explanations for natural language inference tasks?", "answer": ["e-SNLI: Natural Language Inference with Natural Language Explanations"], "answer_arxiv_id": ["1812.01193"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_9129"} +{"question": "Which works tried to synthesise speech in a target audio style given text under the task of controllable text-to-speech synthesis?", "answer": ["Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis", "Expressive Speech Synthesis via Modeling Expressions with Variational Autoencoder", "Hierarchical Generative Modeling for Controllable Speech Synthesis"], "answer_arxiv_id": ["1803.09017", "1804.02135", "1810.07217"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_9130"} +{"question": "What works have developed multi-pass ICL by generating multiple responses from subsets of exemplars?", "answer": ["Natural Language to Code Translation with Execution"], "answer_arxiv_id": ["2204.11454"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_9131"} +{"question": "What papers studied the success of auto ML in many areas and influenced the implementation of AFT?", "answer": ["Automated Machine Learning on Graphs: A Survey", "AutoDS: Towards Human-Centered Automation of Data Science", "Beyond Discrete Selection: Continuous Embedding Space Optimization for Generative Feature Selection", "Self-optimizing Feature Generation via Categorical Hashing Representation and Hierarchical Reinforcement Crossing"], "answer_arxiv_id": ["2103.00742", "2101.05273", "2302.13221", "2309.04612"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_9132"} +{"question": "Which studies are about enforcing strict in-sample learning in the field of offline RL?", "answer": ["Offline RL Without Off-Policy Evaluation", "Offline Reinforcement Learning with Implicit Q-Learning", "Offline RL with No OOD Actions: In-Sample Learning via Implicit Value Regularization", "A Policy-Guided Imitation Approach for Offline Reinforcement Learning", "Offline Multi-Agent Reinforcement Learning with Implicit Global-to-Local Value Regularization"], "answer_arxiv_id": ["2106.08909", "2110.06169", "2303.15810", "2210.08323", "2307.11620"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_9133"} +{"question": "Can you name studies that validate the power-law relationship between model accuracy improvements and growing training sets?", "answer": ["Deep Learning Scaling is Predictable, Empirically"], "answer_arxiv_id": ["1712.00409"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_9134"} +{"question": "Which research was the first to use layout information in pre-training documents?", "answer": ["LayoutLM: Pre-training of Text and Layout for Document Image\n Understanding"], "answer_arxiv_id": ["1912.13318"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_9135"} +{"question": "Which research improved ESRGAN by additional noise injection and proposed ESRGAN+?", "answer": ["ESRGAN+ : Further Improving Enhanced Super-Resolution Generative\n Adversarial Network"], "answer_arxiv_id": ["2001.08073"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_9136"} +{"question": "Could you suggest studies that used depth maps as additional supervision for visual question answering?", "answer": ["Weakly Supervised Relative Spatial Reasoning for Visual Question\n Answering"], "answer_arxiv_id": ["2109.01934"], "source_meta": {"published_time": "20240412"}, "qid": "AutoScholarQuery_train_9137"} +{"question": "Which research works focused on developing a template-matching framework for tracking?", "answer": ["Transformer Tracking", "ATOM: Accurate Tracking by Overlap Maximization", "SiamRPN++: Evolution of Siamese Visual Tracking with Very Deep Networks", "Fully-Convolutional Siamese Networks for Object Tracking", "High-Performance Long-Term Tracking with Meta-Updater", "SiamCAR: Siamese Fully Convolutional Classification and Regression for\n Visual Tracking", "Distractor-aware Siamese Networks for Visual Object Tracking"], "answer_arxiv_id": ["2103.15436", "1811.07628", "1812.11703", "1606.09549", "2004.00305", "1911.07241", "1808.06048"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_9138"} +{"question": "Could you name works which put an effort to refine the predicted density map in dense object counting?", "answer": ["Iterative Crowd Counting", "Indiscernible Object Counting in Underwater Scenes"], "answer_arxiv_id": ["1807.09959", "2304.11677"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_9139"} +{"question": "What study introduced GPT-3, an auto-regressive language foundation model?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_9140"} +{"question": "Can you tell me about studies that research the TTHRESH method which compresses data by decomposing it into lower dimensional tensors?", "answer": ["TTHRESH: Tensor Compression for Multidimensional Visual Data"], "answer_arxiv_id": ["1806.05952"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_9141"} +{"question": "Which research considered large ensembles and multiple gradient-step per timestep regimes when learning online?", "answer": ["Offline-to-Online Reinforcement Learning via Balanced Replay and Pessimistic Q-Ensemble"], "answer_arxiv_id": ["2107.00591"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_9142"} +{"question": "Could you name some studies that fine-tuned open-source Large Multimodal Models (LMMs)?", "answer": ["InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "CogVLM: Visual Expert for Pretrained Language Models", "Visual Instruction Tuning"], "answer_arxiv_id": ["2305.06500", "2311.03079", "2304.08485"], "source_meta": {"published_time": "20240501"}, "qid": "AutoScholarQuery_train_9143"} +{"question": "Can you provide me with paper that explains how BYOL prevents representation collapse?", "answer": ["Wasserstein GAN"], "answer_arxiv_id": ["1701.07875"], "source_meta": {"published_time": "20210529"}, "qid": "AutoScholarQuery_train_9144"} +{"question": "Which works have used deep learning, particularly neural networks, in a supervised manner for DFT?", "answer": ["Neural Message Passing for Quantum Chemistry", "Artificial neural networks for density-functional optimizations in fermionic systems", "Accelerating Finite-Temperature Kohn-Sham Density Functional Theory with Deep Neural Networks"], "answer_arxiv_id": ["1704.01212", "1811.03774", "2010.04905"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_9145"} +{"question": "Which studies demonstrate that instruction tuning is a viable remedy in reducing NLP bias metrics?", "answer": ["Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2210.11416"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_9146"} +{"question": "Could you provide me a study that elaborates on the benefits of an explicit cyclic curricula in image classification?", "answer": ["Challenges of Adversarial Image Augmentations"], "answer_arxiv_id": ["2111.12427"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_9147"} +{"question": "What are some of the extensions for the Chain-of-Thought (CoT) concept for reasoning with LLMs?", "answer": ["Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Text and Patterns: For Effective Chain of Thought, It Takes Two to Tango"], "answer_arxiv_id": ["2205.10625", "2205.11916", "2209.07686"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_9148"} +{"question": "What papers conduct theoretical analysis of the properties of self-training and contrastive learning without negative pairs?", "answer": ["Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data", "Understanding Self-Supervised Learning Dynamics without Contrastive Pairs"], "answer_arxiv_id": ["2010.03622", "2102.06810"], "source_meta": {"published_time": "20220202"}, "qid": "AutoScholarQuery_train_9149"} +{"question": "Which research paper interpret graph convolution networks as a solution to the heat diffusion equation?", "answer": ["GRAND: Graph Neural Diffusion"], "answer_arxiv_id": ["2106.10934"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_9150"} +{"question": "In what papers can I see the use of 2D diffusion models to generate multi-view images from single-view input?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object", "One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image", "Zero123++: a Single Image to Consistent Multi-view Diffusion Base Model", "Wonder3D: Single Image to 3D using Cross-Domain Diffusion", "ConsistNet: Enforcing 3D Consistency for Multi-view Images Diffusion"], "answer_arxiv_id": ["2303.11328", "2306.16928", "2309.03453", "2310.15110", "2310.15008", "2310.10343"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_9151"} +{"question": "What research is there on producing multi-view consistent images to create 3D objects from a single input image?", "answer": ["Zero123++: a Single Image to Consistent Multi-view Diffusion Base Model", "Consistent-1-to-3: Consistent Image to 3D View Synthesis via\n Geometry-aware Diffusion Models", "Consistent123: Improve Consistency for One Image to 3D Object Synthesis", "Wonder3D: Single Image to 3D using Cross-Domain Diffusion", "One-2-3-45++: Fast Single Image to 3D Objects with Consistent Multi-View\n Generation and 3D Diffusion"], "answer_arxiv_id": ["2310.15110", "2310.03020", "2310.08092", "2310.15008", "2311.07885"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_9152"} +{"question": "Which works have investigated how successful contrastive vision-language models are at understanding semantic relationships?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "When and why vision-language models behave like bags-of-words, and what\n to do about it?", "Testing Relational Understanding in Text-Guided Image Generation", "Interpretable Diffusion via Information Decomposition"], "answer_arxiv_id": ["2103.00020", "2210.01936", "2208.00005", "2310.07972"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_9153"} +{"question": "What work contributes to modelling the cross-view 3D dependency, turly inspired by video diffusion models?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2308.16512"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_9154"} +{"question": "Which paper originally introduced the concept of absolute positional encoding in transformers?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20210222"}, "qid": "AutoScholarQuery_train_9155"} +{"question": "What works have contributed to the development of point cloud processing networks?", "answer": ["PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "Point Convolutional Neural Networks by Extension Operators", "RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds"], "answer_arxiv_id": ["1706.02413", "1803.10091", "1911.11236"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_9156"} +{"question": "What research works have been recently conducted to adapt Vision Transformer for video classification?", "answer": ["Video Action Transformer Network", "Video Transformer Network", "Token Shift Transformer for Video Classification", "VidTr: Video Transformer Without Convolutions", "Is Space-Time Attention All You Need for Video Understanding?", "ViViT: A Video Vision Transformer", "VLT: Vision-Language Transformer and Query Generation for Referring Segmentation", "Video Swin Transformer", "AIM: Adapting Image Models for Efficient Video Action Recognition"], "answer_arxiv_id": ["1812.02707", "2102.00719", "2108.02432", "2104.11746", "2102.05095", "2103.15691", "2210.15871", "2106.13230", "2302.03024"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_9157"} +{"question": "What works discuss Bag-of-Visual-Word approaches for visual concept learning as a part of part-prototype-based methods?", "answer": ["Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNet", "Learning Representations by Predicting Bags of Visual Words", "Bag of Visual Words (BoVW) with Deep Features - Patch Classification Model for Limited Dataset of Breast Tumours"], "answer_arxiv_id": ["1904.00760", "2002.12247", "2202.10701"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_9158"} +{"question": "Which paper introduces the concept of rotation density operator used in the calculation of vector fields?", "answer": ["Helmholtz Decomposition and Rotation Potentials in n-dimensional Cartesian Coordinates"], "answer_arxiv_id": ["2012.13157v3"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_9159"} +{"question": "Could you provide me some studies that dynamically alter the conflict gradients in the research of multi-task optimization for MNMT?", "answer": ["Gradient Surgery for Multi-Task Learning", "Gradient Vaccine: Investigating and Improving Multi-task Optimization in Massively Multilingual Models"], "answer_arxiv_id": ["2001.06782", "2010.05874"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_9160"} +{"question": "What papers discuss decoding strategies as an approach for mitigating hallucination during inference time?", "answer": ["Controlling Hallucinations at Word Level in Data-to-Text Generation", "DoLa: Decoding by Contrasting Layers Improves Factuality in Large\n Language Models", "Trusting Your Evidence: Hallucinate Less with Context-aware Decoding", "Inference-Time Intervention: Eliciting Truthful Answers from a Language\n Model"], "answer_arxiv_id": ["2102.02810", "2309.03883", "2305.14739", "2306.03341"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_9161"} +{"question": "Which work discusses architecture-specific XAI methods such as GradCAM?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization"], "answer_arxiv_id": ["1610.02391"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_9162"} +{"question": "Can you name the research papers that highlight the deteriorating quality of uncertainty estimates when moving away from in-distribution data?", "answer": ["Plex: Towards Reliability Using Pretrained Large Model Extensions"], "answer_arxiv_id": ["2207.07411"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_9163"} +{"question": "Which works contributed to the development of self-supervised speech models that have been successfully applied in various speech processing tasks?", "answer": ["HuBERT: Self-Supervised Speech Representation Learning by Masked\n Prediction of Hidden Units", "wav2vec: Unsupervised Pre-training for Speech Recognition", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech\n Representations", "W2v-BERT: Combining Contrastive Learning and Masked Language Modeling\n for Self-Supervised Speech Pre-Training", "XLS-R: Self-supervised Cross-lingual Speech Representation Learning at\n Scale"], "answer_arxiv_id": ["2106.07447", "1904.05862", "2006.11477", "2108.06209", "2111.09296"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_9164"} +{"question": "What research introduced techniques for understanding fine-grained video moments and reasoning with respect to time boundaries?", "answer": ["VTimeLLM: Empower LLM to Grasp Video Moments"], "answer_arxiv_id": ["2311.18445"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_9165"} +{"question": "Which works originally alluded to (ℒ𝖼𝗏𝗑,𝒞)-omniprediction in the context of convex losses?", "answer": ["Omnipredictors", "Loss Minimization through the Lens of Outcome Indistinguishability"], "answer_arxiv_id": ["2109.05389", "2210.08649"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_9166"} +{"question": "Can you list some works that shares similarities with the concept of System Identification?", "answer": ["Robots that can adapt like animals", "ADAPTIVE META-LEARNING FOR IDENTIFICATION OF ROVER-TERRAIN DYNAMICS", "Safe Active Dynamics Learning and Control: A Sequential Exploration-Exploitation Framework"], "answer_arxiv_id": ["1407.3501", "2009.10191", "2008.11700"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_9167"} +{"question": "What work suggested adding stochasticity to ODE-like RNNs and showed a generalization bound?", "answer": ["Noisy Recurrent Neural Networks"], "answer_arxiv_id": ["2102.04877"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_9168"} +{"question": "Which study shows that neural representations can be used effectively for dynamics forecasting in the context of dynamical systems and PDEs?", "answer": ["Continuous PDE Dynamics Forecasting with Implicit Neural Representations"], "answer_arxiv_id": ["2209.14855v2"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_9169"} +{"question": "Which papers have implemented graph or subgraph sampling by dropping nodes for different purposes?", "answer": ["Inductive Representation Learning on Large Graphs", "GCC: Graph Contrastive Coding for Graph Neural Network Pre-Training", "DropEdge: Towards Deep Graph Convolutional Networks on Node Classification"], "answer_arxiv_id": ["1706.02216", "2006.09963", "1907.10903"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_9170"} +{"question": "What papers discuss GAN-based image translation models and note the importance of preserving input semantics in the absence of paired images?", "answer": ["Generative Adversarial Networks", "Image-to-Image Translation with Conditional Adversarial Networks", "Contrastive Learning for Unpaired Image-to-Image Translation", "Instance-wise Hard Negative Example Generation for Contrastive Learning\n in Unpaired Image-to-Image Translation", "Exploring Patch-wise Semantic Relation for Contrastive Learning in\n Image-to-Image Translation Tasks", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "Diverse Image-to-Image Translation via Disentangled Representations", "Multimodal Unsupervised Image-to-Image Translation"], "answer_arxiv_id": ["1406.2661", "1611.07004", "2007.15651", "2108.04547", "2203.01532", "1703.10593", "1808.00948", "1804.04732"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_9171"} +{"question": "Which studies are part of the second category of vision-language models built upon the masked language modeling (MLM) pretraining objective?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "UNITER: UNiversal Image-TExt Representation Learning", "Align before Fuse: Vision and Language Representation Learning with Momentum Distillation"], "answer_arxiv_id": ["1908.02265", "1908.07490", "1909.11740", "2107.07651"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_9172"} +{"question": "What works propose methods for coreset selection?", "answer": ["DeepCore: A Comprehensive Library for Coreset Selection in Deep Learning", "Coresets for Data-efficient Training of Machine Learning Models", "Robust Coreset Construction for Distributed Machine Learning", "Coresets via Bilevel Optimization for Continual Learning and Streaming"], "answer_arxiv_id": ["2204.08499", "1906.01827", "1904.05961", "2006.03875"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_9173"} +{"question": "Which works are about generalizing graph convolutions in hyperbolic space for a better graph or temporal graph representation?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "Inductive Representation Learning on Large Graphs", "FeatureNorm: L2 Feature Normalization for Dynamic Graph Embedding", "BSAL: A Framework of Bi-component Structure and Attribute Learning for Link Prediction", "STAR-GCN: Stacked and Reconstructed Graph Convolutional Networks for Recommender Systems"], "answer_arxiv_id": ["1609.02907", "1710.10903", "1706.02216", "2103.00164", "2204.09508", "1905.13129"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9174"} +{"question": "Which work showed that for deep linear networks, the weight matrices asymptotically align to a rank-1 matrix?", "answer": ["Gradient descent aligns the layers of deep linear networks"], "answer_arxiv_id": ["1810.02032"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_9175"} +{"question": "Which papers discuss text-to-image generation?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "Muse: Text-To-Image Generation via Masked Generative Transformers", "Zero-Shot Text-to-Image Generation", "Generating Diverse High-Fidelity Images with VQ-VAE-2", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2112.10752", "2301.00704", "2102.12092", "1906.00446", "1809.11096", "1812.04948"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_9176"} +{"question": "Which study revealed that UDR has a strong positive bias for low-dimensional representations?", "answer": ["Robust Disentanglement of a Few Factors at a Time"], "answer_arxiv_id": ["2010.13527"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_9177"} +{"question": "Which work proposed the use of a second-order statistics-based correlative mutual information measure for the BSS problem?", "answer": ["AN INFORMATION MAXIMIZATION BASED BLIND SOURCE SEPARATION APPROACH FOR DEPENDENT AND INDEPENDENT SOURCES"], "answer_arxiv_id": ["2205.00794"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_9178"} +{"question": "What papers used implicit functions, such as neural networks, as object representations in volume rendering?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "Local Deep Implicit Functions for 3D Shape"], "answer_arxiv_id": ["2003.08934", "1912.07372", "1812.03828", "1912.06126"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_9179"} +{"question": "In what papers can I find methods that solve a subclass of constrained Variation Inequalities (cVI), specifically constrained zero-sum problems?", "answer": ["Frank-Wolfe Algorithms for Saddle Point Problems", "On the Global Linear Convergence of Frank-Wolfe Optimization Variants"], "answer_arxiv_id": ["1610.07797", "1511.05932"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_9180"} +{"question": "What works highlight attempts on breaking the defenses of randomized ensembles?", "answer": ["Adversarial Vulnerability of Randomized Ensembles", "Building Robust Ensembles via Margin Boosting"], "answer_arxiv_id": ["2206.06737", "2206.03362"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9181"} +{"question": "What studies exist about the established baseline algorithms for safe reinforcement learning, specifically constrained policy optimization and trust region policy optimization?", "answer": ["Constrained Policy Optimization", "Trust Region Policy Optimization"], "answer_arxiv_id": ["1705.10528", "1502.05477"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_9182"} +{"question": "What research applied NeRF on urban dynamic scenes and introduced the Entity-wise Average of Residual Ranks?", "answer": ["RobustNeRF: Ignoring Distractors with Robust Losses"], "answer_arxiv_id": ["2302.00833"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_9183"} +{"question": "What papers introduced permutation-based autoregressive model to bridge the gap between autoregressive language modeling and masked autoencoding?", "answer": ["XLNet: Generalized Autoregressive Pretraining for Language Understanding"], "answer_arxiv_id": ["1906.08237"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_9184"} +{"question": "Which works do the disagreement coefficient based algorithms explored in the research belong to?", "answer": ["Minimax Analysis of Active Learning"], "answer_arxiv_id": ["1410.0996"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_9185"} +{"question": "Which works in exemplar-based image generation capture exemplars’ global styles through an encoder?", "answer": ["Semantic Image Synthesis with Spatially-Adaptive Normalization"], "answer_arxiv_id": ["1903.07291"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_9186"} +{"question": "What works have leveraged structural assumptions on coupling and cost matrices in order to reduce computational and memory complexity for estimating OT maps?", "answer": ["Computing f-Divergences and Distances of High-Dimensional Probability Density Functions — Low-Rank Tensor Approximations —", "Low-Rank Sinkhorn Factorization", "Statistical Optimal Transport via Factored Couplings", "Optimal Transport: Fast Probabilistic Approximation with Exact Solvers", "Linear-Time Gromov Wasserstein Distances using Low Rank Couplings and Costs", "Subspace Detours: Building Transport Plans that are Optimal on Subspace Projections", "Subspace Detours Meet Gromov-Wasserstein"], "answer_arxiv_id": ["2111.07164", "2103.04737", "1806.07348", "1802.05570", "2106.01128", "1905.10099", "2110.10932"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_9187"} +{"question": "What studies use bi-directional transformers for RL?", "answer": ["Masked Autoencoding for Scalable and Generalizable Decision Making", "Uni​[MASK]: Unified Inference in Sequential Decision Problems"], "answer_arxiv_id": ["2211.12740", "2211.10869"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_9188"} +{"question": "Could you provide me some research papers about federated learning used for adapting models to a specific individual or object?", "answer": ["Three Approaches for Personalization with Applications to Federated Learning", "Improving Federated Learning Personalization via Model Agnostic Meta Learning", "Personalized Federated Learning: A Meta-Learning Approach", "Personalized Federated Learning using Hypernetworks"], "answer_arxiv_id": ["2002.10619", "1909.12488", "2002.07948", "2103.04628"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_9189"} +{"question": "What studies considered dot-product attention as approximation of integral transform with non-symmetric learnable kernel function?", "answer": ["Choose a Transformer: Fourier or Galerkin", "Transformers are Deep Infinite-Dimensional Non-Mercer Binary Kernel Machines", "Neural Operator: Learning Maps Between Function Spaces", "Transformer Meets Boundary Value Inverse Problems"], "answer_arxiv_id": ["2105.14995", "2106.01506", "2108.08481v6", "2209.14977"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_9190"} +{"question": "Could you provide me papers where authors introduced an affine transformation to make the log-barrier as a self-concordant barrier of constraint set?", "answer": ["Damped Online Newton Step for Portfolio Selection", "Efficient and Near-Optimal Online Portfolio Selection"], "answer_arxiv_id": ["2202.07574v1", "2209.13932v1"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9191"} +{"question": "What are some representative works that employ the frequency domain for deepfake detection?", "answer": ["Watch your Up-Convolution: CNN Based Generative Deep Neural Networks are Failing to Reproduce Spectral Distributions", "Thinking in Frequency: Face Forgery Detection by Mining Frequency-aware Clues", "Spatial-Phase Shallow Learning: Rethinking Face Forgery Detection in Frequency Domain", "Generalizing Face Forgery Detection with High-frequency Features"], "answer_arxiv_id": ["2003.01826", "2007.09355", "2103.01856", "2103.12376"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_9192"} +{"question": "Could you provide me some examples of studies that use Domain Randomization (DR) in domain generalized semantic segmentation?", "answer": ["Source-Free Open Compound Domain Adaptation in Semantic Segmentation", "Style-Hallucinated Dual Consistency Learning for Domain Generalized\n Semantic Segmentation", "Domain Generalization via Balancing Training Difficulty and Model\n Capability"], "answer_arxiv_id": ["2106.03422", "2204.02548", "2309.00844"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_9193"} +{"question": "Can you suggest some papers about the fine-tuning of LLaMA-2 models on interaction trajectories?", "answer": ["AgentTuning: Enabling Generalized Agent Abilities for LLMs"], "answer_arxiv_id": ["2310.12823v2"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_9194"} +{"question": "Could you provide me the works that train multi-modal models from scratch for Visual Question Answering Systems?", "answer": ["LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "UNITER: UNiversal Image-TExt Representation Learning", "Large-Scale Adversarial Training for Vision-and-Language Representation Learning", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "VinVL: Revisiting Visual Representations in Vision-Language Models", "UNIMO: Towards Unified-Modal Understanding and Generation via Cross-Modal Contrastive Learning"], "answer_arxiv_id": ["1908.07490", "1909.11740", "2006.06195", "2004.06165", "2108.10904", "2101.00529", "2012.15409"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_9195"} +{"question": "Which study utilized an instruction-tuned GPT-3 model aligned with the intention of human users?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_9196"} +{"question": "Could you provide me some studies about unsupervised depth completion methods?", "answer": ["DFuseNet: Deep Fusion of RGB and Sparse Depth Information for Image\n Guided Dense Depth Completion", "Self-supervised Sparse-to-Dense: Self-supervised Depth Completion from\n LiDAR and Monocular Camera", "Learning Topology from Synthetic Data for Unsupervised Depth Completion", "An Adaptive Framework for Learning Unsupervised Depth Completion", "Unsupervised Depth Completion from Visual Inertial Odometry"], "answer_arxiv_id": ["1902.00761", "1807.00275", "2106.02994", "2106.03010", "1905.08616"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_9197"} +{"question": "Which work employs sparse-causal attention for temporal consistency along with latent guidance?", "answer": ["Pix2Video: Video Editing using Image Diffusion"], "answer_arxiv_id": ["2303.12688"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9198"} +{"question": "What research is about exploring enhancing autoregressive code generation by utilizing the model's infilling capability?", "answer": ["Self-Infilling Code Generation"], "answer_arxiv_id": ["2311.17972v3"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_9199"} +{"question": "What work trains LLMs to generate scores for responses through principle-driven synthetic preference data utilizing the SFT model?", "answer": ["SALMON: Self-Alignment with Instructable Reward Models"], "answer_arxiv_id": ["2310.05910"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_9200"} +{"question": "From which research does Greedy Rejection Coding draw inspiration?", "answer": ["Fast Relative Entropy Coding with A* coding"], "answer_arxiv_id": ["2201.12857"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_9201"} +{"question": "Which papers talk about Transformer backbone and its scalability?", "answer": ["Scaling Laws for Neural Language Models", "Scaling Vision Transformers to 22 Billion Parameters"], "answer_arxiv_id": ["2001.08361", "2302.05442"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_9202"} +{"question": "Could you tell me the references that used ratings on a three-point ordinal scale in topic rating task?", "answer": ["Is Automated Topic Model Evaluation Broken?: The Incoherence of\n Coherence"], "answer_arxiv_id": ["2107.02173"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_9203"} +{"question": "Which paper uses ordinary differential equations to re-sample the feature point from the Lyapunov-stable equilibrium points in adversarial training?", "answer": ["Stable Neural ODE with Lyapunov-Stable Equilibrium Points for Defending\n Against Adversarial Attacks"], "answer_arxiv_id": ["2110.12976"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9204"} +{"question": "Which work reported improvements in segmentation when a U-Net is trained using their synthetic data versus real images?", "answer": ["Fake It Till You Make It: Face analysis in the wild using synthetic data\n alone"], "answer_arxiv_id": ["2109.15102"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_9205"} +{"question": "What study proposed a dynamic weighting mechanism to depict illumination conditions of scenes in multimodal fusion?", "answer": ["Fusion of Multispectral Data Through Illumination-aware Deep Neural Networks for Pedestrian Detection"], "answer_arxiv_id": ["1802.09972v1"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_9206"} +{"question": "Which studies developed methods to find primitives in unstructured 3D scenes?", "answer": ["MonteBoxFinder: Detecting and Filtering Primitives to Fit a Noisy Point\n Cloud", "Robust and Accurate Superquadric Recovery: a Probabilistic Approach", "Primitive-based Shape Abstraction via Nonparametric Bayesian Inference"], "answer_arxiv_id": ["2207.14268", "2111.14517", "2203.14714"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_9207"} +{"question": "Which work described topic intrusion task?", "answer": ["Is Automated Topic Model Evaluation Broken?: The Incoherence of\n Coherence"], "answer_arxiv_id": ["2107.02173"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_9208"} +{"question": "In which study did the researchers achieve acceleration in rendering speed by using multi-resolution hash encodings (MHE) with shared shallow MLP networks?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9209"} +{"question": "Can you cite papers discussing the ability of LLMs to mimic human cognition?", "answer": ["Large Pre-trained Language Models Contain Human-like Biases of What is\n Right and Wrong to Do", "Using cognitive psychology to understand GPT-3"], "answer_arxiv_id": ["2103.11790", "2206.14576"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_9210"} +{"question": "Which works are about learning representations of graph-structured data?", "answer": ["Geometric deep learning: going beyond Euclidean data"], "answer_arxiv_id": ["1611.08097v2"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_9211"} +{"question": "What works concentrated on application of the mean-reverting SDE in speech enhancement and speech dereverberation?", "answer": ["Speech Enhancement with Score-Based Generative Models in the Complex STFT Domain", "Speech Enhancement and Dereverberation with Diffusion-based Generative Models"], "answer_arxiv_id": ["2203.17004", "2208.05830"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_9212"} +{"question": "Could you provide the studies that proposed real-world lifelong object learning datasets captured from robotic vision systems?", "answer": ["F-SIOL-310: A Robotic Dataset and Benchmark for Few-Shot Incremental Object Learning", "OpenLORIS-Object: A Robotic Vision Dataset and Benchmark for Lifelong Deep Learning"], "answer_arxiv_id": ["2103.12242", "1911.06487"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_9213"} +{"question": "Which studies introduced an additional loss to penalize inconsistent predictions in neural networks?", "answer": ["A Semantic Loss Function for Deep Learning with Symbolic Knowledge", "Neuro-Symbolic Entropy Regularization"], "answer_arxiv_id": ["1711.11157", "2201.11250v1"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9214"} +{"question": "What papers address autoregression and diffusion in their studies?", "answer": ["Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting"], "answer_arxiv_id": ["2101.12072"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_9215"} +{"question": "What are the research papers that discuss application of stop gradients to Mixture VAEs?", "answer": ["Sticking the Landing: Simple, Lower-Variance Gradient Estimators for Variational Inference"], "answer_arxiv_id": ["1703.09194"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_9216"} +{"question": "Which works developed depth estimators using diffusion-based models?", "answer": ["Unleashing Text-to-Image Diffusion Models for Visual Perception", "Text-image Alignment for Diffusion-based Perception", "EVP: Enhanced Visual Perception using Inverse Multi-Attentive Feature\n Refinement and Regularized Image-Text Alignment"], "answer_arxiv_id": ["2303.02153v1", "2310.00031", "2312.08548"], "source_meta": {"published_time": "20240412"}, "qid": "AutoScholarQuery_train_9217"} +{"question": "What research explored the ability of large language models (LLMs) to mimic human behaviors like reasoning and cognitive tests?", "answer": ["Emergent Analogical Reasoning in Large Language Models", "Using cognitive psychology to understand GPT-3", "Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studies", "From Word Models to World Models: Translating from Natural Language to the Probabilistic Language of Thought"], "answer_arxiv_id": ["2212.09196", "2206.14576", "2208.10264", "2306.12672v2"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_9218"} +{"question": "What research papers discuss skippable networks that can bypass several layers at runtime?", "answer": ["Dual Dynamic Inference: Enabling More Efficient, Adaptive, and Controllable Deep Inference", "FractalNet: Ultra-Deep Neural Networks without Residuals", "BlockDrop: Dynamic Inference Paths in Residual Networks"], "answer_arxiv_id": ["1907.04523", "1605.07648", "1711.08393"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_9219"} +{"question": "Which works engaged in expanding the window size of large language models (LLMs) through staged pre-training?", "answer": ["XGen-7B Technical Report"], "answer_arxiv_id": ["2309.03450"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_9220"} +{"question": "What paper proved the last-iterate convergence rate for any ρ<1/2L, which is a larger range of ρ than in the known results?", "answer": ["Fast Extra Gradient Methods for Smooth Structured Nonconvex-Nonconcave Minimax Problems"], "answer_arxiv_id": ["2106.02326"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_9221"} +{"question": "Which works use customized special tokens to accommodate representations beyond pure text, such as bounding boxes?", "answer": ["Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "Pix2seq: A Language Modeling Framework for Object Detection", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework"], "answer_arxiv_id": ["2206.08916", "2109.10852", "2202.03052"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_9222"} +{"question": "What work discusses the TAPIR algorithm, which combines TAPNet and PIPs strategies?", "answer": ["TAPIR: Tracking Any Point with per-frame Initialization and temporal\n Refinement"], "answer_arxiv_id": ["2306.08637"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_9223"} +{"question": "Could you provide me with some works that have attempted to address the error amplification from temporal dependency?", "answer": ["Learning Near Optimal Policies with Low Inherent Bellman Error", "Bellman-consistent Pessimism for Offline Reinforcement Learning", "Pessimistic Model-based Offline Reinforcement Learning under Partial Coverage"], "answer_arxiv_id": ["2003.00153", "2106.06926", "2107.06226"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_9224"} +{"question": "What papers imply hardness in learning DNF formulas?", "answer": ["Complexity theoretic limitations on learning DNF’s"], "answer_arxiv_id": ["1404.3378"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_9225"} +{"question": "Which papers proposed unified 2D segmentation methods?", "answer": ["3D Instances as 1D Kernels", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2207.07372", "2107.06278", "2112.01527"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_9226"} +{"question": "Which papers propose pretraining options for categorical EHR data?", "answer": ["BEHRT: Transformer for Electronic Health Records", "Language Models Are An Effective Representation Learning Technique For Electronic Health Record Data"], "answer_arxiv_id": ["1907.09538", "2001.05295"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_9227"} +{"question": "What studies propose training LLMs in a supervised multitask fashion?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization", "Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2110.08207", "2109.01652"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_9228"} +{"question": "Could you list some works about descriptor learning in the context of point cloud registration?", "answer": ["CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud\n Registration", "PREDATOR: Registration of 3D Point Clouds with Low Overlap", "Lepard: Learning partial point cloud matching in rigid and deformable\n scenes", "Geometric Transformer for Fast and Robust Point Cloud Registration"], "answer_arxiv_id": ["2110.14076", "2011.13005", "2111.12591", "2202.06688"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_9229"} +{"question": "What papers have outlines methods which train a BiLSTM network to output continuous prompt embeddings?", "answer": ["GPT Understands, Too"], "answer_arxiv_id": ["2103.10385"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_9230"} +{"question": "What papers discuss the advancements and benchmarking performance of models like GPT-4 and Gemini?", "answer": ["GPT-4 Technical Report", "Gemini: A Family of Highly Capable Multimodal Models"], "answer_arxiv_id": ["2303.08774", "2312.11805v4"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_9231"} +{"question": "What works have explored 3D representation methods like voxel and point cloud along with different generation models?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "Learning Descriptor Networks for 3D Shape Synthesis and Analysis", "Sketch and Text Guided Diffusion Model for Colored Point Cloud\n Generation", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "Diffusion-SDF: Text-to-Shape via Voxelized Diffusion"], "answer_arxiv_id": ["1610.07584", "1804.00586", "2308.02874", "2212.08751", "1906.12320", "2212.03293"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9232"} +{"question": "Which research paper discusses the practical popularity of optimizing a weighted version of the MMI objective?", "answer": ["A Diversity-Promoting Objective Function for Neural Conversation Models"], "answer_arxiv_id": ["1510.03055"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_9233"} +{"question": "What research presents BabelCode, a framework for execution-based evaluation, and investigates the effectiveness of language distribution balancing in a training dataset?", "answer": ["Measuring The Impact Of Programming Language Distribution"], "answer_arxiv_id": ["2302.01973"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_9234"} +{"question": "What papers offer a DP parameter learning method for GMMs with unknown mixing weights, means, and covariance matrices?", "answer": ["Differentially Private Algorithms for Learning Mixtures of Separated Gaussians"], "answer_arxiv_id": ["1909.03951"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_9235"} +{"question": "Which works are devoted to the inversion and reconstruction problem in GAN inversion through learning an additional deterministic encoder?", "answer": ["Generative Visual Manipulation on the Natural Image Manifold", "Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation", "E2Style: Improve the Efficiency and Effectiveness of StyleGAN Inversion", "ReStyle: A Residual-Based StyleGAN Encoder via Iterative Refinement"], "answer_arxiv_id": ["1609.03552", "2008.00951", "2104.07661", "2104.02699"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_9236"} +{"question": "Which work achieved state-of-the-art performance in SQL prediction by fine-tuning a T5-3B model and using constrained decoding?", "answer": ["Picard: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models"], "answer_arxiv_id": ["2109.05093"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_9237"} +{"question": "Which studies have advanced in image-language pre-trained models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "PaLM-E: An Embodied Multimodal Language Model", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework", "UNITER: UNiversal Image-TExt Representation Learning", "Hero: Hierarchical Encoder for Video+Language Omni-representation Pre-training", "LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "Less is More: ClipBERT for Video-and-Language Learning via Sparse Sampling", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "FILIP: Fine-grained Interactive Language-Image Pre-Training"], "answer_arxiv_id": ["2103.00020", "2303.03378v1", "2202.03052", "1909.11740", "2005.00200", "1908.07490", "2102.06183", "2201.12086", "2111.07783"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_9238"} +{"question": "What works demonstrate the generation of high-quality samples through the correlation of adjacent pixels in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239", "2102.09672"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_9239"} +{"question": "Could you provide some works of deep-learning-based NVS methods that require a pre-computed depth map?", "answer": ["One Shot 3D Photography", "3D Ken Burns Effect from a Single Image", "Single-View View Synthesis in the Wild with Learned Adaptive Multiplane\n Images"], "answer_arxiv_id": ["2008.12298", "1909.05483", "2205.11733"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_9240"} +{"question": "Can you give examples of research that explored open-vocabulary 3D scene segmentation by encoding 2D open-vocabulary models’ features into 3D scene points?", "answer": ["OpenScene: 3D Scene Understanding with Open Vocabularies", "Feature-Realistic Neural Fusion for Real-Time, Open Set Scene Understanding", "Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Models", "ConceptFusion: Open-set Multimodal 3D Mapping", "CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP"], "answer_arxiv_id": ["2211.15654", "2210.03043", "2207.11514", "2302.07241", "2301.04926"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9241"} +{"question": "Which paper provides proof that FRW gradient descent doesn't get stuck at local minima in an idealized setting with infinite particles?", "answer": ["Learning Gaussian Mixtures Using the Wasserstein-Fisher-Rao Gradient Flow"], "answer_arxiv_id": ["2301.01766"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_9242"} +{"question": "What papers focus on physics-based polarimetric 3D methods?", "answer": ["Perspective Phase Angle Model for Polarimetric 3D Reconstruction"], "answer_arxiv_id": ["2207.09629"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_9243"} +{"question": "Which articles demonstrate the use of in-context learning in classification and question answering?", "answer": ["SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems"], "answer_arxiv_id": ["1905.00537"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_9244"} +{"question": "Which papers used the Alpaca-7B model to collect prompts for safety alignment research?", "answer": ["Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors,\n and Lessons Learned", "BeaverTails: Towards Improved Safety Alignment of LLM via a\n Human-Preference Dataset"], "answer_arxiv_id": ["2209.07858", "2307.04657"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_9245"} +{"question": "What research applied equivariant learning to images?", "answer": ["General E(2) - Equivariant Steerable CNNs"], "answer_arxiv_id": ["1911.08251"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_9246"} +{"question": "What papers highlighted that additive Gaussian noise improves sample quality in autoregressive models?", "answer": ["Improved Autoregressive Modeling with Distribution Smoothing"], "answer_arxiv_id": ["2103.15089"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_9247"} +{"question": "What papers introduced mechanisms or models that used social clues for multi-person trajectory projection?", "answer": ["TRiPOD: Human Trajectory and Pose Dynamics Forecasting in the Wild", "Multi-Person Extreme Motion Prediction", "SoMoFormer: Multi-Person Pose Forecasting with Transformers", "SoMoFormer: Social-Aware Motion Transformer for Multi-Person Motion Prediction", "Stochastic Multi-Person 3D Motion Forecasting"], "answer_arxiv_id": ["2104.04029", "2105.08825", "2208.14023", "2208.09224", "2306.05421"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_9248"} +{"question": "Could you provide me some works on MoE modeling showing significantly reduced training energy and computation cost?", "answer": ["DEMix Layers: Disentangling Domains for Modular Language Modeling"], "answer_arxiv_id": ["2108.05036"], "source_meta": {"published_time": "20230520"}, "qid": "AutoScholarQuery_train_9249"} +{"question": "What works discuss GAN-based methods for controllable generation?", "answer": ["Multimodal Conditional Image Synthesis with Product-of-Experts GANs"], "answer_arxiv_id": ["2112.05130"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_9250"} +{"question": "Who developed the method of exemplar learning for image generation?", "answer": ["In-Context Learning Unlocked for Diffusion Models"], "answer_arxiv_id": ["2305.01115"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9251"} +{"question": "Which studies augment language models with external knowledge during fine-tuning, but also require changes in model architectures or additional training steps on each task and dataset?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "KALA: Knowledge-Augmented Language Model Adaptation", "Large Language Models with Controllable Working Memory", "Atlas: Few-shot Learning with Retrieval Augmented Language Models", "Dialogue Chain-of-Thought Distillation for Commonsense-aware\n Conversational Agents"], "answer_arxiv_id": ["2005.11401", "2204.10555", "2211.05110", "2208.03299", "2310.09343"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_9252"} +{"question": "Could you mention studies about gradient-based methods for crafting or learning signed-distance fields of obstacles?", "answer": ["Newton methods for k-order Markov Constrained Motion Problems", "Continuous-time Gaussian process motion planning via probabilistic inference"], "answer_arxiv_id": ["1407.0414v1", "1707.07383"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_9253"} +{"question": "Which works discuss the teacher-student methodology in the context of self-supervised learning?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2006.07733", "2011.10566"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_9254"} +{"question": "Which works considers Denoising Diffusion Implicit Models or DDIMs?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20220316"}, "qid": "AutoScholarQuery_train_9255"} +{"question": "Could you provide studies that propose various approximation techniques for Bayesian Neural Networks?", "answer": ["Variational Inference: A Review for Statisticians", "Probabilistic Backpropagation for Scalable Learning of Bayesian Neural\n Networks", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in\n Deep Learning", "Uncertainty-guided Source-free Domain Adaptation"], "answer_arxiv_id": ["1601.00670", "1502.05336", "1506.02142", "2208.07591v1"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_9256"} +{"question": "In what research is the concept of 'discount regularization' introduced?", "answer": ["Discount Factor as a Regularizer in Reinforcement Learning"], "answer_arxiv_id": ["2007.02040v1"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_9257"} +{"question": "What study found that deduplicating pre-training data had no clear benefit on language modeling performance?", "answer": ["Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling"], "answer_arxiv_id": ["2304.01373"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_9258"} +{"question": "What works showcase predicting the dereverberated signal from an audio recording and a panoramic image of the recording environment?", "answer": ["Learning Audio-Visual Dereverberation"], "answer_arxiv_id": ["2106.07732"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_9259"} +{"question": "Which works propose the use of neural representations for generating shapes using diffusion models?", "answer": ["HyperDiffusion: Generating Implicit Neural Fields with Weight-Space\n Diffusion"], "answer_arxiv_id": ["2303.17015"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_9260"} +{"question": "Are there any works that utilize distilling diffusion models as a method?", "answer": ["Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2202.00512"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_9261"} +{"question": "Which works generated 3D data in voxel representation?", "answer": ["Learning a Predictable and Generative Vector Representation for Objects", "MarrNet: 3D Shape Reconstruction via 2.5D Sketches", "3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction", "Pix2Vox++: Multi-scale Context-aware 3D Object Reconstruction from Single and Multiple Images", "Pix2Vox: Context-aware 3D Reconstruction from Single and Multi-view Images", "LegoFormer: Transformers for Block-by-Block Multi-view 3D Reconstruction"], "answer_arxiv_id": ["1603.08637", "1711.03129", "1604.00449v1", "2006.12250", "1901.11153", "2106.12102"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_9262"} +{"question": "What studies have looked into the principle of equal opportunity in fairness?", "answer": ["Equality of Opportunity in Supervised Learning"], "answer_arxiv_id": ["1610.02413"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_9263"} +{"question": "What works have been cited as examples of practical applications in Generalized Inductive Learning (GIL)?", "answer": ["Lifelong Graph Learning", "GraphSAIL: Graph Structure Aware Incremental Learning for Recommender Systems", "Continual Learning of Knowledge Graph Embeddings", "Disentangle-based Continual Graph Representation Learning", "Overcoming Catastrophic Forgetting in Graph Neural Networks", "Hierarchical Prototype Networks for Continual Graph Representation Learning", "Overcoming Catastrophic Forgetting in Graph Neural Networks with Experience Replay", "Catastrophic Forgetting in Deep Graph Networks: an Introductory Benchmark for Graph Classification", "Graph Few-shot Class-incremental Learning"], "answer_arxiv_id": ["2009.00647", "2008.13517", "2101.05850", "2010.02565", "2012.06002", "2111.15422", "2003.09908", "2103.11750", "2112.12819"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_9264"} +{"question": "Which works have trained multimodal language models on multi-task objectives?", "answer": ["Unified-IO: A unified model for vision, language, and multi-modal tasks", "CoBIT: A Contrastive Bi-directional Image-Text Generation Model", "GlueGen: Plug and Play Multi-modal Encoders for X-to-image Generation"], "answer_arxiv_id": ["2206.08916", "2303.13455", "2303.10056"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_9265"} +{"question": "What works aims to create tailored models for generalizable customization by training a multimodal encoder and a text-to-image model on dataset-scale images?", "answer": ["BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "Cones: Concept Neurons in Diffusion Models for Customized Generation", "Improving Expressivity of GNNs with Subgraph-specific Factor Embedded\n Normalization"], "answer_arxiv_id": ["2305.14720", "2307.06949", "2302.12228", "2302.13848", "2303.05125", "2305.19903"], "source_meta": {"published_time": "20240522"}, "qid": "AutoScholarQuery_train_9266"} +{"question": "Which papers discuss the optimization of the Shampoo method?", "answer": ["Shampoo: Preconditioned Stochastic Tensor Optimization", "Scalable Second Order Optimization for Deep Learning"], "answer_arxiv_id": ["1802.09568", "2002.09018"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_9267"} +{"question": "Which work proposed kernel-based conditional independence test?", "answer": ["Kernel-based Conditional Independence Test and Application in Causal Discovery"], "answer_arxiv_id": ["1202.3775"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_9268"} +{"question": "Which studies employed adapters for architecture fine-tuning of VLMs?", "answer": ["LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Cheap and Quick: Efficient Vision-Language Instruction Tuning for Large\n Language Models"], "answer_arxiv_id": ["2304.15010", "2305.15023"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_9269"} +{"question": "What work introduces a multimodal encoder for better subject representation?", "answer": ["BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing"], "answer_arxiv_id": ["2305.14720"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_9270"} +{"question": "Do any studies learn a token-level error predictor for machine translation?", "answer": ["Reward Gaming in Conditional Text Generation"], "answer_arxiv_id": ["2211.08714"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9271"} +{"question": "What works propose a neural-style field network to predict the color and displacement of mesh vertices in 3D content generation or editing?", "answer": ["Text2Mesh: Text-Driven Neural Stylization for Meshes"], "answer_arxiv_id": ["2112.03221"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9272"} +{"question": "What works are about joint training of speech and text models?", "answer": ["SLAM: A Unified Encoder for Speech and Language Modeling via Speech-Text Joint Pre-Training", "Mu2SLAM: Multitask, Multilingual Speech and Language Models", "mSLAM: Massively multilingual joint pre-training for speech and text", "Maestro-U: Leveraging joint speech-text representation learning for zero supervised speech ASR", "SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing"], "answer_arxiv_id": ["2110.10329", "2212.09553", "2202.01374", "2210.10027", "2110.07205"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_9273"} +{"question": "Could you provide me references on solving the acquisition function optimization problem with evolutionary algorithms?", "answer": ["Neural Architecture Search with Bayesian Optimisation and Optimal Transport"], "answer_arxiv_id": ["1802.07191"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_9274"} +{"question": "Which papers apply 3D cost volumes in the multi-view stereo (MVS)?", "answer": ["Recurrent MVSNet for High-resolution Multi-view Stereo Depth Inference", "DPSNet: End-to-end Deep Plane Sweep Stereo", "Deep Stereo using Adaptive Thin Volume Representation with Uncertainty\n Awareness", "Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo\n Matching", "Learning Inverse Depth Regression for Multi-View Stereo with Correlation\n Cost Volume", "Cost Volume Pyramid Based Depth Inference for Multi-View Stereo"], "answer_arxiv_id": ["1902.10556", "1905.00538", "1911.12012", "1912.06378", "1912.11746", "1912.08329"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_9275"} +{"question": "Can you list some researches that focus on retrieval from character experiences?", "answer": ["ChatHaruhi: Reviving Anime Character in Reality via Large Language Model"], "answer_arxiv_id": ["2308.09597"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_9276"} +{"question": "Can you name some studies that used adversarial filtering for dataset de-biasing and more faithful model evaluations?", "answer": ["Swag: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference", "WinoGrande: An Adversarial Winograd Schema Challenge at Scale", "Adversarial Filters of Dataset Biases"], "answer_arxiv_id": ["1808.05326", "1907.10641", "2002.04108"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_9277"} +{"question": "Which papers have made significant contributions to building powerful visual perception systems?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Segment Anything", "Emerging Properties in Self-Supervised Vision Transformers", "DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2103.00020", "2401.14159", "2104.14294", "2304.07193"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9278"} +{"question": "Which papers present research on example selection in improving in-context learning with large language models?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?", "Self-Adaptive In-Context Learning: An Information Compression\n Perspective for In-Context Example Selection and Ordering"], "answer_arxiv_id": ["2101.06804", "2212.10375"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_9279"} +{"question": "Which papers discuss the axial attention designed to reduce the computational complexity of original global self-attention?", "answer": ["CCNet: Criss-Cross Attention for Semantic Segmentation", "Axial Attention In Multidimensional Transformers", "Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation"], "answer_arxiv_id": ["1811.11721", "1912.12180", "2003.07853"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9280"} +{"question": "Which work introduced a basic molecular GP library in the GPflow framework?", "answer": ["Gaussian Process Molecule Property Prediction with FlowMO"], "answer_arxiv_id": ["2010.01118v2"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_9281"} +{"question": "What studies proposed dual-encoder architecture in vision-language pre-training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9282"} +{"question": "What paper proposed the fixed Sparse Transformer?", "answer": ["Generating Long Sequences with Sparse Transformers"], "answer_arxiv_id": ["1904.10509"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_9283"} +{"question": "What works emphasize the challenges of training RL agents that can generalize to new environments and tasks?", "answer": ["Towards Generalization and Simplicity in Continuous Control", "Revisiting the Arcade Learning Environment: Evaluation Protocols and Open Problems for General Agents", "Illuminating Generalization in Deep Reinforcement Learning through Procedural Level Generation", "Assessing Generalization in Deep Reinforcement Learning", "A Dissection of Overfitting and Generalization in Continuous Reinforcement Learning", "A Study on Overfitting in Deep Reinforcement Learning", "Gotta Learn Fast: A New Benchmark for Generalization in RL", "Quantifying Generalization in Reinforcement Learning", "Leveraging Procedural Generation to Benchmark Reinforcement Learning", "Obstacle Tower: A Generalization Challenge in Vision, Control, and Planning", "The NetHack Learning Environment", "Measuring Visual Generalization in Continuous Control from Pixels", "Reinforcement Learning Generalization with Surprise Minimization", "Interference and Generalization in Temporal Difference Learning", "Instance-based Generalization in Reinforcement Learning", "Why Generalization in RL is Difficult: Epistemic POMDPs and Implicit Partial Observability", "Learning Dynamics and Generalization in Reinforcement Learning"], "answer_arxiv_id": ["1703.02660", "1709.06009", "1806.10729", "1810.12282", "1806.07937", "1804.06893v2", "1804.03720", "1812.02341", "1912.01588", "1902.01378", "2006.13760", "2010.06740", "2004.12399", "2003.06350", "2011.01089", "2107.06277", "2206.02126"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_9284"} +{"question": "Which papers generate noise to optimize the error during deep neural network poisoning?", "answer": ["Unlearnable Examples: Making Personal Data Unexploitable", "Adversarial Examples Make Strong Poisons"], "answer_arxiv_id": ["2101.04898", "2106.10807"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9285"} +{"question": "Which paper presented TRAK to enhance the efficiency of Datamodel?", "answer": ["TRAK: Attributing Model Behavior at Scale"], "answer_arxiv_id": ["2303.14186v2"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_9286"} +{"question": "What work introduced bidirectional flow embedding layers in the coarse-to-fine scheme?", "answer": ["Bi-PointFlowNet: Bidirectional Learning for Point Cloud Based Scene Flow Estimation"], "answer_arxiv_id": ["2207.07522"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_9287"} +{"question": "Which works have employed KL divergence loss and L2 normalization to address the catastrophic forgetting in Continual Learning?", "answer": ["A continual learning survey: Defying forgetting in classification tasks"], "answer_arxiv_id": ["1909.08383"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_9288"} +{"question": "What papers discussed the use of an adaptor to guide each other inside the encoder for model generalization?", "answer": ["MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2210.03117"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_9289"} +{"question": "What are the papers focusing on instance matting?", "answer": ["Human Instance Matting via Mutual Guidance and Multi-Instance Refinement"], "answer_arxiv_id": ["2205.10767"], "source_meta": {"published_time": "20240424"}, "qid": "AutoScholarQuery_train_9290"} +{"question": "In what paper the test cases were generated to verify the solutions rendered by LLMs?", "answer": ["CodeT: Code Generation with Generated Tests"], "answer_arxiv_id": ["2207.10397"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_9291"} +{"question": "Which studies incorporated the adjoint method into NN training?", "answer": ["Neural Ordinary Differential Equations", "ANODEV2: A Coupled Neural ODE Evolution Framework", "Learning Continuous Models for Continuous Physics"], "answer_arxiv_id": ["1806.07366", "1906.04596", "2202.08494"], "source_meta": {"published_time": "20220718"}, "qid": "AutoScholarQuery_train_9292"} +{"question": "What works proposed methods for improving the structure of latent space activations in supervised setting?", "answer": ["von Mises–Fisher Loss: An Exploration of Embedding Geometries for Supervised Learning"], "answer_arxiv_id": ["2103.15718"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_9293"} +{"question": "Could you provide me studies about 3DGS rendering?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_9294"} +{"question": "Could you provide me some studies that represent parametric calibration methods?", "answer": ["On Calibration of Modern Neural Networks", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration"], "answer_arxiv_id": ["1706.04599", "1910.12656"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_9295"} +{"question": "Which studies adopted parameter-efficient fine-tuning (PEFT) for video generation tasks?", "answer": ["SimDA: Simple Diffusion Adapter for Efficient Video Generation", "ST-Adapter: Parameter-Efficient Image-to-Video Transfer Learning"], "answer_arxiv_id": ["2308.09710", "2206.13559"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9296"} +{"question": "What works propose outsourcing some proof goals to classical provers?", "answer": ["Thor: Wielding Hammers to Integrate Language Models and Automated Theorem Provers"], "answer_arxiv_id": ["2205.10893"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_9297"} +{"question": "Which literature evidence demonstrates the effectiveness of IRLS in low-rank matrix recovery and completion?", "answer": ["Low-rank matrix recovery via iteratively reweighted least squares minimization", "Harmonic Mean Iteratively Reweighted Least Squares for Low-Rank Matrix Recovery", "A Scalable Second Order Method for Ill-Conditioned Matrix Completion from Few Samples"], "answer_arxiv_id": ["1010.2471", "1703.05038", "2106.02119"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_9298"} +{"question": "Which studies have used mechanistic interpretability to reverse engineer circuits in state-of-the-art vision models or transformer models?", "answer": ["Progress measures for grokking via mechanistic interpretability", "Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small"], "answer_arxiv_id": ["2301.05217", "2211.00593"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_9299"} +{"question": "Who proposed the integration of symmetry into model-free RL based on MDP homomorphisms?", "answer": ["MDP Homomorphic Networks: Group Symmetries in Reinforcement Learning"], "answer_arxiv_id": ["2006.16908"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_9300"} +{"question": "Could you give references on several few-shot knowledge distillation methods?", "answer": ["Few Sample Knowledge Distillation for Efficient Network Compression", "Few Shot Network Compression via Cross Distillation", "Compressing Models with Few Samples: Mimicking then Replacing", "Practical Network Acceleration with Tiny Sets"], "answer_arxiv_id": ["1812.01839", "1911.09450", "2201.02620", "2202.07861"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_9301"} +{"question": "What works implement pretrained diffusion models to solve linear and some special non-linear inverse problems?", "answer": ["Denoising Diffusion Restoration Models", "Diffusion Posterior Sampling for General Noisy Inverse Problems"], "answer_arxiv_id": ["2201.11793", "2209.14687"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_9302"} +{"question": "What are the works focused on Language models and their success?", "answer": ["Training language models to follow instructions with human feedback", "GPT-4 Technical Report", "Gemini: A Family of Highly Capable Multimodal Models", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2203.02155", "2303.08774", "2312.11805v4", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_9303"} +{"question": "Could you provide me some works that are about the application of deep learning in software implementation?", "answer": ["Evaluating Large Language Models Trained on Code", "CodeGen: An Open Large Language Model for Code with Multi-Turn Program\n Synthesis", "Self-collaboration Code Generation via ChatGPT"], "answer_arxiv_id": ["2107.03374", "2203.13474", "2304.07590"], "source_meta": {"published_time": "20230716"}, "qid": "AutoScholarQuery_train_9304"} +{"question": "Are there works on human reconstruction employing neural radiance fields from monocular videos?", "answer": ["Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "InstantAvatar: Learning Avatars from Monocular Video in 60 Seconds", "TAVA: Template-free Animatable Volumetric Actors", "H-NeRF: Neural Radiance Fields for Rendering and Temporal Reconstruction\n of Humans in Motion", "ARAH: Animatable Volume Rendering of Articulated Human SDFs", "Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via\n Self-supervised Scene Decomposition"], "answer_arxiv_id": ["2012.15838", "2201.04127", "2212.10550", "2206.08929", "2110.13746", "2210.10036v1", "2302.11566"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9305"} +{"question": "Could you provide me some examples of research referred to as e3nn networks?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials", "Geometric and Physical Quantities improve E(3) Equivariant Message Passing", "MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields", "Learning Local Equivariant Representations for Large-Scale Atomistic Dynamics", "Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs"], "answer_arxiv_id": ["1802.08219", "2101.03164", "2110.02905", "2206.07697", "2204.05249", "2206.11990"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_9306"} +{"question": "Could you provide me the studies that aimed to modify the training process of the prediction model to improve the efficiency of conformal prediction?", "answer": ["Learning Optimal Conformal Classifiers"], "answer_arxiv_id": ["2110.09192"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9307"} +{"question": "Could you give me examples of works that use distribution-based methods in jailbreak research?", "answer": ["GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts", "Jailbreak and Guard Aligned Language Models with Only Few In-Context\n Demonstrations", "Adversarial Demonstration Attacks on Large Language Models", "Scalable and Transferable Black-Box Jailbreaks for Language Models via\n Persona Modulation"], "answer_arxiv_id": ["2309.10253v4", "2310.06387", "2305.14950", "2311.03348"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_9308"} +{"question": "What papers use the alignment and uniformity properties to guide the representation learning?", "answer": ["Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere"], "answer_arxiv_id": ["2005.10242"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_9309"} +{"question": "What studies are about indoor scene synthesis using expensive 3D datasets or CAD retrieval?", "answer": ["Unconstrained Scene Generation with Locally Conditioned Radiance Fields", "ATISS: Autoregressive Transformers for Indoor Scene Synthesis", "SceneFormer: Indoor Scene Generation with Transformers", "GAUDI: A Neural Architect for Immersive 3D Scene Generation"], "answer_arxiv_id": ["2104.00670", "2110.03675v1", "2012.09793", "2207.13751"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_9310"} +{"question": "Which work discusses tracking the rotation and position of 3D Gaussians initialized in the first frame across timesteps?", "answer": ["Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis"], "answer_arxiv_id": ["2308.09713"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_9311"} +{"question": "Which studies have utilized self-attention mechanisms to tackle the issue of capturing long-range dependencies in 3D object detection?", "answer": ["SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds", "Embracing Single Stride 3D Object Detector with Sparse Transformer", "DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets"], "answer_arxiv_id": ["2210.07372", "2112.06375", "2301.06051"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_9312"} +{"question": "What works focus on learning directly from instance-dependent label noise, without explicitly estimating the transition matrix?", "answer": ["A Second-Order Approach to Learning with Instance-Dependent Label Noise", "Learning with Instance-Dependent Label Noise: A Sample Sieve Approach", "Confidence Scores Make Instance-dependent Label-noise Learning Possible", "DivideMix: Learning with Noisy Labels as Semi-supervised Learning"], "answer_arxiv_id": ["2012.11854", "2010.02347", "2001.03772", "2002.07394"], "source_meta": {"published_time": "20220204"}, "qid": "AutoScholarQuery_train_9313"} +{"question": "What works discuss the semantic drift in the feature representations in many-shot class-incremental learning?", "answer": ["Semantic Drift Compensation for Class-Incremental Learning"], "answer_arxiv_id": ["2004.00440"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_9314"} +{"question": "What works discuss the implicit bias of homogeneous networks trained with gradient descent that converges to a Karush-Kuhn-Tucker (KKT) point?", "answer": ["Gradient Descent Maximizes the Margin of Homogeneous Neural Networks", "Directional convergence and alignment in deep learning"], "answer_arxiv_id": ["1906.05890", "2006.06657"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_9315"} +{"question": "Which works provide optimal order regret bounds for kernel-based (non-contextual) bandits?", "answer": ["A Domain-Shrinking based Bayesian Optimization Algorithm with Order-Optimal Regret Performance", "High-Dimensional Experimental Design and Kernel Bandits", "Gaussian Process Bandit Optimization with Few Batches"], "answer_arxiv_id": ["2010.13997", "2105.05806", "2110.07788v4"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_9316"} +{"question": "What research works have been done to address geographical robustness specificly?", "answer": ["Train in Germany, Test in The USA: Making 3D Object Detectors Generalize", "Adaptive Methods for Real-World Domain Generalization", "GIVL: Improving Geographical Inclusivity of Vision-Language Models with\n Pre-Training Methods"], "answer_arxiv_id": ["2005.08139", "2103.15796", "2301.01893"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_9317"} +{"question": "Which paper took a first step towards a more applicable theory of compositional generalization to unseen domains?", "answer": ["First Steps Toward Understanding the Extrapolation of Nonlinear Models to Unseen Domains"], "answer_arxiv_id": ["2211.11719"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_9318"} +{"question": "Which papers discuss point-based transformations for 3D semantic segmentation?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space", "KPConv: Flexible and Deformable Convolution for Point Clouds"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1904.08889"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_9319"} +{"question": "What papers propose adding regularization terms to the loss of DKL to mitigate overfitting?", "answer": ["Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness", "On Feature Collapse and Deep Kernel Learning for Single Forward Pass Uncertainty"], "answer_arxiv_id": ["2006.10108", "2102.11409"], "source_meta": {"published_time": "20220505"}, "qid": "AutoScholarQuery_train_9320"} +{"question": "Can you point out some research that introduced graph convolution based mechanisms for filtering?", "answer": ["Learning Graph-Convolutional Representations for Point Cloud Denoising", "Differentiable Manifold Reconstruction for Point Cloud Denoising", "Deep Point Set Resampling via Gradient Fields", "PD-Flow: A Point Cloud Denoising Framework with Normalizing Flows", "IterativePFN: True Iterative Point Cloud Filtering", "Contrastive Learning for Joint Normal Estimation and Point Cloud Filtering"], "answer_arxiv_id": ["2007.02578", "2007.13551", "2111.02045", "2203.05940", "2304.01529v1", "2208.06811v2"], "source_meta": {"published_time": "20240514"}, "qid": "AutoScholarQuery_train_9321"} +{"question": "What studies propose to use larger and more diverse datasets to mitigate hallucination issues in LLMs?", "answer": ["Factuality Enhanced Language Models for Open-Ended Text Generation"], "answer_arxiv_id": ["2206.04624"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_9322"} +{"question": "What studies explored variations of Lipschitz bandits?", "answer": ["Contextual Bandits with Similarity Information", "Sharp Dichotomies for Regret Minimization in Metric Spaces"], "answer_arxiv_id": ["0907.3986v5", "0911.1174v1"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_9323"} +{"question": "Are there any studies on the use of Proximal Policy Optimization (PPO) in the context of language models and fact-checking?", "answer": ["Proximal Policy Optimization Algorithms", "Secrets of RLHF in Large Language Models Part I: PPO"], "answer_arxiv_id": ["1707.06347", "2307.04964"], "source_meta": {"published_time": "20240718"}, "qid": "AutoScholarQuery_train_9324"} +{"question": "What works established GANs, VAEs, and normalizing flows which are classical approaches for learning deep generative models?", "answer": ["Generative Adversarial Nets", "FuseDream: Training-Free Text-to-Image Generation with Improved CLIP+GAN Space Optimization", "Auto-Encoding Variational Bayes", "Variational Inference with Normalizing Flows", "NICE: Non-linear Independent Components Estimation", "Density estimation using Real NVP"], "answer_arxiv_id": ["1406.2661", "2112.01573", "1312.6114", "1505.05770", "1410.8516", "1605.08803"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_9325"} +{"question": "Could you point out papers that introduce subgraph Graph Neural Networks, which are more expressive than Message Passing Neural Networks?", "answer": ["Nested Graph Neural Networks", "Equivariant Subgraph Aggregation Networks", "Understanding and Extending Subgraph GNNs by Rethinking Their Symmetries", "From Stars to Subgraphs: Uplifting Any GNN with Local Structure Awareness"], "answer_arxiv_id": ["2110.13197", "2110.02910", "2206.11140", "2110.03753"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_9326"} +{"question": "Could you provide me some works that use neural networks to learn the Stein discrepancy?", "answer": ["Stein Neural Sampler", "Learning the Stein Discrepancy for Training and Evaluating Energy-Based Models without Sampling", "Neural Variational Gradient Descent"], "answer_arxiv_id": ["1810.03545", "2002.05616", "2107.10731v2"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_9327"} +{"question": "What papers introduce a different learning setting, separating the learning of structure, motion, and content in text-guided video editing?", "answer": ["MagicEdit: High-Fidelity and Temporally Coherent Video Editing"], "answer_arxiv_id": ["2308.14749"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9328"} +{"question": "Which references discuss the relationship between the pretext task and the downstream task in understanding the success of SSL?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style"], "answer_arxiv_id": ["1902.09229", "2106.04619"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_9329"} +{"question": "What studies have been used in the drug discovery context involving Bayesian optimization?", "answer": ["Scalable Bayesian Optimization Using Deep Neural Networks", "Learning Optimal Interventions", "GeneDisco: A Benchmark for Experimental Design in Drug Discovery"], "answer_arxiv_id": ["1502.05700", "1606.05027v2", "2110.11875v1"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_9330"} +{"question": "What paper is about re-ranking retrieved articles to filter out noise?", "answer": ["GripRank: Bridging the Gap between Retrieval and Generation via the\n Generative Knowledge Improved Passage Ranking"], "answer_arxiv_id": ["2305.18144"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_train_9331"} +{"question": "What works have addressed the issue of reward noise in RL training?", "answer": ["Adversarial Attacks on Neural Network Policies", "Reinforcement Learning with a Corrupted Reward Channel", "Reinforcement Learning with Perturbed Rewards"], "answer_arxiv_id": ["1702.02284", "1705.08417", "1810.01032"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_9332"} +{"question": "Could you provide me some studies where contrastive learning was utilized in vision-language models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "MURAL: Multimodal, Multitask Retrieval Across Languages", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "Combined Scaling for Zero-shot Transfer Learning", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "CoCa: Contrastive Captioners are Image-Text Foundation Models"], "answer_arxiv_id": ["2103.00020", "2109.05125", "2102.05918", "2107.07651", "2111.10050", "2111.07991", "2205.01917"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_9333"} +{"question": "Could you provide me some works that proposed KnowledgeEditor (KE) and MEND for HyperNetwork-based model editing?", "answer": ["Editing Factual Knowledge in Language Models", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2104.08164", "2110.11309"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_9334"} +{"question": "Can you list the works proposing normalizing flows as a method to improve variational autoencoders?", "answer": ["Variational Lossy Autoencoder"], "answer_arxiv_id": ["1611.02731"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_9335"} +{"question": "Which paper presented the Custom Diffusion fine-tuning method for Stable Diffusion?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2212.04488"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9336"} +{"question": "What research publications focus on optimization-based methods for inverse rendering?", "answer": ["Shape, Illumination, and Reflectance from Shading", "Radiometric Scene Decomposition: Scene Reflectance, Illumination, and\n Geometry from RGB-D Images"], "answer_arxiv_id": ["2010.03592", "1604.01354"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_9337"} +{"question": "What papers generate static human-scene interactions?", "answer": ["Putting Humans in a Scene: Learning Affordance in 3D Indoor Environments", "Predictive and Generative Neural Networks for Object Functionality", "Generating 3D People in Scenes without People", "PLACE: Proximity Learning of Articulation and Contact in 3D Environments", "Populating 3D Scenes by Learning Human-Scene Interaction"], "answer_arxiv_id": ["1903.05690", "2006.15520", "1912.02923", "2008.05570", "2012.11581"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_9338"} +{"question": "What research focuses on transferring knowledge only from the trained source models without access to the source data?", "answer": ["Camera On-boarding for Person Re-identification using Hypothesis\n Transfer Learning"], "answer_arxiv_id": ["2007.11149"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_9339"} +{"question": "What work proposed a novel centroid-guided mechanism with a pre-training strategy for intent discovery?", "answer": ["A Clustering Framework for Unsupervised and Semi-supervised New Intent\n Discovery"], "answer_arxiv_id": ["2304.07699"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_9340"} +{"question": "What papers developed specialized OOD detection functions for dense OOD detection?", "answer": ["Scaling Out-of-Distribution Detection for Real-World Settings", "Standardized Max Logits: A Simple yet Effective Approach for Identifying Unexpected Road Obstacles in Urban-Scene Segmentation", "Detecting the Unexpected via Image Resynthesis", "Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentation"], "answer_arxiv_id": ["1911.11132", "2107.11264", "1904.07595", "2003.08440"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_9341"} +{"question": "Could you provide me some studies that focused on the topic of Video Editing with Shape Change?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "Shape-aware Text-driven Layered Video Editing"], "answer_arxiv_id": ["2212.11565", "2303.09535", "2301.13173"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_9342"} +{"question": "What works formulate objects in a scene as a human-centric graph?", "answer": ["Human-centric Indoor Scene Synthesis Using Stochastic Grammar", "Pose2Room: Understanding 3D Scenes from Human Activities"], "answer_arxiv_id": ["1808.08473", "2112.03030"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_9343"} +{"question": "Could you provide me some studies about text-to-image synthesis using diffusion models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10741", "2112.10752"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9344"} +{"question": "What works proposed to use input convex NNs (ICNNs) within the JKO scheme?", "answer": ["Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks", "Proximal Optimal Transport Modeling of Population Dynamics", "Large-Scale Wasserstein Gradient Flows"], "answer_arxiv_id": ["2106.00774", "2106.06345", "2106.00736"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_9345"} +{"question": "Which paper proposed the architecture of DeepONet?", "answer": ["DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators"], "answer_arxiv_id": ["1910.03193"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_9346"} +{"question": "In what papers the researcher developed 3D action-value maps that aligns the 3D task space and the action space?", "answer": ["Coarse-to-Fine Q-attention: Efficient Learning for Visual Robotic\n Manipulation via Discretisation"], "answer_arxiv_id": ["2106.12534"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_9347"} +{"question": "What works inspired the framework of the study and motivated the use of the Gumbel loss?", "answer": ["IQ-Learn: Inverse soft-Q Learning for Imitation"], "answer_arxiv_id": ["2106.12142"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_9348"} +{"question": "What is the established robustness benchmark mentioned in the paper?", "answer": ["RobustBench: a standardized adversarial robustness benchmark"], "answer_arxiv_id": ["2010.09670"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_9349"} +{"question": "Which works introduced activation function, linear layer and recurrent architectures in hyperbolic space?", "answer": ["Hyperbolic Neural Networks++"], "answer_arxiv_id": ["2006.08210"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_9350"} +{"question": "Can you provide studies on self-supervised contrastive learning used for few-shot segmentation?", "answer": ["Learning Representations by Maximizing Mutual Information Across Views"], "answer_arxiv_id": ["1906.00910"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_9351"} +{"question": "Can you provide papers that discuss normative models of adaptive statistical whitening and related transformations using synaptic plasticity mechanisms?", "answer": ["A Normative Theory of Adaptive Dimensionality Reduction in Neural Networks", "Interneurons accelerate learning dynamics in recurrent neural networks for statistical adaptation"], "answer_arxiv_id": ["1511.09426", "2209.10634"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_9352"} +{"question": "What paper fine-tunes the entire UNet backbone with a unique identifier as a method for personalization in text-to-image diffusion models?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_9353"} +{"question": "Could you provide me some works that combined recurrent neural nets with neural ODE dynamics?", "answer": ["GRU-ODE-Bayes: Continuous modeling of sporadically-observed time series"], "answer_arxiv_id": ["1905.12374"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_9354"} +{"question": "Could you name the studies that investigated pretraining approaches by imposing local consistency at pixel or region level?", "answer": ["Unsupervised Learning of Dense Visual Representations", "Propagate Yourself: Exploring Pixel-Level Consistency for Unsupervised Visual Representation Learning", "Dense Contrastive Learning for Self-Supervised Visual Pre-Training", "Spatially Consistent Representation Learning", "Instance Localization for Self-supervised Detection Pretraining", "Region Similarity Representation Learning"], "answer_arxiv_id": ["2011.05499", "2011.10043", "2011.09157", "2103.06122", "2102.08318", "2103.12902"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_9355"} +{"question": "In which paper the user profile for each agent was considered as a fixed during the entire time?", "answer": ["Federated Linear Contextual Bandits"], "answer_arxiv_id": ["2110.14177"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_9356"} +{"question": "Can you provide research papers which focus on the development and enhancement of Large Language Models (LMMs)?", "answer": ["Language Models are Few-Shot Learners", "Evaluating Large Language Models Trained on Code", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2107.03374", "2204.02311"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_9357"} +{"question": "Which papers founded and developed diffusion models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Denoising Diffusion Implicit Models", "Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2112.10752", "2205.11487", "2010.02502", "2006.11239", "2105.05233"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_9358"} +{"question": "Which studies are about adding regularization terms to the loss function in order to pursue wide local minima?", "answer": ["Regularizing Neural Networks by Penalizing Confident Output Distributions", "Deep Mutual Learning", "Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillation", "Entropy-SGD: Biasing Gradient Descent Into Wide Valleys"], "answer_arxiv_id": ["1701.06548", "1706.00384", "1905.08094", "1611.01838"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_9359"} +{"question": "What research talks about the shift in model class from LSTMs to transformers for representation learning in continuous-time chirographic models?", "answer": ["Generating Sequences With Recurrent Neural Networks", "Sketchformer: Transformer-based Representation for Sketched Structure", "CoSE: Compositional Stroke Embeddings"], "answer_arxiv_id": ["1308.0850", "2002.10381", "2006.09930"], "source_meta": {"published_time": "20230407"}, "qid": "AutoScholarQuery_train_9360"} +{"question": "What research constructed a new graph for clustering by pulling nearest neighbors close?", "answer": ["Multi-view Contrastive Graph Clustering"], "answer_arxiv_id": ["2110.11842"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_9361"} +{"question": "What papers generate music that matches the semantic information in the input text?", "answer": ["MusicLM: Generating Music From Text", "Noise2Music: Text-conditioned Music Generation with Diffusion Models"], "answer_arxiv_id": ["2301.11325", "2302.03917"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_9362"} +{"question": "In what papers do the researchers propose to reconstruct the teammate's behaviors by the agent's local observation through an encoder-decoder network?", "answer": ["Agent Modelling under Partial Observability for Deep Reinforcement Learning"], "answer_arxiv_id": ["2006.09447"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_9363"} +{"question": "Which papers grounded responses from question answering models in environment state or physics through environment interactions?", "answer": ["IQA: Visual Question Answering in Interactive Environments", "Embodied Question Answering", "Mind’s Eye: Grounded Language Model Reasoning through Simulation"], "answer_arxiv_id": ["1712.03316", "1711.11543", "2210.05359"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_9364"} +{"question": "What studies utilize a learnable attention map specifically on the eye for gaze estimation?", "answer": ["It's Written All Over Your Face: Full-Face Appearance-Based Gaze\n Estimation"], "answer_arxiv_id": ["1611.08860"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_9365"} +{"question": "Which works proposed methods for uncertainty quantification on graphs using deterministic single-pass GNNs?", "answer": ["Uncertainty Aware Semi-Supervised Learning on Graph Data", "Energy-based Out-of-Distribution Detection for Graph Neural Networks"], "answer_arxiv_id": ["2010.12783", "2302.02914"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_9366"} +{"question": "Which works use the DNN itself to provide uncertainty estimates for its outputs?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples\n in Neural Networks", "Unsupervised Risk Estimation Using Only Conditional Independence\n Structure", "Energy-based Out-of-distribution Detection"], "answer_arxiv_id": ["1610.02136", "1606.05313", "2010.03759"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_9367"} +{"question": "Any works about composing different parameter-efficient modules via simple arithmetic operations?", "answer": ["Composing Parameter-Efficient Modules with Arithmetic Operations"], "answer_arxiv_id": ["2306.14870"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_9368"} +{"question": "Could you provide me with some studies on generating human shapes using scanned data?", "answer": ["gDNA: Towards Generative Detailed Neural Avatars", "MetaAvatar: Learning Animatable Clothed Human Models from Few Depth Images"], "answer_arxiv_id": ["2201.04123", "2106.11944"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9369"} +{"question": "What are the papers about the methodology of lottery ticket for pruning?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Stabilizing the Lottery Ticket Hypothesis", "Comparing Rewinding and Fine-tuning in Neural Network Pruning"], "answer_arxiv_id": ["1803.03635", "1903.01611", "2003.02389"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_9370"} +{"question": "Which papers introduced function approximation to accelerate the speed of game solving?", "answer": ["Turbocharging Solution Concepts: Solving NEs, CEs and CCEs with Neural Equilibrium Solvers"], "answer_arxiv_id": ["2210.09257"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_9371"} +{"question": "What papers explore the universality and modality-agnostic capabilities of transformers?", "answer": ["Pretrained Transformers As Universal Computation Engines", "Language Modeling Is Compression"], "answer_arxiv_id": ["2103.05247", "2309.10668"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9372"} +{"question": "Which papers cover typical applications of equivariant neural network architectures, such as segmentation on spherical manifolds, robotics, and data augmentation?", "answer": ["SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "Equivariant Imaging: Learning Beyond the Range Space", "Rotation-Equivariant Deep Learning for Diffusion MRI", "Deep invariant networks with differentiable augmentation layers", "4D Panoptic Segmentation as Invariant and Equivariant Field Prediction"], "answer_arxiv_id": ["2006.10503", "2103.14756", "2102.06942", "2202.02142", "2303.15651"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_9373"} +{"question": "Do any works conduct fine-grained analyses where distribution shifts are categorized?", "answer": ["OoD-Bench: Quantifying and Understanding Two Dimensions of Out-of-Distribution Generalization"], "answer_arxiv_id": ["2106.03721"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_9374"} +{"question": "Which works have used point clouds as form of intrinsic representations for shapes?", "answer": ["A Point Set Generation Network for 3D Object Reconstruction from a Single Image"], "answer_arxiv_id": ["1612.00603"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_9375"} +{"question": "Could you name some papers that aim at querying LLMs to generate prompts or attributes of categories?", "answer": ["CHiLS: Zero-Shot Image Classification with Hierarchical Label Sets", "Visual Classification via Description from Large Language Models", "What does a platypus look like? Generating customized prompts for\n zero-shot image classification"], "answer_arxiv_id": ["2302.02551", "2210.07183", "2209.03320"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_9376"} +{"question": "Could you provide me some studies about applying contrastive learning to link prediction?", "answer": ["Inductive Representation Learning on Large Graphs", "Deep Graph Contrastive Representation Learning", "Graph Data Augmentation for Graph Machine Learning: A Survey", "Deep Graph Infomax", "Multi-task Self-supervised Graph Neural Networks Enable Stronger Task Generalization", "Automated Self-Supervised Learning for Graphs", "Graph Contrastive Learning with Augmentations", "Prototypical Graph Contrastive Learning"], "answer_arxiv_id": ["1706.02216", "2006.04131", "2202.08871", "1809.10341", "2210.02016", "2106.05470", "2010.13902", "2106.09645"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_9377"} +{"question": "Which papers investigated structured pruning of Visino Transformers?", "answer": ["Vision Transformer Pruning", "Learned Token Pruning for Transformers", "Evo-ViT: Slow-Fast Token Evolution for Dynamic Vision Transformer", "IA-RED2: Interpretability-Aware Redundancy Reduction for Vision Transformers", "CP-ViT: Cascade Vision Transformer Pruning via Progressive Sparsity Prediction", "DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "Multi-Dimensional Model Compression of Vision Transformer"], "answer_arxiv_id": ["2104.08500", "2107.00910", "2108.01390", "2106.12620", "2203.04570", "2106.02034", "2201.00043"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_9378"} +{"question": "What works have proposed semi-supervised learning methods using consistency regularization?", "answer": ["ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring", "MixMatch: A Holistic Approach to Semi-Supervised Learning", "Temporal Ensembling for Semi-Supervised Learning", "Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning", "Unsupervised Data Augmentation for Consistency Training"], "answer_arxiv_id": ["1911.09785", "1905.02249", "1610.02242", "1606.04586", "1904.12848"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_9379"} +{"question": "Are there any studies on simulating human annotation for evaluation?", "answer": ["Discovering Language Model Behaviors with Model-Written Evaluations", "Instruction Tuning with GPT-4", "Visual Instruction Tuning"], "answer_arxiv_id": ["2212.09251", "2304.03277", "2304.08485"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_9380"} +{"question": "Which study introduced Batch Prompting?", "answer": ["Batch Prompting: Efficient Inference with Large Language Model APIs"], "answer_arxiv_id": ["2301.08721"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_9381"} +{"question": "What are some recent studies that use the pose and shape of the parametric SMPL model as a prior for generating humans in text-to-3D human generation?", "answer": ["DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "AvatarCraft: Transforming Text into Neural Human Avatars with\n Parameterized Shape and Pose Control"], "answer_arxiv_id": ["2304.00916", "2303.17606"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_9382"} +{"question": "What papers contribute to the area of 3D Scene Editing of Radiance Fields?", "answer": ["Editing Conditional Radiance Fields", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields", "Decomposing NeRF for Editing via Feature Field Distillation", "Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D\n Image Representations", "SINE: Semantic-driven Image-based NeRF Editing with Prior-guided Editing\n Field", "TextDeformer: Geometry Manipulation using Text Guidance", "Neural Articulated Radiance Field", "NeRF-In: Free-Form NeRF Inpainting with RGB-D Priors", "Deforming Radiance Fields with Cages", "NeuMesh: Learning Disentangled Neural Mesh-based Implicit Field for\n Geometry and Texture Editing", "InstructP2P: Learning to Edit 3D Point Clouds with Text Instructions", "FocalDreamer: Text-driven 3D Editing via Focal-fusion Assembly"], "answer_arxiv_id": ["2105.06466", "2112.03221", "2205.08535", "2112.05139", "2205.15585", "2209.03494", "2303.13277", "2304.13348", "2104.03110", "2206.04901", "2207.12298", "2207.11911", "2306.07154", "2308.10608"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9383"} +{"question": "Could you provide me a study on a model that utilizes a supervised learning strategy and achieved impressive performance on unseen images?", "answer": ["Segment Anything"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_9384"} +{"question": "What papers discuss the aim of GFlowNets to sample composite objects proportionally to a reward function?", "answer": ["Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation", "GFlowNet Foundations"], "answer_arxiv_id": ["2106.04399", "2111.09266"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_9385"} +{"question": "What work indicates that gradient descent on matrix factorization with a small learning rate still maintains the auto-balancing property?", "answer": ["Global Convergence of Gradient Descent for Asymmetric Low-Rank Matrix Factorization"], "answer_arxiv_id": ["2106.14289"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_9386"} +{"question": "Which works show that using videos without annotation allows for an effective multimodal embedding space via contrastive learning?", "answer": ["Self-Supervised MultiModal Versatile Networks", "VATT: Transformers for Multimodal Self-Supervised Learning from Raw\n Video, Audio and Text", "AVLnet: Learning Audio-Visual Language Representations from\n Instructional Videos", "Multimodal Clustering Networks for Self-supervised Learning from\n Unlabeled Videos", "Everything at Once -- Multi-modal Fusion Transformer for Video Retrieval"], "answer_arxiv_id": ["2006.16228", "2104.11178", "2006.09199", "2104.12671", "2112.04446"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_9387"} +{"question": "What works assume task-ids availability at testing time in the continual learning of LLM?", "answer": ["LFPT5: A Unified Framework for Lifelong Few-shot Language Learning Based\n on Prompt Tuning of T5", "Continual Sequence Generation with Adaptive Compositional Modules", "Lifelong Sequence Generation with Dynamic Module Expansion and Adaptation"], "answer_arxiv_id": ["2110.07298", "2203.10652v2", "2310.09886v4"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_9388"} +{"question": "Which papers conduct study on robustness of architectures optimized by Neural Architecture Search methods?", "answer": ["On Adversarial Robustness: A Neural Architecture Search perspective", "When NAS Meets Robustness: In Search of Robust Architectures against Adversarial Attacks", "Adversarially Robust Neural Architectures", "Feature Denoising for Improving Adversarial Robustness", "Bag of Tricks for Adversarial Training", "Smooth Adversarial Training", "Is Robustness the Cost of Accuracy? – A Comprehensive Study on the Robustness of 18 Deep Image Classification Models", "RobustART: Benchmarking Robustness on Architecture Design and Training Techniques", "DSRNA: Differentiable Search of Robust Neural Architectures", "AdvRush: Searching for Adversarially Robust Neural Architectures"], "answer_arxiv_id": ["2007.08428", "1911.10695", "2009.00902", "1812.03411", "2010.00467", "2006.14536", "1808.01688", "2109.05211", "2012.06122", "2108.01289"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_9389"} +{"question": "Which papers have dealt with the process of model or knowledge distillation?", "answer": ["Knowledge distillation: A good teacher is patient and consistent", "Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["2106.05237", "1503.02531"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_9390"} +{"question": "Which papers proved that the training of wide neural networks is equivalent to the optimization of a specific kernel function?", "answer": ["On Exact Computation with an Infinitely Wide Neural Net", "Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent", "On Lazy Training in Differentiable Programming", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks"], "answer_arxiv_id": ["1904.11955", "1902.06720", "1812.07956", "1901.08584"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_9391"} +{"question": "In what research work is the field of 'offline meta-RL' addressed?", "answer": ["Offline Meta-Reinforcement Learning with Advantage Weighting", "model-based offline meta-reinforcement learning with regularization", "FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization"], "answer_arxiv_id": ["2008.06043", "2202.02929", "2010.01112"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_9392"} +{"question": "What research works have proposed methods modeling the learning curves of machine learning algorithms directly?", "answer": ["Learning Curves for Decision Making in Supervised Machine Learning — A Survey"], "answer_arxiv_id": ["2201.12150"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_train_9393"} +{"question": "Could you point me to studies on dynamic scenes that use multiple synchronized cameras?", "answer": ["Neural 3D Video Synthesis from Multi-view Video", "Fourier PlenOctrees for Dynamic Radiance Field Rendering in Real-time", "Streaming Radiance Fields for 3D Video Synthesis", "Mixed Neural Voxels for Fast Multi-view Video Synthesis"], "answer_arxiv_id": ["2103.02597", "2202.08614", "2210.14831", "2212.00190"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_9394"} +{"question": "Which studies propose computationally efficient estimators for phase retrieval based upon the LLL algorithm in the noiseless case?", "answer": ["On the Cryptographic Hardness of Learning Single Periodic Neurons"], "answer_arxiv_id": ["2106.10744"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_9395"} +{"question": "What works propose to reduce extrapolation error in offline RL by inducing policy regularization on the distributional discrepancy with the behavior policy?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "1906.00949", "1911.11361", "2106.06860"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_9396"} +{"question": "What research works focus on enhancing the performance of LLMs by searching for a more effective prompt?", "answer": ["Large Language Models Are Human-Level Prompt Engineers", "Automatic Engineering of Long Prompts", "Connecting Large Language Models with Evolutionary Algorithms Yields\n Powerful Prompt Optimizers", "PromptAgent: Strategic Planning with Language Models Enables\n Expert-level Prompt Optimization"], "answer_arxiv_id": ["2211.01910", "2311.10117", "2309.08532", "2310.16427"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_9397"} +{"question": "Which studies have used deep learning methods based on Convolutional Neural Networks (CNN) for image super-resolution tasks?", "answer": ["Image Super-Resolution Using Deep Convolutional Networks", "Accelerating the Super-Resolution Convolutional Neural Network", "Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Enhanced Deep Residual Networks for Single Image Super-Resolution", "Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution"], "answer_arxiv_id": ["1501.00092", "1608.00367", "1511.04587", "1707.02921", "1704.03915"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_9398"} +{"question": "Which work introduced the idea of a truncated forward process by replacing the last steps in the forward process with an autoencoder for noise generation?", "answer": ["Truncated Diffusion Probabilistic Models and Diffusion-based Adversarial Auto-Encoders"], "answer_arxiv_id": ["2202.09671"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_9399"} +{"question": "Which papers proposed methods to address feature interaction layers in deep sparse networks?", "answer": ["DeepFM: A Factorization-Machine based Neural Network for CTR Prediction", "Product-based Neural Networks for User Response Prediction", "Deep & Cross Network for Ad Click Predictions", "DCN V2: Improved Deep & Cross Network and Practical Lessons for Web-scale Learning to Rank Systems", "Product-based Neural Networks for User Response Prediction over Multi-field Categorical Data"], "answer_arxiv_id": ["1703.04247", "1611.00144", "1708.05123", "2008.13535", "1807.00311"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_9400"} +{"question": "Which work introduced the GP-UCB algorithm for optimizing unknown functions in the field of Bayesian optimization?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design"], "answer_arxiv_id": ["0912.3995v4"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_9401"} +{"question": "What works introduced the feature correspondences to distinguish different classes of objects?", "answer": ["Unsupervised Semantic Segmentation by Distilling Feature Correspondences"], "answer_arxiv_id": ["2203.08414"], "source_meta": {"published_time": "20220919"}, "qid": "AutoScholarQuery_train_9402"} +{"question": "What papers discuss methods for analyzing the complexity of subgradient methods for weakly convex non-smooth unconstrained problems?", "answer": ["Stochastic model-based minimization of weakly convex functions", "Proximally Guided Stochastic Subgradient Method for Nonsmooth, Nonconvex Problems", "Minibatch and Momentum Model-based Methods for Stochastic Weakly Convex Optimization", "Weakly-Convex Concave Min-Max Optimization: Provable Algorithms and Applications in Machine Learning", "On the Convergence Rate of Stochastic Mirror Descent for Nonsmooth Nonconvex Optimization"], "answer_arxiv_id": ["1803.06523", "1707.03505", "2106.03034", "1810.02060", "1806.04781"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9403"} +{"question": "Are there any papers that design efficient KDE algorithms based on the interpolation of kernel density estimators?", "answer": ["Efficient Interpolation of Density Estimators"], "answer_arxiv_id": ["2011.04922"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_9404"} +{"question": "Can you provide some works about knowledge-enhanced urban spatiotemporal prediction utilizing knowledge graphs?", "answer": ["Spatio-Temporal Urban Knowledge Graph Enabled Mobility Prediction", "Out-of-Town Recommendation with Travel Intention Modeling", "Knowledge-driven Site Selection via Urban Knowledge Graph", "Knowledge-infused Contrastive Learning for Urban Imagery-based Socioeconomic Prediction", "MugRep: A Multi-Task Hierarchical Graph Representation Learning Framework for Real Estate Appraisal"], "answer_arxiv_id": ["2111.03465", "2101.12555", "2111.00787", "2302.13094", "2107.05180"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_9405"} +{"question": "What research introduced the idea of mapping discrete tokens into continuous latent variable for more complex controllable text generation?", "answer": ["Diffusion-LM Improves Controllable Text Generation"], "answer_arxiv_id": ["2205.14217"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_9406"} +{"question": "What works use random projection method for dimensionality reduction?", "answer": ["Re-Examining Linear Embeddings for High-Dimensional Bayesian Optimization"], "answer_arxiv_id": ["2001.11659"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9407"} +{"question": "Which studies show that student fails to find a zero-loss solution for a certain fraction of random initializations when the teacher has only a few neurons?", "answer": ["High-dimensional limit theorems for SGD: Effective dynamics and critical scaling", "On Learning Gaussian Multi-index Models with Gradient Flow"], "answer_arxiv_id": ["2206.04030", "2310.19793v2"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_9408"} +{"question": "Which study discussed the existence of a fundamental tradeoff between distortion and realism?", "answer": ["Rethinking Lossy Compression: The Rate-Distortion-Perception Tradeoff"], "answer_arxiv_id": ["1901.07821"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_train_9409"} +{"question": "Which papers explored the relevance or mapping in the latent space between different modalities for summary generation?", "answer": ["Video Summarization using Deep Semantic Features", "Multi-modal Summarization for Video-containing Documents"], "answer_arxiv_id": ["1609.08758", "2009.08018"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_9410"} +{"question": "What work proposed to include context prompts in image classification?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_9411"} +{"question": "What works discussed the importance of fair machine learning models in job market?", "answer": ["A Short-term Intervention for Long-term Fairness in the Labor Market"], "answer_arxiv_id": ["1712.00064"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_9412"} +{"question": "Could you name the studies that have been carried out on fine-grained hand-object interaction and entail contact prediction?", "answer": ["IMos: Intent-Driven Full-Body Motion Synthesis for Human-Object\n Interactions", "Physically Plausible Full-Body Hand-Object Interaction Synthesis", "Affordance Diffusion: Synthesizing Hand-Object Interactions", "CAMS: CAnonicalized Manipulation Spaces for Category-Level Functional\n Hand-Object Manipulation Synthesis", "TOCH: Spatio-Temporal Object-to-Hand Correspondence for Motion\n Refinement", "Task-Oriented Human-Object Interactions Generation with Implicit Neural\n Representations"], "answer_arxiv_id": ["2212.07555", "2309.07907", "2303.12538", "2303.15469", "2205.07982", "2303.13129"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9413"} +{"question": "Are there any studies which implemented Object Detection using vision-language models (VLMs) like CLIP and ALIGN?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_9414"} +{"question": "What studies have recognized and partly modeled the challenges of real-world environments in simulated environments for reinforcement learning?", "answer": ["An empirical investigation of the challenges of real-world reinforcement learning"], "answer_arxiv_id": ["2003.11881"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_9415"} +{"question": "Which papers discussed about learning from observations (LfO) in IL settings?", "answer": ["Imitation Learning by State-Only Distribution Matching", "Better-than-Demonstrator Imitation Learning via Automatically-Ranked Demonstrations", "Imitation from Observation: Learning to Imitate Behaviors from Raw Video via Context Translation", "A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "Generative Adversarial Imitation from Observation"], "answer_arxiv_id": ["2202.04332", "1907.03976", "1707.03374", "1011.0686v3", "1807.06158"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_9416"} +{"question": "Could you provide me some works that conducted training end-to-end models on large-scale web-collected data?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Advancing High-Resolution Video-Language Representation with Large-Scale Video Transcriptions", "MERLOT: Multimodal Neural Script Knowledge Models", "MERLOT Reserve: Neural Script Knowledge through Vision and Language and Sound"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.10337", "2106.02636", "2201.02639"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_train_9417"} +{"question": "Any works about proposing online learning algorithms for classical TPPs?", "answer": ["Tracking Dynamic Point Processes on Networks"], "answer_arxiv_id": ["1409.0031"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_9418"} +{"question": "Which work proposed building zero-shot dense retrieval systems?", "answer": ["Precise Zero-Shot Dense Retrieval without Relevance Labels"], "answer_arxiv_id": ["2212.10496"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_9419"} +{"question": "Which works provide open-source RL frameworks following the Atari breakthrough?", "answer": ["Dopamine: A Research Framework for Deep Reinforcement Learning", "Horizon: Facebook’s Open Source Applied Reinforcement Learning Platform", "Benchmarking Deep Reinforcement Learning for Continuous Control", "Catalyst.RL: A Distributed Framework for Reproducible RL Research", "ChainerRL: A Deep Reinforcement Learning Library", "RLlib: Abstractions for Distributed Reinforcement Learning", "ChainerRL: A Deep Reinforcement Learning Library", "CleanRL: High-quality Single-file Implementations of Deep Reinforcement Learning Algorithms", "Tianshou: A Highly Modularized Deep Reinforcement Learning Library", "rlpyt: A Research Code Base for Deep Reinforcement Learning in PyTorch"], "answer_arxiv_id": ["1812.06110", "1811.00260", "1604.06778", "1903.00027", "1912.03905", "1712.09381", "1912.03905", "2111.08819", "2107.14171", "1909.01500"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_9420"} +{"question": "Which paper provided an evaluation benchmark for diagnosing how much a given system understands about physics by observing its interaction with videos of possible versus impossible events?", "answer": ["IntPhys 2019: A Benchmark for Visual Intuitive Physics Understanding"], "answer_arxiv_id": ["1803.07616"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_9421"} +{"question": "Which resources focus on multilingual text?", "answer": ["MULTITuDE: Large-Scale Multilingual Machine-Generated Text Detection\n Benchmark", "M4: Multi-generator, Multi-domain, and Multi-lingual Black-Box\n Machine-Generated Text Detection"], "answer_arxiv_id": ["2310.13606", "2305.14902"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_9422"} +{"question": "Which studies belongs to weakly supervised object detection method and used multiple instance learning or class activation?", "answer": ["On learning to localize objects with minimal supervision", "Learning Deep Features for Discriminative Localization"], "answer_arxiv_id": ["1403.1024", "1512.04150"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_9423"} +{"question": "Which research incorporated using neural networks in symbolic regression?", "answer": ["AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularity", "AI Feynman: a Physics-Inspired Method for Symbolic Regression", "Neural Symbolic Regression that Scales", "Deep Generative Symbolic Regression"], "answer_arxiv_id": ["2006.10782", "1905.11481", "2106.06427", "2401.00282"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_9424"} +{"question": "Could you list some studies that used neural networks to parameterize the Q-function?", "answer": ["A Finite-Time Analysis of Q-Learning with Neural Network Function Approximation", "Off-Policy Deep Reinforcement Learning without Exploration"], "answer_arxiv_id": ["1912.04511", "1812.02900"], "source_meta": {"published_time": "20221114"}, "qid": "AutoScholarQuery_train_9425"} +{"question": "Are there any research studies that discuss the transformation of tree structure to general graph data structure in the context of Vector Quantization (VQ) indices?", "answer": ["Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs"], "answer_arxiv_id": ["1603.09320v4"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_9426"} +{"question": "Can you suggest some works that discuss the concept of contextual similarity in unsupervised metric learning?", "answer": ["Deep Learning for Person Re-identification: A Survey and Outlook", "Re-ranking Person Re-identification with k-reciprocal Encoding", "Self-Taught Metric Learning without Labels"], "answer_arxiv_id": ["2001.04193", "1701.08398", "2205.01903"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_9427"} +{"question": "Are there any studies that propose defenses against textual backdoor attacks?", "answer": ["Design and Evaluation of a Multi-Domain Trojan Detection Method on Deep Neural Networks", "ONION: A Simple and Effective Defense Against Textual Backdoor Attacks", "RAP: Robustness-Aware Perturbations for Defending against Backdoor Attacks on NLP Models"], "answer_arxiv_id": ["1911.10312", "2011.10369", "2110.07831"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_9428"} +{"question": "Could you mention some researches that examines the effects of scaling model size, data amount, and training budget on the capabilities of deep models?", "answer": ["PaLM: Scaling Language Modeling with Pathways", "GPT-4 Technical Report", "Scaling Language Models: Methods, Analysis & Insights from Training\n Gopher", "Scaling Vision Transformers to 22 Billion Parameters", "Scaling Vision Transformers"], "answer_arxiv_id": ["2204.02311", "2303.08774", "2112.11446", "2302.05442", "2106.04560"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9429"} +{"question": "Do we have research studying the unrestricted adversarial example setting, which aims to find unambiguous examples on which models make mistakes?", "answer": ["Unrestricted Adversarial Examples", "Adversarial training for high-stakes reliability"], "answer_arxiv_id": ["1809.08352", "2205.01663"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_9430"} +{"question": "What research has used BLLMs for generating synthetic training data?", "answer": ["FlexKBQA: A Flexible LLM-Powered Framework for Few-Shot Knowledge Base\n Question Answering", "Data Distribution Bottlenecks in Grounding Language Models to Knowledge Bases"], "answer_arxiv_id": ["2308.12060", "2309.08345v3"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_9431"} +{"question": "Any research about gradient-based methods for estimating the importance measure in feature attribution?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1610.02391", "1703.01365"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_9432"} +{"question": "Which study introduced a patch regularizer to mitigate geometry artifacts and employed a log-likelihood model to ensure multi-view appearance consistency specifically for few-shot reconstruction?", "answer": ["RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs"], "answer_arxiv_id": ["2112.00724"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9433"} +{"question": "Could you provide me some works about the applications of coordinate-based Multilayer Perceptron (MLP) in super-resolution and reconstruction?", "answer": ["Learning Continuous Image Representation with Local Implicit Image Function", "Modulated Periodic Activations for Generalizable Local Functional Representations"], "answer_arxiv_id": ["2012.09161", "2104.03960"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_9434"} +{"question": "There are works about summary chain-of-thought method?", "answer": ["Element-aware Summarization with Large Language Models: Expert-aligned\n Evaluation and Chain-of-Thought Method"], "answer_arxiv_id": ["2305.13412"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_9435"} +{"question": "Which references are related to music generation models that can generate high-quality single-genre music samples?", "answer": ["RAVE: A variational autoencoder for fast and high-quality neural audio synthesis", "Musika! Fast Infinite Waveform Music Generation"], "answer_arxiv_id": ["2111.05011", "2208.08706"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_9436"} +{"question": "Any research papers that focus on applying Federated Learning to imbalanced data tasks?", "answer": ["Communication-Efficient Distributed Stochastic AUC Maximization with Deep Neural Networks"], "answer_arxiv_id": ["2005.02426"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_9437"} +{"question": "What theoretical works exist in model-based reinforcement learning?", "answer": ["Model-based Reinforcement Learning and the Eluder Dimension", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches", "Model-Based Reinforcement Learning with a Generative Model is Minimax Optimal", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Tighter Problem-Dependent Regret Bounds in Reinforcement Learning without Domain Knowledge using Value Function Bounds", "Minimax Regret Bounds for Reinforcement Learning", "PC-MLP: Model-based Reinforcement Learning with Policy Cover Guided Exploration"], "answer_arxiv_id": ["1406.1853", "1811.08540", "1906.03804", "2012.08507", "1901.00210", "1703.05449", "2107.07410"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9438"} +{"question": "What works have applied neural approaches to NL-FOL translation?", "answer": ["Semantic Parsing with Dual Learning", "Formal Specifications from Natural Language", "Logic-Driven Context Extension and Data Augmentation for Logical\n Reasoning of Text", "Exploring Neural Models for Parsing Natural Language into First-Order\n Logic"], "answer_arxiv_id": ["1907.05343", "2206.01962", "2105.03659", "2002.06544"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_9439"} +{"question": "From which papers does the researcher discuss new metrics and understandings that have been suggested to expose the remaining weaknesses of GNNs?", "answer": ["What Do Graph Convolutional Neural Networks Learn?", "Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods", "Revisiting Heterophily For Graph Neural Networks", "Is Homophily a Necessity for Graph Neural Networks?"], "answer_arxiv_id": ["2207.01839", "2110.14446", "2210.07606", "2106.06134"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_9440"} +{"question": "What work projected a Transformer into the space of tree-structured models to uncover its compositionality?", "answer": ["Characterizing Intrinsic Compositionality in Transformers with Tree\n Projections"], "answer_arxiv_id": ["2211.01288"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_9441"} +{"question": "Any works about the incorporation of anti-aliasing techniques into grid-based NeRFs?", "answer": ["Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields", "Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural Radiance Fields"], "answer_arxiv_id": ["2304.06706", "2307.11335"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_9442"} +{"question": "Which studies provided first theoretical justifications behind the empirical findings about the impact of weight collapse on the tightness of variational bounds?", "answer": ["On the Difficulty of Unbiased Alpha Divergence Minimization"], "answer_arxiv_id": ["2010.09541"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_9443"} +{"question": "Could you provide me some works that utilized fuzzy logic in generating complex query embeddings?", "answer": ["FLEX: Feature-Logic Embedding Framework for CompleX Knowledge Graph Reasoning", "Complex Query Answering with Neural Link Predictors", "Logic Embeddings for Complex Query Answering"], "answer_arxiv_id": ["2205.11039", "2011.03459", "2103.00418"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_9444"} +{"question": "What research has been done on increasing zero-shot capabilities of multimodal systems?", "answer": ["Visual Instruction Tuning", "Improved Baselines with Visual Instruction Tuning", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "MiniGPT-v2: large language model as a unified interface for\n vision-language multi-task learning"], "answer_arxiv_id": ["2304.08485", "2310.03744", "2303.16199", "2305.06500", "2304.14178", "2304.10592", "2310.09478"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_9445"} +{"question": "Which contributions are well known in the active setting of BBO, notably Bayesian optimization?", "answer": ["Practical Bayesian Optimization of Machine Learning Algorithms", "Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design"], "answer_arxiv_id": ["1206.2944", "0912.3995"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_9446"} +{"question": "Could you provide a study that proposes a regularization method leveraged from relational information from class word embeddings?", "answer": ["Subspace Regularizers for Few-Shot Class Incremental Learning"], "answer_arxiv_id": ["2110.07059"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_9447"} +{"question": "Could you provide a reference that discusses the effect of bin size in mutual information estimation?", "answer": ["Estimating Information Flow in Deep Neural Networks"], "answer_arxiv_id": ["1810.05728"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_9448"} +{"question": "What works propose penalizing the divergence from the behavioral policy with KL divergence, maximum mean discrepancy (MMD) distance for solving distribution shift and overestimation bias of Q-values?", "answer": ["Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["1906.00949", "1911.11361"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_9449"} +{"question": "Which works identified that deep neural networks are fragile to imperceptible adversarial perturbations?", "answer": ["Intriguing properties of neural networks", "Evasion attacks against machine learning at test time"], "answer_arxiv_id": ["1312.6199", "1708.06131"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9450"} +{"question": "What are the works that link the use of UNet-alike networks with image restoration tasks like image de-raining, and image denoising?", "answer": ["KBNet: Kernel Basis Network for Image Restoration", "Multi-Stage Progressive Image Restoration", "Uformer: A General U-Shaped Transformer for Image Restoration", "Restormer: Efficient Transformer for High-Resolution Image Restoration", "Simple Baselines for Image Restoration", "HINet: Half Instance Normalization Network for Image Restoration", "MemNet: A Persistent Memory Network for Image Restoration"], "answer_arxiv_id": ["2303.02881", "2102.02808", "2106.03106", "2111.09881", "2204.04676", "2105.06086", "1708.02209"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_9451"} +{"question": "What studies proposed supervised methods that utilize attribute classifiers for optimization in GAN's latent space?", "answer": ["GANalyze: Toward Visual Definitions of Cognitive Image Properties", "InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs"], "answer_arxiv_id": ["1906.10112", "2005.09635v2"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9452"} +{"question": "What papers about differentiable rendering and differentiable volumetric rendering in INR for improving realism and control over the generated content?", "answer": ["GeoNeRF: Generalizing NeRF with Geometry Priors", "Advances in Neural Rendering", "Aug-NeRF: Training Stronger Neural Radiance Fields with Triple-Level Physically-Grounded Augmentations", "Control-NeRF: Editable Feature Volumes for Scene Rendering and Manipulation", "NeRF-Editing: Geometry Editing of Neural Radiance Fields", "StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D Mutual Learning", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields"], "answer_arxiv_id": ["2111.13539", "2111.05849", "2207.01164", "2204.10850", "2205.04978", "2205.12183", "2112.05139"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_9453"} +{"question": "Is there a work that mitigates the issue of compromising standard accuracy in AT methods?", "answer": ["Once-for-All Adversarial Training: In-Situ Tradeoff between Robustness and Accuracy for Free"], "answer_arxiv_id": ["2010.11828"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9454"} +{"question": "Which studies introduced the npf model that applies convolutions layers as their encoder?", "answer": ["Convolutional Conditional Neural Processes", "Meta-Learning Stationary Stochastic Process Prediction with Convolutional Neural Processes"], "answer_arxiv_id": ["1910.13556", "2007.01332"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_9455"} +{"question": "Which papers illustrate how exemplar/template retrieval can be advantageous for generating informative responses during dialogue response generation tasks?", "answer": ["Retrieve and Refine: Improved Sequence Generation Models For Dialogue", "Response Generation by Context-aware Prototype Editing", "Skeleton-to-Response: Dialogue Generation Guided by Retrieval Memory"], "answer_arxiv_id": ["1808.04776", "1806.07042", "1809.05296"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_9456"} +{"question": "Which study introduces the clusterness of inlier scores as a quantifying metric?", "answer": ["How to Evaluate the Quality of Unsupervised Anomaly Detection Algorithms?"], "answer_arxiv_id": ["1607.01152"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_9457"} +{"question": "What studies focused on improving the performance of Transformer-based approaches with auxiliary center regression, query initialization and set grouping?", "answer": ["Mask-Attention-Free Transformer for 3D Instance Segmentation"], "answer_arxiv_id": ["2309.01692"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_9458"} +{"question": "Can you provide the papers that first attempted to predict semantic occupancy from image input only?", "answer": ["MonoScene: Monocular 3D Semantic Scene Completion", "Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction"], "answer_arxiv_id": ["2112.00726", "2302.07817"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_9459"} +{"question": "What papers suggest using discrete β-VAE to induce diverse future predictions a policy can condition on in order to solve issues with RCSL?", "answer": ["Addressing Optimism Bias in Sequence Modeling for Reinforcement Learning"], "answer_arxiv_id": ["2207.10295"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_9460"} +{"question": "What research explains the training dynamics of gradient descent in association with the Hessian matrix of the loss function and its sharpness?", "answer": ["Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability"], "answer_arxiv_id": ["2103.00065"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_9461"} +{"question": "Which traditional object counting studies focus on counting cars?", "answer": ["A Large Contextual Dataset for Classification, Detection and Counting of\n Cars with Deep Learning"], "answer_arxiv_id": ["1609.04453"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_9462"} +{"question": "Could you provide me some studies about Prioritized Level Replay?", "answer": ["Replay-Guided Adversarial Environment Design", "Prioritized Level Replay"], "answer_arxiv_id": ["2110.02439", "2010.03934"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_9463"} +{"question": "Could you provide me with studies that established an equivalence between diffusion models and score matching?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_9464"} +{"question": "What papers propose TPV representation for representing the 3D scene?", "answer": ["Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction", "PointOcc: Cylindrical Tri-Perspective View for Point-based 3D Semantic\n Occupancy Prediction"], "answer_arxiv_id": ["2302.07817", "2308.16896"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_9465"} +{"question": "Can you list some works about attention-based pooling for deep Multi-instance learning?", "answer": ["Attention-based Deep Multiple Instance Learning"], "answer_arxiv_id": ["1802.04712"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_9466"} +{"question": "Which papers used perceived signals like gender expression, skin tone, and age in fairness evaluations and bias mitigation techniques in computer vision?", "answer": ["A Step Toward More Inclusive People Annotations for Fairness", "Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigation"], "answer_arxiv_id": ["2105.02317", "1911.11834"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_9467"} +{"question": "What studies introduced Cincer, an interactive label cleaning tool that uses influence functions to identify suspicious example pairs?", "answer": ["Interactive Label Cleaning with Example-based Explanations"], "answer_arxiv_id": ["2106.03922"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_9468"} +{"question": "What research demonstrated that a learned linear transformation is sufficient for BERT to encode image region representations?", "answer": ["What BERT Sees: Cross-Modal Transfer for Visual Question Generation"], "answer_arxiv_id": ["2002.10832"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_9469"} +{"question": "Which paper calls for more diversity in NLP in terms of broader inclusion of other underprivileged people such as those from lower socioeconomic status?", "answer": ["A Survey of Race, Racism, and Anti-Racism in NLP"], "answer_arxiv_id": ["2106.11410"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_9470"} +{"question": "Which paper first discovered adversarial examples?", "answer": ["Intriguing properties of neural networks"], "answer_arxiv_id": ["1312.6199"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_9471"} +{"question": "What work is closest to the researcher's in the context of MARL with communication?", "answer": ["Learning to Communicate with Deep Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1605.06676"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_9472"} +{"question": "What research proposed a pullback metric on the latent space from image space Euclidean metric?", "answer": ["The Riemannian Geometry of Deep Generative Models"], "answer_arxiv_id": ["1711.08014"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_9473"} +{"question": "Who applied Langevin dynamics on the latent space of a GAN to refine its samples?", "answer": ["Refining Deep Generative Models via Discriminator Gradient Flow"], "answer_arxiv_id": ["2012.00780"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_9474"} +{"question": "Which works demonstrate the success of optimistic mirror descent in finding the saddle point?", "answer": ["Optimization, Learning, and Games with Predictable Sequences", "Optimistic Mirror Descent in Saddle-Point Problems: Going the Extra (Gradient) Mile", "Last iterate convergence in no-regret learning: constrained min-max optimization for convex-concave landscapes", "Training GANs with Optimism"], "answer_arxiv_id": ["1311.1869", "1807.02629", "2002.06768", "1711.00141"], "source_meta": {"published_time": "20220619"}, "qid": "AutoScholarQuery_train_9475"} +{"question": "What research proposed Lottery Ticket Hypothesis (LTH) and validated it in other deep learning models?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "The Lottery Ticket Hypothesis for Pre-trained BERT Networks", "Successfully Applying the Stabilized Lottery Ticket Hypothesis to the Transformer Architecture", "A Unified Lottery Ticket Hypothesis for Graph Neural Networks"], "answer_arxiv_id": ["1803.03635", "2007.12223", "2005.03454", "2102.06790"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_9476"} +{"question": "What research papers are about personalized FL (pFL) methods that are based on meta-learning?", "answer": ["Federated Meta-Learning with Fast Convergence and Efficient Communication"], "answer_arxiv_id": ["1802.07876"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_9477"} +{"question": "Which papers focus on new techniques for addressing various Federated Learning (FL) difficulties?", "answer": ["Federated learning with matched averaging", "Advances and Open Problems in Federated Learning", "Federated Learning: Challenges, Methods, and Future Directions", "Federated learning with matched averaging"], "answer_arxiv_id": ["2002.06440", "1912.04977", "1908.07873", "2002.06440"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_9478"} +{"question": "Could you provide me some studies that explore the impact of DP-SGD on gradient direction?", "answer": ["Tempered Sigmoid Activations for Deep Learning with Differential Privacy"], "answer_arxiv_id": ["2007.14191"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_9479"} +{"question": "Are there any studies that performed reconstruction through hardware or additional dedicated scans after the initial capture?", "answer": ["Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2110.07058"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_9480"} +{"question": "What are the recent uplink compression solutions in the context of Federated Learning?", "answer": ["Federated Learning: Strategies for Improving Communication Efficiency", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding"], "answer_arxiv_id": ["1610.05492", "1610.02132"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_9481"} +{"question": "Any studies focused on single-shot image translation using StyleGAN adaptation?", "answer": ["Few-shot Image Generation via Cross-domain Correspondence", "Mind the Gap: Domain Gap Control for Single Shot Domain Adaptation for Generative Adversarial Networks", "One-Shot Adaptation of GAN in Just One CLIP", "JoJoGAN: One Shot Face Stylization"], "answer_arxiv_id": ["2104.06820", "2110.08398", "2203.09301", "2112.11641"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_9482"} +{"question": "What works propose different mathematical tools to model socially interactive behaviors?", "answer": ["CSCNet: Contextual Semantic Consistency Network for Trajectory\n Prediction in Crowded Spaces", "Human Trajectory Prediction via Neural Social Physics"], "answer_arxiv_id": ["2202.08506", "2207.10435"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_9483"} +{"question": "What studies used representations learned from unsupervised contrastive learning to filter out confident samples for noisy label learning?", "answer": ["Jo-SRC: A Contrastive Approach for Combating Noisy Labels", "Multi-Objective Interpolation Training for Robustness to Label Noise", "Selective-Supervised Contrastive Learning with Noisy Labels"], "answer_arxiv_id": ["2103.13029", "2012.04462", "2203.04181"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_9484"} +{"question": "What works utilized the intermediate features of diffusion models for image understanding?", "answer": ["Unsupervised Semantic Correspondence Using Stable Diffusion", "SLiMe: Segment Like Me", "Diffusion Hyperfeatures: Searching Through Time and Space for Semantic\n Correspondence", "A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot\n Semantic Correspondence", "Emergent Correspondence from Image Diffusion"], "answer_arxiv_id": ["2305.15581", "2309.03179", "2305.14334", "2305.15347", "2306.03881"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_9485"} +{"question": "What research papers focus on the earliest NeRF acceleration methods that store precomputed non-view dependent model outputs into finite-resolution structures?", "answer": ["PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Baking Neural Radiance Fields for Real-Time View Synthesis", "FastNeRF: High-Fidelity Neural Rendering at 200FPS", "MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient\n Neural Field Rendering on Mobile Architectures"], "answer_arxiv_id": ["2103.14024", "2103.14645", "2103.10380", "2208.00277"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_9486"} +{"question": "What work proposed a lightweight image encoder known as MobileSAM?", "answer": ["Fast Segment Anything"], "answer_arxiv_id": ["2306.12156"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_9487"} +{"question": "Which studies discuss symmetrization in the context of learning distribution of data augmentations?", "answer": ["Boosting Deep Neural Networks with Geometrical Prior Knowledge: A Survey", "Deep invariant networks with differentiable augmentation layers", "Learning Invariances using the Marginal Likelihood", "Learning Invariant Weights in Neural Networks", "Learning Partial Equivariances from Data", "Invariance Learning in Deep Neural Networks with Differentiable Laplace Approximations"], "answer_arxiv_id": ["2006.16867", "2202.02142", "1808.05563", "2202.12439", "2110.10211", "2202.10638"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_9488"} +{"question": "Could you give me some references of the studies that have been done on large pre-trained vision-language models?", "answer": ["Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "PandaGPT: One Model To Instruction-Follow Them All", "VideoChat: Chat-Centric Video Understanding", "Valley: Video Assistant with Large Language model Enhanced abilitY", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2306.02858", "2305.16355", "2305.06355", "2306.07207", "2301.12597", "2304.10592"], "source_meta": {"published_time": "20240515"}, "qid": "AutoScholarQuery_train_9489"} +{"question": "In which works did they introduce and study the RAP, GEM, and RAP++ algorithms?", "answer": ["Iterative Methods for Private Synthetic Data: Unifying Framework and New Methods", "Private Synthetic Data for Multitask Learning and Marginal Queries"], "answer_arxiv_id": ["2106.07153", "2209.07400"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_9490"} +{"question": "Are there any works using Cleanlab, Data-IQ, and Data Maps for data characterization in tabular data?", "answer": ["Confident Learning: Estimating Uncertainty in Dataset Labels", "Data-IQ: Characterizing subgroups with heterogeneous outcomes in tabular data"], "answer_arxiv_id": ["1911.00068", "2210.13043"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_9491"} +{"question": "What research works explored versions of differentiable top-k operators or sorting functions?", "answer": ["P", "Neural Nearest Neighbors Networks", "Stochastic Optimization of Sorting Networks via Continuous Relaxations", "Fast Differentiable Sorting and Ranking", "Differentiable Sorting Networks for Scalable Sorting and Ranking Supervision", "Monotonic Differentiable Sorting Networks"], "answer_arxiv_id": ["0704.0320", "1810.12575", "1903.08850", "2002.08871", "2105.04019", "2203.09630"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_9492"} +{"question": "Which works use code-based tools to enhance LLMs’ abilities in question answering focusing on tabular and math-related tasks?", "answer": ["PAL: Program-aided Language Models", "MathPrompter: Mathematical Reasoning using Large Language Models"], "answer_arxiv_id": ["2211.10435", "2303.05398"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_9493"} +{"question": "Can you provide me some works studying the concept of self-consistency used in language models?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_9494"} +{"question": "What papers directly address the preservation of the quality in predicted image by matching the patch distribution of images?", "answer": ["Generating natural images with direct Patch Distributions Matching"], "answer_arxiv_id": ["2203.11862"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_9495"} +{"question": "Can you name some studies that proposed distinct types of memory to facilitate long-term object tracking and segmentation?", "answer": ["XMem: Long-Term Video Object Segmentation with an Atkinson-Shiffrin Memory Model"], "answer_arxiv_id": ["2207.07115"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_9496"} +{"question": "Which works are about the application of cascaded diffusion models in image synthesis?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers"], "answer_arxiv_id": ["2205.11487", "2211.01324"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_9497"} +{"question": "Can you name some works that used pretrained vision encoders in large language models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "EVA-CLIP: Improved Training Techniques for CLIP at Scale", "What Makes for Good Visual Tokenizers for Large Language Models?"], "answer_arxiv_id": ["2103.00020", "2303.15389", "2305.12223"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_9498"} +{"question": "Could you provide me some studies that have been proposed to improve the correspondence quality, extend it to more general settings, and learn to generate optimal descriptors?", "answer": ["ZoomOut: Spectral Upsampling for Efficient Shape Correspondence", "Partial Functional Correspondence", "Deep Functional Maps: Structured Prediction for Dense Shape\n Correspondence", "DiffusionNet: Discretization Agnostic Learning on Surfaces"], "answer_arxiv_id": ["1904.07865v4", "1506.05274", "1704.08686", "2012.00888v3"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_9499"} +{"question": "Can you provide works that focus on aligning chess-playing AI systems with human behavior?", "answer": ["Aligning Superhuman AI with Human Behavior: Chess as a Model System"], "answer_arxiv_id": ["2006.01855"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9500"} +{"question": "Could you name some papers related to mitigating discrimination by proposing methods for fair learning with missing values?", "answer": ["Fairness without Imputation: A Decision Tree Approach for Fair Prediction with Missing Values"], "answer_arxiv_id": ["2109.10431"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9501"} +{"question": "Who introduced NeuralODE, a new family of neural network models that parametrizes the continuous dynamics of recurrent neural networks using ordinary differential equations?", "answer": ["Q"], "answer_arxiv_id": ["1611.08152"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_9502"} +{"question": "Can you list the studies which proposed the use of a 'halting unit' in computational modeling?", "answer": ["Adaptive Computation Time for Recurrent Neural Networks"], "answer_arxiv_id": ["1603.08983"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_9503"} +{"question": "Which works have characterized the stabilization pattern of the piece-wise constant step size schedule from the optimization point of view?", "answer": ["Reconciling Modern Deep Learning with Traditional Optimization Analyses: The Intrinsic Learning Rate"], "answer_arxiv_id": ["2010.02916"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_9504"} +{"question": "Which researchers introduce two adapter blocks with bottleneck networks in each Transformer block?", "answer": ["Parameter-Efficient Transfer Learning for NLP"], "answer_arxiv_id": ["1902.00751"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_9505"} +{"question": "In what papers the researchers develop direct Audio-Visual Speech-to-Speech Translation (AV2A)?", "answer": ["AV-TranSpeech: Audio-Visual Robust Speech-to-Speech Translation", "Deep Audio-Visual Speech Recognition", "End-to-end Audio-visual Speech Recognition with Conformers", "Visual Context-driven Audio Feature Enhancement for Robust End-to-End\n Audio-Visual Speech Recognition", "Watch or Listen: Robust Audio-Visual Speech Recognition with Visual\n Corruption Modeling and Reliability Scoring"], "answer_arxiv_id": ["2305.15403", "1809.02108", "2102.06657", "2207.06020", "2303.08536"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_9506"} +{"question": "What research illustrates the issue of overly complex expressions provided by GP methods in practice?", "answer": ["AI Feynman: a Physics-Inspired Method for Symbolic Regression"], "answer_arxiv_id": ["1905.11481"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_9507"} +{"question": "What studies employed the HowTo100M dataset to train a video-language embedding space?", "answer": ["End-to-End Learning of Visual Representations from Uncurated\n Instructional Videos", "HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million\n Narrated Video Clips"], "answer_arxiv_id": ["1912.06430", "1906.03327"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_9508"} +{"question": "Could you provide some works that employ resampling strategy for unsupervised diffusion inpainting?", "answer": ["RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem"], "answer_arxiv_id": ["2201.09865", "2206.04119"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_9509"} +{"question": "Which research introduced fully-supervised 3D detectors?", "answer": ["VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "Center-based 3D Object Detection and Tracking", "SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud", "DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets", "PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud", "From Points to Parts: 3D Object Detection from Point Cloud with\n Part-aware and Part-aggregation Network", "PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection", "Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detection", "STD: Sparse-to-Dense 3D Object Detector for Point Cloud", "Sparse Fuse Dense: Towards High Quality 3D Detection with Depth\n Completion", "Virtual Sparse Convolution for Multimodal 3D Object Detection", "3D Cascade RCNN: High Quality Object Detection in Point Clouds"], "answer_arxiv_id": ["1711.06396", "1812.05784", "2006.11275", "2104.09804", "2301.06051", "1812.04244", "1907.03670", "1912.13192", "2012.15712", "1907.10471", "2203.09780", "2303.02314", "2211.08248"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_9510"} +{"question": "What works developed the method of embedding watermarks during the language model generation phase?", "answer": ["Undetectable Watermarks for Language Models", "Robust Distortion-free Watermarks for Language Models", "A Watermark for Large Language Models"], "answer_arxiv_id": ["2306.09194", "2307.15593", "2301.10226"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_9511"} +{"question": "Could you provide me some studies about deriving upper bounds for r​(g^) when the discrepancy measure is the negative critic loss?", "answer": ["On the Discrimination-Generalization Tradeoff in GANs"], "answer_arxiv_id": ["1711.02771"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_9512"} +{"question": "Which papers used Vision Transformers to predict succeeding pixels or reconstruct missing patches?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers"], "answer_arxiv_id": ["2010.11929", "2103.14030", "2111.06377", "2106.08254"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_9513"} +{"question": "Which toolbox provides MATLAB implementations specifically for six unsupervised signal separation models in remote PPG sensing?", "answer": ["On the Vector Space in Photoplethysmography Imaging"], "answer_arxiv_id": ["1906.04431"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_9514"} +{"question": "Could you provide me some studies about deep learning methods used in SDM?", "answer": ["Deep Multi-Species Embedding", "Multi-Entity Dependence Learning with Rich Context via Conditional Variational Auto-encoder", "Presence-Only Geographical Priors for Fine-Grained Image Classification", "Bird distribution modelling using remote sensing and citizen science data"], "answer_arxiv_id": ["1609.09353v4", "1709.05612", "1906.05272", "2305.01079"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_9515"} +{"question": "Can you list the studies offering a framework to discover latent-specific directions encoding different semantics in diffusion-based modeling?", "answer": ["Understanding the Latent Space of Diffusion Models through the Lens of\n Riemannian Geometry"], "answer_arxiv_id": ["2307.12868"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9516"} +{"question": "What works discuss optimization with compression in distributed and federated settings using quantization technique?", "answer": ["QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "TernGrad: Ternary Gradients to Reduce Communication in Distributed Deep Learning"], "answer_arxiv_id": ["1610.02132", "1705.07878"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_9517"} +{"question": "Which studies examined the use of regularizers to make weight matrices orthogonal during the training of multi-layer perceptron and Convolutional Neural Networks?", "answer": ["Parseval Networks: Improving Robustness to Adversarial Examples"], "answer_arxiv_id": ["1704.08847"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_9518"} +{"question": "Could you provide me with studies that explored attentive contrastive learning on sub-trajectories for pretraining state representations?", "answer": ["Representation Matters: Offline Pretraining for Sequential Decision Making"], "answer_arxiv_id": ["2102.05815"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_9519"} +{"question": "What research focused on multi-source localization within Audio-Visual Sound Localization?", "answer": ["Discriminative Sounding Objects Localization via Self-supervised Audiovisual Matching", "Multiple Sound Sources Localization from Coarse to Fine", "Mix and Localize: Localizing Sound Sources in Mixtures", "Audio-Visual Grouping Network for Sound Localization from Mixtures"], "answer_arxiv_id": ["2010.05466", "2007.06355", "2211.15058", "2303.17056"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_9520"} +{"question": "Which works explored the application of Transformer for model-based RL and multi-task learning?", "answer": ["Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Bootstrapped Transformer for Offline Reinforcement Learning", "A Generalist Agent", "Uni​[MASK]: Unified Inference in Sequential Decision Problems"], "answer_arxiv_id": ["2106.02039", "2206.08569", "2205.06175", "2211.10869"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_9521"} +{"question": "Any works about masked image-text modeling in pretraining strategy for encoder-only vision-language models at small data scale?", "answer": ["Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal Transformers", "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision", "Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "How Much Can CLIP Benefit Vision-and-Language Tasks?", "An Empirical Study of Training End-to-End Vision-and-Language Transformers"], "answer_arxiv_id": ["2004.00849", "2102.03334", "2107.07651", "2107.06383", "2111.02387"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_9522"} +{"question": "What are some studies that augmented a filtered BC objective into the imitation learning procedure?", "answer": ["Watch and Match: Supercharging Imitation with Regularized Optimal Transport"], "answer_arxiv_id": ["2206.15469"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9523"} +{"question": "What research papers used autoregressive generative models for purification methods?", "answer": ["PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examples"], "answer_arxiv_id": ["1710.10766"], "source_meta": {"published_time": "20221101"}, "qid": "AutoScholarQuery_train_9524"} +{"question": "Which studies presented a new simulation benchmark for human-to-robot object handovers?", "answer": ["HandoverSim: A Simulation Framework and Benchmark for Human-to-Robot\n Object Handovers"], "answer_arxiv_id": ["2205.09747"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_9525"} +{"question": "What paper describes the use of CLIP embeddings in relation to diffusion guidance?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_9526"} +{"question": "Which works focus on constructing approximate posteriors or confidence sets in bandit algorithms?", "answer": ["Provably Optimal Algorithms for Generalized Linear Contextual Bandits", "Randomized Exploration in Generalized Linear Bandits", "Neural Contextual Bandits with UCB-based Exploration"], "answer_arxiv_id": ["1703.00048", "1906.08947", "1911.04462"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_9527"} +{"question": "Which work introduced Score Distillation Sampling (SDS) for 3D assets distillation in text-to-3D generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_9528"} +{"question": "What studies explored efficient attention mechanisms for building efficient vision transformers?", "answer": ["Linformer: Self-Attention with Linear Complexity", "Reformer: The Efficient Transformer", "Rethinking Attention with Performers"], "answer_arxiv_id": ["2006.04768", "2001.04451", "2009.14794"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_9529"} +{"question": "What papers propose reduced quantization error in BNNs?", "answer": ["XNOR-Net: ImageNet Classification Using Binary Convolutional Neural\n Networks", "Forward and Backward Information Retention for Accurate Binary Neural\n Networks"], "answer_arxiv_id": ["1603.05279", "1909.10788"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_9530"} +{"question": "Which works develop integrated approaches that combine proposal generation with classification and/or boundary regression in temporal action detection?", "answer": ["End-to-end Learning of Action Detection from Frame Glimpses in Videos", "Single Shot Temporal Action Detection", "Decoupling Localization and Classification in Single Shot Temporal\n Action Detection", "TemporalMaxer: Maximize Temporal Context with only Max Pooling for\n Temporal Action Localization"], "answer_arxiv_id": ["1511.06984", "1710.06236", "1904.07442", "2303.09055"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_9531"} +{"question": "What work can be considered state-of-the-art in the field of DP quantiles?", "answer": ["Differentially Private Approximate Quantiles"], "answer_arxiv_id": ["2110.05429"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_9532"} +{"question": "What research is available regarding compensation for latency and delay in collaborative perception?", "answer": ["V2VNet: Vehicle-to-Vehicle Communication for Joint Perception and Prediction", "V2X-ViT: Vehicle-to-Everything Cooperative Perception with Vision Transformer", "Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting"], "answer_arxiv_id": ["2008.07519", "2203.10638", "1506.04214"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_9533"} +{"question": "What are the relevant works on developing more generalist models that can address diverse problems and quickly adapt to new tasks?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_9534"} +{"question": "Could you provide me studies that sought to explain Graph Neural Networks?", "answer": ["Explainability in Graph Neural Networks: A Taxonomic Survey", "Trustworthy Graph Neural Networks: Aspects, Methods and Trends"], "answer_arxiv_id": ["2012.15445", "2205.07424"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_9535"} +{"question": "Which works does the paper build upon for studying novel view synthesis?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Baking Neural Radiance Fields for Real-Time View Synthesis", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "Differentiable Direct Volume Rendering", "Plenoxels: Radiance Fields without Neural Networks", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Point-NeRF: Point-based Neural Radiance Fields", "ADOP: Approximate Differentiable One-Pixel Point Rendering", "3D Gaussian Splatting for Real-Time Radiance Field Rendering", "SurfelNeRF: Neural Surfel Radiance Fields for Online Photorealistic\n Reconstruction of Indoor Scenes"], "answer_arxiv_id": ["2003.08934", "2103.14645", "2111.11215", "2107.12672", "2112.05131", "2201.05989", "2201.08845", "2110.06635", "2308.04079", "2304.08971"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_9536"} +{"question": "What research has been done on learning textual soft prompts using CLIP-based pseudolabels?", "answer": ["Unsupervised Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2204.03649"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_9537"} +{"question": "Could you provide me some works which employed grid-worlds in their studies on MARL?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments", "Multi-agent Reinforcement Learning in Sequential Social Dilemmas", "Mean Field Multi-Agent Reinforcement Learning", "On the Utility of Learning about Humans for Human-AI Coordination", "Pommerman: A Multi-Agent Playground", "Scalable Evaluation of Multi-Agent Reinforcement Learning with Melting Pot"], "answer_arxiv_id": ["1706.02275", "1702.03037", "1802.05438", "1910.05789", "1809.07124", "2107.06857"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_9538"} +{"question": "What research papers have focused on developing more sample-efficient methods for learning equilibria in two-player zero-sum Markov games?", "answer": ["Provable Self-Play Algorithms for Competitive Reinforcement Learning", "Solving Discounted Stochastic Two-Player Games with Near-Optimal Time and Sample Complexity", "Learning Zero-Sum Simultaneous-Move Markov Games Using Function Approximation and Correlated Equilibrium", "Near-Optimal Reinforcement Learning with Self-Play", "A Sharp Analysis of Model-based Reinforcement Learning with Self-Play", "Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity", "Minimax-Optimal Multi-Agent RL in Markov Games With a Generative Model"], "answer_arxiv_id": ["2002.04017", "1908.11071", "2002.07066", "2006.12007", "2010.01604", "2007.07461", "2208.10458"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_9539"} +{"question": "Which papers replaced prototypes of PSC with predefined fixed ETF?", "answer": ["Neural Collapse with Cross-Entropy Loss", "Inducing Neural Collapse in Imbalanced Learning: Do We Really Need a\n Learnable Classifier at the End of Deep Neural Network?", "Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class\n Incremental Learning"], "answer_arxiv_id": ["2012.08465", "2203.09081", "2302.03004"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_9540"} +{"question": "What studies utilize hierarchical planning methods by generating subgoals in the middle between the existing ones?", "answer": ["Long-Horizon Visual Planning with Goal-Conditioned Hierarchical Predictors", "Divide-and-Conquer Monte Carlo Tree Search For Goal-Directed Planning", "Sub-Goal Trees – a Framework for Goal-Based Reinforcement Learning"], "answer_arxiv_id": ["2006.13205", "2004.11410v1", "2002.12361"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_9541"} +{"question": "Which works discuss the impact of synthetic and natural noise on the translation quality in Neural Machine Translation models?", "answer": ["Synthetic and Natural Noise Both Break Neural Machine Translation", "Findings of the First Shared Task on Machine Translation Robustness", "Evaluating Robustness to Input Perturbations for Neural Machine Translation", "The Unreasonable Volatility of Neural Machine Translation Models"], "answer_arxiv_id": ["1711.02173", "1906.11943", "2005.00580", "2005.12398v1"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_9542"} +{"question": "Could you provide some references that demonstrate the use of autoencoders in Anomaly Detection (AD)?", "answer": ["Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detection", "Learning Memory-guided Normality for Anomaly Detection"], "answer_arxiv_id": ["1904.02639", "2003.13228"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_9543"} +{"question": "Which works discussed the issue of rank collapse of Transformer training?", "answer": ["Attention is not all you need: pure attention loses rank doubly exponentially with depth", "Signal Propagation in Transformers: Theoretical Perspectives and the Role of Rank Collapse"], "answer_arxiv_id": ["2103.03404", "2206.03126"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_9544"} +{"question": "Which works have proposed variants of IoU-based Loss in the context of horizontal detection?", "answer": ["UnitBox: An Advanced Object Detection Network", "Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression", "Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression"], "answer_arxiv_id": ["1608.01471", "1902.09630", "1911.08287"], "source_meta": {"published_time": "20220129"}, "qid": "AutoScholarQuery_train_9545"} +{"question": "Which research papers utilized GANs for text-to-image generation?", "answer": ["Generative Adversarial Networks", "Text to Image Generation with Semantic-Spatial Aware GAN", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image\n Synthesis"], "answer_arxiv_id": ["2203.00667", "2104.00567", "1904.01310"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_9546"} +{"question": "What studies focus on the fully-supervised object detection methods?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", "You Only Look Once: Unified, Real-Time Object Detection", "FCOS: Fully Convolutional One-Stage Object Detection", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["1506.01497", "1506.02640", "1904.01355", "2005.12872"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_9547"} +{"question": "What research introduced FedPop in personalized Federated Learning approach?", "answer": ["FedPop: A Bayesian Approach for Personalised Federated Learning"], "answer_arxiv_id": ["2206.03611"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_9548"} +{"question": "What study combined ShareGPT with Alpaca and then translated the two datasets?", "answer": ["Phoenix: Democratizing ChatGPT across Languages"], "answer_arxiv_id": ["2304.10453"], "source_meta": {"published_time": "20240207"}, "qid": "AutoScholarQuery_train_9549"} +{"question": "What existing works have achieved high accuracy results in the low-label rate regime?", "answer": ["Poisson Learning: Graph Based Semi-Supervised Learning At Very Low Label Rates"], "answer_arxiv_id": ["2006.11184"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_9550"} +{"question": "Can you tell me which researches focus on the fine-grained analysis to locate the neuron associated with knowledge in large language models?", "answer": ["Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2202.05262"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_9551"} +{"question": "What studies have introduced non-neural variants that optimize explicit feature grids?", "answer": ["ReLU Fields: The Little Non-linearity That Could", "Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2205.10824", "2112.05131"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_9552"} +{"question": "What is a work that uses neural implicit in a SLAM system to represent each object in the scene with segmentation priors?", "answer": ["vMAP: Vectorised Object Mapping for Neural Field SLAM"], "answer_arxiv_id": ["2302.01838v2"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_9553"} +{"question": "What paper provides a guarantee for exact low-rank representation of graphs with bounded max degree when using the LPCA factorization model?", "answer": ["Node Embeddings and Exact Low-Rank Representations of Complex Networks"], "answer_arxiv_id": ["2006.05592"], "source_meta": {"published_time": "20211104"}, "qid": "AutoScholarQuery_train_9554"} +{"question": "What work was the most salient in the development of transformer architecture-based summarization models?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_9555"} +{"question": "Which papers presented methods for learning audio-visual representation by establishing a correlation between audio and visual modalities?", "answer": ["SoundNet: Learning Sound Representations from Unlabeled Video", "Ambient Sound Provides Supervision for Visual Learning", "Look, Listen and Learn", "Cooperative Learning of Audio and Video Models from Self-Supervised Synchronization", "Learning to Localize Sound Source in Visual Scenes", "The Sound of Pixels", "The Sound of Motions", "Music Gesture for Visual Sound Separation", "Learning Representations from Audio-Visual Spatial Alignment", "Robust Audio-Visual Instance Discrimination", "Audio-Visual Instance Discrimination with Cross-Modal Agreement", "Looking to Listen at the Cocktail Party: A Speaker-Independent Audio-Visual Model for Speech Separation", "Deep Multimodal Clustering for Unsupervised Audiovisual Learning", "DiffAVA: Personalized Text-to-Audio Generation with Visual Alignment"], "answer_arxiv_id": ["1610.09001", "1608.07017", "1705.08168", "1807.00230", "1803.03849v1", "1804.03160", "1904.05979", "2004.09476", "2011.01819", "2103.15916", "2004.12943", "1804.03619", "1807.03094", "2305.12903"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9556"} +{"question": "What research investigated statistical properties of sliced Wasserstein distance in learning generative models?", "answer": ["Asymptotic Guarantees for Learning Generative Models with the Sliced-Wasserstein Distance"], "answer_arxiv_id": ["1906.04516"], "source_meta": {"published_time": "20220927"}, "qid": "AutoScholarQuery_train_9557"} +{"question": "Which works indicated issues with NLG evaluation metrics through synthetic perturbations?", "answer": ["Perturbation CheckLists for Evaluating NLG Evaluation Metrics", "On the Blind Spots of Model-Based Evaluation Metrics for Text Generation"], "answer_arxiv_id": ["2109.05771", "2212.10020"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_9558"} +{"question": "Can you identify the researches that studied the exploitation semantic relationship between inputs and output in an attention-based model?", "answer": ["aNMM: Ranking Short Answer Texts with Attention-Based Neural Matching Model", "Attentive Neural Processes"], "answer_arxiv_id": ["1801.01641", "1901.05761"], "source_meta": {"published_time": "20220129"}, "qid": "AutoScholarQuery_train_9559"} +{"question": "What studies provide methods for sampling from continuous or differentiable densities, such as Langevin and Hamiltonian MCMC?", "answer": ["MCMC using Hamiltonian dynamics", "The No-U-Turn Sampler: Adaptively Setting Path Lengths in Hamiltonian Monte Carlo"], "answer_arxiv_id": ["1206.1901", "1111.4246"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9560"} +{"question": "Which study introduced the first watermarking method for large language models?", "answer": ["A Watermark for Large Language Models"], "answer_arxiv_id": ["2301.10226"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_9561"} +{"question": "Which works developed distributed optimization algorithms that were provably robust under noise in federated learning?", "answer": ["Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates"], "answer_arxiv_id": ["1803.01498"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_9562"} +{"question": "Which papers propose learning algorithms as solutions to the issue of dull texts generated through MLE?", "answer": ["Negative Training for Neural Dialogue Response Generation", "The Curious Case of Neural Text Degeneration", "Neural Text Generation with Unlikelihood Training"], "answer_arxiv_id": ["1903.02134", "1904.09751", "1908.04319"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_9563"} +{"question": "Which works discuss the use of prompt-tuning methods for adapting vision-language models to downstream tasks?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "PEVL: Position-enhanced Pre-training and Prompt Tuning for Vision-language Models", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention", "Not All Features Matter: Enhancing Few-shot CLIP with Adaptive Prior Refinement", "Personalize Segment Anything Model with One Shot", "Prompt, Generate, then Cache: Cascade of Foundation Models makes Strong Few-shot Learners"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2111.03930", "2110.04544", "2205.11169", "2304.15010", "2303.16199", "2304.01195", "2305.03048", "2303.02151"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_9564"} +{"question": "Which work used images as input data to train a binary classification model for forgery detection, specifically utilizing a straightforward Xception model?", "answer": ["FaceForensics++: Learning to Detect Manipulated Facial Images"], "answer_arxiv_id": ["1901.08971"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_9565"} +{"question": "Which works have leveraged pretrained language models for proposing plans in long-horizon tasks?", "answer": ["Inner Monologue: Embodied Reasoning through Planning with Language Models", "Open-vocabulary Queryable Scene Representations for Real World Planning"], "answer_arxiv_id": ["2207.05608", "2209.09874"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_9566"} +{"question": "Which studies first used proposal-based approaches in video moment retrieval?", "answer": ["TALL: Temporal Activity Localization via Language Query", "Localizing Moments in Video with Natural Language", "Learning 2D Temporal Adjacent Networks for Moment Localization with Natural Language", "Semantic Conditioned Dynamic Modulation for Temporal Sentence Grounding in Videos"], "answer_arxiv_id": ["1705.02101", "1708.01641", "1912.03590", "1910.14303"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_9567"} +{"question": "Which papers discussed the research problems in the AutoML field including model selection, hyperparameter tuning and neural architecture design?", "answer": ["Oboe: Collaborative Filtering for AutoML Model Selection", "Population Based Training of Neural Networks", "Neural Architecture Search with Reinforcement Learning"], "answer_arxiv_id": ["1808.03233", "1711.09846", "1611.01578"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9568"} +{"question": "Could you provide me some studies about generalizing image classification models?", "answer": ["Zero-Shot Learning Through Cross-Modal Transfer"], "answer_arxiv_id": ["1301.3666"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_9569"} +{"question": "What studies showed that language models can provide CoT reasoning through zero-shot prompting?", "answer": ["Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2205.11916"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_9570"} +{"question": "Could you provide me some works proposing contrastive methods and their negative-sample-free variants in the field of unsupervised/self-supervised visual representation learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Representation Learning with Contrastive Predictive Coding", "Contrastive Multiview Coding", "Exploring Simple Siamese Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "On Feature Decorrelation in Self-Supervised Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction"], "answer_arxiv_id": ["2002.05709", "1911.05722", "1807.03748", "1906.05849", "2011.10566", "2006.07733", "2105.00470", "2103.03230"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_9571"} +{"question": "Which works propose model-free regret-based deep learning approaches?", "answer": ["DREAM: Deep Regret Minimization with Advantage Baselines and Model-free Learning", "The Advantage Regret-Matching Actor-Critic"], "answer_arxiv_id": ["2006.10410", "2008.12234"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_9572"} +{"question": "What papers apply logistic regression for the estimation of propensity scores in reweighting-based methods?", "answer": ["Large-scale Causal Approaches to Debiasing Post-click Conversion Rate Estimation with Multi-task Learning", "A Generalized Doubly Robust Learning Framework for Debiasing Post-Click Conversion Rate Prediction", "AutoDebias: Learning to Debias for Recommendation"], "answer_arxiv_id": ["1910.09337", "2211.06684", "2105.04170"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_9573"} +{"question": "Which work introduced the concept of chain-of-thought?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_9574"} +{"question": "Could you show me some researches that applied masked autoencoding in computer vision?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers"], "answer_arxiv_id": ["2111.06377", "2106.08254"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_9575"} +{"question": "Are there any works about how models trained on the poisoned dataset will malfunction?", "answer": ["Wild Patterns Reloaded: A Survey of Machine Learning Security against Training Data Poisoning"], "answer_arxiv_id": ["2205.01992"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_9576"} +{"question": "Could you provide me some studies about enhancing generation controllability in HSI with semantic guidance?", "answer": ["Compositional Human-Scene Interaction Synthesis with Semantic Control", "HUMANISE: Language-conditioned Human Motion Generation in 3D Scenes", "Unified Human-Scene Interaction via Prompted Chain-of-Contacts"], "answer_arxiv_id": ["2207.12824", "2210.09729", "2309.07918"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_9577"} +{"question": "Which papers address evaluating machine unlearning?", "answer": ["Evaluating Machine Unlearning via Epistemic Uncertainty", "On the Necessity of Auditable Algorithmic Definitions for Machine Unlearning", "Unrolling SGD: Understanding Factors Influencing Machine Unlearning"], "answer_arxiv_id": ["2208.10836", "2110.11891v2", "2109.13398"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_9578"} +{"question": "Which works indicate the importance of retrieval technology in the field of large language models (LLMs)?", "answer": ["Siren's Song in the AI Ocean: A Survey on Hallucination in Large\n Language Models", "Retrieval-Augmented Generation for Large Language Models: A Survey"], "answer_arxiv_id": ["2309.01219", "2312.10997"], "source_meta": {"published_time": "20240225"}, "qid": "AutoScholarQuery_train_9579"} +{"question": "Which studies investigated extraction attacks that recover memorized data in neural language models?", "answer": ["Extracting Training Data from Large Language Models", "The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks", "Ethical Challenges in Data-Driven Dialogue Systems", "Understanding Unintended Memorization in Federated Learning"], "answer_arxiv_id": ["2012.07805", "1802.08232", "1711.09050", "2006.07490"], "source_meta": {"published_time": "20220215"}, "qid": "AutoScholarQuery_train_9580"} +{"question": "Which research papers discussed the use of physics simulators in modeling contact dynamics in robotics?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations", "Physics-Based Dexterous Manipulations with Estimated Hand Poses and Residual Reinforcement Learning", "Learning Task-Oriented Grasping from Human Activity Datasets", "Learning to Estimate Pose and Shape of Hand-Held Objects from RGB Images"], "answer_arxiv_id": ["1709.10087", "2008.03285", "1910.11669", "1903.03340v3"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_9581"} +{"question": "What works developed the Score Distillation Sampling method in text-to-3D works?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "EfficientDreamer: High-Fidelity and Robust 3D Creation via\n Orthogonal-view Diffusion Prior"], "answer_arxiv_id": ["2211.10440", "2211.07600", "2303.13873", "2305.16213", "2308.13223"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_9582"} +{"question": "What works have been developed inspired by ChatGPT and GPT-4 in biomedical chatbots?", "answer": ["MedAlpaca - An Open-Source Collection of Medical Conversational AI Models and Training Data", "DoctorGLM: Fine-tuning your Chinese Doctor is not a Herculean Task"], "answer_arxiv_id": ["2304.08247", "2304.01097"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9583"} +{"question": "Which studies are related to improving pseudo-labeling and consistency regularization techniques through loss weighting?", "answer": ["Temporal Ensembling for Semi-Supervised Learning", "Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "Label Propagation for Deep Semi-supervised Learning", "Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning"], "answer_arxiv_id": ["1610.02242v3", "1703.01780", "1904.04717", "2007.01293"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_9584"} +{"question": "Which papers use knowledge distillation-based methods that require a global dataset?", "answer": ["FedMD: Heterogenous Federated Learning via Model Distillation"], "answer_arxiv_id": ["1910.03581"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_9585"} +{"question": "Who worked on fine-tuning the creation of personalized images?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_9586"} +{"question": "Who proposed using the successful practice of stochastic training in nonlinear extensions of CoxPH model?", "answer": ["DeepSurv: Personalized Treatment Recommender System Using A Cox Proportional Hazards Deep Neural Network", "Train faster, generalize better: Stability of stochastic gradient descent"], "answer_arxiv_id": ["1606.00931", "1509.01240"], "source_meta": {"published_time": "20230318"}, "qid": "AutoScholarQuery_train_9587"} +{"question": "Which papers shed light on Energy-based Models (EBMs) and their role in generative modeling?", "answer": ["On the Anatomy of MCMC-Based Maximum Likelihood Learning of Energy-Based Models", "Learning Non-Convergent Non-Persistent Short-Run MCMC Toward Energy-Based Model", "A Theory of Generative ConvNet", "Implicit Generation and Generalization in Energy-Based Models", "Improved Contrastive Divergence Training of Energy Based Models"], "answer_arxiv_id": ["1903.12370", "1904.09770", "1602.03264v3", "1903.08689", "2012.01316v4"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_9588"} +{"question": "Which studies have proposed applying the same threshold to all examples during membership inference?", "answer": ["Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "Membership Inference Attacks Against Machine Learning Models"], "answer_arxiv_id": ["1709.01604", "1610.05820"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_9589"} +{"question": "Which studies try to tackle the issue of huge training and inference costs in diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models"], "answer_arxiv_id": ["2112.10752", "2206.00927", "2211.01095", "2010.02502", "2201.06503", "2206.07309"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_9590"} +{"question": "Which research introduced learnable step-size scale factors for quantization functions?", "answer": ["Learned Step Size Quantization"], "answer_arxiv_id": ["1902.08153"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_9591"} +{"question": "Which work proposed to understand the learning dynamics of end-to-end linear TD?", "answer": ["On The Effect of Auxiliary Tasks on Representation Dynamics"], "answer_arxiv_id": ["2102.13089v1"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_9592"} +{"question": "Any works about employing explicit representations for direct dynamic scene modeling?", "answer": ["TensoRF: Tensorial Radiance Fields", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "Tensor4D : Efficient Neural 4D Decomposition for High-fidelity Dynamic\n Reconstruction and Rendering", "HexPlane: A Fast Representation for Dynamic Scenes"], "answer_arxiv_id": ["2203.09517", "2301.10241", "2211.11610", "2301.09632"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_9593"} +{"question": "What studies observed the epoch-wise double descent behavior of testing loss of deep neural networks?", "answer": ["Deep Double Descent: Where Bigger Models and More Data Hurt"], "answer_arxiv_id": ["1912.02292"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_9594"} +{"question": "Which works proposed using scale-invariant depth losses in monocular depth estimation?", "answer": ["Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer", "Vision Transformers for Dense Prediction", "Depth Map Prediction from a Single Image using a Multi-Scale Deep\n Network", "Digging Into Self-Supervised Monocular Depth Estimation"], "answer_arxiv_id": ["1907.01341v3", "2103.13413", "1406.2283", "1806.01260"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_9595"} +{"question": "Which works introduced various techniques to improve diffusion models for image generation tasks?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_9596"} +{"question": "Are there any works on mitigating the effect of known biases during training in machine learning?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Unlearn Dataset Bias in Natural Language Inference by Fitting the Residual", "Don’t Take the Premise for Granted: Mitigating Artifacts in Natural Language Inference", "Don’t Take the Easy Way Out: Ensemble Based Methods for Avoiding Known Dataset Biases"], "answer_arxiv_id": ["1911.08731", "1908.10763", "1907.04380", "1909.03683"], "source_meta": {"published_time": "20220629"}, "qid": "AutoScholarQuery_train_9597"} +{"question": "Which papers discussed using random access memory in memory-augmented networks?", "answer": ["Associative Long Short-Term Memory", "Neural Random-Access Machines"], "answer_arxiv_id": ["1602.03032", "1511.06392v3"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_9598"} +{"question": "Which papers have discussed the limitations of traditional single-stream models in multi-modal tasks?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "VisualBERT: A Simple and Performant Baseline for Vision and Language"], "answer_arxiv_id": ["1908.02265", "1908.03557"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_9599"} +{"question": "In which studies researchers utilized transformer-based architectures for the interpretation and comprehension of digital documents?", "answer": ["LayoutLMv3: Pre-training for Document AI with Unified Text and Image\n Masking"], "answer_arxiv_id": ["2204.08387"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_9600"} +{"question": "Which studies have represented images through rendering processes with specific parameters such as Neural Radiance Fields, for the purpose of leveraging the 2D prior of diffusion models?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "VectorFusion: Text-to-SVG by Abstracting Pixel-Based Diffusion Models"], "answer_arxiv_id": ["2209.14988", "2212.00774v1", "2211.11319"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_9601"} +{"question": "What research attempts to combine MAPPO with a graph generator in the effort to augment policy-based algorithms with correlated execution?", "answer": ["GCS: Graph-Based Coordination Strategy for Multi-Agent Reinforcement Learning", "The Surprising Effectiveness of PPO in Cooperative Multi-Agent Games"], "answer_arxiv_id": ["2201.06257", "2103.01955"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_9602"} +{"question": "Are there any works which proposed algorithms for correlation clustering on complete graphs?", "answer": ["Parallel Correlation Clustering on Big Graphs", "Correlation Clustering via Strong Triadic Closure Labeling: Fast Approximation Algorithms and Practical Lower Bounds"], "answer_arxiv_id": ["1507.05086", "2111.10699"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_9603"} +{"question": "Which works propose task-specific subnetworks to handle data imbalances in the FSCIL setting?", "answer": ["Few-Shot Lifelong Learning"], "answer_arxiv_id": ["2103.00991"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_9604"} +{"question": "What are some papers that discuss the concept of neural topic models?", "answer": ["Neural Variational Inference for Text Processing", "The Dynamic Embedded Topic Model", "Discriminative Topic Mining via Category-Name Guided Text Embedding", "Contrastive Learning for Neural Topic Model", "InfoCTM: A Mutual Information Maximization Perspective of Cross-Lingual Topic Modeling"], "answer_arxiv_id": ["1511.06038", "1907.05545", "1908.07162", "2110.12764", "2304.03544"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_9605"} +{"question": "What works have conducted theoretical analysis of machine learning algorithms using methods from theoretical physics, especially glassy physics?", "answer": ["Machine learning and the physical sciences", "Learning curves of generic features maps for realistic datasets with a teacher-student model", "Generalization Error Rates in Kernel Regression: The Crossover from the Noiseless to Noisy Regime", "Double Trouble in Double Descent: Bias and Variance(s) in the Lazy Regime", "Asymptotic Errors for Teacher-Student Convex Generalized Linear Models (or : How to Prove Kabashima’s Replica Formula)", "Generalization error in high-dimensional perceptrons: Approaching Bayes error with convex optimization", "High Dimensional Robust M-Estimation: Asymptotic Variance via Approximate Message Passing"], "answer_arxiv_id": ["1903.10563", "2102.08127", "2105.15004", "2003.01054", "2006.06581", "2006.06560", "1310.7320"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_9606"} +{"question": "Can you list some works that introduced time-series Transformers for forecasting tasks?", "answer": ["Deep Transformer Models for Time Series Forecasting: The Influenza Prevalence Case", "A Transformer-based Framework for Multivariate Time Series Representation Learning"], "answer_arxiv_id": ["2001.08317", "2010.02803"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_9607"} +{"question": "Which papers focused on face synthesis via Generative Adversarial Networks (GANs)?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Designing an Encoder for StyleGAN Image Manipulation", "Depth-Aware Generative Adversarial Network for Talking Head Video\n Generation", "Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation", "PIE: Portrait Image Embedding for Semantic Control", "MyStyle: A Personalized Generative Prior"], "answer_arxiv_id": ["1812.04948", "2102.02766", "2203.06605", "2008.00951", "2009.09485", "2203.17272"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9608"} +{"question": "What works present details about curiosity-driven learning through surprise?", "answer": ["Surprise-Based Intrinsic Motivation for Deep Reinforcement Learning"], "answer_arxiv_id": ["1703.01732"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_9609"} +{"question": "Any works about the use of differentiable simulators in GBI?", "answer": ["Robust Bayesian Inference for Simulator-based Models via the MMD Posterior Bootstrap", "MMD-Bayes: Robust Bayesian Estimation via Maximum Mean Discrepancy"], "answer_arxiv_id": ["2202.04744v3", "1909.13339"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_9610"} +{"question": "Could you provide me some works developing methods to fine-tune the CLIP encoder to better recognize regions?", "answer": ["Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2210.04150", "2112.01071"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_9611"} +{"question": "What works have discussed the concept of pruning in training?", "answer": ["The State of Sparsity in Deep Neural Networks", "Rigging the Lottery: Making All Tickets Winners"], "answer_arxiv_id": ["1902.09574", "1911.11134"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_9612"} +{"question": "What research efforts have been made to lessen the dependency on costly manual annotations in 3D instance segmentation?", "answer": ["Weakly Supervised Semantic Point Cloud Segmentation:Towards 10X Fewer\n Labels", "Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts", "SSPC-Net: Semi-supervised Semantic 3D Point Cloud Segmentation Network", "Guided Point Contrastive Learning for Semi-supervised Point Cloud\n Semantic Segmentation", "One Thing One Click: A Self-Training Approach for Weakly Supervised 3D\n Semantic Segmentation", "Weakly Supervised Semantic Point Cloud Segmentation:Towards 10X Fewer\n Labels", "Weakly Supervised Semantic Segmentation for Large-Scale Point Cloud", "Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using\n Bounding Boxes"], "answer_arxiv_id": ["2004.04091", "2012.09165", "2104.07861", "2110.08188", "2104.02246", "2004.04091", "2212.04744", "2206.01203"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_9613"} +{"question": "Which studies are mentioned as part of the research area known as dynamic graph learning?", "answer": ["Lifelong Learning of Graph Neural Networks for Open-World Node Classification", "Streaming Graph Neural Networks via Continual Learning", "Graph Neural Networks with Continual Learning for Fake News Detection from Social Media", "Streaming Graph Neural Networks", "FILDNE: A Framework for Incremental Learning of Dynamic Networks Embeddings"], "answer_arxiv_id": ["2006.14422v4", "2009.10951", "2007.03316", "1810.10627", "1904.03423"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_9614"} +{"question": "Could you provide me some studies about variational optimization?", "answer": ["Variational Optimization"], "answer_arxiv_id": ["1212.4507"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_9615"} +{"question": "Which research papers have explored the use of various geometrical representations like voxels in learning-based 3D reconstruction?", "answer": ["Learning a Multi-View Stereo Machine", "DeepVoxels: Learning Persistent 3D Feature Embeddings"], "answer_arxiv_id": ["1708.05375", "1812.01024"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_9616"} +{"question": "Which works addressed the problem of spatio-temporal grounding?", "answer": ["TubeDETR: Spatio-Temporal Video Grounding with Transformers", "Embracing Consistency: A One-Stage Approach for Spatio-Temporal Video\n Grounding", "Human-centric Spatio-Temporal Video Grounding With Visual Transformers", "Where Does It Exist: Spatio-Temporal Video Grounding for Multi-Form\n Sentences"], "answer_arxiv_id": ["2203.16434", "2209.13306", "2011.05049", "2001.06891"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_9617"} +{"question": "What research uses photometric stereo to estimate normal maps from images under varying light conditions?", "answer": ["Scalable, Detailed and Mask-Free Universal Photometric Stereo"], "answer_arxiv_id": ["2303.15724"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9618"} +{"question": "What paper has considered non-bilinear coupling terms while striving for optimal results?", "answer": ["Sharper Rates for Separable Minimax and Finite Sum Optimization via Primal-Dual Extragradient Methods"], "answer_arxiv_id": ["2202.04640"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_9619"} +{"question": "Which papers provide theoretical analysis for PINNs specifically for equations such as Kolmogorov equations and Navier-Stokes equations?", "answer": ["Error Analysis for Physics Informed Neural Networks (PINNs) approximating Kolmogorov PDEs", "Error estimates for physics-informed neural networks approximating the Navier-Stokes equations"], "answer_arxiv_id": ["2106.14473", "2203.09346"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_9620"} +{"question": "What studies adopt SDF to implicitly represent geometric shapes in the task of new view synthesis?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface\n Reconstruction", "Voxurf: Voxel-based Efficient and Accurate Neural Surface Reconstruction"], "answer_arxiv_id": ["2106.10689", "2206.00665", "2208.12697"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_9621"} +{"question": "In what literature are adversarial imitation learning approaches found?", "answer": ["Generative Adversarial Imitation from Observation", "AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control"], "answer_arxiv_id": ["1807.06158", "2104.02180"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9622"} +{"question": "What studies conducted 2D object detection from the 2D input image?", "answer": ["RegionCLIP: Region-based Language-Image Pretraining", "Detecting Twenty-thousand Classes using Image-level Supervision", "Grounded Language-Image Pre-training", "MDETR - Modulated Detection for End-to-End Multi-Modal Understanding"], "answer_arxiv_id": ["2112.09106", "2201.02605", "2112.03857", "2104.12763"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_9623"} +{"question": "Could you mention some works that developed SSL methods based on masked image modeling?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers", "VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training"], "answer_arxiv_id": ["2111.06377", "2106.08254", "2203.12602"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_9624"} +{"question": "Which research has focused on understanding the double descent phenomenon?", "answer": ["Two models of double descent for weak features", "Surprises in High-Dimensional Ridgeless Least Squares Interpolation", "The generalization error of random features regression: Precise asymptotics and double descent curve", "A Random Matrix Analysis of Random Fourier Features: Beyond the Gaussian Kernel, a Precise Phase Transition, and the Corresponding Double Descent"], "answer_arxiv_id": ["1903.07571", "1903.08560", "1908.05355", "2006.05013"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_9625"} +{"question": "Is there any concurrent work that uses a weighted-mixing strategy for disentangling editing targets?", "answer": ["Uncovering the Disentanglement Capability in Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2212.08698"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_9626"} +{"question": "What research papers discuss helping practitioners understand model’s underlying rationale to debug models in machine learning?", "answer": ["Post hoc Explanations may be Ineffective for Detecting Unknown Spurious Correlation", "Interpretations are Useful: Penalizing Explanations to Align Neural Networks with Prior Knowledge"], "answer_arxiv_id": ["2212.04629", "1909.13584"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_9627"} +{"question": "Which papers explore ODE-based samplers and their usefulness in taking large time steps?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "Fast Sampling of Diffusion Models with Exponential Integrator"], "answer_arxiv_id": ["2112.10752", "2206.00927", "2204.13902"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_9628"} +{"question": "Which work proposed the original PC model for memory that follows a hierarchical and generative structure?", "answer": ["Associative Memories via Predictive Coding"], "answer_arxiv_id": ["2109.08063"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_9629"} +{"question": "What is the work that uses stable diffusion for image watermarking?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_9630"} +{"question": "Which papers proposed topology-aware isometric initialization for model parameters?", "answer": ["Old can be Gold: Better Gradient Flow can Make Vanilla-GCNs Great Again"], "answer_arxiv_id": ["2210.08122"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_9631"} +{"question": "Could you provide me the related work that reason theoretically about the ability of learning under topic-modeling-based distributions?", "answer": ["On some provably correct cases of variational inference for topic models", "Provable Algorithms for Inference in Topic Models", "Contrastive estimation reveals topic posterior information to linear models"], "answer_arxiv_id": ["1503.06567", "1605.08491", "2003.02234"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_9632"} +{"question": "Which works proposed different binning methods for evaluating calibration?", "answer": ["Verified Uncertainty Calibration", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration", "Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning", "A Consistent and Differentiable Lp Canonical Calibration Error Estimator"], "answer_arxiv_id": ["1909.10155", "1910.12656", "2003.07329", "2210.07810v1"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_9633"} +{"question": "Could you provide the reference which proposed an upper confidence bound-based algorithm for LinCBwK with a single anytime constraint?", "answer": ["Stochastic Bandits with Linear Constraints"], "answer_arxiv_id": ["2006.10185"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_9634"} +{"question": "Which works have used fitted value or action-value functions in the study of offline RL methods?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Offline Reinforcement Learning with Implicit Q-Learning", "Conservative Q-Learning for Offline Reinforcement Learning", "MOReL: Model-Based Offline Reinforcement Learning", "Mildly Conservative Q-Learning for Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "2110.06169", "2006.04779", "2005.05951", "2206.04745v3"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_9635"} +{"question": "Can you provide me studies which used Neural Radiance Fields in unsupervised object-centric learning?", "answer": ["Decomposing 3D Scenes into Objects via Unsupervised Volume Segmentation", "Unsupervised Discovery and Composition of Object Light Fields"], "answer_arxiv_id": ["2104.01148", "2205.03923"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_9636"} +{"question": "What research has focused on batched dueling bandits and batched convex optimization in low-adaptivity learning domains?", "answer": ["Batched Dueling Bandits"], "answer_arxiv_id": ["2202.10660v1"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_9637"} +{"question": "What references are about implicit augmentation methods in sequential recommendation?", "answer": ["Contrastive Learning for Representation Degeneration Problem in Sequential Recommendation"], "answer_arxiv_id": ["2110.05730"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_9638"} +{"question": "Could you tell me the studies related to variant of symmetric network, SimSiam?", "answer": ["Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2011.10566"], "source_meta": {"published_time": "20220216"}, "qid": "AutoScholarQuery_train_9639"} +{"question": "What are some previous works on fully quantized training methods?", "answer": ["Mixed Precision Training", "Training Deep Neural Networks with 8-bit Floating Point Numbers", "Scalable Methods for 8-bit Training of Neural Networks", "Training DNNs with Hybrid Block Floating Point", "Faster Neural Network Training with Approximate Tensor Operations", "Training and Inference with Integers in Deep Neural Networks", "Adaptive Precision Training: Quantify Back Propagation in Neural Networks with Fixed-point Numbers", "Deep Learning Training on the Edge with Low-Precision Posits", "Cheetah: Mixed Low-Precision Hardware & Software Co-Design Framework for DNNs on the Edge", "Training High-Performance and Large-Scale Deep Neural Networks with Full 8-bit Integers", "Towards Unified INT8 Training for Convolutional Neural Network"], "answer_arxiv_id": ["1710.03740", "1812.08011", "1805.11046", "1804.01526", "1805.08079", "1802.04680", "1911.00361v2", "1907.13216", "1908.02386", "1909.02384", "1912.12607"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_9640"} +{"question": "Which works utilize language for guiding the learning of image features?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_9641"} +{"question": "What studies contributed to the first category of vision-language models which includes early bililinear pooling and attention based multimodal models?", "answer": ["MUTAN: Multimodal Tucker Fusion for Visual Question Answering", "Bilinear Attention Networks", "Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering", "Dynamic Fusion with Intra- and Inter-modality Attention Flow for Visual Question Answering"], "answer_arxiv_id": ["1705.06676", "1805.07932", "1707.07998", "1812.05252"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_9642"} +{"question": "What papers proposed transformer-based methods for dance synthesis?", "answer": ["AI Choreographer: Music Conditioned 3D Dance Generation with AIST++", "Bailando: 3D Dance Generation by Actor-Critic GPT with Choreographic\n Memory"], "answer_arxiv_id": ["2101.08779", "2203.13055"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_9643"} +{"question": "Which studies discussed the application of contrastive learning in the context of point clouds?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Masked Scene Contrast: A Scalable Framework for Unsupervised 3D\n Representation Learning"], "answer_arxiv_id": ["2007.10985", "2303.14191"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_9644"} +{"question": "Can you give an example of the research that introduced benchmarks for stylistic analysis?", "answer": ["Style is NOT a single variable: Case Studies for Cross-Style Language\n Understanding"], "answer_arxiv_id": ["1911.03663"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_9645"} +{"question": "Can you provide the source that adapted language models for other tasks such as reasoning across discrete sequences and few-shot image classification?", "answer": ["Pretrained Transformers As Universal Computation Engines", "Multimodal Few-Shot Learning with Frozen Language Models"], "answer_arxiv_id": ["2103.05247", "2106.13884"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9646"} +{"question": "What works are based on reconstruction-related geometry understanding tasks in studying 3D point clouds?", "answer": ["Self-Supervised Deep Learning on Point Clouds by Reconstructing Space", "Self-Supervised Learning of Point Clouds via Orientation Estimation", "Global-Local Bidirectional Reasoning for Unsupervised Representation Learning of 3D Point Clouds", "Just Go with the Flow: Self-Supervised Scene Flow Estimation", "Self-Supervised Learning for Domain Adaptation on Point Clouds", "Unsupervised Point Cloud Pre-training via Occlusion Completion"], "answer_arxiv_id": ["1901.08396", "2008.00305", "2003.12971", "1912.00497", "2003.12641", "2010.01089"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_9647"} +{"question": "What papers proposed advanced generative models that can generate 3D molecular conformations?", "answer": ["Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules", "E(n) Equivariant Normalizing Flows", "Equivariant Diffusion for Molecule Generation in 3D", "Diffusion-based Molecule Generationwith Informative Prior Bridges"], "answer_arxiv_id": ["1906.00957", "2105.09016", "2203.17003", "2209.00865"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_9648"} +{"question": "What papers propose a novel architecture that inherently exhibits Lipschitzness properties?", "answer": ["Towards Certifying L-infinity Robustness using Neural Networks with L-inf-dist Neurons"], "answer_arxiv_id": ["2102.05363v4"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_9649"} +{"question": "Which papers introduced the concept of the NeuroLogic decoding?", "answer": ["NeuroLogic Decoding: (Un)supervised Neural Text Generation with Predicate Logic Constraints"], "answer_arxiv_id": ["2010.12884"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_9650"} +{"question": "Which works did ADKF-IFT build upon by generalizing their techniques?", "answer": ["Deep Kernel Learning", "Bayesian Meta-Learning for the Few-Shot Setting via Deep Kernels"], "answer_arxiv_id": ["1511.02222", "1910.05199"], "source_meta": {"published_time": "20220505"}, "qid": "AutoScholarQuery_train_9651"} +{"question": "Could you provide some studies that discussed the performance constraints of constructive models with heavy encoder and light decoder structures?", "answer": ["Learning the Travelling Salesperson Problem Requires Rethinking Generalization"], "answer_arxiv_id": ["2006.07054"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_9652"} +{"question": "Which works explored the paradigm of learning representations through masking and reconstructing in NLP?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "ALBERT: A Lite BERT for Self-supervised Learning of Language\n Representations"], "answer_arxiv_id": ["1810.04805", "1907.11692", "1909.11942"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_9653"} +{"question": "Who mentioned the similarity of the bottleneck structure described here with the Information Bottleneck theory?", "answer": ["Deep Learning and the Information Bottleneck Principle"], "answer_arxiv_id": ["1503.02406"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9654"} +{"question": "Which studies propose adversarial attacks that can fool image classification networks?", "answer": ["Intriguing properties of neural networks", "Adversarial Machine Learning at Scale", "Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks", "Square Attack: a query-efficient black-box adversarial attack via random search", "RobustBench: a standardized adversarial robustness benchmark"], "answer_arxiv_id": ["1312.6199", "1611.01236", "2003.01690", "1912.00049", "2010.09670"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_9655"} +{"question": "Which studies focused on constrained reinforcement learning?", "answer": ["A Reduction-Based Framework for Conservative Bandits and Reinforcement Learning", "Constrained Policy Optimization", "Projection-Based Constrained Policy Optimization", "Reinforcement Learning with Almost Sure Constraints", "Safe Reinforcement Learning with Linear Function Approximation", "Provably Efficient Safe Exploration via Primal-Dual Policy Optimization", "Provably Efficient Model-Free Constrained RL with Linear Function Approximation", "Exploration-Exploitation in Constrained MDPs", "Near-Optimal Sample Complexity Bounds for Constrained MDPs"], "answer_arxiv_id": ["2106.11692", "1705.10528", "2010.03152", "2112.05198v3", "2106.06239", "2003.00534", "2206.11889", "2003.02189", "2206.06270"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_9656"} +{"question": "Could you provide me some works about studying inference scaling in a production setting?", "answer": ["Efficiently Scaling Transformer Inference"], "answer_arxiv_id": ["2211.05102v1"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_9657"} +{"question": "What studies executed event-based LIE, using events as guidance for low-light image enhancement?", "answer": ["Low-Light Video Enhancement with Synthetic Event Guidance"], "answer_arxiv_id": ["2208.11014"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_9658"} +{"question": "Which works introduced a benchmark for testing robustness to group shifts?", "answer": ["Breeds: Benchmarks for Subpopulation Shift", "Wilds: A Benchmark of in-the-Wild Distribution Shifts"], "answer_arxiv_id": ["2008.04859", "2012.07421"], "source_meta": {"published_time": "20221202"}, "qid": "AutoScholarQuery_train_9659"} +{"question": "What research are there on using a multi-objective Q-learning approach that simultaneously learns a set of policies over multiple preferences in MORL?", "answer": ["Dynamic Weights in Multi-Objective Deep Reinforcement Learning", "A Generalized Algorithm for Multi-Objective Reinforcement Learning and Policy Adaptation"], "answer_arxiv_id": ["1809.07803", "1908.08342"], "source_meta": {"published_time": "20220816"}, "qid": "AutoScholarQuery_train_9660"} +{"question": "What studies have discussed a ranked reward mechanism in self-competition?", "answer": ["Ranked Reward: Enabling Self-Play Reinforcement Learning for Combinatorial Optimization"], "answer_arxiv_id": ["1807.01672"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_9661"} +{"question": "What papers developed a generalization of cubic splines in the space of probability measures?", "answer": ["Measure-valued spline curves: an optimal transport viewpoint", "Second order models for optimal transport and cubic splines on the Wasserstein space"], "answer_arxiv_id": ["1801.03186", "1801.04144"], "source_meta": {"published_time": "20220518"}, "qid": "AutoScholarQuery_train_9662"} +{"question": "Could you provide me some works that used functional map paradigm for nonrigid shape matching?", "answer": ["DPFM: Deep Partial Functional Maps", "Continuous and Orientation-preserving Correspondences via Functional Maps", "ZoomOut: Spectral Upsampling for Efficient Shape Correspondence", "Partial Functional Correspondence", "Non-Rigid Puzzles", "Understanding and Improving Features Learned in Deep Functional Maps", "AtomSurf: Surface Representation for Learning on Protein Structures"], "answer_arxiv_id": ["2110.09994v1", "1806.04455v3", "1904.07865v4", "1506.05274", "2011.13076", "2303.16527", "2309.16519"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_9663"} +{"question": "Which studies present methods that model garments as distinct surfaces that interact with the body?", "answer": ["Multi-Garment Net: Learning to Dress 3D People from Images", "BCNet: Learning Body and Cloth Shape from A Single Image", "PERGAMO: Personalized 3D Garments from Monocular Video", "SMPLicit: Topology-aware Generative Model for Clothed People", "DIG: Draping Implicit Garment over the Human Body", "3D Clothed Human Reconstruction in the Wild", "MeshUDF: Fast and Differentiable Meshing of Unsigned Distance Field\n Networks", "ISP: Multi-Layered Garment Draping with Implicit Sewing Patterns"], "answer_arxiv_id": ["1908.06903", "2004.00214", "2210.15040", "2103.06871", "2209.10845", "2207.10053", "2111.14549", "2305.14100"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_9664"} +{"question": "Which work adopted the Bayesian approach to formalize training reconstruction attacks?", "answer": ["Reconstructing Training Data with Informed Adversaries"], "answer_arxiv_id": ["2201.04845"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_9665"} +{"question": "What studies attempted to apply VAE-LSTM-based world models to Atari games leading to increased sample efficiency?", "answer": ["Model Based Reinforcement Learning for Atari"], "answer_arxiv_id": ["1903.00374"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_9666"} +{"question": "Which work generalized the bandit problem to linear MDPs and established a regret bound?", "answer": ["A Provably Efficient Algorithm for Linear Markov Decision Process with Low Switching Cost"], "answer_arxiv_id": ["2101.00494v1"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_9667"} +{"question": "Can you provide me some studies about introduction of fixed-time confidence bands for the off-policy CDF in contextual bandit problems?", "answer": ["Universal Off-Policy Evaluation", "Off-Policy Risk Assessment in Contextual Bandits"], "answer_arxiv_id": ["2104.12820", "2104.08977v2"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_9668"} +{"question": "Can you provide me with some research about Collaborative Fairness in federated fairness?", "answer": ["Collaborative Fairness in Federated Learning", "Fair Federated Medical Image Segmentation via Client Contribution\n Estimation", "Fairness in Federated Learning via Core-Stability"], "answer_arxiv_id": ["2008.12161", "2303.16520", "2211.02091"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_9669"} +{"question": "Could you provide me some studies about programmatic weak supervision in rank aggregation?", "answer": ["A Survey on Programmatic Weak Supervision"], "answer_arxiv_id": ["2202.05433"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_9670"} +{"question": "Which papers use a neural field and a feature grid as their 3D representation in 3D-aware GANs?", "answer": ["Neural Fields in Visual Computing and Beyond", "Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["2111.11426", "2112.07945"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_9671"} +{"question": "In which paper did the scholars use K-Nearest Neighbor Classifier as a default proxy model for data valuation?", "answer": ["Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms"], "answer_arxiv_id": ["1908.08619"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_9672"} +{"question": "What studies discuss the use of hybrid representations in meeting real-time requirements of SLAM challenges?", "answer": ["NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM", "Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural\n Real-Time SLAM"], "answer_arxiv_id": ["2112.12130", "2302.03594", "2304.14377"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_9673"} +{"question": "What paper proposes the construction of a DPP kernel matrix as a Gram Matrix?", "answer": ["GDPP: Learning Diverse Generations using Determinantal Point Processes"], "answer_arxiv_id": ["1812.00068"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_9674"} +{"question": "Could you provide some research papers about AID-based and ITD-based bilevel optimization algorithms?", "answer": ["Hyperparameter optimization with approximate gradient", "On Differentiating Parameterized Argmin and Argmax Problems with Application to Bi-level Optimization", "Approximation Methods for Bilevel Programming", "On the Iteration Complexity of Hypergradient Computation", "Bilevel Optimization: Convergence Analysis and Enhanced Design", "Gradient-based Hyperparameter Optimization through Reversible Learning", "Bilevel Programming for Hyperparameter Optimization and Meta-Learning"], "answer_arxiv_id": ["1602.02355", "1607.05447", "1802.02246", "2006.16218", "2010.07962", "1502.03492", "1806.04910"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_9675"} +{"question": "Which paper introduced the EG3D pretraining framework?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["2112.07945"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_9676"} +{"question": "Could you tell me some examples of recently developed approaches for document retrieval through generative retrieval?", "answer": ["Autoregressive Entity Retrieval", "Transformer Memory as a Differentiable Search Index", "A Neural Corpus Indexer for Document Retrieval"], "answer_arxiv_id": ["2010.00904", "2202.06991", "2206.02743"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_9677"} +{"question": "What papers proposed different mobile UI modeling tasks, datasets and benchmarks?", "answer": ["Screen2Words: Automatic Mobile UI Summarization with Multimodal Learning", "Mapping Natural Language Instructions to Mobile UI Action Sequences", "A Dataset for Interactive Vision-Language Navigation with Unknown Command Feasibility", "UIBert: Learning Generic Multimodal Representations for UI Understanding", "Screen Recognition: Creating Accessibility Metadata for Mobile Applications from Pixels"], "answer_arxiv_id": ["2108.03353", "2005.03776", "2202.02312", "2107.13731", "2101.04893"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_9678"} +{"question": "Which works talked about the shortcomings of existing conditional diffusion models?", "answer": ["Residual Denoising Diffusion Models", "Inversion by Direct Iteration: An Alternative to Denoising Diffusion for\n Image Restoration", "I$^2$SB: Image-to-Image Schr\\\"odinger Bridge"], "answer_arxiv_id": ["2308.13712", "2303.11435", "2302.05872"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_9679"} +{"question": "Can you provide some works that examine the diffusion models in generative modeling framework?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "Understanding Diffusion Models: A Unified Perspective"], "answer_arxiv_id": ["2006.11239", "2011.13456", "2208.11970"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9680"} +{"question": "What papers deal with domain shifts by adjusting the statistics of batch normalization layers using the current test samples?", "answer": ["Improving robustness against common corruptions by covariate shift adaptation", "The Norm Must Go On: Dynamic Unsupervised Domain Adaptation by\n Normalization"], "answer_arxiv_id": ["2006.16971v2", "2112.00463"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_9681"} +{"question": "Which works have studied randomized LSVI (RLSVI), within the scope of the algorithmic framework based on posterior sampling?", "answer": ["Frequentist Regret Bounds for Randomized Least-Squares Value Iteration"], "answer_arxiv_id": ["1911.00567v7"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_9682"} +{"question": "Could you list the studies that explored dialect bias in large language models?", "answer": ["Multi-VALUE: A Framework for Cross-Dialectal English NLP"], "answer_arxiv_id": ["2212.08011"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_9683"} +{"question": "Which papers describe parameter isolation methods in CL that dynamically expand model capacity?", "answer": ["Progressive Neural Networks", "PathNet: Evolution Channels Gradient Descent in Super Neural Networks"], "answer_arxiv_id": ["1606.04671", "1701.08734"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_9684"} +{"question": "Can you suggest research papers discussing fusion of local and global information either by connecting local and global perception modules in a serial manner or by modeling them within a single module?", "answer": ["PVT v2: Improved Baselines with Pyramid Vision Transformer", "LocalViT: Bringing Locality to Vision Transformers", "DaViT: Dual Attention Vision Transformers", "LightViT: Towards Light-Weight Convolution-Free Vision Transformers", "EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers", "MaxViT: Multi-Axis Vision Transformer", "Learned Queries for Efficient Local Attention", "Inception Transformer", "Fast Vision Transformers with HiLo Attention", "Conformer: Local Features Coupling Global Representations for Visual Recognition"], "answer_arxiv_id": ["2106.13797", "2104.05707", "2204.03645", "2207.05557", "2205.03436", "2204.01697", "2112.11435", "2205.12956", "2205.13213", "2105.03889"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9685"} +{"question": "What are some studies on 3D Adversarial Learning, particularly focusing on C&W-based attacks ?", "answer": ["Generating 3D Adversarial Point Clouds", "3D Adversarial Attacks Beyond Point Cloud"], "answer_arxiv_id": ["1809.07016", "2104.12146"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_9686"} +{"question": "Could you provide me some studies that focused on integrating the process of detecting noisy labels and addressing them into the training pipeline?", "answer": ["MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks\n on Corrupted Labels", "Learning to Reweight Examples for Robust Deep Learning", "Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting", "Deep Self-Learning From Noisy Labels", "Distilling Effective Supervision from Severe Label Noise"], "answer_arxiv_id": ["1712.05055", "1803.09050", "1902.07379", "1908.02160", "1910.00701"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_9687"} +{"question": "Are there any researches that addressed the issue of discontinuity artifacts in monocular depth estimation?", "answer": ["AdaBins: Depth Estimation using Adaptive Bins", "Self-supervised Monocular Trained Depth Estimation using Self-attention and Discrete Disparity Volume", "BinsFormer: Revisiting Adaptive Bins for Monocular Depth Estimation", "LocalBins: Improving Depth Estimation by Learning Local Distributions", "Attention Attention Everywhere: Monocular Depth Prediction with Skip Attention"], "answer_arxiv_id": ["2011.14141", "2003.13951", "2204.00987", "2203.15132", "2210.09071"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_9688"} +{"question": "What scientific papers relate to advancements in GANs for text-conditioned image synthesis?", "answer": ["Generative Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Alias-Free Generative Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["2203.00667", "1809.11096", "2106.12423", "1812.04948", "1912.04958"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_9689"} +{"question": "Which research papers propose generalizations of DP to include a distributional assumption over the dataset?", "answer": ["Pufferfish Privacy Mechanisms for Correlated Data", "Bayesian Differential Privacy for Machine Learning"], "answer_arxiv_id": ["1603.03977", "1901.09697"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_9690"} +{"question": "What works have aimed to improve and refine major methods like adversarial training in neural network defense?", "answer": ["Fast is better than free: Revisiting adversarial training", "Adversarial Training for Free!", "Theoretically Principled Trade-off between Robustness and Accuracy", "You Only Propagate Once: Accelerating Adversarial Training via Maximal Principle", "Robust Trajectory Prediction against Adversarial Attacks", "Towards Stable and Efficient Training of Verifiably Robust Neural Networks"], "answer_arxiv_id": ["2001.03994", "1904.12843", "1901.08573", "1905.00877", "2208.00094", "1906.06316"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_9691"} +{"question": "What studies have interpreted MMD GAN and Sobolev GAN as gradient flow approaches?", "answer": ["On the Convergence of Gradient Descent in GANs: MMD GAN As a Gradient Flow"], "answer_arxiv_id": ["2011.02402"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9692"} +{"question": "What studies improved on Socratic Model's approach by treating tool use as a program synthesis problem?", "answer": ["Gorilla: Large Language Model Connected with Massive APIs", "ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs"], "answer_arxiv_id": ["2305.15334", "2307.16789"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_9693"} +{"question": "In which works the generation of CAD models could be influenced by a target B-rep, sketches, images, voxel grids or point clouds?", "answer": ["Fusion 360 Gallery: A Dataset and Environment for Programmatic CAD Construction from Human Design Sequences", "Inferring CAD Modeling Sequences Using Zone Graphs", "Sketch2CAD: Sequential CAD Modeling by Sketching in Context", "Vitruvion: A Generative Model of Parametric CAD Sketches", "Computer-Aided Design as Language", "Reconstructing editable prismatic CAD from rounded voxel models", "ExtrudeNet: Unsupervised Inverse Sketch-and-Extrude for Shape Parsing", "SECAD-Net: Self-Supervised CAD Reconstruction by Learning Sketch-Extrude Operations"], "answer_arxiv_id": ["2010.02392", "2104.03900", "2009.04927", "2109.14124", "2105.02769v1", "2209.01161", "2209.15632", "2303.10613"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_9694"} +{"question": "What papers focus on improving the encoder’s slot-attention module in unsupervised object-centric learning?", "answer": ["Shepherding Slots to Objects: Towards Stable and Robust Object-Centric\n Learning", "Exploring the Role of the Bottleneck in Slot-Based Models Through\n Covariance Regularization", "Improving Object-centric Learning with Query Optimization", "Object Representations as Fixed Points: Training Iterative Refinement\n Algorithms with Implicit Differentiation"], "answer_arxiv_id": ["2303.17842", "2306.02577", "2210.08990", "2207.00787"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_9695"} +{"question": "Could you provide some studies that used feature extractions for domain adaptation tasks?", "answer": ["Deep Visual Domain Adaptation: A Survey"], "answer_arxiv_id": ["1802.03601"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_9696"} +{"question": "Could you provide resources where NVS is enabled from only one or a few images at inference?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural\n Scene Representations", "RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs", "Differentiable Volumetric Rendering: Learning Implicit 3D\n Representations without 3D Supervision", "IBRNet: Learning Multi-View Image-Based Rendering", "Unsupervised Learning of 3D Object Categories from Videos in the Wild", "Learning to Render Novel Views from Wide-Baseline Stereo Pairs", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "Scene Representation Transformer: Geometry-Free Novel View Synthesis\n Through Set-Latent Scene Representations", "ViewFormer: NeRF-free Neural Rendering from Few Images Using\n Transformers", "Dense Depth Priors for Neural Radiance Fields from Sparse Input Views"], "answer_arxiv_id": ["2012.02190", "1906.01618", "2112.00724", "1912.07372", "2102.13090", "2103.16552", "2304.08463", "2103.15595", "2111.13152", "2203.10157", "2112.03288"], "source_meta": {"published_time": "20240626"}, "qid": "AutoScholarQuery_train_9697"} +{"question": "Which works analyze sharpness at different training times in the training regimes?", "answer": ["On the relation between the sharpest directions of DNN loss and the SGD step length", "Emergent properties of the local geometry of neural loss landscapes", "The break-even point on optimization trajectories of deep neural networks", "Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability", "Maximal Initial Learning Rates in Deep ReLU Networks"], "answer_arxiv_id": ["1807.05031", "1910.05929v1", "2002.09572", "2103.00065", "2212.07295"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_9698"} +{"question": "Which papers apply the technique of spectral normalization to regularize the Lipschitz condition of the RL agent’s value function?", "answer": ["Spectral Normalisation for Deep Reinforcement Learning: An Optimisation Perspective", "Towards Deeper Deep Reinforcement Learning with Spectral Normalization"], "answer_arxiv_id": ["2105.05246", "2106.01151"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9699"} +{"question": "Any works with different discretization of point clouds methods such as using grid cells in polar or cylindrical coordinate systems, or polar grid cells with modern attention operations?", "answer": ["PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds\n Semantic Segmentation", "Cylinder3D: An Effective 3D Framework for Driving-scene LiDAR Semantic\n Segmentation", "Spherical Transformer for LiDAR-based 3D Recognition"], "answer_arxiv_id": ["2003.14032", "2008.01550", "2303.12766"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_9700"} +{"question": "Which works combine both time series and clinical notes for improved results?", "answer": ["Machine Learning for Multimodal Electronic Health Records-based Research: Challenges and Perspectives", "Using Clinical Notes with Time Series Data for ICU Management", "How to Leverage Multimodal EHR Data for Better Medical Predictions?", "MUFASA: Multimodal Fusion Architecture Search for Electronic Health Records", "Attention Is All You Need"], "answer_arxiv_id": ["2111.04898v1", "1909.09702", "2110.15763", "2102.02340", "1706.03762"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_9701"} +{"question": "What works used the method of training external explainers based on model behaviors for understanding deep learning models?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1602.04938", "1705.07874"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_9702"} +{"question": "What papers have designed approaches to offload computational workload to the cloud?", "answer": ["Network Offloading Policies for Cloud Robotics: a Learning-based\n Approach", "Clipper: A Low-Latency Online Prediction Serving System", "Cloud-Device Collaborative Adaptation to Continual Changing Environments\n in the Real-world"], "answer_arxiv_id": ["1902.05703", "1612.03079", "2212.00972"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_9703"} +{"question": "What research included tappability prediction for UI objects?", "answer": ["Predicting and Explaining Mobile UI Tappability with Vision Modeling and Saliency Analysis"], "answer_arxiv_id": ["2204.02448"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_9704"} +{"question": "What works exist on data augmentation in computer vision?", "answer": ["How to train your ViT? Data, Augmentation, and Regularization in Vision\n Transformers", "AutoAugment: Learning Augmentation Policies from Data", "AugMix: A Simple Data Processing Method to Improve Robustness and\n Uncertainty", "TrivialAugment: Tuning-free Yet State-of-the-Art Data Augmentation"], "answer_arxiv_id": ["2106.10270", "1805.09501", "1912.02781", "2103.10158"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_9705"} +{"question": "Which encoder-based customization methods are there in the literature?", "answer": ["ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "Taming Encoder for Zero Fine-tuning Image Customization with\n Text-to-Image Diffusion Models", "DreamIdentity: Improved Editability for Efficient Face-identity\n Preserved Image Generation"], "answer_arxiv_id": ["2302.13848", "2305.14720", "2304.03411", "2304.02642", "2307.00300"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_9706"} +{"question": "Which works are related to creating neural operators as approximation for semigroup relationship between the input and output function space?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations", "Transformer for Partial Differential Equations’ Operator Learning", "Lie Point Symmetry Data Augmentation for Neural PDE Solvers", "Towards Multi-spatiotemporal-scale Generalized PDE Modeling"], "answer_arxiv_id": ["2010.08895", "2205.13671", "2202.07643", "2209.15616v2"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9707"} +{"question": "What works align multiple visual modalities in the image representation space?", "answer": ["EventBind: Learning a Unified Representation to Bind Them All for\n Event-based Open-world Understanding", "PointCLIP: Point Cloud Understanding by CLIP", "AudioCLIP: Extending CLIP to Image, Text and Audio", "AVE-CLIP: AudioCLIP-based Multi-window Temporal Transformer for Audio\n Visual Event Localization", "CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "ImageBind: One Embedding Space To Bind Them All", "Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D\n Understanding, Generation, and Instruction Following"], "answer_arxiv_id": ["2308.03135", "2112.02413", "2106.13043", "2210.05060", "2106.11097", "2305.05665v2", "2309.00615"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_9708"} +{"question": "What papers have been dedicated to studying variance-reduced methods in the non-convex setting for standard smooth functions?", "answer": ["Variance Reduction for Faster Non-Convex Optimization", "Nonconvex Finite-Sum Optimization Via SCSG Methods"], "answer_arxiv_id": ["1603.05643", "1706.09156"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_9709"} +{"question": "Could you mention the research papers that explored the idea of learning the learning algorithm itself?", "answer": ["RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "Learning to reinforcement learn", "A Simple Neural Attentive Meta-Learner"], "answer_arxiv_id": ["1611.02779", "1611.05763", "1707.03141v3"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_9710"} +{"question": "Which papers discuss the application of using text to adjust for confounding in causal inference?", "answer": ["Estimating Causal Effects of Tone in Online Debates", "Quantifying the Causal Effects of Conversational Tendencies"], "answer_arxiv_id": ["1906.04177", "2009.03897"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_9711"} +{"question": "What papers are about improving stability and generalization using regularization technique?", "answer": ["Generalization Bounds for Stochastic Saddle Point Problems", "Uniform Stability for First-Order Empirical Risk Minimization"], "answer_arxiv_id": ["2006.02067", "2207.08257v1"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_9712"} +{"question": "Which paper highlights the challenges of reconstructing 3D camera poses from egocentric videos in the EPIC-KITCHENS dataset?", "answer": ["Generative Hybrid Representations for Activity Forecasting with No-Regret Learning", "Ego-Topo: Environment Affordances from Egocentric Video", "EgoDistill: Egocentric Head Motion Distillation for Efficient Video Understanding"], "answer_arxiv_id": ["1904.06250", "2001.04583", "2301.02217"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_9713"} +{"question": "Which works introduced large, capable autoregressive language models on text?", "answer": ["Language Models are Few-Shot Learners", "Scaling Language Models: Methods, Analysis & Insights from Training Gopher"], "answer_arxiv_id": ["2005.14165", "2112.11446"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_9714"} +{"question": "What research papers developed methods for 3D reconstruction from single or few images?", "answer": ["Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "pixelNeRF: Neural Radiance Fields from One or Few Images", "Learning to Render Novel Views from Wide-Baseline Stereo Pairs", "Neural Groundplans: Persistent Neural Scene Representations from a Single Image", "Scene Representation Transformer: Geometry-Free Novel View Synthesis Through Set-Latent Scene Representations"], "answer_arxiv_id": ["1906.01618", "2012.02190", "2304.08463", "2207.11232", "2111.13152"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_9715"} +{"question": "Any researches about the evaluations of VLMs and the benchmarks used?", "answer": ["Microsoft COCO: Common Objects in Context", "Flickr30k Entities: Collecting Region-to-Phrase Correspondences for\n Richer Image-to-Sentence Models", "VQA: Visual Question Answering", "GQA: A New Dataset for Real-World Visual Reasoning and Compositional\n Question Answering"], "answer_arxiv_id": ["1405.0312", "1505.04870", "1505.00468", "1902.09506"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9716"} +{"question": "Can you name research papers that support robustness in fair training, including handling noisy groups or poisoning attacks?", "answer": ["Fair Classification with Noisy Protected Attributes: A Framework with Provable Guarantees", "Robust Optimization for Fairness with Noisy Protected Groups", "Exacerbating Algorithmic Bias through Fairness Attacks", "Poisoning Attacks on Algorithmic Fairness"], "answer_arxiv_id": ["2006.04778", "2002.09343", "2012.08723", "2004.07401"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_9717"} +{"question": "What studies are related to text-to-image generation using diffusion models?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Learning Transferable Visual Models From Natural Language Supervision", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2103.00020", "1910.10683", "2112.10741"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_9718"} +{"question": "Does research on Dog whistle extend to coded language in Chinese and Swedish communication?", "answer": ["Blow the Dog Whistle: A Chinese Dataset for Cant Understanding with\n Common Sense and World Knowledge"], "answer_arxiv_id": ["2104.02704"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_9719"} +{"question": "What research are about the usage of Vision Transformers in image segmentation?", "answer": ["Hierarchical Saliency Detection on Extended CSSD", "SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers"], "answer_arxiv_id": ["1408.5418", "2105.15203"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_9720"} +{"question": "Which studies have focused on 3D scene understanding of indoor environments?", "answer": ["ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "RIO: 3D Object Instance Re-Localization in Changing Indoor Environments"], "answer_arxiv_id": ["1702.04405", "1908.06109"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_9721"} +{"question": "Who has done research into state distribution correction problems, specifically the development of state weighting to match the excursion objective?", "answer": ["An Off-policy Policy Gradient Theorem Using Emphatic Weightings", "Learning Expected Emphatic Traces for Deep RL", "Emphatic Temporal-Difference Learning"], "answer_arxiv_id": ["1811.09013", "2107.05405", "1507.01569"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_9722"} +{"question": "What paper introduced the CPO method as a general-purpose policy search algorithm for SafeRL?", "answer": ["Constrained Policy Optimization"], "answer_arxiv_id": ["1705.10528"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_9723"} +{"question": "Can you give some examples of work on retraining-free sampler strategy for diffusion model acceleration?", "answer": ["On Fast Sampling of Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "gDDIM: Generalized denoising diffusion implicit models"], "answer_arxiv_id": ["2106.00132", "2010.02502", "2206.05564"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9724"} +{"question": "Are there any works on stack-augmented Recurrent Neural Networks?", "answer": ["Learning to Compose Words into Sentences with Reinforcement Learning", "Latent Tree Learning with Differentiable Parsers: Shift-Reduce Parsing and Chart Parsing"], "answer_arxiv_id": ["1611.09100", "1806.00840"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_9725"} +{"question": "What works mentioned learning the reverse covariance as part of the training-based methods?", "answer": ["Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models"], "answer_arxiv_id": ["2206.07309"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_9726"} +{"question": "In what papers the concept of Human-Object Interaction Detection was presented?", "answer": ["Learning to Detect Human-Object Interactions"], "answer_arxiv_id": ["1702.05448"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_9727"} +{"question": "Which paper illustrated great performance in novel view synthesis by representing 3D scenes with a coordinate-based neural network?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_9728"} +{"question": "What papers introduced state-action-dependent baselines to further reduce the variance arithmetic in policy gradients?", "answer": ["Q-Prop: Sample-Efficient Policy Gradient with An Off-Policy Critic", "Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines"], "answer_arxiv_id": ["1611.02247", "1803.07246"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_9729"} +{"question": "What papers have studied on uniform feasibility in constrained resource allocation?", "answer": ["Treatment Allocation under Uncertain Costs"], "answer_arxiv_id": ["2103.11066"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_9730"} +{"question": "What research work has been done on multimodal foundation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Reproducible scaling laws for contrastive language-image learning", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "ImageBind: One Embedding Space To Bind Them All", "Florence: A New Foundation Model for Computer Vision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2103.00020", "2212.07143", "2102.05918", "2305.05665", "2111.11432", "2205.01917", "2204.14198"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_9731"} +{"question": "Which studies used end-task performances on popular tasks to evaluate models in Natural Language Processing?", "answer": ["Pythia: A Suite for Analyzing Large Language Models Across Training and\n Scaling", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2304.01373", "2307.09288"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_9732"} +{"question": "Are there any studies into sound event detection (SED)?", "answer": ["CNN Architectures for Large-Scale Audio Classification"], "answer_arxiv_id": ["1609.09430"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_9733"} +{"question": "Could you provide me studies that have found out LLM prefers memorized text over non-memorized text?", "answer": ["Speak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4"], "answer_arxiv_id": ["2305.00118"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_9734"} +{"question": "What works focus on incremental multi-modal learning?", "answer": ["CLiMB: A Continual Learning Benchmark for Vision-and-Language Tasks", "A Unified Continuous Learning Framework for Multi-modal Knowledge Discovery and Pre-training"], "answer_arxiv_id": ["2206.09059", "2206.05555"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_9735"} +{"question": "Could you provide me some studies that use generative adversarial network (GAN) architectures for video disentanglement?", "answer": ["Decomposing Motion and Content for Natural Video Sequence Prediction", "MoCoGAN: Decomposing Motion and Content for Video Generation"], "answer_arxiv_id": ["1706.08033", "1707.04993"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_9736"} +{"question": "Which work provides a convergence rate for with-replacement simSGDA?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems"], "answer_arxiv_id": ["1906.00331"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_9737"} +{"question": "Which papers state individual fairness notions that require individual points to stay close to points that are similar to themselves in output clusters?", "answer": ["A Notion of Individual Fairness for Clustering", "A Pairwise Fair and Community-preserving Approach to k-Center Clustering"], "answer_arxiv_id": ["2006.04960", "2007.07384v1"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_9738"} +{"question": "What works use EA for approximating the greedy action selection in a continuous action space?", "answer": ["Q-Learning for Continuous Actions with Cross-Entropy Guided Policies", "Soft Actor-Critic with Cross-Entropy Policy Optimization", "GRAC: Self-Guided and Self-Regularized Actor-Critic", "Evolutionary Action Selection for Gradient-based Policy Learning"], "answer_arxiv_id": ["1903.10605", "2112.11115", "2009.08973", "2201.04286"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_9739"} +{"question": "Which works adapted game theory to formulate and solve machine learning problems?", "answer": ["EigenGame: PCA as a Nash Equilibrium"], "answer_arxiv_id": ["2010.00554v2"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_9740"} +{"question": "Which studies introduced parameter-efficient transfer learning for pretrained language models?", "answer": ["Delta Tuning: A Comprehensive Study of Parameter Efficient Methods for Pre-trained Language Models"], "answer_arxiv_id": ["2203.06904"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_9741"} +{"question": "What papers have attempted to measure the social bias of Vision-Language Models (VLMs)?", "answer": ["Evaluating CLIP: Towards Characterization of Broader Capabilities and Downstream Implications", "A Prompt Array Keeps the Bias Away: Debiasing Vision-Language Models with Adversarial Learning", "Debiasing Vision-Language Models via Biased Prompts", "Mitigating Test-Time Bias for Fair Image Retrieval", "Are Gender-Neutral Queries Really Gender-Neutral? Mitigating Gender Bias in Image Search", "Women also Snowboard: Overcoming Bias in Captioning Models", "Balancing the Picture: Debiasing Vision-Language Datasets with Synthetic Contrast Sets"], "answer_arxiv_id": ["2108.02818", "2203.11933", "2302.00070", "2305.19329", "2109.05433", "1803.09797", "2305.15407"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_9742"} +{"question": "Which research proposed a solution for LiDAR scene flow estimation challenges by combining ICP, rigid assumptions, and runtime optimization?", "answer": ["Re-Evaluating LiDAR Scene Flow for Autonomous Driving"], "answer_arxiv_id": ["2304.02150v2"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_9743"} +{"question": "Which papers have been written on the topic of reducing training supervision by exclusively using image-level datasets or using bounding boxes instead of masks?", "answer": ["HODOR: High-level Object Descriptors for Object Re-segmentation in Video Learned from Static Images", "Reducing the Annotation Effort for Video Object Segmentation Datasets"], "answer_arxiv_id": ["2112.09131v2", "2011.01142"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_9744"} +{"question": "What datasets focus on the night driving scenes?", "answer": ["Guided Curriculum Model Adaptation and Uncertainty-Aware Evaluation for\n Semantic Nighttime Image Segmentation", "Dark Model Adaptation: Semantic Image Segmentation from Daytime to\n Nighttime"], "answer_arxiv_id": ["1901.05946", "1810.02575"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_9745"} +{"question": "Which studies discuss the limitations of reward-only manipulation in single-agent RL?", "answer": ["Understanding the Limits of Poisoning Attacks in Episodic Reinforcement Learning"], "answer_arxiv_id": ["2208.13663"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_9746"} +{"question": "What works made use of a method that regresses object properties from 2D bounding boxes using geometric constraints?", "answer": ["3D Bounding Box Estimation Using Deep Learning and Geometry"], "answer_arxiv_id": ["1612.00496"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_9747"} +{"question": "Which previous study performed at the design space level?", "answer": ["Designing Network Design Spaces", "Design Space for Graph Neural Networks", "On Network Design Spaces for Visual Recognition"], "answer_arxiv_id": ["2003.13678", "2011.08843", "1905.13214"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_9748"} +{"question": "Which studies utilized convolutional neural network-based methods in learning-based light field approaches?", "answer": ["Learning-Based View Synthesis for Light Field Cameras", "X-Fields: Implicit Neural View-, Light- and Time-Image Interpolation", "Local Light Field Fusion: Practical View Synthesis with Prescriptive Sampling Guidelines"], "answer_arxiv_id": ["1609.02974", "2010.00450", "1905.00889"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_9749"} +{"question": "Which models integrated visual input information into different architecture language models?", "answer": ["Unifying Vision-and-Language Tasks via Text Generation", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "UniTAB: Unifying Text and Box Outputs for Grounded Vision-Language\n Modeling"], "answer_arxiv_id": ["2102.02779", "2202.03052", "2111.12085"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_9750"} +{"question": "Tell me about the articles that recommend utilizing parameter partitioning strategy to alleviate task interference in multi-task learning?", "answer": ["Attentive Single-Tasking of Multiple Tasks", "Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and\n Generalist Convolution Kernels", "Many Task Learning with Task Routing"], "answer_arxiv_id": ["1904.08918", "1908.09597", "1903.12117"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_9751"} +{"question": "Can you provide examples of DST methods used in pruning during training?", "answer": ["Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science", "Deep Rewiring: Training very sparse deep networks", "Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization", "Sparse Networks from Scratch: Faster Training without Losing Performance", "Selfish Sparse RNN Training", "Rigging the Lottery: Making All Tickets Winners", "Do We Actually Need Dense Over-Parameterization? In-Time Over-Parameterization in Sparse Training", "Sparse Training via Boosting Pruning Plasticity with Neuroregeneration"], "answer_arxiv_id": ["1707.04780v2", "1711.05136", "1902.05967", "1907.04840", "2101.09048", "1911.11134", "2102.02887", "2106.10404"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_9752"} +{"question": "What studies have explored the regularization or initialization of the eigenvalues of the recurrent weight matrix?", "answer": ["A Simple Way to Initialize Recurrent Networks of Rectified Linear Units", "Unitary Evolution Recurrent Neural Networks", "Resurrecting Recurrent Neural Networks for Long Sequences"], "answer_arxiv_id": ["1504.00941", "1511.06464", "2303.06349"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_9753"} +{"question": "What research has derived asymptotic power laws theoretically?", "answer": ["Frequency Bias in Neural Networks for Input of Non-Uniform Density"], "answer_arxiv_id": ["2003.04560"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_9754"} +{"question": "What studies have contributed to the field of context compression?", "answer": ["Compressing Context to Enhance Inference Efficiency of Large Language\n Models", "Optimizing Retrieval-augmented Reader Models via Token Elimination"], "answer_arxiv_id": ["2310.06201", "2310.13682"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_9755"} +{"question": "What literature exists on the use of generative models for image editing?", "answer": ["Q"], "answer_arxiv_id": ["1611.08152"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9756"} +{"question": "Which papers are about Test-time adaptation methods for dataset shift?", "answer": ["Tent: Fully Test-Time Adaptation by Entropy Minimization", "MEMO: Test Time Robustness via Adaptation and Augmentation", "Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation", "Maximum Likelihood with Bias-Corrected Calibration is Hard-To-Beat at Label Shift Adaptation", "Detecting and Correcting for Label Shift with Black Box Predictors", "Regularized Learning for Domain Adaptation under Label Shifts", "Online Adaptation to Label Distribution Shift"], "answer_arxiv_id": ["2006.10726", "2110.09506", "2002.08546", "1901.06852", "1802.03916", "1903.09734", "2107.04520"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_9757"} +{"question": "Which studies used representation learning for the solution in domain generalization?", "answer": ["Invariant Risk Minimization"], "answer_arxiv_id": ["1907.02893"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_9758"} +{"question": "Which works are designed for discrete environments when studying credit assignment problem on a graph from experience?", "answer": ["Topological Experience Replay", "Graph Backup: Data Efficient Backup Exploiting Markovian Transitions"], "answer_arxiv_id": ["2203.15845", "2205.15824"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_9759"} +{"question": "Which papers analyze the generalization in neural networks by bounding the VC-dimension of networks?", "answer": ["Nearly-tight VC-dimension and pseudodimension bounds for piecewise linear neural networks"], "answer_arxiv_id": ["1703.02930v3"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_9760"} +{"question": "Any works incorporate retrieval-based approach for generation, specifically using nearest neighbors (kNN) mechanisms?", "answer": ["Improving language models by retrieving from trillions of tokens", "Memorizing Transformers"], "answer_arxiv_id": ["2112.04426", "2203.08913"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_9761"} +{"question": "Which studies explicitly assign regions of the image to latent vectors?", "answer": ["Object Scene Representation Transformer"], "answer_arxiv_id": ["2206.06922"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_9762"} +{"question": "Which works demonstrated the effectiveness of Vector-Quantized Network in image restoration?", "answer": ["Towards Robust Blind Face Restoration with Codebook Lookup Transformer", "VQFR: Blind Face Restoration with Vector-Quantized Dictionary and\n Parallel Decoder", "RestoreFormer: High-Quality Blind Face Restoration from Undegraded\n Key-Value Pairs"], "answer_arxiv_id": ["2206.11253", "2205.06803", "2201.06374"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_9763"} +{"question": "Are there any studies that performed a systematic evaluations and analysis on existing domain generalization algorithms?", "answer": ["In Search of Lost Domain Generalization", "A Fine-Grained Analysis on Distribution Shift"], "answer_arxiv_id": ["2007.01434", "2110.11328"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_9764"} +{"question": "What research discusses localizing the bounds of the global Lipschitz bound to reduce conservatism, albeit at the expense of computational efficiency?", "answer": ["Training Certifiably Robust Neural Networks with Efficient Local Lipschitz Bounds", "Efficiently Computing Local Lipschitz Constants of Neural Networks via Bound Propagation"], "answer_arxiv_id": ["2111.01395", "2210.07394"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_9765"} +{"question": "Which works showcase applications of multi-modal generative modeling in the field of text-to-image generation?", "answer": ["Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2105.13290", "2204.06125", "2112.10741", "2205.11487", "2206.10789", "2111.14822", "1711.10485", "2112.10752"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_9766"} +{"question": "Is there a study that assumes the adaptive stepsize of Adam optimization algorithm is upper and lower bounded by two constants?", "answer": ["A Novel Convergence Analysis for Algorithms of the Adam Family"], "answer_arxiv_id": ["2112.03459v1"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_9767"} +{"question": "Can you list the papers that first introduced the Equivariant networks as G-Convolution and Steerable CNN?", "answer": ["Group Equivariant Convolutional Networks", "General E(2) - Equivariant Steerable CNNs"], "answer_arxiv_id": ["1602.07576", "1911.08251"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_9768"} +{"question": "Which papers have used sketching as a tool for low-rank approximation in machine learning?", "answer": ["Low Rank Approximation and Regression in Input Sparsity Time", "OSNAP: Faster numerical linear algebra algorithms via sparser subspace embeddings", "Low-distortion Subspace Embeddings in Input-sparsity Time and Applications to Robust Linear Regression", "Optimal CUR Matrix Decompositions", "Subspace Embedding and Linear Regression with Orlicz Norm", "Relative Error Tensor Low Rank Approximation"], "answer_arxiv_id": ["1207.6365", "1211.1002", "1210.3135", "1405.7910v2", "1806.06430", "1704.08246v2"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_9769"} +{"question": "What work referred to the concept of providing initial text or instructions to guide the response of a language model?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2107.13586v1"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_9770"} +{"question": "Which works have shown second-order regret bounds that depend on sample variances in losses?", "answer": ["More Adaptive Algorithms for Adversarial Bandits", "Sparsity, variance and curvature in multi-armed bandits"], "answer_arxiv_id": ["1801.03265", "1711.01037"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_9771"} +{"question": "What research has focused on explanation for AI related to natural language processing (NLP) and computer vision (CV)?", "answer": ["Towards A Rigorous Science of Interpretable Machine Learning", "A Survey of the State of Explainable AI for Natural Language Processing", "Post-hoc Interpretability for Neural NLP: A Survey", "Benchmarking and Survey of Explanation Methods for Black Box Models", "RISE: Randomized Input Sampling for Explanation of Black-box Models", "Explaining Deep Neural Networks and Beyond: A Review of Methods and Applications"], "answer_arxiv_id": ["1702.08608", "2010.00711", "2108.04840", "2102.13076", "1806.07421", "2003.07631"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_9772"} +{"question": "Which papers discuss the use of GAN based methods for unsupervised image super-resolution?", "answer": ["To learn image super-resolution, use a GAN to learn how to do image\n degradation first", "Frequency Separation for Real-World Super-Resolution", "Unsupervised Real-world Image Super Resolution via Domain-distance Aware\n Training", "Unpaired Image Super-Resolution using Pseudo-Supervision"], "answer_arxiv_id": ["1807.11458", "1911.07850", "2004.01178", "2002.11397"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_9773"} +{"question": "Are there any papers extending NeRF to render city-scale scenes?", "answer": ["Block-NeRF: Scalable Large Scene Neural View Synthesis"], "answer_arxiv_id": ["2202.05263"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_9774"} +{"question": "Which studies have been conducted on the enhancement of similarity search and clustering using machine learning?", "answer": ["Learning to Hash for Indexing Big Data - A Survey"], "answer_arxiv_id": ["1509.05472"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_9775"} +{"question": "Any papers focusing on the use of knowledge distillation-based methods in Anomaly Detection (AD)?", "answer": ["Uninformed Students: Student–Teacher Anomaly Detection with Discriminative Latent Embeddings", "Anomaly Detection via Reverse Distillation from One-Class Embedding", "Student-Teacher Feature Pyramid Matching for Anomaly Detection", "A Contrastive Objective for Learning Disentangled Representations"], "answer_arxiv_id": ["1911.02357", "2201.10703", "2103.04257", "2203.11284"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_9776"} +{"question": "Can you list some studies making use of the Mixture of Volumetric Primitives representation?", "answer": ["Mixture of Volumetric Primitives for Efficient Neural Rendering"], "answer_arxiv_id": ["2103.01954"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_9777"} +{"question": "Which works are associated with contemporary deep learning-based super resolution methods?", "answer": ["Image Super-Resolution Using Deep Convolutional Networks", "Real-Time Single Image and Video Super-Resolution Using an Efficient\n Sub-Pixel Convolutional Neural Network", "Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution", "Image Super-Resolution Using Very Deep Residual Channel Attention\n Networks", "Deep Back-Projection Networks For Super-Resolution"], "answer_arxiv_id": ["1501.00092", "1609.05158", "1511.04587", "1704.03915", "1807.02758", "1803.02735"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_9778"} +{"question": "Could you provide me some studies about domain augmentation methodology?", "answer": ["Generalizing Across Domains via Cross-Gradient Training", "Generalizing to Unseen Domains via Adversarial Data Augmentation", "Learning to Learn Single Domain Generalization", "Learning to Generate Novel Domains for Domain Generalization", "Domain Generalization with MixStyle", "Improving Out-of-Distribution Robustness via Selective Augmentation"], "answer_arxiv_id": ["1804.10745", "1805.12018", "2003.13216", "2007.03304", "2104.02008", "2201.00299"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_9779"} +{"question": "Which transformer model demonstrated superior performance on tabular classification/regression tasks?", "answer": ["Revisiting Deep Learning Models for Tabular Data"], "answer_arxiv_id": ["2106.11959"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_9780"} +{"question": "In what study did the researchers propose COBE + GOLF for problems with low Bellman eluder dimension?", "answer": ["A Model Selection Approach for Corruption Robust Reinforcement Learning"], "answer_arxiv_id": ["2110.03580v1"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_9781"} +{"question": "What work introduced a new setting to train a network without triplet data under a zero-shot setting?", "answer": ["Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image\n Retrieval"], "answer_arxiv_id": ["2302.03084"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_9782"} +{"question": "What works employed loss-data curves to evaluate representations by plotting loss against dataset size?", "answer": ["Evaluating representations by the complexity of learning low-loss predictors"], "answer_arxiv_id": ["2009.07368"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_9783"} +{"question": "Can you tell me about the works related to the large-scale scenes of Neural Radiance Fields?", "answer": ["Block-NeRF: Scalable Large Scene Neural View Synthesis", "NeRF++: Analyzing and Improving Neural Radiance Fields"], "answer_arxiv_id": ["2202.05263", "2010.07492"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_9784"} +{"question": "Which works employ generative framework to integrate edge, depth, and semantic information?", "answer": ["Low-Light Image Enhancement via Structure Modeling and Guidance"], "answer_arxiv_id": ["2305.05839"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_9785"} +{"question": "What research used CodeEmb’s approach for EHR code representations?", "answer": ["Unifying Heterogeneous Electronic Health Records Systems via Text-Based Code Embedding"], "answer_arxiv_id": ["2111.09098"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_9786"} +{"question": "Which previous works have studied selection of hard negatives during batch construction for metric and contrastive learning?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Mining on Manifolds: Metric Learning without Labels", "Hard negative examples are hard, but useful", "Working hard to know your neighbor’s margins: Local descriptor learning loss"], "answer_arxiv_id": ["1902.09229", "1803.11095", "2007.12749", "1705.10872"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_9787"} +{"question": "Could you tell me about some works that adopted richer inter-sample relations using the nearest neighbor in feature space for novel class discovery?", "answer": ["Neighborhood Contrastive Learning for Novel Class Discovery", "OpenMix: Reviving Known Knowledge for Discovering Novel Visual\n Categories in An Open World"], "answer_arxiv_id": ["2106.10731", "2004.05551"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_9788"} +{"question": "Could you provide me some works that are trained on various labeled triplet data benchmarks in the field of Composed Image Retrieval?", "answer": ["Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language\n Feedback", "Image Retrieval on Real-life Images with Pre-trained Vision-and-Language\n Models"], "answer_arxiv_id": ["1905.12794", "2108.04024"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_9789"} +{"question": "Which works focused on erasing specific visual concepts from diffusion model in the study of generative models and privacy?", "answer": ["Erasing Concepts from Diffusion Models"], "answer_arxiv_id": ["2303.07345"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_9790"} +{"question": "What studies have employed diversity-based sampling in pool-based active learning?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Deep Active Learning over the Long Tail", "Discriminative Active Learning"], "answer_arxiv_id": ["1708.00489", "1711.00941", "1907.06347"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_9791"} +{"question": "Could you provide a paper on a framework which makes use of various tools and knowledge resources to synthesize programs with LLMs?", "answer": ["Chameleon: Plug-and-Play Compositional Reasoning with Large Language\n Models"], "answer_arxiv_id": ["2304.09842"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_9792"} +{"question": "What studies have developed language models capable of inductive reasoning?", "answer": ["Instruction Induction: From Few Examples to Natural Language Task Descriptions", "Language Models as Inductive Reasoners", "Large Language Models are Human-Level Prompt Engineers", "Describing Differences between Text Distributions with Natural Language", "Explaining Patterns in Data with Language Models via Interpretable Autoprompting"], "answer_arxiv_id": ["2205.10782", "2212.10923", "2211.01910", "2201.12323", "2210.01848"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_9793"} +{"question": "Could you tell me about the studies that rely on agent actions to predict the group element?", "answer": ["Symmetry-Based Disentangled Representation Learning requires Interaction with Environments", "Learning Disentangled Representations and Group Structure of Dynamical Environments"], "answer_arxiv_id": ["1904.00243", "2002.06991"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_9794"} +{"question": "What studies in the combinatorial optimization literature have been applied in the context of machine learning?", "answer": ["Submodular Maximization with Nearly Optimal Approximation, Adaptivity and Query Complexity", "Non-monotone Submodular Maximization with Nearly Optimal Adaptivity and Query Complexity", "Feature Cross Search via Submodular Optimization", "Data-Efficient Structured Pruning via Submodular Optimization", "Submodularity In Machine Learning and Artificial Intelligence"], "answer_arxiv_id": ["1807.07889v3", "1808.06932", "2107.02139", "2203.04940", "2202.00132"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_9795"} +{"question": "Which work combines regression model with latent nearest neighbors for uncertain prediction?", "answer": ["Density estimation in representation space to predict model uncertainty"], "answer_arxiv_id": ["1908.07235"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_9796"} +{"question": "What research works deal with prompt optimization in T2I generation, but does not utilize personalized information?", "answer": ["SUR-adapter: Enhancing Text-to-Image Pre-trained Diffusion Models with\n Large Language Models", "Optimizing Prompts for Text-to-Image Generation"], "answer_arxiv_id": ["2305.05189", "2212.09611"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_9797"} +{"question": "Could you provide me some studies that proposed a two-stage semantic segmentation framework?", "answer": ["A Simple Baseline for Open-Vocabulary Semantic Segmentation with Pre-trained Vision-language Model"], "answer_arxiv_id": ["2112.14757"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_9798"} +{"question": "Which studies relate to the marginalized importance sampling of the offline RL algorithms?", "answer": ["DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections", "OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation", "Finite Sample Analysis of Minimax Offline Reinforcement Learning: Completeness, Fast Rates and First-Order Efficiency", "Offline Reinforcement Learning with Realizability and Single-policy Concentrability", "Offline Reinforcement Learning Under Value and Density-Ratio Realizability: The Power of Gaps", "Beyond the Return: Off-policy Function Estimation under User-specified Error-measuring Distributions", "Revisiting the Linear-Programming Framework for Offline RL with General Function Approximation"], "answer_arxiv_id": ["1906.04733", "2106.10783", "2102.02981", "2202.04634", "2203.13935", "2210.15543", "2212.13861"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_9799"} +{"question": "Can you identify a research work that discusses the Lookahead optimizer?", "answer": ["Lookahead Optimizer: k steps forward, 1 step back"], "answer_arxiv_id": ["1907.08610"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_9800"} +{"question": "Which studies achieved good results using source separation and generative models in Audio-Visual Sound Localization?", "answer": ["The Sound of Pixels", "The Sound of Motions"], "answer_arxiv_id": ["1804.03160", "1904.05979"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_9801"} +{"question": "What papers are applicable without domain information for domain generalization?", "answer": ["Heterogeneous Risk Minimization", "Predicting with High Correlation Features", "No Subclass Left Behind: Fine-Grained Robustness in Coarse-Grained Classification Problems", "Environment Inference for Invariant Learning"], "answer_arxiv_id": ["2105.03818", "1910.00164", "2011.12945", "2010.07249"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_9802"} +{"question": "Could you give me examples of research about models that are tailored for visually-rich document understanding?", "answer": ["mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document\n Understanding", "UReader: Universal OCR-free Visually-situated Language Understanding\n with Multimodal Large Language Model", "OCR-free Document Understanding Transformer", "Pix2Struct: Screenshot Parsing as Pretraining for Visual Language\n Understanding", "Unifying Vision, Text, and Layout for Universal Document Processing"], "answer_arxiv_id": ["2307.02499", "2310.05126", "2111.15664", "2210.03347", "2212.02623"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_9803"} +{"question": "Which works propose algorithms that can incorporate deformations during 3D reconstruction?", "answer": ["Neural 3D Mesh Renderer"], "answer_arxiv_id": ["1711.07566"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_9804"} +{"question": "Could you list some studies in the field of domain adaptation on which our work builds?", "answer": ["Bridging Theory and Algorithm for Domain Adaptation", "Contrastive Adaptation Network for Unsupervised Domain Adaptation", "Domain Adaptation as a Problem of Inference on Graphical Models"], "answer_arxiv_id": ["1904.05801", "1901.00976", "2002.03278"], "source_meta": {"published_time": "20210713"}, "qid": "AutoScholarQuery_train_9805"} +{"question": "What research mentions about the limitations of current approaches being color-agnostic in text attacks?", "answer": ["On the Reliability of Watermarks for Large Language Models"], "answer_arxiv_id": ["2306.04634"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_9806"} +{"question": "Are there any studies that proposed an asymmetric encoder-decoder architecture for better training efficiency?", "answer": ["Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2111.06377"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_9807"} +{"question": "Which research papers focus on generating human motion?", "answer": ["GANimator: Neural Motion Synthesis from a Single Sequence"], "answer_arxiv_id": ["2205.02625"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9808"} +{"question": "Which papers deal with geometric scaffold or mesh construction in novel view synthesis?", "answer": ["Free View Synthesis", "Stable View Synthesis", "Stereo Radiance Fields (SRF): Learning View Synthesis for Sparse Views\n of Novel Scenes"], "answer_arxiv_id": ["2008.05511", "2011.07233", "2104.06935"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_9809"} +{"question": "Which papers discussed the re-labeling strategy employed in CutMix-based methods?", "answer": ["CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features", "TransMix: Attend to Mix for Vision Transformers"], "answer_arxiv_id": ["1905.04899", "2111.09833"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_9810"} +{"question": "What research introduced the Segment Anything Model (SAM) in interactive segmentation?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_9811"} +{"question": "Could you provide me works about using nearest centroid classifiers in DNNs?", "answer": ["Visual Recognition with Deep Nearest Centroids"], "answer_arxiv_id": ["2209.07383"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_9812"} +{"question": "Which works establish a unified perspective to study GNNs representations?", "answer": ["A Unified View on Graph Neural Networks as Graph Signal Denoising", "Interpreting and Unifying Graph Neural Networks with An Optimization Framework"], "answer_arxiv_id": ["2010.01777", "2101.11859"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_9813"} +{"question": "What work first proposed statistical testing for privacy auditing?", "answer": ["Testing Lipschitz Property over Product Distribution and its Applications to Statistical Data Privacy"], "answer_arxiv_id": ["1209.4056"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_9814"} +{"question": "What papers noted risk of pushing away semantically relevant states by repulsion?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning", "Supervised Contrastive Learning"], "answer_arxiv_id": ["2006.07733", "2011.10566", "2004.11362"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_9815"} +{"question": "What studies contributed to the enhancement of functional maps in terms of accuracy, efficiency, and robustness?", "answer": ["Coupled quasi-harmonic bases", "Partial Functional Correspondence"], "answer_arxiv_id": ["1210.0026", "1506.05274"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_9816"} +{"question": "Which works attempt to speed up the learning process of NeRF?", "answer": ["FastNeRF: High-Fidelity Neural Rendering at 200FPS", "Depth-supervised NeRF: Fewer Views and Faster Training for Free"], "answer_arxiv_id": ["2103.10380", "2107.02791"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_9817"} +{"question": "What papers provide evidence that max-margin classifiers tend to offer better generalization guarantees, and clustered embedding spaces are useful for few-shot transfer learning?", "answer": ["Spectrally-normalized margin bounds for neural networks", "Size-Independent Sample Complexity of Neural Networks", "Unraveling Meta-Learning: Understanding Feature Representations for Few-Shot Tasks", "On the Role of Neural Collapse in Transfer Learning"], "answer_arxiv_id": ["1706.08498", "1712.06541", "2002.06753", "2112.15121"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_9818"} +{"question": "Who proposed the method of Successive halving in multi-fidelity HPO?", "answer": ["Non-stochastic Best Arm Identification and Hyperparameter Optimization"], "answer_arxiv_id": ["1502.07943v1"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_9819"} +{"question": "Which study examined online ES estimators but did not consider the time step 𝝉 as a random variable?", "answer": ["Unbiased Gradient Estimation in Unrolled Computation Graphs with Persistent Evolution Strategies"], "answer_arxiv_id": ["2112.13835"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_9820"} +{"question": "Any works addressing interaction in simulated web environments?", "answer": ["WebShop: Towards Scalable Real-World Web Interaction with Grounded\n Language Agents", "WebArena: A Realistic Web Environment for Building Autonomous Agents"], "answer_arxiv_id": ["2207.01206", "2307.13854"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_9821"} +{"question": "Which paper introduced the CUAD dataset that consists of annotated English commercial contracts?", "answer": ["CUAD: An Expert-Annotated NLP Dataset for Legal Contract Review"], "answer_arxiv_id": ["2103.06268"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_9822"} +{"question": "Which works have introduced prior-free methods in the aim of improving category-level pose estimation?", "answer": ["VI-Net: Boosting Category-level 6D Object Pose Estimation via Learning\n Decoupled Rotations on the Spherical Representations", "IST-Net: Prior-free Category-level Pose Estimation with Implicit Space\n Transformation"], "answer_arxiv_id": ["2308.09916", "2303.13479"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_9823"} +{"question": "Could you provide me some works that tried to extend the workflow of diffusion models with more accurate image-conditioning?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object", "Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors", "Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion\n Prior", "DINAR: Diffusion Inpainting of Neural Textures for One-Shot Human\n Avatars"], "answer_arxiv_id": ["2303.11328", "2306.17843", "2303.14184", "2303.09375"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9824"} +{"question": "Which works attempted to improve NeRF performance by baking the components?", "answer": ["Neural Sparse Voxel Fields", "PlenOctrees for Real-time Rendering of Neural Radiance Fields"], "answer_arxiv_id": ["2007.11571", "2103.14024"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_9825"} +{"question": "Which research introduced HyperNetworks as an idea of using an auxiliary neural network to predict network weights in order to change the functioning of a specific neural network?", "answer": ["HyperNetworks"], "answer_arxiv_id": ["1609.09106"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_9826"} +{"question": "What studies showed that diffusion models are less prone to suffer from mode collapse?", "answer": ["Denoising Diffusion Probabilistic Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2006.11239", "1503.03585", "2010.02502", "2105.05233"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_9827"} +{"question": "Which papers discuss the SayCan method that integrates an LLM with a value function for grounding affordances?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances"], "answer_arxiv_id": ["2204.01691"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_9828"} +{"question": "Could you provide me with some examples of works that employed generative methods by learning the distribution of underlying explanatory subgraphs in Graph Neural Networks?", "answer": ["Reinforced Causal Explainer for Graph Neural Networks", "DAG Matters! GFlowNets Enhanced Explainer for Graph Neural Networks", "Parameterized Explainer for Graph Neural Network", "Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanism", "GNNInterpreter: A Probabilistic Generative Model-Level Explanation for Graph Neural Networks", "XGNN: Towards Model-Level Explanations of Graph Neural Networks"], "answer_arxiv_id": ["2204.11028", "2303.02448", "2011.04573", "2201.12987", "2209.07924", "2006.02587"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_9829"} +{"question": "Can you mention the studies that translate English NLI and paraphrase identification samples into other languages?", "answer": ["XNLI: Evaluating Cross-lingual Sentence Representations", "PAWS-X: A Cross-lingual Adversarial Dataset for Paraphrase\n Identification"], "answer_arxiv_id": ["1809.05053", "1908.11828"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_9830"} +{"question": "Do you know any works dealing with the application of pre-trained models on downstream tasks in the field of Visual-Language Representational Learning?", "answer": ["Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models", "Learning to Prompt for Continual Learning", "Novelty Controlled Paraphrase Generation with Retrieval Augmented Conditional Prompt Tuning"], "answer_arxiv_id": ["2209.07511", "2112.08654", "2202.00535"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_9831"} +{"question": "Which research papers suggest data-based approaches for designing intrinsic reward in unsupervised RL?", "answer": ["Behavior From the Void: Unsupervised Active Pre-Training", "Reinforcement Learning with Prototypical Representations"], "answer_arxiv_id": ["2103.04551", "2102.11271v2"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_9832"} +{"question": "What works studied inverse reinforcement learning, a reward function learning method by observing the actions of an agent?", "answer": ["A Survey of Inverse Reinforcement Learning: Challenges, Methods and Progress"], "answer_arxiv_id": ["1806.06877"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_9833"} +{"question": "Which works utilize the ‘sliding window’ approaches in long sequence handling?", "answer": ["Longformer: The Long-Document Transformer", "Big Bird: Transformers for Longer Sequences", "CoLT5: Faster Long-Range Transformers with Conditional Computation"], "answer_arxiv_id": ["2004.05150", "2007.14062", "2303.09752v3"], "source_meta": {"published_time": "20240107"}, "qid": "AutoScholarQuery_train_9834"} +{"question": "What research works have studied the use of Transformer in self-supervised learning?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "An Empirical Study of Training Self-Supervised Vision Transformers", "BEiT: BERT Pre-Training of Image Transformers", "iBOT : Image BERT Pre-Training with Online Tokenizer", "Self-Supervised Learning with Swin Transformers", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2104.14294", "2104.02057", "2106.08254", "2111.07832", "2105.04553", "2111.06377"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_9835"} +{"question": "Which studies considered correlated noise in analysis of SGD?", "answer": ["Random Reshuffling: Simple Analysis with Vast Improvements", "Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond"], "answer_arxiv_id": ["2006.05988", "2110.10342"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9836"} +{"question": "Which researchers have explored and proposed methods for interpretable deep learning models?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based\n Localization", "Striving for Simplicity: The All Convolutional Net", "What Does BERT Look At? An Analysis of BERT's Attention", "Revealing the Dark Secrets of BERT", "Show, Attend and Tell: Neural Image Caption Generation with Visual\n Attention", "Generic Attention-model Explainability for Interpreting Bi-Modal and\n Encoder-Decoder Transformers", "Transformer Interpretability Beyond Attention Visualization", "Beyond Surface Statistics: Scene Representations in a Latent Diffusion\n Model", "Intriguing properties of neural networks"], "answer_arxiv_id": ["1610.02391", "1412.6806", "1906.04341", "1908.08593", "1502.03044", "2103.15679", "2012.09838", "2306.05720", "1312.6199"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_9837"} +{"question": "Which research extended KD approaches from image classification to object detection?", "answer": ["Distilling Object Detectors with Fine-grained Feature Imitation", "Distilling Object Detectors via Decoupled Features", "Focal and Global Knowledge Distillation for Detectors", "Prediction-Guided Distillation for Dense Object Detection"], "answer_arxiv_id": ["1906.03609", "2103.14475", "2111.11837", "2203.05469"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_9838"} +{"question": "What papers have been published about structured commonsense reasoning, where LLMs generate graphs from a natural language input?", "answer": ["WIQA: A dataset for \"What if…\" reasoning over procedural text", "Could you give me a hint? Generating inference graphs for defeasible reasoning", "ExplaGraphs: An Explanation Graph Generation Task for Structured Commonsense Reasoning", "Language Models of Code are Few-Shot Commonsense Learners"], "answer_arxiv_id": ["1909.04739", "2105.05418", "2104.07644", "2210.07128"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_9839"} +{"question": "What works discuss methodologies designed to obtain unified intra-class and discriminative inter-class representation by training CNN network in the field of Person ReID?", "answer": ["Learning Instance-level Spatial-Temporal Patterns for Person\n Re-identification", "Fine-tuning CNN Image Retrieval with No Human Annotation", "Omni-Scale Feature Learning for Person Re-Identification", "Deep Learning for Person Re-identification: A Survey and Outlook"], "answer_arxiv_id": ["2108.00171", "1711.02512", "1905.00953", "2001.04193"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_9840"} +{"question": "What studies are about decision-time planning and expert iteration in the context of two-player zero-sum games?", "answer": ["Solving Imperfect Information Games Using Decomposition", "DeepStack: Expert-Level Artificial Intelligence in Heads-Up No-Limit Poker", "Depth-Limited Solving for Imperfect-Information Games", "Unlocking the Potential of Deep Counterfactual Value Networks", "Combining Deep Reinforcement Learning and Search for Imperfect-Information Games"], "answer_arxiv_id": ["1303.4441", "1701.01724", "1805.08195", "2007.10442", "2007.13544"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_9841"} +{"question": "Who studied the simplicity bias, where Neural Networks (NNs) exhibit a tendency to fit data with simple functions?", "answer": ["A Closer Look at Memorization in Deep Networks", "Shortcut Learning in Deep Neural Networks", "Evading the Simplicity Bias: Training a Diverse Set of Models Discovers\n Solutions with Superior OOD Generalization"], "answer_arxiv_id": ["1706.05394", "2004.07780", "2105.05612"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_9842"} +{"question": "Which research introduced the Generalized Cross-Entropy (GCE) for faster convergence and better accuracy?", "answer": ["Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels"], "answer_arxiv_id": ["1805.07836"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_9843"} +{"question": "Which studies focus on utilizing runtime optimization with self-supervision for scene flow learning?", "answer": ["SCOOP: Self-Supervised Correspondence and Optimization-Based Scene Flow"], "answer_arxiv_id": ["2211.14020"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_9844"} +{"question": "Are there any works about sketch abstraction and its impact on generative modelling of sketches?", "answer": ["A Neural Representation of Sketch Drawings", "Learning Deep Sketch Abstraction", "ChiroDiff: Modelling chirographic data with Diffusion Models", "CLIPasso: Semantically-Aware Object Sketching", "CLIPascene: Scene Sketching with Different Types and Levels of\n Abstraction"], "answer_arxiv_id": ["1704.03477", "1804.04804", "2304.03785", "2202.05822", "2211.17256"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_9845"} +{"question": "Which works present a novel objective function for EBM by extending recovery likelihood framework?", "answer": ["Generalized Denoising Auto-Encoders as Generative Models", "Learning Energy-Based Models by Diffusion Recovery Likelihood"], "answer_arxiv_id": ["1305.6663", "2012.08125"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_9846"} +{"question": "Which papers introduced Key-Value Cache, which minimizes repetitive computation of hidden key-value pairs in LLM decoding process?", "answer": ["Efficiently Scaling Transformer Inference"], "answer_arxiv_id": ["2211.05102"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_9847"} +{"question": "What papers use neural network architectures based on exact inference on deep state space models?", "answer": ["Backprop KF: Learning Discriminative Deterministic State Estimators", "Recurrent Kalman Networks: Factorized Inference in High-Dimensional Deep Feature Spaces", "Action-Conditional Recurrent Kalman Networks For Forward and Inverse Dynamics Learning"], "answer_arxiv_id": ["1605.07148", "1905.07357v1", "2010.10201"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_9848"} +{"question": "What are the research works that apply DRO for traditional fairness in machine learning?", "answer": ["Fairness Without Demographics in Repeated Loss Minimization", "Distributionally Robust Fair Principal Components via Geodesic Descents", "Re-weighting Based Group Fairness Regularization via Classwise Robust\n Optimization"], "answer_arxiv_id": ["1806.08010", "2202.03071", "2303.00442"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_9849"} +{"question": "Could you provide me studies regarding universality among continuous permutation invariant or equivariant functions?", "answer": ["On the Universality of Invariant Networks", "Universal Invariant and Equivariant Graph Neural Networks", "What Graph Neural Networks Cannot Learn: Depth vs Width", "Expressive Power of Invariant and Equivariant Graph Neural Networks", "Expressiveness and Approximation Properties of Graph Neural Networks"], "answer_arxiv_id": ["1901.09342", "1905.04943", "1907.03199", "2006.15646", "2204.04661"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_9850"} +{"question": "Are there any studies on learning non-parametric reward functions using empowerment?", "answer": ["Unsupervised Control through Non-Parametric Discriminative Rewards"], "answer_arxiv_id": ["1811.11359"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_9851"} +{"question": "Which research papers discuss using Large Language Models for learning a language conditioned policy in robotics tasks?", "answer": ["Primal Wasserstein Imitation Learning"], "answer_arxiv_id": ["2006.04678"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_9852"} +{"question": "What are some of the works about domain adversarial learning, a method of Unsupervised Domain Adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks"], "answer_arxiv_id": ["1505.07818"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_9853"} +{"question": "Which works analyze the impact of AI research involving the private sector?", "answer": ["A narrowing of AI research?"], "answer_arxiv_id": ["2009.10385"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_9854"} +{"question": "Which papers present research into various types of 3D representation for 3D segmentation, such as RGBD images, pointcloud, and voxels?", "answer": ["Depth-aware CNN for RGB-D Segmentation", "Malleable 2.5D Convolution: Learning Receptive Fields along the\n Depth-axis for RGB-D Scene Parsing", "GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in\n Point Cloud", "Learning Object Bounding Boxes for 3D Instance Segmentation on Point\n Clouds", "RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds", "Point-Voxel CNN for Efficient 3D Deep Learning", "OccuSeg: Occupancy-aware 3D Instance Segmentation"], "answer_arxiv_id": ["1803.06791", "2007.09365", "1812.03320", "1906.01140", "1911.11236", "1907.03739", "2003.06537v3"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_9855"} +{"question": "Which research address the usage of Federated Learning in aspect sentiment classification and relation extraction?", "answer": ["Distantly Supervised Relation Extraction in Federated Settings"], "answer_arxiv_id": ["2008.05049"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_9856"} +{"question": "Which works utilize image gradients to exploit intensity changes in the camouflaged object from the background?", "answer": ["Deep Gradient Learning for Efficient Camouflaged Object Detection"], "answer_arxiv_id": ["2205.12853"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_9857"} +{"question": "Can you name some studies that have combined neuroscience and AI to improve traditional artificial neural networks (ANNs)?", "answer": ["Learning From Brains How to Regularize Machines"], "answer_arxiv_id": ["1911.05072"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_9858"} +{"question": "What works have been used as stereotype datasets for detecting stereotypes in NLP prediction and classification-based tasks?", "answer": ["StereoSet: Measuring stereotypical bias in pretrained language models", "CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked\n Language Models"], "answer_arxiv_id": ["2004.09456", "2010.00133"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_9859"} +{"question": "What are the works that have extended ImageNet classes translations to other languages?", "answer": ["Contrastive Language-Image Pre-training for the Italian Language", "Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese"], "answer_arxiv_id": ["2108.08688", "2211.01335"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_9860"} +{"question": "Which research papers employ diffusion models for classification and semantic and panoptic segmentation?", "answer": ["Denoising Diffusion Probabilistic Models", "Your Diffusion Model is Secretly a Zero-Shot Classifier", "Label-Efficient Semantic Segmentation with Diffusion Models", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2006.11239", "2303.16203", "2112.03126", "2303.04803"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_9861"} +{"question": "Which papers have contributed to the study of label-flipping and distributional robustness?", "answer": ["Certified Robustness to Label-Flipping Attacks via Randomized Smoothing", "Agnostic Estimation of Mean and Covariance", "Sever: A Robust Meta-Algorithm for Stochastic Optimization"], "answer_arxiv_id": ["2002.03018", "1604.06968", "1803.02815"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_9862"} +{"question": "What studies have focused on the local DP model for distributed frequency estimation?", "answer": ["Local, Private, Efficient Protocols for Succinct Histograms", "Practical Locally Private Heavy Hitters", "Heavy Hitters and the Structure of Local Privacy", "Heavy Hitters and the Structure of Local Privacy", "Frequency Estimation Under Multiparty Differential Privacy: One-shot and Streaming", "Discrete Distribution Estimation under Local Privacy", "Locally Differentially Private Data Collection and Analysis", "Hadamard Response: Estimating Distributions Privately, Efficiently, and with Little Communication", "Communication Complexity in Locally Private Distribution Estimation and Heavy Hitters", "Breaking the Communication-Privacy-Accuracy Trilemma", "Lossless Compression of Efficient Private Local Randomizers", "Optimal Compression of Locally Differentially Private Mechanisms", "Private Frequency Estimation via Projective Geometry"], "answer_arxiv_id": ["1504.04686", "1707.04982", "1711.04740", "1711.04740", "2104.01808", "1602.07387", "1906.01777", "1802.04705v2", "1905.11888", "2007.11707", "2102.12099", "2111.00092", "2203.00194"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_9863"} +{"question": "Which work presents how one can improve the performance of a LLM by combining two probability distributions?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models"], "answer_arxiv_id": ["1911.00172"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_9864"} +{"question": "Who hypothesized that high-inflow words in training data might escalate degeneration issues in neural language models?", "answer": ["A Theoretical Analysis of the Repetition Problem in Text Generation"], "answer_arxiv_id": ["2012.14660v4"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_9865"} +{"question": "Which research highlighted that communication costs increase as the maximum degree increases in decentralized learning?", "answer": ["Exponential Graph is Provably Efficient for Decentralized Deep Training"], "answer_arxiv_id": ["2110.13363"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_9866"} +{"question": "What research primarily uses geometric or dynamic properties to distinguish objects and backgrounds in 3D point clouds?", "answer": ["Multiscale Combinatorial Grouping for Image Segmentation and Object Proposal Generation"], "answer_arxiv_id": ["1503.00848v4"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_9867"} +{"question": "Could you explain the research regarding neural link predictors which learn continuous representations for the entities and relation types in the graph?", "answer": ["Complex Embeddings for Simple Link Prediction", "Embedding Entities and Relations for Learning and Inference in Knowledge Bases", "Convolutional 2D Knowledge Graph Embeddings", "RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space", "TuckER: Tensor Factorization for Knowledge Graph Completion", "LowFER: Low-rank Bilinear Pooling for Link Prediction"], "answer_arxiv_id": ["1606.06357", "1412.6575", "1707.01476", "1902.10197", "1901.09590", "2008.10858"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_9868"} +{"question": "What works propose neural network-based map representations for dense SLAM?", "answer": ["iMAP: Implicit Mapping and Positioning in Real-Time", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "Point-SLAM: Dense Neural Point Cloud-based SLAM", "ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of\n Signed Distance Fields", "NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM"], "answer_arxiv_id": ["2103.12352", "2112.12130", "2304.04278", "2211.11704", "2302.03594"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_9869"} +{"question": "What works consider the distribution discrepancy for accuracy estimation?", "answer": ["Are Labels Always Necessary for Classifier Accuracy Evaluation?", "Predicting Out-of-Distribution Error with the Projection Norm", "Estimating Generalization under Distribution Shifts via Domain-Invariant Representations"], "answer_arxiv_id": ["2007.02915", "2202.05834", "2007.03511"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9870"} +{"question": "Which works propose the idea of text-to-image synthesis with additional conditions in Controllable Generation?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_9871"} +{"question": "Which papers introduced boundary sensitivity in the context of implicit neural shapes?", "answer": ["Controlling Neural Level Sets", "Augmenting Implicit Neural Shape Representations with Explicit Deformation Fields", "MeshSDF: Differentiable Iso-Surface Extraction", "A Level Set Theory for Neural Implicit Evolution under Explicit Flows", "Sketch2Mesh: Reconstructing and Editing 3D Shapes from Sketches"], "answer_arxiv_id": ["1905.11911", "2108.08931", "2006.03997", "2204.07159v2", "2104.00482"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_9872"} +{"question": "Could you give me an example of a language model that was used as a baseline or reference point for other studies?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_9873"} +{"question": "Which works examined the intrinsic dimensionality of neural networks to evaluate their capacity and effectiveness?", "answer": ["Intrinsic Dimensionality Explains the Effectiveness of Language Model\n Fine-Tuning", "The Shape of Learning: Anisotropy and Intrinsic Dimensions in\n Transformer-Based Models"], "answer_arxiv_id": ["2012.13255", "2311.05928"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_9874"} +{"question": "Which projects tried to learn 3D shape from 2D images using various Nerf representations?", "answer": ["EVA3D: Compositional 3D Human Generation from 2D Image Collections", "HumanGen: Generating Human Radiance Fields with Explicit Priors", "Get3DHuman: Lifting StyleGAN-Human into a 3D Generative Model using\n Pixel-aligned Reconstruction Priors", "Chupa: Carving 3D Clothed Humans from Skinned Shape Priors using 2D\n Diffusion Probabilistic Models", "StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation", "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images"], "answer_arxiv_id": ["2210.04888", "2212.05321", "2302.01162", "2305.11870", "2112.11427", "2209.11163"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_9875"} +{"question": "Which studies have explored the complimentarity of outputs by training models to merge multiple candidates generated by LLMs?", "answer": ["LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and\n Generative Fusion", "Small Language Models Improve Giants by Rewriting Their Outputs"], "answer_arxiv_id": ["2306.02561", "2305.13514"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_9876"} +{"question": "What works are about quantization-aware training (QAT) methods?", "answer": ["Learning to Quantize Deep Networks by Optimizing Quantization Intervals with Task Loss", "Cluster-Promoting Quantization with Bit-Drop for Minimizing Network Quantization Loss"], "answer_arxiv_id": ["1808.05779", "2109.02100"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9877"} +{"question": "Which papers elaborated on the safety concerns with the advancement of LLM?", "answer": ["Universal and Transferable Adversarial Attacks on Aligned Language\n Models", "Are aligned neural networks adversarially aligned?", "\"Do Anything Now\": Characterizing and Evaluating In-The-Wild Jailbreak\n Prompts on Large Language Models", "DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT\n Models", "Image Hijacks: Adversarial Images can Control Generative Models at\n Runtime", "ChEF: A Comprehensive Evaluation Framework for Standardized Assessment\n of Multimodal Large Language Models"], "answer_arxiv_id": ["2307.15043", "2306.15447", "2308.03825", "2306.11698", "2309.00236", "2311.02692"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_9878"} +{"question": "What is the study that proposed a SISR-based change detection network with a stacked attention module?", "answer": ["Super-resolution-based Change Detection Network with Stacked Attention Module for Images with Different Resolutions"], "answer_arxiv_id": ["2103.00188v1"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_9879"} +{"question": "Which research articles introduced methods for data augmentation to handle scarcity-tailed data?", "answer": ["MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition", "Remix: Rebalanced Mixup", "Imagine by Reasoning: A Reasoning-Based Implicit Semantic Data\n Augmentation for Long-Tailed Classification"], "answer_arxiv_id": ["2103.12579", "2007.03943", "2112.07928"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_9880"} +{"question": "Which work claims that the inner optimization process of generating adversarial samples could be improved by replacing the hard label with the soft label?", "answer": ["Revisiting Adversarial Robustness Distillation: Robust Soft Labels Make\n Student Better"], "answer_arxiv_id": ["2108.07969"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_9881"} +{"question": "Could you provide me some examples of recent generative models capable of 'dreaming up' novel images from the input views' view-angles?", "answer": ["Sparse3D: Distilling Multiview-Consistent Diffusion for Object\n Reconstruction from Sparse Views", "Consistent View Synthesis with Pose-Guided Diffusion Models"], "answer_arxiv_id": ["2308.14078", "2303.17598"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_9882"} +{"question": "What works provided an interpretation of LoRA?", "answer": ["A Kernel-Based View of Language Model Fine-Tuning"], "answer_arxiv_id": ["2210.05643"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9883"} +{"question": "Which papers have proposed deep GPs to capture nonlinear correlations between fidelities?", "answer": ["Deep Gaussian Processes for Multi-fidelity Modeling", "Multi-fidelity modeling with different input domain definitions using Deep Gaussian Processes"], "answer_arxiv_id": ["1903.07320v1", "2006.15924v1"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_9884"} +{"question": "What research paper proposed the concept of energy-constrained diffusion models?", "answer": ["DIFFormer: Scalable (Graph) Transformers Induced by Energy Constrained Diffusion"], "answer_arxiv_id": ["2301.09474v4"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_9885"} +{"question": "Which papers incorporate the prior into an encoder-decoder architecture?", "answer": ["GLEAN: Generative Latent Bank for Large-Factor Image Super-Resolution", "GLEAN: Generative Latent Bank for Image Super-Resolution and Beyond", "Towards Real-World Blind Face Restoration with Generative Facial Prior", "GAN Prior Embedded Network for Blind Face Restoration in the Wild"], "answer_arxiv_id": ["2012.00739", "2207.14812", "2101.04061", "2105.06070"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_9886"} +{"question": "What research papers proposed methods for semantic segmentation in a fully supervised setting that explore global pixel relations?", "answer": ["Exploring Cross-Image Pixel Contrast for Semantic Segmentation"], "answer_arxiv_id": ["2101.11939"], "source_meta": {"published_time": "20240416"}, "qid": "AutoScholarQuery_train_9887"} +{"question": "What is the study that motivates this work by introducing Contrastive Decoding as an approach to improve the generation quality?", "answer": ["Contrastive Decoding: Open-ended Text Generation as Optimization", "Linear Alignment: A Closed-form Solution for Aligning Human Preferences\n without Tuning and Feedback"], "answer_arxiv_id": ["2210.15097", "2401.11458"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_9888"} +{"question": "Could you provide me studies that aimed to address the sharp minima problem by adjusting the objective to minimize a perturbed loss?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization", "Efficient Sharpness-aware Minimization for Improved Training of Neural\n Networks", "Surrogate Gap Minimization Improves Sharpness-Aware Training", "Towards Efficient and Scalable Sharpness-Aware Minimization", "Sharpness-Aware Gradient Matching for Domain Generalization"], "answer_arxiv_id": ["2010.01412", "2110.03141", "2203.08065", "2203.02714", "2303.10353"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_9889"} +{"question": "Which research shows that certain Graph Neural Networks (GNN) can solve cycle detection and minimum cut problems?", "answer": ["What Graph Neural Networks Cannot Learn: Depth vs Width"], "answer_arxiv_id": ["1907.03199"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_9890"} +{"question": "Which references mentioned a method that directly estimates eigenvectors of the covariance matrix without storing it during training?", "answer": ["Estimating High Order Gradients of the Data Distribution by Denoising"], "answer_arxiv_id": ["2111.04726"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_9891"} +{"question": "What learning-based representations were proposed to train models more adaptively?", "answer": ["End-to-End Learning of Representations for Asynchronous Event-Based Data"], "answer_arxiv_id": ["1904.08245"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9892"} +{"question": "What papers introduced the first matrix estimation and subspace recovery guarantees in ℓ2→∞ and ℓ∞?", "answer": ["Unperturbed: spectral analysis beyond Davis-Kahan"], "answer_arxiv_id": ["1706.06516"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_9893"} +{"question": "What research has found the presence of biases across different demographics in image search results?", "answer": ["Implicit Diversity in Image Summarization"], "answer_arxiv_id": ["1901.10265v3"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9894"} +{"question": "Could you provide some research papers that discussed different penalty terms maintianing the tradeoff between optimality in reward and safety guarantees?", "answer": ["IPO: Interior-point Policy Optimization under Constraints", "Penalized Proximal Policy Optimization for Safe Reinforcement Learning", "Reward Constrained Policy Optimization"], "answer_arxiv_id": ["1910.09615", "2205.11814", "1805.11074"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_9895"} +{"question": "Can you give me examples of research that handle potential shifts between the calibration and test set by reweighting the data points?", "answer": ["Conformal Prediction Under Covariate Shift", "Conformal Prediction Beyond Exchangeability"], "answer_arxiv_id": ["1904.06019", "2202.13415"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_9896"} +{"question": "Could you tell me some works that assume different types of invariance across domains in studying video domain generalization?", "answer": ["In Search of Lost Domain Generalization", "VideoDG: Generalizing Temporal Relations in Videos to Novel Domains", "Domain Generalization through Audio-Visual Relative Norm Alignment in First Person Action Recognition"], "answer_arxiv_id": ["2007.01434", "1912.03716", "2110.10101"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_9897"} +{"question": "Could you provide me some works about tumor synthesis based on various medical modalities?", "answer": ["Is Space-Time Attention All You Need for Video Understanding?", "The Liver Tumor Segmentation Benchmark (LiTS)", "SynthSeg: Segmentation of brain MRI scans of any contrast and resolution\n without retraining", "CancerUniT: Towards a Single Unified Model for Effective Detection,\n Segmentation, and Diagnosis of Eight Major Cancers Using a Large Collection\n of CT Scans"], "answer_arxiv_id": ["2102.05095", "1901.04056", "2107.09559", "2301.12291"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_9898"} +{"question": "What are some works in the field of reinforcement learning that applied an equivariant assumption similar to the physical symmetry?", "answer": ["Equivariant Neural Rendering"], "answer_arxiv_id": ["2006.07630"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_9899"} +{"question": "Which papers have shifted the focus of image synthesis from GANs, VAEs, and flow models to diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "gDDIM: Generalized denoising diffusion implicit models", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2206.05564", "2106.05931"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_9900"} +{"question": "Are there any works formulating graph neural networks using either learnable, fixed ℓ1-norm, or fixed ℓ2-norm pair-wise potentials?", "answer": ["Graph Neural Networks Inspired by Classical Iterative Algorithms", "Graph Unrolling Networks: Interpretable Neural Networks for Graph Signal Denoising", "Elastic Graph Neural Networks", "Predict then Propagate: Graph Neural Networks meet Personalized PageRank"], "answer_arxiv_id": ["2103.06064v4", "2006.01301", "2107.06996", "1810.05997"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_9901"} +{"question": "What studies utilized LLMs to generate both code and candidate unit tests for reranking in code generation?", "answer": ["CodeT: Code Generation with Generated Tests"], "answer_arxiv_id": ["2207.10397"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_9902"} +{"question": "What research works extended the use of Swin Transformer for 3D images?", "answer": ["Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images"], "answer_arxiv_id": ["2201.01266"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_9903"} +{"question": "What works can be referenced regarding the use of the PGD method in generating transfer attacks against unseen victim models?", "answer": ["Intriguing properties of neural networks", "Delving into Transferable Adversarial Examples and Black-box Attacks", "Universal adversarial perturbations"], "answer_arxiv_id": ["1312.6199", "1611.02770", "1610.08401"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_9904"} +{"question": "Could you provide me some studies about robust models that succeed even in conditions of distribution shift?", "answer": ["Deep Domain Confusion: Maximizing for Domain Invariance", "Domain-Adversarial Training of Neural Networks", "Invariant Risk Minimization", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Learning from Failure: Training Debiased Classifier from Biased Classifier", "Environment Inference for Invariant Learning", "Just Train Twice: Improving Group Robustness without Training Group Information"], "answer_arxiv_id": ["1412.3474", "1505.07818", "1907.02893", "1911.08731", "2007.02561", "2010.07249", "2107.09044"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_9905"} +{"question": "Which research works showed that CTC models are effective in speech-to-text generation tasks without a decoder?", "answer": ["A Comparative Study on Non-Autoregressive Modelings for Speech-to-Text\n Generation"], "answer_arxiv_id": ["2110.05249"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_9906"} +{"question": "What studies impose regularization constraints on model parameters or feature representations to handle data heterogeneity in federated learning?", "answer": ["Federated Optimization in Heterogeneous Networks", "FedDANE: A Federated Newton-Type Method", "FedPD: A Federated Learning Framework with Optimal Rates and Adaptivity\n to Non-IID Data", "Federated Learning Based on Dynamic Regularization", "Preservation of the Global Knowledge by Not-True Distillation in\n Federated Learning", "Local-Global Knowledge Distillation in Heterogeneous Federated Learning\n with Non-IID Data", "Multi-Level Branched Regularization for Federated Learning", "Divergence-aware Federated Self-Supervised Learning", "Model-Contrastive Federated Learning", "FedProc: Prototypical Contrastive Federated Learning on Non-IID data", "Federated Learning with Label Distribution Skew via Logits Calibration", "Towards Understanding and Mitigating Dimensional Collapse in\n Heterogeneous Federated Learning", "FedMix: Approximation of Mixup under Mean Augmented Federated Learning", "Acceleration of Federated Learning with Alleviated Forgetting in Local\n Training"], "answer_arxiv_id": ["1812.06127", "2001.01920", "2005.11418", "2111.04263", "2106.03097", "2107.00051", "2207.06936", "2204.04385", "2103.16257", "2109.12273", "2209.00189", "2210.00226", "2107.00233", "2203.02645"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_9907"} +{"question": "Which studies provide theories about exploiting a memory bank for fast learning?", "answer": ["Model-Free Episodic Control", "Neural Episodic Control", "Prioritized Experience Replay", "Self-Imitation Learning"], "answer_arxiv_id": ["1606.04460", "1703.01988", "1511.05952", "1806.05635"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_9908"} +{"question": "What studies propose specialized strategies to perform appropriate normalization layer's parameter updates in continual learning?", "answer": ["Continual Normalization: Rethinking Batch Normalization for Online Continual Learning", "Rebalancing Batch Normalization for Exemplar-based Class-Incremental Learning", "Diagnosing Batch Normalization in Class Incremental Learning"], "answer_arxiv_id": ["2203.16102", "2201.12559", "2202.08025"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_9909"} +{"question": "Which work spearheaded the use of reinforcement learning for improving human alignment in unstructured settings?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_9910"} +{"question": "Can you provide recent studies on multimodal large language models (MLLMs) that focus on modality alignment and instruction tuning?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2304.08485", "2304.10592", "2305.06500"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_9911"} +{"question": "What works have been done on static observation modeling that directly uses self-supervised methods from computer vision?", "answer": ["The Surprising Effectiveness of Representation Learning for Visual Imitation", "A Simple Framework for Contrastive Learning of Visual Representations", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["2112.01511", "2002.05709", "2006.07733"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_9912"} +{"question": "What method used a ResNet-18 encoder with a CNN decoder to estimate 3D bounding box properties of objects in radar images?", "answer": ["Probabilistic Oriented Object Detection in Automotive Radar"], "answer_arxiv_id": ["2004.05310"], "source_meta": {"published_time": "20240428"}, "qid": "AutoScholarQuery_train_9913"} +{"question": "Are there any research papers that investigate the use of synthetic data for training models in optical flow, autonomous driving, semantic segmentation, and human pose estimation?", "answer": ["A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation", "Augmented Reality Meets Computer Vision : Efficient Data Generation for\n Urban Driving Scenes", "Learning Semantic Segmentation from Synthetic Data: A Geometrically\n Guided Input-Output Adaptation Approach", "Learning from Synthetic Humans"], "answer_arxiv_id": ["1512.02134", "1708.01566", "1812.05040", "1701.01370"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_9914"} +{"question": "Which papers considered black-box tuning of (ε,δ)-DP mechanisms?", "answer": ["Private Selection from Private Candidates"], "answer_arxiv_id": ["1811.07971"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_9915"} +{"question": "Could you provide me with follow-up research that achieved state-of-the-art results in audio separtion using dual-path networks?", "answer": ["Dual-Path Transformer Network: Direct Context-Aware Modeling for End-to-End Monaural Speech Separation", "Effective Low-Cost Time-Domain Audio Separation Using Globally Attentive Locally Recurrent Networks", "Sandglasset: A Light Multi-Granularity Self-attentive Network For Time-Domain Speech Separation", "Attention Is All You Need in Speech Separation", "MossFormer: Pushing the Performance Limit of Monaural Speech Separation using Gated Single-head Transformer with Convolution-augmented Joint Self-Attentions"], "answer_arxiv_id": ["2007.13975", "2101.05014", "2103.00819", "2010.13154", "2302.11824"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_9916"} +{"question": "Could you provide me any studies that use priming experiments to comprehend the relationship between different linguistic tasks and recover their hierarchical organization in language models?", "answer": ["Using Priming to Uncover the Organization of Syntactic Representations\n in Neural Language Models", "Structural Persistence in Language Models: Priming as a Window into\n Abstract Language Representations"], "answer_arxiv_id": ["1909.10579", "2109.14989"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_9917"} +{"question": "What studies are related to the development of DCVC series in NVC?", "answer": ["Deep Contextual Video Compression", "Temporal Context Mining for Learned Video Compression", "Hybrid Spatial-Temporal Entropy Modelling for Neural Video Compression", "Neural Video Compression with Diverse Contexts"], "answer_arxiv_id": ["2109.15047", "2111.13850", "2207.05894", "2302.14402"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_9918"} +{"question": "What works introduced InfoGAN which learns interpretable latent codes in generative modeling?", "answer": ["InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets"], "answer_arxiv_id": ["1606.03657"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_9919"} +{"question": "Which studies involve hierarchical IL without planning?", "answer": ["Hierarchical Imitation and Reinforcement Learning", "Explainable Hierarchical Imitation Learning for Robotic Drink Pouring", "CompILE: Compositional Imitation Learning and Execution", "Adversarial Option-Aware Hierarchical Imitation Learning", "Provable Hierarchical Imitation Learning via EM", "Directed-Info GAIL: Learning Hierarchical Policies from Unsegmented Demonstrations using Directed Information", "OptionGAN: Learning Joint Reward-Policy Options using Generative Adversarial Inverse Reinforcement Learning", "Learning from Guided Play: A Scheduled Hierarchical Approach for Improving Exploration in Adversarial Imitation Learning", "Learning from Trajectories via Subgoal Discovery"], "answer_arxiv_id": ["1803.00590", "2105.07348", "1812.01483", "2106.05530", "2010.03133", "1810.01266", "1709.06683", "2112.08932", "1911.07224"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9920"} +{"question": "What studies discuss the topic of instance-dependent complexity measures?", "answer": ["Beyond No Regret: Instance-Dependent PAC Reinforcement Learning", "Instance-Dependent Near-Optimal Policy Identification in Linear MDPs via Online Experiment Design", "Asymptotic Instance-Optimal Algorithms for Interactive Decision Making"], "answer_arxiv_id": ["2108.02717", "2207.02575", "2206.02326"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_9921"} +{"question": "Which research introduced the concept of self-correction for language models?", "answer": ["Automatically Correcting Large Language Models: Surveying the landscape\n of diverse self-correction strategies"], "answer_arxiv_id": ["2308.03188"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_9922"} +{"question": "What works employ a GNN-based dynamics model with object features from a supervised detector for visual simulation?", "answer": ["Compositional Video Prediction", "Propagation Networks for Model-Based Control Under Partial Observation", "Learning Long-term Visual Dynamics with Region Proposal Interaction Networks", "Modular Action Concept Grounding in Semantic Video Prediction"], "answer_arxiv_id": ["1908.08522", "1809.11169", "2008.02265", "2011.11201"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_9923"} +{"question": "What studies have focused on shadow tomography of two-outcome quantum measurements?", "answer": ["Shadow Tomography of Quantum States", "Improved quantum data analysis"], "answer_arxiv_id": ["1711.01053v2", "2011.10908"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_9924"} +{"question": "Which work showed that a 'sparse preconditioning' step can enable Lasso to be statistically efficient for sparse linear regression when the covariates have a Markovian structure?", "answer": ["On the Power of Preconditioning in Sparse Linear Regression"], "answer_arxiv_id": ["2106.09207"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_9925"} +{"question": "Could you provide me with works that use the uncertainty of BNNs to flag adversarial examples?", "answer": ["Adversarial Phenomenon in the Eyes of Bayesian Deep Learning", "Understanding Measures of Uncertainty for Adversarial Example Detection"], "answer_arxiv_id": ["1711.08244", "1803.08533"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_9926"} +{"question": "Which papers examined various forms of bias inherent in recommendation systems?", "answer": ["Bias and Debias in Recommender System: A Survey and Future Directions", "On the Opportunity of Causal Learning in Recommendation Systems: Foundation, Estimation, Prediction and Challenges", "Causal Intervention for Leveraging Popularity Bias in Recommendation", "Asymmetric Tri-training for Debiasing Missing-Not-At-Random Explicit Feedback", "Unbiased Learning to Rank with Unbiased Propensity Estimation"], "answer_arxiv_id": ["2010.03240", "2201.06716", "2105.06067", "1910.01444", "1804.05938"], "source_meta": {"published_time": "20220319"}, "qid": "AutoScholarQuery_train_9927"} +{"question": "What are some significant papers that used the statistical mechanics of learning and HT-SR theory to predict trends in test accuracy of large-scale neural networks?", "answer": ["Predicting trends in the quality of state-of-the-art neural networks without access to training or testing data", "Post-mortem on a deep learning contest: a Simpson’s paradox and the complementary roles of scale metrics versus shape metrics"], "answer_arxiv_id": ["2002.06716v2", "2106.00734"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_9928"} +{"question": "Which works propose that the LLM could generate queries itself?", "answer": ["ReAct: Synergizing Reasoning and Acting in Language Models", "Measuring and Narrowing the Compositionality Gap in Language Models", "Toolformer: Language Models Can Teach Themselves to Use Tools"], "answer_arxiv_id": ["2210.03629", "2210.03350", "2302.04761"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_9929"} +{"question": "What research adopts simple heuristic such as inserting high quality frame per several frames in bit allocation for NVC?", "answer": ["Coarse-to-fine Deep Video Coding with Hyperprior-guided Mode Prediction", "Flexible-Rate Learned Hierarchical Bi-directional Video Compression with Motion Refinement and Frame-Level Bit Allocation", "Neural Video Compression with Diverse Contexts"], "answer_arxiv_id": ["2206.07460", "2206.13613", "2302.14402"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_9930"} +{"question": "What research has been done on zero-shot text-to-3D generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_9931"} +{"question": "Which works propose the use of a camera noise model for synthesizing low-light noisy videos for training neural networks?", "answer": ["A Physics-based Noise Formation Model for Extreme Low-light Raw\n Denoising", "Estimating Fine-Grained Noise Model via Contrastive Learning", "Physics-based Noise Modeling for Extreme Low-light Photography"], "answer_arxiv_id": ["2003.12751", "2204.01716", "2108.02158"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_9932"} +{"question": "Which research papers have used synthetic data to train single-view SVBRDF estimation networks?", "answer": ["Single-Image SVBRDF Capture with a Rendering-Aware Deep Network"], "answer_arxiv_id": ["1810.09718"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_9933"} +{"question": "Which papers discuss the Temporal Video Grounding task in the context of fine-grained video understanding?", "answer": ["TALL: Temporal Activity Localization via Language Query", "Localizing Moments in Video with Natural Language"], "answer_arxiv_id": ["1705.02101", "1708.01641"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9934"} +{"question": "Which papers propose methods for one-shot detection leveraging objectness priors?", "answer": ["One-Shot Object Detection with Co-Attention and Co-Excitation", "One-Shot Instance Segmentation", "Comparison Network for One-Shot Conditional Object Detection"], "answer_arxiv_id": ["1911.12529", "1811.11507", "1904.02317v2"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9935"} +{"question": "What works have been proposed for MDP models with low Bellman rank, low witness rank, bilinear classes, and low Bellman eluder dimension, which can specialize to low-rank MDPs?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches", "Bilinear Classes: A Structural Framework for Provable Generalization in RL", "Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms"], "answer_arxiv_id": ["1610.09512v2", "1811.08540", "2103.10897", "2102.00815"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_9936"} +{"question": "Which studies proposed the task relationship that all tasks share at least a common global minimizer?", "answer": ["How catastrophic can catastrophic forgetting be in linear regression?", "Continual learning: a feature extraction formalization, an efficient algorithm, and fundamental obstructions"], "answer_arxiv_id": ["2205.09588", "2203.14383"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_9937"} +{"question": "Which studies have demonstrated the use of external models to inject features into diffusion models?", "answer": ["Diffusion Models already have a Semantic Latent Space", "Adding Conditional Control to Text-to-Image Diffusion Models", "GLIGEN: Open-Set Grounded Text-to-Image Generation", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models"], "answer_arxiv_id": ["2210.10960", "2302.05543", "2301.07093", "2308.06721"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_9938"} +{"question": "Can you provide studies that handle fine-grained personalized federated learning?", "answer": ["Federated Multi-Task Learning", "Personalized Federated Learning with First Order Model Optimization"], "answer_arxiv_id": ["1705.10467", "2012.08565"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_9939"} +{"question": "Which studies found that probing results are influenced by contextual and frequency biases?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models", "Calibrating Factual Knowledge in Pretrained Language Models"], "answer_arxiv_id": ["2102.09690", "2210.03329"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_9940"} +{"question": "Could you provide studies about aligning the features of local and global models?", "answer": ["Model-Contrastive Federated Learning", "FedFM: Anchor-based Feature Matching for Data Heterogeneity in Federated Learning"], "answer_arxiv_id": ["2103.16257", "2210.07615"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_9941"} +{"question": "Could you inform me about the studies that have made significant progress in video inpainting using patch-based method in the deep learning era?", "answer": ["FuseFormer: Fusing Fine-Grained Information in Transformers for Video Inpainting"], "answer_arxiv_id": ["2109.02974"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_9942"} +{"question": "Which papers introduced lightweight models that can be deployed to low-resource platforms in the DNN-based speech separation methods?", "answer": ["Sudo rm -rf: Efficient Networks for Universal Audio Source Separation"], "answer_arxiv_id": ["2007.06833"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_9943"} +{"question": "What research's work does the Stable Diffusion model base on?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9944"} +{"question": "Could you provide me some works about utilizing ViL models as external knowledge to enhance downstream tasks?", "answer": ["Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "PromptStyler: Prompt-driven Style Generation for Source-free Domain\n Generalization"], "answer_arxiv_id": ["2210.04150", "2307.15199"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_9945"} +{"question": "Any works about unsupervised meta-learning for few-shot learning scenarios in tabular data?", "answer": ["Unsupervised Learning via Meta-Learning", "Revisiting Unsupervised Meta-Learning via the Characteristics of Few-Shot Tasks", "Unsupervised Meta-Learning for Few-Shot Image Classification"], "answer_arxiv_id": ["1810.02334", "2011.14663", "1811.11819"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_9946"} +{"question": "Which study is the first to present a learning-based approach to perform event-based 3D Hand Mesh Reconstruction (HMR)?", "answer": ["EventHands: Real-Time Neural 3D Hand Pose Estimation from an Event\n Stream"], "answer_arxiv_id": ["2012.06475"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_9947"} +{"question": "Which paper follows up with an adaptive bin-width estimator to improve class flexibility in classification for continuous targets?", "answer": ["AdaBins: Depth Estimation using Adaptive Bins"], "answer_arxiv_id": ["2011.14141"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_9948"} +{"question": "What papers demonstrated that incorporating up-to-date, relevant knowledge in the prompt can effectively reduce fact-conflicting hallucination?", "answer": ["FreshLLMs: Refreshing Large Language Models with Search Engine\n Augmentation", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2310.03214", "2005.11401"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_9949"} +{"question": "Could you cite some works about developing computationally efficient estimators that achieve optimal estimation rates in high dimensions?", "answer": ["High-Dimensional Robust Mean Estimation in Nearly-Linear Time"], "answer_arxiv_id": ["1811.09380"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_9950"} +{"question": "Can you name some studies that proposed approaches for improving robustness to hallucinations in LLMs, such as chain-of-thought generation and enforcing self-consistency?", "answer": ["Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2112.00114", "2201.11903", "2203.11171"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9951"} +{"question": "Which studies laid out the theoretical foundations of GFlowNets?", "answer": ["GFlowNet Foundations"], "answer_arxiv_id": ["2111.09266"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_9952"} +{"question": "Which papers focus on interactions with chairs in terms of human motion priors?", "answer": ["COUCH: Towards Controllable Human-Chair Interactions", "NIFTY: Neural Object Interaction Fields for Guided Human Motion\n Synthesis"], "answer_arxiv_id": ["2205.00541", "2307.07511"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_9953"} +{"question": "What papers talk about augmenting language models with additional textual information via using HTML tags for pretraining?", "answer": ["HTLM: Hyper-Text Pre-Training and Prompting of Language Models"], "answer_arxiv_id": ["2107.06955"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_9954"} +{"question": "What are the pioneering works that introduced the Model Inversion Attack (MIA) with shallow models?", "answer": ["Intriguing properties of neural networks"], "answer_arxiv_id": ["1312.6199"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_9955"} +{"question": "What works use Kernel Inception Distance (KID) as a metric to assess image quality?", "answer": ["Demystifying MMD GANs", "An empirical study on evaluation metrics of generative adversarial\n networks"], "answer_arxiv_id": ["1801.01401", "1806.07755"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9956"} +{"question": "What work introduced the CapEval1k dataset for training automatic evaluation metrics?", "answer": ["UMIC: An Unreferenced Metric for Image Captioning via Contrastive\n Learning"], "answer_arxiv_id": ["2106.14019"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_9957"} +{"question": "Who proposed positional encoding to improve reconstruction quality for novel view synthesis?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_9958"} +{"question": "What papers deal with the use of language models to generate increasingly complex programs?", "answer": ["Measuring Coding Challenge Competence With APPS", "[2203.07814] Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2105.09938", "2203.07814"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_9959"} +{"question": "Which study proposed an unsupervised mechanism to generate pseudo-labels for retraining by clustering raw features?", "answer": ["No Subclass Left Behind: Fine-Grained Robustness in Coarse-Grained\n Classification Problems"], "answer_arxiv_id": ["2011.12945"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_9960"} +{"question": "Which works explore novel representations for reconstructing 3D objects?", "answer": ["3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction", "Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["1604.00449v1", "1804.01654v2", "1812.03828", "2003.08934"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_9961"} +{"question": "What studies proposed pruning methods for Vision Transformers?", "answer": ["Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Vision Transformer Slimming: Multi-Dimension Searching in Continuous\n Optimization Space", "Unified Visual Transformer Compression"], "answer_arxiv_id": ["2106.04533", "2201.00814", "2203.08243"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_9962"} +{"question": "Which works explored estimation of hidden variables controlling the data generation process in the context of machine learning?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_9963"} +{"question": "Could you tell me a few examples of methods that have been developed to increase the context length of transformers by employing an attention mechanism that allows tokens to attend to distant tokens sparsely?", "answer": ["Longformer: The Long-Document Transformer", "Big Bird: Transformers for Longer Sequences", "LongT5: Efficient Text-To-Text Transformer for Long Sequences"], "answer_arxiv_id": ["2004.05150", "2007.14062", "2112.07916"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_9964"} +{"question": "What research proposed a model similar to RIR-GRC in the context of RIR-related approaches?", "answer": ["Sliced Recurrent Neural Networks"], "answer_arxiv_id": ["1807.02291v1"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_9965"} +{"question": "What works have found it difficult to inject or update knowledge for LLMs?", "answer": ["Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2202.05262"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_9966"} +{"question": "Could you give me the classic works that studied the 'column subset selection' problem?", "answer": ["An Improved Approximation Algorithm for the Column Subset Selection Problem"], "answer_arxiv_id": ["0812.4293v2"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_9967"} +{"question": "Which researches emphasize on understanding deep learning models by quantifying and visualizing the contribution of image pixels to the model output?", "answer": ["Learning Deep Features for Discriminative Localization", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based\n Localization", "Quantifying Attention Flow in Transformers", "RISE: Randomized Input Sampling for Explanation of Black-box Models", "A Unified Approach to Interpreting Model Predictions", "Layer-wise Relevance Propagation for Neural Networks with Local\n Renormalization Layers", "Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural\n Networks", "Transformer Interpretability Beyond Attention Visualization"], "answer_arxiv_id": ["1512.04150", "1610.02391", "2005.00928", "1806.07421", "1705.07874", "1604.00825", "1910.01279", "2012.09838"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_9968"} +{"question": "What works propose to predict masked subseries-level patches?", "answer": ["A Time Series is Worth 64 Words: Long-term Forecasting with Transformers"], "answer_arxiv_id": ["2211.14730"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_9969"} +{"question": "Can you provide some studies revolving around knowledge distillation?", "answer": ["Distilling the Knowledge in a Neural Network", "DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter", "Collaborative Distillation for Ultra-Resolution Universal Style Transfer"], "answer_arxiv_id": ["1503.02531", "1910.01108", "2003.08436"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_9970"} +{"question": "What works focused on the topic of minimizing the divergence between two distributions over functions through the function-space evidence lower bound?", "answer": ["Functional Variational Bayesian Neural Networks"], "answer_arxiv_id": ["1903.05779"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_9971"} +{"question": "Which study introduced a small number of trainable parameters in the input space to fine-tune the model with high quality and efficiency?", "answer": ["Visual Prompt Tuning for Test-time Domain Adaptation"], "answer_arxiv_id": ["2210.04831"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_9972"} +{"question": "Could you provide examples of studies that took special care of the intermediate-level representations to enhance the adversarial transferability?", "answer": ["Enhancing Adversarial Example Transferability with an Intermediate Level Attack", "An Intermediate-level Attack Framework on The Basis of Linear Regression", "Feature Importance-aware Transferable Adversarial Attacks", "Improving Adversarial Transferability via Neuron Attribution-Based Attacks"], "answer_arxiv_id": ["1907.10823", "2203.10723", "2107.14185", "2204.00008"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_9973"} +{"question": "What are some studies that use Graph Neural Networks (GNNs) to handle different robot morphologies?", "answer": ["Learning to Control Self-Assembling Morphologies: A Study of Generalization via Modularity", "One Policy to Control Them All: Shared Modular Policies for Agent-Agnostic Control"], "answer_arxiv_id": ["1902.05546", "2007.04976"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_9974"} +{"question": "Any works about Diffusion Models with transformer-based architectures?", "answer": ["Scalable Diffusion Models with Transformers"], "answer_arxiv_id": ["2212.09748"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_9975"} +{"question": "Are there any studies where the maximin aggregator solely relies on a random expert's opinion within a robust forecast aggregation?", "answer": ["Robust Merging of Information"], "answer_arxiv_id": ["2106.00088v1"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_9976"} +{"question": "Could you provide me some works that proposed to solve incidental correlations of image background using augmentation methods?", "answer": ["Noise or Signal: The Role of Image Backgrounds in Object Recognition", "Rectifying the Shortcut Learning of Background for Few-Shot Learning", "Few-shot Learning via Saliency-guided Hallucination of Samples"], "answer_arxiv_id": ["2006.09994", "2107.07746", "1904.03472"], "source_meta": {"published_time": "20230930"}, "qid": "AutoScholarQuery_train_9977"} +{"question": "Can you name the papers that focus on applications of Physics-Informed Neural Networks (PINNs)?", "answer": ["Unsupervised Reservoir Computing for Solving Ordinary Differential Equations", "Residual-based adaptivity for two-phase flow simulation in porous media using Physics-informed Neural Networks", "Transfer learning enhanced physics informed neural network for phase-field modeling of fracture", "Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios", "SVD-PINNs: Transfer Learning of Physics-Informed Neural Networks via Singular Value Decomposition", "One-Shot Transfer Learning of Physics-Informed Neural Networks", "Physics-Informed Neural Operator for Learning Partial Differential Equations"], "answer_arxiv_id": ["2108.11417", "2109.14290", "1907.02531v1", "2205.07731", "2211.08760", "2110.11286v2", "2111.03794"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_9978"} +{"question": "Could you provide me research works on scene editing related to NeRF?", "answer": ["NeRF for Outdoor Scene Relighting", "ClimateNeRF: Extreme Weather Synthesis in Neural Radiance Field"], "answer_arxiv_id": ["2112.05140", "2211.13226"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_9979"} +{"question": "Which studies have shown limitations in inductive matrix completion methods due to the inferior expressiveness of the feature space?", "answer": ["Inductive Matrix Completion Based on Graph Neural Networks", "Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach"], "answer_arxiv_id": ["1904.12058", "2007.04833"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_9980"} +{"question": "Could you indicate works that provide a regret lower bound for unknown systems in LQR with full cost feedback?", "answer": ["Logarithmic Regret for Learning Linear Quadratic Regulators Efficiently", "Naive Exploration is Optimal for Online LQR", "Geometric Exploration for Online Control"], "answer_arxiv_id": ["2002.08095", "2001.09576", "2010.13178"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_9981"} +{"question": "Could you provide me some studies about image segmentation models based on visual in-context learning?", "answer": ["Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning", "SegGPT: Segmenting Everything In Context"], "answer_arxiv_id": ["2212.02499", "2304.03284"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_9982"} +{"question": "Could you provide me some studies that introduce new modeling strategies for HD map construction?", "answer": ["End-to-End Vectorized HD-map Construction with Piecewise Bezier Curve", "PivotNet: Vectorized Pivot Learning for End-to-end HD Map Construction"], "answer_arxiv_id": ["2306.09700", "2308.16477"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_9983"} +{"question": "Which works addressed robust learning of monocular depth estimation?", "answer": ["Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer", "MegaDepth: Learning Single-View Depth Prediction from Internet Photos", "Web Stereo Video Supervision for Depth Prediction from Dynamic Scenes", "Learning Single-Image Depth from Videos using Quality Assessment Networks", "Hierarchical Normalization for Robust Monocular Depth Estimation"], "answer_arxiv_id": ["1907.01341", "1804.00607", "1904.11112", "1806.09573", "2210.09670"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_9984"} +{"question": "What works use source images and augmentations but do not rely on pairs of augmentations?", "answer": ["Self-Supervised Learning of Pretext-Invariant Representations"], "answer_arxiv_id": ["1912.01991"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_9985"} +{"question": "Could you provide some references that apply CG in reinforcement-learning?", "answer": ["Language as an Abstraction for Hierarchical Deep Reinforcement Learning", "Neuro-algorithmic Policies enable Fast Combinatorial Generalization"], "answer_arxiv_id": ["1906.07343", "2102.07456"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_9986"} +{"question": "Which works suggested personalizing the Federated learning model with few steps of gradient descent?", "answer": ["Federated Evaluation of On-device Personalization", "Personalized Federated Learning with Moreau Envelopes"], "answer_arxiv_id": ["1910.10252", "2006.08848"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_9987"} +{"question": "Which research work introduced non-linear messages for irreps features?", "answer": ["Geometric and Physical Quantities improve E(3) Equivariant Message Passing"], "answer_arxiv_id": ["2110.02905"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_9988"} +{"question": "Which worksdiscarded pruned token completely?", "answer": ["AdaViT: Adaptive Tokens for Efficient Vision Transformer", "Adaptive Token Sampling For Efficient Vision Transformers", "AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification"], "answer_arxiv_id": ["2112.07658", "2111.15667", "2111.15668", "2106.02034v2"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_9989"} +{"question": "Which studies discuss property inference attacks that capture larger properties of the entire training set?", "answer": ["Hacking Smart Machines with Smarter Ones: How to Extract Meaningful Data from Machine Learning Classifiers"], "answer_arxiv_id": ["1306.4447"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_9990"} +{"question": "What works adopt the approach of knowledge distillation from large-scale pre-trained models for 2D open-vocabulary segmentation?", "answer": ["SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary Semantic Segmentation", "GroupViT: Semantic Segmentation Emerges from Text Supervision", "Language-driven Semantic Segmentation", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models", "Generalized Decoding for Pixel, Image, and Language", "Decoupling Zero-Shot Semantic Segmentation", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP"], "answer_arxiv_id": ["2211.14813", "2202.11094", "2201.03546", "2303.04803", "2212.11270", "2112.07910", "2210.04150"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_9991"} +{"question": "What is the paper that proposed the comparison of image patches with the words in the sentence in the context of contrastive learning?", "answer": ["FILIP: Fine-grained Interactive Language-Image Pre-Training"], "answer_arxiv_id": ["2111.07783"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_9992"} +{"question": "What papers proposed the use of the scheduling ability of LLMs to dynamically integrate pre-trained vision models?", "answer": ["Chameleon: Plug-and-Play Compositional Reasoning with Large Language Models", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face"], "answer_arxiv_id": ["2304.09842", "2303.04671", "2303.17580"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_9993"} +{"question": "Which studies are focused on accelerating training and rendering of NeRF?", "answer": ["Progressively-connected Light Field Network for Efficient View Synthesis", "DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields using Depth Oracle Networks", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "AutoInt: Automatic Integration for Fast Neural Volume Rendering", "Baking Neural Radiance Fields for Real-Time View Synthesis", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "ReLU Fields: The Little Non-linearity That Could", "TensoRF: Tensorial Radiance Fields", "Delicate Textured Mesh Recovery from NeRF via Adaptive Surface Refinement", "MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures", "BakedSDF: Meshing Neural SDFs for Real-Time View Synthesis", "Learning Neural Duplex Radiance Fields for Real-Time View Synthesis", "NeRFMeshing: Distilling Neural Radiance Fields into Geometrically-Accurate 3D Meshes", "Re-ReND: Real-time Rendering of NeRFs across Devices", "R2L: Distilling Neural Radiance Field to Neural Light Field for Efficient Novel View Synthesis", "Real-Time Neural Light Field on Mobile Devices", "Learning Neural Light Fields with Ray-Space Embedding", "Light Field Networks: Neural Scene Representations with Single-Evaluation Rendering"], "answer_arxiv_id": ["2207.04465", "2103.03231", "2103.13744", "2012.01714", "2103.14645", "2103.14024", "2103.13744", "2205.10824", "2203.09517", "2303.02091", "2208.00277", "2302.14859", "2304.10537", "2303.09431", "2303.08717", "2203.17261", "2212.08057", "2112.01523", "2106.02634"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_9994"} +{"question": "What works can be considered as the source of introducing sparsity in deep neural networks mainly by model compression?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding"], "answer_arxiv_id": ["1510.00149"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_9995"} +{"question": "Are there any works that leverage transfer learning to make pre-trained models customized for a specific concept?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion"], "answer_arxiv_id": ["2208.12242", "2208.01618"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_9996"} +{"question": "Which work introduced the SAM optimizer?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization"], "answer_arxiv_id": ["2010.01412"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_9997"} +{"question": "Could you provide me some works about strategic classification?", "answer": ["Strategic Classification"], "answer_arxiv_id": ["1506.06980"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_9998"} +{"question": "Which datasets include depth maps motivated by the Kinect device?", "answer": ["BigHand2.2M Benchmark: Hand Pose Dataset and State of the Art Analysis", "Real-time Joint Tracking of a Hand Manipulating an Object from RGB-D Input"], "answer_arxiv_id": ["1704.02612", "1610.04889"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_9999"} +{"question": "What papers proposed the use of local reconstruction objectives and asymmetric weights to improve semi-supervised learning performance?", "answer": ["Semi-Supervised Learning with Ladder Networks"], "answer_arxiv_id": ["1507.02672"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_10000"} +{"question": "What works focus on developing model pruning methods?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding", "The Lottery Tickets Hypothesis for Supervised and Self-supervised Pre-training in Computer Vision Models", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Linear Mode Connectivity and the Lottery Ticket Hypothesis", "Sanity Checks for Lottery Tickets: Does Your Winning Ticket Really Win the Jackpot?", "Advancing Model Pruning via Bi-level Optimization"], "answer_arxiv_id": ["1510.00149", "2012.06908", "1803.03635", "1912.05671", "2107.00166", "2210.04092"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_10001"} +{"question": "Which works proposed the use of Transformers to capture long-term dependency in Neural Temporal Point Processes (NTPPs)?", "answer": ["Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks", "Self-Attentive Hawkes Process", "Transformer Hawkes Process", "Transformer Embeddings of Irregularly Spaced Events and Their Participants"], "answer_arxiv_id": ["1908.01207", "1907.07561", "2002.09291", "2201.00044"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_10002"} +{"question": "What papers discussed the impact of fine-tuning in deep learning?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1810.04805", "1911.05722"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_10003"} +{"question": "Which studies talk about generalising models to Out of Distribution (OOD) datasets?", "answer": ["Open-World Semi-Supervised Learning"], "answer_arxiv_id": ["2102.03526"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_10004"} +{"question": "Which works contain challenging questions involving temporal commonsense reasoning over the duration, frequency, temporal order, and more aspects of events in question answering?", "answer": ["Torque: A Reading Comprehension Dataset of Temporal Ordering Questions", "A Dataset for Answering Time-Sensitive Questions"], "answer_arxiv_id": ["2005.00242", "2108.06314"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_10005"} +{"question": "Are there works about natural policy gradient in entropy regularized potential games?", "answer": ["Fast Global Convergence of Natural Policy Gradient Methods with Entropy Regularization"], "answer_arxiv_id": ["2007.06558"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_10006"} +{"question": "Which papers explored powerful multi-task model architectures?", "answer": ["MultiMAE: Multi-modal Multi-task Masked Autoencoders", "Multi-Task Learning with Multi-Query Transformer for Dense Prediction", "TaskExpert: Dynamically Assembling Multi-Task Representations with\n Memorial Mixture-of-Experts", "InvPT++: Inverted Pyramid Multi-Task Transformer for Visual Scene\n Understanding", "Universal Representations: A Unified Look at Multiple Task and Domain\n Learning", "NDDR-CNN: Layerwise Feature Fusing in Multi-Task CNNs by Neural\n Discriminative Dimensionality Reduction", "MTL-NAS: Task-Agnostic Neural Architecture Search towards\n General-Purpose Multi-Task Learning", "AutoMTL: A Programming Framework for Automating Efficient Multi-Task\n Learning"], "answer_arxiv_id": ["2204.01678", "2205.14354", "2307.15324", "2306.04842", "2204.02744", "1801.08297", "2003.14058", "2110.13076"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_10007"} +{"question": "Could you provide me some works that suggest the advantage of adaptive algorithms to achieve order-optimal rates without knowledge about problem parameters?", "answer": ["AdaGrad stepsizes: Sharp convergence over nonconvex landscapes", "UniXGrad: A Universal, Adaptive Algorithm with Optimal Guarantees for Constrained Optimization"], "answer_arxiv_id": ["1806.01811", "1910.13857"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_10008"} +{"question": "Which research works can provide information about physics-informed neural networks (PINO) used in Fourier Neural Operators?", "answer": ["Physics-Informed Neural Operator for Learning Partial Differential Equations"], "answer_arxiv_id": ["2111.03794"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_10009"} +{"question": "What works have proposed augmenting training data with synthetically generated X-rays for self-supervised 2D/3D registration?", "answer": ["Self-Supervised 2D/3D Registration for X-Ray to CT Image Fusion", "X-Ray to CT Rigid Registration Using Scene Coordinate Regression"], "answer_arxiv_id": ["2210.07611", "2311.15087"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_10010"} +{"question": "Which studies contributed to the development of image editing?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2108.01073", "2211.09800"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_10011"} +{"question": "Who purposed quantum bandit algorithms for stochastic convex bandits?", "answer": ["Quantum Speedups of Optimizing Approximately Convex Functions with Applications to Logarithmic Regret Stochastic Convex Bandits"], "answer_arxiv_id": ["2209.12897v1"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_10012"} +{"question": "Which works focused on the synthetic-to-real generalization for robust stereo matching?", "answer": ["Domain-invariant Stereo Matching Networks", "Matching-space Stereo Networks for Cross-domain Generalization", "Revisiting Domain Generalized Stereo Matching Networks from a Feature\n Consistency Perspective", "ITSA: An Information-Theoretic Approach to Automatic Shortcut Avoidance\n and Domain Generalization in Stereo Matching Networks", "NeRF-Supervised Deep Stereo"], "answer_arxiv_id": ["1911.13287", "2010.07347", "2203.10887", "2201.02263", "2303.17603"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_10013"} +{"question": "What research papers considered multimodal generation and methods to generate tokens that a VQ-GAN can then decode into an image?", "answer": ["Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "Unified Language-Vision Pretraining in LLM with Dynamic Discrete Visual\n Tokenization", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "Emu: Generative Pretraining in Multimodality", "Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction\n Tuning"], "answer_arxiv_id": ["2206.08916", "2309.04669", "2202.03052", "2307.05222", "2309.02591"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_10014"} +{"question": "Could you mention a study that developed a generalization assessment index for SR networks?", "answer": ["Evaluating the Generalization Ability of Super-Resolution Networks"], "answer_arxiv_id": ["2205.07019"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_10015"} +{"question": "What studies took early steps toward evaluating hallucinations in Language-Visual Language Models (LVLMs)?", "answer": ["Evaluating Object Hallucination in Large Vision-Language Models", "Negative Object Presence Evaluation (NOPE) to Measure Object\n Hallucination in Vision-Language Models"], "answer_arxiv_id": ["2305.10355", "2310.05338"], "source_meta": {"published_time": "20240630"}, "qid": "AutoScholarQuery_train_10016"} +{"question": "What papers discuss the introduction of more difficult math datasets, including high-school and college-level datasets?", "answer": ["Training Verifiers to Solve Math Word Problems", "NumGLUE: A Suite of Fundamental yet Challenging Mathematical Reasoning\n Tasks", "Measuring Mathematical Problem Solving With the MATH Dataset", "ARB: Advanced Reasoning Benchmark for Large Language Models", "MiniF2F: a cross-system benchmark for formal Olympiad-level mathematics"], "answer_arxiv_id": ["2110.14168", "2204.05660", "2103.03874", "2307.13692", "2109.00110"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_10017"} +{"question": "Which studies discuss the concept of invisible information hiding in QR codes?", "answer": ["StegaStamp: Invisible Hyperlinks in Physical Photographs"], "answer_arxiv_id": ["1904.05343"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_10018"} +{"question": "Which works demonstrate how to impose explicit constraints on HNNs and LNNs?", "answer": ["Simplifying Hamiltonian and Lagrangian Neural Networks via Explicit Constraints"], "answer_arxiv_id": ["2010.13581"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10019"} +{"question": "Can you name the study that used 9x9 depthwise convolution to replace the spatial mixer of ViT and MLP-Mixer?", "answer": ["Patches Are All You Need?"], "answer_arxiv_id": ["2201.09792"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_10020"} +{"question": "Which works discuss the tendency of overfitting in FSCIL?", "answer": ["Prototypical Networks for Few-shot Learning", "Learning to Compare: Relation Network for Few-Shot Learning"], "answer_arxiv_id": ["1703.05175", "1711.06025"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_10021"} +{"question": "What studies use scene priors such as the spatial relationship between objects for approximating the scene layout distribution?", "answer": ["Human-centric Indoor Scene Synthesis Using Stochastic Grammar"], "answer_arxiv_id": ["1808.08473"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_10022"} +{"question": "Which papers explore scientific-domain edits and tag edit intentions on ArXiv papers without using feedback?", "answer": ["arXivEdits: Understanding the Human Revision Process in Scientific\n Writing", "Understanding Iterative Revision from Human-Written Text"], "answer_arxiv_id": ["2210.15067", "2203.03802"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_10023"} +{"question": "Can you provide studies that propose methods based on representation learning in the context of imbalanced learning?", "answer": ["Parametric Contrastive Learning", "Generalized Parametric Contrastive Learning"], "answer_arxiv_id": ["2107.12028", "2209.12400"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_10024"} +{"question": "Which works involve the use of Vector-Quantized VAEs in deep generative models?", "answer": ["Neural Discrete Representation Learning"], "answer_arxiv_id": ["1711.00937"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_10025"} +{"question": "Which studies involve the use of Transformer architecture in vision-language pre-training?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10026"} +{"question": "What research studies the problem of mean estimation under probabilistic assumptions on the data generating process?", "answer": ["Finite Sample Differentially Private Confidence Intervals", "Average-Case Averages: Private Algorithms for Smooth Sensitivity and Mean Estimation", "Private Mean Estimation of Heavy-Tailed Distributions", "CoinPress: Practical Private Mean and Covariance Estimation", "Robust and differentially private mean estimation", "Differential privacy and robust statistics in high dimensions"], "answer_arxiv_id": ["1711.03908v1", "1906.02830", "2002.09464", "2006.06618", "2102.09159", "2111.06578"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_10027"} +{"question": "What studies are about the use of nondeterministic stacks in memory-augmented networks?", "answer": ["Learning Context-Free Languages with Nondeterministic Stack RNNs", "Learning Hierarchical Structures with Differentiable Nondeterministic Stacks"], "answer_arxiv_id": ["2010.04674", "2109.01982"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_10028"} +{"question": "Any studies that proposed reinforcement learning applications for Large Language Models (LLMs) in dialogue systems?", "answer": ["Way Off-Policy Batch Deep Reinforcement Learning of Implicit Human\n Preferences in Dialog", "Towards Coherent and Engaging Spoken Dialog Response Generation Using\n Automatic Conversation Evaluators"], "answer_arxiv_id": ["1907.00456", "1904.13015"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_10029"} +{"question": "Which papers have addressed pose estimation for vehicles and people in presence of occlusion?", "answer": ["Robust Category-Level 6D Pose Estimation with Coarse-to-Fine Rendering\n of Neural Features", "NeMo: Neural Mesh Models of Contrastive Features for Robust 3D Pose\n Estimation", "Humans in 4D: Reconstructing and Tracking Humans with Transformers", "PARE: Part Attention Regressor for 3D Human Body Estimation"], "answer_arxiv_id": ["2209.05624", "2101.12378", "2305.20091", "2104.08527"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_10030"} +{"question": "Which research studies use diffusion models for decision-making tasks?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis", "Is Conditional Generative Modeling all you need for Decision-Making?", "Learning Universal Policies via Text-Guided Video Generation", "Diffusion Policy: Visuomotor Policy Learning via Action Diffusion", "Imitating Human Behaviour with Diffusion Models", "AdaptDiffuser: Diffusion Models as Adaptive Self-evolving Planners", "Diffusion-based Generation, Optimization, and Planning in 3D Scenes", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "StructDiffusion: Language-Guided Creation of Physically-Valid Structures using Unseen Objects"], "answer_arxiv_id": ["2205.09991", "2211.15657", "2302.00111", "2303.04137v5", "2301.10677", "2302.01877", "2301.06015", "2208.06193", "2211.04604"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_10031"} +{"question": "Which research paper uses relative distances between nodes to capture positional information?", "answer": ["Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning"], "answer_arxiv_id": ["2009.00142"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10032"} +{"question": "Which studies focused on the federated linear contextual bandit model with a consideration on federated differential privacy?", "answer": ["Differentially-Private Federated Linear Bandits", "Federated Linear Contextual Bandits"], "answer_arxiv_id": ["2010.11425", "2110.14177"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_10033"} +{"question": "Could you provide me some examples of studies about using large generative models for data augmentation?", "answer": ["Data Augmentation for Low-Resource Neural Machine Translation", "Improving Neural Machine Translation Models with Monolingual Data", "GPT3Mix: Leveraging Large-scale Language Models for Text Augmentation", "PromDA: Prompt-based Data Augmentation for Low-Resource NLU Tasks"], "answer_arxiv_id": ["1705.00440", "1511.06709v4", "2104.08826", "2202.12499"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_10034"} +{"question": "Which work introduced the utilization of Generative Adversarial Network in ZSD?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["2203.00667"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_10035"} +{"question": "Can you name some research papers that show the use of in-context learning in code generation?", "answer": ["Natural Language to Code Generation in Interactive Data Science Notebooks"], "answer_arxiv_id": ["2212.09248"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_10036"} +{"question": "What research has studied various f-divergences, such as the Bhattacharyya distance or the Henze-Penrose divergence?", "answer": ["Empirically Estimable Classification Bounds Based on a Nonparametric Divergence Measure", "Learning to Bound the Multi-class Bayes Error"], "answer_arxiv_id": ["1412.6534", "1811.06419"], "source_meta": {"published_time": "20220201"}, "qid": "AutoScholarQuery_train_10037"} +{"question": "Which papers explored the power of representation learning via pre-training in fields like computer vision and natural language processing?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Momentum Contrast for Unsupervised Visual Representation Learning", "Representation Learning with Contrastive Predictive Coding", "Flamingo: a Visual Language Model for Few-Shot Learning", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "Learning Transferable Visual Models From Natural Language Supervision", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["1505.05192", "1911.05722", "1807.03748", "2204.14198", "1810.04805", "2005.14165", "2103.00020", "2204.02311"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_10038"} +{"question": "Which paper showed that unsupervised clustering aids better than MLM pre-training?", "answer": ["Cluster & Tune: Boost Cold Start Performance in Text Classification"], "answer_arxiv_id": ["2203.10581"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_10039"} +{"question": "Any works that use instruction-based editing methods for image editing?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions", "MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image\n Editing"], "answer_arxiv_id": ["2211.09800", "2306.10012"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_10040"} +{"question": "Could you provide the study that designs MOOD, the out-of-distribution molecule generation scheme with score-based diffusion?", "answer": ["Exploring Chemical Space with Score-based Out-of-distribution Generation"], "answer_arxiv_id": ["2206.07632"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_10041"} +{"question": "Could you provide me some papers that propose self-supervised learning for transformers due to their requirement of large amounts of data?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "GPT-4 Technical Report"], "answer_arxiv_id": ["1810.04805", "1907.11692", "2303.08774"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_10042"} +{"question": "Could you provide me references for the most commonly adopted paradigms for solving safe RL problems, called primal-dual approaches?", "answer": ["Risk-Constrained Reinforcement Learning with Percentile Risk Criteria", "Constrained Policy Optimization", "Reward Constrained Policy Optimization", "Responsive Safety in Reinforcement Learning by PID Lagrangian Methods", "Constrained Policy Optimization via Bayesian World Models"], "answer_arxiv_id": ["1512.01629", "1705.10528", "1805.11074", "2007.03964", "2201.09802"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10043"} +{"question": "Could you provide me some studies about the ability of an MPNN to approximate functions over the domain of graphs?", "answer": ["On the Equivalence between Graph Isomorphism Testing and Function Approximation with GNNs", "Expressiveness and Approximation Properties of Graph Neural Networks", "A Simple Proof of the Universality of Invariant/Equivariant Graph Neural Networks"], "answer_arxiv_id": ["1905.12560", "2204.04661", "1910.03802"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_10044"} +{"question": "Could you tell me what studies followed the contrastive learning successful methods to apply them to supervised learning or extend to other domains like NLP or graph?", "answer": ["Supervised Contrastive Learning", "Learning Transferable Visual Models From Natural Language Supervision", "Graph Contrastive Learning with Augmentations"], "answer_arxiv_id": ["2004.11362", "2103.00020", "2010.13902"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10045"} +{"question": "Which work obtained corresponding results for deep linear networks and showed the existence of bad saddles in parameter space for networks with more than three layers?", "answer": ["Deep Learning without Poor Local Minima"], "answer_arxiv_id": ["1605.07110"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_10046"} +{"question": "Can you reference some works that have focused on establishing 2D-3D correspondences through keypoint detection or pixel-wise 3D coordinate estimation?", "answer": ["GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D\n Object Pose Estimation", "PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation", "DPOD: 6D Pose Object Detector and Refiner"], "answer_arxiv_id": ["2102.12145", "1812.11788", "1902.11020"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_10047"} +{"question": "What work proposed an approach to quantify training data influence using Influence Function?", "answer": ["Understanding Black-box Predictions via Influence Functions"], "answer_arxiv_id": ["1703.04730"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_10048"} +{"question": "Which research works have attracted a lot of research interest by being successful large language models?", "answer": ["Language Models are Few-Shot Learners", "Training language models to follow instructions with human feedback", "PaLM: Scaling Language Modeling with Pathways", "PaLM 2 Technical Report", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2005.14165", "2203.02155", "2204.02311", "2305.10403", "2302.13971"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_10049"} +{"question": "What studies have been done on open-ended video question answering?", "answer": ["TVQA: Localized, Compositional Video Question Answering", "Hierarchical Conditional Relation Networks for Video Question Answering", "Invariant Grounding for Video Question Answering"], "answer_arxiv_id": ["1809.01696", "2002.10698", "2206.02349"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_10050"} +{"question": "Who argued for the importance of releasing a Datasheet artifact documenting the motivation, composition, and collection process for a dataset?", "answer": ["Datasheets for Datasets"], "answer_arxiv_id": ["1803.09010"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_10051"} +{"question": "Could you provide me some works that introduced additional vision-to-language adaptation modules to mitigate the challenges in learning visual-language alignment?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models"], "answer_arxiv_id": ["2301.12597"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_10052"} +{"question": "What previous works fall under the category of unsupervised representation learning?", "answer": ["Unsupervised Point Cloud Representation Learning with Deep Neural\n Networks: A Survey"], "answer_arxiv_id": ["2202.13589"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_10053"} +{"question": "Which works use a neural radiance field NeRF for stylization?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "ARF: Artistic Radiance Fields", "StyleRF: Zero-shot 3D Style Transfer of Neural Radiance Fields", "Unified Implicit Neural Stylization"], "answer_arxiv_id": ["2003.08934", "2206.06360", "2303.10598", "2204.01943"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_10054"} +{"question": "What works have been done on pose estimation for animals?", "answer": ["Cross-Domain Adaptation for Animal Pose Estimation"], "answer_arxiv_id": ["1908.05806"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_10055"} +{"question": "What studies used an off-policy algorithm with a behavioral cloning regularization term to achieve state-of-the-art performances?", "answer": ["A Minimalist Approach to Offline Reinforcement Learning", "Addressing Function Approximation Error in Actor-Critic Methods"], "answer_arxiv_id": ["2106.06860", "1802.09477"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_10056"} +{"question": "What works study the incorporation of neural networks with random functions in fine-tuning?", "answer": ["Meta-learning autoencoders for few-shot prediction", "Meta-Learning Probabilistic Inference For Prediction"], "answer_arxiv_id": ["1807.09912", "1805.09921"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_10057"} +{"question": "Which studies introduced approaches similar to the use of VQ-VAE models in Text-to-video (T2V) generation?", "answer": ["GODIVA: Generating Open-DomaIn Videos from nAtural Descriptions"], "answer_arxiv_id": ["2104.14806"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_10058"} +{"question": "Which papers proposed transformer-based approaches for the tasks of numerical time series modeling?", "answer": ["Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecasting", "Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting", "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting", "FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting", "A Transformer-based Framework for Multivariate Time Series Representation Learning", "Anomaly Transformer: Time Series Anomaly Detection with Association Discrepancy", "A Time Series is Worth 64 Words: Long-term Forecasting with Transformers"], "answer_arxiv_id": ["1907.00235", "2012.07436", "2106.13008", "2201.12740", "2010.02803", "2110.02642", "2211.14730"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_10059"} +{"question": "Could you provide some works that generalize classifiers to any energy-based functions in the context of conditional generation with SBDMs?", "answer": ["EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations"], "answer_arxiv_id": ["2207.06635"], "source_meta": {"published_time": "20230804"}, "qid": "AutoScholarQuery_train_10060"} +{"question": "Are there any studies on how StyleGAN 'knows' 3D information?", "answer": ["Do 2D GANs Know 3D Shape? Unsupervised 3D shape reconstruction from 2D Image GANs", "Image GANs meet Differentiable Rendering for Inverse Graphics and Interpretable 3D Neural Rendering", "A Shading-Guided Generative Implicit Model for Shape-Accurate 3D-Aware Image Synthesis", "GAN2X: Non-Lambertian Inverse Rendering of Image GANs", "VoLux-GAN: A Generative Model for 3D Face Synthesis with HDRI Relighting"], "answer_arxiv_id": ["2011.00844", "2010.09125", "2110.15678", "2206.09244", "2201.04873"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10061"} +{"question": "What papers have applied graph-based reasoning to tasks such as scene graph generation, visual question answering, natural language generation, and cross-modal retrieval?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "Graph R-CNN for Scene Graph Generation", "Language-Conditioned Graph Networks for Relational Reasoning", "Relation-Aware Graph Attention Network for Visual Question Answering", "Joint learning of object graph and relation graph for visual question answering", "Say As You Wish: Fine-grained Control of Image Caption Generation with Abstract Scene Graphs", "Fine-grained Video-Text Retrieval with Hierarchical Graph Reasoning"], "answer_arxiv_id": ["1609.02907", "1710.10903", "1808.00191", "1905.04405", "1903.12314", "2205.04188", "2003.00387", "2003.00392"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_10062"} +{"question": "Which works have a focus on a single domain like hyperparameter optimization in combinatorial optimization benchmarks?", "answer": ["HPOBench: A Collection of Reproducible Multi-Fidelity Benchmark Problems for HPO", "HPO-B: A Large-Scale Reproducible Benchmark for Black-Box HPO based on OpenML", "YAHPO Gym - An Efficient Multi-Objective Multi-Fidelity Benchmark for Hyperparameter Optimization"], "answer_arxiv_id": ["2109.06716", "2106.06257", "2109.03670"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10063"} +{"question": "What work studied neural estimators in settings employing Gaussian variables and axis-wise cubic transformations?", "answer": ["On Variational Bounds of Mutual Information"], "answer_arxiv_id": ["1905.06922"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_10064"} +{"question": "Which studies relied on detecting the relation proposals directly using a pre-trained object detector to predict triplet queries?", "answer": ["Learning of Visual Relations: The Devil is in the Tails", "Single-Stage Visual Relationship Learning using Conditional Queries", "Bipartite Graph Network with Adaptive Message Passing for Unbiased Scene\n Graph Generation", "Structured Sparse R-CNN for Direct Scene Graph Generation"], "answer_arxiv_id": ["2108.09668", "2306.05689", "2104.00308", "2106.10815"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_10065"} +{"question": "What works defined the task of predicting the bounding box of the entire object as amodal detection?", "answer": ["Amodal Completion and Size Constancy in Natural Scenes", "TAO-Amodal: A Benchmark for Tracking Any Object Amodally"], "answer_arxiv_id": ["1509.08147", "2312.12433"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_10066"} +{"question": "What works discuss the spatial correspondence between video and audio?", "answer": ["2.5D Visual Sound", "Learning Representations from Audio-Visual Spatial Alignment", "Egocentric Deep Multi-Channel Audio-Visual Active Speaker Localization"], "answer_arxiv_id": ["1812.04204", "2011.01819", "2201.01928"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_10067"} +{"question": "Which papers discuss the development of image-based 4D scene rendering by treating time as an extended input dimension to NeRF?", "answer": ["Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Space-time Neural Irradiance Fields for Free-Viewpoint Video", "Dynamic View Synthesis from Dynamic Monocular Video", "Neural Radiance Flow for 4D View Synthesis and Video Processing", "Neural Volumes: Learning Dynamic Renderable Volumes from Images"], "answer_arxiv_id": ["2011.13084", "2011.12950", "2105.06468", "2012.09790", "1906.07751"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_10068"} +{"question": "What works convert point clouds into pseudo images for 3D object detection?", "answer": ["Range Conditioned Dilated Convolutions for Scale Invariant 3D Object Detection", "RangeDet: In Defense of Range View for LiDAR-based 3D Object Detection", "RSN: Range Sparse Net for Efficient, Accurate LiDAR 3D Object Detection", "Fully Convolutional One-Stage 3D Object Detection on LiDAR Range Images", "FCAF3D: Fully Convolutional Anchor-Free 3D Object Detection"], "answer_arxiv_id": ["2005.09927", "2103.10039", "2106.13365", "2205.13764", "2112.00322"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_10069"} +{"question": "What research works have been done in the area of modeling relations between object and human in two-stage HOI detection?", "answer": ["DRG: Dual Relation Graph for Human-Object Interaction Detection", "Spatially Conditioned Graphs for Detecting Human–Object Interactions", "VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions"], "answer_arxiv_id": ["2008.11714", "2012.06060", "2003.05541"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_10070"} +{"question": "Which studies proposed token pausing or halting approaches in semantic segmentation?", "answer": ["PAUMER: Patch Pausing Transformer for Semantic Segmentation", "Dynamic Token Pruning in Plain Vision Transformers for Semantic\n Segmentation", "Dynamic Token-Pass Transformers for Semantic Segmentation"], "answer_arxiv_id": ["2311.00586v1", "2308.01045", "2308.01944"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_10071"} +{"question": "Could you provide me some studies about the application of LoRA in the fine-tuning of models?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_10072"} +{"question": "Could you provide me some research that focused on which parameters or tasks is better to share in multi-task learning?", "answer": ["Learning to Branch for Multi-Task Learning", "AdaShare: Learning What To Share For Efficient Deep Multi-Task Learning"], "answer_arxiv_id": ["2006.01895", "1911.12423"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_10073"} +{"question": "Could you mention some literature that explores photography-related dimensions in the context of low-level visual aspects?", "answer": ["Exploring Video Quality Assessment on User Generated Contents from\n Aesthetic and Technical Perspectives", "Exploring Opinion-unaware Video Quality Assessment with Semantic\n Affinity Criterion", "Towards Robust Text-Prompted Semantic Criterion for In-the-Wild Video\n Quality Assessment", "Photo Aesthetics Ranking Network with Attributes and Content Adaptation"], "answer_arxiv_id": ["2211.04894", "2302.13269", "2304.14672", "1606.01621"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_10074"} +{"question": "What research demonstrated learning an inverse dynamics model from environment interactions?", "answer": ["Behavioral Cloning from Observation", "Zero-Shot Visual Imitation"], "answer_arxiv_id": ["1805.01954", "1804.08606"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_10075"} +{"question": "What works proposed Domain Generalization (DG) methods focusing on data manipulation such as data augmentation and data generation?", "answer": ["Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Training Deep Networks with Synthetic Data: Bridging the Reality Gap by Domain Randomization", "A Unified Feature Disentangler for Multi-Domain Image Translation and Manipulation", "Learning to Learn Single Domain Generalization"], "answer_arxiv_id": ["1703.06907", "1804.06516", "1809.01361", "2003.13216"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_10076"} +{"question": "Are there any studies about teaching language models to use tools?", "answer": ["Tool Documentation Enables Zero-Shot Tool-Usage with Large Language\n Models", "TALM: Tool Augmented Language Models", "Toolformer: Language Models Can Teach Themselves to Use Tools", "ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs", "ART: Automatic multi-step reasoning and tool-use for large language models", "Augmented Language Models: a Survey"], "answer_arxiv_id": ["2308.00675", "2205.12255", "2302.04761", "2307.16789", "2303.09014v1", "2302.07842"], "source_meta": {"published_time": "20240705"}, "qid": "AutoScholarQuery_train_10077"} +{"question": "Can you cite some studies that optimize dataset distillation using NTK-based ridge regression?", "answer": ["Dataset Distillation with Infinitely Wide Convolutional Networks"], "answer_arxiv_id": ["2107.13034"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_10078"} +{"question": "Could you provide me some research about the use of recurrent networks including autoregressive models in 3D scene synthesis?", "answer": ["GRAINS: Generative Recursive Autoencoders for INdoor Scenes", "ATISS: Autoregressive Transformers for Indoor Scene Synthesis", "Learning 3D Scene Priors with 2D Supervision", "Fast and Flexible Indoor Scene Synthesis via Deep Convolutional\n Generative Models", "SceneFormer: Indoor Scene Generation with Transformers"], "answer_arxiv_id": ["1807.09193", "2110.03675v1", "2211.14157", "1811.12463", "2012.09793"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_10079"} +{"question": "What papers have explored NeRF-based calibration methods?", "answer": ["MOISST: Multimodal Optimization of Implicit Scene for SpatioTemporal\n calibration", "INF: Implicit Neural Fusion for LiDAR and Camera", "Self-Aligning Depth-regularized Radiance Fields for Asynchronous RGB-D\n Sequences"], "answer_arxiv_id": ["2303.03056", "2308.14414", "2211.07459"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_10080"} +{"question": "Could you provide me some papers that develop deep learning-based BOIQA methods?", "answer": ["Blind Omnidirectional Image Quality Assessment with Viewport Oriented Graph Convolutional Networks", "ST360IQ: No-Reference Omnidirectional Image Quality Assessment with Spherical Vision Transformers", "Spatial Attention-based Non-reference Perceptual Quality Prediction Network for Omnidirectional Images"], "answer_arxiv_id": ["2002.09140", "2303.06907", "2103.06116"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_10081"} +{"question": "Can you tell me studies dealing with node selection in the search tree or cutting plane management?", "answer": ["Learning to Cut by Looking Ahead: Cutting Plane Selection via Imitation Learning", "Learning to Select Cuts for Efficient Mixed-Integer Programming", "Reinforcement Learning for Integer Programming: Learning to Cut", "Learning to Use Local Cuts", "Adaptive Cut Selection in Mixed-Integer Linear Programming"], "answer_arxiv_id": ["2206.13414", "2105.13645", "1906.04859", "2206.11618", "2202.10962"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_10082"} +{"question": "What works propose different ways to separate modality-general and modality-specific information to boost multimodal network performance and enhance its generalization ability?", "answer": ["Learning Common and Specific Features for RGB-D Semantic Segmentation with Deconvolutional Networks", "MISA: Modality-Invariant and -Specific Representations for Multimodal Sentiment Analysis"], "answer_arxiv_id": ["1608.01082", "2005.03545"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_10083"} +{"question": "Could you provide some instances of works that treat discrete MDPs and come with provable bounds on the performance?", "answer": ["Safe Policy Improvement with Soft Baseline Bootstrapping"], "answer_arxiv_id": ["1907.05079"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_10084"} +{"question": "Could you name the studies that refined estimates through local optimization on the manifold?", "answer": ["Certifiable Relative Pose Estimation", "Fast and Robust Certifiable Estimation of the Relative Pose Between Two\n Calibrated Cameras"], "answer_arxiv_id": ["2003.13732", "2101.08524"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_10085"} +{"question": "What papers discuss employing RGB image with depth information to enhance anomaly detection performance?", "answer": ["Asymmetric Student-Teacher Networks for Industrial Anomaly Detection"], "answer_arxiv_id": ["2210.07829"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_10086"} +{"question": "What studies employed Graph Neural Networks (GNNs) for feature extraction in object detection with event cameras?", "answer": ["AEGNN: Asynchronous Event-based Graph Neural Networks", "Event-based Asynchronous Sparse Convolutional Networks", "Pushing the Limits of Asynchronous Graph-based Object Detection with\n Event Cameras"], "answer_arxiv_id": ["2203.17149", "2003.09148", "2211.12324"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10087"} +{"question": "Could you provide me some theoretical works which establish properties like convergence to the global optimum?", "answer": ["A Mean Field View of the Landscape of Two-Layer Neural Networks", "On Connected Sublevel Sets in Deep Learning", "Mean Field Analysis of Neural Networks: A Central Limit Theorem", "Mean Field Analysis of Neural Networks: A Law of Large Numbers", "A Rigorous Framework for the Mean Field Limit of Multilayer Neural Networks"], "answer_arxiv_id": ["1804.06561", "1901.07417", "1808.09372", "1805.01053", "2001.11443v3"], "source_meta": {"published_time": "20210830"}, "qid": "AutoScholarQuery_train_10088"} +{"question": "Could you provide me studies about the requirement for generating plausible and diverse samples in VQ-VAEs?", "answer": ["VQ-AR: Vector Quantized Autoregressive Probabilistic Time Series Forecasting", "Generating Diverse High-Fidelity Images with VQ-VAE-2", "The challenge of realistic music generation: modelling raw audio at scale", "Jukebox: A Generative Model for Music"], "answer_arxiv_id": ["2205.15894", "1906.00446", "1806.10474", "2005.00341"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_10089"} +{"question": "Could you provide me with references that detail pre-training strategies in machine learning for improving reasoning ability?", "answer": ["ReasonBERT: Pre-trained to Reason with Distant Supervision", "Logic-Guided Data Augmentation and Regularization for Consistent\n Question Answering", "Reasoning Like Program Executors"], "answer_arxiv_id": ["2109.04912", "2004.10157", "2201.11473"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_10090"} +{"question": "Could you provide me some studies about factorization in MARL?", "answer": ["QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning", "Weighted QMIX: Expanding Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning", "QPLEX: Duplex Dueling Multi-Agent Q-Learning"], "answer_arxiv_id": ["1803.11485", "2006.10800", "2008.01062"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_10091"} +{"question": "Which works utilize mask modeling in 3D pre-training?", "answer": ["Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point\n Modeling", "Masked Discrimination for Self-Supervised Learning on Point Clouds", "PointMamba: A Simple State Space Model for Point Cloud Analysis"], "answer_arxiv_id": ["2111.14819", "2203.11183", "2402.10739"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_10092"} +{"question": "Which works have explored optimizing visual explanations using gradient-based methods?", "answer": ["Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "Weakly-supervised Visual Grounding of Phrases with Linguistic Structures", "Improving Visual Grounding by Encouraging Consistent Gradient-based\n Explanations", "Multi-level Multimodal Common Semantic Space for Image-Phrase Grounding"], "answer_arxiv_id": ["2107.07651", "1705.01371", "2206.15462", "1811.11683"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_10093"} +{"question": "Are there any works that utilized the label propagation to generate pseudo-labels, similar to ALT-OPT?", "answer": ["Label Propagation for Deep Semi-supervised Learning"], "answer_arxiv_id": ["1904.04717"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_10094"} +{"question": "Which works have significantly contributed to the advancement of photorealistic text-to-image generation using diffusion models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers"], "answer_arxiv_id": ["2205.11487", "2211.01324"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_10095"} +{"question": "What papers are about gradient-based approaches in feature attribution methods?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "Learning Important Features Through Propagating Activation Differences", "Axiomatic Attribution for Deep Networks", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "Model Doctor: A Simple Gradient Aggregation Strategy for Diagnosing and Treating CNN Classifiers"], "answer_arxiv_id": ["1312.6034", "1704.02685", "1703.01365", "1610.02391", "2112.04934"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_10096"} +{"question": "Which works exemplify handcrafted planning algorithms incorporated into RL algorithms?", "answer": ["Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model", "Value Prediction Network", "TreeQN and ATreeC: Differentiable Tree-Structured Models for Deep Reinforcement Learning", "Imagination-Augmented Agents for Deep Reinforcement Learning"], "answer_arxiv_id": ["1911.08265", "1707.03497", "1710.11417", "1707.06203"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_10097"} +{"question": "What is the original research on Variational Information Bottleneck?", "answer": ["Deep Variational information bottleneck"], "answer_arxiv_id": ["1612.00410"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_10098"} +{"question": "Which works show that certain ball indicator functions are not approximatable by two-layer networks, yet learnable via GD on a special variant of a three-layer network?", "answer": ["Optimization-Based Separations for Neural Networks"], "answer_arxiv_id": ["2112.02393"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_10099"} +{"question": "Are there research papers where nonlinear CKA was used?", "answer": ["Similarity of Neural Network Representations Revisited", "On the Origins of the Block Structure Phenomenon in Neural Network Representations"], "answer_arxiv_id": ["1905.00414", "2202.07184"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_10100"} +{"question": "What is the first work that introduced a convolutional neural network to compute matching cost and predict disparity maps in stereo matching networks?", "answer": ["Computing the Stereo Matching Cost with a Convolutional Neural Network"], "answer_arxiv_id": ["1409.4326"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_10101"} +{"question": "What works generalized the well-known Johnson-Lindenstrauss lemma to show that random projections of data into certain dimensions preserves the k-means objective?", "answer": ["Performance of Johnson-Lindenstrauss Transform for k-Means and k-Medians Clustering"], "answer_arxiv_id": ["1811.03195"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_10102"} +{"question": "Could you provide some references that dealt with video editing through text-to-image model fine-tuning?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation"], "answer_arxiv_id": ["2212.11565"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_10103"} +{"question": "What works have designed a fully recurrent convolutional neural network for speech separation?", "answer": ["Speech Separation Using an Asynchronous Fully Recurrent Convolutional Neural Network"], "answer_arxiv_id": ["2112.02321"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_10104"} +{"question": "Which research proposes architectures of higher-order GNNs, specifically addressing k-IGN?", "answer": ["On the Universality of Invariant Networks"], "answer_arxiv_id": ["1901.09342"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_10105"} +{"question": "Which studies focus on defending against Federated Learning backdoor attacks?", "answer": ["Can You Really Backdoor Federated Learning?", "Defending against Backdoors in Federated Learning with Robust Learning Rate", "BaFFLe: Backdoor Detection via Feedback-based Federated Learning", "FLTrust: Byzantine-robust Federated Learning via Trust Bootstrapping"], "answer_arxiv_id": ["1911.07963", "2007.03767", "2011.02167", "2012.13995"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_10106"} +{"question": "What studies represent affordances as contact maps for grasping?", "answer": ["Learning Task-Oriented Grasping from Human Activity Datasets", "Hand-Object Contact Consistency Reasoning for Human Grasps Generation", "GenDexGrasp: Generalizable Dexterous Grasping", "Learning Generalizable Dexterous Manipulation from Human Grasp\n Affordance", "Grasp Multiple Objects with One Hand"], "answer_arxiv_id": ["1910.11669", "2104.03304", "2210.00722", "2204.02320", "2310.15599"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_10107"} +{"question": "Which studies present the abundance of adversarial examples in trained networks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Adversarial vulnerability for any classifier", "Are adversarial examples inevitable?", "Adversarially Robust Generalization Requires More Data", "On the Geometry of Adversarial Examples", "Adversarial examples from computational constraints", "Feature Purification: How Adversarial Training Performs Robust Deep Learning", "High-frequency Component Helps Explain the Generalization of Convolutional Neural Networks", "The Pitfalls of Simplicity Bias in Neural Networks", "The Dimpled Manifold Model of Adversarial Examples in Machine Learning", "Shift Invariance Can Reduce Adversarial Robustness", "Adversarial Robustness is at Odds with Lazy Training", "On the (Non-)Robustness of Two-Layer Neural Networks in Different Learning Regimes"], "answer_arxiv_id": ["1412.6572", "1802.08686", "1809.02104", "1804.11285", "1811.00525", "1805.10204", "2005.10190", "1905.13545", "2006.07710", "2106.10151", "2103.02695", "2207.00411", "2203.11864"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_10108"} +{"question": "In what papers the researchers apply another approach which adapts VAE and autoencoder to diffusion models for controllability?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["2112.10752", "2106.05931"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_10109"} +{"question": "Which papers made significant achievements using CLIP-based approaches in cross-modal retrieval?", "answer": ["CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "CenterCLIP: Token Clustering for Efficient Text-Video Retrieval", "X-Pool: Cross-Modal Language-Video Attention for Text-Video Retrieval"], "answer_arxiv_id": ["2106.11097", "2205.00823", "2203.15086"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_10110"} +{"question": "Are there any works on methods of adversarial memorization detection?", "answer": ["The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks"], "answer_arxiv_id": ["1802.08232"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_10111"} +{"question": "What research applies the Realistic Image Pair Rendering (RIPR) in RealFlow?", "answer": ["RealFlow: EM-based Realistic Optical Flow Dataset Generation from Videos"], "answer_arxiv_id": ["2207.11075"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_10112"} +{"question": "Can you name some studies that further develop the use of k-nearest neighbors in re-ranking?", "answer": ["Re-ranking Person Re-identification with k-reciprocal Encoding", "Divide and Fuse: A Re-ranking Approach for Person Re-identification", "A Pose-Sensitive Embedding for Person Re-Identification with Expanded\n Cross Neighborhood Re-Ranking"], "answer_arxiv_id": ["1701.08398", "1708.04169", "1711.10378"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_10113"} +{"question": "Which studies were conducted on 3D diffusion models predominantly concentrated on geometry generation?", "answer": ["Diffusion Probabilistic Models for 3D Point Cloud Generation", "HyperDiffusion: Generating Implicit Neural Fields with Weight-Space\n Diffusion"], "answer_arxiv_id": ["2103.01458", "2303.17015"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_10114"} +{"question": "Which algorithms are notable for adaptively tuning the step size when training neural networks?", "answer": ["ADADELTA: An Adaptive Learning Rate Method"], "answer_arxiv_id": ["1212.5701"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_10115"} +{"question": "What papers provide approaches to test-time adaptation by minimizing the entropy of the model output?", "answer": ["Tent: Fully Test-Time Adaptation by Entropy Minimization", "Test-Time Adaptation via Conjugate Pseudo-labels"], "answer_arxiv_id": ["2006.10726", "2207.09640"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_10116"} +{"question": "Could you provide me studies that presented enhanced alignment between images and text using large pre-trained language models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2205.11487"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_10117"} +{"question": "Which work applies a Transformer architecture on image patches?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_10118"} +{"question": "What are the key studies which characterize SGD under individual smoothness and unbiased function values?", "answer": ["Stochastic Gradient Descent for Nonconvex Learning without Bounded Gradient Assumptions"], "answer_arxiv_id": ["1902.00908"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_10119"} +{"question": "Could you provide me some studies dealing with labeling-trick based methods?", "answer": ["Link Prediction Based on Graph Neural Networks", "Labeling Trick: A Theory of Using Graph Neural Networks for Multi-Node Representation Learning"], "answer_arxiv_id": ["1802.09691", "2010.16103"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10120"} +{"question": "Which research papers proposed early methods for semantic image synthesis?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "High-Resolution Image Synthesis and Semantic Manipulation with\n Conditional GANs", "Semantic Image Synthesis with Spatially-Adaptive Normalization"], "answer_arxiv_id": ["1611.07004", "1711.11585", "1903.07291"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_10121"} +{"question": "Which research papers proposed or studied notable large language models (LLMs) like GPT-3, Claude, GPT-4, and PaLM2?", "answer": ["Language Models are Few-Shot Learners", "GPT-4 Technical Report", "PaLM 2 Technical Report"], "answer_arxiv_id": ["2005.14165", "2303.08774", "2305.10403"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_10122"} +{"question": "What are the studies that conducted deliberate tuning for designing safety in conversational language models?", "answer": ["Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback", "GPT-4 Technical Report"], "answer_arxiv_id": ["2204.05862", "2303.08774"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_10123"} +{"question": "What works contributed to the research about learning parameter initializations that quickly adapt to downstream tasks?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["1703.03400"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_10124"} +{"question": "Which papers discuss obtaining bounds on the test error when using gradient descent in the near-initialization regime?", "answer": ["Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Data", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "Generalization Bounds of Stochastic Gradient Descent for Wide and Deep Neural Networks", "Gradient Descent can Learn Less Over-parameterized Two-layer Neural Networks on Classification Problems", "Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networks", "How Much Over-parameterization Is Sufficient to Learn Deep ReLU Networks?", "Early-stopped neural networks are consistent"], "answer_arxiv_id": ["1808.01204", "1901.08584", "1905.13210", "1905.09870", "1909.12292", "1911.12360", "2106.05932"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_10125"} +{"question": "What studies extensively applied the Reinforcement Learning with Human Feedback (RLHF) technique on Large Language Models alignment?", "answer": ["Reliability and Learnability of Human Bandit Feedback for\n Sequence-to-Sequence Reinforcement Learning", "WebGPT: Browser-assisted question-answering with human feedback", "Training language models to follow instructions with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback", "Constitutional AI: Harmlessness from AI Feedback"], "answer_arxiv_id": ["1805.10627", "2112.09332", "2203.02155", "2204.05862", "2212.08073"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_10126"} +{"question": "Which works introduced cycle consistency in image editing and translation?", "answer": ["Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks"], "answer_arxiv_id": ["1703.10593"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_10127"} +{"question": "Where can I find a work that suggests a direct imposition of a Gaussian prior on the proxy model and model adaptation for robust estimation on the specific set of candidate design space?", "answer": ["RoMA: Robust Model Adaptation for Offline Model-based Optimization"], "answer_arxiv_id": ["2110.14188"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_10128"} +{"question": "Are there any research papers focusing on generating the valid proof to analyze and interpret the reasoning process of language models?", "answer": ["Explaining Answers with Entailment Trees", "ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language"], "answer_arxiv_id": ["2104.08661", "2012.13048"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_10129"} +{"question": "What works propose strategies to reduce the size of the input to the Transformer in image settings?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Visual Transformers: Token-based Image Representation and Processing for Computer Vision"], "answer_arxiv_id": ["2010.11929", "2006.03677"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_10130"} +{"question": "Which work introduced understanding and interpreting model decisions into the form of Concept Activation Vectors (CAVs)?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with\n Concept Activation Vectors (TCAV)"], "answer_arxiv_id": ["1711.11279"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_10131"} +{"question": "What papers simplified the 'prefix tuning' technique and established the standard soft 'prompt-tuning'?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2104.08691"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_10132"} +{"question": "Which research demonstrated a close connection to TD auxiliary tasks?", "answer": ["Understanding Self-Predictive Learning for Reinforcement Learning"], "answer_arxiv_id": ["2212.03319"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10133"} +{"question": "What research explored sign back-translation to construct pseudo-parallel training data for SLT?", "answer": ["Improving Sign Language Translation with Monolingual Data by Sign Back-Translation"], "answer_arxiv_id": ["2105.12397"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_10134"} +{"question": "What works used the technique of self-instruct to collect SFT data?", "answer": ["Self-Instruct: Aligning Language Models with Self-Generated Instructions"], "answer_arxiv_id": ["2212.10560"], "source_meta": {"published_time": "20240207"}, "qid": "AutoScholarQuery_train_10135"} +{"question": "What paper emphasizes that packing relevant documents can enhance language models’ in-context learning and context utilization?", "answer": ["In-context Pretraining: Language Modeling Beyond Document Boundaries"], "answer_arxiv_id": ["2310.10638"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_10136"} +{"question": "What work recently derived finite-sample bounds for distributional off-policy evaluation with maximum likelihood estimation (MLE)?", "answer": ["Distributional Offline Policy Evaluation with Predictive Error Guarantees"], "answer_arxiv_id": ["2302.09456"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10137"} +{"question": "Which studies talk about the gradient-based optimization problems for the standard Lipschitz smooth functions?", "answer": ["Acceleration Methods"], "answer_arxiv_id": ["2101.09545"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_10138"} +{"question": "Could you provide some studies that have worked on leveraging the capability of CLIP and proposed models like ViLD and RegionCLIP?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation", "RegionCLIP: Region-based Language-Image Pretraining"], "answer_arxiv_id": ["2104.13921", "2112.09106"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10139"} +{"question": "What research papers point to the importance of visual information in augmenting our understanding of the world?", "answer": ["Expectation-Maximization Contrastive Learning for Compact\n Video-and-Language Representations", "DiffusionRet: Generative Text-Video Retrieval with Diffusion Model", "Video-Text as Game Players: Hierarchical Banzhaf Interaction for\n Cross-Modal Representation Learning", "Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval", "OmniVL:One Foundation Model for Image-Language and Video-Language Tasks", "Text-Video Retrieval with Disentangled Conceptualization and Set-to-Set\n Alignment"], "answer_arxiv_id": ["2211.11427", "2303.09867", "2303.14369", "2104.00650", "2209.07526", "2305.12218"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_10140"} +{"question": "Which works are related to instance segmentation in an unsupervised manner?", "answer": ["FreeSOLO: Learning to Segment Objects without Annotations", "Cut and Learn for Unsupervised Object Detection and Instance\n Segmentation"], "answer_arxiv_id": ["2202.12181", "2301.11320"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_10141"} +{"question": "What studies propose using a monolithic encoder for predicting the position and scale of objects?", "answer": ["Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects", "SCALOR: Generative World Models with Scalable Object Representations"], "answer_arxiv_id": ["1603.08575", "1806.01794", "1910.02384"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_10142"} +{"question": "What research papers have been sparked by the release of MVTec 3D-AD?", "answer": ["Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection", "Asymmetric Student-Teacher Networks for Industrial Anomaly Detection", "Collaborative Discrepancy Optimization for Reliable Image Anomaly Localization", "Multimodal Industrial Anomaly Detection via Hybrid Fusion", "EasyNet: An Easy Network for 3D Industrial Anomaly Detection"], "answer_arxiv_id": ["2203.05550", "2210.07829", "2302.08769v1", "2303.00601", "2307.13925"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_10143"} +{"question": "Which research proposed the concept of Bilinear Classes?", "answer": ["Bilinear Classes: A Structural Framework for Provable Generalization in RL"], "answer_arxiv_id": ["2103.10897"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_10144"} +{"question": "Could you provide me some works about fine-grained reconstructing accurate human-object interactions?", "answer": ["Grasping Field: Learning Implicit Representations for Human Grasps", "Reconstructing Hand-Object Interactions in the Wild", "3D Human Pose Estimation via Intuitive Physics", "NIFTY: Neural Object Interaction Fields for Guided Human Motion\n Synthesis", "InterDiff: Generating 3D Human-Object Interactions with Physics-Informed\n Diffusion", "InterGen: Diffusion-based Multi-human Motion Generation under Complex\n Interactions", "CG-HOI: Contact-Guided 3D Human-Object Interaction Generation", "HOI-Diff: Text-Driven Synthesis of 3D Human-Object Interactions using\n Diffusion Models"], "answer_arxiv_id": ["2008.04451", "2012.09856", "2303.18246", "2307.07511", "2308.16905", "2304.05684", "2311.16097", "2312.06553"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_10145"} +{"question": "Which papers propose matrix or tensor factorization methods for link prediction?", "answer": ["Temporal Link Prediction using Matrix and Tensor Factorizations"], "answer_arxiv_id": ["1005.4006"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_10146"} +{"question": "Which research works have studied scenarios where rewards from a subset of data can be absent?", "answer": ["How to Leverage Unlabeled Data in Offline Reinforcement Learning", "COG: Connecting New Skills to Past Experience with Offline Reinforcement Learning"], "answer_arxiv_id": ["2202.01741", "2010.14500"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_10147"} +{"question": "Which work discusses the improvement of personalization of Federated Learning via Reptile?", "answer": ["Improving Federated Learning Personalization via Model Agnostic Meta Learning"], "answer_arxiv_id": ["1909.12488"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_10148"} +{"question": "What studies brought significant progress in the field of content-based image retrieval?", "answer": ["Learning Fine-grained Image Similarity with Deep Ranking", "Deep Image Retrieval: Learning global representations for image search", "CNN Image Retrieval Learns from BoW: Unsupervised Fine-Tuning with Hard\n Examples"], "answer_arxiv_id": ["1404.4661", "1604.01325", "1604.02426"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_10149"} +{"question": "Which work applies the Hardness score as regularization during the sampling process in the context of low-density learning?", "answer": ["Generating High Fidelity Data from Low-density Regions using Diffusion Models"], "answer_arxiv_id": ["2203.17260"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_10150"} +{"question": "Could you provide me some studies on LLM-oriented Memory to boost agents’ capabilities?", "answer": ["Reflexion: Language Agents with Verbal Reinforcement Learning", "Voyager: An Open-Ended Embodied Agent with Large Language Models", "CLIN: A Continually Learning Language Agent for Rapid Task Adaptation\n and Generalization", "JARVIS-1: Open-World Multi-task Agents with Memory-Augmented Multimodal\n Language Models"], "answer_arxiv_id": ["2303.11366", "2305.16291", "2310.10134", "2311.05997"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_10151"} +{"question": "Which works propose compositional full-body avatars for expressive control of the human body, hands and face?", "answer": ["X-Avatar: Expressive Human Avatars", "AvatarReX: Real-time Expressive Full-body Avatars"], "answer_arxiv_id": ["2303.04805", "2305.04789"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_10152"} +{"question": "Which references showed that for graphs with certain structures of clusters, spectral clustering with fewer eigenvectors performs better?", "answer": ["A Tighter Analysis of Spectral Clustering, and Beyond"], "answer_arxiv_id": ["2208.01724v1"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_10153"} +{"question": "What study attempted to explicitly separate latent space into semantic representation of an image and its rotation and translation information in disentangled representation learning?", "answer": ["Explicitly disentangling image content from translation and rotation with spatial-VAE"], "answer_arxiv_id": ["1909.11663"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_10154"} +{"question": "Could you provide me some papers that use linear sketching, the second main approach in non-private algorithms in the streaming model?", "answer": ["The Johnson-Lindenstrauss Transform Itself Preserves Differential Privacy", "Local, Private, Efficient Protocols for Succinct Histograms", "Heavy Hitters and the Structure of Local Privacy", "Practical Locally Private Heavy Hitters", "Frequency Estimation Under Multiparty Differential Privacy: One-shot and Streaming"], "answer_arxiv_id": ["1204.2136", "1504.04686", "1711.04740", "1707.04982", "2104.01808"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_10155"} +{"question": "Are there works discussing representation similarity measures based on fixed points topology of internal dynamics in recurrent neural networks?", "answer": ["Universality and individuality in neural dynamics across large populations of recurrent networks"], "answer_arxiv_id": ["1907.08549"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_10156"} +{"question": "Could you provide me some studies about neural network-enhanced latent factor models?", "answer": ["Neural Collaborative Filtering", "Personalized Top-N Sequential Recommendation via Convolutional Sequence\n Embedding", "Session-based Recommendations with Recurrent Neural Networks", "Improving Recommendation Fairness via Data Augmentation"], "answer_arxiv_id": ["1708.05031", "1809.07426", "1511.06939", "2302.06333"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_10157"} +{"question": "Could you provide me some researches about real-time perception and video analytics focusing on accuracy-latency decisions?", "answer": ["Speed/accuracy trade-offs for modern convolutional object detectors", "Towards Streaming Perception"], "answer_arxiv_id": ["1611.10012", "2005.10420"], "source_meta": {"published_time": "20210610"}, "qid": "AutoScholarQuery_train_10158"} +{"question": "Which works conducted research on modern extensions of canonical correlation analysis using kernel methods and deep learning?", "answer": ["Deep Variational Canonical Correlation Analysis"], "answer_arxiv_id": ["1610.03454v3"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_10159"} +{"question": "Which research showed suboptimal log-cubic scaling in cases where the approximation space was either a linear subspace or a nonlinear space of sparse expansions?", "answer": ["Convergence bounds for empirical nonlinear least-squares"], "answer_arxiv_id": ["2001.00639"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10160"} +{"question": "What works introduced techniques or methods for relighting faces and bodies?", "answer": ["Single Image Portrait Relighting", "Neural Light Transport for Relighting and View Synthesis"], "answer_arxiv_id": ["1905.00824", "2008.03806"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_10161"} +{"question": "Which research works refer to training models with limited labelled data for a task, also known as few-shot learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Matching Networks for One Shot Learning"], "answer_arxiv_id": ["1703.03400", "1606.04080"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_10162"} +{"question": "Which research methods use a triplet loss for contrastive learning?", "answer": ["FaceNet: A Unified Embedding for Face Recognition and Clustering"], "answer_arxiv_id": ["1503.03832"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_10163"} +{"question": "Which works extended the attention module of SD to conduct cross-frame attention for global appearance consistency?", "answer": ["Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "ControlVideo: Training-free Controllable Text-to-Video Generation", "TokenFlow: Consistent Diffusion Features for Consistent Video Editing"], "answer_arxiv_id": ["2303.13439", "2305.13077", "2307.10373"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_10164"} +{"question": "What sources talk about the prevalence of zero empirical risk in modern deep learning?", "answer": ["Understanding deep learning requires rethinking generalization"], "answer_arxiv_id": ["1611.03530v2"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_10165"} +{"question": "Which research papers discuss methods to uncover interpretable causal mechanisms in deep learning models?", "answer": ["Neural Natural Language Inference Models Partially Embed Theories of Lexical Entailment and Negation", "Causal Abstractions of Neural Networks", "Causal Proxy Models for Concept-based Model Explanations", "Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small"], "answer_arxiv_id": ["2004.14623", "2106.02997", "2209.14279", "2211.00593"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_10166"} +{"question": "Which papers discussed the role of benchmarks in driving advancements in AI?", "answer": ["SQuAD: 100,000+ Questions for Machine Comprehension of Text", "Utility is in the Eye of the User: A Critique of NLP Leaderboards", "AI and the Everything in the Whole Wide World Benchmark"], "answer_arxiv_id": ["1606.05250", "2009.13888", "2111.15366"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_10167"} +{"question": "Which papers proposed methods for recovering material properties from one image, entirely relying on synthetic data for training?", "answer": ["Modeling Surface Appearance from a Single Photograph using\n Self-augmented Convolutional Neural Networks", "Single-Image SVBRDF Capture with a Rendering-Aware Deep Network", "SurfaceNet: Adversarial SVBRDF Estimation from a Single Image", "ControlMat: A Controlled Generative Approach to Material Capture"], "answer_arxiv_id": ["1809.00886", "1810.09718", "2107.11298", "2309.01700"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_10168"} +{"question": "What works investigated equivariant Transformer and message-passing architectures?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "E(n) Equivariant Graph Neural Networks", "Attentive Group Equivariant Convolutional Networks", "Geometric and Physical Quantities improve E(3) Equivariant Message Passing", "MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields", "E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials", "So3krates: Equivariant attention for interactions on arbitrary length-scales in molecular systems"], "answer_arxiv_id": ["1802.08219", "2006.10503", "2102.09844", "2002.03830", "2110.02905", "2206.07697", "2101.03164", "2205.14276"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_10169"} +{"question": "What neural methods for texture generation have been proposed so far?", "answer": ["Texture Synthesis Using Convolutional Neural Networks", "A Sliced Wasserstein Loss for Neural Texture Synthesis", "Texture Networks: Feed-forward Synthesis of Textures and Stylized Images", "Precomputed Real-Time Texture Synthesis with Markovian Generative\n Adversarial Networks", "Texture Synthesis with Spatial Generative Adversarial Networks", "Learning Texture Manifolds with the Periodic Spatial GAN", "TileGAN: Synthesis of Large-Scale Non-Homogeneous Textures", "Structural-analogy from a Single Image Pair", "SeamlessGAN: Self-Supervised Synthesis of Tileable Texture Maps", "Non-Stationary Texture Synthesis by Adversarial Expansion", "SinGAN: Learning a Generative Model from a Single Natural Image"], "answer_arxiv_id": ["1505.07376", "2006.07229", "1603.03417", "1604.04382", "1611.08207", "1705.06566", "1904.12795", "2004.02222", "2201.05120", "1805.04487", "1905.01164"], "source_meta": {"published_time": "20240105"}, "qid": "AutoScholarQuery_train_10170"} +{"question": "Which studies for instruction tuning of Multimodal LLMs were inspired by the recent success of instruction tuning on LLMs?", "answer": ["MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Valley: Video Assistant with Large Language model Enhanced abilitY", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "VideoChat: Chat-Centric Video Understanding", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models"], "answer_arxiv_id": ["2304.10592", "2306.07207", "2305.06500", "2305.06355", "2306.02858", "2306.05424"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10171"} +{"question": "Could you tell me the works that adopt a time-dependent backward deformation map to query the canonical frame in perception of dynamic scenes?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Video", "Nerfies: Deformable Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2012.12247", "2011.12948"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_10172"} +{"question": "What studies proposed the approach to offload partitioned model states and tensors to CPU memory or NVMe to fully utilize heterogeneous architecture?", "answer": ["ZeRO-Infinity: Breaking the GPU Memory Wall for Extreme Scale Deep\n Learning"], "answer_arxiv_id": ["2104.07857"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10173"} +{"question": "Which research papers proposed using Test-Time Training methods in Test-Time adaptive methods?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts"], "answer_arxiv_id": ["1909.13231"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_10174"} +{"question": "What research works introduced the use of prompts for parameter efficiency?", "answer": ["Visual Prompt Tuning", "Fine-tuning Image Transformers using Learnable Memory", "The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2203.12119", "2203.15243", "2104.08691"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_10175"} +{"question": "Which papers represent the arrangement of objects in scene graph for indoor scene synthesis?", "answer": ["Graph-to-3D: End-to-End Generation and Manipulation of 3D Scenes Using\n Scene Graphs", "SceneGraphNet: Neural Message Passing for 3D Indoor Scene Augmentation"], "answer_arxiv_id": ["2108.08841", "1907.11308"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_10176"} +{"question": "Which works propose compression methods to address efficient communications in distributed algorithms?", "answer": ["QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "Distributed Learning with Compressed Gradient Differences", "On Biased Compression for Distributed Learning", "The Convergence of Sparsified Gradient Methods", "Sparsified SGD with Memory", "EF21: A New, Simpler, Theoretically Better, and Practically Faster Error Feedback", "Distributed Methods with Compressed Communication for Solving Variational Inequalities, with Theoretical Guarantees"], "answer_arxiv_id": ["1610.02132", "1901.09269v3", "2002.12410", "1809.10505", "1809.07599", "2106.05203", "2110.03313v3"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_10177"} +{"question": "Which papers explored the idea of speeding up training through a warm start by reusing parameters of an existing model?", "answer": ["Dota 2 with Large Scale Deep Reinforcement Learning", "Net2Net: Accelerating Learning via Knowledge Transfer", "Scaling Language Models: Methods, Analysis & Insights from Training Gopher"], "answer_arxiv_id": ["1912.06680", "1511.05641", "2112.11446"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_10178"} +{"question": "Which works propose model-based policy updates for non-stationary environments by using short rollouts to prevent model exploitation?", "answer": ["When to Trust Your Model: Model-Based Policy Optimization", "Mastering Diverse Domains through World Models"], "answer_arxiv_id": ["1906.08253", "2301.04104"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_10179"} +{"question": "Could you provide me with studies that employed multi-view videos during training to learn view-invariant features?", "answer": ["View-Invariant Probabilistic Embedding for Human Pose"], "answer_arxiv_id": ["1912.01001"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_10180"} +{"question": "Could you point out some research works on data selection in vision domain?", "answer": ["Beyond neural scaling laws: beating power law scaling via data pruning", "Learning From Less Data: A Unified Data Subset Selection and Active Learning Framework for Computer Vision", "Glister: Generalization based Data Subset Selection for Efficient and Robust Learning", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training", "RETRIEVE: Coreset Selection for Efficient and Robust Semi-Supervised Learning", "Optimizing Data Usage via Differentiable Rewards", "Deep Learning on a Data Diet: Finding Important Examples Early in Training", "Coresets for Data-efficient Training of Machine Learning Models", "Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["2206.14486v6", "1901.01151", "2012.10630", "2103.00123", "2106.07760v2", "1911.10088", "2107.07075", "1906.01827", "1708.00489"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_10181"} +{"question": "Which papers discussed different formulations of Graph Neural Networks?", "answer": ["Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering", "Neural Message Passing for Quantum Chemistry", "Invariant and Equivariant Graph Networks"], "answer_arxiv_id": ["1606.09375", "1704.01212", "1812.09902"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_10182"} +{"question": "Which papers have investigated the potential of using LLMs for automatic reference-free evaluation of generated text?", "answer": ["Investigating Table-to-Text Generation Capabilities of LLMs in\n Real-World Information Seeking Scenarios", "Evaluation Metrics in the Era of GPT-4: Reliably Evaluating Large\n Language Models on Sequence to Sequence Tasks", "GEMBA-MQM: Detecting Translation Quality Error Spans with GPT-4", "Large Language Models Are State-of-the-Art Evaluators of Translation\n Quality", "Is ChatGPT a Good NLG Evaluator? A Preliminary Study", "GPTScore: Evaluate as You Desire"], "answer_arxiv_id": ["2305.14987", "2310.13800", "2310.13988", "2302.14520", "2303.04048", "2302.04166"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_10183"} +{"question": "What are some significant works about atomistic systems that include GP models?", "answer": ["DScribe: Library of Descriptors for Machine Learning in Materials Science", "On representing chemical environments"], "answer_arxiv_id": ["1904.08875", "1209.3140"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_10184"} +{"question": "What works proposed the idea of reparameterizing network weights into task weights and a symmetry matrix for symmetry discovery?", "answer": ["Meta-learning Symmetries by Reparameterization"], "answer_arxiv_id": ["2007.02933"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_10185"} +{"question": "Which works have employed Generative Adversarial Networks (GANs) for image generation in video generation tasks?", "answer": ["Generative Adversarial Networks", "Improved Techniques for Training GANs", "Towards Principled Methods for Training Generative Adversarial Networks", "Wasserstein GAN"], "answer_arxiv_id": ["2203.00667", "1606.03498", "1701.04862", "1701.07875"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_10186"} +{"question": "Could you provide me some studies that focus on improving robustness by altering the training procedures?", "answer": ["EPOpt: Learning Robust Neural Network Policies Using Model Ensembles", "Deep Variational information bottleneck", "Dynamics Generalization via Information Bottleneck in Deep Reinforcement Learning", "Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneck", "Robust Predictable Control"], "answer_arxiv_id": ["1610.01283", "1612.00410", "2008.00614", "1910.12911", "2109.03214"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_10187"} +{"question": "Which works focus on capturing differences in texture between real and generated images in the context of artifact detection?", "answer": ["Attributing Fake Images to GANs: Learning and Analyzing GAN Fingerprints", "Global Texture Enhancement for Fake Face Detection in the Wild"], "answer_arxiv_id": ["1811.08180", "2002.00133"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_10188"} +{"question": "In what research articles are the constraints on the usage of center embedding in language discussed?", "answer": ["Depth-bounding is effective: Improvements and evaluation of unsupervised\n PCFG induction"], "answer_arxiv_id": ["1809.03112"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_10189"} +{"question": "Which works talk about simulated V2X datasets, generated by simulators for autonomous driving?", "answer": ["OPV2V: An Open Benchmark Dataset and Fusion Pipeline for Perception with\n Vehicle-to-Vehicle Communication", "V2X-Sim: Multi-Agent Collaborative Perception Dataset and Benchmark for\n Autonomous Driving", "V2X-ViT: Vehicle-to-Everything Cooperative Perception with Vision\n Transformer", "DeepAccident: A Motion and Accident Prediction Benchmark for V2X\n Autonomous Driving"], "answer_arxiv_id": ["2109.07644", "2202.08449", "2203.10638", "2304.01168"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_10190"} +{"question": "Which study proposed a fully differentiable DAG learning algorithm?", "answer": ["Differentiable DAG Sampling"], "answer_arxiv_id": ["2203.08509"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_10191"} +{"question": "Are there works that discussed the gradient boundedness assumption (GBA) in BLO literature?", "answer": ["Approximation Methods for Bilevel Programming", "A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic", "Bilevel Optimization: Convergence Analysis and Enhanced Design", "A Fully Single Loop Algorithm for Bilevel Optimization without Hessian Inverse", "BOME! Bilevel Optimization Made Easy: A Simple First-Order Approach"], "answer_arxiv_id": ["1802.02246", "2007.05170v4", "2010.07962", "2112.04660", "2209.08709"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_10192"} +{"question": "Could you mention the previous approaches for assessing intrinsic compositionality?", "answer": ["Measuring Compositionality in Representation Learning", "RNNs implicitly implement tensor-product representations"], "answer_arxiv_id": ["1902.07181", "1812.08718"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_10193"} +{"question": "Which paper originally proposed the standard transformer architecture?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_10194"} +{"question": "What studies incorporate projection maintenance data structure in the cutting plane method?", "answer": ["An Improved Cutting Plane Method for Convex Optimization, Convex-Concave Games and its Applications"], "answer_arxiv_id": ["2004.04250v1"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_10195"} +{"question": "What works about Image Super-Resolution have been cited in the text?", "answer": ["Enhanced Deep Residual Networks for Single Image Super-Resolution", "Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution", "Residual Dense Network for Image Super-Resolution", "Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure\n Synthetic Data"], "answer_arxiv_id": ["1707.02921", "1704.03915", "1802.08797", "2107.10833"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_10196"} +{"question": "What are some works that have integrated NeRF into the task of synthesizing talking heads and used audio as the driving signal?", "answer": ["AD-NeRF: Audio Driven Neural Radiance Fields for Talking Head Synthesis", "DFA-NeRF: Personalized Talking Head Generation via Disentangled Face\n Attributes Neural Rendering", "Semantic-Aware Implicit Neural Audio-Driven Video Portrait Generation", "GeneFace: Generalized and High-Fidelity Audio-Driven 3D Talking Face\n Synthesis"], "answer_arxiv_id": ["2103.11078", "2201.00791", "2201.07786", "2301.13430"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10197"} +{"question": "What work propsed expert pruning metrics based on gate statistics?", "answer": ["Memory-efficient NLLB-200: Language-specific Expert Pruning of a\n Massively Multilingual Machine Translation Model"], "answer_arxiv_id": ["2212.09811"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_10198"} +{"question": "What research papers used task or language embeddings as contextual information to generate module parameters in hypernetworks?", "answer": ["Parameter-efficient Multi-task Fine-tuning for Transformers via Shared\n Hypernetworks", "Prototype-based HyperAdapter for Sample-Efficient Multi-task Tuning", "UDapter: Language Adaptation for Truly Universal Dependency Parsing", "Multilingual Machine Translation with Hyper-Adapters"], "answer_arxiv_id": ["2106.04489", "2310.11670", "2004.14327", "2205.10835"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_10199"} +{"question": "Which works demonstrate the use of 2D diffusion models in image synthesis?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Improved Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Cascaded Diffusion Models for High Fidelity Image Generation", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456", "2102.09672", "2105.05233", "2106.15282", "2112.10741", "2112.10752", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_10200"} +{"question": "What papers discuss employing neural network architectures with translational and rotational symmetries to preserve momenta?", "answer": ["Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data", "A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups"], "answer_arxiv_id": ["2002.12880", "2104.09459"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_10201"} +{"question": "What work developed batch normalization through time (BNTT) to estimate the distribution of temporal-variant inputs in SNNs?", "answer": ["Revisiting Batch Normalization for Training Low-latency Deep Spiking Neural Networks from Scratch"], "answer_arxiv_id": ["2010.01729"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_10202"} +{"question": "What studies feature the use of Masked Image Modeling (MIM) in pre-training models for downstream tasks?", "answer": ["Unleashing Vanilla Vision Transformer with Masked Image Modeling for\n Object Detection", "Masked Autoencoders Enable Efficient Knowledge Distillers", "Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud\n Pre-training"], "answer_arxiv_id": ["2204.02964", "2208.12256", "2205.14401"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_10203"} +{"question": "Could you provide me some examples of data sets with reasoning-based commonsense inferences?", "answer": ["CICERO: A Dataset for Contextualized Commonsense Inference in Dialogues"], "answer_arxiv_id": ["2203.13926"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_10204"} +{"question": "What studies focus on improving the generalization performance with a flat loss landscape?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization"], "answer_arxiv_id": ["2010.01412"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_10205"} +{"question": "Is there any research on using CLIP for generating 3D avatars?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars"], "answer_arxiv_id": ["2205.08535"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_10206"} +{"question": "What works extended LLMs to multimodal domains?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "MiniGPT-v2: large language model as a unified interface for\n vision-language multi-task learning", "Visual Instruction Tuning", "Planting a SEED of Vision in Large Language Model"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2305.06500", "2304.10592", "2310.09478", "2304.08485", "2307.08041"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_10207"} +{"question": "Could you tell me some existing methods that encourage optimality in a generative model?", "answer": ["Optimal transport mapping via input convex neural networks", "Building Normalizing Flows with Stochastic Interpolants", "How to Train Your Neural ODE: the World of Jacobian and Kinetic Regularization", "Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow"], "answer_arxiv_id": ["1908.10962", "2209.15571", "2002.02798", "2209.03003"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_10208"} +{"question": "Which works utilise scene graph annotations to guide VLMs' learning on compositional relations?", "answer": ["Incorporating Structured Representations into Pretrained Vision &\n Language Models Using Scene Graphs", "Coarse-to-Fine Contrastive Learning in Image-Text-Graph Space for\n Improved Vision-Language Compositionality"], "answer_arxiv_id": ["2305.06343", "2305.13812"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_10209"} +{"question": "What is the work that relates the non-divergence of Gradient Descent to the presence of a forward invariant set?", "answer": ["Understanding the Unstable Convergence of Gradient Descent"], "answer_arxiv_id": ["2204.01050"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_10210"} +{"question": "Could you provide me some works using agent-centric representation?", "answer": ["Wayformer: Motion Forecasting via Simple & Efficient Attention Networks", "Motion Transformer with Global Intention Localization and Local Movement Refinement", "MultiPath++: Efficient Information Fusion and Trajectory Aggregation for Behavior Prediction", "ProphNet: Efficient Agent-Centric Motion Forecasting with Anchor-Informed Proposals"], "answer_arxiv_id": ["2207.05844", "2209.13508", "2111.14973", "2303.12071"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_10211"} +{"question": "Can you cite the studies involved in weighted behavior cloning methods where they modify an imitation learning algorithm by filtering or weighting the actions in the dataset?", "answer": ["RvS: What is Essential for Offline RL via Supervised Learning?", "BAIL: Best-Action Imitation Learning for Batch Deep Reinforcement Learning", "Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["2112.10751", "1910.12179", "2106.01345"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_10212"} +{"question": "What papers relate LiDAR data to image-based primitive prediction?", "answer": ["Cuboids Revisited: Learning Robust 3D Shape Fitting to Single RGB Images", "Learning Unsupervised Hierarchical Part Decomposition of 3D Objects from\n a Single RGB Image"], "answer_arxiv_id": ["2105.02047", "2004.01176"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_10213"} +{"question": "Are there any studies that perform joint spatial and frequency decomposition by applying Fourier feature encodings on local grid features?", "answer": ["Neural Fourier Filter Bank"], "answer_arxiv_id": ["2212.01735"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_10214"} +{"question": "What papers discussed methods that used image prompts to generate similar content or styles?", "answer": ["StyleDrop: Text-to-Image Generation in Any Style", "Inversion-Based Style Transfer with Diffusion Models", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable Text-to-Image Generation and Editing", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2306.00983", "2211.13203", "2305.14720", "2308.06721"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10215"} +{"question": "Which works have incorporated Monte Carlo dropout for predictive uncertainty?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.02142"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_10216"} +{"question": "What papers analyse curriculum learning strategies in convex models?", "answer": ["Curriculum Learning by Transfer Learning: Theory and Experiments with Deep Networks", "Theory of Curriculum Learning, with Convex Loss Functions"], "answer_arxiv_id": ["1802.03796", "1812.03472"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10217"} +{"question": "What papers propose randomization in the paper assignment process for evaluating different assignment policies?", "answer": ["Mitigating Manipulation in Peer Review via Randomized Reviewer Assignments"], "answer_arxiv_id": ["2006.16437"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_10218"} +{"question": "Which paper proposed a diffusion model in torsion space and used the underlying ODE as a transferable Boltzmann generator?", "answer": ["Torsional Diffusion for Molecular Conformer Generation"], "answer_arxiv_id": ["2206.01729"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10219"} +{"question": "What papers introduced non-uniform quantization?", "answer": ["Post training 4-bit quantization of convolutional networks for rapid-deployment", "Binarized Neural Networks", "BiQGEMM: Matrix Multiplication with Lookup Table For Binary-Coding-based Quantized DNNs", "Quantized Convolutional Neural Networks for Mobile Devices", "LQ-Nets: Learned Quantization for Highly Accurate and Compact Deep Neural Networks", "Convolutional Neural Networks using Logarithmic Data Representation", "Incremental Network Quantization: Towards Lossless CNNs with Low-Precision Weights", "Additive Powers-of-Two Quantization: An Efficient Non-uniform Discretization for Neural Networks"], "answer_arxiv_id": ["1810.05723", "1602.02505", "2005.09904v2", "1512.06473", "1807.10029v1", "1603.01025v2", "1702.03044", "1909.13144"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_10220"} +{"question": "What papers investigated the use of machine learning in recommender systems?", "answer": ["AutoShard: Automated Embedding Table Sharding for Recommender Systems", "Skewness Ranking Optimization for Personalized Recommendation"], "answer_arxiv_id": ["2208.06399", "2005.12971"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_10221"} +{"question": "Any works that show the significance of data de-duplication on models' generalisation?", "answer": ["Deduplicating Training Data Makes Language Models Better", "D4: Improving LLM Pretraining via Document De-Duplication and\n Diversification", "SemDeDup: Data-efficient learning at web-scale through semantic\n deduplication"], "answer_arxiv_id": ["2107.06499", "2308.12284", "2303.09540"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_10222"} +{"question": "What was the first paper to propose asynchronous and decentralized optimization algorithms?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_10223"} +{"question": "What papers showcase the lack of diversity in deterministic decoding methods?", "answer": ["On NMT Search Errors and Model Errors: Cat Got Your Tongue?", "Best-First Beam Search"], "answer_arxiv_id": ["1908.10090", "2007.03909"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_10224"} +{"question": "What works applied the Transformer architecture in MCSR?", "answer": ["Transformer-empowered Multi-scale Contextual Matching and Aggregation\n for Multi-contrast MRI Super-resolution"], "answer_arxiv_id": ["2203.13963"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_10225"} +{"question": "What works discuss the development of positional encodings capable of handling longer text sequences?", "answer": ["A Length-Extrapolatable Transformer", "Train Short, Test Long: Attention with Linear Biases Enables Input\n Length Extrapolation"], "answer_arxiv_id": ["2212.10554", "2108.12409"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_10226"} +{"question": "What is the paper where PixelNeRF, a cross-scene generalizable variant of NeRF, is introduced?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images"], "answer_arxiv_id": ["2012.02190"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_10227"} +{"question": "What papers discuss the use of text-to-image diffusion models for controllable image generation with text prompts?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2102.12092", "2205.11487", "2307.01952"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10228"} +{"question": "What works contain discussions about the relation between compressibility and generalization ability?", "answer": ["Language Modeling Is Compression"], "answer_arxiv_id": ["2309.10668"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_10229"} +{"question": "Which works used CLIP-based supervision for the synthesis of entire 3D scenes?", "answer": ["Zero-Shot Text-Guided Object Generation with Dream Fields", "CLIP-Forge: Towards Zero-Shot Text-to-Shape Generation"], "answer_arxiv_id": ["2112.01455", "2110.02624"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10230"} +{"question": "What studies have designed scalable Transformers to address computational limitations of regular Transformers in L2L systems?", "answer": ["Efficient Transformers: A Survey", "Long Range Arena: A Benchmark for Efficient Transformers"], "answer_arxiv_id": ["2009.06732", "2011.04006"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10231"} +{"question": "What studies utilized NeRFs in perception tasks like panoptic segmentation or object detection?", "answer": ["Panoptic NeRF: 3D-to-2D Label Transfer for Panoptic Urban Scene\n Segmentation", "NeRF-Det: Learning Geometry-Aware Volumetric Representation for\n Multi-View 3D Object Detection", "MonoNeRD: NeRF-like Representations for Monocular 3D Object Detection"], "answer_arxiv_id": ["2203.15224", "2307.14620", "2308.09421"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_10232"} +{"question": "Which papers investigate the learning of conditional probability in a diffusion model with different types of conditions?", "answer": ["Zero-Shot Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2112.10752"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_10233"} +{"question": "Which papers use meshes or point clouds as explicit representations for human reconstruction?", "answer": ["Efficient Meshy Neural Fields for Animatable Human Avatars", "PointAvatar: Deformable Point-based Head Avatars from Videos"], "answer_arxiv_id": ["2303.12965", "2212.08377"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10234"} +{"question": "What studies improved spectral clustering using the power method?", "answer": ["Deterministic Feature Selection for K-means Clustering"], "answer_arxiv_id": ["1109.5664"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_10235"} +{"question": "Which works discuss the role of model and data scale in improving performance in various tasks?", "answer": ["Training Compute-Optimal Large Language Models", "Scaling Laws for Neural Language Models", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2203.15556", "2001.08361", "2204.02311"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_10236"} +{"question": "In what papers the researcher proposed recent approaches to sequential modeling?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces", "It’s Raw! Audio Generation with State-Space Models", "Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layers", "HiPPO: Recurrent Memory with Optimal Polynomial Projections"], "answer_arxiv_id": ["2111.00396", "2202.09729", "2110.13985", "2008.07669"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_10237"} +{"question": "Which works introduced advancements in learning diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2006.11239", "2011.13456"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_10238"} +{"question": "What research give neural networks as the most popular technique for Zero-shot video object segmentation (ZSVOS)?", "answer": ["Zero-Shot Video Object Segmentation via Attentive Graph Neural Networks", "Learning to Segment Moving Objects", "Unsupervised Online Video Object Segmentation with Motion Property\n Understanding"], "answer_arxiv_id": ["2001.06807", "1712.01127", "1810.03783"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_10239"} +{"question": "What studies explore more efficient ways than RL to train imitators?", "answer": ["DiffMimic: Efficient Motion Mimicking with Differentiable Physics"], "answer_arxiv_id": ["2304.03274"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_10240"} +{"question": "Are there any studies that introduced a recognition network-based loss on the decoder to encourage predictable latent traversals?", "answer": ["Learning Disentangled Representations with Latent Variation Predictability"], "answer_arxiv_id": ["2007.12885"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_10241"} +{"question": "Could you provide me some studies about using in-context learning in mathematical reasoning?", "answer": ["Solving Quantitative Reasoning Problems with Language Models"], "answer_arxiv_id": ["2206.14858"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_10242"} +{"question": "Can you reference studies that utilized the Objaverse 1.0 dataset?", "answer": ["Objaverse: A Universe of Annotated 3D Objects"], "answer_arxiv_id": ["2212.08051"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_10243"} +{"question": "What research studies have been conducted on measuring proximity between two distributions using f-divergence in Distributional Robustness (DR)?", "answer": ["Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach", "Distributional Smoothing with Virtual Adversarial Training"], "answer_arxiv_id": ["1610.03425", "1507.00677"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_10244"} +{"question": "Are there any studies that update the affine parameters of batch normalization layers using unsupervised loss, entropy minimization loss, or feature distribution alignments loss?", "answer": ["Tent: Fully Test-time Adaptation by Entropy Minimization", "Efficient Test-Time Model Adaptation without Forgetting", "DELTA: degradation-free fully test-time adaptation", "TTN: A Domain-Shift Aware Batch Normalization in Test-Time Adaptation", "CAFA: Class-Aware Feature Alignment for Test-Time Adaptation"], "answer_arxiv_id": ["2006.10726", "2204.02610", "2301.13018", "2302.05155", "2206.00205"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_10245"} +{"question": "Any works about introducing differentiable rendering functions to optimize neural implicit shape representations?", "answer": ["Differentiable Volumetric Rendering: Learning Implicit 3D\n Representations without 3D Supervision", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural\n Scene Representations"], "answer_arxiv_id": ["1912.07372", "1906.01618"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_10246"} +{"question": "Which research refuted the conjecture of FTRL with the log-barrier (LB-FTRL) achieving the optimal regret?", "answer": ["Pushing the Efficiency-Regret Pareto Frontier for Online Learning of Portfolios and Quantum States"], "answer_arxiv_id": ["2202.02765"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_10247"} +{"question": "Which work propose text module networks to solve complex tasks by decomposing them into simpler ones as a part of neural-symbolic methods?", "answer": ["Text Modular Networks: Learning to Decompose Tasks in the Language of Existing Models"], "answer_arxiv_id": ["2009.00751"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_10248"} +{"question": "Could you provide me some studies that extended earlier works by fine-tuning pretrained text-to-image models for new concept learning?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models", "Multi-Concept Customization of Text-to-Image Diffusion", "Cones: Concept Neurons in Diffusion Models for Customized Generation"], "answer_arxiv_id": ["2208.12242", "2307.06949", "2212.04488", "2303.05125"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10249"} +{"question": "Which papers have used NAS to design network architecture for efficient visual tracking?", "answer": ["LightTrack: Finding Lightweight Neural Networks for Object Tracking via One-Shot Architecture Search"], "answer_arxiv_id": ["2104.14545"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_10250"} +{"question": "Which study showed improvements by optimizing the diffusion model and individual NeRFs for each training object in a joint single stage?", "answer": ["Single-Stage Diffusion NeRF: A Unified Approach to 3D Generation and\n Reconstruction"], "answer_arxiv_id": ["2304.06714"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_10251"} +{"question": "Which papers have adopted the concept of perceptual loss in the approaching advancement of image inpainting?", "answer": ["Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "Toward Multimodal Image-to-Image Translation"], "answer_arxiv_id": ["1603.08155", "1711.11586"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_10252"} +{"question": "Which works proposed fidelity-oriented methods for enhancing the quality of compressed images?", "answer": ["Compression Artifacts Reduction by a Deep Convolutional Network", "Learning Deep CNN Denoiser Prior for Image Restoration", "Early Exit or Not: Resource-Efficient Blind Quality Enhancement for\n Compressed Images", "DAQE: Enhancing the Quality of Compressed Images by Exploiting the\n Inherent Characteristic of Defocus"], "answer_arxiv_id": ["1504.06993", "1704.03264", "2006.16581", "2211.10984"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_10253"} +{"question": "What research do tackle the human-like multimodal features in word learning?", "answer": ["ShapeWorld: A new test methodology for multimodal language understanding", "Abstract Visual Reasoning with Tangram Shapes"], "answer_arxiv_id": ["1704.04517", "2211.16492"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10254"} +{"question": "Which works use Message Passing Neural Networks to encode subgraphs rather than subtrees around graph nodes?", "answer": ["Nested Graph Neural Networks", "From Stars to Subgraphs: Uplifting Any GNN with Local Structure Awareness", "How Powerful are K-hop Message Passing Graph Neural Networks", "Identity-aware Graph Neural Networks"], "answer_arxiv_id": ["2110.13197", "2110.03753", "2205.13328", "2101.10320"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_10255"} +{"question": "Which work used large language models to explain planned actions or adjust the low-level control parameters in self-driving cars?", "answer": ["LanguageMPC: Large Language Models as Decision Makers for Autonomous Driving"], "answer_arxiv_id": ["2310.03026v2"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_10256"} +{"question": "What studies have introduced frequency-based regularization techniques to prevent spectral bias during the training of neural networks?", "answer": ["Learning in the Frequency Domain", "Focal Frequency Loss for Image Reconstruction and Synthesis"], "answer_arxiv_id": ["2002.12416", "2012.12821"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_10257"} +{"question": "Which papers are notable for their contributions to contrastive learning across various domains?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Deep Graph Infomax", "An efficient framework for learning sentence representations", "Learning Representations by Maximizing Mutual Information Across Views"], "answer_arxiv_id": ["2002.05709", "1809.10341", "1803.02893", "1906.00910"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_10258"} +{"question": "What research is about application of retrieval augmented generations (RAG) technique in multimodal image generation?", "answer": ["Retrieval-Augmented Multimodal Language Modeling"], "answer_arxiv_id": ["2211.12561"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_train_10259"} +{"question": "What studies are improving adversarial robustness of models?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Adversarial Weight Perturbation Helps Robust Generalization", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1412.6572", "1706.06083", "2004.05884", "1901.08573"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10260"} +{"question": "Could you provide me papers that utilize the SDS loss to optimize the meshed-based neural field?", "answer": ["DreamEditor: Text-Driven 3D Scene Editing with Neural Fields"], "answer_arxiv_id": ["2306.13455"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_10261"} +{"question": "Could you provide me some works that efficiently applied GCNN for mesh reconstruction?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Pose2Mesh: Graph Convolutional Network for 3D Human Pose and Mesh\n Recovery from a 2D Human Pose", "Mesh Graphormer"], "answer_arxiv_id": ["1609.02907", "2008.09047", "2104.00272"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_10262"} +{"question": "Which papers have been published about modeling object locations in scenes in embodied AI?", "answer": ["Bayesian Relational Memory for Semantic Visual Navigation", "Semantic and Geometric Modeling with Neural Message Passing in 3D Scene Graphs for Hierarchical Mechanical Search", "Learning Object-Based State Estimators for Household Robots", "A Contextual Bandit Approach for Learning to Plan in Environments with Probabilistic Goal Configurations", "Proactive Robot Assistance via Spatio-Temporal Object Modeling"], "answer_arxiv_id": ["1909.04306", "2012.04060", "2011.03183", "2211.16309", "2211.15501"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_10263"} +{"question": "What papers developed visual encoders for visual navigation?", "answer": ["DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion\n Frames", "Where are we in the search for an Artificial Visual Cortex for Embodied\n Intelligence?", "OVRL-V2: A simple state-of-art baseline for ImageNav and ObjectNav", "End-to-End (Instance)-Image Goal Navigation through Correspondence as an\n Emergent Phenomenon"], "answer_arxiv_id": ["1911.00357", "2303.18240", "2303.07798", "2309.16634"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_10264"} +{"question": "Could you provide me with some work related to adversarially trained networks performance in ImageNet?", "answer": ["Do Adversarially Robust ImageNet Models Transfer Better?"], "answer_arxiv_id": ["2007.08489"], "source_meta": {"published_time": "20230111"}, "qid": "AutoScholarQuery_train_10265"} +{"question": "Which work initially presented the diffusion model?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_10266"} +{"question": "Where can we find progress when 'g0' is discontinuous and 'D=1'?", "answer": ["Multilevel nested simulation for efficient risk estimation", "Multilevel Path Branching for Digital Options"], "answer_arxiv_id": ["1802.05016v2", "2209.03017v2"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_10267"} +{"question": "Identify works that used stochasticity to improve differentially private machine learning for models trained on small datasets.", "answer": ["An Empirical Study on the Intrinsic Privacy of SGD"], "answer_arxiv_id": ["1912.02919v4"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_10268"} +{"question": "What are the conventional approaches for transforming raw dialogue text to machine-readable representations?", "answer": ["Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models", "A Hierarchical Latent Variable Encoder-Decoder Model for Generating Dialogues", "Hierarchical Recurrent Attention Network for Response Generation"], "answer_arxiv_id": ["1507.04808", "1605.06069", "1701.07149"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_10269"} +{"question": "Could you provide some works which studied methods on deep degradation modeling?", "answer": ["Noise Flow: Noise Modeling with Conditional Normalizing Flows", "Modeling sRGB Camera Noise with Normalizing Flows", "Dual Adversarial Network: Toward Real-world Noise Removal and Noise\n Generation", "To learn image super-resolution, use a GAN to learn how to do image\n degradation first", "Unsupervised Learning for Real-World Super-Resolution", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "Frequency Separation for Real-World Super-Resolution", "Unsupervised Real-world Image Super Resolution via Domain-distance Aware\n Training", "DeFlow: Learning Complex Image Degradations from Unpaired Data with\n Conditional Flows", "Learn from Unpaired Data for Image Restoration: A Variational Bayes\n Approach"], "answer_arxiv_id": ["1908.08453", "2206.00812", "2007.05946", "1807.11458", "1909.09629", "1703.10593", "1911.07850", "2004.01178", "2101.05796", "2204.10090"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_10270"} +{"question": "Any works in the literature about temporal consistency in video matting?", "answer": ["Deep Video Matting via Spatio-Temporal Alignment and Aggregation", "One-Trimap Video Matting", "Real-Time High-Resolution Background Matting", "Robust High-Resolution Video Matting with Temporal Guidance", "Adaptive Human Matting for Dynamic Videos", "Attention-guided Temporally Coherent Video Object Matting", "VMFormer: End-to-End Video Matting with Transformer"], "answer_arxiv_id": ["2104.11208", "2207.13353", "2012.07810", "2108.11515", "2304.06018", "2105.11427", "2208.12801"], "source_meta": {"published_time": "20240424"}, "qid": "AutoScholarQuery_train_10271"} +{"question": "Which work overcame the challenges of obtaining optimal complexity for projected/proximal SGD within the framework of weakly convex optimization?", "answer": ["Stochastic model-based minimization of weakly convex functions"], "answer_arxiv_id": ["1803.06523"], "source_meta": {"published_time": "20220329"}, "qid": "AutoScholarQuery_train_10272"} +{"question": "Which studies have suggested that language models may not effectively learn from the provided instructions?", "answer": ["Do Prompt-Based Models Really Understand the Meaning of Their Prompts?"], "answer_arxiv_id": ["2109.01247"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_10273"} +{"question": "Any works about patches in learning-based 3D reconstruction?", "answer": ["AtlasNet: A Papier-Mâché Approach to Learning 3D Surface Generation"], "answer_arxiv_id": ["1802.05384"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_10274"} +{"question": "Which works fine-tuned Language Models (LMs) to directly provide answers to logical reasoning questions?", "answer": ["Transformers as Soft Reasoners over Language", "Critical Thinking for Language Models", "RuleBERT: Teaching Soft Rules to Pre-Trained Language Models", "FOLIO: Natural Language Reasoning with First-Order Logic"], "answer_arxiv_id": ["2002.05867", "2009.07185", "2109.13006", "2209.00840"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_10275"} +{"question": "What are the studies that proposed more efficient adversarial training methodologies?", "answer": ["Adversarial Training for Free!", "Fast is better than free: Revisiting adversarial training"], "answer_arxiv_id": ["1904.12843", "2001.03994"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_10276"} +{"question": "Which papers have discussed the use of data augmentation for efficient representation learning in pixel-based RL?", "answer": ["Reinforcement Learning with Augmented Data", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Mastering Visual Continuous Control: Improved Data-Augmented Reinforcement Learning"], "answer_arxiv_id": ["2004.14990", "2004.13649", "2107.09645"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10277"} +{"question": "Which research works have been conducted on specific fallacies like 'Non Sequitur' and 'Ad Hominem'?", "answer": ["Before Name-calling: Dynamics and Triggers of Ad Hominem Fallacies in\n Web Argumentation"], "answer_arxiv_id": ["1802.06613"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_10278"} +{"question": "What works embed bundle adjustment layers in end-to-end differentiable network architectures?", "answer": ["BA-Net: Dense Bundle Adjustment Network", "Pixel-Perfect Structure-from-Motion with Featuremetric Refinement", "Multi-View Optimization of Local Feature Geometry"], "answer_arxiv_id": ["1806.04807v3", "2108.08291", "2003.08348"], "source_meta": {"published_time": "20220808"}, "qid": "AutoScholarQuery_train_10279"} +{"question": "Could you provide me with some Distance-based methods used in Anomaly Detection?", "answer": ["PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and\n Localization", "Sub-Image Anomaly Detection with Deep Pyramid Correspondences", "Towards Total Recall in Industrial Anomaly Detection"], "answer_arxiv_id": ["2011.08785", "2005.02357", "2106.08265"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_10280"} +{"question": "Which studies first proposed to extract visual object features and align vision and language representations in vision-language pre-training?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "LXMERT: Learning Cross-Modality Encoder Representations from Transformers"], "answer_arxiv_id": ["1908.02265", "1908.07490"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_10281"} +{"question": "Which research works explore learning 3D generative models using 2D supervision?", "answer": ["Vision Transformers for Dense Prediction", "Efficient Geometry-aware 3D Generative Adversarial Networks", "GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature\n Fields", "Zero-Shot Text-Guided Object Generation with Dream Fields", "DreamFusion: Text-to-3D using 2D Diffusion", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2103.13413", "2112.07945", "2007.02442", "2011.12100", "2112.01455", "2209.14988", "2305.16213", "2212.00774v1"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_10282"} +{"question": "Who introduced training-free model merging proposing a way to merge multiple tasks into a single multi-task model?", "answer": ["Dataless Knowledge Fusion by Merging Weights of Language Models", "Editing Models with Task Arithmetic"], "answer_arxiv_id": ["2212.09849", "2212.04089"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_10283"} +{"question": "What works developed auxiliary models for efficient updating and refining of LLMs?", "answer": ["Memory-Based Model Editing at Scale", "Editing Factual Knowledge in Language Models", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2206.06520", "2104.08164", "2110.11309"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_10284"} +{"question": "Could you give me some studies aimed at improving the performance of open-surface representation using neural networks?", "answer": ["Deep Implicit Surface Point Prediction Networks", "RangeUDF: Semantic Surface Reconstruction from 3D Point Clouds", "GIFS: Neural Implicit Function for General Shape Representation", "Neural Vector Fields: Implicit Representation by Explicit Learning", "Learning Anchored Unsigned Distance Functions with Gradient Direction\n Alignment for Single-view Garment Reconstruction", "CAP-UDF: Learning Unsigned Distance Functions Progressively from Raw\n Point Clouds with Consistency-Aware Field Optimization"], "answer_arxiv_id": ["2106.05779", "2204.09138", "2204.07126", "2303.04341", "2108.08478", "2210.02757"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_10285"} +{"question": "What studies interpreted ICL as implicit Bayesian inference and developed guarantees when the training distribution is a mixture of Hidden Markov Models?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230117"}, "qid": "AutoScholarQuery_train_10286"} +{"question": "Which pieces of research extracted spectral features in low-rank MDPs with exploration?", "answer": ["FLAMBE: Structural Complexity and Representation Learning of Low Rank MDPs", "Model-free Representation Learning and Exploration in Low-rank MDPs", "Representation Learning for Online and Offline RL in Low-rank MDPs", "Efficient Reinforcement Learning in Block MDPs: A Model-free Representation Learning Approach", "On the Statistical Efficiency of Reward-Free Exploration in Non-Linear RL"], "answer_arxiv_id": ["2006.10814", "2102.07035", "2110.04652", "2202.00063", "2206.10770"], "source_meta": {"published_time": "20220819"}, "qid": "AutoScholarQuery_train_10287"} +{"question": "What studies proposed the physics-based models for motion prediction?", "answer": ["What the Constant Velocity Model Can Teach Us About Pedestrian Motion Prediction", "A Computationally Efficient Model for Pedestrian Motion Prediction"], "answer_arxiv_id": ["1903.07933", "1803.04702v1"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_10288"} +{"question": "What papers focused on training a classifier to distinguish between seen and unseen visual features?", "answer": ["Zero-Shot Semantic Segmentation", "Context-aware Feature Generation for Zero-shot Semantic Segmentation", "A Closer Look at Self-training for Zero-Label Semantic Segmentation"], "answer_arxiv_id": ["1906.00817", "2008.06893", "2104.11692"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_10289"} +{"question": "Which studies extensively contributed to the development of 'diffusion models' in the context of image generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2010.02502"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_10290"} +{"question": "What methods propose ways of faster personalization of Text-to-Image (T2I) models?", "answer": ["ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation"], "answer_arxiv_id": ["2302.13848"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_10291"} +{"question": "Which studies minimize a sound over-approximation of the worst-case loss computed using the Box relaxation?", "answer": ["On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models"], "answer_arxiv_id": ["1810.12715"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_10292"} +{"question": "What are some leading works that implemented deep convolutional neural network on sensors?", "answer": ["A Camera That CNNs: Towards Embedded Neural Networks on Pixel Processor\n Arrays"], "answer_arxiv_id": ["1909.05647"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_10293"} +{"question": "What studies use prompts to guide LLMs in generating complete action sequences for elementary tasks?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents"], "answer_arxiv_id": ["2201.11903", "2201.07207"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_10294"} +{"question": "Can you name the works where non-local attention block is integrated into CNN for the first time?", "answer": ["Non-local Neural Networks"], "answer_arxiv_id": ["1711.07971"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_10295"} +{"question": "Can you name the studies that have proposed different solutions towards addressing the issue of background bias in action recognition models?", "answer": ["VideoMix: Rethinking Data Augmentation for Video Classification", "Learning Representational Invariances for Data-Efficient Action Recognition", "Learn2Augment: Learning to Composite Videos for Data Augmentation in Action Recognition", "Why Can’t I Dance in the Mall? Learning to Mitigate Scene Bias in Action Recognition"], "answer_arxiv_id": ["2012.03457", "2103.16565", "2206.04790", "1912.05534"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_10296"} +{"question": "Which works have verified the efficacy of setting the latent space of VAE to be hyperbolic space?", "answer": ["Continuous Hierarchical Representations with Poincaré Variational Auto-Encoders", "A Wrapped Normal Distribution on Hyperbolic Space for Gradient-Based Learning", "A Rotated Hyperbolic Wrapped Normal Distribution for Hierarchical Representation Learning"], "answer_arxiv_id": ["1901.06033", "1902.02992", "2205.13371"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_10297"} +{"question": "Which works have applied representation learning in RL to generate representations of sensory information?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Learning Action Representations for Reinforcement Learning", "Reinforcement Learning with Prototypical Representations"], "answer_arxiv_id": ["2004.04136", "1902.00183", "2102.11271"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_10298"} +{"question": "Which research papers proposed conversion-based approaches for training Spiking Neural Networks (SNNs)?", "answer": ["Going Deeper in Spiking Neural Networks: VGG and Residual Architectures", "Enabling Deep Spiking Neural Networks with Hybrid Conversion and Spike Timing Dependent Backpropagation"], "answer_arxiv_id": ["1802.02627", "2005.01807"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_10299"} +{"question": "Which papers discussed about learning a 3D-aware representation from a single view image using a standard convolutional autoencoder architecture with deterministic latent vectors?", "answer": ["Multi-view 3D Models from Single Images with a Convolutional Network", "Interpretable Transformations with Encoder-Decoder Networks"], "answer_arxiv_id": ["1511.06702", "1710.07307"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_10300"} +{"question": "Which research papers discuss the idea of using a prior in the context of joint angle reconstruction?", "answer": ["Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time"], "answer_arxiv_id": ["1810.04703"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_10301"} +{"question": "Which research propose methods for self-supervised reinforcement learning where an agent explores the environment through training and proposing novel goals?", "answer": ["Zero-Shot Visual Imitation", "Go-Explore: a New Approach for Hard-Exploration Problems", "Provably Efficient Maximum Entropy Exploration", "Planning to Explore via Self-Supervised World Models", "Discovering and Achieving Goals via World Models"], "answer_arxiv_id": ["1804.08606", "1901.10995", "1812.02690", "2005.05960", "2110.09514"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_10302"} +{"question": "What works aim at obtaining generalization bounds that are dependent on the norms of the weights and Lipschitz continuity properties of the network?", "answer": ["Norm-Based Capacity Control in Neural Networks", "Spectrally-normalized margin bounds for neural networks", "A PAC-Bayesian Approach to Spectrally-Normalized Margin Bounds for Neural Networks", "Size-Independent Sample Complexity of Neural Networks", "Stronger generalization bounds for deep nets via a compression approach", "Deterministic PAC-Bayesian generalization bounds for deep networks via generalizing noise-resilience"], "answer_arxiv_id": ["1503.00036", "1706.08498", "1707.09564", "1712.06541", "1802.05296", "1905.13344"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_10303"} +{"question": "What are the papers that explored the applicability of prompt learning in NLP to computer vision?", "answer": ["Visual Prompt Tuning", "Learning to Prompt for Continual Learning", "Diversity-Aware Meta Visual Prompting", "Explicit Visual Prompting for Low-Level Structure Segmentations"], "answer_arxiv_id": ["2203.12119", "2112.08654", "2303.08138", "2303.10883"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_10304"} +{"question": "What are some works about the use of random walks in constructing node and graph representations?", "answer": ["DeepWalk: Online Learning of Social Representations", "node2vec: Scalable Feature Learning for Networks", "Walk Message Passing Neural Networks and Second-Order Graph Neural Networks"], "answer_arxiv_id": ["1403.6652", "1607.00653", "2006.09499"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_10305"} +{"question": "Which works discuss the use of normalizing flows to minimize KL-divergence over the Wasserstein space?", "answer": ["Variational Inference with Normalizing Flows", "Variational Inference with Continuously-Indexed Normalizing Flows"], "answer_arxiv_id": ["1505.05770", "2007.05426"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_10306"} +{"question": "What studies used Generative Adversarial Networks (GAN) for synthesizing low-light noisy videos for training?", "answer": ["Low Light Video Enhancement using Synthetic Data Produced with an\n Intermediate Domain Mapping"], "answer_arxiv_id": ["2007.09187"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_10307"} +{"question": "What research papers serve as thorough introductions to Conformal Predictions (CP)?", "answer": ["A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification"], "answer_arxiv_id": ["2107.07511"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_10308"} +{"question": "What studies introduce the method of PF-ODEs and DDIM for learning ODEs?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2011.13456", "2010.02502"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_10309"} +{"question": "Are there any papers where a large-scale dataset for category-agnostic pose estimation was developed?", "answer": ["Pose for Everything: Towards Category-Agnostic Pose Estimation"], "answer_arxiv_id": ["2207.10387"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_10310"} +{"question": "What studies have proposed different empirical algorithms to close the sim-to-real gap?", "answer": ["Robust Adversarial Reinforcement Learning", "Transfer from Simulation to Real World through Learning Deep Inverse Dynamics Model", "Sim-to-Real Robot Learning from Pixels with Progressive Nets", "Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World"], "answer_arxiv_id": ["1703.02702", "1610.03518", "1610.04286", "1703.06907"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_10311"} +{"question": "Could you provide me some works that generated textual summaries by taking audio, transcripts, or documents as input along with videos or images?", "answer": ["See, Hear, Read: Leveraging Multimodality with Guided Attention for\n Abstractive Text Summarization"], "answer_arxiv_id": ["2105.09601"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_10312"} +{"question": "Could you provide me some works which involves vision-based manipulation in embodied AI tasks?", "answer": ["AI2-THOR: An Interactive 3D Environment for Visual AI", "VirtualHome: Simulating Household Activities via Programs", "VRGym: A Virtual Testbed for Physical and Interactive AI", "ALFRED A Benchmark for Interpreting Grounded Instructions for Everyday Tasks", "ALFWorld: Aligning Text and Embodied Environments for Interactive Learning", "CLIPort: What and Where Pathways for Robotic Manipulation"], "answer_arxiv_id": ["1712.05474", "1806.07011", "1904.01698", "1912.01734", "2010.03768", "2109.12098"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_10313"} +{"question": "Could you give examples of research that talk about the substantial size of state-of-the-art DNN models?", "answer": ["Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1512.03385"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_10314"} +{"question": "Which work proposed use of a continuous version of shallow fully-connected networks on non-compact symmetric space?", "answer": ["Fully-Connected Network on Noncompact Symmetric Space and Ridgelet Transform based on Helgason-Fourier Analysis"], "answer_arxiv_id": ["2203.01631"], "source_meta": {"published_time": "20220214"}, "qid": "AutoScholarQuery_train_10315"} +{"question": "What works mention large LMs learning in few-shot scenarios?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2204.02311"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_10316"} +{"question": "Could you cite the studies that have considered the two-party learning setting with one features party and one labels party?", "answer": ["Label Leakage and Protection in Two-party Split Learning"], "answer_arxiv_id": ["2102.08504"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_10317"} +{"question": "Any works about using reasoning graphs for generating explanations?", "answer": ["STREET: A Multi-Task Structured Reasoning and Explanation Benchmark"], "answer_arxiv_id": ["2302.06729"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_10318"} +{"question": "Which works have used hessian-based influence functions in gradient-based methods?", "answer": ["Understanding Black-box Predictions via Influence Functions", "TRAK: Attributing Model Behavior at Scale", "Cross-Loss Influence Functions to Explain Deep Network Representations", "Scaling Up Influence Functions", "Rethinking Influence Functions of Neural Networks in the Over-parameterized Regime", "Influence Functions in Deep Learning Are Fragile", "FastIF: Scalable Influence Functions for Efficient Model Interpretation and Debugging", "On Second-Order Group Influence Functions for Black-Box Predictions"], "answer_arxiv_id": ["1703.04730", "2303.14186", "2012.01685v2", "2112.03052", "2112.08297", "2006.14651", "2012.15781", "1911.00418"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_10319"} +{"question": "What papers regressed the reflectance parameters and illumination from a reflectance map?", "answer": ["DeLight-Net: Decomposing Reflectance Maps into Specular Materials and\n Natural Illumination"], "answer_arxiv_id": ["1603.08240"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_10320"} +{"question": "Which studies identified 'robustness disparity' on a wide range of datasets and model architectures?", "answer": ["Fairness Through Robustness: Investigating Robustness Disparity in Deep Learning", "Analysis and Applications of Class-wise Robustness in Adversarial Training"], "answer_arxiv_id": ["2006.12621", "2105.14240"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_10321"} +{"question": "What is the work that uses a diffusion model for Arbitrary-Scale Super-Resolution where the output is at a medium resolution?", "answer": ["Implicit Diffusion Models for Continuous Super-Resolution"], "answer_arxiv_id": ["2303.16491"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_10322"} +{"question": "Which research papers studied the acoustic correspondence between video and audio?", "answer": ["Image2Reverb: Cross-Modal Reverb Impulse Response Synthesis", "Visual Acoustic Matching"], "answer_arxiv_id": ["2103.14201", "2202.06875"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_10323"} +{"question": "What papers have proposed algorithms to achieve the minimax lower bound?", "answer": ["Minimax Regret Bounds for Reinforcement Learning", "Tighter Problem-Dependent Regret Bounds in Reinforcement Learning without Domain Knowledge using Value Function Bounds", "Fast active learning for pure exploration in reinforcement learning"], "answer_arxiv_id": ["1703.05449", "1901.00210", "2007.13442"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_10324"} +{"question": "Which paper introduced the concept of 'fixed points' in the context of non-linear PDE function?", "answer": ["On the Role of Fixed Points of Dynamical Systems in Training Physics-Informed Neural Networks"], "answer_arxiv_id": ["2203.13648"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_10325"} +{"question": "Which works pre-train world models by using offline experience datasets across Atari games?", "answer": ["On the Feasibility of Cross-Task Transfer with Model-Based Reinforcement Learning"], "answer_arxiv_id": ["2210.10763"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_10326"} +{"question": "What works have leveraged soft targets in addition to ground truth labels to enforce consistency regularization across previous and current model predictions?", "answer": ["Measuring and regularizing networks in function space", "Dark Experience for General Continual Learning: a Strong, Simple Baseline", "Learning Fast, Learning Slow: A General Continual Learning Method based on Complementary Learning System"], "answer_arxiv_id": ["1805.08289", "2004.07211", "2201.12604"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_10327"} +{"question": "Which research explored a new reparametrization of the linear forward SDE for refining a VAE’s output?", "answer": ["DiffuseVAE: Efficient, Controllable and High-Fidelity Generation from Low-Dimensional Latents"], "answer_arxiv_id": ["2201.00308"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_10328"} +{"question": "Which works are about optimization of all parameters in Language Models for better performance under the approach prompt-based fine-tuning?", "answer": ["Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference", "Making Pre-trained Language Models Better Few-shot Learners", "GPT Understands, Too", "Differentiable Prompt Makes Pre-trained Language Models Better Few-shot Learners"], "answer_arxiv_id": ["2001.07676", "2012.15723", "2103.10385", "2108.13161"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_10329"} +{"question": "What papers have developed compressed video quality enhancement methods based on existing datasets?", "answer": ["MFQE 2.0: A New Approach for Multi-frame Quality Enhancement on\n Compressed Video", "NTIRE 2021 Challenge on Quality Enhancement of Compressed Video: Dataset\n and Study", "NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of\n Compressed Video: Dataset, Methods and Results"], "answer_arxiv_id": ["1902.09707", "2104.10782", "2204.09314"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10330"} +{"question": "What papers have worked on the development of datasets for hand and object interaction?", "answer": ["HOnnotate: A method for 3D Annotation of Hand and Object Poses", "DexYCB: A Benchmark for Capturing Hand Grasping of Objects", "HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object\n Interaction", "Learning joint reconstruction of hands and manipulated objects"], "answer_arxiv_id": ["1907.01481", "2104.04631", "2203.01577", "1904.05767"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_10331"} +{"question": "Any works in image inpainting that leverage auto-regressive transformer structures?", "answer": ["Diverse Image Inpainting with Bidirectional and Autoregressive Transformers", "High-Fidelity Pluralistic Image Completion with Transformers"], "answer_arxiv_id": ["2104.12335", "2103.14031"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_10332"} +{"question": "What are some studies that have used model rectification-based methods for incremental learning?", "answer": ["Maintaining Discrimination and Fairness in Class Incremental Learning", "Large Scale Incremental Learning", "Semantic Drift Compensation for Class-Incremental Learning", "Mimicking the Oracle: An Initial Phase Decorrelation Approach for Class\n Incremental Learning", "Continual Normalization: Rethinking Batch Normalization for Online\n Continual Learning"], "answer_arxiv_id": ["1911.07053", "1905.13260", "2004.00440", "2112.04731", "2203.16102"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_10333"} +{"question": "What are some examples of latest text-to-image synthesis models built on diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Improved Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2102.09672", "2112.10752", "2205.11487", "2204.06125", "2111.14822", "2302.05543"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_10334"} +{"question": "Which work can be considered as the start of the research on 2D image segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_10335"} +{"question": "Which study contributes the human-preferences datasets about helpfulness and harmlessness?", "answer": ["Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback"], "answer_arxiv_id": ["2204.05862"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_10336"} +{"question": "Which works showed the potential benefit of using the embeddings of pretrained networks for object localization?", "answer": ["Object discovery and representation networks", "Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization", "Self-Supervised Transformers for Unsupervised Object Discovery using Normalized Cut", "Localizing Objects with Self-Supervised Transformers and no Labels", "Large-Scale Unsupervised Object Discovery", "Toward Unsupervised, Multi-Object Discovery in Large-Scale Image Collections"], "answer_arxiv_id": ["2203.08777", "2205.07839", "2202.11539", "2109.14279", "2106.06650", "2007.02662"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_10337"} +{"question": "Which works introduced MatFuse and ControlMat for material generation?", "answer": ["MatFuse: Controllable Material Generation with Diffusion Models", "ControlMat: A Controlled Generative Approach to Material Capture"], "answer_arxiv_id": ["2308.11408", "2309.01700"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_10338"} +{"question": "What works address large action spaces by growing them iteratively from smaller versions?", "answer": ["Growing Action Spaces"], "answer_arxiv_id": ["1906.12266"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_10339"} +{"question": "What papers propose techniques for learning soft textual prompts for adaption in CLIP?", "answer": ["Learning to Prompt for Vision-Language Models", "Visual Prompt Tuning", "MaPLe: Multi-modal Prompt Learning", "Conditional Prompt Learning for Vision-Language Models", "Visual-Language Prompt Tuning with Knowledge-guided Context Optimization"], "answer_arxiv_id": ["2109.01134", "2203.12119", "2210.03117", "2203.05557", "2303.13283"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_10340"} +{"question": "Which works use query-based methods for pose estimation?", "answer": ["Poseur: Direct Human Pose Regression with Transformers", "ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation"], "answer_arxiv_id": ["2201.07412", "2204.12484"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_10341"} +{"question": "Which papers discuss the theoretical aspects of how meta-learning leads to predictors that perform amortized Bayesian inference?", "answer": ["Meta-learning of Sequential Strategies"], "answer_arxiv_id": ["1905.03030"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_10342"} +{"question": "Could you provide me some works that modify the forward rasterization step to make the pipeline differentiable?", "answer": ["A Versatile Scene Model with Differentiable Visibility Applied to Generative Pose Estimation", "Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning", "Pix2Vex: Image-to-Geometry Reconstruction using a Smooth Differentiable Renderer"], "answer_arxiv_id": ["1602.03725v1", "1904.01786", "1903.11149"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_10343"} +{"question": "Which works modify standard multi-headed attention in Transformer architectures to integrate tree information?", "answer": ["Tree Transformer: Integrating Tree Structures into Self-Attention", "Transformer Grammars: Augmenting Transformer Language Models with Syntactic Inductive Biases at Scale"], "answer_arxiv_id": ["1909.06639", "2203.00633"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10344"} +{"question": "Could you mention papers related to the construction of tree-shaped neural networks for NLP tasks?", "answer": ["Improved Semantic Representations From Tree-Structured Long Short-Term Memory Networks", "Recurrent Neural Network Grammars", "Neural Module Networks", "Ordered Neurons: Integrating Tree Structures into Recurrent Neural Networks"], "answer_arxiv_id": ["1503.00075", "1602.07776", "1511.02799v4", "1810.09536"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_10345"} +{"question": "What research has shown instruction tuning to improve the zero-shot generalization of language models to unseen tasks?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization", "Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2110.08207", "2109.01652"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_10346"} +{"question": "What work applies a diffusion model as a trajectory generator?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_10347"} +{"question": "Any papers that express skepticism about the possibility of LMs learning causal strategies from passive data?", "answer": ["Shaking the foundations: delusions in sequence models for interaction and control", "Theoretical Impediments to Machine Learning With Seven Sparks from the Causal Revolution"], "answer_arxiv_id": ["2110.10819", "1801.04016"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10348"} +{"question": "What papers discuss the use of pointmaps in Visual Localization tasks?", "answer": ["DSAC - Differentiable RANSAC for Camera Localization", "Learning Less is More - 6D Camera Localization via 3D Surface Regression", "Visual Camera Re-Localization from RGB and RGB-D Images Using DSAC", "Learning Camera Localization via Dense Scene Matching", "SACReg: Scene-Agnostic Coordinate Regression for Visual Localization"], "answer_arxiv_id": ["1611.05705", "1711.10228", "2002.12324", "2103.16792", "2307.11702"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_10349"} +{"question": "Which studies focus on the second paradigm of augmenting LLMs with tools by giving in-context tool descriptions and demonstrations?", "answer": ["Tool Documentation Enables Zero-Shot Tool-Usage with Large Language\n Models", "Augmented Language Models: a Survey", "Gorilla: Large Language Model Connected with Massive APIs"], "answer_arxiv_id": ["2308.00675", "2302.07842", "2305.15334"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_10350"} +{"question": "Could you provide me some studies about Minimum Bayes Risk (MBR) for reranking approach in NMT systems?", "answer": ["Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural\n Machine Translation"], "answer_arxiv_id": ["2005.10283"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_10351"} +{"question": "In which works were advanced gating functions proposed to alleviate load imbalance, improve speed, and optimize downstream generalization?", "answer": ["Hash Layers For Large Sparse Models", "Tricks for Training Sparse Translation Models", "Taming Sparsely Activated Transformer with Stochastic Experts", "Hard Mixtures of Experts for Large Scale Weakly Supervised Vision", "Mixture-of-Experts with Expert Choice Routing", "Sparse is Enough in Scaling Transformers"], "answer_arxiv_id": ["2106.04426", "2110.08246", "2110.04260", "1704.06363", "2202.09368", "2111.12763"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_10352"} +{"question": "What are some research papers that used 1D-Shapley for feature-based interpretability for black-box models both locally and globally?", "answer": ["A Unified Approach to Interpreting Model Predictions", "Understanding Global Feature Contributions With Additive Importance Measures"], "answer_arxiv_id": ["1705.07874", "2004.00668"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_10353"} +{"question": "Could you list some works that apply patch-wise self-supervised learning for memory reduction in handling large images?", "answer": ["Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology", "Scaling Vision Transformers to Gigapixel Images via Hierarchical Self-Supervised Learning", "Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learning"], "answer_arxiv_id": ["2012.03583", "2206.02647", "2011.08939"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_10354"} +{"question": "Could you provide me papers that studied the interaction between entropy regularization and policy optimization in Markov games?", "answer": ["Regularized Gradient Descent Ascent for Two-Player Zero-Sum Markov Games"], "answer_arxiv_id": ["2205.13746"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_10355"} +{"question": "What works have contributed to task representation for soft body manipulation?", "answer": ["Learning to Manipulate Deformable Objects without Demonstrations", "DiffSkill: Skill Abstraction from Differentiable Physics for Deformable Object Manipulations with Tools", "Learning Deformable Object Manipulation from Expert Demonstrations", "RoboCraft: Learning to See, Simulate, and Shape Elasto-Plastic Objects with Graph Networks", "3D Neural Scene Representations for Visuomotor Control", "AutoBag: Learning to Open Plastic Bags and Insert Objects", "Autonomously Untangling Long Cables", "DexDeform: Dexterous Deformable Object Manipulation with Human Demonstrations and Differentiable Physics", "Learning Deformable Object Manipulation from Expert Demonstrations", "Off-Policy Deep Reinforcement Learning without Exploration", "Dynamics Learning with Cascaded Variational Inference for Multi-Step Manipulation", "Skill-based Model-based Reinforcement Learning", "Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learning", "Learning Latent Plans from Play", "Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Decision Transformer: Reinforcement Learning via Sequence Modeling", "An Optimistic Perspective on Offline Reinforcement Learning", "Is Conditional Generative Modeling all you need for Decision-Making?", "From One Hand to Multiple Hands: Imitation Learning for Dexterous Manipulation from Single-Camera Teleoperation", "Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware", "DexDeform: Dexterous Deformable Object Manipulation with Human Demonstrations and Differentiable Physics", "Scene Editing as Teleoperation: A Case Study in 6DoF Kit Assembly"], "answer_arxiv_id": ["1910.13439", "2203.17275", "2207.10148", "2205.02909", "2107.04004", "2210.17217", "2207.07813", "2304.03223", "2207.10148", "1812.02900", "1910.13395", "2207.07560", "1910.11956", "1903.01973", "2106.02039", "2106.01345", "1907.04543v4", "2211.15657", "2204.12490", "2304.13705", "2304.03223", "2110.04450v4"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_10356"} +{"question": "Which studies discuss effective fine-tuning approaches similar to our work?", "answer": ["Finetune like you pretrain: Improved finetuning of zero-shot vision models", "CLIP Itself is a Strong Fine-tuner: Achieving 85.7% and 88.0% Top-1 Accuracy with ViT-B and ViT-L on ImageNet"], "answer_arxiv_id": ["2212.00638", "2212.06138"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10357"} +{"question": "Which papers delved into neural representations like meshes and point clouds for representing 3D scenes?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Point-NeRF: Point-based Neural Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2201.08845"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_10358"} +{"question": "What works introduce the concept of Concept Activation Vectors (CAVs)?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with\n Concept Activation Vectors (TCAV)"], "answer_arxiv_id": ["1711.11279"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_10359"} +{"question": "Which works are about modifying the attention mechanism?", "answer": ["LongNet: Scaling Transformers to 1,000,000,000 Tokens"], "answer_arxiv_id": ["2307.02486"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_10360"} +{"question": "What papers implicitly address the task of modular customization by representing each concept through unique token embeddings?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models"], "answer_arxiv_id": ["2208.01618", "2211.09794"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_10361"} +{"question": "In what papers the researchers propose truncated-based specialized mechanisms for providing differential privacy guarantees?", "answer": ["Finite Sample Differentially Private Confidence Intervals", "Private and polynomial time algorithms for learning Gaussians and beyond", "A Fast Algorithm for Adaptive Private Mean Estimation", "Generalization for Adaptively-chosen Estimators via Stable Median"], "answer_arxiv_id": ["1711.03908", "2111.11320", "2301.07078v1", "1706.05069"], "source_meta": {"published_time": "20210620"}, "qid": "AutoScholarQuery_train_10362"} +{"question": "What studies demonstrate how ensemble attacks can boost adversarial attacks?", "answer": ["Delving into Transferable Adversarial Examples and Black-box Attacks", "Boosting Adversarial Attacks with Momentum", "Stochastic Variance Reduced Ensemble Adversarial Attack for Boosting the Adversarial Transferability"], "answer_arxiv_id": ["1611.02770", "1710.06081", "2111.10752"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_10363"} +{"question": "Which studies related to using left and right context for infilling sequences in application like editing sentences, restoring ancient text, and fixing bugs in source code?", "answer": ["XL-Editor: Post-editing Sentences with XLNet", "Restoring ancient text using deep learning: a case study on Greek epigraphy"], "answer_arxiv_id": ["1910.10479", "1910.06262v1"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_10364"} +{"question": "Which works have contributed to the development of Bird’s-Eye-View (BEV) representation in multi-view perception for autonomous driving?", "answer": ["BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers", "PETR: Position Embedding Transformation for Multi-View 3D Object\n Detection", "BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object\n Detection", "Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D\n Object Detection", "Far3D: Expanding the Horizon for Surround-view 3D Object Detection"], "answer_arxiv_id": ["2203.17270", "2203.05625", "2206.10092", "2303.11926", "2308.09616"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_10365"} +{"question": "What research papers are primarily related to RL with unsupervised-learning oracles?", "answer": ["Provably efficient RL with Rich Observations via Latent State Decoding", "Provably Efficient Exploration for Reinforcement Learning Using Unsupervised Learning"], "answer_arxiv_id": ["1901.09018", "2003.06898v4"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_10366"} +{"question": "What works utilize text guidance for image generation during training?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_10367"} +{"question": "What studies have been done to improve NeRF’s rendering quality under extremely sparse views?", "answer": ["SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image", "RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from Sparse Inputs"], "answer_arxiv_id": ["2204.00928", "2112.00724"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_10368"} +{"question": "Can you provide works that proposed a general framework to learn the structured sparsity mask on a pre-trained model?", "answer": ["NxMTransformer: Semi-Structured Sparsification for Natural Language Understanding via ADMM"], "answer_arxiv_id": ["2110.15766"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10369"} +{"question": "What are some works in SSDA that align features between domains for knowledge transfer?", "answer": ["Learning Invariant Representations and Risks for Semi-supervised Domain\n Adaptation", "Attract, Perturb, and Explore: Learning a Feature Alignment Network for\n Semi-supervised Domain Adaptation", "CLDA: Contrastive Learning for Semi-Supervised Domain Adaptation", "Semi-supervised Domain Adaptation via Sample-to-Sample Self-Distillation", "Multi-level Consistency Learning for Semi-supervised Domain Adaptation", "ECACL: A Holistic Framework for Semi-Supervised Domain Adaptation"], "answer_arxiv_id": ["2010.04647", "2007.09375", "2107.00085", "2111.14353", "2205.04066", "2104.09136"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_10370"} +{"question": "What references studied the concept of meta reinforcement learning?", "answer": ["RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "Learning to reinforcement learn"], "answer_arxiv_id": ["1611.02779", "1611.05763"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_10371"} +{"question": "What work exists on the development of neural SDEs, where an SDE’s drift and diffusion coefficient are parametrized by neural networks?", "answer": ["Neural Stochastic Differential Equations: Deep Latent Gaussian Models in the Diffusion Limit", "Theoretical guarantees for sampling and inference in generative models with latent diffusions", "Neural Jump Stochastic Differential Equations", "Neural SDE: Stabilizing Neural ODE Networks with Stochastic Noise", "Score-Based Generative Modeling through Stochastic Differential Equations", "Identifying Latent Stochastic Differential Equations"], "answer_arxiv_id": ["1905.09883", "1903.01608", "1905.10403", "1906.02355", "2011.13456", "2007.06075"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_10372"} +{"question": "Any works about generation performance presented by casting detection as a grounding problem via object-word fusion and alignment?", "answer": ["Grounded Language-Image Pre-training", "Universal Instance Perception as Object Discovery and Retrieval", "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection"], "answer_arxiv_id": ["2112.03857", "2303.06674", "2303.05499"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10373"} +{"question": "What studies focus on pixel relation modeling for semantic segmentation?", "answer": ["GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond", "Interlaced Sparse Self-Attention for Semantic Segmentation", "ISNet: Integrate Image-Level and Semantic-Level Context for Semantic Segmentation", "Unified Perceptual Parsing for Scene Understanding", "Non-local Neural Networks", "Disentangled Non-Local Neural Networks", "Multi-Scale Context Aggregation by Dilated Convolutions", "Pyramid Scene Parsing Network"], "answer_arxiv_id": ["1904.11492", "1907.12273v2", "2108.12382", "1807.10221", "1711.07971", "2006.06668", "1511.07122", "1612.01105"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_10374"} +{"question": "Which work initiates the research field of 'food computing'?", "answer": ["A Survey on Food Computing"], "answer_arxiv_id": ["1808.07202"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_10375"} +{"question": "Could you provide me some research on regularization-based methods for managing the catastrophic forgetting problem?", "answer": ["Overcoming catastrophic forgetting in neural networks"], "answer_arxiv_id": ["1612.00796"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_10376"} +{"question": "Who demonstrated an instance of loss of generalization in pre-training a network?", "answer": ["On Warm-Starting Neural Network Training"], "answer_arxiv_id": ["1910.08475"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_10377"} +{"question": "Which studies instigated a shift toward digital avatars with real-time face tracking and authentic face reenactment?", "answer": ["Face2Face: Real-time Face Capture and Reenactment of RGB Videos"], "answer_arxiv_id": ["2007.14808"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10378"} +{"question": "What work projects the fine-tuned weight for every gradient descent update so that it lies within a sphere centered on the initial pre-trained weights?", "answer": ["Distance-Based Regularisation of Deep Networks for Fine-Tuning"], "answer_arxiv_id": ["2002.08253"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_10379"} +{"question": "What works extend FNO to irregular grids?", "answer": ["Fourier Neural Operator with Learned Deformations for PDEs on General Geometries", "NUNO: A General Framework for Learning Parametric PDEs with Non-Uniform Data"], "answer_arxiv_id": ["2207.05209", "2305.18694"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_10380"} +{"question": "Can you identify works that introduced new benchmarks for complex multi-modal tasks?", "answer": ["MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language\n Models", "MMBench: Is Your Multi-modal Model an All-around Player?", "MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities", "SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension", "LAMM: Language-Assisted Multi-Modal Instruction-Tuning Dataset,\n Framework, and Benchmark"], "answer_arxiv_id": ["2306.13394", "2307.06281", "2308.02490", "2307.16125", "2306.06687"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_10381"} +{"question": "What research has been done on data-free pruning using neuron similarity measurements?", "answer": ["Data-free Parameter Pruning for Deep Neural Networks", "RED++ : Data-Free Pruning of Deep Neural Networks via Input Splitting and Output Merging"], "answer_arxiv_id": ["1507.06149", "2110.01397"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_10382"} +{"question": "Could you tell me about the research that developed a policy network for CRS to decide at each conversation turn?", "answer": ["Conversational Recommender System"], "answer_arxiv_id": ["1806.03277"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_10383"} +{"question": "Which studies explored the interplay between the size, structure, and performance of deep RL agents?", "answer": ["D2RL: Deep Dense Architectures in Reinforcement Learning", "Training Larger Networks for Deep Reinforcement Learning"], "answer_arxiv_id": ["2010.09163", "2102.07920"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10384"} +{"question": "Could you provide me some works about enhancing the robustness of DNNs against label noise?", "answer": ["Learning from Noisy Labels with Deep Neural Networks: A Survey", "Masking: A New Perspective of Noisy Supervision", "Deep Learning from Noisy Image Labels with Quality Embedding", "Robust Loss Functions under Label Noise for Deep Neural Networks", "Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe Noise", "Normalized Loss Functions for Deep Learning with Noisy Labels", "MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels", "Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels", "Combating Noisy Labels by Agreement: A Joint Training Method with Co-Regularization", "A Topological Filter for Learning with Label Noise"], "answer_arxiv_id": ["2007.08199", "1805.08193", "1711.00583", "1712.09482", "1802.05300", "2006.13554", "1712.05055", "1804.06872", "2003.02752", "2012.04835"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_10385"} +{"question": "Which papers have leveraged spectral graph neural networks, global graph framelet convolution, and graph wavelets for better dynamic graph representations?", "answer": ["Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting", "Learnable Spectral Wavelets on Dynamic Graphs to Capture Global Interactions"], "answer_arxiv_id": ["2103.07719", "2211.11979"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_10386"} +{"question": "What examples show the use of GPTs in domains outside of event stream data, such as protein sequences or point clouds?", "answer": ["Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling"], "answer_arxiv_id": ["2111.14819"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_10387"} +{"question": "What studies have explored the approach of learning Bayes-optimal agents with a history-based representation using a variational method?", "answer": ["VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning", "Exploration in Approximate Hyper-State Space for Meta Reinforcement Learning"], "answer_arxiv_id": ["1910.08348", "2010.01062"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_10388"} +{"question": "In which studies does it suggest that the activation patterns of neurons significantly influence the behavior of Large Language Models (LLMs)?", "answer": ["Transformer Feed-Forward Layers Build Predictions by Promoting Concepts\n in the Vocabulary Space"], "answer_arxiv_id": ["2203.14680"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_10389"} +{"question": "What are the papers discussing about the ability of Diffusion Probabilistic Models to solve posterior inference problems?", "answer": ["Improving Diffusion Models for Inverse Problems using Manifold Constraints", "Diffusion Posterior Sampling for General Noisy Inverse Problems"], "answer_arxiv_id": ["2206.00941", "2209.14687"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_10390"} +{"question": "Can you mention any research papers focused on personalized fine-tuning of diffusion models?", "answer": ["Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2210.09276"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_10391"} +{"question": "What work proposed to dynamically learn the segmentation in byte sequences as a solution to the problem of choosing the appropriate context window?", "answer": ["Efficient Transformers with Dynamic Token Pooling"], "answer_arxiv_id": ["2211.09761"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10392"} +{"question": "Are there any research works that addressed the limitations of ControlNet, GLIGEN and T2I-Adapter while following the line of fine-tuning adapters?", "answer": ["T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.08453"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10393"} +{"question": "Which studies provide a method assuming specific transformations on the spurious attributes when they are not known?", "answer": ["Neural Networks for Learning Counterfactual G-Invariances from Single Environments"], "answer_arxiv_id": ["2104.10105"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_10394"} +{"question": "What studies employed VNN for policy learning and non-rigid object manipulation in robotics?", "answer": ["EquivAct: SIM(3)-Equivariant Visuomotor Policies beyond Rigid Object\n Manipulation"], "answer_arxiv_id": ["2310.16050"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_10395"} +{"question": "Can you inform me about any recent research that demonstrates better performance for adversarial attack applications by combining the meta-learning design?", "answer": ["Learning to Generate Image Source-Agnostic Universal Adversarial Perturbations"], "answer_arxiv_id": ["2009.13714"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_10396"} +{"question": "Could you list out some research that uses the 'feel-good' modification of the Thompson sampling algorithm?", "answer": ["Feel-Good Thompson Sampling for Contextual Bandits and Reinforcement Learning"], "answer_arxiv_id": ["2110.00871"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_10397"} +{"question": "Could you mention some works that proposed additional data augmentation for improving contrastive learning?", "answer": ["On Compositions of Transformations in Contrastive Self-Supervised Learning", "Multimodal Self-Supervised Learning of General Audio Representations"], "answer_arxiv_id": ["2003.04298", "2104.12807"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_10398"} +{"question": "What studies have proposed to lift text-image diffusion models for open-vocabulary 3D generation in Text-to-3D Generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "EfficientDreamer: High-Fidelity and Robust 3D Creation via\n Orthogonal-view Diffusion Prior", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation"], "answer_arxiv_id": ["2112.10752", "2308.13223", "2305.16213"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_10399"} +{"question": "Which works highlighted the limitations of invariant learning methods in the context of domain generalization?", "answer": ["In Search of Lost Domain Generalization", "Wilds: A Benchmark of in-the-Wild Distribution Shifts", "The Risks of Invariant Risk Minimization"], "answer_arxiv_id": ["2007.01434", "2012.07421", "2010.05761"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_10400"} +{"question": "What studies have been done on partial model personalization in the context of personalized federated learning?", "answer": ["Federated Learning with Partial Model Personalization", "Exploiting Shared Representations for Personalized Federated Learning", "Personalized Federated Learning with Feature Alignment and Classifier\n Collaboration"], "answer_arxiv_id": ["2204.03809", "2102.07078", "2306.11867"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_10401"} +{"question": "Which works propose algorithms for explicit learning of the DAG structure?", "answer": ["Consistency Guarantees for Greedy Permutation-Based Causal Inference Algorithms", "Differentiable Causal Discovery from Interventional Data"], "answer_arxiv_id": ["1702.03530", "2007.01754"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_10402"} +{"question": "Which work introduced a simple approach for the Massart setting?", "answer": ["Learning Halfspaces with Massart Noise Under Structured Distributions"], "answer_arxiv_id": ["2002.05632"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_10403"} +{"question": "Which studies discuss methods that directly learn gaze from face images?", "answer": ["GazeOnce: Real-Time Multi-Person Gaze Estimation", "PureGaze: Purifying Gaze Feature for Generalizable Gaze Estimation"], "answer_arxiv_id": ["2204.09480", "2103.13173"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_10404"} +{"question": "Which studies provide the fundamentals of randomized multilevel Monte Carlo (rMLMC) framework?", "answer": ["A general method for debiasing a Monte Carlo estimator"], "answer_arxiv_id": ["1005.2228"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_10405"} +{"question": "What are some works proposing solutions for the oversmoothing issue in GNNs?", "answer": ["Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning", "Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "Towards Deeper Graph Neural Networks"], "answer_arxiv_id": ["1801.07606", "1810.05997", "2007.09296"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_10406"} +{"question": "Which works also employ LLMs for zero-shot commonsense priors in the Housekeep environment?", "answer": ["Housekeep: Tidying Virtual Households using Commonsense Reasoning"], "answer_arxiv_id": ["2205.10712"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_10407"} +{"question": "Could you reference any studies related to online prediction in control settings?", "answer": ["Learning Linear Dynamical Systems via Spectral Filtering", "Sample Complexity of Kalman Filtering for Unknown Systems"], "answer_arxiv_id": ["1711.00946", "1912.12309"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_10408"} +{"question": "Any works concentrating on learning input-dependent dynamic architectures?", "answer": ["Dynamic Neural Networks: A Survey", "DSelect-k: Differentiable Selection in the Mixture of Experts with Applications to Multi-Task Learning", "Deep Elastic Networks with Model Selection for Multi-Task Learning", "Routing Networks: Adaptive Selection of Non-linear Functions for Multi-Task Learning"], "answer_arxiv_id": ["2102.04906", "2106.03760", "1909.04860", "1711.01239"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_10409"} +{"question": "Which works talk about self-supervised strategy in 3D model training using reconstruction of missing information that is either intentionally hidden or intrinsically missing due to sparse point sampling in Lidar acquisition?", "answer": ["Masked Autoencoder for Self-Supervised Pre-training on Lidar Point\n Clouds", "Masked Autoencoders for Point Cloud Self-supervised Learning", "Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud\n Pre-training", "Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point\n Modeling", "ALSO: Automotive Lidar Self-supervision by Occupancy estimation"], "answer_arxiv_id": ["2207.00531", "2203.06604", "2205.14401", "2111.14819", "2212.05867"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_10410"} +{"question": "Are there any studies about combining self-supervised and supervised learning?", "answer": ["Class-Aware Contrastive Semi-Supervised Learning", "Supervised Contrastive Learning"], "answer_arxiv_id": ["2203.02261", "2004.11362"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_10411"} +{"question": "Could you provide me some works that includes innovations in layout control and attention modules tailored for bounding boxes?", "answer": ["GLIGEN: Open-Set Grounded Text-to-Image Generation", "LayoutDiffuse: Adapting Foundational Diffusion Models for\n Layout-to-Image Generation"], "answer_arxiv_id": ["2301.07093", "2302.08908"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_10412"} +{"question": "Any works that studied a stochastic linear bandits with informed feedback graphs?", "answer": ["Contextual Bandits with Side-Observations"], "answer_arxiv_id": ["2006.03951v2"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_10413"} +{"question": "What research works provide quantitative benchmarks for gradient-based explanations?", "answer": ["RISE: Randomized Input Sampling for Explanation of Black-box Models", "A Benchmark for Interpretability Methods in Deep Neural Networks", "FastSHAP: Real-Time Shapley Value Estimation"], "answer_arxiv_id": ["1806.07421", "1806.10758", "2107.07436"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_10414"} +{"question": "What research works capture visual temporal dynamics from 3D CNN backbone for dynamic scene graph generation?", "answer": ["Target Adaptive Context Aggregation for Video Scene Graph Generation", "Unified Graph Structured Models for Video Understanding"], "answer_arxiv_id": ["2108.08121", "2103.15662"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_10415"} +{"question": "What paper introduced CRAFT and its application in identifying CAVs and localizing the most relevant input regions for each CAV?", "answer": ["CRAFT: Concept Recursive Activation FacTorization for Explainability"], "answer_arxiv_id": ["2211.10154"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_10416"} +{"question": "Which publication showed a correlation between transferability of adversarial images and interactions?", "answer": ["A Unified Approach to Interpreting and Boosting Adversarial\n Transferability"], "answer_arxiv_id": ["2010.04055"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_10417"} +{"question": "Could you provide me some works that used the encoder-decoder structure in sketching models?", "answer": ["Learning to Sketch with Shortcut Cycle Consistency", "SketchEmbedNet: Learning Novel Concepts by Imitating Drawings"], "answer_arxiv_id": ["1805.00247", "2009.04806"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_10418"} +{"question": "What works explore using implicit differentiation to estimate the Jacobian of problem solutions?", "answer": ["Deep Equilibrium Models", "Differential Properties of Sinkhorn Approximation for Learning with Wasserstein Distance", "Nonsmooth Implicit Differentiation for Machine Learning and Optimization", "Implicit differentiation of Lasso-type models for hyperparameter optimization"], "answer_arxiv_id": ["1909.01377", "1805.11897", "2106.04350", "2002.08943"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_10419"} +{"question": "Which works store local features in a hash table when using feature-based approaches to represent the scene in geometric shapes?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_10420"} +{"question": "What research is done about sketching techniques for polynomial kernels?", "answer": ["Near Input Sparsity Time Kernel Embeddings via Adaptive Sampling", "Oblivious Sketching of High-Degree Polynomial Kernels", "Fast Sketching of Polynomial Kernels of Polynomial Degree"], "answer_arxiv_id": ["2007.03927", "1909.01410v5", "2108.09420v1"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_10421"} +{"question": "Which studies found diffusion models to be a powerful generative tool that tend to outperform previous generative methods such as Gaussians or VAEs in terms of behavior modeling?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis", "Know Your Boundaries: The Necessity of Explicit Behavioral Cloning in Offline RL", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "Offline Reinforcement Learning via High-Fidelity Generative Behavior Modeling", "Is Conditional Generative Modeling all you need for Decision-Making?", "Imitating Human Behaviour with Diffusion Models"], "answer_arxiv_id": ["2205.09991", "2206.00695", "2208.06193", "2209.14548", "2211.15657", "2301.10677"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_10422"} +{"question": "What works describe adversarial perturbations' effects on the loss function?", "answer": ["Adversarial Weight Perturbation Helps Robust Generalization"], "answer_arxiv_id": ["2004.05884"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_10423"} +{"question": "Could you provide me some works that rely on parameter efficient methods to extend the capabilities of LMs?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer", "Mini-Model Adaptation: Efficiently Extending Pretrained Models to New\n Languages via Aligned Shallow Training"], "answer_arxiv_id": ["1902.00751", "2005.00052", "2212.10503"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_10424"} +{"question": "What works have studied the convergence of gradient decent for ReLU activated neural networks?", "answer": ["A Convergence Theory for Deep Learning via Over-Parameterization", "An Improved Analysis of Training Over-parameterized Deep Neural Networks"], "answer_arxiv_id": ["1811.03962", "1906.04688"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_10425"} +{"question": "What studies show evidence of MLLMs performing well in vision-centric tasks such as object detection and segmentation?", "answer": ["VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "NExT-Chat: An LMM for Chat, Detection and Segmentation", "u-LLaVA: Unifying Multi-Modal Tasks via Large Language Model", "GLaMM: Pixel Grounding Large Multimodal Model", "SPHINX: The Joint Mixing of Weights, Tasks, and Visual Embeddings for\n Multi-modal Large Language Models"], "answer_arxiv_id": ["2305.11175", "2311.04498", "2311.05348", "2311.03356", "2311.07575"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_10426"} +{"question": "Could you provide me some papers on regularization methods like KL control?", "answer": ["A Theory of Regularized Markov Decision Processes", "Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor", "Maximum a Posteriori Policy Optimisation", "Behavior Regularized Offline Reinforcement Learning", "Way Off-Policy Batch Deep Reinforcement Learning of Implicit Human Preferences in Dialog", "Offline Reinforcement Learning as Anti-Exploration"], "answer_arxiv_id": ["1901.11275", "1801.01290", "1806.06920", "1911.11361", "1907.00456", "2106.06431"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_10427"} +{"question": "Could you provide me some studies investigating CNN variants with a recurrent layer?", "answer": ["Deep Learning Models of the Retinal Response to Natural Scenes"], "answer_arxiv_id": ["1702.01825"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_10428"} +{"question": "Which work is related to the introduction of the Objaverse 1.0 dataset, comprising 800K 3D models of high quality and diverse textures, geometry, and object types?", "answer": ["Objaverse: A Universe of Annotated 3D Objects"], "answer_arxiv_id": ["2212.08051"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_10429"} +{"question": "What papers discussed about various machine learning methods for premise selection from classical models to Transformers?", "answer": ["Premise Selection for Mathematics by Corpus Analysis and Kernel Methods", "Machine-Learned Premise Selection for Lean", "DeepMath - Deep Sequence Models for Premise Selection", "Learning to Prove Theorems by Learning to Generate Theorems", "Magnushammer: A Transformer-based Approach to Premise Selection", "CoProver: A Recommender System for Proof Construction"], "answer_arxiv_id": ["1108.3446", "2304.00994v2", "1606.04442", "2002.07019", "2303.04488", "2304.10486"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_10430"} +{"question": "What study proposed a differentiable ray tracing method combined with deep learning for the learning-based inverse rendering of indoor scenes?", "answer": ["Learning-based Inverse Rendering of Complex Indoor Scenes with Differentiable Monte Carlo Raytracing"], "answer_arxiv_id": ["2211.03017"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_10431"} +{"question": "What work introduced the idea of zero-shot transfer to novel classification tasks?", "answer": ["Learning Visual N-Grams from Web Data"], "answer_arxiv_id": ["1612.09161"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_10432"} +{"question": "Which paper introduced inter-penetration constraints in PROX to prevent collisions?", "answer": ["Resolving 3D Human Pose Ambiguities with 3D Scene Constraints"], "answer_arxiv_id": ["1908.06963"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_10433"} +{"question": "Could you provide me some works that count objects using the zero-shot method in class-agnostic object counting?", "answer": ["Zero-shot Object Counting", "CLIP-Count: Towards Text-Guided Zero-Shot Object Counting"], "answer_arxiv_id": ["2303.02001", "2305.07304"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_10434"} +{"question": "What research initiatives have addressed procedure planning in instructional videos recently?", "answer": ["Procedure Planning in Instructional Videos", "PlaTe: Visually-Grounded Planning with Transformers in Procedural Tasks", "Procedure Planning in Instructional Videos via Contextual Modeling and\n Model-based Policy Learning"], "answer_arxiv_id": ["1907.01172", "2109.04869", "2110.01770"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_10435"} +{"question": "Can you list the optimization-based methods for meta-learning-based few-shot learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-Learning with Differentiable Convex Optimization", "Meta-Learning with Latent Embedding Optimization"], "answer_arxiv_id": ["1703.03400", "1904.03758", "1807.05960"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_10436"} +{"question": "Are there any research papers that focus on learning different state-space geometric structures?", "answer": ["Scalable methods for computing state similarity in deterministic Markov Decision Processes", "Learning Invariant Representations for Reinforcement Learning without Reconstruction", "A Geometric Perspective on Optimal Representations for Reinforcement Learning", "Predictive Information Accelerates Learning in RL", "Learning Task Informed Abstractions", "Denoised MDPs: Learning World Models Better Than the World Itself"], "answer_arxiv_id": ["1911.09291", "2006.10742", "1901.11530", "2007.12401", "2106.15612", "2206.15477"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_10437"} +{"question": "Which research papers have worked with Adobe 3D Assets in the context of acquisition, generation, retrieval and captioning?", "answer": ["TileGen: Tileable, Controllable Material Generation and Capture", "ControlMat: A Controlled Generative Approach to Material Capture", "The Visual Language of Fabrics"], "answer_arxiv_id": ["2206.05649", "2309.01700", "2307.13681"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_10438"} +{"question": "Which work explores the use of a surrogate reward in active recognition?", "answer": ["Learning to View: Decision Transformers for Active Object Detection"], "answer_arxiv_id": ["2301.09544"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_10439"} +{"question": "What papers are about debiasing using an alternative network technical approach?", "answer": ["Learning from Failure: Training Debiased Classifier from Biased Classifier", "Learning Debiased Classifier with Biased Committee"], "answer_arxiv_id": ["2007.02561", "2206.10843"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10440"} +{"question": "Could you provide me some works about deep-learning-based approaches for RRSR?", "answer": ["Super-Resolution by Predicting Offsets: An Ultra-Efficient\n Super-Resolution Network for Rasterized Images"], "answer_arxiv_id": ["2210.04198"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_10441"} +{"question": "Which research identifies challenges in T2I style transfer such as content from style references overshadowing the textual context?", "answer": ["StyleAdapter: A Single-Pass LoRA-Free Model for Stylized Image\n Generation"], "answer_arxiv_id": ["2309.01770"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10442"} +{"question": "Which works demonstrated the success of diffusion models in image generation?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Cascaded Diffusion Models for High Fidelity Image Generation", "Palette: Image-to-Image Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models"], "answer_arxiv_id": ["2102.09672", "2106.15282", "2111.05826", "2210.02303"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_10443"} +{"question": "Could you provide me with studies that predict bounding-boxes to locate objects?", "answer": ["SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition", "Generative Neurosymbolic Machines"], "answer_arxiv_id": ["2001.02407", "2010.12152"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_10444"} +{"question": "Which work proposes to leverage multiple visual features at different granularities during the pre-training phase in unsupervised medical visual representation learning?", "answer": ["Multi-Granularity Cross-modal Alignment for Generalized Medical Visual\n Representation Learning"], "answer_arxiv_id": ["2210.06044"], "source_meta": {"published_time": "20240203"}, "qid": "AutoScholarQuery_train_10445"} +{"question": "Can you provide the paper where the UNet is utilized in Part-A2-Net?", "answer": ["From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network", "U-Net: Convolutional Networks for Biomedical Image Segmentation"], "answer_arxiv_id": ["1907.03670", "1505.04597"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_10446"} +{"question": "Could you list the papers about text-based bot detection techniques including NLP?", "answer": ["Twitter Bot Detection Using Bidirectional Long Short-term Memory Neural\n Networks and Word Embeddings", "Deep Neural Networks for Bot Detection", "SATAR: A Self-supervised Approach to Twitter Account Representation\n Learning and its Application in Bot Detection"], "answer_arxiv_id": ["2002.01336", "1802.04289", "2106.13089"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_10447"} +{"question": "Which studies were focused on novel view synthesis?", "answer": ["ARF: Artistic Radiance Fields", "Ref-NPR: Reference-Based Non-Photorealistic Radiance Fields for Controllable Scene Stylization", "SNeRF: Stylized Neural Implicit Representations for 3D Scenes"], "answer_arxiv_id": ["2206.06360", "2212.02766", "2207.02363"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10448"} +{"question": "Which works developed new hypergraph descriptors?", "answer": ["Hypernetwork Science via High-Order Hypergraph Walks"], "answer_arxiv_id": ["1906.11295v2"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_10449"} +{"question": "Which early dataset has been developed for Table Question Answering?", "answer": ["TabMCQ: A Dataset of General Knowledge Tables and Multiple-choice Questions"], "answer_arxiv_id": ["1602.03960"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_10450"} +{"question": "What work trains a stochastic sampler for image deblurring?", "answer": ["Deblurring via Stochastic Refinement"], "answer_arxiv_id": ["2112.02475"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_10451"} +{"question": "What papers discuss the model exploitation problem in model-based reinforcement learning?", "answer": ["Agnostic System Identification for Model-Based Reinforcement Learning", "When to Trust Your Model: Model-Based Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning", "Lyapunov Density Models: Constraining Distribution Shift in Learning-Based Control"], "answer_arxiv_id": ["1203.1007", "1906.08253", "2005.05951", "2206.10524"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_10452"} +{"question": "Which papers discuss the use of normalization layers such as Batch Normalization in deep neural network architectures?", "answer": ["Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift", "Group Normalization", "Instance Normalization: The Missing Ingredient for Fast Stylization", "Layer Normalization"], "answer_arxiv_id": ["1502.03167", "1803.08494", "1607.08022", "1607.06450"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_10453"} +{"question": "Which work studies the offline policy evaluation (OPE) in linear MDP, circumverting the issue of statistical dependency?", "answer": ["Variance-Aware Off-Policy Evaluation with Linear Function Approximation"], "answer_arxiv_id": ["2106.11960"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_10454"} +{"question": "What works have been carried out on contrastive methods in the field of unsupervised learning?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Momentum Contrast for Unsupervised Visual Representation Learning", "Self-Supervised Pretraining of 3D Features on any Point-Cloud", "What Should Not Be Contrastive in Contrastive Learning", "Spatio-temporal Self-Supervised Representation Learning for 3D Point Clouds", "What Makes for Good Views for Contrastive Learning?"], "answer_arxiv_id": ["2006.07733", "1911.05722", "2101.02691", "2008.05659", "2109.00179", "2005.10243"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_10455"} +{"question": "What works addressed their latent space for image editing using generative adversarial network?", "answer": ["EditGAN: High-Precision Semantic Image Editing", "GANSpace: Discovering Interpretable GAN Controls", "Image-based CLIP-Guided Essence Transfer", "InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs", "LatentCLR: A Contrastive Learning Approach for Unsupervised Discovery of Interpretable Directions", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "Style Transformer: Unpaired Text Style Transfer without Disentangled Latent Representation", "Predict, Prevent, and Evaluate: Disentangled Text-Driven Image Manipulation Empowered by Pre-Trained Vision-Language Model"], "answer_arxiv_id": ["2111.03186", "2004.02546", "2110.12427", "2005.09635", "2104.00820", "2103.17249", "2108.00946", "1905.05621", "2111.13333"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_10456"} +{"question": "Could you provide me a study about the BIG-Bench that contains over 200 tasks drawing on problems involving linguistics, math, common-sense reasoning, and others?", "answer": ["Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models"], "answer_arxiv_id": ["2206.04615"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_10457"} +{"question": "Which works come from the jailbreak community based on attacking LLMs?", "answer": ["Jailbroken: How Does LLM Safety Training Fail?", "Are aligned neural networks adversarially aligned?"], "answer_arxiv_id": ["2307.02483", "2306.15447"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_10458"} +{"question": "Which works demonstrated the effective use of prompts in retrieving knowledge from large language models?", "answer": ["Inducing Relational Knowledge from BERT", "How Can We Know When Language Models Know? On the Calibration of\n Language Models for Question Answering", "Can Generative Pre-trained Language Models Serve as Knowledge Bases for\n Closed-book QA?"], "answer_arxiv_id": ["1911.12753", "2012.00955", "2106.01561"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_10459"} +{"question": "Which studies are relevant to the generative approaches in the field of point clouds with self-supervised learning?", "answer": ["SO-Net: Self-Organizing Network for Point Cloud Analysis", "Learning Representations and Generative Models for 3D Point Clouds", "Self-Supervised Deep Learning on Point Clouds by Reconstructing Space", "Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point\n Modeling", "Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud\n Pre-training"], "answer_arxiv_id": ["1803.04249", "1707.02392", "1901.08396", "2111.14819", "2205.14401"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_10460"} +{"question": "Which studies have been conducted on 3D generation using diffusion models, including shape and point cloud generation?", "answer": ["Diffusion Probabilistic Models for 3D Point Cloud Generation", "MeshDiffusion: Score-based Generative 3D Mesh Modeling", "LION: Latent Point Diffusion Models for 3D Shape Generation", "3D Shape Generation and Completion through Point-Voxel Diffusion"], "answer_arxiv_id": ["2103.01458", "2303.08133", "2210.06978", "2104.03670"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_10461"} +{"question": "Can you point out research that worked on bridging natural language and programming languages?", "answer": ["ERNIE-Code: Beyond English-Centric Cross-lingual Pretraining for\n Programming Languages"], "answer_arxiv_id": ["2212.06742"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_10462"} +{"question": "Could you provide me some studies that apply rejection sampling on sampled responses through the RM to perform self-imitation learning?", "answer": ["RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignment", "Reinforced Self-Training (ReST) for Language Modeling"], "answer_arxiv_id": ["2304.06767", "2308.08998"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_10463"} +{"question": "What studies aimed at improving the brute-force search but encountered limitations because they required noiseless responses?", "answer": ["Hardness and Algorithms for Robust and Sparse Optimization"], "answer_arxiv_id": ["2206.14354"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_10464"} +{"question": "Could you provide some studies that improved the imperceptibility of adversarial perturbations in point clouds?", "answer": ["Geometry-Aware Generation of Adversarial Point Clouds", "Shape-invariant 3D Adversarial Point Clouds"], "answer_arxiv_id": ["1912.11171", "2203.04041"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_10465"} +{"question": "Could you provide me some works about identifying causal structure through interventional data?", "answer": ["Learning Neural Causal Models from Unknown Interventions", "Gradient-Based Neural DAG Learning", "Joint Causal Inference from Multiple Contexts"], "answer_arxiv_id": ["1910.01075", "1906.02226", "1611.10351"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_10466"} +{"question": "What are the works emphasizing the importance of diversity in pre-training data?", "answer": ["On the Effect of Pretraining Corpora on In-context Learning by a\n Large-scale Language Model", "The Pile: An 800GB Dataset of Diverse Text for Language Modeling"], "answer_arxiv_id": ["2204.13509", "2101.00027"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_10467"} +{"question": "Which researches based on loss-based methods of data pruning?", "answer": ["An Empirical Study of Example Forgetting during Deep Neural Network Learning", "Deep Learning on a Data Diet: Finding Important Examples Early in Training", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training", "Coresets for Data-efficient Training of Machine Learning Models"], "answer_arxiv_id": ["1812.05159", "2107.07075", "2103.00123", "1906.01827"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_10468"} +{"question": "What papers propose to tackle counterfactual fairness through optimal transport?", "answer": ["Obtaining fairness using optimal transport theory", "FlipTest: Fairness Testing via Optimal Transport", "Testing Group Fairness via Optimal Transport Projections", "Optimal Transport of Classifiers to Fairness"], "answer_arxiv_id": ["1806.03195", "1906.09218", "2106.01070", "2202.03814"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_10469"} +{"question": "What work introduces a method similar to GradNorm but with two notable distinctions?", "answer": ["How Useful are Gradients for OOD Detection Really?"], "answer_arxiv_id": ["2205.10439"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_10470"} +{"question": "Which works performed advancements in LLMs used in visual assistants?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models", "GPT-4 Technical Report", "PaLM: Scaling Language Modeling with Pathways", "GLM-130B: An Open Bilingual Pre-trained Model"], "answer_arxiv_id": ["2307.09288", "2303.08774", "2204.02311", "2210.02414"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_10471"} +{"question": "Could you provide me some works that have used regular Transformers for tuning hyperparameters in L2L systems?", "answer": ["Towards Learning Universal Hyperparameter Optimizers with Transformers"], "answer_arxiv_id": ["2205.13320"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10472"} +{"question": "Which studies focused on 3D hand modeling using mixture of 3D Gaussians?", "answer": ["Fast and Robust Hand Tracking Using Detection-Guided Optimization"], "answer_arxiv_id": ["1602.04124"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_10473"} +{"question": "Any works explored computing 2D motion with a touch of 3D, through depth-separated layers or quasi-3D space?", "answer": ["Layered Neural Atlases for Consistent Video Editing", "Deformable Sprites for Unsupervised Video Decomposition", "Optical Flow with Semantic Segmentation and Localized Layers", "Tracking Everything Everywhere All at Once"], "answer_arxiv_id": ["2109.11418", "2204.07151", "1603.03911", "2306.05422"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_10474"} +{"question": "Which papers are about the adaptation of the noisy channel model to NMT systems?", "answer": ["Simple and Effective Noisy Channel Modeling for Neural Machine\n Translation", "Language Models not just for Pre-training: Fast Online Neural Noisy\n Channel Modeling"], "answer_arxiv_id": ["1908.05731", "2011.07164"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_10475"} +{"question": "Could you list the studies that regard the soft labels as a regularization method to boost performance?", "answer": ["Understanding and Improving Knowledge Distillation", "When Does Label Smoothing Help?", "Rethinking Soft Labels for Knowledge Distillation: A Bias-Variance Tradeoff Perspective", "Knowledge Distillation as Semiparametric Inference", "Learning from Noisy Labels with Distillation", "When Does Label Smoothing Help?"], "answer_arxiv_id": ["2002.03532", "1906.02629", "2102.00650", "2104.09732", "1703.02391", "1906.02629"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_10476"} +{"question": "Which studies were conducted for multimodal hateful meme detection using conventional fusion?", "answer": ["Exploring Hate Speech Detection in Multimodal Publications"], "answer_arxiv_id": ["1910.03814"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_10477"} +{"question": "Could you provide me some studies about the field of model editing and fairness?", "answer": ["Editing a classifier by rewriting its prediction rules", "Locating and Editing Factual Associations in GPT", "A Near-Optimal Algorithm for Debiasing Trained Machine Learning Models", "Classification with Fairness Constraints: A Meta-Algorithm with Provable Guarantees"], "answer_arxiv_id": ["2112.01008", "2202.05262", "2106.12887", "1806.06055"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_10478"} +{"question": "What works consider an estimation based on Monte Carlo sampling for heteroscedastic classifiers?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "A Simple Probabilistic Method for Deep Classification under Input-Dependent Label Noise", "Correlated Input-Dependent Label Noise in Large-Scale Image Classification"], "answer_arxiv_id": ["1703.04977", "2003.06778v3", "2105.10305"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10479"} +{"question": "What papers studied the zeroth-order online non-convex optimization?", "answer": ["ONLINE NON-CONVEX OPTIMIZATION WITH IMPERFECT FEEDBACK", "ZEROTH-ORDER NON-CONVEX LEARNING VIA HIERARCHICAL DUAL AVERAGING"], "answer_arxiv_id": ["2010.08496", "2109.05829"], "source_meta": {"published_time": "20230807"}, "qid": "AutoScholarQuery_train_10480"} +{"question": "Could you provide me some works on CNNs on programmable sensors?", "answer": ["Cain: Automatic Code Generation for Simultaneous Convolutional Kernels\n on Focal-plane Sensor-processors", "AnalogNet: Convolutional Neural Network Inference on Analog Focal Plane\n Sensor Processors", "Fully Embedding Fast Convolutional Networks on Pixel Processor Arrays"], "answer_arxiv_id": ["2101.08715", "2006.01765", "2004.12525"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_10481"} +{"question": "Which studies propose adversarial training methods for backdoor removal, like Implicit Hypergradient or Adversarial Neuron Pruning?", "answer": ["Adversarial Unlearning of Backdoors via Implicit Hypergradient", "Adversarial Neuron Pruning Purifies Backdoored Deep Models"], "answer_arxiv_id": ["2110.03735", "2110.14430"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_10482"} +{"question": "Can you tell me about research that looks at adapter-based tuning?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "AdapterHub: A Framework for Adapting Transformers", "ZEN: Pre-training Chinese Text Encoder Enhanced by N-gram\n Representations"], "answer_arxiv_id": ["1902.00751", "2007.07779", "1911.00720"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_10483"} +{"question": "Which research papers have successfully applied the attention mechanism in the tracking domain?", "answer": ["Transformer Tracking", "Learning Spatio-Temporal Transformer for Visual Tracking", "Joint Feature Learning and Relation Modeling for Tracking: A One-Stream Framework"], "answer_arxiv_id": ["2103.15436", "2103.17154", "2203.11991"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_10484"} +{"question": "What research discusses adding input perturbations for natural language tasks?", "answer": ["Holistic Evaluation of Language Models", "CondaQA: A Contrastive Reading Comprehension Dataset for Reasoning about Negation", "Adversarial Examples for Evaluating Reading Comprehension Systems", "Learning Visually-Grounded Semantics from Contrastive Adversarial Samples", "TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP", "Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Models"], "answer_arxiv_id": ["2211.09110", "2211.00295", "1707.07328", "1806.10348", "2005.05909", "2111.02840"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10485"} +{"question": "Which research papers have used LLMs, such as GPT-2 or BART, to create paraphrases?", "answer": ["BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension"], "answer_arxiv_id": ["1910.13461"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_10486"} +{"question": "Could you provide me with some studies which have methods for contrastive learning tailored to unsupervised representation learning of time series?", "answer": ["Unsupervised Scalable Representation Learning for Multivariate Time Series", "Neighborhood Contrastive Learning Applied to Online Patient Monitoring", "TS2Vec: Towards Universal Representation of Time Series", "Unsupervised Representation Learning for Time Series with Temporal Neighborhood Coding", "CLOCS: Contrastive Learning of Cardiac Signals Across Space, Time, and Patients", "Time-Series Representation Learning via Temporal and Contextual Contrasting", "Unsupervised Time-Series Representation Learning with Iterative Bilinear Temporal-Spectral Fusion", "Self-Supervised Contrastive Pre-Training for Time Series via Time-Frequency Consistency"], "answer_arxiv_id": ["1901.10738", "2106.05142", "2106.10466", "2106.00750", "2005.13249", "2106.14112", "2202.04770", "2206.08496"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_10487"} +{"question": "Can you provide some studies that explored adversarial training for improving adversarial robustness?", "answer": ["Intriguing properties of neural networks", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1312.6199", "1706.06083"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_10488"} +{"question": "Which papers propose the use of maximum mean discrepancy and Wasserstein distance for reducing domain discrepancy in unsupervised domain adaptation?", "answer": ["Learning Transferable Features with Deep Adaptation Networks", "Wasserstein Distance Guided Representation Learning for Domain Adaptation"], "answer_arxiv_id": ["1502.02791", "1707.01217"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_10489"} +{"question": "What work adopted the PIIF based approaches that was mentioned in this paper?", "answer": ["On Fairness and Stability in Two-Sided Matchings"], "answer_arxiv_id": ["2111.10885"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_10490"} +{"question": "What is the reference for the work known as 'SAM' that has shown remarkable abilities for various vision tasks?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_10491"} +{"question": "Which paper introduced novel GAN architectures like CausalGAN and CausalBEGAN for image generation tasks?", "answer": ["CausalGAN: Learning Causal Implicit Generative Models with Adversarial Training"], "answer_arxiv_id": ["1709.02023"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_10492"} +{"question": "Which works focus on the emergence of pragmatic understanding from large-scale language modeling?", "answer": ["NOPE: A Corpus of Naturally-Occurring Presuppositions in English"], "answer_arxiv_id": ["2109.06987"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_10493"} +{"question": "Which papers are included in the post-training quantization methods for model quantization?", "answer": ["Fixed Point Quantization of Deep Convolutional Networks", "Data-Free Quantization Through Weight Equalization and Bias Correction", "HAWQ: Hessian AWare Quantization of Neural Networks with Mixed-Precision", "Post training 4-bit quantization of convolutional networks for rapid-deployment", "Low-bit Quantization of Neural Networks for Efficient Inference", "ZeroQ: A Novel Zero Shot Quantization Framework", "Up or Down? Adaptive Rounding for Post-Training Quantization", "Brecq: pushing the limit of post-training quantization by block reconstruction"], "answer_arxiv_id": ["1511.06393", "1906.04721", "1905.03696", "1810.05723", "1902.06822", "2001.00281", "2004.10568", "2102.05426"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_10494"} +{"question": "Could you name some works that tested environments for ICRL using Grid-Worlds?", "answer": ["Maximum Likelihood Constraint Inference for Inverse Reinforcement Learning", "Maximum Likelihood Constraint Inference from Stochastic Demonstrations", "Making Human-Like Trade-offs in Constrained Environments by Learning from Demonstrations", "Learning Soft Constraints From Constrained Expert Demonstrations"], "answer_arxiv_id": ["1909.05477", "2102.12554", "2109.11018", "2206.01311"], "source_meta": {"published_time": "20220620"}, "qid": "AutoScholarQuery_train_10495"} +{"question": "What works employed adaptive graphs for improving the capacity of Graph Convolution Networks (GCNs) in skeleton-based action recognition?", "answer": ["Actional-Structural Graph Convolutional Networks for Skeleton-based Action Recognition", "Two-Stream Adaptive Graph Convolutional Networks for Skeleton-Based Action Recognition", "Dynamic GCN: Context-enriched Topology Learning for Skeleton-based Action Recognition", "Channel-wise Topology Refinement Graph Convolution for Skeleton-Based Action Recognition"], "answer_arxiv_id": ["1904.12659", "1805.07694", "2007.14690", "2107.12213"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_10496"} +{"question": "What studies developed methods around set latent representations in the field of codebook learning?", "answer": ["Object Scene Representation Transformer", "Scene Representation Transformer: Geometry-Free Novel View Synthesis Through Set-Latent Scene Representations", "Object-Centric Learning with Slot Attention", "Unsupervised Discovery of Object Radiance Fields"], "answer_arxiv_id": ["2206.06922", "2111.13152", "2006.15055", "2107.07905"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_10497"} +{"question": "Could you provide me some studies on federated learning that focused on the linear speedup for convergence?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "Federated Optimization in Heterogeneous Networks", "Linear Convergence in Federated Learning: Tackling Client Heterogeneity and Sparse Gradients", "Federated Learning Based on Dynamic Regularization", "Local SGD Converges Fast and Communicates Little", "Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning", "Achieving ​ Linear ​ Speedup ​ with ​​ Partial ​​ Worker ​​ Participation ​​ in ​​ Non-IID ​​ Federated ​​ Learning"], "answer_arxiv_id": ["1602.05629", "1812.06127", "2102.07053", "2111.04263", "1805.09767", "1807.06629", "2101.11203"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_10498"} +{"question": "Which papers discussed the effectivity of temperature scaling for model calibration in discriminative settings?", "answer": ["On Calibration of Modern Neural Networks", "Measuring Calibration in Deep Learning", "Calibration of Pre-trained Transformers"], "answer_arxiv_id": ["1706.04599", "1904.01685", "2003.07892"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_10499"} +{"question": "What papers have made proposals in the field of visual reasoning to combine deep representation learning and symbolic program execution?", "answer": ["Inferring and Executing Programs for Visual Reasoning", "Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding"], "answer_arxiv_id": ["1705.03633", "1810.02338"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_10500"} +{"question": "What papers proposed Energy-based GFlowNets for the related problem of fitting a GFlowNet to a nonstationary reward?", "answer": ["Generative Flow Networks for Discrete Probabilistic Modeling"], "answer_arxiv_id": ["2202.01361"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_10501"} +{"question": "Could you provide me some survey papers that discuss about the implicit regularization effect?", "answer": ["Deep learning: a statistical viewpoint", "A Farewell to the Bias-Variance Tradeoff? An Overview of the Theory of Overparameterized Machine Learning", "On the Implicit Bias in Deep-Learning Algorithms"], "answer_arxiv_id": ["2103.09177", "2109.02355", "2208.12591"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_10502"} +{"question": "What works allow users to provide a natural language instruction in writing assistants?", "answer": ["Langsmith: An Interactive Academic Text Revision System", "A Recipe For Arbitrary Text Style Transfer with Large Language Models", "On Improving Summarization Factual Consistency from Natural Language\n Feedback", "CoEdIT: Text Editing by Task-Specific Instruction Tuning"], "answer_arxiv_id": ["2010.04332", "2109.03910", "2212.09968", "2305.09857"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_10503"} +{"question": "Could you tell me some works that evaluated the causal validity of explanations?", "answer": ["CausaLM: Causal Model Explanation Through Counterfactual Language Models", "Amnesic Probing: Behavioral Explanation with Amnesic Counterfactuals", "Explaining Classifiers with Causal Concept Effect (CaCE)"], "answer_arxiv_id": ["2005.13407", "2006.00995", "1907.07165"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_10504"} +{"question": "What papers mentioned traditional multi-view 3D reconstruction methods and applied structure-from-motion techniques?", "answer": ["Capturing, Reconstructing, and Simulating: the UrbanScene3D Dataset"], "answer_arxiv_id": ["2107.04286"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_10505"} +{"question": "What works discuss the use of Knowledge Distillation for transferring knowledge from teacher models to student models?", "answer": ["Distilling the Knowledge in a Neural Network", "Relational Knowledge Distillation"], "answer_arxiv_id": ["1503.02531", "1904.05068"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_10506"} +{"question": "Which works use re-weighting the loss in a more 'smooth' manner?", "answer": ["Exploring the Limits of Weakly Supervised Pretraining"], "answer_arxiv_id": ["1805.00932"], "source_meta": {"published_time": "20221230"}, "qid": "AutoScholarQuery_train_10507"} +{"question": "Which research piece marks the reformulation of INR for video signals to be frame-wise that resulted in the inception of Neural Representations for Videos (NeRV)?", "answer": ["NeRV: Neural Representations for Videos"], "answer_arxiv_id": ["2110.13903"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10508"} +{"question": "What studies are about multi-modality IQA methods?", "answer": ["Improved Baselines with Visual Instruction Tuning", "InternLM-XComposer: A Vision-Language Large Model for Advanced\n Text-image Comprehension and Composition", "Qwen-VL: A Versatile Vision-Language Model for Understanding,\n Localization, Text Reading, and Beyond"], "answer_arxiv_id": ["2310.03744", "2309.15112", "2308.12966"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_10509"} +{"question": "Which research papers conducted comprehensive studies on the properties and limitations of Multilingual BERT (mBERT)?", "answer": ["SemEval-2020 Task 9: Overview of Sentiment Analysis of Code-Mixed Tweets", "Cross-Lingual Ability of Multilingual BERT: An Empirical Study"], "answer_arxiv_id": ["2008.04277", "1912.07840"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_10510"} +{"question": "What studies incorporate residual quantization mechanisms in the context of using VQ-VAE for a unified multimodal language model?", "answer": ["High Fidelity Neural Audio Compression", "SoundStream: An End-to-End Neural Audio Codec"], "answer_arxiv_id": ["2210.13438", "2107.03312"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_10511"} +{"question": "Which paper proposes the idea that model predictions should be 'right for the right reasons'?", "answer": ["Right for the Right Reasons: Training Differentiable Models by Constraining their Explanations"], "answer_arxiv_id": ["1703.03717"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_10512"} +{"question": "Which papers have applied self-competition to maximize a constrained objective?", "answer": ["MuZero with Self-competition for Rate Control in VP9 Video Compression"], "answer_arxiv_id": ["2202.06626"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_10513"} +{"question": "What studies have examined the geometrical properties of latent space in GANs?", "answer": ["Low-Rank Subspaces in GANs", "Do Not Escape From the Manifold: Discovering the Local Coordinates on the Latent Space of GANs"], "answer_arxiv_id": ["2106.04488", "2106.06959"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_10514"} +{"question": "What works use the Umeyama algorithm in predicting 6D pose with NOCS map estimation?", "answer": ["Normalized Object Coordinate Space for Category-Level 6D Object Pose and\n Size Estimation"], "answer_arxiv_id": ["1901.02970"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_10515"} +{"question": "What research papers are associated with Group 1 approaches in text-based image editing, where they operate semantic change on intermediate UNet attention maps?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing", "EDICT: Exact Diffusion Inversion via Coupled Transformations", "Effective Real Image Editing with Accelerated Iterative Diffusion\n Inversion"], "answer_arxiv_id": ["2211.12572", "2304.08465v1", "2211.12446", "2309.04907"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_10516"} +{"question": "What work did further theoretical analysis of Forward Gradient?", "answer": ["Optimization without Backpropagation"], "answer_arxiv_id": ["2209.06302"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_10517"} +{"question": "Any works about calibration in regression?", "answer": ["Accurate Uncertainties for Deep Learning Using Calibrated Regression", "Distribution Calibration for Regression", "Copula Calibration", "Individual Calibration with Randomized Forecasting"], "answer_arxiv_id": ["1807.00263", "1905.06023", "1307.7650", "2006.10288"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_10518"} +{"question": "Which research developed a metric called METEOR for the evaluation of generated captions in dense captioning?", "answer": ["DenseCap: Fully Convolutional Localization Networks for Dense Captioning"], "answer_arxiv_id": ["1511.07571"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10519"} +{"question": "What papers laid the foundational theory behind joint textual-tabular data understanding, specifically in table semantic parsing?", "answer": ["Compositional Semantic Parsing on Semi-Structured Tables", "HybridQA: A Dataset of Multi-Hop Question Answering over Tabular and Textual Data", "Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning", "Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task"], "answer_arxiv_id": ["1508.00305", "2004.07347", "1709.00103", "1809.08887"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_10520"} +{"question": "What papers proposed variational methods to approximate the marginal posterior?", "answer": ["Variational Causal Networks: Approximate Bayesian Inference over Causal Structures", "Differentiable DAG Sampling"], "answer_arxiv_id": ["2106.07635", "2203.08509"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10521"} +{"question": "What works have been designed to tackle the over-squashing problem in GNNs?", "answer": ["Oversquashing in GNNs through the lens of information contraction and graph expansion", "FoSR: First-order Spectral Rewiring for addressing Oversquashing in GNNs"], "answer_arxiv_id": ["2208.03471", "2210.11790"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_10522"} +{"question": "Which study used a 3D autoencoder for generating 3D shapes?", "answer": ["SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation"], "answer_arxiv_id": ["2212.04493"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_10523"} +{"question": "Could you provide a work that uses WikiData to form masked token prediction tasks for encouraging specific answers?", "answer": ["Can Language Models Be Specific? How?"], "answer_arxiv_id": ["2210.05159"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_10524"} +{"question": "Which studies provided evidence that language models can encode geography?", "answer": ["Do Language Models Know the Way to Rome?", "Geographic and Geopolitical Biases of Language Models"], "answer_arxiv_id": ["2109.07971", "2212.10408"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_10525"} +{"question": "Which works focus on node-level and graph-level tasks for KD methods on graph data?", "answer": ["Cold Brew: Distilling Graph Node Representations with Incomplete or Missing Neighborhoods", "Graph-less Neural Networks: Teaching Old MLPs New Tricks via Distillation", "NOSMOG: Learning Noise-robust and Structure-aware MLPs on Graphs", "Graph Representation Learning via Multi-task Knowledge Distillation", "Iterative Graph Self-distillation", "Graph-Free Knowledge Distillation for Graph Neural Networks", "On Representation Knowledge Distillation for Graph Neural Networks"], "answer_arxiv_id": ["2111.04840", "2110.08727", "2208.10010", "1911.05700", "2010.12609", "2105.07519v2", "2111.04964"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_10526"} +{"question": "Which works describe using refinement networks in trajectory prediction?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["1506.01497", "2005.12872"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_10527"} +{"question": "Are there any reported methods of using synthetic gradients to improve the parallelism in training DNNs?", "answer": ["Decoupled Neural Interfaces using Synthetic Gradients"], "answer_arxiv_id": ["1608.05343"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10528"} +{"question": "Are there any papers on evaluating and augmenting the algorithm reasoning abilities of large language models?", "answer": ["Teaching Algorithmic Reasoning via In-context Learning", "Recursion of Thought: A Divide-and-Conquer Approach to Multi-Context Reasoning with Language Models", "Code Execution with Pre-trained Language Models"], "answer_arxiv_id": ["2211.09066", "2306.06891v1", "2305.05383"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_10529"} +{"question": "Which papers talked about the two-stage paradigm in the context of spatio-temporal video grounding?", "answer": ["Where Does It Exist: Spatio-Temporal Video Grounding for Multi-Form\n Sentences", "Object-Aware Multi-Branch Relation Networks for Spatio-Temporal Video\n Grounding", "Human-centric Spatio-Temporal Video Grounding With Visual Transformers"], "answer_arxiv_id": ["2001.06891", "2008.06941", "2011.05049"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_10530"} +{"question": "Which works propose architectural changes to the model for incorporating knowledge graph data into LLMs?", "answer": ["ERNIE: Enhanced Language Representation with Informative Entities", "KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language\n Representation", "Knowledge Enhanced Contextual Word Representations", "LUKE: Deep Contextualized Entity Representations with Entity-aware\n Self-attention", "Improving Multi-hop Knowledge Base Question Answering by Learning\n Intermediate Supervision Signals"], "answer_arxiv_id": ["1905.07129", "1911.06136", "1909.04164", "2010.01057", "2101.03737"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_10531"} +{"question": "Which paper discussed the different behavior of full-batch and mini-batch SAM in theoretical understanding?", "answer": ["How Does Sharpness-Aware Minimization Minimize Sharpness?"], "answer_arxiv_id": ["2211.05729"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_10532"} +{"question": "What studies use an encoder-decoder network to encode images into a 3D volume and decode it with volume rendering?", "answer": ["Neural Volumes: Learning Dynamic Renderable Volumes from Images"], "answer_arxiv_id": ["1906.07751"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_10533"} +{"question": "Which papers proposed models that use the gradient-based meta-learning approach?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-SGD: Learning to Learn Quickly for Few-Shot Learning"], "answer_arxiv_id": ["1703.03400", "1707.09835"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_10534"} +{"question": "What work significantly scaled up training a modern deep learning speech recognition system on several languages?", "answer": ["Massively Multilingual ASR: 50 Languages, 1 Model, 1 Billion Parameters"], "answer_arxiv_id": ["2007.03001"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_10535"} +{"question": "Can you provide me some studies where softmax operation is used as a normalization of the output computation?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2006.09882", "2104.14294"], "source_meta": {"published_time": "20220401"}, "qid": "AutoScholarQuery_train_10536"} +{"question": "What work proposed the Spatial Temporal Graph Convolution Network (STGCN) that this study focus on?", "answer": ["Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition"], "answer_arxiv_id": ["1801.07455"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_10537"} +{"question": "Could you provide me some studies that considered manifolds in the setting of diffusion models for specific applications?", "answer": ["Torsional Diffusion for Molecular Conformer Generation", "DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking"], "answer_arxiv_id": ["2206.01729", "2210.01776"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_10538"} +{"question": "Which papers discuss instance discrimination through the use of contrastive learning?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Learning deep representations by mutual information estimation and maximization", "A Simple Framework for Contrastive Learning of Visual Representations", "Big Self-Supervised Models are Strong Semi-Supervised Learners", "Momentum Contrast for Unsupervised Visual Representation Learning", "Data-Efficient Image Recognition with Contrastive Predictive Coding"], "answer_arxiv_id": ["1807.03748", "1808.06670", "2002.05709", "2006.10029", "1911.05722", "1905.09272"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_10539"} +{"question": "What studies conduct self-supervised pretraining via discriminative tasks in the image domain?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap your own latent: A new approach to self-supervised Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2006.07733", "2104.14294"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_10540"} +{"question": "Which research papers are central to the development of Generative Adversarial Networks (GANs) and their application in 2D image generative methods?", "answer": ["Generative Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Alias-Free Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["2203.00667", "1812.04948", "2106.12423", "1912.04958"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10541"} +{"question": "Could you provide some datasets that predominantly focus on instructional content?", "answer": ["How2: A Large-scale Dataset for Multimodal Language Understanding", "Towards Automatic Learning of Procedures from Web Instructional Videos", "HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million\n Narrated Video Clips"], "answer_arxiv_id": ["1811.00347", "1703.09788", "1906.03327"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_10542"} +{"question": "Any research work proposing a unified view of the existing parameter-efficient fine-tuning strategies?", "answer": ["Towards a Unified View of Parameter-Efficient Transfer Learning", "Delta Tuning: A Comprehensive Study of Parameter Efficient Methods for Pre-trained Language Models"], "answer_arxiv_id": ["2110.04366", "2203.06904"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_10543"} +{"question": "Which works focus on prompt tuning in the context of transfer learning?", "answer": ["Learning to Prompt for Vision-Language Models", "Visual Prompt Tuning", "LION: Implicit Vision Prompt Tuning", "Being Comes from Not-being: Open-vocabulary Text-to-Motion Generation with Wordless Training"], "answer_arxiv_id": ["2109.01134", "2203.12119", "2303.09992", "2210.15929"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_10544"} +{"question": "Which work presents the Anakin architecture?", "answer": ["Podracer architectures for scalable Reinforcement Learning"], "answer_arxiv_id": ["2104.06272"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_10545"} +{"question": "Are there any works that approximate steady-state flow field predictions or predict solutions for unseen flow conditions and geometries with a surrogate CNN-based model?", "answer": ["Prediction of Aerodynamic Flow Fields Using Convolutional Neural Networks"], "answer_arxiv_id": ["1905.13166"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_10546"} +{"question": "Are there any works that discus applications of model-free RL?", "answer": ["Trust Region Policy Optimization", "Proximal Policy Optimization Algorithms", "Asynchronous Methods for Deep Reinforcement Learning", "Playing Atari with Deep Reinforcement Learning", "Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor"], "answer_arxiv_id": ["1502.05477", "1707.06347v2", "1602.01783", "1312.5602", "1801.01290"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_10547"} +{"question": "What studies use domain adaptation methods to mitigate negative transfer?", "answer": ["Characterizing and Avoiding Negative Transfer", "Importance Weighted Adversarial Nets for Partial Domain Adaptation"], "answer_arxiv_id": ["1811.09751", "1803.09210"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10548"} +{"question": "Which papers focus on establishing one-to-one matchings between line segments in the query and the 3D map for line-based localization?", "answer": ["SOLD2: Self-supervised Occlusion-aware Line Description and Detection", "Line as a Visual Sentence: Context-aware Line Descriptor for Visual\n Localization", "Robust Line Segments Matching via Graph Convolution Networks", "VLASE: Vehicle Localization by Aggregating Semantic Edges"], "answer_arxiv_id": ["2104.03362", "2109.04753", "2004.04993", "1807.02536"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_10549"} +{"question": "Could you provide me some studies that specifically designed their approaches for trees or buildings?", "answer": ["City3D: Large-Scale Building Reconstruction from Airborne LiDAR Point\n Clouds"], "answer_arxiv_id": ["2201.10276"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_10550"} +{"question": "In which papers is the concept of Neural Collapse (NC) introduced?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning\n training"], "answer_arxiv_id": ["2008.08186"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_10551"} +{"question": "Which works tried to reduce the dependency on pose in Neural Radiance Field (NeRF) using GANs, SLAM, depth, and coarse annotations?", "answer": ["GNeRF: GAN-based Neural Radiance Field without Posed Camera", "NeRF-SLAM: Real-Time Dense Monocular SLAM with Neural Radiance Fields", "NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in\n the Wild", "NoPe-NeRF: Optimising Neural Radiance Field with No Pose Prior", "SAMURAI: Shape And Material from Unconstrained Real-world Arbitrary Image collections"], "answer_arxiv_id": ["2103.15606", "2210.13641", "2110.07604", "2212.07388", "2205.15768v1"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_10552"} +{"question": "Which research studies the differentially private stochastic convex optimization (DP-SCO) problems in the interpolation regime?", "answer": ["Private optimization in the interpolation regime: faster rates and hardness results"], "answer_arxiv_id": ["2210.17070"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_10553"} +{"question": "Which studies involve merging LMs fine-tuned on different tasks to make a multitask fine-tuned LM in a distributed manner?", "answer": ["ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning"], "answer_arxiv_id": ["2212.01378"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_10554"} +{"question": "Which works first took the approach of training CNFs by reducing the likelihood of the training data?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_10555"} +{"question": "Are there any works that built antibody-specific language models?", "answer": ["Deciphering antibody affinity maturation with language models and weakly supervised learning"], "answer_arxiv_id": ["2112.07782"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_10556"} +{"question": "What benchmarks exist for automatic speech recognition?", "answer": ["Common Voice: A Massively-Multilingual Speech Corpus", "FLEURS: Few-shot Learning Evaluation of Universal Representations of\n Speech"], "answer_arxiv_id": ["1912.06670", "2205.12446"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_10557"} +{"question": "Could you provide me some studies about amortizing and learning the solutions to OT and matching problems?", "answer": ["Faster Matchings via Learned Duals", "Learning Predictions for Algorithms with Predictions", "Meta Optimal Transport"], "answer_arxiv_id": ["2107.09770", "2202.09312", "2206.05262"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_10558"} +{"question": "What works discuss extrapolation in the context of mathematical tasks?", "answer": ["Measuring Arithmetic Extrapolation Performance"], "answer_arxiv_id": ["1910.01888"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_10559"} +{"question": "What is the existing research on matrix solves with Hessian/Fisher matrices in second order/natural grade optimization?", "answer": ["Optimizing Neural Networks with Kronecker-factored Approximate Curvature"], "answer_arxiv_id": ["1503.05671"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_10560"} +{"question": "What studies focus on personalized federated learning techniques?", "answer": ["Towards Personalized Federated Learning"], "answer_arxiv_id": ["2103.00710"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_10561"} +{"question": "Which works documented performance on the Criteo dataset for CTR prediction?", "answer": ["DeepFM: A Factorization-Machine based Neural Network for CTR Prediction", "Deep & Cross Network for Ad Click Predictions"], "answer_arxiv_id": ["1703.04247", "1708.05123"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10562"} +{"question": "Could you provide me some researches which use a teacher-student model to focus on hard instances in multiple instance learning (MIL)?", "answer": ["Multiple Instance Learning Framework with Masked Hard Instance Mining\n for Whole Slide Image Classification"], "answer_arxiv_id": ["2307.15254"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_10563"} +{"question": "What studies proposed Named-entity-recognition (NER) approaches for identifying vulnerable libraries?", "answer": ["Cleaning the NVD: Comprehensive Quality Assessment, Improvements, and\n Analyses"], "answer_arxiv_id": ["2006.15074"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_10564"} +{"question": "Which studies introduce doubly robust estimators and variance reduction techniques to combat the drawbacks of the inverse propensity score method?", "answer": ["Enhanced Doubly Robust Learning for Debiasing Post-Click Conversion Rate Estimation", "StableDR: Stabilized Doubly Robust Learning for Recommendation on Data Missing Not at Random"], "answer_arxiv_id": ["2105.13623", "2205.04701"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_10565"} +{"question": "Which works have built various world models to replicate the real world and serve as virtual test environments for assessing robotic agents?", "answer": ["VirtualHome: Simulating Household Activities via Programs", "Watch-And-Help: A Challenge for Social Perception and Human-AI Collaboration", "AI2-THOR: An Interactive 3D Environment for Visual AI", "VRKitchen: an Interactive 3D Virtual Environment for Task-oriented Learning", "CHALET: Cornell House Agent Learning Environment", "MINOS: Multimodal Indoor Simulator for Navigation in Complex Environments", "Building Generalizable Agents with a Realistic and Rich 3D Environment", "MineRL: A Large-Scale Dataset of Minecraft Demonstrations", "JueWu-MC: Playing Minecraft with Sample-efficient Hierarchical Reinforcement Learning", "MineRL Diamond 2021 Competition: Overview, Results, and Lessons Learned"], "answer_arxiv_id": ["1806.07011", "2010.09890", "1712.05474", "1903.05757", "1801.07357", "1712.03931", "1801.02209", "1907.13440", "2112.04907v1", "2202.10583"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_10566"} +{"question": "What works focus on combining PCs with neural networks to create expressive hybrid models?", "answer": ["HyperSPNs: Compact and Expressive Probabilistic Circuits", "Continuous Mixtures of Tractable Probabilistic Models"], "answer_arxiv_id": ["2112.00914", "2209.10584"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_10567"} +{"question": "Which papers considered SDEs corresponding to the fractional FPE in dimension one?", "answer": ["Fractional Langevin Monte Carlo: Exploring Lévy Driven Stochastic Differential Equations for Markov Chain Monte Carlo", "Non-Asymptotic Analysis of Fractional Langevin Monte Carlo for Non-Convex Optimization"], "answer_arxiv_id": ["1706.03649", "1901.07487"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_10568"} +{"question": "Could you provide me some works about incremental learning in graph data?", "answer": ["Streaming Graph Neural Networks via Continual Learning", "Overcoming Catastrophic Forgetting in Graph Neural Networks with Experience Replay"], "answer_arxiv_id": ["2009.10951", "2003.09908"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_10569"} +{"question": "Are there any studies concerned with environments underpinned by a well-structured schema or ontology from the perspective of grounded language understanding?", "answer": ["RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers"], "answer_arxiv_id": ["1911.04942"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_10570"} +{"question": "Which papers discuss the Minty variational inequality (MVI) assumption in relation to nonconvex-nonconcave min-max optimization problems?", "answer": ["On the Convergence Properties of Non-Euclidean Extragradient Methods for Variational Inequalities with Generalized Monotone Operators", "Towards Better Understanding of Adaptive Gradient Algorithms in Generative Adversarial Nets", "Golden Ratio Algorithms for Variational Inequalities", "Optimistic Dual Extrapolation for Coherent Non-monotone Variational Inequalities", "First-order Convergence Theory for Weakly-Convex-Weakly-Concave Min-max Problems"], "answer_arxiv_id": ["1311.2776v1", "1912.11940", "1803.08832", "2103.04410", "1810.10207"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_10571"} +{"question": "Which works have studied the effectiveness of pretraining representations in Offline RL settings?", "answer": ["Provable Representation Learning for Imitation Learning via Bi-level Optimization", "Pretraining Representations for Data-Efficient Reinforcement Learning", "Provable Representation Learning for Imitation with Contrastive Fourier Features"], "answer_arxiv_id": ["2002.10544", "2106.04799", "2105.12272"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_10572"} +{"question": "What studies have been conducted regarding the susceptibility of LLMs to adversarial interactions, such as red teaming or adversarial prompting?", "answer": ["Red Teaming Language Models with Language Models", "Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned", "Extracting Training Data from Large Language Models", "Analyzing Leakage of Personally Identifiable Information in Language Models", "Universal Adversarial Triggers for Attacking and Analyzing NLP", "Automatically Auditing Large Language Models via Discrete Optimization"], "answer_arxiv_id": ["2202.03286v1", "2209.07858", "2012.07805", "2302.00539", "1908.07125", "2303.04381"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_10573"} +{"question": "What is the first research to integrate an additional EEG encoder to align the pre-trained BART for EEG-to-Text?", "answer": ["Open Vocabulary Electroencephalography-To-Text Decoding and Zero-shot\n Sentiment Classification"], "answer_arxiv_id": ["2112.02690"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_10574"} +{"question": "Any works utilizing environments to induce diversity among a population of agents?", "answer": ["Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions"], "answer_arxiv_id": ["1901.01753"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_10575"} +{"question": "What works apply self supervised learning to text + image in multimodal learning?", "answer": ["Zero-Shot Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["2102.12092", "2112.10752", "2204.06125", "1503.03585"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_10576"} +{"question": "What work first introduced the Decision-Estimation Coefficient as a complexity measure for online RL?", "answer": ["The Statistical Complexity of Interactive Decision Making"], "answer_arxiv_id": ["2112.13487v3"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_10577"} +{"question": "Which papers in Face Anti-spoofing study domain generalization for a model to work effectively on unseen target domains?", "answer": ["Adaptive Transformers for Robust Few-shot Cross-domain Face\n Anti-spoofing"], "answer_arxiv_id": ["2203.12175"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_10578"} +{"question": "Which works use a common transformer framework for unified modeling across different domains?", "answer": ["A Unified Sequence Interface for Vision Tasks", "Pix2seq: A Language Modeling Framework for Object Detection", "Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks", "Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning", "SegGPT: Segmenting Everything In Context", "Visual Prompting via Image Inpainting", "Explore In-Context Learning for 3D Point Cloud Understanding", "Sequential Modeling Enables Scalable Learning for Large Vision Models", "Skeleton-in-Context: Unified Skeleton Sequence Modeling with In-Context\n Learning"], "answer_arxiv_id": ["2206.07669", "2109.10852", "2206.08916", "2212.02499", "2304.03284", "2209.00647", "2306.08659", "2312.00785", "2312.03703"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_10579"} +{"question": "Can you share any studies that delved into object detection employing models with large image diffusion?", "answer": ["DiffusionDet: Diffusion Model for Object Detection"], "answer_arxiv_id": ["2211.09788"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10580"} +{"question": "Could you mention the studies on offline RL in Markov games (MGs)?", "answer": ["Pessimistic Minimax Value Iteration: Provably Efficient Equilibrium Learning from Offline Datasets", "Provably Efficient Offline Multi-agent Reinforcement Learning via Strategy-wise Bonus", "When is Offline Two-Player Zero-Sum Markov Game Solvable?", "Nearly Minimax Optimal Offline Reinforcement Learning with Linear Function Approximation: Single-Agent MDP and Markov Game", "Model-Based Reinforcement Learning Is Minimax-Optimal for Offline Zero-Sum Markov Games", "Offline Learning in Markov Games with General Function Approximation"], "answer_arxiv_id": ["2202.07511", "2206.00159", "2201.03522", "2205.15512", "2206.04044", "2302.02571"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_10581"} +{"question": "What research proposed the Latent Diffusion Model to deal with the computational overhead of generating high-resolution images?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_10582"} +{"question": "Could you provide me a study where the KL divergence in mutual information is replaced with Wasserstein distance for representation learning?", "answer": ["Wasserstein Dependency Measure for Representation Learning"], "answer_arxiv_id": ["1903.11780"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_10583"} +{"question": "Can you tell me about the works that proposed to train certain parts of the solver on a dataset?", "answer": ["Bespoke Solvers for Generative Flow Models", "GENIE: Higher-Order Denoising Diffusion Solvers", "Learning Fast Samplers for Diffusion Models by Differentiating Through\n Sample Quality", "Bilateral Denoising Diffusion Models", "Optimal Linear Subspace Search: Learning to Construct Fast and\n High-Quality Schedulers for Diffusion Models"], "answer_arxiv_id": ["2310.19075", "2210.05475", "2202.05830", "2108.11514", "2305.14677"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_10584"} +{"question": "Which studies have focused on deep representation learning in hyperbolic spaces?", "answer": ["Hyperbolic Deep Neural Networks: A Survey", "Hyperbolic Graph Neural Networks: A Review of Methods and Applications"], "answer_arxiv_id": ["2101.04562", "2202.13852"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_10585"} +{"question": "What studies obtained convergence results for the damped proximal point method?", "answer": ["The Landscape of the Proximal Point Method for Nonconvex-Nonconcave Minimax Optimization"], "answer_arxiv_id": ["2006.08667"], "source_meta": {"published_time": "20221226"}, "qid": "AutoScholarQuery_train_10586"} +{"question": "Which work details about an instruction-editing dataset using GPT-3 and the Prompt-to-Prompt model?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2211.09800"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_10587"} +{"question": "Which papers provide a theoretical perspective to understand in-context learning and cast it as implicit Bayesian inference?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_10588"} +{"question": "What papers are about providing effective control over outputs for diffusion models?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Sketch-Guided Text-to-Image Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2108.01073", "2211.13752", "2302.05543"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_10589"} +{"question": "Could you provide me some studies about testing the robustness in Bayesian models?", "answer": ["Statistical Guarantees for the Robustness of Bayesian Neural Networks", "Probabilistic Safety for Bayesian Neural Networks", "Adv-BNN: Improved Adversarial Defense through Robust Bayesian Neural Network", "Gradient-Free Adversarial Attacks for Bayesian Neural Networks", "Sufficient Conditions for Idealised Models to Have No Adversarial Examples: a Theoretical and Empirical Study with Bayesian Neural Networks"], "answer_arxiv_id": ["1903.01980", "2004.10281", "1810.01279", "2012.12640", "1806.00667"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_10590"} +{"question": "Who were the first to propose the Bernstein-type self-normalized concentration inequality and combine it with variance-weighted regression to achieve minimax optimal regret bound for linear mixture MDPs?", "answer": ["Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes"], "answer_arxiv_id": ["2012.08507"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_10591"} +{"question": "Which works operate on explicit representations such as point clouds in 3D semantic learning?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds"], "answer_arxiv_id": ["1612.00593", "1911.11236"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_10592"} +{"question": "What papers focus on converting information-theoretic quantities to some other algorithmic stability notions?", "answer": ["Reasoning About Generalization via Conditional Mutual Information", "Information-theoretic generalization bounds for black-box learning algorithms", "On Leave-One-Out Conditional Mutual Information For Generalization"], "answer_arxiv_id": ["2001.09122", "2110.01584", "2207.00581"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_10593"} +{"question": "Can you provide some studies that examined convergence rates of the EM algorithm for learning Gaussian mixtures?", "answer": ["Statistical guarantees for the EM algorithm: From population to sample-based analysis", "Ten Steps of EM Suffice for Mixtures of Two Gaussians", "Global Analysis of Expectation Maximization for Mixtures of Two Gaussians", "Convergence Analysis of Gradient EM for Multi-component Gaussian Mixture", "Statistical Convergence of the EM Algorithm on Gaussian Mixture Models", "The EM Algorithm gives Sample-Optimality for Learning Mixtures of Well-Separated Gaussians", "Improved Convergence Guarantees for Learning Gaussian Mixture Models by EM and Gradient EM"], "answer_arxiv_id": ["1408.2156", "1609.00368", "1608.07630", "1705.08530", "1810.04090", "2002.00329", "2101.00575v2"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_10594"} +{"question": "Which study incorporated external knowledge from text-based KGs into the multimodal chain of thought reasoning?", "answer": ["KAM-CoT: Knowledge Augmented Multimodal Chain-of-Thoughts Reasoning"], "answer_arxiv_id": ["2401.12863"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_10595"} +{"question": "Which paper first derived an hierarchical Bayesian approach to nearest neighbour GPs?", "answer": ["Hierarchical Nearest-Neighbor Gaussian Process Models for Large Geostatistical Datasets"], "answer_arxiv_id": ["1406.7343"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_10596"} +{"question": "Can you name some works that use temporal structure for weak-supervision in causal representation learning?", "answer": ["Properties from Mechanisms: An Equivariance Perspective on Identifiable Representation Learning"], "answer_arxiv_id": ["2110.15796"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10597"} +{"question": "Which studies discussed fast sampling of DDIM through knowledge distillation?", "answer": ["Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed", "Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2101.02388", "2202.00512"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10598"} +{"question": "Could you provide me some seminal works on multi-teacher knowledge distillation methods?", "answer": ["Unifying Heterogeneous Classifiers with Distillation"], "answer_arxiv_id": ["1904.06062"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_10599"} +{"question": "Can you specify the works where pre-trained generative models are used to accomplish one-shot reconstructions?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "AG3D: Learning to Generate 3D Avatars from 2D Image Collections", "HumanGen: Generating Human Radiance Fields with Explicit Priors", "Next3D: Generative Neural Texture Rasterization for 3D-Aware Head\n Avatars", "Get3DHuman: Lifting StyleGAN-Human into a 3D Generative Model using\n Pixel-aligned Reconstruction Priors", "Mimic3D: Thriving 3D-Aware GANs via 3D-to-2D Imitation", "GRAM: Generative Radiance Manifolds for 3D-Aware Image Generation"], "answer_arxiv_id": ["2112.07945", "2305.02312", "2212.05321", "2211.11208", "2302.01162", "2303.09036", "2112.08867"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_10600"} +{"question": "Can you name studies that proposed efficient network architectures to reduce the computational complexity of image segmentation models?", "answer": ["BiSeNet: Bilateral Segmentation Network for Real-time Semantic\n Segmentation", "Rethinking BiSeNet For Real-time Semantic Segmentation", "PIDNet: A Real-time Semantic Segmentation Network Inspired by PID\n Controllers", "SOLO: Segmenting Objects by Locations", "Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up\n Panoptic Segmentation", "Real-Time Panoptic Segmentation from Dense Detections", "You Only Segment Once: Towards Real-Time Panoptic Segmentation"], "answer_arxiv_id": ["1808.00897", "2104.13188", "2206.02066", "1912.04488", "1911.10194", "1912.01202", "2303.14651"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_10601"} +{"question": "Which papers have explored graph pretraining to alleviate the generalization problem of graph-based learning in the chemical domain?", "answer": ["Strategies for Pre-training Graph Neural Networks", "Pre-training Molecular Graph Representation with 3D Geometry", "GCC: Graph Contrastive Coding for Graph Neural Network Pre-Training", "3D Infomax improves GNNs for Molecular Property Prediction", "InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization", "Molecular Contrastive Learning of Representations via Graph Neural Networks", "Graph Contrastive Learning Automated", "Graph Contrastive Learning with Augmentations"], "answer_arxiv_id": ["1905.12265", "2110.07728", "2006.09963", "2110.04126", "1908.01000", "2102.10056", "2106.07594", "2010.13902"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_10602"} +{"question": "What papers discussed the learning of invariances by means of MAP, marginal likelihood, BayesOpt, and meta learning?", "answer": ["Learning Invariances in Neural Networks", "Invariance Learning in Deep Neural Networks with Differentiable Laplace Approximations", "On the Importance of Hyperparameters and Data Augmentation for Self-Supervised Learning"], "answer_arxiv_id": ["2010.11882", "2202.10638", "2207.07875v1"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_10603"} +{"question": "Which works extended OLM to generic link functions?", "answer": ["Scalable Generalized Linear Bandits: Online Computation and Hashing"], "answer_arxiv_id": ["1706.00136"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_10604"} +{"question": "Which papers introduced large language models since the emergence of chatGPT?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model"], "answer_arxiv_id": ["2307.09288", "2211.05100"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_train_10605"} +{"question": "Are there any works proposing a new trajectory loss to replace the sharpness definition?", "answer": ["Sharpness-Aware Training for Free"], "answer_arxiv_id": ["2205.14083"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_10606"} +{"question": "What works have been conducted about the 3D object generation using diffusion models?", "answer": ["Diffusion Probabilistic Models for 3D Point Cloud Generation", "3D Shape Generation and Completion through Point-Voxel Diffusion", "DreamFusion: Text-to-3D using 2D Diffusion", "Neural Wavelet-domain Diffusion for 3D Shape Generation"], "answer_arxiv_id": ["2103.01458", "2104.03670", "2209.14988", "2209.08725"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10607"} +{"question": "What are the studies that presented their models' performance on Algebra Question Answering with Rationales (AQuA-RAT) dataset?", "answer": ["Measuring and Improving BERT’s Mathematical Abilities by Predicting the Order of Reasoning", "Program Induction by Rationale Generation: Learning to Solve and Explain Algebraic Word Problems"], "answer_arxiv_id": ["2106.03921", "1705.04146"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10608"} +{"question": "Could you provide me some studies that proposed solutions to the computational challenge of computing the full (inverse) hessian in Optimal Brain Surgeon?", "answer": ["Learning to Prune Deep Neural Networks via Layer-wise Optimal Brain Surgeon", "WoodFisher: Efficient Second-Order Approximation for Neural Network Compression"], "answer_arxiv_id": ["1705.07565", "2004.14340"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_10609"} +{"question": "How about attempts to enhance NeRF’s generalisation ability across diverse scenes?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images"], "answer_arxiv_id": ["2012.02190"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_10610"} +{"question": "What research studies proposed geometric neural networks for point clouds?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space", "Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Framework"], "answer_arxiv_id": ["1612.00593", "1706.02413", "2202.07123"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_10611"} +{"question": "What previous studies investigated performance degradation as a long-term optimization danger in policy optimization algorithms?", "answer": ["Cliff Diving: Exploring Reward Surfaces in Reinforcement Learning Environments"], "answer_arxiv_id": ["2205.07015"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_10612"} +{"question": "What studies have been conducted in the field of visual representation learning under Weakly Supervised Learning (WSL)?", "answer": ["Learning Visual Features from Large Weakly Supervised Data", "Exploring the Limits of Weakly Supervised Pretraining", "Revisiting Unreasonable Effectiveness of Data in Deep Learning Era", "WebVision Database: Visual Learning and Understanding from Web Data", "Learning to Learn from Noisy Labeled Data"], "answer_arxiv_id": ["1511.02251", "1805.00932", "1707.02968", "1708.02862", "1812.05214"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_10613"} +{"question": "What are some studies that used user attributes and utterances from Reddit and Weibo for personalization in dialogue agents?", "answer": ["Training Millions of Personalized Dialogue Agents", "Less is More: Learning to Refine Dialogue History for Personalized\n Dialogue Generation", "Pchatbot: A Large-Scale Dataset for Personalized Chatbot"], "answer_arxiv_id": ["1809.01984", "2204.08128", "2009.13284"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_10614"} +{"question": "Could you provide me some studies that evaluated tool learning in existing LLMs?", "answer": ["ToolEyes: Fine-Grained Evaluation for Tool Learning Capabilities of\n Large Language Models in Real-world Scenarios", "RoTBench: A Multi-Level Benchmark for Evaluating the Robustness of Large\n Language Models in Tool Learning"], "answer_arxiv_id": ["2401.00741", "2401.08326"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_10615"} +{"question": "Which papers present coarse-grained PFL?", "answer": ["Personalized Federated Learning with Moreau Envelopes", "Ditto: Fair and Robust Federated Learning Through Personalization"], "answer_arxiv_id": ["2006.08848", "2012.04221"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_10616"} +{"question": "Could you provide me some studies that developed the first dynamic algorithms for monotone suodular maximization under a matroid constraint?", "answer": ["Dynamic Algorithms for Matroid Submodular Maximization", "Fully Dynamic Submodular Maximization over Matroids"], "answer_arxiv_id": ["2306.00959", "2305.19918"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_10617"} +{"question": "Which works introduced different types of image corruption for Diffusion Models?", "answer": ["Generative Modelling With Inverse Heat Dissipation", "Progressive Deblurring of Diffusion Models for Coarse-to-Fine Image Synthesis", "Soft Diffusion Score Matching For General Corruptions", "Blurring Diffusion Models"], "answer_arxiv_id": ["2206.13397", "2207.11192", "2209.05442", "2209.05557"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10618"} +{"question": "Can you name some works that have explored white-box adversarial attacks specifically on detection architectures?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1412.6572", "1706.06083"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_10619"} +{"question": "What papers explore the implicit regularization in descent algorithms through extra features?", "answer": ["The Implicit Bias of Gradient Descent on Separable Data", "Characterizing Implicit Bias in Terms of Optimization Geometry"], "answer_arxiv_id": ["1710.10345", "1802.08246"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_10620"} +{"question": "What studies involve LaGraph leveraging predictive self-supervised learning of GNNs to predict intractable latent graphs?", "answer": ["Self-Supervised Representation Learning via Latent Graph Prediction"], "answer_arxiv_id": ["2202.08333"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_10621"} +{"question": "What studies establish universal approximation results for transformers?", "answer": ["Are Transformers universal approximators of sequence-to-sequence functions?", "Statistically Meaningful Approximation: a Case Study on Approximating Turing Machines with Transformers"], "answer_arxiv_id": ["1912.10077", "2107.13163"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_10622"} +{"question": "Could you provide me some studies about the task of identifying a causal DAG over latent causal variables?", "answer": ["Learning latent causal graphs via mixture oracles"], "answer_arxiv_id": ["2106.15563"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_10623"} +{"question": "Which papers are about graph analogical reasoning using GNNs?", "answer": ["DeepGAR: Deep Graph Learning for Analogical Reasoning"], "answer_arxiv_id": ["2211.10821"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_10624"} +{"question": "What papers consider the standard diffusion models in their research?", "answer": ["MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "DiffusionDet: Diffusion Model for Object Detection"], "answer_arxiv_id": ["2208.15001", "2211.09788"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_10625"} +{"question": "Are there any works about bias mitigation strategies specifically developed for individual facial recognition technologies?", "answer": ["Understanding bias in facial recognition technologies", "Gender Classification and Bias Mitigation in Facial Images"], "answer_arxiv_id": ["2010.07023v1", "2007.06141"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_10626"} +{"question": "What work made use of a cross-attention layer during image-text matching process?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2201.12086"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_10627"} +{"question": "Could you provide me some studies where the researchers estimate the image-specific degradation model during the test time?", "answer": ["Blind Super-Resolution With Iterative Kernel Correction", "Blind Super-Resolution Kernel Estimation using an Internal-GAN", "Correction Filter for Single Image Super-Resolution: Robustifying Off-the-Shelf Deep Super-Resolvers", "Unfolding the Alternating Optimization for Blind Super Resolution", "Deep Constrained Least Squares for Blind Image Super-Resolution"], "answer_arxiv_id": ["1904.03377", "1909.06581", "1912.00157", "2010.02631", "2202.07508"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_10628"} +{"question": "What works studied scaling trends of zero-shot performance for different bit models?", "answer": ["GLM-130B: An Open Bilingual Pre-trained Model"], "answer_arxiv_id": ["2210.02414"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_10629"} +{"question": "Could you mention some studies where DRO is used for adversarial robustness?", "answer": ["A Unified Wasserstein Distributional Robustness Framework for\n Adversarial Training", "Certified Robust Neural Networks: Generalization and Corruption Resistance"], "answer_arxiv_id": ["2202.13437", "2303.02251v2"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_10630"} +{"question": "Could you provide me some GAN-based methods to address unpaired image-to-image translation problem?", "answer": ["Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks", "DualGAN: Unsupervised Dual Learning for Image-to-Image Translation", "Unsupervised Image-to-Image Translation Networks"], "answer_arxiv_id": ["1703.10593", "1704.02510", "1703.00848"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_10631"} +{"question": "Which research papers offer follow-ups and generalizations to the original work on Min Sum Set Cover?", "answer": ["Ranking with Submodular Valuations", "A note on the generalized min-sum set cover problem"], "answer_arxiv_id": ["1007.2503", "1107.2033"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10632"} +{"question": "What research papers introduced regularizations in fine-tuning to address the few data challenge on domain adaption?", "answer": ["Few-shot Image Generation via Cross-domain Correspondence", "Few Shot Generative Model Adaption via Relaxed Spatial Structural\n Alignment", "A Closer Look at Few-shot Image Generation", "Few-shot Image Generation via Adaptation-Aware Kernel Modulation", "Exploring Incompatible Knowledge Transfer in Few-shot Image Generation"], "answer_arxiv_id": ["2104.06820", "2203.04121", "2205.03805", "2210.16559", "2304.07574"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_10633"} +{"question": "Could you provide me a study about Diffusion GAN that perturbs the data with an adjustable number of steps?", "answer": ["Diffusion-GAN: Training GANs with Diffusion"], "answer_arxiv_id": ["2206.02262"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10634"} +{"question": "Any works about private feature selection in binary features and labels?", "answer": ["Differentially Private Algorithms for Empirical Machine Learning"], "answer_arxiv_id": ["1411.5428"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10635"} +{"question": "What papers have used the properties of the Hessian matrix spectrum to devise training methods?", "answer": ["Optimizing Neural Networks with Kronecker-factored Approximate Curvature", "Finding Approximate Local Minima Faster than Gradient Descent"], "answer_arxiv_id": ["1503.05671", "1611.01146"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_10636"} +{"question": "Could you provide some specific examples of works that use CLIP for high-quality image generation and editing?", "answer": ["VQGAN-CLIP: Open Domain Image Generation and Editing with Natural\n Language Guidance"], "answer_arxiv_id": ["2204.08583"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_10637"} +{"question": "What study introduced a reparameterization trick for μθ and the corresponding loss function to facilitate the training of diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_10638"} +{"question": "What studies have applied diffusion models for image restoration?", "answer": ["Image Super-Resolution via Iterative Refinement", "Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model"], "answer_arxiv_id": ["2104.07636", "2201.11793", "2212.00490"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_10639"} +{"question": "Which works showed that some detected OOD samples can turn out to be semantically similar to training samples?", "answer": ["On the Impact of Spurious Correlation for Out-of-distribution Detection", "Out-of-Distribution Detection with Deep Nearest Neighbors"], "answer_arxiv_id": ["2109.05642", "2204.06507"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_10640"} +{"question": "Can you provide some references about joint training of reader and retriever in retrieval-augmented language models?", "answer": ["REALM: Retrieval-Augmented Language Model Pre-Training"], "answer_arxiv_id": ["2002.08909"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10641"} +{"question": "What studies exist about sampling methods based on diffusion ODEs?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Diffusion-Based Voice Conversion with Fast Maximum Likelihood Sampling Scheme", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models"], "answer_arxiv_id": ["2011.13456", "2109.13821", "2201.06503", "2010.02502", "2211.01095"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_10642"} +{"question": "Could you provide me with a selection of works that propose the use of generative models for context-conditioned sequence generation?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Online Decision Transformer", "Multi-Game Decision Transformers", "Scaling Pareto-Efficient Decision Making via Offline Multi-Objective RL", "Masked Autoencoding for Scalable and Generalizable Decision Making", "Transformer Neural Processes: Uncertainty-Aware Meta Learning Via Sequence Modeling", "Generative Pretraining for Black-box Optimization", "Diffusion Models for Black-Box Optimization", "Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Is Conditional Generative Modeling all you need for Decision-Making?", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "Offline Reinforcement Learning via High-Fidelity Generative Behavior Modeling", "Learning Universal Policies via Text-Guided Video Generation"], "answer_arxiv_id": ["2106.01345", "2106.02039", "2202.05607", "2205.15241", "2305.00567", "2211.12740", "2207.04179", "2206.10786", "2306.07180", "2106.02039", "2211.15657", "2208.06193", "2209.14548", "2302.00111"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_10643"} +{"question": "Which works use explicit or implicit constraints or importance sampling with bounded ratio to ensure that the learned policy is close to the behavior policy?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning", "Way Off-Policy Batch Deep Reinforcement Learning of Implicit Human Preferences in Dialog", "Keep Doing What Worked: Behavioral Modelling Priors for Offline Reinforcement Learning", "EMaQ: Expected-Max Q-Learning Operator for Simple Yet Effective Offline and Online RL", "A Minimalist Approach to Offline Reinforcement Learning", "Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning", "Off-Policy Policy Gradient with State Distribution Correction", "AlgaeDICE: Policy Gradient from Arbitrary Experience", "GradientDICE: Rethinking Generalized Offline Estimation of Stationary Values", "GenDICE: Generalized Offline Estimation of Stationary Values", "OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation"], "answer_arxiv_id": ["1812.02900", "1906.00949", "1911.11361", "1907.00456", "2002.08396", "2007.11091", "2106.06860", "1910.00177", "1904.08473", "1912.02074", "2001.11113", "2002.09072", "2106.10783"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10644"} +{"question": "Any works about the application of strong data-augmentation in discriminative SSL methods?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "What Makes for Good Views for Contrastive Learning?"], "answer_arxiv_id": ["2006.07733", "2006.09882", "2005.10243"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_10645"} +{"question": "What studies proposed datasets generated by formula-driven supervised learning?", "answer": ["Replacing Labeled Real-image Datasets with Auto-generated Contours"], "answer_arxiv_id": ["2206.09132"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_10646"} +{"question": "What work pointed out the limitation of existing meta-learning methods showing optimal performance only when trained on similar curated tasks?", "answer": ["Few-shot Relational Reasoning via Connection Subgraph Pretraining"], "answer_arxiv_id": ["2210.06722"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_10647"} +{"question": "What works discuss the application of GFlowNets in biological sequence design?", "answer": ["Trajectory balance: Improved credit assignment in GFlowNets", "Biological Sequence Design with GFlowNets", "Learning GFlowNets From Partial Episodes For Improved Convergence And Stability"], "answer_arxiv_id": ["2201.13259", "2203.04115", "2209.12782"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_10648"} +{"question": "Which studies utilized the concept of graph neural networks (GNNs) to analyze brain network?", "answer": ["Brain Network Transformer"], "answer_arxiv_id": ["2210.06681"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_10649"} +{"question": "What studies discuss generalization bounds through algorithm stability?", "answer": ["Train faster, generalize better: Stability of stochastic gradient descent", "Generalization Bounds for Uniformly Stable Algorithms", "Generalization Bounds of SGLD for Non-convex Learning: Two Theoretical Viewpoints", "High probability generalization bounds for uniformly stable algorithms with nearly optimal rate", "Sharper bounds for uniformly stable algorithms", "On Generalization Error Bounds of Noisy Gradient Methods for Non-Convex Learning", "Fine-Grained Analysis of Stability and Generalization for Stochastic Gradient Descent", "Stability of Stochastic Gradient Descent on Nonsmooth Convex Losses", "Towards Understanding Generalization via Decomposing Excess Risk Dynamics"], "answer_arxiv_id": ["1509.01240", "1812.09859", "1707.05947", "1902.10710", "1910.07833", "1902.00621", "2006.08157", "2006.06914", "2106.06153"], "source_meta": {"published_time": "20220212"}, "qid": "AutoScholarQuery_train_10650"} +{"question": "What papers addressed the issue of objective mismatch in MBRL?", "answer": ["Minimax Model Learning", "Mismatched No More: Joint Model-Policy Optimization for Model-Based RL"], "answer_arxiv_id": ["2103.02084v1", "2110.02758"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_10651"} +{"question": "Any studies that discuss the L1 norm adaptive bounds and their implication of sparsity in online learning?", "answer": ["No-Regret Algorithms for Unconstrained Online Convex Optimization", "Sparsity regret bounds for individual sequences in online linear regression"], "answer_arxiv_id": ["1211.2260", "1101.1057v3"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10652"} +{"question": "Could you provide me with titles of papers that discuss neural network reassembly and stitching within the context of federated learning?", "answer": ["Revisiting Model Stitching to Compare Neural Representations", "Git Re-Basin: Merging Models Modulo Permutation Symmetries", "Deep Model Reassembly", "Stitchable Neural Networks", "On Cross-Layer Alignment for Model Fusion of Heterogeneous Neural Networks"], "answer_arxiv_id": ["2106.07682", "2209.04836", "2210.17409", "2302.06586", "2110.15538v3"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_10653"} +{"question": "What studies proposed unrolling optimization methods which leverage second-order techniques?", "answer": ["Transformer-Based Learned Optimization"], "answer_arxiv_id": ["2212.01055"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_10654"} +{"question": "What studies initiated the patch-based methods in local feature extractors?", "answer": ["3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions"], "answer_arxiv_id": ["1603.08182"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_10655"} +{"question": "What studies demonstrated improved load balancing of experts following work on SMoE?", "answer": ["BASE Layers: Simplifying Training of Large, Sparse Models"], "answer_arxiv_id": ["2103.16716"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10656"} +{"question": "Which work proposed a hybrid architecture combining grid-based and point-based approaches for radar-only 3D object detection?", "answer": ["Improved Orientation Estimation and Detection with Hybrid Object Detection Networks for Automotive Radar"], "answer_arxiv_id": ["2205.02111v2"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_10657"} +{"question": "Which works proposed the use of a goal-conditioned autotelic agent for the exploration task?", "answer": ["Understanding the World Through Action", "Autotelic Agents with Intrinsically Motivated Goal-Conditioned Reinforcement Learning: A Short Survey"], "answer_arxiv_id": ["2110.12543", "2012.09830"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_10658"} +{"question": "Could you give me examples of papers that aimed to introduce additional generation process controls in text-to-image generation?", "answer": ["BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained\n Diffusion", "Adding Conditional Control to Text-to-Image Diffusion Models", "Composer: Creative and Controllable Image Synthesis with Composable\n Conditions"], "answer_arxiv_id": ["2307.10816", "2302.05543", "2302.09778"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_10659"} +{"question": "What is the reference that considers offline RL with single-policy concentrability and achieves instance-dependent characterization?", "answer": ["Towards Instance-Optimal Offline Reinforcement Learning with Pessimism"], "answer_arxiv_id": ["2110.08695v1"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_10660"} +{"question": "What research papers talk about the impact of transfer learning on downstream tasks?", "answer": ["How transferable are features in deep neural networks?", "Do Better ImageNet Models Transfer Better?", "Taskonomy: Disentangling Task Transfer Learning"], "answer_arxiv_id": ["1411.1792", "1805.08974", "1804.08328"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_10661"} +{"question": "What papers have discussed deep clustering methods?", "answer": ["Unsupervised Deep Embedding for Clustering Analysis", "Towards K-means-friendly Spaces: Simultaneous Deep Learning and\n Clustering", "Deep Clustering for Unsupervised Learning of Visual Features"], "answer_arxiv_id": ["1511.06335", "1610.04794", "1807.05520"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_10662"} +{"question": "Which works have contributed to the study of angular super-resolution?", "answer": ["ShuffleUNet: Super resolution of diffusion-weighted MRIs using deep learning"], "answer_arxiv_id": ["2102.12898"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10663"} +{"question": "What studies found linear trends in the context of cross-benchmark evaluation?", "answer": ["Measuring Robustness to Natural Distribution Shifts in Image Classification", "Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization"], "answer_arxiv_id": ["2007.00644v2", "2107.04649"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_10664"} +{"question": "What works have explored Transformer's usability in the field of action understanding?", "answer": ["Attention Is All You Need", "End-to-End Object Detection with Transformers", "QPIC: Query-Based Pairwise Human-Object Interaction Detection with\n Image-Wide Contextual Information", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1706.03762", "2005.12872", "2103.05399", "2103.00020"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_train_10665"} +{"question": "Which researches proposed replacement options for softmax attention in Transformer models?", "answer": ["cosFormer : Rethinking Softmax in Attention"], "answer_arxiv_id": ["2202.08791"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_10666"} +{"question": "What works discuss relevance maps related to the class-agnostic nature of relevance?", "answer": ["Quantifying Attention Flow in Transformers", "A Survey on the Explainability of Supervised Machine Learning"], "answer_arxiv_id": ["2005.00928", "2011.07876"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_10667"} +{"question": "Which papers applied deep learning in point cloud processing?", "answer": ["PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "Point Transformer", "Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point\n Modeling"], "answer_arxiv_id": ["1706.02413", "2012.09164", "2111.14819"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_10668"} +{"question": "Are there any studies on addressing hallucination issues in MLLMs?", "answer": ["Analyzing and Mitigating Object Hallucination in Large Vision-Language\n Models", "Woodpecker: Hallucination Correction for Multimodal Large Language\n Models"], "answer_arxiv_id": ["2310.00754", "2310.16045"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10669"} +{"question": "Which studies demonstrate extremely quick render times with some achieving over 200 FPS?", "answer": ["FastNeRF: High-Fidelity Neural Rendering at 200FPS"], "answer_arxiv_id": ["2103.10380"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_10670"} +{"question": "What works studied the political bias in large language models?", "answer": ["Whose Opinions Do Language Models Reflect?"], "answer_arxiv_id": ["2303.17548"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_10671"} +{"question": "Which papers improve the retriever by the KL divergence between the retriever and the LLM?", "answer": ["REPLUG: Retrieval-Augmented Black-Box Language Models"], "answer_arxiv_id": ["2301.12652"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_train_10672"} +{"question": "Which works denote the challenges in answering knowledge-intensive questions and the need for robust document retrieval?", "answer": ["Can Pre-trained Vision and Language Models Answer Visual\n Information-Seeking Questions?", "Encyclopedic VQA: Visual questions about detailed properties of\n fine-grained categories"], "answer_arxiv_id": ["2302.11713", "2306.09224"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_10673"} +{"question": "Which research refined these results via a PAC-Bayesian approach?", "answer": ["A PAC-Bayesian Approach to Generalization Bounds for Graph Neural Networks", "Generalization in Graph Neural Networks: Improved PAC-Bayesian Bounds on Graph Diffusion"], "answer_arxiv_id": ["2012.07690", "2302.04451"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_10674"} +{"question": "Which works use a density model to estimate pseudo-count for each state in curiosity-driven exploration?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation"], "answer_arxiv_id": ["1606.01868"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_10675"} +{"question": "What papers talk about using variants of gradient descent in Wasserstein spaces for minimizing KL-divergence?", "answer": ["The Wasserstein Proximal Gradient Algorithm", "Projected Wasserstein gradient descent for high-dimensional Bayesian inference"], "answer_arxiv_id": ["2002.03035", "2102.06350v2"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_10676"} +{"question": "What papers studied representation learning for low-rank MDPs?", "answer": ["Model-free Representation Learning and Exploration in Low-rank MDPs", "FLAMBE: Structural Complexity and Representation Learning of Low Rank MDPs"], "answer_arxiv_id": ["2102.07035", "2006.10814"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10677"} +{"question": "Which publications featured discussions on autoencoders and reservoir computing in relation to deep random feature models?", "answer": ["Deep Randomized Neural Networks"], "answer_arxiv_id": ["2002.12287"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_10678"} +{"question": "Which study introduces a linear function approximation scheme known as embedded linear transition MDPs?", "answer": ["Reinforcement Learning in Feature Space: Matrix Bandit, Kernels, and Regret Bound"], "answer_arxiv_id": ["1905.10389"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10679"} +{"question": "Which papers have explored the concept of improving generalization in reinforcement learning using environment descriptions?", "answer": ["Learning to Win by Reading Manuals in a Monte-Carlo Framework", "Grounding Language to Entities and Dynamics for Generalization in Reinforcement Learning"], "answer_arxiv_id": ["1401.5390", "2101.07393"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_10680"} +{"question": "What research works focused on modifying the activation functions in implicit neural representations?", "answer": ["Implicit Neural Representations with Periodic Activation Functions", "A Structured Dictionary Perspective on Implicit Neural Representations", "Implicit Neural Representations with Periodic Activation Functions"], "answer_arxiv_id": ["2006.09661", "2112.01917", "2006.09661"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_10681"} +{"question": "Any works study the dynamics of non-contrastive learning but only focus on the predictor parameters?", "answer": ["Understanding Self-Supervised Learning Dynamics without Contrastive Pairs"], "answer_arxiv_id": ["2102.06810"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_10682"} +{"question": "Which works propose memory-based methods in continual learning?", "answer": ["Gradient Episodic Memory for Continual Learning", "End-to-End Incremental Learning", "Large Scale Incremental Learning", "Overcoming Catastrophic Forgetting with Unlabeled Data in the Wild"], "answer_arxiv_id": ["1706.08840", "1807.09536", "1905.13260", "1903.12648"], "source_meta": {"published_time": "20230807"}, "qid": "AutoScholarQuery_train_10683"} +{"question": "Could you provide me some works that applied the principles of neural scaling laws to Neural Machine Translation (NMT)?", "answer": ["Scaling Laws for Neural Machine Translation"], "answer_arxiv_id": ["2109.07740"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_10684"} +{"question": "What papers study online RL with additional access to an offline dataset?", "answer": ["Hybrid RL: Using Both Offline and Online Data Can Make RL Efficient", "Leveraging Offline Data in Online Reinforcement Learning"], "answer_arxiv_id": ["2210.06718", "2211.04974"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_10685"} +{"question": "Could you provide me some studies focused on how to specialize language models to the legal field?", "answer": ["When Does Pretraining Help? Assessing Self-Supervised Learning for Law and the CaseHOLD Dataset of 53,000+ Legal Holdings", "Legal Transformer Models May Not Always Help"], "answer_arxiv_id": ["2104.08671", "2109.06862"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_10686"} +{"question": "Which studies deal with scenarios of partially occluded states under common PO tasks?", "answer": ["Memory-based control with recurrent neural networks"], "answer_arxiv_id": ["1512.04455"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_10687"} +{"question": "Could you provide works on utilizing a neural network to help guide search for discrete programs, amortized inference?", "answer": ["CrossBeam: Learning to Search in Bottom-Up Program Synthesis"], "answer_arxiv_id": ["2203.10452"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_10688"} +{"question": "Which papers discuss about creating graph embedding by summing up or averaging all the node embeddings?", "answer": ["Benchmarking Graph Neural Networks"], "answer_arxiv_id": ["2003.00982"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_10689"} +{"question": "Could you give me some examples of studies that explored pretraining both the retriever and the reader on a vast, unlabeled corpus?", "answer": ["End-to-End Training of Neural Retrievers for Open-Domain Question\n Answering"], "answer_arxiv_id": ["2101.00408"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_10690"} +{"question": "Which research papers discuss traditional embodied question answering?", "answer": ["CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning", "Embodied Question Answering", "IQA: Visual Question Answering in Interactive Environments", "Multi-Target Embodied Question Answering"], "answer_arxiv_id": ["1612.06890", "1711.11543", "1712.03316", "1904.04686"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_10691"} +{"question": "Could you list some research papers that used replay methods and stored a generative model trained on data from previous tasks?", "answer": ["Continual Learning with Deep Generative Replay", "Continual Learning with Invertible Generative Models"], "answer_arxiv_id": ["1705.08690", "2202.05694"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_10692"} +{"question": "Which work presents the concept of learning the gradient of the action-distribution score function?", "answer": ["Diffusion Policy: Visuomotor Policy Learning via Action Diffusion"], "answer_arxiv_id": ["2303.04137"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_10693"} +{"question": "What works developed variants of CLIP to enhance the efficiency and performance of multi-modal pretraining?", "answer": ["SLIP: Self-supervision meets Language-Image Pre-training", "Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm", "FILIP: Fine-grained Interactive Language-Image Pre-Training"], "answer_arxiv_id": ["2112.12750", "2110.05208", "2111.07783"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_10694"} +{"question": "What studies have explored an effective strategy for domain generalization by learning domain invariant representations?", "answer": ["Domain Generalization via Invariant Feature Representation", "Domain-Adversarial Training of Neural Networks"], "answer_arxiv_id": ["1301.2115", "1505.07818"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_10695"} +{"question": "Who proposed TF-TAS as the first gradient-based zero-cost proxy specifically for ViT?", "answer": ["Training-free Transformer Architecture Search"], "answer_arxiv_id": ["2203.12217"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_10696"} +{"question": "What work used viewmaker networks in context of feature suppression?", "answer": ["Viewmaker Networks: Learning Views for Unsupervised Representation Learning"], "answer_arxiv_id": ["2010.07432"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_10697"} +{"question": "Can you list the studies that involve benchmarks with easily verifiable correct answers for evaluating recent LLMs?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2307.09288"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_10698"} +{"question": "Which research papers address the design of G-invariant architectures?", "answer": ["Deep Sets", "Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges", "Semi-Supervised Classification with Graph Convolutional Networks", "Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1703.06114", "2104.13478", "1609.02907", "1602.07576"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_10699"} +{"question": "Which work focuses on the improved efficiency of text-to-image generation models using Latent Diffusion Models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_10700"} +{"question": "Could you give me examples of research that have explored memorization and samples reconstruction in generative models, such as autoencoders and large language models?", "answer": ["Memorization in Overparameterized Autoencoders", "Extracting Training Data from Large Language Models", "The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks", "Quantifying Memorization Across Neural Language Models"], "answer_arxiv_id": ["1810.10333", "2012.07805", "1802.08232", "2202.07646"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_10701"} +{"question": "What works combined elements of free-form text and masks to improve image generation practices?", "answer": ["SpaText: Spatio-Textual Representation for Controllable Image Generation", "MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation"], "answer_arxiv_id": ["2211.14305", "2302.08113"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_10702"} +{"question": "Which papers discuss Bayesian methods of uncertainty estimation in neural networks?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samples"], "answer_arxiv_id": ["1703.04977", "1711.09325"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_10703"} +{"question": "Could you provide me some works that use probabilistic dynamics ensemble in MBRL methods?", "answer": ["Bidirectional Model-based Policy Optimization", "Model-Augmented Actor-Critic: Backpropagating through Paths", "How to Learn a Useful Critic? Model-based Action-Gradient-Estimator Policy Optimization", "On Effective Scheduling of Model-based Reinforcement Learning", "On-Policy Model Errors in Reinforcement Learning"], "answer_arxiv_id": ["2007.01995", "2005.08068", "2004.14309", "2111.08550", "2110.07985"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10704"} +{"question": "Which papers focused on policy-based method in the general setting?", "answer": ["Provably Correct Optimization and Exploration with Non-linear Policies"], "answer_arxiv_id": ["2103.11559"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_10705"} +{"question": "What papers investigated understanding neural networks by reverse engineering in the context of mechanistic interpretability?", "answer": ["In-context Learning and Induction Heads", "The Quantization Model of Neural Scaling", "Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small", "Toy Models of Superposition", "Progress measures for grokking via mechanistic interpretability", "A Toy Model of Universality: Reverse Engineering how Networks Learn Group Operations", "Omnigrok: Grokking Beyond Algorithmic Data", "Towards Automated Circuit Discovery for Mechanistic Interpretability"], "answer_arxiv_id": ["2209.11895", "2303.13506v3", "2211.00593", "2209.10652", "2301.05217", "2302.03025", "2210.01117", "2304.14997"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_10706"} +{"question": "Which studies propose the use of expert demonstrations in policy optimization via regularization towards a trained behavioral cloning policy?", "answer": ["Modeling Strong and Human-Like Gameplay with KL-Regularized Search"], "answer_arxiv_id": ["2112.07544"], "source_meta": {"published_time": "20230326"}, "qid": "AutoScholarQuery_train_10707"} +{"question": "Could you provide me some studies about gradient compression techniques used for bi-directional compression in the cross-silo setup or only uplink compression in both setups?", "answer": ["Federated Learning: Strategies for Improving Communication Efficiency", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "Distributed Mean Estimation with Limited Communication", "NUQSGD: Provably Communication-efficient Data-parallel SGD via Nonuniform Quantization", "DRIVE: One-bit Distributed Mean Estimation", "EDEN: Communication-Efficient and Robust Distributed Mean Estimation for Federated Learning", "Uncertainty Principle for Communication Compression in Distributed and Federated Learning and the Search for an Optimal Compressor"], "answer_arxiv_id": ["1610.05492", "1610.02132", "1611.00429", "1908.06077", "2105.08339", "2108.08842", "2002.08958"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_10708"} +{"question": "What studies focus on label inference methods like label propagation approaches, manifold regularization, Poisson learning, and deformed Laplacian regularization?", "answer": ["Poisson Learning: Graph Based Semi-Supervised Learning At Very Low Label Rates"], "answer_arxiv_id": ["2006.11184"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_10709"} +{"question": "Which studies used Recurrent Neural Networks for sequence-to-sequence tasks in human motion prediction?", "answer": ["Recurrent Network Models for Human Dynamics", "Structural-RNN: Deep Learning on Spatio-Temporal Graphs", "QuaterNet: A Quaternion-based Recurrent Model for Human Motion"], "answer_arxiv_id": ["1508.00271", "1511.05298", "1805.06485"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_10710"} +{"question": "What works are devoted to inference-efficient ensembling methods by sharing representations?", "answer": ["Why M Heads are Better than One: Training a Diverse Ensemble of Deep Networks", "An Ensemble with Shared Representations Based on Convolutional Networks for Continually Learning Facial Expressions", "Depth Uncertainty in Neural Networks", "Training independent subnetworks for robust prediction"], "answer_arxiv_id": ["1511.06314", "2103.03934", "2006.08437", "2010.06610"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_10711"} +{"question": "Can you mention any work that employed splitting the image into patches to capture the semantic information in images?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_10712"} +{"question": "What hardware architectures support int4 and int8 quantization?", "answer": ["APNN-TC: Accelerating Arbitrary Precision Neural Networks on Ampere GPU Tensor Cores"], "answer_arxiv_id": ["2106.12169"], "source_meta": {"published_time": "20220328"}, "qid": "AutoScholarQuery_train_10713"} +{"question": "Could you provide me some studies that investigate the behavior of importance weights in gradient-based variational inference?", "answer": ["Filtering Variational Objectives", "Importance Weighting and Variational Inference", "Challenges and Opportunities in High-dimensional Variational Inference", "On the Difficulty of Unbiased Alpha Divergence Minimization"], "answer_arxiv_id": ["1705.09279", "1808.09034", "2103.01085", "2010.09541"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_10714"} +{"question": "Could you tell me about a study related to information loss from information-rich domains to information-poor domains?", "answer": ["Asymmetric GAN for Unpaired Image-to-image Translation"], "answer_arxiv_id": ["1912.11660"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_10715"} +{"question": "Are there any works that introduce a frustum-based sampling strategy for NeRF-based anti-aliasing?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields"], "answer_arxiv_id": ["2103.13415"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_10716"} +{"question": "Any works about using mutual-information-based decoding techniques for promoting diversity or relevance in neural dialogue models?", "answer": ["A Diversity-Promoting Objective Function for Neural Conversation Models"], "answer_arxiv_id": ["1510.03055"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_10717"} +{"question": "What works are utilized in overcoming the major limitation of Neural Radiance Field (NeRF), in its requirement for accurate camera poses?", "answer": ["NeRF--: Neural Radiance Fields Without Known Camera Parameters", "BARF: Bundle-Adjusting Neural Radiance Fields", "Self-Calibrating Neural Radiance Fields", "SiNeRF: Sinusoidal Neural Radiance Fields for Joint Pose Estimation and\n Scene Reconstruction", "NeROIC: Neural Rendering of Objects from Online Image Collections", "Multiview Neural Surface Reconstruction by Disentangling Geometry and\n Appearance", "SPARF: Neural Radiance Fields from Sparse and Noisy Poses"], "answer_arxiv_id": ["2102.07064", "2104.06405", "2108.13826", "2210.04553", "2201.02533v2", "2003.09852", "2211.11738"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_10718"} +{"question": "Which works adapt methods from the centralized setting to handle group fairness in the Federated Learning setting?", "answer": ["Improving Fairness via Federated Learning", "Fairness-aware Agnostic Federated Learning", "Enforcing fairness in private federated learning via the modified method of differential multipliers", "FedFair: Training Fair Models In Cross-Silo Federated Learning", "Addressing Algorithmic Disparity and Performance Inconsistency in Federated Learning"], "answer_arxiv_id": ["2110.15545", "2010.05057", "2109.08604", "2109.05662", "2108.08435"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_10719"} +{"question": "What research papers have explored animatable 3D avatar generation leveraging parametric models for face and body?", "answer": ["OmniAvatar: Geometry-Guided Controllable 3D Head Synthesis", "EVA3D: Compositional 3D Human Generation from 2D Image Collections", "Unsupervised Learning of Efficient Geometry-Aware Neural Articulated Representations", "AvatarGen: A 3D Generative Model for Animatable Human Avatars", "Next3D: Generative Neural Texture Rasterization for 3D-Aware Head Avatars", "AG3D: Learning to Generate 3D Avatars from 2D Image Collections", "GETAvatar: Generative Textured Meshes for Animatable Human Avatars"], "answer_arxiv_id": ["2303.15539", "2210.04888", "2204.08839", "2211.14589", "2211.11208", "2305.02312", "2310.02714"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_10720"} +{"question": "Can you mention some studies about Locate-and-edit editors?", "answer": ["Knowledge Neurons in Pretrained Transformers", "Locating and Editing Factual Associations in GPT", "Mass-Editing Memory in a Transformer", "Knowledge Neurons in Pretrained Transformers"], "answer_arxiv_id": ["2104.08696", "2202.05262", "2210.07229", "2104.08696"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_10721"} +{"question": "What work has been done on proposing a sample-efficient scheme for unbound scenes?", "answer": ["Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields"], "answer_arxiv_id": ["2111.12077"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_10722"} +{"question": "What work presents a method for learning the reward or Q-function from expert data in imitation learning?", "answer": ["Generative Adversarial Imitation Learning", "IQ-Learn: Inverse soft-Q Learning for Imitation"], "answer_arxiv_id": ["1606.03476", "2106.12142"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_10723"} +{"question": "What studies reported the use of massive 3D object datasets in 2D-3D image editing?", "answer": ["OBJECT 3DIT: Language-guided 3D-aware Image Editing"], "answer_arxiv_id": ["2307.11073"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_10724"} +{"question": "Can you provide sources that researched privacy protection for data?", "answer": ["Visual Privacy Protection via Mapping Distortion", "Deep Leakage from Gradients"], "answer_arxiv_id": ["1911.01769", "1906.08935"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_10725"} +{"question": "What works are about supervised protein-ligand binding models trained on binding affinity data from PDBBind?", "answer": ["Protein-Ligand Scoring with Convolutional Neural Networks", "Multi-Scale Representation Learning on Proteins"], "answer_arxiv_id": ["1612.02751", "2204.02337"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_10726"} +{"question": "What studies proposed high weights for pairs of nearby points whenever their corresponding tangent planes are aligned?", "answer": ["Spectral Clustering Based on Local PCA"], "answer_arxiv_id": ["1301.2007"], "source_meta": {"published_time": "20210728"}, "qid": "AutoScholarQuery_train_10727"} +{"question": "Can you cite any papers that integrated attention mechanisms into the task of VSS to better exploit the temporal context?", "answer": ["Learning Local and Global Temporal Contexts for Video Semantic\n Segmentation", "Local Memory Attention for Fast Video Semantic Segmentation", "Mining Relations among Cross-Frame Affinities for Video Semantic\n Segmentation"], "answer_arxiv_id": ["2204.03330", "2101.01715", "2207.10436"], "source_meta": {"published_time": "20240127"}, "qid": "AutoScholarQuery_train_10728"} +{"question": "Which papers discuss the connection between in-context learning in large language models and Bayesian theory?", "answer": ["Meta-learning of Sequential Strategies", "Meta-trained agents implement Bayes-optimal agents", "Transformers Can Do Bayesian Inference", "An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["1905.03030", "2010.11223", "2112.10510", "2111.02080"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_10729"} +{"question": "Could you provide me some studies that unified the reconstruction methods and gave a universal sampling framework for reconstructing nearly all classes of functions with Fourier-based smoothness constraints?", "answer": ["A Universal Sampling Method for Reconstructing Signals with Simple Fourier Transforms"], "answer_arxiv_id": ["1812.08723"], "source_meta": {"published_time": "20220225"}, "qid": "AutoScholarQuery_train_10730"} +{"question": "Can you provide examples of studies that employ a local equivariant feature extractor for object bounding box prediction?", "answer": ["Rotationally Equivariant 3D Object Detection"], "answer_arxiv_id": ["2204.13630"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10731"} +{"question": "What studies investigated the robustness of LLMs under distribution shift?", "answer": ["Generalizing to Unseen Domains: A Survey on Domain Generalization", "GLUE-X: Evaluating Natural Language Understanding Models from an\n Out-of-distribution Generalization Perspective"], "answer_arxiv_id": ["2103.03097", "2211.08073"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_10732"} +{"question": "What research references focus on quantifying the uncertainty of model’s predictions toward next states and rewards?", "answer": ["MOPO: Model-based Offline Policy Optimization", "Offline Reinforcement Learning from Images with Latent Space Models", "COMBO: Conservative Offline Model-Based Policy Optimization"], "answer_arxiv_id": ["2005.13239", "2012.11547", "2102.08363"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_10733"} +{"question": "What research proposed a modification called Variational Structured Dropout using variational inference?", "answer": ["Structured Dropout Variational Inference for Bayesian Neural Networks"], "answer_arxiv_id": ["2102.07927v4"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10734"} +{"question": "Any works about Matching against outliers especially for vision tasks?", "answer": ["Learning to Find Good Correspondences", "Learning Two-View Correspondences and Geometry Using Order-Aware Network"], "answer_arxiv_id": ["1711.05971", "1908.04964"], "source_meta": {"published_time": "20201216"}, "qid": "AutoScholarQuery_train_10735"} +{"question": "What research has been conducted to solve inverse problems using data-driven methods?", "answer": ["Learning to Invert: Signal Recovery via Deep Convolutional Networks"], "answer_arxiv_id": ["1701.03891"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_10736"} +{"question": "What work proposes the use of a Gaussian distribution to generate augmented features in Few-Shot Learning?", "answer": ["Free Lunch for Few-shot Learning: Distribution Calibration"], "answer_arxiv_id": ["2101.06395"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_10737"} +{"question": "Where does the concept of distributionally robust optimization as an approach for conformal prediction under distribution shift has been introduced?", "answer": ["Robust Validation: Confident Predictions Even When Distributions Shift"], "answer_arxiv_id": ["2008.04267"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_10738"} +{"question": "In what papers PC models for memory can be formulated as recurrent networks?", "answer": ["Learning on Arbitrary Graph Topologies via Predictive Coding"], "answer_arxiv_id": ["2201.13180"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_10739"} +{"question": "Which works in visual concept learning discuss concept bottleneck models?", "answer": ["Concept Bottleneck Models", "Discovering and Explaining the Representation Bottleneck of DNNs", "Explainable Neural Network-based Modulation Classification via Concept Bottleneck Models"], "answer_arxiv_id": ["2007.04612", "2111.06236", "2101.01239"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_10740"} +{"question": "Which studies attempted to decrease reliance on labeled data in NR-PCQA?", "answer": ["No-Reference Point Cloud Quality Assessment via Domain Adaptation"], "answer_arxiv_id": ["2112.02851"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10741"} +{"question": "Could you provide some studies that applied clustering algorithms for common region discovery?", "answer": ["Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "PiCIE: Unsupervised Semantic Segmentation using Invariance and Equivariance in Clustering", "Object discovery and representation networks", "Self-Supervised Visual Representation Learning with Semantic Grouping"], "answer_arxiv_id": ["2203.08414", "2103.17070", "2203.08777", "2205.15288"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_10742"} +{"question": "Can you provide me papers that eliminate the requirement of using negative samples in Siamese Framework?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2006.07733", "2011.10566"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_10743"} +{"question": "Which works decompose a neural network into task-specific modules and a shared feature extractor using manually designed heuristics in Multi-Task Learning?", "answer": ["UberNet : Training a ‘Universal’ Convolutional Neural Network for Low-, Mid-, and High-Level Vision using Diverse Datasets and Limited Memory", "Learning Multiple Tasks with Multilinear Relationship Networks", "Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Generalist Convolution Kernels"], "answer_arxiv_id": ["1609.02132", "1506.02117", "1908.09597"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_10744"} +{"question": "Can you provide examples of work that model the motion of dynamic objects for NVS?", "answer": ["DynIBaR: Neural Dynamic Image-Based Rendering"], "answer_arxiv_id": ["2211.11082"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_10745"} +{"question": "What research introduced visual features into textual space by inserting cross-attention layers into LLMs?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_10746"} +{"question": "What work provides the sample complexity of shallow Transformer to study its generalization property?", "answer": ["A Theoretical Understanding of Shallow Vision Transformers: Learning, Generalization, and Sample Complexity"], "answer_arxiv_id": ["2302.06015v3"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_10747"} +{"question": "Can you provide examples of recent research that explored constructing nonlinear manifolds via autoencoder neural networks?", "answer": ["Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders"], "answer_arxiv_id": ["1812.08373"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_10748"} +{"question": "Which studies focus on the use of a disentangled conditional diffusion model for multi-contrast brain MRI SR?", "answer": ["DisC-Diff: Disentangled Conditional Diffusion Model for Multi-Contrast\n MRI Super-Resolution"], "answer_arxiv_id": ["2303.13933"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_10749"} +{"question": "What works proposed NeRF as a novel scene representation method and its variants?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Fourier Features Let Networks Learn High Frequency Functions in Low\n Dimensional Domains", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image\n Synthesis", "BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale\n Scene Rendering", "Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual\n Fly-Throughs", "Block-NeRF: Scalable Large Scene Neural View Synthesis", "RigNeRF: Fully Controllable Neural 3D Portraits", "Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar\n Reconstruction", "Animatable Neural Radiance Fields for Modeling Dynamic Human Bodies", "Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "HDR-NeRF: High Dynamic Range Neural Radiance Fields", "Learning Object-Compositional Neural Radiance Field for Editable Scene\n Rendering", "NeRF-Editing: Geometry Editing of Neural Radiance Fields", "Removing Objects From Neural Radiance Fields", "Clutter Detection and Removal in 3D Scenes with View-Consistent\n Inpainting", "Neural Reflectance Fields for Appearance Acquisition", "NeRD: Neural Reflectance Decomposition from Image Collections"], "answer_arxiv_id": ["2003.08934", "2006.10739", "2110.08985", "2112.05504", "2112.10703", "2202.05263", "2206.06481", "2012.03065", "2105.02872", "2012.15838", "2111.14451", "2109.01847", "2205.04978", "2212.11966", "2304.03763", "2008.03824", "2012.03918"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_10750"} +{"question": "Are there any works that has studied the factors that influences the performance of In-context learning (ICL)?", "answer": ["Systematic Generalization and Emergent Structures in Transformers Trained on Structured Tasks", "On the Effect of Pretraining Corpora on In-context Learning by a Large-scale Language Model", "Language Models are Few-Shot Learners", "An Explanation of In-context Learning as Implicit Bayesian Inference", "Emergent Abilities of Large Language Models", "Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity", "Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?", "What Makes Good In-Context Examples for GPT-3?", "Are Emergent Abilities of Large Language Models a Mirage?"], "answer_arxiv_id": ["2210.00400", "2204.13509", "2005.14165", "2111.02080", "2206.07682", "2104.08786", "2202.12837", "2101.06804", "2304.15004"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10751"} +{"question": "What works in Imitation Learning require the stationary state distribution of the current policy to be close to that of the expert policy?", "answer": ["State-only Imitation with Transition Dynamics Mismatch", "State Alignment-based Imitation Learning"], "answer_arxiv_id": ["2002.11879", "1911.10947"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_10752"} +{"question": "Which papers adopted the unconstrained features model, in which the final layer features are freely modified during training, in the study of MSE-NC?", "answer": ["Neural Collapse Under MSE Loss: Proximity to and Dynamics on the Central Path", "Neural collapse with unconstrained features", "Extended Unconstrained Features Model for Exploring Deep Neural Collapse", "Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training"], "answer_arxiv_id": ["2106.02073", "2011.11619", "2202.08087", "2101.12699"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10753"} +{"question": "Could you give me examples of research that develop language-conditioned policies with Transformers backbones?", "answer": ["Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation"], "answer_arxiv_id": ["2209.05451"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_10754"} +{"question": "Can you name the works that introduced the pipelines, Mask3D and SPFormer?", "answer": ["Superpoint Transformer for 3D Scene Instance Segmentation"], "answer_arxiv_id": ["2211.15766"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_10755"} +{"question": "Could you provide me some works that analyzed the significant disparity of standard accuracy and robust accuracy among different classes or subgroups of data for adversarially trained models?", "answer": ["To be Robust or to be Fair: Towards Fairness in Adversarial Training"], "answer_arxiv_id": ["2010.06121"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_10756"} +{"question": "What researches have proposed the concept of deriving algorithms for offline RL in settings where trajectories might have missing actions?", "answer": ["Semi-Supervised Offline Reinforcement Learning with Action-Free Trajectories"], "answer_arxiv_id": ["2210.06518"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_10757"} +{"question": "What works are related to the Dynamic bootstrapping technique?", "answer": ["Unsupervised Label Noise Modeling and Loss Correction", "Webly Supervised Image Classification with Self-Contained Confidence"], "answer_arxiv_id": ["1904.11238", "2008.11894"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_10758"} +{"question": "Could you provide me some works about syntax knowledge injection in pre-training, such as incorporating dependency relation prediction or syntax tree information?", "answer": ["K-Adapter: Infusing Knowledge into Pre-Trained Models with Adapters", "Syntax-BERT: Improving Pre-trained Transformers with Syntax Trees"], "answer_arxiv_id": ["2002.01808", "2103.04350"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_10759"} +{"question": "Which works propose to train 3D models without 3D supervision via differentiable rendering?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Learning to Infer Implicit Surfaces without 3D Supervision", "Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision"], "answer_arxiv_id": ["2003.08934", "1911.00767", "1912.07372"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_10760"} +{"question": "What works have been focusing on aspects, such as LiDAR representations, in LiDAR segmentation model?", "answer": ["Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution", "Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR\n Segmentation", "PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds\n Semantic Segmentation", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "KPConv: Flexible and Deformable Convolution for Point Clouds", "Scan-based Semantic Segmentation of LiDAR Point Clouds: An Experimental\n Study", "Point Transformer V2: Grouped Vector Attention and Partition-based\n Pooling"], "answer_arxiv_id": ["2007.16100", "2011.10033", "2003.14032", "1904.08755", "1904.08889", "2004.11803", "2210.05666"], "source_meta": {"published_time": "20240502"}, "qid": "AutoScholarQuery_train_10761"} +{"question": "Which studies focus on offline data poisoning attacks?", "answer": ["Poisoning Attacks against Support Vector Machines", "Support Vector Machines under Adversarial Label Contamination", "Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning", "Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization", "Policy Poisoning in Batch Reinforcement Learning and Control"], "answer_arxiv_id": ["1206.6389", "2206.00352", "1712.05526", "1708.08689", "1910.05821"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_10762"} +{"question": "What papers provide a type of molecular representation in 1D string, 2D image, and 3D geometry for drug discovery?", "answer": ["MolGenSurvey: A Systematic Survey in Machine Learning Models for Molecule Design"], "answer_arxiv_id": ["2203.14500"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_10763"} +{"question": "What researches propose rectifying the training procedure for the classification network based on meta-learning framework?", "answer": ["Learning to Rectify for Robust Learning with Noisy Labels", "Dynamic Loss For Robust Learning"], "answer_arxiv_id": ["2111.04239", "2211.12506"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_10764"} +{"question": "What paper introduced the concept of data-dependent dimensionality reduction for preserving pair-wise/multi-wise similarity in indexing big data?", "answer": ["A Survey on Learning to Hash"], "answer_arxiv_id": ["1606.00185"], "source_meta": {"published_time": "20200720"}, "qid": "AutoScholarQuery_train_10765"} +{"question": "Which works proposed the use of generative methods for anomaly detection by using reconstruction as the learning objective?", "answer": ["Generative Adversarial Active Learning for Unsupervised Outlier Detection"], "answer_arxiv_id": ["1809.10816"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10766"} +{"question": "Which works explore the concept of 3D-aware image editing?", "answer": ["StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis", "Efficient Geometry-aware 3D Generative Adversarial Networks", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "DreamFusion: Text-to-3D using 2D Diffusion", "Zero-1-to-3: Zero-shot One Image to 3D Object", "Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion Prior"], "answer_arxiv_id": ["2110.08985", "2112.07945", "2212.00774v1", "2209.14988", "2303.11328v1", "2303.14184"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_10767"} +{"question": "What research papers have proposed methods for automated test case generation for statically typed programming languages?", "answer": ["Many Independent Objective (MIO) Algorithm for Test Suite Generation"], "answer_arxiv_id": ["1901.01541"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_10768"} +{"question": "Can you list papers where optimal transport was used in feature matching?", "answer": ["SuperGlue: Learning Feature Matching with Graph Neural Networks", "LoFTR: Detector-Free Local Feature Matching with Transformers"], "answer_arxiv_id": ["1911.11763", "2104.00680"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_10769"} +{"question": "What works propose translation language model (TLM) task for cross-lingual token alignment?", "answer": ["Cross-lingual Language Model Pretraining"], "answer_arxiv_id": ["1901.07291"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_10770"} +{"question": "What works are about tuning-free methods for text-to-image diffusion models?", "answer": ["InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "Taming Encoder for Zero Fine-tuning Image Customization with\n Text-to-Image Diffusion Models", "FastComposer: Tuning-Free Multi-Subject Image Generation with Localized\n Attention"], "answer_arxiv_id": ["2304.03411", "2304.02642", "2305.10431"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_10771"} +{"question": "Any works that further expands these concepts to present KFLR and KFRA?", "answer": ["Practical Gauss-Newton Optimisation for Deep Learning"], "answer_arxiv_id": ["1706.03662"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_10772"} +{"question": "What works investigate Lipschitz-constrained networks in the context of their robustness?", "answer": ["Sorting Out Lipschitz Function Approximation", "Achieving robustness in classification using optimal transport with hinge regularization"], "answer_arxiv_id": ["1811.05381", "2006.06520"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_10773"} +{"question": "In which studies has the prequential perspective with the block-wise approximation technique been used in the context of image-classification?", "answer": ["Small Data, Big Decisions: Model Selection in the Small-Data Regime"], "answer_arxiv_id": ["2009.12583"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_10774"} +{"question": "Are there studies that used sentence translations from different languages to construct positive pairs in a multilingual scenario?", "answer": ["Contrastive Learning for Many-to-many Multilingual Neural Machine\n Translation", "InfoXLM: An Information-Theoretic Framework for Cross-Lingual Language\n Model Pre-Training", "On Learning Universal Representations Across Languages"], "answer_arxiv_id": ["2105.09501", "2007.07834", "2007.15960"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_10775"} +{"question": "Which paper introduced the DIR algorithm dealing with imbalanced regression?", "answer": ["Delving into Deep Imbalanced Regression"], "answer_arxiv_id": ["2102.09554"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_10776"} +{"question": "Could you provide me some works on privacy preserving action recognition?", "answer": ["Real-world Anomaly Detection in Surveillance Videos", "The Kinetics Human Action Video Dataset", "Privacy-Preserving Human Activity Recognition from Extreme Low\n Resolution", "Extreme Low Resolution Activity Recognition with Multi-Siamese Embedding\n Learning", "Semi-Coupled Two-Stream Fusion ConvNets for Action Recognition at\n Extremely Low Resolutions", "Learning to Anonymize Faces for Privacy Preserving Action Detection", "Privacy-Preserving Deep Action Recognition: An Adversarial Learning\n Framework and A New Dataset", "SPAct: Self-supervised Privacy Preservation for Action Recognition", "PrivHAR: Recognizing Human Actions From Privacy-preserving Lens", "STPrivacy: Spatio-Temporal Privacy-Preserving Action Recognition", "TeD-SPAD: Temporal Distinctiveness for Self-supervised\n Privacy-preservation for video Anomaly Detection", "Privacy-Preserving Action Recognition via Motion Difference Quantization"], "answer_arxiv_id": ["1801.04264", "1705.06950", "1604.03196", "1708.00999", "1610.03898", "1803.11556", "1906.05675", "2203.15205", "2206.03891", "2301.03046", "2308.11072", "2208.02459"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_10777"} +{"question": "Which papers have demonstrated diffusion models that outperform Generative Adversarial Networks on text-to-image generation in the MSCOCO domain?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_10778"} +{"question": "Are there any works that optimize a texture map and render it onto a specified 3D model?", "answer": ["FCA: Learning a 3D Full-coverage Vehicle Camouflage for Multi-view\n Physical Adversarial Attack", "DTA: Physical Camouflage Attacks using Differentiable Transformation\n Network", "ACTIVE: Towards Highly Transferable 3D Physical Camouflage for Universal\n and Robust Vehicle Evasion"], "answer_arxiv_id": ["2109.07193", "2203.09831", "2308.07009"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_10779"} +{"question": "What work proposed a regularization scheme to mitigate discriminator overfitting?", "answer": ["Regularizing Generative Adversarial Networks under Limited Data"], "answer_arxiv_id": ["2104.03310"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_10780"} +{"question": "Which work has made attempts to construct continuous models using recent emergence of implicit neural functions?", "answer": ["ImFace: A Nonlinear 3D Morphable Face Model with Implicit Neural\n Representations", "i3DMM: Deep Implicit 3D Morphable Model of Human Heads", "Learning Neural Parametric Head Models"], "answer_arxiv_id": ["2203.14510", "2011.14143", "2212.02761"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_10781"} +{"question": "What work proposes the use of two complementary prompt pools to encode task-invariant and task-specific information?", "answer": ["DualPrompt: Complementary Prompting for Rehearsal-free Continual\n Learning"], "answer_arxiv_id": ["2204.04799"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_10782"} +{"question": "Which papers discuss different non-stationarity conditions and Continual RL methods?", "answer": ["Towards Continual Reinforcement Learning: A Review and Perspectives"], "answer_arxiv_id": ["2012.13490"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_10783"} +{"question": "Could you provide me with references where causal inference is used to generate counterfactual explanations?", "answer": ["Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR"], "answer_arxiv_id": ["1711.00399v3"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_10784"} +{"question": "Which studies discuss the concept of knowledge distillation in terms of class logit alignment?", "answer": ["Distilling the Knowledge in a Neural Network", "DistilBERT, a distilled version of BERT: smaller, faster, cheaper and\n lighter"], "answer_arxiv_id": ["1503.02531", "1910.01108"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_10785"} +{"question": "Which works showed that automated annotations for fairness evaluations have limitations and can lead to missing annotations for underrepresented groups?", "answer": ["A Step Toward More Inclusive People Annotations for Fairness"], "answer_arxiv_id": ["2105.02317"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_10786"} +{"question": "What works involve the study of poisoning attack on multi-agent reinforcement learners?", "answer": ["Adversarial Policies: Attacking Deep Reinforcement Learning"], "answer_arxiv_id": ["1905.10615"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_10787"} +{"question": "Which works focus on the neural parameterizations of symbolic grammars?", "answer": ["Recurrent Neural Network Grammars", "Unsupervised Recurrent Neural Network Grammars", "Compound Probabilistic Context-Free Grammars for Grammar Induction", "The Return of Lexical Dependencies: Neural Lexicalized PCFGs", "PCFGs Can Do Better: Inducing Probabilistic Context-Free Grammars with Many Symbols", "Neural Bi-Lexicalized PCFG Induction", "Unsupervised Discontinuous Constituency Parsing with Mildly Context-Sensitive Grammars"], "answer_arxiv_id": ["1602.07776", "1904.03746", "1906.10225", "2007.15135", "2104.13727", "2105.15021", "2212.09140v2"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10788"} +{"question": "Any studies proposed hyperbolic kernel SVM for nonlinear classification without resorting to ill-fitting tools developed for Euclidean space?", "answer": ["Large-Margin Classification in Hyperbolic Space"], "answer_arxiv_id": ["1806.00437"], "source_meta": {"published_time": "20220214"}, "qid": "AutoScholarQuery_train_10789"} +{"question": "Which research papers studied the aspect of contextual information in non-stationary multi-armed bandit literature?", "answer": ["Efficient Contextual Bandits in Non-stationary Worlds", "Weighted Linear Bandits for Non-Stationary Environments"], "answer_arxiv_id": ["1708.01799v4", "1909.09146"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_10790"} +{"question": "Could you provide some works that employed assembling snapshots into tensors for factorization?", "answer": ["Embedding Models for Episodic Knowledge Graphs"], "answer_arxiv_id": ["1807.00228"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_10791"} +{"question": "Are there any works that extended the studies of online RL to the case with function approximation?", "answer": ["Provably Efficient Reinforcement Learning with Linear Function Approximation", "Learning Near Optimal Policies with Low Inherent Bellman Error", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Sample-Efficient Reinforcement Learning Is Feasible for Linearly Realizable MDPs with Limited Revisiting", "Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms", "The Statistical Complexity of Interactive Decision Making"], "answer_arxiv_id": ["1907.05388", "2003.00153", "2012.08507", "2105.08024", "2102.00815", "2112.13487v3"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_10792"} +{"question": "Which research papers have mainly focused on symmetries of the joint state-action space of a Markov Decision Process (MDP)?", "answer": ["Plannable Approximations to MDP Homomorphisms: Equivariance under Actions", "Trajectory Prediction using Equivariant Continuous Convolution", "Equivariant Networks for Zero-Shot Coordination", "SO⁢(2)-Equivariant Reinforcement Learning", "On-Robot Learning With Equivariant Models", "Hyperbolic Deep Reinforcement Learning", "Continuous MDP Homomorphisms and Homomorphic Policy Gradient"], "answer_arxiv_id": ["2002.11963", "2010.11344", "2210.12124", "2203.04439", "2203.04923", "2210.01542", "2209.07364v1"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_10793"} +{"question": "Can you list studies that used prototype-based methods in few-shot class incremental learning?", "answer": ["Few-Shot Lifelong Learning", "Constrained Few-shot Class-incremental Learning", "Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class\n Incremental Learning"], "answer_arxiv_id": ["2103.00991", "2203.16588", "2302.03004"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_10794"} +{"question": "Which works attempted to generalize mean-field analysis to deep networks?", "answer": ["A Rigorous Framework for the Mean Field Limit of Multilayer Neural Networks", "Global Convergence of Three-layer Neural Networks in the Mean Field Regime", "A mean-field limit for certain deep neural networks", "Mean Field Analysis of Deep Neural Networks", "Modeling from Features: a Mean-field Framework for Over-parameterized Deep Neural Networks", "Overparameterization of deep ResNet: zero loss and mean-field analysis"], "answer_arxiv_id": ["2001.11443v3", "2105.05228v1", "1906.00193", "1903.04440", "2007.01452", "2105.14417"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_10795"} +{"question": "Any studies about language model inconsistency in explanation generation?", "answer": ["Make Up Your Mind! Adversarial Generation of Inconsistent Natural Language Explanations"], "answer_arxiv_id": ["1910.03065"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_10796"} +{"question": "Which works have proposed vehicular camera-based models for 3D object detection?", "answer": ["DEVIANT: Depth EquiVarIAnt NeTwork for Monocular 3D Object Detection", "MonoPGC: Monocular 3D Object Detection with Pixel Geometry Contexts"], "answer_arxiv_id": ["2207.10758", "2302.10549"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_10797"} +{"question": "Which prior studies have shown that deep networks trained for classification tasks tend to concentrate around their respective class means, which are maximally distant from each other?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning training"], "answer_arxiv_id": ["2008.08186"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_10798"} +{"question": "What papers talked about efficient modality bridging and adaptation via routing and skipping with adapters?", "answer": ["Cheap and Quick: Efficient Vision-Language Instruction Tuning for Large\n Language Models", "Parameter and Computation Efficient Transfer Learning for\n Vision-Language Pre-trained Models"], "answer_arxiv_id": ["2305.15023", "2309.01479"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_10799"} +{"question": "Which studies attempted to extend Armijo line search to the stochastic setting?", "answer": ["Painless Stochastic Gradient: Interpolation, Line-Search, and Convergence Rates"], "answer_arxiv_id": ["1905.09997"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10800"} +{"question": "Which studies have used the diffusion processes for text-to-video (T2V) generation?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data", "VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing"], "answer_arxiv_id": ["2209.14792", "2303.08320", "2303.13439", "2303.09535"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_10801"} +{"question": "Which papers introduced initial work on foundation LLMs, trained on massive amounts of textual data?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2005.14165", "2204.02311", "2302.13971"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_10802"} +{"question": "What papers introduce color-based features of GMs?", "answer": ["Detecting GAN generated Fake Images using Co-occurrence Matrices", "Detecting High-Quality GAN-Generated Face Images using Neural Networks"], "answer_arxiv_id": ["1903.06836", "2203.01716"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_10803"} +{"question": "What works have studied the extrapolation capabilities of graph neural networks?", "answer": ["From Local Structures to Size Generalization in Graph Neural Networks", "How Neural Networks Extrapolate: From Feedforward to Graph Neural Networks"], "answer_arxiv_id": ["2010.08853", "2009.11848"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10804"} +{"question": "Are there any studies that compare their method's performance with CrAM in terms of one-shot pruning?", "answer": ["Soft Threshold Weight Reparameterization for Learnable Sparsity", "Dynamic Model Pruning with Feedback", "Movement Pruning: Adaptive Sparsity by Fine-Tuning", "PLATON: Pruning Large Transformer Models with Upper Confidence Bound of Weight Importance"], "answer_arxiv_id": ["2002.03231", "2006.07253", "2005.07683", "2206.12562"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_10805"} +{"question": "What works dealt with predicting model performance in a shifted target domain?", "answer": ["Leveraging Unlabeled Data to Predict Out-of-Distribution Performance", "Predicting with Confidence on Unseen Distributions", "Mandoline: Model Evaluation under Distribution Shift"], "answer_arxiv_id": ["2201.04234", "2107.03315", "2107.00643"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_10806"} +{"question": "Could you mention some papers that have utilized inpainting for synthesizing anonymous faces?", "answer": ["Natural and Effective Obfuscation by Head Inpainting", "DeepPrivacy: A Generative Adversarial Network for Face Anonymization", "LDFA: Latent Diffusion Face Anonymization for Self-driving Applications"], "answer_arxiv_id": ["1711.09001", "1909.04538", "2302.08931"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_10807"} +{"question": "What research works can you list under video-language pre-trained models?", "answer": ["InternVideo: General Video Foundation Models via Generative and Discriminative Learning", "MERLOT Reserve: Neural Script Knowledge through Vision and Language and Sound", "Temporal Perceiving Video-Language Pre-training", "An Empirical Study of End-to-End Video-Language Transformers with Masked Visual Modeling", "VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding", "All in One: Exploring Unified Video-Language Pre-training", "mPLUG-2: A Modularized Multi-modal Foundation Model Across Text, Image and Video", "HiTeA: Hierarchical Temporal-Aware Video-Language Pre-training", "Lavender: Unifying Video-Language Understanding as Masked Language Modeling"], "answer_arxiv_id": ["2212.03191", "2201.02639", "2301.07463", "2209.01540", "2109.14084", "2203.07303", "2302.00402", "2212.14546", "2206.07160"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_10808"} +{"question": "What research has been done on building Multimodal Large Language Models by connecting visual encoders with powerful LLMs?", "answer": ["LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "Language Is Not All You Need: Aligning Perception with Language Models", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2303.16199", "2304.14178", "2209.06794", "2302.14045", "2305.03726"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_10809"} +{"question": "Are there any research papers on panoptic segmentation?", "answer": ["Panoptic Segmentation", "UPSNet: A Unified Panoptic Segmentation Network", "Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up\n Panoptic Segmentation", "Fully Convolutional Networks for Panoptic Segmentation"], "answer_arxiv_id": ["1801.00868", "1901.03784", "1911.10194", "2012.00720"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_10810"} +{"question": "Which studies discuss the potential harms and benefits of data memorization even in non-verbatim cases?", "answer": ["Preventing Verbatim Memorization in Language Models Gives a False Sense of Privacy"], "answer_arxiv_id": ["2210.17546v3"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_10811"} +{"question": "What research is there on policy learning and evaluation in unconfounded POMDPs?", "answer": ["Closing the Learning-Planning Loop with Predictive State Representations", "Tensor Decompositions for Learning Latent Variable Models", "A PAC RL Algorithm for Episodic POMDPs", "Reinforcement Learning of POMDPs using Spectral Methods", "Sample-Efficient Reinforcement Learning of Undercomplete POMDPs", "RL for Latent MDPs: Regret Guarantees and a Lower Bound"], "answer_arxiv_id": ["0912.2385v1", "1210.7559", "1605.08062v2", "1602.07764", "2006.12484", "2102.04939"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_10812"} +{"question": "Which papers originalized the study of function approximation scenarios in 'sample-efficient RL with function approximation'?", "answer": ["Optimism in Reinforcement Learning with Generalized Linear Function Approximation", "Sample-Optimal Parametric Q-Learning Using Linearly Additive Features", "Provably Efficient Exploration in Policy Optimization", "Provably Efficient Reinforcement Learning with Linear Function Approximation", "Frequentist Regret Bounds for Randomized Least-Squares Value Iteration", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "A Theoretical Analysis of Optimistic Proximal Policy Optimization in Linear Markov Decision Processes"], "answer_arxiv_id": ["1912.04136", "1902.04779", "1912.05830", "1907.05388", "1911.00567v7", "2006.01107", "1910.10597", "2012.08507", "2305.08841"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_10813"} +{"question": "What works have been made on image-to-image translation using diffusion models?", "answer": ["Palette: Image-to-Image Diffusion Models", "Dual Diffusion Implicit Bridges for Image-to-Image Translation", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2111.05826", "2203.08382", "2302.05543"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_10814"} +{"question": "Could you cite the papers that highlight the challenges faced by equivariant architectures, such as limited expressive power?", "answer": ["How Powerful are Graph Neural Networks?", "Provably Powerful Graph Networks", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "Rethinking the Expressive Power of GNNs via Graph Biconnectivity", "On the Expressive Power of Geometric Graph Neural Networks"], "answer_arxiv_id": ["1810.00826", "1905.11136", "1810.02244", "2301.09505", "2301.09308v3"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_10815"} +{"question": "Any papers describing the use of Visual Attention Convolutional AutoEncoders for image-to-image translation in Zero-Shot Sketch-Based Image Retrieval?", "answer": ["A Zero-Shot Framework for Sketch-based Image Retrieval", "Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1807.11724", "1312.6114"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_10816"} +{"question": "Which paper studied the mapping of coordinates to Fourier features to overcome spectral bias in MLPs?", "answer": ["Fourier Features Let Networks Learn High Frequency Functions in Low\n Dimensional Domains"], "answer_arxiv_id": ["2006.10739"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_10817"} +{"question": "What works are about using discrete latent quantities, or tokens, in generative transformer models for image?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "MaskGIT: Masked Generative Image Transformer"], "answer_arxiv_id": ["2112.10752", "2202.04200"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_10818"} +{"question": "Which works propose to extend text with logical formulas for data augmentation in sequence-based models?", "answer": ["Logic-Driven Context Extension and Data Augmentation for Logical\n Reasoning of Text"], "answer_arxiv_id": ["2105.03659"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_10819"} +{"question": "In what works the phase-based methods in flow-agnostic approaches are introduced to predict the phase decomposition of the intermediate frame?", "answer": ["PhaseNet for Video Frame Interpolation"], "answer_arxiv_id": ["1804.00884"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_10820"} +{"question": "What papers introduced the first category of UGC databases collected from real-world media platforms?", "answer": ["YouTube UGC Dataset for Video Compression Research", "Patch-VQ: 'Patching Up' the Video Quality Problem"], "answer_arxiv_id": ["1904.06457", "2011.13544"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_10821"} +{"question": "Which works contributed to deriving generalization bounds for parametric models trained with stochastic gradient descent (SGD)?", "answer": ["Train faster, generalize better: Stability of stochastic gradient descent"], "answer_arxiv_id": ["1509.01240"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_10822"} +{"question": "Are there any papers that proposed independent approaches from marginal contributions in the context of data valuation?", "answer": ["Data Valuation using Reinforcement Learning", "Measuring the Effect of Training Data on Deep Learning Predictions via Randomized Experiments"], "answer_arxiv_id": ["1909.11671", "2206.10013"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_10823"} +{"question": "Which works focus on improving the visual quality through the use of GANs in face swapping and expression reenactment?", "answer": ["SimSwap: An Efficient Framework For High Fidelity Face Swapping", "Smooth-Swap: A Simple Enhancement for Face-Swapping with Smoothness", "FaceShifter: Towards High Fidelity And Occlusion Aware Face Swapping", "Few-Shot Head Swapping in the Wild", "Region-Aware Face Swapping", "BlendFace: Re-designing Identity Encoders for Face-Swapping", "ReenactGAN: Learning to Reenact Faces via Boundary Transfer", "Deep Video Portraits", "GANimation: Anatomically-aware Facial Animation from a Single Image", "Recycle-GAN: Unsupervised Video Retargeting", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "StarGAN: Unified Generative Adversarial Networks for Multi-Domain\n Image-to-Image Translation"], "answer_arxiv_id": ["2106.06340", "2112.05907", "1912.13457", "2204.13100", "2203.04564", "2307.10854", "1807.11079", "1805.11714", "1807.09251", "1808.05174", "1703.10593", "1711.09020"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_10824"} +{"question": "Which papers explore the concept of chain-of-thought prompting in relation to Large Language Models (LLMs)?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2210.03493"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_10825"} +{"question": "Could you provide references that reported the performance of the Perceiver model on optical flow estimation and the GLUE language benchmark?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language\n Understanding"], "answer_arxiv_id": ["1804.07461"], "source_meta": {"published_time": "20240718"}, "qid": "AutoScholarQuery_train_10826"} +{"question": "What papers link Koopman theory with deep learning approaches?", "answer": ["Deep learning for universal linear embeddings of nonlinear dynamics", "Learning Koopman Invariant Subspaces for Dynamic Mode Decomposition", "Learning Deep Neural Network Representations for Koopman Operators of Nonlinear Dynamical Systems"], "answer_arxiv_id": ["1712.09707", "1710.04340", "1708.06850"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10827"} +{"question": "What papers proposed using Transformer Architecture as sequence prediction for layout generation?", "answer": ["LayoutTransformer: Layout Generation and Completion with Self-attention", "BLT: Bidirectional Layout Transformer for Controllable Layout Generation", "Variational Transformer Networks for Layout Generation", "Constrained Graphic Layout Generation via Latent Optimization", "The Layout Generation Algorithm of Graphic Design Based on Transformer-CVAE", "Spot the Error: Non-autoregressive Graphic Layout Generation with\n Wireframe Locator"], "answer_arxiv_id": ["2006.14615", "2112.05112", "2104.02416", "2108.00871", "2110.06794v2", "2401.16375"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_10828"} +{"question": "Which papers discuss leveraging domain-invariant features under self-training methodology of Unsupervised Domain Adaptation?", "answer": ["Exploiting Domain-Specific Features to Enhance Domain Generalization"], "answer_arxiv_id": ["2110.09410"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_10829"} +{"question": "Which work is recognised as a pioneering work in diffusion models trained on multi-view posed data?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2303.11328"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_10830"} +{"question": "What papers discuss the Thompson Sampling Algorithms in the context of regret minimization?", "answer": ["A Tutorial on Thompson Sampling", "Analysis of Thompson Sampling for the multi-armed bandit problem", "Thompson Sampling: An Asymptotically Optimal Finite Time Analysis", "An Information-Theoretic Analysis of Thompson Sampling"], "answer_arxiv_id": ["1707.02038", "1111.1797", "1205.4217", "1403.5341"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_10831"} +{"question": "Could you cite some works where rationales assisted humans in their tasks?", "answer": ["Leveraging Rationales to Improve Human Task Performance", "Are Machine Rationales (Not) Useful to Humans? Measuring and Improving\n Human Utility of Free-Text Rationales"], "answer_arxiv_id": ["2002.04202", "2305.07095"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_train_10832"} +{"question": "Which studies discuss deep generative models in molecule generation?", "answer": ["An Equivariant Generative Framework for Molecular Graph-Structure Co-Design", "Exploring Chemical Space with Score-based Out-of-distribution Generation", "MARS: Markov Molecular Sampling for Multi-objective Drug Discovery"], "answer_arxiv_id": ["2304.12436", "2206.07632", "2103.10432"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_10833"} +{"question": "What research has been done on automatic design of biological sequences using Bayesian optimization techniques in machine learning?", "answer": ["The reparameterization trick for acquisition functions", "BOSS: Bayesian Optimization over String Spaces"], "answer_arxiv_id": ["1712.00424", "2010.00979"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_10834"} +{"question": "Which studies have used a 'digital twin' approach to compare the learning abilities of newborn animals and machines?", "answer": ["Controlled-rearing studies of newborn chicks and deep neural networks", "Modeling Object Recognition in Newborn Chicks using Deep Neural Networks"], "answer_arxiv_id": ["2112.06106", "2106.07185v1"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_10835"} +{"question": "Which research paper collected a new dynamic graph dataset for anomalous node detection in financial networks?", "answer": ["DGraph: A Large-Scale Financial Dataset for Graph Anomaly Detection"], "answer_arxiv_id": ["2207.03579"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_10836"} +{"question": "What papers have worked on pretraining a multimodal encoder-decoder model with early vision-language fusion?", "answer": ["SimVLM: Simple Visual Language Model Pretraining with Weak Supervision"], "answer_arxiv_id": ["2108.10904"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_10837"} +{"question": "Which studies designed auxiliary tasks such as pixel or latent reconstruction to enhance sample efficiency?", "answer": ["Improving Sample Efficiency in Model-Free Reinforcement Learning from Images", "Mask-based Latent Reconstruction for Reinforcement Learning"], "answer_arxiv_id": ["1910.01741", "2201.12096"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_10838"} +{"question": "What studies have pushed the limits of post-training quantization to 4-bit on traditional models by using layer-wise calibration?", "answer": ["Improving Post Training Neural Quantization: Layer-wise Calibration and Integer Programming"], "answer_arxiv_id": ["2006.10518"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_10839"} +{"question": "Could you provide me with any prior works on efficient conformal inferences from pre-trained black-box models in the context of classification?", "answer": ["Classification with Valid and Adaptive Coverage", "Uncertainty Sets for Image Classifiers using Conformal Prediction"], "answer_arxiv_id": ["2006.02544", "2009.14193"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_10840"} +{"question": "Which papers show that small units like characters or bytes often underperform subword models in text representation?", "answer": ["A Multi-dimensional Evaluation of Tokenizer-free Multilingual Pretrained\n Models", "CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language\n Representation"], "answer_arxiv_id": ["2210.07111", "2103.06874"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10841"} +{"question": "Which research introduced behaviors that can be implanted into systems such that a specific trigger feature in an input causes an unexpected output behavior, also known as trojans or backdoors?", "answer": ["Rethinking Backdoor Attacks"], "answer_arxiv_id": ["2307.10163"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_10842"} +{"question": "What recent work presented OmniMotion, an optimization method that relies on a volumetric representation to estimate motion across every pixel and every frame in a video?", "answer": ["Tracking Everything Everywhere All at Once"], "answer_arxiv_id": ["2306.05422"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_10843"} +{"question": "What studies have extended NeRF to model dynamic urban scenes?", "answer": ["Neural Scene Graphs for Dynamic Scenes", "Towards Efficient Neural Scene Graphs by Learning Consistency Fields"], "answer_arxiv_id": ["2011.10379", "2210.04127"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_10844"} +{"question": "Which paper showed that for a truncated mixture of two Gaussians, the output by the EM algorithm converges to the true mean with increasing iterations?", "answer": ["On the Analysis of EM for truncated mixtures of two Gaussians"], "answer_arxiv_id": ["1902.06958"], "source_meta": {"published_time": "20230506"}, "qid": "AutoScholarQuery_train_10845"} +{"question": "Could you provide some works in the field of spatial super-resolution methods in the natural image domain?", "answer": ["Deep Learning for Image Super-resolution: A Survey"], "answer_arxiv_id": ["1902.06068"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_10846"} +{"question": "What papers showcased the recent work on neural network scaling laws?", "answer": ["Explaining Neural Scaling Laws", "Learning Curve Theory"], "answer_arxiv_id": ["2102.06701", "2102.04074"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_10847"} +{"question": "What studies proposed explicit NeRFs to accelerate both training and rendering speed?", "answer": ["Neural Sparse Voxel Fields", "Plenoxels: Radiance Fields without Neural Networks", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction", "Improved Direct Voxel Grid Optimization for Radiance Fields Reconstruction", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2007.11571", "2112.05131", "2111.11215", "2206.05085", "2201.05989", "2203.09517"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_10848"} +{"question": "Which works generate a domain-specific meaning representation by constraining the decoder using a Context-Free Grammar (CFG)?", "answer": ["Few-Shot Semantic Parsing with Language Models Trained on Code"], "answer_arxiv_id": ["2112.08696"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_10849"} +{"question": "Who developed Sound2sight, TATS, and Tempotokens?", "answer": ["Sound2Sight: Generating Visual Dynamics from Sound and Context", "Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive\n Transformer", "Diverse and Aligned Audio-to-Video Generation via Text-to-Video Model\n Adaptation"], "answer_arxiv_id": ["2007.12130", "2204.03638", "2309.16429"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_10850"} +{"question": "Could you show me the work that introduced the first comprehensive benchmark for large language model watermarks?", "answer": ["WaterBench: Towards Holistic Evaluation of Watermarks for Large Language\n Models"], "answer_arxiv_id": ["2311.07138"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_10851"} +{"question": "Which research papers explored the usage of neural fields for training with any-resolution images?", "answer": ["Learning Continuous Image Representation with Local Implicit Image\n Function", "Any-resolution Training for High-resolution Image Synthesis", "Arbitrary-Scale Image Synthesis"], "answer_arxiv_id": ["2012.09161", "2204.07156", "2204.02273"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_10852"} +{"question": "Which pioneering works explore using auxiliary objectives as a self-supervision signal?", "answer": ["Learning to Navigate in Complex Environments", "Reinforcement Learning with Unsupervised Auxiliary Tasks", "Loss is its own Reward: Self-Supervision for Reinforcement Learning", "Playing FPS Games with Deep Reinforcement Learning"], "answer_arxiv_id": ["1611.03673", "1611.05397", "1612.07307", "1609.05521"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_10853"} +{"question": "What studies have adopted Mixup as a component of learning frameworks in various fields like vision, language, and graph?", "answer": ["Manifold Mixup: Better Representations by Interpolating Hidden States", "CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features", "Puzzle Mix: Exploiting Saliency and Local Statistics for Optimal Mixup", "SSMix: Saliency-Based Span Mixup for Text Classification", "Enhancing Cross-lingual Transfer by Manifold Mixup", "GraphMix: Improved Training of GNNs for Semi-Supervised Learning"], "answer_arxiv_id": ["1806.05236", "1905.04899", "2009.06962v2", "2106.08062", "2205.04182", "1909.11715"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_10854"} +{"question": "What works focus on structural pruning to remove redundant channels(filters) of a CNN?", "answer": ["Pruning Filters for Efficient ConvNets", "Importance Estimation for Neural Network Pruning", "CHIP: CHannel Independence-based Pruning for Compact Neural Networks", "Learning Structured Sparsity in Deep Neural Networks", "Discrimination-aware Channel Pruning for Deep Neural Networks", "Compressing Image-to-Image Translation GANs Using Local Density\n Structures on Their Learned Manifold"], "answer_arxiv_id": ["1608.08710", "1906.10771", "2110.13981", "1608.03665", "1810.11809", "2312.14776"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_10855"} +{"question": "Where was it demonstrated that using GPT-3 for generating rich text descriptions improved performance in zero-shot image classification tasks?", "answer": ["What does a platypus look like? Generating customized prompts for\n zero-shot image classification"], "answer_arxiv_id": ["2209.03320"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_10856"} +{"question": "Which paper originally proposed the Mixup technique?", "answer": ["mixup: Beyond Empirical Risk Minimization"], "answer_arxiv_id": ["1710.09412"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_10857"} +{"question": "Which research refined and extended the analysis of Successive Rejects (SR) by involving KL-divergences?", "answer": ["On Best-Arm Identification with a Fixed Budget in Non-Parametric Multi-Armed Bandits"], "answer_arxiv_id": ["2210.00895v2"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_10858"} +{"question": "What papers provide examples for the widespread adoption of VQ-VAE or discrete VAE techniques in multi-modal models?", "answer": ["Zero-Shot Text-to-Image Generation", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Taming Transformers for High-Resolution Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2206.10789", "2012.09841", "2112.10752"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_10859"} +{"question": "What papers proposed the implicit curriculum method in goal relabeling scheme which samples goals from failed trajectories in GCRL?", "answer": ["Hindsight Experience Replay", "Competitive experience replay", "Goal-conditioned Imitation Learning", "Visual Reinforcement Learning with Imagined Goals"], "answer_arxiv_id": ["1707.01495v3", "1902.00528", "1906.05838", "1807.04742"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_10860"} +{"question": "Which papers have discussed fine-tuning as an approach for localization?", "answer": ["F-VLM: Open-Vocabulary Object Detection upon Frozen Vision and Language\n Models", "Simple Open-Vocabulary Object Detection with Vision Transformers"], "answer_arxiv_id": ["2209.15639", "2205.06230"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_10861"} +{"question": "What references repurposed pre-trained text-to-image models for classification, segmentation, or multimodal reasoning?", "answer": ["Text-to-Image Diffusion Models are Zero-Shot Classifiers", "Unleashing Text-to-Image Diffusion Models for Visual Perception", "Are Diffusion Models Vision-And-Language Reasoners?"], "answer_arxiv_id": ["2303.15233", "2303.02153v1", "2305.16397"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10862"} +{"question": "Could you provide a study that involves pretrained language encoder models like BERT?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_10863"} +{"question": "Could you list some works which apply behavior regularization in offline RL?", "answer": ["A Minimalist Approach to Offline Reinforcement Learning", "Behavior Regularized Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["2106.06860", "1911.11361", "2110.06169"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_10864"} +{"question": "Any existing studies that prove the effectiveness of strategy that augments original query question with a context formed by natural language demonstrations?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?", "Learning To Retrieve Prompts for In-Context Learning"], "answer_arxiv_id": ["2101.06804", "2112.08633"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10865"} +{"question": "Which works are relevant to improving the performance of CNNs?", "answer": ["Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs", "Visual Attention Network", "Focal Modulation Networks", "HorNet: Efficient High-Order Spatial Interactions with Recursive Gated\n Convolutions", "More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using\n Sparsity", "MetaFormer Baselines for Vision", "Conv2Former: A Simple Transformer-Style ConvNet for Visual Recognition", "InternImage: Exploring Large-Scale Vision Foundation Models with\n Deformable Convolutions"], "answer_arxiv_id": ["2203.06717", "2202.09741", "2203.11926", "2207.14284", "2207.03620", "2210.13452", "2211.11943", "2211.05778"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_10866"} +{"question": "Which work established a theoretical framework for multiclass boosting generalizing previous learning conditions?", "answer": ["A Theory of Multiclass Boosting"], "answer_arxiv_id": ["1108.2989"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_10867"} +{"question": "Which works introduced the formal notions of replicability that are strongly related to robustness, privacy, and generalization?", "answer": ["Reproducibility in Learning", "Stability is Stable: Connections between Replicability, Privacy, and Adaptive Generalization", "Statistical Indistinguishability of Learning Algorithms"], "answer_arxiv_id": ["2201.08430v2", "2303.12921", "2305.14311"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_10868"} +{"question": "Can you list the papers that have provided magnitude-based methods which outperform other methods at scale?", "answer": ["The State of Sparsity in Deep Neural Networks"], "answer_arxiv_id": ["1902.09574"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_10869"} +{"question": "Which studies focused on addressing open-vocabulary segmentation and object-centric segmentation by leveraging pre-trained text-to-image models?", "answer": ["Diffusion Models for Zero-Shot Open-Vocabulary Segmentation", "DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic\n Segmentation Using Diffusion Models", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models", "DifFSS: Diffusion Model for Few-Shot Semantic Segmentation"], "answer_arxiv_id": ["2306.09316", "2303.11681", "2303.04803", "2307.00773"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_10870"} +{"question": "Could you provide me some studies about online time series forecasting in real-world applications?", "answer": ["Online Learning for Time Series Prediction", "Learning Fast and Slow for Online Time Series Forecasting"], "answer_arxiv_id": ["1302.6927", "2202.11672"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_10871"} +{"question": "Are there papers proposing the scalable graph transformer model?", "answer": ["NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification", "Recipe for a General, Powerful, Scalable Graph Transformer"], "answer_arxiv_id": ["2306.08385", "2205.12454"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_10872"} +{"question": "What works have proposed human-centered metrics for evaluating explanation interpretability?", "answer": ["What I Cannot Predict, I Do Not Understand: A Human-Centered Evaluation\n Framework for Explainability Methods"], "answer_arxiv_id": ["2112.04417"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_10873"} +{"question": "Which research used GPT-4’s adeptness in understanding multimodal textual representations for visual instruction tuning?", "answer": ["Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_10874"} +{"question": "In what works is the problem of limited generalizability of learned hate speech patterns from one dialect to another elaborated?", "answer": ["Detecting Cross-Geographic Biases in Toxicity Modeling on Social Media"], "answer_arxiv_id": ["2104.06999"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_10875"} +{"question": "Which papers proposed recurrent neural networks-based models for Neural Temporal Point Processes (TPPs)?", "answer": ["The Neural Hawkes Process: A Neurally Self-Modulating Multivariate Point Process", "Modeling The Intensity Function Of Point Process Via Recurrent Neural Networks", "Fully Neural Network based Model for General Temporal Point Processes", "Intensity-Free Learning of Temporal Point Processes", "Neural Datalog Through Time: Informed Temporal Modeling via Logical Specification", "User-Dependent Neural Sequence Models for Continuous-Time Event Data"], "answer_arxiv_id": ["1612.09328", "1705.08982", "1905.09690", "1909.12127", "2006.16723", "2011.03231"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_10876"} +{"question": "Which works assume that voters have normalized utilities when studying the distortion?", "answer": ["Optimized Distortion and Proportional Fairness in Voting"], "answer_arxiv_id": ["2205.15760"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_10877"} +{"question": "Could you name the study that constituted a general-purpose visual assistant by exploring diverse multi-modal instruction-following data?", "answer": ["Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_10878"} +{"question": "Which work before ours also disentangled memory and credit assignment in benchmarks?", "answer": ["Behaviour Suite for Reinforcement Learning"], "answer_arxiv_id": ["1908.03568"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_10879"} +{"question": "In which papers were structure-preserving editing methods via Stable Diffusion models attempted?", "answer": ["Imagic: Text-Based Real Image Editing with Diffusion Models", "Prompt-to-Prompt Image Editing with Cross Attention Control"], "answer_arxiv_id": ["2210.09276", "2208.01626"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_10880"} +{"question": "Which works have leveraged audio to enhance activity recognition on video datasets?", "answer": ["EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action\n Recognition", "Attention Bottlenecks for Multimodal Fusion", "Listen to Look: Action Recognition by Previewing Audio"], "answer_arxiv_id": ["1908.08498", "2107.00135", "1912.04487"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_10881"} +{"question": "What works introduced a keypoint estimator to enhance the visual sign representation in the context of SLT?", "answer": ["Neural Sign Language Translation based on Human Keypoint Estimation", "Two-Stream Network for Sign Language Recognition and Translation"], "answer_arxiv_id": ["1811.11436", "2211.01367"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_10882"} +{"question": "Which research combines neural radiance fields (NeRF) with a 3D morphable face model (3DMM) to create diverse expression and perspectives in selfies?", "answer": ["FLAME-in-NeRF : Neural control of Radiance Fields for Free View Face\n Animation"], "answer_arxiv_id": ["2108.04913"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_10883"} +{"question": "What works have suggested the use of varied models at different steps in the diffusion process?", "answer": ["OMS-DPM: Optimizing the Model Schedule for Diffusion Probabilistic\n Models"], "answer_arxiv_id": ["2306.08860"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_10884"} +{"question": "Which papers have studied federated learning as a collaborative learning framework?", "answer": ["Advances and Open Problems in Federated Learning", "Learning Differentially Private Recurrent Language Models"], "answer_arxiv_id": ["1912.04977", "1710.06963"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_10885"} +{"question": "What works focused on changing the optimization at training time by adding regularizers?", "answer": ["RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "CamP: Camera Preconditioning for Neural Radiance Fields", "LOLNeRF: Learn from One Look"], "answer_arxiv_id": ["2112.00724", "2111.12077", "2308.10902", "2111.09996"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_10886"} +{"question": "What papers have employed ensembles as a teacher model for knowledge distillation?", "answer": ["Ensemble Distribution Distillation", "Scaling Ensemble Distribution Distillation to Many Classes with Proxy Targets"], "answer_arxiv_id": ["1905.00076", "2105.06987"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_10887"} +{"question": "Which studies have focused on prompt learning to improve classification accuracy in language models?", "answer": ["Learning to Prompt for Vision-Language Models", "Prompt Distribution Learning"], "answer_arxiv_id": ["2109.01134", "2205.03340"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_10888"} +{"question": "Could you provide me some papers that studied the robustness of trajectory prediction models against various adversarial attacks?", "answer": ["On Adversarial Robustness of Trajectory Prediction for Autonomous\n Vehicles", "AdvDO: Realistic Adversarial Attacks for Trajectory Prediction", "Vehicle trajectory prediction works, but not everywhere"], "answer_arxiv_id": ["2201.05057", "2209.08744", "2112.03909"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_10889"} +{"question": "Are there any works that have tackled the issue of temporal asynchrony in cooperative detection?", "answer": ["Where2comm: Communication-Efficient Collaborative Perception via Spatial Confidence Maps", "Cooperative Perception for 3D Object Detection in Driving Scenarios using Infrastructure Sensors", "Collaboration Helps Camera Overtake LiDAR in 3D Detection"], "answer_arxiv_id": ["2209.12836", "1912.12147", "2303.13560"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_10890"} +{"question": "What works presented the application of Mixture of Experts in Transformer structure of large language models?", "answer": ["GShard: Scaling Giant Models with Conditional Computation and Automatic\n Sharding", "Switch Transformers: Scaling to Trillion Parameter Models with Simple\n and Efficient Sparsity", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts", "Mixture-of-Experts Meets Instruction Tuning:A Winning Combination for\n Large Language Models"], "answer_arxiv_id": ["2006.16668", "2101.03961", "2112.06905", "2305.14705"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_10891"} +{"question": "Which study constructed coresets of size O for general discrete metric?", "answer": ["A Unified Framework for Approximating and Clustering Data"], "answer_arxiv_id": ["1106.1379"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_10892"} +{"question": "What research work involves learning a per-instance free parameter or an MLP network for re-weighting examples and enhancing robust representations?", "answer": ["Learning to Reweight Examples for Robust Deep Learning", "Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting"], "answer_arxiv_id": ["1803.09050", "1902.07379"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_10893"} +{"question": "Which research works propose a frequency enhanced attention technique calculated in the frequency domain using DFT to obtain attentive weights by the spectrums of queries and keys?", "answer": ["FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting"], "answer_arxiv_id": ["2201.12740"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_10894"} +{"question": "Can you name some studies that introduced new image editing methodologies using large-scale diffusion models?", "answer": ["T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Multi-Concept Customization of Text-to-Image Diffusion", "Diffusion Self-Guidance for Controllable Image Generation"], "answer_arxiv_id": ["2302.08453", "2210.09276", "2208.01626", "2212.04488", "2306.00986"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_10895"} +{"question": "What papers showcase the usage of contextual embeddings produced by language models such as BiLSTM and various Transformers?", "answer": ["Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond", "Language-agnostic BERT Sentence Embedding", "Unsupervised Cross-lingual Representation Learning at Scale", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "mT5: A Massively Multilingual Pre-trained Text-to-Text Transformer", "Multilingual Universal Sentence Encoder for Semantic Retrieval"], "answer_arxiv_id": ["1812.10464", "2007.01852", "1911.02116", "1810.04805", "2010.11934", "1907.04307"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_10896"} +{"question": "What is the original work that employed contrastive learning strategy on a web-scale dataset with 400 million image-text pairs?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_10897"} +{"question": "What research papers identified several mislabeled images in ImageNet which hampers robust classification?", "answer": ["When does dough become a bagel? Analyzing the remaining mistakes on ImageNet", "Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarks", "Are we done with ImageNet?", "Re-labeling ImageNet: from Single to Multi-Labels, from Global to Localized Labels"], "answer_arxiv_id": ["2205.04596", "2103.14749", "2006.07159", "2101.05022"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_10898"} +{"question": "What are the works related to the field of mechanistic interpretability aiming to understand neural networks?", "answer": ["Progress measures for grokking via mechanistic interpretability", "Towards Automated Circuit Discovery for Mechanistic Interpretability", "Emergent World Representations: Exploring a Sequence Model Trained on a\n Synthetic Task", "The Geometry of Truth: Emergent Linear Structure in Large Language Model\n Representations of True/False Datasets", "Locating and Editing Factual Associations in GPT", "A Glitch in the Matrix? Locating and Detecting Language Model Grounding\n with Fakepedia"], "answer_arxiv_id": ["2301.05217", "2304.14997", "2210.13382", "2310.06824", "2202.05262", "2312.02073"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_10899"} +{"question": "What are the works that have designed simulators considering performance and realism?", "answer": ["Habitat 2.0: Training Home Assistants to Rearrange their Habitat"], "answer_arxiv_id": ["2106.14405"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_10900"} +{"question": "Are there any texts that raise concerns regarding the usage of logits in federated learning regarding privacy and security?", "answer": ["Privacy and Robustness in Federated Learning: Attacks and Defenses", "Vertical Federated Learning without Revealing Intersection Membership"], "answer_arxiv_id": ["2012.06337", "2106.05508"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_10901"} +{"question": "Are there any studies that proposed frame-based unsupervised optical flow networks to eliminate the need for labeled data?", "answer": ["UnFlow: Unsupervised Learning of Optical Flow with a Bidirectional Census Loss", "SelFlow: Self-Supervised Learning of Optical Flow", "Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimation", "What Matters in Unsupervised Optical Flow", "UPFlow: Upsampling Pyramid for Unsupervised Optical Flow Learning", "SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping"], "answer_arxiv_id": ["1711.07837", "1904.09117", "2003.13045", "2006.04902", "2012.00212", "2105.07014"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_10902"} +{"question": "Which study provided an exponential simple regret bound under an added assumption that the optimality gap is bounded by a semi-metric?", "answer": ["Bayesian Optimization with Exponential Convergence"], "answer_arxiv_id": ["1604.01348"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_10903"} +{"question": "What paper details the performance of the Swin V2 detector achieved through large-scale pre-training?", "answer": ["Swin Transformer V2: Scaling Up Capacity and Resolution"], "answer_arxiv_id": ["2111.09883"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_10904"} +{"question": "Which paper proposes the TRIME training approach for training language models with memory augmentation?", "answer": ["Training Language Models with Memory Augmentation"], "answer_arxiv_id": ["2205.12674"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_10905"} +{"question": "Which study manually crafted regularization terms added to the loss function?", "answer": ["Unifying physical systems’ inductive biases in neural ODE using dynamics constraints"], "answer_arxiv_id": ["2208.02632"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_10906"} +{"question": "Which paper involves rendering 3D textured objects and scenes with radiance fields using MLPs?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_10907"} +{"question": "What research leans on pre-training with large-scale internet data and apply pre-trained visual representations to the robotic domain?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "Simple but Effective: CLIP Embeddings for Embodied AI", "The Unsurprising Effectiveness of Pre-Trained Vision Models for Control", "R3M: A Universal Visual Representation for Robot Manipulation", "Real-World Robot Learning with Masked Visual Pre-training", "Masked Visual Pre-training for Motor Control", "VIP: Towards Universal Visual Reward and Representation via\n Value-Implicit Pre-Training", "Where are we in the search for an Artificial Visual Cortex for Embodied\n Intelligence?", "An Unbiased Look at Datasets for Visuo-Motor Pre-Training", "Language-Driven Representation Learning for Robotics"], "answer_arxiv_id": ["2107.03380", "2111.09888", "2203.03580", "2203.12601", "2210.03109", "2203.06173", "2210.00030", "2303.18240", "2310.09289", "2302.12766"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_10908"} +{"question": "Which studies demonstrated that gradient noise destabilizes the training around sharp minima?", "answer": ["On the diffusion approximation of nonconvex stochastic gradient descent", "On Linear Stability of SGD and Input-Smoothness of Neural Networks"], "answer_arxiv_id": ["1705.07562v2", "2105.13462"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_10909"} +{"question": "What works used RKHS-based methods to compute an approximate solution?", "answer": ["Kernel Methods for the Approximation of Some Key Quantities of Nonlinear Systems", "An exact kernel framework for spatio-temporal dynamics"], "answer_arxiv_id": ["1204.0563v2", "2011.06848"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_10910"} +{"question": "Which studies examined private algorithms for learning univariate and multivariate Gaussians?", "answer": ["Finite Sample Differentially Private Confidence Intervals", "On the Sample Complexity of Privately Learning Unbounded High-Dimensional Gaussians", "A Private and Computationally-Efficient Estimator for Unbounded Gaussians", "FriendlyCore: Practical Differentially Private Aggregation", "Private and polynomial time algorithms for learning Gaussians and beyond", "Private Robust Estimation by Stabilizing Convex Relaxations", "Differential privacy and robust statistics in high dimensions"], "answer_arxiv_id": ["1711.03908v1", "2010.09929v1", "2111.04609v2", "2110.10132", "2111.11320", "2112.03548", "2111.06578"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_10911"} +{"question": "Which study proposed the hybrid interpolation to achieve optimal test loss?", "answer": ["Harmless interpolation of noisy data in regression"], "answer_arxiv_id": ["1903.09139"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_10912"} +{"question": "Which papers describe the use of a placeholder text embedding for visual concept representation in the subject-driven text-to-image generation task?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_10913"} +{"question": "Could you mention some works that proposed models on multi-person motion prediction in 3D?", "answer": ["Multi-Person 3D Motion Prediction with Multi-Range Transformers", "Multi-Person Extreme Motion Prediction", "SoMoFormer: Multi-Person Pose Forecasting with Transformers", "SoMoFormer: Social-Aware Motion Transformer for Multi-Person Motion Prediction", "Stochastic Multi-Person 3D Motion Forecasting"], "answer_arxiv_id": ["2111.12073", "2105.08825", "2208.14023", "2208.09224", "2306.05421"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_10914"} +{"question": "Could you provide me some studies that use deep neural networks in their models for end-to-end motion planning?", "answer": ["NEAT: Neural Attention Fields for End-to-End Autonomous Driving", "End-to-End Urban Driving by Imitating a Reinforcement Learning Coach", "Learning by Cheating", "Learning Interpretable End-to-End Vision-Based Motion Planning for\n Autonomous Driving with Optical Flow Distillation", "Trajectory-guided Control Prediction for End-to-end Autonomous Driving:\n A Simple yet Strong Baseline", "Safety-Enhanced Autonomous Driving Using Interpretable Sensor Fusion\n Transformer"], "answer_arxiv_id": ["2109.04456", "2108.08265", "1912.12294", "2104.12861", "2206.08129", "2207.14024"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_10915"} +{"question": "Which paper describes the DDPM formulation in more detail?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_10916"} +{"question": "Could you provide me studies about Symmetric cross-entropy Learning (SL)?", "answer": ["Symmetric Cross Entropy for Robust Learning with Noisy Labels"], "answer_arxiv_id": ["1908.06112"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_10917"} +{"question": "What works focus on hybrid algorithms that combine constraint-based and score-based methods in Bayesian networks?", "answer": ["A hybrid algorithm for Bayesian network structure learning with application to multi-label learning", "High-dimensional consistency in score-based and hybrid structure learning"], "answer_arxiv_id": ["1506.05692", "1507.02608"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_10918"} +{"question": "Which papers used structural methods to organize neural network units based on their connectivity and edge-weights?", "answer": ["Modular Representation of Layered Neural Networks", "Interpreting Layered Neural Networks via Hierarchical Modular Representation", "Clusterability in Neural Networks"], "answer_arxiv_id": ["1703.00168", "1810.01588", "2103.03386"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_10919"} +{"question": "Could you provide me some studies that utilize nonstationarity in temporal data through contrastive learning for causal representation learning?", "answer": ["Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICA"], "answer_arxiv_id": ["1605.06336"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_10920"} +{"question": "What research works leveraged programming languages in vision tasks?", "answer": ["Language Models of Code are Few-Shot Commonsense Learners", "ViperGPT: Visual Inference via Python Execution for Reasoning", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation\n Models"], "answer_arxiv_id": ["2210.07128", "2303.08128", "2303.04671"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_10921"} +{"question": "Which papers are about generative adversarial networks?", "answer": ["Generative Adversarial Nets", "A New Perspective on Stabilizing GANs training: Direct Adversarial Training"], "answer_arxiv_id": ["1406.2661", "2008.09041"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_10922"} +{"question": "Which papers present foundational behavioral cloning in the context of imitation learning?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret\n Online Learning"], "answer_arxiv_id": ["1011.0686"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_10923"} +{"question": "Could you provide me the references about platforms designed to assess short-term and medium-term forecasting using ML techniques?", "answer": ["ENS-10: A Dataset For Post-Processing Ensemble Weather Forecasts"], "answer_arxiv_id": ["2206.14786"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_10924"} +{"question": "Can you mention some works that gave the non-asymptotic results under certain data assumption for min-ℓ2-norm interpolator?", "answer": ["Benign Overfitting in Linear Regression", "Benign overfitting in ridge regression"], "answer_arxiv_id": ["1906.11300", "2009.14286"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_10925"} +{"question": "Can you provide some references that focused on designing NN architectures tailored to specific problem structures such as, local correlations in features, symmetries in data, convexity, or monotonicity?", "answer": ["Geometric deep learning: going beyond Euclidean data", "Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1611.08097", "1602.07576"], "source_meta": {"published_time": "20220718"}, "qid": "AutoScholarQuery_train_10926"} +{"question": "What methods combine NeRF with time-conditioned latent codes for compactly representing dynamic scenes?", "answer": ["Neural 3D Video Synthesis from Multi-view Video"], "answer_arxiv_id": ["2103.02597"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_10927"} +{"question": "What work introduced TracIn, a gradient-based method to compute training data influence, and what are the limitations of this method?", "answer": ["Estimating Training Data Influence by Tracing Gradient Descent"], "answer_arxiv_id": ["2002.08484"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_10928"} +{"question": "Which work combined rematerialization with paging for memory optimization?", "answer": ["POET: Training Neural Networks on Tiny Devices with Integrated Rematerialization and Paging"], "answer_arxiv_id": ["2207.07697"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_10929"} +{"question": "Could you provide me some studies about using LLMs to detect and counter hate speech?", "answer": ["RAUCG: Retrieval-Augmented Unsupervised Counter Narrative Generation for\n Hate Speech", "Moderating New Waves of Online Hate with Chain-of-Thought Reasoning in Large Language Models", "Generative AI for Hate Speech Detection: Evaluation and Findings", "Detecting and Correcting Hate Speech in Multimodal Memes with Large\n Visual Language Model", "HateRephrase: Zero- and Few-Shot Reduction of Hate Intensity in Online\n Posts using Large Language Models", "Probing LLMs for hate speech detection: strengths and vulnerabilities", "From Dogwhistles to Bullhorns: Unveiling Coded Rhetoric with Language\n Models"], "answer_arxiv_id": ["2310.05650", "2312.15099v2", "2311.09993", "2311.06737", "2310.13985", "2310.12860", "2305.17174"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_10930"} +{"question": "Which study proposed a dynamic graph-based reasoning benchmark?", "answer": ["DyVal: Dynamic Evaluation of Large Language Models for Reasoning Tasks"], "answer_arxiv_id": ["2309.17167"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_10931"} +{"question": "Can you mention some papers that applied diffusion models for text-to-image (T2I) generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_10932"} +{"question": "Could you provide me with some references for an in-depth overview and a how-to for graph transformers?", "answer": ["Recipe for a General, Powerful, Scalable Graph Transformer"], "answer_arxiv_id": ["2205.12454"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10933"} +{"question": "Are there any papers that addressed the drawbacks of population-based RL?", "answer": ["Fast Population-Based Reinforcement Learning on a Single Machine"], "answer_arxiv_id": ["2206.08888"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_10934"} +{"question": "Which studies employed an adversarial training framework for offline RL?", "answer": ["Bellman-consistent Pessimism for Offline Reinforcement Learning", "Adversarially Trained Actor Critic for Offline Reinforcement Learning", "Adversarial Counterfactual Environment Model Learning", "Pessimistic Model-based Offline Reinforcement Learning under Partial Coverage", "RAMBO-RL: Robust Adversarial Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2106.06926", "2202.02446", "2206.04890", "2107.06226", "2204.12581"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_10935"} +{"question": "Which work first presents the concept of LLM alignment?", "answer": ["Artificial Intelligence, Values and Alignment"], "answer_arxiv_id": ["2001.09768v2"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_10936"} +{"question": "What is the first summarization dataset on patents called and who created it?", "answer": ["BigPatent: A Large-Scale Dataset for Abstractive and Coherent Summarization"], "answer_arxiv_id": ["1906.03741"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_10937"} +{"question": "What works have come up with training dynamics-based solutions for mislabel detection?", "answer": ["Unsupervised Label Noise Modeling and Loss Correction", "Identifying Mislabeled Data using the Area Under the Margin Ranking"], "answer_arxiv_id": ["1904.11238", "2001.10528"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10938"} +{"question": "What studies used Chamfer Distance in 3D mesh reconstruction tasks?", "answer": ["Deep Hybrid Self-Prior for Full 3D Mesh Generation", "CD$^2$: Fine-grained 3D Mesh Reconstruction With Twice Chamfer Distance", "MeshMVS: Multi-View Stereo Guided Mesh Reconstruction", "Learning Delaunay Surface Elements for Mesh Reconstruction"], "answer_arxiv_id": ["2108.08017", "2206.00447", "2010.08682", "2012.01203"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_10939"} +{"question": "Could you provide me some studies about SLAM methods that usually run online?", "answer": ["ORB-SLAM: a Versatile and Accurate Monocular SLAM System", "ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAM"], "answer_arxiv_id": ["1502.00956", "2007.11898"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_10940"} +{"question": "Which papers provide explanations of various types of deep learning methods?", "answer": ["Deep Learning for Instance Retrieval: A Survey", "Recent Advance in Content-based Image Retrieval: A Literature Survey", "A Decade Survey of Content Based Image Retrieval using Deep Learning"], "answer_arxiv_id": ["2101.11282", "1706.06064", "2012.00641"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_10941"} +{"question": "Could you provide some papers on truncated estimation of the 'Neumann series' associated with implicit differentiation?", "answer": ["Optimizing Millions of Hyperparameters by Implicit Differentiation", "Scalable Gradient-Based Tuning of Continuous Regularization Hyperparameters"], "answer_arxiv_id": ["1911.02590", "1511.06727"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_10942"} +{"question": "What research can you provide that discusses the differences in filtering heuristics used when creating the datasets?", "answer": ["Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with Text"], "answer_arxiv_id": ["2304.06939"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_10943"} +{"question": "Which research works propose Dirichlet-based uncertainty models (DBU)?", "answer": ["Predictive Uncertainty Estimation via Prior Networks", "Reverse KL-Divergence Training of Prior Networks: Improved Uncertainty and Adversarial Robustness", "Posterior Network: Uncertainty Estimation without OOD Samples via Density-Based Pseudo-Counts", "Evidential Deep Learning to Quantify Classification Uncertainty"], "answer_arxiv_id": ["1802.10501", "1905.13472", "2006.09239", "1806.01768"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_10944"} +{"question": "Which studies proposed different definitions of fairness like group fairness, individual fairness, and counterfactual fairness?", "answer": ["Equality of Opportunity in Supervised Learning", "Quadratic Metric Elicitation for Fairness and Beyond", "Training individually fair ML models with Sensitive Subspace Robustness", "Two Simple Ways to Learn Individual Fairness Metrics from Data", "SenSeI: Sensitive Set Invariance for Enforcing Individual Fairness", "Domain Adaptation meets Individual Fairness. And they get along.", "Individual Fairness in Feature-Based Pricing for Monopoly Markets", "Counterfactual Fairness", "Towards a Unified Framework for Fair and Stable Graph Representation Learning", "Counterfactual Fairness with Partially Known Causal Graph"], "answer_arxiv_id": ["1610.02413", "2011.01516", "1907.00020", "2006.11439", "2006.14168", "2205.00504", "2202.12844", "1703.06856", "2102.13186", "2205.13972"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_10945"} +{"question": "Could you give some references for novel relighting approaches using deep learning?", "answer": ["Deep Photo Style Transfer", "Neural Style Transfer: A Review", "Image-to-Image Translation with Conditional Adversarial Networks", "Single Image Portrait Relighting"], "answer_arxiv_id": ["1703.07511", "1705.04058", "1611.07004", "1905.00824"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10946"} +{"question": "What are some works that combine transformers and convolution to construct efficient Vision Transformers?", "answer": ["MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer", "TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation", "Lite Vision Transformer with Enhanced Self-Attention", "Mobile-Former: Bridging MobileNet and Transformer", "EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers", "Separable Self-attention for Mobile Vision Transformers", "EfficientFormer: Vision Transformers at MobileNet Speed"], "answer_arxiv_id": ["2110.02178", "2204.05525", "2112.10809", "2108.05895", "2205.03436", "2206.02680", "2206.01191"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10947"} +{"question": "Which work used Singular Value Decomposition (SVD) for positional encodings?", "answer": ["Global Self-Attention as a Replacement for Graph Convolution"], "answer_arxiv_id": ["2108.03348"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_10948"} +{"question": "What works provide examples of LLMs-based advancements in image captioning?", "answer": ["ClipCap: CLIP Prefix for Image Captioning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "PaLI-X: On Scaling up a Multilingual Vision and Language Model"], "answer_arxiv_id": ["2111.09734", "2301.12597", "2209.06794", "2305.18565"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_10949"} +{"question": "Which studies discuss Gaussian-sigmoid or Gaussian-softmax integrals, which are a central component of the methodology in question?", "answer": ["Semi-analytical approximations to statistical moments of sigmoid and softmax mappings of normal variables"], "answer_arxiv_id": ["1703.00091v2"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_10950"} +{"question": "What research work designs a network to refine depth estimations and then explicitly warps pixels from source views to novel views?", "answer": ["FWD: Real-time Novel View Synthesis with Forward Warping and Depth"], "answer_arxiv_id": ["2206.08355"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_10951"} +{"question": "What studies proposed methods for image reflection removal that require additional inputs such as multi-frames, polarization or flash-only prior?", "answer": ["Improved Multiple-Image-Based Reflection Removal Algorithm Using Deep Neural Networks", "User-assisted Video Reflection Removal", "Polarized Reflection Removal with Perfect Alignment in the Wild", "Robust Reflection Removal with Reflection-free Flash-only Cues", "Robust Reflection Removal with Flash-only Cues in the Wild"], "answer_arxiv_id": ["2208.04679v2", "2009.03281", "2003.12789", "2103.04273", "2211.02914"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10952"} +{"question": "What research utilized an optimization method to reconstruct the guidance image and employed P2P for real image editing?", "answer": ["Null-text Inversion for Editing Real Images using Guided Diffusion\n Models"], "answer_arxiv_id": ["2211.09794"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_10953"} +{"question": "What research works mention determining the design of a multimodal network based on task objective, data availability and computation budget?", "answer": ["Benchmarking Multimodal AutoML for Tabular Data with Text Fields", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "Learning Factorized Multimodal Representations", "Diverse Image Captioning with Context-Object Split Latent Spaces"], "answer_arxiv_id": ["2111.02705", "2201.12086", "1806.06176", "2011.00966"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_10954"} +{"question": "Can you provide some research papers that discuss the state-of-the-art deep neural network architectures?", "answer": ["Going deeper with convolutions", "Deep Residual Learning for Image Recognition", "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks"], "answer_arxiv_id": ["1409.4842", "1512.03385", "1905.11946"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_10955"} +{"question": "Which research derived a T1/2 worst-case regret depending on the ambient dimensions in the convex cost setting of LQG control?", "answer": ["Adaptive Control and Regret Minimization in Linear Quadratic Gaussian (LQG) Setting"], "answer_arxiv_id": ["2003.05999"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_10956"} +{"question": "What studies have been done on RL with exogenous information?", "answer": ["Discovering and Removing Exogenous State Variables and Rewards for Reinforcement Learning", "Sparsity in Partially Controllable Linear Systems"], "answer_arxiv_id": ["1806.01584", "2110.06150"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_10957"} +{"question": "Could you provide me with studies on integrating voxel and point features to the Region Proposal Network (RPN)?", "answer": ["PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection"], "answer_arxiv_id": ["1912.13192"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_10958"} +{"question": "Can you name the works which were proposed to design selection strategies for detection and reduction of impact from noisy samples during training?", "answer": ["DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "Self-Filtering: A Noise-Aware Sample Selection for Label Noise with\n Confidence Penalization", "Adaptive Sample Selection for Robust Learning under Label Noise"], "answer_arxiv_id": ["2002.07394", "2208.11351", "2106.15292"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_10959"} +{"question": "Which papers have utilized a momentum update mechanism to maintain a long queue of negative samples for contrastive learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1911.05722"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_10960"} +{"question": "What are some pre-training methods using graph data, as compared to 3D point clouds of atom nuclei?", "answer": ["Strategies for Pre-training Graph Neural Networks", "Self-Supervised Learning of Graph Neural Networks: A Unified Review", "Variational Graph Auto-Encoders"], "answer_arxiv_id": ["1905.12265", "2102.10757", "1611.07308"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_10961"} +{"question": "Which works focus on reconstruction-based methods for self-supervised learning?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "Unsupervised Visual Representation Learning by Context Prediction", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2106.08254", "1505.05192", "2111.06377"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_10962"} +{"question": "What are some works that used Transformer-based network for various image restoration tasks?", "answer": ["Pre-Trained Image Processing Transformer"], "answer_arxiv_id": ["2012.00364"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_10963"} +{"question": "Could you provide me with references on the recent efforts to develop foundation language models?", "answer": ["A Survey of Large Language Models"], "answer_arxiv_id": ["2303.18223"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_10964"} +{"question": "Are there research studies that focused on extending IPS for tackling bias in implicit recommendations?", "answer": ["Unbiased Recommender Learning from Missing-Not-At-Random Implicit Feedback"], "answer_arxiv_id": ["1909.03601"], "source_meta": {"published_time": "20220319"}, "qid": "AutoScholarQuery_train_10965"} +{"question": "Can you provide examples of studies that focus on negative sampling strategies to improve the contrastive loss performance in document retrieval?", "answer": ["Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval", "Complement Lexical Retrieval Model with Semantic Residual Embeddings"], "answer_arxiv_id": ["2007.00808", "2004.13969"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_10966"} +{"question": "Any works about deep learning methods for tasks related to tabular data such as data integration and imputation?", "answer": ["Can Foundation Models Wrangle Your Data?"], "answer_arxiv_id": ["2205.09911"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_10967"} +{"question": "Could you provide me examples of works indicating the remarkable effectiveness of Behavior Cloning in robotic manipulation?", "answer": ["What Matters in Learning from Offline Human Demonstrations for Robot\n Manipulation", "One-Shot Visual Imitation Learning via Meta-Learning", "Deep Imitation Learning for Complex Manipulation Tasks from Virtual\n Reality Teleoperation"], "answer_arxiv_id": ["2108.03298", "1709.04905", "1710.04615"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_10968"} +{"question": "What follows up approaches are proposed to improve classification results based on the success of CLIP-like vision-language pretraining?", "answer": ["Learning to Prompt for Vision-Language Models", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2109.01134", "2110.04544", "2203.05557"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_10969"} +{"question": "Which study is about combining a cascade of a diffusion and fixed-size super-resolution models to generate images at higher resolution?", "answer": ["Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2106.15282"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_10970"} +{"question": "What papers study functions which have higher leap complexity?", "answer": ["SGD learning on neural networks: leap complexity and saddle-to-saddle dynamics"], "answer_arxiv_id": ["2302.11055"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_10971"} +{"question": "Who proposed a novel poisoning attack on unlearning systems?", "answer": ["Hard to Forget: Poisoning Attacks on Certified Machine Unlearning"], "answer_arxiv_id": ["2109.08266"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_10972"} +{"question": "Which graphical neural network architectures incorporate shortest path distances as structural features?", "answer": ["Do Transformers Really Perform Bad for Graph Representation?", "Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning", "Position-aware Graph Neural Networks", "Shortest Path Networks for Graph Property Prediction", "Geodesic Graph Neural Network for Efficient Graph Representation Learning"], "answer_arxiv_id": ["2106.05234", "2009.00142", "1906.04817", "2206.01003", "2210.02636"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_10973"} +{"question": "What papers discuss models like Neural CDE and Neural RDE?", "answer": ["Neural Controlled Differential Equations for Irregular Time Series", "Neural Rough Differential Equations for Long Time Series"], "answer_arxiv_id": ["2005.08926", "2009.08295"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_10974"} +{"question": "Which work holds similar findings to ours but applies them in a different setting?", "answer": ["Gradient Flow in Sparse Neural Networks and How Lottery Tickets Win"], "answer_arxiv_id": ["2010.03533"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_10975"} +{"question": "Could you tell me any studies that applied DMs in learning the prior of latent space with a jointly optimized decoder?", "answer": ["Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["2106.05931"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_10976"} +{"question": "Which study generalized the 1-WL test and its various characterizations to graphons?", "answer": ["Fractional Isomorphism of Graphons"], "answer_arxiv_id": ["1909.04122"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_10977"} +{"question": "Could you provide me some works that point out glosses as an incomplete representation of sign language?", "answer": ["Sign Language Transformers: Joint End-to-end Sign Language Recognition\n and Translation", "Considerations for meaningful sign language machine translation based on\n glosses"], "answer_arxiv_id": ["2003.13830", "2211.15464"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_10978"} +{"question": "Any papers about Semantic Scene Completion using point clouds?", "answer": ["Semantic Segmentation-assisted Scene Completion for LiDAR Point Clouds", "SCPNet: Semantic Scene Completion on Point Cloud", "Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning\n Contextual Shape Priors from Scene Completion"], "answer_arxiv_id": ["2109.11453", "2303.06884", "2012.03762"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_10979"} +{"question": "What works have applied mixup regularization techniques in computer vision?", "answer": ["mixup: Beyond Empirical Risk Minimization", "CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features", "Improved Regularization of Convolutional Neural Networks with Cutout"], "answer_arxiv_id": ["1710.09412", "1905.04899", "1708.04552"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_10980"} +{"question": "Which research introduced the curiosity-driven method in the study of exploration-exploitation trade-off in deep RL?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction"], "answer_arxiv_id": ["1705.05363"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_10981"} +{"question": "Which model-based methods have recently matched or surpassed the performance of model-free methods in the Atari game playing benchmark?", "answer": ["Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1911.08265", "2010.02193"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_10982"} +{"question": "What is the HuMoR model cited for in the context of pose estimation?", "answer": ["HuMoR: 3D Human Motion Model for Robust Pose Estimation"], "answer_arxiv_id": ["2105.04668"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_10983"} +{"question": "In which studies did researchers substitute words in a sentence with words from the same syntactic category and then average the attention maps of a BERT model?", "answer": ["Syntactic Substitutability as Unsupervised Dependency Syntax"], "answer_arxiv_id": ["2211.16031"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_10984"} +{"question": "Are there any works about learning policy for specific tasks like folding in garment manipulation?", "answer": ["SpeedFolding: Learning Efficient Bimanual Folding of Garments", "Cloth Funnels: Canonicalized-Alignment for Multi-Purpose Garment\n Manipulation", "UniFolding: Towards Sample-efficient, Scalable, and Generalizable\n Robotic Garment Folding"], "answer_arxiv_id": ["2208.10552", "2210.09347", "2311.01267"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_10985"} +{"question": "Which work trained an image-to-image diffusion model for image editing?", "answer": ["Palette: Image-to-Image Diffusion Models"], "answer_arxiv_id": ["2111.05826"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_10986"} +{"question": "What works represent Scene-Mixture Models explaining a visual scene by a finite mixture of component images?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational\n Inference", "GENESIS: Generative Scene Inference and Sampling with Object-Centric\n Latent Representations"], "answer_arxiv_id": ["1901.11390", "1903.00450", "1907.13052"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_10987"} +{"question": "Which papers discuss extensive studies of pruning in the context of compressing post-training models?", "answer": ["Learning both Weights and Connections for Efficient Neural Networks", "Exploring Sparsity in Recurrent Neural Networks", "Pruning Filters for Efficient ConvNets", "Importance Estimation for Neural Network Pruning", "The State of Sparsity in Deep Neural Networks"], "answer_arxiv_id": ["1506.02626", "1704.05119", "1608.08710", "1906.10771", "1902.09574"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_10988"} +{"question": "Which research provides examples of fine-grained measures of performance within NLP and machine learning?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding", "Bootleg: Chasing the Tail with Self-Supervised Named Entity Disambiguation", "Holistic Evaluation of Language Models", "Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models"], "answer_arxiv_id": ["1804.07461", "2010.10363", "2211.09110", "2206.04615"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_10989"} +{"question": "Could you name the research that employs techniques such as word lists, blocklists of URLs, and Google Safe Search for removing toxic and harmful content during data cleaning process?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora\n with Web Data, and Web Data Only"], "answer_arxiv_id": ["1910.10683", "2306.01116"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_10990"} +{"question": "Which works focus on learning prior knowledge from available scanned data of people interacting with 3D indoor scenes in the field of 3D Human-Scene Interaction Synthesis?", "answer": ["Generating 3D People in Scenes without People", "Populating 3D Scenes by Learning Human-Scene Interaction", "Compositional Human-Scene Interaction Synthesis with Semantic Control"], "answer_arxiv_id": ["1912.02923", "2012.11581", "2207.12824"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_10991"} +{"question": "What paper mentioned the concept of consistency regularization?", "answer": ["Temporal Ensembling for Semi-Supervised Learning"], "answer_arxiv_id": ["1610.02242v3"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_10992"} +{"question": "How has temporal generalization been explored in the context of document classification?", "answer": ["Temporal Adaptation of BERT and Performance on Downstream Document\n Classification: Insights from Social Media", "Improved Multi-label Classification under Temporal Concept Drift:\n Rethinking Group-Robust Algorithms in a Label-Wise Setting"], "answer_arxiv_id": ["2104.08116", "2203.07856"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_10993"} +{"question": "Which paper first proposed to condition the transformation in Normalizing Flows on some parameter?", "answer": ["Semi-Conditional Normalizing Flows for Semi-Supervised Learning"], "answer_arxiv_id": ["1905.00505"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_10994"} +{"question": "What work has been done using CNN based methods in perception tasks?", "answer": ["Semantic Image Segmentation with Deep Convolutional Nets and Fully\n Connected CRFs", "Mask R-CNN", "Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up\n Panoptic Segmentation", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks", "Deeper Depth Prediction with Fully Convolutional Residual Networks", "DeepPose: Human Pose Estimation via Deep Neural Networks"], "answer_arxiv_id": ["1412.7062", "1703.06870", "1911.10194", "1506.01497", "1606.00373", "1312.4659"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_10995"} +{"question": "Which works built textless language models built upon HuBERT units for speech continuation without using text?", "answer": ["On Generative Spoken Language Modeling from Raw Audio", "Text-Free Prosody-Aware Generative Spoken Language Modeling", "Generative Spoken Dialogue Language Modeling"], "answer_arxiv_id": ["2102.01192", "2109.03264", "2203.16502"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_10996"} +{"question": "Are there any studies that introduced GNN for CTDGs?", "answer": ["Inductive representation learning on temporal graphs", "Temporal Graph Networks for Deep Learning on Dynamic Graphs"], "answer_arxiv_id": ["2002.07962", "2006.10637"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_10997"} +{"question": "What studies applied the mean field analysis to the backward dynamics and introduced the concept of depth scales to quantify how deep signals can propagate?", "answer": ["Deep Information Propagation"], "answer_arxiv_id": ["1611.01232"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_10998"} +{"question": "Which work provides a theoretical justification for ensembling diffusion models to achieve the best MSE at different SNR levels?", "answer": ["eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers"], "answer_arxiv_id": ["2211.01324"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_10999"} +{"question": "Which research proposes StyleGAN as a self-supervised learning method for face recognition applications?", "answer": ["How to Boost Face Recognition with StyleGAN?"], "answer_arxiv_id": ["2210.10090"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_11000"} +{"question": "Which work treated self-attention operation as support vector regression without considering the asymmetry in the deployed kernel methods?", "answer": ["A Primal-Dual Framework for Transformers and Neural Networks"], "answer_arxiv_id": ["2406.13781"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11001"} +{"question": "Any works that perform controlled text generation through iterative sampling and maximizing scores computed from various domain-specific modules?", "answer": ["Mix and Match: Learning-free Controllable Text Generation using Energy\n Language Models", "COLD Decoding: Energy-based Constrained Text Generation with Langevin\n Dynamics"], "answer_arxiv_id": ["2203.13299", "2202.11705"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_11002"} +{"question": "Which papers introduced methods that focus on encoding both short-term and medium-term temporal information in video segmentation?", "answer": ["Spatial Feature Calibration and Temporal Fusion for Effective One-stage\n Video Instance Segmentation", "End-to-End Video Instance Segmentation with Transformers", "SeqFormer: Sequential Transformer for Video Instance Segmentation", "In Defense of Online Models for Video Instance Segmentation", "VITA: Video Instance Segmentation via Object Token Association", "MinVIS: A Minimal Video Instance Segmentation Framework without\n Video-based Training", "MDQE: Mining Discriminative Query Embeddings to Segment Occluded\n Instances on Challenging Videos", "A Generalized Framework for Video Instance Segmentation", "TubeFormer-DeepLab: Video Mask Transformer", "Video K-Net: A Simple, Strong, and Unified Baseline for Video\n Segmentation"], "answer_arxiv_id": ["2104.05606", "2011.14503", "2112.08275", "2207.10661", "2206.04403", "2208.02245", "2303.14395", "2211.08834", "2205.15361", "2204.04656"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_11003"} +{"question": "Who have combined topic diversity with coherence scores resulting in topic quality scores?", "answer": ["Topic Modeling in Embedding Spaces"], "answer_arxiv_id": ["1907.04907"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_11004"} +{"question": "What research has proposed Variational Score Distillation (VSD) as a solution to over-saturation, over-smoothing, and low-diversity problems in SDS?", "answer": ["ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation"], "answer_arxiv_id": ["2305.16213"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_11005"} +{"question": "What studies contributed to the progress in AI, vision, NLP, and multimodal models due to large-scale datasets and benchmarks?", "answer": ["ImageNet Large Scale Visual Recognition Challenge", "GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding", "SQuAD: 100,000+ Questions for Machine Comprehension of Text", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "The Pile: An 800GB Dataset of Diverse Text for Language Modeling", "LAION-5B: An open large-scale dataset for training next generation image-text models", "DataComp: In search of the next generation of multimodal datasets"], "answer_arxiv_id": ["1409.0575", "1804.07461", "1606.05250", "1910.10683", "2101.00027", "2210.08402", "2304.14108"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_11006"} +{"question": "What work used subjective human judgments of attributions in evaluating saliency/attribution tools?", "answer": ["Towards Benchmarking Explainable Artificial Intelligence Methods"], "answer_arxiv_id": ["2208.12120"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_11007"} +{"question": "Could you provide examples of works on the modeling of object relations in embodied AI research?", "answer": ["Towards Optimal Correlational Object Search", "Semantic and Geometric Modeling with Neural Message Passing in 3D Scene Graphs for Hierarchical Mechanical Search"], "answer_arxiv_id": ["2110.09991", "2012.04060"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_11008"} +{"question": "What research papers approximated intractable posterior over latent variables by performing amortized variational inference?", "answer": ["Auto-Encoding Variational Bayes", "Stochastic Backpropagation and Approximate Inference in Deep Generative Models"], "answer_arxiv_id": ["1312.6114", "1401.4082v3"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_11009"} +{"question": "Which works presented the concept of learning text embeddings and optimizing diffusion backbone for T2I personalization?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_11010"} +{"question": "Can you mention some works that contributed to monocular 3D detection in driving scenarios?", "answer": ["M3D-RPN: Monocular 3D Region Proposal Network for Object Detection", "Disentangling Monocular 3D Object Detection", "Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving", "FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection", "SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint Estimation", "RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving", "Probabilistic and Geometric Depth: Detecting Objects in Perspective", "Categorical Depth Distribution Network for Monocular 3D Object Detection"], "answer_arxiv_id": ["1907.06038", "1905.12365", "1812.07179", "2104.10956", "2002.10111", "2001.03343", "2107.14160", "2103.01100"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_11011"} +{"question": "Which work improved on the idea from bib.bib4 and achieved a nearly minimax optimal horizon-free regret?", "answer": ["Is Reinforcement Learning More Difficult Than Bandits? A Near-optimal Algorithm Escaping the Curse of Horizon"], "answer_arxiv_id": ["2009.13503"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_11012"} +{"question": "Which studies have investigated large language models(like GPT, LLaMA, Gemini) and their results to a wide range of queries?", "answer": ["Language Models are Few-Shot Learners", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "Gemini: A Family of Highly Capable Multimodal Models"], "answer_arxiv_id": ["2005.14165", "2302.13971", "2307.09288", "2312.11805v4"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_11013"} +{"question": "Could you provide me some papers that study the MFGs under monotone conditions?", "answer": ["Fictitious Play for Mean Field Games: Continuous Time Analysis and Applications", "Concave Utility Reinforcement Learning: the Mean-Field Game Viewpoint", "Scaling up Mean Field Games with Online Mirror Descent"], "answer_arxiv_id": ["2007.03458", "2106.03787", "2103.00623"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_11014"} +{"question": "Which papers have prominently showcased self-supervised learning methods such as Masked Language Modeling in Natural Language Processing and contrastive pre-training in computer vision?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "A Simple Framework for Contrastive Learning of Visual Representations", "ALBERT: A Lite BERT for Self-supervised Learning of Language Representations", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Self-Supervised Learning of Pretext-Invariant Representations"], "answer_arxiv_id": ["1810.04805", "2002.05709", "1909.11942", "1907.11692", "1910.13461", "2002.05709", "1911.05722", "2006.09882", "2105.04906", "1912.01991"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_11015"} +{"question": "Any works about designing an encoder-decoder structure in DLKT to represent the exercise and response embedding sequences?", "answer": ["Towards an Appropriate Query, Key, and Value Computation for Knowledge Tracing"], "answer_arxiv_id": ["2002.07033"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_11016"} +{"question": "Which studies proposed models that demonstrated a remarkable zero-shot ability in creating images from text captions?", "answer": ["Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_11017"} +{"question": "What works have introduced regularization terms in local objective to reduce variance of local training in federated learning?", "answer": ["SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Federated Optimization in Heterogeneous Networks", "Model-Contrastive Federated Learning", "Federated Learning Based on Dynamic Regularization"], "answer_arxiv_id": ["1910.06378", "1812.06127", "2103.16257", "2111.04263"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_11018"} +{"question": "What references can you provide that have employed vision models to connect with Language Learning Models (LLMs)?", "answer": ["Visual Instruction Tuning", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2304.08485", "2201.12086", "2304.10592"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_11019"} +{"question": "Where can I find works that generated photo-realistic images from semantic label maps in the area of Semantic Image Synthesis?", "answer": ["Photographic Image Synthesis with Cascaded Refinement Networks", "Semantic Image Synthesis with Spatially-Adaptive Normalization", "Learning to Predict Layout-to-image Conditional Convolutions for Semantic Image Synthesis", "Shapes and Context: In-the-Wild Image Synthesis & Manipulation", "SEAN: Image Synthesis with Semantic Region-Adaptive Normalization", "SESAME: Semantic Editing of Scenes by Adding, Manipulating or Erasing Objects", "Semantically Multi-modal Image Synthesis", "You Only Need Adversarial Supervision for Semantic Image Synthesis", "Efficient Semantic Image Synthesis via Class-Adaptive Normalization", "Diverse Semantic Image Synthesis via Probability Distribution Modeling", "Semantically Multi-modal Image Synthesis"], "answer_arxiv_id": ["1707.09405", "1903.07291", "1910.06809", "1906.04728", "1911.12861", "2004.04977", "2003.12697", "2012.04781", "2012.04644", "2103.06878", "2003.12697"], "source_meta": {"published_time": "20200331"}, "qid": "AutoScholarQuery_train_11020"} +{"question": "What studies have developed methods for efficient training from pretrained models?", "answer": ["Net2Net: Accelerating Learning via Knowledge Transfer", "Splitting Steepest Descent for Growing Neural Architectures", "Steepest Descent Neural Architecture Optimization: Escaping Local Optimum with Signed Neural Splitting", "Energy-Aware Neural Architecture Optimization with Fast Splitting Steepest Descent", "Firefly Neural Architecture Descent: a General Approach for Growing Neural Networks", "bert2BERT: Towards Reusable Pretrained Language Models", "Learning to Grow Pretrained Models for Efficient Transformer Training"], "answer_arxiv_id": ["1511.05641", "1910.02366", "2003.10392", "1910.03103", "2102.08574", "2110.07143", "2303.00980"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_11021"} +{"question": "What researches applied DEQs in fields like optical flow, image segmentation and inverse imaging problems?", "answer": ["Deep Equilibrium Optical Flow Estimation", "Multiscale Deep Equilibrium Models", "Deep Equilibrium Architectures for Inverse Problems in Imaging"], "answer_arxiv_id": ["2204.08442", "2006.08656", "2102.07944"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_11022"} +{"question": "Which works have contributed to parameter efficient continual learning through prompt tuning in a continual learning scenario?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "LFPT5: A Unified Framework for Lifelong Few-shot Language Learning Based on Prompt Tuning of T5", "Learning to Prompt for Continual Learning", "Continual Prompt Tuning for Dialog State Tracking", "A Unified Continual Learning Framework with General Parameter-Efficient Tuning"], "answer_arxiv_id": ["2104.08691", "2110.07298", "2112.08654", "2203.06654", "2303.10070"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_11023"} +{"question": "Which papers describe 'clean-label' attacks using gradient-based optimization of poisoned data?", "answer": ["Concealed Data Poisoning Attacks on NLP Models"], "answer_arxiv_id": ["2010.12563"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_11024"} +{"question": "Which studies are focused on exploring strategies for optimizing prompts to models through the specific textual features of the prompt?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Maieutic Prompting: Logically Consistent Reasoning with Recursive Explanations"], "answer_arxiv_id": ["2201.11903", "2205.11822"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_11025"} +{"question": "Who designed a pure-transformer architecture for initial image feature extraction as well as feature processing in human pose estimators?", "answer": ["ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation"], "answer_arxiv_id": ["2204.12484"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_11026"} +{"question": "Which research works discussed image captioning to build large multimodal models from separately pretrained vision and language models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "GIT: A Generative Image-to-text Transformer for Vision and Language", "Language Is Not All You Need: Aligning Perception with Language Models", "PaLM-E: An Embodied Multimodal Language Model", "PaLI-X: On Scaling up a Multilingual Vision and Language Model"], "answer_arxiv_id": ["2204.14198", "2209.06794v4", "2205.14100", "2302.14045", "2303.03378", "2305.18565"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_11027"} +{"question": "Which studies suggested filtering out the images with moving objects from training scenes?", "answer": ["Semantically-Guided Representation Learning for Self-Supervised\n Monocular Depth"], "answer_arxiv_id": ["2002.12319"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_11028"} +{"question": "Could you list the papers about building convolutional neural networks for sequential data?", "answer": ["WaveNet: A Generative Model for Raw Audio", "Natural Language Processing (almost) from Scratch", "Neural Machine Translation in Linear Time"], "answer_arxiv_id": ["1609.03499", "1103.0398", "1610.10099"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_11029"} +{"question": "What works have proved rates of convergence assuming the target distribution is a smooth transformation of the uniform distribution on a low-dimensional manifold?", "answer": ["Statistical guarantees for generative models without domination"], "answer_arxiv_id": ["2010.09237"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_11030"} +{"question": "Which work trains LLMs to generate scores for a given response by the chain-of-thought reasoning to construct preference datasets by self-generated responses?", "answer": ["Self-Rewarding Language Models"], "answer_arxiv_id": ["2401.10020"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_11031"} +{"question": "What works operate on the character level to generate adversarial perturbations?", "answer": ["Black-box Generation of Adversarial Text Sequences to Evade Deep\n Learning Classifiers"], "answer_arxiv_id": ["1801.04354"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_11032"} +{"question": "Could you mention any works that use exponential integrators to handle diffusion ODEs?", "answer": ["Fast Sampling of Diffusion Models with Exponential Integrator", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models", "UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2204.13902", "2206.00927", "2211.01095", "2302.04867"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_11033"} +{"question": "Can you cite a paper that describes the CLIP embeddings utilized in the CLIPScore metric?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_11034"} +{"question": "Could you provide me some works that explore additional information based on the 3D structure or reconstruct different modalities such as flow or depth for object recognition?", "answer": ["ROOTS: Object-Centric Representation and Rendering of 3D Scenes", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields", "Unsupervised object-centric video generation and decomposition in 3D", "Conditional Object-Centric Learning from Video", "SAVi++: Towards End-to-End Object-Centric Learning from Real-World Videos"], "answer_arxiv_id": ["2006.06130", "2011.12100", "2007.06705", "2111.12594", "2206.07764"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_11035"} +{"question": "What work proposed a transfer distance based on interpolation of tasks for guiding learning?", "answer": ["An Information-Geometric Distance on the Space of Tasks"], "answer_arxiv_id": ["2011.00613"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_11036"} +{"question": "Could you provide me some studies about convex formulations for convolutions and deeper models?", "answer": ["Vector-output ReLU Neural Network Problems are Copositive Programs: Convex Analysis of Two Layer Networks and Polynomial-time Algorithms", "Convex Regularization behind Neural Reconstruction"], "answer_arxiv_id": ["2012.13329", "2012.05169"], "source_meta": {"published_time": "20221114"}, "qid": "AutoScholarQuery_train_11037"} +{"question": "What papers proposed learning two individual classifiers for CZSL?", "answer": ["Symmetry and Group in Attribute-Object Compositions", "Open World Compositional Zero-Shot Learning", "KG-SP: Knowledge Guided Simple Primitives for Open World Compositional\n Zero-Shot Learning", "Siamese Contrastive Embedding Network for Compositional Zero-Shot\n Learning", "Learning Invariant Visual Representations for Compositional Zero-Shot\n Learning"], "answer_arxiv_id": ["2004.00587", "2101.12609", "2205.06784", "2206.14475", "2206.00415"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_11038"} +{"question": "What works discuss the differences between clean models and backdoored models when detecting poison in a model?", "answer": ["One-Pixel Signature: Characterizing CNN Models for Backdoor Detection", "Practical Detection of Trojan Neural Networks: Data-Limited and Data-Free Cases"], "answer_arxiv_id": ["2008.07711", "2007.15802"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_11039"} +{"question": "Could you provide me some studies about applying 3DMMs in face reconstruction?", "answer": ["Learning an Animatable Detailed 3D Face Model from In-The-Wild Images", "3D face reconstruction with dense landmarks"], "answer_arxiv_id": ["2012.04012", "2204.02776"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_11040"} +{"question": "What research has been done on the theoretical aspects of Contextual MDPs in Reinforcement Learning?", "answer": ["Online learning in MDPs with side information", "Contextual Markov Decision Processes", "Policy Certificates: Towards Accountable Reinforcement Learning", "Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "Markov Decision Processes with Continuous Side Information", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches", "Policy Certificates: Towards Accountable Reinforcement Learning", "Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles"], "answer_arxiv_id": ["1406.6812", "1502.02259", "1811.03056", "1610.09512", "1711.05726", "1811.08540", "1811.03056", "1910.10597"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_11041"} +{"question": "Which study showed that optimistic update is not necessary for achieving linear last-iterate convergence in the presence of regularization?", "answer": ["A Unified Approach to Reinforcement Learning, Quantal Response Equilibria, and Two-Player Zero-Sum Games"], "answer_arxiv_id": ["2206.05825"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_11042"} +{"question": "Could you provide me some studies where the idea of artificial negative sampling has been addressed recently?", "answer": ["Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samples", "Interpretable, Multidimensional, Multimodal Anomaly Detection with Negative Sampling for Detection of Device Failure", "DROCC: Deep Robust One-Class Classification", "G2D: Generate to Detect Anomaly"], "answer_arxiv_id": ["1711.09325", "2007.10088", "2002.12718", "2006.11629"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_11043"} +{"question": "Which paper provides an influential benchmark of deep semi-supervised methods?", "answer": ["Realistic Evaluation of Deep Semi-Supervised Learning Algorithms"], "answer_arxiv_id": ["1804.09170"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_11044"} +{"question": "Can you provide some references that discussed the application of variational autoencoders (VAEs) to conditional generative models?", "answer": ["Auto-Encoding Variational Bayes", "Self-Attention Generative Adversarial Networks", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image\n Synthesis"], "answer_arxiv_id": ["1312.6114", "1805.08318", "1904.01310"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_11045"} +{"question": "Could you provide me some research works about reasoning over time intervals in long-term dialogues?", "answer": ["Mind the Gap Between Conversations for Improved Long-Term Dialogue\n Generation", "Conversation Chronicles: Towards Diverse Temporal and Relational\n Dynamics in Multi-Session Conversations"], "answer_arxiv_id": ["2310.15415", "2310.13420"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_11046"} +{"question": "What studies use autoregressive or reconstruction-based approaches and predict aspects of the input graph for learning representations?", "answer": ["Strategies for Pre-training Graph Neural Networks", "GPT-GNN: Generative Pre-Training of Graph Neural Networks", "Self-Supervised Graph Transformer on Large-Scale Molecular Data", "N-Gram Graph: Simple Unsupervised Representation for Graphs, with Applications to Molecules"], "answer_arxiv_id": ["1905.12265", "2006.15437", "2007.02835", "1806.09206"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_11047"} +{"question": "Could you provide me with some studies that integrated different transformer networks for learning long-range contextual features?", "answer": ["NDT-Transformer: Large-Scale 3D Point Cloud Localisation using the\n Normal Distribution Transform Representation", "PPFNet: Global Context Aware Local Features for Robust 3D Point Matching", "SVT-Net: Super Light-Weight Sparse Voxel Transformer for Large Scale\n Place Recognition", "OverlapTransformer: An Efficient and Rotation-Invariant Transformer\n Network for LiDAR-Based Place Recognition", "AttDLNet: Attention-based DL Network for 3D LiDAR Place Recognition", "CVTNet: A Cross-View Transformer Network for Place Recognition Using LiDAR Data"], "answer_arxiv_id": ["2103.12292", "1802.02669", "2105.00149", "2203.03397", "2106.09637", "2302.01665v2"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_11048"} +{"question": "Which study tackled exemplar-guided image editing as an inpainting task?", "answer": ["Paint by Example: Exemplar-based Image Editing with Diffusion Models"], "answer_arxiv_id": ["2211.13227"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11049"} +{"question": "Which work derived several invariances of the flow and compared the dynamics in parameter and function spaces?", "answer": ["On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization"], "answer_arxiv_id": ["1802.06509"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_11050"} +{"question": "Could you provide me some studies on modifications and regularizations for MLP-based architectures?", "answer": ["On Embeddings for Numerical Features in Tabular Deep Learning"], "answer_arxiv_id": ["2203.05556"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_11051"} +{"question": "Could you provide me some works showing the action of GD in balancing factor matrices in asymmetric low-rank matrix sensing?", "answer": ["Beyond Procrustes: Balancing-Free Gradient Descent for Asymmetric Low-Rank Matrix Sensing"], "answer_arxiv_id": ["2101.05113"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_11052"} +{"question": "Could you provide me some works about the training of unified large multimodal models?", "answer": ["Emu: Generative Pretraining in Multimodality", "Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction\n Tuning", "Planting a SEED of Vision in Large Language Model", "DreamLLM: Synergistic Multimodal Comprehension and Creation"], "answer_arxiv_id": ["2307.05222", "2309.02591", "2307.08041", "2309.11499"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_11053"} +{"question": "Can you provide the works which allow region-specific understanding in LMMs?", "answer": ["Kosmos-2: Grounding Multimodal Large Language Models to the World", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "Ferret: Refer and Ground Anything Anywhere at Any Granularity", "The All-Seeing Project: Towards Panoptic Visual Recognition and\n Understanding of the Open World"], "answer_arxiv_id": ["2306.14824", "2306.15195", "2307.03601", "2305.11175", "2310.07704", "2308.01907"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_11054"} +{"question": "Could you provide me some works that leveraged MLPs for 3D scene representations in novel view synthesis?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_11055"} +{"question": "Could you provide me some studies that advanced diffusion models with different sampling methods and applied conditional control?", "answer": ["Pseudo Numerical Methods for Diffusion Models on Manifolds", "Learning Fast Samplers for Diffusion Models by Differentiating Through\n Sample Quality", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis", "Improved Techniques for Training Score-Based Generative Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2202.09778", "2202.05830", "2010.02502", "2105.05233", "2006.09011", "2112.10741"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_11056"} +{"question": "What are the studies that developed an end-to-end system integrating flow, confidence, and geometric optimization?", "answer": ["DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras"], "answer_arxiv_id": ["2108.10869"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_11057"} +{"question": "What studies created large-scale image datasets for training and evaluating gaze estimation models?", "answer": ["Appearance-Based Gaze Estimation in the Wild"], "answer_arxiv_id": ["1504.02863"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_11058"} +{"question": "What studies consider general Bayesian models, such as BNNs, to represent a probabilistic distribution for the learned model?", "answer": ["Model-Based Active Exploration", "Self-Supervised Exploration via Disagreement", "Planning to Explore via Self-Supervised World Models", "Curious Exploration via Structured World Models Yields Zero-Shot Object Manipulation"], "answer_arxiv_id": ["1810.12162", "1906.04161", "2005.05960", "2206.11403"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_11059"} +{"question": "Are there any works on controllable generation that use diffusion models for image-to-image translation?", "answer": ["Palette: Image-to-Image Diffusion Models", "Sketch-Guided Text-to-Image Diffusion Models", "Pretraining is All You Need for Image-to-Image Translation"], "answer_arxiv_id": ["2111.05826", "2211.13752", "2205.12952"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_11060"} +{"question": "Any studies about the ground sample distance-based positional encoding for optical satellite imagery?", "answer": ["Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial\n Representation Learning"], "answer_arxiv_id": ["2212.14532"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_11061"} +{"question": "Which researchers focused on the field of audio-visual question answering?", "answer": ["Audio Visual Scene-Aware Dialog", "A Simple Baseline for Audio-Visual Scene-Aware Dialog", "Dynamic Graph Representation Learning for Video Dialog via Multi-Modal Shuffled Transformers", "Learning to Answer Questions in Dynamic Audio-Visual Scenarios", "Audio-Visual Scene-Aware Dialog and Reasoning using Audio-Visual Transformers with Joint Student-Teacher Learning"], "answer_arxiv_id": ["1901.09107", "1904.05876", "2007.03848", "2203.14072", "2110.06894"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_11062"} +{"question": "Which works have employed text-based filtering for HTML documents in web navigations?", "answer": ["Mind2Web: Towards a Generalist Agent for the Web", "Multimodal Web Navigation with Instruction-Finetuned Foundation Models", "A Real-World WebAgent with Planning, Long Context Understanding, and\n Program Synthesis"], "answer_arxiv_id": ["2306.06070", "2305.11854", "2307.12856"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_11063"} +{"question": "Could you provide me studies that proposed the negative hard mining criterion for contrastive learning loss?", "answer": ["Debiased Contrastive Learning", "Contrastive Learning with Hard Negative Samples"], "answer_arxiv_id": ["2007.00224", "2010.04592"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_11064"} +{"question": "Which works developed a robust federated learning algorithm by considering a structured affine distribution shift in users’ data?", "answer": ["Robust Federated Learning: The Case of Affine Distribution Shifts"], "answer_arxiv_id": ["2006.08907"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_11065"} +{"question": "What research papers focus on improving the performance of adversarial contrastive learning (ACL)?", "answer": ["When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning?", "Adversarial Contrastive Learning via Asymmetric InfoNCE", "Rethinking the Effect of Data Augmentation in Adversarial Contrastive Learning"], "answer_arxiv_id": ["2111.01124", "2207.08374", "2303.01289"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_11066"} +{"question": "Which works report existing 3D native diffusion models and their limitations?", "answer": ["Shap-E: Generating Conditional 3D Implicit Functions", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "3DGen: Triplane Latent Diffusion for Textured Mesh Generation", "ATT3D: Amortized Text-to-3D Object Synthesis", "3DShape2VecSet: A 3D Shape Representation for Neural Fields and\n Generative Diffusion Models", "Locally Attentional SDF Diffusion for Controllable 3D Shape Generation"], "answer_arxiv_id": ["2305.02463", "2212.08751", "2303.05371", "2306.07349", "2301.11445", "2305.04461"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11067"} +{"question": "Which studies have used sparsity in transformer to reduce memory footprint and achieve efficient attention mechanism?", "answer": ["Generating Long Sequences with Sparse Transformers"], "answer_arxiv_id": ["1904.10509"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_11068"} +{"question": "What studies have been devoted to differentially private adaptive optimization?", "answer": ["Adaptive Bound Optimization for Online Convex Optimization", "On the convergence of Adam and Beyond", "Private Stochastic Non-Convex Optimization: Adaptive Algorithms and Tighter Generalization Bounds", "Private Adaptive Optimization with Side information", "Private Adaptive Gradient Methods for Convex Optimization", "Shampoo: Preconditioned Stochastic Tensor Optimization"], "answer_arxiv_id": ["1002.4908", "1904.09237", "2006.13501", "2202.05963", "2106.13756", "1802.09568"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_11069"} +{"question": "Could you specify the works that train on 2D human videos for novel view synthesis utilizing NeRF?", "answer": ["Neural Articulated Radiance Field", "Neural Actor: Neural Free-view Synthesis of Human Actors with Pose\n Control", "Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "Learning Implicit Fields for Generative Shape Modeling"], "answer_arxiv_id": ["2104.03110", "2106.02019", "2012.15838", "2201.04127", "1812.02822"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_11070"} +{"question": "Which works have focused on regions of disparate performance between human and AI?", "answer": ["Domino: Discovering Systematic Errors with Cross-Modal Embeddings", "Auditing AI models for Verified Deployment under Semantic Specifications", "Discovering Bugs in Vision Models using Off-the-shelf Image Generation and Captioning", "The Spotlight: A General Method for Discovering Systematic Errors in Deep Learning Models", "SEAL: Interactive Tool for Systematic Error Analysis and Labeling", "Distilling Model Failures as Directions in Latent Space"], "answer_arxiv_id": ["2203.14960", "2109.12456", "2208.08831", "2107.00758", "2210.05839", "2206.14754"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_11071"} +{"question": "Which works extended the framework of auto-parsing structures with reinforcement learning?", "answer": ["Cooperative Learning of Disjoint Syntax and Semantics"], "answer_arxiv_id": ["1902.09393"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_11072"} +{"question": "What studies deal with multi-task imitation learning where the task specifications are vector states?", "answer": ["Visual Reinforcement Learning with Imagined Goals"], "answer_arxiv_id": ["1807.04742"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_11073"} +{"question": "Can you name some pioneer works in Self-Supervised Learning that proposed tasks based on data reconstruction like context prediction?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Context Encoders: Feature Learning by Inpainting"], "answer_arxiv_id": ["1505.05192", "1604.07379"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_11074"} +{"question": "What are the studies investigating why and how large language models perform in-context learning?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference", "What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "Transformers Learn In-Context by Gradient Descent", "What learning algorithm is in-context learning? Investigations with linear models", "Looped Transformers as Programmable Computers"], "answer_arxiv_id": ["2111.02080", "2208.01066", "2212.07677", "2211.15661", "2301.13196v1"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_11075"} +{"question": "Are there any papers about non-regressive MIDI-based music generation correlated with video?", "answer": ["Foley Music: Learning to Generate Music from Videos", "Audeo: Audio Generation for a Silent Performance Video", "Multi-Instrumentalist Net: Unsupervised Generation of Music from Body Movements"], "answer_arxiv_id": ["2007.10984", "2006.14348", "2012.03478"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_11076"} +{"question": "What papers have studied auxiliary learning in settings with large quantities of target data?", "answer": ["Weighted Training for Cross-Task Learning", "Unsupervised Cross-Task Generalization via Retrieval Augmentation", "FETA: A Benchmark for Few-Sample Task Transfer in Open-Domain Dialogue"], "answer_arxiv_id": ["2105.14095", "2204.07937", "2205.06262"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_11077"} +{"question": "Which works tune the models on language model generated visual instructions?", "answer": ["Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_11078"} +{"question": "What research has been carried out using text or speech as training signals in self-supervised video features?", "answer": ["HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips", "Look at What I’m Doing: Self-Supervised Spatial Grounding of Narrations in Instructional Videos", "Learning To Recognize Procedural Activities with Distant Supervision", "Egocentric Video-Language Pretraining", "Procedure-Aware Pretraining for Instructional Video Understanding"], "answer_arxiv_id": ["1906.03327", "2110.10596", "2201.10990", "2206.01670", "2303.18230"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_11079"} +{"question": "What studies have been conducted on enhancing Message-Passing Graph Neural Networks (MPNNs) and Graph Transformers with positional encoding (PE) and structural encodings (SE)?", "answer": ["Position-aware Graph Neural Networks", "Graph Attention Networks with Positional Embeddings", "Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning", "Eigen-GNN: A Graph Structure Preserving Plug-in for GNNs", "Benchmarking Graph Neural Networks", "What Graph Neural Networks Cannot Learn: Depth vs Width", "Graph Neural Networks with Learnable Structural and Positional Representations", "Sign and Basis Invariant Networks for Spectral Graph Representation Learning", "Equivariant and Stable Positional Encoding for More Powerful Graph Neural Networks"], "answer_arxiv_id": ["1906.04817", "2105.04037", "2009.00142", "2006.04330", "2003.00982", "1907.03199", "2110.07875", "2202.13013", "2203.00199"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_11080"} +{"question": "Are there any works that have attempted to exploit the notion of weight disentanglement by inducing task-specific subnetworks within a larger network?", "answer": ["Supermasks in Superposition", "Training independent subnetworks for robust prediction", "BatchEnsemble: An alternative approach to Efficient Ensemble and Lifelong Learning", "PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning", "Piggyback: Adapting a Single Network to Multiple Tasks by Learning to Mask Weights"], "answer_arxiv_id": ["2006.14769", "2010.06610", "2002.06715", "1711.05769", "1801.06519"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_11081"} +{"question": "What works have focused on designing efficient 3D CNN-based architectures for action recognition?", "answer": ["X3D: Expanding Architectures for Efficient Video Recognition", "MoViNets: Mobile Video Networks for Efficient Video Recognition"], "answer_arxiv_id": ["2004.04730", "2103.11511"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_11082"} +{"question": "Which studies have attempted to improve the functionality of FNO by merging it with non-equispaced interpolation layers?", "answer": ["Non-equispaced Fourier Neural Solvers for PDEs"], "answer_arxiv_id": ["2212.04689"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_11083"} +{"question": "What studies investigated how to provide finer-grained control over the generative process, such as using regional prompting and additional user-provided image information?", "answer": ["MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation", "SceneComposer: Any-Level Semantic Image Synthesis", "Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.08113", "2211.11742", "2302.05543", "2302.08453"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_11084"} +{"question": "What research papers synthetize 3D-aware images using diffusion models?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "Shap-E: Generating Conditional 3D Implicit Functions", "3D Neural Field Generation using Triplane Diffusion", "3DGen: Triplane Latent Diffusion for Textured Mesh Generation", "AutoDecoding Latent 3D Diffusion Models"], "answer_arxiv_id": ["2209.14988", "2212.08751", "2305.02463", "2211.16677", "2303.05371", "2307.05445"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_11085"} +{"question": "Which works employed contrastive learning strategy in self-supervised learning for remote sensing tasks?", "answer": ["Geography-Aware Self-Supervised Learning", "Seasonal Contrast: Unsupervised Pre-Training from Uncurated Remote\n Sensing Data"], "answer_arxiv_id": ["2011.09980", "2103.16607"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_11086"} +{"question": "Could you list some studies that have employed contrastive learning to prevent representational collapse?", "answer": ["Unsupervised State Representation Learning in Atari", "CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Representation Learning with Contrastive Predictive Coding", "Decoupling Representation Learning from Reinforcement Learning"], "answer_arxiv_id": ["1906.08226", "2004.04136", "1807.03748", "2009.08319"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11087"} +{"question": "Which papers discuss the use of retrieval-augmentation in small language models?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension"], "answer_arxiv_id": ["1910.10683", "1910.13461"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_11088"} +{"question": "Which research introduced the concept of Hierarchical Reinforcement Learning (HRL)?", "answer": ["FeUdal Networks for Hierarchical Reinforcement Learning", "The Option-Critic Architecture"], "answer_arxiv_id": ["1703.01161", "1609.05140"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11089"} +{"question": "What work is most related to the value interval learning in minimax learning?", "answer": ["Minimax Value Interval for Off-Policy Evaluation and Policy Optimization"], "answer_arxiv_id": ["2002.02081"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_11090"} +{"question": "Which paper introduces the Flamingo model in the context of large vision-language models (LVLMs)?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_11091"} +{"question": "What works have examined word-level grounding with detection and segmentation tasks on 3D data?", "answer": ["ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "Matterport3D: Learning from RGB-D Data in Indoor Environments"], "answer_arxiv_id": ["1702.04405", "1709.06158"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_11092"} +{"question": "What papers discussed risk-averse Bayesian Optimization?", "answer": ["Bayesian Optimization of Risk Measures", "Mean-Variance Analysis in Bayesian Optimization under Uncertainty", "Value-at-Risk Optimization with Gaussian Processes"], "answer_arxiv_id": ["2007.05554v3", "2009.08166", "2105.06126"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_11093"} +{"question": "What works have been used to identify Simpson's paradoxes in deep learning contests through the study of the correlation among HT-SR-related generalization metrics, models, and task and test performances?", "answer": ["Post-mortem on a deep learning contest: a Simpson’s paradox and the complementary roles of scale metrics versus shape metrics"], "answer_arxiv_id": ["2106.00734"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_11094"} +{"question": "What papers present approaches for rendering using NeRF?", "answer": ["Light Field Networks: Neural Scene Representations with Single-Evaluation Rendering", "DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields using Depth Oracle Networks", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs"], "answer_arxiv_id": ["2106.02634", "2103.03231", "2103.13744"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_11095"} +{"question": "What are the studies relating to ensemble learning in Transformer inference?", "answer": ["QiaoNing at SemEval-2020 Task 4: Commonsense Validation and Explanation system based on ensemble of language model", "Autoencoding Language Model Based Ensemble Learning for Commonsense Validation and Explanation", "Improving BERT Fine-Tuning via Self-Ensemble and Self-Distillation", "Ensemble Transformer for Efficient and Accurate Ranking Tasks: an Application to Question Answering Systems", "DynE: Dynamic Ensemble Decoding for Multi-Document Summarization"], "answer_arxiv_id": ["2009.02645", "2204.03324", "2002.10345", "2201.05767", "2006.08748"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_11096"} +{"question": "What work analyzed the numerical inversion error of generic normalizing flows via bi-Lipschitz continuity for each flow layer?", "answer": ["Understanding and Mitigating Exploding Inverses in Invertible Neural Networks"], "answer_arxiv_id": ["2006.09347"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_11097"} +{"question": "Which papers can be referred to while studying video-to-video translation?", "answer": ["Video-to-Video Synthesis", "Recycle-GAN: Unsupervised Video Retargeting"], "answer_arxiv_id": ["1808.06601", "1808.05174"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_11098"} +{"question": "Which works discuss the 'Loop Unrolled' architecture in deep learning for solving inverse problems?", "answer": ["Learning to learn by gradient descent by gradient descent", "Learned Primal-dual Reconstruction", "FISTA-Net: Learning A Fast Iterative Shrinkage Thresholding Network for Inverse Problems in Imaging", "HUMUS-Net: Hybrid Unrolled Multi-Scale Network Architecture for Accelerated MRI Reconstruction", "Loop Unrolled Shallow Equilibrium Regularizer (LUSER) - A Memory-Efficient Inverse Problem Solver"], "answer_arxiv_id": ["1606.04474", "1707.06474v3", "2008.02683", "2203.08213", "2210.04987"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_11099"} +{"question": "Any works about functional logistic regression?", "answer": ["On functional logistic regression: some conceptual issues"], "answer_arxiv_id": ["1812.00721v2"], "source_meta": {"published_time": "20230114"}, "qid": "AutoScholarQuery_train_11100"} +{"question": "Which early methods explicitly represent human avatars with deformable templates such as SMPL?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image"], "answer_arxiv_id": ["1904.05866"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_11101"} +{"question": "What are some research papers that propose algorithms for contextual bandits with Lipschitz/Hölder regression functions?", "answer": ["Polynomial Cost of Adaptation for X -Armed Bandits"], "answer_arxiv_id": ["1905.10221v2"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_11102"} +{"question": "Which works propose using smoothness prior for optimization in unsupervised semantic segmentation?", "answer": ["Unsupervised Semantic Segmentation by Distilling Feature Correspondences"], "answer_arxiv_id": ["2203.08414"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_11103"} +{"question": "What studies introduced additional domain knowledge or utilized text generation to enrich the information for training in the field of pathological image analysis?", "answer": ["Text-guided Foundation Model Adaptation for Pathological Image\n Classification", "The Rise of AI Language Pathologists: Exploring Two-level Prompt\n Learning for Few-shot Weakly-supervised Whole Slide Image Classification"], "answer_arxiv_id": ["2307.14901", "2305.17891"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_11104"} +{"question": "What is the proposal provided by 'GLIP' and 'Grounding DINO' in relation to open-vocabulary detection?", "answer": ["Grounded Language-Image Pre-training", "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection"], "answer_arxiv_id": ["2112.03857", "2303.05499"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_11105"} +{"question": "Which works initated the CoPE literature with the single-task and classic MAB setting?", "answer": ["Distributed Exploration in Multi-Armed Bandits", "Collaborative Learning with Limited Interaction: Tight Bounds for Distributed Exploration in Multi-Armed Bandits"], "answer_arxiv_id": ["1311.0800", "1904.03293"], "source_meta": {"published_time": "20211029"}, "qid": "AutoScholarQuery_train_11106"} +{"question": "What studies focused on adversarial training as a method to defend adversarial examples?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Overfitting in adversarially robust deep learning", "Unlabeled Data Improves Adversarial Robustness"], "answer_arxiv_id": ["1706.06083", "2002.11569", "1905.13736"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_11107"} +{"question": "Who introduced latent diffusion for image synthesis?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_11108"} +{"question": "Which paper introduces a method to select the most activated experts for pruning in machine translation MoE models?", "answer": ["Scalable and Efficient MoE Training for Multitask Multilingual Models"], "answer_arxiv_id": ["2109.10465"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_11109"} +{"question": "What papers discuss the role of Vision Language models (VLMs) in zero-shot image classification?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "FLAVA: A Foundational Language And Vision Alignment Model", "Florence: A New Foundation Model for Computer Vision", "CLOOB: Modern Hopfield Networks with InfoLOOB Outperform CLIP"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2112.04482", "2111.11432", "2110.11316"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_11110"} +{"question": "What works have adopted the voting scheme, as introduced by VoteNet, in their point cloud-based methods?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering"], "answer_arxiv_id": ["2102.13090"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_11111"} +{"question": "Could you provide me studies in which RPG was established for s-rectangular reward-robust MDPs?", "answer": ["Twice regularized MDPs and the equivalence between robustness and regularization"], "answer_arxiv_id": ["2110.06267"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_11112"} +{"question": "Which researches present benchmark competitions in data-centric AI?", "answer": ["DataComp: In search of the next generation of multimodal datasets"], "answer_arxiv_id": ["2304.14108"], "source_meta": {"published_time": "20220720"}, "qid": "AutoScholarQuery_train_11113"} +{"question": "Which works showed nonlinear LSNMs are identifiable with Gaussian noise?", "answer": ["Causal Autoregressive Flows"], "answer_arxiv_id": ["2011.02268v2"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_11114"} +{"question": "Which papers surveyed the low-rank approximation problem in machine learning and data science?", "answer": ["Randomized algorithms for matrices and data", "Sketching as a Tool for Numerical Linear Algebra"], "answer_arxiv_id": ["1104.5557v3", "1411.4357"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_11115"} +{"question": "What studies extend Model Inversion for complex DNNs under a white-box setup?", "answer": ["The Secret Revealer: Generative Model-Inversion Attacks Against Deep Neural Networks", "Knowledge-Enriched Distributional Model Inversion Attacks", "Variational Model Inversion Attacks", "Pseudo Label-Guided Model Inversion Attack via Conditional Generative Adversarial Network", "Re-thinking Model Inversion Attacks Against Deep Neural Networks"], "answer_arxiv_id": ["1911.07135", "2010.04092", "2201.10787", "2302.09814", "2304.01669"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_11116"} +{"question": "Could you provide me some studies that combine self-supervised learning with clustering?", "answer": ["Deep Clustering for Unsupervised Learning of Visual Features", "Unsupervised Pre-Training of Image Features on Non-Curated Data", "Prototypical Contrastive Learning of Unsupervised Representations", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Self-labelling via simultaneous clustering and representation learning"], "answer_arxiv_id": ["1807.05520", "1905.01278", "2005.04966", "2006.09882", "1911.05371"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_11117"} +{"question": "Which papers use pruning and probing techniques to demonstrate model capabilities?", "answer": ["Mechanistically analyzing the effects of fine-tuning on procedurally\n defined tasks"], "answer_arxiv_id": ["2311.12786"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_11118"} +{"question": "Which works explored re-calibrating batch normalization statistics to make deep neural networks robust to image corruptions?", "answer": ["Improving robustness against common corruptions by covariate shift adaptation", "Revisiting Batch Normalization for Improving Corruption Robustness", "Evaluating Prediction-Time Batch Normalization for Robustness under\n Covariate Shift"], "answer_arxiv_id": ["2006.16971v2", "2010.03630", "2006.10963"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_11119"} +{"question": "Which research papers introduced machine learning methods for fluid dynamics?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations", "Clifford Neural Layers for PDE Modeling", "Message Passing Neural PDE Solvers"], "answer_arxiv_id": ["2010.08895", "2209.04934", "2202.03376"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_11120"} +{"question": "Which papers discuss data augmentation on outdoor scenes for 3D perception tasks?", "answer": ["Exploring Data Augmentation for Multi-Modality 3D Object Detection", "Exploring Geometric Consistency for Monocular 3D Object Detection", "Augmented Reality Meets Computer Vision : Efficient Data Generation for Urban Driving Scenes", "GeoSim: Realistic Video Simulation via Geometry-Aware Composition for Self-Driving", "3D Data Augmentation for Driving Scenes on Camera"], "answer_arxiv_id": ["2012.12741", "2104.05858", "1708.01566", "2101.06543", "2303.10340"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_11121"} +{"question": "Which work used Pearson correlation to calculate neuron similarity across models?", "answer": ["Universal Neurons in GPT2 Language Models"], "answer_arxiv_id": ["2401.12181"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_11122"} +{"question": "Which are the conventional normalization methods in deep learning the article refers to?", "answer": ["Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift", "Layer Normalization", "Group Normalization"], "answer_arxiv_id": ["1502.03167", "1607.06450", "1803.08494"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_11123"} +{"question": "Could you provide me some research papers which show that language models can be prompted or adapted to produce these facts on-demand?", "answer": ["How Can We Know What Language Models Know?", "Language Models as Knowledge Bases?", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts", "Generate rather than Retrieve: Large Langu-age Models are Strong Context Generators", "COMET: Commonsense Transformers for Automatic Knowledge Graph Construction", "(Comet-)Atomic₂₀²⁰: On Symbolic and Neural Commonsense Knowledge Graphs", "“I’m Not Mad”: Commonsense Implications of Negation and Contradiction", "How Much Knowledge Can You Pack Into the Parameters of a Language Model?"], "answer_arxiv_id": ["1911.12543", "1909.01066", "2010.15980", "2209.10063", "1906.05317", "2010.05953", "2104.06511", "2002.08910"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_11124"} +{"question": "What papers provide information on image vectorization approaches?", "answer": ["Vectorization of Line Drawings via PolyVector Fields", "Towards Layer-wise Image Vectorization"], "answer_arxiv_id": ["1801.01922", "2206.04655"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_11125"} +{"question": "Can you provide examples of studies that use auxiliary tasks and supervisory signals in sequential disentanglement of arbitrary data?", "answer": ["S3VAE: Self-Supervised Sequential VAE for Representation Disentanglement and Data Generation"], "answer_arxiv_id": ["2005.11437"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11126"} +{"question": "Could you provide me studies applying the spatial coupling technique as a refinement to AMP methods?", "answer": ["Probabilistic Reconstruction in Compressed Sensing: Algorithms, Phase Diagrams, and Threshold Achieving Matrices", "Information-Theoretically Optimal Compressed Sensing via Spatial Coupling and Approximate Message Passing", "State Evolution for General Approximate Message Passing Algorithms, with Applications to Spatial Coupling", "Graph-based Approximate Message Passing Iterations"], "answer_arxiv_id": ["1206.3953", "1112.0708", "1211.5164", "2109.11905"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_11127"} +{"question": "Which work also considers the importance of data quantities in collaboration performances and optimization?", "answer": ["Model-sharing Games: Analyzing Federated Learning Under Voluntary Participation"], "answer_arxiv_id": ["2010.00753"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_11128"} +{"question": "Which works in 4D perception research use 3D or higher dimension convolutions particularly on LiDAR point cloud videos for autonomous driving perception system?", "answer": ["4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "Robust Multi-Modality Multi-Object Tracking", "3D Multi-Object Tracking: A Baseline and New Evaluation Metrics"], "answer_arxiv_id": ["1904.08755", "1909.03850", "1907.03961"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_11129"} +{"question": "Which research papers have focused on modeling emerging entities?", "answer": ["Knowledge Transfer for Out-of-Knowledge-Base Entities: A Graph Neural Network Approach"], "answer_arxiv_id": ["1706.05674"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11130"} +{"question": "What studies are there on the more recent methods of document retrieval which use the representative vector embedding?", "answer": ["Embedding-based Retrieval in Facebook Search", "Semantic Product Search", "Embedding-based Product Retrieval in Taobao Search"], "answer_arxiv_id": ["2006.11632", "1907.00937", "2106.09297"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_11131"} +{"question": "Can you provide some works that propose methods to automatically choose task weights in ATL?", "answer": ["Adapting Auxiliary Losses Using Gradient Similarity", "Auxiliary Task Update Decomposition: The Good, The Bad and The Neutral", "Weighted Training for Cross-Task Learning", "Auxiliary Task Reweighting for Minimum-data Learning", "Auxiliary Learning by Implicit Differentiation", "Auxiliary Learning as an Asymmetric Bargaining Game"], "answer_arxiv_id": ["1812.02224", "2108.11346", "2105.14095", "2010.08244", "2007.02693", "2301.13501"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11132"} +{"question": "Which studies question the generality of harmless interpolation for models that incorporate strong structural assumptions?", "answer": ["Augmentation Strategies for Learning with Noisy Labels", "Overfitting in adversarially robust deep learning", "Can we have it all? On the Trade-off between Spatial and Adversarial Robustness of Neural Networks", "How benign is benign overfitting?", "Interpolation can hurt robust generalization even when there is no noise", "How Does a Neural Network’s Architecture Impact Its Robustness to Noisy Labels?", "Provable Benefits of Overparameterization in Model Compression: From Double Descent to Pruning Neural Networks", "Harmless interpolation of noisy data in regression", "Foolish Crowds Support Benign Overfitting"], "answer_arxiv_id": ["2103.02130", "2002.11569", "2002.11318", "2007.04028", "2108.02883", "2012.12896", "2012.08749", "1903.09139", "2110.02914"], "source_meta": {"published_time": "20230118"}, "qid": "AutoScholarQuery_train_11133"} +{"question": "Which papers propose model-based approaches in offline RL?", "answer": ["MOPO: Model-based Offline Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning", "COMBO: Conservative Offline Model-Based Policy Optimization", "Pessimistic Model-based Offline Reinforcement Learning under Partial Coverage", "RAMBO-RL: Robust Adversarial Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2005.13239", "2005.05951", "2102.08363", "2107.06226", "2204.12581"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_11134"} +{"question": "Could you provide me with references about rescaling-invariant sharpness in the context of flat minima generalization?", "answer": ["Fisher-Rao Metric, Geometry, and Complexity of Neural Networks", "Normalized Flat Minima: Exploring Scale-Invariant Definition of Flat Minima for Neural Networks using PAC-Bayesian Analysis", "Relative Flatness and Generalization"], "answer_arxiv_id": ["1711.01530", "1901.04653", "2001.00939"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_11135"} +{"question": "Which works provide a study on the sub-field of object detection, part detection?", "answer": ["Mask R-CNN", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", "Rich feature hierarchies for accurate object detection and semantic segmentation", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["1703.06870", "1506.01497", "1311.2524", "2005.12872"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_11136"} +{"question": "Could you provide me some studies focused on estimating interventional queries using neural methods?", "answer": ["CausalGAN: Learning Causal Implicit Generative Models with Adversarial Training", "The Causal-Neural Connection: Expressiveness, Learnability, and Inference", "Relating Graph Neural Networks to Structural Causal Models"], "answer_arxiv_id": ["1709.02023", "2107.00793", "2109.04173v3"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_11137"} +{"question": "Are there any works that focus on using word embeddings to assist in segmentation?", "answer": ["MIANet: Aggregating Unbiased Instance and General Information for\n Few-Shot Semantic Segmentation"], "answer_arxiv_id": ["2305.13864"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_11138"} +{"question": "What papers mentioned the use of domain knowledge to improve the safety of an RL agent?", "answer": ["Safe Exploration in Continuous Action Spaces", "Towards Safe Reinforcement Learning with a Safety Editor Policy", "Learning Barrier Certificates: Towards Safe Reinforcement Learning with Zero Training-time Violations", "Context-Aware Safe Reinforcement Learning for Non-Stationary Environments", "Trial without Error: Towards Safe Reinforcement Learning via Human Intervention", "Safe Reinforcement Learning via Shielding"], "answer_arxiv_id": ["1801.08757", "2201.12427", "2108.01846", "2101.00531", "1707.05173", "1708.08611"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_11139"} +{"question": "What are some recent works that independently developed variants of V-learning, a decentralized algorithm for learning unknown general-sum Markov games?", "answer": ["V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL", "When Can We Learn General-Sum Markov Games with a Large Number of Players Sample-Efficiently?", "Provably Efficient Reinforcement Learning in Decentralized General-Sum Markov Games"], "answer_arxiv_id": ["2110.14555", "2110.04184", "2110.05682v3"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_11140"} +{"question": "Which papers investigated the application of entropy regularization in extensive-form games?", "answer": ["The Power of Regularization in Solving Extensive-Form Games", "A Unified Approach to Reinforcement Learning, Quantal Response Equilibria, and Two-Player Zero-Sum Games"], "answer_arxiv_id": ["2206.09495", "2206.05825"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_11141"} +{"question": "Can you name the study that introduces a robust multimodal learning method to mitigate the impact of low-quality or noisy modalities?", "answer": ["Provable Dynamic Fusion for Low-Quality Multimodal Data"], "answer_arxiv_id": ["2306.02050"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_11142"} +{"question": "What are some studies on zero-shot learning by utilizing attributes or knowledge graphs?", "answer": ["Zero-shot Recognition via Semantic Embeddings and Knowledge Graphs"], "answer_arxiv_id": ["1803.08035"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_11143"} +{"question": "Which works considered hierarchical assembly in different levels for more effective Neural Architecture Search?", "answer": ["Hierarchical Representations for Efficient Architecture Search", "Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation"], "answer_arxiv_id": ["1711.00436", "1901.02985"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_11144"} +{"question": "Could you provide me with studies that observed a similar trajectory alignment phenomenon for scalar linear networks?", "answer": ["Gradient Descent Monotonically Decreases the Sharpness of Gradient Flow Solutions in Scalar Networks and Beyond"], "answer_arxiv_id": ["2305.13064"], "source_meta": {"published_time": "20230709"}, "qid": "AutoScholarQuery_train_11145"} +{"question": "Could you provide some studies about RLHF's application in gaming and robotics?", "answer": ["Deep reinforcement learning from human preferences", "Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback", "Fine-Tuning Language Models from Human Preferences", "Open Problems and Fundamental Limitations of Reinforcement Learning from\n Human Feedback"], "answer_arxiv_id": ["1706.03741", "2204.05862", "1909.08593", "2307.15217"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_11146"} +{"question": "What is the research that introduced robust OT for improving the accuracy and efficiency of point cloud registration?", "answer": ["Accurate Point Cloud Registration with Robust Optimal Transport"], "answer_arxiv_id": ["2111.00648"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_11147"} +{"question": "Could you provide me a study related to OPE that focuses on the estimation of the expected return at the initial state distribution, rather than recovering the full function?", "answer": ["Beyond the Return: Off-policy Function Estimation under User-specified Error-measuring Distributions"], "answer_arxiv_id": ["2210.15543"], "source_meta": {"published_time": "20230725"}, "qid": "AutoScholarQuery_train_11148"} +{"question": "Are there any studies focusing on the application of Transformer model in vision tasks in the form of the Vision Transformer (ViT)?", "answer": ["Training data-efficient image transformers & distillation through\n attention", "Tokens-to-Token ViT: Training Vision Transformers from Scratch on\n ImageNet", "Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction\n without Convolutions", "PVT v2: Improved Baselines with Pyramid Vision Transformer", "Transformer in Transformer", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2012.12877", "2101.11986", "2102.12122", "2106.13797", "2103.00112", "2103.14030"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_11149"} +{"question": "Can you provide me with works that studied the relevance of the log barrier in graph learning from smooth signals?", "answer": ["Graph Laplacian mixture model", "Multi-modal Graph Learning for Disease Prediction"], "answer_arxiv_id": ["1810.10053", "2203.05880"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_11150"} +{"question": "What research presents the experimental method of canary extraction?", "answer": ["The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks"], "answer_arxiv_id": ["1802.08232"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_11151"} +{"question": "Can you cite works where the researchers investigated causal inference with proxy variables in different settings?", "answer": ["Identifying Causal Effects With Proxy Variables of an Unmeasured Confounder", "Multiply Robust Causal Inference with Double Negative Control Adjustment for Categorical Unmeasured Confounding", "An Introduction to Proximal Causal Learning", "Semiparametric proximal causal inference"], "answer_arxiv_id": ["1609.08816", "1808.04906", "2009.10982v1", "2011.08411"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11152"} +{"question": "Which papers focused on the creation of low-level rotation invariant geometric features?", "answer": ["Learning Rotation-Invariant Representations of Point Clouds Using\n Aligned Edge Convolutional Neural Networks", "Rotation Invariant Point Cloud Classification: Where Local Geometry\n Meets Global Topology"], "answer_arxiv_id": ["2101.00483", "1911.00195"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_11153"} +{"question": "Are there any works which alter the objective to include the safety requirement in reinforcement learning?", "answer": ["Safe Reinforcement Learning with Chance-constrained Model Predictive Control", "Unsolved Problems in ML Safety", "Learning with Safety Constraints: Sample Complexity of Reinforcement Learning for Constrained MDPs", "Provable Safe Reinforcement Learning with Binary Feedback", "Near-Optimal Multi-Agent Learning for Safe Coverage Control", "Constrained Variational Policy Optimization for Safe Reinforcement Learning"], "answer_arxiv_id": ["2112.13941", "2109.13916", "2008.00311", "2210.14492", "2210.06380", "2201.11927"], "source_meta": {"published_time": "20220223"}, "qid": "AutoScholarQuery_train_11154"} +{"question": "What studies replaced the reconstruction loss with an adversarial loss of the GAN framework?", "answer": ["Autoencoding beyond pixels using a learned similarity metric", "Variational Approaches for Auto-Encoding Generative Adversarial Networks", "Implicit Discriminator in Variational Autoencoder"], "answer_arxiv_id": ["1512.09300v2", "1706.04987", "1909.13062"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_11155"} +{"question": "What research utilize diffirentiable rendering to generate 2D images from various views?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "Zero-Shot Text-Guided Object Generation with Dream Fields", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "Understanding Pure CLIP Guidance for Voxel Grid NeRF Models", "Text and Image Guided 3D Avatar Generation and Manipulation", "ClipFace: Text-guided Editing of Textured 3D Morphable Models", "ClipMatrix: Text-controlled Creation of 3D Textured Meshes", "Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and Text-to-Image Diffusion Models", "DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generation", "NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as General Image Priors", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "SparseFusion: Distilling View-conditioned Diffusion for 3D Reconstruction", "DreamBooth3D: Subject-Driven Text-to-3D Generation"], "answer_arxiv_id": ["2205.08535", "2112.01455", "2112.03221", "2209.15172", "2202.06079", "2212.01406", "2109.12922", "2212.14704", "2209.14988", "2211.10440", "2303.07937", "2212.03267", "2212.00774v1", "2211.07600", "2212.00792", "2303.13508"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_11156"} +{"question": "Which studies have addressed the task of precise, local editing for 3D representations?", "answer": ["Text2LIVE: Text-Driven Layered Image and Video Editing", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Prompt-to-Prompt Image Editing with Cross Attention Control", "FocalDreamer: Text-driven 3D Editing via Focal-fusion Assembly", "Vox-E: Text-guided Voxel Editing of 3D Objects", "DreamEditor: Text-Driven 3D Scene Editing with Neural Fields"], "answer_arxiv_id": ["2204.02491", "2211.09800", "2208.01618", "2208.01626", "2308.10608", "2303.12048", "2306.13455"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_11157"} +{"question": "What works have highlighted the mistakes made by machine learning models in new settings regardless of the high performance on IID test sets?", "answer": ["Building Machines That Learn and Think Like People", "Adversarial Examples for Evaluating Reading Comprehension Systems", "Strike (with) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects"], "answer_arxiv_id": ["1604.00289", "1707.07328", "1811.11553"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_11158"} +{"question": "What are the examples of proposed incremental second order methods?", "answer": ["IQN: An Incremental Quasi-Newton Method with Local Superlinear Convergence Rate", "Stochastic Newton and Cubic Newton Methods with Simple Local Linear-Quadratic Rates"], "answer_arxiv_id": ["1702.00709v2", "1912.01597"], "source_meta": {"published_time": "20220717"}, "qid": "AutoScholarQuery_train_11159"} +{"question": "What studies tried to explore the potential of Transformer-based models in time series representation learning?", "answer": ["A Transformer-based Framework for Multivariate Time Series Representation Learning", "Time-Series Representation Learning via Temporal and Contextual Contrasting"], "answer_arxiv_id": ["2010.02803", "2106.14112"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_11160"} +{"question": "Could you provide me some examples of research focused on single-view 3D reconstruction in large datasets?", "answer": ["NeRF-VAE: A Geometry Aware 3D Scene Generative Model", "LOLNeRF: Learn from One Look", "3D generation on ImageNet", "VQ3D: Learning a 3D-Aware Generative Model on ImageNet", "3D-aware Image Generation using 2D Diffusion Models"], "answer_arxiv_id": ["2104.00587", "2111.09996", "2303.01416", "2302.06833", "2303.17905"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_11161"} +{"question": "What is the work where the COCO dataset has been repurposed to address zero-shot detection?", "answer": ["Zero-Shot Object Detection"], "answer_arxiv_id": ["1804.04340"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_11162"} +{"question": "Which research provided fine-tuning model editors using KL-divergence?", "answer": ["Fast Model Editing at Scale"], "answer_arxiv_id": ["2110.11309"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_11163"} +{"question": "Which works propose optimization of the original Transformer to enable processing of longer sequences?", "answer": ["Generating Long Sequences with Sparse Transformers", "Reformer: The Efficient Transformer", "Longformer: The Long-Document Transformer", "Rethinking Attention with Performers", "Linformer: Self-Attention with Linear Complexity", "Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention", "Big Bird: Transformers for Longer Sequences", "cosFormer : Rethinking Softmax in Attention"], "answer_arxiv_id": ["1904.10509", "2001.04451", "2004.05150", "2009.14794", "2006.04768", "2006.16236", "2007.14062", "2202.08791"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_11164"} +{"question": "What papers are about the method of plane sweep volume (PSV) in multi-view stereo?", "answer": ["DeepStereo: Learning to Predict New Views from the World's Imagery", "Stereo Magnification: Learning View Synthesis using Multiplane Images", "DeepView: View Synthesis with Learned Gradient Descent", "Local Light Field Fusion: Practical View Synthesis with Prescriptive\n Sampling Guidelines"], "answer_arxiv_id": ["1506.06825", "1805.09817", "1906.07316", "1905.00889"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_11165"} +{"question": "Which studies have looked into the neural architecture search in the field of AutoML?", "answer": ["Neural Architecture Search with Reinforcement Learning", "Neural Architecture Search: A Survey"], "answer_arxiv_id": ["1611.01578", "1808.05377"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11166"} +{"question": "What works developed the T5 framework in generative transfer and multitask learning?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1910.10683"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_11167"} +{"question": "Could you provide me some research that use decomposed neural field to encode the 4D scenes?", "answer": ["SUDS: Scalable Urban Dynamic Scenes", "D$^2$NeRF: Self-Supervised Decoupling of Dynamic and Static Objects from\n a Monocular Video", "EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via\n Self-Supervision"], "answer_arxiv_id": ["2303.14536", "2205.15838", "2311.02077"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_11168"} +{"question": "What research work is done on CSC methods?", "answer": ["Read, Listen, and See: Leveraging Multimodal Information Helps Chinese\n Spell Checking", "Learning from the Dictionary: Heterogeneous Knowledge Guided Fine-tuning\n for Chinese Spell Checking", "The Past Mistake is the Future Wisdom: Error-driven Contrastive\n Probability Optimization for Chinese Spell Checking", "Correct Like Humans: Progressive Learning Framework for Chinese Text\n Error Correction", "A Frustratingly Easy Plug-and-Play Detection-and-Reasoning Module for\n Chinese Spelling Check"], "answer_arxiv_id": ["2105.12306", "2210.10320", "2203.00991", "2306.17447", "2310.09119"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_11169"} +{"question": "Who developed the method of image-language contrast for pre-training similar to ConVIRT?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "SLIP: Self-supervision meets Language-Image Pre-training", "Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Coarse-to-Fine Vision-Language Pre-training with Fusion in the Backbone", "PyramidCLIP: Hierarchical Feature Alignment for Vision-language Model Pretraining"], "answer_arxiv_id": ["2103.00020", "2112.12750", "2110.05208", "2205.01917", "2206.07643", "2204.14095"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11170"} +{"question": "Which research paper proposes a method for human-AI collaboration via conditional delegation rules that the human can write down?", "answer": ["Human-AI Collaboration via Conditional Delegation: A Case Study of Content Moderation"], "answer_arxiv_id": ["2204.11788"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_11171"} +{"question": "Which papers focus on counterfactual fairness in the context of individual fairness?", "answer": ["Post-processing for Individual Fairness"], "answer_arxiv_id": ["2110.13796"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_11172"} +{"question": "Can you provide some works where reasoning capabilities are used to address tasks beyond reasoning?", "answer": ["ReAct: Synergizing Reasoning and Acting in Language Models", "Inner Monologue: Embodied Reasoning through Planning with Language Models"], "answer_arxiv_id": ["2210.03629", "2207.05608"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_11173"} +{"question": "Are there any gradient-based attribution methods?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based\n Localization", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1610.02391", "1703.01365"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_11174"} +{"question": "Which works focuses on adding specific concepts to diffusion models with very few data samples?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "A Neural Space-Time Representation for Text-to-Image Personalization", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.01618", "2305.15391", "2208.12242"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_11175"} +{"question": "Which studies considered fluctuations and perturbations in the kernels at finite width during inference in Bayesian deep networks?", "answer": ["Asymptotics of representation learning in finite Bayesian neural networks", "Predicting the Outputs of Finite Deep Neural Networks Trained with Noisy Gradients"], "answer_arxiv_id": ["2106.00651", "2004.01190"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_11176"} +{"question": "Could you cite some papers that developed datasets for QA over structured EHR data?", "answer": ["Text-to-SQL Generation for Question Answering on Electronic Medical Records", "EHRSQL: A Practical Text-to-SQL Benchmark for Electronic Health Records"], "answer_arxiv_id": ["1908.01839", "2301.07695"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_11177"} +{"question": "What works developed stochastic bilevel approaches based on Neumann series?", "answer": ["Bilevel Optimization: Convergence Analysis and Enhanced Design", "Amortized Implicit Differentiation for Stochastic Bilevel Optimization"], "answer_arxiv_id": ["2010.07962", "2111.14580"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11178"} +{"question": "Can you list down the papers that use Learning methods in video summarization?", "answer": ["Video Skimming: Taxonomy and Comprehensive Survey", "Summary Transfer: Exemplar-based Subset Selection for Video\n Summarization"], "answer_arxiv_id": ["1909.12948", "1603.03369"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_11179"} +{"question": "Could you provide some studies on the extraction or reconstruction of the training data, particularly in large language models (LLMs)?", "answer": ["Extracting Training Data from Large Language Models"], "answer_arxiv_id": ["2012.07805"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11180"} +{"question": "Could you provide me some studies about the incorporation of hyperbolic geometry in the field of neural networks?", "answer": ["Hyperbolic Deep Neural Networks: A Survey", "Hyperbolic Graph Neural Networks: A Review of Methods and Applications", "Hyperbolic Graph Representation Learning: A Tutorial"], "answer_arxiv_id": ["2101.04562", "2202.13852", "2211.04050"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11181"} +{"question": "Which studies predict labels of unlabeled data by pseudo-labeling?", "answer": ["Billion-scale semi-supervised learning for image classification", "Self-training with Noisy Student improves ImageNet classification", "In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning", "Curriculum Labeling: Revisiting Pseudo-Labeling for Semi-Supervised Learning", "SimPLE: Similar Pseudo Label Exploitation for Semi-Supervised Classification"], "answer_arxiv_id": ["1905.00546", "1911.04252", "2101.06329", "2001.06001", "2103.16725"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_11182"} +{"question": "What studies proposed different assumptions for multi-domain generalization?", "answer": ["Domain Generalization via Multidomain Discriminant Analysis"], "answer_arxiv_id": ["1907.11216"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_11183"} +{"question": "What research explored hierarchical part decomposition in 3D objects?", "answer": ["GRASS: Generative Recursive Autoencoders for Shape Structures", "StructureNet: Hierarchical Graph Networks for 3D Shape Generation", "Learning Unsupervised Hierarchical Part Decomposition of 3D Objects from\n a Single RGB Image"], "answer_arxiv_id": ["1705.02090", "1908.00575", "2004.01176"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_11184"} +{"question": "Which study maps objects into a canonicalized object space for decreasing the gap in object geometry?", "answer": ["CAMS: CAnonicalized Manipulation Spaces for Category-Level Functional\n Hand-Object Manipulation Synthesis"], "answer_arxiv_id": ["2303.15469"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_11185"} +{"question": "What are some works on the traditional two-branch based few-shot object detection pipelines?", "answer": ["Few-Shot Object Detection with Fully Cross-Transformer", "Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector", "Query Adaptive Few-Shot Object Detection with Heterogeneous Graph Convolutional Networks", "Meta Faster R-CNN: Towards Accurate Few-Shot Object Detection with Attentive Feature Alignment", "Few-shot Object Detection via Feature Reweighting"], "answer_arxiv_id": ["2203.15021", "1908.01998", "2112.09791", "2104.07719", "1812.01866"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11186"} +{"question": "Can you name some papers where a KL penalty has been used for language model regularization?", "answer": ["Training language models to follow instructions with human feedback", "Rainier: Reinforced Knowledge Introspector for Commonsense Question Answering", "Recursively Summarizing Books with Human Feedback", "WebGPT: Browser-assisted question-answering with human feedback", "Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["2203.02155", "2210.03078", "2109.10862", "2112.09332", "1909.08593"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_11187"} +{"question": "Could you provide me some papers that proposed approximation algorithms for additive functions?", "answer": ["Approximately EFX Allocations for Indivisible Chores"], "answer_arxiv_id": ["2109.07313"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_11188"} +{"question": "Where can I find the application of PTI, NTI, and SINE in Group 2 approaches for text-based image editing?", "answer": ["Prompt Tuning Inversion for Text-Driven Image Editing Using Diffusion\n Models", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models", "SINE: SINgle Image Editing with Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2305.04441", "2211.09794", "2212.04489"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_11189"} +{"question": "Can you provide works which utilized pseudocode as an input form to evaluate program synthesis models?", "answer": ["SPoC: Search-based Pseudocode to Code"], "answer_arxiv_id": ["1906.04908"], "source_meta": {"published_time": "20220325"}, "qid": "AutoScholarQuery_train_11190"} +{"question": "Could you provide research that proposed approaches to enable synthesis from dynamic monocular video without precisely accurate poses?", "answer": ["Robust Dynamic Radiance Fields", "Progressively Optimized Local Radiance Fields for Robust View Synthesis"], "answer_arxiv_id": ["2301.02239", "2303.13791"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_11191"} +{"question": "What are the state-of-the-art graph SSL frameworks introduced according to a single pretext task with a single philosophy?", "answer": ["When Does Self-Supervision Help Graph Convolutional Networks?"], "answer_arxiv_id": ["2006.09136"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_11192"} +{"question": "Which study challenged the approach that considers adversarial perturbations as artifacts?", "answer": ["Adversarial Examples Are Not Bugs, They Are Features"], "answer_arxiv_id": ["1905.02175"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_11193"} +{"question": "What research works have used discrete Morse complex for image analysis?", "answer": ["Road Network Reconstruction from satellite images with Machine Learning Supported by Topological Methods"], "answer_arxiv_id": ["1909.06728"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_11194"} +{"question": "Which papers are inspired by the optimization methods for inversion in GANs?", "answer": ["Optimizing the Latent Space of Generative Networks", "Inverting The Generator Of A Generative Adversarial Network (II)", "Generative Visual Manipulation on the Natural Image Manifold", "Image Processing Using Multi-Code GAN Prior"], "answer_arxiv_id": ["1707.05776", "1802.05701", "1609.03552", "1912.07116"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_11195"} +{"question": "What works proposed multilingual instruction datasets?", "answer": ["Super-NaturalInstructions: Generalization via Declarative Instructions\n on 1600+ NLP Tasks", "Crosslingual Generalization through Multitask Finetuning", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset"], "answer_arxiv_id": ["2204.07705", "2211.01786", "2110.08207", "2303.03915"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_11196"} +{"question": "Can you provide a study that demonstrated how multilingual and Arabic monolingual LMs reflect trends from Western cultures?", "answer": ["Having Beer after Prayer? Measuring Cultural Bias in Large Language\n Models"], "answer_arxiv_id": ["2305.14456"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_11197"} +{"question": "Could you provide me some papers that propose group convolutions for various domains and more complex continuous groups?", "answer": ["3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data", "General E(2) - Equivariant Steerable CNNs", "Harmonic Networks: Deep Translation and Rotation Equivariance"], "answer_arxiv_id": ["1807.02547", "1911.08251", "1612.04642"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_11198"} +{"question": "Which papers address challenges posed by context window limits during the generation stage in Advanced RAG?", "answer": ["LLMLingua: Compressing Prompts for Accelerated Inference of Large\n Language Models", "Open-source Large Language Models are Strong Zero-shot Query Likelihood\n Models for Document Ranking"], "answer_arxiv_id": ["2310.05736", "2310.13243"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_11199"} +{"question": "Which studies propose strategies to mitigate the challenges due to distribution shifts in Offline Reinforcement Learning?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning", "EMaQ: Expected-Max Q-Learning Operator for Simple Yet Effective Offline and Online RL", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "A Minimalist Approach to Offline Reinforcement Learning", "Off-Policy Deep Reinforcement Learning without Exploration", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2006.04779", "1906.00949", "1911.11361", "2007.11091", "1906.00949", "2106.06860", "1812.02900", "2005.05951"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11200"} +{"question": "What studies discussed the limitations of gradient inversion attacks measured by image similarity?", "answer": ["Evaluating Gradient Inversion Attacks and Defenses in Federated Learning"], "answer_arxiv_id": ["2112.00059"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_11201"} +{"question": "What literature provides examples of the fusion-encoder architecture in Vision-Language Pre-training?", "answer": ["LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "VisualBERT: A Simple and Performant Baseline for Vision and Language", "VinVL: Revisiting Visual Representations in Vision-Language Models", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision"], "answer_arxiv_id": ["1908.07490", "1908.02265", "1908.08530", "1908.03557", "2101.00529", "2108.10904"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_11202"} +{"question": "Who proposed the Slot Attention method for annotation-free object segmentation?", "answer": ["Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["2006.15055"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_11203"} +{"question": "What research has been done on the relationship between the decision-estimation coefficient and the information ratio?", "answer": ["Learning to Optimize Via Information-Directed Sampling", "Information Directed Sampling and Bandits with Heteroscedastic Noise", "Information Directed Sampling for Linear Partial Monitoring", "Asymptotically Optimal Information-Directed Sampling", "Linear Partial Monitoring for Sequential Decision Making Algorithms, Regret Bounds and Applications"], "answer_arxiv_id": ["1403.5556", "1801.09667", "2002.11182", "2011.05944", "2302.03683"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_11204"} +{"question": "Which study applied the amortized formulation to model the conditional SBP for images and state-space models?", "answer": ["Conditional Simulation Using Diffusion Schrödinger Bridges"], "answer_arxiv_id": ["2202.13460"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_11205"} +{"question": "Could you provide me some papers in synthetic data application within the financial and healthcare domain?", "answer": ["Synthetic Data, Real Errors: How (Not) to Publish and Use Synthetic Data", "GT-GAN: General Purpose Time Series Synthesis with Generative Adversarial Networks", "Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs", "C-RNN-GAN: Continuous recurrent neural networks with adversarial training", "Learning to simulate realistic limit order book markets from data as a World Agent", "Towards Realistic Market Simulations: a Generative Adversarial Networks Approach"], "answer_arxiv_id": ["2305.09235", "2210.02040", "1706.02633", "1611.09904", "2210.09897", "2110.13287"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_11206"} +{"question": "Which papers focus on using feature importance in Explainable AI?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "SmoothGrad: removing noise by adding noise", "A Unified Approach to Interpreting Model Predictions", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1602.04938", "1706.03825", "1705.07874", "1703.01365"], "source_meta": {"published_time": "20230325"}, "qid": "AutoScholarQuery_train_11207"} +{"question": "Are there any references that covers advanced techniques of data augmentation?", "answer": ["RandAugment: Practical automated data augmentation with a reduced search\n space", "AutoAugment: Learning Augmentation Policies from Data", "mixup: Beyond Empirical Risk Minimization", "CutMix: Regularization Strategy to Train Strong Classifiers with\n Localizable Features", "Random Erasing Data Augmentation"], "answer_arxiv_id": ["1909.13719", "1805.09501", "1710.09412", "1905.04899", "1708.04896"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_11208"} +{"question": "What works reproduce Whisper-style training using public data and open-source toolkits?", "answer": ["Reproducing Whisper-Style Training Using an Open-Source Toolkit and\n Publicly Available Data"], "answer_arxiv_id": ["2309.13876"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_11209"} +{"question": "What research attempts to unite different modeling approaches through unified frameworks such as prefix modeling, permutation modeling, causal masked modeling or unified language learning?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "XLNet: Generalized Autoregressive Pretraining for Language Understanding", "CM3: A Causal Masked Multimodal Model of the Internet", "UL2: Unifying Language Learning Paradigms"], "answer_arxiv_id": ["1910.10683", "1906.08237", "2201.07520", "2205.05131"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_11210"} +{"question": "Which study showed that a bound is achievable with noisy feedback when Lt is a one-dimensional squared loss?", "answer": ["Online Forecasting of Total-Variation-bounded Sequences"], "answer_arxiv_id": ["1906.03364"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_11211"} +{"question": "What are the references related to the utilization of Gaussian prior in variational autoencoders (VAEs)?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_11212"} +{"question": "Which works introduced a dataset for few-shot tasks in DVS datasets?", "answer": ["N-Omniglot, a large-scale neuromorphic dataset for spatio-temporal sparse few-shot learning"], "answer_arxiv_id": ["2112.13230v3"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_11213"} +{"question": "What works discuss the need for rotation invariance in object pose estimation descriptors?", "answer": ["RIGA: Rotation-Invariant and Globally-Aware Descriptors for Point Cloud\n Registration", "You Only Hypothesize Once: Point Cloud Registration with\n Rotation-equivariant Descriptors", "Rotation-Invariant Transformer for Point Cloud Matching"], "answer_arxiv_id": ["2209.13252", "2109.00182", "2303.08231"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_11214"} +{"question": "Which papers have discussed the underspecified nature of many datasets?", "answer": ["Underspecification Presents Challenges for Credibility in Modern Machine Learning", "Hidden Stratification Causes Clinically Meaningful Failures in Machine Learning for Medical Imaging"], "answer_arxiv_id": ["2011.03395", "1909.12475"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_11215"} +{"question": "Which works attempt to control the underlying factors of disentangled representations by maximizing the mutual information between the images and the latent representations?", "answer": ["InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets"], "answer_arxiv_id": ["1606.03657"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_11216"} +{"question": "Could you name some works that used genetic algorithms for molecular graphs generation?", "answer": ["Augmenting Genetic Algorithms with Deep Neural Networks for Exploring the Chemical Space", "Guiding Deep Molecular Optimization with Genetic Exploration"], "answer_arxiv_id": ["1909.11655", "2007.04897"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_11217"} +{"question": "What research works discuss disentangled representation learning in the context of multi-view learning?", "answer": ["Representation Learning: A Review and New Perspectives", "Towards a Smaller Student: Capacity Dynamic Distillation for Efficient\n Image Retrieval", "Co-advise: Cross Inductive Bias Distillation", "Disentangled Representation Learning"], "answer_arxiv_id": ["1206.5538", "2303.09230", "2106.12378", "2211.11695"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_11218"} +{"question": "Which study proposed a lightweight Querying Transformer to leverage pre-trained image encoder and language model for multimodal tasks?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2301.12597"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_11219"} +{"question": "Which papers have worked on recombination of data, which is similar to lexinvariant language models?", "answer": ["Data Recombination for Neural Semantic Parsing", "Learning to Recombine and Resample Data for Compositional Generalization"], "answer_arxiv_id": ["1606.03622", "2010.03706"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11220"} +{"question": "Could you tell me the research where a numerical integrator has been designed for RMHMC using geodesics?", "answer": ["Geodesic Monte Carlo on Embedded Manifolds"], "answer_arxiv_id": ["1301.6064"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_11221"} +{"question": "Which publications studied 3D constraints via volume rendering following the introduction of Neural Radiance Fields (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_11222"} +{"question": "Could you provide me with some literature about private training of deep ResNets on ImageNet?", "answer": ["Toward Training at ImageNet Scale with Differential Privacy"], "answer_arxiv_id": ["2201.12328"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_11223"} +{"question": "Which studies utilized a UNet for fast or even real-time rendering of point-based models?", "answer": ["Neural Point-Based Graphics", "NPBG++: Accelerating Neural Point-Based Graphics"], "answer_arxiv_id": ["1906.08240", "2203.13318"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_11224"} +{"question": "Which papers have proposed using GANs for video generation?", "answer": ["A Good Image Generator Is What You Need for High-Resolution Video Synthesis", "Temporal Generative Adversarial Nets with Singular Value Clipping", "MoCoGAN: Decomposing Motion and Content for Video Generation", "StyleVideoGAN: A Temporal Generative Model using a Pretrained StyleGAN", "StyleGAN-V: A Continuous Video Generator with the Price, Image Quality and Perks of StyleGAN2", "Generating Videos with Dynamics-aware Implicit Generative Adversarial Networks", "Generating Videos with Scene Dynamics", "Train Sparsely, Generate Densely: Memory-efficient Unsupervised Training of High-resolution Temporal GAN", "Adversarial Video Generation on Complex Datasets", "Latent Image Animator: Learning to animate images via latent space navigation", "Towards Smooth Video Composition", "Generating Long Videos of Dynamic Scenes"], "answer_arxiv_id": ["2104.15069", "1611.06624", "1707.04993", "2107.07224", "2112.14683", "2202.10571", "1609.02612", "1811.09245", "1907.06571", "2203.09043", "2212.07413", "2206.03429"], "source_meta": {"published_time": "20230829"}, "qid": "AutoScholarQuery_train_11225"} +{"question": "In which work was the ability of the DINO model to perform intra- and inter-image correspondence without the need for dense labels demonstrated?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.14294"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11226"} +{"question": "Which works consider more abstract representations in the computational sketching?", "answer": ["A Neural Representation of Sketch Drawings", "Creative Sketch Generation", "DoodleFormer: Creative Sketch Drawing with Transformers", "CLIPDraw: Exploring Text-to-Drawing Synthesis through Language-Image Encoders", "SketchLattice: Latticed Representation for Sketch Manipulation", "CLIPasso: Semantically-Aware Object Sketching"], "answer_arxiv_id": ["1704.03477v4", "2011.10039", "2112.03258", "2106.14843", "2108.11636", "2202.05822"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_11227"} +{"question": "Could you provide me some studies about weak dependency notions when moving beyond linear time-series models?", "answer": ["Ergodic Mirror Descent", "On Empirical Risk Minimization with Dependent and Heavy-Tailed Data"], "answer_arxiv_id": ["1105.4681", "2109.02224"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_11228"} +{"question": "Could you provide some works introduced transformers in end-to-end multi-person pose estimation?", "answer": ["Explicit Box Detection Unifies End-to-End Multi-Person Pose Estimation"], "answer_arxiv_id": ["2302.01593"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_11229"} +{"question": "What are some methods that used a frozen pre-trained image encoder for CLIP pre-training?", "answer": ["LiT: Zero-Shot Transfer with Locked-image text Tuning", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2111.07991", "2103.00020"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11230"} +{"question": "What papers discussed the use of point clouds in 3D representation?", "answer": ["A Point Set Generation Network for 3D Object Reconstruction from a\n Single Image", "Learning Representations and Generative Models for 3D Point Clouds", "Point Transformer"], "answer_arxiv_id": ["1612.00603", "1707.02392", "2012.09164"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_11231"} +{"question": "Are there any efficient variants of the SAM algorithm and could you provide the references for them?", "answer": ["Efficient Sharpness-aware Minimization for Improved Training of Neural Networks", "Towards Efficient and Scalable Sharpness-Aware Minimization"], "answer_arxiv_id": ["2110.03141", "2203.02714"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_11232"} +{"question": "Could you mention some research papers that aimed to extend the context window of LMs by modifying the attention mechanism?", "answer": ["Efficient Streaming Language Models with Attention Sinks", "Unlimiformer: Long-Range Transformers with Unlimited Length Input"], "answer_arxiv_id": ["2309.17453", "2305.01625"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_11233"} +{"question": "Could you provide me some papers where the loss function is made to encourage similar or positive pairs to be drawn together in the embedding space?", "answer": ["Improved Baselines with Momentum Contrastive Learning"], "answer_arxiv_id": ["2003.04297"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_11234"} +{"question": "Which study tackled the irregularity and unordered nature of point cloud data with multi-layer perceptrons (MLP) and max-pooling layer?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation"], "answer_arxiv_id": ["1612.00593"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_11235"} +{"question": "What works attempted to generalize flow matching to Riemannian manifolds?", "answer": ["Flow Matching for Generative Modeling", "Riemannian Flow Matching on General Geometries"], "answer_arxiv_id": ["2210.02747", "2302.03660"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_11236"} +{"question": "Which researches have unified the various versions and diverging approaches to PEFT?", "answer": ["Towards a Unified View of Parameter-Efficient Transfer Learning", "UniPELT: A Unified Framework for Parameter-Efficient Language Model Tuning"], "answer_arxiv_id": ["2110.04366", "2110.07577"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_11237"} +{"question": "Are there any works using a generator in data-free knowledge distillation methods?", "answer": ["Data-Free Adversarial Distillation", "Robust and Resource-Efficient Data-Free Knowledge Distillation by Generative Pseudo Replay"], "answer_arxiv_id": ["1912.11006", "2201.03019"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_11238"} +{"question": "Could you provide me some works that utilize transformers for each modality in Visual-Language models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_11239"} +{"question": "Could you provide me some studies about reducing the gradient error in BNNs?", "answer": ["Bi-Real Net: Enhancing the Performance of 1-bit CNNs With Improved\n Representational Capability and Advanced Training Algorithm", "Forward and Backward Information Retention for Accurate Binary Neural\n Networks", "Learning Frequency Domain Approximation for Binary Neural Networks", "Estimator Meets Equilibrium Perspective: A Rectified Straight Through\n Estimator for Binary Neural Networks Training"], "answer_arxiv_id": ["1808.00278", "1909.10788", "2103.00841", "2308.06689"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_11240"} +{"question": "What are the works that designed ad-hoc solutions for learning models performing equally well in- and out-of-distribution?", "answer": ["Size-Invariant Graph Representations for Graph Classification Extrapolations", "Learning Causally Invariant Representations for Out-of-Distribution Generalization on Graphs", "OOD Link Prediction Generalization Capabilities of Message-Passing GNNs in Larger Test Graphs"], "answer_arxiv_id": ["2103.05045", "2202.05441", "2205.15117"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_11241"} +{"question": "Which work introduced a zero-shot category-level 6D pose estimation task with a self-supervised semantic correspondence learning method?", "answer": ["Zero-Shot Category-Level Object Pose Estimation"], "answer_arxiv_id": ["2204.03635"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_11242"} +{"question": "Which paper introduces DDIM sampling that uses alternative non-Markovian formulation?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_11243"} +{"question": "Which work introduced and solved the adversarial Lipschitz bandit?", "answer": ["Adaptive Discretization for Adversarial Lipschitz Bandits"], "answer_arxiv_id": ["2006.12367v3"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11244"} +{"question": "Which works focus on proposing novel architectures for decoders in the context of Latent space Bayesian optimization?", "answer": ["Grammar Variational Autoencoder", "Junction Tree Variational Autoencoder for Molecular Graph Generation", "NeVAE: A Deep Generative Model for Molecular Graphs∗", "Molecular Hypergraph Grammar with Its Application to Molecular Optimization", "Syntax-Directed Variational Autoencoder for Structured Data"], "answer_arxiv_id": ["1703.01925", "1802.04364", "1802.05283", "1809.02745", "1802.08786"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_11245"} +{"question": "Can you name the work that developed a framework to discover GVF questions with a question network and learn the cumulants?", "answer": ["Discovery of Useful Questions as Auxiliary Tasks"], "answer_arxiv_id": ["1909.04607"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_11246"} +{"question": "Could you provide the study that achieved state-of-the-art results in differentiable point-based methods for real-time rendering of unbounded scenes?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_11247"} +{"question": "What research proposed a multi-step acceleration scheme integrated into the stochastic mirror-prox to enhance the convergence rate?", "answer": ["Accelerated Schemes For A Class of Variational Inequalities"], "answer_arxiv_id": ["1403.4164"], "source_meta": {"published_time": "20220910"}, "qid": "AutoScholarQuery_train_11248"} +{"question": "What research applies a pre-trained text-to-image Diffusion Model for text-driven video editing?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Zero-Shot Video Editing Using Off-The-Shelf Image Diffusion Models", "Pix2Video: Video Editing using Image Diffusion", "Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "TokenFlow: Consistent Diffusion Features for Consistent Video Editing"], "answer_arxiv_id": ["2212.11565", "2303.17599", "2303.12688", "2306.07954", "2303.09535", "2307.10373"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_11249"} +{"question": "What studies focused on exploring notions of reproducibility/replicability in various computational fields?", "answer": ["Replicable Bandits", "Reproducibility in Optimization: Theoretical Framework and Limits"], "answer_arxiv_id": ["2210.01898", "2202.04598"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_11250"} +{"question": "Can you name some research that has studied randomized smoothing in robotic planning and control tasks?", "answer": ["Do Differentiable Simulators Give Better Policy Gradients?", "Bundled Gradients through Contact via Randomized Smoothing", "Leveraging Randomized Smoothing for Optimal Control of Nonsmooth Dynamical Systems", "Global Planning for Contact-Rich Manipulation via Local Smoothing of Quasi-dynamic Contact Models"], "answer_arxiv_id": ["2202.00817", "2109.05143", "2203.03986", "2206.10787"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_11251"} +{"question": "Any works about solving poses using differential bundle adjustment(BA) for end-to-end Structure-from-Motion methods?", "answer": ["BA-Net: Dense Bundle Adjustment Network"], "answer_arxiv_id": ["1806.04807"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_11252"} +{"question": "Which papers have applied bootstrapping in sequence generation?", "answer": ["Revisiting Self-Training for Neural Sequence Generation"], "answer_arxiv_id": ["1909.13788"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_11253"} +{"question": "What works explored the inherent tradeoff between robust and standard accuracy?", "answer": ["Robustness May Be at Odds with Accuracy", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1805.12152", "1901.08573"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_11254"} +{"question": "Which works demonstrate the benefits of mixed-precision training for large language models?", "answer": ["Efficient Large-Scale Language Model Training on GPU Clusters Using\n Megatron-LM", "ZeRO: Memory Optimizations Toward Training Trillion Parameter Models"], "answer_arxiv_id": ["2104.04473", "1910.02054"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_11255"} +{"question": "Which studies apply ANN search for retrieval in document retrieval system?", "answer": ["Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval", "Neural Ranking Models with Weak Supervision"], "answer_arxiv_id": ["2007.00808", "1704.08803"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_11256"} +{"question": "Which prior works prioritize uncertain data to improve uniform sample selection?", "answer": ["Prioritized Experience Replay", "Distributed Prioritized Experience Replay", "Large Batch Experience Replay"], "answer_arxiv_id": ["1511.05952", "1803.00933", "2110.01528"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_11257"} +{"question": "Are there any studies that improved the softmax by normalizing the facial features and adding margins?", "answer": ["SphereFace: Deep Hypersphere Embedding for Face Recognition", "NormFace: L2 Hypersphere Embedding for Face Verification", "CosFace: Large Margin Cosine Loss for Deep Face Recognition", "ArcFace: Additive Angular Margin Loss for Deep Face Recognition", "SphereFace Revived: Unifying Hyperspherical Face Recognition"], "answer_arxiv_id": ["1704.08063", "1704.06369", "1801.09414", "1801.07698", "2109.05565"], "source_meta": {"published_time": "20231104"}, "qid": "AutoScholarQuery_train_11258"} +{"question": "Could you name some studies where preference-based learning with deep neural networks has been explored in the field of Natural Language Processing and Computer Vision?", "answer": ["Deep Reinforcement Learning from Human Preferences", "PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-training", "Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["1706.03741", "2106.05091", "1909.08593"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_11259"} +{"question": "Could you provide me some works that criticized the studies on infinitely wide neural networks?", "answer": ["On Lazy Training in Differentiable Programming"], "answer_arxiv_id": ["1812.07956"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_11260"} +{"question": "Are there any research studies about LLMs completing various conversational tasks such as recommendation, tutoring, and counseling", "answer": ["Large Language Models as Zero-Shot Conversational Recommenders", "Recommender AI Agent: Integrating Large Language Models for Interactive\n Recommendations", "EduChat: A Large-Scale Language Model-based Chatbot System for\n Intelligent Education", "Plug-and-Play Policy Planner for Large Language Model Powered Dialogue\n Agents", "Building Emotional Support Chatbots in the Era of LLMs"], "answer_arxiv_id": ["2308.10053", "2308.16505", "2308.02773", "2311.00262", "2308.11584"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_11261"} +{"question": "Are there studies about procedure planning using multimodal alignment in the context of instructional videos?", "answer": ["Aligning Step-by-Step Instructional Diagrams to Video Demonstrations"], "answer_arxiv_id": ["2303.13800"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_11262"} +{"question": "Which papers introduced transformer-based architectures to vision domains?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Training data-efficient image transformers & distillation through attention", "Incorporating Convolution Designs into Visual Transformers", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2010.11929", "2012.12877", "2103.11816", "2103.14030"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_11263"} +{"question": "Which research papers focus on multilingual language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Multilingual Denoising Pre-training for Neural Machine Translation", "Unsupervised Cross-lingual Representation Learning at Scale", "mT5: A massively multilingual pre-trained text-to-text transformer", "Few-shot Learning with Multilingual Language Models", "mGPT: Few-Shot Learners Go Multilingual", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "PolyLM: An Open Source Polyglot Large Language Model"], "answer_arxiv_id": ["1810.04805", "2001.08210", "1911.02116", "2010.11934", "2112.10668", "2204.07580", "2211.05100", "2307.06018"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_11264"} +{"question": "Which studies have projected the source person and target person into a unified 3D space to capture pixel-level correspondences?", "answer": ["Dense Pose Transfer", "Coordinate-based Texture Inpainting for Pose-Guided Human Image Generation", "Few-Shot Human Motion Transfer by Personalized Geometry and Texture Modeling", "Textured Neural Avatars", "Liquid Warping GAN: A Unified Framework for Human Motion Imitation, Appearance Transfer and Novel View Synthesis", "Disentangled Person Image Generation"], "answer_arxiv_id": ["1809.01995", "1811.11459", "2103.14338", "1905.08776", "1909.12224", "1712.02621"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_11265"} +{"question": "Could you provide me some studies which solved model heterogeneity in Federated Learning using partial training-based methods?", "answer": ["FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout", "HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients"], "answer_arxiv_id": ["2102.13451", "2010.01264"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_11266"} +{"question": "Which works proposed adversarial training strategies for multi-exit networks?", "answer": ["A Panda? No, It’s a Sloth: Slowdown Attacks on Adaptive Multi-Exit Neural Network Inference", "Triple Wins: Boosting Accuracy, Robustness and Efficiency Together by Enabling Input-Adaptive Inference"], "answer_arxiv_id": ["2010.02432", "2002.10025"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_11267"} +{"question": "What research has been done on using Pseudo-labeling or self-training as a semi-supervised learning technique?", "answer": ["Semi-Supervised Learning with Ladder Networks", "Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "Unsupervised Data Augmentation for Consistency Training", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samples"], "answer_arxiv_id": ["1507.02672", "1703.01780", "1904.12848", "2001.07685", "2104.13963"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11268"} +{"question": "Which paper utilizes particle-based optimization directly in the function space?", "answer": ["Function Space Particle Optimization for Bayesian Neural Networks"], "answer_arxiv_id": ["1902.09754"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_11269"} +{"question": "What research focuses on learning data dependencies beyond the given graph structure using linear time transformers?", "answer": ["NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification", "DIFFormer: Scalable (Graph) Transformers Induced by Energy Constrained Diffusion"], "answer_arxiv_id": ["2306.08385", "2301.09474v4"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_11270"} +{"question": "Which paper describes a pioneering work that harnesses prior information from diffusion models for better image fidelity?", "answer": ["Exploiting Diffusion Prior for Real-World Image Super-Resolution"], "answer_arxiv_id": ["2305.07015"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_11271"} +{"question": "What papers proposed lower complexity bounds for optimization over static graphs?", "answer": ["Optimal algorithms for smooth and strongly convex distributed optimization in networks"], "answer_arxiv_id": ["1702.08704"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_11272"} +{"question": "Could you provide me some works that involves the learning and manipulation of prototypes?", "answer": ["Interpretable and Steerable Sequence Learning via Prototypes"], "answer_arxiv_id": ["1907.09728"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_11273"} +{"question": "What research developed multiple soft prompts for cross-domain adaption?", "answer": ["Domain Adaptation via Prompt Learning"], "answer_arxiv_id": ["2202.06687"], "source_meta": {"published_time": "20220407"}, "qid": "AutoScholarQuery_train_11274"} +{"question": "Which papers propose methods for iteratively learning doubly-stochastic matrices?", "answer": ["Ranking via Sinkhorn Propagation"], "answer_arxiv_id": ["1106.1925"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_11275"} +{"question": "Can you provide any references that discuss the saturation effect for the spectral regularized algorithms?", "answer": ["Optimal Convergence for Distributed Learning with Stochastic Gradient Methods and Spectral Algorithms"], "answer_arxiv_id": ["1801.07226"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_11276"} +{"question": "Which study first constructed the audio-visual segmentation benchmark and proposed a baseline method?", "answer": ["Audio-Visual Segmentation with Semantics"], "answer_arxiv_id": ["2301.13190"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11277"} +{"question": "Could you name studies related to the training of hierarchical classifiers?", "answer": ["Making Better Mistakes: Leveraging Class Hierarchies with Deep Networks", "Your \"Flamingo\" is My \"Bird\": Fine-Grained, or Not", "Learning Hierarchy Aware Features for Reducing Mistake Severity"], "answer_arxiv_id": ["1912.09393", "2011.09040", "2207.12646"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_11278"} +{"question": "What studies focus on reconstructing a single 3D scene given dense observations?", "answer": ["Geometry-Aware Recurrent Neural Networks for Active Visual Recognition", "Learning Spatial Common Sense with Geometry-Aware Recurrent Networks", "DeepVoxels: Learning Persistent 3D Feature Embeddings", "Neural Volumes: Learning Dynamic Renderable Volumes from Images", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Multiview Neural Surface Reconstruction by Disentangling Geometry and Appearance", "Advances in Neural Rendering"], "answer_arxiv_id": ["1811.01292v2", "1901.00003", "1812.01024", "1906.07751", "2003.08934", "2003.09852", "2111.05849"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_11279"} +{"question": "What work is cited concerning the use of a pre-trained encoder in the training of latent diffusion models?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11280"} +{"question": "What papers set the ground for contrastive and multiview learning algorithms?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance Discrimination", "Learning Representations by Maximizing Mutual Information Across Views", "Self-Supervised Learning of Pretext-Invariant Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1805.01978", "1906.00910", "1912.01991", "1911.05722", "2002.05709"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_11281"} +{"question": "What researches mainly focus on improving accuracy in video classification?", "answer": ["Two-Stream Convolutional Networks for Action Recognition in Videos", "Temporal Segment Networks: Towards Good Practices for Deep Action Recognition", "Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset", "Learning Spatio-Temporal Features with 3D Residual Networks for Action Recognition", "Rethinking Spatiotemporal Feature Learning: Speed-Accuracy Trade-offs in Video Classification", "Is Space-Time Attention All You Need for Video Understanding?", "MViTv2: Improved Multiscale Vision Transformers for Classification and Detection"], "answer_arxiv_id": ["1406.2199", "1608.00859", "1705.07750", "1708.07632", "1712.04851", "2102.05095", "2112.01526"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_11282"} +{"question": "What studies focus on implementing contrastive learning in the field of representation learning methods?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance Discrimination", "Representation Learning with Contrastive Predictive Coding", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Exploring Cross-Image Pixel Contrast for Semantic Segmentation"], "answer_arxiv_id": ["1805.01978", "1807.03748", "1911.05722", "2002.05709", "2101.11939"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_11283"} +{"question": "Which work is related to program of thought prompting?", "answer": ["Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks"], "answer_arxiv_id": ["2211.12588"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_11284"} +{"question": "Which studies discussed algortihms for label propagation?", "answer": ["Multi-View Matrix Completion for Multi-Label Image Classification", "Label Propagation for Deep Semi-supervised Learning", "Low-shot learning with large-scale diffusion"], "answer_arxiv_id": ["1904.03901", "1904.04717", "1706.02332v3"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_11285"} +{"question": "Which papers originated the learning of long-range dependencies with RNNs from Ising and Hopfield networks?", "answer": ["Hopfield Networks is All You Need"], "answer_arxiv_id": ["2008.02217"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_11286"} +{"question": "What papers have used meta-learning-based methods for domain generalization?", "answer": ["Learning to Generalize: Meta-Learning for Domain Generalization", "Domain Generalization via Model-Agnostic Learning of Semantic Features", "Exploiting Domain-Specific Features to Enhance Domain Generalization", "Hierarchical Variational Memory for Few-shot Learning Across Domains"], "answer_arxiv_id": ["1710.03463", "1910.13580", "2110.09410", "2112.08181"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_11287"} +{"question": "What papers have developed expressive statistical models to represent details beyond major human body?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "SUPR: A Sparse Unified Part-Based Human Representation", "Collaborative Regression of Expressive Bodies using Moderation"], "answer_arxiv_id": ["1904.05866", "2210.13861", "2105.05301"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_11288"} +{"question": "Could you provide me with works that introduced attention-based models in visual question answering?", "answer": ["Hierarchical Question-Image Co-Attention for Visual Question Answering", "Dual Attention Networks for Multimodal Reasoning and Matching", "Stacked Attention Networks for Image Question Answering"], "answer_arxiv_id": ["1606.00061", "1611.00471", "1511.02274"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_11289"} +{"question": "What are the works that have discussed the major pretraining objective for image-language and video-language models after the introduction of CLIP?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks", "Florence: A New Foundation Model for Computer Vision", "VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding", "CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "MERLOT: Multimodal Neural Script Knowledge Models", "OmniVL: One Foundation Model for Image-Language and Video-Language Tasks", "CLIP-ViP: Adapting Pre-trained Image-Text Model to Video-Language Representation Alignment", "InternVideo: General Video Foundation Models via Generative and Discriminative Learning"], "answer_arxiv_id": ["2103.00020", "2201.12086", "2301.12597", "2108.10904", "2208.10442", "2111.11432", "2109.14084", "2106.11097", "2106.02636", "2209.07526", "2209.06430", "2212.03191"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_11290"} +{"question": "Which papers discussed the generative methods in the pretraining paradigm?", "answer": ["GPT-GNN: Generative Pre-Training of Graph Neural Networks", "Motif-based Graph Self-Supervised Learning for Molecular Property Prediction"], "answer_arxiv_id": ["2006.15437", "2110.00987"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_11291"} +{"question": "Are there any works that use a pseudo-likelihood variational framework for node representation learning especially for TAGs?", "answer": ["GMNN: Graph Markov Neural Networks"], "answer_arxiv_id": ["1905.06214"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_11292"} +{"question": "What kind of data does the YouTube-360 dataset provide according to the cited study?", "answer": ["Learning Representations from Audio-Visual Spatial Alignment"], "answer_arxiv_id": ["2011.01819"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11293"} +{"question": "What papers utilized Mask Transformers in semantic segmentation?", "answer": ["Segmenter: Transformer for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2105.05633", "2112.01527"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_11294"} +{"question": "Which works propose methods for Zero-Shot Segmentation?", "answer": ["One-Shot Learning for Semantic Segmentation", "Zero-Shot Semantic Segmentation", "Context-aware Feature Generation for Zero-shot Semantic Segmentation"], "answer_arxiv_id": ["1709.03410", "1906.00817", "2008.06893"], "source_meta": {"published_time": "20230930"}, "qid": "AutoScholarQuery_train_11295"} +{"question": "Which work shows that several data distribution properties can drive in-context learning ability?", "answer": ["Data Distributional Properties Drive Emergent In-Context Learning in\n Transformers"], "answer_arxiv_id": ["2205.05055"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_11296"} +{"question": "What studies used ranking to aggregate answers in NLP tasks?", "answer": ["Ranking Paragraphs for Improving Answer Recall in Open-Domain Question Answering"], "answer_arxiv_id": ["1810.00494"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_11297"} +{"question": "Could you cite some researches that demonstrated the potential of model scaling?", "answer": ["Scaling Laws for Neural Language Models"], "answer_arxiv_id": ["2001.08361"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_11298"} +{"question": "Which work proposed the STORM algorithm that contributed to the development of variance-reduced Adam?", "answer": ["Momentum-Based Variance Reduction in Non-Convex SGD"], "answer_arxiv_id": ["1905.10018"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_11299"} +{"question": "Which studies applied feature perspective transformation to project all view features into a shared plane in the context of multi-view object detection?", "answer": ["Multiview Detection with Feature Perspective Transformation", "Multiview Detection with Shadow Transformer (and View-Coherent Data\n Augmentation)"], "answer_arxiv_id": ["2007.07247", "2108.05888"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_11300"} +{"question": "Which papers introduce diffusion-based audio editing methods?", "answer": ["Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models", "AudioLDM: Text-to-Audio Generation with Latent Diffusion Models"], "answer_arxiv_id": ["2301.12661", "2301.12503"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_11301"} +{"question": "What works utilize large pre-trained models for embodied AI tasks?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Grounded Language-Image Pre-training", "UnifiedQA-v2: Stronger Generalization via Broader Cross-Format Training", "Simple but Effective: CLIP Embeddings for Embodied AI", "CoWs on Pasture: Baselines and Benchmarks for Language-Driven Zero-Shot Object Navigation", "Open-vocabulary Queryable Scene Representations for Real World Planning", "Inner Monologue: Embodied Reasoning through Planning with Language Models", "Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents", "A Persistent Spatial Semantic Representation for High-level Natural Language Instruction Execution", "How Much Can CLIP Benefit Vision-and-Language Tasks?"], "answer_arxiv_id": ["2103.00020", "2112.03857", "2202.12359v1", "2111.09888", "2203.10421", "2209.09874", "2207.05608", "2201.07207", "2107.05612", "2107.06383"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11302"} +{"question": "Which paper shared important properties about ℐ-essential graph?", "answer": ["Characterization and Greedy Learning of Interventional Markov Equivalence Classes of Directed Acyclic Graphs"], "answer_arxiv_id": ["1104.2808"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11303"} +{"question": "Which paper introduced the concept of Automatic Prompt Optimization (APO)?", "answer": ["Automatic Prompt Optimization with \"Gradient Descent\" and Beam Search"], "answer_arxiv_id": ["2305.03495"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_11304"} +{"question": "Could you indicate research papers where the mesh topology of FLAME was finetuned for face animation?", "answer": ["Neural Head Avatars from Monocular RGB Videos", "Realistic One-shot Mesh-based Head Avatars"], "answer_arxiv_id": ["2112.01554", "2206.08343"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_11305"} +{"question": "In which work the author discusses the concept of the output vector field of an SBM being conservative?", "answer": ["Conservativeness of untied auto-encoders"], "answer_arxiv_id": ["1506.07643"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_11306"} +{"question": "Could you provide me some works about feature attribution methodologies used in generating post hoc explanations?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "SmoothGrad: removing noise by adding noise", "Layer-wise Relevance Propagation for Neural Networks with Local Renormalization Layers"], "answer_arxiv_id": ["1312.6034", "1706.03825", "1604.00825"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_11307"} +{"question": "What studies focused on fine-tuning diffusion models on various conditioning signals?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Object-Centric Slot Diffusion"], "answer_arxiv_id": ["2302.05543", "2302.08453", "2208.12242", "2208.01618", "2211.09794", "2210.09276", "2303.10834"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_11308"} +{"question": "Could you provide me some studies about the use of segmentation priors to improve the performance of SLAM systems?", "answer": ["vMAP: Vectorised Object Mapping for Neural Field SLAM", "Neural Implicit Dense Semantic SLAM"], "answer_arxiv_id": ["2302.01838v2", "2304.14560"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_11309"} +{"question": "Which works propose regenerating the faces’ features by using autoencoders?", "answer": ["Semi-Adversarial Networks: Convolutional Autoencoders for Imparting\n Privacy to Face Images"], "answer_arxiv_id": ["1712.00321"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_11310"} +{"question": "Could you provide me some works that considered shuffle-DP?", "answer": ["Amplification by Shuffling: From Local to Central Differential Privacy via Anonymity", "Distributed Differential Privacy via Shuffling", "Separating Local & Shuffled Differential Privacy via Histograms", "Hiding Among the Clones: A Simple and Nearly Optimal Analysis of Privacy Amplification by Shuffling", "On the Power of Multiple Anonymous Messages", "Private Counting from Anonymous Messages: Near-Optimal Accuracy with Vanishing Communication Overhead"], "answer_arxiv_id": ["1811.12469", "1808.01394", "1911.06879", "2012.12803", "1908.11358", "2106.04247"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_11311"} +{"question": "Which works mention the A-OKVQA, a dataset demanding broad knowledge for accurate responses?", "answer": ["A-OKVQA: A Benchmark for Visual Question Answering using World Knowledge", "OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge"], "answer_arxiv_id": ["2206.01718", "1906.00067"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_11312"} +{"question": "Can you point me to the works that have been done for synthesizing images from semantic segmentation maps?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "Semantic Image Synthesis with Spatially-Adaptive Normalization", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "1903.07291", "2302.08453"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_11313"} +{"question": "What is the research that began the exploration of distribution learning with the utilization of both public and private data?", "answer": ["Private Estimation with Public Data"], "answer_arxiv_id": ["2208.07984v2"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_11314"} +{"question": "Which research proposes the use of Adam-type step sizes in bi-level optimization?", "answer": ["BiAdam: Fast Adaptive Bilevel Optimization Methods"], "answer_arxiv_id": ["2106.11396"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11315"} +{"question": "Can you name some studies that show a similar behavior in the landscapes of neural networks trained in supervised learning tasks as found in the study of return landscapes?", "answer": ["Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Essentially No Barriers in Neural Network Energy Landscape", "Topology and Geometry of Half-Rectified Network Optimization", "Linear Mode Connectivity and the Lottery Ticket Hypothesis", "Qualitatively Characterizing Neural Network Optimization Problems"], "answer_arxiv_id": ["1802.10026", "1803.00885", "1611.01540", "1912.05671", "1412.6544"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_11316"} +{"question": "What is the work that revisits the need for explicit regularization in supervised scene flow learning?", "answer": ["Neural Scene Flow Prior"], "answer_arxiv_id": ["2111.01253"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_11317"} +{"question": "Which studies proposed incorporating graph structural information into Transformer architecture by extracting the positional embedding from graph structure?", "answer": ["A Generalization of Transformer Networks to Graphs", "Rethinking Graph Transformers with Spectral Attention"], "answer_arxiv_id": ["2012.09699", "2106.03893"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_11318"} +{"question": "Which works revitalized the Convolutional Neural Networks with deformable convolution?", "answer": ["InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions"], "answer_arxiv_id": ["2211.05778"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_11319"} +{"question": "Which papers contributed to the rapid advance of implicit representations in recent research?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "NeRF++: Analyzing and Improving Neural Radiance Fields"], "answer_arxiv_id": ["1812.03828", "2003.08934", "2010.07492"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11320"} +{"question": "Could you provide me some works about diffusion-based SR methods?", "answer": ["Image Super-Resolution via Iterative Refinement", "DiffIR: Efficient Diffusion Model for Image Restoration", "Scaling Up to Excellence: Practicing Model Scaling for Photo-Realistic\n Image Restoration In the Wild", "Image Restoration with Mean-Reverting Stochastic Differential Equations", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model"], "answer_arxiv_id": ["2104.07636", "2303.09472", "2401.13627", "2301.11699", "2108.02938", "2201.11793", "2212.00490"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_11321"} +{"question": "Which papers establish finite-sample bounds for TD-learning with linear function approximation?", "answer": ["A Finite Time Analysis of Temporal Difference Learning With Linear Function Approximation", "Finite-Time Error Bounds For Linear Stochastic Approximation and TD Learning"], "answer_arxiv_id": ["1806.02450", "1902.00923"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_11322"} +{"question": "What research proposed ATOMIC, a graph of if-then inferences that models social commonsense in daily life events?", "answer": ["ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning"], "answer_arxiv_id": ["1811.00146"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_11323"} +{"question": "What works focus on the existence of adversarial examples and their transferability?", "answer": ["Universal Adversarial Attack on Attention and the Resulting Dataset DAmageNet"], "answer_arxiv_id": ["2001.06325"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_11324"} +{"question": "Any studies about the refinement of directly regressed depth maps by iterative refinement under a local linear model with learned affinity?", "answer": ["Learning Affinity via Spatial Propagation Networks", "Learning Depth with Convolutional Spatial Propagation Network", "CSPN++: Learning Context and Resource Aware Convolutional Spatial\n Propagation Networks for Depth Completion", "Non-Local Spatial Propagation Network for Depth Completion", "Dynamic Spatial Propagation Network for Depth Completion", "GraphCSPN: Geometry-Aware Depth Completion via Dynamic GCNs", "LRRU: Long-short Range Recurrent Updating Networks for Depth Completion"], "answer_arxiv_id": ["1710.01020", "1810.02695", "1911.05377", "2007.10042", "2202.09769", "2210.10758", "2310.08956v1"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_11325"} +{"question": "What papers provide approaches that impose a regularization on model parameters or isolate task-specific parameters to retain previous knowledge?", "answer": ["Progress & Compress: A scalable framework for continual learning", "Continual Learning Through Synaptic Intelligence", "Memory Aware Synapses: Learning what (not) to forget", "PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning", "Overcoming Catastrophic Forgetting with Hard Attention to the Task"], "answer_arxiv_id": ["1805.06370", "1703.04200", "1711.09601", "1711.05769", "1801.01423"], "source_meta": {"published_time": "20230409"}, "qid": "AutoScholarQuery_train_11326"} +{"question": "What studies showed that neural networks typically emphasize intended features over shortcut ones that are spuriously correlated with the classification targets?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_11327"} +{"question": "What works propose sliding window-based VSR techniques?", "answer": ["Video Enhancement with Task-Oriented Flow", "Real-Time Video Super-Resolution with Spatio-Temporal Networks and\n Motion Compensation", "MuCAN: Multi-Correspondence Aggregation Network for Video\n Super-Resolution", "Temporal Modulation Network for Controllable Space-Time Video\n Super-Resolution"], "answer_arxiv_id": ["1711.09078", "1611.05250", "2007.11803", "2104.10642"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_11328"} +{"question": "Could you list the research works that studied compositionality in various contexts such as image generation, video synthesis, etc.?", "answer": ["Learning Canonical Representations for Scene Graph to Image Generation", "Compositional Video Prediction", "Compositional Video Synthesis with Action Graphs", "Generating Videos of Zero-Shot Compositions of Actions and Objects", "CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning", "Compositional Semantic Parsing with Large Language Models", "A Good Prompt Is Worth Millions of Parameters: Low-resource Prompt-based Learning for Vision-Language Models", "Interactive Language Learning by Question Answering"], "answer_arxiv_id": ["1912.07414", "1908.08522", "2006.15327", "1912.02401", "1612.06890", "2209.15003", "2110.08484", "1908.10909"], "source_meta": {"published_time": "20220407"}, "qid": "AutoScholarQuery_train_11329"} +{"question": "Which research papers have discussed rotation invariance through canonicalization of the local frame?", "answer": ["LIFT: Learned Invariant Feature Transform", "Repeatability Is Not Enough: Learning Affine Regions via\n Discriminability", "Self-Supervised Learning of Image Scale and Orientation", "Self-Supervised Equivariant Learning for Oriented Keypoint Detection"], "answer_arxiv_id": ["1603.09114", "1711.06704", "2206.07259", "2204.08613"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_11330"} +{"question": "Which works made additional methodological improvements to the GNN in SEAL?", "answer": ["Labeling Trick: A Theory of Using Graph Neural Networks for Multi-Node Representation Learning", "Nested Graph Neural Networks"], "answer_arxiv_id": ["2010.16103", "2110.13197"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_11331"} +{"question": "Which work proposed generating neural textures using 2D generative models?", "answer": ["StylePeople: A Generative Model of Fullbody Human Avatars"], "answer_arxiv_id": ["2104.08363"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_11332"} +{"question": "What papers propose methods that involve aligning the 3D pixels or points of object proposals in 3D-3D correspondence?", "answer": ["ZeroPose: CAD-Model-based Zero-Shot Pose Estimation"], "answer_arxiv_id": ["2305.17934"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_11333"} +{"question": "Could you provide me some studies on Metric-based methods for few-shot learning?", "answer": ["Matching Networks for One Shot Learning", "Prototypical Networks for Few-shot Learning", "Learning to Compare: Relation Network for Few-Shot Learning", "TapNet: Neural Network Augmented with Task-Adaptive Projection for Few-Shot Learning"], "answer_arxiv_id": ["1606.04080", "1703.05175", "1711.06025", "1905.06549"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_11334"} +{"question": "Are there any studies on segmentation in multi-body systems?", "answer": ["Deep Part Induction from Articulated Object Pairs", "4D Unsupervised Object Discovery", "OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point Clouds", "Dynamic 3D Scene Analysis by Point Cloud Accumulation", "MultiBodySync: Multi-Body Segmentation and Motion Estimation via 3D Scan Synchronization", "Self-Supervised Learning of Lidar Segmentation for Autonomous Indoor Navigation", "4DContrast: Contrastive Learning with Dynamic Correspondences for 3D Scene Understanding"], "answer_arxiv_id": ["1809.07417", "2210.04801", "2210.04458", "2207.12394v1", "2101.06605v3", "2012.05897", "2112.02990"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11335"} +{"question": "What studies propose a method where only a small number of layers are stored in the forward phase of 'Gradient Checkpointing'?", "answer": ["Training Deep Nets with Sublinear Memory Cost"], "answer_arxiv_id": ["1604.06174"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_11336"} +{"question": "Which papers deal with the idea of computing gradients via implicit differentiation instead of the classic forward pass layer-by-layer approach?", "answer": ["Q"], "answer_arxiv_id": ["1611.08152"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_11337"} +{"question": "Could you provide me some studies explaining how pre-trained language models can handle ambiguity and learn reasoning rules implicitly?", "answer": ["Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge", "ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language"], "answer_arxiv_id": ["1803.05457", "2012.13048"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_11338"} +{"question": "Are there any studies on high-dimensional mixture models?", "answer": ["Finite-sample Analysis of Interpolating Linear Classifiers in the Overparameterized Regime", "Binary Classification of Gaussian Mixtures: Abundance of Support Vectors, Benign Overfitting and Regularization", "Risk Bounds for Over-parameterized Maximum Margin Classification on Sub-Gaussian Mixtures", "Benign Overfitting without Linearity: Neural Network Classifiers Trained by Gradient Descent for Noisy Linear Data"], "answer_arxiv_id": ["2004.12019", "2011.09148", "2104.13628", "2202.05928"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_11339"} +{"question": "Could you mention some works on verifying robustness that provide bounds on the robustness of specific classifiers?", "answer": ["A Unified View of Piecewise Linear Neural Network Verification", "Evaluating Robustness of Neural Networks with Mixed Integer Programming", "On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models", "SoK: Certified Robustness for Deep Neural Networks"], "answer_arxiv_id": ["1711.00455", "1711.07356", "1810.12715", "2009.04131v9"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_11340"} +{"question": "What papers proposed reconstructing dynamic scenes by conditioning an implicit representation on time?", "answer": ["Neural Radiance Flow for 4D View Synthesis and Video Processing", "Dynamic View Synthesis from Dynamic Monocular Video", "Space-time Neural Irradiance Fields for Free-Viewpoint Video"], "answer_arxiv_id": ["2012.09790", "2105.06468", "2011.12950"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_11341"} +{"question": "Which study demonstrated the 'Edge of Stability' phenomenon in the context of optimizing neural network sharpness?", "answer": ["Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability"], "answer_arxiv_id": ["2103.00065"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_11342"} +{"question": "Which works focused on referring expression comprehension methods?", "answer": ["TransVG: End-to-End Visual Grounding with Transformers", "Co-Grounding Networks with Semantic Attention for Referring Expression Comprehension in Videos", "Referring Transformer: A One-step Approach to Multi-task Visual Grounding", "PolyFormer: Referring Image Segmentation as Sequential Polygon Generation", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework", "Modeling Relationships in Referential Expressions with Compositional Modular Networks", "Modeling Context in Referring Expressions", "A Real-time Global Inference Network for One-stage Referring Expression Comprehension", "ReCLIP: A Strong Zero-Shot Baseline for Referring Expression Comprehension", "Detector-Free Weakly Supervised Grounding by Separation", "SeqTR: A Simple yet Universal Network for Visual Grounding", "UNITER: UNiversal Image-TExt Representation Learning", "Unified-IO: A unified model for vision, language, and multi-modal tasks"], "answer_arxiv_id": ["2104.08541", "2103.12346", "2106.03089", "2302.07387", "2202.03052", "1611.09978", "1608.00272", "1912.03478", "2204.05991", "2104.09829", "2203.16265", "1909.11740", "2206.08916"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_11343"} +{"question": "Which studies have motivated the use of structured directions in finite-difference methods?", "answer": ["Structured Evolution with Compact Architectures for Scalable Policy Optimization"], "answer_arxiv_id": ["1804.02395v2"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11344"} +{"question": "What work does the researcher refer to when comparing the novelties of their work?", "answer": ["Taxonomizing local versus global structure in neural network loss landscapes"], "answer_arxiv_id": ["2107.11228"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_11345"} +{"question": "What works proposed attention-based models to address issues with RNNs in deep point processes?", "answer": ["Transformer Hawkes Process", "Self-Attentive Hawkes Process"], "answer_arxiv_id": ["2002.09291", "1907.07561"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_11346"} +{"question": "What literature addresses the influence of the adaptation of other agents on an optimal agent's adaptations?", "answer": ["Multiagent Cooperation and Competition with Deep Reinforcement Learning"], "answer_arxiv_id": ["1511.08779"], "source_meta": {"published_time": "20230722"}, "qid": "AutoScholarQuery_train_11347"} +{"question": "Which works discuss query-based methods for robustness evaluation in NLP?", "answer": ["BERT-ATTACK: Adversarial Attack Against BERT Using BERT", "Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment", "BAE: BERT-based Adversarial Examples for Text Classification", "TextBugger: Generating Adversarial Text Against Real-world Applications", "Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Models", "Searching for an Effective Defender: Benchmarking Defense against Adversarial Word Substitution"], "answer_arxiv_id": ["2004.09984", "1907.11932", "2004.01970", "1812.05271", "2111.02840", "2108.12777"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_11348"} +{"question": "What applications have employed multilingual pre-training techniques for ancient languages like Ancient Greek or Latin?", "answer": ["Latin BERT: A Contextual Language Model for Classical Philology", "Unsupervised Cross-lingual Representation Learning at Scale", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model"], "answer_arxiv_id": ["2009.10053", "1911.02116", "2211.05100"], "source_meta": {"published_time": "20240808"}, "qid": "AutoScholarQuery_train_11349"} +{"question": "What studies demonstrate the effectiveness of learning visual representations from the supervision of corresponding text?", "answer": ["Learning Visual Features from Large Weakly Supervised Data", "Learning Visual N-Grams from Web Data", "VirTex: Learning Visual Representations from Textual Annotations", "Multimodal Contrastive Training for Visual Representation Learning"], "answer_arxiv_id": ["1511.02251", "1612.09161v2", "2006.06666", "2104.12836"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11350"} +{"question": "What studies investigated the prototype calorimeter for the International Large Detector (ILD)?", "answer": ["Getting High: High Fidelity Simulation of High Granularity Calorimeters with High Speed", "Decoding Photons: Physics in the Latent Space of a BIB-AE Generative Network"], "answer_arxiv_id": ["2005.05334", "2102.12491"], "source_meta": {"published_time": "20220210"}, "qid": "AutoScholarQuery_train_11351"} +{"question": "Could you provide me some works showing which properties of a minimum leads to better generalization in GD?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "Sharp Minima Can Generalize For Deep Nets", "Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data", "Normalized Flat Minima: Exploring Scale-Invariant Definition of Flat Minima for Neural Networks using PAC-Bayesian Analysis"], "answer_arxiv_id": ["1609.04836", "1703.04933", "1703.11008", "1901.04653"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_11352"} +{"question": "What works propose various optimization techniques such as stochastic weight averaging and gradient regularizer to find flatter minima?", "answer": ["Averaging Weights Leads to Wider Optima and Better Generalization", "Implicit Gradient Regularization"], "answer_arxiv_id": ["1803.05407", "2009.11162"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11353"} +{"question": "What papers applied machine-learning methods as alternative energy predictors followed by optimization steps in crystal structure prediction?", "answer": ["Accelerating crystal structure prediction by machine-learning interatomic potentials with active learning"], "answer_arxiv_id": ["1802.07605"], "source_meta": {"published_time": "20230730"}, "qid": "AutoScholarQuery_train_11354"} +{"question": "Could you point to some papers that followed up on the topic initiated by using market specific Bayesian causal structure learning?", "answer": ["Learning Bayesian Networks: The Combination of Knowledge and Statistical Data", "Causal Discovery from a Mixture of Experimental and Observational Data", "Being Bayesian about Network Structure"], "answer_arxiv_id": ["1302.6815", "1301.6686v1", "1301.3856v1"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11355"} +{"question": "Which works have implemented advanced sampling techniques to enhance diffusion models?", "answer": ["Denoising Diffusion Implicit Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2010.02502", "2206.00927", "2206.00364"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_11356"} +{"question": "Which study is most similar to the authors' work in considering uncertainty in inputs in stable matching?", "answer": ["Stable Matching with Uncertain Linear Preferences"], "answer_arxiv_id": ["1607.02917v1"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_11357"} +{"question": "Which works have analyzed the average sensitivity for maximum matching problems?", "answer": ["Sensitivity Analysis of the Maximum Matching Problem"], "answer_arxiv_id": ["2009.04556"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_11358"} +{"question": "What research work utilized automasking to exclude pixels static in sequential frames to overcome erroneous depth learning in self-supervised monocular DE pipelines?", "answer": ["Digging Into Self-Supervised Monocular Depth Estimation"], "answer_arxiv_id": ["1806.01260"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_11359"} +{"question": "Which works introduced traditional backdoor attacks that modify a fraction of training samples?", "answer": ["Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning"], "answer_arxiv_id": ["1712.05526v1"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_11360"} +{"question": "Could you provide me some research about improving model processing speed using dynamic or static token pruning?", "answer": ["Token Merging: Your ViT But Faster", "Scalable Vision Transformers with Hierarchical Pooling", "Patch Slimming for Efficient Vision Transformers", "PoWER-BERT: Accelerating BERT Inference via Progressive Word-vector Elimination", "IA-RED2: Interpretability-Aware Redundancy Reduction for Vision Transformers"], "answer_arxiv_id": ["2210.09461", "2103.10619", "2106.02852", "2001.08950", "2106.12620"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_11361"} +{"question": "Could you provide any studies that focused on differentiating through linear programs for differentiable optimization?", "answer": ["Melding the Data-Decisions Pipeline: Decision-Focused Learning for Combinatorial Optimization"], "answer_arxiv_id": ["1809.05504"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_11362"} +{"question": "Which works have tackled the issue of missing objects, incorrect spatial relations, and incorrect attributes in text-to-image generation (T2I) models?", "answer": ["Aligning Text-to-Image Models using Human Feedback", "HRS-Bench: Holistic, Reliable and Scalable Benchmark for Text-to-Image Models"], "answer_arxiv_id": ["2302.12192", "2304.05390"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11363"} +{"question": "What works provided comprehensive surveys of language and learning model (LLM) evaluation?", "answer": ["A Survey on Evaluation of Large Language Models", "Evaluating Large Language Models: A Comprehensive Survey", "Holistic Evaluation of Language Models"], "answer_arxiv_id": ["2307.03109", "2310.19736", "2211.09110"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_11364"} +{"question": "What works have been done about designing approximate execution frameworks for real-time perception?", "answer": ["ApproxDet: Content and Contention-Aware Approximate Object Detection for Mobiles"], "answer_arxiv_id": ["2010.10754"], "source_meta": {"published_time": "20210610"}, "qid": "AutoScholarQuery_train_11365"} +{"question": "Are there any works that attempt to learn Deformable NeRFs to model dynamic content?", "answer": ["Nerfies: Deformable Neural Radiance Fields", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Video", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "Text-To-4D Dynamic Scene Generation", "H-NeRF: Neural Radiance Fields for Rendering and Temporal Reconstruction of Humans in Motion", "Neural Articulated Radiance Field", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular Video", "HumanNeRF: Efficiently Generated Human Radiance Field from Sparse Inputs", "HumanRF: High-Fidelity Neural Radiance Fields for Humans in Motion", "LatentAvatar: Learning Latent Expression Code for Expressive Neural Head Avatar", "PoseVocab: Learning Joint-structured Pose Embeddings for Human Avatar Modeling"], "answer_arxiv_id": ["2011.12948", "2011.13961", "2012.12247", "2106.13228", "2301.11280", "2110.13746", "2104.03110", "2201.04127", "2112.02789", "2305.06356", "2305.01190", "2304.13006"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11366"} +{"question": "Which works discussed the problem of tokenization when trying to scale up to a large number of languages?", "answer": ["Unsupervised Cross-lingual Representation Learning at Scale", "How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models"], "answer_arxiv_id": ["1911.02116", "2012.15613"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_11367"} +{"question": "Which papers are there on continual learning that utilized a replay buffer for memory-based methods?", "answer": ["Orthogonal Gradient Descent for Continual Learning"], "answer_arxiv_id": ["1910.07104v1"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_11368"} +{"question": "What research is closest to the presenter's work in terms of using a temporal consistency constraint in the latent space for continuous control tasks?", "answer": ["Temporal Difference Learning for Model Predictive Control"], "answer_arxiv_id": ["2203.04955"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11369"} +{"question": "What studies considered non-Markov processes in the context of diffusion models?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_11370"} +{"question": "What papers examined the application of neural scaling laws to Mixture of Experts (MoE) models?", "answer": ["Unified Scaling Laws for Routed Language Models"], "answer_arxiv_id": ["2202.01169"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_11371"} +{"question": "Could you provide some works that focus on object-level contrastive learning?", "answer": ["Unsupervised Semantic Segmentation by Contrasting Object Mask Proposals", "Unsupervised Semantic Segmentation with Self-supervised Object-centric\n Representations", "CASTing Your Model: Learning to Localize Improves Self-Supervised\n Representations", "Efficient Visual Pretraining with Contrastive Detection", "Self-Supervised Visual Representation Learning from Hierarchical\n Grouping", "Self-Supervised Visual Representation Learning with Semantic Grouping", "Bridging the Gap to Real-World Object-Centric Learning"], "answer_arxiv_id": ["2102.06191", "2207.05027", "2012.04630", "2103.10957", "2012.03044", "2205.15288", "2209.14860"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_11372"} +{"question": "What studies are about predicting the endpoint of heart failure and recommending prescriptions based on diagnoses?", "answer": ["Change Matters: Medication Change Prediction with Recurrent Residual Networks", "SafeDrug: Dual Molecular Graph Encoders for Recommending Effective and Safe Drug Combinations", "GAMENet: Graph Augmented MEmory Networks for Recommending Medication Combination"], "answer_arxiv_id": ["2105.01876", "2105.02711", "1809.01852"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_11373"} +{"question": "Which works discuss on improving the conventional pre-training paragidm in the context of language pre-training?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Forging Multiple Training Objectives for Pre-trained Language Models via Meta-Learning"], "answer_arxiv_id": ["1810.04805", "2210.10293"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_11374"} +{"question": "What papers revealed the difficulty of systematic generalization for standard seq2seq models in semantic parsing, machine translation and algorithmic reasoning?", "answer": ["Improving Text-to-SQL Evaluation Methodology", "On Compositional Generalization of Neural Machine Translation", "Neural Networks and the Chomsky Hierarchy"], "answer_arxiv_id": ["1806.09029", "2105.14802", "2207.02098"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_11375"} +{"question": "What are the issues faced with the WIT benchmark?", "answer": ["LiT: Zero-Shot Transfer with Locked-image text Tuning"], "answer_arxiv_id": ["2111.07991"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_11376"} +{"question": "Which studies found latent knowledge in the internal representations of language models?", "answer": ["Discovering Latent Knowledge in Language Models Without Supervision"], "answer_arxiv_id": ["2212.03827"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_11377"} +{"question": "Which papers are about gradient-based methods in meta-learning?", "answer": ["Recasting Gradient-Based Meta-Learning as Hierarchical Bayes", "On First-Order Meta-Learning Algorithms", "Learning to Generalize: Meta-Learning for Domain Generalization"], "answer_arxiv_id": ["1801.08930v1", "1803.02999", "1710.03463"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_11378"} +{"question": "What model was proposed to decrease training and inference costs by applying the diffusion model in the latent space?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11379"} +{"question": "Are there any studies that discussed about the approaches being too purpose-specific and non-transferable across domains in time series modeling?", "answer": ["A Comprehensive Survey on Transfer Learning"], "answer_arxiv_id": ["1911.02685"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11380"} +{"question": "Which studies have improved upon Frozen's approach by introducing adapters, enhancing visual encoders or finetuning on instructions?", "answer": ["MAGMA – Multimodal Augmentation of Generative Models through Adapter-based Finetuning", "Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models", "Visual Instruction Tuning"], "answer_arxiv_id": ["2112.05253", "2204.14198", "2301.12597", "2304.08485"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_11381"} +{"question": "What research works are about utilizing natural language explanations for commonsense tasks?", "answer": ["Explain Yourself! Leveraging Language Models for Commonsense Reasoning"], "answer_arxiv_id": ["1906.02361"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_11382"} +{"question": "What papers focus on enhancing RAG model robustness during the training of the generation model?", "answer": ["Making Retrieval-Augmented Language Models Robust to Irrelevant Context", "The Power of Noise: Redefining Retrieval for RAG Systems"], "answer_arxiv_id": ["2310.01558", "2401.14887"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_11383"} +{"question": "What papers discussed techniques for entropy coding applied in compression of neural fields representation?", "answer": ["3D Scene Compression through Entropy Penalized Neural Representation\n Functions", "Scalable Model Compression by Entropy Penalized Reparameterization"], "answer_arxiv_id": ["2104.12456", "1906.06624"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_11384"} +{"question": "Which works discuss variance-reduced version of Adam by combining Adam and SVRG?", "answer": ["Divergence Results and Convergence of a Variance Reduced Version of ADAM"], "answer_arxiv_id": ["2210.05607v1"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_11385"} +{"question": "Could you tell me about the researches that provide hardness results in the setting of statistical query (SQ) algorithms?", "answer": ["Hardness of Noise-Free Learning for Two-Hidden-Layer Neural Networks", "Near-Optimal SQ Lower Bounds for Agnostically Learning Halfspaces and ReLUs under Gaussian Marginals", "Statistical-Query Lower Bounds via Functional Gradients", "Statistical Queries and Statistical Algorithms: Foundations and Applications", "On the Complexity of Learning Neural Networks"], "answer_arxiv_id": ["2202.05258", "2006.16200", "2006.15812", "2004.00557v2", "1707.04615"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_11386"} +{"question": "What works proposed adaptive schemes for the frequency of averaging in Local SGD?", "answer": ["Adaptive Communication Strategies to Achieve the Best Error-Runtime Trade-off in Local-Update SGD", "Local SGD with Periodic Averaging: Tighter Analysis and Adaptive Synchronization"], "answer_arxiv_id": ["1810.08313v2", "1910.13598"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_11387"} +{"question": "Which papers have focused on the Frequency Principle in terms of the representation capacity of a DNN?", "answer": ["Training behavior of deep neural network in frequency domain", "On the Spectral Bias of Neural Networks", "Frequency Principle: Fourier Analysis Sheds Light on Deep Neural Networks", "Theory of the Frequency Principle for General Deep Neural Networks"], "answer_arxiv_id": ["1807.01251v6", "1806.08734", "1901.06523", "1906.09235"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_11388"} +{"question": "What are some examples of works that proposed facial gesture co-speech animation models?", "answer": ["FaceFormer: Speech-Driven 3D Facial Animation with Transformers", "CodeTalker: Speech-Driven 3D Facial Animation with Discrete Motion Prior"], "answer_arxiv_id": ["2112.05329", "2301.02379"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_11389"} +{"question": "Which works have adopted retrieval processes to facilitate document annotation within model-generated outputs?", "answer": ["Enabling Large Language Models to Generate Text with Citations"], "answer_arxiv_id": ["2305.14627"], "source_meta": {"published_time": "20240225"}, "qid": "AutoScholarQuery_train_11390"} +{"question": "Which paper demonstrated superior novel view synthesis quality over traditional methods in neural rendering?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_11391"} +{"question": "Could you provide me some studies about decoder-focused methods in multi-task learning for computer vision tasks?", "answer": ["PAD-Net: Multi-Tasks Guided Prediction-and-Distillation Network for\n Simultaneous Depth Estimation and Scene Parsing", "MTI-Net: Multi-Scale Task Interaction Networks for Multi-Task Learning", "Exploring Relational Context for Multi-Task Dense Prediction", "TaskExpert: Dynamically Assembling Multi-Task Representations with\n Memorial Mixture-of-Experts", "Pattern-Affinitive Propagation across Depth, Surface Normal and Semantic\n Segmentation"], "answer_arxiv_id": ["1805.04409", "2001.06902", "2104.13874", "2307.15324", "1906.03525"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_11392"} +{"question": "Could you provide me any comprehensive overview of face identification task that our study primarily focuses?", "answer": ["Deep Face Recognition: A Survey"], "answer_arxiv_id": ["1804.06655"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_11393"} +{"question": "Which studies propose removing contextual cues as a form of data augmentation that can improve the model performance?", "answer": ["GridMask Data Augmentation", "Improved Regularization of Convolutional Neural Networks with Cutout", "Hide-and-Seek: A Data Augmentation Technique for Weakly-Supervised\n Localization and Beyond", "Random Erasing Data Augmentation"], "answer_arxiv_id": ["2001.04086", "1708.04552", "1811.02545", "1708.04896"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_11394"} +{"question": "Who proposed unlearning methods with theoretical guarantees?", "answer": ["Certified Data Removal from Machine Learning Models", "Approximate Data Deletion from Machine Learning Models", "Remember What You Want to Forget: Algorithms for Machine Unlearning"], "answer_arxiv_id": ["1911.03030", "2002.10077", "2103.03279v2"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_11395"} +{"question": "Which works use data augmentation for positive sampling in contrastive learning?", "answer": ["Learning Representations by Maximizing Mutual Information Across Views", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1906.00910", "2002.05709"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11396"} +{"question": "What studies are associated with use of autoregressive rollout in model-based planning for complex domains?", "answer": ["Recurrent Environment Simulators"], "answer_arxiv_id": ["1704.02254"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_11397"} +{"question": "Are there any existing benchmarks for graph learning?", "answer": ["OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphs", "Graph Robustness Benchmark: Benchmarking the Adversarial Robustness of Graph Machine Learning", "GOOD: A Graph Out-of-Distribution Benchmark", "NAS-Bench-Graph: Benchmarking Graph Neural Architecture Search"], "answer_arxiv_id": ["2103.09430", "2111.04314", "2206.08452", "2206.09166"], "source_meta": {"published_time": "20220616"}, "qid": "AutoScholarQuery_train_11398"} +{"question": "What studies work on linear mixture MDPs?", "answer": ["Sample Complexity of Reinforcement Learning using Linearly Combined Model Ensembles", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Model-Based Reinforcement Learning with Value-Targeted Regression", "Provably Efficient Reinforcement Learning for Discounted MDPs with Feature Mapping", "Provably Efficient Exploration in Policy Optimization", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Improved Regret Analysis for Variance-Adaptive Linear Bandits and Horizon-Free Linear Mixture MDPs", "Learning Stochastic Shortest Path with Linear Function Approximation", "Horizon-Free Reinforcement Learning in Polynomial Time: the Power of Stationary Policies", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs"], "answer_arxiv_id": ["1910.10597", "2006.01107", "2006.01107", "2006.13165", "1912.05830", "2012.08507", "2111.03289", "2110.12727", "2203.12922v2", "2205.11507"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_11399"} +{"question": "What papers employ a weakly supervised training combined with test-time fine tuning for stationary material acquisition?", "answer": ["Generative Modelling of BRDF Textures from Flash Images"], "answer_arxiv_id": ["2102.11861"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_11400"} +{"question": "Which articles analyzed causal abstraction in language models?", "answer": ["Causal Abstractions of Neural Networks", "Finding Alignments Between Interpretable Causal Variables and\n Distributed Neural Representations"], "answer_arxiv_id": ["2106.02997", "2303.02536"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_11401"} +{"question": "Which papers focused on improving L2O generalization ability for similar optimization tasks but longer training iterations?", "answer": ["Training Stronger Baselines for Learning to Optimize", "Learning Gradient Descent: Better Generalization and Longer Horizons"], "answer_arxiv_id": ["2010.09089", "1703.03633"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_11402"} +{"question": "Which works investigate spatially-varying lighting estimation by predicting per-pixel spherical lobes or individual environment maps in the input image?", "answer": ["Deep Parametric Indoor Lighting Estimation", "IRISformer: Dense Vision Transformers for Single-Image Inverse Rendering\n in Indoor Scenes", "Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying\n Lighting and SVBRDF from a Single Image", "OpenRooms: An End-to-End Open Framework for Photorealistic Indoor Scene\n Datasets", "Physically-Based Editing of Indoor Scene Lighting from a Single Image", "Fast Spatially-Varying Indoor Lighting Estimation"], "answer_arxiv_id": ["1910.08812", "2206.08423", "1905.02722", "2007.12868", "2205.09343", "1906.03799"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_11403"} +{"question": "What works focus on the iterative learning (ITR) approach for simplifying PBT?", "answer": ["Novel Policy Seeking with Constrained Optimization", "Continuously Discovering Novel Strategies via Reward-Switching Policy Optimization"], "answer_arxiv_id": ["2005.10696", "2204.02246"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_11404"} +{"question": "Could you provide me with literature about improving in-context learning via task formulation?", "answer": ["Surface Form Competition: Why the Highest Probability Answer Isn’t Always Right", "Calibrate Before Use: Improving Few-Shot Performance of Language Models", "Noisy Channel Language Model Prompting for Few-Shot Text Classification"], "answer_arxiv_id": ["2104.08315", "2102.09690", "2108.04106"], "source_meta": {"published_time": "20220905"}, "qid": "AutoScholarQuery_train_11405"} +{"question": "What works train an additional encoder from the unlabeled data with self-supervised learning for ensemble?", "answer": ["SVL-Adapter: Self-Supervised Adapter for Vision-Language Pretrained Models"], "answer_arxiv_id": ["2210.03794"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_11406"} +{"question": "What works apply Graph Neural Networks in traffic forecasting?", "answer": ["Frigate: Frugal Spatio-temporal Forecasting on Road Networks"], "answer_arxiv_id": ["2306.08277"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_11407"} +{"question": "What are some early work on text embeddings?", "answer": ["Efficient Estimation of Word Representations in Vector Space"], "answer_arxiv_id": ["1301.3781"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_11408"} +{"question": "Could you provide me some works about the use of proper scoring rules such as likelihoods and Brier scores to evaluate calibration?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Better Uncertainty Calibration via Proper Scores for Classification and Beyond"], "answer_arxiv_id": ["1612.01474", "2203.07835"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_11409"} +{"question": "Can you specify the work that first formulated adversarial corruption in the context of multi-armed bandit problems?", "answer": ["Stochastic bandits robust to adversarial corruptions"], "answer_arxiv_id": ["1803.09353"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_11410"} +{"question": "What papers have incorporated trigger reversion as a white-box defense method?", "answer": ["Trigger Hunting with a Topological Prior for Trojan Detection"], "answer_arxiv_id": ["2110.08335"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_11411"} +{"question": "What research papers discuss the use of model-based offline RL for dynamics model learning?", "answer": ["MOPO: Model-based Offline Policy Optimization", "COMBO: Conservative Offline Model-Based Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2005.13239", "2102.08363", "2005.05951"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_11412"} +{"question": "Which papers discuss structured latent MDPs and learning short-memory policies as beneficial aspects for sample-efficient learning in POMDPs?", "answer": ["RL for Latent MDPs: Regret Guarantees and a Lower Bound", "Provably Efficient Reinforcement Learning in Partially Observable Dynamical Systems"], "answer_arxiv_id": ["2102.04939", "2206.12020"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_11413"} +{"question": "Could you provide me some research papers that introduce novel approaches for DNN feature selection?", "answer": ["DeepPINK: reproducible feature selection in deep neural networks"], "answer_arxiv_id": ["1809.01185"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_11414"} +{"question": "Could you name some studies that relaxed the unidirectional limitation of autoregressive models using bidirectional models?", "answer": ["DeCoAR 2.0: Deep Contextualized Acoustic Representations with Vector Quantization", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations", "HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units", "w2v-BERT: Combining Contrastive Learning and Masked Language Modeling for Self-Supervised Speech Pre-Training", "WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing"], "answer_arxiv_id": ["2012.06659", "2006.11477", "2106.07447", "2108.06209", "2110.13900"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_11415"} +{"question": "What works have proposed dynamic neural scene flow methods?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2011.12948"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_11416"} +{"question": "Could you provide a study that uses knowledge distillation to aggregate local task models in MTL?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_11417"} +{"question": "Are there any studies that have investigated LLMs to enhance ASR output by error correction?", "answer": ["N-best T5: Robust ASR Error Correction using Multiple Input Hypotheses\n and Constrained Decoding Space", "HyPoradise: An Open Baseline for Generative Speech Recognition with\n Large Language Models", "Generative Speech Recognition Error Correction with Large Language\n Models and Task-Activating Prompting"], "answer_arxiv_id": ["2303.00456", "2309.15701", "2309.15649"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_11418"} +{"question": "What is the reference for LoRA which adopts trainable low-rank decomposition matrices into LLMs' layers?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_11419"} +{"question": "Which work exploits sinusoidal functions of varying frequencies to for input encoding in NeRF?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_11420"} +{"question": "Which papers have contributed to the foundation of template-mesh-based PCA models?", "answer": ["High-Fidelity 3D Digital Human Head Creation from RGB-D Selfies", "HACK: Learning a Parametric Head and Neck Model for High-fidelity\n Animation"], "answer_arxiv_id": ["2010.05562", "2305.04469"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11421"} +{"question": "Could you show me papers using reward models to train LLMs in the context of RALMs?", "answer": ["Teaching language models to support answers with verified quotes", "WebGPT: Browser-assisted question-answering with human feedback", "Language Agent Tree Search Unifies Reasoning Acting and Planning in\n Language Models"], "answer_arxiv_id": ["2203.11147", "2112.09332", "2310.04406"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_11422"} +{"question": "What is the study that makes any loss function symmetric through loss normalization?", "answer": ["Normalized Loss Functions for Deep Learning with Noisy Labels"], "answer_arxiv_id": ["2006.13554"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_11423"} +{"question": "What is the cited work related to the application of Transformer in Audio Classification?", "answer": ["AST: Audio Spectrogram Transformer"], "answer_arxiv_id": ["2104.01778"], "source_meta": {"published_time": "20220117"}, "qid": "AutoScholarQuery_train_11424"} +{"question": "Which papers study how to perform image edits using textual instructions?", "answer": ["Zero-shot Image-to-Image Translation", "Prompt-to-Prompt Image Editing with Cross Attention Control", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2302.03027", "2208.01626", "2211.09800"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_11425"} +{"question": "Could you provide me some studies about modifying or replacing the last linear classification layer for test-data adaptation?", "answer": ["Test-Time Adaptation via Self-Training with Nearest Neighbor Information", "Learning to Generalize across Domains on Single Test Samples"], "answer_arxiv_id": ["2207.10792", "2202.08045"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_11426"} +{"question": "Could you provide me studies about Neural Collapse with Mean Squared Error loss?", "answer": ["Extended Unconstrained Features Model for Exploring Deep Neural Collapse", "On the Optimization Landscape of Neural Collapse under MSE Loss: Global Optimality with Unconstrained Features", "Revealing the Structure of Deep Neural Networks via Convex Duality"], "answer_arxiv_id": ["2202.08087", "2203.01238", "2002.09773"], "source_meta": {"published_time": "20230101"}, "qid": "AutoScholarQuery_train_11427"} +{"question": "Who provided a more comprehensive dataset in the field of GFIQA?", "answer": ["Going the Extra Mile in Face Image Quality Assessment: A Novel Database\n and Model"], "answer_arxiv_id": ["2207.04904"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_11428"} +{"question": "Which paper adopted the Transformer architecture for lane detection?", "answer": ["End-to-end Lane Shape Prediction with Transformers"], "answer_arxiv_id": ["2011.04233"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_11429"} +{"question": "Which works discuss masked language modeling methods in pre-trained Transformer-based large language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Unsupervised Cross-lingual Representation Learning at Scale"], "answer_arxiv_id": ["1810.04805", "1911.02116"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_11430"} +{"question": "Which studies have made attempts to introduce extra conditions as inputs in the text-to-image generation process?", "answer": ["ReCo: Region-Controlled Text-to-Image Generation", "Gligen: Open-Set Grounded Text-to-Image Generation", "Training-Free Layout Control with Cross-Attention Guidance"], "answer_arxiv_id": ["2211.15518", "2301.07093", "2304.03373"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11431"} +{"question": "Could you provide me some studies illustrating the use of 2D diffusion models for novel-view synthesis, 3D generation, and stylistic 3D editing?", "answer": ["Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "MVDream: Multi-view Diffusion for 3D Generation", "Single-Stage Diffusion NeRF: A Unified Approach to 3D Generation and\n Reconstruction", "DreamBooth3D: Subject-Driven Text-to-3D Generation", "Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions", "Vox-E: Text-guided Voxel Editing of 3D Objects", "DreamEditor: Text-Driven 3D Scene Editing with Neural Fields", "Edit-DiffNeRF: Editing 3D Neural Radiance Fields using 2D Diffusion\n Model", "Zero-1-to-3: Zero-shot One Image to 3D Object", "SceneScape: Text-Driven Consistent Scene Generation", "Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models", "TextMesh: Generation of Realistic 3D Meshes From Text Prompts", "Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D\n Generation", "IT3D: Improved Text-to-3D Generation with Explicit View Synthesis", "InstructP2P: Learning to Edit 3D Point Clouds with Text Instructions", "DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2212.00774v1", "2308.16512", "2304.06714", "2303.13508", "2303.12789", "2303.12048", "2306.13455", "2306.09551", "2303.11328", "2302.01133", "2303.11989", "2304.12439", "2303.07937", "2308.11473", "2306.07154", "2209.14988"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_11432"} +{"question": "Which papers provided studies on Sharpness Aware Minimization (SAM) optimizing properties and convergence results for non-convex objectives?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization", "Towards Understanding Sharpness-Aware Minimization"], "answer_arxiv_id": ["2010.01412", "2206.06232"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_11433"} +{"question": "Which works have explored incorporating spatial and stylistic control while generating images from text using diffusion models?", "answer": ["GLIGEN: Open-Set Grounded Text-to-Image Generation", "MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing", "Diffusion Self-Guidance for Controllable Image Generation", "Grounded Text-to-Image Synthesis with Attention Refocusing", "A-STAR: Test-time Attention Segregation and Retention for Text-to-image\n Synthesis", "LLM-grounded Diffusion: Enhancing Prompt Understanding of Text-to-Image\n Diffusion Models with Large Language Models"], "answer_arxiv_id": ["2301.07093", "2304.08465v1", "2306.00986", "2306.05427v2", "2306.14544", "2305.13655"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_11434"} +{"question": "Could you provide me with some studies about anytime inference?", "answer": ["Multi-Scale Dense Networks for Resource Efficient Image Classification", "Anytime Inference with Distilled Hierarchical Neural Ensembles"], "answer_arxiv_id": ["1703.09844", "2003.01474"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_11435"} +{"question": "Could you provide some works on mitigation-based defenses?", "answer": ["Baseline Defenses for Adversarial Attacks Against Aligned Language\n Models", "RAIN: Your Language Models Can Align Themselves without Finetuning", "Jailbreak and Guard Aligned Language Models with Only Few In-Context\n Demonstrations", "Defending Large Language Models Against Jailbreaking Attacks Through\n Goal Prioritization"], "answer_arxiv_id": ["2309.00614", "2309.07124", "2310.06387", "2311.09096"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_11436"} +{"question": "What is the notable work that proposed the restoration loss which is an alternative training objective suitable for categorical probability distributions?", "answer": ["Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise"], "answer_arxiv_id": ["2208.09392"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_11437"} +{"question": "Could you provide me some works on model-based reinforcement learning methods that constrain the policy to the region that is close to the training data?", "answer": ["MOPO: Model-based Offline Policy Optimization", "COMBO: Conservative Offline Model-Based Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2005.13239", "2102.08363", "2005.05951"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_11438"} +{"question": "Could you provide me some studies about observational robustness of an RL agent under observational adversarial attacks?", "answer": ["Robust Deep Reinforcement Learning against Adversarial Perturbations on State Observations", "Robust Reinforcement Learning on State Observations with Learned Optimal Adversary", "Efficient Adversarial Training without Attacking: Worst-Case-Aware Robust Reinforcement Learning", "Deep Reinforcement Learning Policies Learn Shared Adversarial Features Across MDPs"], "answer_arxiv_id": ["2003.08938", "2101.08452", "2210.05927v1", "2112.09025"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_11439"} +{"question": "Which studies focus on the definition of the term 'claim detection' in the field of fact-checking?", "answer": ["Fighting the COVID-19 Infodemic: Modeling the Perspective of\n Journalists, Fact-Checkers, Social Media Platforms, Policy Makers, and the\n Society", "Environmental Claim Detection", "A Benchmark Dataset of Check-worthy Factual Claims", "Towards Automated Factchecking: Developing an Annotation Schema and\n Benchmark for Consistent Automated Claim Detection", "NewsClaims: A New Benchmark for Claim Detection from News with Attribute\n Knowledge"], "answer_arxiv_id": ["2005.00033", "2209.00507", "2004.14425", "1809.08193", "2112.08544"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_11440"} +{"question": "Which works developed heavy-tailed bandit algorithms using truncation and median of means strategies?", "answer": ["Bandits with heavy tail", "Almost Optimal Algorithms for Linear Stochastic Bandits with Heavy-Tailed Payoffs", "Nearly Optimal Regret for Stochastic Linear Bandits with Heavy-Tailed Payoffs", "Adaptive Best-of-Both-Worlds Algorithm for Heavy-Tailed Multi-Armed Bandits"], "answer_arxiv_id": ["1209.1727", "1810.10895", "2004.13465", "2201.11921"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_11441"} +{"question": "What research argues that directly applying the contrastive objective to face images leads to pose-invariant representations?", "answer": ["Pose-disentangled Contrastive Learning for Self-supervised Facial\n Representation"], "answer_arxiv_id": ["2211.13490"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_11442"} +{"question": "What research studies provide probabilistic guarantees for traditional data poisoning using potentially stochastic learners?", "answer": ["Intrinsic Certified Robustness of Bagging against Data Poisoning Attacks", "A Framework of Randomized Selection Based Certified Defenses Against Data Poisoning Attacks"], "answer_arxiv_id": ["2008.04495", "2009.08739"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_11443"} +{"question": "Could you provide me some works about parameter isolation approaches in addressing the problem of catastrophic forgetting?", "answer": ["PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning", "Continual Learning via Neural Pruning", "Expert Gate: Lifelong Learning with a Network of Experts", "Lifelong Learning with Dynamically Expandable Networks"], "answer_arxiv_id": ["1711.05769", "1903.04476", "1611.06194", "1708.01547"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_11444"} +{"question": "Which research proposes an environment optimization method for improving the throughput of the multi-robot system in automated warehouses?", "answer": ["Multi-Robot Coordination and Layout Design for Automated Warehousing"], "answer_arxiv_id": ["2305.06436v3"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_11445"} +{"question": "What works are related to conditonal synthesis of images using GANs?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "Conditional Generative Adversarial Nets", "Conditional Image Synthesis With Auxiliary Classifier GANs"], "answer_arxiv_id": ["1711.10485", "1411.1784", "1610.09585"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_11446"} +{"question": "Which papers focused on reducing hallucinations through attribution?", "answer": ["Survey of Hallucination in Natural Language Generation", "Siren's Song in the AI Ocean: A Survey on Hallucination in Large\n Language Models"], "answer_arxiv_id": ["2202.03629", "2309.01219"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_11447"} +{"question": "Which research papers marked the emergence of diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_11448"} +{"question": "What works have been done in counterfactual explanations or algorithmic recourse in recent years?", "answer": ["Interpretable Predictions of Tree-based Ensembles via Actionable Feature Tweaking", "Inverse Classification for Comparison-based Interpretability in Machine Learning", "Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR", "Actionable Recourse in Linear Classification", "Interpretable Counterfactual Explanations Guided by Prototypes", "Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers", "Explaining Machine Learning Classifiers through Diverse Counterfactual Explanations", "Model-Agnostic Counterfactual Explanations for Consequential Decisions", "Multi-Objective Counterfactual Explanations"], "answer_arxiv_id": ["1706.06691", "1712.08443", "1711.00399", "1809.06514", "1907.02584", "1912.03277", "1905.07697", "1905.11190", "2004.11165v2"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_11449"} +{"question": "What work incorporates NeRF's implicit representation with a hypernetwork for conducting style transfer?", "answer": ["Stylizing 3D Scene via Implicit Representation and HyperNetwork"], "answer_arxiv_id": ["2105.13016"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_11450"} +{"question": "Which papers introduced various tabular neural architectures challenging the dominance of gradient-boosted decision trees?", "answer": ["Neural Oblivious Decision Ensembles for Deep Learning on Tabular Data", "The Tree Ensemble Layer: Differentiability meets Conditional Computation", "Deep Neural Decision Trees", "Gradient Boosting Neural Networks: GrowNet", "SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training", "Revisiting Deep Learning Models for Tabular Data", "TabNet: Attentive Interpretable Tabular Learning", "TabTransformer: Tabular Data Modeling Using Contextual Embeddings", "AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks", "Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning", "Deep & Cross Network for Ad Click Predictions", "DCN V2: Improved Deep & Cross Network and Practical Lessons for Web-scale Learning to Rank Systems", "Self-Normalizing Neural Networks", "Simple Modifications to Improve Tabular Neural Networks", "Hopular: Modern Hopfield Networks for Tabular Data"], "answer_arxiv_id": ["1909.06312", "2002.07772", "1806.06988v1", "2002.07971", "2106.01342", "2106.11959", "1908.07442", "2012.06678", "1810.11921", "2106.02584v2", "1708.05123", "2008.13535", "1706.02515", "2108.03214", "2206.00664"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_11451"} +{"question": "What works demonstrated transformers' ability to recognize Dyck languages?", "answer": ["Self-Attention Networks Can Process Bounded Hierarchical Languages", "Theoretical Limitations of Self-Attention in Neural Sequence Models"], "answer_arxiv_id": ["2105.11115", "1906.06755"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_11452"} +{"question": "Could you provide me some works about recent studies that utilized similar techniques for mobile applications regarding autonomous agents for web and mobile applications?", "answer": ["A Dataset for Interactive Vision-Language Navigation with Unknown Command Feasibility", "Mapping Natural Language Instructions to Mobile UI Action Sequences"], "answer_arxiv_id": ["2202.02312", "2005.03776"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11453"} +{"question": "What research papers have explored meta-learning techniques to accelerate optimization across various tasks?", "answer": ["Learning to learn with quantum neural networks via classical neural networks", "Learning to Optimize Variational Quantum Circuits to Solve Combinatorial Problems", "Optimizing quantum heuristics with meta-learning"], "answer_arxiv_id": ["1907.05415", "1911.11071", "1908.03185"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_11454"} +{"question": "Which research used linear relational networks to model object interactions and learn the policy?", "answer": ["Compositional Multi-Object Reinforcement Learning with Linear Relation Networks"], "answer_arxiv_id": ["2201.13388"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_11455"} +{"question": "Could you provide me some studies about expanding the radiance field of dynamic NeRFs into the temporal domain?", "answer": ["Space-time Neural Irradiance Fields for Free-Viewpoint Video", "Dynamic View Synthesis from Dynamic Monocular Video", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "HexPlane: A Fast Representation for Dynamic Scenes"], "answer_arxiv_id": ["2011.12950", "2105.06468", "2301.10241v2", "2301.09632"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11456"} +{"question": "What efforts have been made to make transformers suitable for online exploration?", "answer": ["Online Decision Transformer"], "answer_arxiv_id": ["2202.05607"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_11457"} +{"question": "Which papers utilize tri-planes for direct 3D generation?", "answer": ["Generative Neural Articulated Radiance Fields", "Unsupervised Learning of Efficient Geometry-Aware Neural Articulated\n Representations", "AG3D: Learning to Generate 3D Avatars from 2D Image Collections", "Rodin: A Generative Model for Sculpting 3D Digital Avatars Using\n Diffusion"], "answer_arxiv_id": ["2206.14314", "2204.08839", "2305.02312", "2212.06135"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_11458"} +{"question": "Which works developed learned search strategies for program synthesis?", "answer": ["A Syntactic Neural Model for General-Purpose Code Generation", "Write, Execute, Assess: Program Synthesis with a REPL", "Learning to Infer Program Sketches", "Neural Sketch Learning for Conditional Program Generation", "Latent Programmer: Discrete Latent Codes for Program Synthesis", "Learning to Combine Per-Example Solutions for Neural Program Synthesis"], "answer_arxiv_id": ["1704.01696", "1906.04604", "1902.06349", "1703.05698", "2012.00377", "2106.07175"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_11459"} +{"question": "Which works leveraged auxiliary memory to decouple computation in the context of architecture design?", "answer": ["Memory Networks", "End-To-End Memory Networks", "Neural Turing Machines", "Inferring Algorithmic Patterns with Stack-Augmented Recurrent Nets", "Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks", "Memory Transformer", "Object-Centric Learning with Slot Attention", "Big Bird: Transformers for Longer Sequences", "Coordination Among Neural Modules Through a Shared Global Workspace"], "answer_arxiv_id": ["1410.3916", "1503.08895", "1410.5401", "1503.01007", "1810.00825", "2006.11527", "2006.15055", "2007.14062", "2103.01197"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_11460"} +{"question": "Could you provide some works which have used transformers in offline RL?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Efficient Planning in a Compact Latent Action Space"], "answer_arxiv_id": ["2106.01345", "2208.10291"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_11461"} +{"question": "Which were the early 3D object detection methods that are extended from 2D detectors to predict additional 3D bounding boxes?", "answer": ["FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection", "DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries"], "answer_arxiv_id": ["2104.10956", "2110.06922"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_11462"} +{"question": "Which work demonstrated that solving multiple optimizations simultaneously leads to significant speedup?", "answer": ["Two-Dimensional Batch Linear Programming on the GPU"], "answer_arxiv_id": ["1902.04995"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_11463"} +{"question": "What studies investigate the effects of various training factors on the flatness of the found minima?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "Three Factors Influencing Minima in SGD", "The Implicit and Explicit Regularization Effects of Dropout"], "answer_arxiv_id": ["1609.04836", "1711.04623", "2002.12915"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_11464"} +{"question": "Which papers have suggested methods to prevent subject modifications in subject-driven generation?", "answer": ["Blended Latent Diffusion", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2206.02779", "2112.10741"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_11465"} +{"question": "Could you provide me some works about measuring the uncertainty using entropy, least confidence, and the margin between the most likely and second most likely labels?", "answer": ["Committee-Based Sample Selection for Probabilistic Classifiers"], "answer_arxiv_id": ["1106.0220v1"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_11466"} +{"question": "Are there any works that focus on AV out-of-domain (OoD) generalization and detection?", "answer": ["Can Autonomous Vehicles Identify, Recover From, and Adapt to\n Distribution Shifts?", "Interpretable Self-Aware Neural Networks for Robust Trajectory\n Prediction", "Multi-Predictor Fusion: Combining Learning-based and Rule-based\n Trajectory Predictors", "VISTA 2.0: An Open, Data-driven Simulator for Multimodal Sensing and\n Policy Learning for Autonomous Vehicles", "Scaling Out-of-Distribution Detection for Real-World Settings", "Task-Relevant Failure Detection for Trajectory Predictors in Autonomous\n Vehicles", "Robustness to Out-of-Distribution Inputs via Task-Aware Generative\n Uncertainty", "Expanding the Deployment Envelope of Behavior Prediction via Adaptive\n Meta-Learning"], "answer_arxiv_id": ["2006.14911", "2211.08701", "2307.01408", "2111.12083", "1911.11132", "2207.12380", "1812.10687", "2209.11820"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_11467"} +{"question": "Which papers propose scoring-based methods, e.g. measures of empathy and the use of counseling-specific dialogue acts, for automated methods to evaluate and improve peer counseling skills?", "answer": ["A Computational Framework for Behavioral Assessment of LLM Therapists"], "answer_arxiv_id": ["2401.00820"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_11468"} +{"question": "What works presented the early methods of unlearning in convolutional neural networks (CNN)?", "answer": ["Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks"], "answer_arxiv_id": ["1911.04933"], "source_meta": {"published_time": "20221015"}, "qid": "AutoScholarQuery_train_11469"} +{"question": "Are there any papers that identify challenges with accurate depth estimation in monocular 3D object detection?", "answer": ["Is Pseudo-Lidar needed for Monocular 3D Object detection?", "Omni3D: A Large Benchmark and Model for 3D Object Detection in the Wild"], "answer_arxiv_id": ["2108.06417", "2207.10660"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_11470"} +{"question": "Which works discussed about reconstructing 3D geometry from pointclouds?", "answer": ["Convolutional Occupancy Networks"], "answer_arxiv_id": ["2003.04618"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_11471"} +{"question": "Which studies focus on object-centric generative modelling that use a compositional image modelling approach?", "answer": ["Neural Expectation Maximization", "Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects", "MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational Inference", "SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition", "Object-Centric Learning with Slot Attention", "Efficient Iterative Amortized Inference for Learning Symmetric and Disentangled Multi-Object Representations", "Illiterate DALL-E Learns to Compose", "Conditional Object-Centric Learning from Video", "Bridging the Gap to Real-World Object-Centric Learning", "Neural Systematic Binder", "SAVi++: Towards End-to-End Object-Centric Learning from Real-World Videos"], "answer_arxiv_id": ["1708.03498", "1806.01794", "1901.11390", "1903.00450", "2001.02407", "2006.15055", "2106.03630", "2110.11405", "2111.12594", "2209.14860v2", "2211.01177", "2206.07764"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_11472"} +{"question": "Can you provide an example of a study that focuses on processing detected error-prone and incoherent regions with transformer?", "answer": ["Mask Transfiner for High-Quality Instance Segmentation"], "answer_arxiv_id": ["2111.13673"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_11473"} +{"question": "What studies are about the optimization of diffusion probabilistic models (DPMs) for few-step sampling?", "answer": ["Denoising Diffusion Implicit Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps", "Thompson Sampling Efficiently Learns to Control Diffusion Processes", "Score-Based Diffusion meets Annealed Importance Sampling", "GENIE: Higher-Order Denoising Diffusion Solvers"], "answer_arxiv_id": ["2010.02502", "2206.00927", "2206.09977", "2208.07698", "2210.05475"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_11474"} +{"question": "Which papers have utilized pessimistic RL algorithm for offline training but incorporated exploration in fine-tuning?", "answer": ["Offline-to-Online Reinforcement Learning via Balanced Replay and Pessimistic Q-Ensemble", "Supported Policy Optimization for Offline Reinforcement Learning"], "answer_arxiv_id": ["2107.00591", "2202.06239"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_11475"} +{"question": "Which work evaluates neural network interpretability methods in a feature selection setup on synthetic datasets?", "answer": ["How good Neural Networks interpretation methods really are? A quantitative benchmark."], "answer_arxiv_id": ["2304.02383"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_11476"} +{"question": "Which papers mentioned have studied template-based methods in pose estimation of unseen objects?", "answer": ["OSOP: A Multi-Stage One Shot Object Pose Estimation Framework", "OVE6D: Object Viewpoint Encoding for Depth-based 6D Object Pose\n Estimation", "Learning Descriptors for Object Recognition and 3D Pose Estimation"], "answer_arxiv_id": ["2203.15533", "2203.01072", "1502.05908"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_11477"} +{"question": "Which works discussed Generative Adversarial Networks (GANs) generation as a method for data-free knowledge distillation?", "answer": ["Data-Free Learning of Student Networks"], "answer_arxiv_id": ["1904.01186"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_11478"} +{"question": "In which studies the authors aimed to train reference models as proxies for the missing bias labels for bias-unsupervised solutions?", "answer": ["Environment Inference for Invariant Learning", "When Does Group Invariant Learning Survive Spurious Correlations?", "Just Train Twice: Improving Group Robustness without Training Group Information"], "answer_arxiv_id": ["2010.07249", "2206.14534", "2107.09044"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_11479"} +{"question": "Which studies are about the differentiable economics?", "answer": ["ProportionNet: Balancing Fairness and Revenue for Auction Design with Deep Learning", "Deep Learning for Two-Sided Matching"], "answer_arxiv_id": ["2010.06398", "2107.03427"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_11480"} +{"question": "Could you provide me some works aiming to perform fully-quantized training of DNNs?", "answer": ["Scalable Methods for 8-bit Training of Neural Networks", "Towards Unified INT8 Training for Convolutional Neural Network"], "answer_arxiv_id": ["1805.11046", "1912.12607"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_11481"} +{"question": "Which works incorporate multi-view context to estimate camera 6D pose?", "answer": ["RelPose++: Recovering 6D Poses from Sparse-view Observations"], "answer_arxiv_id": ["2305.04926"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_11482"} +{"question": "What papers reported techniques to reduce the cost of 3D-aware generation for high-resolution data?", "answer": ["GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "GNeRF: GAN-based Neural Radiance Field without Posed Camera", "EpiGRAF: Rethinking training of 3D GANs", "Generative Multiplane Images: Making a 2D GAN 3D-Aware", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis"], "answer_arxiv_id": ["2007.02442", "2103.15606", "2206.10535", "2207.10642", "2110.08985"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_11483"} +{"question": "Could you provide me some works that demonstrate that language models are able to encode size?", "answer": ["Do Language Embeddings Capture Scales?"], "answer_arxiv_id": ["2010.05345"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_11484"} +{"question": "In what papers the researcher designed an embedding model to bridge the gap between seen and unseen categories in Zero-Shot Learning?", "answer": ["Learning a Deep Embedding Model for Zero-Shot Learning"], "answer_arxiv_id": ["1611.05088"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_11485"} +{"question": "Which papers introduced unsupervised recalibration algorithms for iBCI decoders that work by aligning distributions of neural data?", "answer": ["Adversarial Domain Adaptation For Stable Brain-Machine Interfaces"], "answer_arxiv_id": ["1810.00045"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_11486"} +{"question": "Can you provide works about using homotopy methods in training feedforward networks?", "answer": ["Mollifying Networks", "A homotopy training algorithm for fully connected neural networks"], "answer_arxiv_id": ["1608.04980", "1903.09872"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_11487"} +{"question": "Which paper proposed to parameterize the transition kernel and reproduction distribution of the BA algorithm by neural density estimators?", "answer": ["Towards Empirical Sandwich Bounds on the Rate-Distortion Function"], "answer_arxiv_id": ["2111.12166"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_11488"} +{"question": "What works investigated the generalization theory in deep learning and how the model's performance can be transferred to the population distribution?", "answer": ["Learning and Generalization in Overparameterized Neural Networks, Going Beyond Two Layers", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "Spectrally-normalized margin bounds for neural networks", "Benign Overfitting in Linear Regression", "Rademacher Complexity for Adversarially Robust Generalization"], "answer_arxiv_id": ["1811.04918", "1901.08584", "1706.08498", "1906.11300", "1810.11914"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_11489"} +{"question": "Which research papers discuss convergence guarantees to Nash equilibria in symmetric games?", "answer": ["For Learning in Symmetric Teams, Local Optima are Global Nash Equilibria"], "answer_arxiv_id": ["2207.03470"], "source_meta": {"published_time": "20220803"}, "qid": "AutoScholarQuery_train_11490"} +{"question": "What researches have proposed solutions to the group robustness problem when bias attribute data is available?", "answer": ["Learning Debiased and Disentangled Representations for Semantic Segmentation", "Just Train Twice: Improving Group Robustness without Training Group Information", "Unlearn Dataset Bias in Natural Language Inference by Fitting the Residual", "Avoiding spurious correlations via logit correction", "Don’t Take the Easy Way Out: Ensemble Based Methods for Avoiding Known Dataset Biases"], "answer_arxiv_id": ["2111.00531", "2107.09044", "1908.10763", "2212.01433", "1909.03683"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_11491"} +{"question": "Which works focus on improving documents through human-machine interactions?", "answer": ["Read, Revise, Repeat: A System Demonstration for Human-in-the-loop Iterative Text Revision", "Understanding Iterative Revision from Human-Written Text"], "answer_arxiv_id": ["2204.03685", "2203.03802"], "source_meta": {"published_time": "20220824"}, "qid": "AutoScholarQuery_train_11492"} +{"question": "Which research papers have demonstrated improvements in Language Learning Models (LLM) enabling human-like text generation?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models", "LLM360: Towards Fully Transparent Open-Source LLMs"], "answer_arxiv_id": ["2005.14165", "2204.02311", "2302.13971", "2312.06550"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_11493"} +{"question": "Which works showed that the CD and PCD algorithms are essentially adversarial procedures?", "answer": ["Contrastive Divergence Learning is a Time Reversal Adversarial Game", "On Energy-Based Models with Overparametrized Shallow Neural Networks"], "answer_arxiv_id": ["2012.03295", "2104.07531"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11494"} +{"question": "Could you provide me with some studies about authorship attribution?", "answer": ["GPT-who: An Information Density-based Machine-Generated Text Detector", "Few-Shot Detection of Machine-Generated Text using Style Representations"], "answer_arxiv_id": ["2310.06202", "2401.06712"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_11495"} +{"question": "What are studies where the generation process is treated as an optimization problem over the embedding or token sequences in control of language models?", "answer": ["Controlled Text Generation as Continuous Optimization with Multiple\n Constraints"], "answer_arxiv_id": ["2108.01850"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_11496"} +{"question": "What work involves freezing the A matrix within LoRA to reduce the number of trainable parameters?", "answer": ["LoRA-FA: Memory-efficient Low-rank Adaptation for Large Language Models\n Fine-tuning"], "answer_arxiv_id": ["2308.03303"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_11497"} +{"question": "What papers discuss methods for backpropagating image-space photometric error in relation to 3D scene representations?", "answer": ["RenderNet: A deep convolutional network for differentiable rendering\n from 3D shapes", "DeepVoxels: Learning Persistent 3D Feature Embeddings", "Neural Volumes: Learning Dynamic Renderable Volumes from Images"], "answer_arxiv_id": ["1806.06575", "1812.01024v2", "1906.07751"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_11498"} +{"question": "Could you provide me a study about the development of algorithms for boosting generative models?", "answer": ["Boosted Generative Models"], "answer_arxiv_id": ["1702.08484"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_11499"} +{"question": "What studies have implemented additional layers in the object detectors to collect feature maps?", "answer": ["Feature Pyramid Networks for Object Detection", "EfficientDet: Scalable and Efficient Object Detection"], "answer_arxiv_id": ["1612.03144", "1911.09070"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_11500"} +{"question": "What studies have shown that learning ReLU regression is computationally intractable even in the presence of Gaussian features?", "answer": ["Time/Accuracy Tradeoffs for Learning a ReLU with respect to Gaussian Marginals"], "answer_arxiv_id": ["1911.01462"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_11501"} +{"question": "What researches propose augmenting supervision with AI systems when making language models truthful?", "answer": ["Supervising strong learners by amplifying weak experts", "AI safety via debate", "Scalable agent alignment via reward modeling: a research direction", "Red Teaming Language Models with Language Models"], "answer_arxiv_id": ["1810.08575", "1805.00899", "1811.07871", "2202.03286v1"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_11502"} +{"question": "Which researches focus on learning object motion in 3D?", "answer": ["3D-OES: Viewpoint-Invariant Object-Factorized Environment Simulators", "Learning 3D Dynamic Scene Representations for Robot Manipulation", "Keypoints into the Future: Self-Supervised Correspondence in Model-Based Reinforcement Learning", "3D Neural Scene Representations for Visuomotor Control"], "answer_arxiv_id": ["2011.06464", "2011.01968", "2009.05085", "2107.04004"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_11503"} +{"question": "Who proposed understanding the behaviour of self-predictive learning through its corresponding ODE?", "answer": ["Understanding Self-Predictive Learning for Reinforcement Learning"], "answer_arxiv_id": ["2212.03319"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11504"} +{"question": "Could you provide me some studies which combined consistency regularization and pseudo-labels to improve Semi-Supervised Learning's outcome?", "answer": ["MixMatch: A Holistic Approach to Semi-Supervised Learning", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling"], "answer_arxiv_id": ["1905.02249", "2001.07685", "2110.08263"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_11505"} +{"question": "What papers discuss the efforts to address gaps in representation of biases unique to India?", "answer": ["Socially Aware Bias Measurements for Hindi Language Representations", "Re-contextualizing Fairness in NLP: The Case of India", "SeeGULL: A Stereotype Benchmark with Broad Geo-Cultural Coverage Leveraging Generative Models"], "answer_arxiv_id": ["2110.07871", "2209.12226", "2305.11840"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_11506"} +{"question": "Could you provide papers that looked into token redundancy in the quest to build efficient vision transformers?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "Not All Patches are What You Need: Expediting Vision Transformers via\n Token Reorganizations", "AdaViT: Adaptive Tokens for Efficient Vision Transformer", "Adaptive Token Sampling For Efficient Vision Transformers", "Token Merging: Your ViT But Faster", "DiffRate : Differentiable Compression Rate for Efficient Vision\n Transformers"], "answer_arxiv_id": ["2106.02034v2", "2202.07800", "2112.07658", "2111.15667", "2210.09461", "2305.17997"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_11507"} +{"question": "Which study presented ODE-RNNs, which use Neural Ordinary Differential Equations to model hidden state dynamics of RNNs?", "answer": ["Latent ODEs for Irregularly-Sampled Time Series"], "answer_arxiv_id": ["1907.03907"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_11508"} +{"question": "Which papers proposed heuristic methods for detecting hallucination or estimating the confidence level of the answer in LLMs?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models", "SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for\n Generative Large Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2203.11171", "2303.08896", "2201.11903"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_11509"} +{"question": "Which paper introduced memory banks and segmentation-aware negative sampling methods?", "answer": ["Exploring Cross-Image Pixel Contrast for Semantic Segmentation"], "answer_arxiv_id": ["2101.11939"], "source_meta": {"published_time": "20240416"}, "qid": "AutoScholarQuery_train_11510"} +{"question": "What papers present extrapolating architectures namely Neural Arithmetic Logic Units (NALU) and Neural Arithmetic Units (NAU)?", "answer": ["Neural Arithmetic Logic Units", "iNALU: Improved Neural Arithmetic Logic Unit", "Neural Arithmetic Units"], "answer_arxiv_id": ["1808.00508", "2003.07629", "2001.05016"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_11511"} +{"question": "Could you provide me some studies about PTQ techniques tailored to LLMs?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models", "FlexRound: Learnable Rounding based on Element-wise Division for Post-Training Quantization", "Rethinking Channel Dimensions to Isolate Outliers for Low-bit Weight Quantization of Large Language Models"], "answer_arxiv_id": ["2208.07339", "2206.01861", "2211.10438", "2306.00317", "2309.15531v2"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_11512"} +{"question": "Which papers have contributed to traditional methods of Image compositing focusing on color and lighting consistency?", "answer": ["SSH: A Self-Supervised Framework for Image Harmonization", "DCCF: Deep Comprehensible Color Filter Learning Framework for\n High-Resolution Image Harmonization", "Harmonizer: Learning to Perform White-Box Image and Video Harmonization", "Deep Image Blending", "Deep Image Compositing", "GP-GAN: Towards Realistic High-Resolution Image Blending"], "answer_arxiv_id": ["2108.06805", "2207.04788", "2207.01322", "1910.11495", "2011.02146", "1703.07195"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_11513"} +{"question": "Could you provide some studies in which LSTMs and their extensions were used for automatic diacritization task?", "answer": ["Neural Arabic Text Diacritization: State of the Art Results and a Novel\n Approach for Machine Translation", "Arabic Diacritic Recovery Using a Feature-Rich biLSTM Model"], "answer_arxiv_id": ["1911.03531", "2002.01207"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_train_11514"} +{"question": "What work quantifies the price of anarchy in large games using the smoothness framework?", "answer": ["The Price of Anarchy in Large Games"], "answer_arxiv_id": ["1503.04755"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_11515"} +{"question": "What research introduced simple regret in the context of multi-armed bandit problems?", "answer": ["Pure Exploration in Finitely–Armed and Continuous–Armed Bandits"], "answer_arxiv_id": ["0802.2655"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_11516"} +{"question": "Can you name some studies that use Monte Carlo Tree Search to perform single and multi-objective optimization?", "answer": ["Learning Search Space Partition for Black-box Optimization using Monte Carlo Tree Search", "Sample-Efficient Neural Architecture Search by Learning Actions for Monte Carlo Tree Search", "Learning Space Partitions for Path Planning"], "answer_arxiv_id": ["2007.00708", "1906.06832", "2106.10544"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_11517"} +{"question": "Are there any works that proposed using the base model for a second round of fine-tuning in text summarization?", "answer": ["BRIO: Bringing Order to Abstractive Summarization"], "answer_arxiv_id": ["2203.16804"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_11518"} +{"question": "What works propose neural construction heuristics based on deep models?", "answer": ["Pointer Networks", "Neural Combinatorial Optimization with Reinforcement Learning", "Reinforcement Learning for Solving the Vehicle Routing Problem"], "answer_arxiv_id": ["1506.03134", "1611.09940", "1802.04240"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_11519"} +{"question": "Can you provide works that investigate the use of the student model as a proxy for transfer-based adversarial attacks?", "answer": ["MEGA: Model Stealing via Collaborative Generator-Substitute Networks", "Model Extraction and Adversarial Attacks on Neural Networks using Switching Power Information", "Black-Box Attacks on Sequential Recommenders via Data-Free Model Extraction"], "answer_arxiv_id": ["2202.00008", "2106.08299", "2109.01165"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_11520"} +{"question": "Any researches that introduce predominantly binary models in the context of low-precision quantization?", "answer": ["PokeBNN: A Binary Pursuit of Lightweight Accuracy", "ReActNet: Towards Precise Binary Neural Network with Generalized\n Activation Functions", "Binarizing MobileNet via Evolution-based Searching"], "answer_arxiv_id": ["2112.00133", "2003.03488", "2005.06305"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_11521"} +{"question": "Can you cite studies that first theoretically analyzed contrastive learning?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Contrastive learning, multi-view redundancy, and linear models", "Contrastive estimation reveals topic posterior information to linear models", "Understanding Self-supervised Learning with Dual Deep Networks"], "answer_arxiv_id": ["1902.09229v1", "2008.10150", "2003.02234", "2010.00578"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_11522"} +{"question": "What paper reported learning for dynamical system as well as a Lyapunov function to ensure the stability of the system?", "answer": ["Learning Stable Deep Dynamics Models"], "answer_arxiv_id": ["2001.06116"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_11523"} +{"question": "Can you acknowledge any papers on instance-level segmentation that introduced architectural innovations?", "answer": ["Mask R-CNN", "K-Net: Towards Unified Image Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["1703.06870", "2106.14855", "2112.01527"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_11524"} +{"question": "Which papers propose approaches to deal with hidden confounding by combining both experimental and observational data?", "answer": ["Removing Hidden Confounding by Experimental Grounding", "Combining Experimental and Observational Data to Estimate Treatment Effects on Long Term Outcomes", "Combining Observational and Randomized Data for Estimating Heterogeneous Treatment Effects", "Combining Interventional and Observational Data Using Causal Reductions", "Long-term Causal Inference Under Persistent Confounding via Data Combination"], "answer_arxiv_id": ["1810.11646", "2006.09676", "2202.12891", "2103.04786v3", "2202.07234"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_11525"} +{"question": "Could you provide me some studies about approximately MMS allocations for chores?", "answer": ["Approximation Algorithms for Maximin Fair Division", "An Algorithmic Framework for Approximating Maximin Share Allocation of Chores"], "answer_arxiv_id": ["1703.01851", "1907.04505"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_11526"} +{"question": "Which papers discuss the adoption of the mixture of experts (MoE) in the transformer architecture to reduce computational cost while maintaining a large model capacity?", "answer": ["GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding", "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity"], "answer_arxiv_id": ["2006.16668", "2101.03961"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_11527"} +{"question": "What researches propose Self-Tuning variants which combine contrastive learning with transfer learning?", "answer": ["Self-Tuning for Data-Efficient Deep Learning", "Hyperspherical Consistency Regularization"], "answer_arxiv_id": ["2102.12903", "2206.00845"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_11528"} +{"question": "Which work synthesizes images by composing subjects with random backgrounds?", "answer": ["BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing"], "answer_arxiv_id": ["2305.14720"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_11529"} +{"question": "What are the fundamental models on which the study of the CBMUM problem is based?", "answer": ["Online Clustering of Bandits", "Improved Algorithm on Online Clustering of Bandits", "Learning with Good Feature Representations in Bandits and in RL with a Generative Model"], "answer_arxiv_id": ["1401.8257", "1902.09162", "1911.07676"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_11530"} +{"question": "What papers included word embeddings of actions as textual semantic representations?", "answer": ["Rethinking Zero-shot Video Classification: End-to-end Training for Realistic Applications", "Transductive Zero-Shot Action Recognition by Word-Vector Embedding"], "answer_arxiv_id": ["2003.01455", "1511.04458"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_11531"} +{"question": "What papers apply offline learning to the healthcare industry?", "answer": ["Learning When-to-Treat Policies"], "answer_arxiv_id": ["1905.09751"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_11532"} +{"question": "Do you know any studies about bilevel optimization methods in distributed learning?", "answer": ["FedNest: Federated Bilevel, Minimax, and Compositional Optimization", "Decentralized Gossip-Based Stochastic Bilevel Optimization over Communication Networks"], "answer_arxiv_id": ["2205.02215v3", "2206.10870"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_11533"} +{"question": "Which works discuss the use of Variance-Invariance-Covariance Regularization (VICReg) for SSL training?", "answer": ["VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning"], "answer_arxiv_id": ["2105.04906"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11534"} +{"question": "Which work has respectively planned the problem of handling new relations?", "answer": ["Relational Message Passing for Fully Inductive Knowledge Graph Completion"], "answer_arxiv_id": ["2210.03994"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11535"} +{"question": "Could you provide me some works about deep learning models for point cloud registration?", "answer": ["Deep Closest Point: Learning Representations for Point Cloud Registration", "PRNet: Self-Supervised Learning for Partial-to-Partial Registration", "PointNetLK: Robust & Efficient Point Cloud Registration using PointNet", "Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondences", "Sampling Network Guided Cross-Entropy Method for Unsupervised Point Cloud Registration", "RPM-Net: Robust Point Matching using Learned Features", "Robust Point Cloud Registration Framework Based on Deep Graph Matching", "Reliable Inlier Evaluation for Unsupervised Point Cloud Registration"], "answer_arxiv_id": ["1905.03304", "1910.12240", "1903.05711", "2005.01014", "2109.06619", "2003.13479", "2211.04696", "2202.11292"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_11536"} +{"question": "Tell me about some published articles that deal with the text-driven stylization of a given surface mesh or synthesizing a NeRF from an input text?", "answer": ["Text2Mesh: Text-Driven Neural Stylization for Meshes", "Zero-Shot Text-Guided Object Generation with Dream Fields", "DreamFusion: Text-to-3D using 2D Diffusion", "Text-To-4D Dynamic Scene Generation"], "answer_arxiv_id": ["2112.03221", "2112.01455", "2209.14988", "2301.11280"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_11537"} +{"question": "Can you provide some works on depth prediction in the context of geometric priors from single-view networks?", "answer": ["Depth Map Prediction from a Single Image using a Multi-Scale Deep\n Network", "Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer", "Vision Transformers for Dense Prediction"], "answer_arxiv_id": ["1406.2283", "1907.01341v3", "2103.13413"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_11538"} +{"question": "Can you cite a study that uses a discrete codebook for storing facial motion in the context of Vector-Quantized Network?", "answer": ["Learning to Listen: Modeling Non-Deterministic Dyadic Facial Motion"], "answer_arxiv_id": ["2204.08451"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_11539"} +{"question": "What research papers have focused on integrating Data Augmentation with other representation learning methods to improve sample efficiency?", "answer": ["Data-Efficient Reinforcement Learning with Self-Predictive Representations", "PlayVirtual: Augmenting Cycle-Consistent Virtual Trajectories for Reinforcement Learning", "CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Masked Contrastive Representation Learning for Reinforcement Learning", "Mask-based Latent Reconstruction for Reinforcement Learning"], "answer_arxiv_id": ["2007.05929", "2106.04152", "2004.04136", "2010.07470", "2201.12096"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11540"} +{"question": "Any works about the use of diverse training data to improve generalization in domain generalization research?", "answer": ["Learning to Learn without Forgetting By Maximizing Transfer and Minimizing Interference", "DeceptionNet: Network-Driven Domain Randomization"], "answer_arxiv_id": ["1810.11910", "1904.02750"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_11541"} +{"question": "Which works experimented with the architecture for non-image based domains where the states are concatenated with learned GVFs?", "answer": ["What Should I Know? Using Meta-gradient Descent for Predictive Feature Discovery in a Single Stream of Experience"], "answer_arxiv_id": ["2206.06485"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_11542"} +{"question": "Which research introduced a cyclical system for caption generation and sentence localization in weakly-supervised dense video captioning?", "answer": ["Weakly Supervised Dense Event Captioning in Videos"], "answer_arxiv_id": ["1812.03849"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_11543"} +{"question": "Which papers are regarding the generation of faces in diverse styles with recent advances in the GANs and DMs?", "answer": ["3DAvatarGAN: Bridging Domains for Personalized Editable Avatars", "DATID-3D: Diversity-Preserved Domain Adaptation Using Text-to-Image\n Diffusion for 3D Generative Model", "Dr.3D: Adapting 3D GANs to Artistic Drawings", "Instruct-Video2Avatar: Video-to-Avatar Generation with Instructions"], "answer_arxiv_id": ["2301.02700", "2211.16374", "2211.16798", "2306.02903"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_11544"} +{"question": "Could you provide me some studies about selection bias of Large Language Models in Multiple-Choice Questions?", "answer": ["Large Language Models Sensitivity to The Order of Options in\n Multiple-Choice Questions", "Large Language Models Are Not Robust Multiple Choice Selectors", "Leveraging Large Language Models for Multiple Choice Question Answering"], "answer_arxiv_id": ["2308.11483", "2309.03882", "2210.12353"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_train_11545"} +{"question": "Any works that uses the NeRF technique for generative modeling?", "answer": ["GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields", "Generative Occupancy Fields for 3D Surface-Aware Image Synthesis", "GRAM: Generative Radiance Manifolds for 3D-Aware Image Generation"], "answer_arxiv_id": ["2007.02442", "2011.12100", "2111.00969", "2112.08867"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_11546"} +{"question": "What research applied transformers as encoders to attain feature maps of images for pixel level prediction tasks like segmentation?", "answer": ["Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers", "SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers", "Vision Transformers for Dense Prediction"], "answer_arxiv_id": ["2012.15840", "2105.15203", "2103.13413"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_11547"} +{"question": "Which papers use 3DMM’s blendshapes as predefined motion representations?", "answer": ["HeadGAN: One-shot Neural Head Synthesis and Editing", "StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via\n Pre-trained StyleGAN", "NOFA: NeRF-based One-shot Facial Avatar Reconstruction", "Realistic One-shot Mesh-based Head Avatars"], "answer_arxiv_id": ["2012.08261", "2203.04036", "2307.03441", "2206.08343"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_11548"} +{"question": "Could you provide me some research papers that observed improved convergence and fidelity of GANs when using existing, generic image-based models?", "answer": ["Ensembling Off-the-shelf Models for GAN Training", "Projected GANs Converge Faster", "StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets", "Freeze the Discriminator: a Simple Baseline for Fine-Tuning GANs"], "answer_arxiv_id": ["2112.09130", "2111.01007", "2202.00273", "2002.10964"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_11549"} +{"question": "Which papers discuss adaptive re-estimation when the Polyak stepsize is unknown?", "answer": ["Revisiting the Polyak Step Size"], "answer_arxiv_id": ["1905.00313"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11550"} +{"question": "Which work has used centroid offsets of each point in the predictions?", "answer": ["PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation"], "answer_arxiv_id": ["2004.01658"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_11551"} +{"question": "Which works provided a summary of recent results of global convergences in an overparameterized regime?", "answer": ["Improved Overparametrization Bounds for Global Convergence of SGD for Shallow Neural Networks"], "answer_arxiv_id": ["2201.12052"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_11552"} +{"question": "Could you give me an example of research proposing a Unified Demonstration Retriever across different tasks?", "answer": ["Unified Demonstration Retriever for In-Context Learning"], "answer_arxiv_id": ["2305.04320"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_11553"} +{"question": "What works have discussed leveraging ground truth motions as conditions for video diffusion models?", "answer": ["ControlVideo: Training-free Controllable Text-to-Video Generation"], "answer_arxiv_id": ["2305.13077"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_11554"} +{"question": "Can you provide an example of which variational family is best for distributing q​(z∣x) for mutual information estimation?", "answer": ["Mutual Information Neural Estimation", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1801.04062", "1807.03748"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_11555"} +{"question": "What studies showed that training small transformer models from scratch on scratchpad data could improve length generalization?", "answer": ["What Algorithms can Transformers Learn? A Study in Length Generalization"], "answer_arxiv_id": ["2310.16028"], "source_meta": {"published_time": "20240705"}, "qid": "AutoScholarQuery_train_11556"} +{"question": "Could you list some works that employed machine learning methods in conditionally randomized trials settings?", "answer": ["Recursive Partitioning for Heterogeneous Causal Effects", "Estimation and Inference of Heterogeneous Treatment Effects using Random Forests", "Adapting Neural Networks for the Estimation of Treatment Effects", "Double/Debiased Machine Learning for Treatment and Structural Parameters"], "answer_arxiv_id": ["1504.01132", "1510.04342", "1906.02120", "1608.00060"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_11557"} +{"question": "Could you provide me some works that improved the SimCLR method?", "answer": ["Exploring Simple Siamese Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["2011.10566", "2006.07733"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_11558"} +{"question": "What studies provide insight on reducing energy consumption of neural networks through low-bit fixed-point representations?", "answer": ["FP8 versus INT8 for efficient deep learning inference"], "answer_arxiv_id": ["2303.17951"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_11559"} +{"question": "Could you mention some datasets for evaluating logical reasoning?", "answer": ["On the Paradox of Learning to Reason from Data", "TaxiNLI: Taking a Ride up the NLU Hill", "RuleBert: Teaching Soft Rules to Pre-trained Language Models"], "answer_arxiv_id": ["2205.11502", "2009.14505", "2109.13006"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_11560"} +{"question": "Can you provide the research papers that used adversarial-based methods for text-to-image generations?", "answer": ["StyleGAN-T: Unlocking the Power of GANs for Fast Large-Scale\n Text-to-Image Synthesis"], "answer_arxiv_id": ["2301.09515"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_11561"} +{"question": "What studies have proposed solutions including dense annotations, out-of-domain data, and interaction with annotators to deal with reasoning shortcuts in machine learning?", "answer": ["Right for the Right Reasons: Training Differentiable Models by Constraining their Explanations", "Learning explanations that are hard to vary", "Leveraging Explanations in Interactive Machine Learning: An Overview"], "answer_arxiv_id": ["1703.03717", "2009.00329", "2207.14526"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_11562"} +{"question": "Which papers indicate that LVLMs can be easily fooled and suffer from performance drop due to their over-reliance on a strong language prior?", "answer": ["Mitigating Hallucination in Large Multi-Modal Models via Robust\n Instruction Tuning", "Fool Your (Vision and) Language Model With Embarrassingly Simple\n Permutations"], "answer_arxiv_id": ["2306.14565", "2310.01651"], "source_meta": {"published_time": "20240630"}, "qid": "AutoScholarQuery_train_11563"} +{"question": "Could you mention some works that involve Bongard problems?", "answer": ["Solving Bongard Problems with a Visual Language and Pragmatic Reasoning", "Bongard-LOGO: A New Benchmark for Human-Level Concept Learning and Reasoning"], "answer_arxiv_id": ["1804.04452", "2010.00763"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_11564"} +{"question": "Are there any papers that integrate additional control signals in diffusion models beyond text conditioning?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_11565"} +{"question": "What are some works that discussed transfer-based adversarial attacks?", "answer": ["Boosting Adversarial Attacks with Momentum", "Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks", "Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks"], "answer_arxiv_id": ["1710.06081", "1908.06281", "1904.02884"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_11566"} +{"question": "Can you name the research that has focused on regression-based schemes for 3D human pose and shape estimation?", "answer": ["End-to-end Recovery of Human Shape and Pose", "Learning 3D Human Dynamics from Video", "VIBE: Video Inference for Human Body Pose and Shape Estimation", "Human Mesh Recovery from Monocular Images via a Skeleton-disentangled Representation", "Sim2real transfer learning for 3D human pose estimation: motion to the rescue", "Beyond Static Features for Temporally Consistent 3D Human Pose and Shape from a Video", "End-to-End Human Pose and Mesh Reconstruction with Transformers", "Mesh Graphormer"], "answer_arxiv_id": ["1712.06584", "1812.01601", "1912.05656", "1908.07172", "1907.02499", "2011.08627", "2012.09760", "2104.00272"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_11567"} +{"question": "What work proposed a Transformer-based model that learns individual connectivity strengths between each ROI?", "answer": ["Brain Network Transformer"], "answer_arxiv_id": ["2210.06681"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_11568"} +{"question": "What work categorizes exploration in RL into three types?", "answer": ["Exploration in Deep Reinforcement Learning: A Survey"], "answer_arxiv_id": ["2205.00824"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_11569"} +{"question": "Could you provide me some works that use Score Distillation to directly optimize a neural radiance field?", "answer": ["Compositional 3D Scene Generation using Locally Conditioned Diffusion", "DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2303.12218", "2209.14988"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_11570"} +{"question": "Could you provide me the references about detection task in OCR?", "answer": ["EAST: An Efficient and Accurate Scene Text Detector", "Arbitrary-Oriented Scene Text Detection via Rotation Proposals", "Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusion", "Shape Robust Text Detection with Progressive Scale Expansion Network"], "answer_arxiv_id": ["1704.03155", "1703.01086", "2202.10304", "1806.02559"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_11571"} +{"question": "Could you provide me some works about conversational information seeking without the need to access the dynamic environment for multiple times?", "answer": ["KwaiAgents: Generalized Information-seeking Agent System with Large\n Language Models"], "answer_arxiv_id": ["2312.04889"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_11572"} +{"question": "In what works variance reduction is achieved by circumventing high variance of policy value estimation through techniques like weight clipping or normalization?", "answer": ["Counterfactual Risk Minimization: Learning from Logged Bandit Feedback", "Bayesian Counterfactual Risk Minimization"], "answer_arxiv_id": ["1502.02362", "1806.11500"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_11573"} +{"question": "Which study developed sample-efficient algorithms for determining the von Neumann winner in the case of general preferences?", "answer": ["Human-in-the-loop: Provably Efficient Preference-based Reinforcement Learning with General Function Approximation"], "answer_arxiv_id": ["2205.11140"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_11574"} +{"question": "What papers have been written on mitigating spurious correlations in models caused by bias in the dataset?", "answer": ["Learning to Model and Ignore Dataset Bias with Mixed Capacity Ensembles", "Adversarial Filters of Dataset Biases", "Towards Robustifying NLI Models Against Lexical Dataset Biases", "Just Train Twice: Improving Group Robustness without Training Group\n Information", "CAT: Causal Audio Transformer for Audio Classification"], "answer_arxiv_id": ["2011.03856", "2002.04108", "2005.04732", "2107.09044", "2303.07626"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_11575"} +{"question": "Who demonstrated the hardness of learning depth-333 ReLU networks under the Gaussian distribution even with non-degenerate weight matrices?", "answer": ["Computational Complexity of Learning Neural Networks: Smoothness and Degeneracy"], "answer_arxiv_id": ["2302.07426"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11576"} +{"question": "What are some of the studies on differentiable physics simulation that show great promise?", "answer": ["Differentiable Simulation of Soft Multi-body Systems", "Differentiable Implicit Soft-Body Physics", "ADD: Analytically Differentiable Dynamics for Multi-Body Systems with Frictional Contact", "DiSECt: A Differentiable Simulation Engine for Autonomous Robotic Cutting", "gradSim: Differentiable simulation for system identification and visuomotor control", "RISP: Rendering-Invariant State Predictor with Differentiable Simulation and Rendering for Cross-Domain Parameter Estimation"], "answer_arxiv_id": ["2205.01758v1", "2102.05791", "2007.00987", "2105.12244", "2104.02646v1", "2205.05678"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_11577"} +{"question": "Could you provide me some studies about mapping between two domains using expert demonstrations for proxy tasks?", "answer": ["Domain Adaptive Imitation Learning", "Cross-domain Imitation from Observations"], "answer_arxiv_id": ["1910.00105v2", "2105.10037"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_11578"} +{"question": "What references showed the effectiveness of the PPO algorithm in computer vision?", "answer": ["Proactive Multi-Camera Collaboration for 3D Human Pose Estimation", "AirCapRL: Autonomous Aerial Human Motion Capture using Deep Reinforcement Learning"], "answer_arxiv_id": ["2303.03767", "2007.06343"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_11579"} +{"question": "Which research papers are about improving transformer-based models through model scaling, pretraining, and improved image quantization models?", "answer": ["Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "CogView2: Faster and Better Text-to-Image Generation via Hierarchical Transformers", "Taming Transformers for High-Resolution Image Synthesis", "Vector-quantized Image Modeling with Improved VQGAN"], "answer_arxiv_id": ["2206.10789", "2204.14217", "2012.09841", "2110.04627"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_11580"} +{"question": "Could you provide me some works about domain generalization improvements using pretraining models?", "answer": ["Domain Generalization using Pretrained Models without Fine-tuning", "Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization", "ZooD: Exploiting Model Zoo for Out-of-Distribution Generalization", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time", "Diverse Weight Averaging for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2203.04600", "2110.10832", "2210.09236", "2203.05482", "2205.09739"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_11581"} +{"question": "In what papers the researchers were aimed at estimating HR images from multiple input images in a temporal sliding-window manner?", "answer": ["MuCAN: Multi-Correspondence Aggregation Network for Video\n Super-Resolution", "EDVR: Video Restoration with Enhanced Deformable Convolutional Networks", "Video Super-Resolution Transformer", "VRT: A Video Restoration Transformer"], "answer_arxiv_id": ["2007.11803", "1905.02716", "2106.06847", "2201.12288"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_11582"} +{"question": "Could you provide the paper which proposed conformal risk control for diffusion models integrating quantile regression?", "answer": ["Conffusion: Confidence Intervals for Diffusion Models"], "answer_arxiv_id": ["2211.09795"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_11583"} +{"question": "What research implements diffusion models for 3D point cloud generation?", "answer": ["Diffusion Probabilistic Models for 3D Point Cloud Generation", "LION: Latent Point Diffusion Models for 3D Shape Generation"], "answer_arxiv_id": ["2103.01458", "2210.06978"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_11584"} +{"question": "Which works have achieved bounded regret for general-sum games and variationally stable games?", "answer": ["Near-Optimal No-Regret Learning in General Games", "Adaptive Learning in Continuous Games: Optimal Regret Bounds and Convergence to Nash Equilibrium", "Near-Optimal No-Regret Learning for Correlated Equilibria in Multi-Player General-Sum Games"], "answer_arxiv_id": ["2108.06924", "2104.12761", "2111.06008v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11585"} +{"question": "Which work proposes the Octree for the neural SDF representation?", "answer": ["Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D\n Shapes"], "answer_arxiv_id": ["2101.10994"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_11586"} +{"question": "Could you provide me some studies about reweighting approaches based on bilevel optimization that encounters issues similar to IW and RuLSIF?", "answer": ["MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels", "Learning to Reweight Examples for Robust Deep Learning", "Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting"], "answer_arxiv_id": ["1712.05055", "1803.09050", "1902.07379"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11587"} +{"question": "Could you provide me a study in which an extension of FNO that can handle more flexible geometries was proposed?", "answer": ["Fourier Neural Operator with Learned Deformations for PDEs on General Geometries"], "answer_arxiv_id": ["2207.05209"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_11588"} +{"question": "What work combines temperature scaling with input perturbation?", "answer": ["Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks"], "answer_arxiv_id": ["1706.02690"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_11589"} +{"question": "In what papers the current limitation of latent diffusion models in generating high resolution images has been identified?", "answer": ["SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis"], "answer_arxiv_id": ["2307.01952"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_11590"} +{"question": "What are some works that used sequential Monte Carlo learning algorithms for Bayesian inference over structured latent spaces in probabilistic graphical models?", "answer": ["Sequential Monte Carlo for Graphical Models"], "answer_arxiv_id": ["1402.0330"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_11591"} +{"question": "Can you provide papers about the approaches for bias detection and mitigation for vision-only models?", "answer": ["Towards Fairness in Visual Recognition: Effective Strategies for Bias\n Mitigation", "Bias in Pruned Vision Models: In-Depth Analysis and Countermeasures", "Masking Strategies for Background Bias Removal in Computer Vision Models", "A Multidimensional Analysis of Social Biases in Vision Transformers", "A Survey on Bias in Visual Datasets"], "answer_arxiv_id": ["1911.11834", "2304.12622", "2308.12127", "2308.01948", "2107.07919"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_11592"} +{"question": "Which study provided the surrogate loss originated from image representation learning that was utilised by LaGraph?", "answer": ["Noise2Same: Optimizing A Self-Supervised Bound for Image Denoising"], "answer_arxiv_id": ["2010.11971"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_11593"} +{"question": "What papers discuss neural-based constructive heuristics?", "answer": ["Pointer Networks", "Neural Combinatorial Optimization with Reinforcement Learning", "Reinforcement Learning for Solving the Vehicle Routing Problem", "Attention, Learn to Solve Routing Problems!", "POMO: Policy Optimization with Multiple Optima for Reinforcement Learning", "Simulation-guided Beam Search for Neural Combinatorial Optimization", "The Transformer Network for the Traveling Salesman Problem", "Divide and Conquer Networks", "An Efficient Graph Convolutional Network Technique for the Travelling Salesman Problem", "Generalize a Small Pre-trained Model to Arbitrarily Large TSP Instances", "Learning Generalized Reactive Policies using Deep Neural Networks"], "answer_arxiv_id": ["1506.03134", "1611.09940", "1802.04240", "1803.08475", "2010.16011", "2207.06190", "2103.03012", "1611.02401", "1906.01227", "2012.10658", "1708.07280"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_11594"} +{"question": "Which study proposed a notion of list-replicability?", "answer": ["Replicability and stability in learning", "List and Certificate Complexities in Replicable Learning"], "answer_arxiv_id": ["2304.03757", "2304.02240"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11595"} +{"question": "What researches are about using auxiliary losses and frozen pre-trained layers to deal with the generalization problem in RL?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Data-Efficient Reinforcement Learning with Self-Predictive Representations", "Improving Sample Efficiency in Model-Free Reinforcement Learning from Images", "Decoupling Representation Learning from Reinforcement Learning"], "answer_arxiv_id": ["2004.04136", "2007.05929", "1910.01741", "2009.08319"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_11596"} +{"question": "Could you provide me some papers that adopt Hotflip-based algorithms for automatic construction of prompts?", "answer": ["HotFlip: White-Box Adversarial Examples for Text Classification", "Universal Adversarial Triggers for Attacking and Analyzing NLP", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically\n Generated Prompts", "Automatically Auditing Large Language Models via Discrete Optimization"], "answer_arxiv_id": ["1712.06751", "1908.07125", "2010.15980", "2303.04381"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_11597"} +{"question": "What studies have been performed around restricted dynamic regret?", "answer": ["Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient", "Improved Dynamic Regret for Non-degenerate Functions", "Online Forecasting of Total-Variation-bounded Sequences", "Adaptive Online Estimation of Piecewise Polynomial Trends", "An Optimal Reduction of TV-Denoising to Adaptive Online Learning"], "answer_arxiv_id": ["1605.04638", "1608.03933", "1906.03364", "2010.00073", "2101.09438"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_11598"} +{"question": "Which research work introduces the IER method that the author's study comparison?", "answer": ["Meshing Point Clouds with Predicted Intrinsic-Extrinsic Ratio Guidance"], "answer_arxiv_id": ["2007.09267"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_11599"} +{"question": "Which studies have inferred an adversary directly from the value functions?", "answer": ["Robust Adversarial Reinforcement Learning", "Learning with AMIGo: Adversarially Motivated Intrinsic Goals", "Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Design"], "answer_arxiv_id": ["1703.02702", "2006.12122", "2012.02096"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_11600"} +{"question": "What paper discusses the importance of the choice of interpolation scheme in nCDEs?", "answer": ["Neural Controlled Differential Equations for Online Prediction Tasks"], "answer_arxiv_id": ["2106.11028"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_11601"} +{"question": "Which work introduced hyperbolic convolution neural networks?", "answer": ["Hyperbolic Neural Networks++"], "answer_arxiv_id": ["2006.08210"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_11602"} +{"question": "What researchers propose the establishment of vision dictionaries using a large batch size for contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_11603"} +{"question": "What studies employed contrastive learning methods for object detection?", "answer": ["DetCo: Unsupervised Contrastive Learning for Object Detection", "FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding"], "answer_arxiv_id": ["2102.04803", "2103.05950"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_11604"} +{"question": "Which works developed conditional coding in the scope of NVC?", "answer": ["Conditional Entropy Coding for Efficient Video Compression", "Conditional Coding for Flexible Learned Video Compression", "CANF-VC: Conditional Augmented Normalizing Flows for Video Compression", "VCT: A Video Compression Transformer", "Deep Contextual Video Compression", "Temporal Context Mining for Learned Video Compression", "Hybrid Spatial-Temporal Entropy Modelling for Neural Video Compression", "Neural Video Compression with Diverse Contexts"], "answer_arxiv_id": ["2008.09180", "2104.07930", "2207.05315", "2206.07307", "2109.15047", "2111.13850", "2207.05894", "2302.14402"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_11605"} +{"question": "Which study found the dimensional collapse entanglement among server and client models in federated supervised learning?", "answer": ["Towards Understanding and Mitigating Dimensional Collapse in\n Heterogeneous Federated Learning"], "answer_arxiv_id": ["2210.00226"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_11606"} +{"question": "What studies proposed methodologies for solving ODE by training neural networks?", "answer": ["Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2102.09672"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11607"} +{"question": "What is the paper where a text encoder, CLIP, transforms text prompt into a sequence of intermediate representations?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_11608"} +{"question": "Which papers defined the concept of semantic parsing?", "answer": ["Learning to Map Sentences to Logical Form: Structured Classification with Probabilistic Categorial Grammars"], "answer_arxiv_id": ["1207.1420v1"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_11609"} +{"question": "Could you provide me some research indicating challenges with Logical Language Models (LLMs) in terms of generating hallucination and logical inconsistency?", "answer": ["ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoning", "Consistency Analysis of ChatGPT", "A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on\n Reasoning, Hallucination, and Interactivity", "Evaluating the Logical Reasoning Ability of ChatGPT and GPT-4", "Exploring Self-supervised Logic-enhanced Training for Large Language\n Models"], "answer_arxiv_id": ["2212.07919", "2303.06273", "2302.04023", "2304.03439", "2305.13718"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_11610"} +{"question": "Could you provide me the study that proposed the mean of medians estimator for the super heavy-tailed bandit problem?", "answer": ["Breaking the Moments Condition Barrier: No-Regret Algorithm for Bandits with Super Heavy-Tailed Payoffs"], "answer_arxiv_id": ["2110.13876"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_11611"} +{"question": "Could you show me some works that adopted SAVI to improve rate-distortion (R-D) performance in image compression?", "answer": ["Content Adaptive Optimization for Neural Image Compression", "Improving Inference for Neural Image Compression", "Flexible Neural Image Compression via Code Editing"], "answer_arxiv_id": ["1906.01223", "2006.04240", "2209.09244"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_11612"} +{"question": "Which paper adopted a reinforcement learning based method for optimizing the generator in order to maximize the training accuracy?", "answer": ["Learning To Simulate"], "answer_arxiv_id": ["1810.02513"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_11613"} +{"question": "What studies involve the integration of reinforcement learning into object tracking tasks?", "answer": ["End-to-end Active Object Tracking via Reinforcement Learning", "Deep Reinforcement Learning for Visual Object Tracking in Videos"], "answer_arxiv_id": ["1705.10561", "1701.08936"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_11614"} +{"question": "What works are related to layer dropout?", "answer": ["Deep Networks with Stochastic Depth", "A ConvNet for the 2020s", "Reducing Transformer Depth on Demand with Structured Dropout", "Accelerating Training of Transformer-Based Language Models with\n Progressive Layer Dropping"], "answer_arxiv_id": ["1603.09382", "2201.03545", "1909.11556", "2010.13369"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_11615"} +{"question": "Which work considers an online learning framework that goes beyond a constant memory length?", "answer": ["Online learning with dynamics: A minimax perspective"], "answer_arxiv_id": ["2012.01705"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_11616"} +{"question": "Are there studies that examine the use of descriptive prompts in vision-language models?", "answer": ["Elevater: A Benchmark and Toolkit for Evaluating Language-Augmented Visual Models"], "answer_arxiv_id": ["2204.08790"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_11617"} +{"question": "Which work discussed issues regarding vision encoders in the context of end-to-end MLLMs?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_11618"} +{"question": "Are there any works that reduce training data to enhance the efficiency of models?", "answer": ["Deduplicating Training Data Makes Language Models Better", "EfficientTrain: Exploring Generalized Curriculum Learning for Training\n Visual Backbones", "EfficientNetV2: Smaller Models and Faster Training", "Accelerating Vision Transformer Training via a Patch Sampling Schedule"], "answer_arxiv_id": ["2107.06499", "2211.09703", "2104.00298", "2208.09520"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_11619"} +{"question": "What papers discuss extending random crops and horizontal flips in image processing problems by color operations?", "answer": ["Rethinking the Inception Architecture for Computer Vision", "Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning"], "answer_arxiv_id": ["1512.00567", "1602.07261"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_11620"} +{"question": "Which papers are about large video-language models (VLMs)?", "answer": ["ActBERT: Learning Global-Local Video-Text Representations", "Less is More: ClipBERT for Video-and-Language Learning via Sparse Sampling", "Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval", "VIOLET : End-to-End Video-Language Transformers with Masked Visual-token Modeling", "Just Ask: Learning to Answer Questions from Millions of Narrated Videos", "Flamingo: a Visual Language Model for Few-Shot Learning", "Revisiting the “Video” in Video-Language Understanding", "All in One: Exploring Unified Video-Language Pre-training"], "answer_arxiv_id": ["2011.07231", "2102.06183", "2104.00650", "2111.12681", "2012.00451", "2204.14198", "2206.01720", "2203.07303"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_11621"} +{"question": "What works have explored the adaptation of audio-only speech models into audio-visual models?", "answer": ["MixSpeech: Cross-Modality Self-Learning with Audio-Visual Stream Mixup\n for Visual Speech Translation and Recognition", "AVFormer: Injecting Vision into Frozen Speech Models for Zero-Shot\n AV-ASR", "Audio-visual fine-tuning of audio-only ASR models", "Self-supervised Learning with Random-projection Quantizer for Speech\n Recognition"], "answer_arxiv_id": ["2303.05309", "2303.16501", "2312.09369", "2202.01855"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_11622"} +{"question": "Could you name the works that focus on cross-modal media synthesis?", "answer": ["AD-NeRF: Audio Driven Neural Radiance Fields for Talking Head Synthesis", "Foley Music: Learning to Generate Music from Videos", "Visually Indicated Sounds", "Audeo: Audio Generation for a Silent Performance Video"], "answer_arxiv_id": ["2103.11078", "2007.10984", "1512.08512", "2006.14348"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_11623"} +{"question": "Could you provide examples of evaluations created as benchmarks to evaluate Language Learning Models (LLMs)?", "answer": ["Beyond the Imitation Game: Quantifying and extrapolating the\n capabilities of language models", "Holistic Evaluation of Language Models", "INSTRUCTEVAL: Towards Holistic Evaluation of Instruction-Tuned Large\n Language Models"], "answer_arxiv_id": ["2206.04615", "2211.09110", "2306.04757"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_11624"} +{"question": "What study categorized the recent advancements in reference-based SR?", "answer": ["MASA-SR: Matching Acceleration and Spatial Adaptation for\n Reference-Based Image Super-Resolution"], "answer_arxiv_id": ["2106.02299"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_11625"} +{"question": "Can you provide some studies about continuous GANs for effectively learning representations of real-world images?", "answer": ["Adversarial Generation of Continuous Images", "Image Generators with Conditionally-Independent Pixel Synthesis"], "answer_arxiv_id": ["2011.12026", "2011.13775"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_11626"} +{"question": "Could you name some pre-trained transformer models used to embed raw text into a high-dimensional representation?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "XLNet: Generalized Autoregressive Pretraining for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "1906.08237", "1907.11692"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_11627"} +{"question": "What papers classify examples as challenging according to sensitivity to varying model capacity?", "answer": ["What Do Compressed Deep Neural Networks Forget?"], "answer_arxiv_id": ["1911.05248"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_11628"} +{"question": "Could you provide me some studies that involved synthesizing samples with characteristics similar to BC samples for training a debiased model?", "answer": ["BiaSwap: Removing dataset bias with bias-tailored swapping augmentation", "Learning Debiased Representation via Disentangled Feature Augmentation", "SelecMix: Debiased Learning by Contradicting-pair Sampling"], "answer_arxiv_id": ["2108.10008", "2107.01372", "2211.02291"], "source_meta": {"published_time": "20240430"}, "qid": "AutoScholarQuery_train_11629"} +{"question": "Can you provide examples of papers where cross-modal attention was used to capture correlations between different modalities?", "answer": ["Iterative Answer Prediction with Pointer-Augmented Multimodal Transformers for TextVQA", "Use What You Have: Video Retrieval Using Representations From Collaborative Experts", "Unified Multisensory Perception: Weakly-Supervised Audio-Visual Video Parsing", "Learning to Localize Sound Source in Visual Scenes", "Long-range Multimodal Pretraining for Movie Understanding", "MEGA: Multimodal Alignment Aggregation and Distillation For Cinematic Video Segmentation"], "answer_arxiv_id": ["1911.06258", "1907.13487", "2007.10558", "1803.03849v1", "2308.09775", "2308.11185"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_11630"} +{"question": "Could you provide me papers that tackle the unidentifiability of ICA by incorporating supplementary information?", "answer": ["Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learning", "Disentanglement by Nonlinear ICA with General Incompressible-flow Networks (GIN)", "Hidden Markov Nonlinear ICA: Unsupervised Learning from Nonstationary Time Series", "Variational Autoencoders and Nonlinear ICA: A Unifying Framework"], "answer_arxiv_id": ["1805.08651", "2001.04872", "2006.12107", "1907.04809"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_11631"} +{"question": "Which publications explored methods to handle particular challenges, such as large object scale variations in remote sensing images, by using large kernels?", "answer": ["Large Selective Kernel Network for Remote Sensing Object Detection", "A ConvNet for the 2020s", "Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs", "More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using\n Sparsity"], "answer_arxiv_id": ["2303.09030", "2201.03545", "2203.06717", "2207.03620"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_11632"} +{"question": "What studies extend 2D diffusion models to text-guided video generation?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models", "LAVIE: High-Quality Video Generation with Cascaded Latent Diffusion\n Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Video Diffusion Models", "Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "TokenFlow: Consistent Diffusion Features for Consistent Video Editing", "Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation", "MagicProp: Diffusion-based Video Editing via Motion-aware Appearance\n Propagation"], "answer_arxiv_id": ["2304.08818", "2210.02303", "2309.15103", "2209.14792", "2204.03458", "2212.11565", "2307.10373", "2306.07954", "2309.00908"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_11633"} +{"question": "Which researches utilize Language Learning Models to predict labels for unlabeled questions in In-Context Learning?", "answer": ["MoT: Memory-of-Thought Enables ChatGPT to Self-Improve", "Better Zero-Shot Reasoning with Self-Adaptive Prompting", "Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2305.05181", "2305.14106", "2210.03493"], "source_meta": {"published_time": "20240712"}, "qid": "AutoScholarQuery_train_11634"} +{"question": "Which methods have conventionally been used for user intention understanding?", "answer": ["Bayes and Naive Bayes Classifier", "XGBoost: A Scalable Tree Boosting System"], "answer_arxiv_id": ["1404.0933v1", "1603.02754"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_11635"} +{"question": "Could you show me the papers that stabilize training by using mean-teachers or momentum encoders in self-supervised learning?", "answer": ["Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1703.01780", "1911.05722"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_11636"} +{"question": "What study observed that self-supervised Visual Transformers capture more distinct information than supervised models do?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2104.14294"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11637"} +{"question": "What research studies pre-training on unlabeled protein structures for generalizable representations?", "answer": ["Protein Representation Learning by Geometric Structure Pretraining", "Contrastive Representation Learning for 3D Protein Structures", "Structure-aware Protein Self-supervised Learning"], "answer_arxiv_id": ["2203.06125", "2205.15675", "2204.04213"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_11638"} +{"question": "Which studies showed that deep random networks converge to Gaussian processes?", "answer": ["Deep Neural Networks as Gaussian Processes", "Gaussian Process Behaviour in Wide Deep Neural Networks"], "answer_arxiv_id": ["1711.00165", "1804.11271"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_11639"} +{"question": "Which works have been studying the effectiveness of the external context in tasks such as floorplan reconstruction?", "answer": ["Audio-Visual Floorplan Reconstruction"], "answer_arxiv_id": ["2012.15470"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_11640"} +{"question": "Which papers proposed a framework for learning debiased representations?", "answer": ["Can contrastive learning avoid shortcut solutions?"], "answer_arxiv_id": ["2106.11230"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_11641"} +{"question": "What papers developed a connection between neural collapse (NC) and deep neural network (DNN) performance?", "answer": ["Neural Collapse Inspired Attraction-Repulsion-Balanced Loss for Imbalanced Learning", "Linking Neural Collapse and L2 Normalization with Improved Out-of-Distribution Detection in Deep Neural Networks", "Nearest Class-Center Simplification through Intermediate Layers", "Principled and Efficient Transfer Learning of Deep Models via Neural Collapse", "Understanding Imbalanced Semantic Segmentation Through Neural Collapse", "Inducing Neural Collapse to a Fixed Hierarchy-Aware Frame for Reducing Mistake Severity", "No Fear of Classifier Biases: Neural Collapse Inspired Federated Learning with Synthetic and Fixed Classifier"], "answer_arxiv_id": ["2204.08735", "2209.08378", "2201.08924", "2212.12206", "2301.01100", "2303.05689", "2303.10058"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_11642"} +{"question": "What researches use consecutive blurry inputs to represent the latent motion?", "answer": ["Bringing Alive Blurred Moments", "Blur Interpolation Transformer for Real-World Motion from Blur", "Animation from Blur: Multi-modal Blur Decomposition with Motion Guidance"], "answer_arxiv_id": ["1804.02913", "2211.11423", "2207.10123"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_11643"} +{"question": "Can you cite the studies about Parameter-Efficient Fine-Tuning (PEFT)?", "answer": ["Learning multiple visual domains with residual adapters", "LoRA: Low-Rank Adaptation of Large Language Models", "Scaling & Shifting Your Features: A New Baseline for Efficient Model\n Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Visual Prompt Tuning", "AdaptFormer: Adapting Vision Transformers for Scalable Visual\n Recognition", "Scaling & Shifting Your Features: A New Baseline for Efficient Model\n Tuning"], "answer_arxiv_id": ["1705.08045", "2106.09685", "2210.08823", "2101.00190", "2203.12119", "2205.13535", "2210.08823"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_11644"} +{"question": "What research proposed self-instruct methods for generating diverse and abundant instruction data?", "answer": ["Self-Instruct: Aligning Language Models with Self-Generated Instructions", "WizardLM: Empowering Large Language Models to Follow Complex\n Instructions"], "answer_arxiv_id": ["2212.10560", "2304.12244"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_11645"} +{"question": "Which studies considered Hölder smoothness in contextual bandits?", "answer": ["Smooth Contextual Bandits: Bridging the Parametric and Non-differentiable Regret Regimes", "Smoothness-Adaptive Contextual Bandits"], "answer_arxiv_id": ["1909.02553", "1910.09714"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_11646"} +{"question": "Could you provide me some works about gradient approximation in large-scale distributed training?", "answer": ["Sparse Communication for Distributed Gradient Descent", "Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training", "Sparsified SGD with Memory", "Stochastic Backpropagation: A Memory Efficient Strategy for Training Video Models", "Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout"], "answer_arxiv_id": ["1704.05021", "1712.01887", "1809.07599", "2203.16755", "2010.06808"], "source_meta": {"published_time": "20220228"}, "qid": "AutoScholarQuery_train_11647"} +{"question": "What works use 2D-to-3D lifting methods for monocular 3D human pose estimation?", "answer": ["A simple yet effective baseline for 3d human pose estimation", "Exploiting temporal information for 3D pose estimation", "3D human pose estimation in video with temporal convolutions and\n semi-supervised training", "Anatomy-aware 3D Human Pose Estimation with Bone-based Pose\n Decomposition"], "answer_arxiv_id": ["1705.03098", "1711.08585", "1811.11742", "2002.10322"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_11648"} +{"question": "Which papers focus on data re-sampling method to balance the training data in long-tailed image classification?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition", "MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition", "Balanced Meta-Softmax for Long-Tailed Visual Recognition"], "answer_arxiv_id": ["1910.09217", "2103.12579", "2007.10740"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_11649"} +{"question": "What are some representative datasets that use both pressure and visual data to better understand human motion?", "answer": ["3D Human Pose Estimation via Intuitive Physics"], "answer_arxiv_id": ["2303.18246"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_11650"} +{"question": "Which studies speed up the motion mimicking process using DRL framework through hyper-parameter searching and constraint relaxation?", "answer": ["Efficient Hyperparameter Optimization for Physics-based Character Animation", "Learning and Exploring Motor Skills with Spacetime Bounds"], "answer_arxiv_id": ["2104.12365", "2103.16807"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_11651"} +{"question": "Could you provide some studies concerning diverse policy optimization?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function", "Dynamics-Aware Unsupervised Discovery of Skills", "One Solution is Not All You Need: Few-Shot Extrapolation via Structured MaxEnt RL", "Continuously Discovering Novel Strategies via Reward-Switching Policy Optimization", "Novel Policy Seeking with Constrained Optimization", "Diversity-Inducing Policy Gradient: Using Maximum Mean Discrepancy to Find a Set of Diverse Policies", "Differentiable Quality Diversity", "Effective Diversity in Population Based Reinforcement Learning"], "answer_arxiv_id": ["1802.06070", "1907.01657", "2010.14484", "2204.02246", "2005.10696", "1906.00088", "2106.03894", "2002.00632"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_11652"} +{"question": "What are some examples of research projects on single-stage object detectors?", "answer": ["SSD: Single Shot MultiBox Detector", "YOLOv3: An Incremental Improvement", "FCOS: Fully Convolutional One-Stage Object Detection"], "answer_arxiv_id": ["1512.02325", "1804.02767", "1904.01355"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_11653"} +{"question": "Are there any studies focused on integrating the instance-wise property into an additive framework?", "answer": ["Model Agnostic Supervised Local Explanations", "LIMIS: Locally Interpretable Modeling using Instance-wise Subsampling", "Neural Additive Models: Interpretable Machine Learning with Neural Nets"], "answer_arxiv_id": ["1807.02910", "1909.12367", "2004.13912"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_11654"} +{"question": "Are there any studies about improving the worst-group performance?", "answer": ["Learning Models with Uniform Performance via Distributionally Robust Optimization", "Distributionally Robust Language Modeling", "Just Train Twice: Improving Group Robustness without Training Group Information", "DORO: Distributional and Outlier Robust Optimization"], "answer_arxiv_id": ["1810.08750", "1909.02060", "2107.09044", "2106.06142"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_11655"} +{"question": "Which studies utilized PointNet series to extract geometric features from raw point clouds for 3D object detection?", "answer": ["Frustum PointNets for 3D Object Detection from RGB-D Data", "PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud", "3DSSD: Point-based 3D Single Stage Object Detector", "Deep Hough Voting for 3D Object Detection in Point Clouds", "Fast Point R-CNN"], "answer_arxiv_id": ["1711.08488", "1812.04244", "2002.10187", "1904.09664", "1908.02990"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_11656"} +{"question": "In which papers can I find research about inducing structured sparsity over neural network weights to obtain smaller, pruned networks?", "answer": ["Structured Sparsity Inducing Adaptive Optimizers for Deep Learning"], "answer_arxiv_id": ["2102.03869"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_11657"} +{"question": "Could you provide me some studies that propose methods for identifying difficult data instances?", "answer": ["Understanding Black-box Predictions via Influence Functions", "Co-teaching: Robust Training of Deep Neural Networks with Extremely\n Noisy Labels"], "answer_arxiv_id": ["1703.04730", "1804.06872"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_11658"} +{"question": "Which works use neural networks to search for discrete logical expressions using differentiable reparameterizations?", "answer": ["Learning Algorithms via Neural Logic Networks", "Scalable Rule-Based Representation Learning for Interpretable Classification", "Deep Differentiable Logic Gate Networks"], "answer_arxiv_id": ["1904.01554", "2109.15103", "2210.08277"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_11659"} +{"question": "Which papers used the concept of velocity matching in constructing a flow from a learned velocity field?", "answer": ["Neural Variational Gradient Descent", "Probability Flow Solution of the Fokker-Planck Equation", "Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow", "Flow Matching for Generative Modeling", "Building Normalizing Flows with Stochastic Interpolants"], "answer_arxiv_id": ["2107.10731v2", "2206.04642", "2209.03003", "2210.02747", "2209.15571"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_11660"} +{"question": "Can you cite works that proposed federated bilevel optimization?", "answer": ["Local Stochastic Bilevel Optimization with Momentum-Based Variance Reduction", "FedNest: Federated Bilevel, Minimax, and Compositional Optimization", "Achieving Linear Speedup in Non-IID Federated Bilevel Learning", "Communication-Efficient Federated Hypergradient Computation via Aggregated Iterative Differentiation", "Fast Adaptive Federated Bilevel Optimization", "Communication-Efficient Federated Bilevel Optimization with Local and Global Lower Level Problems"], "answer_arxiv_id": ["2205.01608", "2205.02215v3", "2302.05412", "2302.04969", "2211.01122", "2302.06701"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11661"} +{"question": "Can you point me to a piece of research that discusses properties of the selected solutions for a given initialization in an over-parameterized setting?", "answer": ["In Search of the Real Inductive Bias: On the Role of Implicit Regularization in Deep Learning"], "answer_arxiv_id": ["1412.6614"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_11662"} +{"question": "Which papers utilise dynamic convolution or attention mechanisms for keypoint localization in pose estimation?", "answer": ["FCPose: Fully Convolutional Multi-Person Pose Estimation with Dynamic\n Instance-Aware Convolutions", "InsPose: Instance-Aware Networks for Single-Stage Multi-Person Pose\n Estimation"], "answer_arxiv_id": ["2105.14185", "2107.08982"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_11663"} +{"question": "What previous works were carried out on General contrastive learning, especially those that introduced the idea of aligning the embeddings of two augmented views from the same image?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1911.05722", "2003.04297", "2002.05709"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_11664"} +{"question": "What studies attempted to identify the bias attribute of the training dataset without human supervision?", "answer": ["Discover the Unknown Biased Attribute of an Image Classifier", "Explaining in Style: Training a GAN to explain a classifier in StyleSpace", "UDIS: Unsupervised Discovery of Bias in Deep Visual Recognition Models"], "answer_arxiv_id": ["2104.14556", "2104.13369", "2110.15499"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_11665"} +{"question": "What studies have used a semi-automated technique for instruction datasets fine-tuning?", "answer": ["WizardLM: Empowering Large Language Models to Follow Complex\n Instructions", "A Preliminary Study of the Intrinsic Relationship between Complexity and\n Alignment"], "answer_arxiv_id": ["2304.12244", "2308.05696"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_11666"} +{"question": "Which studies proposed early methods of test-time adaptation by updating batch normalization layers?", "answer": ["Improving robustness against common corruptions by covariate shift adaptation", "Tent: Fully Test-Time Adaptation by Entropy Minimization"], "answer_arxiv_id": ["2006.16971v2", "2006.10726"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_11667"} +{"question": "Any publications that investigated the lower complexity bounds in minimax optimization?", "answer": ["The Complexity of Nonconvex-Strongly-Concave Minimax Optimization", "Complexity Lower Bounds for Nonconvex-Strongly-Concave Min-Max Optimization", "On Lower Iteration Complexity Bounds for the Convex Concave Saddle Point Problems"], "answer_arxiv_id": ["2103.15888", "2104.08708", "1912.07481"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_11668"} +{"question": "Are there any existing metrics like Inception Score (IS) and Frechet Inception Distance (FID) that evaluate the fidelity of synthesized images in the context of text-to-image generation?", "answer": ["Improved Techniques for Training GANs", "GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium"], "answer_arxiv_id": ["1606.03498", "1706.08500"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_11669"} +{"question": "What works have examined greedy feature selection algorithms from a theoretical perspective, particularly in terms of adaptive submodularity?", "answer": ["Submodular meets Spectral: Greedy Algorithms for Subset Selection, Sparse Approximation and Dictionary Selection", "Restricted Strong Convexity Implies Weak Submodularity", "Adaptive Submodularity: Theory and Applications in Active Learning and Stochastic Optimization"], "answer_arxiv_id": ["1102.3975", "1612.00804v2", "1003.3967"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_11670"} +{"question": "Could you provide me some papers that indicated that large-scale pre-trained language models can internalize a sort of implicit “knowledge base”?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "XLNet: Generalized Autoregressive Pretraining for Language Understanding", "Language Models as Knowledge Bases?", "How Can We Know What Language Models Know?", "oLMpics - On what Language Model Pre-training Captures"], "answer_arxiv_id": ["1810.04805", "1906.08237", "1909.01066", "1911.12543", "1912.13283"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_11671"} +{"question": "Which study introduces the Random Deep Vortex Network (RDVN) method for solving the 2D Navier-Stokes equations?", "answer": ["DRVN (Deep Random Vortex Network): A new physics-informed machine learning method for simulating and inferring incompressible fluid flows"], "answer_arxiv_id": ["2206.09571"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_11672"} +{"question": "Can you provide a research paper that first proposed to recover humans from an entire frame in HPS?", "answer": ["Monocular, One-stage, Regression of Multiple 3D People"], "answer_arxiv_id": ["2008.12272"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_11673"} +{"question": "What are some works that examined personalized federated learning from the perspective of meta-learning?", "answer": ["Personalized Federated Learning: A Meta-Learning Approach", "Improving Federated Learning Personalization via Model Agnostic Meta Learning", "Adaptive Gradient-Based Meta-Learning Methods"], "answer_arxiv_id": ["2002.07948", "1909.12488", "1906.02717"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_11674"} +{"question": "What studies focus on curriculum learning to improve sample efficiency, encourage exploration, and solve complex multi-stage tasks?", "answer": ["Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey"], "answer_arxiv_id": ["2003.04960"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_11675"} +{"question": "Which papers demonstrates the application of Large Language Models (LLMs) in tasks such as code generation, open-domain question answering, and multilingual translation?", "answer": ["Code Llama: Open Foundation Models for Code", "Retrieving and Reading: A Comprehensive Survey on Open-domain Question\n Answering", "PaLM: Scaling Language Modeling with Pathways", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2308.12950", "2101.00774", "2204.02311", "2005.14165"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_11676"} +{"question": "Could you provide me with some recent examples of works proposing polynomial-time algorithms for DAG learning?", "answer": ["Learning Quadratic Variance Function (QVF) DAG models via OverDispersion Scoring (ODS)", "Learning Identifiable Gaussian Bayesian Networks in Polynomial Time and Sample Complexity", "Learning linear structural equation models in polynomial time and sample complexity", "A polynomial-time algorithm for learning nonparametric causal graphs", "Structure learning in polynomial time: Greedy algorithms, Bregman information, and exponential families"], "answer_arxiv_id": ["1704.08783v1", "1703.01196", "1707.04673", "2006.11970", "2110.04719"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_11677"} +{"question": "Which works discussed the performance of video-text retrieval models that outperformed existing models pre-trained on video data?", "answer": ["VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding", "Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval", "Advancing High-Resolution Video-Language Representation with Large-Scale Video Transcriptions", "Bridging Video-text Retrieval with Multiple Choice Questions", "All in One: Exploring Unified Video-Language Pre-training"], "answer_arxiv_id": ["2109.14084", "2104.00650", "2111.10337", "2201.04850", "2203.07303"], "source_meta": {"published_time": "20220914"}, "qid": "AutoScholarQuery_train_11678"} +{"question": "Which papers focus on the unconstrained and convex setting in the context of non-smooth optimization?", "answer": ["Tight analyses for non-smooth stochastic gradient descent", "Making SGD Parameter-Free"], "answer_arxiv_id": ["1812.05217", "2205.02160"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_11679"} +{"question": "What are the studies that contain sequential actions but do not cover all types of non-sequential actions?", "answer": ["PET: An Annotated Dataset for Process Extraction from Natural Language\n Text", "Knowing-how & Knowing-that: A New Task for Machine Comprehension of User Manuals"], "answer_arxiv_id": ["2203.04860", "2306.04187v2"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_11680"} +{"question": "What benchmarks have been proposed to evaluate compositionality by binding attributes to and specifying relations between objects?", "answer": ["Benchmarking Spatial Relationships in Text-to-Image Generation", "HRS-Bench: Holistic, Reliable and Scalable Benchmark for Text-to-Image\n Models", "T2I-CompBench: A Comprehensive Benchmark for Open-world Compositional\n Text-to-image Generation", "DALL-Eval: Probing the Reasoning Skills and Social Biases of\n Text-to-Image Generation Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2212.10015", "2304.05390", "2307.06350", "2202.04053", "2205.11487"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_11681"} +{"question": "What research papers provide valuable reference in the history of RL research applied to NLP?", "answer": ["Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation", "Bandit Structured Prediction for Learning from Partial Feedback in Statistical Machine Translation", "Revisiting the Weaknesses of Reinforcement Learning for Neural Machine Translation", "A Deep Reinforced Model for Abstractive Summarization", "Reinforcement Learning for Bandit Neural Machine Translation with Simulated Human Feedback", "A Reinforcement Learning Approach to Interactive-Predictive Neural Machine Translation", "Can Neural Machine Translation be Improved with User Feedback?", "Is Reinforcement Learning (Not) for Natural Language Processing: Benchmarks, Baselines, and Building Blocks for Natural Language Policy Optimization", "Offline RL for Natural Language Generation with Implicit Language Q Learning"], "answer_arxiv_id": ["1609.08144", "1601.04468", "2106.08942", "1705.04304", "1707.07402", "1805.01553", "1804.05958", "2210.01241", "2206.11871"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_11682"} +{"question": "Which papers introduce methods like GAIL, AIRL, or VICE that assign rewards in order to deal with high-dimensional state and action spaces?", "answer": ["Generative Adversarial Imitation Learning", "Learning Robust Rewards with Adversarial Inverse Reinforcement Learning", "Variational Inverse Control with Events: A General Framework for Data-Driven Reward Definition"], "answer_arxiv_id": ["1606.03476", "1710.11248", "1805.11686"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_11683"} +{"question": "What researches focus on text-to-video synthesis?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Structure and Content-Guided Video Synthesis with Diffusion Models", "Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators"], "answer_arxiv_id": ["2210.02303", "2209.14792", "2302.03011", "2212.11565", "2303.13439"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_11684"} +{"question": "What research papers discussed about the methods prefer an end-to-end feature selection for time series modeling?", "answer": ["Distributed and parallel time series feature extraction for industrial big data applications"], "answer_arxiv_id": ["1610.07717v3"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11685"} +{"question": "In what papers can I find applications of DPMs in speech synthesis?", "answer": ["DiffWave: A Versatile Diffusion Model for Audio Synthesis"], "answer_arxiv_id": ["2009.09761v3"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_11686"} +{"question": "Could you mention some works that employs semi-supervised learning for handling noisy annotated data?", "answer": ["A Semi-Supervised Two-Stage Approach to Learning from Noisy Labels", "DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "MixMatch: A Holistic Approach to Semi-Supervised Learning"], "answer_arxiv_id": ["1802.02679", "2002.07394", "1905.02249"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_11687"} +{"question": "What studies set a new benchmark in robust zero-shot depth estimation using multi-source mixed data training?", "answer": ["Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer"], "answer_arxiv_id": ["1907.01341v3"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_11688"} +{"question": "What papers point out the importance of using strong data augmentations when it comes to SSL generalization?", "answer": ["What Makes for Good Views for Contrastive Learning?", "Contrastive Learning with Stronger Augmentations"], "answer_arxiv_id": ["2005.10243", "2104.07713"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_11689"} +{"question": "Could you provide me some works that have limited ability to generalize to high-resolution RGB images and capture long-range correlations?", "answer": ["Structured Uncertainty Prediction Networks", "Estimating High Order Gradients of the Data Distribution by Denoising"], "answer_arxiv_id": ["1802.07079", "2111.04726"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_11690"} +{"question": "Which works focus on 3D reconstruction from set of posed camera images using NeRF?", "answer": ["Nerfies: Deformable Neural Radiance Fields", "​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2011.12948", "2103.13415", "2003.08934"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_11691"} +{"question": "Could you give me examples of research that proposed to use successive halving for multi-fidelity hyperparameter optimization?", "answer": ["Non-stochastic Best Arm Identification and Hyperparameter Optimization"], "answer_arxiv_id": ["1502.07943v1"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_train_11692"} +{"question": "What documents refer to the use of robust mean estimators in linear bandits?", "answer": ["Almost Optimal Algorithms for Linear Stochastic Bandits with Heavy-Tailed Payoffs", "Nearly Optimal Regret for Stochastic Linear Bandits with Heavy-Tailed Payoffs", "Breaking the Moments Condition Barrier: No-Regret Algorithm for Bandits with Super Heavy-Tailed Payoffs"], "answer_arxiv_id": ["1810.10895", "2004.13465", "2110.13876"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_11693"} +{"question": "What are some studies that explored task relationships in Auxiliary-Task Learning (ATL) by grouping positively related tasks together and assigning unrelated tasks to different groups?", "answer": ["Taskonomy: Disentangling Task Transfer Learning", "Efficiently Identifying Task Groupings for Multi-Task Learning"], "answer_arxiv_id": ["1804.08328", "2109.04617"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11694"} +{"question": "Could you cite studies that explored the role of scaling model size in enhancing the calibration of LLMs?", "answer": ["Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "Emergent Abilities of Large Language Models"], "answer_arxiv_id": ["2112.11446", "2206.07682"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_11695"} +{"question": "What research integrates a trainable copy of the encoder to integrate control signals?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_11696"} +{"question": "Which papers present work on graph neural network models for climate modeling?", "answer": ["Forecasting Global Weather with Graph Neural Networks", "GraphCast: Learning skillful medium-range global weather forecasting"], "answer_arxiv_id": ["2202.07575", "2212.12794"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_11697"} +{"question": "Which paper proposed to learn a single-step student model from the output of the original model to reduce the sampling time of diffusion models?", "answer": ["Knowledge Distillation in Iterative Generative Models for Improved\n Sampling Speed"], "answer_arxiv_id": ["2101.02388"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_11698"} +{"question": "Can you suggest any papers that researched over the applications for hashing methods in machine learning, specifically for feature selection?", "answer": ["MISSION: Ultra Large-Scale Feature Selection using Count-Sketches"], "answer_arxiv_id": ["1806.04310"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_11699"} +{"question": "What work can be referenced as a continuation to OpenAI's fill-in-the-middle method which allows flexible generation at arbitrary positions?", "answer": ["FiLM: Fill-in Language Models for Any-Order Generation"], "answer_arxiv_id": ["2310.09930"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_11700"} +{"question": "Which study established the concept of the nearest class-center (NCC) separability?", "answer": ["On the Implicit Bias Towards Minimal Depth of Deep Neural Networks"], "answer_arxiv_id": ["2202.09028"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11701"} +{"question": "Can you point to any papers on the topic of conducting concentrability coefficients by Bellman error?", "answer": ["Bellman-consistent Pessimism for Offline Reinforcement Learning"], "answer_arxiv_id": ["2106.06926"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_11702"} +{"question": "What papers have examined contraction rates in the posterior?", "answer": ["Posterior Variance Analysis of Gaussian Processes with Application to Average Learning Curves"], "answer_arxiv_id": ["1906.01404"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_11703"} +{"question": "What study is most relevant to this paper that discusses how to decompose the Laplacian to obtain the associated eigenfunctions?", "answer": ["The Laplacian in RL: Learning Representations with Efficient Approximations"], "answer_arxiv_id": ["1810.04586"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_11704"} +{"question": "Which works propose to overcome the limits of single view by using multiple views of the material?", "answer": ["Flexible SVBRDF Capture with a Multi-Image Deep Network", "Deep 3D Capture: Geometry and Reflectance from Sparse Multi-View Images"], "answer_arxiv_id": ["1906.11557", "2003.12642"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_11705"} +{"question": "Any works about applying prompt tuning in multimodal VOT?", "answer": ["Visual Prompt Multi-Modal Tracking"], "answer_arxiv_id": ["2303.10826"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_11706"} +{"question": "Which research provided the most suitable and tightest prior framework for analyzing lower bound complexities?", "answer": ["Graph Oracle Models, Lower Bounds, and Gaps for Parallel Stochastic Optimization"], "answer_arxiv_id": ["1805.10222"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_11707"} +{"question": "What papers have the 512x512 and 2K versions of the materials been published?", "answer": ["Flexible SVBRDF Capture with a Multi-Image Deep Network", "Guided Fine-Tuning for Large-Scale Material Transfer"], "answer_arxiv_id": ["1906.11557", "2007.03059"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_11708"} +{"question": "Can you name research works that explore the use of graph-structured data for image synthesis?", "answer": ["Image Generation from Scene Graphs", "Learning Canonical Representations for Scene Graph to Image Generation"], "answer_arxiv_id": ["1804.01622", "1912.07414"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_11709"} +{"question": "Which papers are key contributions in the field of Neural Radiance Fields, or NeRF?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "TensoRF: Tensorial Radiance Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2201.05989", "2203.09517", "2111.12077", "2103.13415", "2304.06706"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_11710"} +{"question": "Which are the works that apply the non-autoregressive model to other sequence-to-sequence tasks?", "answer": ["EMOVIE: A Mandarin Emotion Speech Dataset with a Simple Emotional Text-to-Speech Model", "GeneFace: Generalized and High-Fidelity Audio-Driven 3D Talking Face Synthesis", "Diffsound: Discrete Diffusion Model for Text-to-sound Generation"], "answer_arxiv_id": ["2106.09317", "2301.13430", "2207.09983"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_11711"} +{"question": "Which works employed FedAvg as a benchmark for FL?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "Achieving ​ Linear ​ Speedup ​ with ​​ Partial ​​ Worker ​​ Participation ​​ in ​​ Non-IID ​​ Federated ​​ Learning"], "answer_arxiv_id": ["1602.05629", "2101.11203"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_11712"} +{"question": "Any studies integrating language-guided priors into the novel view synthesis diffusion model?", "answer": ["NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with\n 360{\\deg} Views"], "answer_arxiv_id": ["2211.16431"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_11713"} +{"question": "Which dataset covers programming tasks in ten languages?", "answer": ["Multi-lingual Evaluation of Code Generation Models"], "answer_arxiv_id": ["2210.14868"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_11714"} +{"question": "Could you tell me what studies have been conducted on the construction of a universal object-detection model?", "answer": ["Universal-RCNN: Universal Object Detector via Transferable Graph R-CNN", "Simple Multi-dataset Detection", "Towards Universal Object Detection by Domain Attention"], "answer_arxiv_id": ["2002.07417", "2102.13086", "1904.04402"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_11715"} +{"question": "Could you provide me some studies about Plug-and-Play (PnP) methods for solving a variety of IR tasks?", "answer": ["The Little Engine that Could Regularization by Denoising (RED)", "Learning Deep CNN Denoiser Prior for Image Restoration", "Plug-and-Play Image Restoration with Deep Denoiser Prior", "Plug-and-Play Methods Provably Converge with Properly Trained Denoisers", "Scalable Plug-and-Play ADMM with Convergence Guarantees"], "answer_arxiv_id": ["1611.02862", "1704.03264", "2008.13751", "1905.05406", "2006.03224"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_11716"} +{"question": "Could you provide me studies about detecting highly compressed deepfakes?", "answer": ["ADD: Frequency Attention and Multi-View based Knowledge Distillation to\n Detect Low-Quality Compressed Deepfake Images"], "answer_arxiv_id": ["2112.03553"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_11717"} +{"question": "Can you name some works that model human scene interaction from a single image?", "answer": ["Resolving 3D Human Pose Ambiguities with 3D Scene Constraints", "MoCapDeform: Monocular 3D Human Motion Capture in Deformable Scenes", "HULC: 3D Human Motion Capture with Pose Manifold Sampling and Dense\n Contact Guidance", "Physically Plausible 3D Human-Scene Reconstruction from Monocular RGB\n Image using an Adversarial Learning Approach"], "answer_arxiv_id": ["1908.06963", "2208.08439", "2205.05677", "2307.14570"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_11718"} +{"question": "What works propose using latent diffusion model (LDM) to solve the computational resources problem of diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_11719"} +{"question": "Which studies leverage different kinds of orthogonal polynomials to approximate arbitrary filters?", "answer": ["Adaptive Universal Generalized PageRank Graph Neural Network", "Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering", "Semi-Supervised Classification with Graph Convolutional Networks", "Convolutional Neural Networks on Graphs with Chebyshev Approximation, Revisited", "BernNet: Learning Arbitrary Graph Spectral Filters via Bernstein Approximation", "How Powerful are Spectral Graph Neural Networks"], "answer_arxiv_id": ["2006.07988", "1606.09375", "1609.02907", "2202.03580", "2106.10994", "2205.11172"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_11720"} +{"question": "What paper developed a theory of unbounded, nonsmooth Kurdyka-Lojasiewicz inequalities to prove a stronger result of directional convergence of the parameters?", "answer": ["Directional convergence and alignment in deep learning"], "answer_arxiv_id": ["2006.06657"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_11721"} +{"question": "Which papers proposed report-supervised learning that automatically acquires supervision from free-text radiology reports?", "answer": ["Contrastive Learning of Medical Visual Representations from Paired Images and Text", "Multimodal Representation Learning via Maximization of Local Mutual Information", "Generalized Radiograph Representation Learning via Cross-supervision between Images and Free-text Radiology Reports", "Making the Most of Text Semantics to Improve Biomedical Vision–Language Processing"], "answer_arxiv_id": ["2010.00747", "2103.04537", "2111.03452", "2204.09817"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11722"} +{"question": "Can you list some of the studies that use logit-based methods in knowledge distillation algorithms?", "answer": ["Online Knowledge Distillation with Diverse Peers", "Distilling the Knowledge in a Neural Network", "Online Knowledge Distillation via Multi-branch Diversity Enhancement", "Improved Knowledge Distillation via Teacher Assistant", "Deep Mutual Learning", "Decoupled Knowledge Distillation", "Do Not Blindly Imitate the Teacher: Using Perturbed Loss for Knowledge\n Distillation"], "answer_arxiv_id": ["1912.00350", "1503.02531", "2010.00795", "1902.03393", "1706.00384", "2203.08679", "2305.05010"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_11723"} +{"question": "What work used the NTK formulation of wide neural networks?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_11724"} +{"question": "Could you mention a study that adopts a memory mechanism to perform object-object context reasoning?", "answer": ["Spatial Memory for Context Reasoning in Object Detection"], "answer_arxiv_id": ["1704.04224"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_11725"} +{"question": "Could you provide me some references about Shapley-based data valuation methods?", "answer": ["Data Shapley: Equitable Valuation of Data for Machine Learning", "Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms"], "answer_arxiv_id": ["1904.02868v2", "1908.08619"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_11726"} +{"question": "What papers showed that Behavior Cloning performs effectively when combined with a substantial number of high-quality demonstrations?", "answer": ["BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning", "Bridge Data: Boosting Generalization of Robotic Skills with Cross-Domain\n Datasets"], "answer_arxiv_id": ["2202.02005", "2109.13396"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_11727"} +{"question": "Could you list the studies that have applied contrastive learning to large-scale EEG data?", "answer": ["BENDR: using transformers and a contrastive self-supervised learning task to learn from massive amounts of EEG data."], "answer_arxiv_id": ["2101.12037"], "source_meta": {"published_time": "20230812"}, "qid": "AutoScholarQuery_train_11728"} +{"question": "Which works focus on the satisfaction of the IGM principle which is significant for value factorization in MARL?", "answer": ["QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning", "Qatten: A General Framework for Cooperative Multiagent Reinforcement Learning", "Randomized Entity-wise Factorization for Multi-Agent Reinforcement Learning", "QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement learning", "QPLEX: Duplex Dueling Multi-Agent Q-Learning"], "answer_arxiv_id": ["1803.11485", "2002.03939", "2006.04222", "1905.05408", "2008.01062"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_11729"} +{"question": "What studies incorporated Langevin dynamics and score matching methods into the diffusion model?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_11730"} +{"question": "What research discuss leveraging semantic masks derived from segment models for visual marking?", "answer": ["Segment Anything", "CPT: Colorful Prompt Tuning for Pre-trained Vision-Language Models"], "answer_arxiv_id": ["2401.14159", "2109.11797"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_11731"} +{"question": "Which works can provide examples about Hierarchical IL with Discrete Search?", "answer": ["Subgoal Search For Complex Reasoning Tasks", "Fast and Precise: Adjusting Planning Horizon with Adaptive Subgoal Search"], "answer_arxiv_id": ["2108.11204", "2206.00702"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11732"} +{"question": "Which research works proposed interaction metrics based on game theory?", "answer": ["The Shapley Taylor Interaction Index", "Faith-Shap: The Faithful Shapley Interaction Index"], "answer_arxiv_id": ["1902.05622", "2203.00870"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_11733"} +{"question": "Which papers have developed advanced distillation techniques that aim to enforce greater consistency between the teacher’s and the student’s predictions?", "answer": ["Variational Information Distillation for Knowledge Transfer", "Learning Student Networks via Feature Embedding", "Wasserstein Contrastive Representation Distillation", "Subclass Distillation", "Contrastive Representation Distillation"], "answer_arxiv_id": ["1904.05835", "1812.06597", "2012.08674", "2002.03936", "1910.10699"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_11734"} +{"question": "Are there any studies on weight re-initialization or mask-reinitialization in the context of neural network pruning?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Stabilizing the Lottery Ticket Hypothesis", "Cyclical Pruning for Sparse Neural Networks"], "answer_arxiv_id": ["1803.03635", "1903.01611", "2202.01290"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_11735"} +{"question": "Which work mentioned the use of the GradientNorm heuristic to determine which layer to fine-tune?", "answer": ["Surgical Fine-Tuning Improves Adaptation to Distribution Shifts"], "answer_arxiv_id": ["2210.11466"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_11736"} +{"question": "Are there any studies that analyzed NTK for fully connected (FC) architectures and their associated Reproducing Kernel Hilbert Spaces (RKHS)?", "answer": ["Deep Equals Shallow for ReLU Networks in Kernel Regimes", "On the Inductive Bias of Neural Tangent Kernels", "On the Similarity between the Laplace and Neural Tangent Kernels", "Deep Neural Tangent Kernel and Laplace Kernel Have the Same RKHS"], "answer_arxiv_id": ["2009.14397", "1905.12173", "2007.01580", "2009.10683"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_11737"} +{"question": "Which papers introduced large language models that incorporate multimodal learning?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "InternLM-XComposer: A Vision-Language Large Model for Advanced\n Text-image Comprehension and Composition"], "answer_arxiv_id": ["2302.13971", "2307.09288", "2304.08485", "2304.10592", "2305.06500", "2306.15195", "2309.15112"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_11738"} +{"question": "From which work the MGSM dataset was translated?", "answer": ["Training Verifiers to Solve Math Word Problems"], "answer_arxiv_id": ["2110.14168"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_11739"} +{"question": "Could you point to a work that provided self-supervised representations for sequential data using more complex temporal augmentations?", "answer": ["TS2Vec: Towards Universal Representation of Time Series"], "answer_arxiv_id": ["2106.10466"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_11740"} +{"question": "Which research paper proposed the design of Fourier Neural Operator which uses Fourier space convolution?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations"], "answer_arxiv_id": ["2010.08895"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_11741"} +{"question": "Could you provide a study that explores the prompt tuning on discrete prompts and how these prompts can be shared among models?", "answer": ["RLPrompt: Optimizing Discrete Text Prompts with Reinforcement Learning"], "answer_arxiv_id": ["2205.12548"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_11742"} +{"question": "What studies focused on introducing text conditions into the diffusion process through the use of an unclassified guide?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2112.10741"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_11743"} +{"question": "Any works about infinite-horizon non-episodic RL with provable guarantees?", "answer": ["Learning Infinite-horizon Average-reward MDPs with Linear Function Approximation", "Model-free Reinforcement Learning in Infinite-horizon Average-reward Markov Decision Processes", "Q-learning with UCB Exploration is Sample Efficient for Infinite-Horizon MDP"], "answer_arxiv_id": ["2007.11849v2", "1910.07072", "1901.09311"], "source_meta": {"published_time": "20230106"}, "qid": "AutoScholarQuery_train_11744"} +{"question": "What works discussed the empirical-loss-based reweighting strategy for handling sample reweighting?", "answer": ["AdaLoss: A computationally-efficient and provably convergent adaptive gradient method"], "answer_arxiv_id": ["2109.08282"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_11745"} +{"question": "Could you provide me some VAE-based methods that focus on learning an aligned space between motion and language?", "answer": ["TEMOS: Generating diverse human motions from textual descriptions", "TMR: Text-to-Motion Retrieval Using Contrastive 3D Human Motion\n Synthesis"], "answer_arxiv_id": ["2204.14109", "2305.00976"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_11746"} +{"question": "What works utilize transformer models with masking strategies for sequential recommendation tasks?", "answer": ["BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer"], "answer_arxiv_id": ["1904.06690v2"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_11747"} +{"question": "Could you provide me some studies about the effectiveness of spectrograms autoencoder with reconstruction objective in self-supervision?", "answer": ["SSAST: Self-Supervised Audio Spectrogram Transformer", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2110.09784", "2111.06377"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11748"} +{"question": "What papers replaced Neural-ODE with an MLP and learned local spatio-temporal codes to capture shape and deformations?", "answer": ["Learning Parallel Dense Correspondence from Spatio-Temporal Descriptors\n for Efficient and Robust 4D Reconstruction"], "answer_arxiv_id": ["2103.16341"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_11749"} +{"question": "What researches introduced and used masked language modeling?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_11750"} +{"question": "Which works used cross-attention between a set of activations and learnable latent parameters?", "answer": ["Perceiver: General Perception with Iterative Attention", "End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2103.03206", "2005.12872", "2010.04159"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_11751"} +{"question": "What research used volume rendering to recover material and lighting in image collections under different lighting conditions?", "answer": ["NeRF for Outdoor Scene Relighting", "Neural Radiance Transfer Fields for Relightable Novel-view Synthesis with Global Illumination"], "answer_arxiv_id": ["2112.05140", "2207.13607"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11752"} +{"question": "Which papers have proposed strategies for knowledge transfer from pre-trained 2D foundation models to 3D representations at the object level?", "answer": ["CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D Recognition", "LidarCLIP or: How I Learned to Talk to Point Clouds", "PointCLIP: Point Cloud Understanding by CLIP", "CLIP-FO3D: Learning Free Open-world 3D Scene Representations from 2D Dense CLIP", "OpenScene: 3D Scene Understanding with Open Vocabularies"], "answer_arxiv_id": ["2303.11313", "2212.06858", "2112.02413", "2303.04748", "2211.15654"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_11753"} +{"question": "What works have supplemented sparse-voxel backbones with other operations or informations?", "answer": ["(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection\n for Sparse Semantic Segmentation Network", "RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR\n Point Cloud Segmentation", "2DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Clouds", "Point-to-Voxel Knowledge Distillation for LiDAR Semantic Segmentation"], "answer_arxiv_id": ["2102.04530", "2103.12978", "2207.04397", "2206.02099"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_11754"} +{"question": "Which works have proposed single-policy MORL approaches, transforming a multi-objective problem into a single-objective problem?", "answer": ["A Survey of Multi-Objective Sequential Decision-Making", "A Distributional View on Multi-Objective Policy Optimization"], "answer_arxiv_id": ["1402.0590", "2005.07513"], "source_meta": {"published_time": "20220816"}, "qid": "AutoScholarQuery_train_11755"} +{"question": "What papers include discrete prompts for parameter-efficient transfer learning?", "answer": ["Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference", "Making Pre-trained Language Models Better Few-shot Learners", "Coherence boosting: When your pretrained language model is not paying enough attention"], "answer_arxiv_id": ["2001.07676", "2012.15723", "2110.08294v2"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_11756"} +{"question": "Could you provide me some works about learning shared audio-visual representation?", "answer": ["Learning Sight from Sound: Ambient Sound Provides Supervision for Visual\n Learning", "Labelling unlabelled videos from scratch with multi-modal\n self-supervision", "Audio-Visual Instance Discrimination with Cross-Modal Agreement", "Mix and Localize: Localizing Sound Sources in Mixtures", "iQuery: Instruments as Queries for Audio-Visual Sound Separation", "AVA-AVD: Audio-Visual Speaker Diarization in the Wild", "Conditional Generation of Audio from Video via Foley Analogies", "Sound to Visual Scene Generation by Audio-to-Visual Latent Alignment", "Self-Supervised Video Forensics by Audio-Visual Anomaly Detection"], "answer_arxiv_id": ["1712.07271", "2006.13662", "2004.12943", "2211.15058", "2212.03814", "2111.14448", "2304.08490", "2303.17490", "2301.01767"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_11757"} +{"question": "Can you mention some work that tried to extend the original ViT with hierarchical architectures?", "answer": ["Swin Transformer V2: Scaling Up Capacity and Resolution"], "answer_arxiv_id": ["2111.09883"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_11758"} +{"question": "Could you tell me about the research that works on scaling diffusion models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2112.10752", "2205.11487"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_11759"} +{"question": "Which papers studied linear MDPs where the transition probability function and reward function are linear in some feature mapping over state-action pairs?", "answer": ["Provably Efficient Reinforcement Learning with Linear Function Approximation", "Optimism in Reinforcement Learning with Generalized Linear Function Approximation", "A Unifying View of Optimism in Episodic Reinforcement Learning"], "answer_arxiv_id": ["1907.05388", "1912.04136", "2007.01891"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_11760"} +{"question": "What studies have mainly focused on training data teaching?", "answer": ["Learning to Teach", "Iterative Machine Teaching", "Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting", "Learning to Reweight Examples for Robust Deep Learning", "Learning to Reweight with Deep Interactions"], "answer_arxiv_id": ["1805.03643", "1705.10470", "1902.07379", "1803.09050", "2007.04649v2"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_11761"} +{"question": "In what papers is Federated Learning first introduced?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized\n Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_11762"} +{"question": "Could you list the studies that have used pseudo labeling strategy for node classification tasks in the context of graph data?", "answer": ["Multi-Stage Self-Supervised Learning for Graph Convolutional Networks on Graphs with Few Labeled Nodes", "Be Confident! Towards Trustworthy Graph Neural Networks via Confidence Calibration", "Confidence May Cheat: Self-Training on Graph Neural Networks under Distribution Shift"], "answer_arxiv_id": ["1902.11038", "2109.14285", "2201.11349"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_11763"} +{"question": "Which works represent autoregressive models that can generate videos of varying lengths?", "answer": ["Phenaki: Variable Length Video Generation From Open Domain Textual\n Description"], "answer_arxiv_id": ["2210.02399"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_11764"} +{"question": "Which papers studied the problem of improved compositional generalization in the area of language?", "answer": ["Compositional generalization in a deep seq2seq model by separating syntax and semantics", "Learning to Recombine and Resample Data for Compositional Generalization", "Exploiting Semantics in Neural Machine Translation with Graph Convolutional Networks"], "answer_arxiv_id": ["1904.09708", "2010.03706", "1804.08313"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_11765"} +{"question": "Could you provide me some prior studies on robustness data for text-to-SQL that considered semantic-preserving perturbations?", "answer": ["Towards Robustness of Text-to-SQL Models against Synonym Substitution", "Exploring Underexplored Limitations of Cross-Domain Text-to-SQL Generalization", "Structure-Grounded Pretraining for Text-to-SQL", "Towards Robustness of Text-to-SQL Models Against Natural and Realistic Adversarial Table Perturbation"], "answer_arxiv_id": ["2106.01065", "2109.05157", "2010.12773", "2212.09994"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_11766"} +{"question": "Could you mention some research papers that employ T2I diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2112.10752", "2307.01952", "2205.11487"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_11767"} +{"question": "Which studies present the most recent advancements in Sample-Efficient MARL with guaranteed sample efficiency?", "answer": ["Provable Self-Play Algorithms for Competitive Reinforcement Learning", "Near-Optimal Reinforcement Learning with Self-Play", "A Sharp Analysis of Model-based Reinforcement Learning with Self-Play", "Learning Zero-Sum Simultaneous-Move Markov Games Using Function Approximation and Correlated Equilibrium", "V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL", "When Can We Learn General-Sum Markov Games with a Large Number of Players Sample-Efficiently?", "On Improving Model-Free Algorithms for Decentralized Multi-Agent Reinforcement Learning", "The Complexity of Markov Equilibrium in Stochastic Games", "Breaking the Curse of Multiagents in a Large State Space: RL in Markov Games with Independent Linear Function Approximation"], "answer_arxiv_id": ["2002.04017", "2006.12007", "2010.01604", "2002.07066", "2110.14555", "2110.04184", "2110.05707v2", "2204.03991", "2302.03673"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_11768"} +{"question": "Which research proposed the CKY-like encoder in the composition model?", "answer": ["Jointly Learning Sentence Embeddings and Syntax with Unsupervised\n Tree-LSTMs"], "answer_arxiv_id": ["1705.09189"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_11769"} +{"question": "What studies have been conducted on online learning and control with linear systems, specifically online LQR or LQG with unknown dynamics?", "answer": ["Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification", "Learning Linear Dynamical Systems with Semi-Parametric Least Squares", "Certainty Equivalence is Efficient for Linear Quadratic Control", "On the Sample Complexity of the Linear Quadratic Regulator", "The Power of Predictions in Online Control", "Active Learning for Identification of Linear Dynamical Systems", "Naive Exploration is Optimal for Online LQR", "Improper Learning for Non-Stochastic Control"], "answer_arxiv_id": ["1802.08334", "1902.00768", "1902.07826", "1710.01688", "2006.07569", "2002.00495v2", "2001.09576", "2001.09254"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11770"} +{"question": "Could you tell me about any studies that examined the relations of iterative inference to the 'working memory' of human minds or human visual systems?", "answer": ["Bridging the Gaps Between Residual Learning, Recurrent Neural Networks and Visual Cortex"], "answer_arxiv_id": ["1604.03640"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11771"} +{"question": "Which papers discuss automatic music tagging?", "answer": ["Automatic tagging using deep convolutional neural networks", "Disentangled Multidimensional Metric Learning for Music Similarity", "Sample-level Deep Convolutional Neural Networks for Music Auto-tagging\n Using Raw Waveforms", "musicnn: Pre-trained convolutional neural networks for music audio\n tagging", "Zero-shot Learning for Audio-based Music Classification and Tagging", "Evaluation of CNN-based Automatic Music Tagging Models"], "answer_arxiv_id": ["1606.00298", "2008.03720", "1703.01789", "1909.06654", "1907.02670", "2006.00751"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_11772"} +{"question": "Who analyzed the regularization effect of SAM close to a minimum?", "answer": ["The Dynamics of Sharpness-Aware Minimization: Bouncing Across Ravines and Drifting Towards Wide Minima"], "answer_arxiv_id": ["2210.01513"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11773"} +{"question": "Which papers have studied reward-free exploration?", "answer": ["Reward-Free Exploration for Reinforcement Learning", "Adaptive Reward-Free Exploration", "Fast active learning for pure exploration in reinforcement learning"], "answer_arxiv_id": ["2002.02794", "2006.06294", "2007.13442"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_11774"} +{"question": "Which research papers have used video game benchmarks in reinforcement learning for studying LL?", "answer": ["Playing Atari with Deep Reinforcement Learning", "Open-Ended Learning Leads to Generally Capable Agents", "ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning"], "answer_arxiv_id": ["1312.5602", "2107.12808", "1605.02097"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_11775"} +{"question": "Could you provide me some works about visual instruction tuning in MLLMs?", "answer": ["Qwen-VL: A Versatile Vision-Language Model for Understanding,\n Localization, Text Reading, and Beyond", "What Matters in Training a GPT4-Style Language Model with Multimodal\n Inputs?", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "SVIT: Scaling up Visual Instruction Tuning", "LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding", "An Empirical Study of Scaling Instruct-Tuned Large Multimodal Models"], "answer_arxiv_id": ["2308.12966", "2307.02469", "2305.06500", "2304.08485", "2304.10592", "2307.04087", "2306.17107v2", "2309.09958"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11776"} +{"question": "Which works used state-visitation counts to improve the exploration problem in reinforcement learning?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "Count-Based Exploration with Neural Density Models"], "answer_arxiv_id": ["1606.01868", "1703.01310"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_11777"} +{"question": "What are some of the papers that use Graph Neural Network in molecular graph representation learning?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Inductive Representation Learning on Large Graphs", "Neural Message Passing for Quantum Chemistry", "Graph Attention Networks"], "answer_arxiv_id": ["1609.02907", "1706.02216", "1704.01212", "1710.10903"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_11778"} +{"question": "What studies address examples dealing with aleatoric uncertainty?", "answer": ["Probabilistic Face Embeddings", "Face Image Quality Assessment: A Literature Survey", "MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty Propagation", "Uncertainty Estimation in One-Stage Object Detection", "An Algorithm for Sensor Data Uncertainty Quantification", "Human uncertainty makes classification more robust", "Is one annotation enough? A data-centric image classification benchmark for noisy and ambiguous label estimation", "Benchmarking common uncertainty estimation methods with histopathological images under domain shift and label noise", "Plex: Towards Reliability Using Pretrained Large Model Extensions", "ECCV Caption: Correcting False Negatives by Collecting Machine-and-Human-verified Image-Caption Associations for MS-COCO"], "answer_arxiv_id": ["1904.09658", "2009.01103", "2103.12605", "1905.10296", "2101.02067", "1908.07086", "2207.06214", "2301.01054", "2207.07411", "2204.03359"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_11779"} +{"question": "What works use spatio-temporal interpolation in their video diffusion models?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data"], "answer_arxiv_id": ["2209.14792"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_11780"} +{"question": "What are some task-specific methods in the field of adverse weather restoration?", "answer": ["Multi-Stage Progressive Image Restoration", "HINet: Half Instance Normalization Network for Image Restoration", "Simple Baselines for Image Restoration", "Restormer: Efficient Transformer for High-Resolution Image Restoration", "Efficient and Explicit Modelling of Image Hierarchies for Image\n Restoration"], "answer_arxiv_id": ["2102.02808", "2105.06086", "2204.04676", "2111.09881", "2303.00748"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_11781"} +{"question": "What is the work that introduced mask-based methods for user-prompted/automatically generated masks in image manipulation with DMs?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "DiffEdit: Diffusion-based semantic image editing with mask guidance", "SpaText: Spatio-Textual Representation for Controllable Image Generation"], "answer_arxiv_id": ["2112.10741", "2210.11427", "2211.14305"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_11782"} +{"question": "Could you list some research papers that provide near-optimal sample complexity for best arm identification in linear bandits?", "answer": ["Sequential Experimental Design for Transductive Linear Bandits"], "answer_arxiv_id": ["1906.08399"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_11783"} +{"question": "Can you point to any studies where a large-scale diffusion model was used to generate high quality, high resolution images conditioned on textual input?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "LAION-5B: An open large-scale dataset for training next generation image-text models"], "answer_arxiv_id": ["2112.10752", "2210.08402"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_11784"} +{"question": "Any works focused on the application of prompt tuning under complete black-box conditions?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Visual Prompt Tuning", "MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2203.12119", "2210.03117"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_11785"} +{"question": "What works aim to predict 3D scene graphs from RGB-D and RGB sequence as input?", "answer": ["SceneGraphFusion: Incremental 3D Scene Graph Prediction from RGB-D\n Sequences", "Incremental 3D Semantic Scene Graph Prediction from RGB Sequences"], "answer_arxiv_id": ["2103.14898", "2305.02743"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_11786"} +{"question": "Which work showed the effectiveness of hierarchical behavior cloning for complex tasks in simulated playroom settings?", "answer": ["Imitating Interactive Intelligence"], "answer_arxiv_id": ["2012.05672"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11787"} +{"question": "Could you provide me some works that computed attributions based on perturbations on the input variables?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "Visualizing Deep Neural Network Decisions: Prediction Difference Analysis", "Interpretable Explanations of Black Boxes by Meaningful Perturbation", "A Unified Approach to Interpreting Model Predictions", "Model Agnostic Supervised Local Explanations", "Explaining by Removing: A Unified Framework for Model Explanation", "Mutual Information Preserving Back-propagation: Learn to Invert for Faithful Attribution", "Algorithms to estimate Shapley value feature attributions"], "answer_arxiv_id": ["1602.04938", "1702.04595", "1704.03296", "1705.07874", "1807.02910", "2011.14878", "2104.06629", "2207.07605"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_11788"} +{"question": "What works have applied techniques such as graph and genetic algorithms and flow network to 2D molecular graphs?", "answer": ["Guiding Deep Molecular Optimization with Genetic Exploration", "Junction Tree Variational Autoencoder for Molecular Graph Generation", "MARS: Markov Molecular Sampling for Multi-objective Drug Discovery", "Learning to Extend Molecular Scaffolds with Structural Motifs", "De Novo Molecular Generation via Connection-aware Motif Mining"], "answer_arxiv_id": ["2007.04897", "1802.04364", "2103.10432", "2103.03864", "2302.01129"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_11789"} +{"question": "Could you provide me research that propose Scaling-based methods for confidence calibration?", "answer": ["On Calibration of Modern Neural Networks", "Parameterized Temperature Scaling for Boosting the Expressive Power in Post-Hoc Uncertainty Calibration", "Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning"], "answer_arxiv_id": ["1706.04599", "2102.12182", "2003.07329"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_11790"} +{"question": "What research papers introduces the principle of maximum entropy RL?", "answer": ["Reinforcement Learning and Control as Probabilistic Inference: Tutorial and Review"], "answer_arxiv_id": ["1805.00909"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_11791"} +{"question": "What studies propose a rule-based pruning method to generate a sparse coordination graph?", "answer": ["Context-Aware Sparse Deep Coordination Graphs"], "answer_arxiv_id": ["2106.02886v3"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_11792"} +{"question": "What works categorize network sparsity into structured sparsity, unstructured sparsity and semi-structured sparsity?", "answer": ["Learning Structured Sparsity in Deep Neural Networks", "Channel Pruning for Accelerating Very Deep Neural Networks", "HRank: Filter Pruning using High-Rank Feature Map", "Dynamic Network Surgery for Efficient DNNs", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Rigging the Lottery: Making All Tickets Winners", "Soft Threshold Weight Reparameterization for Learnable Sparsity", "Fast Sparse ConvNets", "Accelerating Sparse Deep Neural Networks", "1xN Pattern for Pruning Convolutional Neural Networks"], "answer_arxiv_id": ["1608.03665", "1707.06168", "2002.10179", "1608.04493", "1803.03635", "1911.11134", "2002.03231", "1911.09723", "2104.08378", "2105.14713"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_11793"} +{"question": "Which works have employed tools from causal inference for studying OPE with unobserved confounders?", "answer": ["Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models", "Confounding-Robust Policy Evaluation in Infinite-Horizon Reinforcement Learning", "Off-policy Evaluation in Infinite-Horizon Reinforcement Learning with Latent Confounders", "Proximal Causal Learning with Kernels: Two-Stage Estimation and Moment Restriction", "Proximal Reinforcement Learning: Efficient Off-Policy Evaluation in Partially Observed Markov Decision Processes", "Dynamic Causal Effects Evaluation in A/B Testing with a Reinforcement Learning Framework"], "answer_arxiv_id": ["1905.05824", "2002.04518", "2007.13893", "2105.04544", "2110.15332", "2002.01711"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_11794"} +{"question": "What studies discussed the application of generative adversarial networks and autoregressive models in T2I capability?", "answer": ["Generative Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Analyzing and Improving the Image Quality of StyleGAN", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2203.00667", "1809.11096", "1912.04958", "2005.14165"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_11795"} +{"question": "What are the major methods to evaluate fairness in deep learning as mentioned in the cited papers?", "answer": ["Verifying Individual Fairness in Machine Learning Models", "Latent Imitator: Generating Natural Individual Discriminatory Instances for Black-Box Fairness Testing", "Learning to Pivot with Adversarial Networks", "Invariant Representations without Adversarial Training", "Controllable Guarantees for Fair Outcomes via Contrastive Information Estimation", "Fairness Metrics: A Comparative Analysis"], "answer_arxiv_id": ["2006.11737", "2305.11602", "1611.01046", "1805.09458", "2101.04108", "2001.07864"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_11796"} +{"question": "What paper also prompts LLMs for priors like in this work?", "answer": ["LMPriors: Pre-Trained Language Models as Task-Specific Priors"], "answer_arxiv_id": ["2210.12530"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_11797"} +{"question": "Can you provide references of works that explored KD variants for life-long learning?", "answer": ["Learning without Forgetting"], "answer_arxiv_id": ["1606.09282"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_11798"} +{"question": "Could you provide me some studies that utilize in-painting in unsupervised AD approaches for analyzing RGB images?", "answer": ["Inpainting Transformer for Anomaly Detection"], "answer_arxiv_id": ["2104.13897"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_11799"} +{"question": "Are there any studies investigating the role of the intermediate attention maps and features in the diffuser?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image\n Diffusion Models", "Zero-shot Image-to-Image Translation", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2208.01626", "2301.13826", "2302.03027", "2211.12572"], "source_meta": {"published_time": "20240105"}, "qid": "AutoScholarQuery_train_11800"} +{"question": "What research discusses learning a policy conditioned on either language or goal images?", "answer": ["Language Conditioned Imitation Learning over Unstructured Data", "CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Tasks"], "answer_arxiv_id": ["2005.07648", "2112.03227"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_11801"} +{"question": "Which paper provides a comprehensive survey on long-range efficient Transformer models?", "answer": ["Efficient Transformers: A Survey"], "answer_arxiv_id": ["2009.06732"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_11802"} +{"question": "What are some works on human motion modeling with multi-modality conditions?", "answer": ["MotionGPT: Human Motion as a Foreign Language", "Weakly-supervised Action Transition Learning for Stochastic Human Motion\n Prediction", "Motion Question Answering via Modular Motion Programs"], "answer_arxiv_id": ["2306.14795", "2205.15608", "2305.08953"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_11803"} +{"question": "Which references provide more information on adapter methods for efficient fine-tuning of visual parameters?", "answer": ["Parameter-Efficient Transfer Learning for NLP"], "answer_arxiv_id": ["1902.00751"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_11804"} +{"question": "What papers developed policy-based methods in tabular and linear settings?", "answer": ["Optimistic Policy Optimization with Bandit Feedback", "Provably Efficient Exploration in Policy Optimization", "PC-PG: Policy Cover Directed Exploration for Provable Policy Gradient Learning"], "answer_arxiv_id": ["2002.08243", "1912.05830", "2007.08459"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11805"} +{"question": "What studies are about Tracking Animals in the Wild (TAO)?", "answer": ["TAO: A Large-Scale Benchmark for Tracking Any Object"], "answer_arxiv_id": ["2005.10356"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_11806"} +{"question": "Which papers focus on estimating Qπ with an off-policy RL algorithm?", "answer": ["Batch Policy Learning under Constraints", "Hyperparameter Selection for Offline Reinforcement Learning"], "answer_arxiv_id": ["1903.08738", "2007.09055"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_11807"} +{"question": "Which studies explored the use of Neural Ordinary Differential Equations (ODEs) or Neural Controlled Differential Equations (CDEs) for treatment effect estimation?", "answer": ["Neural Ordinary Differential Equations for Intervention Modeling", "Predicting the impact of treatments over time with uncertainty aware neural differential equations", "Continuous-Time Modeling of Counterfactual Outcomes Using Neural Controlled Differential Equations"], "answer_arxiv_id": ["2010.08304", "2202.11987v1", "2206.08311"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_11808"} +{"question": "Can you list the works that developed image translation-based methods for face stylization?", "answer": ["WarpGAN: Automatic Caricature Generation", "StyleCariGAN: Caricature Generation via StyleGAN Feature Map Modulation", "Pastiche Master: Exemplar-Based High-Resolution Portrait Style Transfer", "Resolution Dependent GAN Interpolation for Controllable Image Synthesis\n Between Domains", "CariGANs: Unpaired Photo-to-Caricature Translation", "AutoToon: Automatic Geometric Warping for Face Cartoon Generation", "DCT-Net: Domain-Calibrated Translation for Portrait Stylization"], "answer_arxiv_id": ["1811.10100", "2107.04331", "2203.13248", "2010.05334", "1811.00222", "2004.02377", "2207.02426"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_11809"} +{"question": "Which works proposed methods for increasing the personalization of generative models?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "A Neural Space-Time Representation for Text-to-Image Personalization", "Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models", "FastComposer: Tuning-Free Multi-Subject Image Generation with Localized\n Attention", "Key-Locked Rank One Editing for Text-to-Image Personalization", "Photoswap: Personalized Subject Swapping in Images", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2212.04488", "2305.14720", "2305.15391", "2307.11410", "2307.06949", "2305.10431", "2305.01644", "2305.18286", "2302.12228"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_11810"} +{"question": "Which studies proposed solutions for control tasks that may be partially observable due to unmeasured state variables, sensor noise, and unmeasured system parameters?", "answer": ["Variational Recurrent Models for Solving Partially Observable Control Tasks", "Recurrent Off-policy Baselines for Memory-based Continuous Control", "Memory-based control with recurrent neural networks", "Memory-based Deep Reinforcement Learning for POMDPs", "Preparing for the Unknown: Learning a Universal Policy with Online System Identification", "Assessing Generalization in Deep Reinforcement Learning"], "answer_arxiv_id": ["1912.10703", "2110.12628", "1512.04455", "2102.12344", "1702.02453", "1810.12282"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_11811"} +{"question": "What research improved DDIM inversion by adjusting text features?", "answer": ["Null-text Inversion for Editing Real Images using Guided Diffusion\n Models", "Improving Tuning-Free Real Image Editing with Proximal Guidance", "Negative-prompt Inversion: Fast Image Inversion for Editing with\n Text-guided Diffusion Models"], "answer_arxiv_id": ["2211.09794", "2306.05414", "2305.16807"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_11812"} +{"question": "Could you provide me some research that improve on synthesis quality in diffusion models?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2105.05233", "2206.00364v2"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_11813"} +{"question": "Could you provide me with the works that examined model robustness with respect to factors such as pose and size by rendering 3D-objects?", "answer": ["The Robustness Limits of SoTA Vision Models to Natural Variation", "Progress and limitations of deep networks to recognize objects in unusual poses", "Strike (with) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects"], "answer_arxiv_id": ["2210.13604", "2207.08034", "1811.11553"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_11814"} +{"question": "Which works demonstrated that it is possible to teach robot new skills using only a single demonstration?", "answer": ["One-Shot Imitation Learning", "One-Shot Visual Imitation Learning via Meta-Learning", "Zero-Shot Visual Imitation", "One-Shot Imitation from Observing Humans via Domain-Adaptive Meta-Learning", "Watch, Try, Learn: Meta-Learning from Demonstrations and Reward", "One-Shot Imitation from Observing Humans via Domain-Adaptive Meta-Learning", "Language Conditioned Imitation Learning over Unstructured Data", "Language-Conditioned Imitation Learning for Robot Manipulation Tasks"], "answer_arxiv_id": ["1703.07326v3", "1709.04905", "1804.08606", "1802.01557", "1906.03352v4", "1802.01557", "2005.07648", "2010.12083"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_11815"} +{"question": "Which works address the importance of fairness of exposure in a static ranking setting?", "answer": ["Fairness-Aware Ranking in Search & Recommendation Systems with Application to LinkedIn Talent Search", "Fairness in Recommendation Ranking through Pairwise Comparisons", "Measuring Fairness in Ranked Outputs", "Fairness of Exposure in Rankings", "Fair ranking: a critical review, challenges, and future directions", "Fairness in Ranking: A Survey", "Introducing the Expohedron for Efficient Pareto-optimal Fairness-Utility Amortizations in Repeated Rankings", "Evaluating Stochastic Rankings with Expected Exposure", "Optimizing Generalized Gini Indices for Fairness in Rankings", "Joint Multisided Exposure Fairness for Recommendation"], "answer_arxiv_id": ["1905.01989", "1903.00780", "1610.08559", "1802.07281", "2201.12662v1", "2103.14000", "2202.03237v1", "2004.13157", "2204.06521", "2205.00048"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_11816"} +{"question": "Which works assess the social biases in Language Models (LMMs) across gender, race, age, and geographic location?", "answer": ["Evaluating CLIP: Towards Characterization of Broader Capabilities and\n Downstream Implications", "Measuring Representational Harms in Image Captioning", "Stable Bias: Analyzing Societal Representations in Diffusion Models", "Social Biases through the Text-to-Image Generation Lens"], "answer_arxiv_id": ["2108.02818", "2206.07173", "2303.11408", "2304.06034"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_11817"} +{"question": "What researches have been conducted on context-agnostic learning and its impacts on training performance and generalization ability over a distribution of tasks?", "answer": ["Contextualize Me – The Case for Context in Reinforcement Learning"], "answer_arxiv_id": ["2202.04500"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_11818"} +{"question": "What research have been done on exact machine unlearning?", "answer": ["Making AI Forget You: Data Deletion in Machine Learning", "Machine Unlearning for Random Forests", "Machine Unlearning", "Graph Unlearning", "Recommendation Unlearning"], "answer_arxiv_id": ["1907.05012", "2009.05567", "1912.03817v3", "2103.14991", "2201.06820"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_11819"} +{"question": "What work uses reward-free reinforcement learning to tackle the problem of the approachability task for VMDPs?", "answer": ["A Simple Reward-free Approach to Constrained Reinforcement Learning"], "answer_arxiv_id": ["2107.05216"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_11820"} +{"question": "Could you list researches that extended modelling of plausible domains by taking affine combinations beyond the convex hull?", "answer": ["Out-of-Distribution Generalization via Risk Extrapolation (REx)"], "answer_arxiv_id": ["2003.00688v5"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_11821"} +{"question": "Are there any works that used Supermartingales for the analysis of probabilistic programs?", "answer": ["Stochastic Invariants for Probabilistic Termination"], "answer_arxiv_id": ["1611.01063"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_11822"} +{"question": "What are some studies that have explored non-parametric models in causal inference?", "answer": ["Learning directed acyclic graph models based on sparsest permutations", "Consistency Guarantees for Greedy Permutation-Based Causal Inference Algorithms"], "answer_arxiv_id": ["1307.0366", "1702.03530"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_11823"} +{"question": "Which papers explored using diffusion models for discriminative vision tasks?", "answer": ["Unleashing Text-to-Image Diffusion Models for Visual Perception", "Prompting Diffusion Representations for Cross-Domain Semantic\n Segmentation", "DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic\n Segmentation Using Diffusion Models", "One-shot Unsupervised Domain Adaptation with Personalized Diffusion\n Models"], "answer_arxiv_id": ["2303.02153v1", "2307.02138", "2303.11681", "2303.18080"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_11824"} +{"question": "What works are about Differentially Private Stochastic Gradient Descent (DP-SGD) and its variants?", "answer": ["Deep Learning with Differential Privacy", "Differentially Private Empirical Risk Minimization Revisited: Faster and More General", "Bypassing the Ambient Dimension: Private SGD with Gradient Subspace Identification", "Differentially Private Coordinate Descent for Composite Empirical Risk Minimization", "Practical and Private (Deep) Learning Without Sampling or Shuffling"], "answer_arxiv_id": ["1607.00133", "1802.05251v1", "2007.03813", "2110.11688", "2103.00039"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_11825"} +{"question": "Could you provide me some studies about reconstructing full body motion from VR/AR devices?", "answer": ["LoBSTr: Real-time Lower-body Pose Prediction from Sparse Upper-body\n Tracking Signals", "AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion\n Sensing", "Realistic Full-Body Tracking from Sparse Observations via Joint-Level\n Modeling"], "answer_arxiv_id": ["2103.01500", "2207.13784", "2308.08855"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_11826"} +{"question": "Could you provide me some studies about a query-based segmentation architecture to support various segmentation and image-level vision-language understanding tasks?", "answer": ["Generalized Decoding for Pixel, Image, and Language", "Segment Everything Everywhere All at Once"], "answer_arxiv_id": ["2212.11270", "2304.06718"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_11827"} +{"question": "Which studies improved CLIP by connecting the two modalities via cross-modal attention or multi-object representation alignment?", "answer": ["Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual Concepts"], "answer_arxiv_id": ["2107.07651", "2111.08276"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_11828"} +{"question": "What research showed that diffusion can be performed on the latent space of a pretrained Variational Autoencoder?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11829"} +{"question": "Which works enhance non-IID FUSL by generating global supervised signals?", "answer": ["Heterogeneity for the Win: One-Shot Federated Clustering", "Federated Unsupervised Representation Learning"], "answer_arxiv_id": ["2103.00697", "2010.08982"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_11830"} +{"question": "What study suggests resampling of rationales in every iteration so as to avoid overfitting to any specific rationale?", "answer": ["STaR: Bootstrapping Reasoning With Reasoning"], "answer_arxiv_id": ["2203.14465v2"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_11831"} +{"question": "Can you name some self-supervised learning (SSL) based TTA papers?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "MT3: Meta Test-Time Training for Self-Supervised Test-Time Adaption", "Test-Time Training with Masked Autoencoders"], "answer_arxiv_id": ["1909.13231", "2103.16201", "2209.07522"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_11832"} +{"question": "Could you provide me some studies about deep PLL applying variational inference to iteratively recover label distribution?", "answer": ["Instance-Dependent Partial Label Learning"], "answer_arxiv_id": ["2110.12911"], "source_meta": {"published_time": "20220408"}, "qid": "AutoScholarQuery_train_11833"} +{"question": "What research demonstrates that text encoding with large language models is effective at image synthesis?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer"], "answer_arxiv_id": ["2205.11487", "1910.10683"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_11834"} +{"question": "Who reignited interest in the extraordinary generating capacity of diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_11835"} +{"question": "What studies investigated participant behavior in the context of strategic classification without taking into account competition between them?", "answer": ["Alternative Microfoundations for Strategic Classification", "Strategic Classification in the Dark"], "answer_arxiv_id": ["2106.12705", "2102.11592"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_11836"} +{"question": "What works have been done on template updates in the field of video object segmentation?", "answer": ["Video Object Segmentation using Space-Time Memory Networks", "Collaborative Video Object Segmentation by Foreground-Background Integration", "Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation", "XMem: Long-Term Video Object Segmentation with an Atkinson-Shiffrin Memory Model"], "answer_arxiv_id": ["1904.00607", "2003.08333", "2106.05210", "2207.07115"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_11837"} +{"question": "Are there any studies focusing on learning general concepts applicable to a variety of scenarios for diffusion models?", "answer": ["Viewpoint Textual Inversion: Unleashing Novel View Synthesis with\n Pretrained 2D Diffusion Models"], "answer_arxiv_id": ["2309.07986"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_11838"} +{"question": "What studies examined rejection in Support Vector Machines?", "answer": ["Lasso type classifiers with a reject option", "Support vector machines with a reject option"], "answer_arxiv_id": ["0705.2363v1", "1201.1140"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_11839"} +{"question": "Can you provide some works that demonstrate the improvement of model performance through the addition of unified Sequence-to-Sequence NLP training instances to pre-trained language models?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "ZeroPrompt: Scaling Prompt-Based Pretraining to 1,000 Tasks Improves Zero-shot Generalization"], "answer_arxiv_id": ["2109.01652", "2110.08207", "2201.06910"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_11840"} +{"question": "What works feature continuous-time methods that directly learn on the whole dynamic graph?", "answer": ["Inductive representation learning on temporal graphs", "Temporal Graph Networks for Deep Learning on Dynamic Graphs", "Streaming Graph Neural Networks", "APAN: Asynchronous Propagation Attention Network for Real-time Temporal Graph Embedding", "Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks", "Neighborhood-aware Scalable Temporal Network Representation Learning", "Inductive Representation Learning in Temporal Networks via Causal Anonymous Walks", "TCL: Transformer-based Dynamic Graph Modelling via Contrastive Learning", "Do We Really Need Complicated Model Architectures For Temporal Networks?"], "answer_arxiv_id": ["2002.07962", "2006.10637", "1810.10627", "2011.11545", "1908.01207", "2209.01084", "2101.05974", "2105.07944", "2302.11636"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_11841"} +{"question": "Could you give me examples of research that considered the first order approximation to tackle the original DRO problem?", "answer": ["Sensitivity analysis of Wasserstein distributionally robust optimization problems", "Robust Sensitivity Analysis for Stochastic Systems", "On the regularized risk of distributionally robust learning over deep neural networks"], "answer_arxiv_id": ["2006.12022", "1303.0326", "2109.06294"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_11842"} +{"question": "Which papers talk about the advancement of natural language understanding and generation with the use of LLMs?", "answer": ["GPT-4 Technical Report", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2303.08774", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_11843"} +{"question": "Which works introduced datasets for extracting numeric values and semantic descriptions from scientific and financial documents?", "answer": ["FiNER: Financial Numeric Entity Recognition for XBRL Tagging"], "answer_arxiv_id": ["2203.06482"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_11844"} +{"question": "Could you provide me some studies using diffusion models for 3-D shape generation?", "answer": ["LION: Latent Point Diffusion Models for 3D Shape Generation", "DreamFusion: Text-to-3D using 2D Diffusion", "Rodin: A Generative Model for Sculpting 3D Digital Avatars Using Diffusion"], "answer_arxiv_id": ["2210.06978", "2209.14988", "2212.06135"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_11845"} +{"question": "What works build on AMP or NGD techniques for denoising problems related to the implementation of stochastic localization?", "answer": ["Sampling from the Sherrington-Kirkpatrick Gibbs measure via algorithmic stochastic localization", "Posterior Sampling from the Spiked Models via Diffusion Processes", "Sudakov-Fernique post-AMP, and a new proof of the local convexity of the TAP free energy"], "answer_arxiv_id": ["2203.05093", "2304.11449", "2208.09550"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_11846"} +{"question": "Could you name the work instrumental in introducing the concept of Text Decision Transformer (TDT), a variant of Decision Transformer?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["2106.01345"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_11847"} +{"question": "What research works explore the usage of optical flow for deformation regularization?", "answer": ["Neural Radiance Flow for 4D View Synthesis and Video Processing", "Dynamic View Synthesis from Dynamic Monocular Video", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Flow supervision for Deformable NeRF"], "answer_arxiv_id": ["2012.09790", "2105.06468", "2011.13084", "2303.16333"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_11848"} +{"question": "Which studies laid the foundation of later vision-text pertaining models in vision-language pre-training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_11849"} +{"question": "Are there any studies using progressive distillation for reducing the denoising steps in text-to-image generation?", "answer": ["Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2202.00512"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_11850"} +{"question": "What earlier works did not query LLMs along various dimensions such as visual, taxonomy, habitat, and geographic priors and systematically evaluate their effectiveness?", "answer": ["Visual Classification via Description from Large Language Models", "Enhancing CLIP with GPT-4: Harnessing Visual Descriptions as Prompts"], "answer_arxiv_id": ["2210.07183", "2307.11661"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_11851"} +{"question": "Could you provide me some studies that predict light probes or sky models for lighting estimation?", "answer": ["Two-shot Spatially-varying BRDF and Shape Estimation", "DeepLight: Learning Illumination for Unconstrained Mobile Mixed Reality", "Deep Outdoor Illumination Estimation", "Deep Sky Modeling for Single Image Outdoor Lighting Estimation", "All-Weather Deep Outdoor Lighting Estimation"], "answer_arxiv_id": ["2004.00403", "1904.01175", "1611.06403", "1905.03897", "1906.04909"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11852"} +{"question": "Which paper uses CSBMs to study the function of nonlinearity on the node classification performance within GNNs?", "answer": ["Understanding Non-linearity in Graph Neural Networks from the Bayesian-Inference Perspective"], "answer_arxiv_id": ["2207.11311"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_11853"} +{"question": "What papers introduced approaches focusing on the probability density function for dealing with distributional treatment effects?", "answer": ["A unified study of nonparametric inference for monotone functions"], "answer_arxiv_id": ["1806.01928"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_11854"} +{"question": "What papers might benefit from applying the sequential Monte Carlo algorithm template for structure learning?", "answer": ["Learning the Structure of Deep Sparse Graphical Models", "CrossCat: A Fully Bayesian Nonparametric Method for Analyzing Heterogeneous, High Dimensional Data", "Bayesian Symbolic Regression", "Hierarchical Infinite Relational Model"], "answer_arxiv_id": ["1001.0160", "1512.01272", "1910.08892", "2108.07208"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_11855"} +{"question": "Could you provide me some studies on the use of non-Euclidean metrics for constructing proximity graphs for unsupervised learning tasks?", "answer": ["A Path-Based Distance for Street Map Comparison", "Path-Based Spectral Clustering: Guarantees, Robustness to Outliers, and Fast Algorithms", "Balancing Geometry and Density: Path Distances on High-Dimensional Data"], "answer_arxiv_id": ["1309.6131", "1712.06206", "2012.09385"], "source_meta": {"published_time": "20210728"}, "qid": "AutoScholarQuery_train_11856"} +{"question": "Could you provide me some studies on the use of discretization techniques in RL?", "answer": ["Discrete Factorial Representations as an Abstraction for Goal Conditioned Reinforcement Learning", "Choreographer: Learning and Adapting Skills in Imagination"], "answer_arxiv_id": ["2211.00247v1", "2211.13350"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_11857"} +{"question": "What papers have studied the limitations of Transformer models?", "answer": ["Theoretical Limitations of Self-Attention in Neural Sequence Models", "On the Ability and Limitations of Transformers to Recognize Formal Languages"], "answer_arxiv_id": ["1906.06755", "2009.11264"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_11858"} +{"question": "Which work applied classical objective functions to sequence-to-sequence models and showed improved performance?", "answer": ["Classical Structured Prediction Losses for Sequence to Sequence Learning"], "answer_arxiv_id": ["1711.04956"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_11859"} +{"question": "Which work propose the use of a retrieval process in multi-task offline RL settings?", "answer": ["Retrieval-Augmented Reinforcement Learning"], "answer_arxiv_id": ["2202.08417"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_11860"} +{"question": "Which papers discuss 3D generative methods using 3D voxels representation?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Escaping Plato's Cave: 3D Shape From Adversarial Rendering"], "answer_arxiv_id": ["1610.07584", "1503.03585", "1811.11606"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_11861"} +{"question": "Can you specify some papers that implement the online-offline topology in cryptographic protocols?", "answer": ["CryptoNAS: Private Inference on a ReLU Budget"], "answer_arxiv_id": ["2006.08733"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_11862"} +{"question": "Can you mention some studies that proposed to train a 2D viewpoint transformation model and which can be used for generating 3D consistent objects?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2303.11328"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_11863"} +{"question": "What's the work introducing the SSL algorithm SimCLR?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_11864"} +{"question": "What works on Human Motion Synthesis performed based on text?", "answer": ["TEMOS: Generating diverse human motions from textual descriptions", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "Human Motion Diffusion Model", "TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts", "Language2Pose: Natural Language Grounded Pose Forecasting", "FLAME: Free-form Language-based Motion Synthesis & Editing"], "answer_arxiv_id": ["2204.14109", "2208.15001", "2209.14916", "2207.01696", "1907.01108", "2209.00349"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_11865"} +{"question": "What work uses Hungarian algorithm for supervision in same context?", "answer": ["Learning Object Bounding Boxes for 3D Instance Segmentation on Point\n Clouds"], "answer_arxiv_id": ["1906.01140"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_11866"} +{"question": "What studies have shifted from sphere tracing to volume rendering for Signed Distance Function rendering?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces"], "answer_arxiv_id": ["2106.10689", "2106.12052"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_11867"} +{"question": "Which papers investigate varying the action space of MetaBOO methods, including determining the solution update strategy and generating hyperparameters?", "answer": ["Deep Reinforcement Learning Based Parameter Control in Differential Evolution", "Learning Adaptive Differential Evolution Algorithm from Optimization Experiences by Policy Gradient"], "answer_arxiv_id": ["1905.08006", "2102.03572"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_11868"} +{"question": "What studies mention that deblurring models typically use the UNet model?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation"], "answer_arxiv_id": ["1505.04597"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_11869"} +{"question": "Can you cite studies that worked on the multi-modal retrieval task?", "answer": ["WebQA: Multihop and Multimodal QA", "ManyModalQA: Modality Disambiguation and QA over Diverse Inputs", "MultiModalQA: Complex Question Answering over Text, Tables and Images"], "answer_arxiv_id": ["2109.00590", "2001.08034", "2104.06039"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_11870"} +{"question": "What research follow state-of-the-art link prediction literature and take the Hadamard product followed by a MLP as the link prediction Decoder for all methods?", "answer": ["Labeling Trick: A Theory of Using Graph Neural Networks for Multi-Node Representation Learning", "VERSE: Versatile Graph Embeddings from Similarity Measures", "Learning from Counterfactual Links for Link Prediction", "Pairwise Learning for Neural Link Prediction"], "answer_arxiv_id": ["2010.16103", "1803.04742v1", "2106.02172", "2112.02936"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_11871"} +{"question": "What studies proposed fine-grained action recognition datasets in sports?", "answer": ["FineDiving: A Fine-grained Dataset for Procedure-aware Action Quality\n Assessment", "FineGym: A Hierarchical Video Dataset for Fine-grained Action\n Understanding"], "answer_arxiv_id": ["2204.03646", "2004.06704"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_11872"} +{"question": "Which works have presented the result that smaller batch sizes achieves the best training stability and generalization performance?", "answer": ["Revisiting Small Batch Training for Deep Neural Networks"], "answer_arxiv_id": ["1804.07612"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_11873"} +{"question": "Which works proposed methods for learning bipartite noisy-OR BNs?", "answer": ["Unsupervised Learning of Noisy-Or Bayesian Networks"], "answer_arxiv_id": ["1309.6834v1"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_11874"} +{"question": "Which papers are about evolving graph signals by propagating them with graph structures iteratively?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Simplifying Graph Convolutional Networks", "Graph Attention Networks", "Diffusion Improves Graph Learning"], "answer_arxiv_id": ["1609.02907", "1902.07153", "1710.10903", "1911.05485"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_11875"} +{"question": "Which works give lower bounds for the minimax optimization and involve the property that the action sets are invariant under rotation?", "answer": ["Linear Lower Bounds and Conditioning of Differentiable Games", "Lower complexity bounds of first-order methods for convex-concave bilinear saddle-point problems"], "answer_arxiv_id": ["1906.07300", "1808.02901"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_11876"} +{"question": "What studies consider the dynamics of gradient flow on infinite-width homogeneous two-layer networks?", "answer": ["Implicit Bias of Gradient Descent for Wide Two-layer Neural Networks Trained with the Logistic Loss"], "answer_arxiv_id": ["2002.04486"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_11877"} +{"question": "Could you provide me some works about effective data augmentation policies that enhance the generalization ability of DNNs?", "answer": ["mixup: Beyond Empirical Risk Minimization", "Improved Regularization of Convolutional Neural Networks with Cutout", "AutoAugment: Learning Augmentation Strategies from Data", "RandAugment: Practical automated data augmentation with a reduced search space"], "answer_arxiv_id": ["1710.09412", "1708.04552", "1805.09501", "1909.13719"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_11878"} +{"question": "Which articles have been exploring the analyzing, evaluating and mitigating problems related to language model behavior?", "answer": ["Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned"], "answer_arxiv_id": ["2209.07858"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_11879"} +{"question": "What work proposed to do diffusion process in a latent space using Latent Diffusion Model (LDM)?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_11880"} +{"question": "Which research proposed the ISGP-Linkgistic algorithm by using the (u,v)-geometric structure in combined with Legendre functions to learn canonical link functions?", "answer": ["All your loss are belong to Bayes"], "answer_arxiv_id": ["2006.04633"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_11881"} +{"question": "What papers use shared multi-layer perceptron (MLP) in point cloud semantic segmentation?", "answer": ["PointSIFT: A SIFT-like Network Module for 3D Point Cloud Semantic\n Segmentation", "Exploring Spatial Context for 3D Semantic Segmentation of Point Clouds", "PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "Know What Your Neighbors Do: 3D Semantic Segmentation of Point Clouds", "RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds"], "answer_arxiv_id": ["1807.00652", "1802.01500", "1612.00593", "1706.02413", "1810.01151", "1911.11236"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_11882"} +{"question": "What studies have been done on adversarial examples for deep learning models?", "answer": ["Adversarial Attacks and Defences: A Survey", "Adversarial Attacks on Deep Learning Models in Natural Language Processing: A Survey"], "answer_arxiv_id": ["1810.00069", "1901.06796"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_11883"} +{"question": "What studies have focused on learning from symbolic reasoning in neuro-symbolic learning?", "answer": ["DeepProbLog: Neural Probabilistic Logic Programming", "NeurASP: Embracing Neural Networks into Answer Set Programming", "NeuPSL: Neural Probabilistic Soft Logic", "DeepStochLog: Neural Stochastic Logic Programming", "Closed Loop Neural-Symbolic Learning via Integrating Neural Perception, Grammar Parsing, and Symbolic Reasoning"], "answer_arxiv_id": ["1805.10872", "2307.07700", "2205.14268v3", "2106.12574", "2006.06649"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_11884"} +{"question": "Could you provide examples of GNN-based EA methods?", "answer": ["Boosting the Speed of Entity Alignment 10*: Dual Attention Matching\n Network with Normalized Hard Sample Mining"], "answer_arxiv_id": ["2103.15452"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_11885"} +{"question": "Which works explore large-scale upstream training for finetune-free subject-driven generation?", "answer": ["InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "Taming Encoder for Zero Fine-tuning Image Customization with\n Text-to-Image Diffusion Models", "Subject-driven Text-to-Image Generation via Apprenticeship Learning"], "answer_arxiv_id": ["2304.03411", "2304.02642", "2304.00186"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_11886"} +{"question": "What studies have proposed directions in the latent space for specific semantic edits using a pre-trained GAN?", "answer": ["GANSpace: Discovering Interpretable GAN Controls", "Editing in Style: Uncovering the Local Semantics of GANs", "Interpreting the Latent Space of GANs for Semantic Face Editing", "GAN-Control: Explicitly Controllable GANs"], "answer_arxiv_id": ["2004.02546", "2004.14367", "1907.10786", "2101.02477"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_11887"} +{"question": "What works analyze the generalization behavior of overparameterized linear models in the context of regression?", "answer": ["Surprises in High-Dimensional Ridgeless Least Squares Interpolation", "Benign Overfitting in Linear Regression", "Two models of double descent for weak features", "Harmless interpolation of noisy data in regression"], "answer_arxiv_id": ["1903.08560", "1906.11300", "1903.07571", "1903.09139"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_11888"} +{"question": "Which work provided a framework for membership scores in membership inference attacks?", "answer": ["Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "White-box vs Black-box: Bayes Optimal Strategies for Membership Inference", "On the Importance of Difficulty Calibration in Membership Inference Attacks", "Membership Inference Attacks Against Machine Learning Models", "Membership Inference Attacks From First Principles"], "answer_arxiv_id": ["1709.01604", "1908.11229", "2111.08440", "1610.05820", "2112.03570"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_11889"} +{"question": "What research explored the reconstruction methods using generative models (such as AE and GAN) for anomaly detection?", "answer": ["Auto-Encoding Variational Bayes", "Generative Adversarial Networks"], "answer_arxiv_id": ["1312.6114", "2203.00667"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_11890"} +{"question": "What works provide more understanding and analysis of the underlying uncertainties in probabilistic regression?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?"], "answer_arxiv_id": ["1703.04977"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_11891"} +{"question": "Which work extended Decision Transformer to the offline-to-online RL setting?", "answer": ["Online Decision Transformer"], "answer_arxiv_id": ["2202.05607"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_11892"} +{"question": "Can you list studies that propose dynamic confident thresholding in the combination of curriculum learning and pseudo-labeling?", "answer": ["FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning", "Dash: Semi-Supervised Learning with Dynamic Thresholding"], "answer_arxiv_id": ["2205.07246", "2109.00650"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_11893"} +{"question": "Which works have explored positional or structural encodings capturing shortest-path distances in graphs?", "answer": ["Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning"], "answer_arxiv_id": ["2009.00142"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_11894"} +{"question": "What works proposed methods to tailor the SAM for general medical image segmentation?", "answer": ["Segment Anything in Medical Images"], "answer_arxiv_id": ["2304.12306"], "source_meta": {"published_time": "20240501"}, "qid": "AutoScholarQuery_train_11895"} +{"question": "Can you cite research papers that have explored speculative decoding?", "answer": ["Fast Inference from Transformers via Speculative Decoding", "Draft & Verify: Lossless Large Language Model Acceleration via\n Self-Speculative Decoding", "SPEED: Speculative Pipelined Execution for Efficient Decoding"], "answer_arxiv_id": ["2211.17192", "2309.08168", "2310.12072"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_11896"} +{"question": "Are there any research pieces related to numerical reasoning in machine learning?", "answer": ["DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning\n Over Paragraphs"], "answer_arxiv_id": ["1903.00161"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_11897"} +{"question": "Any studies about time-dependent implicit representations for capturing scene dynamics?", "answer": ["Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields"], "answer_arxiv_id": ["2011.12948", "2106.13228"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_11898"} +{"question": "Which works are in the field of constrained action space in reinforcement learning (RL)?", "answer": ["A Review of Safe Reinforcement Learning: Methods, Theory and Applications"], "answer_arxiv_id": ["2205.10330"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_11899"} +{"question": "Could you provide me some studies about text-guided image inpainting?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "SmartBrush: Text and Shape Guided Object Inpainting with Diffusion Model"], "answer_arxiv_id": ["2112.10752", "2212.05034"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11900"} +{"question": "Which works employed an MAE-like approach to achieve camera-based semantic occupancy prediction?", "answer": ["VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene\n Completion"], "answer_arxiv_id": ["2302.12251"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_11901"} +{"question": "Can you provide me some studies about diffusion-based models for performing generative imaging?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "State of the Art on Diffusion Models for Visual Computing"], "answer_arxiv_id": ["2112.10752", "2006.11239", "2010.02502", "2310.07204"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_11902"} +{"question": "What are the models that construct random functions with diverse characteristics?", "answer": ["Attentive Neural Processes", "The Functional Neural Process", "Bootstrapping Neural Processes", "Meta-Learning Stationary Stochastic Process Prediction with Convolutional Neural Processes"], "answer_arxiv_id": ["1901.05761", "1906.08324", "2008.02956", "2007.01332"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_11903"} +{"question": "What research have studied proximal causal inference to identify the value of the target policy in POMDPs?", "answer": ["Identifying Causal Effects With Proxy Variables of an Unmeasured Confounder", "A Confounding Bridge Approach for Double Negative Control Inference on Causal Effects", "Semiparametric proximal causal inference", "An Introduction to Proximal Causal Learning", "Kernel Methods for Unobserved Confounding: Negative Controls, Proxies, and Instruments"], "answer_arxiv_id": ["1609.08816", "1808.04945", "2011.08411", "2009.10982v1", "2012.10315"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_11904"} +{"question": "Are there any works that employ hierarchical labels for training in robust metric learning?", "answer": ["Dynamic Metric Learning: Towards a Scalable Metric Space to Accommodate Multiple Semantic Scales", "Hierarchical Average Precision Training for Pertinent Image Retrieval", "Adaptive Hierarchical Similarity Metric Learning with Noisy Labels", "Hyperbolic Image Embeddings", "Hyperbolic Vision Transformers: Combining Improvements in Metric Learning"], "answer_arxiv_id": ["2103.11781", "2207.04873", "2111.00006", "1904.02239", "2203.10833"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_11905"} +{"question": "Could you provide me some research about EfficientZero, an efficient variant of MuZero?", "answer": ["Mastering Atari Games with Limited Data"], "answer_arxiv_id": ["2111.00210"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11906"} +{"question": "Could you provide works on reducing cover time used for subgoal search?", "answer": ["Discovering Options for Exploration by Minimizing Cover Time"], "answer_arxiv_id": ["1903.00606"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_11907"} +{"question": "Could you provide some studies about the robustness of image classification models?", "answer": ["Measuring Robustness to Natural Distribution Shifts in Image Classification"], "answer_arxiv_id": ["2007.00644"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_11908"} +{"question": "Who conducted work focusing on reasoning on multiple events?", "answer": ["COMET-M: Reasoning about Multiple Events in Complex Sentences"], "answer_arxiv_id": ["2305.14617"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_11909"} +{"question": "Which research paper is the most similar to the current work in terms of side-tuning?", "answer": ["Side-Tuning: A Baseline for Network Adaptation via Additive Side Networks"], "answer_arxiv_id": ["1912.13503"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_11910"} +{"question": "Which works have applied the NeRFs principle to imaging sonar?", "answer": ["Neural Implicit Surface Reconstruction using Imaging Sonar", "Neural Volumetric Reconstruction for Coherent Synthetic Aperture Sonar"], "answer_arxiv_id": ["2209.08221", "2306.09909"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_11911"} +{"question": "What works investigated the utilization of pre-trained models for layout-controllable text-to-image synthesis?", "answer": ["SpaText: Spatio-Textual Representation for Controllable Image Generation", "R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image\n Generation", "Dense Text-to-Image Generation with Attention Modulation", "BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained\n Diffusion", "Grounded Text-to-Image Synthesis with Attention Refocusing"], "answer_arxiv_id": ["2211.14305", "2310.08872", "2308.12964", "2307.10816", "2306.05427v2"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_11912"} +{"question": "Which works apply CNNs or/and GNNs for learning the node and structure features in deep learning of graph matching?", "answer": ["Learning Combinatorial Embedding Networks for Deep Graph Matching"], "answer_arxiv_id": ["1904.00597"], "source_meta": {"published_time": "20201216"}, "qid": "AutoScholarQuery_train_11913"} +{"question": "What works have contributed to the development of benchmarks for evaluating code language models?", "answer": ["Evaluating Large Language Models Trained on Code", "Program Synthesis with Large Language Models", "CodeSearchNet Challenge Evaluating the State of Semantic Code Search", "A Static Evaluation of Code Completion by Large Language Models"], "answer_arxiv_id": ["2107.03374", "2108.07732", "1909.09436", "2306.03203"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_11914"} +{"question": "Which works have been dedicated to deriving theoretical guarantees for GANs and their variants?", "answer": ["Generalization and Equilibrium in Generative Adversarial Nets (GANs)", "On the Discrimination-Generalization Tradeoff in GANs", "How Well Generative Adversarial Networks Learn Distributions", "Nonparametric Density Estimation under Adversarial Losses", "Statistical guarantees for generative models without domination", "Some Theoretical Insights into Wasserstein GANs", "PAC-Bayesian Generalization Bounds for Adversarial Generative Models"], "answer_arxiv_id": ["1703.00573", "1711.02771", "1811.03179", "1805.08836v2", "2010.09237", "2006.02682", "2302.08942"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_11915"} +{"question": "Could you provide me some works in which the approach of Straight-Through Estimator is extended to employ various forms of the ReLU activation function?", "answer": ["Blended Coarse Gradient Descent for Full Quantization of Deep Neural Networks", "Deep Learning with Low Precision by Half-wave Gaussian Quantization"], "answer_arxiv_id": ["1808.05240", "1702.00953"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_11916"} +{"question": "Which papers have used the cross-attention mechanism for efficient multi-modal feature fusion?", "answer": ["AutoAlign: Pixel-Instance Feature Aggregation for Multi-Modal 3D Object\n Detection"], "answer_arxiv_id": ["2201.06493"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_11917"} +{"question": "Could you provide me some works about the first applications of NCLMs on zero-shot TTS?", "answer": ["Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers", "Speak, Read and Prompt: High-Fidelity Text-to-Speech with Minimal\n Supervision"], "answer_arxiv_id": ["2301.02111", "2302.03540"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_11918"} +{"question": "What works have done research on optimizing anomaly detection pipelines?", "answer": ["PyODDS: An End-to-end Outlier Detection System with Automated Machine Learning"], "answer_arxiv_id": ["2003.05602"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_11919"} +{"question": "Which paper explores adaptive estimation of strong convexity constant through the inverse of Polyak step size?", "answer": ["Complexity Guarantees for Polyak Steps with Momentum"], "answer_arxiv_id": ["2002.00915"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_11920"} +{"question": "Which works used graph neural networks for dynamics modeling?", "answer": ["Message Passing Neural PDE Solvers"], "answer_arxiv_id": ["2202.03376"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_11921"} +{"question": "What research synthesizes images using the image space of CLIP?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2204.06125"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_11922"} +{"question": "Is there a study that explores a weakly-supervised setup for learning disentangled representations?", "answer": ["Weakly-Supervised Disentanglement Without Compromises"], "answer_arxiv_id": ["2002.02886"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_11923"} +{"question": "Which works have proposed a learning-based exemplar selection method in class-incremental learning?", "answer": ["Mnemonics Training: Multi-Class Incremental Learning without Forgetting"], "answer_arxiv_id": ["2002.10211"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_11924"} +{"question": "What papers follow the approach of greedy pursuit, adding one coordinate at a time until the required support size is reached?", "answer": ["CoSaMP: Iterative Signal Recovery from Incomplete and Inaccurate Samples", "Signal recovery from incomplete and inaccurate measurements via Regularized Orthogonal Matching Pursuit", "Restricted Strong Convexity Implies Weak Submodularity"], "answer_arxiv_id": ["0803.2392", "0712.1360", "1612.00804v2"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_11925"} +{"question": "What papers introduced loss functions to improve the surrogate for learning the objective function?", "answer": ["High-Dimensional Bayesian Optimisation with Variational Autoencoders and Deep Metric Learning", "Combining Latent Space and Structured Kernels for Bayesian Optimization over Combinatorial Spaces"], "answer_arxiv_id": ["2106.03609v3", "2111.01186"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_11926"} +{"question": "Which works explore methods for building KG context for LLMs using path and neighborhood based search methods?", "answer": ["MindMap: Knowledge Graph Prompting Sparks Graph of Thoughts in Large\n Language Models"], "answer_arxiv_id": ["2308.09729"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_11927"} +{"question": "What studies examine the effect of hyperparameters on sharpness at late training times?", "answer": ["The break-even point on optimization trajectories of deep neural networks"], "answer_arxiv_id": ["2002.09572"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_11928"} +{"question": "What are the relevant papers discussing approaches like DreamBooth and Textual Inversion in text-based image generation?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image Models", "Multi-Concept Customization of Text-to-Image Diffusion", "Null-text Inversion for Editing Real Images using Guided Diffusion Models"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2302.12228", "2212.04488", "2211.09794"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11929"} +{"question": "Which work improved the regressor learning framework by quantifying the uncertainty of the regressed coordinates in human pose estimators?", "answer": ["Human Pose Regression with Residual Log-likelihood Estimation"], "answer_arxiv_id": ["2107.11291"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_11930"} +{"question": "Could you point out studies that cover the topic of learning curves in relation to the expected generalization of learning?", "answer": ["The Shape of Learning Curves: a Review"], "answer_arxiv_id": ["2103.10948"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_11931"} +{"question": "Which work extends NeRF to a parameterization by a point cloud?", "answer": ["Point-NeRF: Point-based Neural Radiance Fields"], "answer_arxiv_id": ["2201.08845"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_11932"} +{"question": "What works introduced classifier- and risk-consistent algorithms under the uniform candidate label generation assumption in Identification-based PLL?", "answer": ["Provably Consistent Partial-Label Learning"], "answer_arxiv_id": ["2007.08929"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_11933"} +{"question": "Could you provide some examples of studies that have developed sophisticated augmentation methods and strategies?", "answer": ["mixup: Beyond Empirical Risk Minimization", "Improved Regularization of Convolutional Neural Networks with Cutout", "CutMix: Regularization Strategy to Train Strong Classifiers with\n Localizable Features", "AugMix: A Simple Data Processing Method to Improve Robustness and\n Uncertainty", "AutoAugment: Learning Augmentation Policies from Data", "RandAugment: Practical automated data augmentation with a reduced search\n space", "TrivialAugment: Tuning-free Yet State-of-the-Art Data Augmentation"], "answer_arxiv_id": ["1710.09412", "1708.04552", "1905.04899", "1912.02781", "1805.09501", "1909.13719", "2103.10158"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_11934"} +{"question": "What papers propose adaptive reweighting methods for PINNs, especially with the use of Neural Tangent Kernel?", "answer": ["When and why PINNs fail to train: A neural tangent kernel perspective"], "answer_arxiv_id": ["2007.14527"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_11935"} +{"question": "Could you give me examples of research that proposed the first VQ-VAE based speech codec", "answer": ["Low bit-rate speech coding with VQ-VAE and a WaveNet decoder"], "answer_arxiv_id": ["1910.06464"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_11936"} +{"question": "Which papers highlighted the emergent abilities of Large Language Models (LLM) in reasoning, planning, and learning?", "answer": ["GLM: General Language Model Pretraining with Autoregressive Blank\n Infilling", "LLaMA: Open and Efficient Foundation Language Models", "GPT-4 Technical Report", "Emergent Abilities of Large Language Models"], "answer_arxiv_id": ["2103.10360", "2302.13971", "2303.08774", "2206.07682"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_11937"} +{"question": "Can you point to the studies that proposed algorithms for the ski-rental problem?", "answer": ["Online Algorithms for Multi-shop Ski Rental with Machine Learned Advice", "Online Computation with Untrusted Advice"], "answer_arxiv_id": ["2002.05808", "1905.05655v4"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_11938"} +{"question": "Which studies connect the sufficiently over-parameterized neural network to linear methods around its initialization?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "An Improved Analysis of Training Over-parameterized Deep Neural Networks", "Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Data", "Gaussian Process Behaviour in Wide Deep Neural Networks", "Stochastic Gradient Descent Optimizes Over-parameterized Deep ReLU Networks", "Overparameterized Nonlinear Learning: Gradient Descent Takes the Shortest Path?", "Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent", "Scaling Limits of Wide Neural Networks with Weight Sharing: Gaussian Process Behavior, Gradient Independence, and Neural Tangent Kernel Derivation", "Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "On Lazy Training in Differentiable Programming", "Generalization Guarantees for Neural Networks via Harnessing the Low-rank Structure of the Jacobian", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "Generalization Bounds of Stochastic Gradient Descent for Wide and Deep Neural Networks", "Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networks", "Towards Understanding the Spectral Bias of Deep Learning", "Disentangling feature and lazy training in deep neural networks", "The Interpolation Phase Transition in Neural Networks: Memorization and Generalization under Lazy Training"], "answer_arxiv_id": ["1806.07572", "1906.04688", "1808.01204", "1804.11271", "1811.08888", "1812.10004", "1902.06720", "1902.04760", "1811.03804", "1811.03962", "1812.07956", "1906.05392v2", "1901.08584", "1905.13210", "1909.12292", "1912.01198", "1906.08034", "2007.12826"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_11939"} +{"question": "What are some works where video generation is done by adding temporal convolutional layers and temporal attention layers to the 2D UNet of a pre-trained text-to-image diffusion models?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "Seer: Language Instructed Video Prediction with Latent Diffusion Models", "ModelScope Text-to-Video Technical Report", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation", "Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models"], "answer_arxiv_id": ["2209.14792", "2211.11018", "2303.14897", "2308.06571", "2306.02018", "2303.08320", "2305.10474v3"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_11940"} +{"question": "Which works contributed to the development of large language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "Attention Is All You Need", "Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "OPT: Open Pre-trained Transformer Language Models", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["1810.04805", "1910.10683", "1706.03762", "2005.14165", "2204.02311", "2205.01068", "2211.05100", "2203.02155"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_11941"} +{"question": "What studies developed the method of gradient checkpointing to reduce the tensor cache?", "answer": ["Training Deep Nets with Sublinear Memory Cost", "Memory-Efficient Backpropagation Through Time", "Optimal Gradient Checkpoint Search for Arbitrary Computation Graphs"], "answer_arxiv_id": ["1604.06174", "1606.03401", "1808.00079"], "source_meta": {"published_time": "20220228"}, "qid": "AutoScholarQuery_train_11942"} +{"question": "Which research works proposed token pruning to achieve adaptive sequence length in transformers?", "answer": ["AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "Adaptive Token Sampling For Efficient Vision Transformers"], "answer_arxiv_id": ["2111.15668", "2111.15667"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_11943"} +{"question": "What are the pioneer works that put into practice vision language foundation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_11944"} +{"question": "What work suggests clamping the activation values above a threshold as OOD examples result in abnormal model activation?", "answer": ["ReAct: Out-of-distribution Detection With Rectified Activations"], "answer_arxiv_id": ["2111.12797"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_11945"} +{"question": "Which works focus on label discretization in formulating regression problems as classification tasks?", "answer": ["Estimating Depth from Monocular Images as Classification Using Deep Fully Convolutional Residual Networks", "Deep Ordinal Regression Network for Monocular Depth Estimation"], "answer_arxiv_id": ["1605.02305", "1806.02446"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_11946"} +{"question": "Which paper established the concept of neural ordinary differential equation that is related to the research?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_11947"} +{"question": "Which research is about matrix bandit in the study of MAB?", "answer": ["Efficient Frameworks for Generalized Low-Rank Matrix Bandit Problems"], "answer_arxiv_id": ["2401.07298v1"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11948"} +{"question": "Which works include the demonstration of models' ability to memorize their training data raising privacy concerns?", "answer": ["Extracting Training Data from Large Language Models", "Extracting Training Data from Diffusion Models", "Diffusion Art or Digital Forgery? Investigating Data Replication in\n Diffusion Models", "Understanding and Mitigating Copying in Diffusion Models"], "answer_arxiv_id": ["2012.07805", "2301.13188", "2212.03860", "2305.20086"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_11949"} +{"question": "Which papers discussed the application of imitation learning in various fields such as manipulation, locomotion, and autonomous driving?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperation", "Grasping in the Wild: Learning 6DoF Closed-Loop Grasping from Low-Cost Demonstrations", "Visual Imitation Made Easy", "End to End Learning for Self-Driving Cars"], "answer_arxiv_id": ["1011.0686v3", "1710.04615v2", "1912.04344", "2008.04899", "1604.07316"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_11950"} +{"question": "Are there any studies that utilized LSTM networks and transformers for non-Markovian reward modeling?", "answer": ["Non-Markovian Reward Modelling from Trajectory Labels via Interpretable Multiple Instance Learning", "Preference Transformer: Modeling Human Preferences using Transformers for RL"], "answer_arxiv_id": ["2205.15367", "2303.00957"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_11951"} +{"question": "Which works provide tools for uncertainty estimation and risk control for machine learning algorithms?", "answer": ["Large-scale probabilistic predictors with and without guarantees of validity", "Distribution-Free Predictive Inference For Regression", "Distribution-free binary classification: prediction sets, confidence intervals and calibration", "Distribution-Free, Risk-Controlling Prediction Sets", "Predictive inference with the jackknife+", "Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control"], "answer_arxiv_id": ["1511.00213v2", "1604.04173", "2006.10564", "2101.02703", "1905.02928", "2110.01052"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_11952"} +{"question": "What papers study the convergence of adaptive methods, particularly indicating a faster convergence in practice?", "answer": ["Language Models are Few-Shot Learners", "Understanding the Difficulty of Training Transformers"], "answer_arxiv_id": ["2005.14165", "2004.08249"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_11953"} +{"question": "Can you point out some references where GAN and DDPM-based approaches were used, conditioning on constraints and physical information, to model the TO problem?", "answer": ["TopologyGAN: Topology Optimization Using Generative Adversarial Networks Based on Physical Fields Over the Initial Domain", "Diffusion Models Beat GANs on Topology Optimization"], "answer_arxiv_id": ["2003.04685", "2208.09591"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11954"} +{"question": "Are there studies that shown BERT’s reliance on word order for grammatical role classification?", "answer": ["When classifying grammatical role, BERT doesn't care about word order...\n except when it matters"], "answer_arxiv_id": ["2203.06204"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_11955"} +{"question": "Any works about using image-to-text transforms for building Visual Question Answering Systems?", "answer": ["Retrieval Augmented Visual Question Answering with Outside Knowledge", "KAT: A Knowledge Augmented Transformer for Vision-and-Language", "REVIVE: Regional Visual Representation Matters in Knowledge-Based Visual Question Answering", "Weakly-Supervised Visual-Retriever-Reader for Knowledge-based Question Answering", "An Empirical Study of GPT-3 for Few-Shot Knowledge-Based VQA", "PromptCap: Prompt-Guided Task-Aware Image Captioning"], "answer_arxiv_id": ["2210.03809", "2112.08614", "2206.01201", "2109.04014", "2109.05014", "2211.09699"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_11956"} +{"question": "Which works design auxiliary objectives to introduce external knowledge or extra temporal constraints in text-based temporal reasoning?", "answer": ["Joint Reasoning for Temporal and Causal Relations", "Temporal Common Sense Acquisition with Minimal Supervision", "Improving Event Duration Prediction via Time-aware Pre-training", "Temporal Reasoning on Implicit Events from Distant Supervision"], "answer_arxiv_id": ["1906.04941", "2005.04304", "2011.02610", "2010.12753"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_11957"} +{"question": "Which work proposed a stream-based deep active learning approach?", "answer": ["Improved Algorithms for Neural Active Learning"], "answer_arxiv_id": ["2210.00423"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_11958"} +{"question": "Which studies extend the CO loss function across continuous supports?", "answer": ["Neural Set Function Extensions: Learning with Discrete Functions in High Dimensions"], "answer_arxiv_id": ["2208.04055"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_11959"} +{"question": "Which studies examined the benign overfitting in nonlinear models?", "answer": ["Benign Overfitting without Linearity: Neural Network Classifiers Trained by Gradient Descent for Noisy Linear Data"], "answer_arxiv_id": ["2202.05928"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_11960"} +{"question": "Could you provide some notable references in the field of Imitation Learning?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "Reinforcement and Imitation Learning via Interactive No-Regret Learning", "Deeply AggreVaTeD: Differentiable Imitation Learning for Sequential Prediction", "Learning to Search Better than Your Teacher", "Active Imitation Learning with Noisy Guidance", "Active Imitation Learning from Multiple Non-Deterministic Teachers: Formulation, Challenges, and Algorithms"], "answer_arxiv_id": ["1011.0686v3", "1406.5979", "1703.01030", "1502.02206", "2005.12801", "2006.07777"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_11961"} +{"question": "Which papers discussed deep learning MTL methods suitable for specific types of data?", "answer": ["An Overview of Multi-Task Learning in Deep Neural Networks"], "answer_arxiv_id": ["1706.05098"], "source_meta": {"published_time": "20211101"}, "qid": "AutoScholarQuery_train_11962"} +{"question": "What studies have actively employed the relation between a low-rank matrix estimation and a nonconvex factored estimation problem?", "answer": ["Estimation of (near) low-rank matrices with noise and high-dimensional scaling", "A Convergent Gradient Descent Algorithm for Rank Minimization and Semidefinite Programming from Random Linear Measurements", "Low-rank Solutions of Linear Matrix Equations via Procrustes Flow", "A Unified Computational and Statistical Framework for Nonconvex Low-Rank Matrix Estimation", "Non-square matrix sensing without spurious local minima via the Burer-Monteiro approach", "Finding Low-Rank Solutions via Non-Convex Matrix Factorization, Efficiently and Provably", "Accelerating Ill-Conditioned Low-Rank Matrix Estimation via Scaled Gradient Descent"], "answer_arxiv_id": ["0912.5100v1", "1506.06081", "1507.03566", "1610.05275v1", "1609.03240", "1606.03168", "2005.08898"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_11963"} +{"question": "Could you provide me some works about minimizing a Lipschitz constant?", "answer": ["Lipschitz Bounds and Provably Robust Training by Laplacian Smoothing"], "answer_arxiv_id": ["2006.03712"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_11964"} +{"question": "Could you tell me about some studies that discussed non-patch-based backdoor attacks?", "answer": ["Clean-Label Backdoor Attacks on Video Recognition Models", "Invisible Backdoor Attack with Sample-Specific Triggers"], "answer_arxiv_id": ["2003.03030", "2012.03816"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_11965"} +{"question": "Which studies focus on grounding tasks that require associating words with objects in the world?", "answer": ["Grounded Language Learning in a Simulated 3D World", "Grounded Language Learning Fast and Slow", "Intra-agent speech permits zero-shot task acquisition", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances"], "answer_arxiv_id": ["1706.06551", "2009.01719", "2206.03139", "2204.01691"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_11966"} +{"question": "What works are there on model robustness against ambiguous images in visual event classification?", "answer": ["Ambiguous Images With Human Judgments for Robust Visual Event\n Classification", "Is one annotation enough? A data-centric image classification benchmark\n for noisy and ambiguous label estimation", "Multi-label Iterated Learning for Image Classification with Label\n Ambiguity"], "answer_arxiv_id": ["2210.03102", "2207.06214", "2111.12172"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_11967"} +{"question": "Could you provide me some works about methods that combined free-form text and bounding boxes for image generation?", "answer": ["No Token Left Behind: Explainability-Aided Image Classification and\n Generation", "GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2204.04908", "2301.07093"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_11968"} +{"question": "What research has been done on the pessimistic exploration problem in the context of unsupervised skill discovery?", "answer": ["Learning more skills through optimistic exploration"], "answer_arxiv_id": ["2107.14226"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_11969"} +{"question": "Which studies have applied graph analysis approaches to Non-fungible tokens (NFTs)?", "answer": ["Mapping the NFT revolution: market trends, trade networks, and visual features", "Networks of Ethereum Non-Fungible Tokens: A graph-based analysis of the ERC-721 ecosystem", "NFT Wash Trading Quantifying suspicious behaviour in NFT markets"], "answer_arxiv_id": ["2106.00647", "2110.12545", "2202.03866"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_11970"} +{"question": "What research papers propose methods that build on DSI to improve the model performance?", "answer": ["A Neural Corpus Indexer for Document Retrieval", "Ultron: An Ultimate Retriever on Corpus with a Model-based Indexer"], "answer_arxiv_id": ["2206.02743", "2208.09257"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_11971"} +{"question": "Could you list some papers that use trajectory proposals as learnable anchors in goal-conditional prediction?", "answer": ["ProphNet: Efficient Agent-Centric Motion Forecasting with\n Anchor-Informed Proposals"], "answer_arxiv_id": ["2303.12071"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_11972"} +{"question": "What studies demonstrate data selection practice based on the values of data sources?", "answer": ["Towards Efficient Data Valuation Based on the Shapley Value", "Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning"], "answer_arxiv_id": ["1902.10275", "2110.14049"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_11973"} +{"question": "Can you identify some works that utilize geometric structures such as rectangles and cones within hyperspace to represent entities in KG?", "answer": ["Beta Embeddings for Multi-Hop Logical Reasoning in Knowledge Graphs", "ConE: Cone Embeddings for Multi-Hop Reasoning over Knowledge Graphs"], "answer_arxiv_id": ["2010.11465", "2110.13715"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_11974"} +{"question": "Which works use the successor representation for transfer, lifelong learning, or learning one representation that solve a set of tasks?", "answer": ["Successor Features for Transfer in Reinforcement Learning", "Deep Reinforcement Learning with Successor Features for Navigation across Similar Environments", "Transfer in Deep Reinforcement Learning Using Successor Features and Generalised Policy Improvement", "Universal Successor Features Approximators", "Universal Successor Representations for Transfer Reinforcement Learning", "Fast Task Inference with Variational Intrinsic Successor Features", "Successor Features Combine Elements of Model-Free and Model-based Reinforcement Learning", "A New Representation of Successor Features for Transfer across Dissimilar Environments", "Learning One Representation to Optimize All Rewards"], "answer_arxiv_id": ["1606.05312", "1612.05533", "1901.10964", "1812.07626", "1804.03758", "1906.05030", "1901.11437", "2107.08426", "2103.07945"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_11975"} +{"question": "What work effectively learns visual concepts through natural language supervision?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_11976"} +{"question": "Which work uses self-supervision targets for learning visual correspondence?", "answer": ["Space-Time Correspondence as a Contrastive Random Walk", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2006.14613", "2104.14294"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_11977"} +{"question": "In what research do they propose an expert selection routing strategy where each token can be assigned to a different number of experts?", "answer": ["Mixture-of-Experts with Expert Choice Routing"], "answer_arxiv_id": ["2202.09368"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_11978"} +{"question": "What works made progress in consistency regularization in semi-supervised semantic segmentation?", "answer": ["Semi-supervised semantic segmentation needs strong, varied perturbations", "Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning", "Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic\n Segmentation", "Semi-Supervised Semantic Segmentation with Cross-Consistency Training", "Perturbed and Strict Mean Teachers for Semi-supervised Semantic\n Segmentation", "Augmentation Matters: A Simple-yet-Effective Approach to Semi-supervised\n Semantic Segmentation", "SFC: Shared Feature Calibration in Weakly Supervised Semantic\n Segmentation"], "answer_arxiv_id": ["1906.01916", "2110.05474", "2208.09910", "2003.09005", "2111.12903", "2212.04976", "2401.11719"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_11979"} +{"question": "What research discusses the use of State Space Models as linear RNNs?", "answer": ["HiPPO: Recurrent Memory with Optimal Polynomial Projections", "Efficiently Modeling Long Sequences with Structured State Spaces"], "answer_arxiv_id": ["2008.07669", "2111.00396"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_11980"} +{"question": "Which works introduce transformers into the backbone for 3D feature extraction?", "answer": ["Voxel Transformer for 3D Object Detection", "Voxel Set Transformer: A Set-to-Set Approach to 3D Object Detection from Point Clouds", "Embracing Single Stride 3D Object Detector with Sparse Transformer", "SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds"], "answer_arxiv_id": ["2109.02497", "2203.10314", "2112.06375", "2210.07372"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_11981"} +{"question": "Which papers discuss self-supervised data collection?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction", "URLB: Unsupervised Reinforcement Learning Benchmark", "Exploration by Random Network Distillation"], "answer_arxiv_id": ["1705.05363", "2110.15191", "1810.12894"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_11982"} +{"question": "Which papers utilized LLMs for instruction data generation?", "answer": ["Instruction Tuning with GPT-4", "Principle-Driven Self-Alignment of Language Models from Scratch with Minimal Human Supervision", "Synthetic Prompting: Generating Chain-of-Thought Demonstrations for Large Language Models"], "answer_arxiv_id": ["2304.03277", "2305.03047", "2302.00618"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_11983"} +{"question": "What studies have used the pretext task of inpainting images in self-supervised learning?", "answer": ["Context Encoders: Feature Learning by Inpainting"], "answer_arxiv_id": ["1604.07379"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_11984"} +{"question": "Which works studied retrieval augmentation techniques to improve data efficiency in T2I generation?", "answer": ["KNN-Diffusion: Image Generation via Large-Scale Retrieval", "Re-Imagen: Retrieval-Augmented Text-to-Image Generator"], "answer_arxiv_id": ["2204.02849", "2209.14491"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_11985"} +{"question": "Which studies propose algorithms trying to match minimax error probability lower bounds?", "answer": ["Minimax Optimal Algorithms for Fixed-Budget Best Arm Identification"], "answer_arxiv_id": ["2206.04646"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_11986"} +{"question": "What studies investigated training data reconstruction attacks where the adversary has access only to final model parameters?", "answer": ["Reconstructing Training Data from Trained Neural Networks"], "answer_arxiv_id": ["2206.07758"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_11987"} +{"question": "Which work theoretically studied Semantics-aware adversarial robustness?", "answer": ["Revisiting Adversarial Risk"], "answer_arxiv_id": ["1806.02924v5"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_11988"} +{"question": "Could you provide me the paper that proposed pessimistic off-policy optimization for learning to rank?", "answer": ["Pessimistic Off-Policy Optimization for Learning to Rank"], "answer_arxiv_id": ["2206.02593"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_11989"} +{"question": "Could you mention the research that used anisotropic 3D Gaussians for 3D scene representation?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_11990"} +{"question": "What works inspired the integration of a retrieval module in various NLP and vision tasks?", "answer": ["Improving language models by retrieving from trillions of tokens", "Relational Memory Augmented Language Models", "Memorizing Transformers", "Retrieval-Based Neural Code Generation", "RetrieveGAN: Image Synthesis via Differentiable Patch Retrieval", "Instance-Conditioned GAN", "Re-Imagen: Retrieval-Augmented Text-to-Image Generator"], "answer_arxiv_id": ["2112.04426", "2201.09680", "2203.08913", "1808.10025", "2007.08513", "2109.05070", "2209.14491"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_11991"} +{"question": "What research papers discusses search and sampling decoding algorithms?", "answer": ["Contrastive Decoding: Open-ended Text Generation as Optimization"], "answer_arxiv_id": ["2210.15097"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_11992"} +{"question": "Which papers showed that sequence to sequence models from supervised learning are competitve or better than reinforcement learning-specific memory methods?", "answer": ["Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPs"], "answer_arxiv_id": ["2110.05038"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_11993"} +{"question": "Could you provide me some works about injecting a small amount of learnable parameters into the frozen CLIP backbone?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling", "Visual Prompt Tuning"], "answer_arxiv_id": ["1902.00751", "2110.04544", "2111.03930", "2203.12119"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_11994"} +{"question": "Are there any studies about binarized vision models?", "answer": ["Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1", "PokeBNN: A Binary Pursuit of Lightweight Accuracy"], "answer_arxiv_id": ["1602.02830", "2112.00133"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_11995"} +{"question": "Could you provide me studies that inserted inductive biases of CNNs on Vision Transformer?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction\n without Convolutions", "CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image\n Classification", "Rethinking Spatial Dimensions of Vision Transformers", "CSWin Transformer: A General Vision Transformer Backbone with\n Cross-Shaped Windows"], "answer_arxiv_id": ["2103.14030", "2102.12122", "2103.14899", "2103.16302", "2107.00652"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_11996"} +{"question": "Could you provide me some studies about applications of flow-based SR methods in specialized SR tasks?", "answer": ["Flow-based Kernel Prior with Application to Blind Super-Resolution", "Blind Super-Resolution for Remote Sensing Images via Conditional\n Stochastic Normalizing Flows", "Flow-based Visual Quality Enhancer for Super-resolution Magnetic\n Resonance Spectroscopic Imaging"], "answer_arxiv_id": ["2103.15977", "2210.07751", "2207.10181"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_11997"} +{"question": "Are there any works that include polynomials as operations within a network in the study of equivariant neural networks?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "Equivariant Polynomials for Graph Neural Networks"], "answer_arxiv_id": ["1802.08219", "2302.11556"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_11998"} +{"question": "Which studies propose methods to generate data for augmentation using large scale pretrained class conditional generative models like BigGAN or VQ-VAE2?", "answer": ["Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Generating Diverse High-Fidelity Images with VQ-VAE-2"], "answer_arxiv_id": ["1809.11096", "1906.00446"], "source_meta": {"published_time": "20220814"}, "qid": "AutoScholarQuery_train_11999"} +{"question": "Which study introduced the model-based metric known as BERTScore?", "answer": ["BERTScore: Evaluating Text Generation with BERT"], "answer_arxiv_id": ["1904.09675"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_12000"} +{"question": "What papers focus on the state-of-the-art membership inference attacks where a carefully calibrated threshold for each example is chosen?", "answer": ["White-box vs Black-box: Bayes Optimal Strategies for Membership Inference", "On the Importance of Difficulty Calibration in Membership Inference Attacks", "Membership Inference Attacks From First Principles"], "answer_arxiv_id": ["1908.11229", "2111.08440", "2112.03570"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_12001"} +{"question": "Which studies initially applied large kernels in ConvNets?", "answer": ["Going Deeper with Convolutions", "Rethinking the Inception Architecture for Computer Vision", "Inception-v4, Inception-ResNet and the Impact of Residual Connections on\n Learning"], "answer_arxiv_id": ["1409.4842", "1512.00567", "1602.07261"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_12002"} +{"question": "Which papers studied robotic navigation in the context of embodied AI?", "answer": ["Embodied Question Answering", "Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments", "Habitat: A Platform for Embodied AI Research", "Touchdown: Natural Language Navigation and Spatial Reasoning in Visual Street Environments", "DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion Frames", "REVERIE: Remote Embodied Visual Referring Expression in Real Indoor Environments", "ProcTHOR: Large-Scale Embodied AI Using Procedural Generation"], "answer_arxiv_id": ["1711.11543", "1711.07280", "1904.01201", "1811.12354", "1911.00357", "1904.10151", "2206.06994"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_12003"} +{"question": "Which papers use RL-based prompt optimization?", "answer": ["RLPrompt: Optimizing Discrete Text Prompts with Reinforcement Learning", "Dynamic Prompt Learning via Policy Gradient for Semi-structured\n Mathematical Reasoning"], "answer_arxiv_id": ["2205.12548", "2209.14610"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_12004"} +{"question": "Which studies have used trajectory optimization for controlling force in robots?", "answer": ["Relaxed-Rigidity Constraints: Kinematic Trajectory Optimization and Collision Avoidance for In-Grasp Manipulation"], "answer_arxiv_id": ["1806.00942v2"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_12005"} +{"question": "Which works proposed an adapter to align the new and external control signal with the original internal representation of the pre-trained text-to-image models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_12006"} +{"question": "Which works proposed approaches that maximize the action entropy to encourage exploration?", "answer": ["Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor"], "answer_arxiv_id": ["1801.01290"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_12007"} +{"question": "Which papers showed both the benefits and drawbacks of using prompt engineering in text-image models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2103.00020", "2109.01134", "2203.05557"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_12008"} +{"question": "What papers discuss protein-text modeling?", "answer": ["Galactica: A Large Language Model for Science", "ProtST: Multi-Modality Learning of Protein Sequences and Biomedical\n Texts", "A Text-guided Protein Design Framework", "OntoProtein: Protein Pretraining With Gene Ontology Embedding"], "answer_arxiv_id": ["2211.09085", "2301.12040", "2302.04611", "2201.11147"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_12009"} +{"question": "What are the works about cross-domain IL?", "answer": ["State Alignment-based Imitation Learning", "Robust Learning from Observation with Model Misspecification"], "answer_arxiv_id": ["1911.10947", "2202.06003"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_12010"} +{"question": "Which studies have proposed solutions for the bi-level optimization problem in dataset distillation?", "answer": ["Dataset Distillation", "Dataset Distillation with Infinitely Wide Convolutional Networks", "Dataset Meta-Learning from Kernel Ridge-Regression", "Dataset Distillation using Neural Feature Regression", "Dataset Condensation with Differentiable Siamese Augmentation", "Dataset Condensation with Gradient Matching", "Dataset Distillation by Matching Training Trajectories"], "answer_arxiv_id": ["1811.10959", "2107.13034", "2011.00050", "2206.00719", "2102.08259", "2006.05929", "2203.11932"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_12011"} +{"question": "Could you name some works about layout generation focusing on a simplified setting?", "answer": ["LayoutGAN: Generating Graphic Layouts with Wireframe Discriminators", "LayoutVAE: Stochastic Scene Layout Generation From a Label Set", "READ: Recursive Autoencoders for Document Layout Generation", "Neural Design Network: Graphic Layout Generation with Constraints", "Attribute-conditioned Layout GAN for Automatic Graphic Design"], "answer_arxiv_id": ["1901.06767", "1907.10719", "1909.00302v4", "1912.09421", "2009.05284"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_12012"} +{"question": "Which papers presented precise polynomial-time attacks against classifiers trained on product distributions?", "answer": ["Can Adversarially Robust Learning Leverage Computational Hardness?"], "answer_arxiv_id": ["1810.01407"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_12013"} +{"question": "Which papers measure data complexity using local neighbor density for sample selection?", "answer": ["Iterative Learning with Open-set Noisy Labels", "CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images"], "answer_arxiv_id": ["1804.00092", "1808.01097"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_12014"} +{"question": "Which work established a connection between noise injection and regularization within the online smooth learning setting?", "answer": ["Randomized Smoothing for Stochastic Optimization"], "answer_arxiv_id": ["1103.4296"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12015"} +{"question": "Which work directly optimizes evaluation metric ROUGE in RL fine-tuning stage?", "answer": ["A Deep Reinforced Model for Abstractive Summarization"], "answer_arxiv_id": ["1705.04304"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12016"} +{"question": "What studies proposed the latent diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Scalable Diffusion Models with Transformers"], "answer_arxiv_id": ["2112.10752", "2212.09748"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_12017"} +{"question": "Could you tell me about works that have proposed methods using dynamic weights in terms of the sample’s training loss values?", "answer": ["Unsupervised Label Noise Modeling and Loss Correction"], "answer_arxiv_id": ["1904.11238"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_12018"} +{"question": "What papers discuss the technique of using a small part of old training data while training on new data to reduce forgetting?", "answer": ["On Tiny Episodic Memories in Continual Learning", "DSI++: Updating Transformer Memory with New Documents"], "answer_arxiv_id": ["1902.10486", "2212.09744"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_12019"} +{"question": "What papers used pre-trained model to help robots to understand complex instructions?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances"], "answer_arxiv_id": ["2204.01691"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_12020"} +{"question": "What works are about bi-level optimization in the context of hyperparameter optimization?", "answer": ["Bilevel Programming for Hyperparameter Optimization and Meta-Learning", "Generalized Data Weighting via Class-level Gradient Manipulation", "Gradient-based Bi-level Optimization for Deep Learning: A Survey", "Unbiased Implicit Feedback via Bi-level Optimization", "Structure-aware Protein Self-supervised Learning"], "answer_arxiv_id": ["1806.04910", "2111.00056", "2207.11719", "2206.00147", "2204.04213"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_12021"} +{"question": "What studies discussed the edge of stability for a two-dimensional function ?", "answer": ["Understanding Edge-of-Stability Training Dynamics with a Minimalist Example"], "answer_arxiv_id": ["2210.03294"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_12022"} +{"question": "Which papers have extended the original Diffuser paper through using a separate inverse dynamics model?", "answer": ["Is Conditional Generative Modeling all you need for Decision-Making?"], "answer_arxiv_id": ["2211.15657"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_12023"} +{"question": "Can you name the studies that explored the meta-policy approach in MORL?", "answer": ["Meta-Learning for Multi-objective Reinforcement Learning"], "answer_arxiv_id": ["1811.03376"], "source_meta": {"published_time": "20220816"}, "qid": "AutoScholarQuery_train_12024"} +{"question": "Which works proposed a basic framework for the training of diffusion models through U-Net?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2011.13456"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_12025"} +{"question": "Could you provide me some studies that make use of voxel grids in 3D volumetric reconstructions?", "answer": ["V2V-PoseNet: Voxel-to-Voxel Prediction Network for Accurate 3D Hand and Human Pose Estimation from a Single Depth Map"], "answer_arxiv_id": ["1711.07399"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_12026"} +{"question": "Which datasets are used for generating basketball reports in the field of D2T generation?", "answer": ["Challenges in Data-to-Document Generation", "Revisiting Challenges in Data-to-Text Generation with Fact Grounding"], "answer_arxiv_id": ["1707.08052", "2001.03830"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_12027"} +{"question": "Could you provide me some works that shown multi-domain training increases robustness and generalization in NLP and computer vision?", "answer": ["Pretrained Transformers Improve Out-of-Distribution Robustness", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2004.06100", "2103.00020"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_12028"} +{"question": "Which works employed the technique of combining 2D image synthesis networks with a controllable 3D head proxy?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "Next3D: Generative Neural Texture Rasterization for 3D-Aware Head\n Avatars", "3DFaceShop: Explicitly Controllable 3D-Aware Portrait Generation", "Generative Neural Articulated Radiance Fields", "Disentangled and Controllable Face Image Generation via 3D\n Imitative-Contrastive Learning", "AniFaceGAN: Animatable 3D-Aware Face Image Generation for Video Avatars", "Deferred Neural Rendering: Image Synthesis using Neural Textures", "3D GAN Inversion for Controllable Portrait Image Animation"], "answer_arxiv_id": ["2112.07945", "2211.11208", "2209.05434", "2206.14314", "2004.11660", "2210.06465", "1904.12356", "2203.13441"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_12029"} +{"question": "Could you provide me the reference that discusses about bounded perturbations for NLP models?", "answer": ["On Guaranteed Optimal Robust Explanations for NLP Models"], "answer_arxiv_id": ["2105.03640"], "source_meta": {"published_time": "20221202"}, "qid": "AutoScholarQuery_train_12030"} +{"question": "What studies use a relaxed message-passing algorithm in modern graph-based approaches to infer the relation-aware context?", "answer": ["RU-Net: Regularized Unrolling Network for Scene Graph Generation"], "answer_arxiv_id": ["2205.01297"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_12031"} +{"question": "Could you provide me with a paper that attempts to align the output of LLMs with human expectations by training the models using human preference datasets?", "answer": ["BeaverTails: Towards Improved Safety Alignment of LLM via a\n Human-Preference Dataset"], "answer_arxiv_id": ["2307.04657"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_12032"} +{"question": "What studies have been conducted in the area of raw image and video enhancement?", "answer": ["Learning to See in the Dark"], "answer_arxiv_id": ["1805.01934v1"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_12033"} +{"question": "Could you provide me some research that combined cross-modal contrastive learning with masked auto-encoding for enhanced results?", "answer": ["Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining"], "answer_arxiv_id": ["2302.02318"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_12034"} +{"question": "Can you name the proactive schemes that add signals like perturbation?", "answer": ["OGAN: Disrupting Deepfakes with an Adversarial Attack that Survives Training"], "answer_arxiv_id": ["2006.12247"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_12035"} +{"question": "What works perform finite-sample analysis for RL in infinite-horizon discounted Markov decision processes?", "answer": ["A Finite Time Analysis of Temporal Difference Learning With Linear Function Approximation", "Finite-Sample Analysis for SARSA with Linear Function Approximation", "Finite-Time Error Bounds For Linear Stochastic Approximation and TD Learning", "Policy Mirror Descent for Reinforcement Learning: Linear Convergence, New Sampling Complexity, and Generalized Problem Classes", "The Efficacy of Pessimism in Asynchronous Q-Learning"], "answer_arxiv_id": ["1806.02450", "1902.02234", "1902.00923", "2102.00135", "2203.07368"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_12036"} +{"question": "What works have proposed Riemannian proximal policy optimization for GMMs and leveraged geometry in policy optimization?", "answer": ["Riemannian Proximal Policy Optimization"], "answer_arxiv_id": ["2005.09195"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_12037"} +{"question": "Which papers are about using Pix2Pix and CycleGAN in image-to-image translation tasks?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks"], "answer_arxiv_id": ["1611.07004", "1703.10593"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_12038"} +{"question": "What works focus on enhancing VLM’s compatibility with downstream tasks by incorporating small adapters?", "answer": ["Learning multiple visual domains with residual adapters", "Efficient parametrization of multi-domain deep neural networks", "Parameter-efficient Multi-task Fine-tuning for Transformers via Shared\n Hypernetworks", "Decomposed Soft Prompt Guided Fusion Enhancing for Compositional\n Zero-Shot Learning"], "answer_arxiv_id": ["1705.08045", "1803.10082", "2106.04489", "2211.10681"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_12039"} +{"question": "Which works provided more background information about the Wigner-D matrix?", "answer": ["Geometric and Physical Quantities improve E(3) Equivariant Message Passing", "Equivalence Between SE(3) Equivariant Networks via Steerable Kernels and Group Convolution"], "answer_arxiv_id": ["2110.02905", "2211.15903"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_12040"} +{"question": "What researches have thoroughly studied the scaling of relevant quantities with width in the standard or neural-tangent parameterizations?", "answer": ["Finite Versus Infinite Neural Networks: an Empirical Study"], "answer_arxiv_id": ["2007.15801"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_12041"} +{"question": "Which works laid the foundation for neural networks using the Neural Tangent Kernel (NTK) theory?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "Theoretical insights into the optimization landscape of over-parameterized shallow neural networks", "Gradient Descent Provably Optimizes Over-parameterized Neural Networks", "On Lazy Training in Differentiable Programming"], "answer_arxiv_id": ["1806.07572", "1707.04926", "1810.02054", "1812.07956"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_12042"} +{"question": "What studies proposed algorithms for scaling attention to longer sequences?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Reformer: The Efficient Transformer", "Rethinking Attention with Performers", "General-purpose, long-context autoregressive modeling with Perceiver AR"], "answer_arxiv_id": ["1901.02860", "2001.04451", "2009.14794", "2202.07765"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_12043"} +{"question": "What works have discussed the issues of gradient imbalance in multi-task learning?", "answer": ["GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks", "Gradient Surgery for Multi-Task Learning"], "answer_arxiv_id": ["1711.02257", "2001.06782"], "source_meta": {"published_time": "20220701"}, "qid": "AutoScholarQuery_train_12044"} +{"question": "What paper is about reducing the inference times through a fewer denoising steps?", "answer": ["Controllable Motion Diffusion Model"], "answer_arxiv_id": ["2306.00416"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_12045"} +{"question": "Which papers discussed about homological obstructions in deep learning?", "answer": ["Topological Constraints on Homeomorphic Auto-Encoding", "Topological Obstructions to Autoencoding"], "answer_arxiv_id": ["1812.10783", "2102.08380"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_12046"} +{"question": "Can you name the work that proposes iterative refinement by regressing a pose difference in template-based methods?", "answer": ["DeepIM: Deep Iterative Matching for 6D Pose Estimation"], "answer_arxiv_id": ["1804.00175"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12047"} +{"question": "What works improved the standard method of adversarial training?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy", "Improving Adversarial Robustness via Promoting Ensemble Diversity", "Self-Adaptive Training: beyond Empirical Risk Minimization", "Adversarial Robustness through Local Linearization", "Overfitting in adversarially robust deep learning", "Fast is better than free: Revisiting adversarial training", "Using Pre-Training Can Improve Model Robustness and Uncertainty", "On the Effectiveness of Adversarial Training Against Common Corruptions"], "answer_arxiv_id": ["1901.08573", "1901.08846", "2002.10319", "1907.02610", "2002.11569", "2001.03994", "1901.09960", "2103.02325"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_12048"} +{"question": "What body of work advances the field of lighting, reflection, and depth-based regression in the context of NeRF?", "answer": ["NeRV: Neural Representations for Videos", "Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance\n Fields", "SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single\n Image", "Depth-supervised NeRF: Fewer Views and Faster Training for Free"], "answer_arxiv_id": ["2110.13903", "2112.03907", "2204.00928", "2107.02791"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_12049"} +{"question": "Which works proposed representing the scene as voxel grids, showcasing faster training and better reconstruction quality than the multi-layer perception(MLP) method?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "ReLU Fields: The Little Non-linearity That Could"], "answer_arxiv_id": ["2112.05131", "2111.11215", "2205.10824"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_12050"} +{"question": "Which studies have investigated the usage of public data in the context of private data analysis and how?", "answer": ["Private Learning and Sanitization: Pure vs. Approximate Differential Privacy", "Limits of Private Learning with Access to Public Data", "Privately Answering Classification Queries in the Agnostic PAC Model", "Private Query Release Assisted by Public Data", "Learning from Mixtures of Private and Public Populations", "Leveraging Public Data for Practical Private Query Release"], "answer_arxiv_id": ["1407.2674v1", "1910.11519", "1907.13553", "2004.10941", "2008.00331", "2102.08598v2"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_12051"} +{"question": "Who introduced Neural Light Field Estimation, a method that effectively models complex lighting conditions for virtual object insertion in street scenes?", "answer": ["Neural Light Field Estimation for Street Scenes with Differentiable Virtual Object Insertion"], "answer_arxiv_id": ["2208.09480v1"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_12052"} +{"question": "Are there any models that employ gradient-induced mechanisms for locating co-salient regions in images?", "answer": ["Gradient-Induced Co-Saliency Detection"], "answer_arxiv_id": ["2004.13364"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12053"} +{"question": "Are there researches investigating the relationship between supp​𝑻 and supp​𝚯⋆, even propose conditions for presence of edges in supp​𝚯⋆?", "answer": ["Maximum likelihood estimation in Gaussian models under total positivity", "Learning Graphs with Monotone Topology Properties and Multiple Connected Components"], "answer_arxiv_id": ["1702.04031", "1705.10934v4"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_12054"} +{"question": "Could you provide me with a work that proposes a continuous dynamical model for position encodings?", "answer": ["Learning to Encode Position for Transformer with Continuous Dynamical Model"], "answer_arxiv_id": ["2003.09229"], "source_meta": {"published_time": "20210222"}, "qid": "AutoScholarQuery_train_12055"} +{"question": "What papers proposed efficient learning algorithms for the original batch setting in strategic classification?", "answer": ["Strategic Classification Made Practical", "Generalized Strategic Classification and the Case of Aligned Incentives"], "answer_arxiv_id": ["2103.01826", "2202.04357"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_12056"} +{"question": "What works argue that a reduced model space resulted from G-invariant architectures can achieve better generalization?", "answer": ["Group Equivariant Convolutional Networks", "On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["1602.07576", "1802.03690", "2104.13478"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_12057"} +{"question": "What studies support the development of LLM-centered AI agents?", "answer": ["A Survey on Large Language Model based Autonomous Agents", "The Rise and Potential of Large Language Model Based Agents: A Survey", "JARVIS-1: Open-World Multi-task Agents with Memory-Augmented Multimodal\n Language Models"], "answer_arxiv_id": ["2308.11432", "2309.07864v3", "2311.05997"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_12058"} +{"question": "Which papers discussed the formulation of segmentation as an in-context coloring problem to cope with various segmentation tasks?", "answer": ["Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning", "SegGPT: Segmenting Everything In Context"], "answer_arxiv_id": ["2212.02499", "2304.03284"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_12059"} +{"question": "Could you provide me with some studies that discussed countermeasures in the domain of face recognition?", "answer": ["Data Poisoning Won’t Save You From Facial Recognition"], "answer_arxiv_id": ["2106.14851"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_12060"} +{"question": "What studies use multi-scale methods to estimate motion from coarse to fine scales in scene flow estimation?", "answer": ["Bi-PointFlowNet: Bidirectional Learning for Point Cloud Based Scene Flow Estimation", "HCRF-Flow: Scene Flow from Point Clouds with Continuous High-order CRFs and Position-aware Flow Embedding", "PointPWC-Net: A Coarse-to-Fine Network for Supervised and Self-Supervised Scene Flow Estimation on 3D Point Clouds"], "answer_arxiv_id": ["2207.07522", "2105.07751", "1911.12408v2"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_12061"} +{"question": "What are the works that implement 2D representations in the diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_12062"} +{"question": "What papers rely on data coverage assumption that is even stronger than all-policy concentrability?", "answer": ["Batch Value-function Approximation with Only Realizability"], "answer_arxiv_id": ["2008.04990"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_12063"} +{"question": "Could you list the works about 3D style transfer by applying style transfer on point clouds or meshes?", "answer": ["Learning to Stylize Novel Views", "3D Photo Stylization: Learning to Generate Stylized Novel Views from a\n Single Image", "StyleMesh: Style Transfer for Indoor 3D Scene Reconstructions"], "answer_arxiv_id": ["2105.13509", "2112.00169", "2112.01530"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_12064"} +{"question": "Could you provide me with some works which involve binary classifier approaches in Membership Inference Attacks (MIAs)?", "answer": ["Membership Inference Attacks on Machine Learning: A Survey", "Membership Inference Attacks Against Machine Learning Models", "Privacy Preservation in Federated Learning: An insightful survey from the GDPR Perspective", "ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models"], "answer_arxiv_id": ["2103.07853", "1610.05820", "2011.05411", "1806.01246"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_12065"} +{"question": "Which studies decompose complicated hand-object contact into spatial relations between each hand joint?", "answer": ["A Skeleton-Driven Neural Occupancy Representation for Articulated Hands", "CAMS: CAnonicalized Manipulation Spaces for Category-Level Functional\n Hand-Object Manipulation Synthesis", "ContactGen: Generative Contact Modeling for Grasp Generation", "DexRepNet: Learning Dexterous Robotic Grasping Network with Geometric\n and Spatial Hand-Object Representations"], "answer_arxiv_id": ["2109.11399", "2303.15469", "2310.03740", "2303.09806"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_12066"} +{"question": "What studies discuss pretrained model learning in NLP?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_12067"} +{"question": "What are the works that attempted to generate 3D faces from text?", "answer": ["High-Fidelity 3D Face Generation from Natural Language Descriptions", "Rodin: A Generative Model for Sculpting 3D Digital Avatars Using\n Diffusion"], "answer_arxiv_id": ["2305.03302", "2212.06135"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_12068"} +{"question": "Are there any studies that emphasize detecting failures under distribution shifts?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations", "Wilds: A Benchmark of in-the-Wild Distribution Shifts", "Open-Set Recognition: a Good Closed-Set Classifier is All You Need?"], "answer_arxiv_id": ["1903.12261", "2012.07421", "2110.06207"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_12069"} +{"question": "What are the papers that learn to compact 3D computer-aided design (CAD) models via CSG operations, not relying on any ground-truth primitive assemblies?", "answer": ["UCSG-Net -- Unsupervised Discovering of Constructive Solid Geometry Tree", "CSG-Stump: A Learning Friendly CSG-Like Representation for Interpretable\n Shape Parsing", "ExtrudeNet: Unsupervised Inverse Sketch-and-Extrude for Shape Parsing", "CAPRI-Net: Learning Compact CAD Shapes with Adaptive Primitive Assembly", "SECAD-Net: Self-Supervised CAD Reconstruction by Learning Sketch-Extrude\n Operations"], "answer_arxiv_id": ["2006.09102", "2108.11305", "2209.15632", "2104.05652", "2303.10613"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_12070"} +{"question": "Which work pre-trained a masked language model on digitally rendered text?", "answer": ["Language Modelling with Pixels"], "answer_arxiv_id": ["2207.06991"], "source_meta": {"published_time": "20240808"}, "qid": "AutoScholarQuery_train_12071"} +{"question": "Which works originally proposed Optimization-based methods in the meta-learning framework for few-shot learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["1703.03400"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_12072"} +{"question": "What studies proposed to use periodic activations to artificially inflate the frequency spectrum of the network for better modeling of fine details?", "answer": ["Implicit Neural Representations with Periodic Activation Functions"], "answer_arxiv_id": ["2006.09661"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_12073"} +{"question": "Are there any studies on the space of functions realizable as infinite-width single hidden-layer ReLU nets with bounded weights norm?", "answer": ["A Function Space View of Bounded Norm Infinite Width ReLU Nets: The Multivariate Case"], "answer_arxiv_id": ["1910.01635"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_12074"} +{"question": "What works adopt semantic parsing for structured knowledge grounding tasks?", "answer": ["A Syntactic Neural Model for General-Purpose Code Generation", "Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning", "Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task", "Picard: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models"], "answer_arxiv_id": ["1704.01696", "1709.00103", "1809.08887", "2109.05093"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_12075"} +{"question": "Any works about developing learning algorithms which are complementary to human expertise?", "answer": ["Predict Responsibly: Improving Fairness and Accuracy by Learning to Defer", "Consistent Estimators for Learning to Defer to an Expert", "Towards Unbiased and Accurate Deferral to Multiple Experts", "P"], "answer_arxiv_id": ["1711.06664", "2006.01862v3", "2102.13004", "0704.0320"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_12076"} +{"question": "What papers on diffusion models in computer vision discuss image generation?", "answer": ["Compositional Visual Generation with Composable Diffusion Models"], "answer_arxiv_id": ["2206.01714"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_12077"} +{"question": "What work proposed a method to improve the transformer-based agent by modeling navigation history?", "answer": ["A Recurrent Vision-and-Language BERT for Navigation"], "answer_arxiv_id": ["2011.13922"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12078"} +{"question": "Which works use InfoNCE as a basis for their self-supervised learning algorithms?", "answer": ["Representation Learning with Contrastive Predictive Coding", "On Mutual Information Maximization for Representation Learning"], "answer_arxiv_id": ["1807.03748", "1907.13625"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_12079"} +{"question": "Which papers have proposed deep ensemble techniques for uncertainty quantification in deep learning?", "answer": ["Uncertainty in Neural Networks: Approximately Bayesian Ensembling", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1810.05546v5", "1612.01474"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_12080"} +{"question": "Can you provide examples of autoregressive counterparts of Masked language modeling?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1810.04805", "2005.14165", "1910.10683"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_12081"} +{"question": "What papers discuss line art colorization using user guidance types such as text?", "answer": ["Tag2Pix: Line Art Colorization Using Text Tag With SECat and Changing\n Loss"], "answer_arxiv_id": ["1908.05840"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12082"} +{"question": "What study has compared the convergence rate of simGDA with nonconvex-strongly-concave problems?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems"], "answer_arxiv_id": ["1906.00331"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_12083"} +{"question": "Which works used uniformly bounded Assumption on the stochastic noise of the gradients in combination with choosing a large enough clipping threshold?", "answer": ["Why gradient clipping accelerates training: A theoretical justification for adaptivity", "Improved Analysis of Clipping Algorithms for Non-convex Optimization", "Normalized/Clipped SGD with Perturbation for Differentially Private Non-Convex Optimization"], "answer_arxiv_id": ["1905.11881", "2010.02519", "2206.13033"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_12084"} +{"question": "Which work connects to the research paper with its internal coordinate-based conformer generation framework?", "answer": ["Torsional Diffusion for Molecular Conformer Generation"], "answer_arxiv_id": ["2206.01729"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12085"} +{"question": "What research works focused on achieving efficient global and local processing in ViTs by removing redundant image tokens?", "answer": ["Making Vision Transformers Efficient from A Token Sparsification View"], "answer_arxiv_id": ["2303.08685"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_12086"} +{"question": "Which papers study the different categories of developing differentially private data analysis techniques?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_12087"} +{"question": "Are there any works that approach explanation of deep neural models using backpropagation-based techniques?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "Learning Important Features Through Propagating Activation Differences", "Full-Gradient Representation for Neural Network Visualization", "Axiomatic Attribution for Deep Networks", "Top-down Neural Attention by Excitation Backprop"], "answer_arxiv_id": ["1312.6034", "1704.02685", "1905.00780", "1703.01365", "1608.00507"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_12088"} +{"question": "Which paper proposed a method to approach the cost problem of data-parallel training using Deep Gradient Compression?", "answer": ["Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training"], "answer_arxiv_id": ["1712.01887"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_12089"} +{"question": "What studies have focused on spectral convergence of graph Laplacians?", "answer": ["Consistency of spectral clustering", "Spectral Convergence of the connection Laplacian from random samples"], "answer_arxiv_id": ["0804.0678v1", "1306.1587"], "source_meta": {"published_time": "20210728"}, "qid": "AutoScholarQuery_train_12090"} +{"question": "Which work established an explained component as a calibration metric?", "answer": ["Beyond calibration: estimating the grouping loss of modern neural networks"], "answer_arxiv_id": ["2210.16315v3"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_12091"} +{"question": "What paper introduced CLIP-Dissect for understanding the roles of hidden layer neurons?", "answer": ["CLIP-Dissect: Automatic Description of Neuron Representations in Deep Vision Networks"], "answer_arxiv_id": ["2204.10965"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_12092"} +{"question": "Which publications discuss approaches to 3D generation inspired by the success of 2D diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Denoising Diffusion Implicit Models", "Fantasia3D: Disentangling Geometry and Appearance for High-quality Text-to-3D Content Creation", "DreamBooth3D: Subject-Driven Text-to-3D Generation", "DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Text2Mesh: Text-Driven Neural Stylization for Meshes"], "answer_arxiv_id": ["2112.10752", "2010.02502", "2303.13873", "2303.13508", "2209.14988", "2211.10440", "2112.03221"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_12093"} +{"question": "What are some initial efforts that focused on generating motion based on pre-defined action categories?", "answer": ["Action2Motion: Conditioned Generation of 3D Human Motions", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE"], "answer_arxiv_id": ["2007.15240", "2104.05670"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_12094"} +{"question": "Which works explored determinant neural network models for neural network quantum states?", "answer": ["Deep neural network solution of the electronic Schrödinger equation", "Backflow Transformations via Neural Networks for Quantum Many-Body Wave-Functions"], "answer_arxiv_id": ["1909.08423", "1807.10770"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_12095"} +{"question": "Which works towards examining the pipeline of training from scratch and fine-tuning?", "answer": ["Rethinking the Value of Network Pruning"], "answer_arxiv_id": ["1810.05270"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_12096"} +{"question": "What works have adapted and extended the Neural Radiance Fields (NeRF) to audio-visual applications?", "answer": ["Learning Neural Acoustic Fields", "Learning Signal-Agnostic Manifolds of Neural Fields"], "answer_arxiv_id": ["2204.00628", "2111.06387"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_12097"} +{"question": "Which works have adopted a socratic approach to employ LLMs in reasoning over video captions?", "answer": ["Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language", "Language Models with Image Descriptors are Strong Few-Shot\n Video-Language Learners"], "answer_arxiv_id": ["2204.00598", "2205.10747"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_12098"} +{"question": "Which works adopt a single transformer-based model for different segmentation tasks?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation", "K-Net: Towards Unified Image Segmentation"], "answer_arxiv_id": ["2107.06278", "2106.14855"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_12099"} +{"question": "What papers discuss sending exact or quantized client centroids directly to the server in Federated Clustering?", "answer": ["Heterogeneity for the Win: One-Shot Federated Clustering", "Making AI Forget You: Data Deletion in Machine Learning"], "answer_arxiv_id": ["2103.00697", "1907.05012"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_12100"} +{"question": "Can you suggest some works that accelerate the reverse process in score-based models?", "answer": ["Accelerating Score-based Generative Models with Preconditioned Diffusion Sampling"], "answer_arxiv_id": ["2207.02196"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_12101"} +{"question": "Which works have focused on pruning the weights in the recent years?", "answer": ["Rigging the Lottery: Making All Tickets Winners", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Rethinking the Value of Network Pruning"], "answer_arxiv_id": ["1911.11134", "1803.03635", "1810.05270"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_12102"} +{"question": "Which works insist on delegating higher-order reasoning tasks to LLM as a solution for visual tasks?", "answer": ["Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation\n Models", "Visual Programming: Compositional visual reasoning without training", "VideoChat: Chat-Centric Video Understanding", "Controllable Text-to-Image Generation with GPT-4", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Free-Bloom: Zero-Shot Text-to-Video Generator with LLM Director and LDM Animator"], "answer_arxiv_id": ["2303.04671", "2211.11559", "2305.06355", "2305.18583", "2304.08485", "2304.10592", "2309.14494v1"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_12103"} +{"question": "What group-based works can be applied to domain generalization problem?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Predict then Interpolate: A Simple Algorithm to Learn Stable Classifiers", "Just Train Twice: Improving Group Robustness without Training Group Information", "Learning from others’ mistakes: Avoiding dataset biases without modeling them", "Focus on the Common Good: Group Distributional Robustness Follows", "Examining and Combating Spurious Features under Distribution Shift"], "answer_arxiv_id": ["1911.08731", "2105.12628", "2107.09044", "2012.01300", "2110.02619", "2106.07171"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_12104"} +{"question": "Which works discuss methods associated with blocking and grouping in the field of quantization methods?", "answer": ["Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation", "Trained Quantization Thresholds for Accurate and Efficient Fixed-Point Inference of Deep Neural Networks", "Data-Free Quantization Through Weight Equalization and Bias Correction", "Quantizing deep convolutional networks for efficient inference: A whitepaper", "Memory-Driven Mixed Low Precision Quantization For Enabling Deep Network Inference On Microcontrollers"], "answer_arxiv_id": ["2004.09602", "1903.08066v3", "1906.04721", "1806.08342", "1905.13082"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_12105"} +{"question": "Can you give examples of research that produced models focusing solely on attribute completion accuracy without considering fairness?", "answer": ["Handling Missing Data with Graph Representation Learning", "Learning on Attribute-Missing Graphs", "Graph Convolutional Networks for Graphs Containing Missing Features"], "answer_arxiv_id": ["2010.16418", "2011.01623", "2007.04583"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_12106"} +{"question": "What research thoroughly explored how to mitigate bias without additional annotation?", "answer": ["Fairness Without Demographics in Repeated Loss Minimization", "Weak Proxies are Sufficient and Preferable for Fairness with Missing\n Sensitive Attributes", "Fair Generative Modeling via Weak Supervision", "Heterogeneous Risk Minimization", "Kernelized Heterogeneous Risk Minimization", "Towards Fair Classifiers Without Sensitive Attributes: Exploring Biases\n in Related Features", "What's in a Name? Reducing Bias in Bios without Access to Protected\n Attributes"], "answer_arxiv_id": ["1806.08010", "2210.03175", "1910.12008", "2105.03818", "2110.12425", "2104.14537", "1904.05233"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_12107"} +{"question": "Could you provide me some studies about point-based editing using StyleGAN?", "answer": ["User-Controllable Latent Transformer for StyleGAN Image Layout Editing", "Drag Your GAN: Interactive Point-based Manipulation on the Generative\n Image Manifold"], "answer_arxiv_id": ["2208.12408", "2305.10973"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_12108"} +{"question": "What works proposed improvements for the efficiency of large kernels?", "answer": ["Very Deep Convolutional Networks for Large-Scale Image Recognition", "Rethinking the Inception Architecture for Computer Vision"], "answer_arxiv_id": ["1409.1556", "1512.00567"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_12109"} +{"question": "Could you name some studies that applied LLMs for modeling the relationship between queries and documents like document ranking?", "answer": ["Is ChatGPT Good at Search? Investigating Large Language Models as\n Re-Ranking Agents", "Fine-Tuning LLaMA for Multi-Stage Text Retrieval", "Open-source Large Language Models are Strong Zero-shot Query Likelihood\n Models for Document Ranking"], "answer_arxiv_id": ["2304.09542", "2310.08319", "2310.13243"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_12110"} +{"question": "What study proposed the first interacting particle systems perspective on Transformers?", "answer": ["Understanding and Improving Transformer From a Multi-Particle Dynamic System Point of View"], "answer_arxiv_id": ["1906.02762"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_12111"} +{"question": "What paper introduced label smoothing?", "answer": ["Rethinking the Inception Architecture for Computer Vision"], "answer_arxiv_id": ["1512.00567"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12112"} +{"question": "Which papers investigated the importance of data points in adversarial training?", "answer": ["Attacks Which Do Not Kill Training Make Adversarial Learning Stronger", "How benign is benign overfitting?", "Exploring Memorization in Adversarial Training"], "answer_arxiv_id": ["2002.11242", "2007.04028", "2106.01606"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_12113"} +{"question": "Can you list the works that applied LLM perplexity for token pruning?", "answer": ["LLMLingua: Compressing Prompts for Accelerated Inference of Large\n Language Models", "LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios\n via Prompt Compression"], "answer_arxiv_id": ["2310.05736", "2310.06839"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_12114"} +{"question": "Who worked on designing TTA objectives using properties of classification problems, for example, entropy minimization and class prototypes?", "answer": ["Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation", "Revisiting Realistic Test-Time Training: Sequential Inference and Adaptation by Anchored Clustering", "Attracting and Dispersing: A Simple Approach for Source-free Domain Adaptation"], "answer_arxiv_id": ["2002.08546", "2206.02721", "2205.04183"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_12115"} +{"question": "Any papers about how small sample size can increase standard error when using adversarial examples?", "answer": ["Understanding and Mitigating the Tradeoff Between Robustness and Accuracy"], "answer_arxiv_id": ["2002.10716"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_12116"} +{"question": "Which papers investigated the dynamics of learning via gradient descent in artificial neural networks?", "answer": ["Exact solutions to the nonlinear dynamics of learning in deep linear neural networks", "The interplay between randomness and structure during learning in RNNs"], "answer_arxiv_id": ["1312.6120", "2006.11036"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_12117"} +{"question": "Which papers study multimodal evaluation tasks like image captioning, image question answering, visual reasoning, and video question answering?", "answer": ["Microsoft COCO Captions: Data Collection and Evaluation Server", "nocaps: novel object captioning at scale", "VQA: Visual Question Answering", "Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering", "GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering"], "answer_arxiv_id": ["1504.00325", "1812.08658v3", "1505.00468", "1612.00837", "1902.09506"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_12118"} +{"question": "Can you list studies devoted to estimating divergences between pairs of distributions in classical machine learning?", "answer": ["Estimating divergence functionals and the likelihood ratio by convex risk minimization"], "answer_arxiv_id": ["0809.0853"], "source_meta": {"published_time": "20221107"}, "qid": "AutoScholarQuery_train_12119"} +{"question": "What research provides a method for global sample complexity specific to the tabular setting?", "answer": ["Policy Mirror Descent for Reinforcement Learning: Linear Convergence, New Sampling Complexity, and Generalized Problem Classes"], "answer_arxiv_id": ["2102.00135"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_12120"} +{"question": "What publications proposed local image editing techniques involving multi-step blending in the masked region?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images"], "answer_arxiv_id": ["2111.14818"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_12121"} +{"question": "Could you name some works that designed powerful positional and structural embeddings for Graph Transformers?", "answer": ["Do Transformers Really Perform Bad for Graph Representation?", "Rethinking Graph Transformers with Spectral Attention", "Recipe for a General, Powerful, Scalable Graph Transformer"], "answer_arxiv_id": ["2106.05234", "2106.03893", "2205.12454"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12122"} +{"question": "Which works elaborate on the application of point-based methods in 3D understanding?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "KPConv: Flexible and Deformable Convolution for Point Clouds", "Rethinking Network Design and Local Geometry in Point Cloud: A Simple\n Residual MLP Framework"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1904.08889", "2202.07123"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_12123"} +{"question": "Which work proposed the intrinsic-curiosity-module (ICM)?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction"], "answer_arxiv_id": ["1705.05363"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12124"} +{"question": "What studies proposed the most related effort of explicit regulations along the entire neural dynamics, interval bound propagation (IBP)?", "answer": ["On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models", "Achieving Verified Robustness to Symbol Substitutions via Interval Bound Propagation", "Towards Stable and Efficient Training of Verifiably Robust Neural Networks"], "answer_arxiv_id": ["1810.12715", "1909.01492", "1906.06316"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_12125"} +{"question": "What works have proposed classifier-free guidance methods to improve text to image alignment?", "answer": ["Classifier-Free Diffusion Guidance", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2207.12598", "2112.10741", "2205.11487"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_12126"} +{"question": "Which works discuss low-dimensional kernel learning with convolutional kernels?", "answer": ["Locality defeats the curse of dimensionality in convolutional teacher-student scenarios", "Approximation and Learning with Deep Convolutional Models: a Kernel Perspective"], "answer_arxiv_id": ["2106.08619v3", "2102.10032"], "source_meta": {"published_time": "20230118"}, "qid": "AutoScholarQuery_train_12127"} +{"question": "Could you provide some references that applied gradient-based optimization on differentiable proxy models for offline model-based optimization?", "answer": ["Conservative Objective Models for Effective Offline Model-Based Optimization", "RoMA: Robust Model Adaptation for Offline Model-based Optimization", "Offline Model-Based Optimization via Normalized Maximum Likelihood Estimation", "Bidirectional Learning for Offline Infinite-width Model-based Optimization"], "answer_arxiv_id": ["2107.06882", "2110.14188", "2102.07970", "2209.07507"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_12128"} +{"question": "Which works proposed some of the earliest fully-convolutional networks that fuse both deep and shallow features for improved segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_12129"} +{"question": "Which works uses Rényi divergence in BCPO paradigm?", "answer": ["Policy Optimization via Importance Sampling"], "answer_arxiv_id": ["1809.06098"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_12130"} +{"question": "Could you provide studies about tensor decomposition using a modified GD?", "answer": ["Beyond Lazy Training for Over-parameterized Tensor Decomposition", "Understanding Deflation Process in Over-parametrized Tensor Decomposition"], "answer_arxiv_id": ["2010.11356", "2106.06573"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_12131"} +{"question": "What studies investigated the optimal granularity of word subtokenization?", "answer": ["A Call for Prudent Choice of Subword Merge Operations in Neural Machine Translation"], "answer_arxiv_id": ["1905.10453"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_12132"} +{"question": "Which work utilizes stacked MLP layers with doubly residual learning for time series forecasting?", "answer": ["N-BEATS: Neural basis expansion analysis for interpretable time series forecasting"], "answer_arxiv_id": ["1905.10437"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_12133"} +{"question": "What papers discussed self-supervised learning with different image augmentations?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_12134"} +{"question": "Which works have applied RL to generate 2D molecular graphs in molecular design?", "answer": ["Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation", "Optimization of Molecules via Deep Reinforcement Learning", "Multi-Objective Molecule Generation using Interpretable Substructures", "Guiding Deep Molecular Optimization with Genetic Exploration", "Hit and Lead Discovery with Explorative RL and Fragment-based Molecule Generation"], "answer_arxiv_id": ["1806.02473", "1810.08678", "2002.03244", "2007.04897", "2110.01219"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_12135"} +{"question": "Which research introduced the concept of Maximum Softmax Probability (MSP) in the context of failure detection?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks"], "answer_arxiv_id": ["1610.02136"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_12136"} +{"question": "What papers propose ways to supplement substitution data in data-free knowledge distillation?", "answer": ["Dream Distillation: A Data-Independent Model Compression Framework", "Large-Scale Generative Data-Free Distillation", "Data-Free Network Quantization With Adversarial Knowledge Distillation", "Explicit and Implicit Knowledge Distillation via Unlabeled Data"], "answer_arxiv_id": ["1905.07072", "2012.05578", "2005.04136", "2302.08771"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_12137"} +{"question": "Any research about the high computational demand challenges in methods using neural fields?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "TensoRF: Tensorial Radiance Fields", "Neural Sparse Voxel Fields"], "answer_arxiv_id": ["2201.05989", "2112.05131", "2203.09517", "2007.11571"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_12138"} +{"question": "Which papers introduced the concept of neural implicit representations?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "Local Implicit Grid Representations for 3D Scenes"], "answer_arxiv_id": ["1901.05103", "1812.03828", "2003.08981"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_12139"} +{"question": "Could you provide me the study which presents significant advancements in the Stable Diffusion model that is fine-tuned for text-guided inpainting?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_12140"} +{"question": "Which works proposed an architecture for image segmentation based on the mask classification approach, thereby improving both semantic and panoptic segmentation settings?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation"], "answer_arxiv_id": ["2107.06278"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_12141"} +{"question": "Could you provide me some studies where Vision Transformers were adopted in image restoration tasks?", "answer": ["SwinIR: Image Restoration Using Swin Transformer", "Activating More Pixels in Image Super-Resolution Transformer", "Dual Aggregation Transformer for Image Super-Resolution"], "answer_arxiv_id": ["2108.10257", "2205.04437", "2308.03364"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12142"} +{"question": "What studies are about the three types of tabular data approaches for neural networks?", "answer": ["Deep Neural Networks and Tabular Data: A Survey"], "answer_arxiv_id": ["2110.01889"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_12143"} +{"question": "Where can I find a more comprehensive overview of quantization methods?", "answer": ["A White Paper on Neural Network Quantization"], "answer_arxiv_id": ["2106.08295"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_12144"} +{"question": "What researches address the issues of sinusoidal function in positional encoding in Transformer architecture?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "1907.11692"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_12145"} +{"question": "Which works combined the method of projection steps after each proposal with the discretization technique to enforce the constraints of the manifold?", "answer": ["Hybrid Monte Carlo methods for sampling probability measures on submanifolds"], "answer_arxiv_id": ["1807.02356"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_12146"} +{"question": "Which works propose enhancing a multi-modal model through stage-wise pre-training using image-only and text-only data?", "answer": ["VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts"], "answer_arxiv_id": ["2111.02358"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_12147"} +{"question": "What research has tried to improve the efficiency of Transformers?", "answer": ["Efficient Transformers: A Survey"], "answer_arxiv_id": ["2009.06732"], "source_meta": {"published_time": "20220117"}, "qid": "AutoScholarQuery_train_12148"} +{"question": "What works have developed oriented bounding box (OBB) detectors for remote sensing object detection?", "answer": ["SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated\n Objects", "R3Det: Refined Single-Stage Detector with Feature Refinement for\n Rotating Object", "Oriented R-CNN for Object Detection", "ReDet: A Rotation-equivariant Detector for Aerial Object Detection", "Dynamic Anchor Learning for Arbitrary-Oriented Object Detection"], "answer_arxiv_id": ["1811.07126", "1908.05612", "2108.05699", "2103.07733", "2012.04150"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_12149"} +{"question": "What works have been conducted on soft decision trees?", "answer": ["Adaptive Neural Trees"], "answer_arxiv_id": ["1807.06699"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_12150"} +{"question": "Which works discuss adaptive stepsize schemes in minimax optimization, particularly in the training of GANs?", "answer": ["NIPS 2016 Tutorial: Generative Adversarial Networks", "A Variational Inequality Perspective on Generative Adversarial Networks"], "answer_arxiv_id": ["1701.00160", "1802.10551"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_12151"} +{"question": "What studies propose a framework for evolving Implicit Neural Spatial Representation weights over time?", "answer": ["A Level Set Theory for Neural Implicit Evolution under Explicit Flows", "Evolutional Deep Neural Network", "Neural Galerkin Scheme with Active Learning for High-Dimensional Evolution Equations"], "answer_arxiv_id": ["2204.07159v2", "2103.09959", "2203.01360"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12152"} +{"question": "What papers propose to work around the shortcut learning problem by shaping the last-layer classifier or introducing penalty terms?", "answer": ["Self-Challenging Improves Cross-Domain Generalization", "Evading the Simplicity Bias: Training a Diverse Set of Models Discovers Solutions with Superior OOD Generalization"], "answer_arxiv_id": ["2007.02454", "2105.05612"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_12153"} +{"question": "What papers propose gp-vae and markovian-gpvae that are scalable especially for time-series data?", "answer": ["GP-VAE: Deep Probabilistic Time Series Imputation", "Markovian Gaussian Process Variational Autoencoders"], "answer_arxiv_id": ["1907.04155", "2207.05543"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_12154"} +{"question": "Which research uses VQ-VAE to discover skills in model-based RL?", "answer": ["Choreographer: Learning and Adapting Skills in Imagination"], "answer_arxiv_id": ["2211.13350"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_12155"} +{"question": "Which papers talked about neural autoencoders generating uncorrelated representations without adding additional loss terms?", "answer": ["VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Isolating Sources of Disentanglement in Variational Autoencoders", "Domain Separation Networks"], "answer_arxiv_id": ["2105.04906", "1802.04942", "1608.06019"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_12156"} +{"question": "Could you give examples of recurrent-free based models proposed in the context of deterministic predictive models?", "answer": ["MIMO Is All You Need : A Strong Multi-In-Multi-Out Baseline for Video\n Prediction"], "answer_arxiv_id": ["2212.04655"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12157"} +{"question": "Can you cite some papers where synthetic scenario databases are generated based on rules?", "answer": ["SUMMIT: A Simulator for Urban Driving in Massive Mixed Traffic", "MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learning", "CityFlow: A Multi-Agent Reinforcement Learning Environment for Large Scale City Traffic Scenario"], "answer_arxiv_id": ["1911.04074", "2109.12674", "1905.05217"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_12158"} +{"question": "What research portrayed the key role of self-attention in the transformer architecture?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_12159"} +{"question": "What studies concerned with the task of dressing-up in the field of garment manipulation?", "answer": ["One Policy to Dress Them All: Learning to Dress People with Diverse\n Poses and Garments"], "answer_arxiv_id": ["2306.12372"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_12160"} +{"question": "Could you provide me with a paper that offers a detailed background about weak supervision frameworks?", "answer": ["A Survey on Programmatic Weak Supervision"], "answer_arxiv_id": ["2202.05433"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_12161"} +{"question": "What works proposed to generate synthetic datasets by matching gradients between two surrogate models trained on distilled dataset and the real dataset?", "answer": ["Dataset Condensation with Gradient Matching", "Dataset Condensation with Differentiable Siamese Augmentation"], "answer_arxiv_id": ["2006.05929", "2102.08259"], "source_meta": {"published_time": "20221119"}, "qid": "AutoScholarQuery_train_12162"} +{"question": "What works propose a conditional neural field for shape and appearance editing in NeRF editing?", "answer": ["Editing Conditional Radiance Fields"], "answer_arxiv_id": ["2105.06466"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_12163"} +{"question": "Which research paper introduced a method called PointFusion, which extracts RGB and depth features independently, for seen object pose estimation?", "answer": ["PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation"], "answer_arxiv_id": ["1711.10871"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_12164"} +{"question": "Which work presented the idea of matching the execution results of generated solutions for minimum Bayes risk selection to mitigate unreliable behavior of LLMs?", "answer": ["Natural Language to Code Translation with Execution"], "answer_arxiv_id": ["2204.11454"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_12165"} +{"question": "Could you provide me some research works that derived exact dynamical equations for a GLN?", "answer": ["The Neural Race Reduction: Dynamics of Abstraction in Gated Networks"], "answer_arxiv_id": ["2207.10430v1"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_12166"} +{"question": "Can you name the works that are about Neural Module Networks?", "answer": ["Neural Module Networks", "Explainable Neural Computation via Stack Neural Module Networks", "Self-Assembling Modular Networks for Interpretable Multi-Hop Reasoning", "Neural Module Networks for Reasoning over Text", "Obtaining Faithful Interpretations from Compositional Neural Networks", "Weakly Supervised Neuro-Symbolic Module Networks for Numerical Reasoning", "VGNMN: Video-grounded Neural Module Networks for Video-Grounded Dialogue Systems"], "answer_arxiv_id": ["1511.02799", "1807.08556", "1909.05803", "1912.04971", "2005.00724", "2101.11802", "2104.07921"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_12167"} +{"question": "What works mention the application of an encoder-decoder network to regress pixel-level dense correspondences?", "answer": ["Pix2Pose: Pixel-Wise Coordinate Regression of Objects for 6D Pose\n Estimation", "DPOD: 6D Pose Object Detector and Refiner"], "answer_arxiv_id": ["1908.07433", "1902.11020"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12168"} +{"question": "What research papers proposed concepts of iterative teaching in machine teaching?", "answer": ["Iterative Machine Teaching", "Towards Black-box Iterative Machine Teaching", "Preference-Based Batch and Sequential Teaching: Towards a Unified View of Models", "Machine Teaching of Active Sequential Learners", "An Optimal Control Approach to Sequential Machine Teaching", "Iterative Teaching by Label Synthesis", "Iterative Teaching by Data Hallucination", "Nonparametric Iterative Machine Teaching"], "answer_arxiv_id": ["1705.10470", "1710.07742", "1910.10944", "1809.02869", "1810.06175", "2110.14432", "2210.17467v2", "2306.03007"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_12169"} +{"question": "What works have explored the procedure of outlier exposure by using OOD data from other sources in OCC and OOD algorithms?", "answer": ["Exploring the Limits of Out-of-Distribution Detection", "Deep Anomaly Detection with Outlier Exposure"], "answer_arxiv_id": ["2106.03004", "1812.04606"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12170"} +{"question": "What works have improved pre-training performance for 3D point clouds through data augmentation or integration of cross-modal information?", "answer": ["Self-Supervised Pretraining of 3D Features on any Point-Cloud", "Masked Scene Contrast: A Scalable Framework for Unsupervised 3D\n Representation Learning", "CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D\n Point Cloud Understanding", "CLIP2Point: Transfer CLIP to Point Cloud Classification with Image-Depth\n Pre-training", "CLIP$^2$: Contrastive Language-Image-Point Pretraining from Real-World\n Point Cloud Data"], "answer_arxiv_id": ["2101.02691", "2303.14191", "2203.00680", "2210.01055", "2303.12417"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_12171"} +{"question": "Can you provide me with papers that discuss multi-modal learning?", "answer": ["Multimodal Machine Learning: A Survey and Taxonomy"], "answer_arxiv_id": ["1705.09406"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_12172"} +{"question": "What research work introduced Poisson Functional Representation (PFR) in the Rate-distortion Efficient Coding (REC) problem?", "answer": ["Strong Functional Representation Lemma and Applications to Coding Theorems"], "answer_arxiv_id": ["1701.02827"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_12173"} +{"question": "Have there been any works that proposed Maximization of expectation of entropy over all samples?", "answer": ["Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learning"], "answer_arxiv_id": ["1908.02983"], "source_meta": {"published_time": "20220515"}, "qid": "AutoScholarQuery_train_12174"} +{"question": "Which studies are based on Mask Transformers for panoptic segmentation?", "answer": ["MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation"], "answer_arxiv_id": ["2012.00759", "2107.06278", "2112.01527", "2206.08948"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_12175"} +{"question": "Can you cite a work that made attempts in one-shot domain adaptation using the capabilities of vision-language networks like CLIP?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_12176"} +{"question": "Which studies expanded MLLMs to visual grounding tasks?", "answer": ["Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Kosmos-2: Grounding Multimodal Large Language Models to the World"], "answer_arxiv_id": ["2306.15195", "2306.14824"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_12177"} +{"question": "Which papers focused on image-to-image translation using text-to-image diffusion?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "Semantic Image Synthesis with Spatially-Adaptive Normalization", "Palette: Image-to-Image Diffusion Models"], "answer_arxiv_id": ["1611.07004", "1903.07291", "2111.05826"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_12178"} +{"question": "What works present methods that quantize the trajectory space to a set of anchors?", "answer": ["CoverNet: Multimodal Behavior Prediction using Trajectory Sets", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction"], "answer_arxiv_id": ["1911.10298", "1910.05449"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_12179"} +{"question": "Are there any studies about scene reconstruction in the context of Neural Implicit Representations?", "answer": ["Convolutional Occupancy Networks", "Local Implicit Grid Representations for 3D Scenes", "Deep Local Shapes: Learning Local SDF Priors for Detailed 3D\n Reconstruction", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM", "MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface\n Reconstruction"], "answer_arxiv_id": ["2003.04618", "2003.08981", "2003.10983", "2112.12130", "2302.03594", "2206.00665"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_12180"} +{"question": "Which works have leveraged IR-based graph representation?", "answer": ["Neural Code Comprehension: A Learnable Representation of Code Semantics", "ProGraML: Graph-based Deep Learning for Program Optimization and Analysis"], "answer_arxiv_id": ["1806.07336", "2003.10536"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_12181"} +{"question": "What papers discuss the design of annotation tasks, considering the socio-cultural backgrounds of annotators?", "answer": ["CrowdWorkSheets: Accounting for Individual and Collective Identities Underlying Crowdsourced Dataset Annotation"], "answer_arxiv_id": ["2206.08931"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_12182"} +{"question": "Which studies have worked on adaptive selection of active learning algorithms for linear models?", "answer": ["Dynamic Ensemble Active Learning: A Non-Stationary Bandit with Expert Advice"], "answer_arxiv_id": ["1810.07778"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_12183"} +{"question": "Which papers proposed methods to mitigate the problem of environment scarcity in Vision-and-Language Navigation?", "answer": ["Learning to Navigate Unseen Environments: Back Translation with Environmental Dropout", "CLEAR: Improving Vision-Language Navigation with Cross-Lingual, Environment-Agnostic Representations", "Environment-agnostic Multitask Learning for Natural Language Grounded Navigation", "EnvEdit: Environment Editing for Vision-and-Language Navigation", "Vision-Language Navigation with Random Environmental Mixup", "Learning from Unlabeled 3D Environments for Vision-and-Language Navigation", "Pathdreamer: A World Model for Indoor Navigation", "Simple and Effective Synthesis of Indoor 3D Scenes"], "answer_arxiv_id": ["1904.04195", "2207.02185", "2003.00443", "2203.15685", "2106.07876", "2208.11781", "2105.08756", "2204.02960"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12184"} +{"question": "Which papers propose tasks evaluating language models against human-like linguistic capabilities?", "answer": ["What do Language Models know about word senses? Zero-Shot WSD with Language Models and Domain Inventories"], "answer_arxiv_id": ["2302.03353"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_12185"} +{"question": "Which papers are about the success of DP-SGD fine-tuning of pretrained large language models?", "answer": ["Large Language Models Can Be Strong Differentially Private Learners", "Differentially Private Fine-tuning of Language Models"], "answer_arxiv_id": ["2110.05679", "2110.06500"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_12186"} +{"question": "Which papers introduce the use of the E-Branchformer as the encoder and Transformer as the decoder?", "answer": ["OWSM v3.1: Better and Faster Open Whisper-Style Speech Models based on\n E-Branchformer"], "answer_arxiv_id": ["2401.16658"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_12187"} +{"question": "Any works about learning different aspects such as actions, skills, macro-actions, rules, and guidance strategies for efficient planning?", "answer": ["Learning Strips Action Models with Classical Planning", "Focused Model-Learning and Planning for Non-Gaussian Continuous State-Action Systems", "Learning compositional models of robot skills for task and motion planning", "Macro-FF: Improving AI Planning with Automatically Learned Macro-Operators", "Learning sparse relational transition models", "PG3: Policy-Guided Planning for Generalized Policy Generation", "Active model learning and diverse action sampling for task and motion planning", "Learning to guide task and motion planning using score-space representation"], "answer_arxiv_id": ["1903.01153", "1607.07762", "2006.06444", "1109.2154", "1810.11177", "2204.10420", "1803.00967", "1807.09962"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12188"} +{"question": "Can you share works that explore the consequences of unchecked marginilization in evaluations?", "answer": ["AI’s Regimes of Representation: A Community-centered Study of Text-to-Image Models in South Asia"], "answer_arxiv_id": ["2305.11844"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_12189"} +{"question": "What works have been done to promote skill diversity using empowerment?", "answer": ["Variational Option Discovery Algorithms", "Diversity is All You Need: Learning Skills without a Reward Function", "Learn Goal-Conditioned Policy with Intrinsic Motivation for Deep Reinforcement Learning"], "answer_arxiv_id": ["1807.10299", "1802.06070", "2104.05043"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12190"} +{"question": "Which research works proposed an approach to delete entire concepts from a diffusion model?", "answer": ["Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models", "Ablating Concepts in Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2211.05105", "2303.13516"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_12191"} +{"question": "Could you provide me with some studies on conformal inference?", "answer": ["Distribution-Free Predictive Inference For Regression", "Testing for Outliers with Conformal p-values"], "answer_arxiv_id": ["1604.04173", "2104.08279"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_12192"} +{"question": "What works proposed to learn an alignment network to encourage invariance for point cloud analysis?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation"], "answer_arxiv_id": ["1612.00593"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_12193"} +{"question": "Can you provide studies that discussed improving training efficiency through approximate state abstractions allowing for a near-optimal policy?", "answer": ["Near Optimal Behavior via Approximate State Abstraction"], "answer_arxiv_id": ["1701.04113v1"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_12194"} +{"question": "Which research papers demonstrated top-down approaches for estimating 3D planes from single images?", "answer": ["PlaneNet: Piece-wise Planar Reconstruction from a Single RGB Image", "PlaneRCNN: 3D Plane Detection and Reconstruction from a Single Image"], "answer_arxiv_id": ["1804.06278", "1812.04072"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_12195"} +{"question": "Could you name some papers that adopt adversarial methods in multi-source domain adaptation?", "answer": ["Deep Cocktail Network: Multi-source Unsupervised Domain Adaptation with Category Shift", "Multi-source Distilling Domain Adaptation"], "answer_arxiv_id": ["1803.00830", "1911.11554"], "source_meta": {"published_time": "20220201"}, "qid": "AutoScholarQuery_train_12196"} +{"question": "Which paper proposed pre-aligning the 3D space with hand-object global poses to support SDF prediction?", "answer": ["AlignSDF: Pose-Aligned Signed Distance Fields for Hand-Object\n Reconstruction"], "answer_arxiv_id": ["2207.12909"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_12197"} +{"question": "Which papers emphasize the importance of using information from the end of training to construct the mask in IMP?", "answer": ["Deconstructing Lottery Tickets: Zeros, Signs, and the Supermask", "Linear Mode Connectivity and the Lottery Ticket Hypothesis", "What’s Hidden in a Randomly Weighted Neural Network?", "Winning the Lottery with Continuous Sparsification", "Rare Gems: Finding Lottery Tickets at Initialization"], "answer_arxiv_id": ["1905.01067", "1912.05671", "1911.13299", "1912.04427", "2202.12002"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_12198"} +{"question": "Which papers claim that recent unsupervised learning methods trained with natural images have great generalization capability on unseen datasets?", "answer": ["A Spatial Guided Self-supervised Clustering Network for Medical Image Segmentation", "Unsupervised Object Segmentation by Redrawing", "Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization"], "answer_arxiv_id": ["2107.04934v1", "1905.13539", "2205.07839"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_12199"} +{"question": "Which studies emphasize the importance of step-by-step reasoning within large language models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Automatic Chain of Thought Prompting in Large Language Models", "Prompting GPT-3 To Be Reliable", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Teaching Algorithmic Reasoning via In-context Learning", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2205.10625", "2210.03493", "2210.09150", "2203.11171", "2211.09066", "2107.13586v1", "2210.03629"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_12200"} +{"question": "In what paper was the technique of utilizing a covariance matrix adaptation to update a population distribution introduced?", "answer": ["Covariance Matrix Adaptation for the Rapid Illumination of Behavior Space"], "answer_arxiv_id": ["1912.02400"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_12201"} +{"question": "Could you provide me some papers that employ a more intricate strategy for an adaptive schedule in curriculum learning?", "answer": ["Adaptive Scheduling for Multi-Task Learning", "Balancing Training for Multilingual Neural Machine Translation", "Bandits Don’t Follow Rules: Balancing Multi-Facet Machine Translation with Multi-Armed Bandits"], "answer_arxiv_id": ["1909.06434", "2004.06748", "2110.06997"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12202"} +{"question": "Which works demonstrate the capability of the attention modules in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2112.10752"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_12203"} +{"question": "Which papers reported the sensitivity of GNNs to the extent of homophily in graph data?", "answer": ["MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing", "Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters", "Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods", "Graph Neural Networks with Heterophily", "Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs"], "answer_arxiv_id": ["1905.00067", "2008.08692", "2110.14446", "2009.13566", "2006.11468v2"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12204"} +{"question": "Which studies focus on vision and language learning using both text and video for representation learning?", "answer": ["End-to-End Learning of Visual Representations from Uncurated\n Instructional Videos", "VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text\n Understanding", "Egocentric Video-Language Pretraining", "HierVL: Learning Hierarchical Video-Language Embeddings", "VAST: A Vision-Audio-Subtitle-Text Omni-Modality Foundation Model and\n Dataset", "VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and\n Dataset"], "answer_arxiv_id": ["1912.06430", "2109.14084", "2206.01670", "2301.02311", "2305.18500", "2304.08345"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_12205"} +{"question": "What studies used methods of aligning support and query samples by exploiting local information on feature maps?", "answer": ["Revisiting Local Descriptor based Image-to-Class Measure for Few-shot Learning", "Cross Attention Network for Few-shot Classification"], "answer_arxiv_id": ["1903.12290", "1910.07677"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_12206"} +{"question": "Which works have integrated some latent evolution under ODEs?", "answer": ["GRU-ODE-Bayes: Continuous modeling of sporadically-observed time series"], "answer_arxiv_id": ["1905.12374"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_12207"} +{"question": "Which papers used ℓ1 regularization to encourage structural properties like piecewise continuity in the context of 'fused lasso'?", "answer": ["The solution path of the generalized lasso", "Optimal rates for total variation denoising", "On the prediction performance of the Lasso"], "answer_arxiv_id": ["1005.1971", "1603.09388", "1402.1700"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_12208"} +{"question": "What are the studies about using labelled grouping of images for providing weak supervision to extend the β-VAE?", "answer": ["Weakly Supervised Disentanglement with Guarantees"], "answer_arxiv_id": ["1910.09772"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_12209"} +{"question": "What research works are related to auctions for identical goods under various settings?", "answer": ["Reducing Inefficiency in Carbon Auctions with Imperfect Competition"], "answer_arxiv_id": ["1912.06428"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_12210"} +{"question": "Could you name some paper that proposed learning-based models in motion prediction?", "answer": ["Learning Lane Graph Representations for Motion Forecasting"], "answer_arxiv_id": ["2007.13732"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_12211"} +{"question": "Which papers are relevant to contrastive methods in the context of the InfoNCE criterion?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "An Empirical Study of Training Self-Supervised Vision Transformers", "Decoupled Contrastive Learning", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2003.04297", "2104.02057", "2110.06848", "1807.03748"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_12212"} +{"question": "What works have been done in time-series forecasting in healthcare and business?", "answer": ["Multitask learning and benchmarking with clinical time series data"], "answer_arxiv_id": ["1703.07771"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_12213"} +{"question": "Could you provide studies about weakly supervised object detection models with knowledge transfer that support label indicating presence or absence of object class in an image?", "answer": ["Boosting Weakly Supervised Object Detection with Progressive Knowledge Transfer", "Revisiting knowledge transfer for training object class detectors"], "answer_arxiv_id": ["2007.07986", "1708.06128"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_12214"} +{"question": "What papers developed batch renormalization and representative batch normalization as approaches for normalizing intermediate activations in neural networks?", "answer": ["Batch Renormalization: Towards Reducing Minibatch Dependence in Batch-Normalized Models"], "answer_arxiv_id": ["1702.03275"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_12215"} +{"question": "What work utilizes random forests in optimizing the cost of the decision problem rather than simply fitting the data?", "answer": ["Stochastic Optimization Forests"], "answer_arxiv_id": ["2008.07473v6"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_12216"} +{"question": "Which studies created a dataset to evaluate the ability of question-answering models to track inconsistent worldviews?", "answer": ["Evaluating Theory of Mind in Question Answering"], "answer_arxiv_id": ["1808.09352"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12217"} +{"question": "Could you provide some studies about steerable group convolutions, a class of equivariant convolutions that operate on feature fields over homogeneous spaces?", "answer": ["Steerable CNNs", "Harmonic Networks: Deep Translation and Rotation Equivariance", "3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data", "Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "A General Theory of Equivariant CNNs on Homogeneous Spaces"], "answer_arxiv_id": ["1612.08498", "1612.04642", "1807.02547", "1802.08219", "1811.02017"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_12218"} +{"question": "What work first showed that loss functions that satisfy the symmetric condition can be robust to label noise in multi-class classification?", "answer": ["Robust Loss Functions under Label Noise for Deep Neural Networks"], "answer_arxiv_id": ["1712.09482"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_12219"} +{"question": "What paper introduces neural implicit scene representation in the field of MVPS?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_12220"} +{"question": "Could you provide me some works about the image-supervised weakly-supervised oriented object detection approaches?", "answer": ["Multiple Instance Detection Network with Online Instance Classifier\n Refinement"], "answer_arxiv_id": ["1704.00138"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_12221"} +{"question": "Which works demonstrate the potential of simple architectures in performing well on time series forecasting?", "answer": ["N-BEATS: Neural basis expansion analysis for interpretable time series forecasting", "Are Transformers Effective for Time Series Forecasting?"], "answer_arxiv_id": ["1905.10437", "2205.13504"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_12222"} +{"question": "What works have investigated the benefits of data augmentation theoretically in the setting of clean labels?", "answer": ["A Group-Theoretic Framework for Data Augmentation", "A Kernel Theory of Modern Data Augmentation", "On the Generalization Effects of Linear Transformations in Data Augmentation", "How Data Augmentation affects Optimization for Linear Regression"], "answer_arxiv_id": ["1907.10905", "1803.06084", "2005.00695", "2010.11171"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_12223"} +{"question": "Which works talk about benchmarks to assess the multimodal capabilities of LLMs?", "answer": ["VQA: Visual Question Answering", "GQA: A New Dataset for Real-World Visual Reasoning and Compositional\n Question Answering", "VizWiz Grand Challenge: Answering Visual Questions from Blind People", "Towards VQA Models That Can Read", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science\n Question Answering", "MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning\n Benchmark for Expert AGI"], "answer_arxiv_id": ["1505.00468", "1902.09506", "1802.08218", "1904.08920", "2209.09513", "2311.16502"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_12224"} +{"question": "What works did apply stochastic rounding in binary or quantized neural networks?", "answer": ["Training Quantized Nets: A Deeper Understanding"], "answer_arxiv_id": ["1706.02379"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_12225"} +{"question": "What non-CL methods are known to use methodologies inspired by knowledge distillation for collapse prevention in self-supervised learning?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_12226"} +{"question": "Which studies developed methods to decompose the scene into static and dynamic parts, and model each dynamic actor with dedicated neural fields?", "answer": ["Neural Scene Graphs for Dynamic Scenes", "Panoptic Neural Fields: A Semantic Object-Aware Neural Scene\n Representation", "UniSim: A Neural Closed-Loop Sensor Simulator"], "answer_arxiv_id": ["2011.10379", "2205.04334", "2308.01898"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_12227"} +{"question": "Can you provide works that use deep learning-based methods in recovering voxel grids for modeling explicit 3D structures?", "answer": ["3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction", "Dense 3D Object Reconstruction from a Single Depth View", "Robust Attentional Aggregation of Deep Feature Sets for Multi-view 3D Reconstruction"], "answer_arxiv_id": ["1604.00449", "1802.00411", "1808.00758"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12228"} +{"question": "What research involve the use of diffusion models for generating human motion distributions from Gaussian noise?", "answer": ["Executing your Commands via Motion Diffusion in Latent Space", "Human Motion Diffusion Model"], "answer_arxiv_id": ["2212.04048", "2209.14916"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_12229"} +{"question": "What studies can be classified into the top-down proposal-based methods for instance segmentation of 3D point clouds?", "answer": ["Top-Down Beats Bottom-Up in 3D Instance Segmentation", "3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans", "GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in\n Point Cloud"], "answer_arxiv_id": ["2302.02871", "1812.07003", "1812.03320"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_12230"} +{"question": "What work pioneers a comprehensive framework for multi-agent collaboration using Large Language Models?", "answer": ["Multi-Agent Collaboration: Harnessing the Power of Intelligent LLM\n Agents"], "answer_arxiv_id": ["2306.03314"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_12231"} +{"question": "Could you provide me some studies that discuss about recent development of reward-free and task-agnostic exploration?", "answer": ["Reward-Free Exploration for Reinforcement Learning", "Task-agnostic Exploration in Reinforcement Learning", "Safe Exploration Incurs Nearly No Additional Sample Complexity for Reward-free RL"], "answer_arxiv_id": ["2002.02794", "2006.09497", "2206.14057"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_12232"} +{"question": "Which papers introduced recurrent models like LSTM, GRUs, LMUs, and Independent RNNs?", "answer": ["Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling", "Independently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN"], "answer_arxiv_id": ["1412.3555", "1803.04831"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_12233"} +{"question": "Could you mention some studies introduced a Score Distillation Sample (SDS) loss to perform text-to-3D synthesis?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_12234"} +{"question": "Could you provide me with the research that established benign overfitting in distributional settings?", "answer": ["Benign Overfitting in Linear Classifiers and Leaky ReLU Networks from KKT Conditions for Margin Maximization"], "answer_arxiv_id": ["2303.01462"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_12235"} +{"question": "Which work initially proposed Transformer for sequence-to-sequence modelling in Natural Language Processing?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20220117"}, "qid": "AutoScholarQuery_train_12236"} +{"question": "Can you provide me research papers that use STAPLE in many medical tasks?", "answer": ["CryoNuSeg: A Dataset for Nuclei Instance Segmentation of Cryosectioned\n H&E-Stained Histological Images"], "answer_arxiv_id": ["2101.00442"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_12237"} +{"question": "What target tasks benefit from attention mechanism according to the mentioned studies?", "answer": ["Dual-Level Collaborative Transformer for Image Captioning", "DFormer: Rethinking RGBD Representation Learning for Semantic\n Segmentation"], "answer_arxiv_id": ["2101.06462", "2309.09668"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_12238"} +{"question": "Which studies developed large multimodal models by incorporating vision encoders?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "LLaMA: Open and Efficient Foundation Language Models", "Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena"], "answer_arxiv_id": ["2010.11929", "2302.13971", "2306.05685"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_train_12239"} +{"question": "What works have advanced the convergence rates of high-probability generalization upper bounds for uniformly stable algorithms?", "answer": ["Generalization Bounds for Uniformly Stable Algorithms", "High probability generalization bounds for uniformly stable algorithms with nearly optimal rate", "Sharper bounds for uniformly stable algorithms"], "answer_arxiv_id": ["1812.09859", "1902.10710", "1910.07833"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_12240"} +{"question": "Which papers discuss the recent success of self-supervised methods, especially in the domain of computer vision?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Whitening for Self-Supervised Representation Learning", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Decoupled Contrastive Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction"], "answer_arxiv_id": ["2002.05709", "2007.06346", "2105.04906", "2110.06848", "2103.03230"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_12241"} +{"question": "What papers discussed the use of control variates for variance reduction in Monte-Carlo simulation of complex systems?", "answer": ["Scalable Discrete Sampling as a Multi-Armed Bandit Problem"], "answer_arxiv_id": ["1506.09039"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_12242"} +{"question": "Can you provide examples of theoretical semi-supervised learning frameworks that demonstrate how constraints on the hypothesis space could reduce the generalization error?", "answer": ["Generalization Bounds for Learning with Linear, Polygonal, Quadratic and Conic Side Knowledge"], "answer_arxiv_id": ["1405.7764"], "source_meta": {"published_time": "20230708"}, "qid": "AutoScholarQuery_train_12243"} +{"question": "Which work investigates using a neural network as a challenging benchmark for LSTM-based seq2seq models?", "answer": ["Learning to Execute"], "answer_arxiv_id": ["1410.4615"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_12244"} +{"question": "Which literature describes continuous sparsification and are related to this study?", "answer": ["Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation"], "answer_arxiv_id": ["1308.3432"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_12245"} +{"question": "What work proposed a regularization method based on contrast learning for image dehazing?", "answer": ["Contrastive Learning for Compact Single Image Dehazing"], "answer_arxiv_id": ["2104.09367"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_12246"} +{"question": "What previous studies showed LLM-based agents' capabilities in task like reasoning and interaction?", "answer": ["A Survey on Large Language Model based Autonomous Agents"], "answer_arxiv_id": ["2308.11432"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_12247"} +{"question": "Which work improves upon TTSA by leveraging an additive correction term in the LL update step while still handling UL constraints?", "answer": ["A Single-Timescale Method for Stochastic Bilevel Optimization"], "answer_arxiv_id": ["2102.04671v4"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_12248"} +{"question": "Could you provide any research where more semantically-similar demonstrations were constructed during evaluation?", "answer": ["What Makes Good In-Context Examples for GPT-3?"], "answer_arxiv_id": ["2101.06804"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_12249"} +{"question": "What papers tackle adversarial attacks to confuse classification models?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Generating Adversarial Examples with Adversarial Networks", "Boosting Adversarial Attacks with Momentum"], "answer_arxiv_id": ["1412.6572", "1706.06083", "1801.02610", "1710.06081"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_12250"} +{"question": "Which paper studies if the principle that humans prefer to spread information evenly during language production can help capture the unique signature of each LLM and human author?", "answer": ["GPT-who: An Information Density-based Machine-Generated Text Detector"], "answer_arxiv_id": ["2310.06202"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_12251"} +{"question": "Could you provide me some works investigating AI alignment?", "answer": ["Artificial Intelligence, Values and Alignment"], "answer_arxiv_id": ["2001.09768"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_12252"} +{"question": "Which adversarial-based methods for Unsupervised Domain Adaptation (UDA) are able to reduce the domain discrepancy via domain discriminator networks?", "answer": ["Domain-Adversarial Training of Neural Networks", "Conditional Adversarial Domain Adaptation", "Adversarial Discriminative Domain Adaptation", "Multi-Adversarial Domain Adaptation", "A DIRT-T Approach to Unsupervised Domain Adaptation", "Adversarial Domain Adaptation with Domain Mixup"], "answer_arxiv_id": ["1505.07818", "1705.10667", "1702.05464", "1809.02176", "1802.08735", "1912.01805"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_12253"} +{"question": "Which works generated lane segments from overhead highway cameras and LiDAR imagery using a recurrent neural network?", "answer": ["Hierarchical Recurrent Attention Networks for Structured Online Maps", "Convolutional Recurrent Network for Road Boundary Extraction"], "answer_arxiv_id": ["2012.12314", "2012.12160"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_12254"} +{"question": "Which works have already investigated dual-gripper or dual-arm manipulation?", "answer": ["Towards Human-Level Bimanual Dexterous Manipulation with Reinforcement Learning", "FabricFlowNet: Bimanual Cloth Manipulation with a Flow-based Policy", "Efficient Bimanual Manipulation Using Learned Task Schemas", "Deep Imitation Learning for Bimanual Robotic Manipulation", "V-MAO: Generative Modeling for Multi-Arm Manipulation of Articulated Objects", "Robot Cooking with Stir-fry: Bimanual Non-prehensile Manipulation of Semi-fluid Objects"], "answer_arxiv_id": ["2206.08686", "2111.05623v3", "1909.13874", "2010.05134", "2111.03987", "2205.05960"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_12255"} +{"question": "Which publication mentions an approach to localize actions using voxel maps?", "answer": ["Egocentric Activity Recognition and Localization on a 3D Map"], "answer_arxiv_id": ["2105.09544"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_12256"} +{"question": "Can you mention some works that execute programs on images but assume API access to a set of available modules with no further training?", "answer": ["Visual Programming: Compositional visual reasoning without training", "ViperGPT: Visual Inference via Python Execution for Reasoning"], "answer_arxiv_id": ["2211.11559", "2303.08128"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_12257"} +{"question": "Does any work define correlation shift when there is a mismatch from the training to the OOD data in the context of distribution?", "answer": ["Marginal Singularity, and the Benefits of Labels in Covariate-Shift"], "answer_arxiv_id": ["1803.01833"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_12258"} +{"question": "Which studies introduced parametric methods in recalibration algorithms such as Platt scaling and temperature scaling?", "answer": ["On Calibration of Modern Neural Networks"], "answer_arxiv_id": ["1706.04599"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_12259"} +{"question": "What papers established GNNs as universal approximators for classes of invariant and equivariant functions?", "answer": ["On the Universality of Invariant Networks", "Universal Invariant and Equivariant Graph Neural Networks"], "answer_arxiv_id": ["1901.09342", "1905.04943"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_12260"} +{"question": "What works have investigated the fundamental limits on the error probability in the minimax setting?", "answer": ["Tight (Lower) Bounds for the Fixed Budget Best Arm Identification Bandit Problem", "On the Existence of a Complexity in Fixed Budget Bandit Identification", "On Uniformly Optimal Algorithms for Best Arm Identification in Two-Armed Bandits with Fixed Budget"], "answer_arxiv_id": ["1605.09004v1", "2303.09468v2", "2308.12000"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_12261"} +{"question": "What papers have investigated the optimization bias manifested as a spectral bias in SGD-based algorithms?", "answer": ["The Convergence Rate of Neural Networks for Learned Functions of\n Different Frequencies", "On the Spectral Bias of Neural Networks"], "answer_arxiv_id": ["1906.00425", "1806.08734"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_12262"} +{"question": "Which papers form the subclass of imitation learning that emphasizes learning behavior from expert datasets without reward labels?", "answer": ["An Algorithmic Perspective on Imitation Learning", "DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills", "AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control", "Generative Adversarial Imitation Learning"], "answer_arxiv_id": ["1811.06711", "1804.02717", "2104.02180", "1606.03476"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12263"} +{"question": "What are some works that have brought feature attribution methods into natural-language processing?", "answer": ["Understanding Neural Networks through Representation Erasure", "Explaining Recurrent Neural Network Predictions in Sentiment Analysis", "Representation of linguistic form and function in recurrent neural networks", "AllenNLP Interpret: A Framework for Explaining Predictions of NLP Models", "How do Decisions Emerge across Layers in Neural Models? Interpretation with Differentiable Masking", "Transformer Interpretability Beyond Attention Visualization", "The Out-of-Distribution Problem in Explainability and Search Methods for Feature Importance Explanations", "Thermostat: A Large Collection of NLP Model Explanations and Analysis Tools"], "answer_arxiv_id": ["1612.08220", "1706.07206", "1602.08952", "1909.09251", "2004.14992", "2012.09838", "2106.00786", "2108.13961v1"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_12264"} +{"question": "Which are the references about sub-linear regret algorithms in the non-tabular POMDP setting?", "answer": ["Regret Minimization in Partially Observable Linear Quadratic Control", "Improper Learning for Non-Stochastic Control"], "answer_arxiv_id": ["2002.00082", "2001.09254"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_12265"} +{"question": "Can you provide me some works that are focused on subjective evaluations of large language models?", "answer": ["AlignBench: Benchmarking Chinese Alignment of Large Language Models", "AlpacaFarm: A Simulation Framework for Methods that Learn from Human\n Feedback"], "answer_arxiv_id": ["2311.18743", "2305.14387"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_12266"} +{"question": "What works applied the binary-tree mechanism to answer range queries?", "answer": ["The Power of Factorization Mechanisms in Local and Central Differential Privacy"], "answer_arxiv_id": ["1911.08339"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_12267"} +{"question": "Which studies focus on studying bias and variance of treatment effects under two randomization schemes for auction experiments?", "answer": ["Randomization and The Pernicious Effects of Limited Budgets on Auction Experiments"], "answer_arxiv_id": ["1605.09171"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_12268"} +{"question": "Which research presents a unified framework for removal-based model explanation methods in machine learning?", "answer": ["Explaining by Removing: A Unified Framework for Model Explanation"], "answer_arxiv_id": ["2011.14878"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_12269"} +{"question": "Could you provide me some studies that apply automatic augmentations in vision through neural architecture search?", "answer": ["Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules", "RandAugment: Practical automated data augmentation with a reduced search space"], "answer_arxiv_id": ["1905.05393", "1909.13719"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_12270"} +{"question": "What paper distills a pretrained diffusion model into a single-step DEQ?", "answer": ["One-Step Diffusion Distillation via Deep Equilibrium Models"], "answer_arxiv_id": ["2401.08639"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_12271"} +{"question": "Are there any studies that focused on seeking references post-generation for attribution?", "answer": ["RARR: Researching and Revising What Language Models Say, Using Language\n Models", "Complex Claim Verification with Evidence Retrieved in the Wild", "Retrieving Supporting Evidence for Generative Question Answering"], "answer_arxiv_id": ["2210.08726", "2305.11859", "2309.11392"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_12272"} +{"question": "Which paper proposed a data-dependent variance-based regularization to address bias-variance tradeoff in machine learning?", "answer": ["Empirical Bernstein Bounds and Sample Variance Penalization"], "answer_arxiv_id": ["0907.3740"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_12273"} +{"question": "Could you provide me some studies about trust region learning-based algorithms?", "answer": ["Continuous control with deep reinforcement learning", "Benchmarking Deep Reinforcement Learning for Continuous Control"], "answer_arxiv_id": ["1509.02971", "1604.06778"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_12274"} +{"question": "Which papers adopted quantization as a technique to enhance communication efficiency?", "answer": ["New Bounds For Distributed Mean Estimation and Variance Reduction", "EDEN: Communication-Efficient and Robust Distributed Mean Estimation for Federated Learning", "Distributed Mean Estimation with Limited Communication", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "signSGD: Compressed Optimisation for Non-Convex Problems", "FedPAQ: A Communication-Efficient Federated Learning Method with Periodic Averaging and Quantization"], "answer_arxiv_id": ["2002.09268", "2108.08842", "1611.00429", "1610.02132", "1802.04434", "1909.13014v4"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_12275"} +{"question": "Which works have enhanced the accuracy of faithfulness evaluation in text summarization?", "answer": ["Asking and Answering Questions to Evaluate the Factual Consistency of\n Summaries", "QAFactEval: Improved QA-Based Factual Consistency Evaluation for\n Summarization"], "answer_arxiv_id": ["2004.04228", "2112.08542"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_train_12276"} +{"question": "What studies have developed strategies to avoid catastrophic forgetting?", "answer": ["The Ideal Continual Learner: An Agent That Never Forgets"], "answer_arxiv_id": ["2305.00316"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_12277"} +{"question": "What studies are relevant to doubly robust estimation for missing data and causal inference?", "answer": ["Double/Debiased Machine Learning for Treatment and Structural Parameters"], "answer_arxiv_id": ["1608.00060"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_12278"} +{"question": "Could you provide me with studies that demonstrate run of several shuffle mechanisms in parallel using a single shuffler?", "answer": ["Shuffle Private Stochastic Convex Optimization", "Differential Privacy in the Shuffle Model: A Survey of Separations"], "answer_arxiv_id": ["2106.09805", "2107.11839"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_12279"} +{"question": "Many studies have conducted research on various approaches for weight quantization. Can you list these studies?", "answer": ["XNOR-Net: ImageNet Classification Using Binary Convolutional Neural\n Networks", "On Quantizing Implicit Neural Representations", "Quantization and Training of Neural Networks for Efficient\n Integer-Arithmetic-Only Inference"], "answer_arxiv_id": ["1603.05279", "2209.01019", "1712.05877"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_12280"} +{"question": "Which research work reported the 'Deep RLSP' algorithm that focuses on sampling past trajectories?", "answer": ["Learning What To Do by Simulating the Past"], "answer_arxiv_id": ["2104.03946"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_12281"} +{"question": "Which paper introduced the concept of 'Neural Manifold ODE'?", "answer": ["Neural Manifold Ordinary Differential Equations"], "answer_arxiv_id": ["2006.10254"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_12282"} +{"question": "What works contribute to the fine-tuning a frozen LLaMA to respond to multi-modality instructions?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Microsoft COCO Captions: Data Collection and Evaluation Server", "Visual Instruction Tuning"], "answer_arxiv_id": ["2302.13971", "1504.00325", "2304.08485"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_12283"} +{"question": "What papers discuss the problems found in using influence functions, particularly in deep learning?", "answer": ["Influence Functions in Deep Learning Are Fragile", "FastIF: Scalable Influence Functions for Efficient Model Interpretation\n and Debugging"], "answer_arxiv_id": ["2006.14651", "2012.15781"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_12284"} +{"question": "What are some large language models that have been trained on massive code data for code generation tasks?", "answer": ["CodeGen: An Open Large Language Model for Code with Multi-Turn Program\n Synthesis", "[2203.07814] Competition-Level Code Generation with AlphaCode", "CodeGeeX: A Pre-Trained Model for Code Generation with Multilingual\n Evaluations on HumanEval-X", "Code Llama: Open Foundation Models for Code"], "answer_arxiv_id": ["2203.13474", "2203.07814", "2303.17568", "2308.12950"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_12285"} +{"question": "Are there any studies on causal estimation with (near) overlap violations?", "answer": ["Overlap in Observational Studies with High-Dimensional Covariates", "Causal Effect Inference with Deep Latent-Variable Models", "Estimating Average Treatment Effects: Supplementary Analyses and Remaining Challenges"], "answer_arxiv_id": ["1711.02582v4", "1705.08821", "1702.01250"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12286"} +{"question": "Could you cite some works that applied pre-trained vision models in diverse control domains?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "Learning Transferable Visual Models From Natural Language Supervision", "LM-Nav: Robotic Navigation with Large Pre-Trained Models of Language, Vision, and Action"], "answer_arxiv_id": ["2107.03380", "2203.03580", "2103.00020", "2207.04429"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_12287"} +{"question": "Are there any works about the design of a pipeline for extracting knowledge from external sources?", "answer": ["K-Lite: Learning Transferable Visual Models with External Knowledge"], "answer_arxiv_id": ["2204.09222"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12288"} +{"question": "What research integrates more sophisticated multi-sample data augmentations techniques such as mixup?", "answer": ["mixup: Beyond Empirical Risk Minimization"], "answer_arxiv_id": ["1710.09412"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_12289"} +{"question": "Could you specify the research that use large amounts of human supervision to teach LMs to use tools?", "answer": ["Internet-Augmented Dialogue Generation", "WebGPT: Browser-assisted question-answering with human feedback", "LaMDA: Language Models for Dialog Applications"], "answer_arxiv_id": ["2107.07566", "2112.09332", "2201.08239"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_12290"} +{"question": "Can you provide me some papers about using Reinforcement Learning in computer vision tasks?", "answer": ["Reinforcement learning based recommender systems: A survey"], "answer_arxiv_id": ["2101.06286"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_12291"} +{"question": "Which studies developed a fast numerical solver to overcome the slow and expensive sampling speed of DDMs?", "answer": ["Gotta Go Fast When Generating Data with Score-Based Models", "Fast Sampling of Diffusion Models with Exponential Integrator", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2105.14080", "2204.13902", "2202.09778"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_12292"} +{"question": "Which work analyzes the features learned by saliency models by aligning activation maps with segmentation for a selection of objects?", "answer": ["Understanding and Visualizing Deep Visual Saliency Models"], "answer_arxiv_id": ["1903.02501"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_12293"} +{"question": "Could you provide me with papers that primarily utilize the pre-trained Q-former in visual instruction training?", "answer": ["InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "VideoChat: Chat-Centric Video Understanding"], "answer_arxiv_id": ["2305.06500", "2304.10592", "2305.06355"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_12294"} +{"question": "Who provided the minimax sample complexity for offline zero-sum Markov games?", "answer": ["Model-Based Reinforcement Learning Is Minimax-Optimal for Offline Zero-Sum Markov Games", "Provably Efficient Offline Multi-agent Reinforcement Learning via Strategy-wise Bonus"], "answer_arxiv_id": ["2206.04044", "2206.00159"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_12295"} +{"question": "Which works introduce techniques to learn different types of constraint functions in continuous domains?", "answer": ["Learning Parametric Constraints in High Dimensions from Demonstrations", "Inverse Constrained Reinforcement Learning"], "answer_arxiv_id": ["1910.03477", "2011.09999"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_12296"} +{"question": "Which paper has shown the capability of Flamingo in supporting in-context learning during the pretraining process?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_12297"} +{"question": "What studies have reported the use and improvement of Spatial transformer network (STN) and its modifications (like Deformable ConvNets - DCN, DCNv2, DCNv3)?", "answer": ["Spatial Transformer Networks", "Deformable Convolutional Networks", "Deformable ConvNets v2: More Deformable, Better Results", "InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions"], "answer_arxiv_id": ["1506.02025", "1703.06211", "1811.11168", "2211.05778"], "source_meta": {"published_time": "20230829"}, "qid": "AutoScholarQuery_train_12298"} +{"question": "Which works discuss finetuning as an influential approach for knowledge transfer in transfer learning?", "answer": ["SpotTune: Transfer Learning through Adaptive Fine-tuning", "Rich feature hierarchies for accurate object detection and semantic segmentation", "Learning Transferable Features with Deep Adaptation Networks"], "answer_arxiv_id": ["1811.08737", "1311.2524", "1502.02791"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_12299"} +{"question": "Which works are about sketch+text joint multi-modal learning?", "answer": ["FS-COCO: Towards Understanding of Freehand Sketches of Common Objects in\n Context", "SceneTrilogy: On Human Scene-Sketch and its Complementarity with Photo\n and Text", "CoGS: Controllable Generation and Search from Sketch and Style", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models", "Sketch-Guided Text-to-Image Diffusion Models", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "SKED: Sketch-guided Text-based 3D Editing", "A Sketch Is Worth a Thousand Words: Image Retrieval with Text and Sketch"], "answer_arxiv_id": ["2203.02113", "2204.11964", "2203.09554", "2302.08453", "2302.05543", "2211.13752", "2003.08934", "2201.05989", "2303.10735", "2208.03354"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_12300"} +{"question": "What prior studies worked on sketch-based image retrieval?", "answer": ["Deep Sketch Hashing: Fast Free-hand Sketch-Based Image Retrieval", "Deep Plastic Surgery: Robust and Controllable Image Editing with\n Human-Drawn Sketches", "LiveSketch: Query Perturbations for Guided Sketch-based Visual Search", "Sketch Less for More: On-the-Fly Fine-Grained Sketch Based Image\n Retrieval", "Adaptive Fine-Grained Sketch-Based Image Retrieval", "CLIP for All Things Zero-Shot Sketch-Based Image Retrieval, Fine-Grained\n or Not", "Text-to-Image Diffusion Models are Great Sketch-Photo Matchmakers", "Cross-Modal Hierarchical Modelling for Fine-Grained Sketch Based Image Retrieval", "StyleMeUp: Towards Style-Agnostic Sketch-Based Image Retrieval", "Partially Does It: Towards Scene-Level FG-SBIR with Partial Input", "SceneTrilogy: On Human Scene-Sketch and its Complementarity with Photo\n and Text", "A Sketch Is Worth a Thousand Words: Image Retrieval with Text and Sketch"], "answer_arxiv_id": ["1703.05605", "2001.02890", "1904.06611", "2002.10310", "2207.01723", "2303.13440", "2403.07214", "2007.15103v2", "2103.15706", "2203.14804", "2204.11964", "2208.03354"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_12301"} +{"question": "Which works discuss cross-image augmentations as a data augmentation technique?", "answer": ["mixup: Beyond Empirical Risk Minimization", "CutMix: Regularization Strategy to Train Strong Classifiers with\n Localizable Features"], "answer_arxiv_id": ["1710.09412", "1905.04899"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_12302"} +{"question": "Which works have proposed in-context few-shot learning by mimicking in-context demonstrations?", "answer": ["On the Effect of Pretraining Corpora on In-context Learning by a Large-scale Language Model"], "answer_arxiv_id": ["2204.13509"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_12303"} +{"question": "Can you list the papers that focused on obtaining camera parameters without using SFM, instead by training both camera parameters and NeRF using only pictures?", "answer": ["GNeRF: GAN-based Neural Radiance Field without Posed Camera", "BARF: Bundle-Adjusting Neural Radiance Fields", "GARF: Gaussian Activated Radiance Fields for High Fidelity\n Reconstruction and Pose Estimation", "FlowCam: Training Generalizable 3D Radiance Fields without Camera Poses\n via Pixel-Aligned Scene Flow", "LU-NeRF: Scene and Pose Estimation by Synchronizing Local Unposed NeRFs"], "answer_arxiv_id": ["2103.15606", "2104.06405", "2204.05735", "2306.00180", "2306.05410"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_12304"} +{"question": "What works first proposed the object detector based on an encoder-decoder architecture?", "answer": ["End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2005.12872"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_12305"} +{"question": "Could you provide me some works which bypass the requirement for predefined human models in human reconstruction?", "answer": ["Video Based Reconstruction of 3D People Models", "PERGAMO: Personalized 3D Garments from Monocular Video", "Human Performance Capture from Monocular Video in the Wild", "3D Clothed Human Reconstruction in the Wild"], "answer_arxiv_id": ["1803.04758", "2210.15040", "2111.14672", "2207.10053"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_12306"} +{"question": "What studies identified uninformative tokens by making intermediate predictions with auxiliary heads?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "SPViT: Enabling Faster Vision Transformers via Soft Token Pruning", "AdaViT: Adaptive Vision Transformers for Efficient Image Recognition"], "answer_arxiv_id": ["2106.02034v2", "2112.13890", "2111.15668"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_12307"} +{"question": "Which paper proposed to maximize the information gain about the agent’s belief of environment dynamics in the context of intrinsic awards design based on information theory?", "answer": ["VIME: Variational Information Maximizing Exploration"], "answer_arxiv_id": ["1605.09674"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12308"} +{"question": "Which studies show that the Online Mirror Descent framework with the (negative) Tsallis entropy regularizer achieves O​(k​T) regret bounds in the adversarial regime?", "answer": ["Fighting Bandits with a New Kind of Smoothness"], "answer_arxiv_id": ["1512.04152"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_12309"} +{"question": "Which studies proposed improvements to steps such as learning a matching cost function or learning to improve subsequent optimization and refinement in deep stereo matching?", "answer": ["Stereo Matching by Training a Convolutional Neural Network to Compare\n Image Patches", "Improved Stereo Matching with Constant Highway Networks and Reflective\n Confidence Learning", "Detect, Replace, Refine: Deep Structured Prediction For Pixel Wise\n Labeling"], "answer_arxiv_id": ["1510.05970", "1701.00165", "1612.04770"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_12310"} +{"question": "What works performed localized editing through the manipulation of attention maps?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2208.01626", "2211.12572"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_12311"} +{"question": "Could you provide me some studies about the acceleration of Neural Radiance Fields?", "answer": ["PlenOctrees for Real-time Rendering of Neural Radiance Fields", "DeRF: Decomposed Radiance Fields"], "answer_arxiv_id": ["2103.14024", "2011.12490"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_12312"} +{"question": "Which papers are about the 'learning to learn' concept in the area of AutoML and meta-learning?", "answer": ["Learning to learn by gradient descent by gradient descent", "Learned Optimizers that Scale and Generalize", "Neural Optimizer Search with Reinforcement Learning", "Understanding and correcting pathologies in the training of learned optimizers"], "answer_arxiv_id": ["1606.04474", "1703.04813", "1709.07417", "1810.10180"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_12313"} +{"question": "Which papers have proposed different methods for solving the percentile criterion in ambiguity sets?", "answer": ["Beyond Confidence Regions: Tight Bayesian Ambiguity Sets for Robust MDPs"], "answer_arxiv_id": ["1902.07605"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_12314"} +{"question": "Could you provide me some studies about confident-conditioned Q-values?", "answer": ["An Optimistic Perspective on Offline Reinforcement Learning", "SUNRISE: A Simple Unified Framework for Ensemble Learning in Deep Reinforcement Learning", "Randomized Ensembled Double Q-Learning: Learning Fast Without a Model", "Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble"], "answer_arxiv_id": ["1907.04543v4", "2007.04938", "2101.05982", "2110.01548"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_12315"} +{"question": "What work proposed the use of EBM to learn both local and global factors of variations from image data?", "answer": ["Unsupervised Learning of Compositional Energy Concepts"], "answer_arxiv_id": ["2111.03042"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_12316"} +{"question": "Are there any research papers about the application of unlearning in unsupervised learning or generative models?", "answer": ["Data Redaction from Pre-trained GANs", "Adapt then Unlearn: Exploiting Parameter Space Semantics for Unlearning\n in Generative Adversarial Networks"], "answer_arxiv_id": ["2206.14389", "2309.14054"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_12317"} +{"question": "What studies have developed hand-crafted 3D local features for point cloud matching, registration, recognition and retrieval?", "answer": ["Rotational Projection Statistics for 3D Local Surface Description and\n Object Recognition"], "answer_arxiv_id": ["1304.3192"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_12318"} +{"question": "Which papers proposed memory-based methods for preventing catastrophic forgetting in Continual Learning?", "answer": ["Continual Learning with Deep Generative Replay", "Towards Robust Evaluations of Continual Learning", "Generative replay with feedback connections as a general strategy for\n continual learning", "Experience Replay for Continual Learning", "Generative Continual Concept Learning", "Orthogonal Gradient Descent for Continual Learning", "Gradient Projection Memory for Continual Learning", "Training Networks in Null Space of Feature Covariance for Continual\n Learning", "TRGP: Trust Region Gradient Projection for Continual Learning", "Sparsity and Heterogeneous Dropout for Continual Learning in the Null\n Space of Neural Activations"], "answer_arxiv_id": ["1705.08690", "1805.09733", "1809.10635", "1811.11682", "1906.03744", "1910.07104", "2103.09762", "2103.07113", "2202.02931", "2203.06514"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_12319"} +{"question": "Could you provide me some research papers which developed provable sample complexity in low-rank MDPs setting?", "answer": ["FLAMBE: Structural Complexity and Representation Learning of Low Rank MDPs", "A Free Lunch from the Noise: Provable and Practical Exploration for Representation Learning", "Representation Learning for Online and Offline RL in Low-rank MDPs", "Model-free Representation Learning and Exploration in Low-rank MDPs", "Efficient Reinforcement Learning in Block MDPs: A Model-free Representation Learning Approach"], "answer_arxiv_id": ["2006.10814", "2111.11485", "2110.04652", "2102.07035", "2202.00063"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_12320"} +{"question": "What works propose Code Large Language Models?", "answer": ["CodeGen: An Open Large Language Model for Code with Multi-Turn Program\n Synthesis", "CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for\n Code Understanding and Generation", "Code Llama: Open Foundation Models for Code"], "answer_arxiv_id": ["2203.13474", "2109.00859", "2308.12950"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_12321"} +{"question": "Could you list studies that used machine learning, including deep learning for tackling similarity in binary files?", "answer": ["Neural Machine Translation Inspired Binary Code Similarity Comparison beyond Function Pairs", "Neural Network-based Graph Embedding for Cross-Platform Binary Code Similarity Detection", "jTrans: Jump-Aware Transformer for Binary Code Similarity Detection"], "answer_arxiv_id": ["1808.04706", "1708.06525v4", "2205.12713"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_12322"} +{"question": "Which papers delve into augmenting LLMs with tools like program executors, translation & QA models?", "answer": ["Program of Thoughts Prompting: Disentangling Computation from Reasoning\n for Numerical Reasoning Tasks", "PAL: Program-aided Language Models", "TALM: Tool Augmented Language Models", "Toolformer: Language Models Can Teach Themselves to Use Tools"], "answer_arxiv_id": ["2211.12588", "2211.10435", "2205.12255", "2302.04761"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_12323"} +{"question": "Are there any research papers that use information-theoretic bounds to achieve the known optimal minimax rates and provide exact characterizations of the generalization error?", "answer": ["Towards a Unified Information-Theoretic Framework for Generalization", "Understanding Generalization via Leave-One-Out Conditional Mutual Information", "Tighter Information-Theoretic Generalization Bounds from Supersamples"], "answer_arxiv_id": ["2111.05275v2", "2206.14800v1", "2302.02432"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_12324"} +{"question": "Could you name some research that applied diffusion models to detection and segmentation tasks?", "answer": ["Label-Efficient Semantic Segmentation with Diffusion Models", "DiffusionDet: Diffusion Model for Object Detection", "DiffusionInst: Diffusion Model for Instance Segmentation"], "answer_arxiv_id": ["2112.03126", "2211.09788", "2212.02773"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_12325"} +{"question": "Which research characterized GR as the slope of the loss surface and depicted its preference for flat regions?", "answer": ["Implicit Gradient Regularization"], "answer_arxiv_id": ["2009.11162"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_12326"} +{"question": "Could you provide me some studies about probing in understanding deep learning models?", "answer": ["What you can cram into a single $&!#⁢* vector: Probing sentence embeddings for linguistic properties", "BERT Rediscovers the Classical NLP Pipeline", "Understanding Learning Dynamics Of Language Models with SVCCA", "A Primer in BERTology: What We Know About How BERT Works"], "answer_arxiv_id": ["1805.01070", "1905.05950", "1811.00225", "2002.12327"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_12327"} +{"question": "Which studies developed a unified formula for scaling laws?", "answer": ["Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2203.15556"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_12328"} +{"question": "What studies use structural causal models to express the transition dynamics in POMDPs?", "answer": ["Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search"], "answer_arxiv_id": ["1811.06272"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_12329"} +{"question": "Could you provide me some works on generating temporally coherent videos?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data", "Video Diffusion Models"], "answer_arxiv_id": ["2209.14792", "2204.03458"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_12330"} +{"question": "What works have been published on the topic of meta-learning that involve learning a good neural network initialization?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "ProMP: Proximal Meta-Policy Search", "On First-Order Meta-Learning Algorithms", "Bayesian Model-Agnostic Meta-Learning"], "answer_arxiv_id": ["1703.03400", "1810.06784", "1803.02999", "1806.03836"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_12331"} +{"question": "What research showcases that nonparametric embeddings show high performance for rare and unseen cases not existing in model parametric space?", "answer": ["Nonparametric Decoding for Generative Retrieval", "Nonparametric Masked Language Modeling"], "answer_arxiv_id": ["2210.02068", "2212.01349"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_12332"} +{"question": "Which works have made advancements in transformer-based models for unordered points in 3D point cloud understanding?", "answer": ["PCT: Point cloud transformer", "Point Transformer"], "answer_arxiv_id": ["2012.09688", "2012.09164"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_12333"} +{"question": "Which papers introduced and extended Score Matching?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_12334"} +{"question": "What papers present approaches that focus on 3D scene reconstruction using standard color videos?", "answer": ["Atlas: End-to-End 3D Scene Reconstruction from Posed Images", "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video", "Neural 3D Scene Reconstruction with the Manhattan-world Assumption"], "answer_arxiv_id": ["2003.10432", "2104.00681", "2205.02836"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12335"} +{"question": "Could you provide me some works that utilized low-rank approximations of the attention matrix in Transformer models?", "answer": ["Linformer: Self-Attention with Linear Complexity", "Luna: Linear Unified Nested Attention"], "answer_arxiv_id": ["2006.04768", "2106.01540"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_12336"} +{"question": "Which papers typically used Negative Log Likelihood (NLL) as a training loss in calibration methods?", "answer": ["On Calibration of Modern Neural Networks", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration", "Intra Order-Preserving Functions for Calibration of Multi-Class Neural Networks", "Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning", "Parameterized Temperature Scaling for Boosting the Expressive Power in Post-Hoc Uncertainty Calibration"], "answer_arxiv_id": ["1706.04599", "1910.12656", "2003.06820", "2003.07329", "2102.12182"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_12337"} +{"question": "Which studies investigated parameter-isolation based methods in continual learning?", "answer": ["Overcoming Catastrophic Forgetting with Hard Attention to the Task", "Scalable and Order-robust Continual Learning with Additive Parameter Decomposition", "GROWN: GRow Only When Necessary for Continual Learning"], "answer_arxiv_id": ["1801.01423", "1902.09432", "2110.00908"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_12338"} +{"question": "What studies have observed the low rank phenomenon in large depth L2-regularized DNNs?", "answer": ["Implicit Regularization Towards Rank Minimization in ReLU Networks"], "answer_arxiv_id": ["2201.12760"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12339"} +{"question": "Could you provide me with papers about automatically constructing knowledge bases from unstructured corpora?", "answer": ["Advanced Semantics for Commonsense Knowledge Extraction", "Refined Commonsense Knowledge from Large-Scale Web Contents"], "answer_arxiv_id": ["2011.00905", "2112.04596"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_12340"} +{"question": "What is the proposed method in the paper by Ho et al. for improving longer video generations?", "answer": ["Video Diffusion Models"], "answer_arxiv_id": ["2204.03458"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_12341"} +{"question": "What early studies utilized motion capture technology for limited dataset collection in dance generation?", "answer": ["Music2Dance: DanceNet for Music-driven Dance Generation"], "answer_arxiv_id": ["2002.03761"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_12342"} +{"question": "Could you provide me some works using a greedy rollout of a periodically updated best-so-far policy as a baseline for routing problems?", "answer": ["Attention, Learn to Solve Routing Problems!"], "answer_arxiv_id": ["1803.08475"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_12343"} +{"question": "Which works present a training-free inference algorithm by estimating the optimal reverse variances under shortened inference steps?", "answer": ["Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in\n Diffusion Probabilistic Models"], "answer_arxiv_id": ["2201.06503"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_12344"} +{"question": "Which works propose successful self-supervised pre-training methodologies?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "Learning Transferable Visual Models From Natural Language Supervision", "DINOv2: Learning Robust Visual Features without Supervision", "Masked Autoencoders Are Scalable Vision Learners", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2104.14294", "2103.00020", "2304.07193", "2111.06377", "1911.05722"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_12345"} +{"question": "Which papers describe the fusion-encoder models where visual and linguistic features are concatenated before feeding them into a self-attentive Transformer?", "answer": ["VisualBERT: A Simple and Performant Baseline for Vision and Language", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "UNITER: UNiversal Image-TExt Representation Learning", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "VinVL: Revisiting Visual Representations in Vision-Language Models", "Large-Scale Adversarial Training for Vision-and-Language Representation Learning", "UNIMO: Towards Unified-Modal Understanding and Generation via Cross-Modal Contrastive Learning", "Unifying Vision-and-Language Tasks via Text Generation", "Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal Transformers", "Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language Representation Learning", "How Much Can CLIP Benefit Vision-and-Language Tasks?", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "MDETR - Modulated Detection for End-to-End Multi-Modal Understanding", "UniTAB: Unifying Text and Box Outputs for Grounded Vision-Language Modeling", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework", "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision", "GIT: A Generative Image-to-text Transformer for Vision and Language", "VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks"], "answer_arxiv_id": ["1908.03557", "1908.08530", "1909.11740", "2004.06165", "2101.00529", "2006.06195", "2012.15409", "2102.02779", "2004.00849", "2104.03135", "2107.06383", "2108.10904", "2104.12763", "2111.12085", "2202.03052", "2102.03334", "2205.14100", "2111.02358", "2208.10442"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_12346"} +{"question": "Could you provide me with studies mentioning single-policy concentrability?", "answer": ["Pessimistic Model-based Offline Reinforcement Learning under Partial Coverage"], "answer_arxiv_id": ["2107.06226"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_12347"} +{"question": "Which study suggests caching the previous context to linearly extend the context with the number of layers in transformers?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context"], "answer_arxiv_id": ["1901.02860"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_12348"} +{"question": "Can you cite some research which considered Dynamic Time Warping (DTW) method for achieving alignment objectives?", "answer": ["D3TW: Discriminative Differentiable Dynamic Time Warping for Weakly\n Supervised Action Alignment and Segmentation", "Learning Discriminative Prototypes with Dynamic Time Warping"], "answer_arxiv_id": ["1901.02598", "2103.09458"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_12349"} +{"question": "What research has been done in the MT-RL setting that observes the benefits of leveraging information gathered from other tasks?", "answer": ["Sample Complexity of Multi-task Reinforcement Learning", "Sharing Knowledge in Multi-Task Deep Reinforcement Learning", "Multi-Task Reinforcement Learning with Context-based Representations", "Provably Efficient Multi-Task Reinforcement Learning with Model Transfer", "Near-Optimal Representation Learning for Linear Bandits and Linear RL"], "answer_arxiv_id": ["1309.6821", "2401.09561", "2102.06177", "2107.08622", "2102.04132"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_12350"} +{"question": "Which works explored the use of deep generative models to learn stable structures from the data in the context of crystal structure prediction?", "answer": ["Data-Driven Approach to Encoding and Decoding 3-D Crystal Structures", "Crystal structure prediction of materials with high symmetry using differential evolution", "Distance Matrix based Crystal Structure Prediction using Evolutionary Algorithms", "Contact map based crystal structure prediction using global optimization", "CrystalGAN: Learning to Discover Crystallographic Structures with Generative Adversarial Networks", "Generative Adversarial Networks for Crystal Structure Prediction"], "answer_arxiv_id": ["1909.00949", "2104.09764", "2009.13955", "2008.07016", "1810.11203", "2004.01396v4"], "source_meta": {"published_time": "20230730"}, "qid": "AutoScholarQuery_train_12351"} +{"question": "What are some studies that underscore the importance of distributional robustness of notions of fairness?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Average Individual Fairness: Algorithms, Generalization and Experiments", "A Distributionally Robust Approach to Fair Classification", "Domain Adaptation meets Individual Fairness. And they get along."], "answer_arxiv_id": ["1911.08731", "1905.10607", "2007.09530", "2205.00504"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_12352"} +{"question": "Could you provide me some works on attempts to improve the use of the latent space?", "answer": ["All-but-the-Top: Simple and Effective Postprocessing for Word Representations", "Learning to Remove: Towards Isotropic Pre-trained BERT Embedding", "How Does Fine-tuning Affect the Geometry of Embedding Space: A Case Study on Isotropy", "Autoencoding Improves Pre-trained Word Embeddings", "Latent Space Cartography: Generalised Metric-Inspired Measures and Measure-Based Transformations for Generative Models", "Learning towards Minimum Hyperspherical Energy", "Representation Degeneration Problem in Training Natural Language Generation Models", "IsoBN: Fine-Tuning BERT with Isotropic Batch Normalization"], "answer_arxiv_id": ["1702.01417", "2104.05274", "2109.04740", "2010.13094", "1902.02113", "1805.09298", "1907.12009", "2005.02178"], "source_meta": {"published_time": "20221215"}, "qid": "AutoScholarQuery_train_12353"} +{"question": "What paper first employed a convolutional network to predict the image coordinates of 2D human joints in regression-based human pose estimators?", "answer": ["DeepPose: Human Pose Estimation via Deep Neural Networks"], "answer_arxiv_id": ["1312.4659"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_12354"} +{"question": "Can you mention some research that developed proxy sensitive attributes by training a second predictor on a different data set containing only sensitive attribute information?", "answer": ["Proxy Fairness", "Assessing Algorithmic Fairness with Unobserved Protected Class Using Data Combination", "Fairness Under Unawareness: Assessing Disparity When Protected Class Is Unobserved", "Evaluating Fairness of Machine Learning Models Under Uncertain and Incomplete Information"], "answer_arxiv_id": ["1806.11212", "1906.00285", "1811.11154", "2102.08410"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_12355"} +{"question": "Which 3D generative model has used Score Distillation Sampling for using a diffusion model to extract a NeRF given text prompts from users?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_12356"} +{"question": "Any works about offline RL in the context of zero-sum Markov games?", "answer": ["When is Offline Two-Player Zero-Sum Markov Game Solvable?", "Model-Based Reinforcement Learning Is Minimax-Optimal for Offline Zero-Sum Markov Games"], "answer_arxiv_id": ["2201.03522", "2206.04044"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_12357"} +{"question": "What work has utilized the implementation of volumetric ray-marching in view synthesis?", "answer": ["Escaping Plato's Cave: 3D Shape From Adversarial Rendering", "DeepVoxels: Learning Persistent 3D Feature Embeddings"], "answer_arxiv_id": ["1811.11606", "1812.01024v2"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_12358"} +{"question": "Any works there which proposed a reward-shaping strategy specifically designed for long-term credit assignment?", "answer": ["RUDDER: Return Decomposition for Delayed Rewards", "Optimizing Agent Behavior over Long Time Scales by Transporting Value", "Synthetic Returns for Long-Term Credit Assignment", "Self-Attentional Credit Assignment for Transfer in Reinforcement Learning", "Learning Long-Term Reward Redistribution via Randomized Return Decomposition", "Reinforcement Learning with Trajectory Feedback", "InferNet for Delayed Reinforcement Tasks: Addressing the Temporal Credit Assignment Problem"], "answer_arxiv_id": ["1806.07857", "1810.06721", "2102.12425", "1907.08027", "2111.13485", "2008.06036", "2105.00568"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_12359"} +{"question": "What studies present the use of Deep lattice networks (DLN) and lattices in learning monotonic functions?", "answer": ["Deep Lattice Networks and Partial Monotonic Functions"], "answer_arxiv_id": ["1709.06680"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_12360"} +{"question": "What works proposed the usage of language to interpret classifications in natural language processing?", "answer": ["Analogs of Linguistic Structure in Deep Representations", "e-SNLI: Natural Language Inference with Natural Language Explanations", "Explain Yourself! Leveraging Language Models for Commonsense Reasoning", "WT5?! Training Text-to-Text Models to Explain their Predictions"], "answer_arxiv_id": ["1707.08139", "1812.01193", "1906.02361", "2004.14546"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_12361"} +{"question": "Which research works focus on the fusion of visible and thermal features to enhance detection accuracy?", "answer": ["Multimodal Object Detection via Probabilistic Ensembling", "Improving Multispectral Pedestrian Detection by Addressing Modality\n Imbalance Problems", "Multi-modal Gated Mixture of Local-to-Global Experts for Dynamic Image\n Fusion"], "answer_arxiv_id": ["2104.02904", "2008.03043", "2302.01392"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_12362"} +{"question": "Which studies used diffusion-based methods in the medical imaging community?", "answer": ["Adaptive Diffusion Priors for Accelerated MRI Reconstruction", "Unsupervised Medical Image Translation with Adversarial Diffusion Models", "CoLa-Diff: Conditional Latent Diffusion Model for Multi-Modal MRI\n Synthesis", "Conversion Between CT and MRI Images Using Diffusion and Score-Matching\n Models"], "answer_arxiv_id": ["2207.05876", "2207.08208", "2303.14081", "2209.12104"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_12363"} +{"question": "What works are available on the usage of scoring rules to define training objectives for generative networks?", "answer": ["A Spectral Energy Distance for Parallel Speech Synthesis"], "answer_arxiv_id": ["2008.01160"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_12364"} +{"question": "Which papers established the concept of training reliable models with label noise?", "answer": ["Image Classification with Deep Learning in the Presence of Noisy Labels: A Survey"], "answer_arxiv_id": ["1912.05170"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12365"} +{"question": "Could you provide me some researches that claim these usual methods to address the catastrophic forgetting are hardly maintain models' accuracy on irrelevant inputs in ME task?", "answer": ["Editing Factual Knowledge in Language Models", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2104.08164", "2110.11309"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_12366"} +{"question": "Which works are related to text-based image editing methods, particularly DreamBooth, Cones2, and Textual inversion?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Cones 2: Customizable Image Synthesis with Multiple Subjects", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.12242", "2305.19327", "2208.01618"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_12367"} +{"question": "Are there any studies that practice meta-learning-based instance reweighting methods?", "answer": ["Learning to Reweight Examples for Robust Deep Learning", "Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting", "Faster Meta Update Strategy for Noise-Robust Deep Learning"], "answer_arxiv_id": ["1803.09050", "1902.07379", "2104.15092"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_12368"} +{"question": "Could you provide me a work that formulates the MIA as the variational inference?", "answer": ["Variational Model Inversion Attacks"], "answer_arxiv_id": ["2201.10787"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_12369"} +{"question": "Could you list some work that solved cross-task class discrimination with replay-based strategy and Gradient Self Adaption?", "answer": ["Dealing with Cross-Task Class Discrimination in Online Continual\n Learning"], "answer_arxiv_id": ["2305.14657"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_12370"} +{"question": "What papers discuss the use of value regularization in offline Reinforcement Learning methods?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Fisher Divergence Critic Regularization", "COMBO: Conservative Offline Model-Based Policy Optimization", "Constraints Penalized Q-learning for Safe Offline Reinforcement Learning"], "answer_arxiv_id": ["2006.04779", "2103.08050", "2102.08363", "2107.09003"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_12371"} +{"question": "What work shows that classifier-guidance can improve the quality of generated samples of Denoising Diffusion Probabilistic Models (DDPM)?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12372"} +{"question": "What works have been done on jointly learning 2D and 3D stylization?", "answer": ["StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D\n Mutual Learning"], "answer_arxiv_id": ["2205.12183"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_12373"} +{"question": "What works proposed an unsupervised cross-modal entity consistency verification method?", "answer": ["Multimodal Analytics for Real-world News using Measures of Cross-modal\n Entity Consistency"], "answer_arxiv_id": ["2003.10421"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_12374"} +{"question": "Which research initially provided framework of PMD that was later modified in the current study?", "answer": ["On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2201.07443"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_12375"} +{"question": "Could you provide me with research that has applied Neural Processes to various practical problems?", "answer": ["Sequential Neural Processes", "Meta-Learning surrogate models for sequential decision making", "Equivariant Learning of Stochastic Fields: Gaussian Processes and Steerable Conditional Neural Processes"], "answer_arxiv_id": ["1906.10264", "1903.11907", "2011.12916"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_12376"} +{"question": "Could you provide me some studies that use the rehearsal methods in Continual Learning?", "answer": ["Gradient Episodic Memory for Continual Learning"], "answer_arxiv_id": ["1706.08840"], "source_meta": {"published_time": "20230326"}, "qid": "AutoScholarQuery_train_12377"} +{"question": "Can you provide me with studies that have utilized the concept of minimizing intra-class variance and maximizing inter-class variance in classification tasks?", "answer": ["On Intra-Class Variance for Deep Learning of Classifiers"], "answer_arxiv_id": ["1901.11186"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_12378"} +{"question": "Could you provide me a research work that focused on using spatially adaptive transformations learned from semantic maps to improve the image quality?", "answer": ["Semantic Image Synthesis with Spatially-Adaptive Normalization"], "answer_arxiv_id": ["1903.07291"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_12379"} +{"question": "Could you mention some studies that investigated the benefits of domain randomization for sim-to-real transfer?", "answer": ["Understanding Domain Randomization for Sim-to-real Transfer"], "answer_arxiv_id": ["2110.03239"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_12380"} +{"question": "Which papers have combined the SO(2) equivalent cylindrical features with fully convolutional backbones?", "answer": ["SpinNet: Learning a General Surface Descriptor for 3D Point Cloud\n Registration"], "answer_arxiv_id": ["2011.12149"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_12381"} +{"question": "What works used CLIP encoders and augmented CLIP representations with additional features to detect hateful memes?", "answer": ["MOMENTA: A Multimodal Framework for Detecting Harmful Memes and Their\n Targets", "Prompting for Multimodal Hateful Meme Classification", "Hate-CLIPper: Multimodal Hateful Meme Classification based on\n Cross-modal Interaction of CLIP Features"], "answer_arxiv_id": ["2109.05184", "2302.04156", "2210.05916"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_12382"} +{"question": "Could you provide studies that use sample quality metrics to enable fast sampling?", "answer": ["Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality", "Fast Inference in Denoising Diffusion Models via MMD Finetuning"], "answer_arxiv_id": ["2202.05830", "2301.07969"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_12383"} +{"question": "Who used a mirror to calibrate the screen plane in the out-of-field view gaze collection system?", "answer": ["MPIIGaze: Real-World Dataset and Deep Appearance-Based Gaze Estimation", "ETH-XGaze: A Large Scale Dataset for Gaze Estimation under Extreme Head\n Pose and Gaze Variation"], "answer_arxiv_id": ["1711.09017", "2007.15837"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_12384"} +{"question": "What are the works in which researchers addressed the need for blind restoration methods?", "answer": ["Unsupervised Degradation Representation Learning for Blind\n Super-Resolution", "Unfolding the Alternating Optimization for Blind Super Resolution", "Flow-based Kernel Prior with Application to Blind Super-Resolution", "Blind Super-Resolution Kernel Estimation using an Internal-GAN", "Efficient and Degradation-Adaptive Network for Real-World Image\n Super-Resolution", "SwinIR: Image Restoration Using Swin Transformer", "Practical Blind Image Denoising via Swin-Conv-UNet and Data Synthesis", "Real-World Blind Super-Resolution via Feature Matching with Implicit\n High-Resolution Priors"], "answer_arxiv_id": ["2104.00416", "2010.02631", "2103.15977", "1909.06581", "2203.14216", "2108.10257", "2203.13278", "2202.13142"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_12385"} +{"question": "What research proposed the method of using generated images as the regularization dataset in model fine-tuning?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_12386"} +{"question": "Can you point me towards some research papers that discuss Improvement in model-generated instances using human feedback?", "answer": ["Training Language Models with Language Feedback at Scale"], "answer_arxiv_id": ["2303.16755"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_12387"} +{"question": "What works contribute to advancements in open-vocabulary segmentation?", "answer": ["A Simple Framework for Open-Vocabulary Segmentation and Detection", "DiscoBox: Weakly Supervised Instance Segmentation and Semantic\n Correspondence from Box Supervision", "Open-Vocabulary Instance Segmentation via Robust Cross-Modal\n Pseudo-Labeling", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2303.08131", "2105.06464", "2111.12698", "2112.12143", "2112.01518", "2303.04803"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_12388"} +{"question": "What research demonstrates the potential of employing diffusion models in video synthesis?", "answer": ["Video Diffusion Models", "Diffusion Probabilistic Modeling for Video Generation", "Flexible Diffusion Modeling of Long Videos", "VideoFusion: Decomposed Diffusion Models for High-Quality Video Generation", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video Generators", "Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2204.03458", "2203.09481", "2205.11495", "2303.08320", "2303.13439", "2304.08818"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_12389"} +{"question": "Could you provide me some works that have used Normalizing Flows for 3D shape generation?", "answer": ["PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows", "Discrete Point Flow Networks for Efficient Point Cloud Generation", "Go with the Flows: Mixtures of Normalizing Flows for Point Cloud Generation and Reconstruction"], "answer_arxiv_id": ["1906.12320", "2007.10170", "2106.03135v3"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_12390"} +{"question": "What works are related to exact unlearning through training a series of suodels?", "answer": ["Machine Unlearning"], "answer_arxiv_id": ["1912.03817"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_12391"} +{"question": "Which studies used this type of analysis in reinforcement learning environments?", "answer": ["Information-Theoretic Confidence Bounds for Reinforcement Learning", "Deciding What to Model: Value-Equivalent Sampling for Reinforcement Learning"], "answer_arxiv_id": ["1911.09724", "2206.02072"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12392"} +{"question": "Which studies demonstrate the application of transformers for running general purpose programs?", "answer": ["Looped Transformers as Programmable Computers"], "answer_arxiv_id": ["2301.13196v1"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12393"} +{"question": "Can you provide some studies that improved or inspired the creation of random features for the Gaussian kernel?", "answer": ["Explicit Approximations of the Gaussian Kernel", "Oblivious Sketching of High-Degree Polynomial Kernels"], "answer_arxiv_id": ["1109.4603", "1909.01410v5"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_12394"} +{"question": "Which works demonstrated the application of sparse MoEs in computer vision?", "answer": ["Scaling Vision with Sparse Mixture of Experts"], "answer_arxiv_id": ["2106.05974"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12395"} +{"question": "Which papers used ground-plane 2D grids for next-frame prediction?", "answer": ["FIERY: Future Instance Prediction in Bird’s-Eye View from Surround Monocular Cameras", "RegFlow: Probabilistic Flow-based Regression for Future Prediction"], "answer_arxiv_id": ["2104.10490", "2011.14620"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_12396"} +{"question": "Could you provide me papers that feature diverse applications of the T2I diffusion model, such as novel view synthesis and text-to-video generation?", "answer": ["Novel View Synthesis with Diffusion Models", "Person Image Synthesis via Denoising Diffusion Model", "Multi-Concept Customization of Text-to-Image Diffusion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Imagen Video: High Definition Video Generation with Diffusion Models"], "answer_arxiv_id": ["2210.04628", "2211.12500", "2212.04488", "2208.12242", "2209.14792", "2210.02303"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_12397"} +{"question": "Could you provide me some studies using probabilistic extensions of IoU in medical imaging?", "answer": ["Learning joint segmentation of tissues and brain lesions from task-specific hetero-modal domain-shifted datasets"], "answer_arxiv_id": ["2009.04009"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_12398"} +{"question": "Which research papers introduced MLPs with functional inputs and neurons, that transform the functional data to scalar values?", "answer": ["Representation of Functional Data in Neural Networks"], "answer_arxiv_id": ["0709.3641v1"], "source_meta": {"published_time": "20230114"}, "qid": "AutoScholarQuery_train_12399"} +{"question": "What papers pioneered the use of convolution networks for monocular depth estimation?", "answer": ["Depth Map Prediction from a Single Image using a Multi-Scale Deep\n Network"], "answer_arxiv_id": ["1406.2283"], "source_meta": {"published_time": "20240412"}, "qid": "AutoScholarQuery_train_12400"} +{"question": "What study result was closest to the researcher’s own work, proving that gradient flow converges to non-robust two-layer ReLU networks under certain assumptions?", "answer": ["Gradient Methods Provably Converge to Non-Robust Networks"], "answer_arxiv_id": ["2202.04347"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12401"} +{"question": "Could you provide me a study that proposed learning object-centric representations used for reward shaping?", "answer": ["Self-supervised Visual Reinforcement Learning with Object-centric Representations"], "answer_arxiv_id": ["2011.14381"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_12402"} +{"question": "What research suggested modeling functional uncertainty with the bootstrap method rather than using a single global latent variable?", "answer": ["Bootstrapping Neural Processes"], "answer_arxiv_id": ["2008.02956"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_12403"} +{"question": "What are some studies that showed interpolating between a model's fine-tuned weights and its pre-trained initialization can lead to improved performance on single tasks?", "answer": ["Robust fine-tuning of zero-shot models", "Merging Models with Fisher-Weighted Averaging", "Linear Mode Connectivity and the Lottery Ticket Hypothesis", "Averaging Weights Leads to Wider Optima and Better Generalization", "Diverse Weight Averaging for Out-of-Distribution Generalization", "Model Ratatouille: Recycling Diverse Models for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2109.01903", "2111.09832", "1912.05671", "1803.05407", "2205.09739", "2212.10445"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_12404"} +{"question": "What studies have focused on empirical research regarding reproducibility issues in the community?", "answer": ["Deep Reinforcement Learning that Matters", "Improving Reproducibility in Machine Learning Research (A Report from the NeurIPS 2019 Reproducibility Program)"], "answer_arxiv_id": ["1709.06560", "2003.12206"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_12405"} +{"question": "Which work proposed Prioritized Experience Replay (PER) to improve efficiency in replaying samples?", "answer": ["Prioritized Experience Replay"], "answer_arxiv_id": ["1511.05952"], "source_meta": {"published_time": "20220822"}, "qid": "AutoScholarQuery_train_12406"} +{"question": "Which works showcase the success of model-based RL on classic board games?", "answer": ["Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model"], "answer_arxiv_id": ["1911.08265"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_12407"} +{"question": "Do any works focus on the training of deep neural networks with label DP?", "answer": ["Deep Learning with Label Differential Privacy"], "answer_arxiv_id": ["2102.06062"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_12408"} +{"question": "Which works propose image manipulation through generative models like GANs and diffusion models?", "answer": ["Generative Adversarial Networks", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1406.2661", "2006.11239"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_12409"} +{"question": "What are some recent advancements in vision-language technologies for text-to-2D content generation?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2103.00020", "2006.11239", "2105.05233", "2112.10752", "2010.02502"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_12410"} +{"question": "Could you provide me with some studies that discussed FQE methodology?", "answer": ["Bootstrapping Fitted Q-Evaluation for Off-Policy Inference", "Batch Policy Learning under Constraints", "Statistical Bootstrapping for Uncertainty Estimation in Off-Policy Evaluation"], "answer_arxiv_id": ["2102.03607", "1903.08738", "2007.13609"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_12411"} +{"question": "Which works discussed the methods for generating adversarial examples without any knowledge of the victim model?", "answer": ["Practical Black-Box Attacks against Machine Learning", "ZOO: Zeroth Order Optimization Based Black-box Attacks to Deep Neural Networks without Training Substitute Models", "Black-box Adversarial Attacks with Limited Queries and Information", "Improving Black-box Adversarial Attacks with a Transfer-based Prior", "Improving Transferability of Adversarial Examples with Input Diversity", "Backpropagating Linearly Improves Transferability of Adversarial Examples"], "answer_arxiv_id": ["1602.02697", "1708.03999", "1804.08598", "1906.06919", "1803.06978", "2012.03528"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_12412"} +{"question": "What works combined uncertainty-based and diversity-based sampling strategies with Active Learning for further improvements?", "answer": ["Contextual Diversity for Active Learning"], "answer_arxiv_id": ["2008.05723"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_12413"} +{"question": "What paper introduced Structure Diffusion that feeds segmented text prompts to the text encoder?", "answer": ["Training-Free Structured Diffusion Guidance for Compositional\n Text-to-Image Synthesis"], "answer_arxiv_id": ["2212.05032"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_12414"} +{"question": "Could you cite studies that extend the vision-language pre-training task with multi-modal text completion and text generation for auxiliary learning?", "answer": ["Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal Skip-connections"], "answer_arxiv_id": ["2107.07651", "2201.12086", "2205.12005"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_12415"} +{"question": "Could you provide me some studies that have adopted risk-sensitive reinforcement learning in MARL?", "answer": ["Risk Perspective Exploration in Distributional Reinforcement Learning", "RMIX: Learning Risk-Sensitive Policies for Cooperative Reinforcement Learning Agents"], "answer_arxiv_id": ["2206.14170", "2102.08159"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_12416"} +{"question": "Could you provide me some works about addressing inverse physics problems in embodied audio-visual AI tasks by utilizing multi-modal information?", "answer": ["Audio-Visual Floorplan Reconstruction", "Semantic Audio-Visual Navigation", "Look, Listen, and Act: Towards Audio-Visual Embodied Navigation", "Finding Fallen Objects Via Asynchronous Audio-Visual Integration"], "answer_arxiv_id": ["2012.15470", "2012.11583", "1912.11684", "2207.03483"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12417"} +{"question": "Which studies can be combined with techniques such as momentum for improved utility?", "answer": ["Momentum Aggregation for Private Non-convex ERM"], "answer_arxiv_id": ["2210.06328"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_12418"} +{"question": "What works have utilized deep learning for unary matching cost computation?", "answer": ["Stereo Matching by Training a Convolutional Neural Network to Compare\n Image Patches", "End-to-End Training of Hybrid CNN-CRF Models for Stereo", "Learning to Compare Image Patches via Convolutional Neural Networks"], "answer_arxiv_id": ["1510.05970", "1611.10229", "1504.03641"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_12419"} +{"question": "Which papers elaborate on the common approaches that evaluate explanation correctness based on feature or pixel perturbation?", "answer": ["Evaluating the visualization of what a Deep Neural Network has learned", "Full-Gradient Representation for Neural Network Visualization", "OpenXAI: Towards a Transparent Evaluation of Model Explanations"], "answer_arxiv_id": ["1509.06321", "1905.00780", "2206.11104"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_12420"} +{"question": "Which research papers discuss the combination of continued pretraining with finetuning in the context of transforming LLMs into bi-encoders?", "answer": ["Text and Code Embeddings by Contrastive Pre-Training", "Large Dual Encoders Are Generalizable Retrievers"], "answer_arxiv_id": ["2201.10005", "2112.07899"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_12421"} +{"question": "What works discuss data structure compatibility scheme?", "answer": ["K-BERT: Enabling Language Representation with Knowledge Graph"], "answer_arxiv_id": ["1909.07606"], "source_meta": {"published_time": "20240203"}, "qid": "AutoScholarQuery_train_12422"} +{"question": "Could you provide me some references that proposed enhancements to the generalization ability of the diffusion model?", "answer": ["AdaptDiffuser: Diffusion Models as Adaptive Self-evolving Planners"], "answer_arxiv_id": ["2302.01877"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12423"} +{"question": "Which papers are related to the recent advances in the field of neural implicit representation?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction"], "answer_arxiv_id": ["2003.08934", "2206.00665"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_12424"} +{"question": "Could you provide me some works that train self-supervised representation learning models by reconstructing speech signals?", "answer": ["Unsupervised speech representation learning using WaveNet autoencoders"], "answer_arxiv_id": ["1901.08810"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_12425"} +{"question": "What papers explore the use of motifs within contrastive learning?", "answer": ["MoCL: Data-driven Molecular Fingerprint via Knowledge-aware Contrastive Learning from Molecular Graph", "Motif-Driven Contrastive Learning of Graph Representations"], "answer_arxiv_id": ["2106.04509", "2012.12533"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_12426"} +{"question": "What paper originally achieved first-order regret?", "answer": ["Tighter Problem-Dependent Regret Bounds in Reinforcement Learning without Domain Knowledge using Value Function Bounds"], "answer_arxiv_id": ["1901.00210"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_12427"} +{"question": "Which papers introduced the first convergence analysis for a simple double-loop procedure in both deterministic and stochastic settings in bilevel optimization?", "answer": ["Approximation Methods for Bilevel Programming"], "answer_arxiv_id": ["1802.02246"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_12428"} +{"question": "What is the paper that proposed initialization for GNNs with pre-trained MLP parameters?", "answer": ["MLPInit: Embarrassingly Simple GNN Training Acceleration with MLP Initialization"], "answer_arxiv_id": ["2210.00102"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_12429"} +{"question": "Are there any works that use graph neural networks to constrain inter-element relationships for conditional layout generation?", "answer": ["Neural Design Network: Graphic Layout Generation with Constraints"], "answer_arxiv_id": ["1912.09421"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_12430"} +{"question": "Can you name some of the works that utilize the concept of simplicial complexes to account for higher-order relations among nodes?", "answer": ["Simplicial Neural Networks", "Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks", "SGAT: Simplicial Graph Attention Network"], "answer_arxiv_id": ["2010.03633", "2103.03212", "2207.11761"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_12431"} +{"question": "Which works use roto-translation invariant features as inputs such as inter-atomic distances, angles or the principal axes of inertia for improving 3D molecular property prediction?", "answer": ["SchNet: A continuous-filter convolutional neural network for modeling quantum interactions", "Directional Message Passing for Molecular Graphs", "Rotation Invariant Graph Neural Networks using Spin Convolutions", "Spherical Message Passing for 3D Molecular Graphs"], "answer_arxiv_id": ["1706.08566", "2003.03123", "2106.09575", "2102.05013"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_12432"} +{"question": "Could you provide me studies about distilling interaction based models into representation based models?", "answer": ["Improving Efficient Neural Ranking Models with Cross-Architecture Knowledge Distillation", "PAIR: Leveraging Passage-Centric Similarity Relation for Improving Dense Passage Retrieval", "RocketQAv2: A Joint Training Method for Dense Passage Retrieval and Passage Re-ranking", "Adversarial Retriever-Ranker for Dense Text Retrieval"], "answer_arxiv_id": ["2010.02666", "2108.06027", "2110.07367", "2110.03611"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_train_12433"} +{"question": "What papers proposed approximation algorithms for subadditive functions?", "answer": ["Almost Envy-Freeness with General Valuations", "Maximin-Aware Allocations of Indivisible Goods"], "answer_arxiv_id": ["1707.04769", "1905.09969"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_12434"} +{"question": "Can you identify the studies that incorporated an additional retraining step after the best architecture is identified in one-shot NAS?", "answer": ["FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search", "Blockwisely Supervised Neural Architecture Search with Knowledge Distillation"], "answer_arxiv_id": ["1812.03443", "1911.13053"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_12435"} +{"question": "Are there any studies that have used transfer learning or distillation for studying multi-task robotic policies?", "answer": ["Actor-Mimic Deep Multitask and Transfer Reinforcement Learning", "Distral: Robust Multitask Reinforcement Learning", "Knowledge Transfer in Multi-Task Deep Reinforcement Learning for Continuous Control", "Policy Distillation"], "answer_arxiv_id": ["1511.06342", "1707.04175", "2010.07494", "1511.06295"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_12436"} +{"question": "Which research explored more on altering the architecture of the neural networks in the context of VMC?", "answer": ["Gold-standard solutions to the Schrödinger equation using deep learning: How much physics do we need?", "Ab-initio quantum chemistry with neural-network wavefunctions"], "answer_arxiv_id": ["2205.09438", "2208.12590"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_12437"} +{"question": "Which works focus on generating 2D novel views based on diffusion models?", "answer": ["Novel View Synthesis with Diffusion Models", "Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2210.04628", "2303.11328"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_12438"} +{"question": "What works are there on predicting multiple structured outputs?", "answer": ["Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypotheses"], "answer_arxiv_id": ["1612.00197"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_12439"} +{"question": "Which studies focus on fairness in a context of federated learning?", "answer": ["Fair Resource Allocation in Federated Learning"], "answer_arxiv_id": ["1905.10497"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_12440"} +{"question": "Which works extends the structure in object variation studies by leveraging shape prior?", "answer": ["Shape Prior Deformation for Categorical 6D Object Pose and Size\n Estimation", "UDA-COPE: Unsupervised Domain Adaptation for Category-level Object Pose\n Estimation", "Learning Canonical Shape Space for Category-Level 6D Object Pose and\n Size Estimation", "ShAPO: Implicit Representations for Multi-Object Shape, Appearance, and\n Pose Optimization"], "answer_arxiv_id": ["2007.08454", "2111.12580", "2001.09322", "2207.13691"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_12441"} +{"question": "Which papers have been published on the topic of text-to-image diffusion models?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_12442"} +{"question": "What papers studied the early-time training phenomenon (ETP) in conventional label noise scenarios?", "answer": ["Early-Learning Regularization Prevents Memorization of Noisy Labels"], "answer_arxiv_id": ["2007.00151"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_12443"} +{"question": "Which studies utilize the sensor point cloud directly for semantic segmentation instead of the mesh-sampled point cloud?", "answer": ["Multi-view PointNet for 3D Scene Understanding", "Bidirectional Projection Network for Cross Dimension Scene Understanding", "Learning Multi-View Aggregation In the Wild for Large-Scale 3D Semantic\n Segmentation"], "answer_arxiv_id": ["1909.13603", "2103.14326", "2204.07548"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_12444"} +{"question": "Which research introduced the replacement of local features by CNN-derived features in image retrieval?", "answer": ["Neural Codes for Image Retrieval"], "answer_arxiv_id": ["1404.1777"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_12445"} +{"question": "What studies incorporated indirect illumination to enhance the fidelity of the estimated BRDF?", "answer": ["Modeling Indirect Illumination for Inverse Rendering", "Shape, Light, and Material Decomposition from Images using Monte Carlo\n Rendering and Denoising", "NeILF: Neural Incident Light Field for Physically-based Material\n Estimation"], "answer_arxiv_id": ["2204.06837", "2206.03380", "2203.07182"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_12446"} +{"question": "Which works discussed basis learning for time series analysis?", "answer": ["Are Transformers Effective for Time Series Forecasting?", "An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling"], "answer_arxiv_id": ["2205.13504", "1803.01271"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_12447"} +{"question": "What works focus on classification and estimate parameters of Dirichlet distribution?", "answer": ["Uncertainty on Asynchronous Time Event Prediction", "Evidential Deep Learning to Quantify Classification Uncertainty", "Posterior Network: Uncertainty Estimation without OOD Samples via Density-Based Pseudo-Counts", "Graph Posterior Network: Bayesian Predictive Uncertainty for Node Classification", "Towards Maximizing the Representation Gap between In-Domain & Out-of-Distribution Examples"], "answer_arxiv_id": ["1911.05503", "1806.01768", "2006.09239", "2110.14012", "2010.10474"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_12448"} +{"question": "Could you provide me a study about multiple features competing in language modelling?", "answer": ["Language Through a Prism: A Spectral Approach for Multiscale Language Representations"], "answer_arxiv_id": ["2011.04823"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_12449"} +{"question": "Which works showed that the NTK limit insufficiently characterizes realistic deep neural networks?", "answer": ["What can linearized neural networks actually say about generalization?", "Limitations of the NTK for Understanding Generalization in Deep Learning"], "answer_arxiv_id": ["2106.06770", "2206.10012"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_12450"} +{"question": "What paper described the use of straight-through estimator in encoder training?", "answer": ["Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation"], "answer_arxiv_id": ["1308.3432"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_12451"} +{"question": "What are the methods that rely on behavior cloning for policy regularization in offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "A Minimalist Approach to Offline Reinforcement Learning", "Behavior Regularized Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["1812.02900", "1906.00949", "2106.06860", "1911.11361", "2110.06169"], "source_meta": {"published_time": "20220812"}, "qid": "AutoScholarQuery_train_12452"} +{"question": "Could you provide me some studies developing tasks involving actions on static internet pages?", "answer": ["Mind2Web: Towards a Generalist Agent for the Web"], "answer_arxiv_id": ["2306.06070"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_12453"} +{"question": "What studies employed QE scores for data filtering, curriculum construction, and assigning quality-related tags to the outputs in NMT systems?", "answer": ["MBR and QE Finetuning: Training-time Distillation of the Best and Most\n Expensive Decoding Methods", "Reinforced Self-Training (ReST) for Language Modeling", "Quality-Aware Translation Models: Efficient Generation and Quality\n Estimation in a Single Model"], "answer_arxiv_id": ["2309.10966", "2308.08998", "2310.06707"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_12454"} +{"question": "Which studies proposed sampling-based scalable graph transformers?", "answer": ["Gophormer: Ego-Graph Transformer for Node Classification"], "answer_arxiv_id": ["2110.13094"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_12455"} +{"question": "Any works about improving the quality and speed of 3D object production in 2D diffusion?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation"], "answer_arxiv_id": ["2211.10440"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_12456"} +{"question": "Which papers proposed learning the optimal covariance of reverse process?", "answer": ["Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models"], "answer_arxiv_id": ["2201.06503", "2206.07309"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_12457"} +{"question": "Any research papers that utilize Concept Bottleneck Models (CBMs) in concept-based interpretable models?", "answer": ["Concept Bottleneck Models"], "answer_arxiv_id": ["2007.04612"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_12458"} +{"question": "Which paper discussed that Fourier Neural Operators are insufficiently expressive for complex configurations like turbulent and multi-physics flows?", "answer": ["U-FNO - an enhanced Fourier neural operator-based deep-learning model for multiphase flow"], "answer_arxiv_id": ["2109.03697"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_12459"} +{"question": "What are some universal tamper localization methods?", "answer": ["Image Manipulation Detection by Multi-View Multi-Scale Supervision", "Learning to Immunize Images for Tamper Localization and Self-Recovery", "From Image to Imuge: Immunized Image Generation", "DRAW: Defending Camera-shooted RAW against Image Manipulation", "RWN: Robust Watermarking Network for Image Cropping Localization"], "answer_arxiv_id": ["2104.06832", "2210.15902", "2110.14196", "2307.16418", "2110.05687v2"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_12460"} +{"question": "Which research first proposed the minimax estimation procedure?", "answer": ["Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation"], "answer_arxiv_id": ["1810.12429"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_12461"} +{"question": "Which studies explored the disparities in practical tasks and the learning of a fair predictor?", "answer": ["Balanced Datasets Are Not Enough: Estimating and Mitigating Gender Bias\n in Deep Image Representations", "Learning Adversarially Fair and Transferable Representations", "Towards Fairness in Visual Recognition: Effective Strategies for Bias\n Mitigation"], "answer_arxiv_id": ["1811.08489", "1802.06309", "1911.11834"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_12462"} +{"question": "Are there any works that learn probabilistic constraints?", "answer": ["Making Human-Like Trade-offs in Constrained Environments by Learning from Demonstrations", "Maximum Likelihood Constraint Inference for Inverse Reinforcement Learning", "Maximum Likelihood Constraint Inference from Stochastic Demonstrations", "Inverse Constrained Reinforcement Learning"], "answer_arxiv_id": ["2109.11018", "1909.05477", "2102.12554", "2011.09999"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_12463"} +{"question": "Which papers focused on the convergence rate of Local SGD in a homogeneous setting?", "answer": ["Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning", "Local SGD Converges Fast and Communicates Little", "Tighter Theory for Local SGD on Identical and Heterogeneous Data", "Is Local SGD Better than Minibatch SGD?"], "answer_arxiv_id": ["1807.06629", "1805.09767", "1909.04746", "2002.07839"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12464"} +{"question": "What papers discussed self-training methods that utilize highly confident predictions as hard pseudo-labels?", "answer": ["FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Self-training with Noisy Student improves ImageNet classification"], "answer_arxiv_id": ["2001.07685v2", "1911.04252"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_12465"} +{"question": "Which papers have used synthetic scans for training human reconstruction methods?", "answer": ["Multi-Garment Net: Learning to Dress 3D People from Images", "SIZER: A Dataset and Model for Parsing 3D Clothing and Learning Size\n Sensitive 3D Clothing", "AGORA: Avatars in Geography Optimized for Regression Analysis"], "answer_arxiv_id": ["1908.06903", "2007.11610", "2104.14643"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_12466"} +{"question": "Could you point me to some research that utilized an ImageNet-pretrained ResNet as a backbone for Continual Learning?", "answer": ["Class-Incremental Learning for Action Recognition in Videos"], "answer_arxiv_id": ["2203.13611"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_12467"} +{"question": "What research papers have studied the effects of freezing earlier layers after training the first task on the performance of the second task?", "answer": ["Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics", "Are All Layers Created Equal?"], "answer_arxiv_id": ["2007.07400", "1902.01996"], "source_meta": {"published_time": "20230409"}, "qid": "AutoScholarQuery_train_12468"} +{"question": "Which papers investigate into the process of integrating deep learning techniques with causal inference, particularly in computer vision?", "answer": ["Causal Intervention for Weakly-Supervised Semantic Segmentation"], "answer_arxiv_id": ["2009.12547"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_12469"} +{"question": "Who leveraged graph Laplacian to learn eigenoptions?", "answer": ["Learning Purposeful Behaviour in the Absence of Rewards", "A Laplacian Framework for Option Discovery in Reinforcement Learning"], "answer_arxiv_id": ["1605.07700", "1703.00956"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12470"} +{"question": "Which papers proposed using one neural network to represent one material for neural based SVBRDF / BTF / BRDF modeling?", "answer": ["Neural BRDF Representation and Importance Sampling", "NeuMIP: Multi-Resolution Neural Materials"], "answer_arxiv_id": ["2102.05963", "2104.02789"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_12471"} +{"question": "Which study first formally introduced the phenomenon of Overkill and constructed a high-quality dataset for it?", "answer": ["XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in\n Large Language Models"], "answer_arxiv_id": ["2308.01263"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_12472"} +{"question": "Which papers introduced the likelihood weighting method for training score-based diffusion models?", "answer": ["Maximum Likelihood Training of Score-Based Diffusion Models"], "answer_arxiv_id": ["2101.09258"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_12473"} +{"question": "Can you give examples of studies that focused on combining labeled training data with SupCon?", "answer": ["Unified Contrastive Learning in Image-Text-Label Space"], "answer_arxiv_id": ["2204.03610"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_12474"} +{"question": "What works focused on the acceleration of NeRF using voxel grid or hash table?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "Neural Sparse Voxel Fields", "TensoRF: Tensorial Radiance Fields", "FastNeRF: High-Fidelity Neural Rendering at 200FPS", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction"], "answer_arxiv_id": ["2201.05989", "2112.05131", "2007.11571", "2203.09517", "2103.10380", "2111.11215"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_12475"} +{"question": "What are some studies that discuss speculative sampling?", "answer": ["Accelerating Large Language Model Decoding with Speculative Sampling", "Fast Inference from Transformers via Speculative Decoding"], "answer_arxiv_id": ["2302.01318", "2211.17192"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_12476"} +{"question": "What paper proposed a convolutional neural network (CNN) approach that outperformed LN and GLM baselines?", "answer": ["Deep Learning Models of the Retinal Response to Natural Scenes"], "answer_arxiv_id": ["1702.01825"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_12477"} +{"question": "Could you mention some works that present summaries of various optimal bounds for the fixed confidence BAI?", "answer": ["Tight (Lower) Bounds for the Fixed Budget Best Arm Identification Bandit Problem", "On the Complexity of Best-Arm Identification in Multi-Armed Bandit Models"], "answer_arxiv_id": ["1605.09004v1", "1407.4443"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_12478"} +{"question": "Which works propose strategies to address the spectral bias issue in INR-based methods?", "answer": ["Fourier Features Let Networks Learn High Frequency Functions in Low\n Dimensional Domains", "Polynomial Implicit Neural Representations For Large Diverse Datasets", "Neural Free-Viewpoint Relighting for Glossy Indirect Illumination", "Neural Fourier Filter Bank"], "answer_arxiv_id": ["2006.10739", "2303.11424", "2307.06335", "2212.01735"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_12479"} +{"question": "Can you name the research works that refer to the challenges of designing interventions in representations?", "answer": ["Towards Best Practices of Activation Patching in Language Models:\n Metrics and Methods"], "answer_arxiv_id": ["2309.16042"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_12480"} +{"question": "What work has empirically demonstrated that recent domain generalization algorithms exhibit no improvement compared with Empirical Risk Minimization (ERM)?", "answer": ["In Search of Lost Domain Generalization"], "answer_arxiv_id": ["2007.01434"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_12481"} +{"question": "Which research papers have introduced mathematical reasoning datasets?", "answer": ["Training Verifiers to Solve Math Word Problems", "NumGLUE: A Suite of Fundamental yet Challenging Mathematical Reasoning\n Tasks", "Are NLP Models really able to Solve Simple Math Word Problems?"], "answer_arxiv_id": ["2110.14168", "2204.05660", "2103.07191"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_12482"} +{"question": "What works dealt with video summarization?", "answer": ["Predicting Important Objects for Egocentric Video Summarization"], "answer_arxiv_id": ["1505.04803"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_12483"} +{"question": "Which works are related to the topic of latent RL?", "answer": ["RL for Latent MDPs: Regret Guarantees and a Lower Bound"], "answer_arxiv_id": ["2102.04939"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_12484"} +{"question": "Could you provide me some works about robust RL where regularized-based methods were used?", "answer": ["Robust Deep Reinforcement Learning against Adversarial Perturbations on State Observations", "Deep Reinforcement Learning with Robust and Smooth Policy", "Robust Deep Reinforcement Learning through Adversarial Loss", "Policy Smoothing for Provably Robust Reinforcement Learning", "Learning Robust Policy against Disturbance in Transition Dynamics via State-Conservative Policy Optimization"], "answer_arxiv_id": ["2003.08938", "2003.09534", "2008.01976", "2106.11420", "2112.10513"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_12485"} +{"question": "What studies talk about spectral GNNs as a subset of graph neural networks?", "answer": ["Spectral Networks and Deep Locally Connected Networks on Graphs", "Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering", "Convolutional Neural Network Architectures for Signals Supported on Graphs", "CayleyNets: Graph Convolutional Neural Networks with Complex Rational Spectral Filters"], "answer_arxiv_id": ["1312.6203", "1606.09375", "1805.00165", "1705.07664"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_12486"} +{"question": "Could you provide me some works that make use of Bayesian Neural Networks (BNNs) for uncertainty estimation?", "answer": ["Probabilistic Backpropagation for Scalable Learning of Bayesian Neural Networks"], "answer_arxiv_id": ["1502.05336"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_12487"} +{"question": "Which works are related to creating sign language translations dataset using ad hoc recorded footage?", "answer": ["Improving Sign Language Translation with Monolingual Data by Sign Back-Translation", "Neural Sign Language Translation based on Human Keypoint Estimation", "How2Sign: A Large-scale Multimodal Dataset for Continuous American Sign Language"], "answer_arxiv_id": ["2105.12397", "1811.11436", "2008.08143"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_12488"} +{"question": "Could you provide me some studies that involve the formulation of a style reward in goal-conditioned RL?", "answer": ["AMP: Adversarial Motion Priors for Stylized Physics-Based Character\n Control"], "answer_arxiv_id": ["2104.02180"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_12489"} +{"question": "What publications discuss the impact of Empirical Risk Minimization (ERM) and OOD objectives on feature learning?", "answer": ["Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization", "On Feature Learning in the Presence of Spurious Correlations"], "answer_arxiv_id": ["2202.06856", "2210.11369"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_12490"} +{"question": "Could you name some works that studied policy optimization under reachability assumptions?", "answer": ["On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift", "Global Optimality Guarantees For Policy Gradient Methods", "Neural Proximal/Trust Region Policy Optimization Attains Globally Optimal Policy", "A Unified View of Entropy-Regularized Markov Decision Processes"], "answer_arxiv_id": ["1908.00261", "1906.01786", "1906.10306", "1705.07798"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_12491"} +{"question": "What works proposed the GradNorm method?", "answer": ["On the Importance of Gradients for Detecting Distributional Shifts in the Wild"], "answer_arxiv_id": ["2110.00218"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_12492"} +{"question": "Which papers introduced Continuous Normalizing Flows?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_12493"} +{"question": "Which studies in the literature focus on learning algorithms for solving specific IP problems?", "answer": ["Learning Combinatorial Optimization Algorithms over Graphs", "Attention, Learn to Solve Routing Problems!", "Learning chordal extensions", "Reinforcement Learning for Solving the Vehicle Routing Problem"], "answer_arxiv_id": ["1704.01665", "1803.08475", "1910.07600v1", "1802.04240"], "source_meta": {"published_time": "20230520"}, "qid": "AutoScholarQuery_train_12494"} +{"question": "What works capture the double descent phenomenon in well-specified models?", "answer": ["A Random Matrix Perspective on Mixtures of Nonlinearities in High Dimensions", "What Causes the Test Error? Going Beyond Bias-Variance via ANOVA"], "answer_arxiv_id": ["1912.00827", "2010.05170"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_12495"} +{"question": "Which works employed sample selection as a method in dealing with noisy labels in learning?", "answer": ["DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "UniCon: Combating Label Noise Through Uniform Selection and Contrastive Learning", "Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels", "How does Disagreement Help Generalization against Label Corruption?"], "answer_arxiv_id": ["2002.07394", "2203.14542", "1804.06872", "1901.04215"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_12496"} +{"question": "Which studies focus on causal reasoning in machine learning?", "answer": ["A Causal View on Robustness of Neural Networks", "Visual Causal Feature Learning", "Representation Learning via Invariant Causal Mechanisms", "Invariance, Causality and Robustness", "Long-Tailed Classification by Keeping the Good and Removing the Bad Momentum Causal Effect", "Counterfactual Generative Networks"], "answer_arxiv_id": ["2005.01095", "1412.2309", "2010.07922", "1812.08233v1", "2009.12991", "2101.06046"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_12497"} +{"question": "Which studies have looked into the tallying settings that are special cases of WTB with m=T?", "answer": ["Rotting Bandits", "Addressing the Long-term Impact of ML Decisions via Policy Regret", "Stochastic Rising Bandits"], "answer_arxiv_id": ["1702.07274", "2106.01325", "2212.03798v1"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_12498"} +{"question": "Which work investigated the challenges of self-supervised learning from continuous data streams and proposed using a minimum-redundancy 'replay' buffer?", "answer": ["The Challenges of Continuous Self-Supervised Learning"], "answer_arxiv_id": ["2203.12710"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_12499"} +{"question": "Could you provide me some studies that introduced a self-consistency decoding strategy to sample multiple outputs of LLMs?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_12500"} +{"question": "Which paper generalizes the variational orthogonal features to the sequential case?", "answer": ["SigGPDE: Scaling Sparse Gaussian Processes on Sequential Data"], "answer_arxiv_id": ["2105.04211"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_12501"} +{"question": "Which work aimed to improve the applicability of existing 3D models by developing machine learning models that predict 3D geometry?", "answer": ["Molecule3D: A Benchmark for Predicting 3D Geometries from Molecular Graphs"], "answer_arxiv_id": ["2110.01717"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_12502"} +{"question": "Could you specify a research study that combined global context information into objects via object pair fusion and an entity-to-relation fusion module?", "answer": ["Structured Sparse R-CNN for Direct Scene Graph Generation"], "answer_arxiv_id": ["2106.10815"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_12503"} +{"question": "Which study proposed to train a singleton example scorer using contrastive learning with signals from LM inferencer?", "answer": ["Learning To Retrieve Prompts for In-Context Learning"], "answer_arxiv_id": ["2112.08633"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_12504"} +{"question": "What papers describe the successful use of GANs in text-to-image generation?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image Synthesis", "Cross-Modal Contrastive Learning for Text-to-Image Generation"], "answer_arxiv_id": ["1711.10485", "1904.01310", "2101.04702"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_12505"} +{"question": "Could you provide me references that parse noisy annotations from corresponding metadata?", "answer": ["Revisiting Unreasonable Effectiveness of Data in Deep Learning Era", "Scaling Vision Transformers", "Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["1707.02968", "2106.04560", "2103.00020", "2102.05918", "2204.14198"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_12506"} +{"question": "Which papers discuss the convergence of homogeneous neural networks to a KKT point of the maximum-margin problem?", "answer": ["Gradient Descent Maximizes the Margin of Homogeneous Neural Networks", "Directional convergence and alignment in deep learning"], "answer_arxiv_id": ["1906.05890", "2006.06657"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_12507"} +{"question": "Which papers pertain to unsupervised domain adaptation that adapts representations across visually different source and target domains?", "answer": ["A Survey of Unsupervised Domain Adaptation for Visual Recognition", "A Survey of Unsupervised Deep Domain Adaptation", "Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation", "Domain-Adversarial Training of Neural Networks", "CyCADA: Cycle-Consistent Adversarial Domain Adaptation"], "answer_arxiv_id": ["2112.06745", "1812.02849", "1711.06969", "1505.07818", "1711.03213"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_12508"} +{"question": "What works establish the theory of causal abstraction?", "answer": ["Causal Consistency of Structural Equation Models", "Abstracting Causal Models"], "answer_arxiv_id": ["1707.00819", "1812.03789"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_12509"} +{"question": "Which works pre-define a fixed node ordering in their autoregressive graph generative models?", "answer": ["GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models", "GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation"], "answer_arxiv_id": ["1802.08773", "2001.09382"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_12510"} +{"question": "Which research introduced the concept of patch-based backdoor attacks?", "answer": ["Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning", "Label-Consistent Backdoor Attacks", "Backdoor Attack in the Physical World"], "answer_arxiv_id": ["1712.05526v1", "1912.02771", "2104.02361"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_12511"} +{"question": "Which works applied dense 3D convolutions after converting a point cloud into a dense voxel grid?", "answer": ["Volumetric and Multi-View CNNs for Object Classification on 3D Data"], "answer_arxiv_id": ["1604.03265"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_12512"} +{"question": "Can you list the works that defined the value function as a function of the number of visits to a state?", "answer": ["Modular Deep Reinforcement Learning for Continuous Motion Planning with Temporal Logic"], "answer_arxiv_id": ["2102.12855"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_12513"} +{"question": "What studies focus on the bottom-up methods in human pose estimation?", "answer": ["PifPaf: Composite Fields for Human Pose Estimation", "HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human\n Pose Estimation", "Rethinking the Heatmap Regression for Bottom-up Human Pose Estimation", "The Center of Attention: Center-Keypoint Grouping via Attention for\n Multi-Person Pose Estimation"], "answer_arxiv_id": ["1903.06593", "1908.10357", "2012.15175", "2110.05132"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_12514"} +{"question": "What work proposed a neural estimator for a lower bound on R​(D)?", "answer": ["Towards Empirical Sandwich Bounds on the Rate-Distortion Function"], "answer_arxiv_id": ["2111.12166"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_12515"} +{"question": "Are there other works in the BERT series methods apart from BERT and BART?", "answer": ["RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1907.11692"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_12516"} +{"question": "What research papers focus on model compression techniques for efficient inference in large language models?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained\n Quantization and Huffman Coding", "A Simple and Effective Pruning Approach for Large Language Models", "Compressing LLMs: The Truth is Rarely Pure and Never Simple", "Compress, Then Prompt: Improving Accuracy-Efficiency Trade-off of LLM\n Inference with Transferable Prompt", "OmniQuant: Omnidirectionally Calibrated Quantization for Large Language\n Models", "AWQ: Activation-aware Weight Quantization for LLM Compression and\n Acceleration", "Atom: Low-bit Quantization for Efficient and Accurate LLM Serving", "Intriguing Properties of Quantization at Scale", "LLM-QAT: Data-Free Quantization Aware Training for Large Language Models", "Norm Tweaking: High-performance Low-bit Quantization of Large Language\n Models"], "answer_arxiv_id": ["1510.00149", "2306.11695", "2310.01382", "2305.11186", "2308.13137", "2306.00978", "2310.19102v3", "2305.19268", "2305.17888v1", "2309.02784"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_12517"} +{"question": "Which papers introduced protein generation based on language modeling?", "answer": ["Accelerating Bayesian Optimization for Biological Sequence Design with Denoising Autoencoders"], "answer_arxiv_id": ["2203.12742"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_12518"} +{"question": "Any papers discussing few-shot 3D reconstruction from RGB images integrating model-based fitting and test-time fine-tuning of model parameters?", "answer": ["H3D-Net: Few-Shot High-Fidelity 3D Head Reconstruction", "Single-Shot Implicit Morphable Faces with Consistent Texture\n Parameterization", "Preface: A Data-driven Volumetric Prior for Few-shot Ultra\n High-resolution Face Synthesis"], "answer_arxiv_id": ["2107.12512", "2305.03043", "2309.16859"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12519"} +{"question": "Which works introduce pre-training of the embedding models with weakly-supervised contrastive learning?", "answer": ["Text Embeddings by Weakly-Supervised Contrastive Pre-training"], "answer_arxiv_id": ["2212.03533"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_12520"} +{"question": "What studies provide evidence that constructing a no-linear-swap regret learner present benefits when compared to other less rational learners?", "answer": ["Bayes correlated equilibria and no-regret dynamics"], "answer_arxiv_id": ["2304.05005"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_12521"} +{"question": "Are there any works that have proposed the use of adapter tuning and mask tuning modules in the adaptation of ASR models to low-resource languages?", "answer": ["Exploiting Adapters for Cross-lingual Low-resource Speech Recognition", "Lightweight Adapter Tuning for Multilingual Speech Translation", "Losses Can Be Blessings: Routing Self-Supervised Speech Representations Towards Efficient Multilingual and Multitask Speech Processing", "PARP: Prune, Adjust and Re-Prune for Self-Supervised Speech Recognition"], "answer_arxiv_id": ["2105.11905", "2106.01463", "2211.01522", "2106.05933"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_12522"} +{"question": "Could you provide me some research about the information theory based generalization bounds that rely on the mutual information of the input and output of the learning algorithm?", "answer": ["How much does your data exploration overfit? Controlling bias via information usage.", "Information-theoretic analysis of generalization capability of learning algorithms"], "answer_arxiv_id": ["1511.05219", "1705.07809"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_12523"} +{"question": "Could you mention some recent works closely related to Neural Processes (NPs)?", "answer": ["Neural Processes", "Attentive Neural Processes", "Neural Diffusion Processes"], "answer_arxiv_id": ["1807.01622", "1901.05761", "2206.03992"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_12524"} +{"question": "Which papers develop proficient conversations grounded in additional modality using the approach of supervised fine-tuning with synthetically generated datasets?", "answer": ["Visual Instruction Tuning", "Video-LLaVA: Learning United Visual Representation by Alignment Before\n Projection", "Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D\n Understanding, Generation, and Instruction Following"], "answer_arxiv_id": ["2304.08485", "2311.10122", "2309.00615"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_12525"} +{"question": "What research papers offer geometric insights for scene understanding using depth maps?", "answer": ["Specificity-preserving RGB-D Saliency Detection", "Deep RGB-D Saliency Detection with Depth-Sensitive Attention and Automatic Multi-Modal Fusion", "Defocus Blur Detection via Depth Distillation"], "answer_arxiv_id": ["2108.08162", "2103.11832v1", "2007.08113"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_12526"} +{"question": "Are there any works where contrastive approaches are not used in self-supervised methods?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "Whitening for Self-Supervised Representation Learning", "Context Autoencoder for Self-Supervised Representation Learning"], "answer_arxiv_id": ["2111.06377", "2007.06346", "2202.03026"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_12527"} +{"question": "What works belong to tensor field-based methods for solutions to equivariance?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural\n networks for 3D point clouds", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks"], "answer_arxiv_id": ["1802.08219", "2006.10503"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_12528"} +{"question": "Could you give me an example of a work that alleviates memory burden of DP optimizer by sharing the space complexity in two rounds of back-propagation?", "answer": ["Scaling up Differentially Private Deep Learning with Fast Per-Example Gradient Clipping"], "answer_arxiv_id": ["2009.03106"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_12529"} +{"question": "Which works focused on learning representations by predicting the augmentation parameters?", "answer": ["Self-Supervised Learning Through Efference Copies", "Unsupervised Representation Learning by Predicting Image Rotations"], "answer_arxiv_id": ["2210.09224", "1803.07728"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_12530"} +{"question": "Which work provides the conventional approach of Generative Adversarial Networks (GANs)?", "answer": ["Generative Adversarial Nets"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_12531"} +{"question": "Which papers focus on the mechanics of In-context learning (ICL)?", "answer": ["What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "What learning algorithm is in-context learning? Investigations with linear models", "In-context Learning and Induction Heads"], "answer_arxiv_id": ["2208.01066", "2211.15661", "2209.11895"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_12532"} +{"question": "Can you tell me about the research that showed the dimensional collapses could improve generalization performance?", "answer": ["Toward a Geometrical Understanding of Self-supervised Contrastive Learning"], "answer_arxiv_id": ["2205.06926"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_12533"} +{"question": "Could you provide me some works about micro-video or long-burst photography?", "answer": ["Shakes on a Plane: Unsupervised Depth Estimation from Unstabilized\n Photography"], "answer_arxiv_id": ["2212.12324"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_12534"} +{"question": "Which works attempted to integrate adapters to align visual and textual representation within LLMs?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Qwen-VL: A Versatile Vision-Language Model for Understanding,\n Localization, Text Reading, and Beyond"], "answer_arxiv_id": ["2301.12597", "2204.14198", "2304.10592", "2303.16199", "2304.15010", "2304.08485", "2305.06500", "2308.12966"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_12535"} +{"question": "What works use the Hessian information as the quantization sensitivity metrics to assist bit-width assignment?", "answer": ["HAWQ: Hessian AWare Quantization of Neural Networks with Mixed-Precision", "HAWQ-V2: Hessian Aware trace-Weighted Quantization of Neural Networks", "Towards Mixed-Precision Quantization of Neural Networks via Constrained\n Optimization"], "answer_arxiv_id": ["1905.03696", "1911.03852", "2110.06554"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_12536"} +{"question": "Which works integrate an uncertainty- or variance-predicting network into the standard mean-predicting regression networks?", "answer": ["Epistemic Neural Networks", "DEUP: Direct Epistemic Uncertainty Prediction"], "answer_arxiv_id": ["2107.08924", "2102.08501"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_12537"} +{"question": "What works undertake soft truncation or standard deviation adjustment for precision enhancement?", "answer": ["Glow: Generative Flow with Invertible 1x1 Convolutions"], "answer_arxiv_id": ["1807.03039"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12538"} +{"question": "Could you provide me some works about quantization-aware training (QAT) methods?", "answer": ["LSQ+: Improving low-bit quantization through learnable offsets and better initialization", "Learned Step Size Quantization", "Deep Learning with Limited Numerical Precision", "Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference", "Quantizing deep convolutional networks for efficient inference: A whitepaper"], "answer_arxiv_id": ["2004.09576", "1902.08153", "1502.02551", "1712.05877", "1806.08342"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_12539"} +{"question": "What research works proposed multi-task HIL based on expert demonstrations?", "answer": ["Intrinsically Motivated Open-Ended Multi-Task Learning Using Transfer Learning to Discover Task Hierarchy", "One-Shot Hierarchical Imitation Learning of Compound Visuomotor Tasks"], "answer_arxiv_id": ["2102.09854v1", "1810.11043"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_12540"} +{"question": "In what dataset can you find the elucidation of relationships between objects and attributes through graph representations extracted from images?", "answer": ["Visual Genome: Connecting Language and Vision Using Crowdsourced Dense\n Image Annotations"], "answer_arxiv_id": ["1602.07332"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_12541"} +{"question": "Which research papers employed iterative inference setting for object discovery?", "answer": ["Binding via Reconstruction Clustering", "Tagger: Deep Unsupervised Perceptual Grouping", "Neural Expectation Maximization", "Relational Neural Expectation Maximization: Unsupervised Discovery of Objects and their Interactions", "Multi-Object Representation Learning with Iterative Variational Inference"], "answer_arxiv_id": ["1511.06418", "1606.06724", "1708.03498", "1802.10353", "1903.00450"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_12542"} +{"question": "What works propose a PAC-Bayesian framework to provide a learning bound on the expected error by the average loss on observed tasks?", "answer": ["A PAC-Bayesian Bound for Lifelong Learning"], "answer_arxiv_id": ["1311.2838"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_12543"} +{"question": "What papers have investigated Model-Agnostic Meta-Learning (MAML)?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["1703.03400"], "source_meta": {"published_time": "20230108"}, "qid": "AutoScholarQuery_train_12544"} +{"question": "Which paper introduced the Vision-Language Transformer (VLT) in the field of referring image segmentation?", "answer": ["Vision-Language Transformer and Query Generation for Referring\n Segmentation"], "answer_arxiv_id": ["2108.05565"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_12545"} +{"question": "What work introduced diffusion model to generate agent actions?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_12546"} +{"question": "Are there any works on physical point tracking methods that maintain good generalization for dynamic objects?", "answer": ["Particle Video Revisited: Tracking Through Occlusions Using Point\n Trajectories", "TAP-Vid: A Benchmark for Tracking Any Point in a Video", "Tracking Everything Everywhere All at Once", "TAPIR: Tracking Any Point with per-frame Initialization and temporal\n Refinement", "CoTracker: It is Better to Track Together"], "answer_arxiv_id": ["2204.04153", "2211.03726", "2306.05422", "2306.08637", "2307.07635"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_12547"} +{"question": "What researches used the Neural Tangent Kernel for deep neural network analysis?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Gradient Descent Finds Global Minima of Deep Neural Networks", "How Much Over-parameterization Is Sufficient to Learn Deep ReLU Networks?", "On Exact Computation with an Infinitely Wide Neural Net", "Generalization Bounds of Stochastic Gradient Descent for Wide and Deep Neural Networks", "Generalization Properties of NAS under Activation and Skip Connection Search", "Tight Bounds on the Smallest Eigenvalue of the Neural Tangent Kernel for Deep ReLU Networks", "Memorization and Optimization in Deep Neural Networks with Minimum Over-parameterization"], "answer_arxiv_id": ["1806.07572", "1811.03962", "1811.03804", "1911.12360", "1904.11955", "1905.13210", "2209.07238", "2012.11654", "2205.10217"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12548"} +{"question": "Which works focus on trying to find matching subnetworks early in training or at initialization?", "answer": ["SNIP: Single-shot Network Pruning based on Connection Sensitivity", "Picking Winning Tickets Before Training by Preserving Gradient Flow", "Pruning neural networks without any data by iteratively conserving synaptic flow"], "answer_arxiv_id": ["1810.02340", "2002.07376", "2006.05467"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_12549"} +{"question": "What work presents generalization difficulty as the needed information to perform a task in addition to any training data provided to a learning system?", "answer": ["On the Measure of Intelligence"], "answer_arxiv_id": ["1911.01547"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_12550"} +{"question": "Are there any works theoretically studying self-supervised learning in other domains such as language modeling?", "answer": ["Why Do Pretrained Language Models Help in Downstream Tasks? An Analysis of Head and Prompt Tuning", "An Explanation of In-context Learning as Implicit Bayesian Inference", "A Mathematical Exploration of Why Language Models Help Solve Downstream Tasks"], "answer_arxiv_id": ["2106.09226", "2111.02080", "2010.03648"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_12551"} +{"question": "What works used the attention mechanism to generalize group convolution?", "answer": ["Attentive Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["2002.03830"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_12552"} +{"question": "Which works indicate the resource intensiveness and inefficiency of top-performing GPT models?", "answer": ["Energy and Policy Considerations for Deep Learning in NLP"], "answer_arxiv_id": ["1906.02243"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_12553"} +{"question": "Which research introduced a weather forecast dataset with tree-structured meaning representations?", "answer": ["Constrained Decoding for Neural NLG from Compositional Representations\n in Task-Oriented Dialogue"], "answer_arxiv_id": ["1906.07220"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_12554"} +{"question": "What studies have been conducted on deep graph generation using GNNs?", "answer": ["Deep Generation of Heterogeneous Networks"], "answer_arxiv_id": ["2206.12336"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_12555"} +{"question": "Which papers demonstrated the reasoning abilities of transformers across a wide range of tasks?", "answer": ["Language Models are Few-Shot Learners", "GPT-4 Technical Report", "Scaling Instruction-Finetuned Language Models", "PaLM: Scaling Language Modeling with Pathways", "Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "Galactica: A Large Language Model for Science", "LaMDA: Language Models for Dialog Applications"], "answer_arxiv_id": ["2005.14165", "2303.08774", "2210.11416", "2204.02311", "2112.11446", "2211.09085", "2201.08239"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_12556"} +{"question": "Which papers proposed LNL strategies such as robust loss design?", "answer": ["Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels", "Symmetric Cross Entropy for Robust Learning with Noisy Labels"], "answer_arxiv_id": ["1805.07836", "1908.06112"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12557"} +{"question": "What was the initial proposal of Knowledge Distillation (KD) used for?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_12558"} +{"question": "What are some of the notable works in image editing using cross-attention maps?", "answer": ["Zero-shot Image-to-Image Translation", "Prompt-to-Prompt Image Editing with Cross Attention Control"], "answer_arxiv_id": ["2302.03027", "2208.01626"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_12559"} +{"question": "What research papers discuss the GBDA attack method for NLP models?", "answer": ["Gradient-based Adversarial Attacks against Text Transformers"], "answer_arxiv_id": ["2104.13733"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_12560"} +{"question": "Which papers explored training a conditional diffusion model for specific image super-resolution tasks?", "answer": ["Image Super-Resolution via Iterative Refinement", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic\n Models", "Deblurring via Stochastic Refinement", "Towards Authentic Face Restoration with Iterative Diffusion Models and\n Beyond"], "answer_arxiv_id": ["2104.07636", "2104.14951", "2112.02475", "2307.08996"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_12561"} +{"question": "Which works contributed to the development of TOD and PhoCaL datasets?", "answer": ["KeyPose: Multi-View 3D Labeling and Keypoint Estimation for Transparent\n Objects", "PhoCaL: A Multi-Modal Dataset for Category-Level Object Pose Estimation\n with Photometrically Challenging Objects"], "answer_arxiv_id": ["1912.02805", "2205.08811"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_12562"} +{"question": "Can you provide a study that proposed convolution Fourier Neural Operator models to address underfitting issues?", "answer": ["U-FNO - an enhanced Fourier neural operator-based deep-learning model for multiphase flow"], "answer_arxiv_id": ["2109.03697"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_12563"} +{"question": "What work improves the stability of existing 2D-lifting methods by modeling the distribution of multi-view images?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2308.16512"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_12564"} +{"question": "Which papers provide theoretically friendly characterizations of many self-supervised learning settings?", "answer": ["Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods", "Joint Embedding Self-Supervised Learning in the Kernel Regime"], "answer_arxiv_id": ["2205.11508", "2209.14884"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_12565"} +{"question": "Any works about the application of local information in 3D coordinates for scene modeling?", "answer": ["ReLU Fields: The Little Non-linearity That Could", "Local Implicit Grid Representations for 3D Scenes"], "answer_arxiv_id": ["2205.10824", "2003.08981"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_12566"} +{"question": "Are there any works that used contrastive self-supervised learning with transformation operations in image data for anomaly detection?", "answer": ["Deep Anomaly Detection Using Geometric Transformations", "CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances", "SSD: A Unified Framework for Self-Supervised Outlier Detection", "CutPaste: Self-Supervised Learning for Anomaly Detection and Localization", "Self-Supervised Predictive Convolutional Attentive Block for Anomaly Detection"], "answer_arxiv_id": ["1805.10917", "2007.08176", "2103.12051", "2104.04015", "2111.09099"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_12567"} +{"question": "Could you give me examples of studies that utilized diffusion models for adversarial purification?", "answer": ["Diffusion Models for Adversarial Purification", "Guided Diffusion Model for Adversarial Purification from Random Noise", "PointDP: Diffusion-driven Purification against Adversarial Attacks on 3D Point Cloud Recognition"], "answer_arxiv_id": ["2205.07460", "2206.10875", "2208.09801"], "source_meta": {"published_time": "20221101"}, "qid": "AutoScholarQuery_train_12568"} +{"question": "Could you provide me some works about the variants of ViT?", "answer": ["End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers", "Cross-Modal Self-Attention Network for Referring Image Segmentation", "Learning Texture Transformer Network for Image Super-Resolution", "Pre-Trained Image Processing Transformer", "VideoBERT: A Joint Model for Video and Language Representation Learning", "Video Action Transformer Network"], "answer_arxiv_id": ["2005.12872", "2010.04159", "2012.15840", "1904.04745", "2006.04139", "2012.00364", "1904.01766", "1812.02707"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12569"} +{"question": "Which works have attempted to solve the problem of tracking any point in videos?", "answer": ["TAP-Vid: A Benchmark for Tracking Any Point in a Video", "Tracking Everything Everywhere All at Once", "CoTracker: It is Better to Track Together", "TAPIR: Tracking Any Point with per-frame Initialization and temporal\n Refinement"], "answer_arxiv_id": ["2211.03726", "2306.05422", "2307.07635", "2306.08637"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_12570"} +{"question": "What is the work that introduced MixGen, a state-of-the-art augmentation procedure for visual-language representation learning?", "answer": ["MixGen: A New Multi-Modal Data Augmentation"], "answer_arxiv_id": ["2206.08358"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_12571"} +{"question": "What works propose the use of pessimism in offline reinforcement learning to compensate for datasets lacking data coverage?", "answer": ["Is Pessimism Provably Efficient for Offline RL?", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Offline Reinforcement Learning with Realizability and Single-policy Concentrability"], "answer_arxiv_id": ["2012.15085", "2103.12021v2", "2202.04634"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_12572"} +{"question": "Could you name the methods that can achieve high WGA by leveraging the group annotation?", "answer": ["Class-Balanced Loss Based on Effective Number of Samples", "Simple data balancing achieves competitive worst-group-accuracy", "Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations", "On Feature Learning in the Presence of Spurious Correlations", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Robust Representation Learning via Perceptual Similarity Metrics"], "answer_arxiv_id": ["1901.05555", "2110.14503", "2204.02937", "2210.11369", "1911.08731", "2106.06620"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_12573"} +{"question": "Which works proposed finding solution for the objective mismatch problem in model-based RL?", "answer": ["Objective Mismatch in Model-based Reinforcement Learning", "The Value Equivalence Principle for Model-Based Reinforcement Learning"], "answer_arxiv_id": ["2002.04523", "2011.03506"], "source_meta": {"published_time": "20220918"}, "qid": "AutoScholarQuery_train_12574"} +{"question": "Which research utilized a generic transformer-based model for optical flow and language modeling?", "answer": ["Perceiver IO: A General Architecture for Structured Inputs & Outputs"], "answer_arxiv_id": ["2107.14795"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_12575"} +{"question": "What studies demonstrated the use of language codes with a shared encoder/decoder architecture?", "answer": ["Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation"], "answer_arxiv_id": ["1611.04558"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_12576"} +{"question": "Which papers studied learning bounding boxes as the shape representation for input geometry?", "answer": ["Learning Shape Abstractions by Assembling Volumetric Primitives"], "answer_arxiv_id": ["1612.00404"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_12577"} +{"question": "What works utilize autoregressive models in the context of adversarial purification methods?", "answer": ["PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examples"], "answer_arxiv_id": ["1710.10766"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_12578"} +{"question": "Which paper observed that models trained on a poisoned dataset tend to learn very different latent representations for backdoor and clean samples?", "answer": ["Spectral Signatures in Backdoor Attacks"], "answer_arxiv_id": ["1811.00636"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_12579"} +{"question": "Which studies successfully used bootstrapping strategy in offline reinforcement learning?", "answer": ["Bootstrapped Transformer for Offline Reinforcement Learning"], "answer_arxiv_id": ["2206.08569"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_12580"} +{"question": "In which works was the asymptotic spectral density of the single-layer conjugate kernel characterized?", "answer": ["On the Spectrum of Random Features Maps of High Dimensional Data"], "answer_arxiv_id": ["1805.11916"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_12581"} +{"question": "Which studies present methods for projecting the features from a vision encoder to the embedding space of a language model?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "OpenFlamingo: An Open-Source Framework for Training Large Autoregressive\n Vision-Language Models", "Multimodal Few-Shot Learning with Frozen Language Models"], "answer_arxiv_id": ["2301.12597", "2204.14198", "2308.01390", "2106.13884"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_12582"} +{"question": "Which research works address averaging on flag manifolds or producing flags as subspace averages?", "answer": ["Chordal Averaging on Flag Manifolds and Its Applications"], "answer_arxiv_id": ["2303.13501"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_12583"} +{"question": "Which works attempted to understand deep neural networks using image perturbation as an explanation technique?", "answer": ["Understanding Deep Networks via Extremal Perturbations and Smooth Masks"], "answer_arxiv_id": ["1910.08485"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_12584"} +{"question": "In which works do researchers utilize pre-trained depth estimators to address the lack of depth information in monocular 3D object detection?", "answer": ["Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving", "Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving", "Learning Depth-Guided Convolutions for Monocular 3D Object Detection", "MonoDTR: Monocular 3D Object Detection with Depth-Aware Transformer"], "answer_arxiv_id": ["2203.02112", "1812.07179", "1912.04799", "2203.10981"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_12585"} +{"question": "What works proposed solutions to the limitations of current latent diffusion models?", "answer": ["MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation", "SyncDiffusion: Coherent Montage via Synchronized Joint Diffusions", "ScaleCrafter: Tuning-free Higher-Resolution Visual Generation with\n Diffusion Models"], "answer_arxiv_id": ["2302.08113", "2306.05178", "2310.07702"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_12586"} +{"question": "Could you provide me some studies that proposed to carefully relax the attacker’s capabilities in order to achieve higher utility from private predictions?", "answer": ["Privacy-preserving Prediction", "Bayesian Differential Privacy for Machine Learning", "Provable Membership Inference Privacy"], "answer_arxiv_id": ["1803.10266", "1901.09697", "2211.06582"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_12587"} +{"question": "What research works proposed methods to condition DDPM?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "Improved Denoising Diffusion Probabilistic Models", "Few-Shot Diffusion Models", "Label-Efficient Semantic Segmentation with Diffusion Models", "GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation", "Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2108.02938", "2011.13456", "2102.09672", "2205.15463", "2112.03126", "2203.02923", "2203.17003"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_12588"} +{"question": "What research works explore the role of augmentation in contrastive learning?", "answer": ["What Makes for Good Views for Contrastive Learning?", "What Should Not Be Contrastive in Contrastive Learning"], "answer_arxiv_id": ["2005.10243", "2008.05659"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_12589"} +{"question": "What works have shown the effectiveness of fusing LiDAR points with features from cameras in 3D object detection?", "answer": ["PointPainting: Sequential Fusion for 3D Object Detection", "MVX-Net: Multimodal VoxelNet for 3D Object Detection", "EPNet: Enhancing Point Features with Image Semantics for 3D Object\n Detection", "FusionPainting: Multimodal Fusion with Adaptive Attention for 3D Object\n Detection"], "answer_arxiv_id": ["1911.10150", "1904.01649", "2007.08856", "2106.12449"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_12590"} +{"question": "Which study used an autoregressive LSTM for graph generation in the context of Bayesian structure learning?", "answer": ["Variational Causal Networks: Approximate Bayesian Inference over Causal Structures"], "answer_arxiv_id": ["2106.07635"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_12591"} +{"question": "What work has used a pretrained text encoder in DFKT?", "answer": ["NAYER: Noisy Layer Data Generation for Efficient and Effective Data-free Knowledge Distillation"], "answer_arxiv_id": ["2310.00258v2"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_12592"} +{"question": "What studies are necessary to achieve single-view reconstruction?", "answer": ["NeRFInvertor: High Fidelity NeRF-GAN Inversion for Single-shot Real\n Image Animation", "Self-Supervised Geometry-Aware Encoder for Style-Based 3D GAN Inversion", "Make Encoder Great Again in 3D GAN Inversion through Geometry and\n Occlusion-Aware Encoding"], "answer_arxiv_id": ["2211.17235", "2212.07409", "2303.12326"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_12593"} +{"question": "Can you list some papers that solve the problems of physical systems' dynamics using a sparse regression model?", "answer": ["Data-driven discovery of coordinates and governing equations"], "answer_arxiv_id": ["1904.02107"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_12594"} +{"question": "Could you provide me some works that involve the use of Graph Neural Networks (GNNs) to model molecules as graphs?", "answer": ["Neural Message Passing for Quantum Chemistry", "Graph Contrastive Learning with Augmentations", "Strategies for Pre-training Graph Neural Networks", "Hyperbolic Graph Neural Networks"], "answer_arxiv_id": ["1704.01212", "2010.13902", "1905.12265", "1910.12892"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_12595"} +{"question": "Could you provide me some studies about adversarial training methods?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_12596"} +{"question": "In which research is the hyperparameter ω dynamically scheduled in adversarial contrastive loss?", "answer": ["Rethinking the Effect of Data Augmentation in Adversarial Contrastive Learning"], "answer_arxiv_id": ["2303.01289"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_12597"} +{"question": "What studies have incorporated viewing directions into scene representations to achieve view-dependent effects in Neural View Synthesis?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields"], "answer_arxiv_id": ["2003.08934", "2103.13415"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_12598"} +{"question": "Which independent work also obtained the optimal rate under general function approximation via the LP framework, similar to the study at hand?", "answer": ["Optimal Conservative Offline RL with General Function Approximation via Augmented Lagrangian"], "answer_arxiv_id": ["2211.00716"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_12599"} +{"question": "What studies have optimized the attention map for a specific point in a source image?", "answer": ["Unsupervised Semantic Correspondence Using Stable Diffusion"], "answer_arxiv_id": ["2305.15581"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_12600"} +{"question": "What papers cover the topic of conditional diffusion models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2112.10752", "2302.08453"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_12601"} +{"question": "Which works try to improve the performance of convolutional sequence models through the use of gating units?", "answer": ["Language Modeling with Gated Convolutional Networks"], "answer_arxiv_id": ["1612.08083"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_12602"} +{"question": "Could you provide me some studies that utilized Faster R-CNN based object detectors and resulted in high inference latency?", "answer": ["Bottom-Up and Top-Down Attention for Image Captioning and Visual\n Question Answering", "VinVL: Revisiting Visual Representations in Vision-Language Models", "ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision"], "answer_arxiv_id": ["1707.07998", "2101.00529", "2102.03334"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_12603"} +{"question": "What research is focused on the reliability of teacher networks during adversarial distillation?", "answer": ["Reliable Adversarial Distillation with Unreliable Teachers"], "answer_arxiv_id": ["2106.04928"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_12604"} +{"question": "Which papers describe the utilization of Vision Transformers for feature extraction?", "answer": ["Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers"], "answer_arxiv_id": ["2012.15840"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_12605"} +{"question": "Which works pioneered the advancements in generative models for creating visually realistic images?", "answer": ["A Style-Based Generator Architecture for Generative Adversarial Networks", "Analyzing and Improving the Image Quality of StyleGAN", "Alias-Free Generative Adversarial Networks", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["1812.04948", "1912.04958", "2106.12423", "2112.10752", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_12606"} +{"question": "Which studies demonstrated that external knowledge is beneficial for detecting factual hallucinations?", "answer": ["A Survey on Automated Fact-Checking", "When Not to Trust Language Models: Investigating Effectiveness of\n Parametric and Non-Parametric Memories"], "answer_arxiv_id": ["2108.11896", "2212.10511"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_12607"} +{"question": "Any works about data augmentation serving as a form of regularization?", "answer": ["mixup: Beyond Empirical Risk Minimization", "RandAugment: Practical automated data augmentation with a reduced search space"], "answer_arxiv_id": ["1710.09412", "1909.13719"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12608"} +{"question": "In which papers was the spectrum of the input-output Jacobian for MLPs studied?", "answer": ["The Emergence of Spectral Universality in Deep Networks", "Dynamical Isometry and a Mean Field Theory of CNNs: How to Train 10,000-Layer Vanilla Convolutional Neural Networks"], "answer_arxiv_id": ["1802.09979", "1806.05393"], "source_meta": {"published_time": "20211123"}, "qid": "AutoScholarQuery_train_12609"} +{"question": "Could you tell me about the work that proposed continuous generative neural networks (CGNNs) and provided the sufficient condition for their global injectivity?", "answer": ["Continuous Generative Neural Networks"], "answer_arxiv_id": ["2205.14627"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_12610"} +{"question": "Which works focus on discovering new design principles for convolutional and graph neural networks?", "answer": ["Designing Network Design Spaces", "Design Space for Graph Neural Networks"], "answer_arxiv_id": ["2003.13678", "2011.08843"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_12611"} +{"question": "What research works have proposed the most direct approach to upsampling low-resolution data by learning a low- to high-resolution mapping via paired data?", "answer": ["Statistics of extreme events in coarse-scale climate simulations via machine learning correction operators trained on nudged datasets"], "answer_arxiv_id": ["2304.02117"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_12612"} +{"question": "Which studies discuss alternate approaches to reinforcement learning from human feedback?", "answer": ["Chain of Hindsight Aligns Language Models with Feedback", "SLiC-HF: Sequence Likelihood Calibration with Human Feedback", "Direct Preference Optimization: Your Language Model is Secretly a Reward\n Model"], "answer_arxiv_id": ["2302.02676", "2305.10425", "2305.18290"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_12613"} +{"question": "What studies have worked on improving the rendering speed of Neural Radiance Fields?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "TensoRF: Tensorial Radiance Fields", "3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2112.05131", "2201.05989", "2203.09517", "2308.04079"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_12614"} +{"question": "Which works introduced the concept of fine-tuning a visual encoder to map images into the embedding space of a LLM?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models"], "answer_arxiv_id": ["2106.13884"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_12615"} +{"question": "Which research papers have deduced the generalization of neural ODEs to Riemannian manifolds?", "answer": ["Neural Manifold Ordinary Differential Equations", "Equivariant Manifold Flows", "Neural Ordinary Differential Equations on Manifolds", "Riemannian Continuous Normalizing Flows"], "answer_arxiv_id": ["2006.10254", "2107.08596", "2006.06663", "2006.10605"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_12616"} +{"question": "What studies have widely applied the permutation equivariance property?", "answer": ["Intriguing Properties of Vision Transformers", "Point Transformer", "The Sensory Neuron as a Transformer: Permutation-Invariant Neural\n Networks for Reinforcement Learning"], "answer_arxiv_id": ["2105.10497", "2012.09164", "2109.02869"], "source_meta": {"published_time": "20230416"}, "qid": "AutoScholarQuery_train_12617"} +{"question": "Which research papers proposed platforms for embodied agents performing household activities in indoor virtual environments?", "answer": ["VirtualHome: Simulating Household Activities via Programs", "ThreeDWorld: A Platform for Interactive Multi-Modal Physical Simulation", "SAPIEN: A SimulAted Part-based Interactive ENvironment"], "answer_arxiv_id": ["1806.07011", "2007.04954", "2003.08515"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12618"} +{"question": "Could you provide me with some references that have inspired the Circuits-style mechanistic interpretability?", "answer": ["In-context Learning and Induction Heads"], "answer_arxiv_id": ["2209.11895"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_12619"} +{"question": "Could you tell me works that have considered differentiable rendering (DR) for reconstructing explicit geometric primitives, meshes, point clouds?", "answer": ["Neural Volumes: Learning Dynamic Renderable Volumes from Images", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning", "Accelerating 3D Deep Learning with PyTorch3D", "Modular Primitives for High-Performance Differentiable Rendering", "Differentiable Surface Rendering via Non-Differentiable Sampling", "Dressi: A Hardware-Agnostic Differentiable Renderer with Reactive Shader\n Packing and Soft Rasterization", "Differentiable Surface Splatting for Point-based Geometry Processing", "ADOP: Approximate Differentiable One-Pixel Point Rendering"], "answer_arxiv_id": ["1906.07751", "2003.08934", "2106.10689", "1904.01786", "2007.08501", "2011.03277", "2108.04886", "2204.01386", "1906.04173v3", "2110.06635"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_12620"} +{"question": "Are there any papers about dynamic reconfiguration, growth and pruning for task-optimized architectures?", "answer": ["Network Morphism", "AutoGrow: Automatic Layer Growing in Deep Convolutional Networks", "Growing Efficient Deep Networks by Structured Continuous Sparsification"], "answer_arxiv_id": ["1603.01670", "1906.02909", "2007.15353"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_12621"} +{"question": "Could you provide some works that introduced voxel grids or planar representations to improve the training efficiency of dynamic NeRFs?", "answer": ["DeVRF: Fast Deformable Voxel Radiance Fields for Dynamic Scenes", "Fast Dynamic Radiance Fields with Time-Aware Neural Voxels", "NeRFPlayer: A Streamable Dynamic Scene Representation with Decomposed\n Neural Radiance Fields", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "HexPlane: A Fast Representation for Dynamic Scenes"], "answer_arxiv_id": ["2205.15723", "2205.15285", "2210.15947", "2301.10241", "2301.09632"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_12622"} +{"question": "Could you provide me some works that connect adversarial robustness and the local Lipschitz constant of the network?", "answer": ["Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation"], "answer_arxiv_id": ["1705.08475"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_12623"} +{"question": "Which studies address the sign ambiguity of eigenvectors by using data augmentation?", "answer": ["Benchmarking Graph Neural Networks", "Graph Neural Networks with Learnable Structural and Positional Representations", "Rethinking Graph Transformers with Spectral Attention", "GraphiT: Encoding Graph Structure in Transformers", "Pure Transformers are Powerful Graph Learners", "A Generalization of ViT/MLP-Mixer to Graphs", "Attending to Graph Transformers"], "answer_arxiv_id": ["2003.00982", "2110.07875", "2106.03893", "2106.05667", "2207.02505", "2212.13350", "2302.04181"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_12624"} +{"question": "Which paper introduces Spatial Dropout that drops entire feature maps in a ConvNet?", "answer": ["Efficient Object Localization Using Convolutional Networks"], "answer_arxiv_id": ["1411.4280"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12625"} +{"question": "What products led to an even higher level of privacy concerns due to the introduction of billion-scale large-language models (LLMs)?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_12626"} +{"question": "Which papers discussed the disentangling objects from images in object-centric learning?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational Inference", "Object-Centric Learning with Slot Attention", "Unsupervised Layered Image Decomposition into Object Prototypes"], "answer_arxiv_id": ["1901.11390", "1903.00450", "2006.15055", "2104.14575"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_12627"} +{"question": "Which papers studied the use of feature decorrelation objectives in unsupervised visual representation learning?", "answer": ["Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "On Feature Decorrelation in Self-Supervised Learning", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning"], "answer_arxiv_id": ["2103.03230", "2105.00470", "2105.04906"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_12628"} +{"question": "What is the study that defines the diffusion process again in image space, and appends the triplane representation to the usual diffusion architecture for 3D view conditioning?", "answer": ["RenderDiffusion: Image Diffusion for 3D Reconstruction, Inpainting and\n Generation"], "answer_arxiv_id": ["2211.09869"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_12629"} +{"question": "Are there any research papers that focus on unsupervised skill discovery?", "answer": ["Variational Intrinsic Control Revisited", "Diversity is All You Need: Learning Skills without a Reward Function", "Explore, Discover and Learn: Unsupervised Discovery of State-Covering Skills", "Dynamics-Aware Unsupervised Discovery of Skills"], "answer_arxiv_id": ["2010.03281", "1802.06070", "2002.03647", "1907.01657"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_12630"} +{"question": "What works propose using dynamically expandable networks for Continual Learning?", "answer": ["Progressive Neural Networks", "Learning without Forgetting"], "answer_arxiv_id": ["1606.04671", "1606.09282"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_12631"} +{"question": "Which works aim at improving knowledge distillation for multi-exit neural networks?", "answer": ["Multi-Scale Dense Networks for Resource Efficient Image Classification", "BranchyNet: Fast Inference via Early Exiting from Deep Neural Networks", "Early Exit or Not: Resource-Efficient Blind Quality Enhancement for Compressed Images", "BERT Loses Patience: Fast and Robust Inference with Early Exit"], "answer_arxiv_id": ["1703.09844", "1709.01686", "2006.16581", "2006.04152"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_12632"} +{"question": "Which studies utilize bootstrapping method to estimate uncertainty in Reinforcement Learning?", "answer": ["Deep Exploration via Bootstrapped DQN"], "answer_arxiv_id": ["1602.04621"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12633"} +{"question": "What papers discuss learning a heatmap for each landmark via image reconstruction?", "answer": ["Unsupervised Learning of Object Landmarks through Conditional Image\n Generation", "Unsupervised Discovery of Object Landmarks as Structural Representations"], "answer_arxiv_id": ["1806.07823", "1804.04412"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_12634"} +{"question": "Could you provide me some studies about diffusion based multi-modal inpainting?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images", "Blended Latent Diffusion", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "SmartBrush: Text and Shape Guided Object Inpainting with Diffusion Model", "Uni-paint: A Unified Framework for Multimodal Image Inpainting with\n Pretrained Diffusion Model", "GLIGEN: Open-Set Grounded Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis"], "answer_arxiv_id": ["2111.14818", "2206.02779", "2112.10741", "2212.05034", "2310.07222", "2301.07093", "2112.10752", "2307.01952"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_12635"} +{"question": "What research discusses hard-constrained RL with schemes designed for specific applications like robotics manipulation, resource allocation problems and others?", "answer": ["OptLayer - Practical Constrained Optimization for Deep Reinforcement Learning in the Real World", "OptNet: Differentiable Optimization as a Layer in Neural Networks", "Differentiable Convex Optimization Layers", "Alternating Differentiation for Optimization Layers", "Resource Constrained Deep Reinforcement Learning"], "answer_arxiv_id": ["1709.07643", "1703.00443", "1910.12430", "2210.01802", "1812.00600"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_12636"} +{"question": "What works proposed ways to reduce the action space during exploration in RL for Large Action Spaces?", "answer": ["Graph Constrained Reinforcement Learning for Natural Language Action Spaces", "Learn What Not to Learn: Action Elimination with Deep Reinforcement Learning", "Aligning to Social Norms and Values in Interactive Narratives"], "answer_arxiv_id": ["2001.08837", "1809.02121", "2205.01975"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_12637"} +{"question": "Which studies propose the pipeline of projecting an image to a joint space of LLM?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "Iterative Prompt Learning for Unsupervised Backlit Image Enhancement"], "answer_arxiv_id": ["2301.12597", "2304.08485", "2303.17569"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_12638"} +{"question": "What is the work where The Segment Anything Model (SAM) is introduced?", "answer": ["Segment Anything"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_12639"} +{"question": "Which study introduced Flash Attention that reduces the GPU memory required during model training and inference?", "answer": ["FlashAttention: Fast and Memory-Efficient Exact Attention with\n IO-Awareness"], "answer_arxiv_id": ["2205.14135"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_12640"} +{"question": "What work introduced the Classifier-Free Guidance approach that enables conditioning in diffusion models?", "answer": ["Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2207.12598"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_12641"} +{"question": "What paper analyzes the computational hardness of non-convex non-concave optimization?", "answer": ["The Complexity of Constrained Min-Max Optimization"], "answer_arxiv_id": ["2009.09623"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_12642"} +{"question": "Which researches are based on 'two-stage' detectors like Fast-RCNN for action localisation models?", "answer": ["Actor-Context-Actor Relation Network for Spatio-Temporal Action\n Localization", "MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient\n Long-Term Video Recognition", "Asynchronous Interaction Aggregation for Action Detection", "Beyond Transfer Learning: Co-finetuning for Action Localisation", "SlowFast Networks for Video Recognition"], "answer_arxiv_id": ["2006.07976", "2201.08383", "2004.07485", "2207.03807", "1812.03982"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_12643"} +{"question": "Could you provide me some studies about training-free fast samplers at inference?", "answer": ["On Fast Sampling of Diffusion Probabilistic Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "Fast Sampling of Diffusion Models with Exponential Integrator", "GENIE: Higher-Order Denoising Diffusion Solvers", "Gotta Go Fast When Generating Data with Score-Based Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2106.00132", "2206.00927", "2204.13902", "2210.05475", "2105.14080", "2202.09778"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_12644"} +{"question": "What are the relevant works addressing confidence thresholding methods?", "answer": ["FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "Dash: Semi-Supervised Learning with Dynamic Thresholding"], "answer_arxiv_id": ["2001.07685v2", "2110.08263", "2109.00650"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12645"} +{"question": "Can you tell me which research first studied the idea of learning disentangled representation?", "answer": ["InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets"], "answer_arxiv_id": ["1606.03657"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_12646"} +{"question": "What papers focus on the application of the DFAC framework and ResZ in distributional MARL?", "answer": ["DFAC Framework: Factorizing the Value Function via Quantile Mixture for Multi-Agent Distributional Q-Learning"], "answer_arxiv_id": ["2102.07936"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_12647"} +{"question": "What works are about post-hoc uncertainty quantification methods?", "answer": ["Uncertainty-guided Source-free Domain Adaptation"], "answer_arxiv_id": ["2208.07591v1"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_12648"} +{"question": "What work is noted for the technique of increasing resolution in the final stage of pretraining?", "answer": ["Fixing the train-test resolution discrepancy"], "answer_arxiv_id": ["1906.06423"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_12649"} +{"question": "Which papers suggest sample selection approaches to enhance the robustness of DNNs against label noise? ", "answer": ["MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels", "Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels", "Combating Noisy Labels by Agreement: A Joint Training Method with Co-Regularization", "A Topological Filter for Learning with Label Noise"], "answer_arxiv_id": ["1712.05055", "1804.06872", "2003.02752", "2012.04835"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_12650"} +{"question": "What works directly apply vanilla conformal prediction to time series without theoretical guarantees or requiring weaker notions of exchangeability?", "answer": ["Copula Conformal Prediction for Multi-step Time Series Forecasting"], "answer_arxiv_id": ["2212.03281"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_12651"} +{"question": "Which dataset is commonly used for deductive logical reasoning?", "answer": ["ProofWriter: Generating Implications, Proofs, and Abductive Statements\n over Natural Language"], "answer_arxiv_id": ["2012.13048"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_12652"} +{"question": "Which works highlighted that left-to-right causal language modeling is a suboptimal objective?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension"], "answer_arxiv_id": ["1910.10683", "1910.13461"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_12653"} +{"question": "Who combined heat dissipation and additive noise and formalized it as a diffusion process with anisotropic noise?", "answer": ["Blurring Diffusion Models"], "answer_arxiv_id": ["2209.05557"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_12654"} +{"question": "Can you identify the works that discovered the bias in raw likelihood score from deep generative models and proposed to correct it?", "answer": ["Likelihood Ratios for Out-of-Distribution Detection", "A Simple Fix to Mahalanobis Distance for Improving Near-OOD Detection", "Out-of-Distribution Detection and Selective Generation for Conditional Language Models"], "answer_arxiv_id": ["1906.02845", "2106.09022", "2209.15558"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_12655"} +{"question": "Which paper applied a diffusion model in the latent space of a variational autoencoder to improve computational efficiency?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_12656"} +{"question": "Which works are about traditional video understanding methods with manually curated video action recognition datasets?", "answer": ["Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset", "HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["1705.07750", "1906.03327", "2110.07058"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_12657"} +{"question": "Which papers provide method to construct high-quality social dialogues using SFT?", "answer": ["SODA: Million-scale Dialogue Distillation with Social Commonsense\n Contextualization"], "answer_arxiv_id": ["2212.10465"], "source_meta": {"published_time": "20240207"}, "qid": "AutoScholarQuery_train_12658"} +{"question": "Which papers introduced the visual storytelling problem and work on creating a text story from a sequence of images?", "answer": ["Visual Storytelling", "Plot and Rework: Modeling Storylines for Visual Storytelling"], "answer_arxiv_id": ["1604.03968", "2105.06950"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_12659"} +{"question": "Which paper proposed an approach for learning a common embedding space between the subject-object pairs and predicates?", "answer": ["Prototype-based Embedding Network for Scene Graph Generation"], "answer_arxiv_id": ["2303.07096"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_12660"} +{"question": "Any works about embedding learnable neural components within synchronous grammars?", "answer": ["Sequence-to-Sequence Learning with Latent Neural Grammars", "Finding Dataset Shortcuts with Grammar Induction", "Hierarchical Phrase-based Sequence-to-Sequence Learning"], "answer_arxiv_id": ["2109.01135v7", "2210.11560", "2211.07906"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12661"} +{"question": "What are the research papers that use metric learning to distinguish various degradations in Blind-SR methods?", "answer": ["Unsupervised Degradation Representation Learning for Blind Super-Resolution", "Metric Learning based Interactive Modulation for Real-World Super-Resolution"], "answer_arxiv_id": ["2104.00416", "2205.05065"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_12662"} +{"question": "Which early studies estimated the 3D human pose from monocular images without explicitly using the corresponding 2D pose?", "answer": ["Structured Prediction of 3D Human Pose with Deep Neural Networks", "Human Pose Estimation in Space and Time using 3D CNN", "Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose", "Integral Human Pose Regression"], "answer_arxiv_id": ["1605.05180", "1609.00036", "1611.07828", "1711.08229"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_12663"} +{"question": "What previous analysis studies on deep saliency models have been conducted?", "answer": ["Understanding and Visualizing Deep Visual Saliency Models"], "answer_arxiv_id": ["1903.02501"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_12664"} +{"question": "Which work was the first to address the imitation learning problem in mean field games?", "answer": ["Learning Deep Mean Field Games for Modeling Large Population Behavior"], "answer_arxiv_id": ["1711.03156"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_12665"} +{"question": "Can you tell which research introduced a novel sampling method based on advances in scheduled optimal transport sampling to create robust learning data?", "answer": ["Flow Matching on General Geometries"], "answer_arxiv_id": ["2302.03660"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_12666"} +{"question": "What research has been done in the field of causality-based domain generalization?", "answer": ["Invariant Risk Minimization", "Heterogeneous Risk Minimization", "Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization", "Invariant Models for Causal Transfer Learning", "Variance Minimization in the Wasserstein Space for Invariant Causal Prediction", "Preventing Failures Due to Dataset Shift: Learning Predictive Models That Transport"], "answer_arxiv_id": ["1907.02893", "2105.03818", "2106.06607", "1507.05333", "2110.07064v2", "1812.04597"], "source_meta": {"published_time": "20210705"}, "qid": "AutoScholarQuery_train_12667"} +{"question": "Can you give examples of research papers that discuss the presumption of a semi-honest cloud-based machine learning service provider?", "answer": ["CryptoGCN: Fast and Scalable Homomorphically Encrypted Graph Convolutional Network Inference"], "answer_arxiv_id": ["2209.11904"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_12668"} +{"question": "Which works proposed to adopt a modular approach to interpretability in RL?", "answer": ["A Survey on Interpretable Reinforcement Learning"], "answer_arxiv_id": ["2112.13112"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_12669"} +{"question": "Which work proposes the Latent Diffusion Model (LDM) for high-resolution image synthesis?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_12670"} +{"question": "Which research focuses on transformation invariance constraints to improve explainability robustness?", "answer": ["Self-Interpretable Model with Transformation Equivariant Interpretation"], "answer_arxiv_id": ["2111.04927"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_12671"} +{"question": "Any works extended NeRF to work with HDR image data?", "answer": ["NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw\n Images", "HDR-NeRF: High Dynamic Range Neural Radiance Fields"], "answer_arxiv_id": ["2111.13679", "2111.14451"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_12672"} +{"question": "Which works involved the study of fundamental trade-offs between adversarial robustness and standard accuracy in machine learning?", "answer": ["Robustness May Be at Odds with Accuracy", "Adversarial Training Can Hurt Generalization", "Is Robustness the Cost of Accuracy? – A Comprehensive Study on the Robustness of 18 Deep Image Classification Models", "Lower Bounds on Adversarial Robustness from Optimal Transport", "More Data Can Expand the Generalization Gap Between Adversarially Robust and Standard Models", "Theoretically Principled Trade-off between Robustness and Accuracy", "Understanding and Mitigating the Tradeoff Between Robustness and Accuracy", "Attacks Which Do Not Kill Training Make Adversarial Learning Stronger", "A Closer Look at Accuracy vs. Robustness"], "answer_arxiv_id": ["1805.12152", "1906.06032", "1808.01688", "1909.12272", "2002.04725", "1901.08573", "2002.10716", "2002.11242", "2003.02460"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_12673"} +{"question": "Could you mention research developing end-to-end encoder-decoder models attending to both image and retrieved caption embeddings?", "answer": ["Retrieval-augmented Image Captioning"], "answer_arxiv_id": ["2302.08268"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_12674"} +{"question": "What papers focus on finding sparse and trainable networks at initialization following the Lottery Ticket Hypothesis?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks"], "answer_arxiv_id": ["1803.03635"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12675"} +{"question": "Which research works excavate on large convolutional kernels?", "answer": ["Large Kernel Matters -- Improve Semantic Segmentation by Global\n Convolutional Network", "Involution: Inverting the Inherence of Convolution for Visual\n Recognition"], "answer_arxiv_id": ["1703.02719", "2103.06255"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_12676"} +{"question": "Which papers discussed the use of deep learning architectures such as auto-encoder, GAN, VAE, and auto-regressive Transformers for image inpainting?", "answer": ["Context Encoders: Feature Learning by Inpainting", "Rethinking Image Inpainting via a Mutual Encoder-Decoder with Feature Equalizations", "StructureFlow: Image Inpainting via Structure-aware Appearance Flow", "Learning Pyramid-Context Encoder Network for High-Quality Image Inpainting", "Pluralistic Image Completion", "Generating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE", "High-Fidelity Pluralistic Image Completion with Transformers", "Diverse Image Inpainting with Bidirectional and Autoregressive Transformers"], "answer_arxiv_id": ["1604.07379", "2007.06929", "1908.03852", "1904.07475", "1903.04227", "2103.10022", "2103.14031", "2104.12335"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_12677"} +{"question": "What work introduces a method to improve subject fidelity and generalization in image generation?", "answer": ["Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning"], "answer_arxiv_id": ["2307.11410"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_12678"} +{"question": "What works have done research on query-based methods in object detection?", "answer": ["End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2005.12872", "2010.04159"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_12679"} +{"question": "Any studies using generative models like invertible networks for the video prediction task?", "answer": ["Stochastic Image-to-Video Synthesis using cINNs"], "answer_arxiv_id": ["2105.04551"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_12680"} +{"question": "Which works developed VQ-VAE based methods for dance generation?", "answer": ["Bailando: 3D Dance Generation by Actor-Critic GPT with Choreographic\n Memory", "GTN-Bailando: Genre Consistent Long-Term 3D Dance Generation based on\n Pre-trained Genre Token Network"], "answer_arxiv_id": ["2203.13055", "2304.12704"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_12681"} +{"question": "Could you provide me some studies about post-processing algorithms that modify the output of the original method to fit fairness constraints?", "answer": ["Equality of Opportunity in Supervised Learning", "A Confidence-Based Approach for Balancing Fairness and Accuracy", "On Fairness and Calibration", "Algorithmic decision making and the cost of fairness", "Multiaccuracy: Black-Box Post-Processing for Fairness in Classification", "HappyMap: A Generalized Multicalibration Method"], "answer_arxiv_id": ["1610.02413", "1601.05764", "1709.02012", "1701.08230", "1805.12317", "2303.04379"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_12682"} +{"question": "What are the research works that have attempted to distill an interpretable policy from a pre-trained neural network policy?", "answer": ["Verifiable Reinforcement Learning via Policy Extraction", "Programmatically Interpretable Reinforcement Learning"], "answer_arxiv_id": ["1805.08328", "1804.02477"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_12683"} +{"question": "What papers discuss the issue of the Shampoo method for rectangular matrices and have proposed workarounds for it?", "answer": ["Scalable Second Order Optimization for Deep Learning"], "answer_arxiv_id": ["2002.09018"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_12684"} +{"question": "In retrosynthesis predictive research, who explored the semi-template-based method by attaching the leaving group?", "answer": ["Learning Graph Models for Retrosynthesis Prediction"], "answer_arxiv_id": ["2006.07038"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12685"} +{"question": "What papers proposed using intrinsic model uncertainty metrics for hallucination detection in LLMs?", "answer": ["The Internal State of an LLM Knows When It's Lying", "On Hallucination and Predictive Uncertainty in Conditional Language\n Generation", "Uncertainty Estimation in Autoregressive Structured Prediction"], "answer_arxiv_id": ["2304.13734", "2103.15025", "2002.07650"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_12686"} +{"question": "What methods utilize the concept of behavior policy in their regularization methods in offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "2106.06860"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_12687"} +{"question": "What papers offer thorough survey on the MPMAB topic?", "answer": ["A survey on multi-player bandits"], "answer_arxiv_id": ["2211.16275"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12688"} +{"question": "Which works have been proposed for generating pseudo-ground truth labels in motion capture datasets?", "answer": ["Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the\n Loop", "NeuralAnnot: Neural Annotator for 3D Human Mesh Training Sets", "On Self-Contact and Human Pose", "Synthetic Training for Accurate 3D Human Pose and Shape Estimation in the Wild"], "answer_arxiv_id": ["1909.12828", "2011.11232", "2104.03176", "2009.10013v2"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_12689"} +{"question": "What research show instances of label distribution learning applications in tasks like facial age estimation, head-pose estimation, and crowd counting?", "answer": ["Deep Differentiable Random Forests for Age Estimation"], "answer_arxiv_id": ["1907.10665"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_12690"} +{"question": "What papers first introduced the concept of 'knowledge neuron' and contributed to the development and understanding of Neuron Attribution?", "answer": ["Knowledge Neurons in Pretrained Transformers", "Locating and Editing Factual Associations in GPT", "Does Localization Inform Editing? Surprising Differences in\n Causality-Based Localization vs. Knowledge Editing in Language Models"], "answer_arxiv_id": ["2104.08696", "2202.05262", "2301.04213"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_12691"} +{"question": "What are some works on two-stage methods in Human-Object Interaction (HOI) detection that focus on exploring human pose information?", "answer": ["Pose-aware Multi-level Feature Network for Human Object Interaction Detection", "Detailed 2D-3D Joint Representation for Human-Object Interaction", "No-Frills Human-Object Interaction Detection: Factorization, Layout Encodings, and Training Techniques"], "answer_arxiv_id": ["1909.08453", "2004.08154", "1811.05967"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_12692"} +{"question": "Could you provide me some works about the application of INRs in image and video compression?", "answer": ["NeRV: Neural Representations for Videos", "E-NeRV: Expedite Neural Video Representation with Disentangled Spatial-Temporal Context", "PS-NeRV: Patch-wise Stylized Neural Representations for Videos", "FFNeRV: Flow-Guided Frame-Wise Neural Representations for Videos", "HNeRV: A Hybrid Neural Representation for Videos"], "answer_arxiv_id": ["2110.13903", "2207.08132", "2208.03742", "2212.12294", "2304.02633"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_12693"} +{"question": "Which datasets are available that analyze human motion through interaction annotations such as human-scene contact?", "answer": ["Populating 3D Scenes by Learning Human-Scene Interaction", "Capturing and Inferring Dense Full-Body Human-Scene Contact"], "answer_arxiv_id": ["2012.11581", "2206.09553"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_12694"} +{"question": "What papers discusses the usage of geometric (or Clifford) algebra in quantum physics?", "answer": ["Clifford Algebras and Spinors"], "answer_arxiv_id": ["1106.3197v2"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_12695"} +{"question": "What works adopted the cross-frame attention on each frame on anchor frame to generate coherent multiple frames?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Zero-Shot Video Editing Using Off-The-Shelf Image Diffusion Models", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "Pix2Video: Video Editing using Image Diffusion", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing"], "answer_arxiv_id": ["2212.11565", "2303.17599", "2303.13439", "2303.12688", "2303.09535"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_12696"} +{"question": "Which papers studied the forms of mutual information-based objectives to learn state-covering skills?", "answer": ["Explore, Discover and Learn: Unsupervised Discovery of State-Covering Skills", "Skew-Fit: State-Covering Self-Supervised Reinforcement Learning"], "answer_arxiv_id": ["2002.03647", "1903.03698"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12697"} +{"question": "What studies utilized neural network-based methods to learn DAG distributions and point estimates for nonlinear model parameters?", "answer": ["Differentiable DAG Sampling", "Deep End-to-end Causal Inference"], "answer_arxiv_id": ["2203.08509", "2202.02195"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_12698"} +{"question": "Could you provide me papers on point cloud forecasting that had applied 3D convolutions?", "answer": ["Self-supervised Point Cloud Prediction Using 3D Spatio-temporal\n Convolutional Networks"], "answer_arxiv_id": ["2110.04076"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_12699"} +{"question": "What works discuss knowledge distillation in multi-modal settings particularly for object detection?", "answer": ["MonoDistill: Learning Spatial Features for Monocular 3D Object Detection", "LIGA-Stereo: Learning LiDAR Geometry Aware Representations for Stereo-based 3D Detector", "Unifying Voxel-based Representation with Transformer for 3D Object Detection"], "answer_arxiv_id": ["2201.10830", "2108.08258", "2206.00630"], "source_meta": {"published_time": "20221117"}, "qid": "AutoScholarQuery_train_12700"} +{"question": "Could you provide me resources on zero-shot Composed Image Retrieval (CIR)?", "answer": ["Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image\n Retrieval", "Zero-Shot Composed Image Retrieval with Textual Inversion", "\"This is my unicorn, Fluffy\": Personalizing frozen vision-language\n representations", "CompoDiff: Versatile Composed Image Retrieval With Latent Diffusion"], "answer_arxiv_id": ["2302.03084", "2303.15247", "2204.01694", "2303.11916"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_12701"} +{"question": "What studies proposed SE(3)-equivariant neural network architectures using SE(3)-invariant feature encoding?", "answer": ["SchNet – a deep learning architecture for molecules and materials", "Directional Message Passing for Molecular Graphs", "GemNet: Universal Directional Graph Neural Networks for Molecules", "Spherical Message Passing for 3D Molecular Graphs"], "answer_arxiv_id": ["1712.06113", "2003.03123", "2106.08903", "2102.05013"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_12702"} +{"question": "Could you provide me some works about iterative optimization methods for discontinuity-preserving normal integration?", "answer": ["Variational Methods for Normal Integration"], "answer_arxiv_id": ["1709.05965"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_12703"} +{"question": "Could you provide me with some papers that discuss the recent exploration of optimal transport in causality, specifically in reweighting-based, matching-based, and representation-based methods?", "answer": ["Optimal transport for causal discovery", "Optimal Transport Weights for Causal Inference", "Optimal Transport for Counterfactual Estimation: A Method for Causal Inference", "Deep Treatment-Adaptive Network for Causal Inference"], "answer_arxiv_id": ["2201.09366", "2109.01991", "2301.07755", "2112.13502"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_12704"} +{"question": "What is the paper that studied data-free black-box distillation with hard labels?", "answer": ["Zero-Shot Knowledge Distillation from a Decision-Based Black-Box Model"], "answer_arxiv_id": ["2106.03310"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_12705"} +{"question": "Are there any works that use large language models to auto-generate intermediate reasoning steps and actions to improve interpretability and problem-solving abilities?", "answer": ["ART: Automatic multi-step reasoning and tool-use for large language models", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2303.09014", "2210.03629"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_12706"} +{"question": "Which papers examined interventions in representations and their impact on a model’s prediction?", "answer": ["Amnesic Probing: Behavioral Explanation with Amnesic Counterfactuals", "Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2006.00995", "2202.05262"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_12707"} +{"question": "Which works proposed the use of ground-plane 2D grids as representations for object detection and segmentation?", "answer": ["Translating Images into Maps", "Lift, Splat, Shoot: Encoding Images from Arbitrary Camera Rigs by Implicitly Unprojecting to 3D", "A Sim2Real Deep Learning Approach for the Transformation of Images from Multiple Vehicle-Mounted Cameras to a Semantically Segmented Image in Bird’s Eye View*", "Orthographic Feature Transform for Monocular 3D Object Detection"], "answer_arxiv_id": ["2110.00966", "2008.05711", "2005.04078", "1811.08188"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_12708"} +{"question": "What are some papers that explore training VLP models with a limited number of image-text pairs?", "answer": ["Unsupervised Multimodal Representation Learning across Medical Images and Reports", "MedCLIP: Contrastive Learning from Unpaired Medical Images and Text"], "answer_arxiv_id": ["1811.08615", "2210.10163"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_12709"} +{"question": "Which works focus on learning a compact representation using a representation learning objective in high-dimensional observations?", "answer": ["Model Based Reinforcement Learning for Atari", "Action-Conditional Video Prediction using Deep Networks in Atari Games", "Learning and Querying Fast Generative Models for Reinforcement Learning", "Recurrent World Models Facilitate Policy Evolution", "Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1903.00374", "1507.08750", "1802.03006", "1809.01999", "1811.04551", "1912.01603", "2010.02193"], "source_meta": {"published_time": "20220918"}, "qid": "AutoScholarQuery_train_12710"} +{"question": "What are some methods to increase memory efficiency in PETL?", "answer": ["Universal Language Model Fine-tuning for Text Classification", "Learning Transferable Visual Models From Natural Language Supervision", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "A Simple Framework for Contrastive Learning of Visual Representations", "Side-Tuning: A Baseline for Network Adaptation via Additive Side\n Networks", "LST: Ladder Side-Tuning for Parameter and Memory Efficient Transfer\n Learning"], "answer_arxiv_id": ["1801.06146", "2103.00020", "1910.10683", "2002.05709", "1912.13503", "2206.06522"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_12711"} +{"question": "Could you provide me some studies about structural re-parameterization to create 2D convolution kernels in computer vision?", "answer": ["ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks", "ExpandNets: Linear Over-parameterization to Train Compact Convolutional Networks", "RepVGG: Making VGG-style ConvNets Great Again", "DO-Conv: Depthwise Over-parameterized Convolutional Layer"], "answer_arxiv_id": ["1908.03930", "1811.10495", "2101.03697", "2006.12030"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_12712"} +{"question": "Can you provide a paper that discusses the use of diffusion models for generating multi-modal trajectories in robot manipulation?", "answer": ["Diffusion Policy: Visuomotor Policy Learning via Action Diffusion"], "answer_arxiv_id": ["2303.04137"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_12713"} +{"question": "What works propose word substitution based methods for data augmentation in low-resource NLP?", "answer": ["DeepStance at SemEval-2016 Task 6: Detecting Stance in Tweets Using Character and Word-Level CNNs", "Character-level Convolutional Networks for Text Classification", "Unsupervised Data Augmentation for Consistency Training", "EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks"], "answer_arxiv_id": ["1606.05694", "1509.01626", "1904.12848", "1901.11196"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_12714"} +{"question": "Could you provide me some works that estimate the underlying MDP using density estimation techniques?", "answer": ["Minimax Model Learning"], "answer_arxiv_id": ["2103.02084"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_12715"} +{"question": "Which paper proposes to operate self-attention on the channel dimension rather than the spatial dimension?", "answer": ["Restormer: Efficient Transformer for High-Resolution Image Restoration"], "answer_arxiv_id": ["2111.09881"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_12716"} +{"question": "What papers proposed techniques for identity blending in multi-subject generation", "answer": ["FastComposer: Tuning-Free Multi-Subject Image Generation with Localized\n Attention"], "answer_arxiv_id": ["2305.10431"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_12717"} +{"question": "Could you list some research articles that discuss the role of chain-of-thought prompting in enhancing complex reasoning in LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Navigate through Enigmatic Labyrinth A Survey of Chain of Thought\n Reasoning: Advances, Frontiers and Future"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2309.15402"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_12718"} +{"question": "What works have proposed methods to estimate or regress the model parameters from a single RGB image?", "answer": ["Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image", "End-to-end Recovery of Human Shape and Pose", "PARE: Part Attention Regressor for 3D Human Body Estimation", "PyMAF: 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment Feedback Loop", "Collaborative Regression of Expressive Bodies using Moderation"], "answer_arxiv_id": ["1607.08128", "1712.06584", "2104.08527", "2103.16507", "2105.05301"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_12719"} +{"question": "What works have been done around object segmentation tasks including image instance segmentation with transformer based models?", "answer": ["Mask R-CNN", "SOLO: Segmenting Objects by Locations", "YOLACT Real-time Instance Segmentation", "SOLOv2: Dynamic and Fast Instance Segmentation", "YOLACT++ Better Real-time Instance Segmentation", "SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation"], "answer_arxiv_id": ["1703.06870", "1912.04488", "1904.02689", "2003.10152", "1912.06218", "2007.14772"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_12720"} +{"question": "Who proposed the EnD regularizer to reduce dataset bias?", "answer": ["EnD: Entangling and Disentangling deep representations for bias correction"], "answer_arxiv_id": ["2103.02023"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_12721"} +{"question": "What research has been done to develop the probabilistic notion of machine unlearning?", "answer": ["Making AI Forget You: Data Deletion in Machine Learning", "Certified Data Removal from Machine Learning Models", "Descent-to-Delete: Gradient-Based Methods for Machine Unlearning", "Machine Unlearning via Algorithmic Stability", "Remember What You Want to Forget: Algorithms for Machine Unlearning"], "answer_arxiv_id": ["1907.05012", "1911.03030", "2007.02923v1", "2102.13179", "2103.03279v2"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_12722"} +{"question": "What recent works managed to fill the gap in the analysis of player dynamics?", "answer": ["Convergence of Learning Dynamics in Information Retrieval Games"], "answer_arxiv_id": ["1806.05359"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_12723"} +{"question": "Are there any studies using human-designed intermediate energy guidance for image-to-image translation and inverse molecular design?", "answer": ["EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations", "Equivariant Energy-Guided SDE for Inverse Molecular Design"], "answer_arxiv_id": ["2207.06635", "2209.15408"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_12724"} +{"question": "Which works discuss improvements in text generation through the use of Conditional Random Fields?", "answer": ["Fast Structured Decoding for Sequence Models", "Non-Autoregressive Text Generation with Pre-trained Language Models"], "answer_arxiv_id": ["1910.11555", "2102.08220"], "source_meta": {"published_time": "20230814"}, "qid": "AutoScholarQuery_train_12725"} +{"question": "What works presented stereo methods in Camera-based 3D object detection?", "answer": ["DSGN: Deep Stereo Geometry Network for 3D Object Detection", "LIGA-Stereo: Learning LiDAR Geometry Aware Representations for\n Stereo-based 3D Detector", "Point-Based Multi-View Stereo Network"], "answer_arxiv_id": ["2001.03398", "2108.08258", "1908.04422"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_12726"} +{"question": "What datasets were used to ensure caption diversity in the development of the Polaris dataset?", "answer": ["nocaps: novel object captioning at scale"], "answer_arxiv_id": ["1812.08658v3"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_12727"} +{"question": "Where AQC has been applied to solve optimization problems for improved traffic flow and in the finance sector for portfolio optimization?", "answer": ["Traffic flow optimization using a quantum annealer", "Hybrid Quantum Investment Optimization with Minimal Holding Period"], "answer_arxiv_id": ["1708.01625", "2012.01091"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_12728"} +{"question": "What are the references that used text-to-image diffusion models in the text-to-video field?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators"], "answer_arxiv_id": ["2209.14792", "2303.13439"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_12729"} +{"question": "Could you provide me some studies introducing diverse masking pipelines for learning audio-visual representations?", "answer": ["MAViL: Masked Audio-Video Learners", "Contrastive Audio-Visual Masked Autoencoder", "Audiovisual Masked Autoencoders"], "answer_arxiv_id": ["2212.08071", "2210.07839", "2212.05922"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_12730"} +{"question": "Could you provide me research about subject-driven image generation?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Cones: Concept Neurons in Diffusion Models for Customized Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242", "2303.05125"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_12731"} +{"question": "What works look into Transfer Learning in Natural Language Processing along with the shift to Zero-Shot and Few-Shot Learning?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "Language Models are Few-Shot Learners", "Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2102.05918", "2112.11446", "2005.14165", "2109.01652"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_12732"} +{"question": "Which works leveraged neural networks for subspace construction and simulation in the field of Neural Physics?", "answer": ["N-Cloth: Predicting 3D Cloth Deformation with Mesh-Based Networks", "DEMEA: Deep Mesh Autoencoders for Non-Rigidly Deforming Objects"], "answer_arxiv_id": ["2112.06397", "1905.10290"], "source_meta": {"published_time": "20240426"}, "qid": "AutoScholarQuery_train_12733"} +{"question": "Could you provide me some research about convolutional neural networks?", "answer": ["Accelerating Eulerian Fluid Simulation With Convolutional Networks", "Prediction of Aerodynamic Flow Fields Using Convolutional Neural Networks"], "answer_arxiv_id": ["1607.03597", "1905.13166"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_12734"} +{"question": "What works conducted neural architecture search for anomaly detection, but do not consider networks beyond Convolutional Neural Networks?", "answer": ["Neural Architecture Search for Visual Anomaly Segmentation"], "answer_arxiv_id": ["2304.08975"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_12735"} +{"question": "Which works have utilized state entropy in unsupervised Reinforcement Learning?", "answer": ["Efficient Exploration via State Marginal Matching", "Provably Efficient Maximum Entropy Exploration", "An Intrinsically-Motivated Approach for Learning Highly Exploring and Fast Mixing Policies", "Behavior From the Void: Unsupervised Active Pre-Training", "Reinforcement Learning with Prototypical Representations", "Exploration by Maximizing Rényi Entropy for Reward-Free RL Framework", "Geometric Entropic Exploration", "The Importance of Non-Markovianity in Maximum State Entropy Exploration", "Unsupervised Reinforcement Learning in Multiple Environments"], "answer_arxiv_id": ["1906.05274", "1812.02690", "1907.04662", "2103.04551", "2102.11271v2", "2006.06193", "2101.02055v2", "2202.03060", "2112.08746v1"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_12736"} +{"question": "What works proposed robust Q-learning algorithms for online setting in the context of distributionally robust RL?", "answer": ["Online Robust Reinforcement Learning with Model Uncertainty"], "answer_arxiv_id": ["2109.14523"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_12737"} +{"question": "Which is the first work that divided the text into discourse units for graph-based reasoning?", "answer": ["DAGN: Discourse-Aware Graph Network for Logical Reasoning"], "answer_arxiv_id": ["2103.14349"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_12738"} +{"question": "What papers recognize visual reference segmentation as a primary task in modern large-scale vision models?", "answer": ["Segment Everything Everywhere All at Once", "Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning", "Visual Prompting via Image Inpainting"], "answer_arxiv_id": ["2304.06718", "2212.02499", "2209.00647"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_12739"} +{"question": "What papers focus on solving the general stochastic bilevel problem assuming the lower-level problem is strongly convex?", "answer": ["A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic", "Approximation Methods for Bilevel Programming", "Provably Faster Algorithms for Bilevel Optimization", "A Near-Optimal Algorithm for Stochastic Bilevel Optimization via Double-Momentum", "Tighter Analysis of Alternating Stochastic Gradient Method for Stochastic Nested Problems", "Projection-Free Stochastic Bi-level Optimization"], "answer_arxiv_id": ["2007.05170", "1802.02246", "2106.04692", "2102.07367", "2106.13781", "2110.11721"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_12740"} +{"question": "Could you provide me with some studies where diffusion models were used for image editing?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Blended Diffusion for Text-driven Editing of Natural Images", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "Paint by Example: Exemplar-based Image Editing with Diffusion Models"], "answer_arxiv_id": ["2208.01626", "2208.12242", "2210.09276", "2111.14818", "2108.01073", "2211.13227"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_12741"} +{"question": "Are there any works which specifically proposed masked point modeling methods?", "answer": ["Masked Autoencoders for Point Cloud Self-supervised Learning", "Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling", "Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-training"], "answer_arxiv_id": ["2203.06604", "2111.14819", "2205.14401"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_12742"} +{"question": "What research indicated that the memory cost of class-prototypes is low compared to using a rehearsal buffer?", "answer": ["The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization"], "answer_arxiv_id": ["2006.16241"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_12743"} +{"question": "What studies focused on reducing local model drifting in local training in FL?", "answer": ["Federated Learning with Non-IID Data", "Federated Optimization in Heterogeneous Networks", "FedDANE: A Federated Newton-Type Method", "Federated Learning Based on Dynamic Regularization", "Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization", "Federated Learning with Unbiased Gradient Aggregation and Controllable Meta Updating"], "answer_arxiv_id": ["1806.00582", "1812.06127", "2001.01920", "2111.04263", "2008.03606", "1910.06378", "2007.07481", "1910.08234"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_12744"} +{"question": "In weakly-supervised grounding, which work utilized a model called Grounding by Separation (GbS)?", "answer": ["Detector-Free Weakly Supervised Grounding by Separation"], "answer_arxiv_id": ["2104.09829"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_12745"} +{"question": "What studies introduce trainable masks in pruning during training?", "answer": ["Dynamic Sparse Training: Find Efficient Sparse Network from scratch with Trainable Masked Layers", "Winning the Lottery with Continuous Sparsification", "Operation-Aware Soft Channel Pruning using Differentiable Masks", "Soft Threshold Weight Reparameterization for Learnable Sparsity", "Training Sparse Neural Networks"], "answer_arxiv_id": ["2005.06870", "1912.04427", "2007.03938", "2002.03231", "1611.06694"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_12746"} +{"question": "Could you provide works on masked image modeling for self-supervised learning?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "Masked Autoencoders Are Scalable Vision Learners", "Hard Patches Mining for Masked Image Modeling", "Context Autoencoder for Self-Supervised Representation Learning", "Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture", "Masked Image Modeling with Local Multi-Scale Reconstruction"], "answer_arxiv_id": ["2106.08254", "2111.06377", "2304.05919", "2202.03026", "2301.08243", "2303.05251"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_12747"} +{"question": "Which papers discuss the vulnerability of LLMs to various forms of adversarial attacks?", "answer": ["Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment", "BERT-ATTACK: Adversarial Attack Against BERT Using BERT", "Benchmarking Robustness of Machine Reading Comprehension Models", "Adversarial Examples for Evaluating Reading Comprehension Systems", "Better Robustness by More Coverage: Adversarial Training with Mixup Augmentation for Robust Fine-tuning"], "answer_arxiv_id": ["1907.11932", "2004.09984", "2004.14004", "1707.07328", "2012.15699v3"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_12748"} +{"question": "What kind of research is conducted to improve the transferability of white-box attacks?", "answer": ["Boosting Adversarial Attacks with Momentum", "Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks", "Enhancing the Transferability of Adversarial Attacks through Variance Tuning", "Boosting Adversarial Transferability through Enhanced Momentum"], "answer_arxiv_id": ["1710.06081", "1908.06281", "2103.15571", "2103.10609"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_12749"} +{"question": "What papers researched interchange intervention, a branch of intervention-based techniques for uncovering causal mechanisms?", "answer": ["Neural Natural Language Inference Models Partially Embed Theories of Lexical Entailment and Negation"], "answer_arxiv_id": ["2004.14623v4"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_12750"} +{"question": "What papers focused on code-to-code generation tasks like automatic program repair (APR) and code translation?", "answer": ["An Empirical Study on Learning Bug-Fixing Patches in the Wild via Neural\n Machine Translation", "CodeXGLUE: A Machine Learning Benchmark Dataset for Code Understanding\n and Generation"], "answer_arxiv_id": ["1812.08693", "2102.04664"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_12751"} +{"question": "Could you provide me with works about the systematic study on local and global geometry properties of loss landscape?", "answer": ["Taxonomizing local versus global structure in neural network loss landscapes"], "answer_arxiv_id": ["2107.11228"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_12752"} +{"question": "What research papers propose methods for text-to-image generation and editing?", "answer": ["Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "CogView: Mastering Text-to-Image Generation via Transformers", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Dreamix: Video Diffusion Models are General Video Editors", "Structure and Content-Guided Video Synthesis with Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2205.11487", "2206.10789", "2204.06125", "2105.13290", "2203.13131", "2112.10741", "2112.10752", "2302.01329", "2302.03011"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_12753"} +{"question": "Which work first proposed interpolation of two images and their labels to augment the training data?", "answer": ["mixup: Beyond Empirical Risk Minimization"], "answer_arxiv_id": ["1710.09412"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_12754"} +{"question": "What works are associated with freezing-based techniques in the context of PEFT methods?", "answer": ["BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models", "What Would Elsa Do? Freezing Layers During Transformer Fine-Tuning"], "answer_arxiv_id": ["2106.10199", "1911.03090"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_12755"} +{"question": "Which work developed the Normalized Object Coordinate Space (NOCS) in the realm of category-level pose estimation?", "answer": ["Normalized Object Coordinate Space for Category-Level 6D Object Pose and\n Size Estimation"], "answer_arxiv_id": ["1901.02970"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_12756"} +{"question": "Which works are considered as noteworthy advances sparked by FCN?", "answer": ["Pyramid Scene Parsing Network", "Unified Perceptual Parsing for Scene Understanding", "U-Net: Convolutional Networks for Biomedical Image Segmentation"], "answer_arxiv_id": ["1612.01105", "1807.10221", "1505.04597"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_12757"} +{"question": "What papers specifically studied permutation symmetry in certain overparametrized networks?", "answer": ["Geometry of the Loss Landscape in Overparameterized Neural Networks: Symmetries and Invariances"], "answer_arxiv_id": ["2105.12221v2"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_12758"} +{"question": "What study combines Siamese networks with masked modeling?", "answer": ["Masked Siamese Networks for Label-Efficient Learning"], "answer_arxiv_id": ["2204.07141"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_12759"} +{"question": "Which work provided the idea of compressing transformer activations into a smaller compressed memory using a learned convolutional operator?", "answer": ["Compressive Transformers for Long-Range Sequence Modelling"], "answer_arxiv_id": ["1911.05507"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_12760"} +{"question": "What papers directly operate on 3D points in point-based methods for point cloud semantic segmentation?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space"], "answer_arxiv_id": ["1612.00593", "1706.02413"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_12761"} +{"question": "Could you provide me some works dealing with diffusion models for high resolutions?", "answer": ["Cascaded Diffusion Models for High Fidelity Image Generation", "ERNIE-ViLG 2.0: Improving Text-to-Image Diffusion Model with Knowledge-Enhanced Mixture-of-Denoising-Experts", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2106.15282", "2210.15257", "2211.01324", "2112.10752"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12762"} +{"question": "Which research introduced SegFix used for both semantic segmentation and instance segmentation?", "answer": ["SegFix: Model-Agnostic Boundary Refinement for Segmentation"], "answer_arxiv_id": ["2007.04269"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_12763"} +{"question": "Which papers introduced the concept of 'In-Context Learning' in LLMs?", "answer": ["Language Models are Few-Shot Learners", "OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["2005.14165", "2205.01068"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_12764"} +{"question": "Can you provide examples of studies that made attempts to transfer deformation utilizing examples or text?", "answer": ["Neural Cages for Detail-Preserving 3D Deformations", "DeformSyncNet: Deformation Transfer via Synchronized Shape Deformation\n Spaces", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "CLIP-Mesh: Generating textured meshes from text using pretrained\n image-text models", "TextDeformer: Geometry Manipulation using Text Guidance"], "answer_arxiv_id": ["1912.06395", "2009.01456", "2112.03221", "2203.13333", "2304.13348"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_12765"} +{"question": "Which research papers have proposed techniques for structural pruning to reduce inference time in diffusion models?", "answer": ["Structural Pruning for Diffusion Models", "SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two\n Seconds"], "answer_arxiv_id": ["2305.10924", "2306.00980"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_12766"} +{"question": "Can you provide some research works that specifically aimed to promote diversity in the decoding process of language models?", "answer": ["A Simple, Fast Diverse Decoding Algorithm for Neural Generation"], "answer_arxiv_id": ["1611.08562"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_12767"} +{"question": "Which studies utilize iterative optimization to estimate the parameters of a human model for human mesh recovery?", "answer": ["Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a\n Single Image", "Unite the People: Closing the Loop Between 3D and 2D Human\n Representations", "Monocular Total Capture: Posing Face, Body, and Hands in the Wild", "Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "Human Body Model Fitting by Learned Gradient Descent", "Decoupling Human and Camera Motion from Videos in the Wild"], "answer_arxiv_id": ["1607.08128", "1701.02468", "1812.01598", "1904.05866", "2008.08474", "2302.12827"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_12768"} +{"question": "Any studies about Physics-Informed Neural Operators (PINO)?", "answer": ["Physics-Informed Neural Operator for Learning Partial Differential Equations"], "answer_arxiv_id": ["2111.03794"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_12769"} +{"question": "What are the studies that proposed a solution for linear PDEs enforcing the differential form of the PDE as a hard constraint?", "answer": ["Learning differentiable solvers for systems with hard constraints", "Adaptive Self-supervision Algorithms for Physics-informed Neural Networks"], "answer_arxiv_id": ["2207.08675", "2207.04084"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_12770"} +{"question": "What papers do ConSpec and employ slot-like attention-based algorithms?", "answer": ["Recurrent Independent Mechanisms", "Object Files and Schemata: Factorizing Declarative and Procedural Knowledge in Dynamical Systems"], "answer_arxiv_id": ["1909.10893", "2006.16225"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_12771"} +{"question": "Which studies propose using temporal average as a tilt-free reference frame for tilt rectification?", "answer": ["Application of Tilt Correlation Statistics to Anisoplanatic Optical\n Turbulence Modeling and Mitigation"], "answer_arxiv_id": ["2108.00528"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_12772"} +{"question": "Are there any studies that show the benefits of weight ensembles and interpolations in models trained from scratch?", "answer": ["Git Re-Basin: Merging Models Modulo Permutation Symmetries", "Model Fusion via Optimal Transport"], "answer_arxiv_id": ["2209.04836", "1910.05653"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_12773"} +{"question": "What works skip forward/backward passes for specific layers based on statistics from prior forward/backward passes?", "answer": ["AutoFreeze: Automatically Freezing Model Blocks to Accelerate Fine-tuning"], "answer_arxiv_id": ["2102.01386v2"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_12774"} +{"question": "What papers observed the linear correlations between the ID (In-Distribution) and OOD (Out-Of-Distribution) performances?", "answer": ["Accuracy on the Line: On the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization", "Do CIFAR-10 Classifiers Generalize to CIFAR-10?", "Do ImageNet Classifiers Generalize to ImageNet?", "Cold Case: the Lost MNIST Digits", "The Effect of Natural Distribution Shift on Question Answering Models"], "answer_arxiv_id": ["2107.04649", "1806.00451", "1902.10811", "1905.10498", "2004.14444"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_12775"} +{"question": "Which works discusses the use of MMLM task in multilingual pre-trained models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Unsupervised Cross-lingual Representation Learning at Scale"], "answer_arxiv_id": ["1810.04805", "1911.02116"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_12776"} +{"question": "Which researchers introduced the notion of counterfactual memorization?", "answer": ["Counterfactual Memorization in Neural Language Models"], "answer_arxiv_id": ["2112.12938"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_12777"} +{"question": "Which studies worked on using VLMs to condition plans on visual inputs for AI agents?", "answer": ["VoxPoser: Composable 3D Value Maps for Robotic Manipulation with\n Language Models", "Distilling Internet-Scale Vision-Language Models into Embodied Agents", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Physically Grounded Vision-Language Models for Robotic Manipulation"], "answer_arxiv_id": ["2307.05973", "2301.12507", "2204.01691", "2309.02561"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_12778"} +{"question": "Which research work investigates the use of transformer-based dialogue modeling methods?", "answer": ["A Simple Language Model for Task-Oriented Dialogue", "Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based Chatbots", "Filling the Gap of Utterance-aware and Speaker-aware Representation for Multi-turn Dialogue", "CSAGN: Conversational Structure Aware Graph Network for Conversational Semantic Role Labeling"], "answer_arxiv_id": ["2005.00796", "2004.03588", "2009.06504", "2109.11541"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_12779"} +{"question": "What studies suggest that achieving graph isomorphism invariance and universal approximation is as challenging as solving the graph isomorphism issue?", "answer": ["On the Equivalence between Graph Isomorphism Testing and Function Approximation with GNNs"], "answer_arxiv_id": ["1905.12560"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_12780"} +{"question": "Which works are about point-based methods for 3D object detection in autonomous driving scenes?", "answer": ["PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud", "Multi-View 3D Object Detection Network for Autonomous Driving", "PIXOR: Real-time 3D Object Detection from Point Clouds"], "answer_arxiv_id": ["1812.04244", "1611.07759", "1902.06326"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_12781"} +{"question": "What studies proposed consistency-based methods in semi-supervised learning and how did they work?", "answer": ["Temporal Ensembling for Semi-Supervised Learning", "Virtual Adversarial Training: A Regularization Method for Supervised and\n Semi-Supervised Learning", "Mean teachers are better role models: Weight-averaged consistency\n targets improve semi-supervised deep learning results"], "answer_arxiv_id": ["1610.02242v3", "1704.03976", "1703.01780"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_12782"} +{"question": "Which studies in cross-lingual representation pre-training have used language modeling and contrastive learning techniques?", "answer": ["Cross-lingual Language Model Pretraining", "InfoXLM: An Information-Theoretic Framework for Cross-Lingual Language Model Pre-Training", "Explicit Alignment Objectives for Multilingual Bidirectional Encoders", "On Learning Universal Representations Across Languages"], "answer_arxiv_id": ["1901.07291", "2007.07834", "2010.07972", "2007.15960"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_12783"} +{"question": "Identify the studies that incorporate ML models into heuristic components in modern solvers.", "answer": ["Learning Cut Selection for Mixed-Integer Linear Programming via Hierarchical Sequence Model"], "answer_arxiv_id": ["2302.00244"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_12784"} +{"question": "Can you mention studies that have documented a trade-off between the robustness of adversarial perturbations and clean image accuracy?", "answer": ["Adversarial Examples Improve Image Recognition", "Feature Denoising for Improving Adversarial Robustness"], "answer_arxiv_id": ["1911.09665", "1812.03411"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_12785"} +{"question": "Can you name some papers that deal with one-stage methods for HOI detection?", "answer": ["UnionDet: Union-Level Detector Towards Real-Time Human-Object Interaction Detection", "PPDM: Parallel Point Detection and Matching for Real-time Human-Object Interaction Detection", "Learning Human-Object Interaction Detection using Interaction Points", "DIRV: Dense Interaction Region Voting for End-to-End Human-Object Interaction Detection"], "answer_arxiv_id": ["2312.12664", "1912.12898", "2003.14023", "2010.01005"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_12786"} +{"question": "What papers introduced the use of a large-scale segmentation dataset and a training framework for prompt-driven segmentation?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_12787"} +{"question": "Could you provide me with the research that utilizes image examples to enhance text descriptions for better open-vocabulary object detection performance?", "answer": ["Multi-modal Queried Object Detection in the Wild"], "answer_arxiv_id": ["2305.18980"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_12788"} +{"question": "What works discussed various ways to remove the structure in structural pruning such as hessian-based estimation?", "answer": ["The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models", "EigenDamage: Structured Pruning in the Kronecker-Factored Eigenbasis"], "answer_arxiv_id": ["2203.07259", "1905.05934"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_12789"} +{"question": "Which papers discuss recent improvements in diffusion models leading to advanced image synthesis?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Zero-Shot Text-to-Image Generation", "Taming Transformers for High-Resolution Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2105.05233", "2102.12092", "2012.09841", "2112.10752", "2203.13131", "2112.10741", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12790"} +{"question": "Which model infers a CLIP image embedding given a text representation while the decoder synthesizes an image?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2204.06125"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_12791"} +{"question": "Any research applied sparse projections on the simplex for marginalizing over discrete variables?", "answer": ["Efficient Marginalization of Discrete and Structured Latent Variables via Sparsity"], "answer_arxiv_id": ["2007.01919"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12792"} +{"question": "Which research works focus on the graph-based approaches to approximate nearest neighbour (ANN) search?", "answer": ["Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs"], "answer_arxiv_id": ["1603.09320v4"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_12793"} +{"question": "Which are the studies that previously considered the sample-based algorithms used in the present study?", "answer": ["On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift", "An Improved Analysis of (Variance-Reduced) Policy Gradient and Natural Policy Gradient Methods"], "answer_arxiv_id": ["1908.00261", "2211.07937"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_12794"} +{"question": "Could you provide me studies where pair-wise node similarity derived from shortest paths, diffusion kernels, random walks, etc. were used in Graph Transformers?", "answer": ["Rethinking Graph Transformers with Spectral Attention", "GraphiT: Encoding Graph Structure in Transformers"], "answer_arxiv_id": ["2106.03893", "2106.05667"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_12795"} +{"question": "What study combined the active learning criterion with data augmentation methods, using the gradient of acquisition function after one-step augmentation as guidance?", "answer": ["LADA: Look-Ahead Data Acquisition via Augmentation for Active Learning"], "answer_arxiv_id": ["2011.04194"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_12796"} +{"question": "Which work builds the vision dictionary with a first-in-first-out queue for contrast architecture regularization?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1911.05722"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_12797"} +{"question": "Who has conducted research to derive meaning from the learned structures in the architecture of transformer-stack models?", "answer": ["Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned", "Are Sixteen Heads Really Better than One?", "BERT Rediscovers the Classical NLP Pipeline", "Revealing the Dark Secrets of BERT", "A Multiscale Visualization of Attention in the Transformer Model", "A Primer in BERTology: What We Know About How BERT Works", "On the Expressive Power of Self-Attention Matrices"], "answer_arxiv_id": ["1905.09418", "1905.10650", "1905.05950", "1908.08593", "1906.05714", "2002.12327", "2106.03764"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_12798"} +{"question": "Which papers focus on generative modeling in the surveillance of variational autoencoders?", "answer": ["Auto-Encoding Variational Bayes", "Stochastic Backpropagation and Approximate Inference in Deep Generative Models"], "answer_arxiv_id": ["1312.6114", "1401.4082v3"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_12799"} +{"question": "What works use volumetric representations to implicitly capture hair?", "answer": ["HVH: Learning a Hybrid Neural Volumetric Representation for Dynamic Hair\n Performance Capture", "Learning Compositional Radiance Fields of Dynamic Human Heads", "NeuWigs: A Neural Dynamic Model for Volumetric Hair Capture and\n Animation", "Neural Strands: Learning Hair Geometry and Appearance from Multi-View\n Images", "Learning Disentangled Avatars with Hybrid 3D Representations"], "answer_arxiv_id": ["2112.06904", "2012.09955", "2212.00613", "2207.14067", "2309.06441"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12800"} +{"question": "Which works propose label-efficient assessment in the context of classifier performance?", "answer": ["Active Bayesian Assessment for Black-Box Classifiers"], "answer_arxiv_id": ["2002.06532"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_12801"} +{"question": "Could you provide me some works that study projection-based transformations for 3D semantic segmentation?", "answer": ["SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloud", "Deep Projective 3D Semantic Segmentation", "2DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Clouds", "Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation", "Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1809.08495", "1705.03428", "2207.04397", "2011.10033", "2007.16100", "1904.08755"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_12802"} +{"question": "What studies have been conducted on datasets for image-based action understanding?", "answer": ["Visual Semantic Role Labeling"], "answer_arxiv_id": ["1505.04474"], "source_meta": {"published_time": "20230402"}, "qid": "AutoScholarQuery_train_12803"} +{"question": "Which works discuss higher-order ODE solvers that generate samples in a few steps?", "answer": ["DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "GENIE: Higher-Order Denoising Diffusion Solvers"], "answer_arxiv_id": ["2206.00927", "2210.05475"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_12804"} +{"question": "Could you list some works that assume the cross-modal paired data is achievable?", "answer": ["Cross-modal knowledge distillation for action recognition", "Learning an Augmented RGB Representation with Cross-Modal Knowledge Distillation for Action Detection"], "answer_arxiv_id": ["1910.04641", "2108.03619v1"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_12805"} +{"question": "Can you name studies where an adaptive token Transformer was proposed to improve the inference efficiency?", "answer": ["MonoATT: Online Monocular 3D Object Detection with Adaptive Token\n Transformer"], "answer_arxiv_id": ["2303.13018"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_12806"} +{"question": "Which work used a scattering transform generative model to perform source separation in a Bayesian framework?", "answer": ["Single frequency CMB B-mode inference with realistic foregrounds from a single training image"], "answer_arxiv_id": ["2111.01138"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_12807"} +{"question": "Which papers discussed methods similar to layer dropping?", "answer": ["Reducing Transformer Depth on Demand with Structured Dropout", "AutoFreeze: Automatically Freezing Model Blocks to Accelerate Fine-tuning"], "answer_arxiv_id": ["1909.11556", "2102.01386v2"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_12808"} +{"question": "What methods adopted the approach of integrating extracted visual features directly through linear layers into pre-trained models?", "answer": ["PaLM-E: An Embodied Multimodal Language Model", "Visual Instruction Tuning", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic"], "answer_arxiv_id": ["2303.03378", "2304.08485", "2306.15195"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_12809"} +{"question": "Could you list the papers that discuss prevention of toxicity as an objective of safety alignment strategies in conversational language models?", "answer": ["RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models"], "answer_arxiv_id": ["2009.11462v2"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_12810"} +{"question": "What are the pioneering vision-language models that adopt contrastive learning paradigms?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_12811"} +{"question": "Which research has employed systems that provide object-centric representations as the input to a transformer network?", "answer": ["Attention over learned object embeddings enables complex visual reasoning", "SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Models"], "answer_arxiv_id": ["2012.08508", "2210.05861"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_12812"} +{"question": "Can you identify research papers that utilize Graph Neural Networks to model similarity and data association in the course of conducting multi-object tracking?", "answer": ["Learning a Proposal Classifier for Multiple Object Tracking", "Learnable Graph Matching: Incorporating Graph Partitioning with Deep\n Feature Learning for Multiple Object Tracking"], "answer_arxiv_id": ["2103.07889", "2103.16178"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_12813"} +{"question": "Can you provide research that utilizes the dataset first introduced in CaloGAN, for the simulation of particle physics?", "answer": ["CaloGAN: Simulating 3D High Energy Particle Showers in Multi-Layer Electromagnetic Calorimeters with Generative Adversarial Networks"], "answer_arxiv_id": ["1712.10321"], "source_meta": {"published_time": "20220210"}, "qid": "AutoScholarQuery_train_12814"} +{"question": "Are there any papers that propose a single-stage pipeline for visual grounding?", "answer": ["3D-SPS: Single-Stage 3D Visual Grounding via Referred Point Progressive\n Selection"], "answer_arxiv_id": ["2204.06272"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_12815"} +{"question": "What papers are related to the first class of non-equilibrium-based CL originating from Contrastive Predictive Coding (CPC)?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Local plasticity rules can learn deep representations using self-supervised contrastive predictions"], "answer_arxiv_id": ["1807.03748", "2010.08262"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_12816"} +{"question": "Which work proposed an efficient algorithm for computing adaptive interventions with provable approximation guarantees on general graphs?", "answer": ["Verification and search algorithms for causal DAGs"], "answer_arxiv_id": ["2206.15374"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_12817"} +{"question": "Which studies popularized large scale contrastive vision-language pretraining?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_12818"} +{"question": "Could you list some works using metric-based methods for deciding bit-width allocation in prime factors decomposition?", "answer": ["HAWQ: Hessian AWare Quantization of Neural Networks with Mixed-Precision", "HAWQ-V2: Hessian Aware trace-Weighted Quantization of Neural Networks", "HAWQV3: Dyadic Neural Network Quantization", "Towards Mixed-Precision Quantization of Neural Networks via Constrained\n Optimization", "OMPQ: Orthogonal Mixed Precision Quantization", "Pruning neural networks without any data by iteratively conserving\n synaptic flow", "Mixed-Precision Neural Network Quantization via Learned Layer-wise\n Importance"], "answer_arxiv_id": ["1905.03696", "1911.03852", "2011.10680", "2110.06554", "2109.07865", "2006.05467", "2203.08368"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_12819"} +{"question": "Any studies about transforming different modalities of data into a unified modality in the HQA task?", "answer": ["Binding Language Models in Symbolic Languages", "MMHQA-ICL: Multimodal In-context Learning for Hybrid Question Answering\n over Text, Tables and Images", "TSQA: Tabular Scenario Based Question Answering"], "answer_arxiv_id": ["2210.02875", "2309.04790", "2101.11429"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_12820"} +{"question": "Any works about shadow synthesis methods applied to image compositing?", "answer": ["Shadow Generation for Composite Image in Real-world Scenes", "PixHt-Lab: Pixel Height Based Light Effect Generation for Image\n Compositing"], "answer_arxiv_id": ["2104.10338", "2303.00137"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_12821"} +{"question": "What paper introduced the Swin Transformer?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2103.14030"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12822"} +{"question": "What research works have proposed methods to compress long contexts in large language models?", "answer": ["LLMLingua: Compressing Prompts for Accelerated Inference of Large\n Language Models", "LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios\n via Prompt Compression"], "answer_arxiv_id": ["2310.05736", "2310.06839"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_12823"} +{"question": "Which work first presented diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_12824"} +{"question": "What works have shown that reinforcement learning algorithms converge to Nash equilibrium in two-player zero-sum Markov games?", "answer": ["Online Reinforcement Learning in Stochastic Games", "Learning Zero-Sum Simultaneous-Move Markov Games Using Function Approximation and Correlated Equilibrium", "Independent Policy Gradient Methods for Competitive Reinforcement Learning", "V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL"], "answer_arxiv_id": ["1712.00579", "2002.07066", "2101.04233v1", "2110.14555"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_12825"} +{"question": "Which research showed a specific gradient-based training can learn polynomials with low-dimensional latent representation?", "answer": ["Neural Networks can Learn Representations with Gradient Descent"], "answer_arxiv_id": ["2206.15144"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_12826"} +{"question": "Which model serves as the extension of the Deep Kalman Filter?", "answer": ["Structured Inference Networks for Nonlinear State Space Models"], "answer_arxiv_id": ["1609.09869v2"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_12827"} +{"question": "In what paper does the researcher rely on a random binary cumulant matrix for their study?", "answer": ["Proto-Value Networks: Scaling Representation Learning with Auxiliary Tasks"], "answer_arxiv_id": ["2304.12567"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_12828"} +{"question": "What works improved the descriptive ability of VLM by using a stronger text encoder or visual encoder?", "answer": ["Attention Is All You Need", "Scaling Vision Transformers", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["1706.03762", "2106.04560", "2301.12597"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_12829"} +{"question": "Could you provide me a study that uses ensembles of predictions for different augmentations of a test sample in Test Time Adaptation?", "answer": ["MEMO: Test Time Robustness via Adaptation and Augmentation"], "answer_arxiv_id": ["2110.09506"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_12830"} +{"question": "Can you provide me prior works that adopted transliteration during the fine-tuning phase for cross-lingual transfer?", "answer": ["IndicBART: A Pre-trained Model for Indic Natural Language Generation", "When Being Unseen from mBERT is just the Beginning: Handling New\n Languages With Multilingual Language Models", "Specializing Multilingual Language Models: An Empirical Study"], "answer_arxiv_id": ["2109.02903", "2010.12858", "2106.09063"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_12831"} +{"question": "Which studies proposed multi-view image-based methods for novel view synthesis?", "answer": ["DeepStereo: Learning to Predict New Views from the World's Imagery", "Learning-Based View Synthesis for Light Field Cameras"], "answer_arxiv_id": ["1506.06825", "1609.02974"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_12832"} +{"question": "What research propounds the one-sided anomaly-focused deviation loss for one-shot anomaly detection (OSAD)?", "answer": ["Explainable Deep Few-shot Anomaly Detection with Deviation Networks"], "answer_arxiv_id": ["2108.00462"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_12833"} +{"question": "Which studies utilize transfer learning to address data heterogeneity in the context of personalized federated learning?", "answer": ["Salvaging Federated Learning by Local Adaptation", "Federated Learning with Non-IID Data"], "answer_arxiv_id": ["2002.04758", "1806.00582"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_12834"} +{"question": "What paper discusses masking patches of images based on the class activation map and refills them from patches of other images?", "answer": ["Masked Images Are Counterfactual Samples for Robust Fine-tuning"], "answer_arxiv_id": ["2303.03052"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_12835"} +{"question": "Which papers introduced the diffusion-based generative models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_12836"} +{"question": "Which research papers applied transformations such as rotation, affinely- and projectively-transformed to each original image in pretext tasks for self-supervised image representation?", "answer": ["Unsupervised Representation Learning by Predicting Image Rotations", "AET vs. AED: Unsupervised Representation Learning by Auto-Encoding\n Transformations rather than Data", "Colorization as a Proxy Task for Visual Understanding"], "answer_arxiv_id": ["1803.07728", "1901.04596", "1703.04044"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_12837"} +{"question": "What papers have developed hypergraph neural network architectures?", "answer": ["UniGNN: a Unified Framework for Graph and Hypergraph Neural Networks"], "answer_arxiv_id": ["2105.00956"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_12838"} +{"question": "What research works have focused on metric-based methods for unsupervised domain adaptation?", "answer": ["Learning Transferable Features with Deep Adaptation Networks", "Deep Domain Confusion: Maximizing for Domain Invariance", "Simultaneous Deep Transfer Across Domains and Tasks", "Central Moment Discrepancy (CMD) for Domain-Invariant Representation Learning"], "answer_arxiv_id": ["1502.02791", "1412.3474", "1510.02192", "1702.08811"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_train_12839"} +{"question": "Which works have trained a single generalist policy to play multiple Atari games with off-policy RL and online data collection?", "answer": ["IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures", "Multi-task Deep Reinforcement Learning with PopArt", "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control"], "answer_arxiv_id": ["1802.01561", "1809.04474", "1909.12238"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_12840"} +{"question": "Which paper proposes an estimator for the gradient of proximal policy optimization (PPO)?", "answer": ["Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1707.06347v2"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_12841"} +{"question": "Could you provide me some works that have followed the approach of incorporating global latent codes and applying test-time tuning to refine these codes?", "answer": ["Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "ShaRF: Shape-conditioned Radiance Fields from a Single View", "CodeNeRF: Disentangled Neural Radiance Fields for Object Categories", "AutoRF: Learning 3D Object Radiance Fields from Single View Observations"], "answer_arxiv_id": ["1906.01618", "2102.08860", "2109.01750", "2204.03593"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_12842"} +{"question": "Could you provide me with some studies that are related to the use of learned world models in zero-shot generalization task?", "answer": ["Planning to Explore via Self-Supervised World Models", "Discovering and Achieving Goals via World Models"], "answer_arxiv_id": ["2005.05960", "2110.09514"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_12843"} +{"question": "Could you provide some studies that proposed various variants of Adam?", "answer": ["On the Convergence of Adaptive Gradient Methods for Nonconvex Optimization", "On the Convergence of A Class of Adam-Type Algorithms for Non-Convex Optimization", "A Simple Convergence Proof of Adam and Adagrad", "Adam Can Converge Without Any Modification On Update Rules"], "answer_arxiv_id": ["1808.05671", "1808.02941", "2003.02395", "2208.09632v5"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_12844"} +{"question": "What papers study early fusion approaches in cooperative perception and adopt LiDAR as a sensor?", "answer": ["Cooper: Cooperative Perception for Connected Autonomous Vehicles based\n on 3D Point Clouds", "Cooperative Perception for 3D Object Detection in Driving Scenarios\n using Infrastructure Sensors"], "answer_arxiv_id": ["1905.05265", "1912.12147"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_12845"} +{"question": "Could you provide me some studies about high-resolution 3D generation with careful training strategy?", "answer": ["EVA3D: Compositional 3D Human Generation from 2D Image Collections"], "answer_arxiv_id": ["2210.04888"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_12846"} +{"question": "Any works which pruned the LLMs relying only on the inference step?", "answer": ["Everybody Prune Now: Structured Pruning of LLMs with only Forward Passes"], "answer_arxiv_id": ["2402.05406"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_12847"} +{"question": "Which studies provide a comprehensive review of incremental learning on Euclidean data?", "answer": ["A continual learning survey: Defying forgetting in classification tasks", "Continual Lifelong Learning with Neural Networks: A Review", "Continual Lifelong Learning in Natural Language Processing: A Survey"], "answer_arxiv_id": ["1909.08383", "1802.07569", "2012.09823"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_12848"} +{"question": "Any works about diffusion models for perceptual tasks?", "answer": ["Unleashing Text-to-Image Diffusion Models for Visual Perception", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models", "MedSegDiff: Medical Image Segmentation with Diffusion Probabilistic Model"], "answer_arxiv_id": ["2303.02153v1", "2303.04803", "2211.00611"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_12849"} +{"question": "Which work applied a synchronized multiview diffusion model to capture the joint probability distribution of multiview images?", "answer": ["SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image"], "answer_arxiv_id": ["2309.03453"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12850"} +{"question": "Which publications have studied the problem when the observation of the learner is partial than that of the expert in POMDP?", "answer": ["Robust Asymmetric Learning in POMDPs"], "answer_arxiv_id": ["2012.15566v3"], "source_meta": {"published_time": "20210617"}, "qid": "AutoScholarQuery_train_12851"} +{"question": "Which work discusses MC-Dropout as a method to improve the confidence estimates of DNNs?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.02142"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_12852"} +{"question": "Which study introduced the MMLU benchmark for evaluating LLMs’ world knowledge and problem solving ability?", "answer": ["Measuring Massive Multitask Language Understanding"], "answer_arxiv_id": ["2009.03300"], "source_meta": {"published_time": "20230913"}, "qid": "AutoScholarQuery_train_12853"} +{"question": "Are there any research regarding the image harmonization methods that aim to rectify the illumination disparity between foreground and background?", "answer": ["Deep Image Harmonization", "Improving the Harmony of the Composite Image by Spatial-Separated\n Attention Module", "DoveNet: Deep Image Harmonization via Domain Verification", "BargainNet: Background-Guided Domain Translation for Image Harmonization", "High-Resolution Image Harmonization via Collaborative Dual\n Transformations"], "answer_arxiv_id": ["1703.00069", "1907.06406", "1911.13239", "2009.09169", "2109.06671"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_12854"} +{"question": "Who proposed the theory about resolution-limited scaling of learning curves that was referred to by ~\\cite{bib.bib5}?", "answer": ["Spectrum Dependent Learning Curves in Kernel Regression and Wide Neural Networks"], "answer_arxiv_id": ["2002.02561"], "source_meta": {"published_time": "20221223"}, "qid": "AutoScholarQuery_train_12855"} +{"question": "Any research that aimed to show the 'saturation effect' in KRR?", "answer": ["On the Saturation Effect of Kernel Ridge Regression", "Kernel interpolation generalizes poorly"], "answer_arxiv_id": ["2405.09362", "2303.15809"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_12856"} +{"question": "What works follows the procedure of query-based methods for crafting adversarial examples in black-box setting?", "answer": ["Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models", "HopSkipJumpAttack: A Query-Efficient Decision-Based Attack", "Subspace Attack: Exploiting Promising Subspaces for Query-Efficient Black-box Attacks", "ZOO: Zeroth Order Optimization Based Black-box Attacks to Deep Neural Networks without Training Substitute Models", "Black-box Adversarial Attacks with Limited Queries and Information", "Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors", "AutoZOOM: Autoencoder-based Zeroth Order Optimization Method for Attacking Black-box Neural Networks"], "answer_arxiv_id": ["1712.04248", "1904.02144", "1906.04392", "1708.03999", "1804.08598", "1807.07978", "1805.11770"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_12857"} +{"question": "What studies proposed alternative models based on linear dynamical systems?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces", "Diagonal State Spaces are as Effective as Structured State Spaces", "On the Parameterization and Initialization of Diagonal State Space Models"], "answer_arxiv_id": ["2111.00396", "2203.14343", "2206.11893"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_12858"} +{"question": "Can you identify a study that proposed the use of heat kernels to construct intrinsic Gaussian Processes on complex constrained domains?", "answer": ["Intrinsic Gaussian processes on complex constrained domains"], "answer_arxiv_id": ["1801.01061"], "source_meta": {"published_time": "20230116"}, "qid": "AutoScholarQuery_train_12859"} +{"question": "What papers propose methods to incorporate context in RL by simply concatenating the context to the state?", "answer": ["Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables", "Multi-Task Reinforcement Learning with Context-based Representations", "Block Contextual MDPs for Continual Learning", "Decomposed Mutual Information Optimization for Generalized Context in Meta-Reinforcement Learning"], "answer_arxiv_id": ["1903.08254", "2102.06177", "2110.06972", "2210.04209"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_12860"} +{"question": "In which research work is the penalization of a model discussed, if it does not prefer the right answers for the right reasons?", "answer": ["Right for the Right Reasons: Training Differentiable Models by Constraining their Explanations"], "answer_arxiv_id": ["1703.03717"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_12861"} +{"question": "Could you name some widely popular LLMs?", "answer": ["LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2302.13971"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_12862"} +{"question": "Could you provide me some works that utilize strategy networks or learn dynamic decisions based on gate functions?", "answer": ["BlockDrop: Dynamic Inference Paths in Residual Networks", "You Look Twice: GaterNet for Dynamic Filter Selection in CNNs", "SBNet: Sparse Blocks Network for Fast Inference"], "answer_arxiv_id": ["1711.08393", "1811.11205", "1801.02108"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_12863"} +{"question": "Which studies utilize uncertainty based approaches to estimate the epistemic uncertainty of Q-values or dynamics in Offline RL?", "answer": ["An Optimistic Perspective on Offline Reinforcement Learning", "Uncertainty Weighted Actor-Critic for Offline Reinforcement Learning", "Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble"], "answer_arxiv_id": ["1907.04543v4", "2105.08140", "2110.01548"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_12864"} +{"question": "What studies or projects are available for automatic audio captioning?", "answer": ["Clotho: An Audio Captioning Dataset"], "answer_arxiv_id": ["1910.09387"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_12865"} +{"question": "Which research papers detail learning pruning budget allocation techniques?", "answer": ["Soft Threshold Weight Reparameterization for Learnable Sparsity", "DSA: More Efficient Budgeted Pruning via Differentiable Sparsity Allocation"], "answer_arxiv_id": ["2002.03231", "2004.02164"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_12866"} +{"question": "What papers employ geometric techniques such as homography, epipolar geometry and SfM to refine optical flow?", "answer": ["Optical Flow with Semantic Segmentation and Localized Layers", "Exploiting Semantic Information and Deep Matching for Optical Flow", "Optical Flow in Mostly Rigid Scenes"], "answer_arxiv_id": ["1603.03911", "1604.01827", "1705.01352"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_12867"} +{"question": "Could you cite examples of studies that are about Generative Visual Dialogues where dialogues about images are generated?", "answer": ["Learning Cooperative Visual Dialog Agents with Deep Reinforcement Learning", "Improving Generative Visual Dialog by Answering Diverse Questions", "Enhancing Visual Dialog Questioner with Entity-based Strategy Learning and Augmented Guesser", "Modeling Explicit Concerning States for Reinforcement Learning in Visual Dialogue"], "answer_arxiv_id": ["1703.06585", "1909.10470", "2109.02297", "2107.05250"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_12868"} +{"question": "Which papers focus on constructing interpretable embeddings to enhance model interpretability?", "answer": ["Learning Interpretable Style Embeddings via Prompting LLMs"], "answer_arxiv_id": ["2305.12696"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_12869"} +{"question": "Which study developed Point·E that trains point-cloud diffusion models?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts"], "answer_arxiv_id": ["2212.08751"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_12870"} +{"question": "Which papers propose learning entire programs end to end with gradient descent in differentiable architectures?", "answer": ["Neural Turing Machines", "Reinforcement Learning Neural Turing Machines - Revised", "Learning Simple Algorithms from Examples"], "answer_arxiv_id": ["1410.5401", "1505.00521", "1511.07275"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_12871"} +{"question": "Which paper proposed the stochastic adjoint sensitivity method for the inference of latent SDEs?", "answer": ["Scalable Gradients for Stochastic Differential Equations"], "answer_arxiv_id": ["2001.01328v6"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_12872"} +{"question": "What researches proposed to accelerate the speed of training implicit layers with a Jacobian-free backpropagation method?", "answer": ["JFB: Jacobian-Free Backpropagation for Implicit Networks"], "answer_arxiv_id": ["2103.12803"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_12873"} +{"question": "Which studies used natural language processing techniques for ICD coding?", "answer": ["Explainable Prediction of Medical Codes from Clinical Text", "Multi-Label Classification of Patient Notes: Case Study on ICD Code Assignment"], "answer_arxiv_id": ["1802.05695", "1709.09587"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_12874"} +{"question": "Which papers have developed the Neural Radiance Fields (NeRF) for 3D reconstruction?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Neural Volumes: Learning Dynamic Renderable Volumes from Images"], "answer_arxiv_id": ["2003.08934", "1906.07751"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_12875"} +{"question": "Any works about the discarding outdated facts technique to keep models up-to-date in the QA domain?", "answer": ["Mitigating Temporal Misalignment by Discarding Outdated Facts"], "answer_arxiv_id": ["2305.14824"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_12876"} +{"question": "Which studies have contributed to the field of anisotropic randomized smoothing?", "answer": ["Certified Defense to Image Transformations via Randomized Smoothing", "ANCER: Anisotropic Certification via Sample-wise Volume Maximization"], "answer_arxiv_id": ["2002.12463", "2107.04570"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_12877"} +{"question": "Could you provide me with some Bayesian Optimization (BO) libraries?", "answer": ["BoTorch: A Framework for Efficient Monte-Carlo Bayesian Optimization", "Tuning Hyperparameters without Grad Students: Scalable and Robust Bayesian Optimisation with Dragonfly"], "answer_arxiv_id": ["1910.06403", "1903.06694"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_12878"} +{"question": "Could you give me examples of research that learned 3D representations?", "answer": ["STaR: Self-supervised Tracking and Reconstruction of Rigid Objects in Motion with Neural Rendering", "NeuralDiff: Segmenting 3D objects that move in egocentric videos"], "answer_arxiv_id": ["2101.01602", "2110.09936"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_12879"} +{"question": "What works use visual prompt tuning to adapt to VLMs?", "answer": ["Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Visual Prompt Tuning", "Exploring Visual Prompts for Adapting Large-Scale Models", "Visual Prompting via Image Inpainting"], "answer_arxiv_id": ["2210.04150", "2203.12119", "2203.17274", "2209.00647"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_12880"} +{"question": "What studies dealt with the insufficiency of the 1-hop scalarization implemented in ClofNet?", "answer": ["SE(3) Equivariant Graph Neural Networks with Complete Local Frames"], "answer_arxiv_id": ["2110.14811"], "source_meta": {"published_time": "20230407"}, "qid": "AutoScholarQuery_train_12881"} +{"question": "Which works demonstrated that the performance bottleneck of deep neural classifiers on long-tailed datasets is improper decision boundaries?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition", "Distribution Alignment: A Unified Framework for Long-tail Visual Recognition"], "answer_arxiv_id": ["1910.09217", "2103.16370"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_12882"} +{"question": "Could you tell me about the studies that have developed a general conformal prediction framework for non-exchangeable data?", "answer": ["Conformal Prediction Beyond Exchangeability"], "answer_arxiv_id": ["2202.13415"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_12883"} +{"question": "Which works use Sharpness-Aware Minimization in their training objective to mitigate forgetting?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization", "DSI++: Updating Transformer Memory with New Documents"], "answer_arxiv_id": ["2010.01412", "2212.09744"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_12884"} +{"question": "What papers demonstrate that Fine Tuning (FT) outperforms Head Probing (HP) when the pretrain and downstream tasks are very different?", "answer": ["Improved Baselines with Momentum Contrastive Learning", "A Large-scale Study of Representation Learning with the Visual Task Adaptation Benchmark", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2003.04297", "1910.04867", "2111.06377"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_12885"} +{"question": "What techniques have been used in neural style transfer to enhance the stylization through semantic correspondence?", "answer": ["Visual Attribute Transfer through Deep Image Analogy", "Cross-domain Correspondence Learning for Exemplar-based Image\n Translation", "Style Mixer: Semantic-aware Multi-Style Transfer Network"], "answer_arxiv_id": ["1705.01088", "2004.05571", "1910.13093"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_12886"} +{"question": "Can you mention the studies that explored the properties and effects of temperature in logit-based knowledge distillation methods?", "answer": ["Revisiting Label Smoothing and Knowledge Distillation Compatibility:\n What was Missing?", "Distilling the Knowledge in a Neural Network", "Meta Knowledge Distillation"], "answer_arxiv_id": ["2206.14532", "1503.02531", "2202.07940"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_12887"} +{"question": "Which papers developed clustering methods in global feature self-supervised learning?", "answer": ["Deep Clustering for Unsupervised Learning of Visual Features", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Online Deep Clustering for Unsupervised Representation Learning"], "answer_arxiv_id": ["1807.05520", "2006.09882", "2006.10645"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_12888"} +{"question": "Which studies demonstrated the use of LLMs in generating step-by-step explanations or rationales?", "answer": ["Rationalization for Explainable NLP: A Survey"], "answer_arxiv_id": ["2301.08912"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_train_12889"} +{"question": "Could you tell me about some variants of DP-SGD that adaptively change the clip norm to improve convergence?", "answer": ["Differentially Private Learning with Adaptive Clipping", "Dynamic Differential-Privacy Preserving SGD", "Adaptive Differentially Private Empirical Risk Minimization"], "answer_arxiv_id": ["1905.03871", "2111.00173", "2110.07435"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_12890"} +{"question": "Which works employed Vision Transformer-based models for object detection?", "answer": ["You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection", "End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2106.00666", "2005.12872", "2010.04159"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_12891"} +{"question": "Which papers discuss variational inference in (deep) state space models and its drawbacks?", "answer": ["Black box variational inference for state space models", "Composing graphical models with neural networks for structured representations and fast inference", "Structured Inference Networks for Nonlinear State Space Models"], "answer_arxiv_id": ["1511.07367", "1603.06277", "1609.09869v2"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_12892"} +{"question": "Any works about fine-tuning the LLM named LLaMA?", "answer": ["Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on\n Self-Chat Data", "Enhancing Chat Language Models by Scaling High-quality Instructional Conversations"], "answer_arxiv_id": ["2304.01196", "2305.14233v1"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_12893"} +{"question": "Could you name the studies that relate to the problem where the learner only has noisy access to gradients?", "answer": ["Non-stationary Stochastic Optimization", "Tracking Moving Agents via Inexact Online Gradient Descent Algorithm"], "answer_arxiv_id": ["1307.5449", "1710.05133v2"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_12894"} +{"question": "What are the papers that tried to encode context information for dense object counting?", "answer": ["Generating High-Quality Crowd Density Maps using Contextual Pyramid CNNs", "Spatiotemporal Modeling for Crowd Counting in Videos"], "answer_arxiv_id": ["1708.00953", "1707.07890"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_12895"} +{"question": "Who studied pseudomonotonicity beyond monotonicity?", "answer": ["Optimal stochastic extragradient schemes for pseudomonotone stochastic variational inequality problems and their variants"], "answer_arxiv_id": ["1410.1628"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_12896"} +{"question": "Which research works demonstrated the value of using external resources such as a morphological analyzer for diacritics?", "answer": ["Adversarial Multitask Learning for Joint Multi-Feature and Multi-Dialect\n Morphological Modeling", "Joint Diacritization, Lemmatization, Normalization, and Fine-Grained\n Morphological Tagging", "Morphosyntactic Tagging with Pre-trained Language Models for Arabic and\n its Dialects", "Camelira: An Arabic Multi-Dialect Morphological Disambiguator"], "answer_arxiv_id": ["1910.12702", "1910.02267", "2110.06852", "2211.16807"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_train_12897"} +{"question": "Which works focused on a robust bandit problem in a challenging setting with strong adversaries who could observe current actions before attacking rewards?", "answer": ["Stochastic Linear Bandits Robust to Adversarial Attacks", "Robust Stochastic Linear Contextual Bandits Under Adversarial Attacks", "Nearly Optimal Algorithms for Linear Contextual Bandits with Adversarial Corruptions", "Linear Contextual Bandits with Adversarial Corruptions", "Corruption-Robust Algorithms with Uncertainty Weighting for Nonlinear Contextual Bandits and Markov Decision Processes"], "answer_arxiv_id": ["2007.03285", "2106.02978v3", "2205.06811", "2110.12615", "2212.05949"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_12898"} +{"question": "What works prolonged the image-text contrastive method to broader spectrum?", "answer": ["Unified Contrastive Learning in Image-Text-Label Space", "Florence: A New Foundation Model for Computer Vision", "INTERN: A New Learning Paradigm Towards General Vision"], "answer_arxiv_id": ["2204.03610", "2111.11432", "2111.08687"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_12899"} +{"question": "Which works have proposed methods to generate synthetic datasets with fairness?", "answer": ["Fair Diffusion: Instructing Text-to-Image Generation Models on Fairness", "Auditing and Generating Synthetic Data with Controllable Trust\n Trade-offs", "Machine Learning for Synthetic Data Generation: A Review"], "answer_arxiv_id": ["2302.10893", "2304.10819", "2302.04062"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_12900"} +{"question": "Can you mention some studies that have explored visual prompts, often in the form of masks overlaid on images?", "answer": ["Exploring Visual Prompts for Adapting Large-Scale Models", "BlackVIP: Black-Box Visual Prompting for Robust Transfer Learning"], "answer_arxiv_id": ["2203.17274", "2303.14773"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_12901"} +{"question": "What studies describe self-supervised multimodal speech models?", "answer": ["Learning Audio-Visual Speech Representation by Masked Multimodal Cluster\n Prediction", "Jointly Learning Visual and Auditory Speech Representations from Raw\n Data", "VATLM: Visual-Audio-Text Pre-Training with Unified Masked Prediction for\n Speech Representation Learning"], "answer_arxiv_id": ["2201.02184", "2212.06246", "2211.11275"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_12902"} +{"question": "What research proposed the federated averaging algorithm (FedAVG) that the researcher used as baseline for implementing federated learning systems?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_12903"} +{"question": "Which works have discussed issues like catastrophic interference, capacity loss, and primacy bias in reinforcement learning?", "answer": ["Interference and Generalization in Temporal Difference Learning", "Understanding and Preventing Capacity Loss in Reinforcement Learning", "The Primacy Bias in Deep Reinforcement Learning", "On Catastrophic Interference in Atari 2600 Games"], "answer_arxiv_id": ["2003.06350", "2204.09560", "2205.07802", "2002.12499"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_12904"} +{"question": "I wonder which works have developed Optimization-based Image Vectorization Methods that utilize a layer-wise optimization framework?", "answer": ["Towards Layer-wise Image Vectorization", "SAMVG: A Multi-stage Image Vectorization Model with the Segment-Anything\n Model"], "answer_arxiv_id": ["2206.04655", "2311.05276"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_12905"} +{"question": "Which papers focus on the reconstruction of two interacting hands?", "answer": ["Interacting Attention Graph for Single Image Two-Hand Reconstruction", "3D Interacting Hand Pose Estimation by Hand De-occlusion and Removal", "Bringing Inputs to Shared Domains for 3D Interacting Hands Recovery in\n the Wild", "Decoupled Iterative Refinement Framework for Interacting Hands\n Reconstruction from a Single RGB Image", "MeMaHand: Exploiting Mesh-Mano Interaction for Single Image Two-Hand\n Reconstruction", "ACR: Attention Collaboration-based Regressor for Arbitrary Two-Hand\n Reconstruction", "Reconstructing Interacting Hands with Interaction Prior from Monocular\n Images"], "answer_arxiv_id": ["2203.09364", "2207.11061", "2303.13652", "2302.02410", "2303.15718", "2303.05938", "2308.14082"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_12906"} +{"question": "Could you provide me some works that design equivariant layers for various permutation symmetries?", "answer": ["Deep Models of Interactions Across Sets", "The general theory of permutation equivarant neural networks and higher order graph variational encoders", "On Learning Sets of Symmetric Elements"], "answer_arxiv_id": ["1803.02879", "2004.03990", "2002.08599"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_12907"} +{"question": "Name the work that attempted training BERT within 24 hours with similar limitations and served as a central point of comparison for BERT training with limited resources?", "answer": ["How to Train BERT with an Academic Budget"], "answer_arxiv_id": ["2104.07705"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_12908"} +{"question": "What prior works use pretrained vision-language models to learn richer vision features for downstream policies?", "answer": ["CLIPort: What and Where Pathways for Robotic Manipulation", "Learning Transferable Visual Models From Natural Language Supervision", "Transporter Networks: Rearranging the Visual World for Robotic Manipulation", "Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation"], "answer_arxiv_id": ["2109.12098", "2103.00020", "2010.14406", "2209.05451"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_12909"} +{"question": "What papers are about Superpixel-Based XAI?", "answer": ["Learning Deep Features for Discriminative Localization", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "RISE: Randomized Input Sampling for Explanation of Black-box Models"], "answer_arxiv_id": ["1512.04150", "1610.02391", "1806.07421"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_12910"} +{"question": "Any studies incorporated the second truncated Taylor method to the PF-ODE in diffusion models?", "answer": ["GENIE: Higher-Order Denoising Diffusion Solvers"], "answer_arxiv_id": ["2210.05475"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_12911"} +{"question": "What studies propose methods for partial identification when interventional queries are non-identifiable?", "answer": ["Bounds on Causal Effects and Application to High Dimensional Data"], "answer_arxiv_id": ["2106.12121"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_12912"} +{"question": "What papers describe assigning the image label to these proposals through multiple instance learning?", "answer": ["Weakly Supervised Deep Detection Networks", "Weakly Supervised Object Localization with Multi-fold Multiple Instance Learning", "C-MIL: Continuation Multiple Instance Learning for Weakly Supervised Object Detection"], "answer_arxiv_id": ["1511.02853", "1503.00949", "1904.05647"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_12913"} +{"question": "What study further develops the vision-language landscape by introducing a model capable of understanding and generation tasks?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2201.12086"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_12914"} +{"question": "Could you name the work that leveraged discretized activations of a masked language model pre-trained on audio to generate syntactically plausible speech or music?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation"], "answer_arxiv_id": ["2209.03143"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_12915"} +{"question": "Which works developed techniques for explainable AI such as Integrated Gradients and Shapley Additive Explanations?", "answer": ["Axiomatic Attribution for Deep Networks", "A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1703.01365", "1705.07874"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_12916"} +{"question": "Any works applied dynamic networks with parameter expansion in continual learning?", "answer": ["DER: Dynamically Expandable Representation for Class Incremental\n Learning", "FOSTER: Feature Boosting and Compression for Class-Incremental Learning", "A Model or 603 Exemplars: Towards Memory-Efficient Class-Incremental\n Learning", "DyTox: Transformers for Continual Learning with DYnamic TOken eXpansion", "Learning to Prompt for Continual Learning"], "answer_arxiv_id": ["2103.16788", "2204.04662", "2205.13218", "2111.11326", "2112.08654"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_12917"} +{"question": "What work proposes a kNN-based method for selecting the top candidates based on their cosine-similarity with the query?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?"], "answer_arxiv_id": ["2101.06804"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_12918"} +{"question": "Could you provide me some studies about Model Inversion that focus on deriving the training dataset based on a network parameter?", "answer": ["The Secret Revealer: Generative Model-Inversion Attacks Against Deep\n Neural Networks", "Do Gradient Inversion Attacks Make Federated Learning Unsafe?", "Knowledge-Enriched Distributional Model Inversion Attacks", "Re-thinking Model Inversion Attacks Against Deep Neural Networks", "See through Gradients: Image Batch Recovery via GradInversion"], "answer_arxiv_id": ["1911.07135", "2202.06924", "2010.04092", "2304.01669", "2104.07586"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_12919"} +{"question": "Could you provide me some examples of MLLMs?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Language Is Not All You Need: Aligning Perception with Language Models"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2304.08485", "2304.10592", "2302.14045"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_12920"} +{"question": "Could you provide me the researches which adopted the principle of pessimism in offline stochastic bandit problems?", "answer": ["On the Optimality of Batch Policy Optimization Algorithms", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism"], "answer_arxiv_id": ["2104.02293", "2103.12021v2"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_12921"} +{"question": "Can you name some recent advancements in deep learning methods for portrait relighting?", "answer": ["Learning Physics-guided Face Relighting under Directional Light", "Single Image Portrait Relighting", "Neural Light Transport for Relighting and View Synthesis", "Neural Video Portrait Relighting in Real-time via Consistency Modeling", "Learning to Relight Portrait Images via a Virtual Light Stage and\n Synthetic-to-Real Adaptation", "LightPainter: Interactive Portrait Relighting with Freehand Scribble", "DiFaReli: Diffusion Face Relighting"], "answer_arxiv_id": ["1906.03355", "1905.00824", "2008.03806", "2104.00484", "2209.10510", "2303.12950", "2304.09479"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_12922"} +{"question": "Could you provide me some studies about conformal prediction for time series data using randomization and ensembles?", "answer": ["Exact and Robust Conformal Inference Methods for Predictive Machine Learning With Dependent Data", "A general framework for multi-step ahead adaptive conformal heteroscedastic time series forecasting"], "answer_arxiv_id": ["1802.06300", "2207.14219"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_12923"} +{"question": "What paper introduced adaptors into the encoder and trained each task individually for various foreground segmentation tasks?", "answer": ["Explicit Visual Prompting for Universal Foreground Segmentations"], "answer_arxiv_id": ["2305.18476"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_12924"} +{"question": "Could you provide me some studies about visual relationship understanding in the context of human-object interaction?", "answer": ["Grounded Situation Recognition", "DRG: Dual Relation Graph for Human-Object Interaction Detection", "Spatially Conditioned Graphs for Detecting Human-Object Interactions", "Efficient Two-Stage Detection of Human-Object Interactions with a Novel\n Unary-Pairwise Transformer", "Exploring Structure-aware Transformer over Interaction Proposals for\n Human-Object Interaction Detection", "Mining the Benefits of Two-stage and One-stage HOI Detection", "HOTR: End-to-End Human-Object Interaction Detection with Transformers", "GEN-VLKT: Simplify Association and Enhance Interaction Understanding for\n HOI Detection", "FGAHOI: Fine-Grained Anchors for Human-Object Interaction Detection", "Relational Context Learning for Human-Object Interaction Detection"], "answer_arxiv_id": ["2003.12058", "2008.11714", "2012.06060", "2112.01838", "2206.06291", "2108.05077", "2104.13682", "2203.13954", "2301.04019", "2304.04997"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_12925"} +{"question": "In what papers was the concept of group-wise correlation introduced for building a more compact cost volume in deep stereo matching?", "answer": ["Group-wise Correlation Stereo Network"], "answer_arxiv_id": ["1903.04025"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_12926"} +{"question": "What work introduced the concept of FISH Mask?", "answer": ["Training Neural Networks with Fixed Sparse Masks"], "answer_arxiv_id": ["2111.09839"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_12927"} +{"question": "Which research has made an analytical survey of the data comprising large language models including personal web pages, social media accounts, etc?", "answer": ["Are Large Pre-Trained Language Models Leaking Your Personal Information?"], "answer_arxiv_id": ["2205.12628"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_12928"} +{"question": "Which works proposed value-based algorithms for linear MDPs?", "answer": ["Sample-Optimal Parametric Q-Learning Using Linearly Additive Features", "Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1902.04779", "1907.05388"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_12929"} +{"question": "Which work extended 3D Gaussians splatting to dynamic scene modeling by direct separate per-frame optimization?", "answer": ["Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis"], "answer_arxiv_id": ["2308.09713"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_12930"} +{"question": "Could you provide me some works about studying annotator group bias in crowdsourcing datasets?", "answer": ["Toward Annotator Group Bias in Crowdsourcing"], "answer_arxiv_id": ["2110.08038"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_12931"} +{"question": "What are some methods that proposed decoupling the two sets of categories for capturing discriminative representations?", "answer": ["Generalized Category Discovery with Decoupled Prototypical Network"], "answer_arxiv_id": ["2211.15115"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_12932"} +{"question": "Which works are foundations for the Self-Supervised Learning framework?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "A Simple Framework for Contrastive Learning of Visual Representations", "Speech2Vec: A Sequence-to-Sequence Framework for Learning Word Embeddings from Speech"], "answer_arxiv_id": ["1810.04805", "2002.05709", "1803.08976"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_12933"} +{"question": "What works have leveraged CLIP to train text-to-image models without text data?", "answer": ["CLIP-GEN: Language-Free Training of a Text-to-Image Generator with CLIP"], "answer_arxiv_id": ["2203.00386"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_12934"} +{"question": "Could you provide me some works on the new direction of LMM development that augment large language models (LLMs) with multimodal comprehension capabilities?", "answer": ["OPT: Open Pre-trained Transformer Language Models", "LLaMA: Open and Efficient Foundation Language Models", "Instruction Tuning with GPT-4"], "answer_arxiv_id": ["2205.01068", "2302.13971", "2304.03277"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_12935"} +{"question": "Which papers investigated training-stage defenses for backdoor attacks?", "answer": ["Backdoor Defense via Decoupling the Training Process"], "answer_arxiv_id": ["2202.03423"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_12936"} +{"question": "Could you provide me works about the effectiveness of U-Net's coarse-and-fine representation?", "answer": ["Defocus Deblurring Using Dual-Pixel Data", "Rethinking Coarse-to-Fine Approach in Single Image Deblurring", "DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better", "Uformer: A General U-Shaped Transformer for Image Restoration", "Dual Adversarial Network: Toward Real-world Noise Removal and Noise Generation", "Multi-Stage Progressive Image Restoration", "Plug-and-Play Image Restoration with Deep Denoiser Prior"], "answer_arxiv_id": ["2005.00305", "2108.05054", "1908.03826", "2106.03106", "2007.05946", "2102.02808", "2008.13751"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_12937"} +{"question": "Could you list the papers discussing the scaling issues with CNFs in high dimensions?", "answer": ["FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models"], "answer_arxiv_id": ["1810.01367"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_12938"} +{"question": "Which papers have worked on layout generation conditioned on element types in graphic layout generation?", "answer": ["Neural Design Network: Graphic Layout Generation with Constraints", "BLT: Bidirectional Layout Transformer for Controllable Layout Generation", "Constrained Graphic Layout Generation via Latent Optimization"], "answer_arxiv_id": ["1912.09421", "2112.05112", "2108.00871"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_12939"} +{"question": "Which papers focused on finding a sparse network by pruning pre-trained dense networks in deep supervised learning?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding", "Training Sparse Neural Networks", "To prune, or not to prune: exploring the efficacy of pruning for model compression"], "answer_arxiv_id": ["1510.00149", "1611.06694", "1710.01878"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_12940"} +{"question": "Which methods are proposed to alleviate over-smoothing in GNNs?", "answer": ["DropEdge: Towards Deep Graph Convolutional Networks on Node Classification", "Dirichlet Energy Constrained Learning for Deep Graph Neural Networks"], "answer_arxiv_id": ["1907.10903", "2107.02392"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_12941"} +{"question": "Which studies have proposed solutions to address the computational complexity in transformer inference?", "answer": ["Reformer: The Efficient Transformer", "Generating Long Sequences with Sparse Transformers", "Rethinking Attention with Performers"], "answer_arxiv_id": ["2001.04451", "1904.10509", "2009.14794"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_12942"} +{"question": "What studies have designed specific algorithms to generate grasping poses?", "answer": ["ArtiBoost: Boosting Articulated 3D Hand-Object Pose Estimation via\n Online Exploration and Synthesis", "AffordPose: A Large-scale Dataset of Hand-Object Interactions with\n Affordance-driven Hand Pose"], "answer_arxiv_id": ["2109.05488", "2309.08942"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_12943"} +{"question": "What studies explored the integration of both generative and extractive readers to further enhance system performance?", "answer": ["UnitedQA: A Hybrid Approach for Open Domain Question Answering", "R2-D2: A Modular Baseline for Open-Domain Question Answering"], "answer_arxiv_id": ["2101.00178", "2109.03502"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_12944"} +{"question": "Could you mention a recent work extending the self-consistency approach to compute the general Wasserstein gradient flow numerically?", "answer": ["Self-Consistent Velocity Matching of Probability Flows"], "answer_arxiv_id": ["2301.13737"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_12945"} +{"question": "What research papers discuss input prompt-based safeguards against jailbreak attacks?", "answer": ["Baseline Defenses for Adversarial Attacks Against Aligned Language\n Models", "Detecting Language Model Attacks with Perplexity", "Jailbreak and Guard Aligned Language Models with Only Few In-Context\n Demonstrations", "Intention Analysis Makes LLMs A Good Jailbreak Defender", "Certifying LLM Safety against Adversarial Prompting", "SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks"], "answer_arxiv_id": ["2309.00614", "2308.14132", "2310.06387", "2401.06561", "2309.02705", "2310.03684"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_12946"} +{"question": "What research has been conducted on application of Federated Learning in healthcare?", "answer": ["Multi-site fMRI Analysis Using Privacy-preserving Federated Learning and Domain Adaptation: ABIDE Results", "The Future of Digital Health with Federated Learning"], "answer_arxiv_id": ["2001.05647", "2003.08119v2"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_12947"} +{"question": "What work exploits multiple scorers to find ⟨hypernym, hyponym⟩ pairs for a given query concept?", "answer": ["Taxonomy Completion via Triplet Matching Network"], "answer_arxiv_id": ["2101.01896"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_12948"} +{"question": "Which works have been conducted on diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_12949"} +{"question": "Can you list some works that use implicit-based approaches for clothed human reconstruction?", "answer": ["PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization", "PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution\n 3D Human Digitization", "SMPLicit: Topology-aware Generative Model for Clothed People", "PaMIR: Parametric Model-Conditioned Implicit Representation for\n Image-based Human Reconstruction", "ICON: Implicit Clothed humans Obtained from Normals", "ECON: Explicit Clothed humans Optimized via Normal integration", "High-fidelity 3D Human Digitization from Single 2K Resolution Images", "ARCH: Animatable Reconstruction of Clothed Humans", "ARCH++: Animation-Ready Clothed Human Reconstruction Revisited", "D-IF: Uncertainty-aware Human Digitization via Implicit Distribution\n Field", "Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing", "High-Fidelity Clothed Avatar Reconstruction from a Single Image", "TeCH: Text-guided Reconstruction of Lifelike Clothed Humans", "X-Avatar: Expressive Human Avatars"], "answer_arxiv_id": ["1905.05172", "2004.00452", "2103.06871", "2007.03858", "2112.09127", "2212.07422", "2303.15108", "2004.04572", "2108.07845", "2308.08857", "2204.08906", "2304.03903", "2308.08545", "2303.04805"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_12950"} +{"question": "Which research presented that closed models underperformed state-of-the-art finetuned models on automatic metrics?", "answer": ["Using Large Language Models for Zero-Shot Natural Language Generation\n from Knowledge Graphs", "Evaluating Generative Models for Graph-to-Text Generation"], "answer_arxiv_id": ["2307.07312", "2307.14712"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_12951"} +{"question": "Could you provide some references that use the same modality for teacher and student models in the context of 3D object detection, specifically LO-to-LO (L2L) and CO-to-CO (C2C)?", "answer": ["PointDistiller: Structured Knowledge Distillation Towards Efficient and\n Compact 3D Detection", "LiDAR Distillation: Bridging the Beam-Induced Domain Gap for 3D Object\n Detection", "Object DGCNN: 3D Object Detection using Dynamic Graphs", "Towards Efficient 3D Object Detection with Knowledge Distillation", "Paint and Distill: Boosting 3D Object Detection with Semantic Passing\n Network", "Structured Knowledge Distillation Towards Efficient and Compact\n Multi-View 3D Detection"], "answer_arxiv_id": ["2205.11098", "2203.14956", "2110.06923", "2205.15156", "2207.05497", "2211.08398"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_12952"} +{"question": "Could you list studies that discuss methods to debias data using extra information?", "answer": ["Women also Snowboard: Overcoming Bias in Captioning Models", "Tell Me Where to Look: Guided Attention Inference Network", "Interpretations are Useful: Penalizing Explanations to Align Neural Networks with Prior Knowledge", "Why Can’t I Dance in the Mall? Learning to Mitigate Scene Bias in Action Recognition", "Learning Not to Learn: Training Deep Neural Networks with Biased Data", "Contrastive Examples for Addressing the Tyranny of the Majority", "Rewriting a Deep Generative Model", "Editing a classifier by rewriting its prediction rules", "On Guiding Visual Attention with Language Specification"], "answer_arxiv_id": ["1803.09797", "1802.10171", "1909.13584", "1912.05534", "1812.10352", "2004.06524", "2007.15646", "2112.01008", "2202.08926"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_12953"} +{"question": "Which studies showcased the impressive few-shot learning capabilities of large pre-trained language models like GPT-3?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Language Models are Few-Shot Learners", "SQA3D: Situated Question Answering in 3D Scenes", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering"], "answer_arxiv_id": ["1902.00751", "2005.14165", "2210.07474", "2209.09513"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_12954"} +{"question": "Could you provide me with any studies of different scheduling problems with predictions?", "answer": ["Scheduling with Predictions and the Price of Misprediction", "Permutation Predictions for Non-Clairvoyant Scheduling", "Scheduling with Speed Predictions", "Strategyproof Scheduling with Predictions", "Scheduling with Predictions", "Speed-Oblivious Online Scheduling: Knowing (Precise) Speeds is not Necessary"], "answer_arxiv_id": ["1902.00732", "2202.10199", "2205.01247", "2209.04058", "2212.10433", "2302.00985"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_12955"} +{"question": "Could you mention studies that have applied knowledge distillation in the context of federated learning?", "answer": ["Data-Free Knowledge Distillation for Heterogeneous Federated Learning"], "answer_arxiv_id": ["2105.10056"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_12956"} +{"question": "Could you provide me some references that have attempted to adoapt diffusion models for cross-modal retrieval?", "answer": ["DiffusionRet: Generative Text-Video Retrieval with Diffusion Model", "MomentDiff: Generative Video Moment Retrieval from Random to Real"], "answer_arxiv_id": ["2303.09867", "2307.02869"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_12957"} +{"question": "What papers directly learn PDE inverse operators from data?", "answer": ["Deep learning architectures for nonlinear operator functions and nonlinear inverse problems"], "answer_arxiv_id": ["1912.11090v3"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_12958"} +{"question": "Which works use transformers in 3D vision?", "answer": ["TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with\n Transformers", "CAPE: Camera View Position Embedding for Multi-View 3D Object Detection", "CenterFormer: Center-based Transformer for 3D Object Detection"], "answer_arxiv_id": ["2203.11496", "2303.10209", "2209.05588"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_12959"} +{"question": "Are there any studies that improved the slot attention module approach by using fixed points as object representations?", "answer": ["Object Representations as Fixed Points: Training Iterative Refinement Algorithms with Implicit Differentiation"], "answer_arxiv_id": ["2207.00787"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_12960"} +{"question": "Could you provide me some studies that deal with potential Markov games and two-player zero-sum games using independent policy gradient algorithms?", "answer": ["Independent Policy Gradient Methods for Competitive Reinforcement Learning", "Decentralized Q-Learning in Zero-sum Markov Games", "Last-iterate Convergence of Decentralized Optimistic Gradient Descent/Ascent in Infinite-horizon Competitive Markov Games", "Global Convergence of Multi-Agent Policy Gradient in Markov Potential Games", "Independent Policy Gradient for Large-Scale Markov Potential Games: Sharper Rates, Function Approximation, and Game-Agnostic Convergence"], "answer_arxiv_id": ["2101.04233v1", "2106.02748v2", "2102.04540", "2106.01969", "2202.04129"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_12961"} +{"question": "Which works propose the method of incorporating position information via advanced positional encoding?", "answer": ["Attention Is All You Need", "Improve Transformer Models with Better Relative Position Embeddings", "Rethinking Positional Encoding in Language Pre-training"], "answer_arxiv_id": ["1706.03762", "2009.13658", "2006.15595"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_12962"} +{"question": "Which research papers are about multiple-choice task in video question answering?", "answer": ["TVQA: Localized, Compositional Video Question Answering", "HERO: Hierarchical Encoder for Video+Language Omni-representation\n Pre-training"], "answer_arxiv_id": ["1809.01696", "2005.00200"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_12963"} +{"question": "What studies have showcased successful use of the contrastive loss?", "answer": ["A Metric Learning Reality Check"], "answer_arxiv_id": ["2003.08505"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_12964"} +{"question": "Which works study the approximation and robustness properties of Lipschitz neural networks?", "answer": ["Pay attention to your loss: understanding misconceptions about 1-Lipschitz neural networks"], "answer_arxiv_id": ["2104.05097"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_12965"} +{"question": "What studies argue that generation can happen deterministically to predict one likely future pose of a human?", "answer": ["Recurrent Network Models for Human Dynamics", "On human motion prediction using recurrent neural networks", "Learning Trajectory Dependencies for Human Motion Prediction", "History Repeats Itself: Human Motion Prediction via Motion Attention"], "answer_arxiv_id": ["1508.00271", "1705.02445", "1908.05436", "2007.11755"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_12966"} +{"question": "Can you name the research works that focus on equipping LLMs with memory capabilities?", "answer": ["Automatically Correcting Large Language Models: Surveying the landscape\n of diverse self-correction strategies", "MemoryBank: Enhancing Large Language Models with Long-Term Memory"], "answer_arxiv_id": ["2308.03188", "2305.10250"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_12967"} +{"question": "Which papers discuss the use of the multiple dispatch functional programming paradigm popularized by Julia?", "answer": ["Julia: A fresh approach to numerical computing"], "answer_arxiv_id": ["1411.1607"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_12968"} +{"question": "What works generate a 3D volume by scattering pixel features to the 3D grid and estimating occupancy probability?", "answer": ["Learning a Multi-View Stereo Machine", "MVSNet: Depth Inference for Unstructured Multi-view Stereo", "Cost Volume Pyramid Based Depth Inference for Multi-View Stereo", "Stereo Matching by Training a Convolutional Neural Network to Compare Image Patches", "Recurrent MVSNet for High-resolution Multi-view Stereo Depth Inference"], "answer_arxiv_id": ["1708.05375", "1804.02505", "1912.08329", "1510.05970", "1902.10556"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_12969"} +{"question": "Which works discuss the usage of vision-language pretraining (VLP) in learning universal representations for texts and images?", "answer": ["Multimodal Research in Vision and Language: A Review of Current and Emerging Trends", "A Survey on Visual Transformer", "Transformers in Vision: A Survey", "A Survey of Vision-Language Pre-Trained Models"], "answer_arxiv_id": ["2010.09522", "2012.12556", "2101.01169", "2202.10936"], "source_meta": {"published_time": "20220901"}, "qid": "AutoScholarQuery_train_12970"} +{"question": "Which research papers cover the usage of LLMs in assessing the factuality of generated text?", "answer": ["AlignScore: Evaluating Factual Consistency with a Unified Alignment\n Function"], "answer_arxiv_id": ["2305.16739"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_12971"} +{"question": "What studies introduced a new low-switching approach in linear MDPs?", "answer": ["A Provably Efficient Algorithm for Linear Markov Decision Process with Low Switching Cost", "Provably Efficient Reinforcement Learning with Linear Function Approximation under Adaptivity Constraints"], "answer_arxiv_id": ["2101.00494v1", "2101.02195"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_12972"} +{"question": "Which study uses a simple linear layer to project image features as one of the projection-based methods?", "answer": ["Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_12973"} +{"question": "Which papers emphasized the importance of augmentations and permutation-invariant architectures for processing neural field weights?", "answer": ["Equivariant Architectures for Learning in Deep Weight Spaces", "Permutation Equivariant Neural Functionals"], "answer_arxiv_id": ["2301.12780", "2302.14040"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_12974"} +{"question": "Can you mention research where the effect of viewport sequences generated in multiple observers’ browsing process on the perception of the ODI quality has been studied?", "answer": ["Perceptual Quality Assessment of Omnidirectional Images as Moving Camera Videos", "Perceptual Quality Assessment of Omnidirectional Images"], "answer_arxiv_id": ["2005.10547", "2207.02674"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_12975"} +{"question": "Which papers proposed methods about Long-Tailed Learning that consider the extremely skewed distribution?", "answer": ["Class-Balanced Loss Based on Effective Number of Samples", "Contrastive Learning with Boosted Memorization", "SMOTE: Synthetic Minority Over-sampling Technique"], "answer_arxiv_id": ["1901.05555", "2205.12693", "1106.1813"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_12976"} +{"question": "Could you provide me some studies on scaling up kernel algorithms using Random feature (RF) map methods?", "answer": ["Random Fourier Features for Kernel Ridge Regression: Approximation Bounds and Statistical Guarantees"], "answer_arxiv_id": ["1804.09893v2"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_12977"} +{"question": "What papers have explored the application of traditional frequency-domain analysis techniques in Deep Neural Networks?", "answer": ["A Fourier Perspective on Model Robustness in Computer Vision", "High Frequency Component Helps Explain the Generalization of\n Convolutional Neural Networks", "Global Filter Networks for Image Classification", "Fourier Neural Operator for Parametric Partial Differential Equations", "Adaptive Fourier Neural Operators: Efficient Token Mixers for\n Transformers", "Adaptive Frequency Filters As Efficient Global Token Mixers", "Deep Frequency Filtering for Domain Generalization", "SPANet: Frequency-balancing Token Mixer using Spectral Pooling\n Aggregation Modulation", "How Do Vision Transformers Work?"], "answer_arxiv_id": ["1906.08988", "1905.13545", "2107.00645v2", "2010.08895", "2111.13587", "2307.14008", "2203.12198", "2308.11568", "2202.06709"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_12978"} +{"question": "Which papers are associated with the study of Tabular RL?", "answer": ["Sample Complexity of Episodic Fixed-Horizon Reinforcement Learning", "Minimax Regret Bounds for Reinforcement Learning", "Tighter Problem-Dependent Regret Bounds in Reinforcement Learning without Domain Knowledge using Value Function Bounds", "Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition"], "answer_arxiv_id": ["1510.08906", "1703.05449", "1901.00210", "2004.10019"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_12979"} +{"question": "What works have utilised insertable differentiable siamese augmentation to improve dataset distillation performance?", "answer": ["Dataset Condensation with Differentiable Siamese Augmentation"], "answer_arxiv_id": ["2102.08259"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_12980"} +{"question": "Which research paper initially proposed the Physics-Informed Neural Networks (PINNs)?", "answer": ["Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equations"], "answer_arxiv_id": ["1711.10561"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_12981"} +{"question": "What research projects have proposed integrating reinforcement learning with diffusion models to enhance the generation speed?", "answer": ["Controllable Motion Diffusion Model"], "answer_arxiv_id": ["2306.00416"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_12982"} +{"question": "Which works propose to learn a unified embedding space for image, text, and point cloud?", "answer": ["ULIP: Learning a Unified Representation of Language, Images, and Point\n Clouds for 3D Understanding", "ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding", "OpenShape: Scaling Up 3D Shape Representation Towards Open-World\n Understanding"], "answer_arxiv_id": ["2212.05171", "2305.08275", "2305.10764"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_12983"} +{"question": "Which works use the pretraining-then-finetune pipeline for multilingual automatic speech recognition (ASR)?", "answer": ["XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations", "HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units"], "answer_arxiv_id": ["2111.09296", "2006.11477", "2106.07447"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_12984"} +{"question": "What papers are about enabling Decision Transformer to explore online and apply it in an online reinforcement learning setting?", "answer": ["Online Decision Transformer"], "answer_arxiv_id": ["2202.05607"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_12985"} +{"question": "Could you provide some studies where researchers are increasingly adopting model architectures that target individual annotators?", "answer": ["Dealing with Disagreements: Looking Beyond the Majority Vote in Subjective Annotations"], "answer_arxiv_id": ["2110.05719"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_12986"} +{"question": "Which research papers deal with semi-supervised object detection?", "answer": ["Unbiased Teacher for Semi-Supervised Object Detection", "End-to-End Semi-Supervised Object Detection with Soft Teacher", "Label Matching Semi-Supervised Object Detection"], "answer_arxiv_id": ["2102.09480", "2106.09018", "2206.06608"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_12987"} +{"question": "Which papers discussed the incorporation of data augmentation and consistency regularization in semi-supervised learning?", "answer": ["Learning with Pseudo-Ensembles", "Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning", "Temporal Ensembling for Semi-Supervised Learning", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "MixMatch: A Holistic Approach to Semi-Supervised Learning"], "answer_arxiv_id": ["1412.4864", "1606.04586", "1610.02242v3", "2001.07685v2", "1905.02249"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_12988"} +{"question": "What papers present sophisticated reasoning architectures beyond simple prompting in LLMs?", "answer": ["Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning", "STaR: Bootstrapping Reasoning With Reasoning", "Faithful Reasoning Using Large Language Models", "Show Your Work: Scratchpads for Intermediate Computation with Language Models"], "answer_arxiv_id": ["2205.09712", "2203.14465", "2208.14271", "2112.00114"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_12989"} +{"question": "What studies are about the masked autoencoder's improved generalization and performance in downstream tasks?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "SimMIM: A Simple Framework for Masked Image Modeling", "What to Hide from Your Students: Attention-Guided Masked Image Modeling", "Revealing the Dark Secrets of Masked Image Modeling", "On Data Scaling in Masked Image Modeling", "Understanding Masked Image Modeling via Learning Occlusion Invariant\n Feature", "Self-Supervised Pre-Training of Swin Transformers for 3D Medical Image\n Analysis"], "answer_arxiv_id": ["2111.06377", "2111.09886", "2203.12719", "2205.13543", "2206.04664", "2208.04164", "2111.14791"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_12990"} +{"question": "Which studies have LVLMs adopted frozen Language Language Models (LLMs) as the language component?", "answer": ["Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2304.08485", "2304.10592"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_12991"} +{"question": "Can you cite a work where curriculum learning was performed at the group level?", "answer": ["Let the Model Decide its Curriculum for Multitask Learning"], "answer_arxiv_id": ["2205.09898"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_12992"} +{"question": "What works discuss the methods for full-body motion capture, integrating the estimation of body, hands, and face models?", "answer": ["Monocular Expressive Body Regression through Body-Driven Attention", "FrankMocap: A Monocular 3D Whole-Body Pose Estimation System via\n Regression and Integration", "Monocular Real-time Full Body Capture with Inter-part Correlations", "Collaborative Regression of Expressive Bodies using Moderation", "Accurate 3D Hand Pose Estimation for Whole-Body 3D Human Mesh Estimation", "PyMAF-X: Towards Well-aligned Full-body Model Regression from Monocular\n Images"], "answer_arxiv_id": ["2008.09062", "2108.06428", "2012.06087", "2105.05301", "2011.11534", "2207.06400"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_12993"} +{"question": "Are there any works that focus on learning latent DAG models?", "answer": ["Anchored Discrete Factor Analysis", "Anchored Causal Inference in the Presence of Measurement Error", "Generalized Independent Noise Condition for Estimating Latent Variable Causal Graphs"], "answer_arxiv_id": ["1511.03299", "1906.00928", "2010.04917"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_12994"} +{"question": "Which works discuss latent models designed for RL policy optimization?", "answer": ["Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Model", "SOLAR: Deep Structured Representations for Model-Based Reinforcement Learning", "Model-Based Reinforcement Learning via Latent-Space Collocation", "Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1907.00953", "1808.09105", "2106.13229", "1811.04551", "1912.01603", "2010.02193"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_12995"} +{"question": "Which works explore the inclusion of depth data in 6-DoF pose estimation?", "answer": ["EPOS: Estimating 6D Pose of Objects with Symmetries", "Coupled Iterative Refinement for 6D Multi-Object Pose Estimation", "ZebraPose: Coarse to Fine Surface Encoding for 6DoF Object Pose\n Estimation", "SurfEmb: Dense and Continuous Correspondence Distributions for Object\n Pose Estimation with Learnt Surface Embeddings"], "answer_arxiv_id": ["2004.00605", "2204.12516", "2203.09418", "2111.13489"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_12996"} +{"question": "What paper introduced the CLIP model?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_12997"} +{"question": "Which papers proposed different variants of ConvRNNs?", "answer": ["Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model", "Stochastic Variational Video Prediction", "PredRNN: A Recurrent Neural Network for Spatiotemporal Predictive Learning", "PredRNN++: Towards A Resolution of the Deep-in-Time Dilemma in Spatiotemporal Predictive Learning", "Convolutional Tensor-Train LSTM for Spatio-temporal Learning", "ContextVP: Fully Context-Aware Video Prediction"], "answer_arxiv_id": ["1706.03458", "1710.11252", "2103.09504", "1804.06300", "2002.09131", "1710.08518"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_12998"} +{"question": "What papers presented instruction-following benchmarks that include a sub-set of test cases relevant to format following?", "answer": ["Instruction-Following Evaluation for Large Language Models", "Benchmarking Large Language Models on Controllable Generation under\n Diversified Instructions"], "answer_arxiv_id": ["2311.07911", "2401.00690"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_12999"} +{"question": "What studies propose a diversity metric based on Determinantal Point Process (DPP)?", "answer": ["Effective Diversity in Population Based Reinforcement Learning"], "answer_arxiv_id": ["2002.00632"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_13000"} +{"question": "Who has done research on generating 3D avatars from text inputs?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars"], "answer_arxiv_id": ["2205.08535"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_13001"} +{"question": "Could you provide me some studies discussing the improvements in the field of DeepSDF since its introduction?", "answer": ["SAL: Sign Agnostic Learning of Shapes from Raw Data", "SALD: Sign Agnostic Learning with Derivatives", "Implicit Neural Representations with Periodic Activation Functions", "Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains", "Implicit Geometric Regularization for Learning Shapes", "DiGS : Divergence guided shape implicit neural representation for unoriented point clouds"], "answer_arxiv_id": ["1911.10414", "2006.05400", "2006.09661", "2006.10739", "2002.10099", "2106.10811"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_13002"} +{"question": "Which paper builds a Text-to-SQL benchmark, evaluating the ability to write broad-domain SQL programs?", "answer": ["Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task"], "answer_arxiv_id": ["1809.08887"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_13003"} +{"question": "What papers worked on k-Nearest Neighbor Language Models (kNN-LMs) in the context of Retrieval-Augmented Language Models (RALMs)?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models", "Training Language Models with Memory Augmentation"], "answer_arxiv_id": ["1911.00172", "2205.12674"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_13004"} +{"question": "What papers are about the method in diffusion-based SR that integrates a low-resolution image directly into the input of existing diffusion models?", "answer": ["Image Super-Resolution via Iterative Refinement", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2104.07636", "2112.10752"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_13005"} +{"question": "Could you ascribe the works that use template-based algorithms for single-step retrosynthesis prediction?", "answer": ["Retro*: Learning Retrosynthetic Planning with Neural Guided A* Search", "Retrosynthesis Prediction with Conditional Graph Logic Network"], "answer_arxiv_id": ["2006.15820", "2001.01408"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13006"} +{"question": "Which paper provides a standard paradigm for training multi-agent reinforcement agents in Dec-POMDPs?", "answer": ["On the Utility of Learning about Humans for Human-AI Coordination"], "answer_arxiv_id": ["1910.05789"], "source_meta": {"published_time": "20220129"}, "qid": "AutoScholarQuery_train_13007"} +{"question": "Which papers proposed to update the weights based on the directional gradient along a random or learned perturbation direction?", "answer": ["Gradients without Backpropagation"], "answer_arxiv_id": ["2202.08587"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_13008"} +{"question": "What work provided a multi-armed bandit problem with context dependent reward distributions?", "answer": ["Contextual Bandits with Latent Confounders: An NMF Approach"], "answer_arxiv_id": ["1606.00119"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_13009"} +{"question": "Which works employ Deformable Attention methods in 3D scene completion?", "answer": ["Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2010.04159"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_13010"} +{"question": "Which researches traced back to plug-and-play controllable text generation methods?", "answer": ["Plug and Play Language Models: a Simple Approach to Controlled Text Generation", "GeDi: Generative Discriminator guided Sequence Generation", "Diffusion-LM Improves Controllable Text Generation"], "answer_arxiv_id": ["1912.02164", "2009.06367", "2205.14217"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_13011"} +{"question": "What reference provides a comprehensive review of the field up to 2020?", "answer": ["A Survey of Privacy Attacks in Machine Learning"], "answer_arxiv_id": ["2007.07646"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_13012"} +{"question": "What papers have been published about learning invariant or diverse features?", "answer": ["Invariant Risk Minimization", "Model Patching: Closing the Subgroup Performance Gap with Data Augmentation", "Rich Feature Construction for the Optimization-Generalization Dilemma", "Controlling directions orthogonal to a classifier"], "answer_arxiv_id": ["1907.02893", "2008.06775", "2203.15516", "2201.11259"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_13013"} +{"question": "Could you provide me some studies about providing recourse to individuals negatively impacted by model predictions?", "answer": ["Interpretable Predictions of Tree-based Ensembles via Actionable Feature Tweaking", "Inverse Classification for Comparison-based Interpretability in Machine Learning", "Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR", "Actionable Recourse in Linear Classification", "Interpretable Counterfactual Explanations Guided by Prototypes", "Learning Model-Agnostic Counterfactual Explanations for Tabular Data", "Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers", "Explaining Machine Learning Classifiers through Diverse Counterfactual Explanations", "Model-Agnostic Counterfactual Explanations for Consequential Decisions", "Algorithmic recourse under imperfect causal knowledge: a probabilistic approach", "Multi-Objective Counterfactual Explanations", "Getting a CLUE: A Method for Explaining Uncertainty Estimates", "Counterfactual Explanations for Arbitrary Regression Models"], "answer_arxiv_id": ["1706.06691", "1712.08443", "1711.00399", "1809.06514", "1907.02584", "1910.09398", "1912.03277", "1905.07697", "1905.11190", "2006.06831", "2004.11165v2", "2006.06848", "2106.15212"], "source_meta": {"published_time": "20220313"}, "qid": "AutoScholarQuery_train_13014"} +{"question": "Any studies demonstrate the properties of disentangled representations associated with larger values of β in VAE?", "answer": ["Information Dropout: Learning Optimal Representations Through Noisy Computation"], "answer_arxiv_id": ["1611.01353v3"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_13015"} +{"question": "What work uses a single reference image and learns a warping between the LQ and reference image in the restoration process?", "answer": ["Learning Warped Guidance for Blind Face Restoration"], "answer_arxiv_id": ["1804.04829"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_13016"} +{"question": "Is there any research about using DM on the latent space to reduce the complexity of iteration steps?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_13017"} +{"question": "Could you provide some research exploring the notion of replicability in learning algorithms?", "answer": ["Reproducibility in Learning"], "answer_arxiv_id": ["2201.08430v2"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_13018"} +{"question": "Which papers introduced the task of 3D occupancy prediction?", "answer": ["Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction", "Semantic Scene Completion from a Single Depth Image"], "answer_arxiv_id": ["2302.07817", "1611.08974"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_13019"} +{"question": "Which research tackled the problems of lack of stability and struggle to capture long-term dependencies in recurrent models via the recent S444 layer and a stable implementation of the actor-critic mechanism?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces"], "answer_arxiv_id": ["2111.00396"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13020"} +{"question": "Which works have applied diffusion models to 3D point cloud generation?", "answer": ["Diffusion Probabilistic Models for 3D Point Cloud Generation"], "answer_arxiv_id": ["2103.01458"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_13021"} +{"question": "What studies are focused on developing graph neural networks for arbitrary geometries that are used for neural simulation?", "answer": ["Learning to Simulate Complex Physics with Graph Networks", "Interaction Networks for Learning about Objects, Relations and Physics", "A Compositional Object-Based Approach to Learning Physical Dynamics", "Learning Mesh-Based Simulation with Graph Networks", "Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids", "Message Passing Neural PDE Solvers", "Physics-Embedded Neural Networks: Graph Neural PDE Solvers with Mixed Boundary Conditions", "Learning rigid dynamics with face interaction graph networks", "Learning Controllable Adaptive Simulation for Multi-resolution Physics"], "answer_arxiv_id": ["2002.09405", "1612.00222", "1612.00341", "2010.03409", "1810.01566", "2202.03376", "2205.11912", "2212.03574", "2305.01122"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13022"} +{"question": "Which studies have proposed methods to overcome the over-smoothing problem in DGNs?", "answer": ["Simple and Deep Graph Convolutional Networks", "Dirichlet Energy Constrained Learning for Deep Graph Neural Networks", "pathGCN: Learning General Graph Spatial Operators from Paths"], "answer_arxiv_id": ["2007.02133", "2107.02392", "2207.07408"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_13023"} +{"question": "What studies have attempted to explain a DNN by distilling the DNN into another interpretable model?", "answer": ["Distilling a Neural Network Into a Soft Decision Tree", "Beyond Sparsity: Tree Regularization of Deep Models for Interpretability", "Interpreting CNN Knowledge Via An Explanatory Graph", "Explainable Neural Networks based on Additive Index Models"], "answer_arxiv_id": ["1711.09784", "1711.06178", "1708.01785", "1806.01933"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_13024"} +{"question": "What studies focus on object-centric 360-degree rendering for NVS?", "answer": ["Vision Transformer for NeRF-Based View Synthesis from a Single Input\n Image"], "answer_arxiv_id": ["2207.05736"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_13025"} +{"question": "Which research papers employed GNNs in action localization for video understanding?", "answer": ["Graph Convolutional Networks for Temporal Action Localization", "Stacked Spatio-Temporal Graph Convolutional Networks for Action\n Segmentation", "Action Graphs: Weakly-supervised Action Localization with Graph\n Convolution Networks"], "answer_arxiv_id": ["1909.03252", "1811.10575", "2002.01449"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_13026"} +{"question": "Could you provide me some works about deep feature reweighting (DFR)?", "answer": ["Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations", "On Feature Learning in the Presence of Spurious Correlations"], "answer_arxiv_id": ["2204.02937", "2210.11369"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_13027"} +{"question": "Could you show me some research papers that focused on the non-iterative computational processes to parse disentangled attributes for objects?", "answer": ["Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition"], "answer_arxiv_id": ["1603.08575", "2001.02407"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_13028"} +{"question": "Among the current research, which works established properties of minimizers to surrogate adversarial risks?", "answer": ["On The Existence of The Adversarial Bayes Classifier (Extended Version)", "The Many Faces of Adversarial Risk", "The Geometry of Adversarial Training in Binary Classification"], "answer_arxiv_id": ["2112.01694", "2201.08956v1", "2111.13613"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_13029"} +{"question": "Any research in policy learning from algorithmic recommendation but without focusing on fairness?", "answer": ["Safe Policy Learning through Extrapolation: Application to Pre-trial Risk Assessment"], "answer_arxiv_id": ["2109.11679v3"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_13030"} +{"question": "What works have aimed to accelerate the convergence and alleviate the data-hungry problem of vision transformer by using convolution layers for patch splitting?", "answer": ["Early Convolutions Help Transformers See Better", "Escaping the Big Data Paradigm with Compact Transformers"], "answer_arxiv_id": ["2106.14881", "2104.05704"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_13031"} +{"question": "Which studies have attempted to learn shared speech and text representation jointly, especially in self-supervised learning scenarios?", "answer": ["SpeechLM: Enhanced Speech Pre-Training with Unpaired Textual Data", "TESSP: Text-Enhanced Self-Supervised Speech Pre-training", "token2vec: A Joint Self-Supervised Pre-training Framework Using Unpaired Speech and Text", "Self-Supervised Audio-and-Text Pre-training with Extremely Low-Resource Parallel Data", "Towards reducing the Need for Speech Training Data To Build Spoken Language Understanding Systems"], "answer_arxiv_id": ["2209.15329", "2211.13443", "2210.16755", "2204.04645", "2203.00006"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_13032"} +{"question": "What papers discuss the use of Fourier Neural Operators in solving PDEs?", "answer": ["Neural Operator: Learning Maps Between Function Spaces With Applications to PDEs"], "answer_arxiv_id": ["2108.08481"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_13033"} +{"question": "Which studies expanded the analysis of a maximum-margin optimization problem to all positive-homogeneous networks?", "answer": ["Lexicographic and Depth-Sensitive Margins in Homogeneous and Non-Homogeneous Deep Models", "Gradient Descent Maximizes the Margin of Homogeneous Neural Networks"], "answer_arxiv_id": ["1905.07325", "1906.05890"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_13034"} +{"question": "Could you mention some studies that focused on improving various aspects of diffusion models such as architecture and sampling speed?", "answer": ["eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers", "Diffusion Models Beat GANs on Image Synthesis", "Elucidating the Design Space of Diffusion-Based Generative Models", "Scalable Diffusion Models with Transformers", "Diffusion Probabilistic Model Made Slim", "On Fast Sampling of Diffusion Probabilistic Models", "Progressive Distillation for Fast Sampling of Diffusion Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2211.01324", "2105.05233", "2206.00364", "2212.09748", "2211.17106", "2106.00132", "2202.00512", "2010.02502"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_13035"} +{"question": "Which paper discussed the geometric aspects of neural collapse for classification tasks?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning training"], "answer_arxiv_id": ["2008.08186"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_13036"} +{"question": "What papers involve in the research of motif-level methods in pretraining?", "answer": ["Self-Supervised Graph Transformer on Large-Scale Molecular Data", "Motif-Driven Contrastive Learning of Graph Representations", "Motif-based Graph Self-Supervised Learning for Molecular Property Prediction"], "answer_arxiv_id": ["2007.02835", "2012.12533", "2110.00987"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_13037"} +{"question": "What works focus on addressing the unique challenges of conversational speech parsing?", "answer": ["Improving Disfluency Detection by Self-Training a Self-Attentive Model", "Parsing Speech: A Neural Approach to Integrating Lexical and\n Acoustic-Prosodic Information", "Assessing the Use of Prosody in Constituency Parsing of Imperfect Transcripts"], "answer_arxiv_id": ["2004.05323", "1704.07287", "2106.07794v1"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_13038"} +{"question": "Which researches illustrate the development of machine-generated multimodal instruction datasets?", "answer": ["Visual Instruction Tuning", "SVIT: Scaling up Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "LAMM: Language-Assisted Multi-Modal Instruction-Tuning Dataset,\n Framework, and Benchmark", "StableLLaVA: Enhanced Visual Instruction Tuning with Synthesized\n Image-Dialogue Data", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2304.08485", "2307.04087", "2304.10592", "2306.06687", "2308.10253", "2304.14178"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_13039"} +{"question": "What studies explore the gradient domination property?", "answer": ["Global Optimality Guarantees For Policy Gradient Methods"], "answer_arxiv_id": ["1906.01786"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_13040"} +{"question": "What works proposed common practices like creating synthesizing datasets and adding knowledge to the prompts to augment PTLMs?", "answer": ["LIME: Learning Inductive Bias for Primitives of Mathematical Reasoning", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Show Your Work: Scratchpads for Intermediate Computation with Language Models"], "answer_arxiv_id": ["2101.06223", "2201.11903", "2112.00114"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_13041"} +{"question": "Can you point to some works on segmentation utilizing vision foundation models?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "Segment Everything Everywhere All at Once"], "answer_arxiv_id": ["2401.14159", "2304.06718"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_13042"} +{"question": "Any works that incorporate the use of DANN or SagNet for domain discrimination?", "answer": ["Domain-Adversarial Training of Neural Networks", "Reducing Domain Gap by Reducing Style Bias"], "answer_arxiv_id": ["1505.07818", "1910.11645"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_13043"} +{"question": "Which researchers analyzed reasoning capabilities of natural language models for mathematical tasks?", "answer": ["Analysing Mathematical Reasoning Abilities of Neural Models", "Deep learning for symbolic mathematics"], "answer_arxiv_id": ["1904.01557", "1912.01412"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_13044"} +{"question": "Which papers conducted research on the influence of adversarial training on model robustness?", "answer": ["Intriguing properties of neural networks", "Evasion Attacks against Machine Learning at Test Time", "Towards Evaluating the Robustness of Neural Networks", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1312.6199", "1708.06131", "1608.04644", "1706.06083"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_13045"} +{"question": "In what works is stochastic gradient descent extended in the distinct problem setting of compositional optimization?", "answer": ["Solving Stochastic Compositional Optimization is Nearly as Easy as Solving Stochastic Optimization"], "answer_arxiv_id": ["2008.10847"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_13046"} +{"question": "Where do researchers use Clifford algebras and Clifford Fourier transforms to solve PDEs numerically?", "answer": ["Geometric algebra generation of molecular surfaces"], "answer_arxiv_id": ["2112.13204"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_13047"} +{"question": "Which papers indicate that the foundational NeRF model could act as the sole representation for simultaneous localization and mapping?", "answer": ["iMAP: Implicit Mapping and Positioning in Real-Time"], "answer_arxiv_id": ["2103.12352"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_13048"} +{"question": "Which works are about attention-based RGB SOD?", "answer": ["DPANet: Depth Potentiality-Aware Gated Attention Network for RGB-D\n Salient Object Detection", "PiCANet: Learning Pixel-wise Contextual Attention for Saliency Detection"], "answer_arxiv_id": ["2003.08608", "1708.06433v2"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_13049"} +{"question": "Which papers have demonstrated that external memory is helpful for various visual and language related tasks?", "answer": ["Generalization through Memorization: Nearest Neighbor Language Models", "Nearest Neighbor Machine Translation", "GNN-LM: Language Modeling based on Global Contexts via GNN", "Improving language models by retrieving from trillions of tokens", "Retrieval Augmented Classification for Long-Tail Visual Recognition", "Large-Scale Long-Tailed Recognition in an Open World", "Dense Passage Retrieval for Open-Domain Question Answering", "Latent Retrieval for Weakly Supervised Open Domain Question Answering", "DeCap: Decoding CLIP Latents for Zero-Shot Captioning via Text-Only\n Training", "SmallCap: Lightweight Image Captioning Prompted with Retrieval\n Augmentation"], "answer_arxiv_id": ["1911.00172", "2010.00710", "2110.08743", "2112.04426", "2202.11233", "1904.05160", "2004.04906", "1906.00300", "2303.03032", "2209.15323"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_13050"} +{"question": "What studies propose the partitioning of parameters into shared and task specific ones to solve task interference?", "answer": ["Attentive Single-Tasking of Multiple Tasks", "Many Task Learning with Task Routing"], "answer_arxiv_id": ["1904.08918", "1903.12117"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_13051"} +{"question": "Which works provide evidence of bias in GNNs due to sensitive node features, and demonstrate such bias in diverse tasks such as recommendations and loan fraud detection?", "answer": ["Say No to the Discrimination: Learning Fair Graph Neural Networks with Limited Sensitive Attribute Information", "Towards a Unified Framework for Fair and Stable Graph Representation Learning", "DeBayes: a Bayesian Method for Debiasing Network Embeddings"], "answer_arxiv_id": ["2009.01454", "2102.13186", "2002.11442"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13052"} +{"question": "Could you provide some studies about discrete prompts?", "answer": ["Universal Adversarial Triggers for Attacking and Analyzing NLP", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically\n Generated Prompts", "How Can We Know What Language Models Know?", "BARTScore: Evaluating Generated Text as Text Generation", "BERTese: Learning to Speak to BERT", "Making Pre-trained Language Models Better Few-shot Learners", "Commonsense Knowledge Mining from Pretrained Models", "Selective Annotation Makes Language Models Better Few-Shot Learners", "Black-box Prompt Learning for Pre-trained Language Models"], "answer_arxiv_id": ["1908.07125", "2010.15980", "1911.12543", "2106.11520", "2103.05327", "2012.15723", "1909.00505", "2209.01975", "2201.08531"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_13053"} +{"question": "Could you provide some studies about learning plausible placement of humans into scenes in context of an object or environment?", "answer": ["PLACE: Proximity Learning of Articulation and Contact in 3D Environments", "Resolving 3D Human Pose Ambiguities with 3D Scene Constraints", "Generating 3D People in Scenes without People", "Populating 3D Scenes by Learning Human-Scene Interaction"], "answer_arxiv_id": ["2008.05570", "1908.06963", "1912.02923", "2012.11581"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_13054"} +{"question": "Who firstly proposed the Split and Rephrase task and created the WebSplit dataset?", "answer": ["Split and Rephrase"], "answer_arxiv_id": ["1707.06971"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_13055"} +{"question": "Could you provide some studies about memory-based approaches in lifelong learning?", "answer": ["Gradient Episodic Memory for Continual Learning", "Efficient Lifelong Learning with A-GEM", "On Tiny Episodic Memories in Continual Learning", "Improved Schemes for Episodic Memory-based Lifelong Learning", "iCaRL: Incremental Classifier and Representation Learning", "Episodic Memory in Lifelong Language Learning", "Efficient Meta Lifelong-Learning with Limited Memory"], "answer_arxiv_id": ["1706.08840", "1812.00420", "1902.10486", "1909.11763", "1611.07725", "1906.01076", "2010.02500"], "source_meta": {"published_time": "20211216"}, "qid": "AutoScholarQuery_train_13056"} +{"question": "Can you tell me about studies that utilize Positive-Unlabeled (PU) learning in tasks like data retrieval, outlier detection, recommendations, and control?", "answer": ["Positive-Unlabeled Learning with Non-Negative Risk Estimator", "PUNCH: Positive UNlabelled Classification based information retrieval in Hyperspectral images", "Positive-Unlabeled Reward Learning"], "answer_arxiv_id": ["1703.00593", "1904.04547", "1911.00459"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_13057"} +{"question": "What works are about video generation and video editing using diffusion models?", "answer": ["Make-A-Video: Text-to-Video Generation without Text-Video Data", "Imagen Video: High Definition Video Generation with Diffusion Models"], "answer_arxiv_id": ["2209.14792", "2210.02303"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_13058"} +{"question": "Which work is notable for successfully reducing the computation cost by diffusing in a low-resolution latent space?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_13059"} +{"question": "Which work proposed a 9D representation for rotation regression?", "answer": ["An Analysis of SVD for Deep Rotation Estimation"], "answer_arxiv_id": ["2006.14616"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_13060"} +{"question": "Which papers discuss using a hyper-network to generate model weights in addition-based PEFT?", "answer": ["Parameter-efficient Multi-task Fine-tuning for Transformers via Shared\n Hypernetworks"], "answer_arxiv_id": ["2106.04489"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_13061"} +{"question": "Which work primarily handles concepts that have clear semantic distinctions?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2212.04488"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_13062"} +{"question": "Which works have been done towards inducing a prior in the model for better capture of the task in pose estimation?", "answer": ["Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields", "T-LEAP: Occlusion-robust pose estimation of walking cows using temporal\n information", "ZoomNAS: Searching for Whole-body Human Pose Estimation in the Wild", "Deep Keypoint-Based Camera Pose Estimation with Geometric Constraints"], "answer_arxiv_id": ["1611.08050", "2104.08029", "2208.11547", "2007.15122"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_13063"} +{"question": "Can you provide studies which adopted differential point rendering technique for scene reconstructions?", "answer": ["Flexible Techniques for Differentiable Rendering with 3D Gaussians", "Approximate Differentiable Rendering with Algebraic Surfaces", "ADOP: Approximate Differentiable One-Pixel Point Rendering"], "answer_arxiv_id": ["2308.14737", "2207.10606", "2110.06635"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_13064"} +{"question": "Can you mention a few studies related to UCB algorithms that reflect the principle of optimism in action selection to encourage exploration?", "answer": ["The KL-UCB Algorithm for Bounded Stochastic Bandits and Beyond"], "answer_arxiv_id": ["1102.2490"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_13065"} +{"question": "What papers are discussing the value underestimation methods in offline RL?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "COMBO: Conservative Offline Model-Based Policy Optimization", "Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble", "Pessimistic Bootstrapping for Uncertainty-Driven Offline Reinforcement Learning", "RORL: Robust Offline Reinforcement Learning via Conservative Smoothing", "Why So Pessimistic? Estimating Uncertainties for Offline RL through Ensembles, and Why Their Independence Matters"], "answer_arxiv_id": ["2006.04779", "2102.08363", "2110.01548", "2202.11566", "2206.02829", "2205.13703"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13066"} +{"question": "What works make an attempt towards introducing weak supervision as an alternative to fully-unsupervised representation learning in object-centric learning?", "answer": ["Conditional Object-Centric Learning from Video", "SAVi++: Towards End-to-End Object-Centric Learning from Real-World Videos", "Discovering Objects that Can Move", "Object Discovery from Motion-Guided Tokens"], "answer_arxiv_id": ["2111.12594", "2206.07764", "2203.10159", "2303.15555"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_13067"} +{"question": "Which work first applied NAS to design the Transformer architecture?", "answer": ["The Evolved Transformer"], "answer_arxiv_id": ["1901.11117"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_13068"} +{"question": "What are some studies that concentrated on context modules or self-attention variations in semantic segmentation?", "answer": ["DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs", "Rethinking Atrous Convolution for Semantic Image Segmentation", "Pyramid Scene Parsing Network", "Segmenter: Transformer for Semantic Segmentation", "Non-local Neural Networks", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["1606.00915", "1706.05587", "1612.01105", "2105.05633", "1711.07971", "2112.01527"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_13069"} +{"question": "Could you provide me the work that proposed to decode tree topology from hyperbolic coordinates in hierarchical clustering?", "answer": ["From Trees to Continuous Embeddings and Back: Hyperbolic Hierarchical Clustering"], "answer_arxiv_id": ["2010.00402"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_13070"} +{"question": "Are there any papers regarding the role of cycle consistency in unpaired I2I translation?", "answer": ["Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks", "Learning to Discover Cross-Domain Relations with Generative Adversarial Networks", "Unsupervised Image-to-Image Translation Networks"], "answer_arxiv_id": ["1703.10593", "1703.05192", "1703.00848"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_13071"} +{"question": "Which works focus on offline model-based optimization through generative modeling?", "answer": ["Autofocused oracles for model-based design", "Model Inversion Networks for Model-Based Optimization", "Bootstrapped Training of Score-Conditioned Generator for Offline Design of Biological Sequences"], "answer_arxiv_id": ["2006.08052", "1912.13464", "2306.03111"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_13072"} +{"question": "Which models have been trained to align pairs of text and image embeddings?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_13073"} +{"question": "What studies have used watermarks to prevent exact duplication of machine learning models?", "answer": ["Adversarial frontier stitching for remote neural network watermarking", "Turning Your Weakness Into a Strength: Watermarking Deep Neural Networks by Backdooring"], "answer_arxiv_id": ["1711.01894", "1802.04633"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13074"} +{"question": "What papers discuss approaches that interleave 3D conv layers with spatial layers to generate temporally smooth videos?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models"], "answer_arxiv_id": ["2304.08818"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13075"} +{"question": "What papers examine the utility of event cameras in 3-D reconstruction?", "answer": ["Semi-Dense 3D Reconstruction with a Stereo Event Camera", "Event-based Stereo Visual Odometry", "EventNeRF: Neural Radiance Fields from a Single Colour Event Camera", "Deformable Neural Radiance Fields using RGB and Event Cameras"], "answer_arxiv_id": ["1807.07429", "2007.15548", "2206.11896", "2309.08416"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_13076"} +{"question": "Could you provide me some works that applied knowledge distillation to reduce model complexity in 2D detection?", "answer": ["Towards Efficient 3D Object Detection with Knowledge Distillation", "Representation Disparity-aware Distillation for 3D Object Detection", "PointDistiller: Structured Knowledge Distillation Towards Efficient and\n Compact 3D Detection"], "answer_arxiv_id": ["2205.15156", "2308.10308", "2205.11098"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_13077"} +{"question": "Which papers discussed the techniques to classify given news as real or fake?", "answer": ["Exploiting Multi-domain Visual Information for Fake News Detection", "Detecting and Grounding Multi-Modal Media Manipulation", "SAFE: Similarity-Aware Multi-Modal Fake News Detection", "Improving Fake News Detection by Using an Entity-enhanced Framework to\n Fuse Diverse Multimodal Clues"], "answer_arxiv_id": ["1908.04472", "2304.02556", "2003.04981", "2108.10509"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_13078"} +{"question": "Any works about 'Video Anomaly Detection' being treated as a supervised, or weakly supervised problem?", "answer": ["Real-world Anomaly Detection in Surveillance Videos", "Generative Cooperative Learning for Unsupervised Video Anomaly Detection", "UBnormal: New Benchmark for Supervised Open-Set Video Anomaly Detection"], "answer_arxiv_id": ["1801.04264", "2203.03962", "2111.08644"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_13079"} +{"question": "What work introduces memory tokens to augment transformer capacity?", "answer": ["Memory Transformer"], "answer_arxiv_id": ["2006.11527"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_13080"} +{"question": "Which paper refines a 3D mesh obtained by propagating the SfM points?", "answer": ["Multi-View Photometric Stereo: A Robust Solution and Benchmark Dataset\n for Spatially Varying Isotropic Materials"], "answer_arxiv_id": ["2001.06659"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_13081"} +{"question": "Could you provide me some studies that implement learning a scoring function to measure cut quality?", "answer": ["Reinforcement Learning for Integer Programming: Learning to Cut", "Learning to Cut by Looking Ahead: Cutting Plane Selection via Imitation Learning", "Learning to Select Cuts for Efficient Mixed-Integer Programming"], "answer_arxiv_id": ["1906.04859", "2206.13414", "2105.13645"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_13082"} +{"question": "Could you list some works that focused on model architectures?", "answer": ["Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks", "Do Wider Neural Networks Really Help Adversarial Robustness?", "Are Transformers More Robust Than CNNs?", "Towards Robust Vision Transformer", "Vision Transformers are Robust Learners"], "answer_arxiv_id": ["2110.03825", "2010.01279", "2111.05464", "2105.07926", "2105.07581"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_13083"} +{"question": "Are there any works that propose to minimize a cost, risk, or error function to find the optimal rejection threshold?", "answer": ["Reliable Multilabel Classification: Prediction with Partial Abstention", "Classification with Rejection Based on Cost-sensitive Classification", "SelectiveNet: A Deep Neural Network with an Integrated Reject Option", "Self-Adaptive Training: beyond Empirical Risk Minimization"], "answer_arxiv_id": ["1904.09235", "2010.11748v5", "1901.09192", "2002.10319"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_13084"} +{"question": "Which studies employed Object Detection-based models for hateful meme detection?", "answer": ["VisualBERT: A Simple and Performant Baseline for Vision and Language", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "UNITER: UNiversal Image-TExt Representation Learning"], "answer_arxiv_id": ["1908.03557", "2004.06165", "1909.11740"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_13085"} +{"question": "Which studies dealt with how the distributions of private test statistics converge to public?", "answer": ["Efficient, Differentially Private Point Estimators", "A statistical framework for differential privacy"], "answer_arxiv_id": ["0809.4794", "0811.2501"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_13086"} +{"question": "Which works leveraged delay differential equations (DDEs) in the context of deep learning?", "answer": ["Delay Differential Neural Networks", "Neural Delay Differential Equations", "Neural Piecewise-Constant Delay Differential Equations", "Neural Ordinary Differential Equations"], "answer_arxiv_id": ["2012.06800v1", "2102.10801", "2201.00960", "1806.07366"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_13087"} +{"question": "Which study introduced Variational Autoencoders (VAEs)?", "answer": ["Efficient-VDVAE: Less is more"], "answer_arxiv_id": ["2203.13751"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_13088"} +{"question": "Are there any researches where minimal projection loss with auto-masking was used?", "answer": ["Digging Into Self-Supervised Monocular Depth Estimation"], "answer_arxiv_id": ["1806.01260"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13089"} +{"question": "Which research is about the stateful losses used in training retrieval models?", "answer": ["Large-Margin Softmax Loss for Convolutional Neural Networks", "SphereFace: Deep Hypersphere Embedding for Face Recognition", "CosFace: Large Margin Cosine Loss for Deep Face Recognition", "ArcFace: Additive Angular Margin Loss for Deep Face Recognition"], "answer_arxiv_id": ["1612.02295", "1704.08063", "1801.09414", "1801.07698"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_13090"} +{"question": "What studies provide deep learning based solutions that allow sampling multiple answers to the ill-posed SR problem?", "answer": ["Explorable Super Resolution", "SRFlow: Learning the Super-Resolution Space with Normalizing Flow", "DeepSEE: Deep Disentangled Semantic Explorative Extreme Super-Resolution", "Denoising Diffusion Restoration Models", "Come-Closer-Diffuse-Faster: Accelerating Conditional Diffusion Models\n for Inverse Problems through Stochastic Contraction", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "SRFlow: Learning the Super-Resolution Space with Normalizing Flow", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of\n Generative Models", "Exploring the solution space of linear inverse problems with GAN latent\n geometry"], "answer_arxiv_id": ["1912.01839", "2006.14200", "2004.04433", "2201.11793", "2112.05146", "2212.00490", "2006.14200", "2108.02938", "2003.03808", "2207.00460"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_13091"} +{"question": "Could you provide me research that conducted studies on dataset distillation or dataset condensation?", "answer": ["Dataset Distillation", "Generative Teaching Networks: Accelerating Neural Architecture Search by Learning to Generate Synthetic Training Data", "Soft-Label Dataset Distillation and Text Dataset Distillation", "Flexible Dataset Distillation: Learn Labels Instead of Images", "Dataset Meta-Learning from Kernel Ridge-Regression", "Dataset Distillation with Infinitely Wide Convolutional Networks", "Dataset Distillation by Matching Training Trajectories", "Dataset Condensation with Gradient Matching", "Dataset Condensation with Differentiable Siamese Augmentation", "Dataset Condensation with Distribution Matching", "CAFE: Learning to Condense Dataset by Aligning Features", "Graph Condensation for Graph Neural Networks", "Condensing Graphs via One-Step Gradient Matching"], "answer_arxiv_id": ["1811.10959", "1912.07768", "1910.02551", "2006.08572", "2011.00050", "2107.13034", "2203.11932", "2006.05929", "2102.08259", "2110.04181", "2203.01531", "2110.07580", "2206.07746"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_13092"} +{"question": "For improving communication efficiency in FL, were there any studies used low-rank approximation?", "answer": ["Atomo: Communication-efficient Learning via Atomic Sparsification", "PowerSGD: Practical Low-Rank Gradient Compression for Distributed Optimization", "DRIVE: One-bit Distributed Mean Estimation", "EDEN: Communication-Efficient and Robust Distributed Mean Estimation for Federated Learning", "Masked Training of Neural Networks with Partial Gradients"], "answer_arxiv_id": ["1806.04090", "1905.13727", "2105.08339", "2108.08842", "2106.08895v3"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13093"} +{"question": "What papers address the data-hungry nature of NeRF by learning shared priors or incorporating additional supervision?", "answer": ["SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single\n Image", "pixelNeRF: Neural Radiance Fields from One or Few Images", "Depth-supervised NeRF: Fewer Views and Faster Training for Free", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "S$^3$-NeRF: Neural Reflectance Field from Shading and Shadow under a\n Single Viewpoint", "GINA-3D: Learning to Generate Implicit Neural Assets in the Wild"], "answer_arxiv_id": ["2204.00928", "2012.02190", "2107.02791", "2103.15595", "2210.08936", "2304.02163"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_13094"} +{"question": "Are there any works that introduced benchmarks across various multilingual tasks?", "answer": ["XTREME-R: Towards More Challenging and Nuanced Multilingual Evaluation"], "answer_arxiv_id": ["2104.07412"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_13095"} +{"question": "What works have proposed algorithmic fairness through the (non-)existence of certain causal paths in a graph?", "answer": ["Avoiding Discrimination through Causal Reasoning", "A Causal Framework for Discovering and Removing Direct and Indirect Discrimination"], "answer_arxiv_id": ["1706.02744", "1611.07509"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_13096"} +{"question": "Could you provide me some references about the unified framework of stochastic beam search?", "answer": ["Language GANs Falling Short", "Stochastic Beams and Where to Find Them: The Gumbel-Top-k Trick for Sampling Sequences Without Replacement", "Conditional Poisson Stochastic Beam Search"], "answer_arxiv_id": ["1811.02549", "1903.06059", "2109.11034"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_13097"} +{"question": "Any works about pretraining to extract visual representations for video understanding?", "answer": ["Omnivore: A Single Model for Many Visual Modalities", "MViTv2: Improved Multiscale Vision Transformers for Classification and Detection", "UniFormer: Unifying Convolution and Self-attention for Visual Recognition", "MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognition", "SlowFast Networks for Video Recognition"], "answer_arxiv_id": ["2201.08377", "2112.01526", "2201.09450", "2201.08383", "1812.03982"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_13098"} +{"question": "What papers have contributions in the field of Ensemble KD?", "answer": ["Knowledge Distillation by On-the-Fly Native Ensemble"], "answer_arxiv_id": ["1806.04606"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_13099"} +{"question": "Are there any works about preventing free-riding as a data sharing incentive?", "answer": ["Mechanisms that Incentivize Data Sharing in Federated Learning"], "answer_arxiv_id": ["2207.04557"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_13100"} +{"question": "Which work directly predicts solutions to economic dispatch problems?", "answer": ["Learning Optimization Proxies for Large-Scale Security-Constrained Economic Dispatch"], "answer_arxiv_id": ["2112.13469"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_13101"} +{"question": "What works touch on the field of meta-learning for RL, including the subtopic of learning policies?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["1703.03400"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_13102"} +{"question": "What papers have investigated techniques for model compression such as pruning or quantization?", "answer": ["Structured Pruning of Deep Convolutional Neural Networks", "Global Vision Transformer Pruning with Hessian-Aware Saliency", "Patch Slimming for Efficient Vision Transformers", "X-Pruner: eXplainable Pruning for Vision Transformers", "A Survey of Quantization Methods for Efficient Neural Network Inference"], "answer_arxiv_id": ["1512.08571v1", "2110.04869", "2106.02852", "2303.04935", "2103.13630"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_13103"} +{"question": "Can you list a few applications of Differentiable Physics simulators?", "answer": ["gradSim: Differentiable simulation for system identification and visuomotor control", "DiffSkill: Skill Abstraction from Differentiable Physics for Deformable Object Manipulations with Tools", "Accelerated Policy Learning with Parallel Differentiable Simulation"], "answer_arxiv_id": ["2104.02646v1", "2203.17275", "2204.07137"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_13104"} +{"question": "Which research paper first introduced periodic functions as activations within neural network architectures?", "answer": ["Implicit Neural Representations with Periodic Activation Functions"], "answer_arxiv_id": ["2006.09661"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_13105"} +{"question": "Which works applied contrastive learning methods in image classification?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Supervised Contrastive Learning"], "answer_arxiv_id": ["2103.00020", "2004.11362"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_13106"} +{"question": "What researches focus on the efficient tuning of image diffusion models?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "Parameter-Efficient Transfer Learning for NLP", "The Power of Scale for Parameter-Efficient Prompt Tuning", "HyperNetworks", "Res-Tuning: A Flexible and Efficient Tuning Paradigm via Unbinding Tuner\n from Backbone"], "answer_arxiv_id": ["2106.09685", "1902.00751", "2104.08691", "1609.09106", "2310.19859"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_13107"} +{"question": "Could you provide some work on SIDE using diffusion architectures?", "answer": ["Monocular Depth Estimation using Diffusion Models", "DDP: Diffusion Model for Dense Visual Prediction", "Unleashing Text-to-Image Diffusion Models for Visual Perception", "Text-image Alignment for Diffusion-based Perception", "DiffusionDepth: Diffusion Denoising Approach for Monocular Depth\n Estimation"], "answer_arxiv_id": ["2302.14816", "2303.17559", "2303.02153v1", "2310.00031", "2303.05021"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_13108"} +{"question": "Which studies use visual-language models that focus on representation learning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_13109"} +{"question": "Which studies use a hierarchical Bayesian approach to meta-learn priors over Neural Network parameters?", "answer": ["A PAC-Bayesian Bound for Lifelong Learning", "Meta-Learning by Adjusting Priors Based on Extended PAC-Bayes Theory", "PACOH: Bayes-Optimal Meta-Learning with PAC-Guarantees"], "answer_arxiv_id": ["1311.2838", "1711.01244", "2002.05551"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_13110"} +{"question": "Could you tell which research finds difficulty in handling negative or composite prompts in some distillation techniques?", "answer": ["On Distillation of Guided Diffusion Models", "Compositional Visual Generation with Composable Diffusion Models"], "answer_arxiv_id": ["2210.03142v3", "2206.01714"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_13111"} +{"question": "What papers have been exploring the use of meta learning for generating better initialization in low-resource adaptation for ASR models?", "answer": ["On First-Order Meta-Learning Algorithms", "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Self-Training for End-to-End Speech Recognition"], "answer_arxiv_id": ["1803.02999", "1703.03400", "1909.09116"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13112"} +{"question": "Could you mention some papers that studied the progress in emergent language from multi-agent communication?", "answer": ["Emergent Multi-Agent Communication in the Deep Learning Era"], "answer_arxiv_id": ["2006.02419"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_13113"} +{"question": "What works discuss point-based methods in 3D detection and contribute to processing point cloud data?", "answer": ["PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud", "3DSSD: Point-based 3D Single Stage Object Detector", "SASA: Semantics-Augmented Set Abstraction for Point-based 3D Object Detection", "Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds"], "answer_arxiv_id": ["1812.04244", "2002.10187", "2201.01976", "2203.11139"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_13114"} +{"question": "What papers focus on image harmonization using pyramid filters or hierarchical structures?", "answer": ["DCCF: Deep Comprehensible Color Filter Learning Framework for\n High-Resolution Image Harmonization", "High-Resolution Image Harmonization via Collaborative Dual\n Transformations", "Hierarchical Dynamic Image Harmonization"], "answer_arxiv_id": ["2207.04788", "2109.06671", "2211.08639"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_13115"} +{"question": "Which work proposed an architecture for an early fusion of multi-sensorial features?", "answer": ["Audio-Visual Scene Analysis with Self-Supervised Multisensory Features"], "answer_arxiv_id": ["1804.03641"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_13116"} +{"question": "What works are there on instance segmentation?", "answer": ["Mask R-CNN", "K-Net: Towards Unified Image Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["1703.06870", "2106.14855", "2112.01527"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_13117"} +{"question": "Are there any studies on distribution generalization of VRPs?", "answer": ["A Game-Theoretic Approach for Improving Generalization Ability of TSP Solvers", "Learning to Solve Travelling Salesman Problem with Hardness-Adaptive Curriculum", "Generalization of Neural Combinatorial Solvers Through the Lens of Adversarial Robustness", "Learning to Solve Routing Problems via Distributionally Robust Optimization", "Learning Generalizable Models for Vehicle Routing Problems via Knowledge Distillation", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["2110.15105", "2204.03236", "2110.10942", "2202.07241", "2210.07686", "1911.08731", "1503.02531"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_13118"} +{"question": "Which papers use Inception Score (IS) for assessing image quality in their studies?", "answer": ["Improved Techniques for Training GANs", "A Note on the Inception Score"], "answer_arxiv_id": ["1606.03498", "1801.01973"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13119"} +{"question": "What research discuss the issues of slow mixing and convergence in Bayesian causal discovery?", "answer": ["Bayesian structure learning using dynamic programming and MCMC"], "answer_arxiv_id": ["1206.5247"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_13120"} +{"question": "What papers assess the performance of ImageNet-pretrained models on medical imaging data?", "answer": ["Transfusion: Understanding Transfer Learning for Medical Imaging", "Comparing Different Deep Learning Architectures for Classification of Chest Radiographs", "CheXtransfer: Performance and Parameter Efficiency of ImageNet Models for Chest X-Ray Interpretation"], "answer_arxiv_id": ["1902.07208", "2002.08991", "2101.06871"], "source_meta": {"published_time": "20230111"}, "qid": "AutoScholarQuery_train_13121"} +{"question": "Could you mention studies about the Target-oriented Hypothesis Adaptation Network?", "answer": ["TOHAN: A One-step Approach towards Few-shot Hypothesis Adaptation"], "answer_arxiv_id": ["2106.06326"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_13122"} +{"question": "Could you provide me some studies about data profiling in Data-Centric AI?", "answer": ["DC-Check: A Data-Centric AI checklist to guide the development of reliable machine learning systems"], "answer_arxiv_id": ["2211.05764"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_13123"} +{"question": "What papers introduced proposal-free methods for video moment retrieval?", "answer": ["To Find Where You Talk: Temporal Sentence Localization in Video with Attention Based Location Regression", "Dense Regression Network for Video Grounding", "Local-Global Video-Text Interactions for Temporal Grounding"], "answer_arxiv_id": ["1804.07014", "2004.03545", "2004.07514"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_13124"} +{"question": "Any studies about assessing multi-dimensional aspects in LLM-based evaluators?", "answer": ["G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment", "Large Language Models are Not Yet Human-Level Evaluators for Abstractive\n Summarization", "Is ChatGPT a Good NLG Evaluator? A Preliminary Study"], "answer_arxiv_id": ["2303.16634", "2305.13091", "2303.04048"], "source_meta": {"published_time": "20240701"}, "qid": "AutoScholarQuery_train_13125"} +{"question": "What work addressed the reconstruction ambiguity by training a permuter model to reorder generated graphs alongside a standard encoder/decoder architecture?", "answer": ["Permutation-Invariant Variational Autoencoder for Graph-Level Representation Learning"], "answer_arxiv_id": ["2104.09856"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_13126"} +{"question": "What research discusses the tendency of SGD to lean toward smaller-norm solutions?", "answer": ["In Search of the Real Inductive Bias: On the Role of Implicit\n Regularization in Deep Learning", "Implicit Regularization in Matrix Factorization"], "answer_arxiv_id": ["1412.6614", "1705.09280"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_13127"} +{"question": "Which research papers outline the usage of vision-language pretraining in improving performance in few-shot classification, cross-modality generation, and visual recognition?", "answer": ["Learning to Prompt for Vision-Language Models", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "Task Residual for Tuning Vision-Language Models", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "ActionCLIP: A New Paradigm for Video Action Recognition"], "answer_arxiv_id": ["2109.01134", "2110.04544", "2211.10277", "2112.10741", "2204.06125", "2103.17249", "2109.08472"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_13128"} +{"question": "Which papers put forward the Area to Point Matching (A2PM) framework?", "answer": ["Searching from Area to Point: A Hierarchical Framework for\n Semantic-Geometric Combined Feature Matching"], "answer_arxiv_id": ["2305.00194"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_13129"} +{"question": "Which studies have used the learning time of examples to determine which examples are memorized by a neural network?", "answer": ["Distribution Density, Tails, and Outliers in Machine Learning: Metrics and Applications", "Characterizing Structural Regularities of Labeled Data in Overparameterized Models"], "answer_arxiv_id": ["1910.13427", "2002.03206"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_13130"} +{"question": "Which papers contain details about the Free Music Archive (FMA) dataset?", "answer": ["FMA: A Dataset For Music Analysis"], "answer_arxiv_id": ["1612.01840"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13131"} +{"question": "Could you provide me with some examples of research on multi-agent systems (MAS) in tasks like model evaluation via multi-agent debates, society simulation, or game playing?", "answer": ["ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate", "Generative Agents: Interactive Simulacra of Human Behavior", "AgentVerse: Facilitating Multi-Agent Collaboration and Exploring\n Emergent Behaviors", "Avalon's Game of Thoughts: Battle Against Deception through Recursive\n Contemplation"], "answer_arxiv_id": ["2308.07201", "2304.03442", "2308.10848", "2310.01320"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_13132"} +{"question": "What studies have applied diffusion models to different decision making problems?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis", "Is Conditional Generative Modeling all you need for Decision-Making?", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "Guided Conditional Diffusion for Controllable Traffic Simulation"], "answer_arxiv_id": ["2205.09991", "2211.15657", "2208.06193", "2210.17366"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_13133"} +{"question": "What work related theoretical guarantees for the single MMD tests using a permutation-based threshold to the Gaussian kernels?", "answer": ["Minimax optimality of permutation tests"], "answer_arxiv_id": ["2003.13208v3"], "source_meta": {"published_time": "20211028"}, "qid": "AutoScholarQuery_train_13134"} +{"question": "Which works have trained multimodal models on image-text pairs datasets using a contrastive objective?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_13135"} +{"question": "What papers are about meta-learning in various stochastic bandit settings?", "answer": ["Sequential Transfer in Multi-armed Bandit with Finite Set of Models", "Meta-Learning Bandit Policies by Gradient Ascent", "Bayesian decision-making under misspecified priors with applications to meta-learning", "Meta-Thompson Sampling", "No Regrets for Learning the Prior in Bandits", "Meta-learning with Stochastic Linear Bandits", "Multi-Environment Meta-Learning in Stochastic Linear Bandits", "Non-stationary Bandits and Meta-Learning with a Small Set of Optimal Arms"], "answer_arxiv_id": ["1307.6887", "2006.05094", "2107.01509", "2102.06129", "2107.06196", "2005.08531", "2205.06326", "2202.13001"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_13136"} +{"question": "Could you provide me some works interpreting transformer models and Large Language Models?", "answer": ["What Does BERT Look At? An Analysis of BERT's Attention", "Neurons in Large Language Models: Dead, N-gram, Positional"], "answer_arxiv_id": ["1906.04341", "2309.04827"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_13137"} +{"question": "Which works suggested employed online RL methods which required a significant number of samples for learning?", "answer": ["Computational Benefits of Intermediate Rewards for Goal-Reaching Policy Learning", "Unpacking Reward Shaping: Understanding the Benefits of Reward Engineering on Sample Complexity", "Understanding the Complexity Gains of Single-Task RL with a Curriculum"], "answer_arxiv_id": ["2107.03961", "2210.09579v1", "2212.12809"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_13138"} +{"question": "Any works about the development of improved interpretable models in the framework of CBM?", "answer": ["A Framework for Learning Ante-hoc Explainable Models via Concepts"], "answer_arxiv_id": ["2108.11761"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_13139"} +{"question": "What papers showed similar types of guarantees to the ground-truth clustering as mentioned in the research paper?", "answer": ["Clustering is difficult only when it does not matter"], "answer_arxiv_id": ["1205.4891v1"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_13140"} +{"question": "In what works are there discussions about fairness issues originating from sensitive node attributes?", "answer": ["FMP: Toward Fair Graph Message Passing against Topology Bias", "FairDrop: Biased Edge Dropout for Enhancing Fairness in Graph Representation Learning", "Fair Representation Learning for Heterogeneous Information Networks"], "answer_arxiv_id": ["2202.04187", "2104.14210", "2104.08769"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_13141"} +{"question": "What work attempts to obtain pixel-level embeddings from CLIP that can be directly utilized for segmentation?", "answer": ["Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2112.01071"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_13142"} +{"question": "In what study was the feature-based fastest point sampling (F-FPS) strategy proposed?", "answer": ["3DSSD: Point-based 3D Single Stage Object Detector"], "answer_arxiv_id": ["2002.10187"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_13143"} +{"question": "Are there any concurrent works that investigate contrastive learning in mask image modeling?", "answer": ["Siamese Image Modeling for Self-Supervised Vision Representation Learning", "Contrastive Masked Autoencoders are Stronger Vision Learners"], "answer_arxiv_id": ["2206.01204", "2207.13532"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_13144"} +{"question": "Which studies attempted to enhance and expand the diffusion framework?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Structured Denoising Diffusion Models in Discrete State-Spaces", "Denoising Diffusion Implicit Models", "Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning", "Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2107.03006", "2010.02502", "2208.04202", "2208.09392"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_13145"} +{"question": "What works used hypernetworks as a conditioning mechanism for representing collections of shapes simultaneously in implicit neural models?", "answer": ["Implicit Neural Representations with Periodic Activation Functions", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "Light Field Networks: Neural Scene Representations with Single-Evaluation Rendering", "MetaAvatar: Learning Animatable Clothed Human Models from Few Depth Images", "Transformers as Meta-Learners for Implicit Neural Representations"], "answer_arxiv_id": ["2006.09661", "1906.01618", "2106.02634", "2106.11944", "2208.02801"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_13146"} +{"question": "What works try to edit real images via user sketches or strokes in the Diffusion Probabilistic Models (DDPM) editing domain?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations"], "answer_arxiv_id": ["2108.01073"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_13147"} +{"question": "Which research investigated a triplane representation that scales well with resolution and lowers the training cost?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["2112.07945"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_13148"} +{"question": "Which papers have studied part correspondences in multi-body systems?", "answer": ["FLOT: Scene Flow on Point Clouds guided by Optimal Transport", "MeteorNet: Deep Learning on Dynamic 3D Point Cloud Sequences", "FlowNet3D: Learning Scene Flow in 3D Point Clouds", "Self-Point-Flow: Self-Supervised Scene Flow Estimation from Point Clouds with Optimal Transport and Random Walk"], "answer_arxiv_id": ["2007.11142", "1910.09165", "1806.01411", "2105.08248"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13149"} +{"question": "Which studies discuss metric-based meta-learning algorithms?", "answer": ["Matching Networks for One Shot Learning", "Prototypical Networks for Few-shot Learning", "Learning to Compare: Relation Network for Few-Shot Learning"], "answer_arxiv_id": ["1606.04080", "1703.05175", "1711.06025"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_13150"} +{"question": "Which work proposes a gated cross-attention mechanism for Multimodal Language Models (MLLMs)?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_13151"} +{"question": "What is an impressive work that first learned class-agnostic box proposals and then refined them to object bounding boxes?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks"], "answer_arxiv_id": ["1506.01497"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_13152"} +{"question": "Which studies made attempts to bridge several time-sensitive video tasks?", "answer": ["Unifying Event Detection and Captioning as Sequence Generation via\n Pre-Training", "QVHighlights: Detecting Moments and Highlights in Videos via Natural\n Language Queries", "UniVTG: Towards Unified Video-Language Temporal Grounding", "UMT: Unified Multi-modal Transformers for Joint Video Moment Retrieval\n and Highlight Detection"], "answer_arxiv_id": ["2207.08625", "2107.09609", "2307.16715", "2203.12745"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_13153"} +{"question": "Which works utilized some form of global preconditioning in gradient-based samplers for latent Gaussian models?", "answer": ["MCMC Methods for Functions: Modifying Old Algorithms to Make Them Faster", "Auxiliary gradient-based sampling algorithms"], "answer_arxiv_id": ["1202.0709", "1610.09641"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13154"} +{"question": "What research demonstrates humans' reliance on shape features for lexical learning and object recognition tasks?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness"], "answer_arxiv_id": ["1811.12231"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_13155"} +{"question": "Which papers focused on improving speed of NeRF training?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2201.05989", "2112.05131"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_13156"} +{"question": "Which studies are performing density estimation in embedding space in the field of KGEs?", "answer": ["TransG : A Generative Model for Knowledge Graph Embedding", "Probabilistic Box Embeddings for Uncertain Knowledge Graph Reasoning"], "answer_arxiv_id": ["1509.05488", "2104.04597"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13157"} +{"question": "Which works proposed typical MARL algorithms such as QMIX and MADDPG, where set input is represented as a concatenation of the m entities' features?", "answer": ["QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning", "Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments"], "answer_arxiv_id": ["1803.11485", "1706.02275"], "source_meta": {"published_time": "20220310"}, "qid": "AutoScholarQuery_train_13158"} +{"question": "Are there any works combine neural implicit mapping with SLAM and improve the scene representation method?", "answer": ["iMAP: Implicit Mapping and Positioning in Real-Time", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural\n Real-Time SLAM", "ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of\n Signed Distance Fields"], "answer_arxiv_id": ["2103.12352", "2112.12130", "2304.14377", "2211.11704"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_13159"} +{"question": "Could you tell me about some works discussing the inadequacy of directly using attention weights in performing well in vision tasks?", "answer": ["Transformer Interpretability Beyond Attention Visualization"], "answer_arxiv_id": ["2012.09838"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_13160"} +{"question": "Could you provide me some studies related to program synthesis from high-level specifications such as input-output examples and natural language descriptions?", "answer": ["A Syntactic Neural Model for General-Purpose Code Generation", "Program Synthesis using Natural Language"], "answer_arxiv_id": ["1704.01696", "1509.00413"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_13161"} +{"question": "What papers proposed methods for decomposing scenes into smaller MLPs to represent local semantics?", "answer": ["Panoptic Neural Fields: A Semantic Object-Aware Neural Scene\n Representation", "Nerflets: Local Radiance Fields for Efficient Structure-Aware 3D Scene\n Representation from 2D Supervision"], "answer_arxiv_id": ["2205.04334", "2303.03361"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13162"} +{"question": "What are the studies that pertain to the methods estimate the distributions of one feature conditioned on the other features and perform iterative imputations for one feature at a time?", "answer": ["On the Stationary Distribution of Iterative Imputations"], "answer_arxiv_id": ["1012.2902"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_13163"} +{"question": "Can you provide some works that make stringent assumptions on the data, namely, tokens are tightly clusterable or can be clearly split into clear relevant and irrelevant sets?", "answer": ["A Theoretical Understanding of Shallow Vision Transformers: Learning, Generalization, and Sample Complexity", "Vision Transformers provably learn spatial structure"], "answer_arxiv_id": ["2302.06015v3", "2210.09221"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13164"} +{"question": "Which papers studied the challenges of applying QAT to GLMs and observed performance declines?", "answer": ["Compression of Generative Pre-trained Language Models via Quantization", "Understanding INT4 Quantization for Language Models: Latency Speedup, Composability, and Failure Cases"], "answer_arxiv_id": ["2203.10705", "2301.12017"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_13165"} +{"question": "Could you tell me about the research that uses CLIP to obtain language tokens that semantically correspond to concepts present in synthetic environments?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_13166"} +{"question": "Can you provide a study where a contrastive adapter has been designed to improve robustness?", "answer": ["Contrastive Adapters for Foundation Model Group Robustness"], "answer_arxiv_id": ["2207.07180"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_13167"} +{"question": "Which papers discuss the computation of the global Lipschitz bound as the product of the spectral norm of linear layers to train robust networks?", "answer": ["Lipschitz-Margin Training: Scalable Certification of Perturbation Invariance for Deep Neural Networks", "Agreement-on-the-Line: Predicting the Performance of Neural Networks under Distribution Shift"], "answer_arxiv_id": ["1802.04034", "2206.13089"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_13168"} +{"question": "What are the works that extended neural ODEs to discrete structure and non-Euclidean setting?", "answer": ["GRAND: Graph Neural Diffusion", "Continuous Graph Neural Networks", "Beltrami Flow and Neural Diffusion on Graphs"], "answer_arxiv_id": ["2106.10934", "1912.00967", "2110.09443"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_13169"} +{"question": "Could you provide me a study that applied smoothed analysis in online learning for classification and regression?", "answer": ["Smoothed Analysis with Adaptive Adversaries", "Smoothed Online Learning is as Easy as Statistical Learning", "Smoothed Analysis of Online and Differentially Private Learning", "Efficient and Near-Optimal Smoothed Online Learning for Generalized Linear Functions"], "answer_arxiv_id": ["2102.08446", "2202.04690", "2006.10129", "2205.13056"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_13170"} +{"question": "Which works study the stability-dependent learning rate?", "answer": ["A Modern Introduction to Online Learning"], "answer_arxiv_id": ["1912.13213"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_13171"} +{"question": "Which works proposed to disambiguate independent motion via leveraging stereo-view information?", "answer": ["PLADE-Net: Towards Pixel-Level Accuracy for Self-Supervised Single-View Depth Estimation with Neural Positional Encoding and Distilled Matting Loss"], "answer_arxiv_id": ["2103.07362"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13172"} +{"question": "What papers focus on implicit representation of shapes?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "Learning Implicit Fields for Generative Shape Modeling", "Deep Marching Tetrahedra: a Hybrid Representation for High-Resolution 3D\n Shape Synthesis", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images", "DiffRF: Rendering-Guided 3D Radiance Field Diffusion", "HyperDiffusion: Generating Implicit Neural Fields with Weight-Space\n Diffusion", "3D Neural Field Generation using Triplane Diffusion", "DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Zero-Shot Text-Guided Object Generation with Dream Fields", "Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["1901.05103", "1812.03828", "1812.02822", "2111.04276", "2212.04493", "2209.11163", "2212.01206", "2303.17015", "2211.16677", "2209.14988", "2211.10440", "2112.01455", "2212.14704"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_13173"} +{"question": "Which works deal with the generation of high-quality pseudo-labels for Semi-Supervised Learning?", "answer": ["Self-training with Noisy Student improves ImageNet classification"], "answer_arxiv_id": ["1911.04252"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_13174"} +{"question": "Which paper used latent interpolation and classifier-free guidance as a method?", "answer": ["Training on Thin Air: Improve Image Classification with Generated Data", "Synthetic Data from Diffusion Models Improves ImageNet Classification"], "answer_arxiv_id": ["2305.15316", "2304.08466"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_13175"} +{"question": "Could you provide me some examples of works that attribute robust overfitting to the training examples with small loss value?", "answer": ["Understanding Robust Overfitting of Adversarial Training and Beyond"], "answer_arxiv_id": ["2206.08675"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_13176"} +{"question": "Could you provide me with some references where OT is utilized in label refinery tasks?", "answer": ["Group-aware Label Transfer for Domain Adaptive Person Re-identification", "SoLar: Sinkhorn Label Refinery for Imbalanced Partial-Label Learning"], "answer_arxiv_id": ["2103.12366", "2209.10365"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_13177"} +{"question": "Which papers discuss learning an EBM by matching the first derivatives of the density function and the data distribution?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["1907.05600", "2011.13456"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_13178"} +{"question": "Could you provide me some studies about decentralized deep training with time-varying network topologies?", "answer": ["Consensus Control for Decentralized Deep Learning", "Exponential Graph is Provably Efficient for Decentralized Deep Training", "A Unified Theory of Decentralized SGD with Changing Topology and Local Updates"], "answer_arxiv_id": ["2102.04828", "2110.13363", "2003.10422"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_13179"} +{"question": "Which works provide concepts of differentially-private architecture for time-series data?", "answer": ["Differentially Private Generative Adversarial Networks for Time Series, Continuous, and Discrete Open Data", "Differentially Private Generative Adversarial Network", "Differentially Private Synthetic Medical Data Generation using Convolutional GANs"], "answer_arxiv_id": ["1901.02477", "1802.06739", "2012.11774"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_13180"} +{"question": "Which works have used masked modeling techniques in the visual domain?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "Scaling Language-Image Pre-training via Masking"], "answer_arxiv_id": ["2111.06377", "2112.09133", "2212.00794"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_13181"} +{"question": "Can you name a few works that have trained feed-forward 3D generative models on a large amount of undisclosed 3D data?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "Shap-E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2212.08751", "2305.02463"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_13182"} +{"question": "Could you provide me some studies that use LLM to generate structured queries for knowledge graphs?", "answer": ["Complex Logical Reasoning over Knowledge Graphs using Large Language\n Models"], "answer_arxiv_id": ["2305.01157"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_13183"} +{"question": "Could you give me some examples of research that characterized how SGD escapes sharp minima?", "answer": ["An Alternative View: When Does SGD Escape Local Minima?", "The Anisotropic Noise in Stochastic Gradient Descent: Its Behavior of Escaping from Sharp Minima and Regularization Effects", "A Diffusion Theory For Deep Learning Dynamics: Stochastic Gradient Descent Exponentially Favors Flat Minima", "Exponential escape efficiency of SGD from sharp minima in non-stationary regime"], "answer_arxiv_id": ["1802.06175", "1803.00195", "2002.03495", "2111.04004"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_13184"} +{"question": "What papers discuss transforming the semantic and acoustic information of speech into multiple tiers of tokens using RVQ-VAE structure?", "answer": ["SpeechTokenizer: Unified Speech Tokenizer for Speech Large Language\n Models"], "answer_arxiv_id": ["2308.16692"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_13185"} +{"question": "Which works focused on the prediction of human motion conditioned on other individuals’ kinesic signals?", "answer": ["Towards Social Artificial Intelligence: Nonverbal Social Signal Prediction in A Triadic Interaction"], "answer_arxiv_id": ["1906.04158v1"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_13186"} +{"question": "What studies deal with adversarial learning-based approaches in unsupervised domain adaptation?", "answer": ["Conditional Adversarial Domain Adaptation"], "answer_arxiv_id": ["1705.10667"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13187"} +{"question": "Which work explores the concept of leveraging arm dependencies in the multi-armed bandit (MAB) model for best-arm identification?", "answer": ["Best-Arm Identification in Correlated Multi-Armed Bandits"], "answer_arxiv_id": ["2109.04941"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_13188"} +{"question": "What is the research paper that analyzed the stochastic gradient descent ascent algorithm and provided a proof for its complexity in nonconvex-strongly-concave setting?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems"], "answer_arxiv_id": ["1906.00331"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_13189"} +{"question": "Which works introduced Polya-Gamma augmentation and utilized different likelihood functions and data augmentation to approximate the softmax function in GP classification?", "answer": ["Bayesian inference for logistic models using Pólya-Gamma latent variables", "Multi-Class Gaussian Process Classification Made Conjugate: Efficient Inference via Data Augmentation", "Bayesian Few-Shot Classification with One-vs-Each Pólya-Gamma Augmented Gaussian Processes"], "answer_arxiv_id": ["1205.0310", "1905.09670", "2007.10417v2"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_13190"} +{"question": "What works focus on the 'majorization-minimization' meta-algorithm, minimizing an upper bound on the objective in each iteration?", "answer": ["Optimization with First-Order Surrogate Functions", "Incremental Majorization-Minimization Optimization with Application to Large-Scale Machine Learning"], "answer_arxiv_id": ["1305.3120v1", "1402.4419v3"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_13191"} +{"question": "Which works introduce the notion of discrepancy in domain adaptation theory?", "answer": ["Domain Adaptation: Learning Bounds and Algorithms"], "answer_arxiv_id": ["0902.3430"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_13192"} +{"question": "What research examples utilize NeRF for textured avatar modeling?", "answer": ["A-NeRF: Articulated Neural Radiance Fields for Learning Human Shape,\n Appearance, and Pose", "Capturing and Animation of Body and Clothing from Monocular Video", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "Neural Capture of Animatable 3D Human from Monocular Video", "SelfNeRF: Fast Training NeRF for Human from Monocular Self-rotating\n Video", "Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via\n Self-supervised Scene Decomposition", "InstantAvatar: Learning Avatars from Monocular Video in 60 Seconds", "NeuMan: Neural Human Radiance Field from a Single Video"], "answer_arxiv_id": ["2102.06199", "2210.01868", "2201.04127", "2208.08728", "2210.01651", "2302.11566", "2212.10550", "2203.12575"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_13193"} +{"question": "What research papers have been published on the development of various generative models?", "answer": ["An Introduction to Variational Autoencoders", "Generative Adversarial Networks", "Neural Discrete Representation Learning", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["1906.02691", "2203.00667", "1711.00937", "2105.05233"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_13194"} +{"question": "Which studies contributed to the development of larger language models?", "answer": ["PaLI-X: On Scaling up a Multilingual Vision and Language Model", "Efficient Methods for Natural Language Processing: A Survey", "PaLM: Scaling Language Modeling with Pathways", "Scaling Language Models: Methods, Analysis & Insights from Training\n Gopher"], "answer_arxiv_id": ["2305.18565", "2209.00099", "2204.02311", "2112.11446"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_13195"} +{"question": "Could you provide me some works that stated the bounds in logistic bandit works?", "answer": ["An Experimental Design Approach for Regret Minimization in Logistic Bandits", "Instance-Wise Minimax-Optimal Algorithms for Logistic Bandits"], "answer_arxiv_id": ["2202.02407", "2010.12642v2"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_13196"} +{"question": "Which works have been performed in the field of generating human-centric assets?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "Relighting4D: Neural Relightable Human from Videos", "PaMIR: Parametric Model-Conditioned Implicit Representation for Image-based Human Reconstruction", "PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization", "HumanLiff: Layer-wise 3D Human Generation with Diffusion Model"], "answer_arxiv_id": ["2205.08535", "2208.15001", "2207.07104", "2007.03858", "1905.05172v3", "2308.09712"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_13197"} +{"question": "Are there works suggesting that LMs can learn implicit world models from the training data?", "answer": ["Implicit Representations of Meaning in Neural Language Models", "Emergent world representations: Exploring a sequence model trained on a synthetic task"], "answer_arxiv_id": ["2106.00737", "2210.13382"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13198"} +{"question": "Which works discussed contrastive-based methods for representation learning in images?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning"], "answer_arxiv_id": ["1911.05722", "2003.04297"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_13199"} +{"question": "Which papers are dedicated to image-text question answering in the HQA task?", "answer": ["MuMuQA: Multimedia Multi-Hop News Question Answering via Cross-Media\n Knowledge Extraction and Grounding"], "answer_arxiv_id": ["2112.10728"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_13200"} +{"question": "What references are about Latent Diffusion Models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_13201"} +{"question": "Could you tell me which research introduced the concept of instruction tuning?", "answer": ["Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2109.01652"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_13202"} +{"question": "What are the research works on the performances of greedy decision tree learning algorithms like ID3, C4.5, and CART?", "answer": ["On the Optimality of Trees Generated by ID3", "ID3 Learns Juntas for Smoothed Product Distributions", "Top-down induction of decision trees: rigorous guarantees and inherent limitations", "Provable guarantees for decision tree induction: the agnostic setting", "Decision tree heuristics can fail, even in the smoothed setting"], "answer_arxiv_id": ["1907.05444", "1906.08654", "1911.07375v1", "2006.00743", "2107.00819"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_13203"} +{"question": "Can you provide some works that explore self-supervised learning involving contrasting multiple views of the data?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Learning Representations by Maximizing Mutual Information Across Views", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1807.03748", "1906.00910", "2002.05709", "1911.05722", "2006.07733", "2104.14548"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_13204"} +{"question": "What works are about optimization by extracting static information from the dynamic DNN?", "answer": ["Just-in-Time Dynamic-Batching"], "answer_arxiv_id": ["1904.07421"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_13205"} +{"question": "What research exists on multi-agent RL algorithms?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments", "Learning Nash Equilibrium for General-Sum Markov Games from Batch Data"], "answer_arxiv_id": ["1706.02275", "1606.08718v4"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13206"} +{"question": "Could you provide me some examples of studies that released models with both Datasheet and Model Card artifacts?", "answer": ["OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["2205.01068"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_13207"} +{"question": "Can you cite a study that simplifies the approach of non-contrastive learning through online predictor?", "answer": ["Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2011.10566"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_13208"} +{"question": "Which research uses the diffusion model to implement the edit based generative processes in text revision?", "answer": ["DiffusER: Discrete Diffusion via Edit-based Reconstruction"], "answer_arxiv_id": ["2210.16886v1"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_13209"} +{"question": "Which papers developed the concept of chain-of-thought prompts?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "WT5?! Training Text-to-Text Models to Explain their Predictions", "Emergent Abilities of Large Language Models", "Teaching Algorithmic Reasoning via In-context Learning", "Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Neural Algorithmic Reasoning", "Learning to Reason and Memorize with Self-Notes"], "answer_arxiv_id": ["2201.11903", "2004.14546", "2206.07682", "2211.09066", "2112.00114", "2105.02761", "2305.00833"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13210"} +{"question": "Can you tell me about any works that utilized FNOs for zero-shot super-resolution?", "answer": ["MAgNet: Mesh Agnostic Neural PDE Solver"], "answer_arxiv_id": ["2210.05495"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_13211"} +{"question": "What works exploited the Karush-Kuhn-Tucker optimality conditions of the embedded optimization model in end-to-end learning?", "answer": ["Differentiable Convex Optimization Layers"], "answer_arxiv_id": ["1910.12430"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_13212"} +{"question": "What papers discuss pretraining on public data as the default choice for large scale private training in NLP tasks?", "answer": ["Differentially Private Fine-tuning of Language Models", "Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping", "Differentially Private Bias-Term only Fine-tuning of Foundation Models", "SubMix: Practical Private Prediction for Large-scale Language Models"], "answer_arxiv_id": ["2110.06500", "2212.01539", "2210.00036", "2201.00971"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_13213"} +{"question": "Could you provide me some works about adding noises to activations, outputs, weights, gradients during training?", "answer": ["Adding Gradient Noise Improves Learning for Very Deep Networks", "Adversarial Noise Layer: Regularize Neural Network By Adding Noise"], "answer_arxiv_id": ["1511.06807", "1805.08000"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_13214"} +{"question": "Which papers studied high-dimensional phenomena that remove the necessity of having more data than dimensions?", "answer": ["Consistency of Interpolation with Laplace Kernels is a High-Dimensional Phenomenon", "Just Interpolate: Kernel “Ridgeless” Regression Can Generalize"], "answer_arxiv_id": ["1812.11167", "1808.00387"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13215"} +{"question": "What works have considered a multi-objective view of ensuring fairness in classifiers?", "answer": ["Minimax Pareto Fairness: A Multi Objective Perspective"], "answer_arxiv_id": ["2011.01821"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_13216"} +{"question": "What studies mention about the application of string kernel Gaussian Process (GP) in Bayes Optimization for discrete molecular spaces?", "answer": ["BOSS: Bayesian Optimization over String Spaces"], "answer_arxiv_id": ["2010.00979"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_13217"} +{"question": "Could you give me some references about the phenomenon of catastrophic forgetting in machine learning?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Measuring Catastrophic Forgetting in Neural Networks", "Sequential mastery of multiple visual tasks: Networks naturally learn to learn and forget to forget", "Understanding Catastrophic Forgetting and Remembering in Continual Learning with Optimal Relevance Mapping"], "answer_arxiv_id": ["1612.00796", "1708.02072", "1905.10837", "2102.11343"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_13218"} +{"question": "Which studies proposed different solutions for Instance Segmentation?", "answer": ["Mask R-CNN", "MaskLab: Instance Segmentation by Refining Object Detection with Semantic and Direction Features", "Mask Scoring R-CNN", "Cascade R-CNN: High Quality Object Detection and Instance Segmentation", "Hybrid Task Cascade for Instance Segmentation", "InstanceCut: from Edges to Instances with MultiCut", "Deep Watershed Transform for Instance Segmentation", "TensorMask: A Foundation for Dense Object Segmentation", "Differentiable Multi-Granularity Human Representation Learning for Instance-Aware Human Semantic Parsing", "SOLO: Segmenting Objects by Locations", "SOLOv2: Dynamic and Fast Instance Segmentation", "Instances as Queries", "SOTR: Segmenting Objects with Transformers", "SOLQ: Segmenting Objects by Learning Queries", "ISTR: End-to-End Instance Segmentation with Transformers", "Sparse Instance Activation for Real-Time Instance Segmentation", "Learning Equivariant Segmentation with Instance-Unique Querying"], "answer_arxiv_id": ["1703.06870", "1712.04837", "1903.00241", "1906.09756", "1901.07518", "1611.08272", "1611.08303", "1903.12174", "2103.04570", "1912.04488", "2003.10152", "2105.01928", "2108.06747", "2106.02351", "2105.00637", "2203.12827", "2210.00911"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_13219"} +{"question": "What studies focus on Subgraph GNNs by encoding node representations from subgraphs rather than subtrees?", "answer": ["MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing", "Counting Substructures with Higher-Order Graph Neural Networks: Possibility and Impossibility Results", "Ego-GNNs: Exploiting Ego Structures in Graph Neural Networks", "k-hop graph neural networks"], "answer_arxiv_id": ["1905.00067", "2012.03174", "2107.10957", "1907.06051"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_13220"} +{"question": "Are there any works that apply multi-view stereo methodology to 3D detection?", "answer": ["Monocular 3D Object Detection with Depth from Motion", "STS: Surround-view Temporal Stereo for Multi-view 3D Detection", "BEVStereo: Enhancing Depth Estimation in Multi-view 3D Object Detection with Dynamic Temporal Stereo"], "answer_arxiv_id": ["2207.12988", "2208.10145", "2209.10248"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_13221"} +{"question": "What works modify the second order momentum so that it is non-decreasing in the context of Adam algorithm?", "answer": ["On the convergence of Adam and Beyond", "On the Convergence of A Class of Adam-Type Algorithms for Non-Convex Optimization"], "answer_arxiv_id": ["1904.09237", "1808.02941"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_13222"} +{"question": "What studies discussed permutation symmetries in the context of neural network loss landscapes?", "answer": ["Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Weight-space symmetry in deep networks gives rise to permutation saddles, connected by equal-loss valleys across the loss landscape", "The Role of Permutation Invariance in Linear Mode Connectivity of Neural Networks"], "answer_arxiv_id": ["1802.10026", "1907.02911", "2110.06296"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_13223"} +{"question": "Which works introduced various video-language datasets for tasks like action recognition and video understanding?", "answer": ["A Dataset for Movie Description", "Localizing Moments in Video with Natural Language", "Towards Automatic Learning of Procedures from Web Instructional Videos", "VATEX: A Large-Scale, High-Quality Multilingual Dataset for\n Video-and-Language Research"], "answer_arxiv_id": ["1501.02530", "1708.01641", "1703.09788", "1904.03493"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_13224"} +{"question": "Which work indicates that gradual pruning, combined with one-shot pruning algorithms, can outperform unstructured pruning methods?", "answer": ["WoodFisher: Efficient Second-Order Approximation for Neural Network Compression"], "answer_arxiv_id": ["2004.14340"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_13225"} +{"question": "What work has been done in the field of contrastive learning-based multimodal models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "SLIP: Self-supervision meets Language-Image Pre-training"], "answer_arxiv_id": ["2103.00020", "2201.12086", "2112.12750"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13226"} +{"question": "Which studies utilized transformer-based trackers for deep tracking methods?", "answer": ["Transformer Tracking", "Joint Feature Learning and Relation Modeling for Tracking: A One-Stream\n Framework", "Efficient Visual Tracking with Exemplar Transformers", "CiteTracker: Correlating Image and Text for Visual Tracking"], "answer_arxiv_id": ["2103.15436", "2203.11991", "2112.09686", "2308.11322"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_13227"} +{"question": "Which papers present a single policy that is trained to be robust to changes that may occur during testing in the RL domain?", "answer": ["Robust Adversarial Reinforcement Learning", "Quantifying Generalization in Reinforcement Learning", "Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "CAD2RL: Real Single-Image Flight Without a Single Real Image", "Sim-to-Real Transfer of Robotic Control with Dynamics Randomization", "Solving Rubik’s Cube with a Robot Hand", "Learning Dexterous In-Hand Manipulation", "Zero-Shot Terrain Generalization for Visual Locomotion Policies", "Sim-to-Real Transfer in Deep Reinforcement Learning for Robotics: a Survey"], "answer_arxiv_id": ["1703.02702", "1812.02341", "1703.06907", "1611.04201", "1710.06537", "1910.07113", "1808.00177", "2011.05513", "2009.13303"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_13228"} +{"question": "Which recent works show that utilizing out of domain data can be more effective than filtering them?", "answer": ["An Empirical Study and Analysis on Open-Set Semi-Supervised Learning", "On Pseudo-Labeling for Class-Mismatch Semi-Supervised Learning", "They are Not Completely Useless: Towards Recycling Transferable Unlabeled Data for Class-Mismatched Semi-Supervised Learning"], "answer_arxiv_id": ["2101.08237", "2301.06010", "2011.13529"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_13229"} +{"question": "Which works combine forms of Ewald summation with MPNN models?", "answer": ["Efficient Approximations of Complete Interatomic Potentials for Crystal Property Prediction"], "answer_arxiv_id": ["2306.10045"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_13230"} +{"question": "What works have applied domain adaptation methods in legal text classification tasks?", "answer": ["Improved Multi-label Classification under Temporal Concept Drift:\n Rethinking Group-Robust Algorithms in a Label-Wise Setting", "Zero-shot Transfer of Article-aware Legal Outcome Classification for\n European Court of Human Rights Cases"], "answer_arxiv_id": ["2203.07856", "2302.00609"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_13231"} +{"question": "Can you cite some works that use contrastive methods to learn representations from various views of the input graph?", "answer": ["InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization", "Graph Contrastive Learning with Augmentations", "Pre-training Molecular Graph Representation with 3D Geometry"], "answer_arxiv_id": ["1908.01000", "2010.13902", "2110.07728"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_13232"} +{"question": "Which work indicates LLMs may encounter difficulties in generating planning goals when it comes to numerical or spatial reasoning?", "answer": ["Translating Natural Language to Planning Goals with Large-Language\n Models"], "answer_arxiv_id": ["2302.05128"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_13233"} +{"question": "Which research uses each variable as a graph and uses the graph-Cartesian product to represent the search space in combinatorial BO?", "answer": ["Combinatorial Bayesian Optimization using the Graph Cartesian Product"], "answer_arxiv_id": ["1902.00448"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_13234"} +{"question": "Which papers build on the lottery ticket existence theory to prove that sparse random source networks contain strong lottery tickets?", "answer": ["Proving the Lottery Ticket Hypothesis: Pruning is All You Need", "Optimal Lottery Tickets via SubsetSum: Logarithmic Over-Parameterization is Sufficient", "Logarithmic Pruning is All You Need", "On the Existence of Universal Lottery Tickets", "Most Activation Functions Can Win the Lottery Without Excessive Depth", "A General Framework For Proving The Equivariant Strong Lottery Ticket Hypothesis"], "answer_arxiv_id": ["2002.00585", "2006.07990", "2006.12156", "2111.11146", "2205.02321", "2206.04270"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_13235"} +{"question": "Which studies focus on obtaining high-quality task-related tokens?", "answer": ["Prompt Distribution Learning", "Distribution-Aware Prompt Tuning for Vision-Language Models"], "answer_arxiv_id": ["2205.03340", "2309.03406"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13236"} +{"question": "Could you refer me to works that discuss the trade-off between fairness and classification accuracy?", "answer": ["Conditional Learning of Fair Representations"], "answer_arxiv_id": ["1910.07162"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_13237"} +{"question": "What sources are commonly used to retrieve bounding boxes for pre-training?", "answer": ["Microsoft COCO: Common Objects in Context", "Visual Genome: Connecting Language and Vision Using Crowdsourced Dense\n Image Annotations"], "answer_arxiv_id": ["1405.0312", "1602.07332"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_13238"} +{"question": "Which paper proposes using a powerful detector to generate bounding boxes for each instance in instance segmentation?", "answer": ["Mask R-CNN", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks"], "answer_arxiv_id": ["1703.06870", "1506.01497"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13239"} +{"question": "What works have been done related to sharpness-aware minimization?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization", "ASAM: Adaptive Sharpness-Aware Minimization for Scale-Invariant Learning of Deep Neural Networks", "Regularizing Neural Networks via Adversarial Model Perturbation", "Efficient Sharpness-aware Minimization for Improved Training of Neural Networks", "When Vision Transformers Outperform ResNets without Pre-training or Strong Data Augmentations"], "answer_arxiv_id": ["2010.01412", "2102.11600", "2010.04925", "2110.03141", "2106.01548"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_13240"} +{"question": "What papers have led to a wave of innovation in universal segmentation by developing mask classification architectures?", "answer": ["K-Net: Towards Unified Image Segmentation", "MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "Masked-attention Mask Transformer for Universal Image Segmentation", "Panoptic-PartFormer: Learning a Unified Model for Panoptic Part\n Segmentation"], "answer_arxiv_id": ["2106.14855", "2012.00759", "2112.01527", "2204.04655"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_13241"} +{"question": "Which studies propose a fully reversible network that can reverse back to the input without any information loss?", "answer": ["i-RevNet: Deep Invertible Networks"], "answer_arxiv_id": ["1802.07088"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_13242"} +{"question": "Which work establishes a connection between diffusion models and denoising score-based models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_13243"} +{"question": "Which papers present alternatives to handle infinitely long sequences without overwhelming computation or memory overhead in self-attention computation?", "answer": ["Longformer: The Long-Document Transformer", "Efficient Streaming Language Models with Attention Sinks"], "answer_arxiv_id": ["2004.05150", "2309.17453"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_13244"} +{"question": "What papers are about designing ranking-based loss function for object detection tasks?", "answer": ["DR Loss: Improving Object Detection by Distributional Ranking", "AP-Loss for Accurate One-Stage Object Detection", "A Ranking-based, Balanced Loss Function Unifying Classification and Localisation in Object Detection", "Rank & Sort Loss for Object Detection and Instance Segmentation", "Correlation Loss: Enforcing Correlation between Classification and Localization"], "answer_arxiv_id": ["1907.10156", "2008.07294", "2009.13592", "2107.11669", "2301.01019"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_13245"} +{"question": "Could you provide me some studies that focus on improving Monte Carlo Tree Search?", "answer": ["Learning to Search with MCTSnets", "Solving Hard AI Planning Instances Using Curriculum-Driven Deep Reinforcement Learning", "Thinking Fast and Slow with Deep Learning and Tree Search"], "answer_arxiv_id": ["1802.04697", "2006.02689", "1705.08439"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_13246"} +{"question": "What research papers propose weak supervision frameworks using multiple sources for noisy label estimates?", "answer": ["Neural Ranking Models with Weak Supervision", "Snorkel: Rapid Training Data Creation with Weak Supervision"], "answer_arxiv_id": ["1704.08803", "1711.10160"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_13247"} +{"question": "Which studies utilize dynamic convolution methods or self-attention methods in efficient video processing?", "answer": ["Dynamic Kernel Distillation for Efficient Pose Estimation in Videos", "Temporally Distributed Networks for Fast Video Semantic Segmentation", "Learning Trajectory-Aware Transformer for Video Super-Resolution"], "answer_arxiv_id": ["1908.09216", "2004.01800", "2204.04216"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_13248"} +{"question": "Could you provide me some works about point cloud segmentation?", "answer": ["Analyzing Deep Learning Representations of Point Clouds for Real-Time In-Vehicle LiDAR Perception", "Towards Semantic Segmentation of Urban-Scale 3D Point Clouds: A Dataset, Benchmarks and Challenges", "RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds", "KPConv: Flexible and Deformable Convolution for Point Clouds", "Learning Semantic Segmentation of Large-Scale Point Clouds with Random Sampling", "PIDS: Joint Point Interaction-Dimension Search for 3D Point Cloud", "Using a Waffle Iron for Automotive Point Cloud Semantic Segmentation", "SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation", "Scan-based Semantic Segmentation of LiDAR Point Clouds: An Experimental Study", "FIDNet: LiDAR Point Cloud Semantic Segmentation with Fully Interpolation Decoding", "CENet: Toward Concise and Efficient LiDAR Semantic Segmentation for Autonomous Driving", "Rethinking Range View Representation for LiDAR Segmentation", "PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation", "Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentation", "PolarStream: Streaming Lidar Object Detection and Segmentation with Polar Pillars", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation", "LiDAR-based Panoptic Segmentation via Dynamic Shifting Network", "AMVNet: Assertion-based Multi-View Fusion Network for LiDAR Semantic Segmentation", "Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution", "RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR Point Cloud Segmentation", "Perception-Aware Multi-Sensor Fusion for 3D LiDAR Semantic Segmentation", "GFNet: Geometric Flow Network for 3D Point Cloud Semantic Segmentation"], "answer_arxiv_id": ["2210.14612", "2009.03137", "1911.11236", "1904.08889", "2107.02389", "2211.15759", "2301.10100v2", "2004.01803", "2004.11803", "2109.03787v1", "2207.12691", "2303.05367", "2003.14032", "2103.14962", "2106.07545", "1904.08755", "2011.10033", "2011.11964", "2012.04934", "2007.16100", "2103.12978", "2106.15277", "2207.02605"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_13249"} +{"question": "Any examples of studies that use causal inference to improve the performance of models for visual question answering?", "answer": ["Counterfactual VQA: A Cause-Effect Look at Language Bias"], "answer_arxiv_id": ["2006.04315v4"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_13250"} +{"question": "What study shows the effectiveness of integrated positional encoding when applied to the multi-scale hybrid tri-plane representation?", "answer": ["Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural\n Radiance Fields"], "answer_arxiv_id": ["2307.11335"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_13251"} +{"question": "Can you point to the publications that have approached parameter and resource efficient adaptation using input prefix fine-tuning?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2104.08691", "2101.00190"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_13252"} +{"question": "What paper used gradient penalty in the objective function to avoid local equilibrium in GANs training?", "answer": ["On Convergence and Stability of GANs"], "answer_arxiv_id": ["1705.07215"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_13253"} +{"question": "Which works used diffusion models for super-resolution application?", "answer": ["Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2106.15282"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_13254"} +{"question": "What research has been done on establishing information-theoretic lower bounds in online RL?", "answer": ["Episodic Reinforcement Learning in Finite MDPs: Minimax Lower Bounds Revisited"], "answer_arxiv_id": ["2010.03531"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_13255"} +{"question": "Where meta-learning under adversarial bandit feedback has been studied?", "answer": ["Non-stationary Bandits and Meta-Learning with a Small Set of Optimal Arms"], "answer_arxiv_id": ["2202.13001"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_13256"} +{"question": "In which studies large-scale image-text encoder models e.g. CLIP and ALIGN were used for semantic interaction between vision and language?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_13257"} +{"question": "Which research papers discuss the DIstribution Correction Estimation (DICE) method for offline imitation learning?", "answer": ["LobsDICE: Offline Learning from Observation via Stationary Distribution Correction Estimation", "Imitation Learning via Off-Policy Distribution Matching", "OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation"], "answer_arxiv_id": ["2202.13536", "1912.05032", "2106.10783"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_13258"} +{"question": "Which paper is adopted by most recent work in supervised skeleton-based action recognition?", "answer": ["Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition"], "answer_arxiv_id": ["1801.07455"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_13259"} +{"question": "What research discusses the complications of extending libraries to cover molecular representations?", "answer": ["Using Bayesian Optimization to Accelerate Virtual Screening for the Discovery of Therapeutics Appropriate for Repurposing for COVID-19", "Self-focusing virtual screening with active design space pruning"], "answer_arxiv_id": ["2005.07121v1", "2205.01753"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_13260"} +{"question": "Which papers focus on the development of partial identification bounds for policy learning and evaluation?", "answer": ["Confounding-Robust Policy Evaluation in Infinite-Horizon Reinforcement Learning", "Off-policy Policy Evaluation For Sequential Decisions Under Unobserved Confounding", "Estimating and Improving Dynamic Treatment Regimes With a Time-Varying Instrumental Variable"], "answer_arxiv_id": ["2002.04518", "2003.05623", "2104.07822"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_13261"} +{"question": "Which references focused on utilizing instruction tuning or reinforcement learning to generate better critiques?", "answer": ["Learning to Simulate Natural Language Feedback for Interactive Semantic\n Parsing", "Self-critiquing models for assisting human evaluators", "Making Language Models Better Tool Learners with Execution Feedback", "Constitutional AI: Harmlessness from AI Feedback"], "answer_arxiv_id": ["2305.08195", "2206.05802v2", "2305.13068", "2212.08073"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_13262"} +{"question": "Which works trade off between efficiency and performance in model design by restricting transformers blocks?", "answer": ["MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer", "Mobile-Former: Bridging MobileNet and Transformer", "TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation", "EfficientFormer: Vision Transformers at MobileNet Speed"], "answer_arxiv_id": ["2110.02178", "2108.05895", "2204.05525", "2206.01191"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_13263"} +{"question": "Are there any studies focusing on the estimation of low-rank transition matrices of Markov chains?", "answer": ["Spectral State Compression of Markov Processes", "Estimation of Markov Chain via Rank-Constrained Likelihood", "Learning Markov models via low-rank optimization"], "answer_arxiv_id": ["1802.02920v3", "1804.00795v2", "1907.00113v2"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_13264"} +{"question": "What research implicitly improves policies for multiple preferences in multi-objective RL problems?", "answer": ["Dynamic Weights in Multi-Objective Deep Reinforcement Learning", "A Distributional View on Multi-Objective Policy Optimization"], "answer_arxiv_id": ["1809.07803", "2005.07513"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_13265"} +{"question": "Are there studies that assumed the presence of honest nodes in Federated Learning and its applicability in some scenarios?", "answer": ["Collaborative Machine Learning with Incentive-Aware Model Rewards", "Incentivizing Collaboration in Machine Learning via Synthetic Data Rewards"], "answer_arxiv_id": ["2010.12797", "2112.09327"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_13266"} +{"question": "What studies have tried to enhance the radiance field performance through the application of the anti-aliased grid-based technique?", "answer": ["Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2304.06706"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_13267"} +{"question": "What are the works that study NeuralODEs and neuralSDEs as continuous-time models?", "answer": ["Neural Ordinary Differential Equations", "Scalable Gradients for Stochastic Differential Equations"], "answer_arxiv_id": ["1806.07366", "2001.01328"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_13268"} +{"question": "Which works pointed out that LLMs have verbosity and self-enhancement bias?", "answer": ["Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena"], "answer_arxiv_id": ["2306.05685"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_13269"} +{"question": "Which studies have proposed uncertainty quantification methods for both classification and regression?", "answer": ["On Calibration of Modern Neural Networks", "Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning", "Calibration of Neural Networks using Splines", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration", "A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Accurate Uncertainties for Deep Learning Using Calibrated Regression", "Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift", "What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?"], "answer_arxiv_id": ["1706.04599", "2003.07329", "2006.12800", "1910.12656", "2011.06225", "1506.02142", "1612.01474", "1807.00263", "1906.02530", "1703.04977"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13270"} +{"question": "Are there any studies that improved object detection abilities through optimization of event representations and incorporation of swin transformer architecture?", "answer": ["From Chaos Comes Order: Ordering Event Representations for Object\n Recognition and Detection"], "answer_arxiv_id": ["2304.13455"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_13271"} +{"question": "Are there any proposals for high-quality LiDAR data synthesis using neural fields?", "answer": ["Neural LiDAR Fields for Novel View Synthesis"], "answer_arxiv_id": ["2305.01643v2"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13272"} +{"question": "What papers have further refined the approach of aligning CNN-based features of images with textual embeddings?", "answer": ["VSE++: Improving Visual-Semantic Embeddings with Hard Negatives"], "answer_arxiv_id": ["1707.05612"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13273"} +{"question": "Which study described an algorithm exploiting truncation strategy, for example by using TOFU as an instance?", "answer": ["Almost Optimal Algorithms for Linear Stochastic Bandits with Heavy-Tailed Payoffs"], "answer_arxiv_id": ["1810.10895"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_13274"} +{"question": "Which paper does the current work extend by allowing any non-negative violation tolerance?", "answer": ["Iterative Methods for Private Synthetic Data: Unifying Framework and New Methods"], "answer_arxiv_id": ["2106.07153"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_13275"} +{"question": "What is the first study that tackled the large city-scale outdoor scene localization task?", "answer": ["Text2Pos: Text-to-Point-Cloud Cross-Modal Localization"], "answer_arxiv_id": ["2203.15125"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_13276"} +{"question": "What research papers discussed privacy-preserving filtering techniques in action recognition?", "answer": ["Privacy-Preserving Action Recognition for Smart Hospitals using Low-Resolution Depth Images", "Privacy-Preserving Human Activity Recognition from Extreme Low Resolution"], "answer_arxiv_id": ["1811.09950", "1604.03196"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_13277"} +{"question": "Which works related to data augmentation techniques were used to improve graphing training distribution?", "answer": ["Data Augmentation for Deep Graph Learning: A Survey", "Graph Data Augmentation for Graph Machine Learning: A Survey", "Model-Agnostic Augmentation for Accurate Graph Classification"], "answer_arxiv_id": ["2202.08235", "2202.08871", "2202.10107"], "source_meta": {"published_time": "20221105"}, "qid": "AutoScholarQuery_train_13278"} +{"question": "In which study was the encoder-decoder layers modified by letting the decoders attend to all encoder layers?", "answer": ["Training Deeper Neural Machine Translation Models with Transparent Attention"], "answer_arxiv_id": ["1808.07561"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_13279"} +{"question": "What papers have investigated learning diverse classifiers?", "answer": ["Improving the Adversarial Robustness and Interpretability of Deep Neural Networks by Regularizing their Input Gradients", "Evading the Simplicity Bias: Training a Diverse Set of Models Discovers Solutions with Superior OOD Generalization", "Controlling directions orthogonal to a classifier", "Diversify and Disambiguate: Learning From Underspecified Data"], "answer_arxiv_id": ["1711.09404", "2105.05612", "2201.11259", "2202.03418v3"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_13280"} +{"question": "Which research papers have proposed methods for detecting and alleviating the negative impact from pseudo-labels for a better self-training scheme?", "answer": ["Robust Target Training for Multi-Source Domain Adaptation"], "answer_arxiv_id": ["2210.01676"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13281"} +{"question": "Which works have shown that deterministic next-step forecasting models may result in poor or unstable rollouts due to compounding prediction errors?", "answer": ["Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge", "Forecasting Global Weather with Graph Neural Networks", "ClimaX: A foundation model for weather and climate"], "answer_arxiv_id": ["1711.07970", "2202.07575", "2301.10343"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_13282"} +{"question": "Which papers brought theoretical results on various aspects of reinforcement learning?", "answer": ["Provably efficient RL with Rich Observations via Latent State Decoding", "Is a Good Representation Sufficient for Sample Efficient Reinforcement Learning?", "Provably Efficient Exploration in Policy Optimization"], "answer_arxiv_id": ["1901.09018", "1910.03016", "1912.05830"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_13283"} +{"question": "What research adopts 3D graph convolution to obtain geometric sensitivity in object variation studies?", "answer": ["FS-Net: Fast Shape-based Network for Category-Level 6D Object Pose\n Estimation with Decoupled Rotation Mechanism"], "answer_arxiv_id": ["2103.07054"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_13284"} +{"question": "Which papers considered the curvature of loss function for improvising Adam optimizers?", "answer": ["AdaBelief Optimizer: Adapting Stepsizes by the Belief in Observed Gradients"], "answer_arxiv_id": ["2010.07468"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_13285"} +{"question": "What references are there where decision-making is based on actual physical evaluation when using surrogate models?", "answer": ["RoMA: Robust Model Adaptation for Offline Model-based Optimization", "Conservative Objective Models for Effective Offline Model-Based Optimization", "Global Optimization Networks"], "answer_arxiv_id": ["2110.14188", "2107.06882", "2202.01277"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_13286"} +{"question": "What research papers propose the use of contrastive learning in representation learning in deep learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1911.05722", "2006.09882", "2006.07733", "2103.00020"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_13287"} +{"question": "Which paper introduced LERF, a method for training a feature field of the Vision-Language Model together with the radiance field?", "answer": ["LERF: Language Embedded Radiance Fields"], "answer_arxiv_id": ["2303.09553"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_13288"} +{"question": "What studies utilized Gromov-Wasserstein (GW) couplings in tasks such as graph node matching and partitioning?", "answer": ["Scalable Gromov-Wasserstein Learning for Graph Partitioning and Matching", "Gromov-Wasserstein Learning for Graph Matching and Node Embedding", "Generalized Spectral Clustering via Gromov-Wasserstein Learning"], "answer_arxiv_id": ["1905.07645", "1901.06003", "2006.04163"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_13289"} +{"question": "Which study investigates the balancedness of subset sizes of RPMs?", "answer": ["Why the Rich Get Richer? On the Balancedness of Random Partition Models"], "answer_arxiv_id": ["2201.12697"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_13290"} +{"question": "Which studies propose gradient-based post-hoc interpretation methods?", "answer": ["Learning Deep Features for Discriminative Localization", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "Axiomatic Attribution for Deep Networks", "Learning Important Features Through Propagating Activation Differences"], "answer_arxiv_id": ["1512.04150", "1610.02391", "1703.01365", "1704.02685"], "source_meta": {"published_time": "20221030"}, "qid": "AutoScholarQuery_train_13291"} +{"question": "Which works propose a bundle adjustment method in dense visual SLAM algorithms to optimize keyframe poses and construct the dense 3D structure?", "answer": ["BundleFusion: Real-time Globally Consistent 3D Reconstruction using\n On-the-fly Surface Re-integration", "BA-Net: Dense Bundle Adjustment Network"], "answer_arxiv_id": ["1604.01093", "1806.04807"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_13292"} +{"question": "What papers are about the application of NeRFs in 3D panoptic segmentation?", "answer": ["Panoptic Neural Fields: A Semantic Object-Aware Neural Scene Representation", "Panoptic Lifting for 3D Scene Understanding with Neural Fields", "Panoptic NeRF: 3D-to-2D Label Transfer for Panoptic Urban Scene Segmentation", "DM-NeRF: 3D Scene Geometry Decomposition and Manipulation from 2D Images", "Instance Neural Radiance Field"], "answer_arxiv_id": ["2205.04334", "2212.09802", "2203.15224", "2208.07227", "2304.04395"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_13293"} +{"question": "What studies have been made on data augmentation by interpolation as a way to mitigate the trade-off phenomenon?", "answer": ["Adversarial Vertex Mixup: Toward Better Adversarially Robust Generalization", "Interpolated Adversarial Training: Achieving Robust Neural Networks without Sacrificing Too Much Accuracy", "Adversarial Domain Adaptation with Domain Mixup"], "answer_arxiv_id": ["2003.02484", "1906.06784v7", "1912.01805"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_13294"} +{"question": "Which studies have explored the topic of open vocabulary in the context of 3D scene graphs?", "answer": ["Context-Aware Entity Grounding with Open-Vocabulary 3D Scene Graphs", "ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and Planning"], "answer_arxiv_id": ["2309.15940", "2309.16650v1"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_13295"} +{"question": "Which papers discuss the usage of input regularization in neural network visualization?", "answer": ["Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations"], "answer_arxiv_id": ["2201.12961"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_13296"} +{"question": "Could you provide me some works about application of image-language models to zero-shot text-to-3D generation tasks?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "Zero-Shot Text-Guided Object Generation with Dream Fields", "Text2Mesh: Text-Driven Neural Stylization for Meshes", "CLIP-Forge: Towards Zero-Shot Text-to-Shape Generation", "Understanding Pure CLIP Guidance for Voxel Grid NeRF Models", "Text and Image Guided 3D Avatar Generation and Manipulation", "ClipFace: Text-guided Editing of Textured 3D Morphable Models", "ClipMatrix: Text-controlled Creation of 3D Textured Meshes", "Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2205.08535", "2112.01455", "2112.03221", "2110.02624", "2209.15172", "2202.06079", "2212.01406", "2109.12922", "2212.14704"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_13297"} +{"question": "What research papers feature the development of differentiable simulators?", "answer": ["DiffTaichi: Differentiable Programming for Physical Simulation", "Learning to Simulate Complex Physics with Graph Networks", "Additive manufacturing process design with differentiable simulations", "Automated shape differentiation in the Unified Form Language"], "answer_arxiv_id": ["1910.00935", "2002.09405", "2107.10919", "1808.08083"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_13298"} +{"question": "What work showcased an underperformance of existing methods in sample sizing on large datasets with the help of data pruning?", "answer": ["Beyond neural scaling laws: beating power law scaling via data pruning"], "answer_arxiv_id": ["2206.14486v6"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_13299"} +{"question": "Where is MDKAE, which disentangles dominant factors using Koopman spectral analysis, discussed?", "answer": ["Multifactor Sequential Disentanglement via Structured Koopman Autoencoders"], "answer_arxiv_id": ["2303.17264"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13300"} +{"question": "Who employed a transformer-based encoder-decoder architecture with skip connections from encoders to decoders for depth prediction?", "answer": ["Attention Attention Everywhere: Monocular Depth Prediction with Skip\n Attention"], "answer_arxiv_id": ["2210.09071"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_13301"} +{"question": "What research papers have studied the use of designated gates and introduced the idea of specialization in the context of multi-task learning?", "answer": ["Expert Gate: Lifelong Learning with a Network of Experts", "Continual Learning Through Synaptic Intelligence", "Superposition of many models into one"], "answer_arxiv_id": ["1611.06194", "1703.04200", "1902.05522"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_13302"} +{"question": "Which studies were primarily concerned with the conventional semantic segmentation methods?", "answer": ["Rethinking Atrous Convolution for Semantic Image Segmentation", "SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers", "Learning Statistical Texture for Semantic Segmentation", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "FVP: Fourier Visual Prompting for Source-Free Unsupervised Domain\n Adaptation of Medical Image Segmentation"], "answer_arxiv_id": ["1706.05587", "2105.15203", "2103.04133", "2107.06278", "2304.13672"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_13303"} +{"question": "Which work introduced key components for complex-valued deep neural networks?", "answer": ["Deep Complex Networks"], "answer_arxiv_id": ["1705.09792"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_13304"} +{"question": "What research papers are about creating solutions for optimization instability and over-smoothing in deep transformers?", "answer": ["Improving Deep Transformer with Depth-Scaled Initialization and Merged Attention", "Learning Deep Transformer Models for Machine Translation", "Understanding the Difficulty of Training Transformers", "ReZero is All You Need: Fast Convergence at Large Depth", "Revisiting Over-smoothing in BERT from the Perspective of Graph", "Anti-Oversmoothing in Deep Vision Transformers via the Fourier Domain Analysis: From Theory to Practice", "Attention is not all you need: pure attention loses rank doubly exponentially with depth"], "answer_arxiv_id": ["1908.11365", "1906.01787", "2004.08249", "2003.04887", "2202.08625", "2203.05962", "2103.03404"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_13305"} +{"question": "Could you provide me some of the studies about enhancing long context understanding through efficient attention?", "answer": ["Big Bird: Transformers for Longer Sequences", "LongLoRA: Efficient Fine-tuning of Long-Context Large Language Models"], "answer_arxiv_id": ["2007.14062", "2309.12307"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_13306"} +{"question": "Are there any research papers that build vision-centric frameworks with pre-trained LLMs using autoregressive modeling?", "answer": ["VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks"], "answer_arxiv_id": ["2305.11175"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_13307"} +{"question": "What studies apply intra-layer parallelism to deep neural networks in combination with other strategies?", "answer": ["One weird trick for parallelizing convolutional neural networks", "Beyond Data and Model Parallelism for Deep Neural Networks"], "answer_arxiv_id": ["1404.5997", "1807.05358"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_13308"} +{"question": "Any research that describes the use of natural language to derive potential-based shaping rewards?", "answer": ["Using Natural Language for Reward Shaping in Reinforcement Learning"], "answer_arxiv_id": ["1903.02020"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_13309"} +{"question": "Which works focus on extracting and representing causal relationships among entities in text based on commonsense knowledge?", "answer": ["Atomic: An Atlas of Machine Commonsense for If-Then Reasoning", "COMET: Commonsense Transformers for Automatic Knowledge Graph Construction", "CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge"], "answer_arxiv_id": ["1811.00146", "1906.05317", "1811.00937"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_13310"} +{"question": "What works popularized the term double descent?", "answer": ["To Understand Deep Learning We Need to Understand Kernel Learning"], "answer_arxiv_id": ["1802.01396"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_13311"} +{"question": "What work is known for finding the lottery tickets prior to Federated Learning training?", "answer": ["LotteryFL: Personalized and Communication-Efficient Federated Learning with Lottery Ticket Hypothesis on Non-IID Datasets"], "answer_arxiv_id": ["2008.03371"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13312"} +{"question": "Could you provide me some works that have significantly contributed to the advancements in natural language processing through instruction tuning?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "The Flan Collection: Designing Data and Methods for Effective\n Instruction Tuning"], "answer_arxiv_id": ["2109.01652", "2110.08207", "2301.13688"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_13313"} +{"question": "Which research papers described a detector type based on meta-classification for backdoor detection?", "answer": ["Detecting AI Trojans Using Meta Neural Analysis", "Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs"], "answer_arxiv_id": ["1910.03137", "1906.10842"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_13314"} +{"question": "What studies first discovered the barren plateau of Quantum Neural Networks (QNNs)?", "answer": ["Barren plateaus in quantum neural network training landscapes"], "answer_arxiv_id": ["1803.11173v1"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_13315"} +{"question": "What were some approaches that formalized the task of learning logical rules as a constrained optimization procedure?", "answer": ["OptNet: Differentiable Optimization as a Layer in Neural Networks", "SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver", "DeepLogic: Towards End-to-End Differentiable Logical Reasoning", "An Integer Linear Programming Framework for Mining Constraints from Data", "Diff-Explainer: Differentiable Convex Optimization for Explainable Multi-hop Inference", "Fast Differentiable Sorting and Ranking"], "answer_arxiv_id": ["1703.00443", "1905.12149", "1805.07433", "2006.10836", "2105.03417", "2002.08871"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_13316"} +{"question": "What research papers provide examples of using multi-modal tasks in instruction tuning?", "answer": ["MultiInstruct: Improving Multi-Modal Zero-Shot Learning via Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models", "Visual Instruction Tuning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning"], "answer_arxiv_id": ["2212.10773", "2304.10592", "2304.08485", "2305.03726"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_13317"} +{"question": "Are there any studies that analyze the effect of learning rate on SGD using a stochastic differential equation?", "answer": ["On Learning Rates and Schrödinger Operators"], "answer_arxiv_id": ["2004.06977"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_13318"} +{"question": "What works propose Gradient Matching methods for creating compact synthetic datasets?", "answer": ["Dataset Condensation with Gradient Matching", "Dataset Condensation with Differentiable Siamese Augmentation", "Dataset Condensation with Contrastive Signals", "Dataset Condensation via Efficient Synthetic-Data Parameterization"], "answer_arxiv_id": ["2006.05929", "2102.08259", "2202.02916", "2205.14959v2"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_13319"} +{"question": "Which research introduces interaction networks, a model that performs simulations by combining an object-centric and relation-centric component?", "answer": ["Interaction Networks for Learning about Objects, Relations and Physics"], "answer_arxiv_id": ["1612.00222"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_13320"} +{"question": "Which works use CLIP as the backbone for few-shot settings?", "answer": ["Learning to Prompt for Vision-Language Models", "Unified Vision and Language Prompt Learning"], "answer_arxiv_id": ["2109.01134", "2210.07225"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_13321"} +{"question": "Could you provide me some studies on VLN models for ALFRED that employ modular setups?", "answer": ["FILM: Following Instructions in Language with Modular Methods", "Prompter: Utilizing Large Language Model Prompting for a Data Efficient Embodied Instruction Following"], "answer_arxiv_id": ["2110.07342", "2211.03267"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_13322"} +{"question": "Could you provide me studies on the construction of a more flexible Bayesian pseudocoreset variational posterior?", "answer": ["Bayesian Inference via Sparse Hamiltonian Flows"], "answer_arxiv_id": ["2203.05723"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_13323"} +{"question": "What works improved the performance of meta-learning methods using Kernel ridge regression (KRR)?", "answer": ["Dataset Meta-Learning from Kernel Ridge-Regression", "Dataset Distillation with Infinitely Wide Convolutional Networks", "Efficient Dataset Distillation using Random Feature Approximation"], "answer_arxiv_id": ["2011.00050", "2107.13034", "2210.12067"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_13324"} +{"question": "What are the studies that focus on regret minimization problem in the multi-armed bandit model?", "answer": ["Regret Analysis of Stochastic and Nonstochastic Multi-armed Bandit Problems", "Analysis of Thompson Sampling for the multi-armed bandit problem"], "answer_arxiv_id": ["1204.5721v2", "1111.1797"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_13325"} +{"question": "Who proposed an algorithm to achieve a nearly minimax optimal regret bound in episodic MDP?", "answer": ["Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes"], "answer_arxiv_id": ["2012.08507"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_13326"} +{"question": "What work extends DMFAL to support batch active-learning using greedy approach?", "answer": ["Batch Multi-Fidelity Active Learning with Budget Constraints"], "answer_arxiv_id": ["2210.12704"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_13327"} +{"question": "Which papers introduce different methods for novel view synthesis from a single input image?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis", "SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image", "Stereo Magnification: Learning view synthesis using multiplane images", "Single-View View Synthesis with Multiplane Images", "SynSin: End-to-end View Synthesis from a Single Image", "PixelSynth: Generating a 3D-Consistent Experience from a Single Image", "Worldsheet: Wrapping the World in a 3D Sheet for View Synthesis from a Single Image", "Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis"], "answer_arxiv_id": ["2012.02190", "2104.00677", "2204.00928", "1805.09817", "2004.11364", "1912.08804", "2108.05892", "2012.09854", "2104.00677"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13328"} +{"question": "Could you provide examples of research that apply the influence function to measure the influential contribution of training data?", "answer": ["Understanding Black-box Predictions via Influence Functions"], "answer_arxiv_id": ["1703.04730"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_13329"} +{"question": "Which studies focused on image generation from structured layouts with fixed classes for content control?", "answer": ["Semantic Image Synthesis with Spatially-Adaptive Normalization", "Image-to-Image Translation with Conditional Adversarial Networks", "Object-Centric Image Generation from Layouts", "Context-Aware Layout to Image Generation with Enhanced Object Appearance", "Attribute-guided image generation from layout", "Prompt-Free Diffusion: Taking \"Text\" out of Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["1903.07291", "1611.07004", "2003.07449", "2103.11897", "2008.11932", "2305.16223"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_13330"} +{"question": "Could you list some papers discussing dense retrieval using bi-encoder model?", "answer": ["Dense Passage Retrieval for Open-Domain Question Answering", "Latent Retrieval for Weakly Supervised Open Domain Question Answering", "RocketQA: An Optimized Training Approach to Dense Passage Retrieval for Open-Domain Question Answering", "Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval"], "answer_arxiv_id": ["2004.04906", "1906.00300", "2010.08191", "2007.00808"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_13331"} +{"question": "Which studies use CAD models as a supplementary data source for monocular 3D object detection?", "answer": ["AutoShape: Real-Time Shape-Aware Monocular 3D Object Detection"], "answer_arxiv_id": ["2108.11127"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_13332"} +{"question": "Which works introduced perturbation-based methods for the model self-detection approach?", "answer": ["DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability\n Curvature"], "answer_arxiv_id": ["2301.11305"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_13333"} +{"question": "Could you provide works that developed customized adversarial samples with natural-style?", "answer": ["Adversarial Camouflage: Hiding Physical-World Attacks with Natural Styles", "Generating Natural Adversarial Examples", "AdvART: Adversarial Art for Camouflaged Object Detection Attacks"], "answer_arxiv_id": ["2003.08757", "1710.11342", "2303.01734v2"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13334"} +{"question": "Which papers discuss the limitations of supervised methods used for MAR, including issues with out-of-domain problems and requirement of large-scale training dataset?", "answer": ["Solving Inverse Problems in Medical Imaging with Score-Based Generative Models", "ADN: Artifact Disentanglement Network for Unsupervised Metal Artifact Reduction"], "answer_arxiv_id": ["2111.08005", "1908.01104"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_13335"} +{"question": "What works are related to confident model acceleration?", "answer": ["Consistent Accelerated Inference via Confident Adaptive Transformers", "Confident Adaptive Language Modeling"], "answer_arxiv_id": ["2104.08803", "2207.07061"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_13336"} +{"question": "What studies are about exploiting similar ideas in full-observation matrix sensing?", "answer": ["Implicit Regularization of Discrete Gradient Dynamics in Linear Neural Networks", "The Implicit Bias of Depth: How Incremental Learning Drives Generalization"], "answer_arxiv_id": ["1904.13262", "1909.12051"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_13337"} +{"question": "What works apply layer-wise message passing scheme in Graph Neural Networks?", "answer": ["Neural Message Passing for Quantum Chemistry", "Adaptive Kernel Graph Neural Network"], "answer_arxiv_id": ["1704.01212", "2112.04575"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_13338"} +{"question": "Could you provide some references related to dynamic sparse training?", "answer": ["Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science", "Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization", "Deep Rewiring: Training very sparse deep networks", "Rigging the Lottery: Making All Tickets Winners", "Sparse Networks from Scratch: Faster Training without Losing Performance", "Do We Actually Need Dense Over-Parameterization? In-Time Over-Parameterization in Sparse Training", "Sparse Training via Boosting Pruning Plasticity with Neuroregeneration", "MEST: Accurate and Fast Memory-Economic Sparse Training Framework on the Edge", "AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural Networks", "Top-KAST: Top-K Always Sparse Training", "Powerpropagation: A sparsity inducing weight reparameterisation", "CHEX: CHannel EXploration for CNN Model Compression", "More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using Sparsity", "How Well Do Sparse ImageNet Models Transfer?"], "answer_arxiv_id": ["1707.04780v2", "1902.05967", "1711.05136", "1911.11134", "1907.04840", "2102.02887", "2106.10404", "2110.14032", "2106.12379", "2106.03517", "2110.00296", "2203.15794", "2207.03620", "2111.13445"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13339"} +{"question": "Which paper introduced the NTK to demonstrate the training of a deep and infinitely wide neural network from a random initialization?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_13340"} +{"question": "What paper is about the blended model that improves the stealth of triggers in Backdoor attacks?", "answer": ["Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning"], "answer_arxiv_id": ["1712.05526v1"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_13341"} +{"question": "Can you name some studies focused on demonstrating the feasibility of turbulent super-resolution?", "answer": ["Super-resolution reconstruction of turbulent flows with machine learning"], "answer_arxiv_id": ["1811.11328"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_13342"} +{"question": "Which works initially employed Parameter-Efficient Fine-Tuning (PEFT) in natural language processing?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "The Power of Scale for Parameter-Efficient Prompt Tuning", "LoRA: Low-Rank Adaptation of Large Language Models", "BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models"], "answer_arxiv_id": ["1902.00751", "2104.08691", "2106.09685", "2106.10199"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_13343"} +{"question": "Could you provide me with studies where an on-policy bisimulation metric for policy evaluation was developed?", "answer": ["Scalable methods for computing state similarity in deterministic Markov Decision Processes"], "answer_arxiv_id": ["1911.09291"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_13344"} +{"question": "Could you provide me some studies that use a neural gas network to address problems in few-shot class-incremental learning?", "answer": ["Few-Shot Class-Incremental Learning"], "answer_arxiv_id": ["2004.10956"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13345"} +{"question": "Which papers discuss the core idea or use cases of unsupervised contrastive learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1911.05722"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_13346"} +{"question": "What works have evaluated feature selection methods on synthetic datasets?", "answer": ["How good Neural Networks interpretation methods really are? A quantitative benchmark."], "answer_arxiv_id": ["2304.02383"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_13347"} +{"question": "Which research works used an approximate density model as exploration bonus in the reward signal?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "#Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning"], "answer_arxiv_id": ["1606.01868", "1611.04717"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_13348"} +{"question": "Which papers extended the nonstochastic control algorithm to unknown systems?", "answer": ["The Nonstochastic Control Problem"], "answer_arxiv_id": ["1911.12178"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_13349"} +{"question": "Are there any studies about applying bi-level optimization in the form of meta-learning and the use of differentiable closed-form solvers in time series forecasting?", "answer": ["Meta-Forecasting by combining Global Deep Representations with Local Adaptation"], "answer_arxiv_id": ["2111.03418"], "source_meta": {"published_time": "20220713"}, "qid": "AutoScholarQuery_train_13350"} +{"question": "What are the works that provide the first high probability guarantees in the non-convex setting for algorithms with adaptive step size?", "answer": ["A High Probability Analysis of Adaptive SGD with Momentum", "High Probability Bounds for a Class of Nonconvex Algorithms with AdaGrad Stepsize"], "answer_arxiv_id": ["2007.14294", "2204.02833"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_13351"} +{"question": "What papers suggest going beyond pixel-level importance scores to concept-level scores for the ease of human intervention?", "answer": ["Right for the Right Concept: Revising Neuro-Symbolic Concepts by Interacting with their Explanations"], "answer_arxiv_id": ["2011.12854"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_13352"} +{"question": "Which works underscored the significance of generative priors in Image Restoration through the emergence of GANs?", "answer": ["Generative Adversarial Networks", "Unsupervised Representation Learning with Deep Convolutional Generative\n Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Progressive Growing of GANs for Improved Quality, Stability, and\n Variation"], "answer_arxiv_id": ["1406.2661", "1511.06434", "1812.04948", "1710.10196"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_13353"} +{"question": "Could you provide me some studies which used the diffusion models to restore predefined corruptions on image or audio data?", "answer": ["Universal Speech Enhancement with Score-based Diffusion", "Denoising Diffusion Restoration Models"], "answer_arxiv_id": ["2206.03065", "2201.11793"], "source_meta": {"published_time": "20220912"}, "qid": "AutoScholarQuery_train_13354"} +{"question": "Which works focused on causal tracing to locate factual knowledge?", "answer": ["Locating and Editing Factual Associations in GPT", "Dissecting Recall of Factual Associations in Auto-Regressive Language\n Models"], "answer_arxiv_id": ["2202.05262", "2304.14767"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_13355"} +{"question": "Which works propose the use of visual attributes to enhance zero-shot recognition?", "answer": ["Waffling around for Performance: Visual Classification with Random Words\n and Broad Concepts", "What does a platypus look like? Generating customized prompts for\n zero-shot image classification", "Visual Classification via Description from Large Language Models"], "answer_arxiv_id": ["2306.07282", "2209.03320", "2210.07183"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_13356"} +{"question": "What are the studies that have evaluated explainability on object-detection models like DETR and ViT?", "answer": ["End-to-End Object Detection with Transformers", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2005.12872", "2010.11929"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_13357"} +{"question": "What studies explored the use of probabilistic predictive models to enhance the realism of predictions based on adversarial training?", "answer": ["STRPM: A Spatiotemporal Residual Predictive Model for High-Resolution\n Video Prediction", "MoCoGAN: Decomposing Motion and Content for Video Generation"], "answer_arxiv_id": ["2203.16084", "1707.04993"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_13358"} +{"question": "What are some of the works relating to classifier-free conditioned generation?", "answer": ["Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2207.12598"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_13359"} +{"question": "Which papers propose methods to improve the transferability of attacks?", "answer": ["Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks", "Enhancing the Transferability of Adversarial Attacks through Variance Tuning", "Improving Adversarial Transferability via Neuron Attribution-Based Attacks"], "answer_arxiv_id": ["1904.02884", "2103.15571", "2204.00008"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13360"} +{"question": "Where can I find research works that addressed the extension of 3D Gaussian Splatting (3DGS) scenes to controlled dynamic scenes and multi-camera capture setup?", "answer": ["4D Gaussian Splatting for Real-Time Dynamic Scene Rendering", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis"], "answer_arxiv_id": ["2310.08528", "2308.09713"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_13361"} +{"question": "Which studies involve modifying rotary positional encoding?", "answer": ["RoFormer: Enhanced Transformer with Rotary Position Embedding"], "answer_arxiv_id": ["2104.09864"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_13362"} +{"question": "Which work proposed using patches as input in Transformer-based models?", "answer": ["BEiT: BERT Pre-Training of Image Transformers"], "answer_arxiv_id": ["2106.08254"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_13363"} +{"question": "Can you provide some references about the concept of parameter-efficient fine-tuning?", "answer": ["Efficient Methods for Natural Language Processing: A Survey", "Parameter-Efficient Transfer Learning for NLP", "Simple, Scalable Adaptation for Neural Machine Translation", "Learning multiple visual domains with residual adapters", "AdapterHub: A Framework for Adapting Transformers", "LoRA: Low-Rank Adaptation of Large Language Models", "Training Neural Networks with Fixed Sparse Masks"], "answer_arxiv_id": ["2209.00099", "1902.00751", "1909.08478", "1705.08045", "2007.07779", "2106.09685", "2111.09839"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_13364"} +{"question": "Could you mention some methods that add regularization terms to the value loss in order to implicitly regulate the distribution shift?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Fisher Divergence Critic Regularization", "Critic Regularized Regression"], "answer_arxiv_id": ["2006.04779", "2103.08050", "2006.15134"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_13365"} +{"question": "Could you give me examples of research in which INRs have been applied to spatio-temporal forecasting and reduced-order modeling?", "answer": ["Continuous PDE Dynamics Forecasting with Implicit Neural Representations", "CROM: Continuous Reduced-Order Modeling of PDEs Using Implicit Neural Representations"], "answer_arxiv_id": ["2209.14855", "2206.02607"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_13366"} +{"question": "What papers are about exploring template-based methods specifically tailored for unseen objects?", "answer": ["Multi-path Learning for Object Pose Estimation Across Domains", "Templates for 3D Object Pose Estimation Revisited: Generalization to New Objects and Robustness to Occlusions", "Fusing Local Similarities for Retrieval-based 3D Orientation Estimation\n of Unseen Objects"], "answer_arxiv_id": ["1908.00151", "2203.17234v1", "2203.08472"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_13367"} +{"question": "What works relate to the tuning paradigm in generative LLM-based approaches?", "answer": ["Do LLMs Understand User Preferences? Evaluating LLMs On User Rating\n Prediction", "TALLRec: An Effective and Efficient Tuning Framework to Align Large\n Language Model with Recommendation", "Towards Unified Conversational Recommender Systems via\n Knowledge-Enhanced Prompt Learning", "Recommendation as Language Processing (RLP): A Unified Pretrain,\n Personalized Prompt & Predict Paradigm (P5)", "M6-Rec: Generative Pretrained Language Models are Open-Ended Recommender\n Systems"], "answer_arxiv_id": ["2305.06474", "2305.00447", "2206.09363", "2203.13366", "2205.08084"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_13368"} +{"question": "Could you provide me some studies about how a better numerical solver can accelerate diffusion inference from existing research?", "answer": ["Pseudo Numerical Methods for Diffusion Models on Manifolds", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "Fast Sampling of Diffusion Models with Exponential Integrator", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2202.09778", "2211.01095", "2206.00927", "2204.13902", "2206.00364"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_13369"} +{"question": "Could you provide me papers that mentioned pseudo-labeling in the field of semi-supervised learning?", "answer": ["Unsupervised Data Augmentation for Consistency Training", "Self-training with Noisy Student improves ImageNet classification", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling"], "answer_arxiv_id": ["1904.12848", "1911.04252", "2001.07685v2", "2110.08263"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_13370"} +{"question": "Which papers shed light on adversarial attacks on text that involve adding typos, swapping synonyms and other semantics-preserving transformations?", "answer": ["HotFlip: White-Box Adversarial Examples for Text Classification", "Generating Natural Language Adversarial Examples", "BERT-ATTACK: Adversarial Attack Against BERT Using BERT", "Gradient-based Adversarial Attacks against Text Transformers"], "answer_arxiv_id": ["1712.06751", "1804.07998", "2004.09984", "2104.13733"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_13371"} +{"question": "Could you provide me some studies that focus on question answering tasks on 3D data?", "answer": ["3D Question Answering", "ScanQA: 3D Question Answering for Spatial Scene Understanding"], "answer_arxiv_id": ["2112.08359", "2112.10482"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_13372"} +{"question": "What are some papers about text-guided image editing?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control"], "answer_arxiv_id": ["2208.01626"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_13373"} +{"question": "Which research introduced Invertible Concept-based Explanation (ICE) for concept-based explainability?", "answer": ["Invertible Concept-based Explanations for CNN Models with Non-negative Concept Activation Vectors"], "answer_arxiv_id": ["2006.15417"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_13374"} +{"question": "Can you provide some studies that introduce alternative quantisation schemes in the context of INRs?", "answer": ["On Quantizing Implicit Neural Representations", "Implicit Neural Representations for Image Compression"], "answer_arxiv_id": ["2209.01019", "2112.04267v2"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_13375"} +{"question": "Which work originally discovered the Neural Collapse phenomenon?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning training"], "answer_arxiv_id": ["2008.08186"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_13376"} +{"question": "Which works use the expected hypervolume improvement and max-value entropy search as acquisition function in Multi-Objective Bayesian Optimization (MOBO)?", "answer": ["Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization", "Parallel Bayesian Optimization of Multiple Noisy Objectives with Expected Hypervolume Improvement"], "answer_arxiv_id": ["2006.05078", "2105.08195"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_13377"} +{"question": "What studies belong to the glimpse approach in object-centric learning, which sequentially extract patches from the input?", "answer": ["SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition", "Generative Neurosymbolic Machines"], "answer_arxiv_id": ["2001.02407", "2010.12152"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_13378"} +{"question": "Could you specify studies that designed task-aware methods such as uncertainty-aware policy optimization?", "answer": ["MOPO: Model-based Offline Policy Optimization", "Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models"], "answer_arxiv_id": ["2005.13239", "1805.12114"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_13379"} +{"question": "What methods rely on the created virtual data and force local features to be similar to the features of the same-class virtual data?", "answer": ["Virtual Homogeneity Learning: Defending against Data Heterogeneity in Federated Learning"], "answer_arxiv_id": ["2206.02465"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_13380"} +{"question": "Which studies have shed light on the problem of overthinking by neural networks?", "answer": ["Shallow-Deep Networks: Understanding and Mitigating Network Overthinking"], "answer_arxiv_id": ["1810.07052"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13381"} +{"question": "Could you provide me some studies that addressed the batched RL setting?", "answer": ["Near-Optimal Regret Bounds for Multi-batch Reinforcement Learning"], "answer_arxiv_id": ["2210.08238"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_13382"} +{"question": "Which works have explored the optimization of impressions in long-form answers by maximizing tokens from a particular paragraph in output?", "answer": ["GEO: Generative Engine Optimization"], "answer_arxiv_id": ["2311.09735"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_13383"} +{"question": "What work introduces the notion of Bellman-consistent pessimism, a technique used in offline RL?", "answer": ["Bellman-consistent Pessimism for Offline Reinforcement Learning"], "answer_arxiv_id": ["2106.06926"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13384"} +{"question": "Which work integrates both autoregressive and autoencoding methods in the field of natural language processing?", "answer": ["GLM: General Language Model Pretraining with Autoregressive Blank\n Infilling"], "answer_arxiv_id": ["2103.10360"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_13385"} +{"question": "What research had been done on using neural distance fields in 3D shape reconstruction and representation?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Neural Unsigned Distance Fields for Implicit Function Learning", "Neural-Pull: Learning Signed Distance Functions from Point Clouds by\n Learning to Pull Space onto Surfaces", "INeRF: Inverting Neural Radiance Fields for Pose Estimation", "StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation"], "answer_arxiv_id": ["1901.05103", "2010.13938", "2011.13495", "2012.05877", "2112.11427"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_13386"} +{"question": "What are the studies regarding model-agnostic feature selection methods such as Anchors and LIME?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier"], "answer_arxiv_id": ["1602.04938"], "source_meta": {"published_time": "20221202"}, "qid": "AutoScholarQuery_train_13387"} +{"question": "Could you provide me some studies about utilizing disentangled style and content representation in diffusion models?", "answer": ["Diffusion-based Image Translation using Disentangled Style and Content Representation"], "answer_arxiv_id": ["2209.15264"], "source_meta": {"published_time": "20220316"}, "qid": "AutoScholarQuery_train_13388"} +{"question": "Which studies demonstrate the exceptional zero-shot and out-of-distribution capabilities of foundation models like CLIP and ALIGN?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "LiT: Zero-Shot Transfer with Locked-image text Tuning"], "answer_arxiv_id": ["2103.00020", "2205.01917", "2111.07991"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_13389"} +{"question": "Which works discuss the concept of knowledge distillation and the original methods in it?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_13390"} +{"question": "Which studies are related to utilizing a displacement field to represent the motion in dynamic scene representation?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View\n Synthesis of a Dynamic Scene From Monocular Video"], "answer_arxiv_id": ["2011.13961", "2012.12247"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_13391"} +{"question": "Could you provide me some research papers about the convergence of TD?", "answer": ["Finite Sample Analyses for TD(0) with Function Approximation", "A Finite Time Analysis of Temporal Difference Learning With Linear Function Approximation", "Finite-Time Error Bounds For Linear Stochastic Approximation and TD Learning"], "answer_arxiv_id": ["1704.01161", "1806.02450", "1902.00923"], "source_meta": {"published_time": "20220214"}, "qid": "AutoScholarQuery_train_13392"} +{"question": "In which papers is learning visual correspondence applied in grasping?", "answer": ["NeRF-Supervision: Learning Dense Object Descriptors from Neural Radiance\n Fields", "DGCM-Net: Dense Geometrical Correspondence Matching Network for\n Incremental Experience-based Robotic Grasping", "USEEK: Unsupervised SE(3)-Equivariant 3D Keypoints for Generalizable\n Manipulation"], "answer_arxiv_id": ["2203.01913", "2001.05279", "2209.13864"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_13393"} +{"question": "Which papers talk about using model-based methods that construct a world model in RL?", "answer": ["Masked World Models for Visual Control", "Dream to Control: Learning Behaviors by Latent Imagination", "DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations", "Mastering Atari with Discrete World Models", "Mastering Diverse Domains through World Models", "DayDreamer: World Models for Physical Robot Learning"], "answer_arxiv_id": ["2206.14244", "1912.01603", "2110.14565", "2010.02193", "2301.04104", "2206.14176"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13394"} +{"question": "Could you provide me some studies focus on breaking the symmetries between slots by learning specialized slots for different object types?", "answer": ["Towards causal generative scene models via competition of experts", "Self-Supervised Visual Representation Learning with Semantic Grouping"], "answer_arxiv_id": ["2004.12906", "2205.15288"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_13395"} +{"question": "Which research papers proposed transformer-based methods that use mask-attention?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2107.06278", "2112.01527"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_13396"} +{"question": "Which papers use semidefinite programming for network verification?", "answer": ["Semidefinite relaxations for certifying robustness to adversarial examples", "Enabling certification of verification-agnostic networks via memory-efficient semidefinite programming"], "answer_arxiv_id": ["1811.01057", "2010.11645"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_13397"} +{"question": "Could you provide me some papers where diffusion-based vocoders were introduced?", "answer": ["DiffWave: A Versatile Diffusion Model for Audio Synthesis"], "answer_arxiv_id": ["2009.09761v3"], "source_meta": {"published_time": "20230802"}, "qid": "AutoScholarQuery_train_13398"} +{"question": "Which paper is about simultaneous queries of multiple arms in the context of graph-based bandit problems?", "answer": ["Leveraging Side Observations in Stochastic Bandits"], "answer_arxiv_id": ["1210.4839"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_13399"} +{"question": "Can you name the studies that have studied the problem of over-smoothing in Message Passing Neural Networks (MPNNs) from an asymptotic perspective?", "answer": ["Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning", "Graph Neural Networks Exponentially Lose Expressive Power for Node Classification"], "answer_arxiv_id": ["1801.07606", "1905.10947"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_13400"} +{"question": "Which works tokenize various input modalities into sequences and employ a single Transformer for joint learning in multi-modal representation learning?", "answer": ["OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework", "VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts"], "answer_arxiv_id": ["2202.03052", "2111.02358"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_13401"} +{"question": "Could you provide me some references about visual reasoning tasks?", "answer": ["VLUE: A Multi-Task Benchmark for Evaluating Vision-Language Models", "VQA: Visual Question Answering", "Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering", "Being Negative but Constructively: Lessons Learnt from Creating Better Visual Question Answering Datasets", "Vision Meets Definitions: Unsupervised Visual Word Sense Disambiguation Incorporating Gloss Information", "CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning", "COVR: A Test-Bed for Visually Grounded Compositional Generalization with Real Images", "Understanding Image and Text Simultaneously: a Dual Vision-Language Machine Comprehension Task", "Words aren’t enough, their order matters: On the Robustness of Grounding Visual Referring Expressions"], "answer_arxiv_id": ["2205.15237", "1505.00468", "1612.00837", "1704.07121", "2305.01788", "1612.06890", "2109.10613", "1612.07833", "2005.01655"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_13402"} +{"question": "What studies have talked about the specialized use of these agents in solving data science problems?", "answer": ["ChatDev: Communicative Agents for Software Development", "MLCopilot: Unleashing the Power of Large Language Models in Solving\n Machine Learning Tasks"], "answer_arxiv_id": ["2307.07924", "2304.14979"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_13403"} +{"question": "What studies have investigated the relationship between sharpness and model generalization from the lens of loss surface geometry?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp\n Minima", "Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural\n Networks with Many More Parameters than Training Data", "Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Averaging Weights Leads to Wider Optima and Better Generalization", "Sharpness-Aware Minimization for Efficiently Improving Generalization", "SWAD: Domain Generalization by Seeking Flat Minima"], "answer_arxiv_id": ["1609.04836", "1703.11008", "1802.10026", "1803.05407", "2010.01412", "2102.08604"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_13404"} +{"question": "Which works introduced the original Precision and Recall for improving the interpretation of fidelity and diversity?", "answer": ["Assessing Generative Models via Precision and Recall"], "answer_arxiv_id": ["1806.00035"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_13405"} +{"question": "Can you tell me the papers where PLRNNs were used for DS analysis?", "answer": ["Identifying nonlinear dynamical systems with multiple time scales and long-range dependencies", "Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI", "Existence of n-cycles and border-collision bifurcations in piecewise-linear continuous maps with applications to recurrent neural networks"], "answer_arxiv_id": ["1910.03471", "1902.07186v2", "1911.04304v2"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_13406"} +{"question": "What works are about noise scale learning that necessitate re-training?", "answer": ["Variational Diffusion Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2107.00630", "2102.09672"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_13407"} +{"question": "In what works is text prompt incorporated as condition to control the output image during the generative process using Diffusion models?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10741", "2102.12092", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_13408"} +{"question": "Which papers improved diffusion models with adversarial learning?", "answer": ["Adversarial score matching and improved sampling for image generation", "Refining Generative Process with Discriminator Guidance in Score-based Diffusion Models"], "answer_arxiv_id": ["2009.05475", "2211.17091v4"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_13409"} +{"question": "What was the work that introduced the notion of average sensitivity for the first time?", "answer": ["Average Sensitivity of Graph Algorithms"], "answer_arxiv_id": ["1904.03248"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_13410"} +{"question": "Which research papers are related to adversarial attacks in NLP particularly focused on LLMs?", "answer": ["Explaining and Harnessing Adversarial Examples", "Open Sesame! Universal Black Box Jailbreaking of Large Language Models", "White-Box Multi-Objective Adversarial Attack on Dialogue Generation", "Survey of Vulnerabilities in Large Language Models Revealed by\n Adversarial Attacks"], "answer_arxiv_id": ["1412.6572", "2309.01446", "2305.03655", "2310.10844"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_13411"} +{"question": "Which research papers provided finite-sample statistical analyses for linear dynamical systems in system identification?", "answer": ["Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification", "On the Sample Complexity of the Linear Quadratic Regulator", "Non-asymptotic Identification of LTI Systems from a Single Trajectory", "Finite Sample Analysis of Stochastic System Identification", "Statistical Learning Theory for Control: A Finite Sample Perspective"], "answer_arxiv_id": ["1802.08334", "1710.01688", "1806.05722", "1903.09122", "2209.05423v2"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_13412"} +{"question": "What work introduced a method for image retrieval through dialogue between a user and an automated system?", "answer": ["Chatting Makes Perfect: Chat-based Image Retrieval"], "answer_arxiv_id": ["2305.20062"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_13413"} +{"question": "Which works focused on training the retrievers and Language Models (LMs) jointly in an end-to-end setting for document selection in RALMs?", "answer": ["REALM: Retrieval-Augmented Language Model Pre-Training", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Atlas: Few-shot Learning with Retrieval Augmented Language Models"], "answer_arxiv_id": ["2002.08909", "2005.11401", "2208.03299"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_13414"} +{"question": "Can you list some research that discussed the dynamic regret for convex Lipschitz losses and strongly convex losses?", "answer": ["Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient", "Online Forecasting of Total-Variation-bounded Sequences"], "answer_arxiv_id": ["1605.04638", "1906.03364"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_13415"} +{"question": "Which researches proposed using positional embeddings to overcome the limitations of MPNNs?", "answer": ["Benchmarking Graph Neural Networks", "A Generalization of Transformer Networks to Graphs"], "answer_arxiv_id": ["2003.00982", "2012.09699"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_13416"} +{"question": "Which works proposed end-to-end solutions for reconstructing room layout, object bounding boxes, and meshes from a single image?", "answer": ["Total3DUnderstanding: Joint Layout, Object Pose and Mesh Reconstruction for Indoor Scenes from a Single Image", "Holistic 3D Scene Understanding from a Single Image with Implicit Representation"], "answer_arxiv_id": ["2002.12212", "2103.06422"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13417"} +{"question": "Which studies signal the advent of contrastive learning in deep clustering?", "answer": ["Contrastive Clustering", "Twin Contrastive Learning for Online Clustering", "You Never Cluster Alone"], "answer_arxiv_id": ["2009.09687", "2210.11680", "2106.01908"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_13418"} +{"question": "What works have made significant strides in the large vision and language models?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "Learning Transferable Visual Models From Natural Language Supervision", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2005.14165", "2204.02311", "2103.00020", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_13419"} +{"question": "Are there any studies that show the universal sequence to sequence approximation capability of transformer models?", "answer": ["Are Transformers universal approximators of sequence-to-sequence functions?"], "answer_arxiv_id": ["1912.10077"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13420"} +{"question": "Could you provide me the work that found the limitations of MPNNs in imitating complex graph algorithms?", "answer": ["What Can Neural Networks Reason About?"], "answer_arxiv_id": ["1905.13211"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_13421"} +{"question": "What works utilize hash grids and point clouds to reduce computational costs of large neural networks?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Point-NeRF: Point-based Neural Radiance Fields"], "answer_arxiv_id": ["2201.05989", "2201.08845"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13422"} +{"question": "What work introduced an interpretable and easy plug-in spatial-temporal attention mechanism for video action recognition?", "answer": ["Interpretable Spatio-temporal Attention for Video Action Recognition"], "answer_arxiv_id": ["1810.04511"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_13423"} +{"question": "Which papers are about multi-trial NAS algorithms on GNNs?", "answer": ["Auto-GNN: Neural Architecture Search of Graph Neural Networks"], "answer_arxiv_id": ["1909.03184"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_13424"} +{"question": "Which works focus on few-view reconstruction required to output 3D geometries from feed-forward methods?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision"], "answer_arxiv_id": ["2308.16512", "2306.07881", "2306.11719"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_13425"} +{"question": "Can you provide studies employing hybrid methods, using implicit fields with explicit shape proxies, for human rendering?", "answer": ["STAR: Sparse Trained Articulated Human Body Regressor", "Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "Total Capture: A 3D Deformation Model for Tracking Faces, Hands, and\n Bodies"], "answer_arxiv_id": ["2008.08535", "1904.05866", "1801.01615"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_13426"} +{"question": "Which work indicated that error-minimizing perturbations cause overfitting in the context of unlearnable datasets?", "answer": ["Unlearnable Examples: Making Personal Data Unexploitable"], "answer_arxiv_id": ["2101.04898"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13427"} +{"question": "What works propose plausible CEs by considering proximity to the data manifold?", "answer": ["FACE: Feasible and Actionable Counterfactual Explanations"], "answer_arxiv_id": ["1909.09369"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_13428"} +{"question": "Which papers discuss instruction following models that use Reinforcement Learning from Human Feedback (RLHF)?", "answer": ["Deep Reinforcement Learning from Human Preferences"], "answer_arxiv_id": ["1706.03741"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13429"} +{"question": "What studies have used normal maps as auxiliary information for neural scene reconstruction?", "answer": ["MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface\n Reconstruction", "NeuRIS: Neural Reconstruction of Indoor Scenes Using Normal Priors"], "answer_arxiv_id": ["2206.00665", "2206.13597"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13430"} +{"question": "Which papers are about reducing the computational burden in training by collecting ensemble members efficiently?", "answer": ["Snapshot Ensembles: Train 1, get M for free", "Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Loss Surface Simplexes for Mode Connecting Volumes and Fast Ensembling"], "answer_arxiv_id": ["1704.00109", "1802.10026", "2102.13042"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_13431"} +{"question": "What are the examples of works where the integrity of the meta-test is maintained during meta-training?", "answer": ["Meta-Learning Representations for Continual Learning"], "answer_arxiv_id": ["1905.12588"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_13432"} +{"question": "Could you provide me some works about four-IMU setting for inertial motion capture systems?", "answer": ["LiDAR-aid Inertial Poser: Large-scale Human Motion Capture by Sparse\n Inertial and LiDAR Sensors"], "answer_arxiv_id": ["2205.15410"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_13433"} +{"question": "What papers contain theories or applications of on-policy RL setting?", "answer": ["Proximal Policy Optimization Algorithms", "Trust Region Policy Optimization"], "answer_arxiv_id": ["1707.06347v2", "1502.05477"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_13434"} +{"question": "Can you identify the works that focus on the RL with general utilities in multi-agent systems?", "answer": ["Variational Policy Gradient Method for Reinforcement Learning with General Utilities", "On the Convergence and Sample Efficiency of Variance-Reduced Policy Gradient Method", "MARL with General Utilities via Decentralized Shadow Reward Actor-Critic", "Reward is Enough for Convex MDPs", "Concave Utility Reinforcement Learning: the Mean-Field Game Viewpoint", "Policy-based Primal-Dual Methods for Convex Constrained Markov Decision Processes", "Scalable Multi-Agent Reinforcement Learning with General Utilities"], "answer_arxiv_id": ["2007.02151", "2102.08607", "2106.00543", "2106.00661", "2106.03787", "2205.10715", "2302.07938"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_13435"} +{"question": "Which study introduces 3D Gaussian splatting that improves rendering speed and fidelity?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_13436"} +{"question": "Who proposed the generalization of ORC for undirected hypergraphs?", "answer": ["Curvature of Hypergraphs via Multi-Marginal Optimal Transport", "On the spectrum of hypergraphs"], "answer_arxiv_id": ["1803.08584", "1711.09356"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_13437"} +{"question": "Which studies involve the use of category-specific template shapes while optimizing the 3D shape from an image or video collection?", "answer": ["Birds of a Feather: Capturing Avian Shape Models from Images"], "answer_arxiv_id": ["2105.09396"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_13438"} +{"question": "Could you point me to studies that discuss efficient algorithms for learning one-hidden-layer ReLU networks with Gaussian inputs?", "answer": ["Algorithms and SQ Lower Bounds for PAC Learning One-Hidden-Layer ReLU Networks", "Small Covers for Near-Zero Sets of Polynomials and Learning Latent Variable Models"], "answer_arxiv_id": ["2006.12476", "2012.07774v1"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13439"} +{"question": "What works reported that different CNNs tend to learn qualitatively similar filters in the first layer?", "answer": ["Gabor Convolutional Networks"], "answer_arxiv_id": ["1705.01450v4"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_13440"} +{"question": "Could you provide me some works about learning representations by artificially masking parts of the input and training a network to reconstruct the hidden content?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers", "data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language"], "answer_arxiv_id": ["2111.06377", "2106.08254", "2202.03555"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_13441"} +{"question": "What are the papers that have contributed to the understanding of the 'catastrophic forgetting' problem using the concept of 'personalized federated learning'?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Federated Learning with Partial Model Personalization"], "answer_arxiv_id": ["1612.00796", "2204.03809"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_13442"} +{"question": "Which papers focused on the application of Graph Network Simulators in particle-based simulations?", "answer": ["Learning to Simulate Complex Physics with Graph Networks"], "answer_arxiv_id": ["2002.09405"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_13443"} +{"question": "What papers demonstrate that attacks in the frequency domain are more effective than those in the spatial domain?", "answer": ["Low Frequency Adversarial Perturbation", "On the Effectiveness of Low Frequency Perturbations"], "answer_arxiv_id": ["1809.08758", "1903.00073"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_13444"} +{"question": "Can you cite studies about common supervised multi-modal tasks?", "answer": ["Balanced Multimodal Learning via On-the-fly Gradient Modulation", "Audiovisual SlowFast Networks for Video Recognition", "AdaMML: Adaptive Multi-Modal Learning for Efficient Video Recognition", "What Makes Training Multi-modal Classification Networks Hard?", "Don’t Just Assume; Look and Answer: Overcoming Priors for Visual Question Answering", "ACNet: Attention Based Network to Exploit Complementary Features for RGBD Semantic Segmentation", "Efficient RGB-D Semantic Segmentation for Indoor Scene Analysis"], "answer_arxiv_id": ["2203.15332", "2001.08740", "2105.05165", "1905.12681", "1712.00377", "1905.10089", "2011.06961"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_13445"} +{"question": "What research proposed a heuristic that relies on self-supervised features to localize the most salient object in the image?", "answer": ["Localizing Objects with Self-Supervised Transformers and no Labels"], "answer_arxiv_id": ["2109.14279"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_13446"} +{"question": "Can you list the bottom-up grouping-based methods utilized in earlier approaches for 3D point cloud instance segmentation?", "answer": ["SoftGroup for 3D Instance Segmentation on Point Clouds", "DyCo3D: Robust Instance Segmentation of 3D Point Clouds through Dynamic\n Convolution", "Hierarchical Aggregation for 3D Instance Segmentation", "Instance Segmentation in 3D Scenes using Semantic Superpoint Tree\n Networks", "PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation"], "answer_arxiv_id": ["2203.01509", "2011.13328", "2108.02350", "2108.07478", "2004.01658"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_13447"} +{"question": "Which paper developed 'MotionCLIP', a transformer-based auto-encoder with text-to-motion capabilities?", "answer": ["MotionCLIP: Exposing Human Motion Generation to CLIP Space"], "answer_arxiv_id": ["2203.08063"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_13448"} +{"question": "Are there studies on diffusion models for image synthesis tasks?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Variational Diffusion Models", "Diffusion Models Beat GANs on Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models", "Scalable Diffusion Models with Transformers"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2107.00630", "2105.05233", "2112.10752", "2212.09748"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_13449"} +{"question": "Could you provide me some works that adapt convolution kernels to the underlying local geometries in point cloud segmentation?", "answer": ["A-CNN: Annularly Convolutional Neural Networks on Point Clouds", "KPConv: Flexible and Deformable Convolution for Point Clouds", "Tangent Convolutions for Dense Prediction in 3D", "ShellNet: Efficient Point Cloud Convolutional Neural Networks using\n Concentric Shells Statistics", "Point-Voxel CNN for Efficient 3D Deep Learning", "SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional\n Filters", "Point Convolutional Neural Networks by Extension Operators", "Relation-Shape Convolutional Neural Network for Point Cloud Analysis", "Pointwise Convolutional Neural Networks"], "answer_arxiv_id": ["1904.08017", "1904.08889", "1807.02443", "1908.06295", "1907.03739", "1803.11527", "1803.10091", "1904.07601", "1712.05245"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_13450"} +{"question": "Which work introduced an encoder-only DETR without using a decoder?", "answer": ["Rethinking Transformer-based Set Prediction for Object Detection"], "answer_arxiv_id": ["2011.10881"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_13451"} +{"question": "Which paper demonstrated the effectiveness of using adversarial transformations in image classification?", "answer": ["Adversarial AutoAugment"], "answer_arxiv_id": ["1912.11188"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_13452"} +{"question": "What works have proposed methods for leveraging prior human knowledge to improve the robustness of neural networks?", "answer": ["Concept Bottleneck Models", "Right for the Right Reasons: Training Differentiable Models by Constraining their Explanations"], "answer_arxiv_id": ["2007.04612", "1703.03717"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_13453"} +{"question": "Could you provide a reference that proved convergence guarantees for deterministic SAM without requiring decaying perturbation size?", "answer": ["Towards Understanding Sharpness-Aware Minimization"], "answer_arxiv_id": ["2206.06232"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_13454"} +{"question": "Which works proposed policy regularization methods for handling the distribution shift challenge in offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "AWAC: Accelerating Online Reinforcement Learning with Offline Datasets", "Believe What You See: Implicit Constraint Approach for Offline Multi-Agent Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "2006.09359", "2106.03400", "2106.06860"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13455"} +{"question": "Could you provide me some works that employed butterfly-based methods in approximating matrix multiplication?", "answer": ["Pixelated Butterfly: Simple and Efficient Sparse Training for Neural Network Models", "Monarch: Expressive Structured Matrices for Efficient and Accurate Training"], "answer_arxiv_id": ["2112.00029", "2204.00595"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_13456"} +{"question": "What research papers employed L1 or L2 norms as objectives in PSNR-oriented methods for single image super-resolution?", "answer": ["Accelerating the Super-Resolution Convolutional Neural Network", "Real-Time Single Image and Video Super-Resolution Using an Efficient\n Sub-Pixel Convolutional Neural Network", "Image Super-Resolution Using Deep Convolutional Networks", "Residual Dense Network for Image Super-Resolution", "Enhanced Deep Residual Networks for Single Image Super-Resolution"], "answer_arxiv_id": ["1608.00367", "1609.05158", "1501.00092", "1802.08797", "1707.02921"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_13457"} +{"question": "Any works showing that calibration can be improved by directly applying methods, such as ensembling and pre-training?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Using Pre-Training Can Improve Model Robustness and Uncertainty"], "answer_arxiv_id": ["1612.01474", "1901.09960"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_13458"} +{"question": "What are some recent FR models that performed well on datasets with discernable facial attributes?", "answer": ["CosFace: Large Margin Cosine Loss for Deep Face Recognition", "ArcFace: Additive Angular Margin Loss for Deep Face Recognition", "SphereFace: Deep Hypersphere Embedding for Face Recognition", "CurricularFace: Adaptive Curriculum Learning Loss for Deep Face\n Recognition", "MagFace: A Universal Representation for Face Recognition and Quality\n Assessment"], "answer_arxiv_id": ["1801.09414", "1801.07698", "1704.08063", "2004.00288", "2103.06627"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_13459"} +{"question": "What studies have been conducted on two-human interaction synthesis?", "answer": ["Human Motion Diffusion as a Generative Prior", "InterGen: Diffusion-based Multi-human Motion Generation under Complex\n Interactions", "Generative Proxemics: A Prior for 3D Social Interaction from Images"], "answer_arxiv_id": ["2303.01418", "2304.05684", "2306.09337"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_13460"} +{"question": "Could you provide examples of research that worked on creating diverse input patterns with data augmentation-based methods for image attacks?", "answer": ["Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks", "Improving Transferability of Adversarial Examples with Input Diversity"], "answer_arxiv_id": ["1904.02884", "1803.06978"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_13461"} +{"question": "Which research improves DSVAE by replacing Euclidean distance with Wasserstein distance?", "answer": ["Disentangled Recurrent Wasserstein Autoencoder"], "answer_arxiv_id": ["2101.07496v1"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13462"} +{"question": "Which works propose the Outlier channel splitting method?", "answer": ["Improving Neural Network Quantization without Retraining using Outlier Channel Splitting"], "answer_arxiv_id": ["1901.09504"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_13463"} +{"question": "What models have been proposed to train visual representation using ViT from scratch?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2108.10904", "2205.01917", "2204.14198", "2301.12597"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_13464"} +{"question": "Which works made use of the Bayesian interpretation of effective parameters based on the Hessian of the training loss for studying neural networks ?", "answer": ["Rethinking Parameter Counting in Deep Models: Effective Dimensionality Revisited"], "answer_arxiv_id": ["2003.02139"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13465"} +{"question": "Can you provide me with some studies on monocular 3D pose estimation?", "answer": ["VIBE: Video Inference for Human Body Pose and Shape Estimation", "PLIKS: A Pseudo-Linear Inverse Kinematic Solver for 3D Human Body\n Estimation", "GLA-GCN: Global-local Adaptive Graph Convolutional Network for 3D Human\n Pose Estimation from Monocular Video", "Diffusion-Based 3D Human Pose Estimation with Multi-Hypothesis\n Aggregation"], "answer_arxiv_id": ["1912.05656", "2211.11734", "2307.05853", "2303.11579"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_13466"} +{"question": "Which works have contributed towards building long-form video understanding datasets?", "answer": ["Generic Event Boundary Detection: A Benchmark for Event Segmentation", "MAD: A Scalable Dataset for Language Grounding in Videos from Movie\n Audio Descriptions", "MovieQA: Understanding Stories in Movies through Question-Answering"], "answer_arxiv_id": ["2101.10511", "2112.00431", "1512.02902"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_13467"} +{"question": "Which paper suggests that the microbatch clipping method could be used instead?", "answer": ["A General Approach to Adding Differential Privacy to Iterative Training Procedures"], "answer_arxiv_id": ["1812.06210"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_13468"} +{"question": "Which studies involved in Speech to Audio-Visual Speech Translation (A2AV)?", "answer": ["TRAVID: An End-to-End Video Translation Framework", "Technology Pipeline for Large Scale Cross-Lingual Dubbing of Lecture\n Videos into Multiple Indian Languages", "Face-Dubbing++: Lip-Synchronous, Voice Preserving Translation of Videos", "Towards Automatic Face-to-Face Translation"], "answer_arxiv_id": ["2309.11338", "2211.01338", "2206.04523", "2003.00418"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_13469"} +{"question": "What works have used language models and zero-shot prompting to understand user intention?", "answer": ["Human-Centric Research for NLP: Towards a Definition and Guiding\n Questions", "Zero-Shot Prompting for Implicit Intent Prediction and Recommendation\n with Commonsense Reasoning"], "answer_arxiv_id": ["2207.04447", "2210.05901"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_13470"} +{"question": "Are there any works exploring class prior estimation algorithms specifically for PUL?", "answer": ["Class-prior Estimation for Learning from Positive and Unlabeled Data", "Mixture Proportion Estimation and PU Learning: A Modern Approach"], "answer_arxiv_id": ["1611.01586", "2111.00980"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_13471"} +{"question": "What are the papers that purposed the VQA dataset to evaluate the hallucinations of MLLMs?", "answer": ["Holistic Analysis of Hallucination in GPT-4V(ision): Bias and\n Interference Challenges", "Evaluating Object Hallucination in Large Vision-Language Models"], "answer_arxiv_id": ["2311.03287", "2305.10355"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_13472"} +{"question": "Which researches deal with poisoned-label backdoor attacks?", "answer": ["Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning", "Untargeted Backdoor Watermark: Towards Harmless and Stealthy Dataset Copyright Protection"], "answer_arxiv_id": ["1712.05526v1", "2210.00875"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_13473"} +{"question": "Which papers proposed methods to improve the sample efficiency in imitation learning?", "answer": ["Discriminator-Actor-Critic: Addressing Sample Inefficiency and Reward Bias in Adversarial Imitation Learning"], "answer_arxiv_id": ["1809.02925"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_13474"} +{"question": "Which paper introduces the FILM pipeline for language-based colorization?", "answer": ["Learning to Color from Language"], "answer_arxiv_id": ["1804.06026v1"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_13475"} +{"question": "Which studies integrated a pretrained image encoder with autoregressive language model?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "PaLI-X: On Scaling up a Multilingual Vision and Language Model", "PaLI-3 Vision Language Models: Smaller, Faster, Stronger", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2106.13884", "2209.06794", "2305.18565", "2310.09199", "2301.12597", "2303.03378"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_13476"} +{"question": "Could you provide some works about masked image inpainting?", "answer": ["Images Speak in Images: A Generalist Painter for In-Context Visual\n Learning", "SegGPT: Segmenting Everything In Context"], "answer_arxiv_id": ["2212.02499", "2304.03284"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_13477"} +{"question": "What papers employed the encoder-decoder structure to refine the low-resolution coarse predictions in scene parsing?", "answer": ["SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image\n Segmentation", "RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic\n Segmentation", "Encoder-Decoder with Atrous Separable Convolution for Semantic Image\n Segmentation", "SPGNet: Semantic Prediction Guidance for Scene Parsing"], "answer_arxiv_id": ["1511.00561", "1611.06612", "1802.02611", "1908.09798"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_13478"} +{"question": "What research explored using GPT-3 to generate rich textual prompts for each class to improve zero-shot accuracy?", "answer": ["What does a platypus look like? Generating customized prompts for zero-shot image classification"], "answer_arxiv_id": ["2209.03320"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13479"} +{"question": "What studies offer solutions in designing acquisition functions for 𝒙t(2) adopts standard BO?", "answer": ["A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning", "PREFERENTIAL BATCH BAYESIAN OPTIMIZATION"], "answer_arxiv_id": ["1012.2599", "2003.11435"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_13480"} +{"question": "What works have developed specific architectures to capture the objects related to semantic labels in MLL?", "answer": ["Combining Metric Learning and Attention Heads For Accurate and Efficient Multilabel Image Classification", "General Multi-label Image Classification with Transformers", "Query2Label: A Simple Transformer Way to Multi-Label Classification"], "answer_arxiv_id": ["2209.06585", "2011.14027", "2107.10834"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_13481"} +{"question": "Which studies performed pixel-level labeling of an image in semantic segmentation?", "answer": ["Image Segmentation Using Deep Learning: A Survey"], "answer_arxiv_id": ["2001.05566"], "source_meta": {"published_time": "20240127"}, "qid": "AutoScholarQuery_train_13482"} +{"question": "What works focused on enabling highly realistic rendering and novel view synthesis by implicitly encoding the scene’s appearance and geometry?", "answer": ["A-NeRF: Articulated Neural Radiance Fields for Learning Human Shape, Appearance, and Pose", "HDR-NeRF: High Dynamic Range Neural Radiance Fields", "NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw Images"], "answer_arxiv_id": ["2102.06199v3", "2111.14451", "2111.13679"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_13483"} +{"question": "What papers reflect the use of different types of neural networks for generative models?", "answer": ["LayoutGAN: Generating Graphic Layouts with Wireframe Discriminators", "Attribute-conditioned Layout GAN for Automatic Graphic Design", "Constrained Graphic Layout Generation via Latent Optimization", "LayoutVAE: Stochastic Scene Layout Generation From a Label Set", "Variational Transformer Networks for Layout Generation", "LayoutTransformer: Layout Generation and Completion with Self-attention", "LayoutFormer++: Conditional Graphic Layout Generation via Constraint\n Serialization and Decoding Space Restriction", "BLT: Bidirectional Layout Transformer for Controllable Layout Generation", "LayoutDM: Discrete Diffusion Model for Controllable Layout Generation", "LayoutDM: Transformer-based Diffusion Model for Layout Generation", "LayoutDiffusion: Improving Graphic Layout Generation by Discrete\n Diffusion Probabilistic Models", "DLT: Conditioned layout generation with Joint Discrete-Continuous\n Diffusion Layout Transformer"], "answer_arxiv_id": ["1901.06767", "2009.05284", "2108.00871", "1907.10719", "2104.02416", "2006.14615", "2208.08037", "2112.05112", "2303.08137", "2305.02567", "2303.11589", "2303.03755"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_13484"} +{"question": "What are the works that have made progress in joint hand pose and object reconstruction?", "answer": ["Learning joint reconstruction of hands and manipulated objects", "Leveraging Photometric Consistency over Time for Sparsely Supervised\n Hand-Object Reconstruction", "Collaborative Learning for Hand and Object Reconstruction with\n Attention-guided Graph Convolution", "ArtiBoost: Boosting Articulated 3D Hand-Object Pose Estimation via\n Online Exploration and Synthesis"], "answer_arxiv_id": ["1904.05767", "2004.13449", "2204.13062", "2109.05488"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13485"} +{"question": "Which studies have used Large Language Models (LLMs) prompting in the development of web agents?", "answer": ["ReAct: Synergizing Reasoning and Acting in Language Models", "Reflexion: Language Agents with Verbal Reinforcement Learning", "LASER: LLM Agent with State-Space Exploration for Web Navigation"], "answer_arxiv_id": ["2210.03629", "2303.11366", "2309.08172"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_13486"} +{"question": "What studies deal with algorithms that prune at initialization or follow a computationally expensive cycle of pruning and retraining for multiple iterations?", "answer": ["SNIP: Single-shot Network Pruning based on Connection Sensitivity", "Picking Winning Tickets Before Training by Preserving Gradient Flow", "Pruning neural networks without any data by iteratively conserving synaptic flow", "Progressive Skeletonization: Trimming more fat from a network at initialization", "The State of Sparsity in Deep Neural Networks", "Winning the Lottery with Continuous Sparsification", "Comparing Rewinding and Fine-tuning in Neural Network Pruning"], "answer_arxiv_id": ["1810.02340", "2002.07376", "2006.05467", "2006.09081", "1902.09574", "1912.04427", "2003.02389"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_13487"} +{"question": "Which works use LMs to generate test cases instead of directly judging the correctness of output programs?", "answer": ["Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2203.07814"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_13488"} +{"question": "Could you provide some examples of studies that investigated the efficacy of adapting large-scale pre-trained models in the vision field by modifying input images at the pixel level?", "answer": ["Visual Prompt Tuning", "Exploring Visual Prompts for Adapting Large-Scale Models"], "answer_arxiv_id": ["2203.12119", "2203.17274"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13489"} +{"question": "Can you provide me some examples of text-guided works that apply contrastive losses for NeRF Stylization?", "answer": ["NeRF-Art: Text-Driven Neural Radiance Fields Stylization", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields"], "answer_arxiv_id": ["2212.08070", "2112.05139"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_13490"} +{"question": "Which papers proposed design of single-step denoising distributions as conditional energy-based models?", "answer": ["Learning Energy-Based Models by Diffusion Recovery Likelihood"], "answer_arxiv_id": ["2012.08125"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_13491"} +{"question": "Can you mention the works that focus on adversarial linear MDPs methods?", "answer": ["Policy Optimization in Adversarial MDPs: Improved Exploration via Dilated Bonuses", "Policy Optimization in Adversarial MDPs: Improved Exploration via Dilated Bonuses", "Refined Regret for Adversarial MDPs with Linear Function Approximation", "Improved Regret for Efficient Online Reinforcement Learning with Linear Function Approximation", "A Theoretical Analysis of Optimistic Proximal Policy Optimization in Linear Markov Decision Processes"], "answer_arxiv_id": ["2107.08346", "2107.08346", "2301.12942", "2301.13087", "2305.08841"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_13492"} +{"question": "Can you mention the work analyzing the empirical Rademacher complexity of a specific, simple GNN architecture?", "answer": ["Generalization and Representational Limits of Graph Neural Networks"], "answer_arxiv_id": ["2002.06157"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_13493"} +{"question": "What works introduced VAE-based methods for paired data in text-to-shape generation?", "answer": ["ShapeCrafter: A Recursive Text-Conditioned 3D Shape Generation Model", "AutoSDF: Shape Priors for 3D Completion, Reconstruction and Generation"], "answer_arxiv_id": ["2207.09446", "2203.09516"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_13494"} +{"question": "Could you provide the studies demonstrating various successes of autoregressive and masked models across many tasks and domains?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "Robust Speech Recognition via Large-Scale Weak Supervision", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "MaskGIT: Masked Generative Image Transformer"], "answer_arxiv_id": ["2111.06377", "2212.04356", "2206.10789", "2202.04200"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_13495"} +{"question": "Could you list works where soft Q-functions are used in offline RL to make the learned policy and behavior policy sufficiently similar?", "answer": ["Behavior Regularized Offline Reinforcement Learning", "Continuous Doubly Constrained Batch Reinforcement Learning"], "answer_arxiv_id": ["1911.11361", "2102.09225"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_13496"} +{"question": "What studies have demonstrated that Large Language Models can complete tasks using in-context learning?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_13497"} +{"question": "Which papers aim to address the issue of visual generalization in reinforcement learning?", "answer": ["Generalization in Reinforcement Learning by Soft Data Augmentation", "Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation", "Reinforcement Learning with Augmented Data", "Mastering Visual Continuous Control: Improved Data-Augmented Reinforcement Learning", "Self-Supervised Policy Adaptation during Deployment", "Pre-Trained Image Encoder for Generalizable Visual Reinforcement Learning", "Network Randomization: A Simple Technique for Generalization in Deep Reinforcement Learning", "Improving Generalization in Reinforcement Learning with Mixture Regularization", "Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Sim-to-Real Transfer of Robotic Control with Dynamics Randomization", "BayesSim: adaptive domain randomization via probabilistic inference for robotics simulators"], "answer_arxiv_id": ["2011.13389", "2107.00644v2", "2004.14990", "2107.09645", "2007.04309", "2212.08860", "1910.05396", "2010.10814", "1703.06907", "1710.06537", "1906.01728"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_13498"} +{"question": "Could you provide me some studies that applied symbolic regression in automated scientific discovery from data?", "answer": ["Machine Learning Topological Invariants with Neural Networks", "Discovering conservation laws from trajectories via machine learning", "AI Poincaré 2.0: Machine Learning Conservation Laws from Differential Equations", "AI Feynman: a Physics-Inspired Method for Symbolic Regression"], "answer_arxiv_id": ["1708.09401", "2102.04008", "2203.12610", "1905.11481"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_13499"} +{"question": "Which studies focused on the application and effect of mixup technology in long-tail learning?", "answer": ["Improving Calibration for Long-Tailed Recognition", "mixup: Beyond Empirical Risk Minimization", "Remix: Rebalanced Mixup"], "answer_arxiv_id": ["2104.00466", "1710.09412", "2007.03943"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_13500"} +{"question": "What works have used word embeddings as classifier weights in order to project region features into the text embedding space?", "answer": ["Zero-Shot Object Detection", "Zero-Shot Object Detection by Hybrid Region Embedding"], "answer_arxiv_id": ["1804.04340", "1805.06157"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_13501"} +{"question": "What papers propose QA datasets designed to test reasoning ability similar to the task formulation in the proposed dataset?", "answer": ["ProofWriter: Generating Implications, Proofs, and Abductive Statements\n over Natural Language", "FOLIO: Natural Language Reasoning with First-Order Logic", "Language Models Are Greedy Reasoners: A Systematic Formal Analysis of Chain-of-Thought"], "answer_arxiv_id": ["2012.13048", "2209.00840", "2210.01240v4"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_13502"} +{"question": "What work seeks to incorporate RL into the entailment trees?", "answer": ["RLET: A Reinforcement Learning Based Approach for Explainable QA with\n Entailment Trees"], "answer_arxiv_id": ["2210.17095"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_13503"} +{"question": "Which works propose classifier-guided method in the field of image generation using guided diffusions?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_13504"} +{"question": "Could you provide references on the works that used diffusion-based models for video generation and editing?", "answer": ["Video Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "Pix2Video: Video Editing using Image Diffusion", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "Text2LIVE: Text-Driven Layered Image and Video Editing", "Structure and Content-Guided Video Synthesis with Diffusion Models", "Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "SweetDreamer: Aligning Geometric Priors in 2D Diffusion for Consistent\n Text-to-3D"], "answer_arxiv_id": ["2204.03458", "2210.02303", "2209.14792", "2211.11018", "2303.12688", "2303.13439", "2204.02491", "2302.03011", "2212.11565", "2310.02596"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_13505"} +{"question": "Could you provide me some studies that design unlearning algorithms for simple machine learning algorithms?", "answer": ["Certifiable Machine Unlearning for Linear Models", "Approximate Data Deletion from Machine Learning Models", "Machine Unlearning: Linear Filtration for Logit-based Classifiers", "Machine Unlearning for Random Forests"], "answer_arxiv_id": ["2106.15093", "2002.10077", "2002.02730", "2009.05567"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_13506"} +{"question": "Which papers have introduced Meta-Model Matching methods in the context of data condensation or distillation?", "answer": ["Dataset Distillation with Infinitely Wide Convolutional Networks", "Efficient Dataset Distillation using Random Feature Approximation", "Dataset Distillation using Neural Feature Regression", "Remember the Past: Distilling Datasets into Addressable Memories for Neural Networks"], "answer_arxiv_id": ["2107.13034", "2210.12067", "2206.00719", "2206.02916"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_13507"} +{"question": "Could you provide me with works that study offline RL in a function approximation setting?", "answer": ["Information-Theoretic Considerations in Batch Reinforcement Learning", "Q* Approximation Schemes for Batch Reinforcement Learning: A Theoretical Comparison", "Batch Value-function Approximation with Only Realizability", "Bellman-consistent Pessimism for Offline Reinforcement Learning", "Is Pessimism Provably Efficient for Offline RL?", "Provable Benefits of Actor-Critic Methods for Offline Reinforcement Learning", "Pessimistic Model-based Offline Reinforcement Learning under Partial Coverage", "Near-optimal Offline Reinforcement Learning with Linear Representation: Leveraging Variance Information with Pessimism", "Offline Reinforcement Learning with Realizability and Single-policy Concentrability"], "answer_arxiv_id": ["1905.00360", "2003.03924", "2008.04990", "2106.06926", "2012.15085", "2108.08812", "2107.06226", "2203.05804v1", "2202.04634"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13508"} +{"question": "What papers stated transfer learning to be unsuccessful due to the difference between general-domain text and biomedical text?", "answer": ["Domain-Specific Language Model Pretraining for Biomedical Natural Language Processing"], "answer_arxiv_id": ["2007.15779"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_13509"} +{"question": "What papers discussed the approach of adversarial training in adversarial defenses?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_13510"} +{"question": "What are some studies that incorporate explicit language supervision in training machine learning models for language?", "answer": ["Deep Learning for Sentiment Analysis : A Survey", "A Survey on Deep Learning for Named Entity Recognition", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "Attention Is All You Need"], "answer_arxiv_id": ["1801.07883", "1812.09449", "1810.04805", "2005.14165", "1706.03762"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_13511"} +{"question": "What studies demonstrated that only evaluating average performance of a model can be deceptive?", "answer": ["ID and OOD Performance Are Sometimes Inversely Correlated on Real-world Datasets", "Deconstructing Distributions: A Pointwise Framework of Learning", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Does Progress On Object Recognition Benchmarks Improve Real-World Generalization?"], "answer_arxiv_id": ["2209.00613", "2202.09931", "1911.08731", "2307.13136"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_13512"} +{"question": "What studies employed tactile feedback for cloth manipulation?", "answer": ["Learning to Singulate Layers of Cloth using Tactile Feedback", "Visuotactile Affordances for Cloth Manipulation with Local Control"], "answer_arxiv_id": ["2207.11196", "2212.05108"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_13513"} +{"question": "Which research made progress in understanding the implicit bias in the regression setting of non-linear networks?", "answer": ["Gradient flow dynamics of shallow ReLU networks for square loss and orthogonal inputs"], "answer_arxiv_id": ["2206.00939"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13514"} +{"question": "What paper used a pretrained Stable Diffusion model as an image feature extractor with additional text input?", "answer": ["Unleashing Text-to-Image Diffusion Models for Visual Perception"], "answer_arxiv_id": ["2303.02153v1"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_13515"} +{"question": "What papers show that vulnerabilities exist in the Proof-of-Learning?", "answer": ["“Adversarial Examples” for Proof-of-Learning"], "answer_arxiv_id": ["2108.09454"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_13516"} +{"question": "Could you provide me some studies about maximizing agreement on the latent space using contrastive loss?", "answer": ["Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1807.03748"], "source_meta": {"published_time": "20211202"}, "qid": "AutoScholarQuery_train_13517"} +{"question": "Which studies investigated computational subgraphs for model interpretability?", "answer": ["Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2202.05262"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_13518"} +{"question": "Which paper is the first to propose the transformer-based object detection approach known as DETR?", "answer": ["End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2005.12872"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_13519"} +{"question": "Could you provide some papers about applying machine learning to identify and localize errors in source code?", "answer": ["A Survey of Machine Learning for Big Code and Naturalness"], "answer_arxiv_id": ["1709.06182"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_13520"} +{"question": "Which studies introduced conventional fairness metrics like Equalised Odds, Equalised Opportunity and Demographic Parity that cannot be applied to generative models?", "answer": ["Equality of Opportunity in Supervised Learning", "Certifying and removing disparate impact"], "answer_arxiv_id": ["1610.02413", "1412.3756"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_13521"} +{"question": "Which work empirically observed that the choice of h=g produces better sample generation quality for the generative process?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_13522"} +{"question": "What research papers assumed a linear non-Gaussian acyclic model (LiNGAM)?", "answer": ["DirectLiNGAM: A direct method for learning a linear non-Gaussian structural equation model"], "answer_arxiv_id": ["1101.2489v3"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_13523"} +{"question": "Which studies have focused on structured pruning on different substructures of Vision Transformer (ViT) models?", "answer": ["Unified Visual Transformer Compression", "Vision Transformer Pruning", "Chasing Sparsity in Vision Transformers: An End-to-End Exploration"], "answer_arxiv_id": ["2203.08243", "2104.08500", "2106.04533"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_13524"} +{"question": "What do the papers referenced discuss on the topic of audio-visual event localization?", "answer": ["Audio-Visual Event Localization in Unconstrained Videos", "dual-modality seq2seq network for Audio-Visual event localization", "Positive Sample Propagation along the Audio-Visual Event Line", "AVE-CLIP: AudioCLIP-based Multi-window Temporal Transformer for Audio Visual Event Localization"], "answer_arxiv_id": ["1803.08842", "1902.07473", "2104.00239", "2210.05060"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_13525"} +{"question": "What studies demonstrated that trained diffusion models have internal representations suitable for image segmentation?", "answer": ["Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models", "Diffusion models as plug-and-play priors", "DifFSS: Diffusion Model for Few-Shot Semantic Segmentation"], "answer_arxiv_id": ["2303.04803", "2206.09012", "2307.00773"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_13526"} +{"question": "Which work converts the entire diffusion process into a parallel one by considering the entire diffusion trajectory as a single sample and solving for the fixed point of the trajectory?", "answer": ["Deep Equilibrium Approaches to Diffusion Models"], "answer_arxiv_id": ["2210.12867"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_13527"} +{"question": "Which works use policy optimization for safe RL?", "answer": ["Constrained Policy Optimization", "Reward Constrained Policy Optimization", "IPO: Interior-point Policy Optimization under Constraints", "Responsive Safety in Reinforcement Learning by PID Lagrangian Methods"], "answer_arxiv_id": ["1705.10528", "1805.11074", "1910.09615", "2007.03964"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_13528"} +{"question": "In what papers did researchers explore the theoretical limitations of end-to-end gradients based learning?", "answer": ["Failures of Gradient-Based Deep Learning"], "answer_arxiv_id": ["1703.07950v2"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_13529"} +{"question": "What are the studies that focus on proposing linear models that propagate node information as the discretization of the graph heat equation?", "answer": ["Dissecting the Diffusion Process in Linear Graph Convolutional Networks", "Simplifying Graph Convolutional Networks"], "answer_arxiv_id": ["2102.10739", "1902.07153"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_13530"} +{"question": "What study gave a complete characterization of min-cost solutions in the case of shallow univariate ReLU networks with unregularized bias?", "answer": ["Spontaneous Symmetry Breaking in Generative Diffusion Models"], "answer_arxiv_id": ["2305.19693"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_13531"} +{"question": "Can you list some of the studies that have raised concerns about the robustness of adversarial training methods?", "answer": ["Investigating Vulnerabilities of Deep Neural Policies", "Deep Reinforcement Learning Policies Learn Shared Adversarial Features Across MDPs", "Adversarial Robust Deep Reinforcement Learning Requires Redefining Robustness"], "answer_arxiv_id": ["2108.13093", "2112.09025", "2301.07487"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_13532"} +{"question": "Which studies pioneered the concept of membership inference attacks (MIAs)?", "answer": ["Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting"], "answer_arxiv_id": ["1709.01604"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_13533"} +{"question": "Any papers that represent the scene geometric shapes in terms of local feature stoarge in a dictionary?", "answer": ["Variable Bitrate Neural Fields"], "answer_arxiv_id": ["2206.07707"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_13534"} +{"question": "What studies have extended the work on DragGAN, which enables interactive point-based image editing in diffusion models?", "answer": ["DragDiffusion: Harnessing Diffusion Models for Interactive Point-based\n Image Editing", "DragonDiffusion: Enabling Drag-style Manipulation on Diffusion Models"], "answer_arxiv_id": ["2306.14435", "2307.02421"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_13535"} +{"question": "What works are there in the field of uncertainty estimation in deep networks, especially for safety-critical tasks?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "Deep Evidential Regression", "Evidential Deep Learning to Quantify Classification Uncertainty", "Image-to-Image Regression with Distribution-Free Uncertainty Quantification and Applications in Imaging", "Amortized Conditional Normalized Maximum Likelihood: Reliable Out of Distribution Uncertainty Estimation"], "answer_arxiv_id": ["1506.02142", "1703.04977", "1910.02600", "1806.01768", "2202.05265", "2011.02696"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_13536"} +{"question": "Which study presented the idea of Post-hoc CBM (PCBM) that involves learning a concept subspace in the embedding space of the pre-trained model?", "answer": ["Post-hoc Concept Bottleneck Models"], "answer_arxiv_id": ["2205.15480"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_13537"} +{"question": "Could you give me examples of research that use deep neural networks-based methods for out-of-distribution detection and derive confidence scores based on the output?", "answer": ["Towards Open Set Deep Networks", "A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Generalized ODIN: Detecting Out-of-distribution Image without Learning from Out-of-distribution Data", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "Energy-based Out-of-distribution Detection", "ReAct: Out-of-distribution Detection With Rectified Activations", "POEM: Out-of-Distribution Detection with Posterior Sampling"], "answer_arxiv_id": ["1511.06233", "1610.02136", "2002.11297", "1706.02690", "2010.03759", "2111.12797", "2206.13687"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_13538"} +{"question": "What research papers are concerned with coefficient schedule design, variance strategy optimization, superimposed image decomposition, curve integration, stochastic differential equations, and residual learning for image restoration?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Improved Denoising Diffusion Probabilistic Models", "Variational Diffusion Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in\n Diffusion Probabilistic Models", "Estimating the Optimal Covariance with Imperfect Mean in Diffusion\n Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Deep Residual Learning for Image Recognition", "Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image\n Denoising", "Residual Dense Network for Image Restoration", "Densely Residual Laplacian Super-Resolution", "Multi-Stage Progressive Image Restoration", "MAXIM: Multi-Axis MLP for Image Processing"], "answer_arxiv_id": ["2112.10752", "2102.09672", "2107.00630", "2201.06503", "2206.07309", "2011.13456", "1512.03385", "1608.03981", "1812.10477", "1906.12021", "2102.02808", "2201.02973"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_13539"} +{"question": "What papers discuss the application of zeroth-order optimization in robotics?", "answer": ["Stochastic Zeroth-order Riemannian Derivative Estimation and Optimization"], "answer_arxiv_id": ["2003.11238"], "source_meta": {"published_time": "20220927"}, "qid": "AutoScholarQuery_train_13540"} +{"question": "Can you provide the references that describe the common structure of most recent generative Visual Language Models (VLMs)?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "GIT: A Generative Image-to-text Transformer for Vision and Language", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Qwen-VL: A Versatile Vision-Language Model for Understanding,\n Localization, Text Reading, and Beyond", "mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with\n Modality Collaboration", "CogVLM: Visual Expert for Pretrained Language Models", "Unified-IO: A Unified Model for Vision, Language, and Multi-Modal Tasks"], "answer_arxiv_id": ["2204.14198", "2205.14100", "2301.12597", "2304.08485", "2304.10592", "2308.12966", "2311.04257", "2311.03079", "2206.08916"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_13541"} +{"question": "Where can I find applications of deep learning techniques in the domain of federated RL?", "answer": ["Efficient Parallel Methods for Deep Reinforcement Learning", "IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures", "Distributed Prioritized Experience Replay", "Massively Parallel Methods for Deep Reinforcement Learning", "Federated Deep Reinforcement Learning"], "answer_arxiv_id": ["1705.04862", "1802.01561", "1803.00933", "1507.04296", "1901.08277"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_13542"} +{"question": "Is there any research where OT was used in the context of object detection?", "answer": ["OTA: Optimal Transport Assignment for Object Detection", "Self-supervised object detection from audio-visual correspondence"], "answer_arxiv_id": ["2103.14259", "2104.06401"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_13543"} +{"question": "What are some data-driven methods that used multimodal datasets for text-to-3D generation?", "answer": ["Text2Shape: Generating Shapes from Natural Language by Learning Joint\n Embeddings"], "answer_arxiv_id": ["1803.08495"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_13544"} +{"question": "Could you provide me some studies about the emotional intelligence and empathy of LLMs?", "answer": ["Large Language Models Understand and Can be Enhanced by Emotional\n Stimuli", "Emotionally Numb or Empathetic? Evaluating How LLMs Feel Using\n EmotionBench", "An Appraisal-Based Chain-Of-Emotion Architecture for Affective Language\n Model Game Agents"], "answer_arxiv_id": ["2307.11760", "2308.03656", "2309.05076"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_13545"} +{"question": "What works have been done on determining object-level correspondence, visual object tracking?", "answer": ["Tracking without bells and whistles"], "answer_arxiv_id": ["1903.05625"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13546"} +{"question": "Which papers are about designing learning modules for training a more powerful feature extractor in few-shot class-incremental learning?", "answer": ["Forward Compatible Few-Shot Class-Incremental Learning"], "answer_arxiv_id": ["2203.06953"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13547"} +{"question": "Could you list some studies providing gap-dependent results for multi-armed bandits and RL?", "answer": ["Fine-Grained Gap-Dependent Bounds for Tabular MDPs via Adaptive Multi-Step Bootstrap"], "answer_arxiv_id": ["2102.04692"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_13548"} +{"question": "Is there any paper that formulates the problem of interval estimation for various functionals as a constrained optimization problem over a confidence band?", "answer": ["Universal Off-Policy Evaluation"], "answer_arxiv_id": ["2104.12820"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_13549"} +{"question": "What are the works that proposed to incorporate the Rational Speech Act framework in tasks such as image captioning, translation, dialogue, and text generation?", "answer": ["Pragmatic Inference with a CLIP Listener for Contrastive Captioning", "Pragmatically Informative Image Captioning with Character-Level\n Inference", "Lost in Machine Translation: A Method to Reduce Meaning Loss", "Will I Sound Like Me? Improving Persona Consistency in Dialogues through\n Pragmatic Self-Consciousness", "Perspective-taking and Pragmatics for Generating Empathetic Responses\n Focused on Emotion Causes", "Pragmatically Informative Text Generation"], "answer_arxiv_id": ["2306.08818", "1804.05417", "1902.09514", "2004.05816", "2109.08828", "1904.01301"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_13550"} +{"question": "Which papers discuss the effects of data encoding as input to NLP models on language disparity?", "answer": ["Do All Languages Cost the Same? Tokenization in the Era of Commercial\n Language Models", "Language Model Tokenizers Introduce Unfairness Between Languages"], "answer_arxiv_id": ["2305.13707", "2305.15425v2"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_13551"} +{"question": "Which works contributed to the development and variants of Variational autoencoders and Generative Adversarial Models (GANs)?", "answer": ["Neural Discrete Representation Learning", "Generative Adversarial Networks", "Taming Transformers for High-Resolution Image Synthesis", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Large Scale Adversarial Representation Learning"], "answer_arxiv_id": ["1711.00937", "2203.00667", "2012.09841", "1812.04948", "1809.11096", "1907.02544"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_13552"} +{"question": "What work conducted an analysis of the one-inclusion graph algorithm in the realizable PAC regression setting using the V𝛾-dimension?", "answer": ["Optimal PAC Bounds Without Uniform Convergence"], "answer_arxiv_id": ["2304.09167"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_13553"} +{"question": "Which works established pioneering concepts of binary neural network?", "answer": ["BinaryConnect: Training Deep Neural Networks with binary weights during\n propagations", "Binarized Neural Networks"], "answer_arxiv_id": ["1511.00363", "1602.02505v3"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_13554"} +{"question": "What works don't use DDIM inversion but rather utilize model optimization based on the target text prompt?", "answer": ["DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2110.02711", "2210.09276"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_13555"} +{"question": "What papers are about English-centric LMs leading to a discrepancy between high-resource and low-resource languages?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2302.13971", "2307.09288"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_13556"} +{"question": "Could you provide me some examples of methods that control which components inside the backbone to be updated or fixed during training?", "answer": ["Training Neural Networks with Fixed Sparse Masks", "Parameter-Efficient Transfer Learning with Diff Pruning"], "answer_arxiv_id": ["2111.09839", "2012.07463"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_13557"} +{"question": "Which works employ Multi-layer perceptron (MLP) to predict attributes such as signed distance to scene surface (SDF)?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction"], "answer_arxiv_id": ["1901.05103", "2106.10689", "2106.12052", "2104.10078"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13558"} +{"question": "Which works use various uncertainty estimation techniques, such as Gaussian processes or model ensembles, for model-based RL?", "answer": ["EPOpt: Learning Robust Neural Network Policies Using Model Ensembles", "Model-Based Reinforcement Learning via Meta-Policy Optimization", "Model-Ensemble Trust-Region Policy Optimization", "Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models", "Deep Dynamics Models for Learning Dexterous Manipulation", "MOPO: Model-based Offline Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["1610.01283", "1809.05214", "1802.10592", "1805.12114", "1909.11652", "2005.13239", "2005.05951"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_13559"} +{"question": "Which papers provide information about the foundations of Siamese networks in contrastive learning?", "answer": ["Learning Gradient Fields for Shape Generation"], "answer_arxiv_id": ["2008.06520"], "source_meta": {"published_time": "20210529"}, "qid": "AutoScholarQuery_train_13560"} +{"question": "What works focus on ensemble-based approaches for Bayesian neural networks?", "answer": ["Bootstrapped Thompson Sampling and Deep Exploration", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1507.00300", "1612.01474"], "source_meta": {"published_time": "20210719"}, "qid": "AutoScholarQuery_train_13561"} +{"question": "Are there any papers using contrastive learning in similar domains?", "answer": ["Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learning", "Contrastive Learning Inverts the Data Generating Process", "Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style"], "answer_arxiv_id": ["1805.08651", "2102.08850v4", "2106.04619v4"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_13562"} +{"question": "What papers aimed at improving the training speed of neural radiance fields?", "answer": ["PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2103.14024", "2201.05989", "2112.05131"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_13563"} +{"question": "Which papers introduced and developed the concept of knowledge distillation?", "answer": ["Distilling the Knowledge in a Neural Network", "Contrastive Representation Distillation", "Knowledge Distillation: A Survey", "Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning"], "answer_arxiv_id": ["1503.02531", "1910.10699", "2006.05525", "2012.09816"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_13564"} +{"question": "Which work first introduced prompt tuning to computer vision?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_13565"} +{"question": "What works have proposed alternating minimization as a solution approach to MLR?", "answer": ["Alternating Minimization for Mixed Linear Regression", "Solving a Mixture of Many Random Linear Equations by Tensor Decomposition and Alternating Minimization", "Alternating Minimization Converges Super-Linearly for Mixed Linear Regression", "On Learning Mixture of Linear Regressions in the Non-Realizable Setting"], "answer_arxiv_id": ["1310.3745", "1608.05749v1", "2004.10914v2", "2205.13166"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_13566"} +{"question": "Which works have extended feature visualizations to ViTs?", "answer": ["What do Vision Transformers Learn? A Visual Exploration"], "answer_arxiv_id": ["2212.06727"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_13567"} +{"question": "Who shows the effectiveness of PPO in several standard multi-agent environments?", "answer": ["The Surprising Effectiveness of PPO in Cooperative Multi-Agent Games"], "answer_arxiv_id": ["2103.01955"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_13568"} +{"question": "What studies have been done on Bayesian Neural Networks?", "answer": ["All You Need is a Good Functional Prior for Bayesian Deep Learning", "Model Selection in Bayesian Neural Networks via Horseshoe Priors", "Hands-on Bayesian Neural Networks – A Tutorial for Deep Learning Users"], "answer_arxiv_id": ["2011.12829", "1705.10388", "2007.06823"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_13569"} +{"question": "Which studies highlighted the need for privacy protection in machine learning?", "answer": ["YFCC100M: The New Data in Multimedia Research", "Ego4D: Around the World in 3,000 Hours of Egocentric Video", "Extracting Training Data from Large Language Models", "When Machine Learning Meets Privacy: A Survey and Outlook"], "answer_arxiv_id": ["1503.01817", "2110.07058", "2012.07805", "2011.11819"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_13570"} +{"question": "What articles proposed solutions regarding the problem of training with synthetic data?", "answer": ["Classification Accuracy Score for Conditional Generative Models", "Ensembles of GANs for synthetic training data generation"], "answer_arxiv_id": ["1905.10887", "2104.11797"], "source_meta": {"published_time": "20210503"}, "qid": "AutoScholarQuery_train_13571"} +{"question": "Could you provide me some studies that discussed MAE-based methods for representation learning?", "answer": ["ConvMAE: Masked Convolution Meets Masked Autoencoders", "Designing BERT for Convolutional Networks: Sparse and Hierarchical\n Masked Modeling"], "answer_arxiv_id": ["2205.03892", "2301.03580"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_13572"} +{"question": "Which studies propose an improved framework on Contrastive Hebbian Learning known as equilibrium propagation?", "answer": ["Equilibrium Propagation: Bridging the Gap Between Energy-Based Models and Backpropagation", "Generalization of Equilibrium Propagation to Vector Field Dynamics"], "answer_arxiv_id": ["1602.05179", "1808.04873"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13573"} +{"question": "Could you provide me the examples of papers that report on using the connectivity properties of neural networks for ensembling?", "answer": ["Loss Surface Simplexes for Mode Connecting Volumes and Fast Ensembling", "Averaging Weights Leads to Wider Optima and Better Generalization", "Learning Neural Network Subspaces", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time"], "answer_arxiv_id": ["2102.13042", "1803.05407", "2102.10472", "2203.05482"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_13574"} +{"question": "What studies provide results on sample complexity and generalization bounds for fine-tuning and linear probing in transfer learning?", "answer": ["On the Theory of Transfer Learning: The Importance of Task Diversity", "Few-Shot Learning via Learning the Representation, Provably", "How Fine-Tuning Allows for Effective Meta-Learning", "A Theoretical Analysis of Fine-tuning with Linear Teachers", "Improved Regularization and Robustness for Fine-tuning in Neural Networks"], "answer_arxiv_id": ["2006.11650", "2002.09434", "2105.02221v1", "2107.01641", "2111.04578"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_13575"} +{"question": "What studies incorporate temporal attention layers into pretrained models to capture motion dynamics?", "answer": ["CogVideo: Large-scale Pretraining for Text-to-Video Generation via\n Transformers"], "answer_arxiv_id": ["2205.15868"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_13576"} +{"question": "What studies are about creating a foundation model for Minecraft?", "answer": ["Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online\n Videos"], "answer_arxiv_id": ["2206.11795"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_13577"} +{"question": "Which papers have focused on developing equivariant architectures for data types associated with permutation and Euclidean group symmetries?", "answer": ["Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges", "Symmetry Group Equivariant Architectures for Physics", "Group Equivariant Convolutional Networks", "Steerable CNNs", "Relational inductive biases, deep learning, and graph networks", "Invariant and Equivariant Graph Networks", "Equivariant Subgraph Aggregation Networks", "Vector Neurons: A General Framework for SO(3)-Equivariant Networks", "E(n) Equivariant Graph Neural Networks", "Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds"], "answer_arxiv_id": ["2104.13478", "2203.06153v1", "1602.07576", "1612.08498", "1806.01261", "1812.09902", "2110.02910", "2104.12229", "2102.09844", "1802.08219"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_13578"} +{"question": "Are there any works on panoptic scene graph generation (PSG) and its extension into the video domain?", "answer": ["Panoptic Scene Graph Generation", "Pair then Relation: Pair-Net for Panoptic Scene Graph Generation", "TextPSG: Panoptic Scene Graph Generation from Textual Descriptions", "HiLo: Exploiting High Low Frequency Relations for Unbiased Panoptic Scene Graph Generation", "Haystack: A Panoptic Scene Graph Dataset to Evaluate Rare Predicate Classes", "Panoptic Video Scene Graph Generation", "Video K-Net: A Simple, Strong, and Unified Baseline for Video Segmentation"], "answer_arxiv_id": ["2207.11247", "2307.08699", "2310.07056", "2303.15994", "2309.02286", "2311.17058", "2204.04656"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_13579"} +{"question": "Can you list the studies associated with posthoc methods in concept visualization particularly TCAV and ACE?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)", "Towards Automatic Concept-based Explanations"], "answer_arxiv_id": ["1711.11279", "1902.03129"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_13580"} +{"question": "What work replaced the Fourier basis with wavelet basis?", "answer": ["Graph wavelet neural network"], "answer_arxiv_id": ["1904.07785"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_13581"} +{"question": "Could you cite works that utilized the PPO algorithm in robotics?", "answer": ["Solving Rubik’s Cube with a Robot Hand", "Learning robust perceptive locomotion for quadrupedal robots in the wild", "Learning Dexterous In-Hand Manipulation"], "answer_arxiv_id": ["1910.07113", "2201.08117", "1808.00177"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_13582"} +{"question": "Which work introduced the concept of likelihood ratio attack (LiRA) in the context of membership inference attacks?", "answer": ["Membership Inference Attacks From First Principles"], "answer_arxiv_id": ["2112.03570"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_13583"} +{"question": "What research works encode agent states directly using RNNs for trajectory forecasting?", "answer": ["MultiPath++: Efficient Information Fusion and Trajectory Aggregation for Behavior Prediction", "Trajectron++: Dynamically-Feasible Trajectory Forecasting With Heterogeneous Data", "Multi-Head Attention for Multi-Modal Joint Vehicle Motion Forecasting"], "answer_arxiv_id": ["2111.14973", "2001.03093", "1910.03650"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_13584"} +{"question": "What works utilise deep generative models for 1D sequence prediction in proteins and antibodies?", "answer": ["Fold2Seq: A Joint Sequence(1D)-Fold(3D) Embedding-based Generative Model for Protein Design"], "answer_arxiv_id": ["2106.13058"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_13585"} +{"question": "What work introduced a consistency loss based on KL divergence for few-shot GDA?", "answer": ["Few-shot Image Generation via Cross-domain Correspondence", "Few Shot Generative Model Adaption via Relaxed Spatial Structural Alignment"], "answer_arxiv_id": ["2104.06820", "2203.04121"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13586"} +{"question": "Is there a study that examined which datapoints are most strongly memorized during training using influence functions?", "answer": ["What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation"], "answer_arxiv_id": ["2008.03703"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_13587"} +{"question": "Which studies have shown that the sample complexity is affected by strong moderate confidence terms in the non-asymptotic regime?", "answer": ["Nearly Instance Optimal Sample Complexity Bounds for Top-k Arm Selection", "The Simulator: Understanding Adaptive Sampling in the Moderate-Confidence Regime"], "answer_arxiv_id": ["1702.03605", "1702.05186"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_13588"} +{"question": "Which papers have proposed different model architectures and pre-training datasets for VLMs?", "answer": ["LAION-5B: An open large-scale dataset for training next generation image-text models"], "answer_arxiv_id": ["2210.08402"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_13589"} +{"question": "Any works about building foundation models through task unification?", "answer": ["Florence: A New Foundation Model for Computer Vision", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "CoCa: Contrastive Captioners are Image-Text Foundation Models"], "answer_arxiv_id": ["2111.11432", "2208.10442", "2108.10904", "2205.01917"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_13590"} +{"question": "Is there any work discussing the data-collection and transcription challenges in the setting of low-resource language translation?", "answer": ["Neural Machine Translation for Low-Resource Languages: A Survey"], "answer_arxiv_id": ["2106.15115"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_13591"} +{"question": "Which study developed 4dcontrast that augments scenes with moving synthetic objects in the 3D domain?", "answer": ["4DContrast: Contrastive Learning with Dynamic Correspondences for 3D\n Scene Understanding"], "answer_arxiv_id": ["2112.02990"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_13592"} +{"question": "Could you provide me some references about chain-of-thoughts prompting for in-context learning?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2201.11903", "2205.10625", "2203.11171"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_13593"} +{"question": "Could you provide me some studies that converted the approximation of Shapley values to a weighted least squares problem?", "answer": ["A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1705.07874"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_13594"} +{"question": "Are there any studies that do not update the models but modify the pixels of adversarial samples for self-supervised objective?", "answer": ["Adversarial Attacks are Reversible with Natural Supervision", "Robust Perception through Equivariance"], "answer_arxiv_id": ["2103.14222", "2212.06079"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_13595"} +{"question": "What researches provided datasets for benchmarking methods in the field of 3D interactions from monocular inputs?", "answer": ["BEHAVE: Dataset and Method for Tracking Human Object Interactions", "InterCap: Joint Markerless 3D Tracking of Humans and Objects in\n Interaction", "NeuralDome: A Neural Modeling Pipeline on Multi-View Human-Object\n Interactions"], "answer_arxiv_id": ["2204.06950", "2209.12354", "2212.07626"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_13596"} +{"question": "In what studies was GNN/Transformer module, a hybrid MPNN+Transformer architecture, utilized?", "answer": ["Exphormer: Sparse Transformers for Graphs"], "answer_arxiv_id": ["2303.06147"], "source_meta": {"published_time": "20221227"}, "qid": "AutoScholarQuery_train_13597"} +{"question": "Which works involve in testing the model’s ability to deduct an ambiguous entity or ask clarification questions?", "answer": ["Asking Clarifying Questions in Open-Domain Information-Seeking\n Conversations", "Contrastive Multi-document Question Generation"], "answer_arxiv_id": ["1907.06554", "1911.03047"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_13598"} +{"question": "Which papers focused on solving the slow sampling problem of Diffusion Probabilistic Models (DPMs)?", "answer": ["Denoising Diffusion Implicit Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps", "Elucidating the Design Space of Diffusion-Based Generative Models", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic\n Models"], "answer_arxiv_id": ["2010.02502", "2206.00927", "2206.00364v2", "2211.01095"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_13599"} +{"question": "Which research works showed that pre-trained large language models didn't help to improve the performance of ICD coding?", "answer": ["Towards BERT-based Automatic ICD Coding: Limitations and Opportunities", "Does the Magic of BERT Apply to Medical Code Assignment? A Quantitative Study"], "answer_arxiv_id": ["2104.06709", "2103.06511"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_13600"} +{"question": "Which works are about visual recognition models that make predictions based on a fixed set of classes?", "answer": ["Microsoft COCO: Common Objects in Context"], "answer_arxiv_id": ["1405.0312"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_13601"} +{"question": "Which papers have investigated subject customization in terms of customized generation?", "answer": ["SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "Subject-driven Text-to-Image Generation via Apprenticeship Learning", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "Continual Diffusion: Continual Customization of Text-to-Image Diffusion\n with C-LoRA", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2303.11305", "2304.00186", "2302.13848", "2304.03411", "2304.06027", "2307.06949"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_13602"} +{"question": "Which studies generate the weights of implicit neural representations such as radiance fields or signed distance functions?", "answer": ["From data to functa: Your data point is a function and you can treat it\n like one", "Shap-E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2201.12204", "2305.02463"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_13603"} +{"question": "Are there any studies that relax or break the order of play assumption in strategic classification?", "answer": ["Strategic Representation"], "answer_arxiv_id": ["2206.08542"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_13604"} +{"question": "What works have annotated attributes of X-ray images and skin disease images?", "answer": ["MTNeuro: A Benchmark for Evaluating Representations of Brain Structure Across Multiple Levels of Abstraction", "Chest ImaGenome Dataset for Clinical Reasoning"], "answer_arxiv_id": ["2301.00345", "2108.00316"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13605"} +{"question": "What are references that attempted using cognitive tests on language learning models like GPT-3 and GPT-4?", "answer": ["Sparks of Artificial General Intelligence: Early experiments with GPT-4"], "answer_arxiv_id": ["2303.12712v5"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_13606"} +{"question": "Could you provide me with papers in which Schrödinger Bridges are utilized in image generation?", "answer": ["Solving Schrödinger Bridges via Maximum Likelihood"], "answer_arxiv_id": ["2106.02081"], "source_meta": {"published_time": "20220316"}, "qid": "AutoScholarQuery_train_13607"} +{"question": "What works improved or built upon the concept of PointNet?", "answer": ["PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "Dynamic Graph CNN for Learning on Point Clouds", "KPConv: Flexible and Deformable Convolution for Point Clouds", "PCT: Point cloud transformer", "Point Transformer"], "answer_arxiv_id": ["1706.02413", "1801.07829", "1904.08889", "2012.09688", "2012.09164"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_13608"} +{"question": "Which papers propose theories explaining language models (LMs) perform instance conditional learning (ICL) through induction heads?", "answer": ["In-context Learning and Induction Heads", "Label Words are Anchors: An Information Flow Perspective for\n Understanding In-Context Learning"], "answer_arxiv_id": ["2209.11895v1", "2305.14160"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_13609"} +{"question": "Which research attempted to capture temporal information in the videos by using time-contrastive learning?", "answer": ["R3M: A Universal Visual Representation for Robot Manipulation", "Time-Contrastive Networks: Self-Supervised Learning from Video"], "answer_arxiv_id": ["2203.12601", "1704.06888"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13610"} +{"question": "Which research papers consider approximations resulting from a Taylor expansion?", "answer": ["Semi-analytical approximations to statistical moments of sigmoid and softmax mappings of normal variables", "Correlated Input-Dependent Label Noise in Large-Scale Image Classification"], "answer_arxiv_id": ["1703.00091v2", "2105.10305"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_13611"} +{"question": "What works have implemented deep generative models in physics to simulate particle showers, reducing the dependence on computational resources?", "answer": ["CaloDVAE : Discrete Variational Autoencoders for Fast Calorimeter Shower Simulation", "CaloFlow: Fast and Accurate Generation of Calorimeter Showers with Normalizing Flows", "CaloGAN: Simulating 3D High Energy Particle Showers in Multi-Layer Electromagnetic Calorimeters with Generative Adversarial Networks"], "answer_arxiv_id": ["2210.07430", "2106.05285", "1712.10321"], "source_meta": {"published_time": "20220210"}, "qid": "AutoScholarQuery_train_13612"} +{"question": "Can you provide research that discusses how the capabilities of LLMs are unlocked through instruction tuning?", "answer": ["Training language models to follow instructions with human feedback", "Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2203.02155", "2210.11416"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_13613"} +{"question": "What papers are about neural network-based level generation for Super Mario?", "answer": ["Procedural Content Generation using Neuroevolution and Novelty Search for Diverse Video Game Levels", "Dungeon and Platformer Level Blending and Generation using Conditional VAEs", "Conditional Level Generation and Game Blending", "Generating and Blending Game Levels via Quality-Diversity in the Latent Space of a Variational Autoencoder"], "answer_arxiv_id": ["2204.06934", "2106.12692", "2010.07735", "2102.12463"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_13614"} +{"question": "What studies address the limitations in the offline setting under unknown data collection?", "answer": ["Safe Policy Improvement with Baseline Bootstrapping", "Off-Policy Deep Reinforcement Learning without Exploration"], "answer_arxiv_id": ["1712.06924", "1812.02900"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_13615"} +{"question": "Can you provide me some studies using an additional constraint like momentum encoder, stop gradient method, and reconstruction loss to optimize the performance of the Siamese network?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Exploring Simple Siamese Representation Learning", "Leveraging Shape Completion for 3D Siamese Tracking", "Relighting Images in the Wild with a Self-Supervised Siamese Auto-Encoder"], "answer_arxiv_id": ["2006.07733", "2011.10566", "1903.01784", "2012.06444"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_13616"} +{"question": "Which papers discussed approaches using recurrent neural network layers for object detection?", "answer": ["Learning to Detect Objects with a 1 Megapixel Event Camera"], "answer_arxiv_id": ["2009.13436"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_13617"} +{"question": "What papers proposed methods to enforce specific properties on neural networks?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "POLICE: Provably Optimal Linear Constraint Enforcement for Deep Neural Networks"], "answer_arxiv_id": ["1612.00593", "2211.01340"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_13618"} +{"question": "Could you provide examples of invariant GNNs used in the prediction of molecular properties?", "answer": ["SchNet – a deep learning architecture for molecules and materials", "Directional Message Passing for Molecular Graphs", "Spherical Message Passing for 3D Molecular Graphs", "ComENet: Towards Complete and Efficient Message Passing for 3D Molecular Graphs"], "answer_arxiv_id": ["1712.06113", "2003.03123", "2102.05013", "2206.08515"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13619"} +{"question": "Any works that extend the label models of data programming for use in multitask models?", "answer": ["Training Complex Models with Multi-Task Weak Supervision"], "answer_arxiv_id": ["1810.02840v2"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_13620"} +{"question": "What papers discuss parametric models used for 3D human modeling?", "answer": ["Total Capture: A 3D Deformation Model for Tracking Faces, Hands, and\n Bodies", "End-to-end Recovery of Human Shape and Pose", "STAR: Sparse Trained Articulated Human Body Regressor"], "answer_arxiv_id": ["1801.01615", "1712.06584", "2008.08535"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_13621"} +{"question": "Which papers developed approaches for subject-driven image generation?", "answer": ["Instance-Conditioned GAN", "MyStyle: A Personalized Generative Prior", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion", "KNN-Diffusion: Image Generation via Large-Scale Retrieval", "Re-Imagen: Retrieval-Augmented Text-to-Image Generator"], "answer_arxiv_id": ["2109.05070", "2203.17272", "2208.12242", "2208.01618", "2204.02849", "2209.14491"], "source_meta": {"published_time": "20230401"}, "qid": "AutoScholarQuery_train_13622"} +{"question": "What is the research that introduced the empathetic listener mixture model?", "answer": ["MoEL: Mixture of Empathetic Listeners"], "answer_arxiv_id": ["1908.07687"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_13623"} +{"question": "What papers report about the use of priors in the context of offline-RL?", "answer": ["Keep Doing What Worked: Behavioral Modelling Priors for Offline Reinforcement Learning", "Behavior Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["2002.08396", "1911.11361"], "source_meta": {"published_time": "20220120"}, "qid": "AutoScholarQuery_train_13624"} +{"question": "Could you tell me some studies that have explored temporal cues for robust pose estimation in crowded scenes?", "answer": ["VNect: Real-time 3D Human Pose Estimation with a Single RGB Camera"], "answer_arxiv_id": ["1705.01583"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_13625"} +{"question": "Which works describe the use of object-centric representations to improve visual understanding?", "answer": ["Joint Hand Motion and Interaction Hotspots Prediction from Egocentric Videos", "VIOLA: Imitation Learning for Vision-Based Manipulation with Object Proposal Priors", "Shaping embodied agent behavior with activity-context priors from egocentric video", "Ego-Exo: Transferring Visual Representations from Third-person to First-person Videos"], "answer_arxiv_id": ["2204.01696", "2210.11339", "2110.07692", "2104.07905"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13626"} +{"question": "Can you mention the work that developed the Trained Mixture Classifier (TMC) based on Dempster’s combination rule?", "answer": ["Trusted Multi-View Classification"], "answer_arxiv_id": ["2102.02051"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_13627"} +{"question": "Which work is similar to self-training and uses a part of the test nodes for training?", "answer": ["Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning"], "answer_arxiv_id": ["1801.07606"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_13628"} +{"question": "What works explored higher-dimensional sheaf-based neural networks?", "answer": ["Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs", "Sheaf Neural Networks with Connection Laplacians"], "answer_arxiv_id": ["2202.04579", "2206.08702"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_13629"} +{"question": "Which work proposed leveraging multiple scales of features for supervised contrastive learning?", "answer": ["Multi-scale and Cross-scale Contrastive Learning for Semantic\n Segmentation"], "answer_arxiv_id": ["2203.13409"], "source_meta": {"published_time": "20240416"}, "qid": "AutoScholarQuery_train_13630"} +{"question": "Any works about the use of CLIP as a classifier with a correct prompt template in vision-language models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_13631"} +{"question": "Could you provide me research where the researchers pre-finetune language models on multiple datasets?", "answer": ["Muppet: Massive Multi-task Representations with Pre-Finetuning"], "answer_arxiv_id": ["2101.11038"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_13632"} +{"question": "What studies can you refer to for LiDAR-based 3D object detection models?", "answer": ["PointPillars: Fast Encoders for Object Detection from Point Clouds", "Real-Time And Robust 3D Object Detection with Roadside LiDARs"], "answer_arxiv_id": ["1812.05784", "2207.05200"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_13633"} +{"question": "Which research papers use the MCD principle for domain adaptation, aligning a source model to unlabeled target data?", "answer": ["Maximum Classifier Discrepancy for Unsupervised Domain Adaptation", "Unsupervised Visual Domain Adaptation: A Deep Max-Margin Gaussian Process Approach"], "answer_arxiv_id": ["1712.02560", "1902.08727"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_13634"} +{"question": "Could you provide works which proposed variational autoencoder-based methods for conditional generation of protein sequences?", "answer": ["Design of metalloproteins and novel protein folds using variational autoencoders"], "answer_arxiv_id": ["1806.09900"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_13635"} +{"question": "What work resolved the small-loss regret for tabular MDPs?", "answer": ["Bias no more: high-probability data-dependent regret bounds for adversarial bandits and MDPs"], "answer_arxiv_id": ["2006.08040"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13636"} +{"question": "Which paper explored the essential balance between distortion and realism for reconstructed images?", "answer": ["Rethinking Lossy Compression: The Rate-Distortion-Perception Tradeoff"], "answer_arxiv_id": ["1901.07821"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_13637"} +{"question": "What is the reference for the claim that safety concerns take precedence while training RL agents in the real world?", "answer": ["MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learning", "Trustworthy Reinforcement Learning Against Intrinsic Vulnerabilities: Robustness, Safety, and Generalizability", "A Review of Safe Reinforcement Learning: Methods, Theory and Applications"], "answer_arxiv_id": ["2109.12674", "2209.08025", "2205.10330"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_13638"} +{"question": "Which study conducts evaluations of methods to extract confidence scores from probabilities output by LLMs trained via Reinforcement Learning with Human Feedback?", "answer": ["Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence\n Scores from Language Models Fine-Tuned with Human Feedback"], "answer_arxiv_id": ["2305.14975"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_13639"} +{"question": "Can you mention some studies that work on the Anomaly Detection problem under domain or distribution shift?", "answer": ["Cross-Domain Video Anomaly Detection without Target Domain Adaptation", "Few-shot Scene-adaptive Anomaly Detection", "Anomaly Detection under Distribution Shift", "Anomaly Heterogeneity Learning for Open-set Supervised Anomaly Detection", "Catching Both Gray and Black Swans: Open-set Supervised Anomaly\n Detection"], "answer_arxiv_id": ["2212.07010", "2007.07843", "2303.13845", "2310.12790", "2203.14506"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_13640"} +{"question": "What research is involved in introducing a diverse range of modalities that can be used as conditions to control the image generation process in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_13641"} +{"question": "Which works discuss the problems of probability density in cumulative intensity models with a neural network?", "answer": ["Fully Neural Network based Model for General Temporal Point Processes"], "answer_arxiv_id": ["1905.09690"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_13642"} +{"question": "Which papers discuss the extension of GANs into 3D space using voxel-based representations?", "answer": ["HoloGAN: Unsupervised learning of 3D representations from natural images", "3D Shape Induction from 2D Views of Multiple Objects", "Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling"], "answer_arxiv_id": ["1904.01326", "1612.05872", "1610.07584"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_13643"} +{"question": "Which papers discussed the application of pre-trained Transformers in Natural Language Processing (NLP)?", "answer": ["XLNet: Generalized Autoregressive Pretraining for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge", "PaLM: Scaling Language Modeling with Pathways", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["1906.08237", "1907.11692", "1811.00937", "2204.02311", "2201.11903"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_13644"} +{"question": "What papers were referenced regarding the usage of composite tokens in code?", "answer": ["InCoder: A Generative Model for Code Infilling and Synthesis", "Learning Programmatic Idioms for Scalable Semantic Parsing", "Program Synthesis and Semantic Parsing with Learned Code Idioms"], "answer_arxiv_id": ["2204.05999", "1904.09086", "1906.10816"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_13645"} +{"question": "Which SLAM techniques operate on raw image intensities and maintain maps represented by point clouds?", "answer": ["Direct Sparse Odometry"], "answer_arxiv_id": ["1607.02565"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_13646"} +{"question": "Which work is responsible for creating the MovieCLIP dataset for visual scene recognition in movies?", "answer": ["MovieCLIP: Visual Scene Recognition in Movies"], "answer_arxiv_id": ["2210.11065"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_13647"} +{"question": "What works have been conducted on the paradigm of noisy correspondence rectification or calibration?", "answer": ["Noisy Correspondence Learning with Meta Similarity Correction", "BiCro: Noisy Correspondence Rectification for Multi-modality Data via Bi-directional Cross-modal Similarity Consistency"], "answer_arxiv_id": ["2304.06275", "2303.12419v2"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_13648"} +{"question": "Are there any papers regarding the design of robust algorithms for OCC algorithms based on deep neural nets?", "answer": ["DROCC: Deep Robust One-Class Classification"], "answer_arxiv_id": ["2002.12718"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_13649"} +{"question": "What research work unifies object detection and visual grounding into a grounded language-image pre-training model?", "answer": ["Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2112.03857"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_13650"} +{"question": "Which works proposed the concept of single-shot pruning (SSP) for offline RL?", "answer": ["Single-Shot Pruning for Offline Reinforcement Learning"], "answer_arxiv_id": ["2112.15579"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_13651"} +{"question": "Which paper reveals that Soft Q-learning can perform badly in general (non-tree) DAGs?", "answer": ["Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation"], "answer_arxiv_id": ["2106.04399"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_13652"} +{"question": "Whose research employed influence functions to detect adversarial examples?", "answer": ["Detecting Adversarial Samples Using Influence Functions and Nearest Neighbors"], "answer_arxiv_id": ["1909.06872"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_13653"} +{"question": "What researches aim to improve learned representations by exploring features coding in contrastive learning?", "answer": ["Data-Efficient Image Recognition with Contrastive Predictive Coding", "Self-supervised Representation Learning with Relative Predictive Coding"], "answer_arxiv_id": ["1905.09272", "2103.11275"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_13654"} +{"question": "What are some of the works that explore tackling training-time poisoning attacks in offline RL?", "answer": ["Policy Poisoning in Batch Reinforcement Learning and Control", "COPA: Certifying Robust Policies for Offline Reinforcement Learning against Poisoning Attacks", "Corruption-Robust Offline Reinforcement Learning"], "answer_arxiv_id": ["1910.05821", "2203.08398", "2106.06630"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_13655"} +{"question": "Which studies propose methods that learn a joint distribution of all features to impute real-valued missing data?", "answer": ["Variational Autoencoder with Arbitrary Conditioning", "MIWAE: Deep Generative Modelling and Imputation of Incomplete Data Sets", "Handling Incomplete Heterogeneous Data using VAEs", "Variational Selective Autoencoder: Learning from Partially-Observed Heterogeneous Data", "Missing Data Imputation and Acquisition with Deep Hierarchical Models and Hamiltonian Monte Carlo", "GAIN: Missing Data Imputation using Generative Adversarial Nets", "MisGAN: Learning from Incomplete Data with Generative Adversarial Networks", "Multiple Imputation via Generative Adversarial Network for High-dimensional Blockwise Missing Value Problems", "FragmGAN: Generative Adversarial Nets for Fragmentary Data Imputation and Prediction", "Handling Missing Data with Graph Representation Learning", "A Graph-based Imputation Method for Sparse Medical Records", "GEDI: A Graph-based End-to-end Data Imputation Framework", "Simultaneous Missing Value Imputation and Structure Learning with Groups", "Handling Missing Data via Max-Entropy Regularized Graph Autoencoder", "MCFlow: Monte Carlo Flow Models for Data Imputation", "EMFlow: Data Imputation in Latent Space via EM and Deep Flow Models", "Multiple Imputation with Neural Network Gaussian Process for High-dimensional Incomplete Data"], "answer_arxiv_id": ["1806.02382", "1812.02633", "1807.03653", "2102.12679v1", "2202.04599", "1806.02920", "1902.09599", "2112.11507v1", "2203.04692", "2010.16418", "2111.09084", "2208.06573", "2110.08223", "2211.16771", "2003.12628", "2106.04804", "2211.13297"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_13656"} +{"question": "Are there any works that lift the assumption of sender and receiver sharing a prior state distribution?", "answer": ["Learning to Persuade on the Fly: Robustness Against Ignorance"], "answer_arxiv_id": ["2102.10156"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_13657"} +{"question": "Which paper postulates that an atomic intervention set is a minimal sized verifying set for G if and only if the set is a minimum vertex cover of covered edges?", "answer": ["Verification and search algorithms for causal DAGs"], "answer_arxiv_id": ["2206.15374"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_13658"} +{"question": "Which studies are examples of similarity-based data-driven metrics in machine learning?", "answer": ["CLIPScore: A Reference-free Evaluation Metric for Image Captioning", "BERTScore: Evaluating Text Generation with BERT", "Mutual Information Divergence: A Unified Metric for Multimodal\n Generative Models"], "answer_arxiv_id": ["2104.08718", "1904.09675", "2205.13445"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_13659"} +{"question": "What studies fall into the sample selection approach and aims to exploit the memorization effect of DNNs for robust training?", "answer": ["DivideMix: Learning with Noisy Labels as Semi-supervised Learning"], "answer_arxiv_id": ["2002.07394"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_13660"} +{"question": "What works focused on identifying settings from non-independent and identical data in disentanglement representation learning?", "answer": ["Weakly-Supervised Disentanglement Without Compromises", "Learning disentangled representations via product manifold projection", "Recurrent Independent Mechanisms", "CITRIS: Causal Identifiability from Temporal Intervened Sequences", "Weakly supervised causal representation learning", "Learning Temporally Causal Latent Processes from General Temporal Data", "Invariant Risk Minimization Games", "Linear Causal Disentanglement via Interventions", "Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICA"], "answer_arxiv_id": ["2002.02886", "2103.01638", "1909.10893v6", "2202.03169", "2203.16437", "2110.05428", "2002.04692", "2211.16467v3", "2107.10098"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_13661"} +{"question": "What are some papers about enhancing vision transformers using efficient local-attention modules?", "answer": ["CCNet: Criss-Cross Attention for Semantic Segmentation", "Axial Attention In Multidimensional Transformers", "Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Twins: Revisiting the Design of Spatial Attention in Vision Transformers", "Focal Self-attention for Local-Global Interactions in Vision Transformers", "Glance-and-Gaze Vision Transformer", "CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows", "MaxViT: Multi-Axis Vision Transformer"], "answer_arxiv_id": ["1811.11721", "1912.12180", "2003.07853", "2103.14030", "2104.13840", "2107.00641", "2106.02277", "2107.00652", "2204.01697"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_13662"} +{"question": "What research proposed to use NeRF and mesh for high-resolution 3D content generation?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation"], "answer_arxiv_id": ["2211.10440"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_13663"} +{"question": "What study proposed Paint-By-Example to inpaint a masked image?", "answer": ["Paint by Example: Exemplar-based Image Editing with Diffusion Models"], "answer_arxiv_id": ["2211.13227"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_13664"} +{"question": "Which work demonstrated the superior performance of the visual reasoning method based on object-centric representation and self-attention on various visual reasoning domains?", "answer": ["Attention over learned object embeddings enables complex visual reasoning"], "answer_arxiv_id": ["2012.08508"], "source_meta": {"published_time": "20220618"}, "qid": "AutoScholarQuery_train_13665"} +{"question": "What research proposed methods to correct vision classifiers using language inputs?", "answer": ["On Guiding Visual Attention with Language Specification"], "answer_arxiv_id": ["2202.08926"], "source_meta": {"published_time": "20230408"}, "qid": "AutoScholarQuery_train_13666"} +{"question": "Can you mention some works that have achieved optimal results reaching the lower bound?", "answer": ["Accelerated Primal-Dual Gradient Method for Smooth and Convex-Concave Saddle-Point Problems with Bilinear Coupling", "Lifted Primal-Dual Method for Bilinearly Coupled Smooth Minimax Optimization", "Sharper Rates for Separable Minimax and Finite Sum Optimization via Primal-Dual Extragradient Methods"], "answer_arxiv_id": ["2112.15199", "2201.07427v1", "2202.04640"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_13667"} +{"question": "Which works established the graph convolutional networks (GCN) or graph attention networks (GAT)?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Simple and Deep Graph Convolutional Networks", "Graph Attention Networks"], "answer_arxiv_id": ["1609.02907", "2007.02133", "1710.10903"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_13668"} +{"question": "What research used regression by classification for problems with sharp borders in motion, such as stereo disparity?", "answer": ["Wasserstein Distances for Stereo Disparity Estimation"], "answer_arxiv_id": ["2007.03085"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_13669"} +{"question": "What research extended the results on implicit bias to other algorithms like momentum-based GD and SGD?", "answer": ["Characterizing Implicit Bias in Terms of Optimization Geometry", "Fast Margin Maximization via Dual Acceleration", "Stochastic Gradient Descent on Separable Data: Exact Convergence with a Fixed Learning Rate"], "answer_arxiv_id": ["1802.08246", "2107.00595", "1806.01796"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_13670"} +{"question": "Could you provide me studies that demonstrated significantly reducing compute and memory in the model personalization?", "answer": ["Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2302.12228"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_13671"} +{"question": "Which works utilized NeRF for applications such as large-scale urban reconstruction, human face or body reconstruction, and robotic applications?", "answer": ["MARS: An Instance-aware, Modular and Realistic Simulator for Autonomous Driving", "PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes", "Render for CNN: Viewpoint Estimation in Images Using CNNs Trained with Rendered 3D Model Views", "iNeRF: Inverting Neural Radiance Fields for Pose Estimation", "NeRRF: 3D Reconstruction and View Synthesis for Transparent and Specular Objects with Neural Refractive-Reflective Fields", "iMAP: Implicit Mapping and Positioning in Real-Time", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "LATITUDE: Robotic Global Localization with Truncated Dynamic Low-pass Filter in City-scale NeRF"], "answer_arxiv_id": ["2307.15058", "1711.00199", "1505.05641", "2012.05877", "2309.13039", "2103.12352", "2112.12130", "2209.08498"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_13672"} +{"question": "Which survey papers provide an overview of semantic parsing?", "answer": ["A Survey on Semantic Parsing"], "answer_arxiv_id": ["1812.00978"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13673"} +{"question": "Any studies about binary mask-based matting?", "answer": ["Mask Guided Matting via Progressive Refinement Network"], "answer_arxiv_id": ["2012.06722"], "source_meta": {"published_time": "20240424"}, "qid": "AutoScholarQuery_train_13674"} +{"question": "Which works proposed the concept of virtual node in MPNN and observed improvements in performance through the use of virtual node?", "answer": ["Neural Message Passing for Quantum Chemistry", "OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphs", "Open Graph Benchmark: Datasets for Machine Learning on Graphs"], "answer_arxiv_id": ["1704.01212", "2103.09430", "2005.00687"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_13675"} +{"question": "What are the studies that the separation rank bounds extend?", "answer": ["Inductive Bias of Deep Convolutional Networks through Pooling Geometry", "Deep Learning and Quantum Entanglement: Fundamental Connections with Implications to Network Design", "Expressive power of recurrent neural networks"], "answer_arxiv_id": ["1605.06743", "1704.01552v2", "1711.00811"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_13676"} +{"question": "Which works mention the use of Bilevel Optimization (BLO) in the context of machine learning?", "answer": ["Investigating Bi-Level Optimization for Learning and Vision from a Unified Perspective: A Survey and Beyond", "Gradient-based Bi-level Optimization for Deep Learning: A Survey"], "answer_arxiv_id": ["2101.11517", "2207.11719"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_13677"} +{"question": "Could you provide me some works related to reducing the complexity of Online Newton Step (ONS) using sketching?", "answer": ["Efficient Second Order Online Learning by Sketching", "Robust Frequent Directions with Application in Online Learning"], "answer_arxiv_id": ["1602.02202", "1705.05067"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_13678"} +{"question": "Have any studies tried to recapture the favorable properties of state space models while improving the performance of attention models?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Block-Recurrent Transformers"], "answer_arxiv_id": ["1901.02860", "2203.07852"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_13679"} +{"question": "Where can I find more about the Physics-Informed Neural Operator that makes the data-driven Fourier Neural Operator physics-informed?", "answer": ["Physics-Informed Neural Operator for Learning Partial Differential Equations", "Physics-Informed Deep Neural Operator Networks"], "answer_arxiv_id": ["2111.03794", "2207.05748"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_13680"} +{"question": "Are there any studies masterminding advancements in multi-level smooth SCO?", "answer": ["Multi-Level Composite Stochastic Optimization via Nested Variance Reduction"], "answer_arxiv_id": ["1908.11468"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_13681"} +{"question": "Could you provide me references about action recognition researches using multi-modal data?", "answer": ["EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action Recognition", "Audiovisual SlowFast Networks for Video Recognition", "SCSampler: Sampling Salient Clips from Video for Efficient Action Recognition", "Listen to Look: Action Recognition by Previewing Audio", "AdaMML: Adaptive Multi-Modal Learning for Efficient Video Recognition"], "answer_arxiv_id": ["1908.08498", "2001.08740", "1904.04289", "1912.04487", "2105.05165"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_13682"} +{"question": "Can you provide the work that studied policy iteration with function approximation in discounted-reward Markov Decision Processes (MDPs)?", "answer": ["The Role of Lookahead and Approximate Policy Evaluation in Reinforcement Learning with Linear Value Function Approximation"], "answer_arxiv_id": ["2109.13419v7"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13683"} +{"question": "Which paper proposed using a disagreement-based notion of uncertainty to construct an importance weighted predictor with theoretical guarantees?", "answer": ["Importance Weighted Active Learning"], "answer_arxiv_id": ["0812.4952"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_13684"} +{"question": "What studies used Fourier layers to replace spatial self-attention for computer vision tasks?", "answer": ["Adaptive Fourier Neural Operators: Efficient Token Mixers for Transformers"], "answer_arxiv_id": ["2111.13587"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_13685"} +{"question": "Can you name me research papers that the researchers compared with their proposed methods but focused only on classification tasks?", "answer": ["Training language models to follow instructions with human feedback", "MAGMA – Multimodal Augmentation of Generative Models through Adapter-based Finetuning", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2203.02155", "2112.05253", "2201.12086"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_13686"} +{"question": "What papers can be referenced regarding VLM's role in visual question answering?", "answer": ["VQA: Visual Question Answering"], "answer_arxiv_id": ["1505.00468"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_13687"} +{"question": "What studies decode coordinates and recognition results directly by predicting sequence in scene text spotting?", "answer": ["SPTS: Single-Point Text Spotting", "SPTS v2: Single-Point Scene Text Spotting", "Towards Unified Scene Text Spotting based on Sequence Generation", "DeepSolo++: Let Transformer Decoder with Explicit Points Solo for\n Multilingual Text Spotting"], "answer_arxiv_id": ["2112.07917", "2301.01635", "2304.03435", "2305.19957"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_13688"} +{"question": "Which works introduced new score functions to distinguish In-distribution (ID) samples from Out-of-distribution (OOD) samples?", "answer": ["Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "Energy-based Out-of-distribution Detection"], "answer_arxiv_id": ["1706.02690", "1807.03888", "2010.03759"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_13689"} +{"question": "Which work proposed the mask transformer framework for end-to-end panoptic segmentation?", "answer": ["MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers"], "answer_arxiv_id": ["2012.00759"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_13690"} +{"question": "What are some QD-RL algorithms that leverage both on-policy and off-policy RL?", "answer": ["Approximating Gradients for Differentiable Quality Diversity in Reinforcement Learning", "Diversity Policy Gradient for Sample Efficient Quality-Diversity Optimization"], "answer_arxiv_id": ["2202.03666", "2006.08505"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13691"} +{"question": "Are there any research papers about vision-language models that support in-context learning?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "OpenFlamingo: An Open-Source Framework for Training Large Autoregressive\n Vision-Language Models"], "answer_arxiv_id": ["2204.14198", "2305.03726", "2308.01390"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_13692"} +{"question": "Could you mention some studies focusing on the collaboration in federated learning?", "answer": ["One for One, or All for All: Equilibria and Optimality of Collaboration in Federated Learning", "Mechanisms that Incentivize Data Sharing in Federated Learning"], "answer_arxiv_id": ["2103.03228", "2207.04557v1"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_13693"} +{"question": "Could you provide the studies about the application of a video transformer to encode video features?", "answer": ["VideoChat: Chat-Centric Video Understanding"], "answer_arxiv_id": ["2305.06355"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_13694"} +{"question": "What research incorporated the deployment of CLIP embedding cosine similarity in retrieval-based diffusion models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_13695"} +{"question": "What are some previous approaches proposed to address the issues in multi-concept generation?", "answer": ["Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image Diffusion Models", "Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesis"], "answer_arxiv_id": ["2301.13826", "2212.05032"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_13696"} +{"question": "Which works are about GAN-based and VAE-based models for text-conditioned image generation?", "answer": ["DAE-GAN: Dynamic Aspect-aware GAN for Text-to-Image Synthesis", "Controllable Text-to-Image Generation", "Taming Transformers for High-Resolution Image Synthesis", "Muse: Text-To-Image Generation via Masked Generative Transformers"], "answer_arxiv_id": ["2108.12141", "1909.07083", "2012.09841", "2301.00704"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_13697"} +{"question": "Could you provide me some works where DDPM has been used for image inpainting?", "answer": ["RePaint: Inpainting using Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2201.09865"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_13698"} +{"question": "Which papers discuss image-language pretraining with web images paired with alt-text?", "answer": ["LAION-5B: An open large-scale dataset for training next generation\n image-text models", "GIT: A Generative Image-to-text Transformer for Vision and Language", "PaLI: A Jointly-Scaled Multilingual Language-Image Model", "EVA-CLIP: Improved Training Techniques for CLIP at Scale"], "answer_arxiv_id": ["2210.08402", "2205.14100", "2209.06794", "2303.15389"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_13699"} +{"question": "Are there any works on the development of advanced solution pipelines in Neural Combinatorial Optimization?", "answer": ["An Efficient Graph Convolutional Network Technique for the Travelling Salesman Problem", "Learning Collaborative Policies to Solve NP-hard Routing Problems", "Learning the Travelling Salesperson Problem Requires Rethinking Generalization", "Simulation-guided Beam Search for Neural Combinatorial Optimization", "DIFUSCO: Graph-based Diffusion Solvers for Combinatorial Optimization", "H-TSP: Hierarchically Solving the Large-Scale Travelling Salesman Problem", "Learning to Schedule Heuristics in Branch-and-Bound"], "answer_arxiv_id": ["1906.01227", "2110.13987", "2006.07054", "2207.06190", "2302.08224", "2304.09395", "2103.10294"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_13700"} +{"question": "Could you provide some papers discussing an approach to semantic image editing that involves training an end-to-end architecture based on transformer models?", "answer": ["ManiTrans: Entity-Level Text-Guided Image Manipulation via Token-wise Semantic Alignment and Generation", "End-to-End Visual Editing with a Generatively Pre-Trained Artist", "EdiBERT: a generative model for image editing"], "answer_arxiv_id": ["2204.04428", "2205.01668", "2111.15264"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_13701"} +{"question": "What study proposed training a detector with CLIP text encoder as language embedding with additional knowledge distillation, named ViLD?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation"], "answer_arxiv_id": ["2104.13921"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_13702"} +{"question": "What papers contributed to understanding SSL from the perspective of learning theory?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Understanding Contrastive Learning Requires Incorporating Inductive Biases", "Understanding Negative Samples in Instance Discriminative Self-supervised Representation Learning", "Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data"], "answer_arxiv_id": ["1902.09229", "2202.14037", "2102.06866", "2010.03622"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_13703"} +{"question": "Which papers tried to reduce the computational cost of SAM by computing the perturbations on a subset of the parameters?", "answer": ["Efficient Sharpness-aware Minimization for Improved Training of Neural Networks"], "answer_arxiv_id": ["2110.03141"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_13704"} +{"question": "Can you indicate some papers that focus on Mixup for multi-modal learning?", "answer": ["STEMM: Self-learning with Speech-text Manifold Mixup for Speech Translation", "MixSpeech: Cross-Modality Self-Learning with Audio-Visual Stream Mixup for Visual Speech Translation and Recognition"], "answer_arxiv_id": ["2203.10426", "2303.05309"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_13705"} +{"question": "Which studies use traditional supervised approaches with rejection option in the detector?", "answer": ["PAC-Wrap: Semi-Supervised PAC Anomaly Detection"], "answer_arxiv_id": ["2205.10798"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_13706"} +{"question": "Which studies have used images generated by GLIDE for training to improve the performance on the corresponding image classification tasks?", "answer": ["Is synthetic data from generative models ready for image recognition?"], "answer_arxiv_id": ["2210.07574"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_13707"} +{"question": "Could you provide me some works that use atrous convolutional layers for enlarging the receptive field in image semantic segmentation?", "answer": ["Multi-Scale Context Aggregation by Dilated Convolutions", "DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs", "Rethinking Atrous Convolution for Semantic Image Segmentation", "Large Kernel Matters —— Improve Semantic Segmentation by Global Convolutional Network"], "answer_arxiv_id": ["1511.07122", "1606.00915", "1706.05587", "1703.02719"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13708"} +{"question": "Which studies are linked to unsupervised meta-learning by constructing synthetic tasks and extracting the meaningful information from unlabeled data?", "answer": ["Unsupervised Learning via Meta-Learning", "Unsupervised Meta-Learning for Few-Shot Image Classification", "Unsupervised Meta-Learning through Latent-Space Interpolation in Generative Models"], "answer_arxiv_id": ["1810.02334", "1811.11819", "2006.10236"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_13709"} +{"question": "Could you give me example of a study that introduces methods for fine-tuning a pre-trained diffusion model?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2106.09685", "2208.12242", "2208.01618"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_13710"} +{"question": "Coule you provide me some studies about masked autoencoder for videos?", "answer": ["Masked Autoencoders As Spatiotemporal Learners", "VideoMAE: Masked Autoencoders are Data-Efficient Learners for\n Self-Supervised Video Pre-Training", "OmniMAE: Single Model Masked Pretraining on Images and Videos", "MAR: Masked Autoencoders for Efficient Action Recognition", "VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking", "AdaMAE: Adaptive Masking for Efficient Spatiotemporal Learning with\n Masked Autoencoders", "MGMAE: Motion Guided Masking for Video Masked Autoencoding", "Keeping Your Eye on the Ball: Trajectory Attention in Video Transformers", "Masked Motion Encoding for Self-Supervised Video Representation Learning", "SemMAE: Semantic-Guided Masking for Learning Masked Autoencoders", "MaskViT: Masked Visual Pre-Training for Video Prediction"], "answer_arxiv_id": ["2205.09113", "2203.12602", "2206.08356", "2207.11660", "2303.16727v2", "2211.09120", "2308.10794", "2106.05392", "2210.06096", "2206.10207", "2206.11894"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_13711"} +{"question": "Which study first introduced the 'double descent' performance curve in overparametrized models?", "answer": ["Reconciling modern machine learning practice and the bias-variance trade-off"], "answer_arxiv_id": ["1812.11118"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13712"} +{"question": "What papers discuss using the steady-state property of stationary Markov processes in Off-Policy Policy Evaluation?", "answer": ["Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation", "DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections"], "answer_arxiv_id": ["1810.12429", "1906.04733"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_13713"} +{"question": "What studies have introduced semi-synthetic sketch datasets to address the lack of data for sketch segmentation?", "answer": ["SketchyScene: Richly-Annotated Scene Sketches", "SketchyCOCO: Image Generation from Freehand Scene Sketches", "COCO-Stuff: Thing and Stuff Classes in Context", "DeepLab: Semantic Image Segmentation with Deep Convolutional Nets,\n Atrous Convolution, and Fully Connected CRFs", "A Neural Representation of Sketch Drawings"], "answer_arxiv_id": ["1808.02473", "2003.02683", "1612.03716", "1606.00915", "1704.03477"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_13714"} +{"question": "What studies illustrate the application of sentence embeddings for semantic search of relevant document from a large vector database?", "answer": ["MS MARCO: A Human Generated MAchine Reading COmprehension Dataset", "BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information\n Retrieval Models"], "answer_arxiv_id": ["1611.09268", "2104.08663"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_13715"} +{"question": "Which works motivated the usage of PSC in DNNs?", "answer": ["Deep Residual Learning for Image Recognition", "Distributed Representations of Sentences and Documents"], "answer_arxiv_id": ["1512.03385", "1405.4053"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_13716"} +{"question": "What research has been done on appearance-based gaze estimation?", "answer": ["Appearance-based Gaze Estimation With Deep Learning: A Review and\n Benchmark"], "answer_arxiv_id": ["2104.12668"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_13717"} +{"question": "Which study combined SimCLR and CLIP for training vision encoder using image-language pairs, similar to the approach of this research?", "answer": ["SLIP: Self-supervision meets Language-Image Pre-training"], "answer_arxiv_id": ["2112.12750"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_13718"} +{"question": "Which work(s) have mentioned about the concept of underlying geometry of NeRF as an overfit in the context of novel view synthesis?", "answer": ["Volume Rendering of Neural Implicit Surfaces", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2106.12052", "2003.08934"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_13719"} +{"question": "Which papers discuss the contrastive learning objective in self-supervised learning?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1807.03748", "2002.05709", "1911.05722"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_13720"} +{"question": "Which works provide a theoretical study about SSL through the lens of statistical learning theory?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data", "Understanding Negative Samples in Instance Discriminative Self-supervised Representation Learning", "Investigating the Role of Negatives in Contrastive Representation Learning", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss"], "answer_arxiv_id": ["1902.09229v1", "2010.03622", "2102.06866", "2106.09943", "2106.04156"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_13721"} +{"question": "Which research uses an adapted Rotary Embedding in large language models?", "answer": ["RoFormer: Enhanced Transformer with Rotary Position Embedding"], "answer_arxiv_id": ["2104.09864"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_13722"} +{"question": "Which works attempted to attribute causality to specific components of neural networks’ internal representations?", "answer": ["Null It Out: Guarding Protected Attributes by Iterative Nullspace Projection", "Identifying and Controlling Important Neurons in Neural Machine Translation", "Causal Distillation for Language Models", "Amnesic Probing: Behavioral Explanation with Amnesic Counterfactuals", "Low-Complexity Probing via Finding Subnetworks"], "answer_arxiv_id": ["2004.07667", "1811.01157", "2112.02505", "2006.00995", "2104.03514"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_13723"} +{"question": "Which papers formally justify a sampling-importance-resampling proposal for parameters in a transdimensional move as an involutive MCMC step?", "answer": ["Involutive MCMC: a Unifying Framework"], "answer_arxiv_id": ["2006.16653"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_13724"} +{"question": "Which works used unsupervised concept discovery in concept-based interpretability?", "answer": ["Towards Automatic Concept-based Explanations", "CRAFT: Concept Recursive Activation FacTorization for Explainability"], "answer_arxiv_id": ["1902.03129", "2211.10154"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_13725"} +{"question": "Which studies involve improvements in SDF reconstruction for point clouds by using the Eikonal equation?", "answer": ["Implicit Neural Representations with Periodic Activation Functions"], "answer_arxiv_id": ["2006.09661"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_13726"} +{"question": "Which studies demonstrated replicative counter-examples where online RvS can diverge in stochastic environments?", "answer": ["Planning from Pixels using Inverse Dynamics Models", "Upside-Down Reinforcement Learning Can Diverge in Stochastic Environments With Episodic Resets"], "answer_arxiv_id": ["2012.02419", "2205.06595"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_13727"} +{"question": "What research presents continuous relaxation techniques for spanning trees?", "answer": ["Leveraging Recursive Gumbel-Max Trick for Approximate Inference in Combinatorial Spaces"], "answer_arxiv_id": ["2110.15072"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_13728"} +{"question": "Which studies evaluated methods on predicting outcomes of specific physical events?", "answer": ["Physion: Evaluating Physical Prediction from Vision in Humans and Machines"], "answer_arxiv_id": ["2106.08261"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_13729"} +{"question": "Which studies demonstrate the use of Majorization-Minimization technique in learning-to-rank (LTR)?", "answer": ["StochasticRank: Global Optimization of Scale-Free Discrete Functions"], "answer_arxiv_id": ["2003.02122"], "source_meta": {"published_time": "20220404"}, "qid": "AutoScholarQuery_train_13730"} +{"question": "Which research papers employed BERT, LSTM, GRU, or RNN as context encoders for dialogue state tracking in dialogues?", "answer": ["Scalable Multi-Domain Dialogue State Tracking", "An End-to-End Dialogue State Tracking System with Machine Reading\n Comprehension and Wide & Deep Classification", "SUMBT: Slot-Utterance Matching for Universal and Scalable Belief\n Tracking", "BERT-DST: Scalable End-to-End Dialogue State Tracking with Bidirectional\n Encoder Representations from Transformer", "From Machine Reading Comprehension to Dialogue State Tracking: Bridging\n the Gap", "TripPy: A Triple Copy Strategy for Value Independent Neural Dialog State\n Tracking", "Domain-Aware Dialogue State Tracker for Multi-Domain Dialogue Systems"], "answer_arxiv_id": ["1712.10224", "1912.09297", "1907.07421", "1907.03040", "2004.05827", "2005.02877", "2001.07526"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_13731"} +{"question": "Where was activation patching used as an intervention-based technique for LMs investigations?", "answer": ["Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2202.05262"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_13732"} +{"question": "Can you point out the work that introduced Top-k Sampling in language generation models?", "answer": ["Hierarchical Neural Story Generation"], "answer_arxiv_id": ["1805.04833"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_13733"} +{"question": "Which studies utilize auto-encoding frameworks in unsupervised object-centric learning?", "answer": ["Multi-Object Representation Learning with Iterative Variational\n Inference", "MONet: Unsupervised Scene Decomposition and Representation", "GENESIS: Generative Scene Inference and Sampling with Object-Centric\n Latent Representations", "SPACE: Unsupervised Object-Oriented Scene Representation via Spatial\n Attention and Decomposition", "Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "Object-Centric Learning with Slot Attention", "Illiterate DALL-E Learns to Compose", "Complex-Valued Autoencoders for Object Discovery", "Rotating Features for Object Discovery"], "answer_arxiv_id": ["1903.00450", "1901.11390", "1907.13052", "2001.02407", "1603.08575", "2006.15055", "2110.11405", "2204.02075", "2306.00600"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_13734"} +{"question": "Can you identify the studies that aim to improve layout encoding and network architecture in segmentation-based works?", "answer": ["DuLa-Net: A Dual-Projection Network for Estimating Room Layouts from a\n Single RGB Panorama"], "answer_arxiv_id": ["1811.11977"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13735"} +{"question": "What works adapt the representation of instances for new classes in the context of few-shot class-incremental learning?", "answer": ["Few-Shot Incremental Learning with Continually Evolved Classifiers", "Self-Promoted Prototype Refinement for Few-Shot Class-Incremental Learning", "Subspace Regularizers for Few-Shot Class Incremental Learning"], "answer_arxiv_id": ["2104.03047", "2107.08918", "2110.07059"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13736"} +{"question": "Which work was the first convolutional neural network used to directly predict optical flow from image pairs?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks"], "answer_arxiv_id": ["1504.06852", "1612.01925"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_13737"} +{"question": "Which work demonstrated that the bottleneck of the U-Net can be used as a semantic latent space?", "answer": ["Diffusion Models already have a Semantic Latent Space"], "answer_arxiv_id": ["2210.10960v2"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_13738"} +{"question": "Can you provide any literature that studies watermarking as a type of attack where feature vectors are perturbed?", "answer": ["Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networks"], "answer_arxiv_id": ["1804.00792"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_13739"} +{"question": "Could you provide me some works about action recognition with the help of video representation learning", "answer": ["Omnivore: A Single Model for Many Visual Modalities", "MViTv2: Improved Multiscale Vision Transformers for Classification and\n Detection", "UniFormer: Unifying Convolution and Self-attention for Visual\n Recognition", "MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient\n Long-Term Video Recognition", "SlowFast Networks for Video Recognition"], "answer_arxiv_id": ["2201.08377", "2112.01526", "2201.09450", "2201.08383", "1812.03982"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_13740"} +{"question": "Are there any papers that focus on loss supervision for enhancing monocular depth estimation?", "answer": ["Single Image Depth Prediction Made Better: A Multivariate Gaussian Take"], "answer_arxiv_id": ["2303.18164"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_13741"} +{"question": "Can you provide works that employ the embedding-based multiple instance learning approach for WSI analysis?", "answer": ["Attention-based Deep Multiple Instance Learning", "Data Efficient and Weakly Supervised Computational Pathology on Whole\n Slide Images", "Dual-stream Multiple Instance Learning Network for Whole Slide Image\n Classification with Self-supervised Contrastive Learning"], "answer_arxiv_id": ["1802.04712", "2004.09666", "2011.08939"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_13742"} +{"question": "Which paper discussed a general procedure for converting a non-compressed mechanism to a compressed one that is computationally differentially private?", "answer": ["Lossless Compression of Efficient Private Local Randomizers"], "answer_arxiv_id": ["2102.12099"], "source_meta": {"published_time": "20221108"}, "qid": "AutoScholarQuery_train_13743"} +{"question": "What papers propose modifications in the internal self-attention mechanism of Transformers for efficiency?", "answer": ["Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks", "Coordination Among Neural Modules Through a Shared Global Workspace", "Perceiver: General Perception with Iterative Attention", "Perceiver IO: A General Architecture for Structured Inputs & Outputs"], "answer_arxiv_id": ["1810.00825", "2103.01197", "2103.03206", "2107.14795"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_13744"} +{"question": "Which papers presented the methods to obtain an ensemble prediction by a single forward pass?", "answer": ["Depth Uncertainty in Neural Networks", "Training independent subnetworks for robust prediction"], "answer_arxiv_id": ["2006.08437", "2010.06610"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_13745"} +{"question": "What works deal with scalable link prediction with distillation, decoder, and sketching designs?", "answer": ["Linkless Link Prediction via Relational Distillation", "Flashlight : Scalable Link Prediction with Effective Decoders", "Graph Neural Networks for Link Prediction with Subgraph Sketching"], "answer_arxiv_id": ["2210.05801", "2209.10100", "2209.15486"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_13746"} +{"question": "Could you list the studies that assesses the mathematical reasoning abilities of external calculators and Python interpreters using computation-intensive QA datasets?", "answer": ["Training Verifiers to Solve Math Word Problems", "Dynamic Prompt Learning via Policy Gradient for Semi-structured Mathematical Reasoning"], "answer_arxiv_id": ["2110.14168", "2209.14610"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13747"} +{"question": "Which papers focus on creating interpretable prediction processes using a concept-based explanation approach?", "answer": ["Concept Bottleneck Models", "This Looks Like That: Deep Learning for Interpretable Image Recognition"], "answer_arxiv_id": ["2007.04612", "1806.10574"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_13748"} +{"question": "Which work came up with a new training method by formulating backward-compatible regularization as a contrastive loss?", "answer": ["Asymmetric metric learning for knowledge transfer"], "answer_arxiv_id": ["2006.16331"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_13749"} +{"question": "Which paper first introduced a recurrent method for non-rigid scene flow estimation?", "answer": ["FlowStep3D: Model Unrolling for Self-Supervised Scene Flow Estimation"], "answer_arxiv_id": ["2011.10147"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_13750"} +{"question": "What work established a link between specific dataset distillation methods and optimizing certain divergence measures associated with Bayesian pseudocoresets?", "answer": ["On Divergence Measures for Bayesian Pseudocoresets"], "answer_arxiv_id": ["2210.06205"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_13751"} +{"question": "Which papers presented Normalizing Flows for efficient and accurate variational inference?", "answer": ["Variational Inference with Normalizing Flows"], "answer_arxiv_id": ["1505.05770"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_13752"} +{"question": "Which studies discuss the creation of printable adversarial patches and invisibility cloaks for fooling person detectors?", "answer": ["Fooling automated surveillance cameras: adversarial patches to attack\n person detection", "On Physical Adversarial Patches for Object Detection", "Making an Invisibility Cloak: Real World Adversarial Attacks on Object\n Detectors"], "answer_arxiv_id": ["1904.08653", "1906.11897", "1910.14667"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_13753"} +{"question": "Which work mentions that offline algorithms lack guaranteed performance improvement over the behavior policy?", "answer": ["Offline Reinforcement Learning with Soft Behavior Regularization"], "answer_arxiv_id": ["2110.07395"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_13754"} +{"question": "Can you tell me the research that discussed that the optimization of the entropic approaches being unstable for small values?", "answer": ["Large-Scale Optimal Transport and Mapping Estimation"], "answer_arxiv_id": ["1711.02283"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_13755"} +{"question": "Could you list some researches that focus on distilling knowledge through feature-level consistency or contrastive learning in 3D understanding methodologies?", "answer": ["CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP", "Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Data"], "answer_arxiv_id": ["2301.04926", "2203.16258"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_13756"} +{"question": "What works present methods of graph-based bot detection?", "answer": ["BotRGCN: Twitter Bot Detection with Relational Graph Convolutional\n Networks", "Heterogeneity-aware Twitter Bot Detection with Relational Graph\n Transformers", "BotMoE: Twitter Bot Detection with Community-Aware Mixtures of Modal-Specific Experts", "BotPercent: Estimating Bot Populations in Twitter Communities"], "answer_arxiv_id": ["2106.13092", "2109.02927", "2304.06280v2", "2302.00381"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_13757"} +{"question": "Are there any studies that apply iterative inference to static learning problems such as image analysis and pointcloud segmentation?", "answer": ["Rich feature hierarchies for accurate object detection and semantic segmentation", "Fast R-CNN", "CNN-RNN: A Unified Framework for Multi-label Image Classification", "Mask R-CNN", "3DCNN-DQN-RNN: A Deep Reinforcement Learning Framework for Semantic Parsing of Large-scale 3D Point Clouds"], "answer_arxiv_id": ["1311.2524", "1504.08083", "1604.04573v1", "1703.06870", "1707.06783"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13758"} +{"question": "Could you provide me some research in the direction of nonstationary RL outside policy optimization and value-based methods", "answer": ["Non-stationary Reinforcement Learning without Prior Knowledge: An Optimal Black-box Approach"], "answer_arxiv_id": ["2102.05406"], "source_meta": {"published_time": "20230810"}, "qid": "AutoScholarQuery_train_13759"} +{"question": "Which papers are the earlier studies about group fairness in machine learning?", "answer": ["Equality of Opportunity in Supervised Learning", "Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatment"], "answer_arxiv_id": ["1610.02413", "1610.08452"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_13760"} +{"question": "What works implemented different backbones such as biRNN and Transformer encoder architecture in body motion prediction?", "answer": ["Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time"], "answer_arxiv_id": ["1810.04703"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_13761"} +{"question": "Which studies utilize evolutionary algorithms for biological sequence design?", "answer": ["AdaLead: A simple and robust adaptive greedy search algorithm for sequence design"], "answer_arxiv_id": ["2010.02141"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_13762"} +{"question": "Can you point me to the literature where zero-shot diffusion-based methods for image restoration were proposed?", "answer": ["Denoising Diffusion Restoration Models", "Improving Diffusion Models for Inverse Problems using Manifold\n Constraints", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "Diffusion Posterior Sampling for General Noisy Inverse Problems", "Generative Diffusion Prior for Unified Image Restoration and Enhancement", "Denoising Diffusion Models for Plug-and-Play Image Restoration"], "answer_arxiv_id": ["2201.11793", "2206.00941", "2212.00490", "2209.14687", "2304.01247", "2305.08995"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_13763"} +{"question": "Which studies use Google Football as a complex and stochastic benchmark?", "answer": ["Google Research Football: A Novel Reinforcement Learning Environment"], "answer_arxiv_id": ["1907.11180"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_13764"} +{"question": "What studies utilized variance-weighted regression for tight analyses in offline RL?", "answer": ["Near-optimal Offline Reinforcement Learning with Linear Representation: Leveraging Variance Information with Pessimism", "Nearly Minimax Optimal Offline Reinforcement Learning with Linear Function Approximation: Single-Agent MDP and Markov Game"], "answer_arxiv_id": ["2203.05804", "2205.15512"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_13765"} +{"question": "Which works proposed the use of contrastive objectives in self-supervised learning for visual representation training?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "An Empirical Study of Training Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2002.05709", "2104.02057"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_13766"} +{"question": "Are there studies for video semantic segmentation that employ recurrent units or design an attention propagation module to improve prediction accuracy and consistency?", "answer": ["Semantic Video Segmentation by Gated Recurrent Flow Propagation", "Temporally Distributed Networks for Fast Video Semantic Segmentation"], "answer_arxiv_id": ["1612.08871", "2004.01800"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13767"} +{"question": "What are some of the foundational works on generative adversarial networks (GANs) used for conditional, controllable generation?", "answer": ["Generative Adversarial Networks", "Generative Adversarial Text to Image Synthesis", "RetrieveGAN: Image Synthesis via Differentiable Patch Retrieval", "Semantics Disentangling for Text-to-Image Generation"], "answer_arxiv_id": ["1406.2661", "1605.05396", "2007.08513", "1904.01480"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_13768"} +{"question": "Which papers discuss the use of simple geometric primitives to approximate a 3D shape?", "answer": ["Learning Shape Abstractions by Assembling Volumetric Primitives", "3D-PRNN: Generating Shape Primitives with Recurrent Neural Networks", "Im2Struct: Recovering 3D Shape Structure from a Single RGB Image", "Learning Shape Templates with Structured Implicit Functions"], "answer_arxiv_id": ["1612.00404", "1708.01648", "1804.05469", "1904.06447"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_13769"} +{"question": "What works tackle the contextual linear bandit setting where the arm set changes over time?", "answer": ["Robust Stochastic Linear Contextual Bandits Under Adversarial Attacks", "Nearly Optimal Algorithms for Linear Contextual Bandits with Adversarial Corruptions"], "answer_arxiv_id": ["2106.02978v3", "2205.06811"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_13770"} +{"question": "Which researches proposed Gaussian mixture model frameworks to study the self-attention in transformers?", "answer": ["Probabilistic Attention for Interactive Segmentation", "Modeling Concentrated Cross-Attention for Neural Machine Translation with Gaussian Mixture Model"], "answer_arxiv_id": ["2106.15338", "2109.05244"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_13771"} +{"question": "Could you provide works where the researchers achieved efficient results by combining BBSD-Softmax with a Kolmogorov-Smirnov (KS) statistical test and using the Bonferroni correction?", "answer": ["Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift"], "answer_arxiv_id": ["1810.11953"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_13772"} +{"question": "Which works are focused on enhancing the text-to-image alignment in text-to-image generation?", "answer": ["Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image\n Diffusion Models", "Training-Free Structured Diffusion Guidance for Compositional\n Text-to-Image Synthesis", "Grounded Text-to-Image Synthesis with Attention Refocusing", "Compositional Visual Generation with Composable Diffusion Models", "Aligning Text-to-Image Models using Human Feedback", "Training Diffusion Models with Reinforcement Learning"], "answer_arxiv_id": ["2301.13826", "2212.05032", "2306.05427v2", "2206.01714", "2302.12192", "2305.13301"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_13773"} +{"question": "Are there any known works on speech enhancement in speech generative models?", "answer": ["Real Time Speech Enhancement in the Waveform Domain", "Universal Speech Enhancement with Score-based Diffusion"], "answer_arxiv_id": ["2006.12847", "2206.03065"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13774"} +{"question": "What are some works that considered imperfect demonstrations when exploiting the teacher policy?", "answer": ["Confidence-Aware Imitation Learning from Demonstrations with Varying Optimality", "Discriminator-Weighted Offline Imitation Learning from Suboptimal Demonstrations"], "answer_arxiv_id": ["2110.14754", "2207.10050"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_13775"} +{"question": "Which papers mention the contribution of the open-source community in releasing large language models like LLaMA and CodeLLaMA?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models", "Code Llama: Open Foundation Models for Code"], "answer_arxiv_id": ["2307.09288", "2308.12950"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_13776"} +{"question": "What studies helped in building the foundation for proof techniques in VR-IWAE bound methodology?", "answer": ["Doubly Reparameterized Gradient Estimators for Monte Carlo Objectives", "Importance Weighting and Variational Inference"], "answer_arxiv_id": ["1810.04152", "1808.09034"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_13777"} +{"question": "Any works about imperceptible poisoning which modifies examples slightly and does not damage their semantic information?", "answer": ["Unlearnable Examples: Making Personal Data Unexploitable", "Adversarial Examples Make Strong Poisons", "MetaPoison: Practical General-purpose Clean-label Data Poisoning", "Witches’ Brew: Industrial Scale Data Poisoning via Gradient Matching", "Self-Ensemble Protection: Training Checkpoints Are Good Data Protectors"], "answer_arxiv_id": ["2101.04898", "2106.10807", "2004.00225", "2009.02276", "2211.12005"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_13778"} +{"question": "Which studies have generated pseudo labels for self-training in the context of unsupervised domain adaptation?", "answer": ["Decoupled Adaptation for Cross-Domain Object Detection"], "answer_arxiv_id": ["2110.02578"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_13779"} +{"question": "Which work is most comparable to current research?", "answer": ["Continuous diffusion for categorical data"], "answer_arxiv_id": ["2211.15089"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13780"} +{"question": "Can you provide examples of research that deals with editing techniques in generative models?", "answer": ["Closed-Form Factorization of Latent Semantics in GANs", "Unsupervised Discovery of Interpretable Directions in the GAN Latent Space", "StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation", "InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs", "In-Domain GAN Inversion for Real Image Editing", "StyleGAN of All Trades: Image Manipulation with Only Pretrained StyleGAN", "Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation", "JoJoGAN: One Shot Face Stylization", "Make It So: Steering StyleGAN for Any Image Inversion and Editing", "StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated Images using Conditional Continuous Normalizing Flows", "GAN-Control: Explicitly Controllable GANs"], "answer_arxiv_id": ["2007.06600v4", "2002.03754", "2011.12799", "2005.09635", "2004.00049", "2111.01619", "2008.00951", "2112.11641", "2304.14403", "2008.02401", "2101.02477"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_13781"} +{"question": "Which work introduced Ladder VAE (LVAE) with a top-down inference path?", "answer": ["Ladder Variational Autoencoders"], "answer_arxiv_id": ["1602.02282v3"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_13782"} +{"question": "Are there any studies that demonstrated the implicit bias in two-layer leaky ReLU networks trained on linearly separable and symmetric data?", "answer": ["Gradient Descent on Two-layer Nets: Margin Maximization and Simplicity Bias"], "answer_arxiv_id": ["2110.13905"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13783"} +{"question": "What studies observed that local minima found by SGD from different initializations can be connected via a piece-wise linear path of low loss?", "answer": ["Essentially No Barriers in Neural Network Energy Landscape", "Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs"], "answer_arxiv_id": ["1803.00885", "1802.10026"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_13784"} +{"question": "What studies proposed agents to abstain from answering in cases of ambiguity?", "answer": ["Rich Knowledge Sources Bring Complex Knowledge Conflicts: Recalibrating\n Models to Reflect Conflicting Evidence"], "answer_arxiv_id": ["2210.13701"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_13785"} +{"question": "What was the approach used to calibrate model predictions in order to improve the ability of the models for in-context learning?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models"], "answer_arxiv_id": ["2102.09690"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_13786"} +{"question": "Can you provide examples of studies that utilize exemplar-based local explanations?", "answer": ["Efficient Data Representation by Selecting Prototypes with Importance Weights"], "answer_arxiv_id": ["1707.01212"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_13787"} +{"question": "What publications have researched biases in VLMs used for text-to-image generation?", "answer": ["DALL-Eval: Probing the Reasoning Skills and Social Biases of\n Text-to-Image Generation Models", "T2IAT: Measuring Valence and Stereotypical Biases in Text-to-Image\n Generation", "Social Biases through the Text-to-Image Generation Lens"], "answer_arxiv_id": ["2202.04053", "2306.00905", "2304.06034"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13788"} +{"question": "What studies have explored the behavior of maximum-margin classifiers, SGD training dynamics and inductive biases of neural network models in the presence of spurious features?", "answer": ["Understanding the failure modes of out-of-distribution generalization", "Gradient Starvation: A Learning Proclivity in Neural Networks", "On the Spectral Bias of Neural Networks"], "answer_arxiv_id": ["2010.15775", "2011.09468", "1806.08734"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_13789"} +{"question": "Could you cite some studies that attempted to characterize the 'double descent' phenomenon in various models?", "answer": ["Surprises in High-Dimensional Ridgeless Least Squares Interpolation", "Asymptotics of Ridge (less) Regression under General Source Condition"], "answer_arxiv_id": ["1903.08560", "2006.06386"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13790"} +{"question": "Which works fine-tuned language models with crawled conversational data for open-domain dialogues?", "answer": ["DialoGPT: Large-Scale Generative Pre-training for Conversational\n Response Generation", "Recipes for building an open-domain chatbot"], "answer_arxiv_id": ["1911.00536", "2004.13637"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_13791"} +{"question": "Which paper derived a Lévy-driven stochastic differential equation for modeling non-gaussianity of noise?", "answer": ["Towards Theoretically Understanding Why Sgd Generalizes Better Than Adam in Deep Learning"], "answer_arxiv_id": ["2010.05627"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_13792"} +{"question": "Are there any research papers that explore how modifying the given data can boost the training speed?", "answer": ["Partition Speeds Up Learning Implicit Neural Representations Based on\n Exponential-Increase Hypothesis"], "answer_arxiv_id": ["2310.14184"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_13793"} +{"question": "In what work can we find evidence about the global and exponential instability of the closed-loop control based GAN models?", "answer": ["Training Generative Adversarial Networks with Limited Data"], "answer_arxiv_id": ["2006.06676"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_13794"} +{"question": "What studies approached regret minimization as a saddle point problem in the structured bandit setting?", "answer": ["Exploration by Optimisation in Partial Monitoring", "Mirror Descent and the Information Ratio", "Asymptotically Optimal Information-Directed Sampling"], "answer_arxiv_id": ["1907.05772", "2009.12228", "2011.05944"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_13795"} +{"question": "Which works have explored the causes of spurious correlations in Neural Networks (NNs)?", "answer": ["An investigation of why overparameterization exacerbates spurious correlations"], "answer_arxiv_id": ["2005.04345"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_13796"} +{"question": "What is the model that introduces a multi-head attention mechanism to the graph version of the supervised classification tasks?", "answer": ["Masked Label Prediction: Unified Message Passing Model for Semi-Supervised Classification"], "answer_arxiv_id": ["2009.03509"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_13797"} +{"question": "Which work extends the basic idea of K-FAC to convolutional layers?", "answer": ["A Kronecker-factored approximate Fisher matrix for convolution layers"], "answer_arxiv_id": ["1602.01407"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_13798"} +{"question": "What work proposed a MIM method that works with pairs of scene images based on the cross-view completion task (CroCo)?", "answer": ["CroCo: Self-Supervised Pre-training for 3D Vision Tasks by Cross-View\n Completion"], "answer_arxiv_id": ["2210.10716"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_13799"} +{"question": "Which work developed a strategy to utilize strong and weak labels simultaneously for better HOI detection?", "answer": ["Detecting Human-Object Interaction with Mixed Supervision"], "answer_arxiv_id": ["2011.04971"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_13800"} +{"question": "What works addressed distributional robustness for group fairness?", "answer": ["Fairness without Demographics through Adversarially Reweighted Learning"], "answer_arxiv_id": ["2006.13114"], "source_meta": {"published_time": "20221202"}, "qid": "AutoScholarQuery_train_13801"} +{"question": "Could you provide me some studies about the cross-attention operation based on the attention mechanism for object tracking?", "answer": ["AiATrack: Attention in Attention for Transformer Visual Tracking", "Transformer Tracking with Cyclic Shifting Window Attention", "MixFormer: End-to-End Tracking with Iterative Mixed Attention", "STMTrack: Template-free Visual Tracking with Space-time Memory Networks", "Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual Tracking", "Joint Feature Learning and Relation Modeling for Tracking: A One-Stream Framework", "Transformer Tracking", "Transforming Model Prediction for Tracking", "Correlation-Aware Deep Tracking", "Learning Spatio-Temporal Transformer for Visual Tracking"], "answer_arxiv_id": ["2207.09603", "2205.03806", "2203.11082", "2104.00324", "2103.11681", "2203.11991", "2103.15436", "2203.11192", "2203.01666", "2103.17154"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_13802"} +{"question": "Any works about learning general foundation models for control in RL?", "answer": ["A Generalist Agent", "Foundation Models for Decision Making: Problems, Methods, and Opportunities"], "answer_arxiv_id": ["2205.06175", "2303.04129"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_13803"} +{"question": "Could you provide me some papers that proposed alternate definitions and extensions of the standard notions of multicalibration?", "answer": ["Omnipredictors", "Low-Degree Multicalibration", "HappyMap: A Generalized Multicalibration Method"], "answer_arxiv_id": ["2109.05389", "2203.01255", "2303.04379"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_13804"} +{"question": "Whose work created the evaluation dataset 'TrustfulQA' and what is it for?", "answer": ["TruthfulQA: Measuring How Models Mimic Human Falsehoods"], "answer_arxiv_id": ["2109.07958"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_13805"} +{"question": "Which papers studied RL with general function approximation, especially under the low Bellman-rank assumption?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "On Oracle-Efficient PAC RL with Rich Observations", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches", "Provably efficient RL with Rich Observations via Latent State Decoding", "Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms"], "answer_arxiv_id": ["1610.09512v2", "1803.00606", "1811.08540", "1901.09018", "2102.00815"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_13806"} +{"question": "What work has described the use of unsupervised energy-based concepts?", "answer": ["Unsupervised Learning of Compositional Energy Concepts"], "answer_arxiv_id": ["2111.03042"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_13807"} +{"question": "Could you provide me studies that employed imitation learning to teach an LLM to use search engines, for example, WebGPT?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback"], "answer_arxiv_id": ["2112.09332"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_13808"} +{"question": "What works learn 3D occupancy in a self-supervised manner in monocular scenarios?", "answer": ["Behind the Scenes: Density Fields for Single View Reconstruction", "SceneRF: Self-Supervised Monocular 3D Scene Reconstruction with Radiance\n Fields"], "answer_arxiv_id": ["2301.07668", "2212.02501"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_13809"} +{"question": "What are the researches that explore the connection between OT and information theory?", "answer": ["Information Constrained Optimal Transport: From Talagrand, to Marton, to Cover", "When Optimal Transport Meets Information Geometry"], "answer_arxiv_id": ["2008.10249", "2206.14791"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_13810"} +{"question": "Could you provide me some works about off-the-shelf instruction-tuning datasets?", "answer": ["The Flan Collection: Designing Data and Methods for Effective\n Instruction Tuning"], "answer_arxiv_id": ["2301.13688"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_13811"} +{"question": "What papers train a transformer decoder to predict occupancy and color from RGB-D inputs?", "answer": ["Multiview Compressive Coding for 3D Reconstruction"], "answer_arxiv_id": ["2301.08247"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_13812"} +{"question": "Could you give me studies that have made recent progress in improving the sampling speed of CNFs/diffusion models?", "answer": ["Improving and generalizing flow-based generative models with minibatch optimal transport", "Consistency Models"], "answer_arxiv_id": ["2302.00482", "2303.01469"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_13813"} +{"question": "What papers proposed different ideas on how to learn causal features from data?", "answer": ["A Unified View of Causal and Non-causal Feature Selection"], "answer_arxiv_id": ["1802.05844"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_13814"} +{"question": "Could you give examples of research that proposed metrics constructed on top of pretrained neural models to score translation outputs?", "answer": ["BERTScore: Evaluating Text Generation with BERT", "Automatic Machine Translation Evaluation in Many Languages via Zero-Shot\n Paraphrasing", "COMET: A Neural Framework for MT Evaluation", "BLEURT: Learning Robust Metrics for Text Generation"], "answer_arxiv_id": ["1904.09675", "2004.14564", "2009.09025", "2004.04696"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_13815"} +{"question": "What studies have specifically addressed the representation of articulated shapes like human bodies or hands?", "answer": ["TAVA: Template-free Animatable Volumetric Actors", "Neural Actor: Neural Free-view Synthesis of Human Actors with Pose\n Control", "Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "Animatable Neural Radiance Fields for Modeling Dynamic Human Bodies", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "LISA: Learning Implicit Shape and Appearance of Hands", "NIMBLE: A Non-rigid Hand Model with Bones and Muscles", "LiveHand: Real-time and Photorealistic Neural Hand Rendering", "HARP: Personalized Hand Reconstruction from a Monocular RGB Video"], "answer_arxiv_id": ["2206.08929", "2106.02019", "2012.15838", "2105.02872", "2201.04127", "2204.01695", "2202.04533", "2302.07672", "2212.09530"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_13816"} +{"question": "Which research papers addressed NSWC SCO problems?", "answer": ["Distributionally Robust Learning with Weakly Convex Losses: Convergence Rates and Finite-Sample Guarantees"], "answer_arxiv_id": ["2301.06619"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_13817"} +{"question": "Which work introduced residual connections in Convolutional Neural Networks (CNNs)?", "answer": ["Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1512.03385"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_13818"} +{"question": "What papers focus on using personas to elicit toxic content?", "answer": ["Toxicity in ChatGPT: Analyzing Persona-assigned Language Models", "Revealing Persona Biases in Dialogue Systems", "Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona\n Biases in Dialogue Systems"], "answer_arxiv_id": ["2304.05335", "2104.08728", "2310.05280"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_13819"} +{"question": "Which paper uses an off-the-shelf expression generator to sample expressions from a prior distribution?", "answer": ["Deep learning for symbolic mathematics"], "answer_arxiv_id": ["1912.01412"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_13820"} +{"question": "Have there been researches that leveraked reinforcement learning in molecular graphs generation?", "answer": ["Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation", "MolecularRNN: Generating realistic molecular graphs with optimized properties"], "answer_arxiv_id": ["1806.02473", "1905.13372"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_13821"} +{"question": "In what papers can I find the generalized Sinkhorn’s algorithm for solving unbalanced Optimal Transport?", "answer": ["Scaling Algorithms for Unbalanced Transport Problems"], "answer_arxiv_id": ["1607.05816"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_13822"} +{"question": "Which papers exploit the property of code execution for reranking in code generation?", "answer": ["Natural Language to Code Translation with Execution", "[2203.07814] Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2204.11454", "2203.07814"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_13823"} +{"question": "Could you provide a work that updates its rationales by potentially replacing a rationale from the model with a rationale from a surrogate?", "answer": ["STaR: Bootstrapping Reasoning With Reasoning"], "answer_arxiv_id": ["2203.14465v2"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_13824"} +{"question": "What works have contributed to Offline RL algorithms?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Pessimistic Q-Learning for Offline Reinforcement Learning: Towards Optimal Sample Complexity", "Doubly Robust Off-policy Value Evaluation for Reinforcement Learning", "Off-Policy Deep Reinforcement Learning without Exploration", "Offline Reinforcement Learning: Fundamental Barriers for Value Function Approximation"], "answer_arxiv_id": ["2006.04779", "2202.13890", "1511.03722", "1812.02900", "2111.10919v2"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_13825"} +{"question": "What work has been done on 2D open-vocabulary segmentation without fine-tuning on a closed set of classes?", "answer": ["Scaling Open-Vocabulary Image Segmentation with Image-Level Labels"], "answer_arxiv_id": ["2112.12143"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13826"} +{"question": "Could you cite the works that focused on developing new knowledge models for learning contextual commonsense generation?", "answer": ["Think Before You Speak: Explicitly Generating Implicit Commonsense\n Knowledge for Response Generation", "Knowledge Graph Generation From Text"], "answer_arxiv_id": ["2110.08501", "2211.10511"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_13827"} +{"question": "What studies showed the possibility of optimization challenges in RL?", "answer": ["A Geometric Perspective on Self-Supervised Policy Adaptation", "Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation", "Masked Visual Pre-training for Motor Control", "VRL3: A Data-Driven Framework for Visual Deep Reinforcement Learning"], "answer_arxiv_id": ["2011.07318", "2107.00644v2", "2203.06173", "2202.10324"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_13828"} +{"question": "What methods have been used for high-dimensional black-box optimization in continuous spaces?", "answer": ["Bayesian Optimization in a Billion Dimensions via Random Embeddings"], "answer_arxiv_id": ["1301.1942"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_13829"} +{"question": "What studies explore Mahalanobis distance for out-of-distribution detection with generative classifiers?", "answer": ["A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "A Multi-Head Model for Continual Learning via Out-of-Distribution Replay", "On Generalizing Beyond Domains in Cross-Domain Continual Learning"], "answer_arxiv_id": ["1807.03888", "2208.09734", "2203.03970"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_13830"} +{"question": "Are there any studies that have performed a rigorous analysis of approximating operators arising in PDEs with discontinuous solutions?", "answer": ["Error estimates for DeepONets: A deep learning framework in infinite dimensions"], "answer_arxiv_id": ["2102.09618"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_13831"} +{"question": "What papers discuss problems that arise when the testing degradation does not match the training degradation?", "answer": ["Blind Super-Resolution With Iterative Kernel Correction", "Blind Image Super-Resolution: A Survey and Beyond", "Reflash Dropout in Image Super-Resolution", "Toward Convolutional Blind Denoising of Real Photographs", "Masked Image Training for Generalizable Deep Image Denoising"], "answer_arxiv_id": ["1904.03377", "2107.03055", "2112.12089", "1807.04686", "2303.13132"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_13832"} +{"question": "What studies have examined the reproduction of training data in AI-generated art settings?", "answer": ["Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models", "Benchmarking Deepart Detection"], "answer_arxiv_id": ["2212.03860", "2302.14475"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_13833"} +{"question": "Which studies have explored the addition of auxiliary learning objectives to aid the learning of the target task?", "answer": ["Auxiliary Learning by Implicit Differentiation", "AANG: Automating Auxiliary learniNG"], "answer_arxiv_id": ["2007.02693", "2205.14082"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_13834"} +{"question": "Can you tell me about the papers that researched prediction calibration as a method to improve in-context learning's effectiveness?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models"], "answer_arxiv_id": ["2102.09690"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_13835"} +{"question": "Which studies have proposed different architectures to incorporate the task context into the policy in reinforcement learning?", "answer": ["Context Meta-Reinforcement Learning via Neuromodulation", "Contextualize Me – The Case for Context in Reinforcement Learning", "Multi-Task Reinforcement Learning with Soft Modularization", "Multi-Task Reinforcement Learning with Context-based Representations", "Combining Modular Skills in Multitask Learning", "Linear Representation Meta-Reinforcement Learning for Instant Adaptation", "Recomposing the Reinforcement Learning Building Blocks with Hypernetworks", "Hypernetworks in Meta-Reinforcement Learning", "Hypernetworks for Zero-shot Transfer in Reinforcement Learning"], "answer_arxiv_id": ["2111.00134", "2202.04500", "2003.13661", "2102.06177", "2202.13914", "2101.04750", "2106.06842", "2210.11348", "2211.15457"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_13836"} +{"question": "Which paper introduces the popular self-driving car simulator called CARLA?", "answer": ["CARLA: An Open Urban Driving Simulator"], "answer_arxiv_id": ["1711.03938"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_13837"} +{"question": "Any studies that provided rigorous guarantees for learning system identification and control?", "answer": ["Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification", "Non-asymptotic Identification of LTI Systems from a Single Trajectory", "Online Control with Adversarial Disturbances", "Naive Exploration is Optimal for Online LQR"], "answer_arxiv_id": ["1802.08334", "1806.05722", "1902.08721", "2001.09576"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_13838"} +{"question": "Which works have proposed or incorporated architectural models for semantic sketch segmentation tasks?", "answer": ["DeepLab: Semantic Image Segmentation with Deep Convolutional Nets,\n Atrous Convolution, and Fully Connected CRFs", "A Neural Representation of Sketch Drawings"], "answer_arxiv_id": ["1606.00915", "1704.03477"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_13839"} +{"question": "Which work documents the use of reading vectors and contrast vectors to extract high-level concepts in LLMs?", "answer": ["Representation Engineering: A Top-Down Approach to AI Transparency"], "answer_arxiv_id": ["2310.01405"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_13840"} +{"question": "What works have studied the topic of stochastically delayed feedback in multi-armed bandit and contextual bandit problems?", "answer": ["Distributed Delayed Stochastic Optimization", "Efficient Optimal Learning for Contextual Bandits", "Online Learning under Delayed Feedback", "Linear Bandits with Stochastic Delayed Feedback", "Stochastic Bandit Models for Delayed Conversions", "Stochastic bandits with arm-dependent delays", "Banker Online Mirror Descent: A Universal Approach for Delayed Online Bandit Learning"], "answer_arxiv_id": ["1104.5525v1", "1106.2369", "1306.0686", "1807.02089v3", "1706.09186", "2006.10459", "2301.10500"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_13841"} +{"question": "What papers improve the efficiency of representation learning using auxiliary tasks in RL?", "answer": ["Can Increasing Input Dimensionality Improve Deep Reinforcement Learning?"], "answer_arxiv_id": ["2003.01629"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_13842"} +{"question": "What recent works have leveraged Transformer-based approaches for action recognition?", "answer": ["ViViT: A Video Vision Transformer", "Is Space-Time Attention All You Need for Video Understanding?", "Object-Region Video Transformers", "Keeping Your Eye on the Ball: Trajectory Attention in Video Transformers", "MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognition", "Multiscale Vision Transformers", "Multiview Transformers for Video Recognition"], "answer_arxiv_id": ["2103.15691", "2102.05095", "2110.06915", "2106.05392", "2201.08383", "2104.11227", "2201.04288"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_13843"} +{"question": "Which papers focussed on generating video streams for the lip region in the field of talking head synthesis?", "answer": ["A Lip Sync Expert Is All You Need for Speech to Lip Generation In The\n Wild", "DINet: Deformation Inpainting Network for Realistic Face Visually\n Dubbing on High Resolution Video", "Seeing What You Said: Talking Face Generation Guided by a Lip Reading\n Expert", "Masked Lip-Sync Prediction by Audio-Visual Contextual Exploitation in\n Transformers", "StyleSync: High-Fidelity Generalized and Personalized Lip Sync in\n Style-based Generator", "Identity-Preserving Talking Face Generation with Landmark and Appearance\n Priors"], "answer_arxiv_id": ["2008.10010", "2303.03988", "2303.17480", "2212.04970", "2305.05445", "2305.08293"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_13844"} +{"question": "Which study demonstrated that with an advanced training setup, the performance of canonical ResNet-50 could be boosted?", "answer": ["ResNet strikes back: An improved training procedure in timm"], "answer_arxiv_id": ["2110.00476"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_13845"} +{"question": "What research papers investigated higher-dimensional parameterizations of 𝐒𝐎(3) in point cloud registration?", "answer": ["On the Continuity of Rotation Representations in Neural Networks", "A Smooth Representation of Belief over SO(3) for Deep Rotation Learning\n with Uncertainty"], "answer_arxiv_id": ["1812.07035", "2006.01031"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_13846"} +{"question": "Which papers have explored the use of differentiable quadratic program (QP) layer, OptNet, for safety guarantees in neural network controllers?", "answer": ["OptNet: Differentiable Optimization as a Layer in Neural Networks"], "answer_arxiv_id": ["1703.00443"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_13847"} +{"question": "Which works enforce control on the Lipschitz constant through layer orthogonalization?", "answer": ["Orthogonalizing Convolutional Layers with the Cayley Transform", "Orthogonal Convolutional Neural Networks", "Controllable Orthogonalization in Training DNNs", "Sorting Out Lipschitz Function Approximation"], "answer_arxiv_id": ["2104.07167", "1911.12207", "2004.00917", "1811.05381"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13848"} +{"question": "Which published articles contributed to the field of data-driven initialization methods?", "answer": ["GradInit: Learning to Initialize Neural Networks for Stable and Efficient Training", "Towards Theoretically Inspired Neural Initialization Optimization"], "answer_arxiv_id": ["2102.08098", "2210.05956"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_13849"} +{"question": "Could you provide me a study about the classifier-free method in a conditional diffusion model?", "answer": ["Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2207.12598"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_13850"} +{"question": "What works focus on embodied localization in contexts like estimating agent's position?", "answer": ["Where Are You? Localization from Embodied Dialog", "Transformer-based Localization from Embodied Dialog with Large-scale\n Pre-training", "Vision-and-Dialog Navigation", "The RobotSlang Benchmark: Dialog-guided Robot Localization and\n Navigation", "Text2Pos: Text-to-Point-Cloud Cross-Modal Localization"], "answer_arxiv_id": ["2011.08277", "2210.04864", "1907.04957", "2010.12639", "2203.15125"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_13851"} +{"question": "Which works extend 2D diffusion models from single-view images to multi-view images?", "answer": ["Novel View Synthesis with Diffusion Models", "NerfDiff: Single-image View Synthesis with NeRF-guided Distillation from\n 3D-aware Diffusion", "NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as\n General Image Priors", "SparseFusion: Distilling View-conditioned Diffusion for 3D\n Reconstruction", "Consistent View Synthesis with Pose-Guided Diffusion Models", "Generative Novel View Synthesis with 3D-Aware Diffusion Models", "Long-Term Photometric Consistent Novel View Synthesis with Diffusion\n Models", "Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision", "DreamSparse: Escaping from Plato's Cave with 2D Frozen Diffusion Model\n Given Sparse Views", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "MVDiffusion: Enabling Holistic Multi-view Image Generation with\n Correspondence-Aware Diffusion", "3D-aware Image Generation using 2D Diffusion Models", "Deceptive-NeRF: Enhancing NeRF Reconstruction using Pseudo-Observations\n from Diffusion Models"], "answer_arxiv_id": ["2210.04628", "2302.10109", "2212.03267", "2212.00792", "2303.17598", "2304.02602", "2304.10700", "2306.11719", "2306.03414", "2306.07881", "2307.01097", "2303.17905", "2305.15171"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_13852"} +{"question": "Which works focus on traditional FL methods that are based on update-correction in FedAvg?", "answer": ["SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Federated Learning for Face Recognition with Gradient Correction", "FedDC: Federated Learning with Non-IID Data via Local Drift Decoupling and Correction"], "answer_arxiv_id": ["1910.06378", "2112.07246", "2203.11751"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_13853"} +{"question": "Which researchers proposed an out-of-distribution graph learning benchmark with molecular tasks?", "answer": ["GOOD: A Graph Out-of-Distribution Benchmark"], "answer_arxiv_id": ["2206.08452"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_13854"} +{"question": "Which works performed local K-means clustering to efficiently generate superpixels?", "answer": ["SLIC: Self-Supervised Learning with Iterative Clustering for Human\n Action Videos"], "answer_arxiv_id": ["2206.12534"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_13855"} +{"question": "Which works introduced the concept of chain-of-thought (CoT) prompting?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2201.11903", "2205.11916"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_13856"} +{"question": "What are some research papers that discuss the inherent flaws of probing in interpreting language models?", "answer": ["Probing Classifiers: Promises, Shortcomings, and Advances"], "answer_arxiv_id": ["2102.12452"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_13857"} +{"question": "Which works have been conducted in the area of information theory for explaining phenomena in multimodal learning?", "answer": ["Multimodal Co-learning: Challenges, Applications with Datasets, Recent Advances and Future Directions", "Self-supervised Learning from a Multi-view Perspective", "What Makes Multi-modal Learning Better than Single (Provably)"], "answer_arxiv_id": ["2107.13782v3", "2006.05576", "2106.04538"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13858"} +{"question": "What research exists on generating summaries containing more entities while keeping length constant?", "answer": ["From Sparse to Dense: GPT-4 Summarization with Chain of Density\n Prompting"], "answer_arxiv_id": ["2309.04269"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_13859"} +{"question": "What papers elaborate manipulation studies with exploration for interaction?", "answer": ["AdaAfford: Learning to Adapt Manipulation Affordance for 3D Articulated\n Objects via Few-shot Interactions", "Where2Explore: Few-shot Affordance Learning for Unseen Novel Categories\n of Articulated Objects"], "answer_arxiv_id": ["2112.00246", "2309.07473"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_13860"} +{"question": "Which research proposed an approach that uses only the non-descendant variables of the sensitive attribute as model inputs for attaining perfect counterfactual fairness?", "answer": ["Counterfactual Fairness"], "answer_arxiv_id": ["1703.06856"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_13861"} +{"question": "Which research referenced operator methods for analyzing SGD in linear regression?", "answer": ["Benign Overfitting of Constant-Stepsize SGD for Linear Regression", "Last Iterate Risk Bounds of SGD with Decaying Stepsize for Overparameterized Linear Regression"], "answer_arxiv_id": ["2103.12692", "2110.06198"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_13862"} +{"question": "What are some studies that utilize large-scale image-text pairs to expand detection vocabulary?", "answer": ["Open-Vocabulary Object Detection Using Captions", "RegionCLIP: Region-based Language-Image Pretraining", "Open-Vocabulary DETR with Conditional Matching", "PromptDet: Towards Open-vocabulary Detection using Uncurated Images", "Grounded Language-Image Pre-training", "Learning Object-Language Alignments for Open-Vocabulary Object Detection", "DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via Word-Region Alignment"], "answer_arxiv_id": ["2011.10678", "2112.09106", "2203.11876", "2203.16513", "2112.03857", "2211.14843", "2304.04514"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_13863"} +{"question": "What works addressed the computation time limitations in the reverse process of image generation?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13864"} +{"question": "What papers introduced spectral information-based approaches in graphical neural networks?", "answer": ["Spectral Networks and Deep Locally Connected Networks on Graphs", "Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering", "Convolutional Neural Network Architectures for Signals Supported on Graphs", "Semi-Supervised Classification with Graph Convolutional Networks", "CayleyNets: Graph Convolutional Neural Networks with Complex Rational Spectral Filters", "Geometric deep learning on graphs and manifolds using mixture model CNNs"], "answer_arxiv_id": ["1312.6203", "1606.09375", "1805.00165", "1609.02907", "1705.07664", "1611.08402v3"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_13865"} +{"question": "What works achieved good performance without relying on manually annotated word alignment datasets?", "answer": ["SimAlign: High Quality Word Alignments without Parallel Training Data\n using Static and Contextualized Embeddings", "Word Alignment by Fine-tuning Embeddings on Parallel Corpora", "Multilingual Sentence Transformer as A Multilingual Word Aligner"], "answer_arxiv_id": ["2004.08728", "2101.08231", "2301.12140"], "source_meta": {"published_time": "20240716"}, "qid": "AutoScholarQuery_train_13866"} +{"question": "Which references establish the usefulness of injectivity or flow networks for the utility of injectivity and bijectivity for downstream applications?", "answer": ["The Reversible Residual Network: Backpropagation Without Storing Activations", "Non-Euclidean Universal Approximation", "Globally Injective ReLU Networks", "Density estimation using Real NVP", "Faster Uncertainty Quantification for Inverse Problems with Conditional Normalizing Flows", "Residual Flows for Invertible Generative Modeling", "NICE: Non-linear Independent Components Estimation", "Improving Variational Inference with Inverse Autoregressive Flow"], "answer_arxiv_id": ["1707.04585", "2006.02341", "2006.08464", "1605.08803", "2007.07985", "1906.02735", "1410.8516", "1606.04934"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_13867"} +{"question": "Can you provide some examples of works related to online learning?", "answer": ["Online Learning: A Comprehensive Survey", "On Kernelized Multi-armed Bandits", "Online Adaptive Asymmetric Active Learning with Limited Budgets"], "answer_arxiv_id": ["1802.02871", "1704.00445", "1911.07498"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_13868"} +{"question": "What studies used input perplexity as a detection mechanism to defend against optimization-based attacks?", "answer": ["Baseline Defenses for Adversarial Attacks Against Aligned Language\n Models", "Detecting Language Model Attacks with Perplexity"], "answer_arxiv_id": ["2309.00614", "2308.14132"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_13869"} +{"question": "What researchers introduced Non-autoregressive Generative Transformers and proved its efficiency in parallel token generation?", "answer": ["MaskGIT: Masked Generative Image Transformer", "MAGVIT: Masked Generative Video Transformer"], "answer_arxiv_id": ["2202.04200", "2212.05199"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_13870"} +{"question": "What papers describe the utilization of Self-imitation learning for training goal conditioned policies?", "answer": ["Self-Imitation Learning", "Goal-conditioned Imitation Learning", "Goal-Conditioned Reinforcement Learning with Imagined Subgoals", "Learning to Reach Goals via Iterated Supervised Learning"], "answer_arxiv_id": ["1806.05635", "1906.05838", "2107.00541", "1912.06088"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_13871"} +{"question": "What research works studied variance-aware regret in linear bandits with light-tailed rewards?", "answer": ["Information Directed Sampling and Bandits with Heteroscedastic Noise", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs"], "answer_arxiv_id": ["1801.09667", "2205.11507"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_13872"} +{"question": "Which works have proposed methods to generate adversarial examples for DMs?", "answer": ["Adversarial Example Does Good: Preventing Painting Imitation from\n Diffusion Models via Adversarial Examples", "Mist: Towards Improved Adversarial Examples for Diffusion Models", "Anti-DreamBooth: Protecting users from personalized text-to-image\n synthesis"], "answer_arxiv_id": ["2302.04578", "2305.12683", "2303.15433"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_13873"} +{"question": "What research papers analyzed the training dynamics at large widths and training times, using the top eigenvalue of the neural tangent kernel (NTK) as a proxy for sharpness?", "answer": ["The large learning rate phase of deep learning: the catapult mechanism"], "answer_arxiv_id": ["2003.02218"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_13874"} +{"question": "What research provides a convergence rate for with-replacement altSGDA for nonconvex-PŁ objectives?", "answer": ["Faster Single-loop Algorithms for Minimax Optimization without Strong Concavity"], "answer_arxiv_id": ["2112.05604"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_13875"} +{"question": "What studies have previously proposed or used stochastic activation pruning (SAP)?", "answer": ["Stochastic activation pruning for robust adversarial defense"], "answer_arxiv_id": ["1803.01442"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_13876"} +{"question": "Could you provide me references where sketching is applied in distributed problems?", "answer": ["Distributed Low Rank Approximation of Implicit Functions of a Matrix", "Optimal Principal Component Analysis in Distributed and Streaming Models"], "answer_arxiv_id": ["1601.07721", "1504.06729"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_13877"} +{"question": "In what study the researchers conducted linear probing on k-shot samples from the next task, revealing a strong correlation between retaining past information and learning efficiency on new tasks?", "answer": ["Is forgetting less a good inductive bias for forward transfer?"], "answer_arxiv_id": ["2303.08207"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_13878"} +{"question": "Can you provide some references that describe instances where LLMs invoke external tools or models?", "answer": ["Toolformer: Language Models Can Teach Themselves to Use Tools", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face", "ChatGPT for Robotics: Design Principles and Model Abilities"], "answer_arxiv_id": ["2302.04761", "2303.04671", "2303.17580", "2306.17582"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_13879"} +{"question": "Which papers introduced spherical CNNs and later extended the equivariance to conformal transformations?", "answer": ["Learning SO(3) Equivariant Representations with Spherical CNNs"], "answer_arxiv_id": ["1711.06721"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13880"} +{"question": "What papers focus on features extracted by CLIP to enhance the capability of few-shot learning?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling", "Not All Features Matter: Enhancing Few-shot CLIP with Adaptive Prior Refinement", "Visual Prompt Tuning"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2111.03930", "2304.01195", "2203.12119"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_13881"} +{"question": "What paper proposed methods based on linear 3DMMs for face reconstruction?", "answer": ["MoFA: Model-based Deep Convolutional Face Autoencoder for Unsupervised\n Monocular Reconstruction", "Learning an Animatable Detailed 3D Face Model from In-The-Wild Images", "Face Alignment in Full Pose Range: A 3D Total Solution", "Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From\n Single Image to Image Set", "EMOCA: Emotion Driven Monocular Face Capture and Animation", "Learning to Regress 3D Face Shape and Expression from an Image without\n 3D Supervision", "Towards High Fidelity Monocular Face Reconstruction with Rich\n Reflectance using Self-supervised Learning and Ray Tracing", "HiFace: High-Fidelity 3D Face Reconstruction by Learning Static and\n Dynamic Details", "A Hierarchical Representation Network for Accurate and Detailed Face\n Reconstruction from In-The-Wild Images"], "answer_arxiv_id": ["1703.10580", "2012.04012", "1804.01005", "1903.08527", "2204.11312", "1905.06817", "2103.15432", "2303.11225", "2302.14434"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_13882"} +{"question": "Which papers have applied linearized models for predicting fine-tuning generalization?", "answer": ["A linearized framework and a new benchmark for model selection for fine-tuning"], "answer_arxiv_id": ["2102.00084"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_13883"} +{"question": "What paper propose a gradient-free prompt optimization method known as Genetic Prompt Search (GPS)?", "answer": ["GPS: Genetic Prompt Search for Efficient Few-shot Learning"], "answer_arxiv_id": ["2210.17041"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_13884"} +{"question": "What works examined the capability of prompting PaLM for translation?", "answer": ["Prompting PaLM for Translation: Assessing Strategies and Performance", "In-context Examples Selection for Machine Translation"], "answer_arxiv_id": ["2211.09102", "2212.02437"], "source_meta": {"published_time": "20230117"}, "qid": "AutoScholarQuery_train_13885"} +{"question": "Which papers focus on image composition that overlay a foreground object on a background image to yield a composite result?", "answer": ["ST-GAN: Spatial Transformer Generative Adversarial Networks for Image\n Compositing", "GP-GAN: Towards Realistic High-Resolution Image Blending", "Towards Realistic 3D Embedding via View Alignment"], "answer_arxiv_id": ["1803.01837", "1703.07195", "2007.07066v3"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_13886"} +{"question": "Which papers discussed algorithms such as FQI or DAgger for Agnostic MBRL in the context of offline RL?", "answer": ["Information-Theoretic Considerations in Batch Reinforcement Learning", "Agnostic System Identification for Model-Based Reinforcement Learning"], "answer_arxiv_id": ["1905.00360", "1203.1007"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_13887"} +{"question": "Which studies proposed various model architectures for Vision-and-Language Navigation (VLN)?", "answer": ["Reinforced Cross-Modal Matching and Self-Supervised Imitation Learning\n for Vision-Language Navigation", "Self-Monitoring Navigation Agent via Auxiliary Progress Estimation", "Evolving Graphical Planner: Contextual Global Planning for\n Vision-and-Language Navigation", "Object-and-Action Aware Model for Visual Language Navigation"], "answer_arxiv_id": ["1811.10092", "1901.03035", "2007.05655", "2007.14626"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_13888"} +{"question": "What studies utilize tensor factorization to obtain efficient explicit representation in the field of implicit radiance?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "TensoRF: Tensorial Radiance Fields", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "Canonical Factors for Hybrid Neural Fields"], "answer_arxiv_id": ["2112.07945", "2203.09517", "2301.10241", "2308.15461"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_13889"} +{"question": "Can you provide some research papers that implemented listwise reranking into large language models (LLMs)?", "answer": ["Is ChatGPT Good at Search? Investigating Large Language Models as\n Re-Ranking Agents", "RankVicuna: Zero-Shot Listwise Document Reranking with Open-Source Large\n Language Models", "Zero-Shot Listwise Document Reranking with a Large Language Model", "RankZephyr: Effective and Robust Zero-Shot Listwise Reranking is a\n Breeze!", "Large Language Models are Effective Text Rankers with Pairwise Ranking\n Prompting"], "answer_arxiv_id": ["2304.09542", "2309.15088", "2305.02156", "2312.02724", "2306.17563"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_13890"} +{"question": "What works utilise training neural networks to map 3D spatial coordinates to signed distance functions for implicit 3D representation learning?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation"], "answer_arxiv_id": ["1901.05103"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_13891"} +{"question": "Could you provide me examples of research papers that designed monocular 3D detectors for camera-based 3D object detection?", "answer": ["3D Bounding Box Estimation Using Deep Learning and Geometry", "ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", "FCOS: Fully Convolutional One-Stage Object Detection", "Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving", "MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty Propagation", "Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimation", "BEVStereo: Enhancing Depth Estimation in Multi-view 3D Object Detection with Dynamic Temporal Stereo"], "answer_arxiv_id": ["1612.00496", "1812.02781", "1506.01497", "1904.01355", "1812.07179", "2103.12605", "1901.02970", "2209.10248"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_13892"} +{"question": "In which papers have influence functions been used to analyze neural network predictions?", "answer": ["Understanding Black-box Predictions via Influence Functions", "On the Accuracy of Influence Functions for Measuring Group Effects"], "answer_arxiv_id": ["1703.04730", "1905.13289"], "source_meta": {"published_time": "20211224"}, "qid": "AutoScholarQuery_train_13893"} +{"question": "Which papers discuss the use of weight-sharing in neural architecture search (NAS)?", "answer": ["Searching for MobileNetV3", "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", "Neural Architecture Search with Reinforcement Learning", "Single Path One-Shot Neural Architecture Search with Uniform Sampling", "ElasticViT: Conflict-aware Supernet Training for Deploying Fast Vision\n Transformer on Diverse Mobile Devices", "DARTS: Differentiable Architecture Search"], "answer_arxiv_id": ["1905.02244", "1905.11946", "1611.01578", "1904.00420", "2303.09730", "1806.09055"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_13894"} +{"question": "Which papers studied the optimization of functional gradients in the space of probability distributions?", "answer": ["Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithm", "Stein Variational Gradient Descent as Gradient Flow", "Provable Bayesian Inference via Particle Mirror Descent"], "answer_arxiv_id": ["1608.04471", "1704.07520", "1506.03101"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_13895"} +{"question": "Could you inform me about works that have discussed the Mixture of Experts (MOE) concept?", "answer": ["Scaling Vision with Sparse Mixture of Experts", "Pushing Mixture of Experts to the Limit: Extremely Parameter Efficient\n MoE for Instruction Tuning", "SiRA: Sparse Mixture of Low Rank Adaptation", "Mixtral of Experts", "Beyond Distillation: Task-level Mixture-of-Experts for Efficient\n Inference"], "answer_arxiv_id": ["2106.05974", "2309.05444", "2311.09179", "2401.04088", "2110.03742"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_13896"} +{"question": "Which papers are focused on the application of satellite-to-ground synthesis in cross-view image generation?", "answer": ["Sat2Vid: Street-view Panoramic Video Synthesis from a Single Satellite Image", "InfiniCity: Infinite-Scale City Synthesis", "Sat2Density: Faithful Density Learning from Satellite-Ground Image Pairs"], "answer_arxiv_id": ["2012.06628v3", "2301.09637", "2303.14672"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_13897"} +{"question": "Which works generate layout boxes auto-regressively?", "answer": ["Diverse Multimedia Layout Generation with Multi Choice Learning", "LayoutTransformer: Layout Generation and Completion with Self-attention"], "answer_arxiv_id": ["2301.06629", "2006.14615"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_13898"} +{"question": "What works have focused on tasks like visual question answering, VL navigation, or VL manipulation under the topic of Vision-language (VL) grounding?", "answer": ["VQA: Visual Question Answering", "Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments", "CLIPort: What and Where Pathways for Robotic Manipulation", "Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation", "How Much Can CLIP Benefit Vision-and-Language Tasks?", "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision"], "answer_arxiv_id": ["1505.00468v7", "1711.07280", "2109.12098", "2209.05451", "2107.06383", "2102.03334"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_13899"} +{"question": "What works propose various strategies for learning control with bi-level formulations?", "answer": ["Learning Sampling Distributions for Robot Motion Planning", "Differentiable MPC for End-to-end Planning and Control", "Learning Convex Optimization Control Policies", "Task-based End-to-end Model Learning in Stochastic Optimization"], "answer_arxiv_id": ["1709.05448", "1810.13400", "1912.09529", "1703.04529"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_13900"} +{"question": "Which research works predominantly concentrated on the generation of low-resolution images due to the substantial computational burden associated with performing iterative denoising on high-resolution images?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_13901"} +{"question": "What papers reported the addition of a low-cost language-specific module as a solution for multilingual ASR models?", "answer": ["Losses Can Be Blessings: Routing Self-Supervised Speech Representations Towards Efficient Multilingual and Multitask Speech Processing", "Exploiting Adapters for Cross-lingual Low-resource Speech Recognition", "Lightweight Adapter Tuning for Multilingual Speech Translation"], "answer_arxiv_id": ["2211.01522", "2105.11905", "2106.01463"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13902"} +{"question": "Could you give me papers about building a foundation model for promptable segmentation with strong generalization?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "Faster Segment Anything: Towards Lightweight SAM for Mobile Applications", "Semantic-SAM: Segment and Recognize Anything at Any Granularity"], "answer_arxiv_id": ["2401.14159", "2306.14289", "2307.04767"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_13903"} +{"question": "What works use perturbation-based methods to evaluate the change in model outputs with respect to perturbed inputs when quantifying feature importance?", "answer": ["A Unified Approach to Interpreting Model Predictions", "“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "Feature Removal Is A Unifying Principle For Model Explanation Methods"], "answer_arxiv_id": ["1705.07874", "1602.04938", "2011.03623"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_13904"} +{"question": "Can you provide some references on conditional generation using diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Diffusion Models in Vision: A Survey", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2209.04747", "2006.11239"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_13905"} +{"question": "What are some studies regarding quantum algorithms for optimization problems?", "answer": ["Quantum SDP-Solvers: Better upper and lower bounds", "A Quantum Interior Point Method for LPs and SDPs", "Quantum algorithms and lower bounds for convex optimization", "Convex optimization using quantum oracles"], "answer_arxiv_id": ["1705.01843", "1808.09266", "1809.01731", "1809.00643"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_13906"} +{"question": "What studies discuss trajectory optimization over the latent space in RL?", "answer": ["Model-Based Reinforcement Learning via Latent-Space Collocation"], "answer_arxiv_id": ["2106.13229"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_13907"} +{"question": "What studies argue that LLMs, like GPT-4 and GPT-4V, fail to form abstractions and reason in contexts not previously seen in their training data?", "answer": ["Comparing Humans, GPT-4, and GPT-4V On Abstraction and Reasoning Tasks"], "answer_arxiv_id": ["2311.09247"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_13908"} +{"question": "Which papers have studied multi-task reinforcement learning?", "answer": ["Sample Complexity of Multi-task Reinforcement Learning", "Sharing Knowledge in Multi-Task Deep Reinforcement Learning", "Distral: Robust Multitask Reinforcement Learning", "Multi-task Deep Reinforcement Learning with PopArt", "Multi-Task Reinforcement Learning with Context-based Representations"], "answer_arxiv_id": ["1309.6821", "2401.09561", "1707.04175", "1809.04474", "2102.06177"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_13909"} +{"question": "What papers explore the use of active protection methods like encryption for models?", "answer": ["A Survey of Machine and Deep Learning Methods for Internet of Things (IoT) Security"], "answer_arxiv_id": ["1807.11023v1"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_13910"} +{"question": "Which works successfully applied VLMs for image generation tasks?", "answer": ["Diffusion-Based Scene Graph to Image Generation with Masked Contrastive Pre-Training"], "answer_arxiv_id": ["2211.11138"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_13911"} +{"question": "What studies discussed the connections between E2D, MOPS, and OMLE in the setting of partially observable RL?", "answer": ["Unified Algorithms for RL with Decision-Estimation Coefficients: No-Regret, PAC, and Reward-Free Learning"], "answer_arxiv_id": ["2209.11745"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_13912"} +{"question": "Which works established the idea of using a predictive model for better self-supervised learning?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Recurrent Latent Variable Model for Sequential Data"], "answer_arxiv_id": ["1807.03748", "1506.02216"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_13913"} +{"question": "What works use distributions of positive and negative samples as a confidence measure in pseudo labeling?", "answer": ["Addressing Failure Prediction by Learning Model Confidence"], "answer_arxiv_id": ["1910.04851"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_13914"} +{"question": "Which methods advanced visual representation by integrating a Perceiver Resampler with vision encoders?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "OpenFlamingo: An Open-Source Framework for Training Large Autoregressive\n Vision-Language Models"], "answer_arxiv_id": ["2204.14198", "2308.01390"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_13915"} +{"question": "What papers discuss the initial attempts at building ToM representations in neural network based models?", "answer": ["Modeling Others using Oneself in Multi-Agent Reinforcement Learning", "Machine Theory of Mind"], "answer_arxiv_id": ["1802.09640", "1802.07740v2"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_13916"} +{"question": "Which papers are related to 1-bit low-rank matrix completion?", "answer": ["1-Bit Matrix Completion"], "answer_arxiv_id": ["1209.3672"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13917"} +{"question": "Could you mention studies exploring the use of an external memory module for tackling long-horizon planning problems in RL-based navigation?", "answer": ["Memory-Augmented Reinforcement Learning for Image-Goal Navigation", "Scene Memory Transformer for Embodied Agents in Long-Horizon Tasks", "Learning to plan with uncertain topological maps", "Semi-parametric Topological Memory for Navigation", "Topological Semantic Graph Memory for Image-Goal Navigation", "Weakly-Supervised Multi-Granularity Map Learning for Vision-and-Language Navigation"], "answer_arxiv_id": ["2101.05181", "1903.03878", "2007.05270", "1803.00653", "2209.08274", "2210.07506"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_13918"} +{"question": "Which works successfully demonstrated using pretrained representations across areas such as computer vision, natural language processing and audio?", "answer": ["Rich feature hierarchies for accurate object detection and semantic segmentation", "Unsupervised Visual Representation Learning by Context Prediction", "Momentum Contrast for Unsupervised Visual Representation Learning", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1311.2524", "1505.05192", "1911.05722", "1810.04805", "2005.14165", "1807.03748"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_13919"} +{"question": "What studies give insights about post-processing backdoor defense strategies for mitigating the effect of backdoors?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks", "Adversarial Neuron Pruning Purifies Backdoored Deep Models", "Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks", "Adversarial Unlearning of Backdoors via Implicit Hypergradient"], "answer_arxiv_id": ["1805.12185", "2110.14430", "2101.05930", "2110.03735"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_13920"} +{"question": "Which works present strategies to estimate scaling law parameters in deep learning?", "answer": ["Revisiting Neural Scaling Laws in Language and Vision"], "answer_arxiv_id": ["2209.06640"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_13921"} +{"question": "What research papers can you cite about attention manipulation methods in layout-aware text-to-image generation?", "answer": ["eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Zero-shot spatial layout conditioning for text-to-image diffusion models", "BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained\n Diffusion", "Training-Free Location-Aware Text-to-Image Synthesis", "Guided Image Synthesis via Initial Image Editing in Diffusion Model", "Dense Text-to-Image Generation with Attention Modulation", "R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image\n Generation", "Grounded Text-to-Image Synthesis with Attention Refocusing"], "answer_arxiv_id": ["2211.01324", "2306.13754", "2307.10816", "2304.13427", "2305.03382", "2308.12964", "2310.08872", "2306.05427v2"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_13922"} +{"question": "What papers have demonstrated properties of transfer between self-supervised pre-training and supervised FT?", "answer": ["Exploring the Limits of Large Scale Pre-training", "Understanding Contrastive Learning Requires Incorporating Inductive Biases"], "answer_arxiv_id": ["2110.02095", "2202.14037"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_13923"} +{"question": "Which works discuss the gradient flow properties of Deep Neural Networks?", "answer": ["Exact solutions to the nonlinear dynamics of learning in deep linear neural networks", "Dynamical Isometry and a Mean Field Theory of CNNs: How to Train 10,000-Layer Vanilla Convolutional Neural Networks", "Initialization of ReLUs for Dynamical Isometry"], "answer_arxiv_id": ["1312.6120", "1806.05393", "1806.06362"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_13924"} +{"question": "Could you provide me with the papers that apply the masked modeling paradigm to time series analysis?", "answer": ["A Transformer-based Framework for Multivariate Time Series Representation Learning", "A Time Series is Worth 64 Words: Long-term Forecasting with Transformers", "Ti-MAE: Self-Supervised Masked Time Series Autoencoders"], "answer_arxiv_id": ["2010.02803", "2211.14730", "2301.08871"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13925"} +{"question": "Which work tries to solve by building MLLMs by combining models of other modalities?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "MultiModal-GPT: A Vision and Language Model for Dialogue with Humans", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model"], "answer_arxiv_id": ["2204.14198", "2304.10592", "2305.03726", "2301.12597", "2305.04790", "2304.14178", "2305.06500", "2305.11175", "2306.05424", "2304.15010"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_13926"} +{"question": "Which works discuss triggerless data poisoning attacks that can be done in a label-only fashion?", "answer": ["Label Sanitization against Label Flipping Poisoning Attacks", "Just How Toxic is Data Poisoning? A Unified Benchmark for Backdoor and Data Poisoning Attacks"], "answer_arxiv_id": ["1803.00992", "2006.12557"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_13927"} +{"question": "Which work proposed the single-step revealing condition in under-complete POMDPs?", "answer": ["Sample-Efficient Reinforcement Learning of Undercomplete POMDPs"], "answer_arxiv_id": ["2006.12484"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13928"} +{"question": "In what paper is the skip-convolution approach proposed for efficient video processing?", "answer": ["Skip-Convolutions for Efficient Video Processing"], "answer_arxiv_id": ["2104.11487"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_13929"} +{"question": "Which researchers first proposed Score Distillation Sampling (SDS)?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_13930"} +{"question": "Which works propose pre-training vision-language models through reconstructing masked inputs?", "answer": ["VisualBERT: A Simple and Performant Baseline for Vision and Language", "UNITER: UNiversal Image-TExt Representation Learning", "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks"], "answer_arxiv_id": ["1908.03557", "1909.11740", "2102.03334", "2108.10904", "2208.10442"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_13931"} +{"question": "What papers have adopted reinforcement learning for automatic augmentations in vision?", "answer": ["AutoAugment: Learning Augmentation Strategies from Data"], "answer_arxiv_id": ["1805.09501"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_13932"} +{"question": "In what papers are Diffusion models trained using image-text pairs in order to produce semantically rich embeddings for generation and editing of images?", "answer": ["Imagic: Text-Based Real Image Editing with Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "PRedItOR: Text Guided Image Editing with Diffusion Prior", "DiffEdit: Diffusion-based semantic image editing with mask guidance", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2210.09276", "2205.11487", "2302.07979", "2210.11427", "2211.09800"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_13933"} +{"question": "What study presents the first server-less, peer-to-peer Federated Learning (FL) approach?", "answer": ["BrainTorrent: A Peer-to-Peer Environment for Decentralized Federated Learning"], "answer_arxiv_id": ["1905.06731"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_13934"} +{"question": "Which research introduced the concept of Knowledge Distillation?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13935"} +{"question": "What studies reformed the task as data-rich tasks like NLI or QA for indirect supervision of low-resource IE models?", "answer": ["Can NLI Provide Proper Indirect Supervision for Low-resource Biomedical\n Relation Extraction?", "Textual Entailment for Event Argument Extraction: Zero- and Few-Shot\n with Multi-Source Learning", "Parameter-Efficient Low-Resource Dialogue State Tracking by Prompt\n Tuning", "Summarization as Indirect Supervision for Relation Extraction"], "answer_arxiv_id": ["2212.10784", "2205.01376", "2301.10915", "2205.09837"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_13936"} +{"question": "Could you provide a study which has proposed using instance-independent prompt-tuning?", "answer": ["IDPG: An Instance-Dependent Prompt Generation Method"], "answer_arxiv_id": ["2204.04497"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_13937"} +{"question": "Which works propose ways to improve dense retrieval models?", "answer": ["ERNIE-Search: Bridging Cross-Encoder with Dual-Encoder via Self On-the-fly Distillation for Dense Passage Retrieval"], "answer_arxiv_id": ["2205.09153"], "source_meta": {"published_time": "20230409"}, "qid": "AutoScholarQuery_train_13938"} +{"question": "What is the work that proposed a method to utilize multiple diffusion vector volumes as input and learn volume-wise denoising in MRI?", "answer": ["Patch2Self: Denoising Diffusion MRI with Self-Supervised Learning"], "answer_arxiv_id": ["2011.01355"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_13939"} +{"question": "Which study showed that GRW and ERM are equivalent under certain conditions, particularly under the zero-one loss?", "answer": ["Does Distributionally Robust Supervised Learning Give Robust Classifiers?"], "answer_arxiv_id": ["1611.02041"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_13940"} +{"question": "Which language models were shown to achieve state-of-the-art performance on many tasks in natural language processing such as text generation, summarization, or (supervised) machine translation?", "answer": ["Language Models are Few-Shot Learners", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "1810.04805", "2204.02311"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_13941"} +{"question": "What studies provided a theoretical analysis of GNNs based on the Stochastic Block Model?", "answer": ["Analysis of spectral clustering algorithms for community detection: the general bipartite setting", "Consistency of spectral clustering in stochastic block models", "Convergence and Stability of Graph Convolutional Networks on Large Random Graphs"], "answer_arxiv_id": ["1803.04547", "1312.2050", "2006.01868"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_13942"} +{"question": "Can you provide papers that have discussed social fairness in the field of algorithm and machine learning fairness?", "answer": ["Fair Clustering via Equitable Group Representations", "Approximation Algorithms for Socially Fair Clustering", "Approximating Fair Clustering with Cascaded Norm Objectives"], "answer_arxiv_id": ["2006.11009", "2103.02512v2", "2111.04804"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_13943"} +{"question": "What references provide examples of 3D GNN models that utilize roto-translational invariant 3D information?", "answer": ["Equivariant message passing for the prediction of tensorial properties and molecular spectra", "Learning from Protein Structure with Geometric Vector Perceptrons"], "answer_arxiv_id": ["2102.03150", "2009.01411"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_13944"} +{"question": "Are there any works that mention listwise reranking contributing to the 'lost in the middle' problem in LLMs?", "answer": ["Found in the Middle: Permutation Self-Consistency Improves Listwise\n Ranking in Large Language Models"], "answer_arxiv_id": ["2310.07712"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_13945"} +{"question": "Could you provide me with studies that explored differentiable relaxations for dynamic programming based losses in deep learning?", "answer": ["Soft-DTW: a Differentiable Loss Function for Time-Series", "Differentiable Dynamic Programming for Structured Prediction and Attention", "Differentiable Divergences Between Time Series"], "answer_arxiv_id": ["1703.01541", "1802.03676", "2010.08354v3"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_13946"} +{"question": "Could you provide me some works that used deep learning-based methods in snapshot compressive imaging?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation", "Generative Adversarial Networks", "Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "Deep Residual Learning for Image Recognition", "Memory-Efficient Network for Large-scale Video Compressive Sensing", "MetaSCI: Scalable and Adaptive Reconstruction for Video Compressive\n Sensing", "Plug-and-Play Algorithms for Large-scale Snapshot Compressive Imaging", "Plug-and-Play Algorithms for Video Snapshot Compressive Imaging", "Spatial-Temporal Transformer for Video Snapshot Compressive Imaging", "EfficientSCI: Densely Connected Network with Space-time Factorization\n for Large-scale Video Snapshot Compressive Imaging", "Attention Is All You Need"], "answer_arxiv_id": ["1505.04597", "1406.2661", "1603.08155", "1512.03385", "2103.03089", "2103.01786", "2003.13654", "2101.04822", "2209.01578", "2305.10006", "1706.03762"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_13947"} +{"question": "Which papers are closest to this work in their study of convergence on multi-layer neural networks?", "answer": ["Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "An Improved Analysis of Training Over-parameterized Deep Neural Networks", "Loss landscapes and optimization in over-parameterized non-linear systems and neural networks"], "answer_arxiv_id": ["1811.03804", "1811.03962", "1906.04688", "2003.00307"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_13948"} +{"question": "What studies have focused on the task of image-to-image translation?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "StarGAN: Unified Generative Adversarial Networks for Multi-Domain\n Image-to-Image Translation", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial\n Networks", "Learning to Discover Cross-Domain Relations with Generative Adversarial\n Networks", "Palette: Image-to-Image Diffusion Models", "StyleDrop: Text-to-Image Generation in Any Style", "Pretraining is All You Need for Image-to-Image Translation", "Tuning-Free Inversion-Enhanced Control for Consistent Image Editing"], "answer_arxiv_id": ["1611.07004", "1711.09020", "1703.10593", "1703.05192", "2111.05826", "2306.00983", "2205.12952", "2312.14611"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_13949"} +{"question": "What studies have discussed adjusting time schedule in text-to-3D generation?", "answer": ["Diffusion models as plug-and-play priors", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2206.09012", "2212.00774v1"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13950"} +{"question": "Which works considered transformer-based architectures for vision tasks in resource-limited settings?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Training data-efficient image transformers & distillation through\n attention", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2010.11929", "2012.12877", "2103.14030"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_13951"} +{"question": "What studies introduced and explored the concept of federated bandits with a differential privacy guarantee?", "answer": ["Federated Recommendation System via Differential Privacy", "Federated Bandit: A Gossiping Approach"], "answer_arxiv_id": ["2005.06670v2", "2010.12763"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_13952"} +{"question": "Which work focused on finding optimized textual embeddings for custom concepts to generate concept-reflecting images in T2I models?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_13953"} +{"question": "Could you provide me some studies that discuss relevance maps related to the class-specific nature of relevance?", "answer": ["Learning Important Features Through Propagating Activation Differences", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1704.02685", "1703.01365"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_13954"} +{"question": "What studies utilized external knowledge and tools to assist models in accomplishing reasoning tasks?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback", "PAL: Program-aided Language Models", "Toolformer: Language Models Can Teach Themselves to Use Tools"], "answer_arxiv_id": ["2112.09332", "2211.10435", "2302.04761"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_13955"} +{"question": "Are there any ML-PDE works that do not require training data and can function as solvers themselves?", "answer": ["Learning Incompressible Fluid Dynamics from Scratch - Towards Fast, Differentiable Fluid Models that Generalize", "Learning the solution operator of parametric partial differential equations with physics-informed DeepOnets"], "answer_arxiv_id": ["2006.08762", "2103.10974"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_13956"} +{"question": "Which work shows that minimizing squared loss for sufficiently deep neural nets yields multicalibration with regard to smaller neural nets of a fixed complexity?", "answer": ["Loss Minimization Yields Multicalibration for Large Neural Networks"], "answer_arxiv_id": ["2304.09424"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_13957"} +{"question": "Can you identify the works that used Monte Carlo approximation of data-dependent parameter distributions for quicker kernel approximation?", "answer": ["On the Equivalence between Kernel Quadrature Rules and Random Feature Expansions", "Fast Approximation and Estimation Bounds of Kernel Quadrature for Infinitely Wide Models"], "answer_arxiv_id": ["1502.06800", "1902.00648"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_13958"} +{"question": "Which works first sample the number of dimensions and then run the diffusion process in this fixed dimension?", "answer": ["Equivariant Diffusion for Molecule Generation in 3D", "Equivariant 3D-Conditional Diffusion Models for Molecular Linker Design"], "answer_arxiv_id": ["2203.17003", "2210.05274"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_13959"} +{"question": "Could you provide me some works that utilize shallow classifiers with extracted text features?", "answer": ["Origin Tracing and Detecting of LLMs", "Intrinsic Dimension Estimation for Robust Detection of AI-Generated\n Texts"], "answer_arxiv_id": ["2304.14072", "2306.04723"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_13960"} +{"question": "Are there any studies about employing a debugging component used to improve existing implementations based on feedback from a code execution environment?", "answer": ["Teaching Large Language Models to Self-Debug"], "answer_arxiv_id": ["2304.05128v2"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_13961"} +{"question": "What works suggested extracting pairwise similarities from tree-based ensembles?", "answer": ["Geometry- and Accuracy-Preserving Random Forest Proximities"], "answer_arxiv_id": ["2201.12682"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_13962"} +{"question": "Can you provide the studies that first proposed knowledge distillation?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_13963"} +{"question": "Which works lay the foundation for diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1503.03585", "1907.05600"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_13964"} +{"question": "What are some of the transformer based models proposed for object detection tasks?", "answer": ["End-to-End Object Detection with Transformers", "Deformable DETR: Deformable Transformers for End-to-End Object Detection", "Sparse DETR: Efficient End-to-End Object Detection with Learnable Sparsity", "DETRs with Hybrid Matching", "DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection", "DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETR"], "answer_arxiv_id": ["2005.12872", "2010.04159", "2111.14330", "2207.13080", "2203.03605", "2201.12329"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_13965"} +{"question": "For a thorough understanding of GNN concepts, what papers would you recommend?", "answer": ["Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges", "Theory of Graph Neural Networks: Representation and Learning"], "answer_arxiv_id": ["2104.13478", "2204.07697"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_13966"} +{"question": "What papers have been proposed regarding dynamic scene modeling?", "answer": ["Neural Human Video Rendering by Learning Dynamic Textures and\n Rendering-to-Video Translation", "AvatarReX: Real-time Expressive Full-body Avatars", "Efficient Neural Radiance Fields for Interactive Free-viewpoint Video", "CloSET: Modeling Clothed Humans on Continuous Surface with Explicit\n Template Decomposition", "NeuralHumanFVV: Real-Time Neural Volumetric Human Performance Rendering\n using RGB Cameras", "Neural Free-Viewpoint Performance Rendering under Complex Human-object\n Interactions", "Artemis: Articulated Neural Pets with Appearance and Motion synthesis"], "answer_arxiv_id": ["2001.04947", "2305.04789", "2112.01517", "2304.03167", "2103.07700", "2108.00362", "2202.05628"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_13967"} +{"question": "Which papers propose the use of DDMs as priors for image restoration adhering to the DDIM scheme?", "answer": ["Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "Denoising Diffusion Models for Plug-and-Play Image Restoration"], "answer_arxiv_id": ["2201.11793", "2212.00490", "2305.08995"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_13968"} +{"question": "Which papers present non-contrastive, joint-embedding methods in the field of self-supervised learning?", "answer": ["How Does SimSiam Avoid Collapse Without Negative Samples? A Unified Understanding with Self-supervised Contrastive Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2203.16262", "2103.03230", "2011.10566"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_13969"} +{"question": "Which paper used ChatGPT to generate prompts for multimodal tasks?", "answer": ["Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models"], "answer_arxiv_id": ["2303.04671"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_13970"} +{"question": "Which papers introduced the concept of demographic parity in the context of fairness in machine learning?", "answer": ["Fairness Through Awareness"], "answer_arxiv_id": ["1104.3913"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_13971"} +{"question": "What studies utilized Sparse MoEs in the domain of natural language processing?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer", "BASE Layers: Simplifying Training of Large, Sparse Models", "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity"], "answer_arxiv_id": ["1701.06538", "2103.16716", "2101.03961"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_13972"} +{"question": "Which works propose the use of GAN for purifying adversarial inputs in adversarial purification methods?", "answer": ["Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models"], "answer_arxiv_id": ["1805.06605"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_13973"} +{"question": "Can you name the studies that count the episodic state visitations?", "answer": ["RIDE: Rewarding Impact-Driven Exploration for Procedurally-Generated Environments"], "answer_arxiv_id": ["2002.12292"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_13974"} +{"question": "Which works proposed alternate representations based on implicit surfaces for 3D GANs?", "answer": ["Multiview Neural Surface Reconstruction by Disentangling Geometry and\n Appearance", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction"], "answer_arxiv_id": ["2003.09852", "2104.10078", "2106.10689"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_13975"} +{"question": "Which datasets are created through a comprehensive human annotation process?", "answer": ["OpenAssistant Conversations -- Democratizing Large Language Model\n Alignment"], "answer_arxiv_id": ["2304.07327"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_13976"} +{"question": "Can you tell me about the works that achieved the convergence rate for log-linear policies through the off-policy NAC?", "answer": ["Finite-Sample Analysis of Off-Policy Natural Actor-Critic with Linear Function Approximation"], "answer_arxiv_id": ["2105.12540"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_13977"} +{"question": "Which works showcased remarkable novel view synthesis quality by volume rendering a neural implicit field?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance\n Fields", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2103.13415", "2112.03907", "2304.06706"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_13978"} +{"question": "Which study first adopted the CNN architecture to image denoising?", "answer": ["Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising"], "answer_arxiv_id": ["1608.03981"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_13979"} +{"question": "Which papers belong to the online conformal prediction group of work?", "answer": ["Adaptive Conformal Inference Under Distribution Shift", "Conformal Inference for Online Prediction with Arbitrary Distribution Shifts", "Adaptive Conformal Predictions for Time Series"], "answer_arxiv_id": ["2106.00170", "2208.08401", "2202.07282"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_13980"} +{"question": "What works propose to finetune language models using either human demonstrations or reinforcement learning from human feedback?", "answer": ["UnifiedQA: Crossing Format Boundaries With a Single QA System", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "Adapting Language Models for Zero-shot Learning by Meta-tuning on Dataset and Prompt Collections", "Finetuned Language Models Are Zero-Shot Learners", "Deep Reinforcement Learning from Human Preferences", "Learning to summarize from human feedback", "A General Language Assistant as a Laboratory for Alignment", "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2005.00700", "2110.08207", "2104.04670", "2109.01652", "1706.03741", "2009.01325", "2112.00861", "2204.05862", "2203.02155"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_13981"} +{"question": "Which studies establish that deeper networks can behave similar to very shallow networks?", "answer": ["Deep Linear Networks can Benignly Overfit when Shallow Ones Do"], "answer_arxiv_id": ["2209.09315"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_13982"} +{"question": "Any works that investigated the task similarity from a theoretical perspective, using synthetic tasks with a controlled setting?", "answer": ["Continual Learning in the Teacher-Student Setup: Impact of Task\n Similarity"], "answer_arxiv_id": ["2107.04384"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_13983"} +{"question": "Any studies about a model that employs a spatiotemporal grid mini-cube sampling method to extract fragments from original videos, enabling end-to-end training of the VQA model?", "answer": ["FAST-VQA: Efficient End-to-end Video Quality Assessment with Fragment\n Sampling"], "answer_arxiv_id": ["2207.02595"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_13984"} +{"question": "Which papers observed heavy-tailed weight spectra in the pre-trained models?", "answer": ["Traditional and Heavy-Tailed Self Regularization in Neural Network Models", "Implicit Self-Regularization in Deep Neural Networks: Evidence from Random Matrix Theory and Implications for Learning"], "answer_arxiv_id": ["1901.08276", "1810.01075"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_13985"} +{"question": "Are there any works that analyzed the limitations of networks in nonlinear function approximation within the scope of extrapolation theory?", "answer": ["How Neural Networks Extrapolate: From Feedforward to Graph Neural Networks"], "answer_arxiv_id": ["2009.11848"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_13986"} +{"question": "Which research works propose policy regularization methods in offline RL?", "answer": ["Offline Reinforcement Learning with Fisher Divergence Critic Regularization", "Off-Policy Deep Reinforcement Learning without Exploration"], "answer_arxiv_id": ["2103.08050", "1812.02900"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_13987"} +{"question": "Can you point me to studies that have studied decentralized learning in multi-queue multi-server settings?", "answer": ["Stability of Decentralized Queueing Networks Beyond Complete Bipartite Cases", "Decentralized Learning in Online Queuing Systems", "Efficient decentralized multi-agent learning in asymmetric bipartite queueing systems"], "answer_arxiv_id": ["2210.07632", "2106.04228", "2206.03324"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_13988"} +{"question": "Who extended k-nearest neighbors to deep networks?", "answer": ["Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning", "Visual correspondence-based explanations improve AI robustness and human-AI team accuracy"], "answer_arxiv_id": ["1803.04765", "2208.00780"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_13989"} +{"question": "Which works indicate that S4 and Transformers have complementary strengths?", "answer": ["Efficient Long Sequence Modeling via State Space Augmented Transformer", "Long Range Language Modeling via Gated State Spaces", "Hungry Hungry Hippos: Towards Language Modeling with State Space Models", "Long Movie Clip Classification with State-Space Video Models", "Simplifying and Understanding State Space Models with Diagonal Linear RNNs"], "answer_arxiv_id": ["2212.08136", "2206.13947", "2212.14052", "2204.01692v3", "2212.00768"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_13990"} +{"question": "What study proposed the use of EAS in solving CO problems?", "answer": ["Efficient Active Search for Combinatorial Optimization Problems"], "answer_arxiv_id": ["2106.05126"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_13991"} +{"question": "What papers discussed reinforcement learning-based approaches for fine-tuning the LLM to maximize the reward model?", "answer": ["Training language models to follow instructions with human feedback", "Proximal Policy Optimization Algorithms", "Safe RLHF: Safe Reinforcement Learning from Human Feedback"], "answer_arxiv_id": ["2203.02155", "1707.06347", "2310.12773"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_13992"} +{"question": "What papers provide examples of research that combined various GNNs with different scopes?", "answer": ["MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing", "N-GCN: Multi-scale Graph Convolution for Semi-supervised Node Classification"], "answer_arxiv_id": ["1905.00067", "1802.08888"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_13993"} +{"question": "What studies propose approaches for adaptation in damaged body and perturbed action space scenarios?", "answer": ["Learning Fast Adaptation with Meta Strategy Optimization", "Learning to Adapt in Dynamic, Real-World Environments through Meta-Reinforcement Learning", "Preparing for the Unknown: Learning a Universal Policy with Online System Identification"], "answer_arxiv_id": ["1909.12995", "1803.11347", "1702.02453"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_13994"} +{"question": "Could you provide me some studies about causal inference?", "answer": ["Invariant Risk Minimization", "Towards Non-I.I.D. Image Classification: A Dataset and Baselines", "Heterogeneous Risk Minimization", "Domain Generalization using Causal Matching", "Out-of-distribution Generalization with Causal Invariant Transformations", "Towards a Theoretical Framework of Out-of-Distribution Generalization"], "answer_arxiv_id": ["1907.02893", "1906.02899", "2105.03818", "2006.07500", "2203.11528", "2106.04496"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_13995"} +{"question": "Which works have tried to design a lightweight and effective point-based detector?", "answer": ["3DSSD: Point-based 3D Single Stage Object Detector", "SASA: Semantics-Augmented Set Abstraction for Point-based 3D Object Detection", "Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds"], "answer_arxiv_id": ["2002.10187", "2201.01976", "2203.11139"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_13996"} +{"question": "Which studies have reported on retina models for increasing adversarial robustness of the systems?", "answer": ["Foveation-based Mechanisms Alleviate Adversarial Examples", "Biologically Inspired Mechanisms for Adversarial Robustness"], "answer_arxiv_id": ["1511.06292", "2006.16427"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_13997"} +{"question": "Could you provide me some papers in which the PAQ is viewed as extending the set of items to a continuous spectrum when using a generative model like GAN?", "answer": ["Generative Adversarial Nets", "Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["1406.2661", "1912.04958"], "source_meta": {"published_time": "20230908"}, "qid": "AutoScholarQuery_train_13998"} +{"question": "What work relates the denoising paradigm to the discretization of a stochastic differential equation?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_13999"} +{"question": "What studies have looked at adversarially delayed feedback in research?", "answer": ["Learning Adversarial Markov Decision Processes with Delayed Feedback", "Near-Optimal Regret for Adversarial MDP with Delayed Bandit Feedback", "Nonstochastic Multiarmed Bandits with Unrestricted Delays", "An Optimal Algorithm for Adversarial Bandits with Arbitrary Delays", "Adapting to Delays and Data in Adversarial Multi-Armed Bandits", "Nonstochastic Bandits and Experts with Arm-Dependent Delays"], "answer_arxiv_id": ["2012.14843", "2201.13172", "1906.00670", "1910.06054v2", "2010.06022", "2111.01589v2"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_14000"} +{"question": "Which study proposed the method 'Reinforcement Learning by Simulating the Past', also known as 'IRL from a single state'?", "answer": ["Preferences Implicit in the State of the World"], "answer_arxiv_id": ["1902.04198"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14001"} +{"question": "What papers proposed joint representation framework for motion generation?", "answer": ["Human Motion Diffusion Model", "T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete\n Representations"], "answer_arxiv_id": ["2209.14916", "2301.06052"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_14002"} +{"question": "Can you give an example of a paper that integrated self-evaluation and rewind mechanisms to promote harmless responses?", "answer": ["RAIN: Your Language Models Can Align Themselves without Finetuning"], "answer_arxiv_id": ["2309.07124"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_14003"} +{"question": "Which studies have elucidated the relationship between structural properties of a graph and ORC?", "answer": ["Comparative analysis of two discretizations of Ricci curvature for complex networks"], "answer_arxiv_id": ["1712.07600"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_14004"} +{"question": "Any research focused on unsupervised multi-object image segmentation?", "answer": ["Object-Centric Learning with Slot Attention", "MarioNette: Self-Supervised Sprite Learning", "Unsupervised Layered Image Decomposition into Object Prototypes", "Unsupervised Discovery of Object Radiance Fields"], "answer_arxiv_id": ["2006.15055", "2104.14553", "2104.14575", "2107.07905"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_14005"} +{"question": "What research extended the application of DDPM to the point cloud completion task by training a point-voxel CNN?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion"], "answer_arxiv_id": ["2104.03670"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_14006"} +{"question": "Which work features a modification to the spatio-temporal 3D convolution for enabling efficient continual inference in existing 3D CNNs?", "answer": ["Continual 3D Convolutional Neural Networks for Real-time Processing of Videos"], "answer_arxiv_id": ["2106.00050"], "source_meta": {"published_time": "20220117"}, "qid": "AutoScholarQuery_train_14007"} +{"question": "Who studied the infinite width limit for convergence analysis?", "answer": ["One-pass Stochastic Gradient Descent in Overparametrized Two-layer Neural Networks"], "answer_arxiv_id": ["2105.00262"], "source_meta": {"published_time": "20221114"}, "qid": "AutoScholarQuery_train_14008"} +{"question": "Could you provide me some works about producing policies within a family of tasks related to φ𝜑?", "answer": ["Successor Features for Transfer in Reinforcement Learning", "Universal Successor Features Approximators", "Deep Reinforcement Learning with Successor Features for Navigation across Similar Environments", "Disentangled Cumulants Help Successor Representations Transfer to New Tasks"], "answer_arxiv_id": ["1606.05312", "1812.07626", "1612.05533", "1911.10866"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_14009"} +{"question": "Could you provide me some works about predicting the corresponding position on the object CAD model for every point?", "answer": ["GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D\n Object Pose Estimation"], "answer_arxiv_id": ["2102.12145"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_14010"} +{"question": "What studies have applied reinforcement learning in recommender systems?", "answer": ["Recommendations with Negative Feedback via Pairwise Deep Reinforcement Learning", "Reinforcement Learning to Optimize Long-term User Engagement in Recommender Systems", "Jointly Learning to Recommend and Advertise", "Two-Stage Constrained Actor-Critic for Short Video Recommendation", "Reinforcing User Retention in a Billion Scale Short Video Recommender System"], "answer_arxiv_id": ["1802.06501", "1902.05570", "2003.00097", "2302.01680", "2302.01724"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_14011"} +{"question": "Do you have any sources about in-sample algorithms such as AWR, IQL, and AWAC method?", "answer": ["Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning", "AWAC: Accelerating Online Reinforcement Learning with Offline Datasets"], "answer_arxiv_id": ["1910.00177", "2110.06169", "2006.09359"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_14012"} +{"question": "Which paper presents a class of unbiased estimators for kernel density in high dimensions for a variety of commonly used kernels?", "answer": ["Hashing-Based-Estimators for Kernel Density in High Dimensions"], "answer_arxiv_id": ["1808.10530v1"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_14013"} +{"question": "Which studies discuss data augmentation as a method to mitigate spurious correlations?", "answer": ["Is BERT Really Robust? A Strong Baseline for Natural Language Attack on\n Text Classification and Entailment", "Generating Natural Language Adversarial Examples", "Identifying and Mitigating Spurious Correlations for Improving\n Robustness in NLP Models", "Automatic Shortcut Removal for Self-Supervised Representation Learning"], "answer_arxiv_id": ["1907.11932", "1804.07998", "2110.07736", "2002.08822"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_14014"} +{"question": "What research efforts have been made in merging the vocabularies of semantic segmentation datasets?", "answer": ["MSeg: A Composite Dataset for Multi-domain Semantic Segmentation", "LMSeg: Language-guided Multi-dataset Segmentation", "Learning Semantic Segmentation from Multiple Datasets with Label Shifts"], "answer_arxiv_id": ["2112.13762", "2302.13495", "2202.14030"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14015"} +{"question": "Which studies provide variations of Variational Autoencoder for improving the disentanglement of the learned representation?", "answer": ["Auto-Encoding Variational Bayes", "Disentangling by Factorising"], "answer_arxiv_id": ["1312.6114", "1802.05983"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_14016"} +{"question": "What works employ periodic functions for 3D-aware image synthesis to represent scenes as view-consistent radiance fields?", "answer": ["pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis", "StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation"], "answer_arxiv_id": ["2012.00926", "2112.11427"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_14017"} +{"question": "What studies propose different metrics for goal choosing like learning progress and goal difficulty in goal-directed exploration?", "answer": ["Active Learning of Inverse Models with Intrinsically Motivated Goal Exploration in Robots", "Many-Goals Reinforcement Learning", "Automatic Goal Generation for Reinforcement Learning Agents", "Keeping Your Distance: Solving Sparse Reward Tasks Using Self-Balancing Shaped Rewards", "Automatic Curriculum Learning through Value Disagreement"], "answer_arxiv_id": ["1301.4862", "1806.09605", "1705.06366", "1911.01417", "2006.09641"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_14018"} +{"question": "Which studies focus on masked language modeling and autoregressive language modeling tasks in model editing?", "answer": ["Knowledge Neurons in Pretrained Transformers", "LM-Debugger: An Interactive Tool for Inspection and Intervention in Transformer-Based Language Models"], "answer_arxiv_id": ["2104.08696", "2204.12130"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_14019"} +{"question": "What papers are about applying diffusion models in the context of image generation?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation", "DiffCollage: Parallel Generation of Large Content with Diffusion Models", "Text2Light: Zero-Shot Text-Driven HDR Panorama Generation"], "answer_arxiv_id": ["2204.06125", "2112.10741", "2112.10752", "2205.11487", "2302.08113", "2303.17076", "2209.09898"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_14020"} +{"question": "Which works looked at constraining the adversary to be smooth in the context of online learning?", "answer": ["Smoothed Analysis with Adaptive Adversaries", "Oracle-Efficient Online Learning for Beyond Worst-Case Adversaries", "Smoothed Online Learning is as Easy as Statistical Learning"], "answer_arxiv_id": ["2102.08446", "2202.08549v3", "2202.04690"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_14021"} +{"question": "Can you cite works that are involved in exploring two-stage design for estimating 3D poses using hand-crafted affinity matrix?", "answer": ["Fast and Robust Multi-Person 3D Pose Estimation from Multiple Views"], "answer_arxiv_id": ["1901.04111"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_14022"} +{"question": "Which research included per-pixel lane mark labeling in 35 classes with multiple sensors?", "answer": ["The ApolloScape Open Dataset for Autonomous Driving and its Application"], "answer_arxiv_id": ["1803.06184"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_14023"} +{"question": "Which studies primarily focus on extracting sequential actions in procedural graph extraction?", "answer": ["Constructing Flow Graphs from Procedural Cybersecurity Texts"], "answer_arxiv_id": ["2105.14357"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_14024"} +{"question": "Are there any research papers that study the relation between multicalibration and algorithmic fairness considerations?", "answer": ["Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness"], "answer_arxiv_id": ["1711.05144"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14025"} +{"question": "Which work performed studies on both size and distribution generalization of VRPs?", "answer": ["On the Generalization of Neural Combinatorial Optimization Heuristics", "On First-Order Meta-Learning Algorithms"], "answer_arxiv_id": ["2206.00787", "1803.02999"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_14026"} +{"question": "What are the studies where LID tools like 'cld3' have been employed for language detection?", "answer": ["Natural Language Processing with Small Feed-Forward Networks"], "answer_arxiv_id": ["1708.00214"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_14027"} +{"question": "Which works focus on adaptivity to the norm of the comparator in the context of online optimization?", "answer": ["Coin Betting and Parameter-Free Online Learning", "Black-Box Reductions for Parameter-free Online Learning in Banach Spaces", "Lipschitz and Comparator-Norm Adaptivity in Online Learning"], "answer_arxiv_id": ["1602.04128", "1802.06293", "2002.12242"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_14028"} +{"question": "What works discussed ASGNet that adopts superpixel-guided clustering for few-shot segmentation?", "answer": ["Adaptive Prototype Learning and Allocation for Few-Shot Segmentation"], "answer_arxiv_id": ["2104.01893"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14029"} +{"question": "Which works in robot learning have explored leveraging large language models?", "answer": ["Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition"], "answer_arxiv_id": ["2307.14535"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_14030"} +{"question": "What are some studies that used large language models as evaluation and verification tools?", "answer": ["Solving General Arithmetic Word Problems", "Generate & Rank: A Multi-task Framework for Math Word Problems", "Sparks of Artificial General Intelligence: Early experiments with GPT-4", "Training Verifiers to Solve Math Word Problems", "Faithful Reasoning Using Large Language Models", "REFINER: Reasoning Feedback on Intermediate Representations", "Training language models to follow instructions with human feedback", "GPT-4 Technical Report", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Self-Refine: Iterative Refinement with Self-Feedback", "Teaching Large Language Models to Self-Debug"], "answer_arxiv_id": ["1608.01413", "2109.03034", "2303.12712v5", "2110.14168", "2208.14271", "2304.01904", "2203.02155", "2303.08774", "2203.11171", "2303.17651", "2304.05128v2"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_14031"} +{"question": "Could you give examples of research discussing black-box Membership Inference Attacks?", "answer": ["Membership Inference Attacks Against Machine Learning Models", "ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models", "Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "White-box vs Black-box: Bayes Optimal Strategies for Membership Inference", "Systematic Evaluation of Privacy Risks of Machine Learning Models", "Label-Only Membership Inference Attacks", "Practical Blind Membership Inference Attack via Differential Comparisons", "Demystifying Membership Inference Attacks in Machine Learning as a Service", "ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models"], "answer_arxiv_id": ["1610.05820", "1806.01246", "1709.01604", "1908.11229", "2003.10595", "2007.14321", "2101.01341", "1807.09173", "1806.01246"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_14032"} +{"question": "What two recent works combined NPG with bonus-based optimism to handle exploration?", "answer": ["Optimistic Policy Optimization with Bandit Feedback", "Provably Efficient Exploration in Policy Optimization"], "answer_arxiv_id": ["2002.08243", "1912.05830"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_14033"} +{"question": "What works introduced Reinforcement Learning from Human Feedback (RLHF) for better alignment of Large Language Models?", "answer": ["Deep Reinforcement Learning from Human Preferences", "Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["1706.03741", "1909.08593"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_14034"} +{"question": "What's the study that increases the resolution of the vision backbone during its generative training?", "answer": ["PaLI: A Jointly-Scaled Multilingual Language-Image Model"], "answer_arxiv_id": ["2209.06794v4"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_14035"} +{"question": "Have there been studies that aimed at a similar hindsight problem motivated by resource allocation?", "answer": ["Hindsight Learning for MDPs with Exogenous Inputs"], "answer_arxiv_id": ["2207.06272"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_14036"} +{"question": "Which works study the MAB problems in different online settings?", "answer": ["Cascading Bandits: Learning to Rank in the Cascade Model", "Causal Bandits: Learning Good Interventions via Causal Inference", "Linear Contextual Bandits with Knapsacks", "Product Ranking for Revenue Maximization with Multiple Purchases"], "answer_arxiv_id": ["1502.02763", "1606.03203", "1507.06738", "2210.08268"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14037"} +{"question": "Are there any works that pre-train the diffusion model on domain-specific images to improve efficiency?", "answer": ["Taming Encoder for Zero Fine-tuning Image Customization with Text-to-Image Diffusion Models", "InstantBooth: Personalized Text-to-Image Generation without Test-Time Finetuning"], "answer_arxiv_id": ["2304.02642", "2304.03411"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_14038"} +{"question": "Is there any work indicating that parameters can move away from their initial configuration in the mean-field regime and learn useful features?", "answer": ["On Lazy Training in Differentiable Programming"], "answer_arxiv_id": ["1812.07956"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_14039"} +{"question": "In which studies is the text-guided version of image-inpainting developed?", "answer": ["SmartBrush: Text and Shape Guided Object Inpainting with Diffusion Model", "Imagen Editor and EditBench: Advancing and Evaluating Text-Guided Image\n Inpainting", "Uni-paint: A Unified Framework for Multimodal Image Inpainting with\n Pretrained Diffusion Model"], "answer_arxiv_id": ["2212.05034", "2212.06909", "2310.07222"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_14040"} +{"question": "What works have attempted to make predictions on in-distribution generalization with the use of unlabeled data?", "answer": ["Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach", "RATT: Leveraging Unlabeled Data to Guarantee Generalization", "Distributional Generalization: A New Kind of Generalization", "Assessing Generalization of SGD via Disagreement"], "answer_arxiv_id": ["1705.07086", "2105.00303", "2009.08092", "2106.13799"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_14041"} +{"question": "Could you provide me some papers about Spatio-Temporal Graph Neural Networks (STGNNs)?", "answer": ["Deep Learning for Spatio-Temporal Data Mining: A Survey", "Spatio-Temporal Graph Neural Networks for Predictive Learning in Urban Computing: A Survey"], "answer_arxiv_id": ["1906.04928", "2303.14483"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_14042"} +{"question": "Which works have proven that there could be no relation between the explanation methods (E) and the model's decision (Y)?", "answer": ["A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations", "Rethinking the Role of Gradient-based Attribution Methods for Model Interpretability"], "answer_arxiv_id": ["1805.07039", "2006.09128"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_14043"} +{"question": "Could you give me examples of research that proposed combining information from frames for video-based 3D HPS estimation?", "answer": ["Beyond Static Features for Temporally Consistent 3D Human Pose and Shape\n from a Video", "Capturing Humans in Motion: Temporal-Attentive 3D Human Pose and Shape\n Estimation from Monocular Video", "Encoder-decoder with Multi-level Attention for 3D Human Shape and Pose\n Estimation", "Global-to-Local Modeling for Video-based 3D Human Pose and Shape\n Estimation"], "answer_arxiv_id": ["2011.08627", "2203.08534", "2109.02303", "2303.14747"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_14044"} +{"question": "Can you provide examples of studies that adopt the encoder-decoder architecture for infilling tasks?", "answer": ["MASS: Masked Sequence to Sequence Pre-training for Language Generation", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer"], "answer_arxiv_id": ["1905.02450", "1910.13461", "1910.10683"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_14045"} +{"question": "What particular works concentrate on video object segmentation (VOS)?", "answer": ["Self-supervised Video Object Segmentation by Motion Grouping", "MAST: A Memory-Augmented Self-Supervised Tracker", "Self-supervised Learning for Video Correspondence Flow", "Video Object Segmentation Without Temporal Information", "Video Object Segmentation using Space-Time Memory Networks", "Learning to Segment Moving Objects", "FEELVOS: Fast End-to-End Embedding Learning for Video Object Segmentation", "Tracking Emerges by Colorizing Videos", "Anchor Diffusion for Unsupervised Video Object Segmentation"], "answer_arxiv_id": ["2104.07658", "2002.07793", "1905.00875", "1709.06031v2", "1904.00607", "1712.01127", "1902.09513", "1806.09594", "1910.10895"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_14046"} +{"question": "What research works exist about model refinement as a fast sampling method based on extra training or optimization?", "answer": ["Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow"], "answer_arxiv_id": ["2209.03003"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_14047"} +{"question": "Who worked on the fusion-encoder architecture in VLP field?", "answer": ["LXMERT: Learning Cross-Modality Encoder Representations from Transformers", "Align before Fuse: Vision and Language Representation Learning with Momentum Distillation"], "answer_arxiv_id": ["1908.07490", "2107.07651"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14048"} +{"question": "Which works explored the relationship between score matching and Fisher information, and Langevin dynamics?", "answer": ["Bayesian model comparison with the Hyvärinen score: computation and consistency"], "answer_arxiv_id": ["1711.00136"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_14049"} +{"question": "Which work is related to identifying a near-optimal item under a random utility-based discrete choice model?", "answer": ["Best-item Learning in Random Utility Models with Subset Choices"], "answer_arxiv_id": ["2002.07994"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_14050"} +{"question": "Any works that propose to swap with the same person’s identity to reach a high-realistic face-swapping for detecting deepfakes?", "answer": ["Detecting Deepfakes with Self-Blended Images"], "answer_arxiv_id": ["2204.08376"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_14051"} +{"question": "Any research on applying VLMs to cross-modal retrieval tasks?", "answer": ["Masked Contrastive Pre-Training for Efficient Video-Text Retrieval", "WenLan: Bridging Vision and Language by Large-Scale Multi-Modal Pre-Training"], "answer_arxiv_id": ["2212.00986", "2103.06561"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_14052"} +{"question": "Can you provide papers on using Siamese networks for visual object tracking?", "answer": ["Fully-Convolutional Siamese Networks for Object Tracking", "Siamese Instance Search for Tracking", "Fast Online Object Tracking and Segmentation: A Unifying Approach", "Siamese Box Adaptive Network for Visual Tracking", "Ocean: Object-aware Anchor-free Tracking", "Learn to Match: Automatic Matching Network Design for Visual Tracking", "ATOM: Accurate Tracking by Overlap Maximization", "Learning Discriminative Model Prediction for Tracking", "Probabilistic Regression for Visual Tracking"], "answer_arxiv_id": ["1606.09549", "1605.05863", "1812.05050", "2003.06761", "2006.10721", "2108.00803", "1811.07628", "1904.07220", "2003.12565"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_14053"} +{"question": "Which publications did the researcher refer to when discussing learning the continuous and discrete part of their process?", "answer": ["A Continuous Time Framework for Discrete Denoising Models"], "answer_arxiv_id": ["2205.14987"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14054"} +{"question": "Which studies propose that larger models are better at memorization?", "answer": ["Quantifying Memorization Across Neural Language Models"], "answer_arxiv_id": ["2202.07646"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_14055"} +{"question": "What are the Transformers-based methods introduced for HSI reconstruction?", "answer": ["Mask-guided Spectral-wise Transformer for Efficient Hyperspectral Image\n Reconstruction", "Coarse-to-Fine Sparse Transformer for Hyperspectral Image Reconstruction", "Degradation-Aware Unfolding Half-Shuffle Transformer for Spectral\n Compressive Imaging", "Residual Degradation Learning Unfolding Framework with Mixing Priors\n across Spectral and Spatial for Compressive Spectral Imaging", "Pixel Adaptive Deep Unfolding Transformer for Hyperspectral Image\n Reconstruction"], "answer_arxiv_id": ["2111.07910", "2203.04845", "2205.10102", "2211.06891", "2308.10820"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_14056"} +{"question": "Can you mention any work about pretrained autoregressive transformers in the domain of text-to-image generation?", "answer": ["Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers"], "answer_arxiv_id": ["2102.12092", "2105.13290"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_14057"} +{"question": "Could you provide me with studies about improving one-shot models with Normalizing flow models?", "answer": ["MoFlow: An Invertible Flow Model for Generating Molecular Graphs", "GraphNVP: An Invertible Flow Model for Generating Molecular Graphs"], "answer_arxiv_id": ["2006.10137", "1905.11600"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_14058"} +{"question": "Are there any works that show direct application of invariant learning to complicated non-Euclidean molecular structure does not yield promising results?", "answer": ["DrugOOD: Out-of-Distribution (OOD) Dataset Curator and Benchmark for AI-aided Drug Discovery – A Focus on Affinity Prediction Problems with Noise Annotations", "Learning Causally Invariant Representations for Out-of-Distribution Generalization on Graphs"], "answer_arxiv_id": ["2201.09637", "2202.05441"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_14059"} +{"question": "Which works proposed models that generate the pseudo segmentation for training by solving alignment objectives through CTC?", "answer": ["Connectionist Temporal Modeling for Weakly Supervised Action Labeling"], "answer_arxiv_id": ["1607.08584"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_14060"} +{"question": "Can you name some studies that proposed the use of large kernel depthwise convolution?", "answer": ["Visual Attention Network"], "answer_arxiv_id": ["2202.09741"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_14061"} +{"question": "What is the reference for the GradCAM method that operates at intermediate network layers?", "answer": ["Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization"], "answer_arxiv_id": ["1610.02391"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_14062"} +{"question": "What papers proposed modality-agnostic models that can take any modality as input?", "answer": ["VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text", "Perceiver: General Perception with Iterative Attention", "Omnivore: A Single Model for Many Visual Modalities"], "answer_arxiv_id": ["2104.11178", "2103.03206", "2201.08377"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_14063"} +{"question": "What works have included the pessimism principle in their designs to overcome the need for the offline dataset to be highly explorative?", "answer": ["Provably Good Batch Reinforcement Learning Without Great Exploration", "Conservative Q-Learning for Offline Reinforcement Learning", "Is Pessimism Provably Efficient for Offline RL?", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Settling the Sample Complexity of Model-Based Offline Reinforcement Learning", "Near-Optimal Offline Reinforcement Learning via Double Variance Reduction", "Pessimistic Q-Learning for Offline Reinforcement Learning: Towards Optimal Sample Complexity"], "answer_arxiv_id": ["2007.08202", "2006.04779", "2012.15085", "2103.12021v2", "2204.05275", "2102.01748", "2202.13890"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_14064"} +{"question": "Which research papers studied the convergence of gradient descent algorithms focussing on the Bures-Wasserstein barycenter?", "answer": ["Gradient descent algorithms for Bures-Wasserstein barycenters"], "answer_arxiv_id": ["2001.01700"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_14065"} +{"question": "Which paper discusses the role of model size in improving the ability of Large Language Models (LLMs)?", "answer": ["Language Models are Few-Shot Learners", "Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2112.11446", "2204.02311"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_14066"} +{"question": "Which works addressed the issue of image quality and resolution in 3D-aware generation by developing efficient 3D representation?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "GRAM: Generative Radiance Manifolds for 3D-Aware Image Generation", "VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids", "Generative Multiplane Images: Making a 2D GAN 3D-Aware"], "answer_arxiv_id": ["2112.07945", "2112.08867", "2206.07695", "2207.10642"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_14067"} +{"question": "Have there been any studies that propose using inverse dynamics modeling in different settings?", "answer": ["Provable RL with Exogenous Distractors via Multistep Inverse Dynamics", "Agent-Controller Representations: Principled Offline RL with Rich Exogenous Information", "Curiosity-driven Exploration by Self-supervised Prediction", "Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos", "Multi-Environment Pretraining Enables Transfer to Action Limited Datasets"], "answer_arxiv_id": ["2110.08847", "2211.00164v2", "1705.05363", "2206.11795", "2211.13337"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14068"} +{"question": "What researches predict heatmaps at the pixel level in images for possible future paths?", "answer": ["Looking to Relations for Future Trajectory Forecast", "From Goals, Waypoints & Paths To Long Term Human Trajectory Forecasting"], "answer_arxiv_id": ["1905.08855", "2012.01526"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_14069"} +{"question": "Which papers focused on the exploration of 1-bit Transformers?", "answer": ["BinaryBERT: Pushing the Limit of BERT Quantization", "BiBERT: Accurate Fully Binarized BERT", "BiT: Robustly Binarized Multi-distilled Transformer"], "answer_arxiv_id": ["2012.15701", "2203.06390", "2205.13016"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_14070"} +{"question": "What previous empirical studies have investigated the impact of different factors on domain generalization datasets?", "answer": ["A Fine-Grained Analysis on Distribution Shift", "Reappraising Domain Generalization in Neural Networks"], "answer_arxiv_id": ["2110.11328", "2110.07981"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_14071"} +{"question": "Are there any research papers on incorporating robustness against various factors in Reinforcement Learning?", "answer": ["Who Is the Strongest Enemy? Towards Optimal and Efficient Evasion Attacks in Deep RL", "Adversarial Robust Deep Reinforcement Learning Requires Redefining Robustness", "Action Robust Reinforcement Learning and Applications in Continuous Control", "Reinforcement Learning with Perturbed Rewards", "Model-based Adversarial Meta-Reinforcement Learning", "Maximum Entropy RL (Provably) Solves Some Robust RL Problems", "Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Robust Adversarial Reinforcement Learning", "Robust Reinforcement Learning as a Stackelberg Game via Adaptively-Regularized Adversarial Training"], "answer_arxiv_id": ["2106.05087", "2301.07487", "1901.09184", "1810.01032", "2006.08875", "2103.06257", "1703.06907", "1703.02702", "2202.09514"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_14072"} +{"question": "Could you provide me some works that evaluate InfoNCE from a geometric perspective of the embedding space?", "answer": ["Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere", "Towards the Generalization of Contrastive Self-Supervised Learning", "Understanding Dimensional Collapse in Contrastive Self-supervised Learning"], "answer_arxiv_id": ["2005.10242", "2111.00743", "2110.09348"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_14073"} +{"question": "Which papers established benchmarks for evaluating question-answering tasks?", "answer": ["SQuAD: 100,000+ Questions for Machine Comprehension of Text", "Think you have Solved Question Answering? Try ARC, the AI2 Reasoning\n Challenge", "Breaking NLI Systems with Sentences that Require Simple Lexical\n Inferences"], "answer_arxiv_id": ["1606.05250v3", "1803.05457", "1805.02266"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_14074"} +{"question": "Is there any research that introduces positional encodings based on heat kernels and other graph kernels?", "answer": ["Weisfeiler and Leman Go Infinite: Spectral and Combinatorial Pre-Colorings"], "answer_arxiv_id": ["2201.13410"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_14075"} +{"question": "What research has been done on defense strategies for enhancing the robustness of neural networks?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Theoretically Principled Trade-off between Robustness and Accuracy", "Adversarial Weight Perturbation Helps Robust Generalization", "Adversarial Training for Free!"], "answer_arxiv_id": ["1706.06083", "1901.08573", "2004.05884", "1904.12843"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_14076"} +{"question": "Could you tell me about any papers that explore secure aggregation and differential privacy in the context of Federated Learning?", "answer": ["Towards Federated Learning at Scale: System Design", "FedML: A Research Library and Benchmark for Federated Machine Learning", "Can You Really Backdoor Federated Learning?", "Learning Differentially Private Recurrent Language Models"], "answer_arxiv_id": ["1902.01046v2", "2007.13518", "1911.07963", "1710.06963"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_14077"} +{"question": "What research has been done with the application of visual navigation?", "answer": ["Semantic Visual Navigation by Watching YouTube Videos"], "answer_arxiv_id": ["2006.10034"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14078"} +{"question": "Can you list the works that incorporate additional constraints into the loss function in regularization-based CL methods?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Learning without Forgetting", "Orthogonal Gradient Descent for Continual Learning"], "answer_arxiv_id": ["1612.00796", "1606.09282", "1910.07104"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_14079"} +{"question": "What research has been conducted to examine if ImageNet classification accuracy correlates with accuracy as measured using higher-quality human labels?", "answer": ["From ImageNet to Image Classification: Contextualizing Progress on Benchmarks", "Are we done with ImageNet?"], "answer_arxiv_id": ["2005.11295", "2006.07159"], "source_meta": {"published_time": "20230111"}, "qid": "AutoScholarQuery_train_14080"} +{"question": "Which research papers propose latent continuous-time autoencoder models relevant to this work?", "answer": ["Neural Ordinary Differential Equations", "ODE2VAE: Deep generative second order ODEs with Bayesian neural networks"], "answer_arxiv_id": ["1806.07366", "1905.10994"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_14081"} +{"question": "Which studies deal with the use of GRUs, LSTMs, and DeepAR for modelling sequence data?", "answer": ["Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling", "DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks"], "answer_arxiv_id": ["1412.3555", "1704.04110"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_14082"} +{"question": "What papers introduced a new approach to counterfactual datasets, involving generating new documents altogether?", "answer": ["Hallucination Augmented Recitations for Language Models"], "answer_arxiv_id": ["2311.07424"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14083"} +{"question": "What studies aim to improve local training to tackle data heterogeneity challenge in Federated Learning?", "answer": ["Model-Contrastive Federated Learning", "Federated Optimization in Heterogeneous Networks", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Federated Learning Based on Dynamic Regularization", "Federated Learning via Posterior Averaging: A New Perspective and Practical Algorithms", "Addressing Class Imbalance in Federated Learning"], "answer_arxiv_id": ["2103.16257", "1812.06127", "1910.06378", "2111.04263", "2010.05273v4", "2008.06217"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_14084"} +{"question": "Which works contribute to the field of Referring Expression Segmentation (RES), a process targeting segments in an image based on language expressions?", "answer": ["ImageSpirit: Verbal Guided Image Parsing", "Segmentation from Natural Language Expressions", "YouRefIt: Embodied Reference Understanding with Language and Gesture", "Generation and Comprehension of Unambiguous Object Descriptions"], "answer_arxiv_id": ["1310.4389", "1603.06180", "2109.03413", "1511.02283"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_14085"} +{"question": "What research papers focus on representation learning under multi-modal correspondence?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Vision-Language Pre-Training with Triple Contrastive Learning", "FLAVA: A Foundational Language And Vision Alignment Model"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2202.10401", "2112.04482"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_14086"} +{"question": "What work has been done on RL techniques specifically for machine translation?", "answer": ["Adversarial Neural Machine Translation", "Revisiting the Weaknesses of Reinforcement Learning for Neural Machine Translation"], "answer_arxiv_id": ["1704.06933", "2106.08942"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_14087"} +{"question": "Could you give me examples of research papers that attempted to improve both speed and accuracy of object detection?", "answer": ["Path Aggregation Network for Instance Segmentation", "CSPNet: A New Backbone that can Enhance Learning Capability of CNN", "RepVGG: Making VGG-style ConvNets Great Again"], "answer_arxiv_id": ["1803.01534", "1911.11929", "2101.03697"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_14088"} +{"question": "What papers adopted the variance reduction and momentum techniques into stochastic bilevel programming to achieve better complexity results?", "answer": ["Provably Faster Algorithms for Bilevel Optimization", "A Near-Optimal Algorithm for Stochastic Bilevel Optimization via Double-Momentum", "A framework for bilevel optimization that enables stochastic and global variance reduction algorithms"], "answer_arxiv_id": ["2106.04692", "2102.07367", "2201.13409"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_14089"} +{"question": "Can you provide some examples of supervised approaches trained on synthetic data?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "Kubric: A scalable dataset generator", "A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation"], "answer_arxiv_id": ["1504.06852", "2203.03570", "1512.02134"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_14090"} +{"question": "Which study proposed using handcrafted features for small scale problems to improve the utility-privacy tradeoff?", "answer": ["Differentially Private Learning Needs Better Features (or Much More Data)"], "answer_arxiv_id": ["2011.11660"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_14091"} +{"question": "What works are associated with meta-learning a learning rule that allows RNNs to modify its parameters at each time step?", "answer": ["Using Fast Weights to Attend to the Recent Past", "Differentiable plasticity: training plastic neural networks with backpropagation"], "answer_arxiv_id": ["1610.06258", "1804.02464"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_14092"} +{"question": "What studies gave an asymptotic last-iterate convergence result for Optimistic Mirror Worst-case Update (OMWU) in Normal-form Games (NFGs)?", "answer": ["Last-Iterate Convergence: Zero-Sum Games and Constrained Min-Max Optimization"], "answer_arxiv_id": ["1807.04252"], "source_meta": {"published_time": "20220619"}, "qid": "AutoScholarQuery_train_14093"} +{"question": "Are there any papers about compositional counting?", "answer": ["On the Practical Computational Power of Finite Precision RNNs for\n Language Recognition"], "answer_arxiv_id": ["1805.04908"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_14094"} +{"question": "What works proposed leveraging the neural renderer directly instead of handcrafting the rendering equations?", "answer": ["Neural Volumes: Learning Dynamic Renderable Volumes from Images", "SSN: Soft Shadow Network for Image Compositing", "Controllable Shadow Generation Using Pixel Height Maps", "PixHt-Lab: Pixel Height Based Light Effect Generation for Image\n Compositing"], "answer_arxiv_id": ["1906.07751", "2007.08211", "2207.05385", "2303.00137"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_14095"} +{"question": "Which studies try to reduce the discrepancy of local training tasks among clients in regards to pFL methods?", "answer": ["On Bridging Generic and Personalized Federated Learning for Image Classification"], "answer_arxiv_id": ["2107.00778"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_14096"} +{"question": "Are there any studies that leverage weight averaging along the training trajectory for intermediate performance speedups?", "answer": ["Stop Wasting My Time! Saving Days of ImageNet and BERT Training with Latest Weight Averaging"], "answer_arxiv_id": ["2209.14981"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_14097"} +{"question": "What research studies touch upon accuracy prediction under distribution shift?", "answer": ["Are Labels Always Necessary for Classifier Accuracy Evaluation?", "Predicting with Confidence on Unseen Distributions", "Leveraging Unlabeled Data to Predict Out-of-Distribution Performance"], "answer_arxiv_id": ["2007.02915", "2107.03315", "2201.04234"], "source_meta": {"published_time": "20220220"}, "qid": "AutoScholarQuery_train_14098"} +{"question": "Which datasets include large-scale videos of 3D hands assembling several objects?", "answer": ["Assembly101: A Large-Scale Multi-View Video Dataset for Understanding Procedural Activities"], "answer_arxiv_id": ["2203.14712v2"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_14099"} +{"question": "Which works introduced neural tangent kernels (NTKs) and proved its constancy property in the infinite width limit?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_14100"} +{"question": "What papers propose novel approaches for semantic feature extraction and prediction in 3D semantic and instance segmentation?", "answer": ["PointConv: Deep Convolutional Networks on 3D Point Clouds", "KPConv: Flexible and Deformable Convolution for Point Clouds", "PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on\n Point Clouds", "Submanifold Sparse Convolutional Networks", "3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "Stratified Transformer for 3D Point Cloud Segmentation"], "answer_arxiv_id": ["1811.07246", "1904.08889", "2103.14635", "1706.01307", "1711.10275", "1904.08755", "2203.14508"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_14101"} +{"question": "What work attempted to utilize a 4D representation for video editing?", "answer": ["CoDeF: Content Deformation Fields for Temporally Consistent Video\n Processing"], "answer_arxiv_id": ["2308.07926"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_14102"} +{"question": "Can you identify a paper that describes the use of a StyleGAN2-like generator in the EG3D framework?", "answer": ["Analyzing and Improving the Image Quality of StyleGAN"], "answer_arxiv_id": ["1912.04958"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_14103"} +{"question": "Which studies belong to the line of research developing better Transformer-based architectures for Visual Relationship Detection tasks?", "answer": ["Reformulating HOI Detection as Adaptive Set Prediction", "SGTR: End-to-end Scene Graph Generation with Transformer", "Iterative Scene Graph Generation", "Visual Relationship Detection Using Part-and-Sum Transformers with\n Composite Queries", "HOTR: End-to-End Human-Object Interaction Detection with Transformers", "QPIC: Query-Based Pairwise Human-Object Interaction Detection with\n Image-Wide Contextual Information", "Mining the Benefits of Two-stage and One-stage HOI Detection", "GEN-VLKT: Simplify Association and Enhance Interaction Understanding for\n HOI Detection", "Consistency Learning via Decoding Path Augmentation for Transformers in\n Human Object Interaction Detection"], "answer_arxiv_id": ["2103.05983", "2112.12970", "2207.13440", "2105.02170", "2104.13682", "2103.05399", "2108.05077", "2203.13954", "2204.04836"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_14104"} +{"question": "Which studies replaced traditional autoregressive entropy models with checkerboard-based designs to improve the efficacy of the entropy coding?", "answer": ["Checkerboard Context Model for Efficient Learned Image Compression", "ELIC: Efficient Learned Image Compression with Unevenly Grouped\n Space-Channel Contextual Adaptive Coding"], "answer_arxiv_id": ["2103.15306", "2203.10886"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14105"} +{"question": "What papers showcased the new capabilities of Language Model-based Learning (LLM)?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Teaching Algorithmic Reasoning via In-context Learning", "Large Language Models Can Self-Improve", "Large Language Models Are Reasoning Teachers", "Language Models Can Teach Themselves to Program Better"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2211.09066", "2210.11610", "2212.10071", "2207.14502"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_14106"} +{"question": "What papers have advocated the use of task interpolation strategy in few-task meta-learning?", "answer": ["Meta-Learning with Fewer Tasks through Task Interpolation", "Set-based Meta-Interpolation for Few-Task Meta-Learning"], "answer_arxiv_id": ["2106.02695", "2205.09990"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_14107"} +{"question": "Which works give a brief introduction to representation learning for nodes in supervised or semi-supervised classification tasks?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Graph Attention Networks", "Inductive Representation Learning on Large Graphs", "Simplifying Graph Convolutional Networks"], "answer_arxiv_id": ["1609.02907", "1710.10903", "1706.02216", "1902.07153"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_14108"} +{"question": "Which studies proposed using exam tasks for assessing a model's knowledge and expertise?", "answer": ["Measuring Massive Multitask Language Understanding", "AGIEval: A Human-Centric Benchmark for Evaluating Foundation Models", "C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for\n Foundation Models"], "answer_arxiv_id": ["2009.03300", "2304.06364", "2305.08322"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_14109"} +{"question": "Which studies utilise Follow-The-Regularized-Leader (FTRL) and Online Mirror Descent (OMD) approaches for achieving a BOBW guarantee?", "answer": ["A Second-order Bound with Excess Losses", "More Adaptive Algorithms for Adversarial Bandits", "Tsallis-INF: An Optimal Algorithm for Stochastic and Adversarial Bandits"], "answer_arxiv_id": ["1402.2044", "1801.03265", "1807.07623"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14110"} +{"question": "What works discussed objective function weights in PINN method?", "answer": ["Meta-learning PINN loss functions"], "answer_arxiv_id": ["2107.05544"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_14111"} +{"question": "Which research papers proposed effective parametric human prior models for monocular human reconstruction?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "STAR: Sparse Trained Articulated Human Body Regressor"], "answer_arxiv_id": ["1904.05866", "2008.08535"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_14112"} +{"question": "What papers introduce Fourier neural operators (FNOs)?", "answer": ["Neural Operator: Graph Kernel Network for Partial Differential Equations", "Fourier Neural Operator with Learned Deformations for PDEs on General Geometries", "Neural Operator: Learning Maps Between Function Spaces With Applications to PDEs"], "answer_arxiv_id": ["2003.03485", "2207.05209", "2108.08481"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_14113"} +{"question": "Which studies use reinforcement learning for vision tasks?", "answer": ["Tuning computer vision models with task rewards", "HIVE: Harnessing Human Feedback for Instructional Visual Editing"], "answer_arxiv_id": ["2302.08242", "2303.09618"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_14114"} +{"question": "Could you provide some studies about data watermarking methods on Diffusion Models?", "answer": ["A Recipe for Watermarking Diffusion Models", "DiffusionShield: A Watermark for Copyright Protection against Generative\n Diffusion Models", "Tree-Ring Watermarks: Fingerprints for Diffusion Images that are\n Invisible and Robust"], "answer_arxiv_id": ["2303.10137", "2306.04642", "2305.20030"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_14115"} +{"question": "Are there any studies discuss about the optimal policy needing to cover only the dataset's optimal policy in single-agent bandits and reinforcement learning?", "answer": ["Is Pessimism Provably Efficient for Offline RL?", "Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning", "Bellman-consistent Pessimism for Offline Reinforcement Learning"], "answer_arxiv_id": ["2012.15085", "2103.12021v2", "2106.04895", "2106.06926"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_14116"} +{"question": "Which models introduced the concept of Q-Former to align visual and linguistic modalities?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2301.12597"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_14117"} +{"question": "Which papers utilized reinforcement learning for protein sequence design?", "answer": ["Biological Sequence Design with GFlowNets"], "answer_arxiv_id": ["2203.04115"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_14118"} +{"question": "Could you provide me with some references of works focusing on regulating the value-aware model error using Spectral Normalization?", "answer": ["Is Model Ensemble Necessary? Model-based RL via a Single Model with Lipschitz Regularized Value Function"], "answer_arxiv_id": ["2302.01244"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14119"} +{"question": "Could you provide me a work that provides a benchmark for pointwise lpsubscript𝑙𝑝l_{p}-robustness of neural networks?", "answer": ["Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks"], "answer_arxiv_id": ["2003.01690"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_14120"} +{"question": "Any studies about forming valid prediction sets with small f-divergence of the discrepancy?", "answer": ["Robust Validation: Confident Predictions Even When Distributions Shift"], "answer_arxiv_id": ["2008.04267v3"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_14121"} +{"question": "What are some works that introduced channel attention modules in Convolutional Neural Networks (CNNs)?", "answer": ["Squeeze-and-Excitation Networks", "ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks"], "answer_arxiv_id": ["1709.01507", "1910.03151"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_14122"} +{"question": "Which publications discuss the photorealistic image generation capabilities of Denoising Diffusion Probabalistic Models (DDPMs)?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Probabilistic Models", "Cascaded Diffusion Models for High Fidelity Image Generation", "Classifier-Free Diffusion Guidance", "Generative Modeling by Estimating Gradients of the Data Distribution", "Elucidating the Design Space of Diffusion-Based Generative Models", "Scaling up GANs for Text-to-Image Synthesis", "VQGAN-CLIP: Open Domain Image Generation and Editing with Natural\n Language Guidance"], "answer_arxiv_id": ["2105.05233", "2006.11239", "2106.15282", "2207.12598", "1907.05600", "2206.00364v2", "2303.05511", "2204.08583"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14123"} +{"question": "Which studies extended vector neurons to SE3 equivariance?", "answer": ["3D Equivariant Graph Implicit Functions"], "answer_arxiv_id": ["2203.17178"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_14124"} +{"question": "Which papers discussed the use of state-of-the-art LLMs?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14125"} +{"question": "Which works construct 3D parametric head models to represent 3D faces?", "answer": ["Learning to Regress 3D Face Shape and Expression from an Image without\n 3D Supervision", "Learning an Animatable Detailed 3D Face Model from In-The-Wild Images", "EMOCA: Emotion Driven Monocular Face Capture and Animation", "Towards Metrical Reconstruction of Human Faces"], "answer_arxiv_id": ["1905.06817", "2012.04012", "2204.11312", "2204.06607"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_14126"} +{"question": "What papers are related to the application of generative adversarial networks (GANs) in the field of text-to-image generation?", "answer": ["Generative Adversarial Text to Image Synthesis", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "DF-GAN: A Simple and Effective Baseline for Text-to-Image Synthesis", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image\n Synthesis"], "answer_arxiv_id": ["1605.05396", "1711.10485", "2008.05865", "1904.01310"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14127"} +{"question": "Which paper introduced the Deep Evidential Regression method for estimating both types of uncertainty in regression tasks?", "answer": ["Deep Evidential Regression"], "answer_arxiv_id": ["1910.02600"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_14128"} +{"question": "Which work first uncovered robust semantic point correspondences in diffusion models?", "answer": ["Emergent Correspondence from Image Diffusion"], "answer_arxiv_id": ["2306.03881"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14129"} +{"question": "What studies have been conducted to compress the VLT models?", "answer": ["Compressing Visual-linguistic Model via Knowledge Distillation", "UPop: Unified and Progressive Pruning for Compressing Vision-Language\n Transformers", "ELIP: Efficient Language-Image Pre-training with Fewer Vision Tokens", "CrossGET: Cross-Guided Ensemble of Tokens for Accelerating\n Vision-Language Transformers"], "answer_arxiv_id": ["2104.02096", "2301.13741", "2309.16738", "2305.17455"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14130"} +{"question": "Which studies have emphasized discriminative audio-visual feature learning through contrastive learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Improved Baselines with Momentum Contrastive Learning", "Learning to Localize Sound Source in Visual Scenes", "Localizing Visual Sounds the Hard Way", "Mix and Localize: Localizing Sound Sources in Mixtures", "A Closer Look at Weakly-Supervised Audio-Visual Source Localization", "Localizing Visual Sounds the Easy Way", "Learning Audio-Visual Source Localization via False Negative Aware\n Contrastive Learning"], "answer_arxiv_id": ["1911.05722", "2002.05709", "2003.04297", "1803.03849", "2104.02691", "2211.15058", "2209.09634", "2203.09324", "2303.11302"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_14131"} +{"question": "Can you list some works on standard text-to-image diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Zero-Shot Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2112.10752", "2102.12092", "2204.06125", "2206.10789"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_14132"} +{"question": "Could you provide me some works that developed models for VideoQA tasks like NExT-QA, iVQA, EgoSchema, and ActivityNet-QA?", "answer": ["NExT-QA:Next Phase of Question-Answering to Explaining Temporal Actions", "Just Ask: Learning to Answer Questions from Millions of Narrated Videos", "EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language\n Understanding", "ActivityNet-QA: A Dataset for Understanding Complex Web Videos via\n Question Answering"], "answer_arxiv_id": ["2105.08276", "2012.00451", "2308.09126", "1906.02467"], "source_meta": {"published_time": "20240409"}, "qid": "AutoScholarQuery_train_14133"} +{"question": "Are there any works about optimizing individual instances over discrete spaces like hypercubes, graphs, and MILP?", "answer": ["Bayesian Optimization of Combinatorial Structures", "R"], "answer_arxiv_id": ["1806.08838", "1210.6589"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_14134"} +{"question": "Which study revisited CNN-based methods by stacking heatmaps as 3D volumes?", "answer": ["Revisiting Skeleton-based Action Recognition"], "answer_arxiv_id": ["2104.13586"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_14135"} +{"question": "What papers are about vision-based foundation models?", "answer": ["Segment Anything", "SegGPT: Segmenting Everything In Context", "Segment Everything Everywhere All at Once"], "answer_arxiv_id": ["2401.14159", "2304.03284", "2304.06718"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14136"} +{"question": "What papers describe Decentralized Parallel SGD?", "answer": ["Mini-batch Stochastic Approximation Methods for Nonconvex Stochastic Composite Optimization", "Asynchronous Parallel Stochastic Gradient for Nonconvex Optimization", "Asynchronous Decentralized Parallel Stochastic Gradient Descent"], "answer_arxiv_id": ["1308.6594", "1506.08272v5", "1710.06952"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_14137"} +{"question": "Can you provide examples of work that proposed to optimize compact weight spaces in T2I models?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion", "SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "LoRA: Low-Rank Adaptation of Large Language Models", "StyleDrop: Text-to-Image Generation in Any Style"], "answer_arxiv_id": ["2212.04488", "2303.11305", "2106.09685", "2306.00983"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_14138"} +{"question": "What research works extended the federated linear bandits to generalized linear bandits?", "answer": ["Communication Efficient Federated Learning for Generalized Linear Bandits"], "answer_arxiv_id": ["2202.01087"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_14139"} +{"question": "What work proposes generating images effectively in latent space to lower computational costs?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_14140"} +{"question": "Which works propose ways to transfer knowledge across tasks in AutoML?", "answer": ["Transfer Learning with Neural AutoML", "Design Space for Graph Neural Networks", "Dataset Condensation with Differentiable Siamese Augmentation", "Task-Adaptive Neural Network Search with Meta-Contrastive Learning"], "answer_arxiv_id": ["1803.02780", "2011.08843", "2102.08259", "2103.01495"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_14141"} +{"question": "Which papers found that the performance of Language Learning Models (LLMs) is poor on abstract induction tasks like the Abstraction and Reasoning Corpus?", "answer": ["Large Language Models as General Pattern Machines"], "answer_arxiv_id": ["2307.04721"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_14142"} +{"question": "Which research works focused on improving accuracy of word distribution prediction through retrieving similar training contexts?", "answer": ["Improving language models by retrieving from trillions of tokens", "Generalization through Memorization: Nearest Neighbor Language Models"], "answer_arxiv_id": ["2112.04426", "1911.00172"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_14143"} +{"question": "Which works depict how to integrate visual features into the linguistic space of LMMs using a Linear Layer?", "answer": ["Linearly Mapping from Image to Text Space", "Visual Instruction Tuning", "A Multi-Modal Context Reasoning Approach for Conditional Inference on\n Joint Textual and Visual Clues"], "answer_arxiv_id": ["2209.15162", "2304.08485", "2305.04530"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_14144"} +{"question": "What study proposes to learn an EBM as a prior model in the latent space of Deep Latent Variable Models (DLVMs)?", "answer": ["Learning Latent Space Energy-Based Prior Model"], "answer_arxiv_id": ["2006.08205"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_14145"} +{"question": "Which papers deal with the challenges of local inconsistent solution and client-drifts in federated optimization?", "answer": ["Federated Learning: Strategies for Improving Communication Efficiency", "From Distributed Machine Learning to Federated Learning: A Survey", "Improving the Model Consistency of Decentralized Federated Learning", "Enhance Local Consistency in Federated Learning: A Multi-Step Inertial Momentum Approach"], "answer_arxiv_id": ["1610.05492", "2104.14362", "2302.04083", "2302.05726"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_14146"} +{"question": "Could you provide me some studies on high probability convergence for Adaptive SGD?", "answer": ["High Probability Bounds for a Class of Nonconvex Algorithms with AdaGrad Stepsize"], "answer_arxiv_id": ["2204.02833"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_14147"} +{"question": "What works focus on novel view image synthesis based on many multi-view ray samples of the scene?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Multiview Neural Surface Reconstruction by Disentangling Geometry and\n Appearance", "Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance\n Fields", "Hybrid Neural Rendering for Large-Scale Scenes with Motion Blur", "NeRO: Neural Geometry and BRDF Reconstruction of Reflective Objects from\n Multiview Images"], "answer_arxiv_id": ["2003.08934", "2103.13415", "2003.09852", "2112.03907", "2304.12652", "2305.17398"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14148"} +{"question": "Which work used 𝔾3,0,0 geometric products to compute rotation-invariant features from small point clouds in machine learning?", "answer": ["Geometric Algebra Attention Networks for Small Point Clouds"], "answer_arxiv_id": ["2110.02393"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_14149"} +{"question": "What works were proposed to handle the data efficiency and scalability issues of Vision Transformer?", "answer": ["Training data-efficient image transformers & distillation through\n attention", "Going deeper with Image Transformers"], "answer_arxiv_id": ["2012.12877", "2103.17239"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_14150"} +{"question": "Which studies have addressed the issue of language models being prone to predict specific labels due to intrinsic bias or demonstration permutations?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models", "Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2102.09690", "2104.08786"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_14151"} +{"question": "What papers have extended the ERM task with the generic convex loss and ridge regularisation in the high-dimensional regime?", "answer": ["High Dimensional Classification via Regularized and Unregularized Empirical Risk Minimization: Precise Error and Optimal Loss", "The role of regularization in classification of high-dimensional noisy Gaussian mixture"], "answer_arxiv_id": ["1905.13742", "2002.11544"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_14152"} +{"question": "Which works explored the curriculum generation methods based on task difficulty?", "answer": ["Automatic Goal Generation for Reinforcement Learning Agents"], "answer_arxiv_id": ["1705.06366"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14153"} +{"question": "In what paper the persistent independent particles (PIPs) are proposed?", "answer": ["Particle Video Revisited: Tracking Through Occlusions Using Point\n Trajectories"], "answer_arxiv_id": ["2204.04153"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_14154"} +{"question": "Which work has shown improvement in the architecture and learning procedure on the CLRS algorithmic benchmark?", "answer": ["A Generalist Neural Algorithmic Learner"], "answer_arxiv_id": ["2209.11142"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_14155"} +{"question": "Could you provide me some works in which interpretability at the WSI-level is primarily achieved through attention maps?", "answer": ["A graph-transformer for whole slide image classification", "Weakly Supervised Joint Whole-Slide Segmentation and Classification in\n Prostate Cancer"], "answer_arxiv_id": ["2205.09671", "2301.02933"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_14156"} +{"question": "What studies introduced benchmarks for bilingual lexicon induction?", "answer": ["Word translation without parallel data"], "answer_arxiv_id": ["1710.04087"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_14157"} +{"question": "What studies consider non-parametric mixing and linear Gaussian latent causal model?", "answer": ["Learning Linear Causal Representations from Interventions under General Nonlinear Mixing"], "answer_arxiv_id": ["2306.02235"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14158"} +{"question": "What studies investigated large models for tasks like language translation, video generation, and image retrieval?", "answer": ["VideoFusion: Decomposed Diffusion Models for High-Quality Video Generation", "Granularity-aware Adaptation for Image Retrieval over Multiple Tasks"], "answer_arxiv_id": ["2303.08320", "2210.02254"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_14159"} +{"question": "Which work introduced Noisy Nodes, a method that used denoising noisy node information as an auxiliary task for addressing oversmoothing in GNNs?", "answer": ["Simple GNN Regularisation for 3D Molecular Property Prediction & Beyond"], "answer_arxiv_id": ["2106.07971"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_14160"} +{"question": "Could you provide me studies that attempt to obtain interaction through attention-based or GNN-based method?", "answer": ["Multi-Head Attention for Multi-Modal Joint Vehicle Motion Forecasting", "Social Attention: Modeling Attention in Human Crowds", "SCOUT: Socially-COnsistent and UndersTandable Graph Attention Network for Trajectory Prediction of Vehicles and VRUs", "Spectral Temporal Graph Neural Network for Trajectory Prediction", "LaneRCNN: Distributed Representations for Graph-Centric Motion Forecasting", "Implicit Latent Variable Model for Scene-Consistent Motion Forecasting", "Learning Lane Graph Representations for Motion Forecasting", "VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representation"], "answer_arxiv_id": ["1910.03650", "1710.04689", "2102.06361", "2106.02930", "2101.06653", "2007.12036", "2007.13732", "2005.04259"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_14161"} +{"question": "What work proves that gradient descent beyond Edge of Stability (EoS) follows an optimization trajectory subjected to a sharpness constraint so that a flatter region is found?", "answer": ["Self-Stabilization: The Implicit Bias of Gradient Descent at the Edge of Stability"], "answer_arxiv_id": ["2209.15594"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_14162"} +{"question": "In what paper was Edge Convolution, a variant of graph neural network, introduced?", "answer": ["Dynamic Graph CNN for Learning on Point Clouds"], "answer_arxiv_id": ["1801.07829"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_14163"} +{"question": "What work proposed a metric to indicate an early point to end the dense model training and start pruning-retraining?", "answer": ["When to Prune? A Policy towards Early Structural Pruning"], "answer_arxiv_id": ["2110.12007"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_14164"} +{"question": "What papers extend neural relational graphs and relational inference to incorporate latent interaction variables?", "answer": ["Neural Relational Inference for Interacting Systems"], "answer_arxiv_id": ["1802.04687"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_14165"} +{"question": "Which work studied the repercussions stemming from the presence of reasoning shortcuts in multi-hop knowledge editing?", "answer": ["MQuAKE: Assessing Knowledge Editing in Language Models via Multi-Hop\n Questions"], "answer_arxiv_id": ["2305.14795"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_14166"} +{"question": "Are there any studies to demonstrate that in the early phase of training, using lower learning rates may result in worse conditioning of kernel and Hessian matrices?", "answer": ["The break-even point on optimization trajectories of deep neural networks"], "answer_arxiv_id": ["2002.09572"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_14167"} +{"question": "Which works propose methods to improve the efficacy and data efficiency in training vision-language models?", "answer": ["SLIP: Self-supervision meets Language-Image Pre-training", "Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm", "FILIP: Fine-grained Interactive Language-Image Pre-Training", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models"], "answer_arxiv_id": ["2112.12750", "2110.05208", "2111.07783", "2205.01917", "2111.07991", "2201.12086", "2301.12597"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_14168"} +{"question": "What previous works hold significant contributions in enhancing the transferability of adversarial examples by modifying the gradient computation on substitute models?", "answer": ["Backpropagating Linearly Improves Transferability of Adversarial Examples"], "answer_arxiv_id": ["2012.03528"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_14169"} +{"question": "Could you provide the studies making use of small language models to measure prompt mutual information or perplexity, finding the highest-scoring elements?", "answer": ["Unlocking Context Constraints of LLMs: Enhancing Context Efficiency of\n LLMs with Self-Information-Based Content Filtering", "LLMLingua: Compressing Prompts for Accelerated Inference of Large\n Language Models"], "answer_arxiv_id": ["2304.12102", "2310.05736"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_train_14170"} +{"question": "Which works deal with direct learned inversion methods in the context of neural adjoint methods?", "answer": ["Bathymetry Inversion using a Deep-Learning-Based Surrogate for Shallow Water Equations Solvers"], "answer_arxiv_id": ["2203.02821"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_14171"} +{"question": "Could you provide me with some works on refining prompting strategies or fine-tuning models using datasets annotated with explicit reasoning steps?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Recursion of Thought: A Divide-and-Conquer Approach to Multi-Context Reasoning with Language Models", "MetaMath: Bootstrap Your Own Mathematical Questions for Large Language\n Models", "MAmmoTH: Building Math Generalist Models through Hybrid Instruction\n Tuning"], "answer_arxiv_id": ["2203.11171", "2305.10601", "2306.06891v1", "2309.12284", "2309.05653"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_14172"} +{"question": "Could you provide me some studies on token pruning in Vision Transformers?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification", "AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "TokenLearner: What Can 8 Learned Tokens Do for Images and Videos?", "Evo-ViT: Slow-Fast Token Evolution for Dynamic Vision Transformer", "Token Merging: Your ViT But Faster", "Prune Spatio-temporal Tokens by Semantic-aware Temporal Accumulation"], "answer_arxiv_id": ["2106.02034", "2111.15668", "2106.11297", "2108.01390", "2210.09461", "2308.04549"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_14173"} +{"question": "Could you provide me some studies that have been conducted on the permutation symmetries of neurons ?", "answer": ["Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Weight-space symmetry in deep networks gives rise to permutation saddles, connected by equal-loss valleys across the loss landscape", "Optimizing Mode Connectivity via Neuron Alignment", "The Role of Permutation Invariance in Linear Mode Connectivity of Neural Networks", "Git Re-Basin: Merging Models Modulo Permutation Symmetries"], "answer_arxiv_id": ["1802.10026", "1907.02911", "2009.02439", "2110.06296", "2209.04836"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_14174"} +{"question": "Who conducted empirical analysis on models that lack linear connectivity and found their different generalization behaviors?", "answer": ["Linear Connectivity Reveals Generalization Strategies"], "answer_arxiv_id": ["2205.12411"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_14175"} +{"question": "What works propose the MLE method for learning neural ODEs for generative models?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_14176"} +{"question": "Could you provide me some works that studied Reward-Conditioned Reinforcement Learning?", "answer": ["Offline Reinforcement Learning as One Big Sequence Modeling Problem", "Decision Transformer: Reinforcement Learning via Sequence Modeling", "Reward-Conditioned Policies", "RvS: What is Essential for Offline RL via Supervised Learning?", "Training Agents using Upside-Down Reinforcement Learning", "Is Conditional Generative Modeling all you need for Decision-Making?"], "answer_arxiv_id": ["2106.02039", "2106.01345", "1912.13465", "2112.10751", "1912.02877", "2211.15657"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_14177"} +{"question": "Any works about using point cloud templates across various point cloud tasks?", "answer": ["Shape Prior Deformation for Categorical 6D Object Pose and Size\n Estimation", "Category-Level 6D Object Pose and Size Estimation using Self-Supervised\n Deep Prior Deformation Networks", "Learning 3D Object Shape and Layout without 3D Supervision", "Towards High-Fidelity Single-view Holistic Reconstruction of Indoor\n Scenes", "Learning 3D Scene Priors with 2D Supervision", "Mask3D: Mask Transformer for 3D Semantic Instance Segmentation", "SemAffiNet: Semantic-Affine Transformation for Point Cloud Segmentation"], "answer_arxiv_id": ["2007.08454", "2207.05444", "2206.07028", "2207.08656", "2211.14157", "2210.03105", "2205.13490"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14178"} +{"question": "What works showed that unrelated tasks in Multitask Learning may have a negative impact on each other when trained together?", "answer": ["Taskonomy: Disentangling Task Transfer Learning"], "answer_arxiv_id": ["1804.08328"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_14179"} +{"question": "Any studies has proposed different acquisition functions for Bayesian Optimization?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design", "Entropy Search for Information-Efficient Global Optimization", "Predictive Entropy Search for Efficient Global Optimization of Black-box Functions"], "answer_arxiv_id": ["0912.3995v4", "1112.1217v1", "1406.2541"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_14180"} +{"question": "Which papers discussed the counterfactual risk minimization methods?", "answer": ["Counterfactual Risk Minimization: Learning from Logged Bandit Feedback", "Empirical Bernstein Bounds and Sample Variance Penalization"], "answer_arxiv_id": ["1502.02362", "0907.3740"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_14181"} +{"question": "What works employ the transformer as backbone to extract deep features in transformer-based semantic segmentation methods?", "answer": ["SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers", "Multi-Scale High-Resolution Vision Transformer for Semantic Segmentation", "Vision Transformer Adapter for Dense Predictions", "Content-aware Token Sharing for Efficient Semantic Segmentation with\n Vision Transformers"], "answer_arxiv_id": ["2105.15203", "2111.01236", "2205.08534", "2306.02095"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14182"} +{"question": "Which research has demonstrated the effectiveness of quantization for various convolutional neural networks (CNNs)?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding", "Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference", "Data-Free Quantization Through Weight Equalization and Bias Correction", "HAQ: Hardware-Aware Automated Quantization with Mixed Precision", "MCUNet: Tiny Deep Learning on IoT Devices"], "answer_arxiv_id": ["1510.00149", "1712.05877", "1906.04721", "1811.08886", "2007.10319"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_14183"} +{"question": "Can you provide me the work on the construction of a unified embedding space for texts and images using textual prompts?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_14184"} +{"question": "Could you provide some works that focused on language modeling of CAD sketches?", "answer": ["Computer-Aided Design as Language", "Vitruvion: A Generative Model of Parametric CAD Sketches"], "answer_arxiv_id": ["2105.02769v1", "2109.14124"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14185"} +{"question": "Which dataset involves logical reasoning questions collected from the National Civil Servants Examination of China?", "answer": ["LogiQA: A Challenge Dataset for Machine Reading Comprehension with\n Logical Reasoning"], "answer_arxiv_id": ["2007.08124"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_14186"} +{"question": "Which studies have been conducted on making deep neural models inherently explainable?", "answer": ["This Looks Like That: Deep Learning for Interpretable Image Recognition", "Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNet", "Convolutional Dynamic Alignment Networks for Interpretable Classifications", "B-cos Networks: Alignment is All We Need for Interpretability", "Deformable ProtoPNet: An Interpretable Image Classifier Using Deformable Prototypes", "A Framework for Learning Ante-hoc Explainable Models via Concepts", "A Framework to Learn with Interpretation"], "answer_arxiv_id": ["1806.10574", "1904.00760", "2104.00032", "2205.10268", "2111.15000", "2108.11761", "2010.09345"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_14187"} +{"question": "What paper proposed the mean of medians method corresponding to heavy-tailed bandit?", "answer": ["Breaking the Moments Condition Barrier: No-Regret Algorithm for Bandits with Super Heavy-Tailed Payoffs"], "answer_arxiv_id": ["2110.13876"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_14188"} +{"question": "Can you name some works focused on using a dynamic decoder to focus on important regions from multiple feature levels?", "answer": ["Dynamic Head: Unifying Object Detection Heads with Attentions"], "answer_arxiv_id": ["2106.08322"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_14189"} +{"question": "What works investigated single-view reconstruction with meshes?", "answer": ["GAMesh: Guided and Augmented Meshing for Deep Point Networks"], "answer_arxiv_id": ["2010.09774"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_14190"} +{"question": "Which work introduced a deformation network to model the motion of Gaussians in dynamic 3D-GS?", "answer": ["Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene\n Reconstruction"], "answer_arxiv_id": ["2309.13101"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_14191"} +{"question": "Could you provide me researches in low-resource ASR methodologies that utilized fine-tuning high-resource ASR models and self-supervised speech models?", "answer": ["Multilingual sequence-to-sequence speech recognition: architecture,\n transfer learning, and language modeling", "wav2vec 2.0: A Framework for Self-Supervised Learning of Speech\n Representations", "XLS-R: Self-supervised Cross-lingual Speech Representation Learning at\n Scale"], "answer_arxiv_id": ["1810.03459", "2006.11477", "2111.09296"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_14192"} +{"question": "What papers focus on the analysis of the convergence of NPG and Q-NPG for log-linear policies?", "answer": ["Linear Convergence for Natural Policy Gradient with Log-linear Policy Parametrization", "Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies"], "answer_arxiv_id": ["2209.15382", "2210.01400v3"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_14193"} +{"question": "What works searched network architectures to enhance efficiency in model computation?", "answer": ["Improving Transformer Models by Reordering their Sublayers"], "answer_arxiv_id": ["1911.03864"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_14194"} +{"question": "Which paper studied the adaptive adjustment of the clipping threshold of DP-SGD during training?", "answer": ["AdaCliP: Adaptive Clipping for Private SGD", "Private Adaptive Gradient Methods for Convex Optimization"], "answer_arxiv_id": ["1908.07643", "2106.13756"], "source_meta": {"published_time": "20221203"}, "qid": "AutoScholarQuery_train_14195"} +{"question": "Could you provide me some studies that utilized pre-trained networks for extracting instance features in MIL for WSI classification?", "answer": ["Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learning", "DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning for Histopathology Whole Slide Image Classification", "TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classification", "Bi-directional Weakly Supervised Knowledge Distillation for Whole Slide Image Classification", "Scaling Vision Transformers to Gigapixel Images via Hierarchical Self-Supervised Learning", "Self-Supervised Vision Transformers Learn Visual Concepts in Histopathology"], "answer_arxiv_id": ["2011.08939", "2203.12081", "2106.00908", "2210.03664", "2206.02647", "2203.00585"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14196"} +{"question": "Which study first implemented hard-label textual adversarial attack method?", "answer": ["Generating Natural Language Attacks in a Hard Label Black Box Setting"], "answer_arxiv_id": ["2012.14956"], "source_meta": {"published_time": "20240202"}, "qid": "AutoScholarQuery_train_14197"} +{"question": "Which studies specifically focus on quantizing a pre-trained diffusion model without re-training?", "answer": ["Post-training Quantization on Diffusion Models", "Q-Diffusion: Quantizing Diffusion Models", "Brecq: pushing the limit of post-training quantization by block reconstruction"], "answer_arxiv_id": ["2211.15736", "2302.04304", "2102.05426"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_14198"} +{"question": "Which studies have proposed nucleus and typical decoding as example of sampling methods?", "answer": ["The Curious Case of Neural Text Degeneration", "Locally Typical Sampling"], "answer_arxiv_id": ["1904.09751", "2202.00666"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_14199"} +{"question": "What papers have contributions to the Instance Attribution method within the field of neurons network?", "answer": ["Input Similarity from the Neural Network Perspective", "Estimating Training Data Influence by Tracing Gradient Descent"], "answer_arxiv_id": ["2102.05262", "2002.08484"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_14200"} +{"question": "Which language models were reported to be successful in real-world programming competition?", "answer": ["Competition-Level Code Generation with AlphaCode", "Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2203.07814", "2107.03374"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_14201"} +{"question": "Could you provide some studies about text-driven image generation and processing that use the CLIP model?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2103.17249", "2110.02711", "2112.10752"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14202"} +{"question": "Which studies proposed handling larger, unbounded scenes and unconstrained photo collections with the NeRF method?", "answer": ["NeRF++: Analyzing and Improving Neural Radiance Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale Scene Rendering", "Block-NeRF: Scalable Large Scene Neural View Synthesis", "NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections"], "answer_arxiv_id": ["2010.07492", "2111.12077", "2112.05504", "2202.05263", "2008.02268"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_14203"} +{"question": "What research investigated benign overfitting of neural networks?", "answer": ["On the Multiple Descent of Minimum-Norm Interpolants and Restricted Lower Isometry of Kernels", "The Neural Tangent Kernel in High Dimensions: Triple Descent and a Multi-Scale Theory of Generalization", "Towards an Understanding of Benign Overfitting in Neural Networks", "The Interpolation Phase Transition in Neural Networks: Memorization and Generalization under Lazy Training", "Deep Linear Networks can Benignly Overfit when Shallow Ones Do"], "answer_arxiv_id": ["1908.10292", "2008.06786", "2106.03212v1", "2007.12826", "2209.09315"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_14204"} +{"question": "Could you provide me some works studying concept drift?", "answer": ["New Analysis and Algorithm for Learning with Drifting Distributions"], "answer_arxiv_id": ["1205.4343"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_14205"} +{"question": "Can you provide some studies where parameter-free methods are developed, particularly in the deterministic setting?", "answer": ["Making SGD Parameter-Free", "DoG is SGD’s Best Friend: A Parameter-Free Dynamic Step Size Schedule", "Learning-Rate-Free Learning by D-Adaptation", "Adaptive Gradient Descent without Descent", "Adaptive proximal algorithms for convex optimization under local Lipschitz continuity of the gradient"], "answer_arxiv_id": ["2205.02160", "2302.12022", "2301.07733v5", "1910.09529", "2301.04431v4"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14206"} +{"question": "What research focuses on proceeding with model selection using an efficient online learning regime?", "answer": ["NASOA: Towards Faster Task-oriented Online Fine-tuning with a Zoo of\n Models"], "answer_arxiv_id": ["2108.03434"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14207"} +{"question": "Could you provide some studies that incorporate light field rendering with volumetric integration to achieve high fidelity rendering?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering", "Neural Lumigraph Rendering", "Light Field Neural Rendering", "Learning Neural Light Fields with Ray-Space Embedding Networks", "NeX: Real-time View Synthesis with Neural Basis Expansion"], "answer_arxiv_id": ["2102.13090", "2103.11571", "2112.09687", "2112.01523", "2103.05606"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_14208"} +{"question": "Which works proposed methods to construct spatially hierarchical GNN framework for joint representation learning on tissue and cell structures?", "answer": ["HACT-Net: A Hierarchical Cell-to-Tissue Graph Neural Network for\n Histopathological Image Classification"], "answer_arxiv_id": ["2007.00584"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14209"} +{"question": "What papers deal with semantics information from different views and conduct coupled contrastive learning for effective domain adaptation?", "answer": ["The Norm Must Go On: Dynamic Unsupervised Domain Adaptation by Normalization", "MetaAlign: Coordinating Domain Alignment and Classification for Unsupervised Domain Adaptation"], "answer_arxiv_id": ["2112.00463", "2103.13575"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_14210"} +{"question": "Can you provide the work that originally discussed the concept of conditional GFlowNet?", "answer": ["GFlowNet Foundations"], "answer_arxiv_id": ["2111.09266"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_14211"} +{"question": "What are the studies that focused on determining the extent to which recourses remain invariant to factors like model choice, data distribution shifts, and perturbations?", "answer": ["On Counterfactual Explanations under Predictive Multiplicity", "Consistent Counterfactuals for Deep Models", "Algorithmic Recourse in the Wild: Understanding the Impact of Data and Model Shifts", "Towards Robust and Reliable Algorithmic Recourse", "Evaluating Robustness of Counterfactual Explanations", "On the Adversarial Robustness of Causal Algorithmic Recourse", "Counterfactual Explanations Can Be Manipulated"], "answer_arxiv_id": ["2006.13132", "2110.03109", "2012.11788", "2102.13620", "2103.02354v3", "2112.11313", "2106.02666"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_14212"} +{"question": "Could you provide me some works that enhance MLLMs' perception of fine-grained information by introducing additional image local feature extraction modules?", "answer": ["NExT-Chat: An LMM for Chat, Detection and Segmentation", "Ferret: Refer and Ground Anything Anywhere at Any Granularity"], "answer_arxiv_id": ["2311.04498", "2310.07704"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_14213"} +{"question": "Which works present on the subsampled Newton methods which require large batch sizes?", "answer": ["Exact and Inexact Subsampled Newton Methods for Optimization", "Convergence of Newton-MR under Inexact Hessian Information", "Convergence rates of sub-sampled Newton methods", "Sub-sampled Cubic Regularization for Non-convex Optimization"], "answer_arxiv_id": ["1609.08502", "1909.06224", "1508.02810", "1705.05933"], "source_meta": {"published_time": "20220717"}, "qid": "AutoScholarQuery_train_14214"} +{"question": "What research work has been conducted around Task and Motion Planning?", "answer": ["Integrated Task and Motion Planning", "Task and Motion Planning with Large Language Models for Object Rearrangement", "ProgPrompt: Generating Situated Robot Task Plans using Large Language Models"], "answer_arxiv_id": ["2010.01083", "2303.06247", "2209.11302"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_14215"} +{"question": "Could you provide me some research that extended tool-use large language models to other modalities and domains?", "answer": ["MM-ReAct : Prompting ChatGPT for Multimodal Reasoning and Action", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models", "GeneGPT: Augmenting Large Language Models with Domain Tools for Improved Access to Biomedical Information"], "answer_arxiv_id": ["2303.11381", "2303.04671", "2304.09667"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_14216"} +{"question": "Are there any research works focused on proposing holistic evaluations for text-to-image models?", "answer": ["Holistic Evaluation of Text-To-Image Models"], "answer_arxiv_id": ["2311.04287"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_14217"} +{"question": "Any works specific to multimodal clustering?", "answer": ["Multimodal Clustering Networks for Self-supervised Learning from\n Unlabeled Videos"], "answer_arxiv_id": ["2104.12671"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_14218"} +{"question": "Whose research introduced a node-absorbing autoregressive diffusion process?", "answer": ["Autoregressive Diffusion Model for Graph Generation"], "answer_arxiv_id": ["2307.08849"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_14219"} +{"question": "Who proposed a soft rasterization pipeline that allows for a better gradient flow?", "answer": ["SynSin: End-to-end View Synthesis from a Single Image"], "answer_arxiv_id": ["1912.08804"], "source_meta": {"published_time": "20220512"}, "qid": "AutoScholarQuery_train_14220"} +{"question": "What works are about applying Implicit Neural Representations (INRs) to represent images?", "answer": ["Implicit Neural Representations with Periodic Activation Functions", "Learning Continuous Image Representation with Local Implicit Image Function"], "answer_arxiv_id": ["2006.09661", "2012.09161"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_14221"} +{"question": "What papers show that adversarial contrastive learning (ACL) has better robustness against adversarial attacks and common corruptions on downstream tasks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks", "Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["1412.6572", "2003.01690", "1903.12261"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_14222"} +{"question": "Can you indicate works that proposed Concept Bottleneck Models?", "answer": ["Concept Bottleneck Models", "Interpretability Beyond Classification Output: Semantic Bottleneck Networks"], "answer_arxiv_id": ["2007.04612", "1907.10882"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_14223"} +{"question": "What papers have used simulators such as CARLA, SUMO, and Flow for training and evaluating autonomous driving planners?", "answer": ["End-to-end Driving via Conditional Imitation Learning", "Learning by cheating", "Learning to drive from a world on rails"], "answer_arxiv_id": ["1710.02410", "1912.12294", "2105.00636"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_14224"} +{"question": "Which works use YOLOX due to its powerful detection performance in the tracking-by-detection framework?", "answer": ["ByteTrack: Multi-Object Tracking by Associating Every Detection Box", "Observation-Centric SORT: Rethinking SORT for Robust Multi-Object\n Tracking", "StrongSORT: Make DeepSORT Great Again", "Quo Vadis: Is Trajectory Forecasting the Key Towards Long-Term\n Multi-Object Tracking?", "MotionTrack: Learning Robust Short-term and Long-term Motions for\n Multi-Object Tracking", "MOTRv2: Bootstrapping End-to-End Multi-Object Tracking by Pretrained\n Object Detectors"], "answer_arxiv_id": ["2110.06864", "2203.14360", "2202.13514", "2210.07681", "2303.10404", "2211.09791"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14225"} +{"question": "Could you mention some studies that have been ubiquitous in Natural Language Processing (NLP) regarding pretraining?", "answer": ["Distributed Representations of Words and Phrases and their Compositionality", "Semi-supervised Sequence Learning", "Unsupervised Pretraining for Sequence to Sequence Learning", "Deep contextualized word representations", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1310.4546", "1511.01432", "1611.02683", "1802.05365", "1810.04805", "2005.14165"], "source_meta": {"published_time": "20220401"}, "qid": "AutoScholarQuery_train_14226"} +{"question": "Could you mention the works that have found similar phenomena to the research being discussed?", "answer": ["Graph Neural Networks are Inherently Good Generalizers: Insights by Bridging GNNs and MLPs"], "answer_arxiv_id": ["2212.09034"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_14227"} +{"question": "Which work proposed a differentiable constraint on directed acyclic graphs (DAGs)?", "answer": ["DAGs with NO TEARS: Continuous Optimization for Structure Learning"], "answer_arxiv_id": ["1803.01422"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_14228"} +{"question": "What papers propose a neural network algorithm that jointly optimizes imputation and regression?", "answer": ["Linear predictor on linearly-generated data with missing values: non consistency and solutions", "NeuMiss networks: differentiable programming for supervised learning with missing values"], "answer_arxiv_id": ["2002.00658v2", "2007.01627"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14229"} +{"question": "Which studies developed crystal property predictors by representing crystals with chemical formulas?", "answer": ["IRNet: A General Purpose Deep Residual Regression Framework for Materials Discovery", "Predicting materials properties without crystal structure: Deep representation learning from stoichiometry"], "answer_arxiv_id": ["1907.03222v1", "1910.00617"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_14230"} +{"question": "Which paper did ICQ employ for central critic?", "answer": ["QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1803.11485"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_14231"} +{"question": "Could you provide me some papers that enhance the OOD detection by manipulating various aspects?", "answer": ["Mitigating Neural Network Overconfidence with Logit Normalization", "DICE: Leveraging Sparsification for Out-of-Distribution Detection", "Extremely Simple Activation Shaping for Out-of-Distribution Detection"], "answer_arxiv_id": ["2205.09310", "2111.09805", "2209.09858"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_14232"} +{"question": "Which papers discuss methods for improving robustness using methods like minimizing the worst-group loss?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the\n Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14233"} +{"question": "Any works that use the concept of augmentation overlap to formulate how the positive samples are aligned?", "answer": ["Chaos is a Ladder: A New Theoretical Understanding of Contrastive Learning via Augmentation Overlap"], "answer_arxiv_id": ["2203.13457"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_14234"} +{"question": "What research papers contributed to the development of models with very long context window lengths?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14235"} +{"question": "Which papers cover the adaptation and improvement of deep generative models for stochastic human motion forecasting?", "answer": ["The Pose Knows: Video Forecasting by Generating Pose Futures", "MT-VAE: Learning Motion Transformations to Generate Multimodal Human Dynamics", "HP-GAN: Probabilistic 3D human motion prediction via GAN"], "answer_arxiv_id": ["1705.00053", "1808.04545", "1711.09561"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_14236"} +{"question": "Which papers focus on communication efficiency, one of the efficiency aspects of Federated Learning?", "answer": ["Federated Learning With Quantized Global Model Updates", "Communication Efficiency in Federated Learning: Achievements and Challenges", "FedChain: Chained Algorithms for Near-Optimal Communication Cost in Federated Learning", "FedPara: Low-rank Hadamard Product for Communication-Efficient Federated Learning"], "answer_arxiv_id": ["2006.10672", "2107.10996", "2108.06869", "2108.06098"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_14237"} +{"question": "What recent studies indicate that safety alignment from RLHF may come undone with a few examples of finetuning?", "answer": ["Removing RLHF Protections in GPT-4 via Fine-Tuning"], "answer_arxiv_id": ["2311.05553"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_14238"} +{"question": "What papers investigated generating task-specific arithmetic operations for reading comprehension tasks?", "answer": ["Giving BERT a Calculator: Finding Operations and Arguments with Reading Comprehension"], "answer_arxiv_id": ["1909.00109"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_14239"} +{"question": "Which papers have tried to maintain the symmetrical structure of ETF in imbalanced datasets by fixing the classifier as ETF?", "answer": ["Inducing Neural Collapse in Imbalanced Learning: Do We Really Need a Learnable Classifier at the End of Deep Neural Network?", "A Geometric Analysis of Neural Collapse with Unconstrained Features"], "answer_arxiv_id": ["2203.09081", "2105.02375"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_14240"} +{"question": "Which work demonstrates the use of gradient matching in gradient inversion?", "answer": ["Deep Leakage from Gradients"], "answer_arxiv_id": ["1906.08935"], "source_meta": {"published_time": "20220912"}, "qid": "AutoScholarQuery_train_14241"} +{"question": "Which studies focus on the development of novel tabular architectures based on transformer models?", "answer": ["TabTransformer: Tabular Data Modeling Using Contextual Embeddings", "Revisiting Deep Learning Models for Tabular Data", "SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training", "TabNet: Attentive Interpretable Tabular Learning"], "answer_arxiv_id": ["2012.06678", "2106.11959", "2106.01342", "1908.07442"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_14242"} +{"question": "Which papers discuss potential improvements to the over-smoothing effects of Score Distillation Sampling (SDS)?", "answer": ["ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Text-to-3D with Classifier Score Distillation", "Noise-Free Score Distillation"], "answer_arxiv_id": ["2305.16213", "2310.19415", "2310.17590"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_14243"} +{"question": "Can you name some research about model merging for federated learning?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "On the Convergence of FedAvg on Non-IID Data"], "answer_arxiv_id": ["1602.05629", "1907.02189"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14244"} +{"question": "Which research introduced the concept of neural radiance field (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_14245"} +{"question": "What studies proposed an implicit solution for DS problem, using (advantage-weighted) variants of behavioral cloning?", "answer": ["Critic Regularized Regression", "BAIL: Best-Action Imitation Learning for Batch Deep Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["2006.15134", "1910.12179", "2106.06860"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_14246"} +{"question": "Could you provide me some studies about the task of causal language modeling in in-context learning?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models", "LaMDA: Language Models for Dialog Applications"], "answer_arxiv_id": ["2005.14165", "2204.02311", "2302.13971", "2201.08239"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_14247"} +{"question": "What research papers show the development of Fully-connected Graph transformer and its performance improvements?", "answer": ["A Generalization of Transformer Networks to Graphs", "Rethinking Graph Transformers with Spectral Attention", "GraphiT: Encoding Graph Structure in Transformers", "Sign and Basis Invariant Networks for Spectral Graph Representation Learning"], "answer_arxiv_id": ["2012.09699", "2106.03893", "2106.05667", "2202.13013"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_14248"} +{"question": "Could you tell me about the study that constructed the Waterbirds dataset?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_14249"} +{"question": "Which study proposed a differentiable soft quantization approach?", "answer": ["Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit\n Neural Networks"], "answer_arxiv_id": ["1908.05033"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_14250"} +{"question": "Can you provide a paper about overcoming the limitations of lower reconstruction quality and increased model size in neural rendering?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_14251"} +{"question": "What research has been done to control CTC alignment prediction?", "answer": ["Advancing Multi-Accented LSTM-CTC Speech Recognition using a Domain Specific Student-Teacher Learning Paradigm", "Guiding CTC Posterior Spike Timings for Improved Posterior Fusion and Knowledge Distillation", "A New Training Pipeline for an Improved Neural Transducer", "Towards Real-time Mispronunciation Detection in Kids’ Speech"], "answer_arxiv_id": ["1809.06833", "1904.08311", "2005.09319", "2003.01765"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_14252"} +{"question": "Which papers discussed the projection strategies that fail to maintain global coherence?", "answer": ["RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "Improving Diffusion Models for Inverse Problems using Manifold Constraints"], "answer_arxiv_id": ["2201.09865", "2206.00941"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_14253"} +{"question": "What studies have investigated and found these distillation methods may lead to the student mimicking the style of the teacher, but not the reasoning abilities?", "answer": ["Orca: Progressive Learning from Complex Explanation Traces of GPT-4", "The False Promise of Imitating Proprietary LLMs"], "answer_arxiv_id": ["2306.02707v1", "2305.15717"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_14254"} +{"question": "Which paper proposed the MIWAE bound for training DLVMs in the context of missing data under the MAR assumption?", "answer": ["MIWAE: Deep Generative Modelling and Imputation of Incomplete Data Sets"], "answer_arxiv_id": ["1812.02633"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_14255"} +{"question": "Could you provide some works that examined overparameterization in low-rank matrix recovery from a small random initialization with noisy measurements?", "answer": ["Global Convergence of Sub-gradient Method for Robust Matrix Recovery: Small Initialization, Noisy Measurements, and Over-parameterization"], "answer_arxiv_id": ["2202.08788"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_14256"} +{"question": "Which works have brought in visual instruction tuning to improve the instruction-following ability of VLLMs?", "answer": ["GPT-4 Technical Report", "Visual Instruction Tuning", "An Empirical Study of Scaling Instruct-Tuned Large Multimodal Models", "Improved Baselines with Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2303.08774", "2304.08485", "2309.09958", "2310.03744", "2304.10592"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_14257"} +{"question": "Could you provide me with a study that proposes a slot attention-based classifier for transparent and accurate classification?", "answer": ["SCOUTER: Slot Attention-based Classifier for Explainable Image\n Recognition"], "answer_arxiv_id": ["2009.06138"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_14258"} +{"question": "Any studies related to domain adaptation and taxonomies?", "answer": ["TACS: Taxonomy Adaptive Cross-Domain Semantic Segmentation", "Domain Adaptation with Asymmetrically-Relaxed Distribution Alignment", "Graph-Relational Domain Adaptation", "Continuously Indexed Domain Adaptation", "Domain-Indexing Variational Bayes: Interpretable Domain Index for Domain Adaptation", "Collaborative Deep Learning for Recommender Systems", "Towards Bayesian Deep Learning: A Framework and Some Existing Methods", "Bidirectional Inference Networks: A Class of Deep Bayesian Networks for Health Profiling", "A Survey on Bayesian Deep Learning", "Zero-Shot Recommender Systems"], "answer_arxiv_id": ["2109.04813", "1903.01689", "2202.03628", "2007.01807", "2302.02561", "1409.2944", "1608.06884", "1902.02037", "1604.01662", "2105.08318"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_14259"} +{"question": "What studies are about GAN-based unpaired image-to-image translations that map two domains to the same metric space?", "answer": ["One-Sided Unsupervised Domain Mapping", "Geometry-Consistent Generative Adversarial Networks for One-Sided Unsupervised Domain Mapping", "Contrastive Learning for Unpaired Image-to-Image Translation", "The Spatially-Correlative Loss for Various Image Translation Tasks"], "answer_arxiv_id": ["1706.00826v2", "1809.05852", "2007.15651", "2104.00854"], "source_meta": {"published_time": "20230804"}, "qid": "AutoScholarQuery_train_14260"} +{"question": "What studies deal with learning prototypes representing distinctive properties of each class in an unsupervised manner?", "answer": ["Towards Robust Interpretability with Self-Explaining Neural Networks", "This Looks Like That: Deep Learning for Interpretable Image Recognition"], "answer_arxiv_id": ["1806.07538", "1806.10574"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14261"} +{"question": "Which works extended the pre-training of LVLMs with instruction-following data for better performance?", "answer": ["InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning"], "answer_arxiv_id": ["2305.06500"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_14262"} +{"question": "What works have used data augmentation in the context of code generation by synthesizing tests for human-authored code?", "answer": ["Competition-Level Code Generation with AlphaCode", "Leveraging Automated Unit Tests for Unsupervised Code Translation"], "answer_arxiv_id": ["2203.07814", "2110.06773"], "source_meta": {"published_time": "20220729"}, "qid": "AutoScholarQuery_train_14263"} +{"question": "Which papers explored the use of Laplace distribution in probabilistic regression?", "answer": ["Human Pose Regression with Residual Log-likelihood Estimation", "Maximum Likelihood Uncertainty Estimation: Robustness to Outliers"], "answer_arxiv_id": ["2107.11291", "2202.03870"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_14264"} +{"question": "Which EG3D inversion method is used in the paper?", "answer": ["Real-Time Radiance Fields for Single-Image Portrait View Synthesis"], "answer_arxiv_id": ["2305.02310"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14265"} +{"question": "What papers discussed speeding up the training and rendering of NeRFs?", "answer": ["TensoRF: Tensorial Radiance Fields", "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "Neural Sparse Voxel Fields"], "answer_arxiv_id": ["2203.09517", "2103.13744", "2201.05989", "2112.05131", "2007.11571"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14266"} +{"question": "Can you provide some references about using Graph Neural Networks in learning combinatorial algorithms?", "answer": ["GREED: A Neural Framework for Learning Graph Distance Functions", "Lifelong Learning for Neural powered Mixed Integer Programming"], "answer_arxiv_id": ["2112.13143", "2208.12226v3"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_14267"} +{"question": "What papers have enhanced the joint vision-and-language feature space?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "SLIP: Self-supervision meets Language-Image Pre-training", "FLAVA: A Foundational Language And Vision Alignment Model"], "answer_arxiv_id": ["2103.00020", "2201.12086", "2112.12750", "2112.04482"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_14268"} +{"question": "What techniques proposed neural reference-based IQA using direct supervision from human judgments?", "answer": ["The Unreasonable Effectiveness of Deep Features as a Perceptual Metric", "PieAPP: Perceptual Image-Error Assessment through Pairwise Preference", "Image Quality Assessment: Unifying Structure and Texture Similarity", "SinGAN: Learning a Generative Model from a Single Natural Image", "DreamSim: Learning New Dimensions of Human Visual Similarity using\n Synthetic Data"], "answer_arxiv_id": ["1801.03924", "1806.02067", "2004.07728", "1905.01164", "2306.09344"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_14269"} +{"question": "Are there works showcasing emerging properties of large multimodal models?", "answer": ["The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)", "GPT-4V(ision) as a Generalist Evaluator for Vision-Language Tasks", "Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena", "CLIPScore: A Reference-free Evaluation Metric for Image Captioning"], "answer_arxiv_id": ["2309.17421", "2311.01361", "2306.05685", "2104.08718"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_14270"} +{"question": "What studies showcase a combination of both contrastive learning and self-prediction in protein representation?", "answer": ["Protein Representation Learning by Geometric Structure Pretraining"], "answer_arxiv_id": ["2203.06125"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14271"} +{"question": "What study uses Prior Networks for uncertainty estimation in regression tasks?", "answer": ["Regression Prior Networks"], "answer_arxiv_id": ["2006.11590"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_14272"} +{"question": "Which works focus on exploring the abilities to control characteristics in generated levels by utilizing a trained generative model?", "answer": ["Generating and Blending Game Levels via Quality-Diversity in the Latent Space of a Variational Autoencoder", "Mario Plays on a Manifold: Generating Functional Content in Latent Space through Differential Geometry"], "answer_arxiv_id": ["2102.12463", "2206.00106"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_14273"} +{"question": "In the study of Vision Language Models, what papers have explored ways to comprehend visual and language representations?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "Visual Dialog", "Modulating early visual processing by language", "ImageBERT: Cross-modal Pre-training with Large-scale Weak-supervised\n Image-Text Data", "Large-Scale Adversarial Training for Vision-and-Language Representation\n Learning", "ERNIE-ViL: Knowledge Enhanced Vision-Language Representations Through\n Scene Graph", "Unicoder-VL: A Universal Encoder for Vision and Language by Cross-modal\n Pre-training"], "answer_arxiv_id": ["1908.02265", "1611.08669", "1707.00683", "2001.07966", "2006.06195", "2006.16934", "1908.06066"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_14274"} +{"question": "What research appended the soft prompt to the hidden representations at each layer in both the text and image encoder?", "answer": ["MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2210.03117"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_14275"} +{"question": "Which paper demonstrates that SCN has surpassed invariant GNNs on the large-scale OC-20 dataset?", "answer": ["Spherical Channels for Modeling Atomic Interactions"], "answer_arxiv_id": ["2206.14331"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_14276"} +{"question": "Which works proposed extracting 3D dance pose from 2D dance video with tracking and pose estimation tools?", "answer": ["BRACE: The Breakdancing Competition Dataset for Dance Motion Synthesis", "Dance with You: The Diversity Controllable Dancer Generation via\n Diffusion Models", "Music-Driven Group Choreography", "AI Choreographer: Music Conditioned 3D Dance Generation with AIST++"], "answer_arxiv_id": ["2207.10120", "2308.13551", "2303.12337", "2101.08779"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_14277"} +{"question": "What works have demonstrated the importance of cost aggregation modules due to its favorable generalization ability?", "answer": ["AdaStereo: A Simple and Efficient Approach for Adaptive Stereo Matching", "GraftNet: Towards Domain Generalized Stereo Matching with a\n Broad-Spectrum and Task-Oriented Feature"], "answer_arxiv_id": ["2004.04627", "2204.00179"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_14278"} +{"question": "What studies focus on the setting where the total contamination is bound in robust MABs?", "answer": ["Stochastic bandits robust to adversarial corruptions"], "answer_arxiv_id": ["1803.09353"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_14279"} +{"question": "Could you provide me some works about online gradient-based Reinforcement Learning?", "answer": ["Online Decision Transformer"], "answer_arxiv_id": ["2202.05607"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_14280"} +{"question": "Can you list some works that have designed elaborate strategies to replay exemplars and distill knowledge in the CIL scenario?", "answer": ["RMM: Reinforced Memory Management for Class-Incremental Learning", "Class-Incremental Learning by Knowledge Distillation with Adaptive Feature Consolidation", "PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning"], "answer_arxiv_id": ["2301.05792", "2204.00895", "2004.13513"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_14281"} +{"question": "Which work uses intra-class variance from head classes for feature augmentation for tail classes through feature transfer learning?", "answer": ["Feature Transfer Learning for Face Recognition with Under-Represented Data"], "answer_arxiv_id": ["1803.09014"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_14282"} +{"question": "Could you provide me the reference in which prefix tuning was proposed, which is relevant to the transfer learning?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2101.00190"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_14283"} +{"question": "Which research papers explore stereotypes related to demographic identity in multiple areas?", "answer": ["StereoSet: Measuring stereotypical bias in pretrained language models", "CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked\n Language Models", "The Woman Worked as a Babysitter: On Biases in Language Generation"], "answer_arxiv_id": ["2004.09456", "2010.00133", "1909.01326v2"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14284"} +{"question": "What works utilize BERT-style models for retrieval-based dialogue models?", "answer": ["Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based\n Chatbots"], "answer_arxiv_id": ["2004.03588"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_14285"} +{"question": "Which studies introduced two-stage detectors?", "answer": ["Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks", "R-FCN: Object Detection via Region-based Fully Convolutional Networks"], "answer_arxiv_id": ["1506.01497", "1605.06409"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_14286"} +{"question": "Which papers have demonstrated the use of Diffusion Models in tasks such as image generation, restoration, speech generation, and video generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2006.11239", "2011.13456"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_14287"} +{"question": "Which research introduced the use of the Transformer as a hypernetwork in TransINR?", "answer": ["Transformers as Meta-Learners for Implicit Neural Representations", "Attention Is All You Need"], "answer_arxiv_id": ["2208.02801", "1706.03762"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_14288"} +{"question": "Which work also modeled the transition matrices as a third-order tensor but didn't have an ARHMM structure or make connections to the LDS and SLDS?", "answer": ["High-dimensional vector autoregressive time series modeling via tensor decomposition"], "answer_arxiv_id": ["1909.06624"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_14289"} +{"question": "Could you provide me some recent research works on flow matching?", "answer": ["Flow Matching for Generative Modeling", "Building Normalizing Flows with Stochastic Interpolants", "Rectified Flow: A Marginal Preserving Approach to Optimal Transport"], "answer_arxiv_id": ["2210.02747", "2209.15571", "2209.14577v1"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_14290"} +{"question": "Can you provide a study where a learned metric using human annotation was proposed due to the observed poor correlation of BLEURT-20 to human ratings?", "answer": ["TaTa: A Multilingual Table-to-Text Dataset for African Languages"], "answer_arxiv_id": ["2211.00142"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_14291"} +{"question": "What works showed superior results of diffusion models compared to GAN and VAE based methods?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2105.05233", "2207.12598"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14292"} +{"question": "Can you list papers that studied the pure exploration problem in the multi-armed bandit model?", "answer": ["Optimal Best Arm Identification with Fixed Confidence"], "answer_arxiv_id": ["1602.04589v2"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_14293"} +{"question": "What research papers focus on designing and analyzing PnP algorithms?", "answer": ["Plug-and-Play ADMM for Image Restoration: Fixed Point Convergence and\n Applications", "The Little Engine that Could: Regularization by Denoising (RED)", "Learning Deep CNN Denoiser Prior for Image Restoration", "Image Restoration by Iterative Denoising and Backward Projections", "Regularization by Denoising: Clarifications and New Interpretations", "An Online Plug-and-Play Algorithm for Regularized Image Reconstruction"], "answer_arxiv_id": ["1605.01710", "1611.02862", "1704.03264", "1710.06647", "1806.02296", "1809.04693"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_14294"} +{"question": "Which papers constructed task taxonomies for computer-vision tasks based on the tasks' specific representations?", "answer": ["Taskonomy: Disentangling Task Transfer Learning"], "answer_arxiv_id": ["1804.08328"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_14295"} +{"question": "What research papers discuss the technique known as background modeling used for handling label ambiguity in partially labeled data?", "answer": ["Multi-organ Segmentation over Partially Labeled Datasets with\n Multi-scale Feature Abstraction", "Marginal loss and exclusion loss for partially supervised multi-organ\n segmentation"], "answer_arxiv_id": ["2001.00208", "2007.03868"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_14296"} +{"question": "Which work shows how incorporating causal dependencies between environment entities can improve generalization to out-of-distribution states?", "answer": ["Causal Dynamics Learning for Task-Independent State Abstraction"], "answer_arxiv_id": ["2206.13452"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_14297"} +{"question": "What papers explored GAN-based methods for speech synthesis?", "answer": ["HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis", "MelGAN: Generative Adversarial Networks for Conditional Waveform Synthesis"], "answer_arxiv_id": ["2010.05646", "1910.06711"], "source_meta": {"published_time": "20230802"}, "qid": "AutoScholarQuery_train_14298"} +{"question": "Which papers established the foundation for self and cross-attention mechanisms in data aggregation?", "answer": ["Attention Is All You Need", "Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks", "Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "Attention-Based Models for Speech Recognition"], "answer_arxiv_id": ["1706.03762", "1810.00825", "2107.07651", "1506.07503"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_14299"} +{"question": "What research considered variance-constrained best arm identification but with feedback noise depending only on the agent action?", "answer": ["Almost Optimal Variance-Constrained Best Arm Identification"], "answer_arxiv_id": ["2201.10142"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_14300"} +{"question": "Which papers discussed the use of kernel properties for improving the softmax approximation?", "answer": ["Hybrid Random Features", "Chefs’ Random Tables: Non-Trigonometric Random Features"], "answer_arxiv_id": ["2110.04367", "2205.15317"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_14301"} +{"question": "Which studies have used GANs for text guided image generation?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks", "DF-GAN: A Simple and Effective Baseline for Text-to-Image Synthesis", "Improving Text-to-Image Synthesis Using Contrastive Learning", "Cross-Modal Contrastive Learning for Text-to-Image Generation", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image Synthesis"], "answer_arxiv_id": ["1711.10485", "2008.05865", "2107.02423", "2101.04702", "1904.01310"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_14302"} +{"question": "Which papers studied the global convergence of GD on the symmetric matrix sensing with noiseless measurements and overestimated rank?", "answer": ["Algorithmic Regularization in Over-parameterized Matrix Sensing and Neural Networks with Quadratic Activations", "Small random initialization is akin to spectral learning: Optimization and generalization guarantees for overparameterized low-rank matrix reconstruction"], "answer_arxiv_id": ["1712.09203", "2106.15013"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_14303"} +{"question": "Which works have been done to improve the performance of Lasso in situations where some clusters of covariates are highly correlated?", "answer": ["Correlated variables in regression: clustering and sparse estimation"], "answer_arxiv_id": ["1209.5908"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14304"} +{"question": "Which works demonstrate that Transformers can simulate practical computer programs?", "answer": ["Looped Transformers as Programmable Computers", "Statistically Meaningful Approximation: a Case Study on Approximating Turing Machines with Transformers", "Transformers Learn Shortcuts to Automata", "Self-Attention Networks Can Process Bounded Hierarchical Languages"], "answer_arxiv_id": ["2301.13196v1", "2107.13163", "2210.10749", "2105.11115"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_14305"} +{"question": "What research claims that their work does not automatically adapt to the margin region?", "answer": ["Efficient Active Learning with Abstention"], "answer_arxiv_id": ["2204.00043"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_14306"} +{"question": "What works propose methods based on fixed keypoint detection for instance-level 6D object pose estimation?", "answer": ["PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation", "Vote from the Center: 6 DoF Pose Estimation in RGB-D Images by Radial\n Keypoint Voting", "MV6D: Multi-View 6D Pose Estimation on RGB-D Frames Using a Deep\n Point-wise Voting Network", "Keypoint Cascade Voting for Point Cloud Based 6DoF Pose Estimation"], "answer_arxiv_id": ["1812.11788", "2104.02527", "2208.01172", "2210.08123"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_14307"} +{"question": "Which study proposed the Transitive closure time-distillation (TRACT) method?", "answer": ["TRACT: Denoising Diffusion Models with Transitive Closure Time-Distillation"], "answer_arxiv_id": ["2303.04248"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_14308"} +{"question": "What works use pruning as part of a larger interpretability effort?", "answer": ["NeuroSurgeon: A Toolkit for Subnetwork Analysis"], "answer_arxiv_id": ["2309.00244"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_14309"} +{"question": "Which works are about text-guided diffusion models?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2211.01324"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_14310"} +{"question": "Could you provide me some work that analyse the training dynamics in multi-layer linear neural networks?", "answer": ["A Convergence Analysis of Gradient Descent for Deep Linear Neural Networks", "Gradient descent with identity initialization efficiently learns positive definite linear transformations by deep residual networks"], "answer_arxiv_id": ["1810.02281", "1802.06093"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14311"} +{"question": "Which studies conducted a theoretical analysis on reconstruction from a federated-learning setup?", "answer": ["Reconstructing Training Data from Model Gradient, Provably"], "answer_arxiv_id": ["2212.03714v3"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_14312"} +{"question": "Which research works are about soft-parameter sharing in MTL?", "answer": ["Cross-stitch Networks for Multi-task Learning", "Progressive Neural Networks", "PathNet: Evolution Channels Gradient Descent in Super Neural Networks", "End-to-End Multi-Task Learning with Attention", "Attentive Single-Tasking of Multiple Tasks"], "answer_arxiv_id": ["1604.03539", "1606.04671", "1701.08734", "1803.10704", "1904.08918"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14313"} +{"question": "Which works have explored the explainability of temporal graph neural networks?", "answer": ["An Explainer for Temporal Graph Neural Networks"], "answer_arxiv_id": ["2209.00807"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14314"} +{"question": " Which papers are about generative adversarial nets (GANs) which contribute to text-to-image generative models?", "answer": ["Generative Adversarial Networks", "Generative Adversarial Text to Image Synthesis", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked\n Generative Adversarial Networks", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661", "1605.05396", "1612.03242", "1711.10485"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_14315"} +{"question": "What papers considered shuffle DP ERM and SCO?", "answer": ["Encode, Shuffle, Analyze Privacy Revisited: Formalizations and Empirical Evaluation", "Shuffle Private Stochastic Convex Optimization"], "answer_arxiv_id": ["2001.03618", "2106.09805"], "source_meta": {"published_time": "20210617"}, "qid": "AutoScholarQuery_train_14316"} +{"question": "What studies mentioned a source-specific poison-only attack that can reduce latent separation?", "answer": ["Demon in the Variant: Statistical Analysis of DNNs for Robust Backdoor Contamination Detection"], "answer_arxiv_id": ["1908.00686"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_14317"} +{"question": "Which papers have reported the first meaningful MIM visual pre-training results?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_14318"} +{"question": "Are there any papers about the use of OT in long-tailed recognition?", "answer": ["SoLar: Sinkhorn Label Refinery for Imbalanced Partial-Label Learning"], "answer_arxiv_id": ["2209.10365"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_14319"} +{"question": "What is the name of the research that introduced Transformers?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_14320"} +{"question": "Which studies focus on using benchmark functions in black-box optimization?", "answer": ["Bayesian Optimization is Superior to Random Search for Machine Learning Hyperparameter Tuning: Analysis of the Black-Box Optimization Challenge 2020"], "answer_arxiv_id": ["2104.10201"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_14321"} +{"question": "Could you tell me about some research that applied variational mechanisms in text-based motion generation research?", "answer": ["TEMOS: Generating diverse human motions from textual descriptions"], "answer_arxiv_id": ["2204.14109"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_14322"} +{"question": "What papers reflect on the usage of CoT reasoning in the context of LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2201.11903", "2205.11916"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_14323"} +{"question": "Which research proposed the concept of Ordered Neurons often associated with LSTM-based language models?", "answer": ["Ordered Neurons: Integrating Tree Structures into Recurrent Neural Networks"], "answer_arxiv_id": ["1810.09536"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_14324"} +{"question": "What studies have made enhancements in model architecture and layout representation?", "answer": ["HoHoNet: 360 Indoor Holistic Understanding with Latent Horizontal\n Features"], "answer_arxiv_id": ["2011.11498"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_14325"} +{"question": "Which studies generate multi-turn instructions for tuning a chat model?", "answer": ["Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on\n Self-Chat Data", "Enhancing Chat Language Models by Scaling High-quality Instructional Conversations"], "answer_arxiv_id": ["2304.01196", "2305.14233v1"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_14326"} +{"question": "Any studies that conducted experiments on imposing priors on feature visualizations?", "answer": ["Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space"], "answer_arxiv_id": ["1612.00005"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_14327"} +{"question": "Which papers utilized SC2LE for their studies on centralized unit micromanagement in StarCraft?", "answer": ["StarCraft II: A New Challenge for Reinforcement Learning"], "answer_arxiv_id": ["1708.04782"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_14328"} +{"question": "Could you provide me some works where non-local cross-scale attention to explore cross-scale feature correlations are used?", "answer": ["Image Super-Resolution with Cross-Scale Non-Local Attention and\n Exhaustive Self-Exemplars Mining"], "answer_arxiv_id": ["2006.01424"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_14329"} +{"question": "What studies were focused on using the optimization problem for the inversion and reconstruction in GAN inversion?", "answer": ["Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?", "Image2StyleGAN++: How to Edit the Embedded Images?", "Transforming and Projecting Images into Class-conditional Generative Networks", "Inverting The Generator Of A Generative Adversarial Network"], "answer_arxiv_id": ["1904.03189", "1911.11544", "2005.01703", "1802.05701"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_14330"} +{"question": "What works introduce the concept of Zero-Shot Learning (ZSL) where semantic knowledge is transferred from seen classes to unseen ones?", "answer": ["Label-Embedding for Image Classification", "Feature Generating Networks for Zero-Shot Learning", "f-VAEGAN-D2: A Feature Generating Framework for Any-Shot Learning", "FREE: Feature Refinement for Generalized Zero-Shot Learning", "Contrastive Embedding for Generalized Zero-Shot Learning", "EGANS: Evolutionary Generative Adversarial Network Search for Zero-Shot\n Learning", "Align Your Prompts: Test-Time Prompting with Distribution Alignment for\n Zero-Shot Generalization"], "answer_arxiv_id": ["1503.08677", "1712.00981", "1903.10132", "2107.13807", "2103.16173", "2308.09915", "2311.01459"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_14331"} +{"question": "Can you provide me with research papers that utilise information-theoretic approaches to quantify distance to the optimal occupancy measure?", "answer": ["Learning to Optimize Via Information-Directed Sampling", "Regret Bounds for Information-Directed Reinforcement Learning"], "answer_arxiv_id": ["1403.5556", "2206.04640"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_14332"} +{"question": "Can you name some papers discussing semantic segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation", "Learning Deconvolution Network for Semantic Segmentation", "SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image\n Segmentation", "U-Net: Convolutional Networks for Biomedical Image Segmentation", "DeepLab: Semantic Image Segmentation with Deep Convolutional Nets,\n Atrous Convolution, and Fully Connected CRFs", "Multi-Scale Context Aggregation by Dilated Convolutions", "ParseNet: Looking Wider to See Better", "Pyramid Scene Parsing Network", "ICNet for Real-Time Semantic Segmentation on High-Resolution Images", "Dual Attention Network for Scene Segmentation", "CCNet: Criss-Cross Attention for Semantic Segmentation", "Asymmetric Non-local Neural Networks for Semantic Segmentation", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Semi-supervised Semantic Segmentation with Directional Context-aware\n Consistency", "Learning Context-aware Classifier for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038", "1505.04366", "1511.00561", "1505.04597", "1606.00915", "1511.07122", "1506.04579", "1612.01105", "1704.08545", "1809.02983", "1811.11721", "1908.07678", "2107.06278", "2106.14133", "2303.11633"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_14333"} +{"question": "Are there studies that have researched on methods to cause less data loss after compression than quantization methods?", "answer": ["Embedding Compression with Isotropic Iterative Quantization", "Spreading vectors for similarity search"], "answer_arxiv_id": ["2001.05314", "1806.03198"], "source_meta": {"published_time": "20221215"}, "qid": "AutoScholarQuery_train_14334"} +{"question": "What papers provided a theoretical understanding of the benign overfitting phenomenon in linear models through regression problems?", "answer": ["Two models of double descent for weak features", "Surprises in High-Dimensional Ridgeless Least Squares Interpolation", "Benign Overfitting in Linear Regression", "Benign Overfitting of Constant-Stepsize SGD for Linear Regression"], "answer_arxiv_id": ["1903.07571", "1903.08560", "1906.11300", "2103.12692"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_14335"} +{"question": "Which papers analyze neural network expressive power by determining which equivariant polynomials a given architecture can compute?", "answer": ["Deep Sets", "On Universal Equivariant Set Networks", "On the Universality of Invariant Networks", "On Learning Sets of Symmetric Elements", "Can Graph Neural Networks Count Substructures?", "On the Universality of Rotation Equivariant Point Cloud Networks", "Equivariant Polynomials for Graph Neural Networks"], "answer_arxiv_id": ["1703.06114", "1910.02421", "1901.09342", "2002.08599", "2002.04025", "2010.02449", "2302.11556"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14336"} +{"question": "What are some studies that accelerate rendering by intelligent sample placement?", "answer": ["DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields\n using Depth Oracle Networks", "TermiNeRF: Ray Termination Prediction for Efficient Neural Rendering", "AdaNeRF: Adaptive Sampling for Real-time Rendering of Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields"], "answer_arxiv_id": ["2103.03231", "2111.03643", "2207.10312", "2111.12077"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14337"} +{"question": "Could you provide me some studies about using Large Language Models (LLMs) in building generalist VL models?", "answer": ["Scaling Instruction-Finetuned Language Models", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2210.11416", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_14338"} +{"question": "Which work optimizes the rank of states preserving order defined by the true cost-to-goal?", "answer": ["Learning to Rank for Synthesizing Planning Heuristics"], "answer_arxiv_id": ["1608.01302"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14339"} +{"question": "Which works primarily addressed the issue of adversarial examples in neural networks?", "answer": ["Intriguing properties of neural networks", "Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1312.6199", "1412.6572", "1706.06083"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_14340"} +{"question": "What works have made use of the reconstruction of masked subgraphs in molecules for molecular topology pretraining?", "answer": ["Strategies for Pre-training Graph Neural Networks", "GPT-GNN: Generative Pre-Training of Graph Neural Networks"], "answer_arxiv_id": ["1905.12265", "2006.15437"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_14341"} +{"question": "Which work employs textual inversion to extract the concept from a style image?", "answer": ["Inversion-Based Style Transfer with Diffusion Models"], "answer_arxiv_id": ["2211.13203"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14342"} +{"question": "Which works are essentials to improving the efficiency of ODE-based solvers?", "answer": ["DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "Pseudo Numerical Methods for Diffusion Models on Manifolds", "GENIE: Higher-Order Denoising Diffusion Solvers"], "answer_arxiv_id": ["2206.00927", "2202.09778", "2210.05475"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_14343"} +{"question": "Could you provide me some studies that are about improving the compression performance for single videos?", "answer": ["HNeRV: A Hybrid Neural Representation for Videos", "HiNeRV: Video Compression with Hierarchical Encoding-based Neural\n Representation", "FFNeRV: Flow-Guided Frame-Wise Neural Representations for Videos", "PS-NeRV: Patch-wise Stylized Neural Representations for Videos", "E-NeRV: Expedite Neural Video Representation with Disentangled\n Spatial-Temporal Context"], "answer_arxiv_id": ["2304.02633", "2306.09818", "2212.12294", "2208.03742", "2207.08132"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14344"} +{"question": "Which studies have addressed RL with corrupted rewards?", "answer": ["Reinforcement Learning with a Corrupted Reward Channel", "Reinforcement Learning with Perturbed Rewards"], "answer_arxiv_id": ["1705.08417", "1810.01032"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_14345"} +{"question": "Are there any works that proposed the welfare fairness to control the trade-off between fairness and total influence by an inequality aversion parameter?", "answer": ["Fair Influence Maximization: A Welfare Optimization Approach"], "answer_arxiv_id": ["2006.07906"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_14346"} +{"question": "What studies have proposed ways to evaluate the multilingual reasoning capabilities of large language models?", "answer": ["Language Models are Multilingual Chain-of-Thought Reasoners"], "answer_arxiv_id": ["2210.03057"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_14347"} +{"question": "Could you provide me some works related to feature-wise linear modulation layer for deep neural networks?", "answer": ["FiLM: Visual Reasoning with a General Conditioning Layer"], "answer_arxiv_id": ["1709.07871"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_14348"} +{"question": "Which papers are related to program synthesis from the description and I/O pairs or other modality inputs?", "answer": ["Towards Synthesizing Complex Programs from Input-Output Examples", "Leveraging Grammar and Reinforcement Learning for Neural Program Synthesis", "RobustFill: Neural Program Learning under Noisy I/O", "Synthesizing Programs for Images using Reinforced Adversarial Learning", "Latent Programmer: Discrete Latent Codes for Program Synthesis", "CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning", "Learning to Infer and Execute 3D Shape Programs"], "answer_arxiv_id": ["1706.01284v4", "1805.04276", "1703.07469", "1804.01118", "2012.00377", "2207.01780", "1901.02875"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_14349"} +{"question": "Can you name some methods that concurrently estimate depth/normals, reflectance, and illumination from a single scene image?", "answer": ["Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying\n Lighting and SVBRDF from a Single Image", "Learning Indoor Inverse Rendering with 3D Spatially-Varying Lighting", "Learning-based Inverse Rendering of Complex Indoor Scenes with\n Differentiable Monte Carlo Raytracing", "Physically-Based Editing of Indoor Scene Lighting from a Single Image", "IRISformer: Dense Vision Transformers for Single-Image Inverse Rendering\n in Indoor Scenes"], "answer_arxiv_id": ["1905.02722", "2109.06061", "2211.03017", "2205.09343", "2206.08423"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14350"} +{"question": "Could you mention some studies that performed the largest-scale AD benchmark?", "answer": ["ADBench: Anomaly Detection Benchmark"], "answer_arxiv_id": ["2206.09426"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_14351"} +{"question": "What papers discuss introducing additional guidance information in fine-tuning-based TIE methods for image synthesis?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2302.05543", "2302.08453", "2211.09800"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_14352"} +{"question": "Are there any studies on computationally tractable Gaussian process regression similar to the proposed method?", "answer": ["Gaussian Processes for Big Data", "Doubly Sparse Variational Gaussian Processes"], "answer_arxiv_id": ["1309.6835", "2001.05363v1"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_14353"} +{"question": "What papers focused on image editing techniques using cross-attention maps as an editing medium?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Diffusion Self-Guidance for Controllable Image Generation", "Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of Code"], "answer_arxiv_id": ["2208.01626", "2306.00986", "2310.01506"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_14354"} +{"question": "Can you list some of the studies that focus on Markov potential games?", "answer": ["Global Convergence of Multi-Agent Policy Gradient in Markov Potential Games", "Independent Policy Gradient for Large-Scale Markov Potential Games: Sharper Rates, Function Approximation, and Game-Agnostic Convergence", "Gradient play in stochastic games: stationary points, convergence, and sample complexity", "Convergence and Price of Anarchy Guarantees of the Softmax Policy Gradient in Markov Potential Games", "Independent Natural Policy Gradient Always Converges in Markov Potential Games"], "answer_arxiv_id": ["2106.01969", "2202.04129", "2106.00198", "2206.07642", "2110.10614"], "source_meta": {"published_time": "20220803"}, "qid": "AutoScholarQuery_train_14355"} +{"question": "Which studies focused on behavioral analysis of various types of learners in the context of machine teaching?", "answer": ["The Teaching Dimension of Linear Learners", "Teaching Multiple Concepts to a Forgetful Learner", "Towards Black-box Iterative Machine Teaching", "Understanding the Role of Adaptivity in Machine Teaching: The Case of Version Space Learners", "Preference-Based Batch and Sequential Teaching: Towards a Unified View of Models", "Interactive Teaching Algorithms for Inverse Reinforcement Learning", "The Sample Complexity of Teaching-by-Reinforcement on Q-Learning"], "answer_arxiv_id": ["1512.02181", "1805.08322", "1710.07742", "1802.05190", "1910.10944", "1905.11867", "2006.09324"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_14356"} +{"question": "What researchs incorporate morphology information into the controller to improve learning performance?", "answer": ["MetaMorph: Learning Universal Controllers with Transformers", "AnyMorph: Learning Transferable Polices By Inferring Agent Morphology"], "answer_arxiv_id": ["2203.11931", "2206.12279"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_14357"} +{"question": "Which research papers focus on global explainability through the use of knowledge transfer?", "answer": ["Distilling the Knowledge in a Neural Network", "Improving Simple Models with Confidence Profiles", "Enhancing Simple Models by Exploiting What They Already Know"], "answer_arxiv_id": ["1503.02531", "1807.07506", "1905.13565"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_14358"} +{"question": "Which works have proposed solutions to 2D anomaly detection?", "answer": ["EasyNet: An Easy Network for 3D Industrial Anomaly Detection", "Anomaly Detection via Reverse Distillation from One-Class Embedding", "Towards Total Recall in Industrial Anomaly Detection", "Pushing the Limits of Few-shot Anomaly Detection in Industry Vision: Graphcore", "What Makes a Good Data Augmentation for Few-Shot Unsupervised Image Anomaly Detection?", "Segment Any Anomaly without Training via Hybrid Prompt Regularization", "SoftPatch: Unsupervised Anomaly Detection with Noisy Data"], "answer_arxiv_id": ["2307.13925", "2201.10703", "2106.08265", "2301.12082", "2304.03294", "2305.10724", "2403.14233"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_14359"} +{"question": "Which studies proposed modifications to the original model training to make the model more amenable to unlearning?", "answer": ["Unrolling SGD: Understanding Factors Influencing Machine Unlearning"], "answer_arxiv_id": ["2109.13398"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_14360"} +{"question": "What papers are seminal in the field of human pose estimation?", "answer": ["Deep Learning-Based Human Pose Estimation: A Survey"], "answer_arxiv_id": ["2012.13392"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14361"} +{"question": "Which papers introduce spatial-location-dependent weights in relaxed G𝐺G-steerable group convolution?", "answer": ["Approximately Equivariant Networks for Imperfectly Symmetric Dynamics"], "answer_arxiv_id": ["2201.11969"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_14362"} +{"question": "What papers made extensions on the work of bib.bib29?", "answer": ["Training Verifiers to Solve Math Word Problems"], "answer_arxiv_id": ["2110.14168"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_14363"} +{"question": "What research have applied joint embedding-weight tuning methods in concept customization?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2208.12242", "2212.04488"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14364"} +{"question": "What research touched on the limitations of 'hard attention' Transformers in recognizing various formal languages and circuit classes?", "answer": ["Theoretical Limitations of Self-Attention in Neural Sequence Models", "Formal Language Recognition by Hard Attention Transformers: Perspectives from Circuit Complexity"], "answer_arxiv_id": ["1906.06755", "2204.06618"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_14365"} +{"question": "Which works related to DocVQA employ OCR to input both text and layout information?", "answer": ["SCATTER: Selective Context Attentional Scene Text Recognizer", "Sequence-to-Sequence Contrastive Learning for Text Recognition", "TextAdaIN: Paying Attention to Shortcut Learning in Text Recognizers", "Multimodal Semi-Supervised Learning for Text Recognition", "CLIPTER: Looking at the Bigger Picture in Scene Text Recognition"], "answer_arxiv_id": ["2003.11288", "2012.10873", "2105.03906", "2205.03873", "2301.07464v2"], "source_meta": {"published_time": "20240107"}, "qid": "AutoScholarQuery_train_14366"} +{"question": "Could you suggest some datasets for diacritization in classical Arabic?", "answer": ["Arabic Text Diacritization Using Deep Neural Networks"], "answer_arxiv_id": ["1905.01965"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_train_14367"} +{"question": "Could you provide me some studies about methods combining pre-computed solutions and physics informed losses?", "answer": ["Physics-Informed Neural Operator for Learning Partial Differential Equations"], "answer_arxiv_id": ["2111.03794"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_14368"} +{"question": "Which works are about sequence to sequence formulation in many tasks in natural language processing (NLP)?", "answer": ["Sequence to Sequence Learning with Neural Networks"], "answer_arxiv_id": ["1409.3215"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_14369"} +{"question": "What works study the use of pre-trained representations in robot learning algorithms?", "answer": ["R3M: A Universal Visual Representation for Robot Manipulation", "Masked Visual Pre-training for Motor Control", "Reinforcement Learning with Action-Free Pre-Training from Videos", "VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training"], "answer_arxiv_id": ["2203.12601", "2203.06173", "2203.13880", "2210.00030v2"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_14370"} +{"question": "What are the works that modify the sampling grid of the convolution kernel, which relates to content-adaptive networks?", "answer": ["Deformable Convolutional Networks", "Deformable ConvNets v2: More Deformable, Better Results", "InternImage: Exploring Large-Scale Vision Foundation Models with\n Deformable Convolutions", "Dilated convolution with learnable spacings"], "answer_arxiv_id": ["1703.06211", "1811.11168", "2211.05778", "2112.03740"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_14371"} +{"question": "Are there any studies suggesting the use of diffusion models to generate plans in order to solve robotic tasks?", "answer": ["End-to-End Robotic Reinforcement Learning without Reward Engineering"], "answer_arxiv_id": ["1904.07854"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_14372"} +{"question": "Which papers propose the construction of specific multi-stage ML pipelines such as (data generation)–(architecture search)–(classification) and (data reweighting)–(finetuning)–(pretraining)?", "answer": ["Learning from Mistakes - A Framework for Neural Architecture Search"], "answer_arxiv_id": ["2111.06353"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_14373"} +{"question": "What works observed that a fully-connected layer leaks the input through the gradients, leading to linear layer leakage (LLL) attacks?", "answer": ["Rethinking Privacy Preserving Deep Learning: How to Evaluate and Thwart\n Privacy Attacks"], "answer_arxiv_id": ["2006.11601"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_14374"} +{"question": "Which works proposed MCMC samplers that are developed to be sample efficient with gradients?", "answer": ["Stochastic Gradient Hamiltonian Monte Carlo", "Preconditioned Stochastic Gradient Langevin Dynamics for Deep Neural Networks", "Global Convergence of Langevin Dynamics Based Algorithms for Nonconvex Optimization"], "answer_arxiv_id": ["1402.4102", "1512.07666", "1707.06618"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_14375"} +{"question": "What study formulates an optimization problem with the goal of obtaining misclassification with the fewest possible perturbations in adversarial attacks?", "answer": ["Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1608.04644"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_14376"} +{"question": "What works involved forcing Q values to be low for Out Of Distribution (OOD) state-action pairs?", "answer": ["Bridging the Gap Between Value and Policy Based Reinforcement Learning", "Conservative Q-Learning for Offline Reinforcement Learning", "COMBO: Conservative Offline Model-Based Policy Optimization"], "answer_arxiv_id": ["1702.08892", "2006.04779", "2102.08363"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_14377"} +{"question": "Who introduced the problem of Proof-of-Learning?", "answer": ["Proof-of-Learning: Definitions and Practice"], "answer_arxiv_id": ["2103.05633v1"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_14378"} +{"question": "What research works enhance vision perception tasks by utilizing diffusion-generated images?", "answer": ["Not Just Pretty Pictures: Toward Interventional Data Augmentation Using\n Text-to-Image Generators", "Leaving Reality to Imagination: Robust Classification via Generated\n Datasets", "Synthetic Data from Diffusion Models Improves ImageNet Classification", "StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual\n Representation Learners"], "answer_arxiv_id": ["2212.11237", "2302.02503", "2304.08466", "2306.00984"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_14379"} +{"question": "Could you provide me works on the development of long-form video understanding datasets?", "answer": ["Towards Long-Form Video Understanding", "Online Model Distillation for Efficient Video Inference", "Generic Event Boundary Detection: A Benchmark for Event Segmentation"], "answer_arxiv_id": ["2106.11310", "1812.02699v2", "2101.10511"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_14380"} +{"question": "What study uses a generative model of the behavior policy for sampling perturbed actions during policy optimization?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration"], "answer_arxiv_id": ["1812.02900"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_14381"} +{"question": "What are some of the early works that utilize auxiliary information in Zero-Shot Learning (ZSL)?", "answer": ["Latent Embeddings for Zero-shot Classification"], "answer_arxiv_id": ["1603.08895"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14382"} +{"question": "Could you give me examples of research that used maximum softmax probability as the metric to detect OOD samples?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks"], "answer_arxiv_id": ["1610.02136"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_14383"} +{"question": "What recent method proposes to bake any NeRF model into a mesh structure to achieve real-time rendering?", "answer": ["NeRFMeshing: Distilling Neural Radiance Fields into\n Geometrically-Accurate 3D Meshes"], "answer_arxiv_id": ["2303.09431"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_14384"} +{"question": "Could you provide me some studies about extending DINO to utilize patch-level representations?", "answer": ["Efficient Self-supervised Vision Transformers for Representation Learning"], "answer_arxiv_id": ["2106.09785"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_14385"} +{"question": "Could you provide me with examples of works that proposed proposal-learnable methods to adaptively predict video segments?", "answer": ["Adaptive Proposal Generation Network for Temporal Sentence Localization in Videos", "Multilevel Language and Vision Integration for Text-to-Clip Retrieval", "Natural Language Video Localization with Learnable Moment Proposals", "QVHighlights: Detecting Moments and Highlights in Videos via Natural Language Queries", "UMT: Unified Multi-modal Transformers for Joint Video Moment Retrieval and Highlight Detection"], "answer_arxiv_id": ["2109.06398", "1804.05113", "2109.10678", "2107.09609", "2203.12745"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_14386"} +{"question": "Could you cite the works related to multi-agent task solving paradigm?", "answer": ["Generative Agents: Interactive Simulacra of Human Behavior", "Agents: An Open-source Framework for Autonomous Language Agents", "ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate", "GameGPT: Multi-agent Collaborative Framework for Game Development", "LM vs LM: Detecting Factual Errors via Cross Examination", "MetaAgents: Simulating Interactions of Human Behaviors for LLM-based\n Task-oriented Coordination via Collaborative Generative Agents"], "answer_arxiv_id": ["2304.03442", "2309.07870", "2308.07201", "2310.08067", "2305.13281", "2310.06500"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_14387"} +{"question": "What study proposed the first label-only Model Inversion using a black-box Boundary Repelling search?", "answer": ["Label-Only Model Inversion Attacks via Boundary Repulsion"], "answer_arxiv_id": ["2203.01925"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14388"} +{"question": "Could you provide me with the references about using question-answering capabilities to boost EAE?", "answer": ["Event Extraction by Answering (Almost) Natural Questions", "Prompt for Extraction? PAIE: Prompting Argument Interaction for Event\n Argument Extraction", "Event Extraction as Question Generation and Answering"], "answer_arxiv_id": ["2004.13625", "2202.12109", "2307.05567"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_14389"} +{"question": "Which works discuss the development of single-encoder multi-head models?", "answer": ["Med3D: Transfer Learning for 3D Medical Image Analysis", "Continual Segment: Towards a Single, Unified and Accessible Continual\n Segmentation Model of 143 Whole-body Organs in CT Scans"], "answer_arxiv_id": ["1904.00625", "2302.00162"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_14390"} +{"question": "What research propose the utilization of large kernels in convolution?", "answer": ["Going deeper with convolutions", "Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning", "Large Kernel Matters —— Improve Semantic Segmentation by Global Convolutional Network", "Rethinking the Inception Architecture for Computer Vision", "Very Deep Convolutional Networks for Large-Scale Image Recognition", "Involution: Inverting the Inherence of Convolution for Visual Recognition", "On the Connection between Local Attention and Dynamic Depth-wise Convolution", "A ConvNet for the 2020s", "Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs", "CKConv: Continuous Kernel Convolution For Sequential Data", "Towards a General Purpose CNN for Long Range Dependencies in ND"], "answer_arxiv_id": ["1409.4842", "1602.07261", "1703.02719", "1512.00567", "1409.1556", "2103.06255", "2106.04263", "2201.03545", "2203.06717", "2102.02611", "2206.03398"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_14391"} +{"question": "What research provided insights about the relationship between flatness of the minimum and its generalization capability?", "answer": ["Exploring Generalization in Deep Learning", "Averaging Weights Leads to Wider Optima and Better Generalization", "Fantastic Generalization Measures and Where to Find Them"], "answer_arxiv_id": ["1706.08947", "1803.05407", "1912.02178"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_14392"} +{"question": "Which studies exhibit black-box attacks on machine learning algorithms for membership inference?", "answer": ["Membership Inference Attacks Against Machine Learning Models", "Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting", "Evaluating Differentially Private Machine Learning in Practice", "Enhanced Membership Inference Attacks against Machine Learning Models"], "answer_arxiv_id": ["1610.05820", "1709.01604", "1902.08874", "2111.09679"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14393"} +{"question": "Could you provide me studies that researched implicit bias in two-layer leaky-ReLU networks with linearly-separable data?", "answer": ["Gradient Descent on Two-layer Nets: Margin Maximization and Simplicity Bias", "Towards Understanding Learning in Neural Networks with Linear Teachers", "Implicit Bias in Leaky ReLU Networks Trained on High-Dimensional Data"], "answer_arxiv_id": ["2110.13905", "2101.02533", "2210.07082"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_14394"} +{"question": "What research papers highlighted that pruning often achieves best results and dynamic sparse training methods like SET and RigL improve over static sparse training?", "answer": ["The State of Sparse Training in Deep Reinforcement Learning"], "answer_arxiv_id": ["2206.10369"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_14395"} +{"question": "Which paper found sets of parameters in which any critical point is a global minimizer and any outside critical point is a saddle point?", "answer": ["Global optimality conditions for deep neural networks"], "answer_arxiv_id": ["1707.02444"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_14396"} +{"question": "What studies have been shown to exploit character-aware language models for image generation?", "answer": ["Character-Aware Models Improve Visual Text Rendering"], "answer_arxiv_id": ["2212.10562"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14397"} +{"question": "Can you cite some studies that explored the emulation of iterations of proximal gradient descent using neural networks?", "answer": ["DeepSTI: Towards Tensor Reconstruction using Fewer Orientations in Susceptibility Tensor Imaging"], "answer_arxiv_id": ["2209.04504"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_14398"} +{"question": "What studies have explored performance variations over a broad model scale and its applications in LLMs?", "answer": ["OPT: Open Pre-trained Transformer Language Models", "GLM-130B: An Open Bilingual Pre-trained Model"], "answer_arxiv_id": ["2205.01068", "2210.02414"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_14399"} +{"question": "What research utilized a multi-layer multiplicative filter network for signal reconstruction?", "answer": ["BACON: Band-limited Coordinate Networks for Multiscale Scene\n Representation"], "answer_arxiv_id": ["2112.04645"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_14400"} +{"question": "Could you provide me some studies that explored attacks on gradient-based explanations?", "answer": ["Interpretation of Neural Networks is Fragile", "Explanations can be manipulated and geometry is to blame", "Fooling Neural Network Interpretations via Adversarial Model Manipulation"], "answer_arxiv_id": ["1710.10547", "1906.07983", "1902.02041"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_14401"} +{"question": "What studies connect ridge regression with early stopping in gradient descent and its variants?", "answer": ["Iterate averaging as regularization for stochastic gradient descent", "The Implicit Regularization of Stochastic Gradient Flow for Least Squares"], "answer_arxiv_id": ["1802.08009", "2003.07802"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14402"} +{"question": "What papers discuss debiasing models according to individual fairness criteria?", "answer": ["Training individually fair ML models with Sensitive Subspace Robustness", "Individual Fairness Guarantees for Neural Networks", "CertiFair: A Framework for Certified Global Fairness of Neural Networks"], "answer_arxiv_id": ["1907.00020", "2205.05763", "2205.09927"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_14403"} +{"question": "What studies have explored accelerated algorithms with communication compression?", "answer": ["Acceleration for Compressed Gradient Descent in Distributed and Federated Optimization", "CANITA: Faster Rates for Distributed Convex Optimization with Communication Compression", "Error Compensated Distributed SGD Can Be Accelerated"], "answer_arxiv_id": ["2002.11364", "2107.09461", "2010.00091"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14404"} +{"question": "Which work introduced the concept of neural collapse?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning\n training"], "answer_arxiv_id": ["2008.08186"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_14405"} +{"question": "Could you provide me some studies which developed various variations of chain-of-thought to prompt models’ reasoning ability?", "answer": ["Large Language Models are Zero-Shot Reasoners", "Iteratively Prompt Pre-trained Language Models for Chain of Thought", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Graph of Thoughts: Solving Elaborate Problems with Large Language Models", "RecMind: Large Language Model Powered Agent For Recommendation"], "answer_arxiv_id": ["2205.11916", "2203.08383", "2305.10601", "2308.09687", "2308.14296"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_14406"} +{"question": "Could you provide me some works that propose to prune the parameters based on their sensitivity or contribution to the network output?", "answer": ["SNIP: Single-shot Network Pruning based on Connection Sensitivity", "Network Trimming: A Data-Driven Neuron Pruning Approach towards Efficient Deep Architectures"], "answer_arxiv_id": ["1810.02340", "1607.03250"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_14407"} +{"question": "Which research has proposed a trust-region variational inference for Gaussian Mixture Models (GMMs) to approximate multimodal distributions?", "answer": ["Trust-Region Variational Inference with Gaussian Mixture Models"], "answer_arxiv_id": ["1907.04710"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_14408"} +{"question": "Could you provide me some studies about the concept of Policy Pruning and Shrinking (PoPs) in DRL?", "answer": ["PoPS: Policy Pruning and Shrinking for Deep Reinforcement Learning"], "answer_arxiv_id": ["2001.05012"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_14409"} +{"question": "Which papers discusses about conventional super-resolution models?", "answer": ["Image Super-Resolution Using Deep Convolutional Networks", "Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Enhanced Deep Residual Networks for Single Image Super-Resolution", "Image Super-Resolution Using Very Deep Residual Channel Attention\n Networks", "Residual Dense Network for Image Super-Resolution", "Fast, Accurate, and Lightweight Super-Resolution with Cascading Residual\n Network", "Attention in Attention Network for Image Super-Resolution", "Residual Non-local Attention Networks for Image Restoration", "LAPAR: Linearly-Assembled Pixel-Adaptive Regression Network for Single\n Image Super-Resolution and Beyond", "Best-Buddy GANs for Highly Detailed Image Super-Resolution", "SwinIR: Image Restoration Using Swin Transformer", "Deeply-Recursive Convolutional Network for Image Super-Resolution", "On Efficient Transformer-Based Image Pre-training for Low-Level Vision", "Dual Aggregation Transformer for Image Super-Resolution", "Recursive Generalization Transformer for Image Super-Resolution", "Blueprint Separable Residual Network for Efficient Image\n Super-Resolution", "Efficient Image Super-Resolution using Vast-Receptive-Field Attention"], "answer_arxiv_id": ["1501.00092", "1511.04587", "1707.02921", "1807.02758", "1802.08797", "1803.08664", "2104.09497", "1903.10082", "2105.10422", "2103.15295", "2108.10257", "1511.04491", "2112.10175", "2308.03364", "2303.06373", "2205.05996", "2210.05960"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14410"} +{"question": "What works discuss optimization-based meta-learning use in domain generalization?", "answer": ["Domain Generalization via Model-Agnostic Learning of Semantic Features", "Learning to Generalize: Meta-Learning for Domain Generalization", "Episodic Training for Domain Generalization"], "answer_arxiv_id": ["1910.13580", "1710.03463", "1902.00113"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_14411"} +{"question": "What research works in LLaVA series aligned vision encoders and LMMs for better image-text understanding?", "answer": ["Improved Baselines with Visual Instruction Tuning", "Visual Instruction Tuning"], "answer_arxiv_id": ["2310.03744", "2304.08485"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_14412"} +{"question": "Which works are about heterogeneous causal discovery?", "answer": ["Causal Discovery in Heterogeneous Environments Under the Sparse Mechanism Shift Hypothesis"], "answer_arxiv_id": ["2206.02013"], "source_meta": {"published_time": "20210705"}, "qid": "AutoScholarQuery_train_14413"} +{"question": "Can you tell me about the works that focus on generating full-body geometric representations of humans?", "answer": ["The Wanderings of Odysseus in 3D Scenes", "Weakly-supervised Action Transition Learning for Stochastic Human Motion\n Prediction", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE", "TEMOS: Generating diverse human motions from textual descriptions", "MotionCLIP: Exposing Human Motion Generation to CLIP Space", "Stochastic Multi-Person 3D Motion Forecasting", "We are More than Our Joints: Predicting how 3D Bodies Move"], "answer_arxiv_id": ["2112.09251", "2205.15608", "2104.05670", "2204.14109", "2203.08063", "2306.05421", "2012.00619"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14414"} +{"question": "Which works discuss the parallelizing of the gradient computation in relation to data?", "answer": ["Effective Federated Adaptive Gradient Methods with Non-IID Decentralized Data"], "answer_arxiv_id": ["2009.06557"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_14415"} +{"question": "What papers are there on efforts to accelerate the training or inference processes for MoE?", "answer": ["FastMoE: A Fast Mixture-of-Expert Training System", "GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding"], "answer_arxiv_id": ["2103.13262", "2006.16668"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14416"} +{"question": "What research refers to social meaning as the meaning that emerges through human interaction on social media in the form of emotion, sarcasm, irony and so on?", "answer": ["Improving Social Meaning Detection with Pragmatic Masking and Surrogate\n Fine-Tuning"], "answer_arxiv_id": ["2108.00356"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_train_14417"} +{"question": "What paper introduces a top-down pathway into its transformer backbone?", "answer": ["SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds"], "answer_arxiv_id": ["2210.07372"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_14418"} +{"question": "Could you provide some works that employ trainable adapters for Text-to-Image models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_14419"} +{"question": "In what studies are end-to-end solutions based on the idea of mask classification presented for panoptic segmentation?", "answer": ["MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "Segmenter: Transformer for Semantic Segmentation", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation", "OneFormer: One Transformer to Rule Universal Image Segmentation", "Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentation", "ReMaX: Relaxing for Better Training on Efficient Panoptic Segmentation"], "answer_arxiv_id": ["2012.00759", "2105.05633", "2107.06278", "2112.01527", "2206.08948", "2211.06220", "2206.02777", "2306.17319"], "source_meta": {"published_time": "20230804"}, "qid": "AutoScholarQuery_train_14420"} +{"question": "Could you list some studies that focused on fine-tuning LLMs on large-scale translation corpora?", "answer": ["A Paradigm Shift in Machine Translation: Boosting Translation\n Performance of Large Language Models"], "answer_arxiv_id": ["2309.11674"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_14421"} +{"question": "Which work uses the method of mining answer aliases from a Knowledge Graph (KG) for 'answer expansion'?", "answer": ["What's in a Name? Answer Equivalence For Open-Domain Question Answering"], "answer_arxiv_id": ["2109.05289"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_14422"} +{"question": "What research presented methods by which the QAP deconstructs into smaller problems and solves each with a series of LAPs?", "answer": ["Isometric Multi-Shape Matching", "Efficient Deformable Shape Correspondence via Kernel Matching"], "answer_arxiv_id": ["2012.02689", "1707.08991"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_14423"} +{"question": "Which papers developed algorithms for the kernelised bandit problems?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design", "Finite-Time Analysis of Kernelised Contextual Bandits", "On Kernelized Multi-armed Bandits", "High-Dimensional Experimental Design and Kernel Bandits", "A Domain-Shrinking based Bayesian Optimization Algorithm with Order-Optimal Regret Performance", "Gaussian Process Bandit Optimization with Few Batches"], "answer_arxiv_id": ["0912.3995v4", "1309.6869", "1704.00445", "2105.05806", "2010.13997", "2110.07788v4"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_14424"} +{"question": "What are some studies that analyzed the convergence of the stochastic proximal point method?", "answer": ["Nonasymptotic convergence of stochastic proximal point algorithms for constrained convex optimization", "Accelerated, Optimal, and Parallel: Some results on model-based stochastic optimization"], "answer_arxiv_id": ["1706.06297v1", "2101.02696"], "source_meta": {"published_time": "20220906"}, "qid": "AutoScholarQuery_train_14425"} +{"question": "Could you provide me some studies about the use of neural network Ansatzes in VMC to improve the accuracy of many-electron calculations?", "answer": ["Ab-Initio Solution of the Many-Electron Schrödinger Equation with Deep Neural Networks", "Deep neural network solution of the electronic Schrödinger equation", "Fermionic neural-network states for ab-initio electronic structure", "Solving Many-Electron Schrödinger Equation Using Deep Neural Networks", "Backflow Transformations via Neural Networks for Quantum Many-Body Wave-Functions", "Iterative backflow renormalization procedure for many-body ground state wave functions of strongly interacting normal Fermi liquids"], "answer_arxiv_id": ["1909.02487", "1909.08423", "1909.12852", "1807.07014", "1807.10770", "1501.02199"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_14426"} +{"question": "Can you name some papers that utilized Trust region methods?", "answer": ["Scalable Global Optimization via Local Bayesian Optimization", "Think Global and Act Local: Bayesian Optimisation over High-Dimensional Categorical and Mixed Search Spaces"], "answer_arxiv_id": ["1910.01739", "2102.07188"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14427"} +{"question": "Which papers reported that Re-labeling methods with self-consistency regularization obtained state-of-the art performance in real-world noisy datasets?", "answer": ["Unsupervised Data Augmentation for Consistency Training"], "answer_arxiv_id": ["1904.12848"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_14428"} +{"question": "Could you list the references for the multi-stream models that use multiple spatio-temporal views?", "answer": ["Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset", "SlowFast Networks for Video Recognition", "Broaden Your Views for Self-Supervised Video Learning"], "answer_arxiv_id": ["1705.07750", "1812.03982", "2103.16559"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_14429"} +{"question": "Any works that illustrated the use of natural language inference modules for checking information consistency to detect factual hallucinations?", "answer": ["Adversarial NLI for Factual Correctness in Text Summarisation Models"], "answer_arxiv_id": ["2005.11739"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_14430"} +{"question": "What research papers were focused on studies of textual entailment?", "answer": ["A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference", "Swag: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference"], "answer_arxiv_id": ["1704.05426", "1808.05326"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_14431"} +{"question": "Can you list the works that used planes as parametric primitives in their research?", "answer": ["PlaneNet: Piece-wise Planar Reconstruction from a Single RGB Image", "PlaneRCNN: 3D Plane Detection and Reconstruction from a Single Image", "PlaneRecTR: Unified Query Learning for 3D Plane Recovery from a Single\n View"], "answer_arxiv_id": ["1804.06278", "1812.04072", "2307.13756"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_14432"} +{"question": "What studies are there on the approximation properties of the Random Features Regression (RFR) model?", "answer": ["Linearized two-layers neural networks in high dimension", "The curse of overparametrization in adversarial training: Precise analysis of robust generalization for random features regression"], "answer_arxiv_id": ["1904.12191", "2201.05149"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_14433"} +{"question": "Can you name the studies that consider molecules as 2D topology graphs and design graph generation models?", "answer": ["Junction Tree Variational Autoencoder for Molecular Graph Generation", "Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation", "GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation", "GraphDF: A Discrete Flow Model for Molecular Graph Generation", "GraphEBM: Molecular Graph Generation with Energy-Based Models"], "answer_arxiv_id": ["1802.04364", "1806.02473", "2001.09382", "2102.01189", "2102.00546"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_14434"} +{"question": "Which studies have explored embedding inversion attacks across computer vision?", "answer": ["High Fidelity Visualization of What Your Self-Supervised Representation\n Knows About", "Inverting Visual Representations with Convolutional Networks", "Understanding Invariance via Feedforward Inversion of Discriminatively\n Trained Classifiers"], "answer_arxiv_id": ["2112.09164", "1506.02753", "2103.07470"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_14435"} +{"question": "Which papers showcased the ability of diffusion models to learn the multimodal distribution of offline policies or human behaviors?", "answer": ["Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "Imitating Human Behaviour with Diffusion Models", "Diffusion Policy: Visuomotor Policy Learning via Action Diffusion", "Offline Reinforcement Learning via High-Fidelity Generative Behavior Modeling"], "answer_arxiv_id": ["2208.06193", "2301.10677", "2303.04137v5", "2209.14548"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14436"} +{"question": "Any work has utilized attention mechanism to learn contributions of neighboring nodes to the target nodes?", "answer": ["Graph Attention Networks", "How Attentive are Graph Attention Networks?"], "answer_arxiv_id": ["1710.10903", "2105.14491"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14437"} +{"question": "What are some works that proposed structural re-parameterization techniques to scale up kernel size?", "answer": ["Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs", "DiracNets: Training Very Deep Neural Networks Without Skip-Connections", "RepVGG: Making VGG-style ConvNets Great Again"], "answer_arxiv_id": ["2203.06717", "1706.00388", "2101.03697"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_14438"} +{"question": "In what papers are deep-learning based approaches used to replace explicit multi-view geometry estimation?", "answer": ["NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video", "TransformerFusion: Monocular RGB Scene Reconstruction using Transformers", "SimpleRecon: 3D Reconstruction Without 3D Convolutions"], "answer_arxiv_id": ["2104.00681", "2107.02191", "2208.14743"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_14439"} +{"question": "Which works tackled the issue of improving performance by model ensembling?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1612.01474"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_14440"} +{"question": "In which paper can I find a taxonomy of predictors distinguishing their performance on different types of data in context of DNNs?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_14441"} +{"question": "What studies applied learning visual correspondence in functional perception?", "answer": ["The Functional Correspondence Problem"], "answer_arxiv_id": ["2109.01097"], "source_meta": {"published_time": "20240511"}, "qid": "AutoScholarQuery_train_14442"} +{"question": "Could you provide me with works that have researched on transformer models in the context of theory of mind and planning?", "answer": ["Neural Theory-of-Mind? On the Limits of Social Intelligence in Large LMs"], "answer_arxiv_id": ["2210.13312"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14443"} +{"question": "Which research papers applied few-shot meta-learning to computer vision and natural language processing?", "answer": ["Meta Learning for Natural Language Processing: A Survey"], "answer_arxiv_id": ["2205.01500"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_14444"} +{"question": "Which research suggests fine-tuning by updating either a subset of model parameters?", "answer": ["BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-models"], "answer_arxiv_id": ["2106.10199"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14445"} +{"question": "Could you give me examples of studies that proposed value-based offline RL algorithms?", "answer": ["Playing Atari with Deep Reinforcement Learning"], "answer_arxiv_id": ["1312.5602"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_14446"} +{"question": "What researches have been made about predicting the expected model output considering randomness stemming from model initialisation?", "answer": ["Datamodels: Predicting Predictions from Training Data", "TRAK: Attributing Model Behavior at Scale"], "answer_arxiv_id": ["2202.00622", "2303.14186"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_14447"} +{"question": "Which works explored the potential of MoE for scaling the models with trillion-size parameters in NLP?", "answer": ["Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity", "M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts", "GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding"], "answer_arxiv_id": ["2101.03961", "2110.03888", "2112.06905", "2006.16668"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_14448"} +{"question": "Who revisited the consistency and generalization of many surrogate loss-based algorithms with respect to the ranking loss measure?", "answer": ["Rethinking and Reweighting the Univariate Losses for Multi-Label Ranking: Consistency and Generalization"], "answer_arxiv_id": ["2105.05026"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_14449"} +{"question": "Could you provide me some studies that discovered knowledge neurons in BERT?", "answer": ["Knowledge Neurons in Pretrained Transformers"], "answer_arxiv_id": ["2104.08696"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_14450"} +{"question": "Could you provide me some studies about neural network repair?", "answer": ["Towards Repairing Neural Networks Correctly", "SpecRepair: Counter-Example Guided Safety Repair of Deep Neural Networks"], "answer_arxiv_id": ["2012.01872", "2106.01917v5"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_14451"} +{"question": "Could you provide some references for open datasets generated due to the increased availability of aerial LiDAR technology?", "answer": ["SUM: A Benchmark Dataset of Semantic Urban Meshes", "DublinCity: Annotated LiDAR Point Cloud and its Applications"], "answer_arxiv_id": ["2103.00355", "1909.03613"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_14452"} +{"question": "Are there any studies which integrates rendering of an intermediate 3D representation into the denoising step of a 2D diffusion model?", "answer": ["Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data"], "answer_arxiv_id": ["2306.11719", "2306.07881"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_14453"} +{"question": "Which work combined prediction with example-level rejection and LLMs on the text decontextualization problem?", "answer": ["Learning to Reject with a Fixed Predictor: Application to\n Decontextualization"], "answer_arxiv_id": ["2301.09044"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_14454"} +{"question": "Which studies propose using masks, explanatory graphs, and probabilistic models to interpret the internal layers of CNNs?", "answer": ["Interpretable CNNs for Object Classification", "Inference Graphs for CNN Interpretation", "Dataset Distillation via Factorization", "Dataset Distillation: A Comprehensive Review"], "answer_arxiv_id": ["1901.02413", "2110.10568", "2210.16774", "2301.07014"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_14455"} +{"question": "What are some research papers that offer computationally efficient alternatives to Sharpness Aware Minimization (SAM)?", "answer": ["Towards Efficient and Scalable Sharpness-Aware Minimization"], "answer_arxiv_id": ["2203.02714"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_14456"} +{"question": "Which works explored classifier-guidance and classifier-free guidance in controlling the output of diffusion models?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2207.12598", "2112.10752"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_14457"} +{"question": "What papers explain ABC with arbitrary probabilistic acceptance kernels and introduce generalized posteriors for ABC?", "answer": ["Generalized Posteriors in Approximate Bayesian Computation"], "answer_arxiv_id": ["2011.08644"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_14458"} +{"question": "Could you provide me with some references on encoding images into latent representations using VPD and the Stable Diffusion model?", "answer": ["Unleashing Text-to-Image Diffusion Models for Visual Perception"], "answer_arxiv_id": ["2303.02153v1"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_14459"} +{"question": "What recent benchmarks introduced complex visual reasoning to 3D data?", "answer": ["ScanQA: 3D Question Answering for Spatial Scene Understanding", "3D Question Answering"], "answer_arxiv_id": ["2112.10482", "2112.08359"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_14460"} +{"question": "Which works show the effectiveness of TEE in model protection?", "answer": ["ShadowNet: A Secure and Efficient On-device Model Inference System for Convolutional Neural Networks"], "answer_arxiv_id": ["2011.05905"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_14461"} +{"question": "Which studies aim to equip language models with the ability to use search engines?", "answer": ["Internet-Augmented Dialogue Generation", "LaMDA: Language Models for Dialog Applications", "Internet-augmented language models through few-shot prompting for open-domain question answering", "BlenderBot 3: a deployed conversational agent that continually learns to responsibly engage", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2107.07566", "2201.08239", "2203.05115", "2208.03188", "2210.03629"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_14462"} +{"question": "Which works are recognized in the area of visually-grounded language learning?", "answer": ["VidLanKD: Improving Language Understanding via Video-Distilled Knowledge Transfer", "Imagination-Augmented Natural Language Understanding"], "answer_arxiv_id": ["2107.02681", "2204.08535"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_14463"} +{"question": "Which studies discussed the impact of varying types of noise and intrinsic biases on the performance of current motion forecasting models?", "answer": ["Towards Robust and Adaptive Motion Forecasting: A Causal Representation Perspective", "Human Trajectory Prediction via Counterfactual Analysis"], "answer_arxiv_id": ["2111.14820", "2107.14202"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_14464"} +{"question": "Which studies have made strides in generating 3D models from text prompts or 2D references in the area of 3D reconstruction?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "HoloDiffusion: Training a 3D Diffusion Model using 2D Images", "One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization", "Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and\n Text-to-Image Diffusion Models", "Magic3D: High-Resolution Text-to-3D Content Creation", "3D Neural Field Generation using Triplane Diffusion"], "answer_arxiv_id": ["2209.14988", "2303.16509", "2306.16928", "2306.07881", "2212.04493", "2212.14704", "2211.10440", "2211.16677"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_14465"} +{"question": "What researches have been done on defending deep neural networks against adversarial examples?", "answer": ["Opportunities and Challenges in Deep Learning Adversarial Robustness: A\n Survey", "Adversarial Examples: Attacks and Defenses for Deep Learning", "Threat of Adversarial Attacks on Deep Learning in Computer Vision: A\n Survey", "Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning", "Adversarial Attacks and Defenses in Images, Graphs and Text: A Review", "Universal Adversarial Perturbations: A Survey", "Bag of Tricks for Adversarial Training", "Uncovering the Limits of Adversarial Training against Norm-Bounded\n Adversarial Examples"], "answer_arxiv_id": ["2007.00753", "1712.07107", "1801.00553", "1712.03141", "1909.08072", "2005.08087v1", "2010.00467", "2010.03593"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_14466"} +{"question": "Which works focused on the scaling of the model that contributed to the success of CLIP?", "answer": ["CoCa: Contrastive Captioners are Image-Text Foundation Models", "GPT-4 Technical Report", "EVA-CLIP: Improved Training Techniques for CLIP at Scale"], "answer_arxiv_id": ["2205.01917", "2303.08774", "2303.15389"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_14467"} +{"question": "Which works are involved in language-supervised segmentation using hierarchical grouping?", "answer": ["GroupViT: Semantic Segmentation Emerges from Text Supervision", "ViewCo: Discovering Text-Supervised Segmentation Masks via Multi-View Semantic Consistency", "Learning Open-vocabulary Semantic Segmentation Models From Natural Language Supervision", "Learning to Generate Text-grounded Mask for Open-world Semantic Segmentation from Only Image-Text Pairs", "Open Vocabulary Semantic Segmentation with Patch Aligned Contrastive Learning"], "answer_arxiv_id": ["2202.11094", "2302.10307", "2301.09121", "2212.00785", "2212.04994"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_14468"} +{"question": "Could you provide me some studies about visual instruction tuning?", "answer": ["Visual Prompt Tuning"], "answer_arxiv_id": ["2203.12119"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_14469"} +{"question": "Could you provide me with some studies that have introduced instruction-based capabilities to image editing?", "answer": ["InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2211.09800"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14470"} +{"question": "Can you cite the studies that apply concatenation methods for efficient video processing?", "answer": ["Accel: A Corrective Fusion Network for Efficient Semantic Segmentation\n on Video"], "answer_arxiv_id": ["1807.06667"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_14471"} +{"question": "Could you provide me a study on a distributed training framework that assembles independent trainers?", "answer": ["Simplifying Distributed Neural Network Training on Massive Graphs: Randomized Partitions Improve Model Aggregation"], "answer_arxiv_id": ["2305.09887"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_14472"} +{"question": "What works defined prompts as pixels and added to each image?", "answer": ["Exploring Visual Prompts for Adapting Large-Scale Models", "Diversity-Aware Meta Visual Prompting"], "answer_arxiv_id": ["2203.17274", "2303.08138"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_14473"} +{"question": "Who conducted research on sample diagnosis as a method of backdoor defense?", "answer": ["STRIP: A Defence Against Trojan Attacks on Deep Neural Networks", "SCALE-UP: An Efficient Black-box Input-level Backdoor Detection via Analyzing Scaled Prediction Consistency"], "answer_arxiv_id": ["1902.06531", "2302.03251"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_14474"} +{"question": "Which work originally proposed the Feature Pyramid Networks (FPN)?", "answer": ["Feature Pyramid Networks for Object Detection"], "answer_arxiv_id": ["1612.03144"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_14475"} +{"question": "Which research focused on performing distillation on language model tasks specifically for BERT-based models?", "answer": ["Understanding BERT Rankers Under Distillation"], "answer_arxiv_id": ["2007.11088"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_14476"} +{"question": "Which works studied sparse projections on the simplex and non-negative orthant?", "answer": ["Sparse projections onto the simplex"], "answer_arxiv_id": ["1206.1529"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_14477"} +{"question": "What research has been conducted on making legal text more understandable, specifically through legal text simplification?", "answer": ["Unsupervised Simplification of Legal Texts"], "answer_arxiv_id": ["2209.00557"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_14478"} +{"question": "Which paper introduces a voxel-based sparse convolution and point transformer feature extraction module for scene flow estimation?", "answer": ["SCTN: Sparse Convolution-Transformer Network for Scene Flow Estimation"], "answer_arxiv_id": ["2105.04447"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_14479"} +{"question": "Which works framed the structure search as a continuous optimization problem?", "answer": ["DAGs with NO TEARS: Continuous Optimization for Structure Learning", "DAG-GNN: DAG Structure Learning with Graph Neural Networks", "Gradient-Based Neural DAG Learning"], "answer_arxiv_id": ["1803.01422", "1904.10098", "1906.02226"], "source_meta": {"published_time": "20220411"}, "qid": "AutoScholarQuery_train_14480"} +{"question": "Which paper introduced the SRCNN model for single-image super-resolution?", "answer": ["Image Super-Resolution Using Deep Convolutional Networks"], "answer_arxiv_id": ["1501.00092"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_14481"} +{"question": "What studies present upper bounds for semi-supervised regression that are contingent on the degree to which the marginal informs the labeling function?", "answer": ["Density-sensitive semisupervised inference"], "answer_arxiv_id": ["1204.1685"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_14482"} +{"question": "Could you provide me the studies about solving the inverse rendering problem using neural field representations?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "NeRV: Neural Reflectance and Visibility Fields for Relighting and View\n Synthesis", "NeRD: Neural Reflectance Decomposition from Image Collections", "PhySG: Inverse Rendering with Spherical Gaussians for Physics-based\n Material Editing and Relighting", "Extracting Triangular 3D Models, Materials, and Lighting From Images", "NeRFactor: Neural Factorization of Shape and Reflectance Under an\n Unknown Illumination"], "answer_arxiv_id": ["2003.08934", "2012.03927", "2012.03918", "2104.00674", "2111.12503", "2106.01970"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_14483"} +{"question": "Which papers discuss Transformers based approaches to object detection, instance segmentation, and human poses?", "answer": ["Pix2seq: A Language Modeling Framework for Object Detection", "A Unified Sequence Interface for Vision Tasks"], "answer_arxiv_id": ["2109.10852", "2206.07669"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_14484"} +{"question": "What are some studies that have used experience replay and representation consolidation in rehearsal-based methods?", "answer": ["Experience Replay for Continual Learning", "Task Agnostic Representation Consolidation: a Self-supervised based\n Continual Learning Approach"], "answer_arxiv_id": ["1811.11682", "2207.06267"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_14485"} +{"question": "What are some works that discussed the advantages of human motion representation?", "answer": ["NTU RGB+D 120: A Large-Scale Benchmark for 3D Human Activity\n Understanding", "Skeleton-Based Mutually Assisted Interacted Object Localization and\n Human Action Recognition"], "answer_arxiv_id": ["1905.04757", "2110.14994"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_14486"} +{"question": "What papers are about integrating an additional modality into speech recognition?", "answer": ["Deep Audio-Visual Speech Recognition", "Robust Self-Supervised Audio-Visual Speech Recognition"], "answer_arxiv_id": ["1809.02108", "2201.01763"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_14487"} +{"question": "Could you provide me a paper that proposed a method utilizing sine-based MLPs in neural rendering?", "answer": ["SiNeRF: Sinusoidal Neural Radiance Fields for Joint Pose Estimation and Scene Reconstruction"], "answer_arxiv_id": ["2210.04553"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_14488"} +{"question": "What works have addressed holistic scene decomposition by splitting albedo and shading?", "answer": ["Direct Intrinsics: Learning Albedo-Shading Decomposition by\n Convolutional Regression", "CGIntrinsics: Better Intrinsic Image Decomposition through\n Physically-Based Rendering"], "answer_arxiv_id": ["1512.02311", "1808.08601"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_14489"} +{"question": "What works talk about the amazing performance of autoregressive models in the context of image generation?", "answer": ["Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "CogView: Mastering Text-to-Image Generation via Transformers", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2203.13131", "2105.13290", "2206.10789"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_14490"} +{"question": "What study asserts language as a suitable medium for compression to summarize past events in the context of RL with incomplete state information?", "answer": ["History Compression via Language Models in Reinforcement Learning"], "answer_arxiv_id": ["2205.12258"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_14491"} +{"question": "Which works involved option-first methods in variational option discovery?", "answer": ["Variational Intrinsic Control", "Diversity is All You Need: Learning Skills without a Reward Function", "Variational Option Discovery Algorithms", "Dynamics-Aware Unsupervised Discovery of Skills"], "answer_arxiv_id": ["1611.07507", "1802.06070", "1807.10299", "1907.01657"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_14492"} +{"question": "Could you provide me any studies that examined the implicit bias in smoothed homogeneous neural network trained by gradient descent with exponentially-tailed losses?", "answer": ["Gradient Descent Maximizes the Margin of Homogeneous Neural Networks"], "answer_arxiv_id": ["1906.05890"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_14493"} +{"question": "Which papers discussed the use of causal auto-regressive recurrent networks for sequential data modalities?", "answer": ["Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation"], "answer_arxiv_id": ["1406.1078"], "source_meta": {"published_time": "20230407"}, "qid": "AutoScholarQuery_train_14494"} +{"question": "Which research works have aimed to reduce the storage and computation demands of RAD tensor data?", "answer": ["RADDet: Range-Azimuth-Doppler based Radar Object Detection for Dynamic Road Users", "CNN based Road User Detection using the 3D Radar Cube"], "answer_arxiv_id": ["2105.00363", "2004.12165"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_14495"} +{"question": "What are some works related to the use of hypernetworks for learning best response functions?", "answer": ["Self-Tuning Networks: Bilevel Optimization of Hyperparameters using Structured Best-Response Functions", "Learning the Pareto Front with Hypernetworks"], "answer_arxiv_id": ["1903.03088", "2010.04104"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_14496"} +{"question": "Who pioneered the work on the use of disentanglement for controlling syntactic-level generative factors in sentence representations?", "answer": ["Disentangling Generative Factors in Natural Language with Discrete\n Variational Autoencoders"], "answer_arxiv_id": ["2109.07169"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_14497"} +{"question": "What studies discuss sampling-based algorithms for ANN problem?", "answer": ["A Bandit Approach to Maximum Inner Product Search"], "answer_arxiv_id": ["1812.06360"], "source_meta": {"published_time": "20230104"}, "qid": "AutoScholarQuery_train_14498"} +{"question": "What datasets are used for the virtualized versions of real rooms?", "answer": ["Matterport3D: Learning from RGB-D Data in Indoor Environments", "The Replica Dataset: A Digital Replica of Indoor Spaces"], "answer_arxiv_id": ["1709.06158", "1906.05797"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_14499"} +{"question": "Do you know of works proposing alternative methods to combine syntactic control and attention-based optimization?", "answer": ["Harnessing the Spatial-Temporal Attention of Diffusion Models for High-Fidelity Text-to-Image Synthesis"], "answer_arxiv_id": ["2304.03869"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_14500"} +{"question": "Which works proposed Bayesian approaches for estimating the predictive uncertainty of neural network models?", "answer": ["Bayesian Neural Networks: An Introduction and Survey"], "answer_arxiv_id": ["2006.12024"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_14501"} +{"question": "Which are the studies in which the idea of encoding an HDR image into a multi-exposure LDR image stack for LDR IQA has been proposed?", "answer": ["How to cheat with metrics in single-image HDR reconstruction"], "answer_arxiv_id": ["2108.08713"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_14502"} +{"question": "What paper proposed a reinforcement learning approach to design 3-stage amplifier circuits from threshold specification?", "answer": ["Learning to Design Circuits"], "answer_arxiv_id": ["1812.02734"], "source_meta": {"published_time": "20230725"}, "qid": "AutoScholarQuery_train_14503"} +{"question": "What papers employ curriculum learning in the field of RL?", "answer": ["Reverse Curriculum Generation for Reinforcement Learning", "ScreenerNet: Learning Self-Paced Curriculum for Deep Neural Networks", "Learning to Teach in Cooperative Multiagent Reinforcement Learning", "BaRC: Backward Reachability Curriculum for Robotic Reinforcement Learning", "Solving Rubik’s Cube with a Robot Hand", "Teacher algorithms for curriculum learning of Deep RL in continuously parameterized environments", "Curriculum Learning with a Progression Function", "Adaptive Procedural Task Generation for Hard-Exploration Problems", "Self-Paced Deep Reinforcement Learning", "Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Design", "Evolving Curricula with Regret-Based Environment Design", "REvolveR: Continuous Evolutionary Models for Robot-to-robot Policy Transfer"], "answer_arxiv_id": ["1707.05300", "1801.00904", "1805.07830", "1806.06161", "1910.07113", "1910.07224", "2008.00511", "2007.00350", "2004.11812", "2012.02096", "2203.01302", "2202.05244"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_14504"} +{"question": "Could you provide me some papers that propose alterations to adversarial training that do not strictly attempt to find a maximal adversarial attack at every iteration?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy", "On the Convergence and Robustness of Adversarial Training", "Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning", "Adversarial Logit Pairing"], "answer_arxiv_id": ["1901.08573", "2112.08304", "1704.03976", "1803.06373"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_14505"} +{"question": "Which work introduces the use of high-low and low-high pseudo-labels to learn diverse relation triplets?", "answer": ["HiLo: Exploiting High Low Frequency Relations for Unbiased Panoptic\n Scene Graph Generation"], "answer_arxiv_id": ["2303.15994"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_14506"} +{"question": "What are some research papers on the usage of deep learning in multi-fidelity modeling?", "answer": ["Multi-fidelity regression using artificial neural networks: efficient approximation of parameter-dependent output quantities", "A composite neural network that learns from multi-fidelity data: Application to function approximation and inverse PDE problems", "Multi-fidelity Bayesian Neural Networks: Algorithms and Applications", "On Transfer Learning of Neural Networks using Bi-fidelity Data for Uncertainty Propagation"], "answer_arxiv_id": ["2102.13403", "1903.00104", "2012.13294", "2002.04495"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_14507"} +{"question": "Which studies in Vision-Language Navigation (VLN) use transformer-based architectures for joint visual and textual representations?", "answer": ["History Aware Multimodal Transformer for Vision-and-Language Navigation", "A Recurrent Vision-and-Language BERT for Navigation"], "answer_arxiv_id": ["2110.13309", "2011.13922"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_14508"} +{"question": "Which studies started to recognize the challenges of egocentric object tracking?", "answer": ["Is First Person Vision Challenging for Object Tracking?", "Visual Object Tracking in First Person Vision"], "answer_arxiv_id": ["2011.12263", "2209.13502"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_14509"} +{"question": "Which papers applied Neural ODEs in graphics?", "answer": ["Neural Ordinary Differential Equation Control of Dynamics on Graphs"], "answer_arxiv_id": ["2006.09773v5"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_14510"} +{"question": "Can you point me towards research that optimizes NeRF from scratch in few-shot settings?", "answer": ["Depth-supervised NeRF: Fewer Views and Faster Training for Free", "Dense Depth Priors for Neural Radiance Fields from Sparse Input Views"], "answer_arxiv_id": ["2107.02791", "2112.03288"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_14511"} +{"question": "Which work prompts large language models to self-evaluate and output the likelihood of their answers being true?", "answer": ["Language Models (Mostly) Know What They Know"], "answer_arxiv_id": ["2207.05221"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_14512"} +{"question": "Who proposed a class of poison attacks that applies to language models, aiming to cause information leakage in the training data?", "answer": ["Truth Serum: Poisoning Machine Learning Models to Reveal Their Secrets"], "answer_arxiv_id": ["2204.00032"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_14513"} +{"question": "Which works directly applied the off-the-shelf multi-view contrastive losses for out-of-distribution detection?", "answer": ["SSD: A Unified Framework for Self-Supervised Outlier Detection", "Contrastive Training for Improved Out-of-Distribution Detection"], "answer_arxiv_id": ["2103.12051", "2007.05566"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_14514"} +{"question": "Can you provide any research that combines state space models with other sequence modeling primitives?", "answer": ["Liquid Structural State-Space Models", "Mega: Moving Average Equipped Gated Attention", "Long Range Language Modeling via Gated State Spaces", "Hungry Hungry Hippos: Towards Language Modeling with State Space Models", "It’s Raw! Audio Generation with State-Space Models"], "answer_arxiv_id": ["2209.12951", "2209.10655", "2206.13947", "2212.14052", "2202.09729"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_14515"} +{"question": "Which studies discuss the potential issues of divergence with the decaying moving average algorithm RMSProp and Adam?", "answer": ["On the convergence of Adam and Beyond"], "answer_arxiv_id": ["1904.09237"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_14516"} +{"question": "Which works discussed the use of manually annotated keysteps in instructional videos for the task of keystep recognition?", "answer": ["COIN: A Large-scale Dataset for Comprehensive Instructional Video Analysis", "Cross-task weakly supervised learning from instructional videos", "HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips"], "answer_arxiv_id": ["1903.02874", "1903.08225", "1906.03327"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_14517"} +{"question": "Which research provided the black box shift estimation (BBSE) method for importance weight estimation?", "answer": ["Detecting and Correcting for Label Shift with Black Box Predictors"], "answer_arxiv_id": ["1802.03916"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14518"} +{"question": "Which research paper is known for establishing the theoretical consistency of the Physics-Informed Neural Networks (PINN) method?", "answer": ["On the convergence of physics informed neural networks for linear second-order elliptic and parabolic type PDEs"], "answer_arxiv_id": ["2004.01806"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_14519"} +{"question": "In the context of diffusion-based virtual try-on, which studies have been using the prior of large-scale pre-trained diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Learning Transferable Visual Models From Natural Language Supervision", "Paint by Example: Exemplar-based Image Editing with Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2112.10752", "2103.00020", "2211.13227"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14520"} +{"question": "What research papers have showcased synthesizing novel views from single images or texts using pre-trained 2D diffusion models?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2303.11328"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_14521"} +{"question": "Are there any studies that extend 2D-UNet into 3D-UNet by injecting temporal layers for text-to-video generation?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "ModelScope Text-to-Video Technical Report"], "answer_arxiv_id": ["2304.08818", "2308.06571"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_14522"} +{"question": "What studies have investigated the psychopathic tendencies and moral reasoning in models like GPT-3 and GPT-4?", "answer": ["Evaluating Psychological Safety of Large Language Models", "Do Large GPT Models Discover Moral Dimensions in Language\n Representations? A Topological Study Of Sentence Embeddings"], "answer_arxiv_id": ["2212.10529", "2309.09397"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_14523"} +{"question": "Excuse me, any theoretical analysis of the generalization error of learning with coarse labels?", "answer": ["Efficient Algorithms for Learning from Coarse Labels"], "answer_arxiv_id": ["2108.09805"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_14524"} +{"question": "Could you provide some studies that employ learnable perturbations in the embedding space or pixel space?", "answer": ["Visual Prompt Tuning"], "answer_arxiv_id": ["2203.12119"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_14525"} +{"question": "Which works have extended the use of diffusion models for realistic image editing?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Imagic: Text-Based Real Image Editing with Diffusion Models", "Prompt-to-Prompt Image Editing with Cross Attention Control", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations"], "answer_arxiv_id": ["2208.12242", "2210.09276", "2208.01626", "2108.01073"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14526"} +{"question": "Are there any research papers that focus on selecting and combining from a set of expert models for transfer learning?", "answer": ["A linearized framework and a new benchmark for model selection for fine-tuning", "No One Representation to Rule Them All: Overlapping Features of Training Methods"], "answer_arxiv_id": ["2102.00084", "2110.12899"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_14527"} +{"question": "What papers are about learned options in Hierarchical Reinforcement Learning?", "answer": ["FeUdal Networks for Hierarchical Reinforcement Learning", "Diversity is All You Need: Learning Skills without a Reward Function", "Data-Efficient Hierarchical Reinforcement Learning", "Hierarchical Reinforcement Learning by Discovering Intrinsic Options"], "answer_arxiv_id": ["1703.01161", "1802.06070", "1805.08296", "2101.06521"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_14528"} +{"question": "What papers mentioned that crowdsourcing of paraphrases is used for dataset building and augmentation?", "answer": ["Outlier Detection for Improved Data Quality and Diversity in Dialog\n Systems", "Directed Diversity: Leveraging Language Embedding Distances for Collective Creativity in Crowd Ideation"], "answer_arxiv_id": ["1904.03122", "2101.06030v1"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_14529"} +{"question": "What are some works that have provided solutions to improve the optimization process of ZSL methods?", "answer": ["Class Normalization for (Continual)? Generalized Zero-Shot Learning"], "answer_arxiv_id": ["2006.11328"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_14530"} +{"question": "Which works are known for introducing classical methods of Image-to-Image Registration?", "answer": ["Benchmarking 6DOF Outdoor Visual Localization in Changing Conditions"], "answer_arxiv_id": ["1707.09092"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14531"} +{"question": "Which studies used the auto-encoder to learn low-dimensional representation space in deep clustering?", "answer": ["Unsupervised Deep Embedding for Clustering Analysis"], "answer_arxiv_id": ["1511.06335"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_14532"} +{"question": "Are there any information retrieval approaches used in question answering over a knowledge graph?", "answer": ["Large-scale Simple Question Answering with Memory Networks", "UHop: An Unrestricted-Hop Relation Extraction Framework for\n Knowledge-Based Question Answering"], "answer_arxiv_id": ["1506.02075", "1904.01246"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_14533"} +{"question": "Which work used vision and language inputs as prompts and achieved remarkable few-shot results?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_14534"} +{"question": "Could you list research papers which utilized the 'median-of-means' approach to address heavy tails?", "answer": ["Geometric median and robust estimation in Banach spaces", "Loss Minimization and Parameter Estimation with Heavy Tails"], "answer_arxiv_id": ["1308.1334", "1307.1827"], "source_meta": {"published_time": "20230908"}, "qid": "AutoScholarQuery_train_14535"} +{"question": "Could you name some studies that focused on the convergence of graph neural networks to theoretical limit networks?", "answer": ["Convergence and Stability of Graph Convolutional Networks on Large Random Graphs", "Graphon Neural Networks and the Transferability of Graph Neural Networks"], "answer_arxiv_id": ["2006.01868", "2006.03548"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14536"} +{"question": "Which works have been mentioned discussing latent-variable generative models trained to transform Gaussian noise into a sample from data distribution?", "answer": ["Denoising Diffusion Probabilistic Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["2006.11239", "1503.03585"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14537"} +{"question": "Could you provide me with research about mitigating hallucinations by constructing high-quality data?", "answer": ["HalluciDoctor: Mitigating Hallucinatory Toxicity in Visual Instruction\n Data", "Mitigating Hallucination in Large Multi-Modal Models via Robust\n Instruction Tuning", "Mitigating Fine-Grained Hallucination by Fine-Tuning Large\n Vision-Language Models with Caption Rewrites", "Detecting and Preventing Hallucinations in Large Vision Language Models"], "answer_arxiv_id": ["2311.13614", "2306.14565", "2312.01701", "2308.06394"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_14538"} +{"question": "Could you provide me studies that analysed self-attention layers in the nature language processing field?", "answer": ["Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy\n Lifting, the Rest Can Be Pruned", "On the Relationship between Self-Attention and Convolutional Layers", "A Unified Multiscale Encoder-Decoder Transformer for Video Segmentation", "Are Sixteen Heads Really Better than One?"], "answer_arxiv_id": ["1905.09418", "1911.03584", "2304.05930", "1905.10650"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_14539"} +{"question": "Which paper introduced the method TTP as a leading perturbation generative method?", "answer": ["On Generating Transferable Targeted Perturbations"], "answer_arxiv_id": ["2103.14641"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_14540"} +{"question": "Could you provide me some studies that learn refiners using model generations?", "answer": ["Generating Sequences by Learning to [Self-]Correct", "Check Your Facts and Try Again: Improving Large Language Models with External Knowledge and Automated Feedback"], "answer_arxiv_id": ["2211.00053", "2302.12813"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_14541"} +{"question": "Which research papers are about approaches based on randomized smoothing in certified defenses?", "answer": ["Certified Robustness to Adversarial Examples with Differential Privacy", "Certified Adversarial Robustness via Randomized Smoothing", "Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers", "Consistency Regularization for Certified Robustness of Smoothed Classifiers", "MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius", "Boosting Randomized Smoothing with Variance Reduced Classifiers", "SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness", "Denoised Smoothing: A Provable Defense for Pretrained Classifiers", "(Certified!!) Adversarial Robustness for Free!"], "answer_arxiv_id": ["1802.03471", "1902.02918", "1906.04584", "2006.04062v4", "2001.02378", "2106.06946", "2111.09277", "2003.01908", "2206.10550"], "source_meta": {"published_time": "20221101"}, "qid": "AutoScholarQuery_train_14542"} +{"question": "What are some task-independent models that use VAEs?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image"], "answer_arxiv_id": ["1904.05866"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14543"} +{"question": "Which research used crowdsourced human evaluation to assess aspects of image generation models?", "answer": ["HYPE: A Benchmark for Human eYe Perceptual Evaluation of Generative Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Toward Verifiable and Reproducible Human Evaluation for Text-to-Image Generation", "Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation"], "answer_arxiv_id": ["1904.01121", "2205.11487", "2304.01816", "2305.01569"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_14544"} +{"question": "Which study first predicts primitive labels to infer unseen compositions in CZSL?", "answer": ["Simple Primitives with Feasibility- and Contextuality-Dependence for\n Open-World Compositional Zero-shot Learning"], "answer_arxiv_id": ["2211.02895"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14545"} +{"question": "Could you provide me some papers about DICE family of estimators and VPM which can be behavioral-agnostic?", "answer": ["GradientDICE: Rethinking Generalized Offline Estimation of Stationary Values", "GenDICE: Generalized Offline Estimation of Stationary Values", "Offline Policy Selection under Uncertainty", "Off-Policy Evaluation via the Regularized Lagrangian", "DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections", "CoinDICE: Off-Policy Confidence Interval Estimation", "Batch Stationary Distribution Estimation"], "answer_arxiv_id": ["2001.11113", "2002.09072", "2012.06919", "2007.03438", "1906.04733", "2010.11652", "2003.00722"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_14546"} +{"question": "Could you provide me some papers where 3D geometric consistency and depth prediction consistency were imposed across consecutive frames?", "answer": ["Unsupervised Learning of Depth and Ego-Motion from Monocular Video Using 3D Geometric Constraints", "Unsupervised Scale-consistent Depth and Ego-motion Learning from Monocular Video"], "answer_arxiv_id": ["1802.05522", "1908.10553"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_14547"} +{"question": "Which paper discusses the limitation of spatial attention compared to spectral attention?", "answer": ["How Expressive are Transformers in Spectral Domain for Graphs?"], "answer_arxiv_id": ["2201.09332"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_14548"} +{"question": "Which papers propose high-level subgoal selection procedures based on shortest paths between states?", "answer": ["Search on the Replay Buffer: Bridging Planning and Reinforcement Learning", "Semi-parametric Topological Memory for Navigation"], "answer_arxiv_id": ["1906.05253", "1803.00653"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_14549"} +{"question": "Who extended GCN to relation-aware GCN for KGs?", "answer": ["Modeling Relational Data with Graph Convolutional Networks", "Composition-based Multi-Relational Graph Convolutional Networks"], "answer_arxiv_id": ["1703.06103v4", "1911.03082"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_14550"} +{"question": "Could you provide me some works which used GFlowNets in Bayesian structure learning?", "answer": ["Bayesian Structure Learning with Generative Flow Networks"], "answer_arxiv_id": ["2202.13903"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_14551"} +{"question": "What works proposed using sparse and compressed communication to address communication overhead in large scale deep neural network training?", "answer": ["ScaleCom: Scalable Sparsified Gradient Compression for Communication-Efficient Distributed Training", "1-bit Adam: Communication Efficient Large-Scale Training with Adam’s Convergence Speed", "PowerSGD: Practical Low-Rank Gradient Compression for Distributed Optimization"], "answer_arxiv_id": ["2104.11125", "2102.02888", "1905.13727"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_14552"} +{"question": "Can you provide studies that design an architecture offering increased robustness to occlusions in 3D hand pose estimation?", "answer": ["HandOccNet: Occlusion-Robust 3D Hand Mesh Estimation Network"], "answer_arxiv_id": ["2203.14564"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_14553"} +{"question": "What studies have incorporated implicit neural representations or differentiable neural rendering into GANs for 3D-aware image generation?", "answer": ["GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis", "Lifting 2D StyleGAN for 3D-Aware Face Generation"], "answer_arxiv_id": ["2007.02442", "2011.12100", "2012.00926", "2011.13126"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_14554"} +{"question": "Any research papers that use invertible networks in designing end-to-end neural networks?", "answer": ["Analyzing Inverse Problems with Invertible Neural Networks"], "answer_arxiv_id": ["1808.04730"], "source_meta": {"published_time": "20211207"}, "qid": "AutoScholarQuery_train_14555"} +{"question": "Which work initially proposed the idea of enriching the few-shot examples with reasoning steps, which is an approach towards chain-of-thought prompting?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_14556"} +{"question": "What research papers curated multiagent datasets in simulated environments?", "answer": ["OPV2V: An Open Benchmark Dataset and Fusion Pipeline for Perception with\n Vehicle-to-Vehicle Communication", "V2X-Sim: Multi-Agent Collaborative Perception Dataset and Benchmark for\n Autonomous Driving"], "answer_arxiv_id": ["2109.07644", "2202.08449"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_14557"} +{"question": "What works use heuristic strategies to determine which layers to freeze and when?", "answer": ["FreezeOut: Accelerate Training by Progressively Freezing Layers", "Layer Freezing & Data Sieving: Missing Pieces of a Generic Framework for Sparse Training"], "answer_arxiv_id": ["1706.04983", "2209.11204"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_14558"} +{"question": "Could you provide me some references about methods enhancing the robustness of feature extractors to geometric misalignment?", "answer": ["Shift-tolerant Perceptual Similarity Metric", "Making Convolutional Networks Shift-Invariant Again", "The Unreasonable Effectiveness of Deep Features as a Perceptual Metric"], "answer_arxiv_id": ["2207.13686", "1904.11486", "1801.03924"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_14559"} +{"question": "Could you provide me with some studies that examine how datasets and training shape representations in models?", "answer": ["What shapes feature representations? Exploring datasets, architectures, and training"], "answer_arxiv_id": ["2006.12433"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_14560"} +{"question": "What research utilized CKA to establish the impact of parameter initialization and characteristics of last layers of overparameterized models' representations?", "answer": ["Do Wide and Deep Networks Learn the Same Things? Uncovering How Neural Network Representations Vary with Width and Depth"], "answer_arxiv_id": ["2010.15327"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_14561"} +{"question": "Which works have made use of neural fields in the study of representations for hand-object interaction modeling?", "answer": ["Neural Fields in Visual Computing and Beyond", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Neural Volumes: Learning Dynamic Renderable Volumes from Images", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural\n Scene Representations", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Learning Implicit Fields for Generative Shape Modeling", "Volume Rendering of Neural Implicit Surfaces", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "TensoRF: Tensorial Radiance Fields", "3D Gaussian Splatting for Real-Time Radiance Field Rendering", "Novel View Synthesis of Dynamic Scenes with Globally Coherent Depths\n from a Monocular Camera", "NeRF-DS: Neural Radiance Fields for Dynamic Specular Objects", "Mixed Neural Voxels for Fast Multi-view Video Synthesis", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis"], "answer_arxiv_id": ["2111.11426", "1812.03828", "2003.08934", "1906.07751", "1906.01618", "1901.05103", "1812.02822", "2106.12052", "2106.10689", "2201.05989", "2112.05131", "2203.09517", "2308.04079", "2004.01294", "2303.14435", "2212.00190", "2301.10241", "2308.09713"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14562"} +{"question": "What research avoids the requirement of having a strictly safe policy known for reducing the constraint violation?", "answer": ["A Provably-Efficient Model-Free Algorithm for Constrained Markov Decision Processes"], "answer_arxiv_id": ["2106.01577"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_14563"} +{"question": "Which works use online interactions for pretraining the agent to improve sample efficiency in RL?", "answer": ["Diversity is All You Need: Learning Skills without a Reward Function", "URLB: Unsupervised Reinforcement Learning Benchmark"], "answer_arxiv_id": ["1802.06070", "2110.15191"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14564"} +{"question": "What studies utilized multi-dataset training for developing robust computer vision models?", "answer": ["Towards Universal Object Detection by Domain Attention", "Simple Multi-dataset Detection", "MSeg: A Composite Dataset for Multi-domain Semantic Segmentation", "Detection Hub: Unifying Object Detection Datasets via Query Adaptation on Language Embedding", "The Missing Link: Finding label relations across datasets", "Universal-RCNN: Universal Object Detector via Transferable Graph R-CNN"], "answer_arxiv_id": ["1904.04402", "2102.13086", "2112.13762", "2206.03484", "2206.04453", "2002.07417"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14565"} +{"question": "Which paper adopted permutation invariant training (PIT) to achieve successful results on separating arbitrary sounds with a fixed number of sources?", "answer": ["Universal Sound Separation"], "answer_arxiv_id": ["1905.03330"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_14566"} +{"question": "Any works about the use of codebooks in NLP?", "answer": ["Supervised Topic Models"], "answer_arxiv_id": ["1003.0783"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_14567"} +{"question": "Can you provide me some works about applying OT on auto-encoder models?", "answer": ["VAE with a VampPrior", "Sinkhorn AutoEncoders"], "answer_arxiv_id": ["1705.07120", "1810.01118"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_14568"} +{"question": "Which studies propose a direct solution to the application of offline-to-online reinforcement learning?", "answer": ["AWAC: Accelerating Online Reinforcement Learning with Offline Datasets"], "answer_arxiv_id": ["2006.09359"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_14569"} +{"question": "What studies extend the concept by ensuring both cross and self-attention maps accurately represent objects?", "answer": ["Grounded Text-to-Image Synthesis with Attention Refocusing"], "answer_arxiv_id": ["2306.05427v2"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_14570"} +{"question": "What model is UMIC fine-tuned on?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning"], "answer_arxiv_id": ["1909.11740"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_14571"} +{"question": "Are there any studies that addressed issues related to AI ethics?", "answer": ["Erasing Concepts from Diffusion Models", "Ablating Concepts in Text-to-Image Diffusion Models", "Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Models", "Adapt then Unlearn: Exploiting Parameter Space Semantics for Unlearning\n in Generative Adversarial Networks"], "answer_arxiv_id": ["2303.07345", "2303.13516", "2303.17591", "2309.14054"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_14572"} +{"question": "Which papers developed distributed versions of RL algorithms to accelerate training?", "answer": ["Asynchronous Methods for Deep Reinforcement Learning", "IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures", "Gossip-based Actor-Learner Architectures for Deep Reinforcement Learning"], "answer_arxiv_id": ["1602.01783", "1802.01561", "1906.04585"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_14573"} +{"question": "Which studies report creating neural representations invariant to random perturbation?", "answer": ["Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction", "Drop, Swap, and Generate: A Self-Supervised Approach for Generating Neural Activity"], "answer_arxiv_id": ["2102.10106", "2111.02338"], "source_meta": {"published_time": "20230812"}, "qid": "AutoScholarQuery_train_14574"} +{"question": "What references talk about the long-form generation capability of large language models?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Progressive Generation of Long Text with Pretrained Language Models", "Re3: Generating Longer Stories With Recursive Reprompting and Revision"], "answer_arxiv_id": ["1901.02860", "2006.15720", "2210.06774"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_14575"} +{"question": "What studies address the bias of deep neural networks towards simple functions in the context of underspecification?", "answer": ["A Closer Look at Memorization in Deep Networks", "Implicit Bias of Gradient Descent on Linear Convolutional Networks", "The Pitfalls of Simplicity Bias in Neural Networks", "Shortcut Learning in Deep Neural Networks", "Gradient Starvation: A Learning Proclivity in Neural Networks"], "answer_arxiv_id": ["1706.05394", "1806.00468", "2006.07710", "2004.07780", "2011.09468"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_14576"} +{"question": "Which studies introduced stochastic rank-1 matrix bandits for multi-dimensional online decision making problems?", "answer": ["Stochastic Rank-1 Bandits", "Bernoulli Rank-1 Bandits for Click Feedback", "Solving Bernoulli Rank-One Bandits with Unimodal Thompson Sampling"], "answer_arxiv_id": ["1608.03023v3", "1703.06513", "1912.03074"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_14577"} +{"question": "What works conduct research on open-vocabulary panoptic segmentation?", "answer": ["Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models", "Generalized Decoding for Pixel, Image, and Language", "Segment Everything Everywhere All at Once", "Universal Instance Perception as Object Discovery and Retrieval"], "answer_arxiv_id": ["2303.04803", "2212.11270", "2304.06718", "2303.06674"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_14578"} +{"question": "Which papers directly regularize Q-values by lowering return estimates for out-of-distribution actions?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Why So Pessimistic? Estimating Uncertainties for Offline RL through Ensembles, and Why Their Independence Matters", "Offline Reinforcement Learning as Anti-Exploration"], "answer_arxiv_id": ["2006.04779", "2205.13703", "2106.06431"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_14579"} +{"question": "What research articles showcase the significance of symbols in leveraging external tools?", "answer": ["PAL: Program-aided Language Models", "LLM+P: Empowering Large Language Models with Optimal Planning\n Proficiency", "Logic-LM: Empowering Large Language Models with Symbolic Solvers for\n Faithful Logical Reasoning"], "answer_arxiv_id": ["2211.10435", "2304.11477", "2305.12295"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_14580"} +{"question": "Can you name the study that explores an MMI-like prompting approach for reasoning tasks?", "answer": ["Guess the Instruction! Flipped Learning Makes Language Models Stronger Zero-Shot Learners"], "answer_arxiv_id": ["2210.02969"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_14581"} +{"question": "What literature exists on extending the original setting to handle utilities that are unknown in strategic classification?", "answer": ["Strategic Classification from Revealed Preferences"], "answer_arxiv_id": ["1710.07887"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_14582"} +{"question": "What early methods focused on Eulerian perspective in the context of hand-crafted magnification filters?", "answer": ["Video Acceleration Magnification"], "answer_arxiv_id": ["1704.04186"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14583"} +{"question": "Which study evolved the Top-k Sampling strategy into the Top-p (Nucleus) Sampling?", "answer": ["The Curious Case of Neural Text Degeneration"], "answer_arxiv_id": ["1904.09751"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14584"} +{"question": "Which papers focused on vision-based tasks in the field of Continual learning?", "answer": ["A continual learning survey: Defying forgetting in classification tasks", "Incremental Object Detection via Meta-Learning", "SSUL: Semantic Segmentation with Unknown Label for Exemplar-based Class-Incremental Learning", "SS-IL: Separated Softmax for Incremental Learning"], "answer_arxiv_id": ["1909.08383", "2003.08798", "2106.11562", "2003.13947"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_14585"} +{"question": "What works discuss the use of auto-decoding in generalizable Implicit Neural Representations (INRs)?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation"], "answer_arxiv_id": ["1812.03828", "1901.05103"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_14586"} +{"question": "Could you name the works which proposed computing the stochastic reach-avoid set together with the probability in a manageable way to address the curse of dimensionality?", "answer": ["Stochastic reachability of a target tube: Theory and computation", "Risk-sensitive safety analysis using Conditional Value-at-Risk*"], "answer_arxiv_id": ["1810.05217", "2101.12086"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_14587"} +{"question": "Can you cite some studies that have applied box embedding in various fields?", "answer": ["Predicting Visual Overlap of Images Through Interpretable Non-Metric Box\n Embeddings", "Modeling Fine-Grained Entity Types with Box Embeddings", "Word2Box: Capturing Set-Theoretic Semantics of Words using Box\n Embeddings", "Temporal Knowledge Graph Completion using Box Embeddings"], "answer_arxiv_id": ["2008.05785", "2101.00345", "2106.14361", "2109.08970"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_14588"} +{"question": "Which datasets have been used for object anomaly detection in industrial vision?", "answer": ["The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localization", "The Eyecandies Dataset for Unsupervised Multimodal Anomaly Detection and Localization", "SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentation"], "answer_arxiv_id": ["2112.09045", "2210.04570", "2207.14315"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_14589"} +{"question": "What literature focuses on using pseudo-labeling in Unsupervised Domain Adaptation?", "answer": ["Prototypical Pseudo Label Denoising and Target Structure Learning for\n Domain Adaptive Semantic Segmentation"], "answer_arxiv_id": ["2101.10979"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14590"} +{"question": "Can you provide examples of research that focused on pre-training models for object detection and segmentation?", "answer": ["Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "RegionCLIP: Region-based Language-Image Pretraining", "Grounded Language-Image Pre-training", "PhraseCut: Language-based Image Segmentation in the Wild", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "Image Segmentation Using Text and Image Prompts", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting"], "answer_arxiv_id": ["2104.13921", "2112.09106", "2112.03857", "2008.01187", "2112.12143", "2112.10003", "2112.01518"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_14591"} +{"question": "Which papers discussed how language models can rely on undesirable different linguistic features to solve a task?", "answer": ["Annotation Artifacts in Natural Language Inference Data", "Language Through a Prism: A Spectral Approach for Multiscale Language Representations"], "answer_arxiv_id": ["1803.02324", "2011.04823"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_14592"} +{"question": "Which works demonstrate performance improvements on knowledge-intensive tasks but lack in generalization ability across domains?", "answer": ["ERNIE: Enhanced Language Representation with Informative Entities", "LUKE: Deep Contextualized Entity Representations with Entity-aware\n Self-attention", "ERICA: Improving Entity and Relation Understanding for Pre-trained\n Language Models via Contrastive Learning", "Improving language models by retrieving from trillions of tokens"], "answer_arxiv_id": ["1905.07129", "2010.01057", "2012.15022", "2112.04426"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_14593"} +{"question": "Which studies have emerged to enhance the established federated learning aggregation method, FedAvg?", "answer": ["Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classification", "Federated Optimization in Heterogeneous Networks", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "FedExP: Speeding Up Federated Averaging via Extrapolation"], "answer_arxiv_id": ["1909.06335", "1812.06127", "1910.06378", "2301.09604"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_14594"} +{"question": "What studies analyze expressivity focused on the frequency response of GNNs as well as their capacity to compute specific graph functions?", "answer": ["Revisiting Graph Neural Networks: All We Have is Low-Pass Filters", "Understanding the Representation Power of Graph Neural Networks in Learning Graph Topology", "Generalization and Representational Limits of Graph Neural Networks", "What Graph Neural Networks Cannot Learn: Depth vs Width", "Can Graph Neural Networks Count Substructures?", "On Graph Neural Networks versus Graph-Augmented MLPs", "Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting"], "answer_arxiv_id": ["1905.09550", "1907.05008", "2002.06157", "1907.03199", "2002.04025", "2010.15116", "2006.09252"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_14595"} +{"question": "Which papers originally introduced and further advanced Denoising diffusion probabilistic models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_14596"} +{"question": "Could you provide me some works about policy gradients that address multi-agent credit assignment?", "answer": ["Counterfactual Multi-Agent Policy Gradients", "Difference Rewards Policy Gradients"], "answer_arxiv_id": ["1705.08926", "2012.11258"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_14597"} +{"question": "What frameworks have been proposed for language model evaluation?", "answer": ["Holistic Evaluation of Language Models"], "answer_arxiv_id": ["2211.09110"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_14598"} +{"question": "Which papers have tackled text-to-image generation using diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Diffusion Models Beat GANs on Image Synthesis", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2112.10752", "2102.12092", "2205.11487", "2105.05233", "2112.10741", "2207.12598"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_14599"} +{"question": "Which works adopted the concept of meta-benchmarks and holistic evaluation across multiple situations or tasks in natural language processing (NLP)?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding", "Wilds: A Benchmark of in-the-Wild Distribution Shifts", "Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models", "Is ChatGPT a General-Purpose Natural Language Processing Task Solver?", "Holistic Evaluation of Language Models"], "answer_arxiv_id": ["1804.07461", "2012.07421", "2206.04615", "2302.06476", "2211.09110"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_14600"} +{"question": "What study proposes the voxel representation that naturally fits the Eulerian stage of the Material Point Method?", "answer": ["Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction"], "answer_arxiv_id": ["2111.11215"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_14601"} +{"question": "Which studies proposed the use of implicit methods for 3D representation?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "Implicit Geometric Regularization for Learning Shapes", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Implicit Functions in Feature Space for 3D Shape Reconstruction and\n Completion", "Deep Local Shapes: Learning Local SDF Priors for Detailed 3D\n Reconstruction", "PatchNets: Patch-Based Generalizable Deep Implicit 3D Shape\n Representations", "3DShape2VecSet: A 3D Shape Representation for Neural Fields and\n Generative Diffusion Models", "GridPull: Towards Scalability in Learning Implicit Representations from\n 3D Point Clouds", "SA-ConvONet: Sign-Agnostic Optimization of Convolutional Occupancy\n Networks"], "answer_arxiv_id": ["1812.02822", "2002.10099", "1812.03828", "1901.05103", "2003.01456", "2003.10983", "2008.01639", "2301.11445", "2308.13175", "2105.03582"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_14602"} +{"question": "Which papers focus on the design of randomized exploration bandit algorithms that achieve adaptive guarantees?", "answer": ["Tsallis-INF: An Optimal Algorithm for Stochastic and Adversarial Bandits"], "answer_arxiv_id": ["1807.07623v6"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_14603"} +{"question": "Which works pioneered Privacy-Preserving Machine Learning (PPML) via Homomorphic Encryption (HE)?", "answer": ["CryptoNAS: Private Inference on a ReLU Budget", "CrypTen: Secure Multi-Party Computation Meets Machine Learning"], "answer_arxiv_id": ["2006.08733", "2109.00984"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_14604"} +{"question": "Could you provide me some works about the applications of diffusion models to image restoration?", "answer": ["Denoising Diffusion Restoration Models"], "answer_arxiv_id": ["2201.11793"], "source_meta": {"published_time": "20220316"}, "qid": "AutoScholarQuery_train_14605"} +{"question": "What works introduced advances in multi-agent reinforcement learning (MARL)?", "answer": ["An Overview of Multi-agent Reinforcement Learning from Game Theoretical Perspective", "Multi-Agent Reinforcement Learning: A Selective Overview of Theories and Algorithms"], "answer_arxiv_id": ["2011.00583", "1911.10635"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_14606"} +{"question": "Could you provide me some studies discussing Direct-Answer-Prediction methods for KBQA?", "answer": ["PullNet: Open Domain Question Answering with Iterative Retrieval on Knowledge Bases and Text", "Sequence-to-Sequence Knowledge Graph Completion and Question Answering", "UniK-QA: Unified Representations of Structured and Unstructured Knowledge for Open-Domain Question Answering"], "answer_arxiv_id": ["1904.09537", "2203.10321", "2012.14610"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_14607"} +{"question": "What researches used diffusion models for adversarial purification?", "answer": ["Diffusion Models for Adversarial Purification", "Threat Model-Agnostic Adversarial Defense using Diffusion Models", "Ada3Diff: Defending against 3D Adversarial Point Clouds via Adaptive Diffusion", "DensePure: Understanding Diffusion Models towards Adversarial Robustness", "PointDP: Diffusion-driven Purification against Adversarial Attacks on 3D Point Cloud Recognition", "Guided Diffusion Model for Adversarial Purification", "Guided Diffusion Model for Adversarial Purification from Random Noise"], "answer_arxiv_id": ["2205.07460", "2207.08089", "2211.16247", "2211.00322", "2208.09801", "2205.14969", "2206.10875"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_14608"} +{"question": "What works considered the diversity in data subset selection algorithms?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Learning From Less Data: A Unified Data Subset Selection and Active Learning Framework for Computer Vision", "Coresets for Data-efficient Training of Machine Learning Models", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training"], "answer_arxiv_id": ["1708.00489", "1901.01151", "1906.01827", "2103.00123"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_14609"} +{"question": "What is the approach proposed by Lv et al. to handle noisy samples in noisy PLL?", "answer": ["On the Robustness of Average Losses for Partial-Label Learning"], "answer_arxiv_id": ["2106.06152"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_14610"} +{"question": "What paper presents the significance of randomly generated environments in training general agents and proposed ProcGen with pixel-space observations?", "answer": ["Leveraging Procedural Generation to Benchmark Reinforcement Learning"], "answer_arxiv_id": ["1912.01588"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_14611"} +{"question": "Could you provide me some studies about designing polynomial-time methods for efficient computation of exact correlated equilibria?", "answer": ["Polynomial-time Computation of Exact Correlated Equilibrium in Compact Games"], "answer_arxiv_id": ["1011.0253v1"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_14612"} +{"question": "Could you provide papers that have used the framework of sequential Monte-Carlo methods in the context of generative models?", "answer": ["Auto-Encoding Sequential Monte Carlo", "Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMC", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Score-Based Diffusion meets Annealed Importance Sampling"], "answer_arxiv_id": ["1705.10306", "2302.11552", "1503.03585", "2208.07698"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14613"} +{"question": "What works generate training data instances given a known task output space through analogous input?", "answer": ["Data Augmentation using Pre-trained Transformer Models", "Neural Data Augmentation via Example Extrapolation"], "answer_arxiv_id": ["2003.02245", "2102.01335"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14614"} +{"question": "What studies can I refer to that have combined adversarial optimization with NeRF for various purposes?", "answer": ["Aug-NeRF: Training Stronger Neural Radiance Fields with Triple-Level Physically-Grounded Augmentations", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields"], "answer_arxiv_id": ["2207.01164", "2011.12100", "2112.05139"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_14615"} +{"question": "Which studies proposed trajectory prediction models using convolutional neural networks in Autonomous Driving (AD)?", "answer": ["ChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing\n the Worst", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for\n Behavior Prediction", "Trajectron++: Dynamically-Feasible Trajectory Forecasting With\n Heterogeneous Data"], "answer_arxiv_id": ["1812.03079", "1910.05449", "2001.03093"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_14616"} +{"question": "What are some studies focused on encoder-decoder training in the context of pre-trained Transformer-based large language models?", "answer": ["Multilingual Denoising Pre-training for Neural Machine Translation", "Evaluating Large Language Models Trained on Code", "mT5: A Massively Multilingual Pre-trained Text-to-Text Transformer"], "answer_arxiv_id": ["2001.08210", "2107.03374", "2010.11934"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_14617"} +{"question": "Which studies proposed that the number of connected components in supp​ 𝑻 is identical to that in supp​ 𝚯⋆?", "answer": ["Learning Graphs with Monotone Topology Properties and Multiple Connected Components"], "answer_arxiv_id": ["1705.10934v4"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_14618"} +{"question": "Which models incorporate position coordinates into training data to achieve fine-grained image understanding and visual grounding?", "answer": ["Kosmos-2: Grounding Multimodal Large Language Models to the World", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic"], "answer_arxiv_id": ["2306.14824", "2306.15195"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_14619"} +{"question": "Could you provide me some studies that present fine-grained characterizations of the expressive power for certain functions in different settings, sometimes with statistical analyses?", "answer": ["Inductive Biases and Variable Creation in Self-Attention Mechanisms", "On the Expressive Power of Self-Attention Matrices", "What learning algorithm is in-context learning? Investigations with linear models", "Do Transformers Parse while Predicting the Masked Word?", "Self-Attention Networks Can Process Bounded Hierarchical Languages", "Exploring Length Generalization in Large Language Models", "Hidden Progress in Deep Learning: SGD Learns Parities Near the Computational Limit"], "answer_arxiv_id": ["2110.10090", "2106.03764", "2211.15661", "2303.08117", "2105.11115", "2207.04901", "2207.08799"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14620"} +{"question": "Could you provide me some literature about enhancing unified representation in FUSL with non-IID data?", "answer": ["Collaborative Unsupervised Visual Representation Learning from\n Decentralized Data", "Divergence-aware Federated Self-Supervised Learning", "L-DAWA: Layer-wise Divergence Aware Weight Aggregation in Federated\n Self-Supervised Visual Representation Learning", "ProtoFL: Unsupervised Federated Learning via Prototypical Distillation", "Orchestra: Unsupervised Federated Learning via Globally Consistent\n Clustering"], "answer_arxiv_id": ["2108.06492", "2204.04385", "2307.07393", "2307.12450", "2205.11506"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14621"} +{"question": "Could you present works that use sparse document and query representations in document retrieval?", "answer": ["A Deep Relevance Matching Model for Ad-hoc Retrieval"], "answer_arxiv_id": ["1711.08611"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_14622"} +{"question": "Can you name some works that employ recurrent-based methods in VSR?", "answer": ["Frame-Recurrent Video Super-Resolution", "Recurrent Back-Projection Network for Video Super-Resolution", "Efficient Video Super-Resolution through Recurrent Latent Space\n Propagation", "Video Super-Resolution with Recurrent Structure-Detail Network", "Revisiting Temporal Modeling for Video Super-resolution", "FDAN: Flow-guided Deformable Alignment Network for Video\n Super-Resolution", "BasicVSR: The Search for Essential Components in Video Super-Resolution\n and Beyond", "BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation\n and Alignment"], "answer_arxiv_id": ["1801.04590", "1903.10128", "1909.08080", "2008.00455", "2008.05765", "2105.05640", "2012.02181", "2104.13371"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_14623"} +{"question": "Could you provide me studies that use full eNTK for computation?", "answer": ["Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent"], "answer_arxiv_id": ["1902.06720"], "source_meta": {"published_time": "20220625"}, "qid": "AutoScholarQuery_train_14624"} +{"question": "Could you provide me some works that are associated with the concept of dataset constraint in offline RL?", "answer": ["Retrieval-Augmented Reinforcement Learning", "Large-Scale Retrieval for Reinforcement Learning"], "answer_arxiv_id": ["2202.08417", "2206.05314"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_14625"} +{"question": "What are works where normalising flow models that are estimated from incomplete data with VGI and MCEM?", "answer": ["Variational Inference with Normalizing Flows", "Normalizing Flows for Probabilistic Modeling and Inference"], "answer_arxiv_id": ["1505.05770", "1912.02762"], "source_meta": {"published_time": "20211125"}, "qid": "AutoScholarQuery_train_14626"} +{"question": "Which research works have used test-time distillation with pre-trained image generators in the application of diffusion to 3D?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Zero-1-to-3: Zero-shot One Image to 3D Object", "SparseFusion: Distilling View-conditioned Diffusion for 3D\n Reconstruction"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2303.11328", "2212.00792"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_14627"} +{"question": "Can you identify the works that reported the use of Stable Diffusion as the 2D prior and viewpoint-conditioned diffusion model in generating 3D content from a given image?", "answer": ["Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors", "High-Resolution Image Synthesis with Latent Diffusion Models", "Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2306.17843", "2112.10752", "2303.11328"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_14628"} +{"question": "Which papers discuss that negative log-likelihood based evaluation suffers from pitfalls in high dimensions and may not correlate with higher sample quality?", "answer": ["A note on the evaluation of generative models", "f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization", "Do Deep Generative Models Know What They Don’t Know?", "Perfect density models cannot guarantee anomaly detection"], "answer_arxiv_id": ["1511.01844", "1606.00709v1", "1810.09136", "2012.03808v3"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_14629"} +{"question": "What works studied adversarial training for Graph Neural Networks (GNNs) under structure perturbations?", "answer": ["Topology Attack and Defense for Graph Neural Networks: An Optimization Perspective", "Towards an Efficient and General Framework of Robust Training for Graph Neural Networks", "Spectral Adversarial Training for Robust Graph Neural Network", "Learning Robust Representation through Graph Adversarial Contrastive Learning"], "answer_arxiv_id": ["1906.04214", "2002.10947", "2211.10896", "2201.13025"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_14630"} +{"question": "Which papers discuss about the development of large language models (LLMs)?", "answer": ["GPT-4 Technical Report", "Scaling Instruction-Finetuned Language Models", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2303.08774", "2210.11416", "2302.13971"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_14631"} +{"question": "Which works discuss about the concept of task ambiguity in few-shot learning settings?", "answer": ["Probabilistic Model-Agnostic Meta-Learning", "Understanding the Capabilities, Limitations, and Societal Impact of Large Language Models"], "answer_arxiv_id": ["1806.02817", "2102.02503"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_14632"} +{"question": "Which works propose calibration-based approaches to understand what language models know?", "answer": ["Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation\n in Natural Language Generation", "Batch Calibration: Rethinking Calibration for In-Context Learning and\n Prompt Engineering"], "answer_arxiv_id": ["2302.09664", "2309.17249"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_14633"} +{"question": "Which works discussed minimizing individual regret in general-sum games?", "answer": ["Fast Convergence of Regularized Learning in Games", "Hedging in games: Faster convergence of external and swap regrets", "Near-Optimal No-Regret Learning in General Games", "Near-Optimal No-Regret Learning for Correlated Equilibria in Multi-Player General-Sum Games"], "answer_arxiv_id": ["1507.00407", "2006.04953", "2108.06924", "2111.06008v3"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14634"} +{"question": "Which works initially called for an increase in geographic diversity in visual datasets?", "answer": ["No Classification without Representation: Assessing Geodiversity Issues in Open Data Sets for the Developing World", "Does Object Recognition Work for Everyone?"], "answer_arxiv_id": ["1711.08536", "1906.02659"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_14635"} +{"question": "Which studies used diffusion models in pixel-level segmentation by leveraging pre-trained stable diffusion?", "answer": ["Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image\n Diffusion Models", "Peekaboo: Text to Image Diffusion Models are Zero-Shot Segmentors", "VGDiffZero: Text-to-image Diffusion Models Can Be Zero-shot Visual\n Grounders", "Diffuse, Attend, and Segment: Unsupervised Zero-Shot Segmentation using\n Stable Diffusion", "DiffusionSeg: Adapting Diffusion Towards Unsupervised Object Discovery"], "answer_arxiv_id": ["2301.13826", "2211.13224", "2309.01141", "2308.12469", "2303.09813"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14636"} +{"question": "Which papers propose point-based representations for radiance fields utilizing fast point rasterization pipelines?", "answer": ["Point-Based Neural Rendering with Per-View Optimization", "ADOP: Approximate Differentiable One-Pixel Point Rendering"], "answer_arxiv_id": ["2109.02369", "2110.06635"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_14637"} +{"question": "Are there any studies that utilized embedding-based scores for unsupervised accuracy estimation?", "answer": ["Transferability Estimation using Bhattacharyya Class Separability", "PACTran: PAC-Bayesian Metrics for Estimating the Transferability of Pretrained Models to Classification Tasks"], "answer_arxiv_id": ["2111.12780", "2203.05126"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_14638"} +{"question": "Which works argue that negatively curved edges are responsible for oversquashing?", "answer": ["Understanding over-squashing and bottlenecks on graphs via curvature"], "answer_arxiv_id": ["2111.14522"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_14639"} +{"question": "What papers deal with the risk oscillation in deep learning that is often referred to as the edge of stability (EoS)?", "answer": ["A Walk with SGD", "The large learning rate phase of deep learning: the catapult mechanism"], "answer_arxiv_id": ["1802.08770", "2003.02218"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_14640"} +{"question": "Could you provide some works about finite-sample tests for novelty detection?", "answer": ["Predictive inference with the jackknife+"], "answer_arxiv_id": ["1905.02928"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_14641"} +{"question": "What work applied the VSDL model?", "answer": ["Taxonomizing local versus global structure in neural network loss landscapes"], "answer_arxiv_id": ["2107.11228"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_14642"} +{"question": "Are there any works that explore transformer-based architectures as an alternative approach in operator learning?", "answer": ["Transformer for Partial Differential Equations’ Operator Learning", "GNOT: A General Neural Operator Transformer for Operator Learning"], "answer_arxiv_id": ["2205.13671", "2302.14376"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_14643"} +{"question": "What papers investigate the use of GPT3 as a writing assistant?", "answer": ["CoAuthor: Designing a Human-AI Collaborative Writing Dataset for Exploring Language Model Capabilities"], "answer_arxiv_id": ["2201.06796"], "source_meta": {"published_time": "20220824"}, "qid": "AutoScholarQuery_train_14644"} +{"question": "What research argues that LLMs can convert tables to text through prompting alone, thus motivating the use of LLM to classify tabular data?", "answer": ["Murmur: Modular Multi-Step Reasoning for Semi-Structured Data-to-Text Generation"], "answer_arxiv_id": ["2212.08607"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_14645"} +{"question": "Could you provide me some works that used isoperimetric inequalities to loosen the strong log-concavity assumption?", "answer": ["Proximal Langevin Algorithm: Rapid Convergence Under Isoperimetry"], "answer_arxiv_id": ["1911.01469"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_14646"} +{"question": "What research is there on leveraging inter-class similarity information in novel class discovery?", "answer": ["Modeling Inter-Class and Intra-Class Constraints in Novel Class\n Discovery", "Class-relation Knowledge Distillation for Novel Class Discovery"], "answer_arxiv_id": ["2210.03591", "2307.09158"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_14647"} +{"question": "Which amodal segmentation study operates only on a closed-world set of classes in Amodal COCO?", "answer": ["Self-Supervised Scene De-occlusion"], "answer_arxiv_id": ["2004.02788"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_14648"} +{"question": "Who proposed the implementation of SOS1 formulation or the big-M formulation for lower bound calculation in optimal methods?", "answer": ["Best Subset Selection via a Modern Optimization Lens", "Learning Sparse Nonlinear Dynamics via Mixed-Integer Optimization"], "answer_arxiv_id": ["1507.03133", "2206.00176"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_14649"} +{"question": "What studies analyze and propose generating new graph data by connecting subgraphs from different input graphs?", "answer": ["Graph Transplant: Node Saliency-Guided Graph Mixup with Local Structure Preservation", "Model-Agnostic Augmentation for Accurate Graph Classification"], "answer_arxiv_id": ["2111.05639", "2202.10107"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_14650"} +{"question": "Could you provide me some research that synthesizes a source audio at a new camera pose in the room?", "answer": ["Novel-View Acoustic Synthesis"], "answer_arxiv_id": ["2301.08730"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_14651"} +{"question": "What are some studies involving contextual auctions, particularly those incorporating public information about bidders and items?", "answer": ["Dynamic Incentive-aware Learning: Robust Pricing in Contextual Auctions", "A Context-Integrated Transformer-Based Neural Network for Auction Design", "Optimizing Multiple Performance Metrics with Deep GSP Auctions for E-commerce Advertising", "Neural Auction: End-to-End Learning of Auction Mechanisms for E-Commerce Advertising"], "answer_arxiv_id": ["2002.11137v1", "2201.12489", "2012.02930", "2106.03593"], "source_meta": {"published_time": "20230520"}, "qid": "AutoScholarQuery_train_14652"} +{"question": "Which work proposed an approach in which concept conjunctions are achieved by adding estimated scored from a parallel set of diffusion processes?", "answer": ["Compositional Visual Generation with Composable Diffusion Models"], "answer_arxiv_id": ["2206.01714"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_14653"} +{"question": "Can you provide works where recent uncertainty-based multimodal fusion methods were used for classification tasks?", "answer": ["Trusted Multi-View Classification", "COLD Fusion: Calibrated and Ordinal Latent Distribution Fusion for Uncertainty-Aware Multimodal Emotion Recognition", "Uncertainty-aware Audiovisual Activity Recognition using Deep Bayesian Variational Inference"], "answer_arxiv_id": ["2102.02051", "2206.05833", "1811.10811"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_14654"} +{"question": "What studies delve into the application of retrieval techniques for machine translation?", "answer": ["Search Engine Guided Neural Machine Translation", "Neural Machine Translation with Monolingual Translation Memory", "Nearest Neighbor Machine Translation", "Neural Machine Translation with Contrastive Translation Memories"], "answer_arxiv_id": ["1705.07267", "2105.11269v2", "2010.00710", "2212.03140"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_14655"} +{"question": "Could you provide me some studies about pre-training in RL that used dynamics learning-based representation learning?", "answer": ["Pretraining Representations for Data-Efficient Reinforcement Learning", "Reinforcement Learning with Action-Free Pre-Training from Videos"], "answer_arxiv_id": ["2106.04799", "2203.13880"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_14656"} +{"question": "In what papers can I find the proposition of introducing trainable prompt embeddings as an alternative to fine-tuning?", "answer": ["GPT Understands, Too", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2103.10385", "2101.00190", "2104.08691"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_14657"} +{"question": "What are some works that have successfully used reinforcement learning to improve the alignment between generated text and human preferences?", "answer": ["Learning to summarize from human feedback", "Training language models to follow instructions with human feedback", "Is Reinforcement Learning (Not) for Natural Language Processing: Benchmarks, Baselines, and Building Blocks for Natural Language Policy Optimization"], "answer_arxiv_id": ["2009.01325", "2203.02155", "2210.01241"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_14658"} +{"question": "Can you list some works that employed data augmentation-based methods for improving uncertainty estimates?", "answer": ["mixup: Beyond Empirical Risk Minimization", "OpenMix: Exploring Outlier Samples for Misclassification Detection"], "answer_arxiv_id": ["1710.09412", "2303.17093"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_14659"} +{"question": "Are there any studies that discussed the generalizability of methods to unknown norms?", "answer": ["Adversarial Training and Robustness for Multiple Perturbations", "Perceptual Adversarial Robustness: Defense Against Unseen Threat Models"], "answer_arxiv_id": ["1904.13000", "2006.12655"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_14660"} +{"question": "Which papers discuss algorithms that prune after training and are effective in speeding up inference but rely on a computationally expensive training procedure?", "answer": ["Pruning Convolutional Neural Networks for Resource Efficient Inference", "Learning to Prune Deep Neural Networks via Layer-wise Optimal Brain Surgeon", "The Combinatorial Brain Surgeon: Pruning Weights That Cancel One Another in Neural Networks"], "answer_arxiv_id": ["1611.06440", "1705.07565", "2203.04466"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_14661"} +{"question": "Which papers incorporated neural networks in the rendering pipeline of point clouds?", "answer": ["Neural Point-Based Graphics", "NPBG++: Accelerating Neural Point-Based Graphics", "Neural Point Light Fields", "Point-Based Neural Rendering with Per-View Optimization"], "answer_arxiv_id": ["1906.08240", "2203.13318", "2112.01473", "2109.02369"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_14662"} +{"question": "Could you mention some studies that evaluate models' capability to solve competitive Python programming problems?", "answer": ["Measuring Coding Challenge Competence With APPS"], "answer_arxiv_id": ["2105.09938"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_14663"} +{"question": "Which work discusses the method of encoding an input set with a recurrent neural network referred to as LSTMAgg?", "answer": ["Inductive Representation Learning on Large Graphs"], "answer_arxiv_id": ["1706.02216"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_14664"} +{"question": "What studies used RNN-based models for time series forecasting?", "answer": ["Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networks", "DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks"], "answer_arxiv_id": ["1703.07015", "1704.04110"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14665"} +{"question": "Which works are essential to knowledge distillation for time and memory-efficient deep learning?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_14666"} +{"question": "Which papers studied the importance of studying how variables change with scale in model performance?", "answer": ["Scaling Laws for Neural Language Models", "A Constructive Prediction of the Generalization Error Across Scales", "Deep Learning Scaling is Predictable, Empirically"], "answer_arxiv_id": ["2001.08361", "1909.12673v2", "1712.00409"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_14667"} +{"question": "Which studies utilized question generation pipelines to evaluate the faithfulness of summaries?", "answer": ["Answers Unite! Unsupervised Metrics for Reinforced Summarization Models", "QuestEval: Summarization Asks for Fact-based Evaluation", "QAFactEval: Improved QA-Based Factual Consistency Evaluation for\n Summarization"], "answer_arxiv_id": ["1909.01610", "2103.12693", "2112.08542"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_14668"} +{"question": "Which papers decompose time-dependent neural fields into an inverse displacement field and canonical time-invariant neural fields for NeRF framework?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2011.12948"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_14669"} +{"question": "What prior works discuss the use of isotropic points being substituted by Gaussian point modeling to enhance differential rendering?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis"], "answer_arxiv_id": ["2308.04079", "2308.09713"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14670"} +{"question": "What are the papers that focus on exploiting auxiliary networks for self-knowledge distillation?", "answer": ["Knowledge Distillation by On-the-Fly Native Ensemble", "Be Your Own Teacher: Improve the Performance of Convolutional Neural\n Networks via Self Distillation"], "answer_arxiv_id": ["1806.04606", "1905.08094"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_14671"} +{"question": "Could you provide me some studies about post-processing outputs from differentially private GAN-based models?", "answer": ["Private Post-GAN Boosting"], "answer_arxiv_id": ["2007.11934"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_14672"} +{"question": "Could you give me some examples of studies that apply in-context learning to various language tasks?", "answer": ["Using natural language prompts for machine translation", "MetaICL: Learning to Learn In Context", "Measuring and Narrowing the Compositionality Gap in Language Models"], "answer_arxiv_id": ["2202.11822", "2110.15943", "2210.03350"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_14673"} +{"question": "What paper discussed manipulating cross-attention layers for controlling the output of diffusion models in tuning-free TIE approaches?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control"], "answer_arxiv_id": ["2208.01626"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_14674"} +{"question": "Which works propose diffusing in a latent space to improve the generation quality and computational costs?", "answer": ["Score-based Generative Modeling in Latent Space", "Subspace Diffusion Generative Models"], "answer_arxiv_id": ["2106.05931", "2205.01490"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_14675"} +{"question": "Which papers have presented using polarimetric imaging to capture polarization cues of transparent objects?", "answer": ["Polarized Reflection Removal with Perfect Alignment in the Wild"], "answer_arxiv_id": ["2003.12789"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_14676"} +{"question": "Could you specify the studies that proposed learning safety guaranteed neural network controllers through verification-in-the-loop training?", "answer": ["Neural Certificates for Safe Control Policies", "Learning Safe Neural Network Controllers with Barrier Certificates", "ShieldNN: A Provably Safe NN Filter for Unsafe NN Controllers"], "answer_arxiv_id": ["2006.08465", "2009.09826", "2006.09564"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_14677"} +{"question": "Could you provide studies related to private mean estimation with respect to Mahalanobis distance?", "answer": ["A Fast Algorithm for Adaptive Private Mean Estimation"], "answer_arxiv_id": ["2301.07078"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_14678"} +{"question": "Which studies focus on fine-tuning-based TIE methods with the intention of synthesizing new images by model fine-tuning over domain-specific data?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Imagic: Text-Based Real Image Editing with Diffusion Models", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2210.09276", "2106.09685"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_14679"} +{"question": "Could you provide some studies on the approach where the choice of multi-dimensional discrete action is converted into a sequential choice of action across the dimensions at each time step?", "answer": ["Efficient Entropy for Policy Gradient with Multidimensional Action Space"], "answer_arxiv_id": ["1806.00589"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_14680"} +{"question": "Could you provide researches that have used data augmentation methods to diversify training data in Text-to-Audio Generation?", "answer": ["AudioGen: Textually Guided Audio Generation"], "answer_arxiv_id": ["2209.15352"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_14681"} +{"question": "What are the references that discuss the standard volume rendering technique used by NeRF and its successors?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains", "NeRF++: Analyzing and Improving Neural Radiance Fields", "Block-NeRF: Scalable Large Scene Neural View Synthesis", "TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2103.13415", "2111.12077", "2006.10739", "2010.07492", "2202.05263", "2203.09517"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14682"} +{"question": "Could you provide me some works about the methods based on episodic memory to reduce catastrophic forgetting?", "answer": ["Gradient Episodic Memory for Continual Learning", "Efficient Lifelong Learning with A-GEM", "Memory Efficient Experience Replay for Streaming Learning", "Learning to Learn without Forgetting By Maximizing Transfer and Minimizing Interference"], "answer_arxiv_id": ["1706.08840", "1812.00420", "1809.05922", "1810.11910"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_14683"} +{"question": "Could you provide me some works about black-box attack in Model Inversion?", "answer": ["GAMIN: An Adversarial Approach to Black-Box Model Inversion"], "answer_arxiv_id": ["1909.11835"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14684"} +{"question": "What works simultaneously achieved same approximation guarantee with amortized update time where k is the cardinality constraint?", "answer": ["Fully Dynamic Algorithm for Constrained Submodular Optimization"], "answer_arxiv_id": ["2006.04704v2"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_14685"} +{"question": "Could you provide me some studies about α-contamination model?", "answer": ["Mean-based Best Arm Identification in Stochastic Bandits under Reward Contamination"], "answer_arxiv_id": ["2111.07458"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_14686"} +{"question": "Which works were early approaches for Text-to-Video (T2V) generation?", "answer": ["Video Generation From Text", "Sync-DRAW: Automatic Video Generation using Deep Recurrent Attentive\n Architectures", "To Create What You Tell: Generating Videos from Captions"], "answer_arxiv_id": ["1710.00421", "1611.10314", "1804.08264"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_14687"} +{"question": "Which research empirically studies the training dynamics of a two-layer single neuron?", "answer": ["Implicit Regularization in ReLU Networks with the Square Loss"], "answer_arxiv_id": ["2012.05156"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_14688"} +{"question": "What works utilize techniques to constrain the Lipschitz constants of neural networks in practice?", "answer": ["Improved Training of Wasserstein GANs", "On Convergence and Stability of GANs", "Improving the Improved Training of Wasserstein GANs: A Consistency Term and Its Dual Effect", "Lipschitz Generative Adversarial Nets", "Spectral Normalization for Generative Adversarial Networks"], "answer_arxiv_id": ["1704.00028", "1705.07215", "1803.01541", "1902.05687", "1802.05957"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_14689"} +{"question": "What papers are about learning neural implicit functions from 3D ground truth?", "answer": ["Local Implicit Grid Representations for 3D Scenes", "Deep Local Shapes: Learning Local SDF Priors for Detailed 3D Reconstruction", "Convolutional Occupancy Networks", "acorn: Adaptive Coordinate Networks for Neural Scene Representation", "Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes", "Deep Implicit Moving Least-Squares Functions for 3D Reconstruction", "SA-ConvONet: Sign-Agnostic Optimization of Convolutional Occupancy Networks"], "answer_arxiv_id": ["2003.08981", "2003.10983", "2003.04618", "2105.02788", "2101.10994", "2103.12266", "2105.03582"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_14690"} +{"question": "What studies discuss the concept of shortcut learning?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_14691"} +{"question": "Which works present an improvement on the FedAvg algorithm through global aggregation methods?", "answer": ["Federated Learning with Matched Averaging", "Bayesian Nonparametric Federated Learning of Neural Networks", "FedBE: Making Bayesian Model Ensemble Applicable to Federated Learning"], "answer_arxiv_id": ["2002.06440", "1905.12022", "2009.01974"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_14692"} +{"question": "Any papers where numerical simulation was used for precise GT flows in real-world phenomena?", "answer": ["ScalarFlow: A Large-Scale Volumetric Data Set of Real-world Scalar Transport Flows for Computer Animation and Machine Learning", "Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge"], "answer_arxiv_id": ["2011.10284", "1711.07970"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_14693"} +{"question": "Can you provide me some papers focused on multi-modality large language models (MLLMs)?", "answer": ["Otter: A Multi-Modal Model with In-Context Instruction Tuning", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "InternLM-XComposer: A Vision-Language Large Model for Advanced\n Text-image Comprehension and Composition"], "answer_arxiv_id": ["2305.03726", "2304.15010", "2304.08485", "2305.06500", "2309.15112"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_14694"} +{"question": "Are there any studies on uncalibrated two-view pose estimation?", "answer": ["Scalability in Perception for Autonomous Driving: Waymo Open Dataset", "GRelPose: Generalizable End-to-End Relative Camera Pose Regression", "Relative Camera Pose Estimation Using Convolutional Neural Networks"], "answer_arxiv_id": ["1912.04838", "2211.14950", "1702.01381"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_14695"} +{"question": "What papers expanded on NeRF by improving scaling in unbounded scenes?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields"], "answer_arxiv_id": ["2103.13415", "2111.12077"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_14696"} +{"question": "Which study proposed the regularized learning of label shift (RLLS) method?", "answer": ["Regularized Learning for Domain Adaptation under Label Shifts"], "answer_arxiv_id": ["1903.09734"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14697"} +{"question": "Could you provide me some literature on supervised approaches to video summarization?", "answer": ["CLIP-It! Language-Guided Video Summarization", "Reconstructive Sequence-Graph Network for Video Summarization", "Video Summarization with Attention-Based Encoder-Decoder Networks", "Query-Focused Video Summarization: Dataset, Evaluation, and A Memory\n Network Based Approach", "Video Summarization with Long Short-term Memory"], "answer_arxiv_id": ["2107.00650", "2105.04066", "1708.09545", "1707.04960", "1605.08110"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_14698"} +{"question": "Is there a paper that discusses the computation of minimal sized atomic verifying set in polynomial time?", "answer": ["Verification and search algorithms for causal DAGs"], "answer_arxiv_id": ["2206.15374"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_14699"} +{"question": "Which studies utilized programming languages in the field of robotics?", "answer": ["Code as Policies: Language Model Programs for Embodied Control", "ProgPrompt: Generating Situated Robot Task Plans using Large Language\n Models", "VoxPoser: Composable 3D Value Maps for Robotic Manipulation with\n Language Models"], "answer_arxiv_id": ["2209.07753", "2209.11302", "2307.05973"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14700"} +{"question": "What research papers considered utility-based preferences, and assumed that the per-trajectory reward is linear in trajectory features?", "answer": ["Dueling Posterior Sampling for Preference-Based Reinforcement Learning", "Dueling RL: Reinforcement Learning with Trajectory Preferences"], "answer_arxiv_id": ["1908.01289", "2111.04850"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_14701"} +{"question": "Are there any papers that leverage Sign Distance Function (SDF) and represent the 3D scene by implicit surface?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "Improving neural implicit surfaces geometry with patch warping", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction", "Voxurf: Voxel-based Efficient and Accurate Neural Surface Reconstruction"], "answer_arxiv_id": ["2106.10689", "2106.12052", "2112.09648", "2104.10078", "2208.12697"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_14702"} +{"question": "Can you name some research that addresses the issue of different feature distributions for labeled and unlabeled data in semi-supervised learning?", "answer": ["An Information-theoretical Approach to Semi-supervised Learning under Covariate-shift", "Semi-Supervised Domain Generalization with Stochastic StyleMatch"], "answer_arxiv_id": ["2202.12123v1", "2106.00592"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_train_14703"} +{"question": "Can you provide me some studies that discuss the challenges when ViT-based segmentation approaches are implemented in a semi-supervised setting?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_14704"} +{"question": "Could you provide some references that discuss atom-by-atom generation in graph-based models?", "answer": ["Multi-Objective De Novo Drug Design with Conditional Graph Generative Model"], "answer_arxiv_id": ["1801.07299v3"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_14705"} +{"question": "Could you provide me a paper about a training-free method to reduce efforts of training with encoders for CLIP Model?", "answer": ["Tip-Adapter: Training-free Adaption of CLIP for Few-shot Classification"], "answer_arxiv_id": ["2207.09519"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14706"} +{"question": "What papers are about parameter tuning in transfer learning?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "FacT: Factor-Tuning for Lightweight Adaptation on Vision Transformer", "LST: Ladder Side-Tuning for Parameter and Memory Efficient Transfer Learning"], "answer_arxiv_id": ["2106.09685", "2212.03145", "2206.06522"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_14707"} +{"question": "What papers discuss learning to pick between different domain-specific QA models?", "answer": ["MetaQA: Combining Expert Agents for Multi-Skill Question Answering", "TWEAC: Transformer with Extendable QA Agent Classifiers", "UKP-SQuARE v3: A Platform for Multi-Agent QA Research"], "answer_arxiv_id": ["2112.01922", "2104.07081", "2303.18120"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_14708"} +{"question": "What work proposed to handle basis ambiguity through BasisNet?", "answer": ["Sign and Basis Invariant Networks for Spectral Graph Representation Learning"], "answer_arxiv_id": ["2202.13013"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_14709"} +{"question": "Which works convert the graph-structured skeleton into regular image-like input for skeleton-based action recognition task?", "answer": ["A New Representation of Skeleton Sequences for 3D Action Recognition", "2D/3D Pose Estimation and Action Recognition using Multitask Deep Learning", "Revisiting Skeleton-based Action Recognition"], "answer_arxiv_id": ["1703.03492", "1802.09232", "2104.13586"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_14710"} +{"question": "What works discuss the generation of an RGB image using a pre-trained text and depth-conditioned diffusion model?", "answer": ["Text2Tex: Text-driven Texture Synthesis via Diffusion Models", "TEXTure: Text-Guided Texturing of 3D Shapes"], "answer_arxiv_id": ["2303.11396", "2302.01721"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_14711"} +{"question": "What are some studies that focus on LLM output response-based defenses?", "answer": ["Self-Guard: Empower the LLM to Safeguard Itself", "LLM Self Defense: By Self Examination, LLMs Know They Are Being Tricked", "Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations", "Chain-of-Verification Reduces Hallucination in Large Language Models"], "answer_arxiv_id": ["2310.15851", "2308.07308", "2312.06674", "2309.11495"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_14712"} +{"question": "What works proposed training diffusion models on the latent space?", "answer": ["D2C: Diffusion-Denoising Models for Few-shot Conditional Generation", "Score-based Generative Modeling in Latent Space", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2106.06819", "2106.05931", "2112.10752"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_14713"} +{"question": "What are the studies that use concepts like Hypersphere Embedding and Max-Mahalanobis center (MMC) loss in adversarial training?", "answer": ["Rethinking Softmax Cross-Entropy Loss for Adversarial Robustness", "Boosting Adversarial Training with Hypersphere Embedding"], "answer_arxiv_id": ["1905.10626", "2002.08619"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14714"} +{"question": "What researchers analyzed the identifiability of fairness violations under general assumptions on the distribution and classifiers?", "answer": ["Assessing Algorithmic Fairness with Unobserved Protected Class Using Data Combination", "Evaluating Fairness of Machine Learning Models Under Uncertain and Incomplete Information", "Equalized odds postprocessing under imperfect group information"], "answer_arxiv_id": ["1906.00285", "2102.08410", "1906.03284"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_14715"} +{"question": "What work focuses on enhancing 3D object detection performance at the expense of reconstruction performance?", "answer": ["NeRF-RPN: A general framework for object detection in NeRFs"], "answer_arxiv_id": ["2211.11646"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_14716"} +{"question": "Can you list some studies that follow the paradigm of learning latent embeddings of nodes or networks for IM?", "answer": ["DISCO: Influence Maximization Meets Network Embedding and Deep Learning", "ToupleGDD: A Fine-Designed Solution of Influence Maximization by Deep Reinforcement Learning"], "answer_arxiv_id": ["1906.07378", "2210.07500"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_14717"} +{"question": "Could you provide the studies that discuss ensemble learning methods that take into account a present distribution shift?", "answer": ["Domain Adaptation for Statistical Classifiers", "Self-ensembling for visual domain adaptation", "Temporal Ensembling for Semi-Supervised Learning", "Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "There are Many Consistent Explanations of Unlabeled Data: Why You Should Average", "Algorithms and Theory for Multiple-Source Adaptation", "Deep Cocktail Network: Multi-source Unsupervised Domain Adaptation with Category Shift", "Domain Adaptive Ensemble Learning", "Asymmetric Tri-training for Unsupervised Domain Adaptation"], "answer_arxiv_id": ["1109.6341", "1706.05208", "1610.02242v3", "1703.01780", "1806.05594", "1805.08727", "1803.00830", "2003.07325", "1702.08400"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_14718"} +{"question": "Regarding 3D object detection, could you mention studies which have used V2V communication?", "answer": ["Collaboration Helps Camera Overtake LiDAR in 3D Detection", "Collaborative Semantic Occupancy Prediction with Hybrid Feature Fusion\n in Connected Automated Vehicles"], "answer_arxiv_id": ["2303.13560", "2402.07635"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_14719"} +{"question": "Could you provide me some works on contrastive learning and masked modeling?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["1911.05722", "2002.05709", "2006.09882", "1810.04805", "2111.06377"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_14720"} +{"question": "Which research papers are dealing with spatial understanding in large multimodal models?", "answer": ["Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest", "Ferret: Refer and Ground Anything Anywhere at Any Granularity", "MiniGPT-v2: large language model as a unified interface for\n vision-language multi-task learning", "Macaw-LLM: Multi-Modal Language Modeling with Image, Audio, Video, and\n Text Integration", "LISA: Reasoning Segmentation via Large Language Model"], "answer_arxiv_id": ["2306.15195", "2307.03601", "2310.07704", "2310.09478", "2306.09093", "2308.00692"], "source_meta": {"published_time": "20240617"}, "qid": "AutoScholarQuery_train_14721"} +{"question": "Could you name the papers that proposed methods to interpret each CAV’s role in decision-making?", "answer": ["From Attribution Maps to Human-Understandable Explanations through\n Concept Relevance Propagation", "A Holistic Approach to Unifying Automatic Concept Extraction and Concept\n Importance Estimation"], "answer_arxiv_id": ["2206.03208", "2306.07304"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_14722"} +{"question": "Are there works focusing on enhancing the perceptual quality of compressed images?", "answer": ["IEGAN: Multi-purpose Perceptual Quality Image Enhancement Using\n Generative Adversarial Network"], "answer_arxiv_id": ["1811.09134"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14723"} +{"question": "What research has proven that there exist 2-dimensional instances of their model where EoS convergence occur?", "answer": ["Second-order regression models exhibit progressive sharpening to the edge of stability"], "answer_arxiv_id": ["2210.04860"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_14724"} +{"question": "What work is more closely related to the layout-proposal-based method for scene layout proposals according to HOI triplets?", "answer": ["Exploiting Relationship for Complex-scene Image Generation"], "answer_arxiv_id": ["2104.00356"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_14725"} +{"question": "Can you cite the studies that worked on the functional map framework for shape correspondence?", "answer": ["Deep Functional Maps: Structured Prediction for Dense Shape\n Correspondence", "Deep Geometric Functional Maps: Robust Feature Learning for Shape\n Correspondence", "DPFM: Deep Partial Functional Maps", "Unsupervised Learning of Robust Spectral Shape Matching", "On Partial Shape Correspondence and Functional Maps"], "answer_arxiv_id": ["1704.08686", "2003.14286", "2110.09994", "2304.14419v1", "2310.14692"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_14726"} +{"question": "Could you provide me some works that display the superior generative capabilities of diffusion models in image generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Improved Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2102.09672", "2112.10752"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_14727"} +{"question": "Who are the researchers that focused on studying shared audio-language representation?", "answer": ["AudioCLIP: Extending CLIP to Image, Text and Audio", "Large-scale Contrastive Language-Audio Pretraining with Feature Fusion\n and Keyword-to-Caption Augmentation", "CLAP: Learning Audio Concepts From Natural Language Supervision"], "answer_arxiv_id": ["2106.13043", "2211.06687", "2206.04769"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_14728"} +{"question": "Which research papers discussed the difference between Grammatical Error Correction (GEC) and ASR error correction?", "answer": ["Generative Speech Recognition Error Correction with Large Language Models and Task-Activating Prompting", "Putting Natural in Natural Language Processing"], "answer_arxiv_id": ["2309.15649", "2305.04572"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_14729"} +{"question": "Which works are related to studying ambiguity in question answering?", "answer": ["AmbigQA: Answering Ambiguous Open-domain Questions", "SituatedQA: Incorporating Extra-Linguistic Contexts into QA"], "answer_arxiv_id": ["2004.10645", "2109.06157"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_14730"} +{"question": "Which papers exemplify the use of WikiHow dataset for teaching language models?", "answer": ["WikiHow: A Large Scale Text Summarization Dataset", "Reasoning about Goals, Steps, and Temporal Ordering with WikiHow"], "answer_arxiv_id": ["1810.09305", "2009.07690"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14731"} +{"question": "Could you name the studies using hashing methods for dimensionality reduction in machine learning?", "answer": ["Sketching as a Tool for Numerical Linear Algebra", "Low Rank Approximation and Regression in Input Sparsity Time"], "answer_arxiv_id": ["1411.4357", "1207.6365"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_14732"} +{"question": "What are the works on Generative Adversarial Networks (GANs) for super-resolution tasks?", "answer": ["Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network", "ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks", "Recovering Realistic Texture in Image Super-resolution by Deep Spatial Feature Transform", "RankSRGAN: Generative Adversarial Networks with Ranker for Image Super-Resolution", "Fourier Space Losses for Efficient Perceptual Image Super-Resolution", "SROBB: Targeted Perceptual Loss for Single Image Super-Resolution", "Structure-Preserving Super Resolution with Gradient Guidance", "Deep Unfolding Network for Image Super-Resolution", "Metric Learning based Interactive Modulation for Real-World Super-Resolution"], "answer_arxiv_id": ["1609.04802", "1809.00219", "1804.02815", "1908.06382", "2106.00783", "1908.07222", "2003.13081", "2003.10428", "2205.05065"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_14733"} +{"question": "Are there any works that used multi-class N-pair loss in the context of Clip training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20220423"}, "qid": "AutoScholarQuery_train_14734"} +{"question": "Which papers originally studied robust optimization that eventually led to the development of DRO?", "answer": ["Theory and Applications of Robust Optimization", "Holistic Robust Data-Driven Decisions"], "answer_arxiv_id": ["1010.5445", "2207.09560v3"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14735"} +{"question": "What research has utilized more advanced pre-trained generative models, such as the denoising diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14736"} +{"question": "Could you provide me some studies employing a mean-teacher method for more stable pseudo-labels training?", "answer": ["Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results"], "answer_arxiv_id": ["1703.01780"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_14737"} +{"question": "Which studies feature point-based methods in surface reconstruction?", "answer": ["Recurrent MVSNet for High-resolution Multi-view Stereo Depth Inference"], "answer_arxiv_id": ["1902.10556"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_14738"} +{"question": "What paper proposes method to improve model robustness on out-of-distribution data?", "answer": ["The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution\n Generalization"], "answer_arxiv_id": ["2006.16241"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_14739"} +{"question": "In which papers can I find that the training of a neural network by using gradient descent also defines a kernel named the neural tangent kernel (NTK)?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent"], "answer_arxiv_id": ["1806.07572", "1902.06720"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_14740"} +{"question": "What are some works that used box-supervised approaches for instance segmentation?", "answer": ["BoxInst: High-Performance Instance Segmentation with Box Annotations", "Learning to Segment via Cut-and-Paste"], "answer_arxiv_id": ["2012.02310", "1803.06414"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14741"} +{"question": "Which works utilize the filtering-based method for event denoising?", "answer": ["ProgressiveMotionSeg: Mutually Reinforced Framework for Event-Based\n Motion Segmentation", "A Unifying Contrast Maximization Framework for Event Cameras, with\n Applications to Motion, Depth, and Optical Flow Estimation"], "answer_arxiv_id": ["2203.11732", "1804.01306"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_14742"} +{"question": "In which paper the latest transformer-based architecture for pose and shape estimation from images is discussed?", "answer": ["Humans in 4D: Reconstructing and Tracking Humans with Transformers"], "answer_arxiv_id": ["2305.20091"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_14743"} +{"question": "Can you name the papers that reconstruct a semantic map incrementally using SLAM?", "answer": ["CNN-SLAM: Real-time dense monocular SLAM with learned depth prediction", "PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff and Things"], "answer_arxiv_id": ["1704.03489", "1903.01177"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_14744"} +{"question": "What works are about detecting and mitigating social biases in word embeddings?", "answer": ["Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings", "Gender Bias in Meta-Embeddings"], "answer_arxiv_id": ["1607.06520", "2205.09867"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_14745"} +{"question": "Which papers introduced a new approach that optimises a non-linear feature extractor with a single gradient step at each iteration in DTW?", "answer": ["Learning Discriminative Prototypes with Dynamic Time Warping"], "answer_arxiv_id": ["2103.09458"], "source_meta": {"published_time": "20230319"}, "qid": "AutoScholarQuery_train_14746"} +{"question": "What studies contributed to the development of Low-Rank Adaptation (LoRA) that reduces parameters in T2I personalization?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14747"} +{"question": "In which work is a dynamic metric learning method developed?", "answer": ["Dynamic Metric Learning: Towards a Scalable Metric Space to Accommodate Multiple Semantic Scales"], "answer_arxiv_id": ["2103.11781"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_14748"} +{"question": "Which work developed a pillar-based radar association module to associate radar features with corresponding detection results?", "answer": ["CenterFusion: Center-based Radar and Camera Fusion for 3D Object\n Detection"], "answer_arxiv_id": ["2011.04841"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14749"} +{"question": "What studies modify the conformal procedure to achieve approximate X-conditional coverage?", "answer": ["Localized Conformal Prediction: A Generalized Inference Framework for Conformal Prediction", "Conformal Prediction With Conditional Guarantees"], "answer_arxiv_id": ["2106.08460v2", "2305.12616"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_14750"} +{"question": "What are some examples of works that are categorized under detection-by-enhancement in low-light object detection?", "answer": ["Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement", "Toward Fast, Flexible, and Robust Low-Light Image Enhancement"], "answer_arxiv_id": ["2001.06826", "2204.10137"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_14751"} +{"question": "Could you specify the studies in which priors have been used to improve planning in model-based RL?", "answer": ["Latent Skill Planning for Exploration and Transfer", "Skill-based Model-based Reinforcement Learning"], "answer_arxiv_id": ["2011.13897", "2207.07560"], "source_meta": {"published_time": "20220120"}, "qid": "AutoScholarQuery_train_14752"} +{"question": "Could you provide me with references about neural attacks?", "answer": ["Attacking Recommender Systems with Augmented User Profiles", "Shilling Black-box Recommender Systems by Learning to Generate Fake User\n Profiles", "Targeted Data Poisoning Attack on News Recommendation System by Content\n Perturbation"], "answer_arxiv_id": ["2005.08164", "2206.11433", "2203.03560"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_14753"} +{"question": "Can you name a software library that offers functionalities similar to higher but requires users to use its own stateless modules?", "answer": ["Torchmeta: A Meta-Learning library for PyTorch"], "answer_arxiv_id": ["1909.06576"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_14754"} +{"question": "What are some follow-up studies to NeRF which explore dynamic and deformable scenes?", "answer": ["Nerfies: Deformable Neural Radiance Fields", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Video", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "TöRF: Time-of-Flight Radiance Fields for Dynamic Scene View Synthesis"], "answer_arxiv_id": ["2011.12948", "2012.12247", "2011.13961", "2109.15271"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_14755"} +{"question": "What studies have employed task augmentation as a strategy for few-task meta-learning?", "answer": ["Improving Generalization in Meta-learning via Task Augmentation", "STraTA: Self-Training with Task Augmentation for Better Few-shot Learning", "FlipDA: Effective and Robust Data Augmentation for Few-Shot Learning", "Cross-Domain Few-Shot Classification via Adversarial Task Augmentation"], "answer_arxiv_id": ["2007.13040", "2109.06270", "2108.06332", "2104.14385"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_14756"} +{"question": "Which papers discussed range-view-based methods to accommodate the LiDAR scanning mode?", "answer": ["Range Conditioned Dilated Convolutions for Scale Invariant 3D Object\n Detection", "RangeDet:In Defense of Range View for LiDAR-based 3D Object Detection", "To the Point: Efficient 3D Object Detection in the Range Image with\n Graph Convolution Kernels"], "answer_arxiv_id": ["2005.09927", "2103.10039", "2106.13381"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14757"} +{"question": "Could you please list some papers about the applications of INRs in microscopy imaging?", "answer": ["Recovery of Continuous 3D Refractive Index Maps from Discrete\n Intensity-Only Measurements using Neural Fields"], "answer_arxiv_id": ["2112.00002"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14758"} +{"question": "Are there any research papers on few-shot learning for instance segmentation?", "answer": ["One-Shot Instance Segmentation", "FGN: Fully Guided Network for Few-Shot Instance Segmentation"], "answer_arxiv_id": ["1811.11507", "2003.13954"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_14759"} +{"question": "Which study uses AI feedback in simulations for model development to improve harmlessness and helpfulness?", "answer": ["Constitutional AI: Harmlessness from AI Feedback"], "answer_arxiv_id": ["2212.08073"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_14760"} +{"question": "Which works use graph-based deep learning approaches to analyze the tumor microenvironment?", "answer": ["A Survey on Graph-Based Deep Learning for Computational Histopathology"], "answer_arxiv_id": ["2107.00272"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14761"} +{"question": "What works contributed to the development of Llama2?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models", "Root Mean Square Layer Normalization", "GLU Variants Improve Transformer", "RoFormer: Enhanced Transformer with Rotary Position Embedding"], "answer_arxiv_id": ["2307.09288", "1910.07467", "2002.05202", "2104.09864"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_14762"} +{"question": "Which studies focused on excluding occlusions while predicting facial features for photometric loss in 3D face reconstruction?", "answer": ["Self-Supervised Monocular 3D Face Reconstruction by Occlusion-Aware\n Multi-view Geometry Consistency"], "answer_arxiv_id": ["2007.12494"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_14763"} +{"question": "Which works have developed visualization tools for saliency maps given a trained DNN?", "answer": ["Understanding Deep Image Representations by Inverting Them", "Visualizing and Understanding Convolutional Networks", "Inverting Visual Representations with Convolutional Networks", "Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps"], "answer_arxiv_id": ["1412.0035", "1311.2901", "1506.02753", "1312.6034"], "source_meta": {"published_time": "20230106"}, "qid": "AutoScholarQuery_train_14764"} +{"question": "Which works have contributed to the field of LLMs and tool-use?", "answer": ["Language Models are Few-Shot Learners", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora\n with Web Data, and Web Data Only", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Language Models are Few-Shot Learners", "Training language models to follow instructions with human feedback", "Toolformer: Language Models Can Teach Themselves to Use Tools", "REVEAL: Retrieval-Augmented Visual-Language Pre-Training with\n Multi-Source Multimodal Knowledge Memory", "LMEye: An Interactive Perception Network for Large Language Models", "IdealGPT: Iteratively Decomposing Vision and Language Reasoning via\n Large Language Models"], "answer_arxiv_id": ["2005.14165", "1810.04805", "2307.09288", "2306.01116", "2201.11903", "2005.14165", "2203.02155", "2302.04761", "2212.05221", "2305.03701", "2305.14985"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14765"} +{"question": "What is one of the neural response retrieval systems that produce a single context embedding per turn?", "answer": ["ConveRT: Efficient and Accurate Conversational Representations from\n Transformers"], "answer_arxiv_id": ["1911.03688"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_14766"} +{"question": "Could you provide me some studies about the usage of large kernels in Swin Transformer?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2103.14030"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14767"} +{"question": "Which studies utilized querying the expert to obtain the optimal value function for guiding the expert intervention?", "answer": ["Reinforcement and Imitation Learning via Interactive No-Regret Learning", "Deeply AggreVaTeD: Differentiable Imitation Learning for Sequential Prediction"], "answer_arxiv_id": ["1406.5979", "1703.01030"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_14768"} +{"question": "What research has been conducted that utilizes implicit knowledge from pre-trained diffusion models?", "answer": ["Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "DifFace: Blind Face Restoration with Diffused Error Contraction"], "answer_arxiv_id": ["2201.11793", "2212.00490", "2212.06512"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14769"} +{"question": "Which research focuses on applications of LLMs in areas such as image and video processing?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_14770"} +{"question": "Are there any works about surgical workflow recognition and analysis in the medical field?", "answer": ["Comparative Validation of Machine Learning Algorithms for Surgical Workflow and Skill Analysis with the HeiChole Benchmark", "EndoNet: A Deep Architecture for Recognition Tasks on Laparoscopic\n Videos"], "answer_arxiv_id": ["2109.14956v1", "1602.03012"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_14771"} +{"question": "What studies have improved NeRF’s rendering quality by addressing issues such as aliasing artifacts, scalability, and capturing specular reflections?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance\n Fields"], "answer_arxiv_id": ["2103.13415", "2111.12077", "2112.03907"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_14772"} +{"question": "What papers proposed algorithms that maintain a buffer of experience from prior tasks to mitigate forgetfulness in continual reinforcement learning?", "answer": ["Gradient Episodic Memory for Continual Learning", "Selective Experience Replay for Lifelong Learning", "Scalable Recollections for Continual Lifelong Learning", "Experience Replay for Continual Learning", "Online Learned Continual Compression with Adaptive Quantization Modules"], "answer_arxiv_id": ["1706.08840", "1802.10269", "1711.06761", "1811.11682", "1911.08019"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_14773"} +{"question": "What papers utilized embedding features in devising OOD scoring functions?", "answer": ["A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "Extremely Simple Activation Shaping for Out-of-Distribution Detection", "Provable Guarantees for Understanding Out-of-distribution Detection"], "answer_arxiv_id": ["1807.03888", "2209.09858", "2112.00787"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_14774"} +{"question": "What studies adopt auxiliary tasks to encode inductive biases in image processing?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Deep Reinforcement and InfoMax Learning", "Contrastive Behavioral Similarity Embeddings for Generalization in Reinforcement Learning"], "answer_arxiv_id": ["2004.04136", "2006.07217", "2101.05265"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_14775"} +{"question": "Could you provide some works that used score distillation sampling, also utilized in AYG?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "TextMesh: Generation of Realistic 3D Meshes From Text Prompts", "DreamTime: An Improved Optimization Strategy for Diffusion-Guided 3D\n Generation", "IT3D: Improved Text-to-3D Generation with Explicit View Synthesis", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as\n General Image Priors", "NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with\n 360{\\deg} Views", "ATT3D: Amortized Text-to-3D Object Synthesis"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2303.13873", "2212.00774v1", "2304.12439", "2306.12422", "2308.11473", "2305.16213", "2211.07600", "2212.03267", "2211.16431", "2306.07349"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_14776"} +{"question": "Which works investigate the expressive power of classic GNNs and develop a more powerful structure?", "answer": ["How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["1810.00826"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14777"} +{"question": "Which works undertook the central DP model in distributed frequency estimation?", "answer": ["Universally Utility-Maximizing Privacy Mechanisms", "Concentrated Differential Privacy: Simplifications, Extensions, and Lower Bounds", "Differential Privacy on Finite Computers", "Federated Heavy Hitters Discovery with Differential Privacy", "Sample and Threshold Differential Privacy: Histograms and applications"], "answer_arxiv_id": ["0811.2841", "1605.02065", "1709.05396", "1902.08534", "2112.05693"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_14778"} +{"question": "Is there any research that proposed an attention-based architecture to learn in-context causal relationships between entities?", "answer": ["Learning to Induce Causal Structure"], "answer_arxiv_id": ["2204.04875"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14779"} +{"question": "Which studies have integrated Large Language Models (LLMs) for cross-modal compositional search with a single image and textual intent?", "answer": ["Grounding Language Models to Images for Multimodal Inputs and Outputs"], "answer_arxiv_id": ["2301.13823"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_14780"} +{"question": "What studies about uncertainty estimation use straightforward deep ensembles?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep\n Ensembles", "Maximizing Overall Diversity for Improved Uncertainty Estimates in Deep\n Ensembles"], "answer_arxiv_id": ["1612.01474", "1906.07380"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_14781"} +{"question": "Could you provide me some studies about the enhancement of Sequential Monte Carlo with NFs?", "answer": ["Annealed Flow Transport Monte Carlo"], "answer_arxiv_id": ["2102.07501"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_14782"} +{"question": "What paper elaborates on parameter-efficient tuning in language learning models?", "answer": ["Delta Tuning: A Comprehensive Study of Parameter Efficient Methods for Pre-trained Language Models"], "answer_arxiv_id": ["2203.06904v2"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_14783"} +{"question": "Which papers improved dense semantic correspondence used in CoCosNet?", "answer": ["CoCosNet v2: Full-Resolution Correspondence Learning for Image\n Translation"], "answer_arxiv_id": ["2012.02047"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14784"} +{"question": "Which work introduced the transformer architecture?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_14785"} +{"question": "What works are related to the study of self-supervised learning, focusing on contrastive methods?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "FaceNet: A Unified Embedding for Face Recognition and Clustering", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["2002.05709", "1503.03832", "1807.03748"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_14786"} +{"question": "What works adopt NCLMs for music generation?", "answer": ["MusicLM: Generating Music From Text", "SingSong: Generating musical accompaniments from singing", "VampNet: Music Generation via Masked Acoustic Token Modeling", "Simple and Controllable Music Generation"], "answer_arxiv_id": ["2301.11325", "2301.12662", "2307.04686", "2306.05284"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14787"} +{"question": "Which research papers have discussed the attributes of CNNs reflecting the spatial structure of natural images?", "answer": ["Deep Learning"], "answer_arxiv_id": ["1901.10233"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14788"} +{"question": "What research papers involve the use of a bi-level learning framework for data condensation?", "answer": ["Dataset Distillation"], "answer_arxiv_id": ["1811.10959"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_14789"} +{"question": "Could you provide me any works about Prompt Learning where learnable prompts are appended to the input to fine-tune models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "MaPLe: Multi-modal Prompt Learning", "Knowledge-Aware Prompt Tuning for Generalizable Vision-Language Models", "Read-only Prompt Optimization for Vision-Language Few-shot Learning", "Self-regulating Prompts: Foundational Model Adaptation without Forgetting"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2109.01134", "2203.05557", "2210.03117", "2308.11186", "2308.14960", "2307.06948v2"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14790"} +{"question": "Are there any studies using BERT embeddings for part-of-speech tagging?", "answer": ["Small and Practical BERT Models for Sequence Labeling"], "answer_arxiv_id": ["1909.00100"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_14791"} +{"question": "What works proposed methods based on conditional independence constraints for removing known spurious attributes?", "answer": ["Modeling the Data-Generating Process is Necessary for Out-of-Distribution Generalization", "Causally motivated Shortcut Removal Using Auxiliary Labels"], "answer_arxiv_id": ["2206.07837", "2105.06422"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_14792"} +{"question": "What research used latent vectors from full history in a cross-attention layer for image generation in the LDM pipeline?", "answer": ["Make-A-Story: Visual Memory Conditioned Consistent Story Generation"], "answer_arxiv_id": ["2211.13319"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_14793"} +{"question": "Any works about diffusion models applied in the field of text-driven motion generation?", "answer": ["MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model"], "answer_arxiv_id": ["2208.15001"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_14794"} +{"question": "Which works used additional criteria of targetness in Active Learning for Domain Adaptation?", "answer": ["Active Adversarial Domain Adaptation", "Active Domain Adaptation via Clustering Uncertainty-weighted Embeddings", "Discrepancy-Based Active Learning for Domain Adaptation", "Active Learning for Domain Adaptation: An Energy-Based Approach"], "answer_arxiv_id": ["1904.07848", "2010.08666", "2103.03757", "2112.01406"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_14795"} +{"question": "What works used structural dropout as a computationally efficient way to regularize the model?", "answer": ["Deep Networks with Stochastic Depth", "Reducing Transformer Depth on Demand with Structured Dropout"], "answer_arxiv_id": ["1603.09382", "1909.11556"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_14796"} +{"question": "Which papers used contrastive learning-based pretext tasks such as SimCLR or MOCO in deep clustering?", "answer": ["SCAN: Learning to Classify Images without Labels", "Contrastive Clustering", "You Never Cluster Alone"], "answer_arxiv_id": ["2005.12320", "2009.09687", "2106.01908"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_14797"} +{"question": "What works presented a framework named CFA to customize specific training configurations for each class?", "answer": ["CFA: Class-wise Calibrated Fair Adversarial Training"], "answer_arxiv_id": ["2303.14460"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14798"} +{"question": "Which papers first theoretically identified and empirically discovered the existence of a single, continuous manifold connecting global minimizers?", "answer": ["Topology and Geometry of Half-Rectified Network Optimization", "On Connected Sublevel Sets in Deep Learning", "Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Essentially No Barriers in Neural Network Energy Landscape"], "answer_arxiv_id": ["1611.01540", "1901.07417", "1802.10026", "1803.00885"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_14799"} +{"question": "Which works are considered to have been explicitly designed for federated learning settings, utilizing a client-centric method of adaptivity?", "answer": ["AdaScale SGD: A User-Friendly Algorithm for Distributed Training", "Adaptive Learning Rates for Faster Stochastic Gradient Methods"], "answer_arxiv_id": ["2007.05105", "2208.05287v1"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_14800"} +{"question": "What research introduced position embeddings to address length extrapolation in LLMs?", "answer": ["Train Short, Test Long: Attention with Linear Biases Enables Input\n Length Extrapolation", "A Length-Extrapolatable Transformer"], "answer_arxiv_id": ["2108.12409", "2212.10554"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14801"} +{"question": "Which paper proposed the generalized smoothness condition called (L0,L1) smoothness?", "answer": ["Why gradient clipping accelerates training: A theoretical justification for adaptivity"], "answer_arxiv_id": ["1905.11881"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_14802"} +{"question": "What work considered learning and stabilizing HGs with noisy preferences?", "answer": ["Noise Robust Core-stable Coalitions of Hedonic Games"], "answer_arxiv_id": ["2109.07738"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_14803"} +{"question": "Could you give me examples of articles suggesting that the sample efficiency of MuZero can be improved by using additional training signals?", "answer": ["Mastering Atari Games with Limited Data", "Procedural generalization by planning with self-supervised world models"], "answer_arxiv_id": ["2111.00210", "2111.01587"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_14804"} +{"question": "What studies considered dual-domain methods in sparse-view CT reconstruction?", "answer": ["Deep-neural-network based sinogram synthesis for sparse-view CT image reconstruction", "DuDoNet: Dual Domain Network for CT Metal Artifact Reduction"], "answer_arxiv_id": ["1803.00694v2", "1907.00273"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_14805"} +{"question": "Could you provide me some papers that analyse how adversarial attacks are carried out in the real world?", "answer": ["SoK: On the Semantic AI Security in Autonomous Driving", "Invisible for both Camera and LiDAR: Security of Multi-Sensor Fusion\n based Perception in Autonomous Driving Under Physical-World Attacks", "Does Physical Adversarial Example Really Matter to Autonomous Driving?\n Towards System-Level Effect of Adversarial Object Evasion Attack"], "answer_arxiv_id": ["2203.05314", "2106.09249", "2308.11894"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_14806"} +{"question": "Which researches have integrated generic SMILES grammars with neural networks for molecule generation?", "answer": ["Grammar Variational Autoencoder", "Syntax-Directed Variational Autoencoder for Structured Data"], "answer_arxiv_id": ["1703.01925", "1802.08786"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14807"} +{"question": "What studies have highlighted situations where model-based RL provides a significant benefit?", "answer": ["On the Expressivity of Neural Networks for Deep Reinforcement Learning"], "answer_arxiv_id": ["1910.05927"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_14808"} +{"question": "Which research improved variant function prediction performance by integrating retrieval-based methods with protein language models?", "answer": ["Tranception: protein fitness prediction with autoregressive transformers and inference-time retrieval"], "answer_arxiv_id": ["2205.13760"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_14809"} +{"question": "What paper introduced other transformations like image rotations for learning representations?", "answer": ["Equivariant Contrastive Learning"], "answer_arxiv_id": ["2111.00899"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_14810"} +{"question": "Could you provide me some studies that applied Machine Learning to improve BnB by learning to select nodes to expand?", "answer": ["Learning to Compare Nodes in Branch and Bound with Graph Neural Networks"], "answer_arxiv_id": ["2210.16934"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_14811"} +{"question": "Which papers introduced Machine Learning predictions to online caching optimization problems?", "answer": ["Near-Optimal Bounds for Online Caching with Machine Learned Advice", "Competitive caching with machine learned advice", "Parsimonious Learning-Augmented Caching"], "answer_arxiv_id": ["1910.12172", "1802.05399", "2202.04262"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_14812"} +{"question": "What papers conducted research on the neural network approximation of parametric nonlinear hyperbolic PDEs?", "answer": ["Error analysis for deep neural network approximations of parametric hyperbolic conservation laws", "Deep learning observables in computational fluid dynamics", "Iterative Surrogate Model Optimization (ISMO): An active learning algorithm for PDE constrained optimization with deep neural networks."], "answer_arxiv_id": ["2207.07362", "1903.03040", "2008.05730"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_14813"} +{"question": "Could you provide me some references proposing OoD scores using CLIP’s textual information?", "answer": ["Delving into Out-of-Distribution Detection with Vision-Language Representations", "Exploring the Limits of Out-of-Distribution Detection", "Zero-Shot Out-of-Distribution Detection Based on the Pre-trained Model CLIP"], "answer_arxiv_id": ["2211.13445", "2106.03004", "2109.02748"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_14814"} +{"question": "Which work introduces the DP-SGD algorithm with clipping?", "answer": ["Deep Learning with Differential Privacy"], "answer_arxiv_id": ["1607.00133"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_14815"} +{"question": "Are there studies discussing association between steep cliffs in RNN loss functions and bifurcations?", "answer": ["On the difficulty of training Recurrent Neural Networks"], "answer_arxiv_id": ["1211.5063"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_14816"} +{"question": "Could you give me some research that developed gradient-based methods for solving bilevel optimization problems?", "answer": ["Approximation Methods for Bilevel Programming", "A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic", "Reviving and Improving Recurrent Back-Propagation"], "answer_arxiv_id": ["1802.02246", "2007.05170v4", "1803.06396v4"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_14817"} +{"question": "Can you mention works which focused on the impact of irrelevant input context on natural language benchmarks?", "answer": ["Adversarial Examples for Evaluating Reading Comprehension Systems", "Learning to Attend On Essential Terms: An Enhanced Retriever-Reader Model for Open-domain Question Answering", "comps: Conceptual Minimal Pair Sentences for testing Robust Property Knowledge and its Inheritance in Pre-trained Language Models", "Large Language Models with Controllable Working Memory", "Capturing Failures of Large Language Models via Human Cognitive Biases", "P"], "answer_arxiv_id": ["1707.07328", "1808.09492", "2210.01963", "2211.05110", "2202.12299", "0704.0320"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_14818"} +{"question": "What studies propose auxiliary exploration mechanisms or training techniques to tackle the problem of state coverage?", "answer": ["Learning more skills through optimistic exploration", "The Option Keyboard Combining Skills in Reinforcement Learning", "Explore, Discover and Learn: Unsupervised Discovery of State-Covering Skills"], "answer_arxiv_id": ["2107.14226", "2106.13105", "2002.03647"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_14819"} +{"question": "Which papers explored the learning of 3D priors directly from 3D datasets using diffusion models?", "answer": ["LION: Latent Point Diffusion Models for 3D Shape Generation", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "Shap-E: Generating Conditional 3D Implicit Functions", "HyperDiffusion: Generating Implicit Neural Fields with Weight-Space\n Diffusion", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "MeshDiffusion: Score-based Generative 3D Mesh Modeling"], "answer_arxiv_id": ["2210.06978", "2212.08751", "2305.02463", "2303.17015", "2212.04493", "2303.08133"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14820"} +{"question": "Which work applied LLMs to enhance narrative coherence through prompt engineering and a simulated memory system in AI Dungeon3?", "answer": ["RecurrentGPT: Interactive Generation of (Arbitrarily) Long Text"], "answer_arxiv_id": ["2305.13304"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_14821"} +{"question": "Which papers focus on the extraction of TimeX and temporal relations?", "answer": ["Fine-Grained Temporal Relation Extraction"], "answer_arxiv_id": ["1902.01390"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14822"} +{"question": "Which studies contributed to accelerated sampling in diffusion probabilistic models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_14823"} +{"question": "What work explored the importance of understanding the static background for the downstream planning module in autonomous driving?", "answer": ["End-to-end Autonomous Driving: Challenges and Frontiers", "Planning-oriented Autonomous Driving", "DriveAdapter: Breaking the Coupling Barrier of Perception and Planning in End-to-End Autonomous Driving", "Scene as Occupancy"], "answer_arxiv_id": ["2306.16927", "2212.10156", "2308.00398", "2306.02851"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_14824"} +{"question": "What work proposed the Planning KL Divergence (PKL) metric used in task-aware motion prediction?", "answer": ["Learning to Evaluate Perception Models Using Planner-Centric Metrics"], "answer_arxiv_id": ["2004.08745"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_14825"} +{"question": "What work is closely related to the researcher's approach on performing agglomerative clustering of correlated covariates and on applying Lasso?", "answer": ["Correlated variables in regression: clustering and sparse estimation"], "answer_arxiv_id": ["1209.5908"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14826"} +{"question": "What studies applied the diffusion model to improve the perceptual quality in an autoencoder network?", "answer": ["A Residual Diffusion Model for High Perceptual Quality Codec Augmentation"], "answer_arxiv_id": ["2301.05489v3"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_14827"} +{"question": "Which works introduced the concept of attention mechanism in neural machine translation?", "answer": ["Neural Machine Translation by Jointly Learning to Align and Translate"], "answer_arxiv_id": ["1409.0473"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_14828"} +{"question": "Any works that heavily rely on knowing the exact cost function in optimal transport approaches?", "answer": ["Neural Optimal Transport with General Cost Functionals"], "answer_arxiv_id": ["2205.15403"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_14829"} +{"question": "What paper is about Differentiable Siamese Augmentation (DSA)?", "answer": ["Dataset Condensation with Differentiable Siamese Augmentation"], "answer_arxiv_id": ["2102.08259"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_14830"} +{"question": "What studies discuss the issue of capacity bottleneck or the curse of multilinguality in multilingual models?", "answer": ["Unsupervised Cross-lingual Representation Learning at Scale"], "answer_arxiv_id": ["1911.02116"], "source_meta": {"published_time": "20230418"}, "qid": "AutoScholarQuery_train_14831"} +{"question": "Are there any research papers focused on prompt tuning?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2107.13586v1", "2104.08691"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_14832"} +{"question": "What paper centered around building a billion image-text paired dataset for training large-scale vision-language models?", "answer": ["LAION-5B: An open large-scale dataset for training next generation\n image-text models"], "answer_arxiv_id": ["2210.08402"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14833"} +{"question": "What are the papers that discuss downstream video-language tasks?", "answer": ["Invariant Grounding for Video Question Answering", "Equivariant and Invariant Grounding for Video Question Answering", "Vision-and-Language Navigation: Interpreting visually-grounded\n navigation instructions in real environments"], "answer_arxiv_id": ["2206.02349", "2207.12783", "1711.07280"], "source_meta": {"published_time": "20240703"}, "qid": "AutoScholarQuery_train_14834"} +{"question": "Could you provide some studies that have used diffusion models for the conditional and unconditional image, video, and 3D object generation?", "answer": ["Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Imagen Video: High Definition Video Generation with Diffusion Models", "Novel View Synthesis with Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2112.10752", "2206.00927", "2209.14792", "2210.02303", "2210.04628"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_14835"} +{"question": "What works have exploited the intrinsic difference between the learning of clean and mislabeled examples to detect and correct misclassification errors?", "answer": ["Learning From Noisy Large-Scale Datasets With Minimal Supervision", "DivideMix: Learning with Noisy Labels as Semi-supervised Learning", "Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels", "Learning with Instance-Dependent Label Noise: A Sample Sieve Approach"], "answer_arxiv_id": ["1701.01619", "2002.07394", "1804.06872", "2010.02347"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_14836"} +{"question": "Which papers discuss representation-based methods for OOD detection?", "answer": ["SSD: A Unified Framework for Self-Supervised Outlier Detection", "Partial and Asymmetric Contrastive Learning for Out-of-Distribution Detection in Long-Tailed Recognition", "CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances", "Out-of-distribution Detection Learning with Unreliable Out-of-distribution Sources", "How to exploit hyperspherical embeddings for out-of-distribution detection?", "Mitigating Neural Network Overconfidence with Logit Normalization"], "answer_arxiv_id": ["2103.12051", "2207.01160", "2007.08176", "2311.03236", "2203.04450", "2205.09310"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_14837"} +{"question": "Which studies have tackled object manipulation tasks in presence of multiple objects in a single-task reinforcement learning setting?", "answer": ["COBRA: Data-Efficient Model-Based RL through Unsupervised Object Discovery and Curiosity-Driven Exploration", "A Perspective on Objects and Systematic Generalization in Model-Based RL", "Entity Abstraction in Visual Model-Based Reinforcement Learning"], "answer_arxiv_id": ["1905.09275", "1906.01035", "1910.12827"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_14838"} +{"question": "Which papers studied Unsupervised Domain Adaptation (UDA) in computer vision tasks?", "answer": ["Both Style and Distortion Matter: Dual-Path Unsupervised Domain\n Adaptation for Panoramic Semantic Segmentation", "Look at the Neighbor: Distortion-aware Unsupervised Domain Adaptation\n for Panoramic Semantic Segmentation", "Category Contrast for Unsupervised Domain Adaptation in Visual Tasks", "Lifelong Unsupervised Domain Adaptive Person Re-identification with\n Coordinated Anti-forgetting and Adaptation", "UDA-COPE: Unsupervised Domain Adaptation for Category-level Object Pose\n Estimation", "The Norm Must Go On: Dynamic Unsupervised Domain Adaptation by\n Normalization", "Domain Adaptive Semantic Segmentation Using Weak Labels", "Unsupervised Domain Adaptation for Nighttime Aerial Tracking", "Spectral Unsupervised Domain Adaptation for Visual Recognition", "Semantics, Distortion, and Style Matter: Towards Source-free UDA for\n Panoramic Segmentation"], "answer_arxiv_id": ["2303.14360", "2308.05493", "2106.02885", "2112.06632", "2111.12580", "2112.00463", "2007.15176", "2203.10541", "2106.06112v3", "2403.12505"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_14839"} +{"question": "Which studies proposed learning first-order logical rules for inductive knowledge graph completion?", "answer": ["Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "DRUM: End-To-End Differentiable Rule Mining On Knowledge Graphs"], "answer_arxiv_id": ["1702.08367", "1911.00055"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_14840"} +{"question": "Who pointed out that the sensitivity of adding or removing one example with microbatch gradient clipping is actually 2*C?", "answer": ["How to DP-fy ML: A Practical Guide to Machine Learning with Differential Privacy"], "answer_arxiv_id": ["2303.00654"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_14841"} +{"question": "Can you name studies related to locally-informed proposals and adaptive MCMC in structure learning algorithms?", "answer": ["Adaptive random neighbourhood informed Markov chain Monte Carlo for high-dimensional Bayesian variable selection"], "answer_arxiv_id": ["2110.11747"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_14842"} +{"question": "What works observed that training dynamics with large learning rates differ from the small learning rate regime?", "answer": ["The Two Regimes of Deep Network Training"], "answer_arxiv_id": ["2002.10376"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_14843"} +{"question": "Could you provide me some works that implemented storage optimizations in improving rendering speed?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Variable Bitrate Neural Fields", "Plenoxels: Radiance Fields without Neural Networks", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction", "Neural Sparse Voxel Fields"], "answer_arxiv_id": ["2201.05989", "2206.07707", "2112.05131", "2111.11215", "2007.11571"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_14844"} +{"question": "Could you provide some researches that investigated 3D semantic segmentation?", "answer": ["JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Clouds", "PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space", "PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection", "3DMV: Joint 3D-Multi-View Prediction for 3D Semantic Scene Segmentation", "OctNet: Learning Deep 3D Representations at High Resolutions"], "answer_arxiv_id": ["2007.06888", "1612.00593", "1706.02413", "1912.13192", "1803.10409", "1611.05009"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_14845"} +{"question": "Which paper presents a memory-guided semantic segmentation method that abstracts the conceptual knowledge of semantic classes into the memory?", "answer": ["Pin the Memory: Learning to Generalize Semantic Segmentation"], "answer_arxiv_id": ["2204.03609"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_14846"} +{"question": "Which studies apply FastMap for efficient search in graph optimization and shortest path?", "answer": ["The FastMap Algorithm for Shortest Path Computations", "Fast Rates for Contextual Linear Optimization"], "answer_arxiv_id": ["1706.02792", "2011.03030v3"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_14847"} +{"question": "Could you provide me some studies regarding example-based explanation methods utilised for AI-assisted decision making?", "answer": ["Towards a Science of Human-AI Decision Making: A Survey of Empirical Studies"], "answer_arxiv_id": ["2112.11471"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_14848"} +{"question": "What research focused on agent’s performance to reason about plasticity?", "answer": ["An empirical study of implicit regularization in deep offline RL"], "answer_arxiv_id": ["2207.02099"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_14849"} +{"question": "Which works revealed principles and implications of peripheral vision mechanisms?", "answer": ["Can Peripheral Representations Improve Clutter Metrics on Complex\n Scenes?", "Emergent Properties of Foveated Perceptual Systems"], "answer_arxiv_id": ["1608.04042", "2006.07991"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14850"} +{"question": "What papers have techniques for extracting video features used to synthesize sounds as a part of video-to-audio generation?", "answer": ["Generating Visually Aligned Sound from Videos", "Taming Visually Guided Sound Generation", "I hear your true colors: Image Guided Audio Generation"], "answer_arxiv_id": ["2008.00820", "2110.08791v1", "2211.03089"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_14851"} +{"question": "Which papers studied the convergence of gradient decent for smooth activations using the Neural Tangent Kernel (NTK) parameterization?", "answer": ["Gradient Descent Finds Global Minima of Deep Neural Networks"], "answer_arxiv_id": ["1811.03804"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_14852"} +{"question": "What works proposed fine-tuning and employing small task-specific adapters for cross-domain classification?", "answer": ["A Broader Study of Cross-Domain Few-Shot Learning", "Boosting the Generalization Capability in Cross-Domain Few-shot Learning\n via Noise-enhanced Supervised Autoencoder", "Cross-domain Few-shot Learning with Task-specific Adapters"], "answer_arxiv_id": ["1912.07200", "2108.05028", "2107.00358"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_14853"} +{"question": "Could you provide me some works that proposed temporally-extended exploration techniques to enhance simple exploration strategies?", "answer": ["Deep Exploration via Bootstrapped DQN", "TAAC: Temporally Abstract Actor-Critic for Continuous Control"], "answer_arxiv_id": ["1602.04621", "2104.06521"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_14854"} +{"question": "Which works drew a connection between self-attention and nonparametric kernel regression?", "answer": ["Improving Transformers with Probabilistic Attention Keys"], "answer_arxiv_id": ["2110.08678"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_14855"} +{"question": "Which works discuss the utilization of instructional videos for multimodal video understanding tasks?", "answer": ["Towards Automatic Learning of Procedures from Web Instructional Videos", "COIN: A Large-scale Dataset for Comprehensive Instructional Video Analysis", "Cross-task weakly supervised learning from instructional videos", "HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips"], "answer_arxiv_id": ["1703.09788", "1903.02874", "1903.08225", "1906.03327"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_14856"} +{"question": "Which research paper is about instruction tuning of language model on a diverse instruction dataset?", "answer": ["Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2210.11416"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_14857"} +{"question": "What studies utilized the SketchGraphs dataset for generative model of CAD sketches and other applications of learning in physical design?", "answer": ["Engineering Sketch Generation for Computer-Aided Design", "Vitruvion: A Generative Model of Parametric CAD Sketches", "SketchGen: Generating Constrained CAD Sketches"], "answer_arxiv_id": ["2104.09621", "2109.14124", "2106.02711"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_14858"} +{"question": "Which papers used point clouds as the 3D representation?", "answer": ["A Point Set Generation Network for 3D Object Reconstruction from a\n Single Image"], "answer_arxiv_id": ["1612.00603"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_14859"} +{"question": "Are there studies that extended causal Bandit Optimization utilizing data from different intervention targets?", "answer": ["Multi-task Causal Learning with Gaussian Processes"], "answer_arxiv_id": ["2009.12821v1"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_14860"} +{"question": "What work developed training methods that involved training with stylized images to disconnect texture information from the class label?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness"], "answer_arxiv_id": ["1811.12231"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_14861"} +{"question": "What are some studies of conservative exploration in bandits and tabular MDPs?", "answer": ["Conservative Bandits", "Conservative Contextual Linear Bandits", "Improved Algorithms for Conservative Exploration in Bandits", "Conservative Exploration in Reinforcement Learning"], "answer_arxiv_id": ["1602.04282", "1611.06426", "2002.03221", "2002.03218v2"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_14862"} +{"question": "What studies combine MaskFormer with CLIP to improve open vocabulary semantic segmentation?", "answer": ["Decoupling Zero-Shot Semantic Segmentation", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2112.07910", "2303.04803"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_14863"} +{"question": "Any works introduced the task of Active Speaker Localization in an egocentric scene?", "answer": ["Egocentric Deep Multi-Channel Audio-Visual Active Speaker Localization"], "answer_arxiv_id": ["2201.01928"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_14864"} +{"question": "What works focus on dynamic Bayesian networks in a setting where temporal information is available?", "answer": ["Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICA", "Partial Disentanglement via Mechanism Sparsity", "Learning Temporally Causal Latent Processes from General Temporal Data", "CITRIS: Causal Identifiability from Temporal Intervened Sequences", "Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systems"], "answer_arxiv_id": ["2107.10098", "2207.07732", "2110.05428", "2202.03169", "2206.06169"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14865"} +{"question": "Could you provide me studies that emphasized the impact of synonym perturbations on text detection performance?", "answer": ["Red Teaming Language Model Detectors with Language Models"], "answer_arxiv_id": ["2305.19713"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_14866"} +{"question": "Which studies refined full-body motions prediction using physics-based optimization and egocentric SLAM system?", "answer": ["Physical Inertial Poser (PIP): Physics-aware Real-time Human Motion\n Tracking from Sparse Inertial Sensors", "EgoLocate: Real-time Motion Capture, Localization, and Mapping with\n Sparse Body-mounted Sensors"], "answer_arxiv_id": ["2203.08528", "2305.01599"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_14867"} +{"question": "Which studies focus on developing discrete environment in Vision-and-Language Navigation (VLN)?", "answer": ["Vision-and-Language Navigation: Interpreting visually-grounded\n navigation instructions in real environments", "Room-Across-Room: Multilingual Vision-and-Language Navigation with Dense\n Spatiotemporal Grounding", "REVERIE: Remote Embodied Visual Referring Expression in Real Indoor\n Environments"], "answer_arxiv_id": ["1711.07280", "2010.07954", "1904.10151"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_14868"} +{"question": "What studies discussed the over-smoothing phenomenon in GNNs?", "answer": ["Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning", "A Survey on Oversmoothing in Graph Neural Networks", "A Note on Over-Smoothing for Graph Neural Networks"], "answer_arxiv_id": ["1801.07606", "2303.10993", "2006.13318"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14869"} +{"question": "Could you provide me some works that model explicitly 3D flow and reduce the reconstruction to a canonical (static) one in NeRF extensions?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Novel View Synthesis of Dynamic Scenes with Globally Coherent Depths from a Monocular Camera", "Nerfies: Deformable Neural Radiance Fields", "Neural Trajectory Fields for Dynamic Novel View Synthesis", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Video", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Neural Radiance Flow for 4D View Synthesis and Video Processing", "STaR: Self-supervised Tracking and Reconstruction of Rigid Objects in Motion with Neural Rendering", "NeRFPlayer: A Streamable Dynamic Scene Representation with Decomposed Neural Radiance Fields", "Neural Deformable Voxel Grid for Fast Optimization of Dynamic View Synthesis", "Fast Dynamic Radiance Fields with Time-Aware Neural Voxels", "DynIBaR: Neural Dynamic Image-Based Rendering", "DeVRF: Fast Deformable Voxel Radiance Fields for Dynamic Scenes"], "answer_arxiv_id": ["2011.13961", "2004.01294", "2011.12948", "2105.05994", "2012.12247", "2011.13084", "2012.09790", "2101.01602", "2210.15947", "2206.07698v2", "2205.15285", "2211.11082", "2205.15723"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_14870"} +{"question": "Which papers present SQ lower bounds for learning Gaussian mixture models and list-decodable linear regression?", "answer": ["Statistical Query Lower Bounds for Robust Estimation of High-Dimensional Gaussians and Gaussian Mixtures", "Statistical Query Lower Bounds for List-Decodable Linear Regression"], "answer_arxiv_id": ["1611.03473", "2106.09689"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_14871"} +{"question": "Which papers introduced large-scale generative diffusion models like Stable Diffusion (SD) and DALLE-2?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2204.06125"], "source_meta": {"published_time": "20240105"}, "qid": "AutoScholarQuery_train_14872"} +{"question": "What study introduced iterative render-and-compare strategy for 6D pose estimation?", "answer": ["DeepIM: Deep Iterative Matching for 6D Pose Estimation"], "answer_arxiv_id": ["1804.00175"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_14873"} +{"question": "What study improves for allowing greater flexibility for reading and writing in the context of building differentiable versions of storage components?", "answer": ["End-To-End Memory Networks"], "answer_arxiv_id": ["1503.08895"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_14874"} +{"question": "What papers are about finding the optimal couplings in high dimensional continuous measures?", "answer": ["Large-Scale Optimal Transport and Mapping Estimation", "Do Neural Optimal Transport Solvers Work? A Continuous Wasserstein-2 Benchmark", "Neural Optimal Transport", "Optimal transport mapping via input convex neural networks", "Generative Modeling with Optimal Transport Maps", "Score-based Generative Neural Networks for Large-Scale Optimal Transport"], "answer_arxiv_id": ["1711.02283", "2106.01954", "2201.12220", "1908.10962", "2110.02999", "2110.03237"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_14875"} +{"question": "Which works introduced Retrieval-Augmented Generation (RAG) in LLMs by using external database knowledge?", "answer": ["Retrieval-Augmented Generation for Large Language Models: A Survey", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Toolformer: Language Models Can Teach Themselves to Use Tools", "Generalization through Memorization: Nearest Neighbor Language Models"], "answer_arxiv_id": ["2312.10997", "2005.11401", "2302.04761", "1911.00172"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_14876"} +{"question": "Which work benchmarks ImageNet classification performance across different backbones?", "answer": ["ResNet strikes back: An improved training procedure in timm"], "answer_arxiv_id": ["2110.00476"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_14877"} +{"question": "In what research has the method of speculative decoding to accelerate LLM decoding been discussed?", "answer": ["Accelerating Large Language Model Decoding with Speculative Sampling", "Fast Inference from Transformers via Speculative Decoding", "Inference with Reference: Lossless Acceleration of Large Language Models", "Language Model Cascades: Token-level uncertainty and beyond"], "answer_arxiv_id": ["2302.01318", "2211.17192", "2304.04487", "2404.10136v1"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_14878"} +{"question": "Could you tell me about the research papers that investigated nonparametric generative models using kernel methods?", "answer": ["A Spectral Approach to Gradient Estimation for Implicit Distributions", "Gradient Estimators for Implicit Models", "Nonparametric Score Estimators"], "answer_arxiv_id": ["1806.02925", "1705.07107", "2005.10099"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_14879"} +{"question": "What works propose or study the (L0,L1)-smoothness condition?", "answer": ["Why gradient clipping accelerates training: A theoretical justification for adaptivity", "Improved Analysis of Clipping Algorithms for Non-convex Optimization", "Robustness to Unbounded Smoothness of Generalized SignSGD", "Variance-reduced Clipping for Non-convex Optimization"], "answer_arxiv_id": ["1905.11881", "2010.02519", "2208.11195", "2303.00883"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14880"} +{"question": "Could you provide me some studies about ensemble distillation methods used to improve the generalization of global models?", "answer": ["Ensemble Distillation for Robust Model Fusion in Federated Learning"], "answer_arxiv_id": ["2006.07242"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_14881"} +{"question": "What studies about the transformers being used in various fields of computer vision?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2010.11929", "2103.14030"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_14882"} +{"question": "Which studies first extracted image and language features separately in the field of referring image segmentation?", "answer": ["Segmentation from Natural Language Expressions", "Recurrent Multimodal Interaction for Referring Image Segmentation"], "answer_arxiv_id": ["1603.06180", "1703.07939"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14883"} +{"question": "What papers focus on post-training quantization (PTQ) methods?", "answer": ["Post-training 4-bit quantization of convolution networks for rapid-deployment", "ZeroQ: A Novel Zero Shot Quantization Framework", "Low-bit Quantization of Neural Networks for Efficient Inference", "Improving Post Training Neural Quantization: Layer-wise Calibration and Integer Programming", "Quantizing deep convolutional networks for efficient inference: A whitepaper", "Brecq: pushing the limit of post-training quantization by block reconstruction", "Same, Same But Different - Recovering Neural Network Quantization Error Through Weight Factorization", "Data-Free Quantization Through Weight Equalization and Bias Correction", "Up or Down? Adaptive Rounding for Post-Training Quantization", "Improving Neural Network Quantization without Retraining using Outlier Channel Splitting"], "answer_arxiv_id": ["1810.05723v3", "2001.00281", "1902.06822", "2006.10518", "1806.08342", "2102.05426", "1902.01917", "1906.04721", "2004.10568", "1901.09504"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_14884"} +{"question": "Could you provide me some examples of datasets that were collected with specialized hardware?", "answer": ["Google Scanned Objects: A High-Quality Dataset of 3D Scanned Household\n Items", "OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic\n Perception, Reconstruction and Generation"], "answer_arxiv_id": ["2204.11918", "2301.07525"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_14885"} +{"question": "What studies estimate the gradient at the global step using Taylor expansion?", "answer": ["Asynchronous Stochastic Gradient Descent with Delay Compensation"], "answer_arxiv_id": ["1609.08326"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14886"} +{"question": "Which papers mentioned bootstrapping strategy in the context of fully-labeled classification?", "answer": ["Self-training with Noisy Student improves ImageNet classification"], "answer_arxiv_id": ["1911.04252"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_14887"} +{"question": "In what work showed that allocation 2≤w⋆​(μ)i⋆−1≤K−1+12 for Gaussian distributions?", "answer": ["A Non-asymptotic Approach to Best-Arm Identification for Gaussian Bandits"], "answer_arxiv_id": ["2105.12978"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_14888"} +{"question": "Could you provide me some works contributing to creating smaller and more efficient models through knowledge distillation?", "answer": ["Distilling the Knowledge in a Neural Network", "Do Deep Convolutional Nets Really Need to be Deep and Convolutional?", "Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion"], "answer_arxiv_id": ["1503.02531", "1603.05691", "1912.08795"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_14889"} +{"question": "Could you give some references that introduced the npf model which applies a graph neural network and employed local latent variables instead of a global one?", "answer": ["The Functional Neural Process"], "answer_arxiv_id": ["1906.08324"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_14890"} +{"question": "What works provided quadrature nodes that achieve spectral accuracy?", "answer": ["Kernel interpolation with continuous volume sampling"], "answer_arxiv_id": ["2002.09677"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_14891"} +{"question": "Which study focused on tabular Markov Decision Process (MDP) with scalar rewards and their algorithm can learn a pessimistic estimate of the true inverse CDF of the return?", "answer": ["Conservative Offline Distributional Reinforcement Learning"], "answer_arxiv_id": ["2107.06106"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_14892"} +{"question": "Could you mention some works that provided efficient algorithms to achieve specific error under Gaussian and log-concave marginals in the zero-bias setting of ReLU regression problem?", "answer": ["Learning Neural Networks with Two Nonlinear Layers in Polynomial Time"], "answer_arxiv_id": ["1709.06010"], "source_meta": {"published_time": "20220804"}, "qid": "AutoScholarQuery_train_14893"} +{"question": "What works focused on utilizing pre-trained computer vision models to approximate human perceptual similarity judgments over images?", "answer": ["Evaluating (and improving) the correspondence between deep neural networks and human representations"], "answer_arxiv_id": ["1706.02417"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_14894"} +{"question": "Which papers used diffusion models for super-resolution?", "answer": ["Image Super-Resolution via Iterative Refinement", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic Models"], "answer_arxiv_id": ["2104.07636", "2104.14951"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_14895"} +{"question": "Which references introduce methods for likelihood metrics measurement, particularly average negative log likelihood?", "answer": ["Convolutional Social Pooling for Vehicle Trajectory Prediction", "The Trajectron: Probabilistic Multi-Agent Trajectory Modeling With Dynamic Spatiotemporal Graphs"], "answer_arxiv_id": ["1805.06771", "1810.05993"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_14896"} +{"question": "Are there any works that make use of 222D anchor points in DETR?", "answer": ["Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2010.04159"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_14897"} +{"question": "Which works obtained small-return regret for tabular and linear MDPs via concentration bounds that scale with the variance?", "answer": ["Reward-Free Exploration for Reinforcement Learning", "First-Order Regret in Reinforcement Learning with Linear Function Approximation: A Robust Estimation Approach"], "answer_arxiv_id": ["2002.02794", "2112.03432"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_14898"} +{"question": "What is the study where the BRAC was used as an offline backbone algorithm?", "answer": ["Behavior Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["1911.11361"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_14899"} +{"question": "What work proposed the idea of Post-hoc CBM?", "answer": ["Post-hoc Concept Bottleneck Models"], "answer_arxiv_id": ["2205.15480"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_14900"} +{"question": "Are there any works that explore eliciting values and personal traits via prompt engineering?", "answer": ["Identifying and Manipulating the Personality Traits of Language Models", "Can ChatGPT Assess Human Personalities? A General Evaluation Framework"], "answer_arxiv_id": ["2212.10276", "2303.01248"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_14901"} +{"question": "What study incorporates the DreamBooth fine-tuning process of Stable Diffusion and designs a bi-level min-max optimization process to generate protective perturbations?", "answer": ["Anti-DreamBooth: Protecting users from personalized text-to-image\n synthesis"], "answer_arxiv_id": ["2303.15433"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_14902"} +{"question": "What work introduces a hierarchical masking strategy to learn multi-scale features for a hybrid convolution-transformer encoder?", "answer": ["ConvMAE: Masked Convolution Meets Masked Autoencoders"], "answer_arxiv_id": ["2205.03892"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_14903"} +{"question": "What are the reference papers for the powerful LLMs, such as FLAN-PaLM-540B and GPT-4?", "answer": ["Scaling Instruction-Finetuned Language Models", "GPT-4 Technical Report"], "answer_arxiv_id": ["2210.11416", "2303.08774"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_14904"} +{"question": "Can you mention some classical works in Audio-Visual Sound Localization that achieved improving performance over time?", "answer": ["Learning to Localize Sound Source in Visual Scenes", "Learning to Localize Sound Sources in Visual Scenes: Analysis and Applications", "Deep Audio-Visual Learning: A Survey", "Localizing Visual Sounds the Hard Way", "Localizing Visual Sounds the Easy Way", "Self-Supervised Predictive Learning: A Negative-Free Method for Sound Source Localization in Visual Scenes", "Exploiting Transformation Invariance and Equivariance for Self-supervised Sound Localisation"], "answer_arxiv_id": ["1803.03849v1", "1911.09649", "2001.04758", "2104.02691", "2203.09324", "2203.13412v1", "2206.12772"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_14905"} +{"question": "What works have applied regularization techniques during the training of NeRF?", "answer": ["Depth-supervised NeRF: Fewer Views and Faster Training for Free", "Dense Depth Priors for Neural Radiance Fields from Sparse Input Views", "DINER: Depth-aware Image-based NEural Radiance fields", "SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis", "Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis", "Learning Transferable Visual Models From Natural Language Supervision", "InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Rendering", "RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs"], "answer_arxiv_id": ["2107.02791", "2112.03288", "2211.16630", "2303.16196", "2104.00677", "2103.00020", "2112.15399", "2112.00724"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14906"} +{"question": "What are some examples of model-based disentangled methods in the multi-view learning?", "answer": ["Multi-level Feature Learning for Contrastive Multi-view Clustering", "Reconsidering Representation Alignment for Multi-view Clustering", "Contrastive Multiview Coding"], "answer_arxiv_id": ["2106.11193", "2103.07738", "1906.05849"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_14907"} +{"question": "Could you tell me about a few studies that have used reinforcement learning to improve LLM's text generation capabilities?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_14908"} +{"question": "Which dataset has been used in 3D model training?", "answer": ["ShapeNet: An Information-Rich 3D Model Repository", "Objaverse: A Universe of Annotated 3D Objects"], "answer_arxiv_id": ["1512.03012", "2212.08051"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_14909"} +{"question": "Which papers focused on developing methods for segmenting 3D scenes into planes?", "answer": ["Fully-Convolutional Point Networks for Large-Scale Point Clouds", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff\n and Things", "Panoptic Multi-TSDFs: a Flexible Representation for Online\n Multi-resolution Volumetric Mapping and Long-term Dynamic Scene Consistency", "PanopticNDT: Efficient and Robust Panoptic Mapping"], "answer_arxiv_id": ["1808.06840", "1904.08755", "1903.01177", "2109.10165", "2309.13635"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_14910"} +{"question": "What works on diffusion models use them for representation learning?", "answer": ["Label-Efficient Semantic Segmentation with Diffusion Models", "Diffusion Autoencoders: Toward a Meaningful and Decodable Representation"], "answer_arxiv_id": ["2112.03126", "2111.15640"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_14911"} +{"question": "Can you name some research works that used SSL techniques in image and video analysis?", "answer": ["VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "A Cookbook of Self-Supervised Learning"], "answer_arxiv_id": ["2105.04906", "2304.12210"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_14912"} +{"question": "Can you mention some research papers that discuss Federated Learning that aim at solving single-objective minimization problems?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_14913"} +{"question": "Which papers have proposed efficient estimation of OT maps using neural OT models?", "answer": ["Wasserstein-2 Generative Networks", "Do Neural Optimal Transport Solvers Work? A Continuous Wasserstein-2 Benchmark", "Large-Scale Wasserstein Gradient Flows", "Continuous Wasserstein-2 Barycenter Estimation without Minimax Optimization"], "answer_arxiv_id": ["1909.13082v4", "2106.01954", "2106.00736", "2102.01752"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_14914"} +{"question": "What studies examined the MSMO task?", "answer": ["VMSMO: Learning to Generate Multimodal Summary for Video-based News\n Articles", "Multi-modal Summarization for Video-containing Documents", "MHMS: Multimodal Hierarchical Multimedia Summarization", "Hierarchical Cross-Modality Semantic Correlation Learning Model for\n Multimodal Summarization", "Semantics-Consistent Cross-domain Summarization via Optimal Transport\n Alignment", "UniMS: A Unified Framework for Multimodal Summarization with Knowledge\n Distillation", "Align and Attend: Multimodal Summarization with Dual Contrastive Losses", "TLDW: Extreme Multimodal Summarisation of News Videos"], "answer_arxiv_id": ["2010.05406", "2009.08018", "2204.03734", "2112.12072", "2210.04722", "2109.05812", "2303.07284", "2210.08481"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_14915"} +{"question": "What is the original architecture which motivates the inter-dependence among words of a sentence in NLP tasks?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_14916"} +{"question": "Which works incorporated T5 language model into their image generation approach?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers"], "answer_arxiv_id": ["2205.11487", "2211.01324"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_14917"} +{"question": "Which research papers have utilized constrained or regularized dynamic programming for Offline Reinforcement Learning?", "answer": ["Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning", "Critic Regularized Regression", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Off-Policy Deep Reinforcement Learning without Exploration", "A Policy-Guided Imitation Approach for Offline Reinforcement Learning"], "answer_arxiv_id": ["1910.00177", "2006.15134", "1906.00949", "1812.02900", "2210.08323"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_14918"} +{"question": "Which papers proposed representation learning in block MDPs?", "answer": ["Provably efficient RL with Rich Observations via Latent State Decoding", "Kinematic State Abstraction and Provably Efficient Rich-Observation Reinforcement Learning"], "answer_arxiv_id": ["1901.09018", "1911.05815"], "source_meta": {"published_time": "20220819"}, "qid": "AutoScholarQuery_train_14919"} +{"question": "What studies have been conducted on classical instance segmentation methods?", "answer": ["Learning to Segment Object Candidates", "Mask R-CNN", "Hybrid Task Cascade for Instance Segmentation", "Cascade R-CNN: High Quality Object Detection and Instance Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["1506.06204", "1703.06870", "1901.07518", "1906.09756", "2112.01527"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_14920"} +{"question": "Which studies did the researcher use as a basis for exploring recent advances in diffusion models for images?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Variational Diffusion Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2112.10752", "2107.00630", "2105.05233"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14921"} +{"question": "Are there any works which implement unconditional generation in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2011.13456"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_14922"} +{"question": "What papers are associated with the development of multimodal language models (MLLMs)?", "answer": ["A Survey on Multimodal Large Language Models"], "answer_arxiv_id": ["2306.13549"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_14923"} +{"question": "Which works investigated 3D generative modeling with 3D voxel grids?", "answer": ["3D Shape Induction from 2D Views of Multiple Objects", "Escaping Plato's Cave: 3D Shape From Adversarial Rendering", "Inverse Graphics GAN: Learning to Generate 3D Shapes from Unstructured\n 2D Data"], "answer_arxiv_id": ["1612.05872", "1811.11606", "2002.12674"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_14924"} +{"question": "Which study used the exact form of likelihood to construct confidence sets for parameters of exponential families?", "answer": ["Bregman Deviations of Generic Exponential Families"], "answer_arxiv_id": ["2201.07306"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_14925"} +{"question": "Did any research revisit the capabilities of the SVAE similar to the work presented by the authors?", "answer": ["Revisiting Structured Variational Autoencoders"], "answer_arxiv_id": ["2305.16543"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_14926"} +{"question": "Can you give me examples of research that aims to improve the efficiency of adversarial contrastive learning (ACL)?", "answer": ["Efficient Adversarial Contrastive Learning via Robustness-Aware Coreset Selection"], "answer_arxiv_id": ["2302.03857"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_14927"} +{"question": "What studies have explored the correlation and impact of shortcuts in neural networks trained with SGD?", "answer": ["Understanding the failure modes of out-of-distribution generalization", "What shapes feature representations? Exploring datasets, architectures, and training"], "answer_arxiv_id": ["2010.15775", "2006.12433"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14928"} +{"question": "Could you list the works that evaluate different aspects of performance of generative models separately?", "answer": ["Assessing Generative Models via Precision and Recall", "Improved Precision and Recall Metric for Assessing Generative Models", "Reliable Fidelity and Diversity Metrics for Generative Models"], "answer_arxiv_id": ["1806.00035", "1904.06991", "2002.09797"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_14929"} +{"question": "What research works have delved into VT2V for query-guided summarization?", "answer": ["CLIP-It! Language-Guided Video Summarization", "Query-controllable Video Summarization"], "answer_arxiv_id": ["2107.00650", "2004.03661"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_14930"} +{"question": "Who studied the model’s activations for understanding its prediction?", "answer": ["Network Dissection: Quantifying Interpretability of Deep Visual Representations", "Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)", "On Completeness-aware Concept-Based Explanations in Deep Neural Networks", "Leveraging Sparse Linear Layers for Debuggable Deep Networks"], "answer_arxiv_id": ["1704.05796", "1711.11279", "1910.07969", "2105.04857"], "source_meta": {"published_time": "20220629"}, "qid": "AutoScholarQuery_train_14931"} +{"question": "Which studies explore the training of NeRF with a single image for view synthesis from sparse inputs?", "answer": ["Novel View Synthesis with Diffusion Models", "SparseFusion: Distilling View-conditioned Diffusion for 3D\n Reconstruction", "NerfDiff: Single-image View Synthesis with NeRF-guided Distillation from\n 3D-aware Diffusion"], "answer_arxiv_id": ["2210.04628", "2212.00792", "2302.10109"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_14932"} +{"question": "What studies have been conducted on extending free-viewpoint video in sequences of textured meshes?", "answer": ["Learning Locally Editable Virtual Humans", "HDHumans: A Hybrid Approach for High-fidelity Digital Humans"], "answer_arxiv_id": ["2305.00121", "2210.12003"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_14933"} +{"question": "Are there any works that provide a taxonomy of LLM pathologies?", "answer": ["Language Models Are Greedy Reasoners: A Systematic Formal Analysis of Chain-of-Thought", "Survey of Hallucination in Natural Language Generation"], "answer_arxiv_id": ["2210.01240", "2202.03629"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_14934"} +{"question": "Which paper introduced the continuized framework that allows obtaining simpler proofs and flexibility of asynchronous algorithms?", "answer": ["A Continuized View on Nesterov Acceleration for Stochastic Gradient Descent and Randomized Gossip"], "answer_arxiv_id": ["2106.07644"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_14935"} +{"question": "Which publications adopted the classic approach to 6D object pose estimation problem by establishing 3D-to-2D correspondences and computing the pose with a Pn𝑛nP algorithm?", "answer": ["BB8: A Scalable, Accurate, Robust to Partial Occlusion Method for\n Predicting the 3D Poses of Challenging Objects without Using Depth", "Real-Time Seamless Single Shot 6D Object Pose Prediction", "Segmentation-driven 6D Object Pose Estimation", "Single-Stage 6D Object Pose Estimation", "DPOD: 6D Pose Object Detector and Refiner", "PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation"], "answer_arxiv_id": ["1703.10896", "1711.08848", "1812.02541", "1911.08324", "1902.11020", "1812.11788"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_14936"} +{"question": "Which works extended pruning methods to LLMs?", "answer": ["A Simple and Effective Pruning Approach for Large Language Models", "SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot"], "answer_arxiv_id": ["2306.11695", "2301.00774"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_14937"} +{"question": "Any works about adding an ℓ1-regularization term to produce sparser adversarial examples?", "answer": ["EAD: Elastic-Net Attacks to Deep Neural Networks via Adversarial Examples"], "answer_arxiv_id": ["1709.04114"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_14938"} +{"question": "What studies discuss vision-language models (VLMs) used for image captioning?", "answer": ["Show and Tell: A Neural Image Caption Generator"], "answer_arxiv_id": ["1411.4555"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_14939"} +{"question": "What are the studies about modeling potential interactions using an ego-centric representation?", "answer": ["Populating 3D Scenes by Learning Human-Scene Interaction"], "answer_arxiv_id": ["2012.11581"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14940"} +{"question": "In what paper has been observed that LLM-based evaluators prefer longer responses?", "answer": ["Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena"], "answer_arxiv_id": ["2306.05685"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_14941"} +{"question": "What work is referenced as capitalizing on contrastive learning and vast data quantities to align images and texts in a semantic space?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_14942"} +{"question": "What papers provide defenses against backdoor attacks in federated learning?", "answer": ["Can You Really Backdoor Federated Learning?", "BaFFLe: Backdoor Detection via Feedback-based Federated Learning", "Local and Central Differential Privacy for Robustness and Privacy in Federated Learning", "Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates", "DeepSight: Mitigating Backdoor Attacks in Federated Learning Through Deep Model Inspection", "Shielding Collaborative Learning: Mitigating Poisoning Attacks through Client-Side Detection"], "answer_arxiv_id": ["1911.07963", "2011.02167", "2009.03561v5", "1803.01498", "2201.00763", "1910.13111"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_14943"} +{"question": "In what studies do the researchers incorporate transformer networks into the MIL paradigm?", "answer": ["TransMIL: Transformer based Correlated Multiple Instance Learning for\n Whole Slide Image Classification"], "answer_arxiv_id": ["2106.00908"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14944"} +{"question": "What works have shown success with attention-based aggregation in point-based networks?", "answer": ["Point2Sequence: Learning the Shape Representation of 3D Point Clouds\n with an Attention-based Sequence to Sequence Network", "Modeling Point Clouds with Self-Attention and Gumbel Subset Sampling", "PCAN: 3D Attention Map Learning Using Contextual Information for Point\n Cloud Based Retrieval", "Point Transformer", "An End-to-End Transformer Model for 3D Object Detection", "Point Transformer V2: Grouped Vector Attention and Partition-based\n Pooling"], "answer_arxiv_id": ["1811.02565", "1904.03375", "1904.09793", "2012.09164", "2109.08141", "2210.05666"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_14945"} +{"question": "Which works used the information bottleneck to promote ensemble diversity?", "answer": ["DICE: Diversity in Deep Ensembles via Conditional Redundancy Adversarial Estimation"], "answer_arxiv_id": ["2101.05544"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_14946"} +{"question": "Which works address the issue of Open Set Domain Generalization?", "answer": ["Open Domain Generalization with Domain-Augmented Meta-Learning", "Simple Domain Generalization Methods are Strong Baselines for Open\n Domain Generalization", "Generalizable Decision Boundaries: Dualistic Meta-Learning for Open Set\n Domain Generalization", "Activate and Reject: Towards Safe Domain Generalization under Category\n Shift"], "answer_arxiv_id": ["2104.03620", "2303.18031", "2308.09391", "2310.04724"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_14947"} +{"question": "Which studies highlight the use of averaging for invariant function estimation?", "answer": ["Learning Invariances using the Marginal Likelihood", "Stochastic Latent Residual Video Prediction"], "answer_arxiv_id": ["1808.05563", "2002.09219"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_14948"} +{"question": "What are some studies on Post Training Quantization (PTQ) techniques targeted specifically for language models?", "answer": ["Understanding and Overcoming the Challenges of Efficient Transformer Quantization", "Towards Efficient Post-training Quantization of Pre-trained Language Models", "ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers", "LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models"], "answer_arxiv_id": ["2109.12948", "2109.15082", "2206.01861", "2208.07339", "2211.10438"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_14949"} +{"question": "What works have attempted to use question-answering for the understanding of stories in visual media?", "answer": ["MovieQA: Understanding Stories in Movies through Question-Answering", "TVQA: Localized, Compositional Video Question Answering", "TVQA+: Spatio-Temporal Grounding for Video Question Answering"], "answer_arxiv_id": ["1512.02902", "1809.01696", "1904.11574"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_14950"} +{"question": "What papers studied large-kernel settings in the 2010s?", "answer": ["Local Relation Networks for Image Recognition", "Rethinking the Inception Architecture for Computer Vision", "Inception-v4, Inception-ResNet and the Impact of Residual Connections on\n Learning", "Large Kernel Matters -- Improve Semantic Segmentation by Global\n Convolutional Network"], "answer_arxiv_id": ["1904.11491", "1512.00567", "1602.07261", "1703.02719"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_14951"} +{"question": "Which works focus on the application of a noise-contrastive objective in self-supervised learning?", "answer": ["Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1807.03748"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_14952"} +{"question": "What papers describe the two-step process of language model applications?", "answer": ["Code Llama: Open Foundation Models for Code", "Llemma: An Open Language Model For Mathematics"], "answer_arxiv_id": ["2308.12950", "2310.10631"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_14953"} +{"question": "Could you provide studies on how Equal Error Rate Minimization (ERM) amplifies the unfairness of a classifier in the long run?", "answer": ["Fairness Without Demographics in Repeated Loss Minimization"], "answer_arxiv_id": ["1806.08010"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_14954"} +{"question": "Could you provide me studies about prompting in text attacks?", "answer": ["Large Language Models can be Guided to Evade AI-Generated Text Detection"], "answer_arxiv_id": ["2305.10847"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_14955"} +{"question": "Which papers propose methods for Class Incremental Learning (CIL) in continual learning context?", "answer": ["On the Effectiveness of LayerNorm Tuning for Continual Learning in\n Vision Transformers", "Class-Incremental Learning using Diffusion Model for Distillation and\n Replay", "DER: Dynamically Expandable Representation for Class Incremental\n Learning", "Class Incremental Learning with Pre-trained Vision-Language Models"], "answer_arxiv_id": ["2308.09610", "2306.17560", "2103.16788", "2310.20348"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_14956"} +{"question": "Could you provide me the papers that introduce a fairness notion stating that points should have centers within a distance R if there are n/k points around it within R?", "answer": ["A Center in Your Neighborhood: Fairness in Facility Location"], "answer_arxiv_id": ["1908.09041"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_14957"} +{"question": "Can you name any research suggesting the use of a Udr model-selection metric?", "answer": ["Unsupervised Model Selection for Variational Disentangled Representation Learning"], "answer_arxiv_id": ["1905.12614"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_14958"} +{"question": "What is the paper that uses an over-sampling method and translates head-class samples to replace the duplicated tail-class samples?", "answer": ["M2m: Imbalanced Classification via Major-to-minor Translation"], "answer_arxiv_id": ["2004.00431"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_14959"} +{"question": "What are the studies that incorporate intrinsic motivation into reinforcement learning?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction", "Exploration by Random Network Distillation"], "answer_arxiv_id": ["1705.05363", "1810.12894"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_14960"} +{"question": "What work introduced a pure vision model pre-trained to fill missing patches in images?", "answer": ["Visual Prompting via Image Inpainting"], "answer_arxiv_id": ["2209.00647"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_14961"} +{"question": "Could you cite some works on input-driven MDPs, where additional assumptions were made to factorize rewards or transitions to filter out the exogenous process?", "answer": ["Discovering and Removing Exogenous State Variables and Rewards for Reinforcement Learning", "Sample-Efficient Reinforcement Learning in the Presence of Exogenous Information"], "answer_arxiv_id": ["1806.01584", "2206.04282v1"], "source_meta": {"published_time": "20220713"}, "qid": "AutoScholarQuery_train_14962"} +{"question": "What work combines graph neural networks with a pointer network to handle user-defined type prediction?", "answer": ["LambdaNet: Probabilistic Type Inference using Graph Neural Networks"], "answer_arxiv_id": ["2005.02161"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_14963"} +{"question": "What is the Chinese benchmark that designs adversarial questions spanning multiple domains?", "answer": ["Evaluating Hallucinations in Chinese Large Language Models"], "answer_arxiv_id": ["2310.03368"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_14964"} +{"question": "Are there any studies that have turned their focus on the development of unified multimodal language models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models", "Language Is Not All You Need: Aligning Perception with Language Models"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2302.14045"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_14965"} +{"question": "Which studies achieve instance-dependent regret in bandits and RL?", "answer": ["Information Directed Sampling and Bandits with Heteroscedastic Noise", "Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes", "Improved Variance-Aware Confidence Sets for Linear Bandits and Linear Mixture MDP", "Improved Regret Analysis for Variance-Adaptive Linear Bandits and Horizon-Free Linear Mixture MDPs", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs", "Variance-Dependent Regret Bounds for Linear Bandits and Reinforcement Learning: Adaptivity and Computational Efficiency", "Tighter Problem-Dependent Regret Bounds in Reinforcement Learning without Domain Knowledge using Value Function Bounds", "Sharp Variance-Dependent Bounds in Reinforcement Learning: Best of Both Worlds in Stochastic and Deterministic Environments", "First-Order Regret in Reinforcement Learning with Linear Function Approximation: A Robust Estimation Approach"], "answer_arxiv_id": ["1801.09667", "2012.08507", "2101.12745", "2111.03289", "2205.11507", "2302.10371", "1901.00210", "2301.13446", "2112.03432"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_14966"} +{"question": "Which research papers are about the generation of animatable 3D assets for heads and bodies?", "answer": ["Articulated 3D Head Avatar Generation using Text-to-Image Diffusion\n Models", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "Dancing Avatar: Pose and Text-Guided Human Motion Videos Synthesis with\n Image Diffusion Model", "AvatarCraft: Transforming Text into Neural Human Avatars with\n Parameterized Shape and Pose Control"], "answer_arxiv_id": ["2307.04859", "2304.00916", "2308.07749", "2303.17606"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_14967"} +{"question": "Can you indicate the papers where Stable Diffusion was used for image editing?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_14968"} +{"question": "Can you name some works that solved bit-width allocation problem using differentiable variables, thus employing optimization-based methods in mixed-precision quantization?", "answer": ["Estimating or Propagating Gradients Through Stochastic Neurons for\n Conditional Computation", "Mixed Precision DNNs: All you need is a good parametrization", "FracBits: Mixed Precision Quantization via Fractional Bit-Widths", "Differentiable Dynamic Quantization with Mixed Precision and Adaptive\n Resolution", "Categorical Reparameterization with Gumbel-Softmax", "Rethinking Differentiable Search for Mixed-Precision Neural Networks", "SDQ: Stochastic Differentiable Quantization with Mixed Precision", "HMQ: Hardware Friendly Mixed Precision Quantization Block for CNNs"], "answer_arxiv_id": ["1308.3432", "1905.11452v3", "2007.02017", "2106.02295", "1611.01144", "2004.05795", "2206.04459", "2007.09952"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_14969"} +{"question": "Which papers proposed various training schemes to overcome drawbacks of Class Activation Map (CAM) in weakly supervised semantic segmentation?", "answer": ["Revisiting Dilated Convolution: A Simple Approach for Weakly- and Semi-\n Supervised Semantic Segmentation", "Railroad is not a Train: Saliency as Pseudo-pixel Supervision for Weakly\n Supervised Semantic Segmentation"], "answer_arxiv_id": ["1805.04574", "2105.08965"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_14970"} +{"question": "Which papers provide theoretical analyses of learning dynamics in transformers?", "answer": ["Vision Transformers provably learn spatial structure", "How Do Transformers Learn Topic Structure: Towards a Mechanistic Understanding", "Approximating How Single Head Attention Learns", "Scan and Snap: Understanding Training Dynamics and Token Composition in 1-layer Transformer"], "answer_arxiv_id": ["2210.09221", "2303.04245", "2103.07601", "2305.16380"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_14971"} +{"question": "Could you provide some works that theoretically and empirically investigated the relationship between generalization ability and the width of minima?", "answer": ["Exploring Generalization in Deep Learning", "Sharp Minima Can Generalize For Deep Nets", "Emergent properties of the local geometry of neural loss landscapes"], "answer_arxiv_id": ["1706.08947", "1703.04933", "1910.05929v1"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_14972"} +{"question": "Which works have dealt with video summarization?", "answer": ["Video Summarization with Long Short-term Memory", "Video Summarization Using Deep Neural Networks: A Survey"], "answer_arxiv_id": ["1605.08110", "2101.06072"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_14973"} +{"question": "Can you provide some works that use adversarial training to improve performance in out-of-distribution scenarios?", "answer": ["Robust Reinforcement Learning using Adversarial Populations", "A Regret Minimization Approach to Iterative Learning Control"], "answer_arxiv_id": ["2008.01825", "2102.13478"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_14974"} +{"question": "Can you name the papers that studied score-based methods such as Greedy Equivalent Search (GES)?", "answer": ["Learning Gaussian Networks"], "answer_arxiv_id": ["1302.6808"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_14975"} +{"question": "Could you name the works that propose hybrid approaches involving pretrained autoregressive models?", "answer": ["Latent Diffusion for Language Generation"], "answer_arxiv_id": ["2212.09462"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_14976"} +{"question": "Which studies employed batch normalization statistics to update the model for test-time adaptation?", "answer": ["Revisiting Batch Normalization For Practical Domain Adaptation"], "answer_arxiv_id": ["1603.04779"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_14977"} +{"question": "What is an example of a method that enhances the efficiency of parameter-efficient training techniques further?", "answer": ["QLoRA: Efficient Finetuning of Quantized LLMs"], "answer_arxiv_id": ["2305.14314"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_14978"} +{"question": "What is the pioneering research that introduced whole-body parametric models?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image"], "answer_arxiv_id": ["1904.05866"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_14979"} +{"question": "What works utilize prototype-based methods in federated learning?", "answer": ["Prototype Guided Federated Learning of Visual Feature Representations", "FedProc: Prototypical Contrastive Federated Learning on Non-IID data", "FedFM: Anchor-based Feature Matching for Data Heterogeneity in Federated Learning"], "answer_arxiv_id": ["2105.08982", "2109.12273", "2210.07615"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_14980"} +{"question": "Which works reflect initial efforts of large multimodal models involving the combination of computer vision with large language models?", "answer": ["Multimodal Few-Shot Learning with Frozen Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "Language Is Not All You Need: Aligning Perception with Language Models", "PaLM-E: An Embodied Multimodal Language Model", "OpenFlamingo: An Open-Source Framework for Training Large Autoregressive\n Vision-Language Models"], "answer_arxiv_id": ["2106.13884", "2204.14198", "2301.12597", "2201.12086", "2302.14045", "2303.03378", "2308.01390"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_14981"} +{"question": "Can you name the studies that have applied low-rank approximation to each feature map of Convolutional Neural Networks?", "answer": ["Network Decoupling: From Regular to Depthwise Separable Convolutions", "Convolutional neural networks with low-rank regularization", "Accelerating Very Deep Convolutional Networks for Classification and\n Detection"], "answer_arxiv_id": ["1808.05517", "1511.06067", "1505.06798"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_14982"} +{"question": "Which works have used Clifford algebras to formulate quantum neural networks?", "answer": ["Clifford Algebras, Quantum Neural Networks and Generalized Quantum Fourier Transform"], "answer_arxiv_id": ["2206.01808v1"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_14983"} +{"question": "Which papers showed that CNTKs can outperform standard CNNs?", "answer": ["Harnessing the Power of Infinitely Wide Deep Nets on Small-data Tasks"], "answer_arxiv_id": ["1910.01663"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_14984"} +{"question": "Which works propose different architecture modifications to address task interference?", "answer": ["End-to-End Multi-Task Learning with Attention", "Reparameterizing Convolutions for Incremental Multi-Task Learning without Task Interference", "Cross-stitch Networks for Multi-task Learning", "UniNet: A Unified Scene Understanding Network and Exploring Multi-Task Relationships through the Lens of Adversarial Attacks"], "answer_arxiv_id": ["1803.10704", "2007.12540", "1604.03539", "2108.04584"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_14985"} +{"question": "What are the papers about ASR realizing multilingual goals through pretraining on large multilingual corpora?", "answer": ["Unsupervised Cross-lingual Representation Learning for Speech\n Recognition"], "answer_arxiv_id": ["2006.13979"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_14986"} +{"question": "Which study presents a new architecture combining CNNs and ViT for mobile semantic segmentation?", "answer": ["TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation"], "answer_arxiv_id": ["2204.05525"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_14987"} +{"question": "What research discussed how adversarial examples printed on papers were effective in fooling DNN classifiers?", "answer": ["Adversarial examples in the physical world"], "answer_arxiv_id": ["1607.02533"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_14988"} +{"question": "Which works explored functionality enhancement of Generative Adversarial Networks (GAN)?", "answer": ["InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets", "Conditional Generative Adversarial Nets", "Conditional Image Synthesis with Auxiliary Classifier GANs", "Adversarial Feature Learning"], "answer_arxiv_id": ["1606.03657", "1411.1784", "1610.09585", "1605.09782"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_14989"} +{"question": "Are there any works focusing on fine-grained reward modeling using secondary neural models?", "answer": ["Fine-Grained Human Feedback Gives Better Rewards for Language Model\n Training"], "answer_arxiv_id": ["2306.01693"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_14990"} +{"question": "What papers discussed self-distillation, a form of knowledge distillation within one model?", "answer": ["Be Your Own Teacher: Improve the Performance of Convolutional Neural\n Networks via Self Distillation"], "answer_arxiv_id": ["1905.08094"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_14991"} +{"question": "Could you provide me some works about global-matching methods for TIReID?", "answer": ["See Finer, See More: Implicit Modality Alignment for Text-based Person\n Retrieval", "Dual-Path Convolutional Image-Text Embeddings with Instance Loss", "DSSL: Deep Surroundings-person Separation Learning for Text-based Person\n Retrieval"], "answer_arxiv_id": ["2208.08608", "1711.05535", "2109.05534"], "source_meta": {"published_time": "20230819"}, "qid": "AutoScholarQuery_train_14992"} +{"question": "In which paper was the Lookahead algorithm first introduced for minimization?", "answer": ["Lookahead Optimizer: k steps forward, 1 step back"], "answer_arxiv_id": ["1907.08610"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_14993"} +{"question": "Could you provide me some works that introduced one-shot capabilities in person-agnostic methods for talking-head synthesis?", "answer": ["One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing", "HeadGAN: One-shot Neural Head Synthesis and Editing", "First Order Motion Model for Image Animation", "Fast Bi-layer Neural Synthesis of One-Shot Realistic Head Avatars", "MegaPortraits: One-shot Megapixel Neural Head Avatars", "StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via\n Pre-trained StyleGAN", "NOFA: NeRF-based One-shot Facial Avatar Reconstruction", "HyperReenact: One-Shot Reenactment via Jointly Learning to Refine and\n Retarget Faces"], "answer_arxiv_id": ["2011.15126", "2012.08261", "2003.00196", "2008.10174", "2207.07621", "2203.04036", "2307.03441", "2307.10797"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_14994"} +{"question": "Can you mention some works that investigate the improvement of privacy/utility/computation trade-offs for DP-SGD?", "answer": ["Unlocking High-Accuracy Differentially Private Image Classification through Scale", "Public Data-Assisted Mirror Descent for Private Model Training", "Bypassing the Ambient Dimension: Private SGD with Gradient Subspace Identification", "Practical and Private (Deep) Learning Without Sampling or Shuffling", "Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams"], "answer_arxiv_id": ["2204.13650", "2112.00193", "2007.03813", "2103.00039", "2202.08312"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_14995"} +{"question": "What research focuses on object grasping and language-guided robotic manipulation, which are crucial skills for robotic manipulation?", "answer": ["Deep Learning Approaches to Grasp Synthesis: A Review", "Language-Conditioned Imitation Learning for Robot Manipulation Tasks", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Interactive Language: Talking to Robots in Real Time"], "answer_arxiv_id": ["2207.02556", "2010.12083", "2204.01691", "2210.06407"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_14996"} +{"question": "Which papers presented a roto-scale-translation equivariant CNN?", "answer": ["Deformation Robust Roto-Scale-Translation Equivariant CNNs"], "answer_arxiv_id": ["2111.10978"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_14997"} +{"question": "Could you provide me some studies that focus on the challenge of compounding error in model-based RL?", "answer": ["Lipschitz Continuity in Model-based Reinforcement Learning", "Investigating Compounding Prediction Errors in Learned Dynamics Models"], "answer_arxiv_id": ["1804.07193", "2203.09637v1"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_14998"} +{"question": "Could you provide me some works that discuss the extension of the expressive power of GNNs?", "answer": ["Relational Pooling for Graph Representations", "Building powerful and equivariant graph neural networks with structural message-passing", "The Surprising Power of Graph Neural Networks with Random Node Initialization", "Coloring graph neural networks for node disambiguation", "Random Features Strengthen Graph Neural Networks", "Equivariant Polynomials for Graph Neural Networks", "Graph Neural Networks with Local Graph Parameters", "Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting", "Graph Homomorphism Convolution", "Breaking the Limits of Message Passing Graph Neural Networks", "Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks", "Weisfeiler and Lehman Go Cellular: CW Networks", "Topological Graph Neural Networks", "Agent-based Graph Neural Networks", "Neural Trees for Learning on Graphs", "Weisfeiler and Leman Go Relational", "Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning", "Directional Graph Networks", "Equivariant Subgraph Aggregation Networks", "Reconstruction for Powerful Graph Representations", "How Powerful are K-hop Message Passing Graph Neural Networks", "Understanding and Extending Subgraph GNNs by Rethinking Their Symmetries", "Weisfeiler and Leman go Machine Learning: The Story so far", "DropGNN: Random Dropouts Increase the Expressiveness of Graph Neural Networks", "A Theoretical Comparison of Graph Neural Network Extensions", "Ordered Subgraph Aggregation Networks", "Autobahn: Automorphism-based Graph Neural Nets", "Identity-aware Graph Neural Networks", "Nested Graph Neural Networks", "From Stars to Subgraphs: Uplifting Any GNN with Local Structure Awareness", "A Complete Expressiveness Hierarchy for Subgraph GNNs via Subgraph Weisfeiler-Lehman Tests", "Rethinking the Expressive Power of GNNs via Graph Biconnectivity"], "answer_arxiv_id": ["1903.02541", "2006.15107", "2010.01179v2", "1912.06058", "2002.03155", "2302.11556", "2106.06707", "2006.09252", "2005.01214", "2106.04319", "2103.03212", "2106.12575", "2102.07835v4", "2206.11010", "2105.07264", "2211.17113", "2009.00142", "2010.02863", "2110.02910", "2110.00577", "2205.13328", "2206.11140", "2112.09992", "2111.06283", "2201.12884v1", "2206.11168", "2103.01710", "2101.10320", "2110.13197", "2110.03753", "2302.07090", "2301.09505"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_14999"} +{"question": "What works reveal Nash equilibria's computational intractability in two-player general-sum games?", "answer": ["Settling the Complexity of Computing Two-Player Nash Equilibria"], "answer_arxiv_id": ["0704.1678"], "source_meta": {"published_time": "20220803"}, "qid": "AutoScholarQuery_train_15000"} +{"question": "Which works have explored the self-supervised learning approach in computational pathology?", "answer": ["Multi-modal Masked Autoencoders Learn Compositional Histopathological Representations", "Scaling Vision Transformers to Gigapixel Images via Hierarchical Self-Supervised Learning"], "answer_arxiv_id": ["2209.01534", "2206.02647"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_15001"} +{"question": "Which paper introduced the PSD models?", "answer": ["Non-parametric Models for Non-negative Functions"], "answer_arxiv_id": ["2007.03926"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_15002"} +{"question": "Which studies developed and used the BiSECT method and corpus for the Split and Rephrase task?", "answer": ["BiSECT: Learning to Split and Rephrase Sentences with Bitexts"], "answer_arxiv_id": ["2109.05006"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_15003"} +{"question": "Can you mention the studies in trajectory prediction that focused on the environmental shifts between training and testing?", "answer": ["Adaptive Trajectory Prediction via Transferable GNN", "Expanding the Deployment Envelope of Behavior Prediction via Adaptive\n Meta-Learning", "Vehicle trajectory prediction works, but not everywhere"], "answer_arxiv_id": ["2203.05046", "2209.11820", "2112.03909"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_15004"} +{"question": "What papers present knowledge transfer methodologies for reinforcement learning?", "answer": ["Actor-Mimic Deep Multitask and Transfer Reinforcement Learning", "Policy Improvement via Imitation of Multiple Oracles", "MULTIPOLAR: Multi-Source Policy Aggregation for Transfer Reinforcement Learning between Diverse Environmental Dynamics", "Transfer in Deep Reinforcement Learning Using Successor Features and Generalised Policy Improvement", "Selective Experience Replay for Lifelong Learning", "Generalized Hindsight for Reinforcement Learning", "Conservative Data Sharing for Multi-Task Offline Reinforcement Learning", "REPAINT: Knowledge Transfer in Deep Reinforcement Learning"], "answer_arxiv_id": ["1511.06342", "2007.00795", "1909.13111", "1901.10964", "1802.10269", "2002.11708", "2109.08128", "2011.11827"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_15005"} +{"question": "Which works demonstrate the application of pre-trained vision models for motor control?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "Pre-Trained Image Encoder for Generalizable Visual Reinforcement Learning", "Masked Visual Pre-training for Motor Control", "The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "R3M: A Universal Visual Representation for Robot Manipulation", "VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training", "On Pre-Training for Visuo-Motor Control: Revisiting a Learning-from-Scratch Baseline"], "answer_arxiv_id": ["2107.03380", "2212.08860", "2203.06173", "2203.03580", "2203.12601", "2210.00030", "2212.05749"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_15006"} +{"question": "Could you provide me some works about proposal-based techniques for temporal video grounding?", "answer": ["MAN: Moment Alignment Network for Natural Language Moment Retrieval via\n Iterative Graph Adjustment", "Semantic Conditioned Dynamic Modulation for Temporal Sentence Grounding\n in Videos"], "answer_arxiv_id": ["1812.00087", "1910.14303"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15007"} +{"question": "Are there any works that 'learn to segment instances into more fine-grained masks' through parts segmentation?", "answer": ["Part-aware Panoptic Segmentation", "Multi-task Fusion for Efficient Panoptic-Part Segmentation"], "answer_arxiv_id": ["2106.06351", "2212.07671"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_15008"} +{"question": "Could you provide some references about few-shot domain adaptation?", "answer": ["Few-Shot Adversarial Domain Adaptation", "Multi-source Few-shot Domain Adaptation"], "answer_arxiv_id": ["1711.02536", "2109.12391"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_15009"} +{"question": "Any works about modeling hard constraints such as lexical constraints by a differentiable n-gram matching function in the field of score-based sampling methods?", "answer": ["Don’t Take It Literally: An Edit-Invariant Sequence Loss for Text Generation"], "answer_arxiv_id": ["2106.15078"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_15010"} +{"question": "What papers have been released about perturbation-based interpretation techniques?", "answer": ["Real Time Image Saliency for Black Box Classifiers", "Interpretable Explanations of Black Boxes by Meaningful Perturbation", "“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "RISE: Randomized Input Sampling for Explanation of Black-box Models", "Visualizing and Understanding Convolutional Networks"], "answer_arxiv_id": ["1705.07857", "1704.03296", "1602.04938", "1806.07421", "1311.2901"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_15011"} +{"question": "Could you provide some works about the practical study of feature interactions in Machine Learning?", "answer": ["Beyond Word Importance: Contextual Decomposition to Extract Interactions from LSTMs", "Feature Interaction Interpretability: A Case for Explaining Ad-Recommendation Systems via Neural Interaction Detection", "Detecting Statistical Interactions From Neural Network Weights", "Learning Global Pairwise Interactions with Bayesian Neural Networks", "Hierarchical interpretations for neural network predictions", "Explaining Explanations: Axiomatic Feature Interactions for Deep Networks"], "answer_arxiv_id": ["1801.05453", "2006.10966", "1705.04977", "1901.08361", "1806.05337", "2002.04138"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_15012"} +{"question": "What prior studies adopted cascaded approaches in DMs?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Cascaded Diffusion Models for High Fidelity Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2102.09672", "2106.15282", "2205.11487"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_15013"} +{"question": "What papers discuss the use of an iterative feedback collection approach for enhanced reasoning tasks with large language models?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2305.10601", "2210.03629"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_15014"} +{"question": "What research developed high-quality reconstruction from multi-view images using point cloud-based representations?", "answer": ["ADOP: Approximate Differentiable One-Pixel Point Rendering", "3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2110.06635", "2308.04079"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_15015"} +{"question": "What studies are about the advancements in sparse MoE?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer", "GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding", "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity", "Unified Scaling Laws for Routed Language Models", "Hash Layers For Large Sparse Models"], "answer_arxiv_id": ["1701.06538", "2006.16668", "2101.03961", "2202.01169", "2106.04426"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_15016"} +{"question": "What works have demonstrated success in modeling text-to-image generation with diffusion models?", "answer": ["KNN-Diffusion: Image Generation via Large-Scale Retrieval", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2204.02849", "2204.06125", "2205.11487", "2112.10741"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_15017"} +{"question": "What are some studies of methods for fine-tuning models without group knowledge?", "answer": ["Just Train Twice: Improving Group Robustness without Training Group\n Information", "Learning from Failure: Training Debiased Classifier from Biased\n Classifier"], "answer_arxiv_id": ["2107.09044", "2007.02561"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_15018"} +{"question": "What works have been done on backdoor attacks that assume control over the weights of deployed models?", "answer": ["Subnet Replacement: Deployment-stage backdoor attack against deep neural networks in gray-box setting", "Towards Practical Deployment-Stage Backdoor Attack on Deep Neural Networks"], "answer_arxiv_id": ["2107.07240", "2111.12965"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_15019"} +{"question": "Which studies discuss the use of Test-time data augmentation (TTA) in estimating uncertainty and improving classification accuracy?", "answer": ["Understanding Measures of Uncertainty for Adversarial Example Detection", "Classification Confidence Estimation with Test-Time Data-Augmentation", "Going deeper with convolutions", "Deep Residual Learning for Image Recognition", "Mixup Inference: Better Exploiting Mixup to Defend Adversarial Attacks", "Greedy Policy Search: A Simple Baseline for Learnable Test-Time Augmentation", "Better Aggregation in Test-Time Augmentation", "Learning Loss for Test-Time Augmentation", "Tent: Fully Test-Time Adaptation by Entropy Minimization", "MEMO: Test Time Robustness via Adaptation and Augmentation"], "answer_arxiv_id": ["1803.08533", "2006.16705", "1409.4842", "1512.03385", "1909.11515", "2002.09103", "2011.11156", "2010.11422", "2006.10726", "2110.09506"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_15020"} +{"question": "What research assumes having proxy datasets to train proxy models?", "answer": ["Evaluating Fairness of Machine Learning Models Under Uncertain and Incomplete Information", "Multiaccurate Proxies for Downstream Fairness"], "answer_arxiv_id": ["2102.08410", "2107.04423v2"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_15021"} +{"question": "What works provided the information-theoretical interpretation of beta-VAEs?", "answer": ["Fixing a Broken ELBO"], "answer_arxiv_id": ["1711.00464"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_15022"} +{"question": "Can you provide references where the problem of non-identifiability is approached using additional auxiliary labels?", "answer": ["Variational Autoencoders and Nonlinear ICA: A Unifying Framework", "Weakly Supervised Disentangled Generative Causal Representation Learning"], "answer_arxiv_id": ["1907.04809", "2010.02637"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_15023"} +{"question": "Which studies performed deep learning for global weather forecasting tasks?", "answer": ["Forecasting Global Weather with Graph Neural Networks", "MetNet: A Neural Weather Model for Precipitation Forecasting"], "answer_arxiv_id": ["2202.07575", "2003.12140"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_15024"} +{"question": "What are some examples of person-specific methodologies in the domain of neural avatars?", "answer": ["Pose-Controllable Talking Face Generation by Implicitly Modularized\n Audio-Visual Representation", "AvatarMAV: Fast 3D Head Avatar Reconstruction Using Motion-Aware Neural\n Voxels", "Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar\n Reconstruction", "Neural Head Avatars from Monocular RGB Videos"], "answer_arxiv_id": ["2104.11116", "2211.13206", "2012.03065", "2112.01554"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_15025"} +{"question": "What work used tensor-train decomposition on dense weight matrices for making neural networks more parameter efficient?", "answer": ["Tensorizing Neural Networks"], "answer_arxiv_id": ["1509.06569"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_15026"} +{"question": "Which studies involve pruning the attention matrix as a method to tackle the complexity problem in Transformers?", "answer": ["Big Bird: Transformers for Longer Sequences", "Sparse and Continuous Attention Mechanisms"], "answer_arxiv_id": ["2007.14062", "2006.07214"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15027"} +{"question": "Could you provide me research papers that explain the vulnerability of the loss function to small noises near a sharp minimum?", "answer": ["Visualizing the Loss Landscape of Neural Nets", "On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "Sharp Minima Can Generalize For Deep Nets"], "answer_arxiv_id": ["1712.09913", "1609.04836", "1703.04933"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_15028"} +{"question": "What is the initial work that proposed the concept of Residual Networks (ResNets)?", "answer": ["Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1512.03385"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_15029"} +{"question": "Which work critically examines the limitations of causal tracing in determining the specific layers to be edited in LLMs?", "answer": ["Does Localization Inform Editing? Surprising Differences in\n Causality-Based Localization vs. Knowledge Editing in Language Models"], "answer_arxiv_id": ["2301.04213"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_15030"} +{"question": "Which studies focus on unsupervised representation learning techniques that pursue disentangled and compositional latent representations for visual concepts?", "answer": ["InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets", "Early Visual Concept Learning with Unsupervised Deep Learning", "Learning Disentangled Representations with Semi-Supervised Deep Generative Models", "Learning to Manipulate Individual Objects in an Image"], "answer_arxiv_id": ["1606.03657", "1606.05579", "1706.00400", "2004.05495"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_15031"} +{"question": "What papers are about feature sparsity and neural networks?", "answer": ["LassoNet: A Neural Network with Feature Sparsity", "Group Sparse Regularization for Deep Neural Networks", "Sparse-Input Neural Networks for High-dimensional Nonparametric Regression and Classification"], "answer_arxiv_id": ["1907.12207", "1607.00485", "1711.07592"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_15032"} +{"question": "What research has been done that showed benefits from transfers from spoken language or other sign language data?", "answer": ["SLTUnet: A Simple Unified Model for Sign Language Translation"], "answer_arxiv_id": ["2305.01778"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_15033"} +{"question": "What are the works that augmented high-level actions with location labels to support fine-grained scenario analysis?", "answer": ["ROAD: The ROad event Awareness Dataset for Autonomous Driving"], "answer_arxiv_id": ["2102.11585"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15034"} +{"question": "Are there references discussing compression, specifically compression progress, as a key factor in human curiosity?", "answer": ["Driven by Compression Progress: A Simple Principle Explains Essential Aspects of Subjective Beauty, Novelty, Surprise, Interestingness, Attention, Curiosity, Creativity, Art, Science, Music, Jokes"], "answer_arxiv_id": ["0812.4360"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_15035"} +{"question": "What studies use a contrastive loss to align 2D and 3D representations on large indoor or outdoor point clouds?", "answer": ["Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Data", "Self-Supervised Image-to-Point Distillation via Semantically Tolerant\n Contrastive Loss"], "answer_arxiv_id": ["2203.16258", "2301.05709"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_15036"} +{"question": "Are there any recent studies that applied prompt tuning to CLIP?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2109.01134", "2203.05557"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_15037"} +{"question": "Which works are popular fine-grained image classification datasets?", "answer": ["Fine-grained Visual-textual Representation Learning", "Fine-Grained Visual Classification of Aircraft"], "answer_arxiv_id": ["1709.00340", "1306.5151"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_15038"} +{"question": "Could you provide papers reporting results using 1% of the validation images in academic research on certifying ImageNet with RS?", "answer": ["Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers", "SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness", "Denoised Smoothing: A Provable Defense for Pretrained Classifiers"], "answer_arxiv_id": ["1906.04584", "2111.09277", "2003.01908"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_15039"} +{"question": "Which papers entail extending customization in denoising diffusion probabilistic models?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.12242", "2208.01618"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_15040"} +{"question": "Any works about the activation function SeLU being theoretically motivated?", "answer": ["Self-Normalizing Neural Networks"], "answer_arxiv_id": ["1706.02515"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_15041"} +{"question": "What papers focus on optimizing the backward Markovian process to approximate the non-Markovian forward process?", "answer": ["Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models"], "answer_arxiv_id": ["2201.06503"], "source_meta": {"published_time": "20220429"}, "qid": "AutoScholarQuery_train_15042"} +{"question": "Any works about tackling robustness to model misspecification for neural likelihood?", "answer": ["Misspecification-robust Sequential Neural Likelihood", "Robust Neural Posterior Estimation and Statistical Model Criticism"], "answer_arxiv_id": ["2301.13368", "2210.06564"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15043"} +{"question": "Could you provide me works about improving policy performance through the application of data augmentation techniques to pixel-based inputs?", "answer": ["An Empirical Investigation of Representation Learning for Imitation", "Challenges and Opportunities in Offline Reinforcement Learning from Visual Observations"], "answer_arxiv_id": ["2205.07886", "2206.04779"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_15044"} +{"question": "Which works discuss the use of more expressive approximate posteriors including normalizing flows?", "answer": ["Posterior Refinement Improves Sample Efficiency in Bayesian Neural Networks", "Improving predictions of Bayesian neural nets via local linearization", "Variational Boosting: Iteratively Refining Posterior Approximations", "Mixtures of Laplace Approximations for Improved Post-Hoc Uncertainty in Deep Learning"], "answer_arxiv_id": ["2205.10041", "2008.08400v3", "1611.06585v2", "2111.03577"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_15045"} +{"question": "What studies apply diffusion models to fields like multi-modal generation, language generation, music generation and object detection?", "answer": ["Versatile Diffusion: Text, Images and Variations All in One Diffusion Model", "DreamFusion: Text-to-3D using 2D Diffusion", "Text-To-4D Dynamic Scene Generation", "Diffusion-LM Improves Controllable Text Generation", "NerfDiff: Single-image View Synthesis with NeRF-guided Distillation from\n 3D-aware Diffusion", "NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with\n 360{\\deg} Views", "Symbolic Music Generation with Diffusion Models", "DiffusionDet: Diffusion Model for Object Detection"], "answer_arxiv_id": ["2211.08332v4", "2209.14988", "2301.11280", "2205.14217", "2302.10109", "2211.16431", "2103.16091", "2211.09788"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_15046"} +{"question": "What works utilize prompt and prefix tunings for learning new classes while freezing the pre-trained ViT?", "answer": ["Learning to Prompt for Continual Learning", "DualPrompt: Complementary Prompting for Rehearsal-free Continual\n Learning"], "answer_arxiv_id": ["2112.08654", "2204.04799"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_15047"} +{"question": "Who clarified why deep convolutional networks trained with DFA and FA fail to learn efficiently?", "answer": ["Align, then memorise: the dynamics of learning with feedback alignment"], "answer_arxiv_id": ["2011.12428"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_15048"} +{"question": "Can you suggest some works focused on leveraging the rich embeddings learnt by contrastive models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "Scaling Open-Vocabulary Object Detection", "Image Segmentation Using Text and Image Prompts", "CRIS: CLIP-Driven Referring Image Segmentation", "ZeroCap: Zero-Shot Image-to-Text Generation for Visual-Semantic\n Arithmetic", "Language Models Can See: Plugging Visual Controls in Text Generation", "ClipCap: CLIP Prefix for Image Captioning", "Frozen CLIP Models are Efficient Video Learners", "CLIP4Caption: CLIP for Video Caption", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2103.00020", "2106.11097", "2306.09683", "2112.10003", "2111.15174", "2111.14447", "2205.02655", "2111.09734", "2208.03550", "2110.06615", "2204.06125", "2112.10741"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15049"} +{"question": "What are some studies that aim to provide a better planning strategy by integrating low-level actions and planning trajectory?", "answer": ["Trajectory-guided Control Prediction for End-to-end Autonomous Driving:\n A Simple yet Strong Baseline"], "answer_arxiv_id": ["2206.08129"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_15050"} +{"question": "Which study introduced Graph Convolutional Networks (GCNs), a type of graph neural network that applies convolutional operations on graphs?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks"], "answer_arxiv_id": ["1609.02907"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_15051"} +{"question": "Which works contributed to the advances in diffusion models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2102.09672", "2010.02502", "2006.11239", "2105.05233"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_15052"} +{"question": "Which papers discuss the success of attention mechanisms in machine learning?", "answer": ["End-to-End Object Detection with Transformers", "Language Models are Few-Shot Learners", "Masked Autoencoders Are Scalable Vision Learners", "DMAP: a Distributed Morphological Attention Policy for Learning to\n Locomote with a Changing Body", "Masked Autoencoders As Spatiotemporal Learners"], "answer_arxiv_id": ["2005.12872", "2005.14165", "2111.06377", "2209.14218", "2205.09113"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_15053"} +{"question": "What works represent the paradigm shift from traditional methodologies to more sophisticated models in deep clustering?", "answer": ["Unsupervised Deep Learning by Neighbourhood Discovery", "GATCluster: Self-Supervised Gaussian-Attention Network for Image Clustering", "Multi-Modal Deep Clustering: Unsupervised Partitioning of Images"], "answer_arxiv_id": ["1904.11567", "2002.11863", "1912.02678"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_15054"} +{"question": "What papers study the relationship between subsampling and regularization for ridge ensembles in the overparameterized regime?", "answer": ["The Implicit Regularization of Ordinary Least Squares Ensembles", "Bagging in overparameterized learning: Risk characterization and risk monotonization", "Subsample Ridge Ensembles: Equivalences and Generalized Cross-Validation"], "answer_arxiv_id": ["1910.04743v2", "2210.11445", "2304.13016"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15055"} +{"question": "Which paper used S4 in the components of a sequential VAE?", "answer": ["Deep Latent State Space Models for Time-Series Generation"], "answer_arxiv_id": ["2212.12749v3"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15056"} +{"question": "Which works minimize the quantization error between a latent weight and its binary variant in binary neural networks (BNNs)?", "answer": ["XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks", "XNOR-Net++: Improved binary neural networks"], "answer_arxiv_id": ["1603.05279", "1909.13863"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_15057"} +{"question": "Are there any benchmark papers that emphasize the need for developing and testing single-step models for the multi-step domain?", "answer": ["Mind the Retrosynthesis Gap: Bridging the divide between Single-step and Multi-step Retrosynthesis Prediction"], "answer_arxiv_id": ["2212.11809"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_15058"} +{"question": "What research has been done on evidence that the average node degrees of deep networks are highly correlated with their convergence speeds in deep network as graph model?", "answer": ["How does topology influence gradient propagation and model performance of deep networks with DenseNet-type skip connections?"], "answer_arxiv_id": ["1910.00780"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_15059"} +{"question": "Any works about scene flow that are un- or self-supervised method, benefit from using multiple frames?", "answer": ["Competitive Collaboration: Joint Unsupervised Learning of Depth, Camera Motion, Optical Flow and Motion Segmentation", "Self-Supervised Monocular Scene Flow Estimation", "Self-Supervised Multi-Frame Monocular Scene Flow"], "answer_arxiv_id": ["1805.09806", "2004.04143", "2105.02216"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_15060"} +{"question": "What works could you recommend on model-based hand pose estimation, which utilize statistical models with low-dimensional parameters?", "answer": ["3D Hand Shape and Pose from Images in the Wild", "FrankMocap: Fast Monocular 3D Hand and Body Motion Capture by Regression\n and Integration", "Fast and Robust Hand Tracking Using Detection-Guided Optimization", "End-to-end Hand Mesh Recovery from a Monocular RGB Image", "Monocular Real-time Hand Shape and Motion Capture using Multi-modal Data", "Embodied Hands: Modeling and Capturing Hands and Bodies Together"], "answer_arxiv_id": ["1902.03451", "2008.08324", "1602.04124", "1902.09305", "2003.09572", "2201.02610"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_15061"} +{"question": "Which studies proposed classical methods in safe RL such as CPO, Lyapunov-based approaches, FOCOPS, CUP, and others based on various strategies?", "answer": ["Constrained Policy Optimization", "A Lyapunov-based Approach to Safe Reinforcement Learning", "Projection-Based Constrained Policy Optimization", "First Order Constrained Optimization in Policy Space", "Constrained Variational Policy Optimization for Safe Reinforcement Learning", "Enforcing Hard Constraints with Soft Barriers: Safe Reinforcement Learning in Unknown Stochastic Environments"], "answer_arxiv_id": ["1705.10528", "1805.07708", "2010.03152", "2002.06506", "2201.11927", "2209.15090"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_15062"} +{"question": "Could you provide me some works about kernel fusion in the aim of reducing data movement?", "answer": ["PyTorch: An Imperative Style, High-Performance Deep Learning Library"], "answer_arxiv_id": ["1912.01703"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_15063"} +{"question": "What work proposed the idea of viewing image patches as tokens and processing these tokens by Transformers?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_15064"} +{"question": "What works focus on embedding-based models?", "answer": ["Convolutional 2D Knowledge Graph Embeddings", "End-to-end Structure-Aware Convolutional Networks for Knowledge Base\n Completion"], "answer_arxiv_id": ["1707.01476", "1811.04441"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_15065"} +{"question": "Which paper formulated in-context learning as an instance of implicit Bayesian inference?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_15066"} +{"question": "What works focus on Neural Radiance Fields (NeRFs) and similar volumetric representations in inverse rendering?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Deep Reflectance Volumes: Relightable Reconstructions from Multi-View Photometric Images", "NeRD: Neural Reflectance Decomposition from Image Collections", "Neural-PIL: Neural Pre-Integrated Lighting for Reflectance Decomposition", "SAMURAI: Shape And Material from Unconstrained Real-world Arbitrary Image collections", "NeRFactor: Neural Factorization of Shape and Reflectance Under an Unknown Illumination", "NeRF for Outdoor Scene Relighting", "Learning Object-Centric Neural Scattering Functions for Free-Viewpoint Relighting and Scene Composition"], "answer_arxiv_id": ["2003.08934", "2007.09892", "2012.03918", "2110.14373", "2205.15768v1", "2106.01970", "2112.05140", "2303.06138"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_15067"} +{"question": "What papers have leveraged estimation of dynamics in offline RL methods?", "answer": ["MOReL: Model-Based Offline Reinforcement Learning", "MOPO: Model-based Offline Policy Optimization", "Model-Based Offline Planning", "RAMBO-RL: Robust Adversarial Model-Based Offline Reinforcement Learning", "Model-Based Offline Planning with Trajectory Pruning"], "answer_arxiv_id": ["2005.05951", "2005.13239", "2008.05556", "2204.12581", "2105.07351"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_15068"} +{"question": "Which studies demonstrated that denoising diffusion can better capture the score than simple denoising?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_15069"} +{"question": "Which works use 3D CNN for object/human-centric motion as video representations?", "answer": ["Towards Long-Form Video Understanding", "Generating Descriptions with Grounded and Co-Referenced People"], "answer_arxiv_id": ["2106.11310", "1704.01518"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_15070"} +{"question": "Which studies addressed the use of differentiable renderer for vector graphics?", "answer": ["ClipGen: A Deep Generative Model for Clipart Vectorization and Synthesis", "Rethinking Style Transfer: From Pixels to Parameterized Brushstrokes", "Modern Evolution Strategies for Creativity: Fitting Concrete Images and Abstract Concepts", "Im2Vec: Synthesizing Vector Graphics without Vector Supervision", "Towards Layer-wise Image Vectorization"], "answer_arxiv_id": ["2106.04912", "2103.17185", "2109.08857", "2102.02798", "2206.04655"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_15071"} +{"question": "Which works discuss the energy-based and other mathematical methods for Pixel-Wise Out-of-Distribution (OoD) detection?", "answer": ["A Simple Unified Framework for Detecting Out-of-Distribution Samples and\n Adversarial Attacks", "Generalized Out-of-Distribution Detection: A Survey", "Enhancing The Reliability of Out-of-distribution Image Detection in\n Neural Networks", "Energy-based Out-of-distribution Detection"], "answer_arxiv_id": ["1807.03888", "2110.11334", "1706.02690", "2010.03759"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_15072"} +{"question": "Which studies utilize 3D voxel grids in 3D generative methods?", "answer": ["3D Shape Induction from 2D Views of Multiple Objects", "Escaping Plato's Cave: 3D Shape From Adversarial Rendering", "Inverse Graphics GAN: Learning to Generate 3D Shapes from Unstructured\n 2D Data", "Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling"], "answer_arxiv_id": ["1612.05872", "1811.11606", "2002.12674", "1610.07584"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_15073"} +{"question": "Which works introduced diffusion models as critical in text-to-image generation?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_15074"} +{"question": "Which studies cover the SimFS method, a special lossless compression method that compresses the simulation data by storing simulation checkpoints?", "answer": ["SimFS: A Simulation Data Virtualizing File System Interface"], "answer_arxiv_id": ["1902.03154"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_15075"} +{"question": "Which works explored learning a latent variable representation of offline data for assisted planning?", "answer": ["Learning Latent Plans from Play"], "answer_arxiv_id": ["1903.01973"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_15076"} +{"question": "What papers discuss efficient transfer learning for generalizable synthetic image detection?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "A Comprehensive Survey on Transfer Learning", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "Visual Prompt Tuning", "AdaptFormer: Adapting Vision Transformers for Scalable Visual\n Recognition", "Conditional Prompt Learning for Vision-Language Models", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2111.06377", "1911.02685", "2104.08691", "1902.00751", "2106.09685", "2203.12119", "2205.13535", "2203.05557", "2109.01134"], "source_meta": {"published_time": "20231227"}, "qid": "AutoScholarQuery_train_15077"} +{"question": "Which work identified a subset of 8 DMLab tasks related to memory?", "answer": ["Stabilizing Transformers for Reinforcement Learning"], "answer_arxiv_id": ["1910.06764"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_15078"} +{"question": "Which papers introduced the concept of feature decorrelation in the context of self-supervised learning, boosting generalization and class incremental learning?", "answer": ["VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "On Feature Decorrelation in Self-Supervised Learning", "Reducing Overfitting in Deep Networks by Decorrelating Representations", "Decorrelated Batch Normalization", "Mimicking the Oracle: An Initial Phase Decorrelation Approach for Class Incremental Learning"], "answer_arxiv_id": ["2105.04906", "2103.03230", "2105.00470", "1511.06068", "1804.08450", "2112.04731"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_15079"} +{"question": "What are the works focusing on panoramic images for predicting distinct scene structures in layout estimation?", "answer": ["LayoutNet: Reconstructing the 3D Room Layout from a Single RGB Image", "HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch\n Data Augmentation", "HoHoNet: 360 Indoor Holistic Understanding with Latent Horizontal\n Features", "Corners for Layout: End-to-End Layout Recovery from 360 Images"], "answer_arxiv_id": ["1803.08999", "1901.03861", "2011.11498", "1903.08094"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15080"} +{"question": "Are there any papers that discuss the integration of extra learnable modules like Adapter, LoRA, and VPT during the finetuning stage?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "AdapterFusion: Non-Destructive Task Composition for Transfer Learning", "AdapterHub: A Framework for Adapting Transformers", "LoRA: Low-Rank Adaptation of Large Language Models", "Visual Prompt Tuning"], "answer_arxiv_id": ["1902.00751", "2005.00247", "2007.07779", "2106.09685", "2203.12119"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_15081"} +{"question": "Which studies have attempted to leverage 2D diffusion models to generate 3D assets from a text prompt in 3D generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2212.00774v1"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_15082"} +{"question": "What research works proposed confidence-based thresholding techniques to ensure the quality of pseudo labels?", "answer": ["Unsupervised Data Augmentation for Consistency Training", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "Dash: Semi-Supervised Learning with Dynamic Thresholding"], "answer_arxiv_id": ["1904.12848", "2001.07685v2", "2110.08263", "2109.00650"], "source_meta": {"published_time": "20220515"}, "qid": "AutoScholarQuery_train_15083"} +{"question": "What research used foreground object labels to predict foreground Depth Maps for depth guidance?", "answer": ["MonoDETR: Depth-guided Transformer for Monocular 3D Object Detection"], "answer_arxiv_id": ["2203.13310"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_15084"} +{"question": "Any works addressing the high bias and variance issue of IPS, especially in large action spaces?", "answer": ["Counterfactual Risk Minimization: Learning from Logged Bandit Feedback", "Off-policy Bandits with Deficient Support", "CAB: Continuous Adaptive Blending for Policy Evaluation and Learning", "Doubly robust off-policy evaluation with shrinkage"], "answer_arxiv_id": ["1502.02362", "2006.09438", "1811.02672", "1907.09623"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_15085"} +{"question": "Are there any studies that focus on providing sufficient global conditions for retraining to converge or proposing general optimization algorithms?", "answer": ["Performative Prediction", "Outside the Echo Chamber: Optimizing the Performative Risk", "Performative Prediction in a Stateful World", "Stochastic Optimization for Performative Prediction", "How to Learn when Data Reacts to Your Model: Performative Gradient Descent", "Stochastic optimization with decision-dependent distributions", "Zeroth-Order Methods for Convex-Concave Minmax Problems: Applications to Decision-Dependent Risk Minimization"], "answer_arxiv_id": ["2002.06673", "2102.08570", "2011.03885", "2006.06887", "2102.07698", "2011.11173", "2106.09082"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_15086"} +{"question": "Are there any studies that proposed low-rank structural conditions in function approximation?", "answer": ["Contextual Decision Processes with Low Bellman Rank are PAC-Learnable", "Model-based RL in Contextual Decision Processes: PAC bounds and Exponential Improvements over Model-free Approaches"], "answer_arxiv_id": ["1610.09512v2", "1811.08540"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_15087"} +{"question": "Could you provide me some works where synthetic samples are generated using data augmentations or pretrained generative networks like BigBiGAN?", "answer": ["Unsupervised Meta-Learning for Few-Shot Image Classification", "Unsupervised Meta-Learning through Latent-Space Interpolation in Generative Models"], "answer_arxiv_id": ["1811.11819", "2006.10236"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_15088"} +{"question": "In which works is the Transformer block utilized to aggregate instance features?", "answer": ["Attention Is All You Need", "TransMIL: Transformer based Correlated Multiple Instance Learning for\n Whole Slide Image Classification"], "answer_arxiv_id": ["1706.03762", "2106.00908"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_15089"} +{"question": "Which studies have constructed 3D priors based on collections of 3D primitives using an encoder-decoder approach?", "answer": ["DISN: Deep Implicit Surface Network for High-quality Single-view 3D\n Reconstruction", "Occupancy Networks: Learning 3D Reconstruction in Function Space"], "answer_arxiv_id": ["1905.10711", "1812.03828"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_15090"} +{"question": "Could you provide examples of research that employed diversity-based methods in Active Learning (AL)?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds"], "answer_arxiv_id": ["1708.00489", "1906.03671"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_15091"} +{"question": "Which study detailed the PIPs++ modifications and their impact on occlusion events?", "answer": ["PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point\n Tracking"], "answer_arxiv_id": ["2307.15055"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_15092"} +{"question": "Which works utilize point trajectories to regularize network optimization in the field of temporal domain expansion of NeRF?", "answer": ["Neural Radiance Flow for 4D View Synthesis and Video Processing", "Neural Trajectory Fields for Dynamic Novel View Synthesis"], "answer_arxiv_id": ["2012.09790", "2105.05994"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15093"} +{"question": "Which works studied the effect of architecture design choices in ResNets on adversarial robustness?", "answer": ["Do Wider Neural Networks Really Help Adversarial Robustness?", "Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks"], "answer_arxiv_id": ["2010.01279", "2110.03825"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_15094"} +{"question": "Any studies that suggest modeling uncertainty in classification by using a heatmap target?", "answer": ["Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimation", "Stacked Hourglass Networks for Human Pose Estimation"], "answer_arxiv_id": ["1406.2984", "1603.06937"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_15095"} +{"question": "Which works involved adapting MT LMs to align with human preferences through reinforcement learning?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_15096"} +{"question": "Which research shows the same for the lottery ticket setting in which weights are rewound to their values from early in training after pruning?", "answer": ["The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks"], "answer_arxiv_id": ["1803.03635"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_15097"} +{"question": "Could you list the papers utilizing vector-based methods for solving equivariance?", "answer": ["Vector Neurons: A General Framework for SO(3)-Equivariant Networks", "Learning from Protein Structure with Geometric Vector Perceptrons", "E(n) Equivariant Graph Neural Networks", "Equivariant message passing for the prediction of tensorial properties\n and molecular spectra", "SVNet: Where SO(3) Equivariance Meets Binarization on Point Cloud\n Representation"], "answer_arxiv_id": ["2104.12229", "2009.01411", "2102.09844", "2102.03150", "2209.05924"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_15098"} +{"question": "Could you mention some works that focused on shape reconstruction from sparse multi-view images?", "answer": ["On the generalization of learning-based 3D reconstruction", "Pix2Vox++: Multi-scale Context-aware 3D Object Reconstruction from\n Single and Multiple Images", "FvOR: Robust Joint Shape and Pose Optimization for Few-view Object\n Reconstruction", "3D-R2N2: A Unified Approach for Single and Multi-view 3D Object\n Reconstruction", "SSDNeRF: Semantic Soft Decomposition of Neural Radiance Fields"], "answer_arxiv_id": ["2006.15427", "2006.12250", "2205.07763", "1604.00449", "2212.03406"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_15099"} +{"question": "Can you provide some studies that have been focusing on acceleration of NeRF variants?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2201.05989", "2112.05131", "2203.09517"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_15100"} +{"question": "Any works that discuss the DINO model and its various uses?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "DINOv2: Learning Robust Visual Features without Supervision", "NeRF-SOS: Any-View Self-supervised Object Segmentation on Complex Scenes", "Splicing ViT Features for Semantic Appearance Transfer", "Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "Deep ViT Features as Dense Visual Descriptors"], "answer_arxiv_id": ["2104.14294", "2304.07193", "2209.08776", "2201.00424", "2203.08414", "2112.05814"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_15101"} +{"question": "What is the novel use of embeddings in Embroid's counterparts?", "answer": ["Training Subset Selection for Weak Supervision", "Label Propagation with Weak Supervision"], "answer_arxiv_id": ["2206.02914", "2210.03594"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_15102"} +{"question": "Are there any works that introduced FedUL, an approach for training a global model with only unlabeled clients?", "answer": ["Federated Learning from Only Unlabeled Data with Class-Conditional-Sharing Clients"], "answer_arxiv_id": ["2204.03304"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_15103"} +{"question": "Which works have leveraged 2D foundation models for open-world 3D understanding?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP", "CLIP2Point: Transfer CLIP to Point Cloud Classification with Image-Depth\n Pre-training", "OpenScene: 3D Scene Understanding with Open Vocabularies", "ConceptFusion: Open-set Multimodal 3D Mapping", "OpenIns3D: Snap and Lookup for 3D Open-vocabulary Instance Segmentation", "OpenMask3D: Open-Vocabulary 3D Instance Segmentation"], "answer_arxiv_id": ["2112.02413", "2210.01055", "2211.15654", "2302.07241", "2309.00616", "2306.13631"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_15104"} +{"question": "In what papers was the denoising procedure viewed as parameterizing the gradients of the data distribution?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_15105"} +{"question": "What research has examined quantization techniques focused on low-precision data types for model weights and activations?", "answer": ["PTQD: Accurate Post-Training Quantization for Diffusion Models", "Post-training Quantization on Diffusion Models"], "answer_arxiv_id": ["2305.10657", "2211.15736"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_15106"} +{"question": "Which research papers focus on discrete-time treatment–outcome setups in counterfactual reasoning?", "answer": ["Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models", "Decisions, Counterfactual Explanations and Strategic Behavior", "Counterfactual Explanations in Sequential Decision Making Under Uncertainty", "Meaningfully Debugging Model Mistakes using Conceptual Counterfactual Explanations"], "answer_arxiv_id": ["1905.05824", "2002.04333", "2107.02776", "2106.12723"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_15107"} +{"question": "Can you name some works that proposed other uncertainty measures from a trained model?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Energy-based Out-of-distribution Detection"], "answer_arxiv_id": ["1506.02142", "1612.01474", "2010.03759"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_15108"} +{"question": "Any works analyzed the case where the efficiency of graph Laplacians is decreased to zero?", "answer": ["Convergence of Laplacian spectra from random samples", "A graph discretization of the Laplace-Beltrami operator", "Graph approximations to the Laplacian spectra"], "answer_arxiv_id": ["1507.00151", "1301.2222", "1910.09224"], "source_meta": {"published_time": "20210728"}, "qid": "AutoScholarQuery_train_15109"} +{"question": "What works explored the potential of mesh agnostic networks for learning 3D information?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "MeshCNN: A Network with an Edge", "DiffusionNet: Discretization Agnostic Learning on Surfaces"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1809.05910", "2012.00888v3"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_15110"} +{"question": "What studies introduce a candidate zero-shot RL method called forward-backward representations?", "answer": ["Learning Successor States and Goal-Dependent Values: A Mathematical Viewpoint", "Learning One Representation to Optimize All Rewards"], "answer_arxiv_id": ["2101.07123", "2103.07945"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_15111"} +{"question": "What research used the notion of Lorenz efficiency to generate rankings that increase the utility of both the worst-off users and producers?", "answer": ["Two-sided fairness in rankings via Lorenz dominance"], "answer_arxiv_id": ["2110.15781"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_15112"} +{"question": "Could you provide me works in which exemplar-guided image editing is performed conditioning on stylized images or layouts?", "answer": ["AdaAttN: Revisit Attention Mechanism in Arbitrary Neural Style Transfer", "StyTr2:Image Style Transfer with Transformers", "ReCo: Region-Controlled Text-to-Image Generation", "Gligen: Open-Set Grounded Text-to-Image Generation", "High-Resolution Complex Scene Synthesis with Transformers"], "answer_arxiv_id": ["2108.03647", "2105.14576", "2211.15518", "2301.07093", "2105.06458"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15113"} +{"question": "What papers used diffusion process to model motion sequences?", "answer": ["Human Motion Diffusion Model", "MoFusion: A Framework for Denoising-Diffusion-based Motion Synthesis"], "answer_arxiv_id": ["2209.14916", "2212.04495"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_15114"} +{"question": "Which works employed spike cameras for image reconstruction by analysing spike intervals and counts?", "answer": ["A Retina-inspired Sampling Method for Visual Texture Reconstruction"], "answer_arxiv_id": ["1907.08769"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_15115"} +{"question": "Which works proposed a coupling of classical reconstruction methods and CNN in operator learning?", "answer": ["Beltrami-Net: Domain Independent Deep D-bar Learning for Absolute Imaging with Electrical Impedance Tomography (a-EIT)", "Deep D-bar: Real time Electrical Impedance Tomography Imaging with Deep Neural Networks", "BCR-Net: a neural network based on the nonstandard wavelet form", "Solving Electrical Impedance Tomography with Deep Learning"], "answer_arxiv_id": ["1811.12830", "1711.03180", "1810.08754", "1906.03944"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_15116"} +{"question": "Which papers found the geometric-based data augmentation to be especially beneficial?", "answer": ["RandAugment: Practical automated data augmentation with a reduced search\n space"], "answer_arxiv_id": ["1909.13719"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15117"} +{"question": "Are there any studies that applied Distributional Robustness to improve model robustness?", "answer": ["Global-Local Regularization Via Distributional Robustness", "A Unified Wasserstein Distributional Robustness Framework for Adversarial Training", "Certifying Some Distributional Robustness with Principled Adversarial Training"], "answer_arxiv_id": ["2203.00553v3", "2202.13437", "1710.10571v5"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_15118"} +{"question": "What paper is associated with projecting CLIP features into 3D meshes?", "answer": ["ConceptFusion: Open-set Multimodal 3D Mapping"], "answer_arxiv_id": ["2302.07241"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_15119"} +{"question": "Any papers about active learning approaches specifically for image classification and object detection?", "answer": ["Crowdsourcing in Computer Vision", "Active Learning for Deep Object Detection via Probabilistic Modeling", "Active Learning for Deep Object Detection", "Active Learning for Deep Neural Networks on Edge Devices"], "answer_arxiv_id": ["1611.02145", "2103.16130", "1809.09875", "2106.10836"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_15120"} +{"question": "Which papers discussed the generalizability of pre-trained vision models on various downstream tasks?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "ImageBind: One Embedding Space To Bind Them All"], "answer_arxiv_id": ["2103.00020", "2201.12086", "2102.05918", "2305.05665v2"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_15121"} +{"question": "What works have explored code generation with neural networks?", "answer": ["Latent Predictor Networks for Code Generation", "Mapping Language to Code in Programmatic Context", "A Syntactic Neural Model for General-Purpose Code Generation", "Graph-based, Self-Supervised Program Repair from Diagnostic Feedback"], "answer_arxiv_id": ["1603.06744", "1808.09588", "1704.01696", "2005.10636"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_15122"} +{"question": "Which papers initially introduced the concept of coreset selection?", "answer": ["Practical Coreset Constructions for Machine Learning", "Super-Samples from Kernel Herding", "Active Learning for Convolutional Neural Networks: A Core-Set Approach"], "answer_arxiv_id": ["1703.06476v2", "1203.3472", "1708.00489"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_15123"} +{"question": "What works proposed a modularized model based on mPLUG-Owl for OCR-free document understanding?", "answer": ["mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document\n Understanding"], "answer_arxiv_id": ["2307.02499"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_15124"} +{"question": "Could you provide me some studies that discuss the influence of eliminating overlapping data between pre-training and downstream tasks in CLIP on the performance?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15125"} +{"question": "Which papers discuss the application of bilevel optimization in neural architecture search?", "answer": ["DARTS: Differentiable Architecture Search", "iDARTS: Differentiable Architecture Search with Stochastic Implicit Gradients"], "answer_arxiv_id": ["1806.09055", "2106.10784"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_15126"} +{"question": "Which works aim to understand unsupervised anomaly detection models?", "answer": ["Learning Robust Representations via Multi-View Information Bottleneck", "Discovering Invariant Rationales for Graph Neural Networks", "Towards Self-Explainable Graph Neural Network", "Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanism"], "answer_arxiv_id": ["2002.07017", "2201.12872", "2108.12055", "2201.12987"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_15127"} +{"question": "Could you highlight works that discussed the advantages of the Wasserstein distance?", "answer": ["Distributionally Robust Stochastic Optimization with Wasserstein Distance"], "answer_arxiv_id": ["1604.02199v3"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_15128"} +{"question": "Can you mention some works that have proposed meta-learning algorithms comparable to ADKF-IFT?", "answer": ["Matching Networks for One Shot Learning", "Conditional Neural Processes", "Meta-Dataset: A Dataset of Datasets for Learning to Learn from Few Examples", "Meta-Curvature", "Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?", "Self-supervised learning for few-shot image classification", "Meta Two-Sample Testing: Learning Kernels for Testing with Limited Data", "Few-Shot Bayesian Optimization with Deep Kernel Surrogates", "Contextual Squeeze-and-Excitation for Efficient Few-Shot Image Classification"], "answer_arxiv_id": ["1606.04080", "1807.01613", "1903.03096", "1902.03356", "2003.11539", "1911.06045", "2106.07636", "2101.07667", "2206.09843"], "source_meta": {"published_time": "20220505"}, "qid": "AutoScholarQuery_train_15129"} +{"question": "What research revolutionized the field by utilizing dense contextualized vectors for document indexing?", "answer": ["Latent Retrieval for Weakly Supervised Open Domain Question Answering", "Dense Passage Retrieval for Open-Domain Question Answering"], "answer_arxiv_id": ["1906.00300", "2004.04906"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_15130"} +{"question": "Could you list down the studies which proposed methods for Depth Completion?", "answer": ["Learning Joint 2D-3D Representations for Depth Completion", "HMS-Net: Hierarchical Multi-scale Sparsity-invariant Network for Sparse\n Depth Completion", "Sparse and Dense Data with CNNs: Depth Completion and Semantic\n Segmentation", "Self-supervised Sparse-to-Dense: Self-supervised Depth Completion from\n LiDAR and Monocular Camera", "Dense Depth Posterior (DDP) from Single Image and Sparse Range", "PENet: Towards Precise and Efficient Image Guided Depth Completion", "CSPN++: Learning Context and Resource Aware Convolutional Spatial\n Propagation Networks for Depth Completion", "Non-Local Spatial Propagation Network for Depth Completion", "Dynamic Spatial Propagation Network for Depth Completion", "Uncertainty-Aware CNNs for Depth Completion: Uncertainty from Beginning\n to End", "Propagating Confidences through CNNs for Sparse Data Regression", "Bayesian Deep Basis Fitting for Depth Completion with Uncertainty", "Depth Completion via Deep Basis Fitting", "Sparse and noisy LiDAR completion with RGB guidance and uncertainty", "DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene\n from Sparse LiDAR Data and Single Color Image", "Depth Completion from Sparse LiDAR Data with Depth-Normal Constraints", "Deep Depth Completion of a Single RGB-D Image", "CompletionFormer: Depth Completion with Convolutions and Vision\n Transformers"], "answer_arxiv_id": ["2012.12402", "1808.08685", "1808.00769", "1807.00275", "1901.10034", "2103.00783", "1911.05377", "2007.10042", "2202.09769", "2006.03349", "1805.11913", "2103.15254v1", "1912.10336", "1902.05356", "1812.00488", "1910.06727", "1803.09326", "2304.13030"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_15131"} +{"question": "Which works have discussed major advances in learning useful representations from unlabelled data?", "answer": ["Self-supervised Learning: Generative or Contrastive", "Self-supervised Pretraining of Visual Features in the Wild"], "answer_arxiv_id": ["2006.08218", "2103.01988"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_15132"} +{"question": "What works optimized a voxel grid instead of an MLP to accelerate both training and rendering in Neural Radiance Fields?", "answer": ["Plenoxels: Radiance Fields without Neural Networks", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction"], "answer_arxiv_id": ["2112.05131", "2111.11215"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15133"} +{"question": "Can you give references of the papers that dealt with Q-learning with pessimism for offline RL?", "answer": ["Pessimistic Q-Learning for Offline Reinforcement Learning: Towards Optimal Sample Complexity", "The Efficacy of Pessimism in Asynchronous Q-Learning"], "answer_arxiv_id": ["2202.13890", "2203.07368"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_15134"} +{"question": "Which works in literature explored the techniques of transforming time series data into different types of images?", "answer": ["Imaging Time-Series to Improve Classification and Imputation", "Classification of Time-Series Images Using Deep Convolutional Neural Networks", "Spatially Encoding Temporal Correlations to Classify Temporal Data Using Convolutional Neural Networks"], "answer_arxiv_id": ["1506.00327", "1710.00886", "1509.07481v1"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_15135"} +{"question": "Can you provide some studies that proposed Reinforcement Learning from AI Feedback (RLAIF) to address the annotation cost issue in RLHF?", "answer": ["Constitutional AI: Harmlessness from AI Feedback", "RLAIF: Scaling Reinforcement Learning from Human Feedback with AI\n Feedback", "SALMON: Self-Alignment with Instructable Reward Models"], "answer_arxiv_id": ["2212.08073", "2309.00267", "2310.05910"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_15136"} +{"question": "Could you provide me some studies about Out-of-Distribution (OOD) objectives aiming to regularize ERM for learning invariant features?", "answer": ["Invariant Risk Minimization", "Out-of-Distribution Generalization via Risk Extrapolation (REx)", "Gradient Starvation: A Learning Proclivity in Neural Networks", "On Calibration and Out-of-domain Generalization", "Fishr: Invariant Gradient Variances for Out-of-distribution Generalization"], "answer_arxiv_id": ["1907.02893", "2003.00688v5", "2011.09468", "2102.10395", "2109.02934"], "source_meta": {"published_time": "20230422"}, "qid": "AutoScholarQuery_train_15137"} +{"question": "Which work successfully used PPO technique to fine tune their LLM and achieved to generate more human-aligned outputs?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_15138"} +{"question": "Could you tell me the names of some papers that study large vision-language models (LVLMs)?", "answer": ["ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "UNITER: UNiversal Image-TExt Representation Learning", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks"], "answer_arxiv_id": ["1908.02265", "1909.11740", "2004.06165"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15139"} +{"question": "Which papers have focused on improving truthfulness of model by editing all internal representations of LLM?", "answer": ["Does Localization Inform Editing? Surprising Differences in\n Causality-Based Localization vs. Knowledge Editing in Language Models"], "answer_arxiv_id": ["2301.04213"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_15140"} +{"question": "What work decoupled feature generation and neural rendering using a tri-plane representation?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["2112.07945"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_15141"} +{"question": "Are there any multimodal benchmarks established to evaluate advanced LMMs?", "answer": ["MMBench: Is Your Multi-modal Model an All-around Player?", "SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension"], "answer_arxiv_id": ["2307.06281", "2307.16125"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_15142"} +{"question": "Which work centers the computation of cost volume on the reference view and uses a 3D CNN for refinement?", "answer": ["MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo"], "answer_arxiv_id": ["2103.15595"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_15143"} +{"question": "Can you list the papers proposing GFlowNets as a means of describing distributions over composite objects that can be constructed sequentially?", "answer": ["GFlowNet Foundations"], "answer_arxiv_id": ["2111.09266"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15144"} +{"question": "Can you name some works focussing on the use of superquadratics for geometric abstraction?", "answer": ["Superquadrics Revisited: Learning 3D Shape Parsing beyond Cuboids", "Learning Unsupervised Hierarchical Part Decomposition of 3D Objects from a Single RGB Image"], "answer_arxiv_id": ["1904.09970", "2004.01176"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_15145"} +{"question": "Could you provide me some works about reconstruction of compositional scenes using pixel-aligned conditioning?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "GRF: Learning a General Radiance Field for 3D Representation and Rendering"], "answer_arxiv_id": ["2012.02190", "2010.04595"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_15146"} +{"question": "Which studies extended the robustness of policies to more tractable formulations and structured uncertainty sets?", "answer": ["Action Robust Reinforcement Learning and Applications in Continuous Control", "Robust Reinforcement Learning for Continuous Control with Model Misspecification", "Robust Adversarial Reinforcement Learning", "Robust Reinforcement Learning on State Observations with Learned Optimal Adversary", "Scaling Up Robust MDPs by Reinforcement Learning"], "answer_arxiv_id": ["1901.09184", "1906.07516", "1703.02702", "2101.08452", "1306.6189"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_15147"} +{"question": "What are the works that employed dataset distillation for tasks like continual learning, federated learning and neural architecture search?", "answer": ["Dataset Condensation via Efficient Synthetic-Data Parameterization", "Meta Knowledge Condensation for Federated Learning", "FedDM: Iterative Distribution Matching for Communication-Efficient\n Federated Learning", "Generative Teaching Networks: Accelerating Neural Architecture Search by\n Learning to Generate Synthetic Training Data"], "answer_arxiv_id": ["2205.14959v2", "2209.14851", "2207.09653", "1912.07768"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_15148"} +{"question": "Which papers studied the concept of representation collapse in the context of metric learning?", "answer": ["Revisiting Training Strategies and Generalization Performance in Deep\n Metric Learning"], "answer_arxiv_id": ["2002.08473"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_15149"} +{"question": "Are there papers that discuss the use of projection maintenance data structure in semidefinite programming?", "answer": ["An Improved Cutting Plane Method for Convex Optimization, Convex-Concave Games and its Applications", "A Faster Interior Point Method for Semidefinite Programming", "Solving SDP Faster: A Robust IPM Framework and Efficient Implementation", "A Faster Quantum Algorithm for Semidefinite Programming via Robust IPM Framework", "A Faster Small Treewidth SDP Solver"], "answer_arxiv_id": ["2004.04250v1", "2009.10217", "2101.08208", "2207.11154", "2211.06033"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_15150"} +{"question": "Can you list the studies about data poisoning?", "answer": ["Generative Poisoning Attack Method Against Neural Networks", "Certified Defenses for Data Poisoning Attacks"], "answer_arxiv_id": ["1703.01340", "1706.03691"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_15151"} +{"question": "Could you provide me some studies about end-to-end training with a Vision Transformer (ViT) as the visual encoder in the VLM paradigm?", "answer": ["Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks"], "answer_arxiv_id": ["2107.07651", "2201.12086", "2208.10442"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_15152"} +{"question": "Which research papers have shown that neural networks struggle to generalize to problems requiring a combination of reasoning steps not seen during training?", "answer": ["Generalization without systematicity: On the compositional skills of\n sequence-to-sequence recurrent networks", "Unobserved Local Structures Make Compositional Generalization Hard", "Data Factors for Better Compositional Generalization"], "answer_arxiv_id": ["1711.00350", "2201.05899", "2311.04420"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_15153"} +{"question": "Which work found that the memorization ability of large language models significantly grows with various factors?", "answer": ["Quantifying Memorization Across Neural Language Models"], "answer_arxiv_id": ["2202.07646"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_15154"} +{"question": "Any works that employ higher-order logic rules for exploring new propositions in cumulative reasoning?", "answer": ["Cumulative Reasoning with Large Language Models"], "answer_arxiv_id": ["2308.04371"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_15155"} +{"question": "Which work proposed a novel cross-language transfer prompt method?", "answer": ["Boosting Cross-lingual Transferability in Multilingual Models via\n In-Context Learning"], "answer_arxiv_id": ["2305.15233"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_15156"} +{"question": "Which works illustrate that count-based exploration provides intrinsic rewards?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "Count-Based Exploration with Neural Density Models", "#Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning", "Count-Based Exploration with the Successor Representation", "LECO: Learnable Episodic Count for Task-Specific Intrinsic Reward"], "answer_arxiv_id": ["1606.01868", "1703.01310", "1611.04717", "1807.11622", "2210.05409"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_15157"} +{"question": "Which works utilize the hard parameter sharing paradigm in Multi-Task Learning (MTL)?", "answer": ["UberNet: Training a `Universal' Convolutional Neural Network for Low-,\n Mid-, and High-Level Vision using Diverse Datasets and Limited Memory", "GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep\n Multitask Networks", "Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics", "Multi-Task Learning as Multi-Objective Optimization"], "answer_arxiv_id": ["1609.02132", "1711.02257", "1705.07115v3", "1810.04650"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_15158"} +{"question": "Could you tell me which papers did a convergence analysis of gradient descent relying on the near constancy of NTK for wide neural networks?", "answer": ["Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "An Improved Analysis of Training Over-parameterized Deep Neural Networks", "Loss landscapes and optimization in over-parameterized non-linear systems and neural networks"], "answer_arxiv_id": ["1811.03804", "1811.03962", "1906.04688", "2003.00307"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_15159"} +{"question": "Which works obtained conditional lower bounds for pattern matching problems using the orthogonal vector conjecture?", "answer": ["Why walking the dog takes time: Frechet distance has no strongly subquadratic algorithms unless SETH fails", "Edit Distance Cannot Be Computed in Strongly Subquadratic Time (unless SETH is false)", "Which Regular Expression Patterns are Hard to Match?", "A Dichotomy for Regular Expression Membership Testing", "Multivariate Fine-Grained Complexity of Longest Common Subsequence", "An Equivalence Class for Orthogonal Vectors"], "answer_arxiv_id": ["1404.1448v2", "1412.0348", "1511.07070", "1611.00918", "1803.00938v1", "1811.12017v1"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_15160"} +{"question": "Could you provide me some works that have proposed recurrent versions of transformers?", "answer": ["Universal Transformers", "Staircase Attention for Recurrent Processing of Sequences", "Block-Recurrent Transformers"], "answer_arxiv_id": ["1807.03819", "2106.04279", "2203.07852"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_15161"} +{"question": "Are there any works that employed instruction tuning for list-level recommendation generation?", "answer": ["PALR: Personalization Aware LLMs for Recommendation"], "answer_arxiv_id": ["2305.07622v3"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_15162"} +{"question": "Which research papers theoretically analyze the benefits of data augmentation?", "answer": ["A Kernel Theory of Modern Data Augmentation", "On the Generalization Effects of Linear Transformations in Data Augmentation", "How Data Augmentation affects Optimization for Linear Regression", "Sample Efficiency of Data Augmentation Consistency Regularization", "Data Augmentation vs. Equivariant Networks: A Theory of Generalization on Dynamics Forecasting", "A Group-Theoretic Framework for Data Augmentation", "Learning with invariances in random features and kernel models"], "answer_arxiv_id": ["1803.06084", "2005.00695", "2010.11171", "2202.12230", "2206.09450v1", "1907.10905", "2102.13219"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_15163"} +{"question": "Could you provide me some works on graph-based generative models for molecule topological structures generation?", "answer": ["Learning Multimodal Graph-to-Graph Translation for Molecular Optimization", "Junction Tree Variational Autoencoder for Molecular Graph Generation", "Multi-Objective De Novo Drug Design with Conditional Graph Generative Model"], "answer_arxiv_id": ["1812.01070", "1802.04364", "1801.07299v3"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_train_15164"} +{"question": "What are some studies that focus on reducing LLM inferences by pruning or compressing KV cache data?", "answer": ["H$_2$O: Heavy-Hitter Oracle for Efficient Generative Inference of Large\n Language Models"], "answer_arxiv_id": ["2306.14048"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_15165"} +{"question": "What are some works employing the effectiveness of diffusion-based text-guided image manipulation?", "answer": ["DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "Diffusion Models already have a Semantic Latent Space"], "answer_arxiv_id": ["2110.02711", "2210.10960"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_15166"} +{"question": "What works proposed variational autoencoders for compression of visual observations into latent vectors?", "answer": ["Recurrent World Models Facilitate Policy Evolution"], "answer_arxiv_id": ["1809.01999"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15167"} +{"question": "Which work discusses a setting to adapt the conformal bounds through a scaling factor in the nonconformity function?", "answer": ["Conformal Prediction: a Unified Review of Theory and New Challenges"], "answer_arxiv_id": ["2005.07972"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_15168"} +{"question": "Which contributions proposed enhancements to diffusion models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance"], "answer_arxiv_id": ["2102.09672", "2105.05233", "2207.12598"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_15169"} +{"question": "What papers investigate score and displacement based methods inspired by diffusive processes?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "Flow Straight and Fast: Learning to Generate and Transfer Data with\n Rectified Flow", "Fast Point Cloud Generation with Straight Flows"], "answer_arxiv_id": ["1907.05600", "2103.01458", "2209.03003", "2212.01747"], "source_meta": {"published_time": "20240514"}, "qid": "AutoScholarQuery_train_15170"} +{"question": "What work focuses on feature-level smoothing to reduce the effects of flickering in a video?", "answer": ["TokenFlow: Consistent Diffusion Features for Consistent Video Editing"], "answer_arxiv_id": ["2307.10373"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15171"} +{"question": "What work presented a policy expansion strategy that adaptively selects a policy from a set?", "answer": ["Policy Expansion for Bridging Offline-to-Online Reinforcement Learning"], "answer_arxiv_id": ["2302.00935"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_15172"} +{"question": "What studies tackled the task of visual grounding in 3D question answering?", "answer": ["Neural Modular Control for Embodied Question Answering", "3D Question Answering"], "answer_arxiv_id": ["1810.11181", "2112.08359"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_15173"} +{"question": "What study uses Chain-of-Thought tasks during instruction tuning?", "answer": ["Scaling Instruction-Finetuned Language Models"], "answer_arxiv_id": ["2210.11416"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_15174"} +{"question": "What studies have shown that geometric ensembling techniques can improve the quality of image super-resolution?", "answer": ["Seven ways to improve example-based single image super resolution"], "answer_arxiv_id": ["1511.02228"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15175"} +{"question": "What works have helped in characterizing the structure for deep fully-connected networks (FCNs), influencing the Hessian rank of arbitrary sized networks?", "answer": ["Analytic Insights into Structure and Rank of Neural Network Hessian Maps"], "answer_arxiv_id": ["2106.16225v2"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_15176"} +{"question": "Are there studies that apply different divergence functions like KL-divergence in BCPO paradigm?", "answer": ["Behavior Regularized Offline Reinforcement Learning", "Way Off-Policy Batch Deep Reinforcement Learning of Implicit Human Preferences in Dialog"], "answer_arxiv_id": ["1911.11361", "1907.00456"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_15177"} +{"question": "What is the pioneering work in knowledge distillation that distills knowledge from a larger teacher model to a smaller student model?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_15178"} +{"question": "What papers are about the seminal works in Vision Transformers?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Swin Transformer V2: Scaling Up Capacity and Resolution", "CvT: Introducing Convolutions to Vision Transformers", "Multiscale Vision Transformers", "MViTv2: Improved Multiscale Vision Transformers for Classification and\n Detection"], "answer_arxiv_id": ["2010.11929", "2103.14030", "2111.09883", "2103.15808", "2104.11227", "2112.01526"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15179"} +{"question": "What is the study where the slot attention mechanism was used?", "answer": ["Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["2006.15055"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_15180"} +{"question": "Which work is relevant for self-consistent chain-of-thought?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_15181"} +{"question": "Which works combine GANs and NeRF to synthesize high-fidelity novel views?", "answer": ["pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis", "GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature\n Fields", "Efficient Geometry-aware 3D Generative Adversarial Networks", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image\n Synthesis", "GRAM: Generative Radiance Manifolds for 3D-Aware Image Generation"], "answer_arxiv_id": ["2012.00926", "2007.02442", "2011.12100", "2112.07945", "2110.08985", "2112.08867"], "source_meta": {"published_time": "20240411"}, "qid": "AutoScholarQuery_train_15182"} +{"question": "Which study analysed the single-call variant?", "answer": ["Solving Nonconvex-Nonconcave Min-Max Problems exhibiting Weak Minty Solutions"], "answer_arxiv_id": ["2201.12247"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_15183"} +{"question": "What were the early attempts to apply the diffusion model to text-driven motion generation?", "answer": ["MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "Human Motion Diffusion Model", "FLAME: Free-form Language-based Motion Synthesis & Editing"], "answer_arxiv_id": ["2208.15001", "2209.14916", "2209.00349"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15184"} +{"question": "Which research studies have combined techniques like consistency regularization and pseudo-label to uncrease performance in semi-supervised learning?", "answer": ["Dash: Semi-Supervised Learning with Dynamic Thresholding", "AdaMatch: A Unified Approach to Semi-Supervised Learning and Domain Adaptation", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "SoftMatch: Addressing the Quantity-Quality Trade-off in Semi-supervised Learning", "MixMatch: A Holistic Approach to Semi-Supervised Learning"], "answer_arxiv_id": ["2109.00650", "2106.04732", "2001.07685v2", "2205.07246", "2110.08263", "2301.10921", "1905.02249"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_train_15185"} +{"question": "What studies have successfully applied generative flow networks on tasks like molecule generation?", "answer": ["Trajectory balance: Improved credit assignment in GFlowNets", "Biological Sequence Design with GFlowNets"], "answer_arxiv_id": ["2201.13259", "2203.04115"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_15186"} +{"question": "Could you show me some works about the study of convergence of an off-policy actor critic?", "answer": ["Doubly Robust Off-Policy Actor-Critic: Convergence and Optimality"], "answer_arxiv_id": ["2102.11866"], "source_meta": {"published_time": "20211104"}, "qid": "AutoScholarQuery_train_15187"} +{"question": "Can you point me to the research that leveraged the pretrained CLIP model for video-language tasks?", "answer": ["CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "X-CLIP: End-to-End Multi-grained Contrastive Learning for Video-Text\n Retrieval"], "answer_arxiv_id": ["2106.11097", "2207.07285"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_15188"} +{"question": "Which papers proposed techniques in geometric deep learning that represent shapes and scenes as level-sets of continuous functions?", "answer": ["Controlling Neural Level Sets"], "answer_arxiv_id": ["1905.11911"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_15189"} +{"question": "Are there any papers discussing multiple solution discovery methodologies such as niching, parallel multi-starts and deflation?", "answer": ["Computing multiple solutions of topology optimization problems"], "answer_arxiv_id": ["2004.11797v2"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_15190"} +{"question": "Could you provide me some works that researched on differentially private exact posterior sampling?", "answer": ["Privacy for Free: Posterior Sampling and Stochastic Gradient Monte Carlo", "On the Theory and Practice of Privacy-Preserving Bayesian Data Analysis", "On the Differential Privacy of Bayesian Inference"], "answer_arxiv_id": ["1502.07645", "1603.07294", "1512.06992"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_15191"} +{"question": "What research papers explore the utilization of adversarial training for representation learning and image synthesis?", "answer": ["Adversarial Robustness as a Prior for Learned Representations", "Decoder-free Robustness Disentanglement without (Additional) Supervision", "Image Synthesis with a Single (Robust) Classifier", "A Unified Contrastive Energy-based Model for Understanding the Generative Ability of Adversarial Training"], "answer_arxiv_id": ["1906.00945", "2007.01356", "1906.09453", "2203.13455"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15192"} +{"question": "Can you name some works that have proposed popular methods for confidence calibration?", "answer": ["Q", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration", "Non-Parametric Calibration for Classification"], "answer_arxiv_id": ["1611.08152", "1910.12656", "1906.04933"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15193"} +{"question": "Which works propose novel graph autoencoders for unsupervised representation learning on tabular multi-view data collected ITW?", "answer": ["Multimodal Deep Learning", "Auto-Encoding Variational Bayes", "Joint Multimodal Learning with Deep Generative Models", "Multimodal Generative Models for Scalable Weakly-Supervised Learning", "Variational Mixture-of-Experts Autoencoders for Multi-Modal Deep Generative Models"], "answer_arxiv_id": ["2301.04856v1", "1312.6114", "1611.01891", "1802.05335", "1911.03393"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15194"} +{"question": "What kinds of work have been done in the field of scene relighting both multi-view and single-view?", "answer": ["Free-viewpoint Indoor Neural Relighting from Multi-view Stereo", "Self-supervised Outdoor Scene Relighting", "OutCast: Outdoor Single-image Relighting with Cast Shadows"], "answer_arxiv_id": ["2106.13299v1", "2107.03106", "2204.09341"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_15195"} +{"question": "What studies have applied the interpretation of neural ODEs in graph neural networks?", "answer": ["Discrete and Continuous Deep Residual Learning Over Graphs", "Graph Neural Ordinary Differential Equations", "Continuous Graph Neural Networks"], "answer_arxiv_id": ["1911.09554", "1911.07532", "1912.00967"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_15196"} +{"question": "Which studies focused on neural collapse using MSE loss functions?", "answer": ["On the Optimization Landscape of Neural Collapse under MSE Loss: Global Optimality with Unconstrained Features", "Neural collapse with unconstrained features", "Extended Unconstrained Features Model for Exploring Deep Neural Collapse"], "answer_arxiv_id": ["2203.01238v2", "2011.11619", "2202.08087"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_15197"} +{"question": "Could you provide me some studies about the application of offline RL in the NLP domain?", "answer": ["Text Generation by Learning from Demonstrations", "CHAI: A CHatbot AI for Task-Oriented Dialogue with Offline Reinforcement Learning", "Way Off-Policy Batch Deep Reinforcement Learning of Implicit Human Preferences in Dialog", "Human-centric dialog training via offline reinforcement learning", "Offline RL for Natural Language Generation with Implicit Language Q Learning", "Quark: Controllable Text Generation with Reinforced [Un]learning"], "answer_arxiv_id": ["2009.07839", "2204.08426", "1907.00456", "2010.05848", "2206.11871", "2205.13636"], "source_meta": {"published_time": "20230723"}, "qid": "AutoScholarQuery_train_15198"} +{"question": "What works propose privacy-preserving Q-learning, actor critic algorithm with differentially private critic, and adaptive control of differentially private linear quadratic systems?", "answer": ["Privacy-preserving Q-Learning with Functional Noise in Continuous Spaces", "Locally Private Distributed Reinforcement Learning", "Actor Critic with Differentially Private Critic", "Privacy-Constrained Policies via Mutual Information Regularized Policy Gradients", "Adaptive Control of Differentially Private Linear Quadratic Systems"], "answer_arxiv_id": ["1901.10634", "2001.11718", "1910.05876", "2012.15019", "2108.11563"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_15199"} +{"question": "Can you mention any studies that worked on distilling a large teacher model into a much smaller student architecture?", "answer": ["Diffusion Probabilistic Model Made Slim"], "answer_arxiv_id": ["2211.17106"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_15200"} +{"question": "What studies focused on Lipschitz regularization on neural networks?", "answer": ["Learning Smooth Neural Functions via Lipschitz Regularization"], "answer_arxiv_id": ["2202.08345"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_15201"} +{"question": "What research works discuss personalization in the field of Federated Learning?", "answer": ["Personalized Federated Learning with Gaussian Processes", "Self-Aware Personalized Federated Learning"], "answer_arxiv_id": ["2106.15482", "2204.08069v1"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_15202"} +{"question": "Which studies focus on the enhancement of embedding structures for encoding sets of answers in complex logical query answering on KG?", "answer": ["Embedding Logical Queries on Knowledge Graphs"], "answer_arxiv_id": ["1806.01445"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_15203"} +{"question": "Could you provide me some studies that proposed improvements for Adaptive Polyphase Sampling?", "answer": ["Truly shift-invariant convolutional neural networks", "Learnable Polyphase Sampling for Shift Invariant and Equivariant\n Convolutional Networks"], "answer_arxiv_id": ["2011.14214", "2210.08001"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15204"} +{"question": "Any recent work that further investigated the univariate case of one-hidden layer ReLU networks?", "answer": ["Penalising the biases in norm regularisation enforces sparsity"], "answer_arxiv_id": ["2303.01353"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_15205"} +{"question": "Can you provide some works that have designed a network and corresponding loss function to manage spectral bias?", "answer": ["Spectral Unsupervised Domain Adaptation for Visual Recognition"], "answer_arxiv_id": ["2106.06112v3"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_15206"} +{"question": "What works discuss the dimension of check-worthiness of factual claims?", "answer": ["A Benchmark Dataset of Check-worthy Factual Claims", "Overview of the CLEF--2021 CheckThat! Lab on Detecting Check-Worthy\n Claims, Previously Fact-Checked Claims, and Fake News"], "answer_arxiv_id": ["2004.14425", "2109.12987"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_15207"} +{"question": "Which research work proposed the use of image modality in MAGMA and how it works by using a CLIP vision encoder?", "answer": ["MAGMA – Multimodal Augmentation of Generative Models through Adapter-based Finetuning", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2112.05253", "2103.00020"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_15208"} +{"question": "Could you provide some studies focusing on entropy regularization in BNNs?", "answer": ["Equal Bits: Enforcing Equally Distributed Binary Network Weights"], "answer_arxiv_id": ["2112.03406"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_15209"} +{"question": "Could you provide me some references where they proposed methods like training disconnected distributions or deriving rejection mechanisms from pre-trained generators to address misspecification issues?", "answer": ["DeLiGAN : Generative Adversarial Networks for Diverse and Limited Data", "Disconnected Manifold Learning for Generative Adversarial Networks", "Discriminator Rejection Sampling", "Learning Disconnected Manifolds: a no GAN’s land", "Polarity Sampling: Quality and Diversity Control of Pre-Trained Generative Networks via Singular Values"], "answer_arxiv_id": ["1706.02071", "1806.00880", "1810.06758", "2006.04596", "2203.01993"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_15210"} +{"question": "Could you provide me some studies that focus on the intersection of reinforcement learning and PID control?", "answer": ["A Model-Based Reinforcement Learning Approach for PID Design", "Reinforcement Learning based Design of Linear Fixed Structure Controllers"], "answer_arxiv_id": ["2206.03567v1", "2005.04537"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_15211"} +{"question": "What works introduced Music Language Models (MusicLM)?", "answer": ["MusicLM: Generating Music From Text"], "answer_arxiv_id": ["2301.11325"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_15212"} +{"question": "Could you list the works that have focused on conditional GANs using 3D variables as conditions?", "answer": ["Disentangled and Controllable Face Image Generation via 3D\n Imitative-Contrastive Learning", "GIF: Generative Interpretable Faces"], "answer_arxiv_id": ["2004.11660", "2009.00149"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_15213"} +{"question": "What papers cover the trending shift in semantic segmentation towards transformer-based architectures and attention mechanisms?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective\n with Transformers", "Segmenter: Transformer for Semantic Segmentation"], "answer_arxiv_id": ["2010.11929", "2012.15840", "2105.05633"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_15214"} +{"question": "What studies propose to control the risk of disentangled factors in the latent space of StyleGAN?", "answer": ["Semantic uncertainty intervals for disentangled latent spaces"], "answer_arxiv_id": ["2207.10074"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_15215"} +{"question": "Can you provide a reference that used upper bounds on the gradient of the estimate of the average reward function?", "answer": ["A general sample complexity analysis of vanilla policy gradient"], "answer_arxiv_id": ["2107.11433v5"], "source_meta": {"published_time": "20221114"}, "qid": "AutoScholarQuery_train_15216"} +{"question": "In which studies has feature extraction from satellite imagery been used for retrieval itself?", "answer": ["Asymmetric Hash Code Learning for Remote Sensing Image Retrieval"], "answer_arxiv_id": ["2201.05772"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_15217"} +{"question": "Which papers proposed to train rendered image-to-shape generation?", "answer": ["CLIP-Forge: Towards Zero-Shot Text-to-Shape Generation", "CLIP-Sculptor: Zero-Shot Generation of High-Fidelity and Diverse Shapes from Natural Language", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation"], "answer_arxiv_id": ["2110.02624", "2211.01427", "2212.04493"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_15218"} +{"question": "What research learns a spatiotemporal irradiance field directly from a single video and resolves the shape-motion ambiguities in monocular inputs?", "answer": ["Space-time Neural Irradiance Fields for Free-Viewpoint Video"], "answer_arxiv_id": ["2011.12950"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15219"} +{"question": "Which works proposed to use the causal effect of non-labeled spurious attributes in pre-training to prevent spurious correlations?", "answer": ["Causal Transportability for Visual Recognition"], "answer_arxiv_id": ["2204.12363"], "source_meta": {"published_time": "20230408"}, "qid": "AutoScholarQuery_train_15220"} +{"question": "Which work defined an instance's computational difficulty based on the number of hidden layers after which the network's prediction matches the output?", "answer": ["Deep Learning Through the Lens of Example Difficulty"], "answer_arxiv_id": ["2106.09647"], "source_meta": {"published_time": "20230103"}, "qid": "AutoScholarQuery_train_15221"} +{"question": "What papers are about the development of attention networks in single image super-resolution?", "answer": ["Image Super-Resolution Using Very Deep Residual Channel Attention Networks"], "answer_arxiv_id": ["1807.02758"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15222"} +{"question": "Could you provide examples of works that contributed to the concept of invariant learning in domain generalization?", "answer": ["Deep CORAL: Correlation Alignment for Deep Domain Adaptation", "Domain-Adversarial Training of Neural Networks", "Invariant Risk Minimization"], "answer_arxiv_id": ["1607.01719", "1505.07818", "1907.02893"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_15223"} +{"question": "What are some key contributions to Federated Continual Learning?", "answer": ["Federated Continual Learning with Weighted Inter-client Transfer", "Federated Class-Incremental Learning", "FedSpeech: Federated Text-to-Speech with Continual Learning", "Continual Distributed Learning for Crisis Management", "A distillation-based approach integrating continual learning and\n federated learning for pervasive services"], "answer_arxiv_id": ["2003.03196", "2203.11473", "2110.07216", "2104.12876", "2109.04197"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_15224"} +{"question": "Could you name some works that have focused on automated explanation quality evaluation?", "answer": ["Investigating the Benefits of Free-Form Rationales", "Are Machine Rationales (Not) Useful to Humans? Measuring and Improving\n Human Utility of Free-Text Rationales", "Measuring Association Between Labels and Free-Text Rationales", "REV: Information-Theoretic Evaluation of Free-Text Rationales", "Do Models Explain Themselves? Counterfactual Simulatability of Natural\n Language Explanations"], "answer_arxiv_id": ["2206.11083", "2305.07095", "2010.12762", "2210.04982", "2307.08678"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_15225"} +{"question": "What work introduced the concept of 'Algorithm-of-Thoughts' to handle reasoning pathways?", "answer": ["Algorithm of Thoughts: Enhancing Exploration of Ideas in Large Language\n Models"], "answer_arxiv_id": ["2308.10379"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_15226"} +{"question": "Could you provide me the paper that introduced the RDN model?", "answer": ["Residual Dense Network for Image Super-Resolution"], "answer_arxiv_id": ["1802.08797"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_15227"} +{"question": "Which research papers utilized image and text encoder in CLIP-Adapter?", "answer": ["CLIP-Adapter: Better Vision-Language Models with Feature Adapters"], "answer_arxiv_id": ["2110.04544"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_15228"} +{"question": "What works have attempted to mitigate the impact of compounding error in Learning from Observation?", "answer": ["Generative Adversarial Imitation from Observation"], "answer_arxiv_id": ["1807.06158"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_15229"} +{"question": "Are there any papers where neural network-based models are used for hand grasp generation?", "answer": ["Learning joint reconstruction of hands and manipulated objects", "GRAB: A Dataset of Whole-Body Human Grasping of Objects", "Hand-Object Contact Consistency Reasoning for Human Grasps Generation", "ContactPose: A Dataset of Grasps with Object Contact and Hand Pose", "A Skeleton-Driven Neural Occupancy Representation for Articulated Hands", "ContactGen: Generative Contact Modeling for Grasp Generation"], "answer_arxiv_id": ["1904.05767", "2008.11200", "2104.03304", "2007.09545", "2109.11399", "2310.03740"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_15230"} +{"question": "Could you provide me a list of papers that developed higher resolution heatmaps as an improvement to CAM?", "answer": ["Dilated Residual Networks", "F-CAM: Full Resolution Class Activation Maps via Guided Parametric Upscaling"], "answer_arxiv_id": ["1705.09914", "2109.07069v2"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_15231"} +{"question": "In which papers have quantum natural gradient been discussed?", "answer": ["Quantum Natural Gradient"], "answer_arxiv_id": ["1909.02108"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_15232"} +{"question": "Which studies adopt vector-based encoding schemes in trajectory prediction?", "answer": ["MultiPath++: Efficient Information Fusion and Trajectory Aggregation for\n Behavior Prediction", "Path-Aware Graph Attention for HD Maps in Motion Prediction", "ProphNet: Efficient Agent-Centric Motion Forecasting with\n Anchor-Informed Proposals", "ScenarioNet: Open-Source Platform for Large-Scale Traffic Scenario\n Simulation and Modeling", "TrafficGen: Learning to Generate Diverse and Realistic Traffic Scenarios"], "answer_arxiv_id": ["2111.14973", "2202.13772", "2303.12071", "2306.12241", "2210.06609"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_15233"} +{"question": "Could you name the research that does periodic resetting of current policy to enhance network plasticity in RL?", "answer": ["Transient Non-stationarity and Generalisation in Deep Reinforcement Learning"], "answer_arxiv_id": ["2006.05826"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_15234"} +{"question": "What is the work known for introducing a likelihood-based strategy in appearance modeling for tracking?", "answer": ["Unified Sequence-to-Sequence Learning for Single- and Multi-Modal Visual\n Object Tracking"], "answer_arxiv_id": ["2304.14394"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_15235"} +{"question": "What paper first investigated using the diffusion process for data distribution learning in generative tasks?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_15236"} +{"question": "What are the works discussing feature-based methods in social media bot-detection?", "answer": ["Scalable and Generalizable Social Bot Detection through Data Selection", "Deep Neural Networks for Bot Detection", "RTbust: Exploiting Temporal Patterns for Botnet Detection on Twitter", "SATAR: A Self-supervised Approach to Twitter Account Representation\n Learning and its Application in Bot Detection"], "answer_arxiv_id": ["1911.09179", "1802.04289", "1902.04506", "2106.13089"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_15237"} +{"question": "Is there any research on providing a better starting point for subsequent fine-tuning in multi-task settings?", "answer": ["Fusing finetuned models for better pretraining", "ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning"], "answer_arxiv_id": ["2204.03044", "2212.01378"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_15238"} +{"question": "What paper proposed capturing fine structures from the exemplar image through dense correspondence learning in the context of image translation?", "answer": ["Cross-domain Correspondence Learning for Exemplar-based Image Translation"], "answer_arxiv_id": ["2004.05571"], "source_meta": {"published_time": "20230802"}, "qid": "AutoScholarQuery_train_15239"} +{"question": "Which studies use the projection of the 3D point clouds into 2D grids as a segmentation method?", "answer": ["SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time\n Road-Object Segmentation from 3D LiDAR Point Cloud", "SqueezeSegV2: Improved Model Structure and Unsupervised Domain\n Adaptation for Road-Object Segmentation from a LiDAR Point Cloud", "SalsaNext: Fast, Uncertainty-aware Semantic Segmentation of LiDAR Point\n Clouds for Autonomous Driving", "RangeViT: Towards Vision Transformers for 3D Semantic Segmentation in\n Autonomous Driving", "Rethinking Range View Representation for LiDAR Segmentation"], "answer_arxiv_id": ["1710.07368", "1809.08495", "2003.03653", "2301.10222", "2303.05367"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_15240"} +{"question": "Which research studies have focused on generating longer videos?", "answer": ["NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generation", "Phenaki: Variable Length Video Generation From Open Domain Textual\n Description"], "answer_arxiv_id": ["2303.12346", "2210.02399"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15241"} +{"question": "In what papers did the researchers discuss using contrastive loss in training models?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["2002.05709", "1807.03748"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15242"} +{"question": "Could you provide me some works on interpretability for deep low-level networks?", "answer": ["Interpreting Super-Resolution Networks with Local Attribution Maps", "Finding Discriminative Filters for Specific Degradations in Blind Super-Resolution", "Rethinking Alignment in Video Super-Resolution Transformers"], "answer_arxiv_id": ["2011.11036", "2108.01070", "2207.08494"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15243"} +{"question": "What works have employed diffusion model based methods for image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2204.06125"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15244"} +{"question": "What work is related to Unlimiformer, encodes long inputs in chunks but has decoder attend to all inputs at the same time?", "answer": ["Efficient Long-Text Understanding with Short-Text Models"], "answer_arxiv_id": ["2208.00748"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_15245"} +{"question": "Which papers discuss the semidefinite programming (SDP) approach and its local version to provide accurate numerical bounds?", "answer": ["Efficient and Accurate Estimation of Lipschitz Constants for Deep Neural Networks"], "answer_arxiv_id": ["1906.04893"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_15246"} +{"question": "Can you tell me about the study that computes cascaded cost volumes with multi-head attention for refining?", "answer": ["GeoNeRF: Generalizing NeRF with Geometry Priors"], "answer_arxiv_id": ["2111.13539"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_15247"} +{"question": "Which research papers study task ambiguity?", "answer": ["Probabilistic Model-Agnostic Meta-Learning", "Active Learning Helps Pretrained Models Learn the Intended Task"], "answer_arxiv_id": ["1806.02817", "2204.08491"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_15248"} +{"question": "Could you reference studies that have used VLMs in image captioning?", "answer": ["mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal Skip-connections"], "answer_arxiv_id": ["2205.12005"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_15249"} +{"question": "What papers investigated the theoretical properties of EM for discrete LVMs?", "answer": ["On Convergence Properties of the Monte Carlo EM Algorithm"], "answer_arxiv_id": ["1206.4768"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_15250"} +{"question": "Which works discuss the challenges faced by early-stage datasets in providing nontemporal consistent images?", "answer": ["I Like to Move It: 6D Pose Estimation as an Action Decision Process", "HomebrewedDB: RGB-D Dataset for 6D Pose Estimation of 3D Objects", "Recovering 6D Object Pose and Predicting Next-Best-View in the Crowd", "A Dataset for Improved RGBD-based Object Detection and Pose Estimation\n for Warehouse Pick-and-Place"], "answer_arxiv_id": ["2009.12678", "1904.03167", "1512.07506", "1509.01277"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_15251"} +{"question": "What works discuss the topic of ensemble methods in the context of learning diversified set of functions?", "answer": ["Improving Adversarial Robustness via Promoting Ensemble Diversity", "Ridge Rider: Finding Diverse Solutions by Following Eigenvectors of the Hessian", "Learning Neural Network Subspaces", "DICE: Diversity in Deep Ensembles via Conditional Redundancy Adversarial Estimation"], "answer_arxiv_id": ["1901.08846", "2011.06505", "2102.10472", "2101.05544"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_15252"} +{"question": "Could you provide me some works that proposed a learning algorithm for dynamics by leveraging spectral learning?", "answer": ["Closing the Learning-Planning Loop with Predictive State Representations", "Hilbert Space Embeddings of Predictive State Representations"], "answer_arxiv_id": ["0912.2385v1", "1309.6819v1"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_15253"} +{"question": "Which papers provides information on Distributionally Robust Optimization (DRO) methods for deep learning?", "answer": ["Distributionally Robust Language Modeling", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1909.02060", "1911.08731"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_15254"} +{"question": "What papers tried to address the shape (point cloud) completion tasks directly on point cloud or have combined graph convolutions?", "answer": ["SnowflakeNet: Point Cloud Completion by Snowflake Point Deconvolution\n with Skip-Transformer", "PoinTr: Diverse Point Cloud Completion with Geometry-Aware Transformers", "Deformable Shape Completion with Graph Convolutional Autoencoders"], "answer_arxiv_id": ["2108.04444", "2108.08839", "1712.00268"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_15255"} +{"question": "Which papers implemented concatenation based conditioning first with the use of single global latent codes?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "Learning Implicit Fields for Generative Shape Modeling", "DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation"], "answer_arxiv_id": ["1812.03828", "1812.02822", "1901.05103"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_15256"} +{"question": "Could you mention which research papers propose ParaRel and LPAQA for MLMs?", "answer": ["Measuring and Improving Consistency in Pretrained Language Models", "How Can We Know What Language Models Know?"], "answer_arxiv_id": ["2102.01017", "1911.12543"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_15257"} +{"question": "Could you provide some works that introduced new methods for producing 3D position-aware features?", "answer": ["PETR: Position Embedding Transformation for Multi-View 3D Object\n Detection"], "answer_arxiv_id": ["2203.05625"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_15258"} +{"question": "Could you provide me some works that use special tokens for personalized concepts in image editing?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2208.01618", "2208.12242", "2212.04488"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_15259"} +{"question": "What works have been done in the field of RL for Visual Navigation tasks?", "answer": ["Object Goal Navigation using Goal-Oriented Semantic Exploration", "Visual Navigation with Spatial Attention", "What do navigation agents learn about their environment?"], "answer_arxiv_id": ["2007.00643", "2104.09807", "2206.08500"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_15260"} +{"question": "Could you provide me a research which introduces a stitch operator to create a graph directly by adding new transitions?", "answer": ["BATS: Best Action Trajectory Stitching"], "answer_arxiv_id": ["2204.12026"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_15261"} +{"question": "Any works about creating top-performing algorithms by choosing the base learners as well as the ensembling method?", "answer": ["XGBoost: A Scalable Tree Boosting System", "Deep Ensembles: A Loss Landscape Perspective", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1603.02754", "1912.02757", "1612.01474"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15262"} +{"question": "Which works discuss techniques to address the computational bottleneck in transformers, using sparsity-based methods?", "answer": ["Longformer: The Long-Document Transformer", "Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context"], "answer_arxiv_id": ["2004.05150", "1901.02860"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_15263"} +{"question": "Which research papers have discussed about the enforcement of affine constraints in CPWA NNs?", "answer": ["POLICE: Provably Optimal Linear Constraint Enforcement for Deep Neural Networks"], "answer_arxiv_id": ["2211.01340"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_15264"} +{"question": "Which works focused on vision-language alignment to enhance predictions in referring image segmentation?", "answer": ["Recurrent Multimodal Interaction for Referring Image Segmentation", "Referring Expression Object Segmentation with Caption-Aware Consistency", "Locate then Segment: A Strong Pipeline for Referring Image Segmentation", "Dynamic Multimodal Instance Segmentation guided by natural language\n queries", "CRIS: CLIP-Driven Referring Image Segmentation", "Multi-Modal Mutual Attention and Iterative Interaction for Referring\n Image Segmentation", "MDETR -- Modulated Detection for End-to-End Multi-Modal Understanding", "Vision-Language Transformer and Query Generation for Referring\n Segmentation"], "answer_arxiv_id": ["1703.07939", "1910.04748", "2103.16284", "1807.02257", "2111.15174", "2305.15302", "2104.12763", "2108.05565"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_15265"} +{"question": "Could you provide me some studies about the cross entropy soft risk optimization as proposed for improving sample efficiency in CVaR optimization?", "answer": ["Efficient Risk-Averse Reinforcement Learning"], "answer_arxiv_id": ["2205.05138"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_15266"} +{"question": "Which studies introduced the concept of federated learning, learning split across multiple devices and datasets?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data"], "answer_arxiv_id": ["1602.05629"], "source_meta": {"published_time": "20220911"}, "qid": "AutoScholarQuery_train_15267"} +{"question": "Could you list some works that focused on evaluating and mitigating hallucination in image captioning in MLLMs?", "answer": ["Evaluation and Analysis of Hallucination in Large Vision-Language Models", "OPERA: Alleviating Hallucination in Multi-Modal Large Language Models\n via Over-Trust Penalty and Retrospection-Allocation"], "answer_arxiv_id": ["2308.15126", "2311.17911"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_15268"} +{"question": "What work is it that represents MILP instances as bipartite graphs and uses graph neural networks to capture features for branching decisions?", "answer": ["Exact Combinatorial Optimization with Graph Convolutional Neural Networks"], "answer_arxiv_id": ["1906.01629"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_15269"} +{"question": "Could you list the works that used values predicted by a distinct model using Monte Carlo Tree Search for expanding a set of candidates for future steps?", "answer": ["Machine Translation Decoding beyond Beam Search", "Don't throw away your value model! Generating more preferable text with\n Value-Guided Monte-Carlo Tree Search decoding", "PPL-MCTS: Constrained Textual Generation Through Discriminator-Guided\n MCTS Decoding"], "answer_arxiv_id": ["2104.05336", "2309.15028", "2109.13582"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_15270"} +{"question": "Could you provide studies that use classifiers for guiding LMs towards generating desired textual attributes?", "answer": ["Fudge: Controlled Text Generation With Future Discriminators", "Learning to Write with Cooperative Discriminators", "Learning to Decode for Future Success"], "answer_arxiv_id": ["2104.05218", "1805.06087", "1701.06549"], "source_meta": {"published_time": "20220605"}, "qid": "AutoScholarQuery_train_15271"} +{"question": "Which works are about detecting 3D boxes directly from single images?", "answer": ["Learning Depth-Guided Convolutions for Monocular 3D Object Detection", "FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection", "Probabilistic and Geometric Depth: Detecting Objects in Perspective", "Objects are Different: Flexible Monocular 3D Object Detection"], "answer_arxiv_id": ["1912.04799", "2104.10956", "2107.14160", "2104.02323"], "source_meta": {"published_time": "20221117"}, "qid": "AutoScholarQuery_train_15272"} +{"question": "Which works demonstrated that implicit radiance fields can effectively learn scene representations and synthesize high-quality novel views?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "NeRF++: Analyzing and Improving Neural Radiance Fields"], "answer_arxiv_id": ["2003.08934", "2103.13415", "2010.07492"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_15273"} +{"question": "Could you provide examples of notable foundation models in language and vision-language domains?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "Florence: A New Foundation Model for Computer Vision", "Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks"], "answer_arxiv_id": ["1810.04805", "2005.14165", "2204.02311", "2111.11432", "2208.10442"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_15274"} +{"question": "Can you provide me with some studies that introduced sparse convolution based on sparse voxels to improve the performance of 3D object detection?", "answer": ["Generative Sparse Detection Networks for 3D Single-shot Object Detection", "FCAF3D: Fully Convolutional Anchor-Free 3D Object Detection"], "answer_arxiv_id": ["2006.12356", "2112.00322"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_15275"} +{"question": "What studies are about DDPM which models complex data distributions through discrete steps?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_15276"} +{"question": "Could you provide me some works about the methods in contrastive learning that propose diversity in representations?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["2006.07733"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_15277"} +{"question": "Which works handle rotations in point cloud data by converting to canonical poses using PCA?", "answer": ["Deep Positional and Relational Feature Learning for Rotation-Invariant\n Point Cloud Analysis", "Endowing Deep 3D Models with Rotation Invariance Based on Principal\n Component Analysis"], "answer_arxiv_id": ["2011.09080", "1910.08901"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_15278"} +{"question": "Which study presented the RevBiFPN, a fully reversible bidirectional feature pyramid network?", "answer": ["RevBiFPN: The Fully Reversible Bidirectional Feature Pyramid Network"], "answer_arxiv_id": ["2206.14098"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_15279"} +{"question": "What papers are about using vision transformers in conjunction with masked image modelling?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "Efficient Self-supervised Learning with Contextualized Target Representations for Vision, Speech and Language", "Masked Contrastive Representation Learning"], "answer_arxiv_id": ["2111.06377", "2212.07525", "2211.06012"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_15280"} +{"question": "Which studies initially proposed tokenizing speech signals into sequences of learnable, discrete units for textless NLP?", "answer": ["Text-Free Image-to-Speech Synthesis Using Learned Segmental Units", "Generative Spoken Language Modeling from Raw Audio", "Text-Free Prosody-Aware Generative Spoken Language Modeling", "Generative Spoken Dialogue Language Modeling"], "answer_arxiv_id": ["2012.15454", "2102.01192", "2109.03264", "2203.16502"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_15281"} +{"question": "What researches have applied 3D CNNs in action recognition?", "answer": ["Learning Spatiotemporal Features with 3D Convolutional Networks", "Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset", "A Closer Look at Spatiotemporal Convolutions for Action Recognition", "Non-local Neural Networks", "X3D: Expanding Architectures for Efficient Video Recognition"], "answer_arxiv_id": ["1412.0767", "1705.07750", "1711.11248", "1711.07971", "2004.04730"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15282"} +{"question": "What works present a comprehensive overview of the common stages of Few-Shot Learning (FSL)?", "answer": ["A Closer Look at Few-shot Classification"], "answer_arxiv_id": ["1904.04232"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_15283"} +{"question": "What research introduced negative samples for contrast in contrastive learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1911.05722", "2002.05709"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_15284"} +{"question": "What works use energies such as As-Rigid-As-Possible (ARAP), Laplacian, or edge contraction to restrict the decoder in shape deformation?", "answer": ["3D-CODED : 3D Correspondences by Deep Deformation"], "answer_arxiv_id": ["1806.05228"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_15285"} +{"question": "What paper proposes generating images in a pyramidal fashion with additional reconstruction guidance?", "answer": ["Pyramidal Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2208.01864"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_15286"} +{"question": "Which work carried out the test of closed models on faithfulness using modified table-to-text generation datasets?", "answer": ["Investigating Table-to-Text Generation Capabilities of LLMs in\n Real-World Information Seeking Scenarios"], "answer_arxiv_id": ["2305.14987"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_15287"} +{"question": "Can you point me to the work discussing the first oblivious sketch for logistic regression?", "answer": ["Oblivious sketching for logistic regression"], "answer_arxiv_id": ["2107.06615"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_15288"} +{"question": "Which studies used self-information for token pruning in black-box compression?", "answer": ["Compressing Context to Enhance Inference Efficiency of Large Language\n Models"], "answer_arxiv_id": ["2310.06201"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_15289"} +{"question": "Which benchmark recognized and addressed the issue of lacking out-of-distribution queries in existing benchmarks?", "answer": ["Google Landmarks Dataset v2 A Large-Scale Benchmark for Instance-Level Recognition and Retrieval"], "answer_arxiv_id": ["2004.01804"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_15290"} +{"question": "Which studies have introduced SLAM system into human pose estimation to reconstruct the 4D human pose?", "answer": ["4D Human Body Capture from Egocentric Video via 3D Scene Grounding", "BodySLAM: Joint Camera Localisation, Mapping, and Human Motion Tracking", "The One Where They Reconstructed 3D Humans and Environments in TV Shows", "Decoupling Human and Camera Motion from Videos in the Wild"], "answer_arxiv_id": ["2011.13341", "2205.02301", "2207.14279", "2302.12827"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15291"} +{"question": "Which papers have applied GAN modeling for action-conditioned motion generation?", "answer": ["Deep Video Generation, Prediction and Completion of Human Action\n Sequences", "Learning Diverse Stochastic Human-Action Generators by Learning Smooth\n Latent Transitions"], "answer_arxiv_id": ["1711.08682", "1912.10150"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15292"} +{"question": "Could you provide me some works related to onboarding humans to work with AI models?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "“Why is ‘Chicago’ deceptive?” Towards Building Model-Driven Tutorials for Humans", "Teaching Humans When To Defer to a Classifier via Exemplars", "Improving Human-AI Collaboration with Descriptions of AI Behavior", "Training Towards Critical Use: Learning to Situate AI Predictions Relative to Human Knowledge"], "answer_arxiv_id": ["1602.04938", "2001.05871", "2111.11297", "2301.06937", "2308.15700"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_15293"} +{"question": "What works are used in PAGED design to demonstrate the opportunities and gaps of LLMs in procedural graphs extraction task?", "answer": ["Scaling Instruction-Finetuned Language Models", "Training language models to follow instructions with human feedback", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2210.11416", "2203.02155", "2307.09288"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_15294"} +{"question": "Which papers suggest learning the optimal prediction model parameter to minimize the task loss?", "answer": ["Task-based End-to-end Model Learning in Stochastic Optimization", "Few-shot Object Detection via Feature Reweighting", "Melding the Data-Decisions Pipeline: Decision-Focused Learning for Combinatorial Optimization", "Control-Oriented Model-Based Reinforcement Learning with Implicit Differentiation", "Value Gradient weighted Model-Based Reinforcement Learning", "Control-Oriented Model-Based Reinforcement Learning with Implicit Differentiation", "Leaving the Nest : Going Beyond Local Loss Functions for Predict-Then-Optimize"], "answer_arxiv_id": ["1703.04529", "1812.01866", "1809.05504", "2106.03273", "2204.01464", "2106.03273", "2305.16830"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_15295"} +{"question": "What research works have explored length extrapolation for transformer-based models?", "answer": ["Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation", "SHAPE: Shifted Absolute Position Embedding for Transformers", "Exploring Length Generalization in Large Language Models"], "answer_arxiv_id": ["2108.12409", "2109.05644", "2207.04901"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_15296"} +{"question": "What research papers deal with hierarchical methods for the generation of long videos?", "answer": ["Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive\n Transformer", "Flexible Diffusion Modeling of Long Videos", "NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generation"], "answer_arxiv_id": ["2204.03638", "2205.11495", "2303.12346"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_15297"} +{"question": "Which papers have worked on learning from multiple unlabeled datasets?", "answer": ["On the Minimal Supervision for Training Any Binary Classifier from Only Unlabeled Data", "Binary Classification from Multiple Unlabeled Datasets via Surrogate Set Classification"], "answer_arxiv_id": ["1808.10585", "2102.00678"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_15298"} +{"question": "What work uses the classifier-guided fashion of conditioning in a conditional diffusion model?", "answer": ["Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2102.09672"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15299"} +{"question": "Which works improve the representation ability and surface reconstruction by introducing more flexible and deformable primitives?", "answer": ["CvxNet: Learnable Convex Decomposition", "3DIAS: 3D Shape Reconstruction with Implicit Algebraic Surfaces", "Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networks"], "answer_arxiv_id": ["1909.05736", "2108.08653", "2103.10429v1"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_15300"} +{"question": "What studies have explored image retrieval for LLMs?", "answer": ["Grounding Language Models to Images for Multimodal Inputs and Outputs"], "answer_arxiv_id": ["2301.13823"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_15301"} +{"question": "Which papers addressed label distribution skew in federated learning using knowledge distillation?", "answer": ["Federated Mutual Learning", "FedMD: Heterogenous Federated Learning via Model Distillation", "Ensemble Distillation for Robust Model Fusion in Federated Learning", "Parameterized Knowledge Transfer for Personalized Federated Learning", "The Best of Both Worlds: Accurate Global and Personalized Models through Federated Learning with Data-Free Hyper-Knowledge Distillation", "CD2-pFed: Cyclic Distillation-guided Channel Decoupling for Model Personalization in Federated Learning"], "answer_arxiv_id": ["2006.16765", "1910.03581", "2006.07242", "2111.02862", "2301.08968", "2204.03880"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_15302"} +{"question": "What papers consider the non-episodic setting where the system can be continuously monitored?", "answer": ["Feedback Linearization based on Gaussian Processes with event-triggered Online Learning"], "answer_arxiv_id": ["1911.06565"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15303"} +{"question": "What works use backpropagation of a target class score to the input image for gradient visualization methods?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps"], "answer_arxiv_id": ["1312.6034"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_15304"} +{"question": "Which works discuss DP-trained generative models?", "answer": ["Differentially Private Diffusion Models Generate Useful Synthetic Images"], "answer_arxiv_id": ["2302.13861"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_15305"} +{"question": "Could you mention the works that applied the multi-fidelity idea in hyperparameter optimization (HPO) for automated machine learning (autoML)?", "answer": ["Non-stochastic Best Arm Identification and Hyperparameter Optimization", "Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization", "BOHB: Robust and Efficient Hyperparameter Optimization at Scale", "A System for Massively Parallel Hyperparameter Tuning"], "answer_arxiv_id": ["1502.07943v1", "1603.06560", "1807.01774", "1810.05934v5"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_15306"} +{"question": "What work proposed the method of obtaining discriminative representations by enlarging the receptive field for shallow graph embeddings based clustering methods?", "answer": ["Attributed Graph Clustering via Adaptive Graph Convolution"], "answer_arxiv_id": ["1906.01210"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_15307"} +{"question": "Which work first introduced event cameras to the semantic segmentation task using deep learning?", "answer": ["EV-SegNet: Semantic Segmentation for Event-based Cameras"], "answer_arxiv_id": ["1811.12039"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_15308"} +{"question": "Could you provide me some studies on transformation-based approaches such as real-valued non-volume preserving transformations (NVP) and Fourier Flows (FF) for time series data?", "answer": ["Density estimation using Real NVP"], "answer_arxiv_id": ["1605.08803"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_15309"} +{"question": "Which papers proposed vision-language instruction tuning without a pretrained LLM?", "answer": ["MultiInstruct: Improving Multi-Modal Zero-Shot Learning via Instruction Tuning"], "answer_arxiv_id": ["2212.10773"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_15310"} +{"question": "Could you provide me some works about the use of pretrained models as feature extractors?", "answer": ["Evaluating (and improving) the correspondence between deep neural networks and human representations", "Transforming Neural Network Visual Representations to Predict Human Judgments of Similarity"], "answer_arxiv_id": ["1706.02417", "2010.06512"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_15311"} +{"question": "Could you name the studies where cross-attention was used for attending between different sets of activations?", "answer": ["Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["2107.07651", "1810.04805"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_15312"} +{"question": "Are there any studies that show human-AI teams do not perform better than the maximum performance of the human or AI alone even with AI explanations?", "answer": ["Does the Whole Exceed its Parts? The Effect of AI Explanations on Complementary Team Performance", "Understanding the Effect of Out-of-distribution Examples and Interactive Explanations on Human-AI Decision Making"], "answer_arxiv_id": ["2006.14779", "2101.05303"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_15313"} +{"question": "Could you provide some examples of GAN-based methods for NeRF editing?", "answer": ["HoloGAN: Unsupervised learning of 3D representations from natural images", "BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled\n Images", "Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image\n Synthesis", "Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["1904.01326", "2002.08988", "1610.07584", "2110.08985", "2112.07945"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15314"} +{"question": "Any works that confirms the validity of using MAUVE?", "answer": ["On the Usefulness of Embeddings, Clusters and Strings for Text Generator\n Evaluation"], "answer_arxiv_id": ["2205.16001"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_15315"} +{"question": "What works first revealed that the issue of robust fairness occurs in conventional adversarial training?", "answer": ["To be Robust or to be Fair: Towards Fairness in Adversarial Training"], "answer_arxiv_id": ["2010.06121"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_15316"} +{"question": "What papers provide insights on the sample complexity of reinforcement learning in POMDPs?", "answer": ["PAC Reinforcement Learning with Rich Observations"], "answer_arxiv_id": ["1602.02722"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_15317"} +{"question": "What are some researches that studied the sensitivity of neural network models particularly in regards to adversarial attacks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Intriguing properties of neural networks", "Adversarial Examples for Evaluating Reading Comprehension Systems"], "answer_arxiv_id": ["1412.6572", "1312.6199", "1707.07328"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_15318"} +{"question": "Which study introduced the notion of components separation via a gradient descent in signal space with indirect constraints?", "answer": ["A new approach for the statistical denoising of Planck interstellar dust polarization data"], "answer_arxiv_id": ["2102.03160v2"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_15319"} +{"question": "Can you cite studies that made contributions to the area of 3D Occupancy Prediction through dense prediction methods?", "answer": ["Joint 2D-3D-Semantic Data for Indoor Scene Understanding", "Semantic Scene Completion from a Single Depth Image", "ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "KITTI-360: A Novel Dataset and Benchmarks for Urban Scene Understanding\n in 2D and 3D", "MonoScene: Monocular 3D Semantic Scene Completion", "VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene\n Completion", "OpenOccupancy: A Large Scale Benchmark for Surrounding Semantic\n Occupancy Perception", "Occ3D: A Large-Scale 3D Occupancy Prediction Benchmark for Autonomous\n Driving", "Scene as Occupancy"], "answer_arxiv_id": ["1702.01105", "1611.08974", "1702.04405", "2109.13410", "2112.00726", "2302.12251", "2303.03991", "2304.14365", "2306.02851"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_15320"} +{"question": "Which papers propose novel datasets for GUI-related tasks?", "answer": ["Wireframe-Based UI Design Search Through Image Autoencoder", "Unblind Your Apps: Predicting Natural-Language Labels for Mobile GUI Components by Deep Learning", "Screen2Words: Automatic Mobile UI Summarization with Multimodal Learning"], "answer_arxiv_id": ["2103.07085", "2003.00380", "2108.03353"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_15321"} +{"question": "What study focused on LLMs fine-tuned with an auxiliary dataset enriched with PII?", "answer": ["Analyzing Leakage of Personally Identifiable Information in Language Models"], "answer_arxiv_id": ["2302.00539"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_15322"} +{"question": "Can you name some research that have used transformer architectures for ERP images to model long-range dependencies?", "answer": ["PanoFormer: Panorama Transformer for Indoor 360 Depth Estimation", "EGformer: Equirectangular Geometry-biased Transformer for 360 Depth\n Estimation"], "answer_arxiv_id": ["2203.09283", "2304.07803"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_15323"} +{"question": "Which studies are based on modifying the texture of the 3D mesh alongside MeshAdv?", "answer": ["Adversarial Attacks Beyond the Image Space", "FCA: Learning a 3D Full-coverage Vehicle Camouflage for Multi-view\n Physical Adversarial Attack", "DTA: Physical Camouflage Attacks using Differentiable Transformation\n Network", "ACTIVE: Towards Highly Transferable 3D Physical Camouflage for Universal\n and Robust Vehicle Evasion"], "answer_arxiv_id": ["1711.07183", "2109.07193", "2203.09831", "2308.07009"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_15324"} +{"question": "What research has been conducted on language-conditioned imitation learning that would potentially benefit from Vision-and-Language Transformer (VTL)?", "answer": ["Language-Conditioned Imitation Learning for Robot Manipulation Tasks", "Language Conditioned Imitation Learning over Unstructured Data", "What Matters in Language Conditioned Robotic Imitation Learning over Unstructured Data"], "answer_arxiv_id": ["2010.12083", "2005.07648", "2204.06252"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15325"} +{"question": "Could you tell me which studies provided results for modern Rectified Linear Units (ReLU) networks including FNNs and Convolutional Neural Networks(CNNs)?", "answer": ["Identity Matters in Deep Learning", "Optimization Landscape and Expressivity of Deep CNNs"], "answer_arxiv_id": ["1611.04231", "1710.10928"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15326"} +{"question": "Which study in Off-Policy RL estimates the stationary state distributions of both the current policy and the mixed buffer policy?", "answer": ["Off-Policy Policy Gradient Algorithms by Constraining the State Distribution Shift"], "answer_arxiv_id": ["1911.06970"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_15327"} +{"question": "Which works have proposed to adapt an intrinsic reward with mutual information (MI) for skill discovery, similarly to diayn?", "answer": ["Variational Intrinsic Control"], "answer_arxiv_id": ["1611.07507"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_15328"} +{"question": "Which study pioneered the coarse-to-fine network approach for end-to-end monocular depth estimation?", "answer": ["Depth Map Prediction from a Single Image using a Multi-Scale Deep\n Network"], "answer_arxiv_id": ["1406.2283"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_15329"} +{"question": "Which studies present the use of machine learning in guiding combinatorial algorithms?", "answer": ["Learning Combinatorial Optimization Algorithms over Graphs", "Attention, Learn to Solve Routing Problems!", "Reinforcement Learning for Solving the Vehicle Routing Problem", "AutoShard: Automated Embedding Table Sharding for Recommender Systems", "DreamShard: Generalizable Embedding Table Placement for Recommender Systems"], "answer_arxiv_id": ["1704.01665", "1803.08475", "1802.04240", "2208.06399", "2210.02023"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_15330"} +{"question": "Could you provide me some studies that suggest gradient-based features perform well on visual downstream tasks?", "answer": ["Gradients as Features for Deep Representation Learning", "Fast Adaptation with Linearized Neural Networks", "LQF: Linear Quadratic Fine-Tuning"], "answer_arxiv_id": ["2004.05529", "2103.01439", "2012.11140"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15331"} +{"question": "What works have shown the relation between robustness, logit margins and Lipschitz constants?", "answer": ["Parseval Networks: Improving Robustness to Adversarial Examples", "Lipschitz-Margin Training: Scalable Certification of Perturbation Invariance for Deep Neural Networks", "Towards Certifying L-infinity Robustness using Neural Networks with L-inf-dist Neurons"], "answer_arxiv_id": ["1704.08847", "1802.04034", "2102.05363v4"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_15332"} +{"question": "What are the notable papers that discuss Arbitrary-Scale Image Generation using MLPs in CIPS and INR-GAN?", "answer": ["Image Generators with Conditionally-Independent Pixel Synthesis", "Adversarial Generation of Continuous Images"], "answer_arxiv_id": ["2011.13775", "2011.12026"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_15333"} +{"question": "Which work focuses on conducting semantic image synthesis at the feature map level?", "answer": ["Linear Semantics in Generative Adversarial Networks"], "answer_arxiv_id": ["2104.00487"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_15334"} +{"question": "Can you mention some papers that present methods for mitigating bias in machine learning models?", "answer": ["Equality of Opportunity in Supervised Learning", "Counterfactual Fairness"], "answer_arxiv_id": ["1610.02413", "1703.06856"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_15335"} +{"question": "What works are about the application of diffusion in semantic image synthesis?", "answer": ["Semantic Image Synthesis via Diffusion Models"], "answer_arxiv_id": ["2207.00050"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_15336"} +{"question": "Could you provide me some studies that examined the methods of intelligently selecting subsets of examples for evaluation?", "answer": ["Comparing Test Sets with Item Response Theory"], "answer_arxiv_id": ["2106.00840"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_15337"} +{"question": "Could you give me an example of a study that extends a multi-task diffusion model to metric depth prediction?", "answer": ["Monocular Depth Estimation using Diffusion Models"], "answer_arxiv_id": ["2302.14816"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15338"} +{"question": "What references were cited when discussing non-IID data causing weight divergence and performance drop in federated learning?", "answer": ["On the Convergence of FedAvg on Non-IID Data"], "answer_arxiv_id": ["1907.02189"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_15339"} +{"question": "What papers propose an extension of recalibration along with theoretical guarantees?", "answer": ["Modular Conformal Calibration"], "answer_arxiv_id": ["2206.11468"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_15340"} +{"question": "Could you provide me some works on semantic segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation", "Pyramid Scene Parsing Network", "Per-Pixel Classification is Not All You Need for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038", "1612.01105", "2107.06278"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_15341"} +{"question": "Which papers are related to instance-wise AUC optimization?", "answer": ["Stochastic AUC Maximization with Deep Neural Networks"], "answer_arxiv_id": ["1908.10831"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_15342"} +{"question": "What papers are about VT2T modalities, specifically in generating video captions?", "answer": ["Coherent Multi-Sentence Video Description with Variable Level of Detail", "How2: A Large-scale Dataset for Multimodal Language Understanding", "An overview on the evaluated video retrieval tasks at TRECVID 2022"], "answer_arxiv_id": ["1403.6173", "1811.00347", "2306.13118"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_15343"} +{"question": "What studies select examples in an online fashion by taking the top k examples in a minibatch according to excess loss?", "answer": ["Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt"], "answer_arxiv_id": ["2206.07137"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_15344"} +{"question": "What works propose shape exaggeration blocks for additional control while using StyleGAN for caricature synthesis?", "answer": ["StyleCariGAN: Caricature Generation via StyleGAN Feature Map Modulation"], "answer_arxiv_id": ["2107.04331"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15345"} +{"question": "Which work introduced a retrieval augmentation based on preceding tokens into language models?", "answer": ["REALM: Retrieval-Augmented Language Model Pre-Training"], "answer_arxiv_id": ["2002.08909"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_15346"} +{"question": "What recent works delved into improving the detailed aspects of image comprehension?", "answer": ["Kosmos-2: Grounding Multimodal Large Language Models to the World", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks"], "answer_arxiv_id": ["2306.14824", "2305.11175"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15347"} +{"question": "Which paper presents deep learning for set structured data?", "answer": ["Deep Sets"], "answer_arxiv_id": ["1703.06114"], "source_meta": {"published_time": "20220826"}, "qid": "AutoScholarQuery_train_15348"} +{"question": "What research first proposed a data-driven method with pixel-aligned features for recovering 3D human shapes, garments, and textures?", "answer": ["PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization"], "answer_arxiv_id": ["1905.05172"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_15349"} +{"question": "What are the works that propose to regularize the learned policy to stay close to the behavior policy for addressing the distribution shift in offline RL?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Behavior Regularized Offline Reinforcement Learning"], "answer_arxiv_id": ["1812.02900", "1906.00949", "1911.11361"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_15350"} +{"question": "Can you provide an example of a work that investigates the robustness of contrastive-SSL methods against extreme spurious correlation?", "answer": ["Can contrastive learning avoid shortcut solutions?"], "answer_arxiv_id": ["2106.11230"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_15351"} +{"question": "Could you provide me some studies that explained GNN predictions on static graphs using mutual information?", "answer": ["GNNExplainer: Generating Explanations for Graph Neural Networks"], "answer_arxiv_id": ["1903.03894"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_15352"} +{"question": "What works have used curiosity-based intrinsic motivation to solve sparse-reward reinforcement learning problems?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Curiosity-driven Exploration by Self-supervised Prediction", "Exploration by Random Network Distillation"], "answer_arxiv_id": ["1507.00814", "1705.05363", "1810.12894"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_15353"} +{"question": "Which works use massively parallel simulation for quadruped locomotion?", "answer": ["Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learning", "Rapid Locomotion via Reinforcement Learning"], "answer_arxiv_id": ["2109.11978", "2205.02824"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_15354"} +{"question": "Which work indicates that the Gradient Descent method achieves the optimal complexity for non-convex settings?", "answer": ["Lower Bounds for Finding Stationary Points I"], "answer_arxiv_id": ["1710.11606"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_15355"} +{"question": "Is there any research examining the use of atomic norms for the recovery of simultaneously sparse and low-rank matrices?", "answer": ["Tight convex relaxations for sparse matrix factorization"], "answer_arxiv_id": ["1407.5158"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_15356"} +{"question": "What papers focus on efficient architecture design for 3D HPE?", "answer": ["A simple yet effective baseline for 3d human pose estimation", "3D human pose estimation in video with temporal convolutions and\n semi-supervised training", "XNect: Real-time Multi-Person 3D Motion Capture with a Single RGB Camera"], "answer_arxiv_id": ["1705.03098", "1811.11742", "1907.00837"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_15357"} +{"question": "Could you provide me some studies about inductive biases and reasoning principles acquired by sequential predictors?", "answer": ["Language Models are Few-Shot Learners", "RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "Learning to reinforcement learn"], "answer_arxiv_id": ["2005.14165", "1611.02779", "1611.05763"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_15358"} +{"question": "What prior works have leveraged feature maps to conduct local editing tasks like enlarging the eyes or changing the shape of ears without manual annotations?", "answer": ["PandA: Unsupervised Learning of Parts and Appearances in the Feature Maps of GANs"], "answer_arxiv_id": ["2206.00048"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_15359"} +{"question": "Which studies contend that Deep Sets remain universal if a hidden layer as large as the input set is required?", "answer": ["Universal Approximation of Functions on Sets"], "answer_arxiv_id": ["2107.01959"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_15360"} +{"question": "What are the studies that proposed hard-parameter sharing in Multi-Task Learning (MTL)?", "answer": ["UberNet : Training a ‘Universal’ Convolutional Neural Network for Low-, Mid-, and High-Level Vision using Diverse Datasets and Limited Memory", "Mask R-CNN"], "answer_arxiv_id": ["1609.02132", "1703.06870"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15361"} +{"question": "What works introduced prompt learning in the realm of computer vision?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "Language Models as Knowledge Bases?", "Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "PLOT: Prompt Learning with Optimal Transport for Vision-Language Models"], "answer_arxiv_id": ["2107.13586v1", "1909.01066", "2109.01134", "2203.05557", "2210.01253"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_15362"} +{"question": "Which papers discussed the use of diffusion models for text-to-image generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2205.11487", "2204.06125", "2112.10752"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_15363"} +{"question": "Which paper was the research direction of prompt-based learning significantly inspired by?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_15364"} +{"question": "Which works studied how language models generalize across linguistically related constructions, implying an implicit task hierarchy?", "answer": ["Language Modelling as a Multi-Task Problem"], "answer_arxiv_id": ["2101.11287"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_15365"} +{"question": "What studies suggest a memory-efficient approach in the context of multi-scale supervision?", "answer": ["MINER: Multiscale Implicit Neural Representations"], "answer_arxiv_id": ["2202.03532"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_15366"} +{"question": "Which work first introduced graph hypernetworks for architecture search in image classification?", "answer": ["Parameter Prediction for Unseen Deep Architectures"], "answer_arxiv_id": ["2110.13100"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15367"} +{"question": "Are there any studies examining what contributes to robustness during the downstream task performance?", "answer": ["Fine-Tuning can Distort Pretrained Features and Underperform\n Out-of-Distribution"], "answer_arxiv_id": ["2202.10054"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_15368"} +{"question": "Which paper is about recursive correction that can avoid the challenges faced by static correction methods?", "answer": ["Bias-Variance Reduced Local SGD for Less Heterogeneous Federated Learning"], "answer_arxiv_id": ["2102.03198"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_15369"} +{"question": "What is the research paper that reveals the potential of using synthetic data generated from text-to-image models to train visual representations?", "answer": ["StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners"], "answer_arxiv_id": ["2306.00984"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_15370"} +{"question": "Which study utilized a hierarchical design and a shifted-window strategy to imitate a CNN-based model?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2103.14030"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_15371"} +{"question": "Who proposed the robust adversarial training algorithm?", "answer": ["Robust Adversarial Reinforcement Learning"], "answer_arxiv_id": ["1703.02702"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_15372"} +{"question": "Which work extends NeRF with an additional branch for encoding 3D semantic labels?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene\n Representation"], "answer_arxiv_id": ["2103.15875"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_15373"} +{"question": "Could you provide the paper that demonstrated the relatedness of Lipschitz constant and fully connected layers?", "answer": ["Towards Fast Computation of Certified Robustness for ReLU Networks"], "answer_arxiv_id": ["1804.09699"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_15374"} +{"question": "Could you recommend some works that specifically address the issue of false positives and false negatives in self-supervised learning?", "answer": ["A Closer Look at Weakly-Supervised Audio-Visual Source Localization", "Visual Sound Localization in the Wild by Cross-Modal Interference Erasing", "Robust Audio-Visual Instance Discrimination", "Audio-Visual Instance Discrimination with Cross-Modal Agreement"], "answer_arxiv_id": ["2209.09634", "2202.06406", "2103.15916", "2004.12943"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_15375"} +{"question": "Can you provide some studies that deal with combining an SDF-based 3D representation with a style-based 2D generator?", "answer": ["StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation"], "answer_arxiv_id": ["2112.11427"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_15376"} +{"question": "Which works are based on contrastive learning in the field of self-supervised representation learning from visual data?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1911.05722", "2003.04297", "2002.05709", "1807.03748"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_15377"} +{"question": "What paper proposed the Decision-Estimation Coefficient (DEC) in relation to sample-efficient interactive learning?", "answer": ["The Statistical Complexity of Interactive Decision Making"], "answer_arxiv_id": ["2112.13487v3"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_15378"} +{"question": "Can you list some of the works that have focused on enhancing the rendering speed of neural implicit representations?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "FastNeRF: High-Fidelity Neural Rendering at 200FPS"], "answer_arxiv_id": ["2201.05989", "2103.10380"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_15379"} +{"question": "What early studies attempt to model temporal dependencies in recurrent-free models by using 3D convolutional networks?", "answer": ["Video Frame Synthesis using Deep Voxel Flow", "FutureGAN: Anticipating the Future Frames of Video Sequences using Spatio-Temporal 3d Convolutions in Progressively Growing GANs"], "answer_arxiv_id": ["1702.02463", "1810.01325"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_15380"} +{"question": "Could you provide me some research texts that constructed views through augmentations to single-modal or multi-modal data in context of contrastive representation learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "An Empirical Study of Training Self-Supervised Vision Transformers", "Learning Transferable Visual Models From Natural Language Supervision", "Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2104.02057", "2103.00020", "2112.03857"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_15381"} +{"question": "Can you quote for me the work that uses fewer patches obtained by splitting images during training?", "answer": ["Accelerating Vision Transformer Training via a Patch Sampling Schedule"], "answer_arxiv_id": ["2208.09520"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_15382"} +{"question": "Which are the studies that proposed a debiased dataset using human labor?", "answer": ["Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering", "Rel3D: A Minimally Contrastive Benchmark for Grounding Spatial Relations in 3D"], "answer_arxiv_id": ["1612.00837", "2012.01634v1"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_15383"} +{"question": "What research works developed LiDAR panoptic methods based on well-designed semantic segmentation networks?", "answer": ["Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentation", "GP-S3Net: Graph-based Panoptic Sparse Semantic Segmentation Network", "LiDAR-based Panoptic Segmentation via Dynamic Shifting Network"], "answer_arxiv_id": ["2103.14962", "2108.08401", "2011.11964"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_15384"} +{"question": "Which studies introduced high quality and realistic image synthesis models, such as DALL-E and Cogview?", "answer": ["Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers"], "answer_arxiv_id": ["2102.12092", "2105.13290"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_15385"} +{"question": "Are there any studies known to utilize vector quantization to move from lossy to lossless compression in federated learning?", "answer": ["UVeQFed: Universal Vector Quantization for Federated Learning"], "answer_arxiv_id": ["2006.03262"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15386"} +{"question": "What are some of the works related to long-form video understanding?", "answer": ["Towards Long-Form Video Understanding", "Long-Form Video-Language Pre-Training with Multimodal Temporal\n Contrastive Learning", "HierVL: Learning Hierarchical Video-Language Embeddings", "MIST: Multi-modal Iterative Spatial-Temporal Transformer for Long-form\n Video Question Answering", "MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient\n Long-Term Video Recognition"], "answer_arxiv_id": ["2106.11310", "2210.06031", "2301.02311", "2212.09522", "2201.08383"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_15387"} +{"question": "What studies introduced classifier-guided diffusion?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_15388"} +{"question": "Which work proposed to use coordinate based networks for modeling full head geometries?", "answer": ["i3DMM: Deep Implicit 3D Morphable Model of Human Heads"], "answer_arxiv_id": ["2011.14143"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_15389"} +{"question": "Which works introduced the first BAI algorithms under the assumption of bounded support?", "answer": ["lil’ UCB : An Optimal Exploration Algorithm for Multi-Armed Bandits"], "answer_arxiv_id": ["1312.7308"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15390"} +{"question": "Who used an INN to learn a bijective transformation between a word surface and its morphemes?", "answer": ["Two Birds with One Stone: Investigating Invertible Neural Networks for\n Inverse Problems in Morphology"], "answer_arxiv_id": ["1912.05274"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_15391"} +{"question": "Which benchmark included reasoning tasks like visual QA?", "answer": ["IGLUE: A Benchmark for Transfer Learning across Modalities, Tasks, and\n Languages"], "answer_arxiv_id": ["2201.11732"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_15392"} +{"question": "Could you list any works covered the study of learning in games?", "answer": ["Near-Optimal No-Regret Learning in General Games", "Hardness of Independent Learning and Sparse Equilibrium Computation in Markov Games", "How and Why to Manipulate Your Own Agent: On the Incentives of Users of Learning Agents", "Fast Convergence of Regularized Learning in Games"], "answer_arxiv_id": ["2108.06924", "2303.12287", "2112.07640", "1507.00407"], "source_meta": {"published_time": "20230709"}, "qid": "AutoScholarQuery_train_15393"} +{"question": "What paper discusses the concept of Personalized Federated Learning (PFL)?", "answer": ["Survey of Personalization Techniques for Federated Learning"], "answer_arxiv_id": ["2003.08673"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_15394"} +{"question": "Which works analyze the use of CVaR in policy gradient in reinforcement learning?", "answer": ["Optimizing the CVaR via Sampling", "Learning Robust Options by Conditional Value at Risk Optimization", "Efficient Risk-Averse Reinforcement Learning"], "answer_arxiv_id": ["1404.3862", "1905.09191", "2205.05138"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_15395"} +{"question": "What papers used retrieved documents for supplementary context in generative tasks?", "answer": ["Contextualized Representations Using Textual Encyclopedic Knowledge", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2004.12006", "2005.11401"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_15396"} +{"question": "Who first proposed diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_15397"} +{"question": "In which studies is the issue of over-squashing in MPNNs discussed?", "answer": ["On the Bottleneck of Graph Neural Networks and its Practical Implications", "Understanding over-squashing and bottlenecks on graphs via curvature"], "answer_arxiv_id": ["2006.05205", "2111.14522"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_15398"} +{"question": "Which papers elaborate on the application of transfer learning across different environments?", "answer": ["Adapting Deep Visuomotor Representations with Weak Pairwise Constraints"], "answer_arxiv_id": ["1511.07111"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_15399"} +{"question": "Can you cite the studies that used pseudo labels based on class-dependent confidence thresholds and energy score thresholds?", "answer": ["InPL: Pseudo-labeling the Inliers First for Imbalanced Semi-supervised\n Learning"], "answer_arxiv_id": ["2303.07269"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_15400"} +{"question": "Could you mention some studies that have proposed machine learning solutions for the BPTP and related problems?", "answer": ["Reinforcement learning of rare diffusive dynamics", "Transferable neural networks for enhanced sampling of protein dynamics"], "answer_arxiv_id": ["2105.04321", "1801.00636v1"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_15401"} +{"question": "What paper studied the efficiency of extrapolation methods on signals that exhibit an intensive stationary component?", "answer": ["Any-resolution Training for High-resolution Image Synthesis"], "answer_arxiv_id": ["2204.07156"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_15402"} +{"question": "Are there any works that discussed the embedding of part-whole hierarchies in capsule networks?", "answer": ["Dynamic Routing Between Capsules"], "answer_arxiv_id": ["1710.09829"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_15403"} +{"question": "Could you provide me some studies about methods that consider partial correlations between uncertain transition probabilities?", "answer": ["A Bayesian Approach to Robust Reinforcement Learning", "Robust Markov Decision Process: Beyond Rectangularity", "Lightning Does Not Strike Twice: Robust MDPs with Coupled Uncertainty"], "answer_arxiv_id": ["1905.08188", "1811.00215", "1206.4643"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_15404"} +{"question": "What are the studies that adopt a CLIP-based contrastive loss and score distillation sampling to supervise the optimization of editing NeRF?", "answer": ["NeRF-Art: Text-Driven Neural Radiance Fields Stylization", "DreamEditor: Text-Driven 3D Scene Editing with Neural Fields"], "answer_arxiv_id": ["2212.08070", "2306.13455"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_15405"} +{"question": "What papers refer to the initial idea of subgraph GNNs?", "answer": ["Reconstruction for Powerful Graph Representations", "DropGNN: Random Dropouts Increase the Expressiveness of Graph Neural Networks"], "answer_arxiv_id": ["2110.00577", "2111.06283"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_15406"} +{"question": "Are there any works that modified the discriminator in GANs to improve stability?", "answer": ["Generative Multi-Adversarial Networks", "PacGAN: The power of two samples in generative adversarial networks", "Energy-based Generative Adversarial Network"], "answer_arxiv_id": ["1611.01673", "1712.04086", "1609.03126v4"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_15407"} +{"question": "Can you mention a study that uses an active learning approach to identify unknown parameters and aims to minimize the Euclidean distance in the parameter space?", "answer": ["Active Learning for Nonlinear System Identification with Guarantees"], "answer_arxiv_id": ["2006.10277"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_15408"} +{"question": "What studies have explored the composition of energy-based models and diffusion models?", "answer": ["Compositional Visual Generation and Inference with Energy Based Models", "Learning to Compose Visual Relations", "Compositional Visual Generation with Composable Diffusion Models", "Reduce, Reuse, Recycle: Compositional Generation with Energy-Based\n Diffusion Models and MCMC", "Compositional Sculpting of Iterative Generative Processes", "Implicit Generation and Generalization in Energy-Based Models"], "answer_arxiv_id": ["2004.06030", "2111.09297", "2206.01714", "2302.11552", "2309.16115", "1903.08689"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15409"} +{"question": "Are there any works that incorporate a memory mechanism on an MViT backbone to model temporal interactions?", "answer": ["MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient\n Long-Term Video Recognition"], "answer_arxiv_id": ["2201.08383"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_15410"} +{"question": "Which research provides theoretical explanations for the observation that modern deep learning methods can perfectly fit the training data while still performing well on test data?", "answer": ["Understanding deep learning requires rethinking generalization"], "answer_arxiv_id": ["1611.03530v2"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15411"} +{"question": "Could you provide me a research paper which designed an encoder-decoder architecture totally based on multi-head self-attention for dehazing?", "answer": ["Vision Transformers for Single Image Dehazing", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2204.03883", "2103.14030"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_15412"} +{"question": "Which paper explored a combination of weight decay and learning rate schedule to improve performance in adversarial training?", "answer": ["Self-ensemble Adversarial Training for Improved Robustness"], "answer_arxiv_id": ["2203.09678"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15413"} +{"question": "Can you cite some works on Test-Time Adaptation (TTA) methods that focus on improving the model's performance especially when the test data distribution differs from the training data distribution?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "Test-Time Training with Masked Autoencoders", "Tent: Fully Test-Time Adaptation by Entropy Minimization", "MEMO: Test Time Robustness via Adaptation and Augmentation", "Test-Time Training Can Close the Natural Distribution Shift Performance Gap in Deep Learning Based Compressed Sensing", "Evaluating the Adversarial Robustness of Adaptive Test-time Defenses", "Test time Adaptation through Perturbation Robustness", "Continual Test-Time Domain Adaptation"], "answer_arxiv_id": ["1909.13231", "2209.07522", "2006.10726", "2110.09506", "2204.07204v2", "2202.13711", "2110.10232", "2203.13591"], "source_meta": {"published_time": "20240531"}, "qid": "AutoScholarQuery_train_15414"} +{"question": "Can you name some papers where hyperedge features are considered and a message-passing framework is used, interpreted as GNNs applied to the star expansion graph?", "answer": ["Be More with Less: Hypergraph Attention Networks for Inductive Text Classification", "HNHN: Hypergraph Networks with Hyperedge Neurons", "HyperSAGE: Generalizing Inductive Representation Learning on Hypergraphs", "UniGNN: a Unified Framework for Graph and Hypergraph Neural Networks", "Learning over Families of Sets - Hypergraph Representation Learning for Higher Order Tasks", "You are AllSet: A Multiset Function Framework for Hypergraph Neural Networks", "Message Passing Neural Networks for Hypergraphs", "A Hypergraph Neural Network Framework for Learning Hyperedge-Dependent Node Embeddings", "HEAT: Hyperedge Attention Networks"], "answer_arxiv_id": ["2011.00387", "2006.12278", "2010.04558", "2105.00956", "2101.07773", "2106.13264", "2203.16995v2", "2212.14077", "2201.12113"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_15415"} +{"question": "Can you mention the researches that have conducted an evaluation of the zero-shot prompting capability of LLMs on text-to-SQL?", "answer": ["Evaluating the Text-to-SQL Capabilities of Large Language Models", "A comprehensive evaluation of ChatGPT’s zero-shot Text-to-SQL capability"], "answer_arxiv_id": ["2204.00498", "2303.13547"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_15416"} +{"question": "What works have proposed model-based methods for learning control policies that provide reachability, safety and reach-avoid specifications?", "answer": ["Formal Synthesis of Lyapunov Neural Networks", "Stabilizing Neural Control Using Self-Learned Almost Lyapunov Critics", "Neural Lyapunov Control", "Learning Certified Control Using Contraction Metric", "AMYTISS: Parallelized Automated Controller Synthesis for Large-Scale Stochastic Systems", "A Barrier Function Approach to Finite-Time Stochastic System Verification and Control", "Safety Guarantees for Neural Network Dynamic Systems via Stochastic Barrier Functions", "Safety Certification for Stochastic Systems via Neural Barrier Functions"], "answer_arxiv_id": ["2003.08910", "2107.04989", "2005.00611", "2011.12569", "2005.06191", "1909.05109", "2206.07811", "2206.01463"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_15417"} +{"question": "Are there any researches that propose using an adaptive step size for NPG?", "answer": ["On the Linear convergence of Natural Policy Gradient Algorithm"], "answer_arxiv_id": ["2105.01424v1"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_15418"} +{"question": "Which studies used witnesses and percolation as valuable features for graph matching?", "answer": ["An efficient reconciliation algorithm for social networks"], "answer_arxiv_id": ["1307.1690"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_15419"} +{"question": "What works discuss the possibility to transfer the edit of one image to another?", "answer": ["EditGAN: High-Precision Semantic Image Editing", "Rewriting a Deep Generative Model"], "answer_arxiv_id": ["2111.03186", "2007.15646"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_15420"} +{"question": "Could you list studies in the literature that introduced the concept and application of unrolled/unfolded networks?", "answer": ["Theoretical Linear Convergence of Unfolded ISTA and its Practical Weights and Thresholds", "Learning step sizes for unfolded sparse coding"], "answer_arxiv_id": ["1808.10038", "1905.11071"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_15421"} +{"question": "Which works developed deep models for learning effective features for dense correspondence?", "answer": ["D2-Net: A Trainable CNN for Joint Description and Detection of Local Features", "LF-Net: Learning Local Features from Images", "SuperGlue: Learning Feature Matching with Graph Neural Networks", "DISK: Learning local features with policy gradient", "LIFT: Learned Invariant Feature Transform"], "answer_arxiv_id": ["1905.03561", "1805.09662", "1911.11763", "2006.13566", "1603.09114"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15422"} +{"question": "Could you show examples of studies that explored the idea of approximating equivariance while learning from data?", "answer": ["Learning to Convolve: A Generalized Weight-Tying Approach", "Variational Autoencoder with Learned Latent Structure", "Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding", "Group Equivariant Generative Adversarial Networks", "Topographic VAEs learn Equivariant Capsules"], "answer_arxiv_id": ["1905.04663", "2006.10597v2", "2007.10930", "2005.01683", "2109.01394"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_15423"} +{"question": "Could you tell me about works that applied Siamese-based trackers in deep tracking methods?", "answer": ["Fully-Convolutional Siamese Networks for Object Tracking", "Siam R-CNN: Visual Tracking by Re-Detection", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1606.09549", "1911.12836", "2002.05709"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_15424"} +{"question": "Could you provide me some works about dynamic pruning of Transformers during inference time?", "answer": ["Reducing Transformer Depth on Demand with Structured Dropout", "DynaBERT: Dynamic BERT with Adaptive Width and Depth", "DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference", "BERT Loses Patience: Fast and Robust Inference with Early Exit"], "answer_arxiv_id": ["1909.11556", "2004.04037", "2004.12993", "2006.04152"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_15425"} +{"question": "Which papers have identified trustworthiness as a key challenge in developing automated systems?", "answer": ["The Next Decade in AI: Four Steps Towards Robust Artificial Intelligence", "Perspectives on the State and Future of Deep Learning - 2023"], "answer_arxiv_id": ["2002.06177v3", "2312.09323"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_15426"} +{"question": "Which papers discuss audio-visual understanding tasks?", "answer": ["Active Contrastive Learning of Audio-Visual Video Representations", "Cyclic Co-Learning of Sounding Object Visual Grounding and Sound Separation"], "answer_arxiv_id": ["2009.09805", "2104.02026"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_15427"} +{"question": "Which works proposed versions of NeRF that address the issue of unbounded scenes?", "answer": ["NeRF++: Analyzing and Improving Neural Radiance Fields", "F$^{2}$-NeRF: Fast Neural Radiance Field Training with Free Camera\n Trajectories", "MERF: Memory-Efficient Radiance Fields for Real-time View Synthesis in\n Unbounded Scenes", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields"], "answer_arxiv_id": ["2010.07492", "2303.15951", "2302.12249", "2111.12077"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_15428"} +{"question": "Could you provide any research that has focused on Inverse Reinforcement Learning from Observation?", "answer": ["MobILE: Model-Based Imitation Learning From Observation Alone", "Imitation from Observation: Learning to Imitate Behaviors from Raw Video via Context Translation", "Generative Adversarial Imitation from Observation", "Off-Policy Imitation Learning from Observations", "Provably Efficient Imitation Learning from Observation Alone"], "answer_arxiv_id": ["2102.10769", "1707.03374", "1807.06158", "2102.13185", "1905.10948"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_15429"} +{"question": "Any studies about audio captioning?", "answer": ["Audio Captioning Transformer"], "answer_arxiv_id": ["2107.09817"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_15430"} +{"question": "Which paper is about a classifier-guided conditional diffusion model where the pre-trained reward model is used as a form of classifier?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15431"} +{"question": "Which papers propose primal-dual algorithms for CMDPs that achieve square root of K bound on regrets and constraint violations?", "answer": ["Exploration-Exploitation in Constrained MDPs", "Provably Efficient Safe Exploration via Primal-Dual Policy Optimization"], "answer_arxiv_id": ["2003.02189", "2003.00534"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_15432"} +{"question": "What studies have built simulator environments for low-cost experimentation in reinforcement learning and robotics?", "answer": ["OpenAI Gym", "DeepMind Control Suite", "Lyceum: An efficient and scalable ecosystem for robot learning", "Unity: A General Platform for Intelligent Agents", "Brax - A Differentiable Physics Engine for Large Scale Rigid Body Simulation"], "answer_arxiv_id": ["1606.01540", "1801.00690", "2001.07343v1", "1809.02627", "2106.13281"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_15433"} +{"question": "What papers have proposed or discuss the use of a robust min-max extension of Multi-task Learning (MTL) and Reinforcement Learning (RL)?", "answer": ["Minimax Multi-Task Learning and a Generalized Loss-Compositional Paradigm for MTL", "Task-Robust Model-Agnostic Meta-Learning"], "answer_arxiv_id": ["1209.2784", "2002.04766"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_15434"} +{"question": "What research papers have studied fairness when predictive labels are missing?", "answer": ["Fairness-aware Model-agnostic Positive and Unlabeled Learning", "Can I Trust My Fairness Metric? Assessing Fairness with Unlabeled Data and Bayesian Inference"], "answer_arxiv_id": ["2206.09346", "2010.09851"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15435"} +{"question": "What papers focus on learning atomic forces and predict energy based on numeric integration?", "answer": ["Machine Learning of Accurate Energy-Conserving Molecular Force Fields", "Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields", "Accurate global machine learning force fields for molecules with hundreds of atoms", "Machine learning force fields: Construction, validation, and outlook"], "answer_arxiv_id": ["1611.04678", "1802.09238", "2209.14865", "1610.02098"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_15436"} +{"question": "Which papers have adopted Wasserstein distance at the pixel level in the context of multi-class segmentation?", "answer": ["Generalised Wasserstein Dice Score for Imbalanced Multi-class Segmentation using Holistic Convolutional Networks", "Generalized Wasserstein Dice Score, Distributionally Robust Deep Learning, and Ranger for brain tumor segmentation: BraTS 2020 challenge"], "answer_arxiv_id": ["1707.00478", "2011.01614"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_15437"} +{"question": "Which studies have focused on 3D object detection from multi-view images for outdoor scenes by projecting the features to Bird’s Eye View?", "answer": ["BEVDet: High-performance Multi-camera 3D Object Detection in\n Bird-Eye-View", "BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers", "FrustumFormer: Adaptive Instance-aware Resampling for Multi-view 3D\n Detection", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets"], "answer_arxiv_id": ["2112.11790", "2203.17270", "2301.04467", "1812.05784", "2301.06051"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_15438"} +{"question": "Which works addressed the issue of the weak interaction of bi-encodings in retrieval systems?", "answer": ["Modularized Transfomer-based Ranking Framework", "Efficient Document Re-Ranking for Transformers by Precomputing Term\n Representations", "Poly-encoders: Transformer Architectures and Pre-training Strategies for\n Fast and Accurate Multi-sentence Scoring"], "answer_arxiv_id": ["2004.13313", "2004.14255", "1905.01969"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_15439"} +{"question": "Could you provide examples of studies examining graph neural networks (GNNs) and graph convolutional networks (GCNs)?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks"], "answer_arxiv_id": ["1609.02907"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_15440"} +{"question": "What is the reference of the study that presents new capabilities of language models such as in-context learning when these models scale up?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_15441"} +{"question": "What papers used the GAIL method for environment model learning?", "answer": ["Virtual-Taobao: Virtualizing Real-world Online Retail Environment for Reinforcement Learning", "Generative Adversarial User Model for Reinforcement Learning Based Recommendation System"], "answer_arxiv_id": ["1805.10000", "1812.10613"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_15442"} +{"question": "Could you provide me some works where researchers addressed the learning of OT maps through saddle point optimization problems?", "answer": ["Generative Modeling with Optimal Transport Maps", "Neural Optimal Transport with General Cost Functionals", "An Optimal Transport Perspective on Unpaired Image Super-Resolution", "Entropic Neural Optimal Transport via Diffusion Processes"], "answer_arxiv_id": ["2110.02999", "2205.15403", "2202.01116", "2211.01156"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15443"} +{"question": "What works propose an adaptive collaboration of flows and enhanced deformable separable convolution for intermediate frame synthesis by filtering operations?", "answer": ["AdaCoF: Adaptive Collaboration of Flows for Video Frame Interpolation", "Multiple Video Frame Interpolation via Enhanced Deformable Separable\n Convolution"], "answer_arxiv_id": ["1907.10244", "2006.08070"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_15444"} +{"question": "What works considered learning lower-bounded Q values in RL?", "answer": ["Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "Efficient Diffusion Policies for Offline Reinforcement Learning"], "answer_arxiv_id": ["2110.01548", "2208.06193", "2305.20081"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_15445"} +{"question": "Which papers employed local features from a 2D feature map to improve the detail of the predicted shapes and generalization to unseen categories?", "answer": ["Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images", "DISN: Deep Implicit Surface Network for High-quality Single-view 3D\n Reconstruction", "Fostering Generalization in Single-view 3D Reconstruction by Learning a\n Hierarchy of Local and Global Shape Priors"], "answer_arxiv_id": ["1804.01654", "1905.10711", "2104.00476"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_15446"} +{"question": "Are there any papers that utilize Signed Distance Functions (SDF) to recover garment meshes from images and segmentation masks?", "answer": ["SMPLicit: Topology-aware Generative Model for Clothed People", "DIG: Draping Implicit Garment over the Human Body", "3D Clothed Human Reconstruction in the Wild"], "answer_arxiv_id": ["2103.06871", "2209.10845", "2207.10053"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_15447"} +{"question": "What studies have proposed adding noises to input data as a method of enhancing robustness?", "answer": ["AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty"], "answer_arxiv_id": ["1912.02781"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_15448"} +{"question": "Which research works relay the critic-guided approach to reuse previously learned policies?", "answer": ["CUP: Critic-Guided Policy Reuse"], "answer_arxiv_id": ["2210.08153"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_15449"} +{"question": "Which paper studied the MDPs with linear approximation and unknown feature vectors and parameters?", "answer": ["FLAMBE: Structural Complexity and Representation Learning of Low Rank MDPs"], "answer_arxiv_id": ["2006.10814"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_15450"} +{"question": "Could you mention a work that provides similar guarantees for linear MDPs in the realm of instance-dependent algorithms?", "answer": ["Instance-Dependent Near-Optimal Policy Identification in Linear MDPs via Online Experiment Design"], "answer_arxiv_id": ["2207.02575"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_15451"} +{"question": "Could you provide me some works about architecture specific distillation or model compression?", "answer": ["Understanding BERT Rankers Under Distillation"], "answer_arxiv_id": ["2007.11088"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_15452"} +{"question": "What papers are about the sample complexity of Q-learning under synchronous sampling?", "answer": ["Is Q-Learning Minimax Optimal? A Tight Sample Complexity Analysis"], "answer_arxiv_id": ["2102.06548"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_15453"} +{"question": "Which work introduced a control-aware metric (CAPO) for motion prediction?", "answer": ["Control-Aware Prediction Objectives for Autonomous Driving"], "answer_arxiv_id": ["2204.13319"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_15454"} +{"question": "What paper propose the multi-step GDA method with a certain iteration complexity?", "answer": ["Solving a Class of Non-Convex Min-Max Games Using Iterative First Order Methods"], "answer_arxiv_id": ["1902.08297"], "source_meta": {"published_time": "20221226"}, "qid": "AutoScholarQuery_train_15455"} +{"question": "What work globally ranks examples to identify examples in a single failure mode?", "answer": ["Distilling Model Failures as Directions in Latent Space"], "answer_arxiv_id": ["2206.14754"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15456"} +{"question": "Can you name some human video datasets widely used for benchmarking self-supervised video representation learning?", "answer": ["The Kinetics Human Action Video Dataset", "UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild"], "answer_arxiv_id": ["1705.06950", "1212.0402"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_15457"} +{"question": "What papers have utilized a balanced sampling strategy to handle online and offline data?", "answer": ["Overcoming Exploration in Reinforcement Learning with Demonstrations", "MoDem: Accelerating Visual Model-Based Reinforcement Learning with Demonstrations", "Policy Expansion for Bridging Offline-to-Online Reinforcement Learning"], "answer_arxiv_id": ["1709.10089", "2212.05698", "2302.00935"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_15458"} +{"question": "Are there any research papers about using Transformer architecture for tracking?", "answer": ["Attention Is All You Need", "Transformer Tracking", "Learning Spatio-Temporal Transformer for Visual Tracking", "Transforming Model Prediction for Tracking", "Correlation-Aware Deep Tracking"], "answer_arxiv_id": ["1706.03762", "2103.15436", "2103.17154", "2203.11192", "2203.01666"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_15459"} +{"question": "What are some of the studies related to domain adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks"], "answer_arxiv_id": ["1505.07818"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_15460"} +{"question": "Which work is aimed to accelerate video modelling by applying a hierarchical grid sampling schedule and is most closely related to multigrid training?", "answer": ["A Multigrid Method for Efficiently Training Video Models"], "answer_arxiv_id": ["1912.00998"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_15461"} +{"question": "Which studies focus on developing benchmarks for specific knowledge types and task families in the context of LLMs?", "answer": ["Measuring Massive Multitask Language Understanding", "AGIEval: A Human-Centric Benchmark for Evaluating Foundation Models", "Evaluating Large Language Models Trained on Code", "Training Verifiers to Solve Math Word Problems"], "answer_arxiv_id": ["2009.03300", "2304.06364", "2107.03374", "2110.14168"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15462"} +{"question": "Could you provide me some of the recent advancements that demonstrate the success of simpler model architectures when subjected to extensive large-scale pretraining?", "answer": ["PaLI: A Jointly-Scaled Multilingual Language-Image Model", "Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction\n Tuning", "Emu: Generative Pretraining in Multimodality"], "answer_arxiv_id": ["2209.06794", "2309.02591", "2307.05222"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_15463"} +{"question": "What research has been done on the use of LLMs in multi-hop question answering or knowledge probing?", "answer": ["Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning", "Crepe: Open-Domain Question Answering with False Presuppositions", "An Empirical Study on Few-shot Knowledge Probing for Pretrained Language Models"], "answer_arxiv_id": ["2205.09712", "2211.17257", "2109.02772"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_15464"} +{"question": "What works examine bounding the cumulative dynamic regret?", "answer": ["Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient"], "answer_arxiv_id": ["1605.04638"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_15465"} +{"question": "Could you provide me studies that proposed new variance reduction methods and achieved optimal complexity?", "answer": ["Spider: Near-Optimal Non-Convex Optimization via Stochastic Path Integrated Differential Estimator", "Hybrid Stochastic Gradient Descent Algorithms for Stochastic Nonconvex Optimization", "An Optimal Hybrid Variance-Reduced Algorithm for Stochastic Composite Nonconvex Optimization", "PAGE: A Simple and Optimal Probabilistic Gradient Estimator for Nonconvex Optimization", "Momentum-Based Variance Reduction in Non-Convex SGD"], "answer_arxiv_id": ["1807.01695", "1905.05920", "2008.09055", "2008.10898", "1905.10018"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_15466"} +{"question": "Which papers discuss approaches for video domain adaptation that aim to transfer knowledge from a label-sufficient source domain?", "answer": ["Temporal Attentive Alignment for Large-Scale Video Domain Adaptation"], "answer_arxiv_id": ["1907.12743"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_15467"} +{"question": "Which paper introduces a Lagrange multiplier in the Monge formulation for the choice of neural OT architectures?", "answer": ["Scalable Computations of Wasserstein Barycenter via Input Convex Neural Networks"], "answer_arxiv_id": ["2007.04462"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_15468"} +{"question": "What studies have used Autoencoders for achieving disentangled representations?", "answer": ["Disentangling Factors of Variation with Cycle-Consistent Variational Auto-Encoders", "Learning to Factorize and Relight a City", "Swapping Autoencoder for Deep Image Manipulation", "Adversarial Latent Autoencoders"], "answer_arxiv_id": ["1804.10469", "2008.02796v1", "2007.00653", "2004.04467"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_15469"} +{"question": "Could you provide me a reference that has used likelihood as a regularization in the training objectives for generative models?", "answer": ["Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models"], "answer_arxiv_id": ["1705.08868"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_15470"} +{"question": "Could you provide me some studies about 3D mesh generation?", "answer": ["SpiralNet++: A Fast and Highly Efficient Mesh Convolution Operator", "DiffusionNet: Discretization Agnostic Learning on Surfaces", "Primal-Dual Mesh Convolutional Neural Networks", "Subdivision-Based Mesh Convolution Networks", "A Simple Approach to Intrinsic Correspondence Learning on Unstructured\n 3D Meshes", "MeshCNN: A Network with an Edge", "HodgeNet: Learning Spectral Geometry on Triangle Meshes", "Scan2Mesh: From Unstructured Range Scans to 3D Meshes", "BSP-Net: Generating Compact Meshes via Binary Space Partitioning", "PolyGen: An Autoregressive Generative Model of 3D Meshes"], "answer_arxiv_id": ["1911.05856", "2012.00888v3", "2010.12455", "2106.02285", "1809.06664", "1809.05910", "2104.12826", "1811.10464", "1911.06971", "2002.10880"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_15471"} +{"question": "Could you provide me some works addressing the issue of spatial heterogeneity in urban events forecasting?", "answer": ["HintNet: Hierarchical Knowledge Transfer Networks for Traffic Accident Forecasting on Heterogeneous Spatio-Temporal Data"], "answer_arxiv_id": ["2203.03100"], "source_meta": {"published_time": "20230930"}, "qid": "AutoScholarQuery_train_15472"} +{"question": "Could you provide me some studies that have developed learned sim agents with different assumptions?", "answer": ["SpAGNN: Spatially-Aware Graph Neural Networks for Relational Behavior Forecasting from Sensor Data", "Implicit Latent Variable Model for Scene-Consistent Motion Forecasting", "Narrowing the coordinate-frame gap in behavior prediction models: Distillation for efficient and accurate scene-centric motion forecasting", "TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors", "Multi-Agent Imitation Learning for Driving Simulation", "Symphony: Learning Realistic and Diverse Agents for Autonomous Driving Simulation", "BITS: Bi-level Imitation for Traffic Simulation"], "answer_arxiv_id": ["1910.08233", "2007.12036", "2206.03970", "2101.06557", "1803.01044v1", "2205.03195", "2208.12403"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_15473"} +{"question": "What studies use a confounded MDP model and handle the memoryless unobserved confounding through optimal balancing or proxy variables?", "answer": ["Off-policy Evaluation in Infinite-Horizon Reinforcement Learning with Latent Confounders", "Instrumental Variable Value Iteration for Causal Offline Reinforcement Learning", "Provably Efficient Causal Reinforcement Learning with Confounded Observational Data", "Off-Policy Confidence Interval Estimation with Confounded Markov Decision Process", "Offline Reinforcement Learning with Instrumental Variables in Confounded Markov Decision Processes"], "answer_arxiv_id": ["2007.13893", "2102.09907", "2006.12311", "2202.10589", "2209.08666"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_15474"} +{"question": "Are there studies that emphasize the socially situated nature of fairness and harms in India?", "answer": ["Re-imagining Algorithmic Fairness in India and Beyond"], "answer_arxiv_id": ["2101.09995"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_15475"} +{"question": "What studies proposed efficient popular techniques like sparse attention patterns for Transformer models?", "answer": ["Image Transformer", "Longformer: The Long-Document Transformer", "Reformer: The Efficient Transformer"], "answer_arxiv_id": ["1802.05751", "2004.05150", "2001.04451"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_15476"} +{"question": "Can you provide me some works about sender and receiver structures in draw & guess game for emergent communication?", "answer": ["Learning to Draw: Emergent Communication through Sketching"], "answer_arxiv_id": ["2106.02067"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_15477"} +{"question": "Can you specify some comprehensive reviews of Normalizing Flows and their different architectures?", "answer": ["Normalizing Flows for Probabilistic Modeling and Inference", "Normalizing Flows: An Introduction and Review of Current Methods"], "answer_arxiv_id": ["1912.02762", "1908.09257"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_15478"} +{"question": "What study uses a mixed INT8/FP16 decomposition to solve the activation outliers problem in LLM. int8?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale"], "answer_arxiv_id": ["2208.07339"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_15479"} +{"question": "What papers illustrate examples of white-box scenarios in embedding inversion attacks, where the attacker has access to the full model weights?", "answer": ["InvBERT: Reconstructing Text from Contextualized Word Embeddings by\n inverting the BERT pipeline"], "answer_arxiv_id": ["2109.10104"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_15480"} +{"question": "What works introduce equivariance to the self-attention mechanism?", "answer": ["SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks"], "answer_arxiv_id": ["2006.10503"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_15481"} +{"question": "What works developed approximation methods for Shapley values to lessen the computation burden?", "answer": ["A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1705.07874"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_15482"} +{"question": "Can you name the research work that enhanced roadside perception performance by fusing the height and depth representation?", "answer": ["BEVHeight++: Toward Robust Visual Centric 3D Object Detection"], "answer_arxiv_id": ["2309.16179"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_15483"} +{"question": "What research works have made contributions in improving efficiency and sampling speed for diffusion models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Denoising Diffusion Implicit Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "Gotta Go Fast When Generating Data with Score-Based Models", "Learning to Efficiently Sample from Diffusion Probabilistic Models", "On Fast Sampling of Diffusion Probabilistic Models", "Progressive Distillation for Fast Sampling of Diffusion Models", "On Distillation of Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Diffusion Autoencoders: Toward a Meaningful and Decodable Representation", "DiffuseVAE: Efficient, Controllable and High-Fidelity Generation from Low-Dimensional Latents"], "answer_arxiv_id": ["2102.09672", "2010.02502", "2010.02502", "2201.06503", "2105.14080", "2106.03802", "2106.00132", "2202.00512", "2210.03142", "2112.10752", "2111.15640", "2201.00308"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15484"} +{"question": "Could you provide me some works that used Neural Processes for modeling stochastic processes?", "answer": ["Conditional Neural Processes", "Attentive Neural Processes", "The Functional Neural Process", "Sequential Neural Processes"], "answer_arxiv_id": ["1807.01613", "1901.05761", "1906.08324", "1906.10264"], "source_meta": {"published_time": "20230507"}, "qid": "AutoScholarQuery_train_15485"} +{"question": "Can you provide studies about feature-based methods for visual SLAM?", "answer": ["Ultimate SLAM? Combining Events, Images, and IMU for Robust Visual SLAM\n in HDR and High Speed Scenarios", "PL-EVIO: Robust Monocular Event-based Visual Inertial Odometry with\n Point and Line Features"], "answer_arxiv_id": ["1709.06310", "2209.12160"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_15486"} +{"question": "Which paper discussed Symplectic ODE-Net (SymODEN) and how it makes the learned dynamics symplectic by construction?", "answer": ["Symplectic ODE-Net: Learning Hamiltonian Dynamics with Control"], "answer_arxiv_id": ["1909.12077"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_15487"} +{"question": "What studies have been done on unsupervised learning for random graph models?", "answer": ["Asymptotic analysis of the stochastic block model for modular networks and its algorithmic applications", "Community detection thresholds and the weak Ramanujan property", "A Proof Of The Block Model Threshold Conjecture", "Consistency Thresholds for the Planted Bisection Model", "Exact Recovery in the Stochastic Block Model", "Non-backtracking spectrum of random graphs: community detection and non-regular Ramanujan graphs", "Semidefinite Programs on Sparse Random Graphs and their Application to Community Detection", "Information-theoretic thresholds for community detection in sparse networks", "Optimizing Generalized PageRank Methods for Seed-Expansion Community Detection", "Block Models and Personalized PageRank", "Exact Community Recovery in Correlated Stochastic Block Models"], "answer_arxiv_id": ["1109.3041", "1311.3085", "1311.4115", "1407.1591v5", "1405.3267", "1501.06087", "1504.05910", "1607.01760", "1905.10881", "1607.03483", "2203.15736v1"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_train_15488"} +{"question": "Which papers talk about the use of node or edge dropout as a method of graph rewiring?", "answer": ["DropEdge: Towards Deep Graph Convolutional Networks on Node Classification", "DropGNN: Random Dropouts Increase the Expressiveness of Graph Neural Networks"], "answer_arxiv_id": ["1907.10903", "2111.06283"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_15489"} +{"question": "Which papers discuss channel pruning as a technique for network compression?", "answer": ["Pruning Convolutional Neural Networks for Resource Efficient Inference", "meProp: Sparsified Back Propagation for Accelerated Deep Learning with\n Reduced Overfitting", "Discrimination-aware Channel Pruning for Deep Neural Networks", "Discrimination-aware Network Pruning for Deep Model Compression", "Wavelet Knowledge Distillation: Towards Efficient Image-to-Image\n Translation"], "answer_arxiv_id": ["1611.06440", "1706.06197", "1810.11809", "2001.01050", "2203.06321"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_15490"} +{"question": "Are there any biomedical datasets used for prompt-based learning and evaluation of few and zero-shot classification?", "answer": ["Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks", "In-BoXBART: Get Instructions into Biomedical Multi-Task Learning"], "answer_arxiv_id": ["2204.07705", "2204.07600"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_15491"} +{"question": "Could you list the studies that applied methods, other than linear programming, for finding NE?", "answer": ["Exploration-Exploitation in Multi-Agent Competition: Convergence with Bounded Rationality", "On Last-Iterate Convergence Beyond Zero-Sum Games", "Asynchronous Gradient Play in Zero-Sum Multi-agent Games"], "answer_arxiv_id": ["2106.12928", "2203.12056", "2211.08980v1"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_15492"} +{"question": "What papers focus on contrastive learning in self-supervised learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Self-Supervised Visual Representation Learning with Semantic Grouping", "Patch-level Representation Learning for Self-supervised Vision Transformers", "Unsupervised Feature Learning via Non-Parametric Instance Discrimination"], "answer_arxiv_id": ["1911.05722", "2002.05709", "2006.07733", "2203.03884", "2006.09882", "2205.15288", "2206.07990", "1805.01978"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_15493"} +{"question": "What work introduces a practical episodic model-based algorithm in continuous time?", "answer": ["Continuous-Time Model-Based Reinforcement Learning"], "answer_arxiv_id": ["2102.04764"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15494"} +{"question": "Could you provide me some papers that developed efficient frameworks to enhance the speed of NeRFs?", "answer": ["ActiveRMAP: Radiance Field for Active Mapping And Planning"], "answer_arxiv_id": ["2211.12656"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_15495"} +{"question": "What researches provide datasets capturing humans in complex environments?", "answer": ["Novel View Synthesis of Dynamic Scenes with Globally Coherent Depths from a Monocular Camera", "Neural 3D Video Synthesis from Multi-view Video"], "answer_arxiv_id": ["2004.01294", "2103.02597"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_15496"} +{"question": "Which research paper introduces the LVS method in sound source localization (SSL)?", "answer": ["Localizing Visual Sounds the Hard Way"], "answer_arxiv_id": ["2104.02691"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_15497"} +{"question": "Which papers developed techniques for reducing attribution noise?", "answer": ["Guided Integrated Gradients: an Adaptive Path Method for Removing Noise", "IDGI: A Framework to Eliminate Explanation Noise from Integrated Gradients"], "answer_arxiv_id": ["2106.09788", "2303.14242"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_15498"} +{"question": "What papers have started to show promise for solving inverse problems using score-based generative models?", "answer": ["Robust Compressed Sensing MRI with Deep Generative Priors", "Solving Inverse Problems in Medical Imaging with Score-Based Generative Models"], "answer_arxiv_id": ["2108.01368", "2111.08005"], "source_meta": {"published_time": "20211207"}, "qid": "AutoScholarQuery_train_15499"} +{"question": "Can you provide some works that applied dropout for compressing networks?", "answer": ["Variational Dropout Sparsifies Deep Neural Networks", "Learning Sparse Networks Using Targeted Dropout"], "answer_arxiv_id": ["1701.05369", "1905.13678"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_15500"} +{"question": "What works cover the pretrain-then-finetune paradigm in the context of auxiliary learning?", "answer": ["What makes ImageNet good for transfer learning?", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "wav2vec: Unsupervised Pre-training for Speech Recognition", "Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks"], "answer_arxiv_id": ["1608.08614", "1810.04805", "1904.05862", "2004.10964"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_15501"} +{"question": "Which study introduced Last Layer Ensemble (LLE) to mitigate multiple shortcuts?", "answer": ["A Whac-A-Mole Dilemma: Shortcuts Come in Multiples Where Mitigating One\n Amplifies Others"], "answer_arxiv_id": ["2212.04825"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_15502"} +{"question": "Which paper first introduced the MIG as a detector for disentanglement?", "answer": ["Isolating Sources of Disentanglement in Variational Autoencoders"], "answer_arxiv_id": ["1802.04942"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_15503"} +{"question": "What papers propose the use of (structured) sparsity as a criterion to identify and remove redundant structures in a DNN?", "answer": ["Learning Structured Sparsity in Deep Neural Networks", "Group Sparsity: The Hinge Between Filter Pruning and Decomposition for Network Compression", "Neural Network Compression via Sparse Optimization", "Highly Efficient Salient Object Detection with 100K Parameters", "Pruning Filter in Filter", "DeepHoyer: Learning Sparser Neural Network with Differentiable Scale-Invariant Sparsity Measures"], "answer_arxiv_id": ["1608.03665", "2003.08935", "2011.04868", "2003.05643", "2009.14410", "1908.09979"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_15504"} +{"question": "What is the first work that proposed a deep-learning-based approach for automatic colorization?", "answer": ["Deep Colorization"], "answer_arxiv_id": ["1605.00075v1"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15505"} +{"question": "What are some studies in the field of generative models applied to tasks such as image-to-image translation and image reconstruction?", "answer": ["The Medical Segmentation Decathlon", "ViViT: A Video Vision Transformer"], "answer_arxiv_id": ["2106.05735", "2103.15691"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_15506"} +{"question": "Are there any papers that proposed modified versions of WDRO, allowing for broader generalization guarantees but with error terms that vanish only when ρ goes to zero?", "answer": ["Finite-Sample Guarantees for Wasserstein Distributionally Robust Optimization: Breaking the Curse of Dimensionality"], "answer_arxiv_id": ["2009.04382v3"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_15507"} +{"question": "Can you specify research papers that developed hybrid representations in NeRF methods?", "answer": ["Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction", "Improved Direct Voxel Grid Optimization for Radiance Fields Reconstruction", "Neural Sparse Voxel Fields", "TensoRF: Tensorial Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "NeRFPlayer: A Streamable Dynamic Scene Representation with Decomposed Neural Radiance Fields", "Point-NeRF: Point-based Neural Radiance Fields", "HexPlane: A Fast Representation for Dynamic Scenes", "Factor Fields: A Unified Framework for Neural Fields and Beyond"], "answer_arxiv_id": ["2111.11215", "2206.05085", "2007.11571", "2203.09517", "2201.05989", "2210.15947", "2201.08845", "2301.09632", "2302.01226"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_15508"} +{"question": "What papers have proposed deep RL algorithms that employ auxiliary prediction tasks?", "answer": ["Reinforcement Learning with Unsupervised Auxiliary Tasks", "Eigenoption Discovery through the Deep Successor Representation", "A Geometric Perspective on Optimal Representations for Reinforcement Learning", "DeepMDP: Learning Continuous Latent Space Models for Representation Learning", "Hyperbolic Discounting and Learning over Multiple Horizons", "The Value-Improvement Path: Towards Better Representations for Reinforcement Learning", "Understanding and Preventing Capacity Loss in Reinforcement Learning"], "answer_arxiv_id": ["1611.05397", "1710.11089", "1901.11530", "1906.02736", "1902.06865", "2006.02243", "2204.09560"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_15509"} +{"question": "Any work about models that introduced trainable temporal layers to construct Text-to-Video generative models?", "answer": ["SimDA: Simple Diffusion Adapter for Efficient Video Generation", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "MagicEdit: High-Fidelity and Temporally Coherent Video Editing"], "answer_arxiv_id": ["2308.09710", "2307.04725", "2308.14749"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_15510"} +{"question": "Could you provide some studies that used functional methods to organize neural network units based on unit activation patterns?", "answer": ["Clustering units in neural networks: upstream vs downstream information"], "answer_arxiv_id": ["2203.11815"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_15511"} +{"question": "What works are about super-resolution methods which exploit the high frequency information of images?", "answer": ["Meta-SR: A Magnification-Arbitrary Network for Super-Resolution", "Learning Continuous Image Representation with Local Implicit Image Function", "Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data", "ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks"], "answer_arxiv_id": ["1903.00875", "2012.09161", "2107.10833", "1809.00219"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_15512"} +{"question": "What works are associated with specific initial parameters that are predicted by a GHN given a computational graph of the neural network?", "answer": ["Breaking the Architecture Barrier: A Method for Efficient Knowledge Transfer Across Networks", "Net2Net: Accelerating Learning via Knowledge Transfer", "GradMax: Growing Neural Networks using Gradient Information"], "answer_arxiv_id": ["2212.13970", "1511.05641", "2201.05125"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_15513"} +{"question": "Which studies have adopted convolutional neural networks for feature extraction and learning in RGB camera based tracking?", "answer": ["Learning Multi-Domain Convolutional Neural Networks for Visual Tracking"], "answer_arxiv_id": ["1510.07945"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_15514"} +{"question": "Which works propose dealing with multiple loss minimization tasks?", "answer": ["Classification with Fairness Constraints: A Meta-Algorithm with Provable Guarantees", "A Reductions Approach to Fair Classification", "Average Individual Fairness: Algorithms, Generalization and Experiments"], "answer_arxiv_id": ["1806.06055", "1803.02453", "1905.10607"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_15515"} +{"question": "Which studies used model-agnostic approaches as a way of improving the quality of masks in pre-existing segmentation models?", "answer": ["Look Closer to Segment Better: Boundary Patch Refinement for Instance Segmentation", "SegFix: Model-Agnostic Boundary Refinement for Segmentation", "DeepStrip: High Resolution Boundary Refinement", "CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement", "High Quality Segmentation for Ultra High-resolution Images", "PolyTransform: Deep Polygon Transformer for Instance Segmentation"], "answer_arxiv_id": ["2104.05239", "2007.04269", "2003.11670", "2005.02551", "2111.14482", "1912.02801"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_15516"} +{"question": "Any works about how abductive reasoning is framed within natural language understanding or multimodal vision-language integration?", "answer": ["Abductive Commonsense Reasoning", "The Abduction of Sherlock Holmes: A Dataset for Visual Abductive Reasoning", "Visual Abductive Reasoning"], "answer_arxiv_id": ["1908.05739", "2202.04800", "2203.14040v1"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_15517"} +{"question": "Could you provide me some research that discuss the representation-based approach in semi-supervised learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1911.05722"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_15518"} +{"question": "What works have been done on global explanation methods that aim to explain the prediction strategy learned by the machine across the population?", "answer": ["One Explanation Does Not Fit All: A Toolkit and Taxonomy of AI Explainability Techniques"], "answer_arxiv_id": ["1909.03012"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_15519"} +{"question": "Can you provide a recent study that introduced the copy mechanism in text generation tasks to select text segments from other documents?", "answer": ["Copy Is All You Need"], "answer_arxiv_id": ["2307.06962"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_15520"} +{"question": "What studies proposed hierarchical classification-regression frameworks for visual localization?", "answer": ["Hierarchical Scene Coordinate Classification and Regression for Visual\n Localization", "Large Scale Joint Semantic Re-Localisation and Scene Understanding via\n Globally Unique Instance Coordinate Regression", "LoFTR: Detector-Free Local Feature Matching with Transformers"], "answer_arxiv_id": ["1909.06216", "1909.10239", "2104.00680"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15521"} +{"question": "What works use the classifier as a reward model for reinforcement learning?", "answer": ["Reinforcement Learning Based Text Style Transfer without Parallel Training Corpus"], "answer_arxiv_id": ["1903.10671"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_15522"} +{"question": "Can you point out the major works in the past decade on communication-efficient variants of Data-Parallel SGD?", "answer": ["QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "PowerSGD: Practical Low-Rank Gradient Compression for Distributed Optimization", "DoubleSqueeze: Parallel Stochastic Gradient Descent with Double-pass Error-Compensated Compression", "Atomo: Communication-efficient Learning via Atomic Sparsification"], "answer_arxiv_id": ["1610.02132", "1905.13727", "1905.05957", "1806.04090"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_15523"} +{"question": "Which works describe learning one type of correspondences by manually creating misalignment, such as shifting audio temporally to create temporal supervision?", "answer": ["Audio-Visual Scene Analysis with Self-Supervised Multisensory Features", "Cooperative Learning of Audio and Video Models from Self-Supervised\n Synchronization", "Look, Listen and Learn"], "answer_arxiv_id": ["1804.03641", "1807.00230", "1705.08168"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_15524"} +{"question": "Can you provide examples of research that encourage the algorithm to pick actions that reduce the uncertainty in its knowledge of the environment?", "answer": ["VIME: Variational Information Maximizing Exploration", "Variational Information Maximisation for Intrinsically Motivated Reinforcement Learning", "Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models"], "answer_arxiv_id": ["1605.09674", "1509.08731v1", "1507.00814"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15525"} +{"question": "Which works are related to actor-critic approaches in model-free reinforcement learning?", "answer": ["Distributed Distributional Deterministic Policy Gradients", "Continuous control with deep reinforcement learning"], "answer_arxiv_id": ["1804.08617", "1509.02971"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_15526"} +{"question": "Which works employed a GAN to represent an implicit function with a shared MLP in the field of Neural Fields?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "SDF-StyleGAN: Implicit SDF-Based StyleGAN for 3D Shape Generation"], "answer_arxiv_id": ["1812.02822", "2206.12055"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_15527"} +{"question": "Which papers discuss Latent Diffusion Models that perform the diffusion process in the latent space of a Variational AutoEncoder?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Neural Discrete Representation Learning"], "answer_arxiv_id": ["2112.10752", "1711.00937"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_15528"} +{"question": "Which research proposed dual and semi-dual methods for dealing with strongly convex regularizers in the Sinkhorn algorithm?", "answer": ["Smooth and Sparse Optimal Transport", "Quadratically regularized optimal transport"], "answer_arxiv_id": ["1710.06276", "1903.01112"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15529"} +{"question": "What studies are about using world model in model-based RL?", "answer": ["World Models", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1803.10122", "2010.02193"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_15530"} +{"question": "What works proposed augmenting a single positive of an anchor video with a RGB stream?", "answer": ["A Large-Scale Study on Unsupervised Spatiotemporal Representation Learning", "Spatiotemporal Contrastive Video Representation Learning", "Broaden Your Views for Self-Supervised Video Learning"], "answer_arxiv_id": ["2104.14558", "2008.03800", "2103.16559"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_15531"} +{"question": "What studies introduced modern feature selection methods that utilize feature inclusion weights?", "answer": ["Interpretable Explanations of Black Boxes by Meaningful Perturbation", "LassoNet: A Neural Network with Feature Sparsity", "A Rate-Distortion Framework for Explaining Black-box Model Decisions"], "answer_arxiv_id": ["1704.03296", "1907.12207", "2110.08252"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_15532"} +{"question": "Which papers discuss approaches using a lightweight backbone and feature aggregation module in mobile semantic segmentation?", "answer": ["DFANet: Deep Feature Aggregation for Real-Time Semantic Segmentation"], "answer_arxiv_id": ["1904.02216"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15533"} +{"question": "What works introduced methods like NFD, 3DGen, and DiffTF that employ diffusion on the triplane representation?", "answer": ["3D Neural Field Generation using Triplane Diffusion", "3DGen: Triplane Latent Diffusion for Textured Mesh Generation", "Large-Vocabulary 3D Diffusion Model with Transformer"], "answer_arxiv_id": ["2211.16677", "2303.05371", "2309.07920"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_15534"} +{"question": "What works enforce conservative estimates of future rewards as a solution for DS problem?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "COMBO: Conservative Offline Model-Based Policy Optimization", "Adversarially Trained Actor Critic for Offline Reinforcement Learning"], "answer_arxiv_id": ["2006.04779", "2102.08363", "2202.02446"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_15535"} +{"question": "Which research studies have utilized regularization as a method for creating personalized models in FL?", "answer": ["Ditto: Fair and Robust Federated Learning Through Personalization", "Personalized Federated Learning with Moreau Envelopes"], "answer_arxiv_id": ["2012.04221", "2006.08848"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_15536"} +{"question": "What works have been done on reconstruction in multi-body systems?", "answer": ["Ditto: Building Digital Twins of Articulated Objects from Interaction", "Unsupervised pose-aware part decomposition for 3D articulated objects", "ScrewNet: Category-Independent Articulation Model Estimation From Depth Images Using Screw Theory", "A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representation", "CaDeX: Learning Canonical Deformation Coordinate Space for Dynamic Surface Representation via Neural Homeomorphism"], "answer_arxiv_id": ["2202.08227", "2110.04411", "2008.10518", "2104.07645", "2203.16529"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15537"} +{"question": "Which works introduced graph coarsening to machine learning as a methodology of graph summarization?", "answer": ["Graph Summarization Methods and Applications: A Survey"], "answer_arxiv_id": ["1612.04883"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_15538"} +{"question": "Which studies discuss the success of vision-language models in Few-shot Classification?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "FILIP: Fine-grained Interactive Language-Image Pre-Training"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.07783"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15539"} +{"question": "Which works in the field of Generative Adversarial Networks (GANs) manipulated the latent space to achieve the desired effect in the generated images?", "answer": ["A Spectral Regularizer for Unsupervised Disentanglement", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated Images using Conditional Continuous Normalizing Flows", "GANSpace: Discovering Interpretable GAN Controls", "Closed-Form Factorization of Latent Semantics in GANs", "LatentCLR: A Contrastive Learning Approach for Unsupervised Discovery of Interpretable Directions", "Drag Your GAN: Interactive Point-based Manipulation on the Generative Image Manifold"], "answer_arxiv_id": ["1812.01161v2", "2103.17249", "2008.02401", "2004.02546", "2007.06600v4", "2104.00820", "2305.10973"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_15540"} +{"question": "Which research work focused on improving LLMs for multi-step computation through finetuning a LM on intermediate computation steps?", "answer": ["Show Your Work: Scratchpads for Intermediate Computation with Language Models"], "answer_arxiv_id": ["2112.00114"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_15541"} +{"question": "What works discuss the integration of inductive logic programming (ILP) with neural networks?", "answer": ["End-to-End Differentiable Proving", "Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "Learning Explanatory Rules from Noisy Data", "Neural Logic Reinforcement Learning", "DRUM: End-To-End Differentiable Rule Mining On Knowledge Graphs", "Learn to Explain Efficiently via Neural Logic Inductive Learning"], "answer_arxiv_id": ["1705.11040", "1702.08367", "1711.04574", "1904.10729", "1911.00055", "1910.02481"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_15542"} +{"question": "What works talk about how LLMs exploit shortcuts or spurious correlations?", "answer": ["Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference", "PAWS: Paraphrase Adversaries from Word Scrambling", "Hypothesis Only Baselines in Natural Language Inference", "Annotation Artifacts in Natural Language Inference Data"], "answer_arxiv_id": ["1902.01007", "1904.01130", "1805.01042", "1803.02324"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_15543"} +{"question": "Could you provide me the work that generated robust representations by contrasting the predicted next frame in Contrastive Predictive Coding?", "answer": ["Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1807.03748"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15544"} +{"question": "Could you provide me some works about Neural Operators?", "answer": ["PDE-Net: Learning PDEs from Data", "PDE-Net 2.0: Learning PDEs from Data with A Numeric-Symbolic Hybrid Deep Network", "Towards physics-informed deep learning for turbulent flow prediction", "Combining Differentiable PDE Solvers and Graph Neural Networks for Fluid Flow Prediction", "Physics-Informed Neural Operator for Learning Partial Differential Equations"], "answer_arxiv_id": ["1710.09668", "1812.04426", "1911.08655", "2007.04439", "2111.03794"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_15545"} +{"question": "Which works study the two-sample problem with sample sizes following independent Poisson processes?", "answer": ["The two-sample problem for Poisson processes: adaptive tests with a non-asymptotic wild bootstrap approach"], "answer_arxiv_id": ["1203.3572"], "source_meta": {"published_time": "20211028"}, "qid": "AutoScholarQuery_train_15546"} +{"question": "What is the work that successfully created desired objects within specific bounding boxes?", "answer": ["GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2301.07093"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_15547"} +{"question": "Which papers centered on the analysis of variants of AdaGrad for different classes of functions?", "answer": ["Scale-Free Algorithms for Online Linear Optimization", "AdaGrad stepsizes: Sharp convergence over nonconvex landscapes", "Linear Convergence of Adaptive Stochastic Gradient Descent"], "answer_arxiv_id": ["1502.05744", "1806.01811", "1908.10525v2"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_15548"} +{"question": "Can you tell me about studies that extended the standard multi-armed bandit setting to multiple-point bandit feedback?", "answer": ["Regret in Online Combinatorial Optimization"], "answer_arxiv_id": ["1204.4710"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_15549"} +{"question": "Which research papers utilized Recurrent Neural Networks for Keyphrase Generation?", "answer": ["Deep Keyphrase Generation", "Does Order Matter? An Empirical Study on Generating Multiple Keyphrases\n as a Sequence", "An Empirical Study on Neural Keyphrase Generation", "One Size Does Not Fit All: Generating and Evaluating Variable Number of\n Keyphrases"], "answer_arxiv_id": ["1704.06879", "1909.03590", "2009.10229", "1810.05241"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_15550"} +{"question": "Which study extends its open-vocabulary object detection capability beyond text by using both the image encoder and text encoder from CLIP?", "answer": ["Open-Vocabulary DETR with Conditional Matching"], "answer_arxiv_id": ["2203.11876"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_15551"} +{"question": "Could you name the works that improved the iterative render-and-compare method?", "answer": ["CosyPose: Consistent multi-view multi-object 6D pose estimation", "On the Continuity of Rotation Representations in Neural Networks", "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", "Normalized Object Coordinate Space for Category-Level 6D Object Pose and\n Size Estimation", "Disentangling Monocular 3D Object Detection"], "answer_arxiv_id": ["2008.08465", "1812.07035", "1905.11946", "1901.02970", "1905.12365"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_15552"} +{"question": "Could you provide me some studies about reward-free RL that considered a class of Markov Decision Processes (MDPs) with low inherent Bellman error?", "answer": ["Provably Efficient Reward-Agnostic Navigation with Linear Value Iteration"], "answer_arxiv_id": ["2008.07737"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_15553"} +{"question": "What studies introduce the Masked Image Modeling (MIM) in the field of pre-training in computer vision?", "answer": ["BEiT: BERT Pre-Training of Image Transformers"], "answer_arxiv_id": ["2106.08254"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_15554"} +{"question": "Do any research papers demonstrate the effectiveness of a relatively light model with retrieval augmentations in knowledge-intensive NLP tasks?", "answer": ["Few-shot Learning with Retrieval Augmented Language Models"], "answer_arxiv_id": ["2208.03299"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_15555"} +{"question": "What studies have addressed utterance-level speech tasks such as speaker recognition or language identification?", "answer": ["Robust Speech Recognition via Large-Scale Weak Supervision", "Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis", "AutoVC: Zero-Shot Voice Style Transfer with Only Autoencoder Loss"], "answer_arxiv_id": ["2212.04356", "1806.04558", "1905.05879"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_15556"} +{"question": "Which works studied text-guided image synthesis in the context of GANs?", "answer": ["Generative Adversarial Nets"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_15557"} +{"question": "What are some works that proposed fixed policies for simultaneous machine translation?", "answer": ["STACL: Simultaneous Translation with Implicit Anticipation and\n Controllable Latency using Prefix-to-Prefix Framework", "Efficient Wait-k Models for Simultaneous Machine Translation", "Universal Simultaneous Machine Translation with Mixture-of-Experts\n Wait-k Policy"], "answer_arxiv_id": ["1810.08398", "2005.08595v2", "2109.05238"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_15558"} +{"question": "What study introduced the concept of a feedforward network that takes RGB frames of a fixed temporal window as input and predicts the motion for any given query point through iterative updates?", "answer": ["Particle Video Revisited: Tracking Through Occlusions Using Point\n Trajectories"], "answer_arxiv_id": ["2204.04153"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_15559"} +{"question": "Which works provide references to the use of graph matching in 3D vision tasks?", "answer": ["Self-Supervised Learning for Multimodal Non-Rigid 3D Shape Matching"], "answer_arxiv_id": ["2303.10971"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_15560"} +{"question": "Could you provide me some researches that proposed methods to solve the trajectory stitching problem in offline RL?", "answer": ["Q-learning Decision Transformer: Leveraging Dynamic Programming for Conditional Sequence Modelling in Offline RL", "Model-based Trajectory Stitching for Improved Offline Reinforcement Learning", "BATS: Best Action Trajectory Stitching"], "answer_arxiv_id": ["2209.03993", "2211.11603", "2204.12026"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_15561"} +{"question": "Could you provide some studies that used model pruning to accelerate inference at lower precision while retaining quality?", "answer": ["Squeezing Large-Scale Diffusion Models for Mobile", "SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two\n Seconds"], "answer_arxiv_id": ["2307.01193", "2306.00980"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_15562"} +{"question": "What subsequent works incorporate representation learning for NCD?", "answer": ["Automatically Discovering and Learning New Visual Categories with Ranking Statistics", "Neighborhood Contrastive Learning for Novel Class Discovery", "A Unified Objective for Novel Class Discovery"], "answer_arxiv_id": ["2002.05714", "2106.10731", "2108.08536"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_15563"} +{"question": "Which research revisits discriminator as reward in the context of imitation learning?", "answer": ["Offline Learning from Demonstrations and Unlabeled Experience", "Discriminator-Actor-Critic: Addressing Sample Inefficiency and Reward Bias in Adversarial Imitation Learning", "Generative Adversarial Imitation Learning", "Learning Robust Rewards with Adversarial Inverse Reinforcement Learning", "LobsDICE: Offline Learning from Observation via Stationary Distribution Correction Estimation", "Imitation Learning from Imperfect Demonstration", "Discriminator-Weighted Offline Imitation Learning from Suboptimal Demonstrations"], "answer_arxiv_id": ["2011.13885", "1809.02925", "1606.03476", "1710.11248", "2202.13536", "1901.09387", "2207.10050"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_15564"} +{"question": "Which papers propose forming pseudo-image-text pairs using an image-text retrieval alignment?", "answer": ["Unsupervised Vision-and-Language Pre-training via Retrieval-based Multi-Granular Alignment"], "answer_arxiv_id": ["2203.00242"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_15565"} +{"question": "Which papers explored scaling of the Vision Transformer models?", "answer": ["Scaling Vision Transformers", "EVA: Exploring the Limits of Masked Visual Representation Learning at\n Scale", "Scaling Vision Transformers to 22 Billion Parameters"], "answer_arxiv_id": ["2106.04560", "2211.07636", "2302.05442"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_15566"} +{"question": "Any works that applies prompt-based tuning methods on pre-trained models using edge prediction?", "answer": ["GraphPrompt: Unifying Pre-Training and Downstream Tasks for Graph Neural Networks"], "answer_arxiv_id": ["2302.08043"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_15567"} +{"question": "Could you give me examples of research that use the model’s epistemic uncertainty as an intrinsic reward for planning?", "answer": ["Self-Supervised Exploration via Disagreement", "Planning to Explore via Self-Supervised World Models"], "answer_arxiv_id": ["1906.04161", "2005.05960"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_15568"} +{"question": "What papers discuss the application of multi-objective optimization (MOO) in multi-task learning (MTL)?", "answer": ["Multi-Task Learning for Dense Prediction Tasks: A Survey", "A Joint Many-Task Model: Growing a Neural Network for Multiple NLP Tasks", "An Overview of Multi-Task Learning in Deep Neural Networks", "A Survey on Multi-Task Learning"], "answer_arxiv_id": ["2004.13379", "1611.01587", "1706.05098", "1707.08114"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_15569"} +{"question": "What work first proposed the diffusion probabilistic model?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_15570"} +{"question": "Which works focus on Neural Network Gaussian Processes (NNGPs)?", "answer": ["Deep Neural Networks as Gaussian Processes", "Gaussian Process Behaviour in Wide Deep Neural Networks", "Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes", "Deep Convolutional Networks as shallow Gaussian Processes", "Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1711.00165", "1804.11271", "1810.05148", "1808.05587", "1806.07572"], "source_meta": {"published_time": "20210830"}, "qid": "AutoScholarQuery_train_15571"} +{"question": "Which research shows that Codex performs better in following task instructions and exploiting algorithmic patterns based on the prompt exemplars?", "answer": ["Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them"], "answer_arxiv_id": ["2210.09261v1"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_15572"} +{"question": "Can you tell me about any papers that have studied VLMs in the video domain trained on both image and video data paired with text?", "answer": ["Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval"], "answer_arxiv_id": ["2104.00650"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_15573"} +{"question": "What papers proposed the use of CTC-based methods in language-free scene text recognition?", "answer": ["Rosetta: Large scale system for text detection and recognition in images"], "answer_arxiv_id": ["1910.05085"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_15574"} +{"question": "Which studies claim that intervening allows to infer the edge orientation of any edge cut?", "answer": ["Learning Causal Graphs with Small Interventions", "Cost-Optimal Learning of Causal Graphs"], "answer_arxiv_id": ["1511.00041", "1703.02645"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15575"} +{"question": "Could you provide the works where an RNN has been used to retrieve reasoning paths over the Wikipedia graph?", "answer": ["Learning to Retrieve Reasoning Paths over Wikipedia Graph for Question Answering"], "answer_arxiv_id": ["1911.10470"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_15576"} +{"question": "What papers applied deep learning in iterative disparity refinement?", "answer": ["End-to-End Learning of Geometry and Context for Deep Stereo Regression", "A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation", "Learning for Disparity Estimation through Feature Constancy", "StereoNet: Guided Hierarchical Refinement for Real-Time Edge-Aware Depth\n Prediction", "Practical Deep Stereo (PDS): Toward applications-friendly deep stereo\n matching", "AANet: Adaptive Aggregation Network for Efficient Stereo Matching", "StereoDRNet: Dilated Residual Stereo Net", "HITNet: Hierarchical Iterative Tile Refinement Network for Real-time\n Stereo Matching", "RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching", "Practical Stereo Matching via Cascaded Recurrent Network with Adaptive\n Correlation", "Uncertainty Guided Adaptive Warping for Robust and Efficient Stereo\n Matching", "Iterative Geometry Encoding Volume for Stereo Matching"], "answer_arxiv_id": ["1703.04309", "1512.02134", "1712.01039", "1807.08865", "1806.01677", "2004.09548", "1904.02251", "2007.12140", "2109.07547", "2203.11483", "2307.14071", "2303.06615"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_15577"} +{"question": "What are the studies about the inversion of images into the latent space of GANs?", "answer": ["Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?", "Image2StyleGAN++: How to Edit the Embedded Images?", "HyperStyle: StyleGAN Inversion with HyperNetworks for Real Image Editing"], "answer_arxiv_id": ["1904.03189", "1911.11544", "2111.15666"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_15578"} +{"question": "What paper indicates that successful advances in deep learning, such as diffusion models, are supported by ideas from differential equations?", "answer": ["A Variational Perspective on Diffusion-Based Generative Models and Score Matching"], "answer_arxiv_id": ["2106.02808"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_15579"} +{"question": "Which works are related to the methodologies of Masked image modeling?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Data-Efficient Image Recognition with Contrastive Predictive Coding", "Context Encoders: Feature Learning by Inpainting"], "answer_arxiv_id": ["1505.05192", "1905.09272", "1604.07379"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_15580"} +{"question": "Are there any papers about building animator-edited datasets from anime communities or by hiring animators to annotate dance motions?", "answer": ["DanceFormer: Music Conditioned 3D Dance Generation with Parametric\n Motion Transformer"], "answer_arxiv_id": ["2103.10206"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_15581"} +{"question": "What studies proposed replicable algorithms in the context of stochastic bandits and clustering?", "answer": ["Replicable Bandits", "Replicable Clustering"], "answer_arxiv_id": ["2210.01898v2", "2302.10359v3"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_15582"} +{"question": "Could you provide me some studies that use autoencoder, VAE, or GAN-based deep learning techniques for the task of inpainting?", "answer": ["Context Encoders: Feature Learning by Inpainting", "Structural inpainting", "Image Inpainting for Irregular Holes Using Partial Convolutions", "SPG-Net: Segmentation Prediction and Guidance Network for Image Inpainting", "Progressive Image Inpainting with Full-Resolution Residual Network", "Deep Inception Generative Network for Cognitive Image Inpainting", "Deep Fusion Network for Image Completion", "Pluralistic Image Completion", "Large Scale Image Completion via Co-Modul-ated Generative Adversarial Networks", "Generating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE"], "answer_arxiv_id": ["1604.07379", "1803.10348", "1804.07723", "1805.03356", "1907.10478", "1812.01458", "1904.08060", "1903.04227", "2103.10428", "2103.10022"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_15583"} +{"question": "Can you cite some studies that incorporated intermediate execution results to guide program search at training and inference time during code generation?", "answer": ["Robust Text-to-SQL Generation with Execution-Guided Decoding"], "answer_arxiv_id": ["1807.03100"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_15584"} +{"question": "Could you provide me some works about diffusion models for image synthesis?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance", "High-Resolution Image Synthesis with Latent Diffusion Models", "DiffuseVAE: Efficient, Controllable and High-Fidelity Generation from\n Low-Dimensional Latents", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2105.05233", "2207.12598", "2112.10752", "2201.00308", "2102.09672"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_15585"} +{"question": "What papers highlight the use of U-Net architecture of diffusion models for pixel-level prediction tasks?", "answer": ["Feature Pyramid Networks for Object Detection", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models", "Prompting Diffusion Representations for Cross-Domain Semantic\n Segmentation"], "answer_arxiv_id": ["1612.03144", "2303.04803", "2307.02138"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_15586"} +{"question": "Any research about ensemble techniques in semi-supervised learning?", "answer": ["Temporal Ensembling for Semi-Supervised Learning"], "answer_arxiv_id": ["1610.02242v3"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_15587"} +{"question": "What are the research articles which have formulated the adaptation to the single test sample as a variational Bayesian inference problem?", "answer": ["Learning to Generalize across Domains on Single Test Samples"], "answer_arxiv_id": ["2202.08045"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_15588"} +{"question": "Could you provide me some studies about the adoption of self-supervised large language models in NLP?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "LLaMA: Open and Efficient Foundation Language Models", "Sparks of Artificial General Intelligence: Early experiments with GPT-4"], "answer_arxiv_id": ["1810.04805", "2005.14165", "2204.02311", "2302.13971", "2303.12712v5"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15589"} +{"question": "What works have investigated federated learning for kernelized contextual bandits in synchronous and asynchronous settings?", "answer": ["Communication Efficient Distributed Learning for Kernelized Contextual Bandits"], "answer_arxiv_id": ["2206.04835"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_15590"} +{"question": "What works focused on timing strategies in the adversarial formulation?", "answer": ["Tactics of Adversarial Attack on Deep Reinforcement Learning Agents", "Stealthy and Efficient Adversarial Attacks against Deep Reinforcement Learning"], "answer_arxiv_id": ["1703.06748", "2005.07099"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_15591"} +{"question": "Could you provide me some works about implementing hierarchical reasoning and complex open-ended context understanding?", "answer": ["Zero-Shot Task Generalization with Multi-Task Deep Reinforcement\n Learning", "Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online\n Videos", "MineDojo: Building Open-Ended Embodied Agents with Internet-Scale\n Knowledge"], "answer_arxiv_id": ["1706.05064", "2206.11795", "2206.08853"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_15592"} +{"question": "Which papers developed methods for program repair by focusing on grammatical features and semantic information of programs?", "answer": ["Dynamic Neural Program Embeddings for Program Repair", "Neural Program Repair with Execution-based Backpropagation"], "answer_arxiv_id": ["1711.07163", "2105.04123"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_15593"} +{"question": "What studies discuss kernel herding schemes?", "answer": ["Super-Samples from Kernel Herding", "On the Equivalence between Herding and Conditional Gradient Algorithms", "Optimally-Weighted Herding is Bayesian Quadrature"], "answer_arxiv_id": ["1203.3472", "1203.4523", "1408.2049v2"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_15594"} +{"question": "Can you list the works that focus on designing category-specific networks for 3D single-view object reconstruction?", "answer": ["Category-Specific Object Reconstruction from a Single Image", "CCD-3DR: Consistent Conditioning in Diffusion for Single-Image 3D Reconstruction", "Shape and Viewpoint without Keypoints", "Learning Category-Specific Mesh Reconstruction from Image Collections", "Self-supervised Single-view 3D Reconstruction via Semantic Consistency", "Shelf-Supervised Mesh Prediction in the Wild", "U-RED: Unsupervised 3D Shape Retrieval and Deformation for Partial Point Clouds"], "answer_arxiv_id": ["1411.6069", "2308.07837", "2007.10982", "1803.07549", "2003.06473", "2102.06195", "2308.06383"], "source_meta": {"published_time": "20230816"}, "qid": "AutoScholarQuery_train_15595"} +{"question": "Which papers are concerned with the study of PuB-AMG in two-player zero-sum games?", "answer": ["Structure in the Value Function of Two-Player Zero-Sum Games of Incomplete Information", "On Bellman’s Optimality Principle for zs-POSGs"], "answer_arxiv_id": ["1606.06888v1", "2006.16395"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_15596"} +{"question": "Which works applied language models in the context of purely informal mathematics?", "answer": ["Deep learning for symbolic mathematics", "Measuring Mathematical Problem Solving With the MATH Dataset", "NaturalProofs: Mathematical Theorem Proving in Natural Language", "A Neural Network Solves, Explains, and Generates University Math Problems by Program Synthesis and Few-Shot Learning at Human Level", "NaturalProver: Grounded Mathematical Proof Generation with Language Models"], "answer_arxiv_id": ["1912.01412", "2103.03874", "2104.01112", "2112.15594v4", "2205.12910"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_15597"} +{"question": "What is the pretraining strategy proposed by SoCo for convolutional detectors?", "answer": ["Aligning Pretraining for Detection via Object-Level Contrastive Learning"], "answer_arxiv_id": ["2106.02637"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_15598"} +{"question": "Which studies have implemented hyper-complex layers in the context of cross-block orchestration?", "answer": ["Quaternion Recurrent Neural Networks"], "answer_arxiv_id": ["1806.04418"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_15599"} +{"question": "What research has been done on learning deeper ReLU networks?", "answer": ["Learning Deep ReLU Networks Is Fixed-Parameter Tractable"], "answer_arxiv_id": ["2009.13512"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15600"} +{"question": "Could you provide me references about the studies on multi-task transfer learning?", "answer": ["The Natural Language Decathlon: Multitask Learning as Question Answering", "UnifiedQA: Crossing Format Boundaries With a Single QA System", "CrossFit: A Few-shot Learning Challenge for Cross-task Generalization in NLP", "Exploring and Predicting Transferability across NLP Tasks"], "answer_arxiv_id": ["1806.08730", "2005.00700", "2104.08835", "2005.00770"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_15601"} +{"question": "What research proposed the incorporation of Latent Diffusion Models (LDMs) in high-resolution text-to-image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_15602"} +{"question": "Which papers utilize Graph neural networks in Session-based Recommendation Systems?", "answer": ["Session-based Recommendation with Graph Neural Networks", "Rethinking the Item Order in Session-based Recommendation with Graph Neural Networks", "Global Context Enhanced Graph Neural Networks for Session-based Recommendation", "Learning Multi-granularity Consecutive User Intent Unit for Session-based Recommendation"], "answer_arxiv_id": ["1811.00855", "1911.11942", "2106.05081", "2112.13197"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_15603"} +{"question": "Could you tell me which research utilized the FS-COCO dataset of freehand sketches with their textual descriptions?", "answer": ["FS-COCO: Towards Understanding of Freehand Sketches of Common Objects in\n Context"], "answer_arxiv_id": ["2203.02113"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_15604"} +{"question": "Which works propose to quantify uncertainty directly from language model generations?", "answer": ["Teaching Models to Express Their Uncertainty in Words", "Language Models (Mostly) Know What They Know"], "answer_arxiv_id": ["2205.14334", "2207.05221"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_15605"} +{"question": "What papers found that normalizing flow-based methods outperform GANs in image generation tasks?", "answer": ["Low-Light Image Enhancement with Normalizing Flow", "SRFlow: Learning the Super-Resolution Space with Normalizing Flow", "Glow: Generative Flow with Invertible 1x1 Convolutions"], "answer_arxiv_id": ["2109.05923", "2006.14200", "1807.03039"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_15606"} +{"question": "Which papers discuss learning invariances across training and testing domains in the context of learning a predictor from multiple domains?", "answer": ["In Search of Lost Domain Generalization", "Invariant Risk Minimization", "Deep CORAL: Correlation Alignment for Deep Domain Adaptation"], "answer_arxiv_id": ["2007.01434", "1907.02893", "1607.01719"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_15607"} +{"question": "Which papers discussed purification of adversarial examples using generative models?", "answer": ["PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examples", "(Certified!!) Adversarial Robustness for Free!", "Adversarial Purification with Score-based Generative Models", "Diffusion Models for Adversarial Purification", "Guided Diffusion Model for Adversarial Purification", "DensePure: Understanding Diffusion Models towards Adversarial Robustness"], "answer_arxiv_id": ["1710.10766", "2206.10550", "2106.06041", "2205.07460", "2205.14969", "2211.00322"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_15608"} +{"question": "Which works repurposed the CLIP model to operate as an open-vocabulary detector?", "answer": ["RegionCLIP: Region-based Language-Image Pretraining", "Simple Open-Vocabulary Object Detection with Vision Transformers", "Scaling Open-Vocabulary Object Detection"], "answer_arxiv_id": ["2112.09106", "2205.06230", "2306.09683"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15609"} +{"question": "What previous study discusses the applicability of DPMs in 3D protein modeling?", "answer": ["SE(3) diffusion model with application to protein backbone generation"], "answer_arxiv_id": ["2302.02277"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_15610"} +{"question": "Could you provide me some examples of studies that implemented multiple explainers for selecting diverse rationales?", "answer": ["MGR: Multi-generator Based Rationalization"], "answer_arxiv_id": ["2305.04492"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_15611"} +{"question": "Which paper proposed Sharpness-Aware Minimization (SAM), a state-of-the-art optimization methodology?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization"], "answer_arxiv_id": ["2010.01412"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_15612"} +{"question": "Which research papers discuss the area of contrastive learning as a paradigm in instance discrimination?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Prototypical Contrastive Learning of Unsupervised Representations"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2005.04966"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_15613"} +{"question": "Which studies use 2D convolutions on 3D cost volume for cost construction and aggregation in stereo matching networks?", "answer": ["A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation", "Learning for Disparity Estimation through Feature Constancy", "Hierarchical Discrete Distribution Decomposition for Match Density\n Estimation", "Real-time self-adaptive deep stereo", "AANet: Adaptive Aggregation Network for Efficient Stereo Matching", "Continual Adaptation for Deep Stereo"], "answer_arxiv_id": ["1512.02134", "1712.01039", "1812.06264", "1810.05424", "2004.09548", "2007.05233"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_15614"} +{"question": "In what research was the idea of positional or structural encodings capturing identity-awareness proposed?", "answer": ["Identity-aware Graph Neural Networks"], "answer_arxiv_id": ["2101.10320"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_15615"} +{"question": "What studies incorporated diffusion models into their conditional image generation methods?", "answer": ["ECON: Explicit Clothed humans Optimized via Normal integration", "Dense Text-to-Image Generation with Attention Modulation", "Harnessing the Spatial-Temporal Attention of Diffusion Models for\n High-Fidelity Text-to-Image Synthesis", "Freestyle Layout-to-Image Synthesis", "Multiscale Structure Guided Diffusion for Image Deblurring", "Chop & Learn: Recognizing and Generating Object-State Compositions"], "answer_arxiv_id": ["2212.07422", "2308.12964", "2304.03869", "2303.14412", "2212.01789", "2309.14339"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15616"} +{"question": "Which papers discuss searching over the manifold directly in the context of archive distillation?", "answer": ["Discovering Representations for Black-box Optimization", "Policy Manifold Search: Exploring the Manifold Hypothesis for Diversity-based Neuroevolution"], "answer_arxiv_id": ["2003.04389", "2104.13424v1"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15617"} +{"question": "Are there any works that showed the impact of trivial alterations to test samples on LLM’s ToM performance?", "answer": ["Large Language Models Fail on Trivial Alterations to Theory-of-Mind\n Tasks", "Clever Hans or Neural Theory of Mind? Stress Testing Social Reasoning in\n Large Language Models"], "answer_arxiv_id": ["2302.08399", "2305.14763"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_15618"} +{"question": "Can you provide a work that addresses the data scarcity in specific domains by Pre-training LLMs?", "answer": ["TaPEx: Table Pre-training via Learning a Neural SQL Executor"], "answer_arxiv_id": ["2107.07653"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_15619"} +{"question": "What studies utilize separate networks for spatial correlations and temporal dependencies in multivariate time series forecasting?", "answer": ["Connecting the Dots: Identifying Network Structure via Graph Signal Processing†", "Graph WaveNet for Deep Spatial-Temporal Graph Modeling", "Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting", "Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting"], "answer_arxiv_id": ["1810.13066", "1906.00121", "2103.07719", "2007.02842"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_15620"} +{"question": "Which models have shown promising outcomes in generating high-quality talking heads?", "answer": ["StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via\n Pre-trained StyleGAN"], "answer_arxiv_id": ["2203.04036"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_15621"} +{"question": "Which papers focus on optimizing the model runtime on devices?", "answer": ["Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models"], "answer_arxiv_id": ["2211.02048"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_15622"} +{"question": "Which studies do research on bi-encoders with a focus on negative sample mining?", "answer": ["Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", "Approximate Nearest Neighbor Negative Contrastive Learning for Dense\n Text Retrieval", "Dense Passage Retrieval for Open-Domain Question Answering", "SimCSE: Simple Contrastive Learning of Sentence Embeddings", "Sparse, Dense, and Attentional Representations for Text Retrieval", "RocketQA: An Optimized Training Approach to Dense Passage Retrieval for\n Open-Domain Question Answering", "Optimizing Dense Retrieval Model Training with Hard Negatives", "DiffCSE: Difference-based Contrastive Learning for Sentence Embeddings"], "answer_arxiv_id": ["1908.10084", "2007.00808", "2004.04906", "2104.08821", "2005.00181v3", "2010.08191", "2104.08051", "2204.10298"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_15623"} +{"question": "Are there any research findings that indicate that saddle-point optimization can be challenging when scaling to high dimensions because it may exhibit instabilities and hyperparameter sensitivity?", "answer": ["What Matters for Adversarial Imitation Learning?"], "answer_arxiv_id": ["2106.00672"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15624"} +{"question": "Which work introduced the usage of cascaded diffusion models for high-fidelity, high-resolution generation?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2112.10741"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15625"} +{"question": "What works have improved on the graph hypernetworks (GHNs) model?", "answer": ["Parameter Prediction for Unseen Deep Architectures"], "answer_arxiv_id": ["2110.13100"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_15626"} +{"question": "Could you provide me some studies that model their decoder using a Gaussian mixture distribution in the context of synthetic data generation?", "answer": ["Modeling Tabular Data using Conditional GAN", "CTAB-GAN: Effective Table Data Synthesizing"], "answer_arxiv_id": ["1907.00503", "2102.08369"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_15627"} +{"question": "What studies have shown that RNNs or related models can represent context-free languages like counter and Dyck languages?", "answer": ["On the Practical Computational Power of Finite Precision RNNs for\n Language Recognition", "Sequential Neural Networks as Automata", "A Formal Hierarchy of RNN Architectures", "RNNs can generate bounded hierarchical languages with optimal memory"], "answer_arxiv_id": ["1805.04908", "1906.01615", "2004.08500", "2010.07515"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_15628"} +{"question": "Which papers proposed the use of Super-resolution models to enhance the images produced by latent diffusion models?", "answer": ["Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2106.15282"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_15629"} +{"question": "Which work utilizes the multi-head self-attention of ViT to select representative local patches?", "answer": ["TransFG: A Transformer Architecture for Fine-grained Recognition"], "answer_arxiv_id": ["2103.07976"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_15630"} +{"question": "Who achieved effective training of the 3 billion parameter model, SwinV2-G, and what techniques did they use?", "answer": ["Swin Transformer V2: Scaling Up Capacity and Resolution"], "answer_arxiv_id": ["2111.09883"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_15631"} +{"question": "Which research introduced learned prompt representation in the context of Vision-Language models?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15632"} +{"question": "Which papers present applications of text-guided diffusion models?", "answer": ["RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "SmartBrush: Text and Shape Guided Object Inpainting with Diffusion Model", "Adding Conditional Control to Text-to-Image Diffusion Models", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "Imagic: Text-Based Real Image Editing with Diffusion Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2201.09865", "2212.05034", "2302.05543", "2211.09800", "2210.09276", "2208.12242"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_15633"} +{"question": "Are there any works applying compression with error compensation in communication of model parameters in the decentralized setting?", "answer": ["DeepSqueeze: Decentralization Meets Error-Compensated Compression"], "answer_arxiv_id": ["1907.07346"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_15634"} +{"question": "Which studies argue about overparameterization being necessary for robustness and for interpolating training data using a neural network with a small Lipschitz constant?", "answer": ["A law of robustness for two-layers neural networks", "A Universal Law of Robustness via Isoperimetry"], "answer_arxiv_id": ["2009.14444", "2105.12806"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_15635"} +{"question": "Which works propose more challenging tasks to evaluate the human-imitative capabilities of modern LLMs?", "answer": ["Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models", "Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2206.04615", "2107.03374"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_15636"} +{"question": "Which work used a regression-by-classification loss for absolute pose regression?", "answer": ["Neural Reprojection Error: Merging Feature Learning and Camera Pose\n Estimation"], "answer_arxiv_id": ["2103.07153"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15637"} +{"question": "What are other datasets and benchmarks related to temporal reasoning?", "answer": ["TimeBench: A Comprehensive Evaluation of Temporal Reasoning Abilities in\n Large Language Models"], "answer_arxiv_id": ["2311.17667"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_15638"} +{"question": "Which papers propose to fine-tune the pre-trained text-to-image generation model for customized generation?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion", "LoRA: Low-Rank Adaptation of Large Language Models", "Controlling Text-to-Image Diffusion by Orthogonal Finetuning"], "answer_arxiv_id": ["2208.12242", "2212.04488", "2106.09685", "2306.07280"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_15639"} +{"question": "What research aimed to perform semantic segmentation in an unsupervised manner?", "answer": ["Invariant Information Clustering for Unsupervised Image Classification\n and Segmentation", "Autoregressive Unsupervised Image Segmentation", "InfoSeg: Unsupervised Semantic Image Segmentation with Mutual\n Information Maximization", "PiCIE: Unsupervised Semantic Segmentation using Invariance and\n Equivariance in Clustering", "TransFGU: A Top-down Approach to Fine-Grained Unsupervised Semantic\n Segmentation", "Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "Leveraging Hidden Positives for Unsupervised Semantic Segmentation"], "answer_arxiv_id": ["1807.06653", "2007.08247", "2110.03477", "2103.17070", "2112.01515", "2203.08414", "2303.15014"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_15640"} +{"question": "What are the works investigating biases mitigation with reweighting?", "answer": ["Last Layer Re-Training is Sufficient for Robustness to Spurious\n Correlations"], "answer_arxiv_id": ["2204.02937"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15641"} +{"question": "Could you provide some works that utilize attention features during inversion for spatio-temporal preservation and blending?", "answer": ["FateZero: Fusing Attentions for Zero-shot Text-based Video Editing"], "answer_arxiv_id": ["2303.09535"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15642"} +{"question": "Which paper proposed a balanced k-means clustering algorithm suitable for an AQC?", "answer": ["Balanced k-Means Clustering on an Adiabatic Quantum Computer"], "answer_arxiv_id": ["2008.04419"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_15643"} +{"question": "Which works focus on post-hoc explainability methods that explicitly produce perturbations?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1602.04938", "1705.07874"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_15644"} +{"question": "What are the variants of the neural radiance field (NeRF)?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Neural Sparse Voxel Fields", "SurfelNeRF: Neural Surfel Radiance Fields for Online Photorealistic\n Reconstruction of Indoor Scenes", "NeRFusion: Fusing Radiance Fields for Large-Scale Scene Reconstruction"], "answer_arxiv_id": ["2103.13415", "2007.11571", "2304.08971", "2203.11283"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_15645"} +{"question": "Which papers provide theoretical studies of approaches based on the low-density separation and analysis of self-training?", "answer": ["Multi-class Probabilistic Bounds for Self-learning", "How does unlabeled data improve generalization in self-training? A one-hidden-layer theoretical analysis"], "answer_arxiv_id": ["2109.14422", "2201.08514"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_15646"} +{"question": "What are the works about applying a transformer-based architecture for difficult-aware learning, uncertainty modeling, and temporal consistency in Camouflaged Object Detection?", "answer": ["Implicit Motion Handling for Video Camouflaged Object Detection"], "answer_arxiv_id": ["2203.07363"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_15647"} +{"question": "What studies have successfully applied the expectation-maximization algorithm in natural language processing?", "answer": ["Expectation-Maximization Attention Networks for Semantic Segmentation", "Mixture Models for Diverse Machine Translation: Tricks of the Trade"], "answer_arxiv_id": ["1907.13426", "1902.07816"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_15648"} +{"question": "Can you name studies that used large-scale pretrained vision-language models in a zero-shot manner?", "answer": ["CRIS: CLIP-Driven Referring Image Segmentation", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels"], "answer_arxiv_id": ["2111.15174", "2112.12143"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_15649"} +{"question": "What research proposes the Decision Transformer, a Transformer-based approach to learn a return-conditioned policy for offline RL?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["2106.01345"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_15650"} +{"question": "What papers introduce scale-free algorithms for achieving adaptive regret guarantees?", "answer": ["Scale-Free Online Learning"], "answer_arxiv_id": ["1601.01974"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_15651"} +{"question": "What research attempts to disentangle speaker representation with the removal of irrelevant information like devices, noise, and channel information in speaker recognition?", "answer": ["Adversarial Speaker Verification", "Disentangled Speaker Representation Learning via Mutual Information Minimization"], "answer_arxiv_id": ["1904.12406", "2208.08012"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_15652"} +{"question": "What works build their framework using complete equivariant polynomial basis with the help of spherical harmonics and tensor product in the context of 3D GNNs?", "answer": ["MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields"], "answer_arxiv_id": ["2206.07697"], "source_meta": {"published_time": "20230407"}, "qid": "AutoScholarQuery_train_15653"} +{"question": "What research used task-related features in abundant bias-aligned samples by synthesizing new images?", "answer": ["BiaSwap: Removing Dataset Bias with Bias-Tailored Swapping Augmentation"], "answer_arxiv_id": ["2108.10008"], "source_meta": {"published_time": "20211202"}, "qid": "AutoScholarQuery_train_15654"} +{"question": "What research papers introduced synthetic tasks for better understanding and interpretation of transformers?", "answer": ["What is my math transformer doing? Three results on interpretability and generalization", "Towards Understanding Grokking: An Effective Theory of Representation Learning", "Progress measures for grokking via mechanistic interpretability", "Unveiling Transformers with LEGO: a synthetic reasoning task"], "answer_arxiv_id": ["2211.00170", "2205.10343", "2301.05217", "2206.04301v3"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_15655"} +{"question": "What study provided the Neural Radiance Fields approach of defining the world as an RGB color and transparency for each position and viewing angle?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_15656"} +{"question": "Are there any papers that obtained convergence with high probability for normalized SGD under heavy-tailed noise?", "answer": ["High-probability bounds for Non-Convex Stochastic Optimization with Heavy Tails"], "answer_arxiv_id": ["2106.14343"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_15657"} +{"question": "Which works demonstrate the use of LLMs in natural language processing tasks?", "answer": ["Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference", "Language Models are Few-Shot Learners", "Multitask Prompted Training Enables Zero-Shot Task Generalization"], "answer_arxiv_id": ["2001.07676", "2005.14165", "2110.08207"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_15658"} +{"question": "Which paper describes the training of Point-NeRF MLPs on multiple objects for better generalization?", "answer": ["SimNP: Learning Self-Similarity Priors Between Neural Points"], "answer_arxiv_id": ["2309.03809"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_15659"} +{"question": "Could you list studies that developed binning and its variants for calibration?", "answer": ["Verified Uncertainty Calibration"], "answer_arxiv_id": ["1909.10155"], "source_meta": {"published_time": "20220215"}, "qid": "AutoScholarQuery_train_15660"} +{"question": "Which paper provides a comprehensive survey on survival analysis based on the proportional hazard assumption?", "answer": ["Machine Learning for Survival Analysis: A Survey"], "answer_arxiv_id": ["1708.04649"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_15661"} +{"question": "Which studies introduced deformable human NeRFs that unwrap the posed space to a shared canonicalized space?", "answer": ["A-NeRF: Articulated Neural Radiance Fields for Learning Human Shape,\n Appearance, and Pose", "Vid2Actor: Free-viewpoint Animatable Person Synthesis from Video in the\n Wild", "Animatable Neural Radiance Fields from Monocular RGB Videos", "Neural Articulated Radiance Field", "ARAH: Animatable Volume Rendering of Articulated Human SDFs", "Generative Neural Articulated Radiance Fields", "InstantAvatar: Learning Avatars from Monocular Video in 60 Seconds", "TAVA: Template-free Animatable Volumetric Actors", "DANBO: Disentangled Articulated Neural Body Representations via Graph\n Neural Networks", "Capturing and Animation of Body and Clothing from Monocular Video", "SHERF: Generalizable Human NeRF from a Single Image"], "answer_arxiv_id": ["2102.06199", "2012.12884", "2106.13629", "2104.03110", "2210.10036v1", "2206.14314", "2212.10550", "2206.08929", "2205.01666", "2210.01868", "2303.12791"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_15662"} +{"question": "What research works can provide information about the application of soft Q-functions in online RL settings?", "answer": ["Taming the Noise in Reinforcement Learning via Soft Updates", "Equivalence Between Policy Gradients and Soft Q-Learning", "Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor"], "answer_arxiv_id": ["1512.08562", "1704.06440", "1801.01290"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_15663"} +{"question": "Which research used the positional encoding ability of CNN to overcome problems identified in previous investigations?", "answer": ["Conditional Positional Encodings for Vision Transformers"], "answer_arxiv_id": ["2102.10882"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_15664"} +{"question": "What papers present a generative model that minimizes the Sinkhorn divergence between data distribution and generative distribution?", "answer": ["Learning Generative Models with Sinkhorn Divergences"], "answer_arxiv_id": ["1706.00292v3"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_15665"} +{"question": "What research works focus on human-centric scene understanding in 3D large-scale scenarios?", "answer": ["HSC4D: Human-centered 4D Scene Capture in Large-scale Indoor-outdoor\n Space Using Wearable IMUs and LiDAR", "LiDAR-aid Inertial Poser: Large-scale Human Motion Capture by Sparse\n Inertial and LiDAR Sensors", "LiDARCap: Long-range Marker-less 3D Human Motion Capture with LiDAR\n Point Clouds"], "answer_arxiv_id": ["2203.09215", "2205.15410", "2203.14698"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_15666"} +{"question": "What are the papers that explore conditional generative models approaches to solve super-resolution tasks?", "answer": ["Explorable Super Resolution", "SRFlow: Learning the Super-Resolution Space with Normalizing Flow", "Image Super-Resolution via Iterative Refinement", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic\n Models"], "answer_arxiv_id": ["1912.01839", "2006.14200", "2104.07636", "2104.14951"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_15667"} +{"question": "Which papers proposed techniques to handle the 'oversmoothing' issue in deep GNNs (Graph Neural Networks)?", "answer": ["Representation Learning on Graphs with Jumping Knowledge Networks", "Simple and Deep Graph Convolutional Networks", "Tackling Over-Smoothing for General Graph Convolutional Networks", "PairNorm: Tackling Oversmoothing in GNNs"], "answer_arxiv_id": ["1806.03536v2", "2007.02133", "2008.09864", "1909.12223"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_15668"} +{"question": "What studies propose optimistic policy optimization algorithms based on the natural policy gradient (NPG) algorithm and the policy-cover technique?", "answer": ["PC-PG: Policy Cover Directed Exploration for Provable Policy Gradient Learning", "Provably Correct Optimization and Exploration with Non-linear Policies", "Cautiously Optimistic Policy Optimization and Exploration with Linear Function Approximation"], "answer_arxiv_id": ["2007.08459", "2103.11559", "2103.12923"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_15669"} +{"question": "Which papers provide an overview of SDM (Species distribution modeling)?", "answer": ["Species Distribution Modeling for Machine Learning Practitioners: A Review"], "answer_arxiv_id": ["2107.10400"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_15670"} +{"question": "What study argued about computing the best response policy in zero-sum Markov games?", "answer": ["Near-Optimal Reinforcement Learning with Self-Play"], "answer_arxiv_id": ["2006.12007"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_15671"} +{"question": "What papers used ODE models to study Stochastic Adaptive Momentum (SAM)?", "answer": ["How Does Sharpness-Aware Minimization Minimize Sharpness?", "Towards Understanding Sharpness-Aware Minimization"], "answer_arxiv_id": ["2211.05729", "2206.06232"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_15672"} +{"question": "What works optimise data embedding to enhance the quality of indexed content in the development of Advanced RAG?", "answer": ["Structure-Aware Language Model Pretraining Improves Dense Retrieval on\n Structured Data"], "answer_arxiv_id": ["2305.19912"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_15673"} +{"question": "Can you provide the work that shows hardness in learning depth-222 networks under the uniform distribution based on the hardness of the Learning with Rounding (LWR) problem?", "answer": ["Hardness of Noise-Free Learning for Two-Hidden-Layer Neural Networks"], "answer_arxiv_id": ["2202.05258"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_15674"} +{"question": "Could you provide me some studies about recent approaches that formulate the radiance fields and implicit surface representations in a unified model?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction", "Improving neural implicit surfaces geometry with patch warping", "NeRFusion: Fusing Radiance Fields for Large-Scale Scene Reconstruction", "Neural Sparse Voxel Fields", "DeepVoxels: Learning Persistent 3D Feature Embeddings"], "answer_arxiv_id": ["2106.10689", "2106.12052", "2104.10078", "2112.09648", "2203.11283", "2007.11571", "1812.01024"], "source_meta": {"published_time": "20220826"}, "qid": "AutoScholarQuery_train_15675"} +{"question": "What works introduced a large-scale, fine-grained image classification and detection datasets and emphasized the importance of few-shot learning or self-supervised learning methods?", "answer": ["The iNaturalist Species Classification and Detection Dataset", "Benchmarking Representation Learning for Natural World Image Collections"], "answer_arxiv_id": ["1707.06642", "2103.16483"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_15676"} +{"question": "What paper makes a significant shift toward diversity by introducing MAUVE?", "answer": ["MAUVE: Measuring the Gap Between Neural Text and Human Text using\n Divergence Frontiers"], "answer_arxiv_id": ["2102.01454"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_15677"} +{"question": "Which papers discussed the usage of SDFs to represent open surfaces like garments as thin volumes?", "answer": ["SMPLicit: Topology-aware Generative Model for Clothed People", "DIG: Draping Implicit Garment over the Human Body"], "answer_arxiv_id": ["2103.06871", "2209.10845v2"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_15678"} +{"question": "Which works are related to scientific language models that are able to process chemical inputs?", "answer": ["Galactica: A Large Language Model for Science", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Translation between Molecules and Natural Language", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["2211.09085", "1810.04805", "2204.11817", "1910.10683"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_15679"} +{"question": "What are some works which used the self-distillation method for image representation learning?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2006.07733", "2104.14294"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_15680"} +{"question": "Could you list the research works that employ Fourier neural operators and Picard iterations for parallel decoding?", "answer": ["Fast Sampling of Diffusion Models via Operator Learning", "Parallel Sampling of Diffusion Models"], "answer_arxiv_id": ["2211.13449", "2305.16317"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_15681"} +{"question": "Can you provide examples of prototype-based methods in FSS?", "answer": ["Prototypical Networks for Few-shot Learning", "PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment", "Prior Guided Feature Enrichment Network for Few-Shot Segmentation", "Adaptive Prototype Learning and Allocation for Few-Shot Segmentation", "Learning What Not to Segment: A New Perspective on Few-Shot Segmentation"], "answer_arxiv_id": ["1703.05175", "1908.06391", "2008.01449", "2104.01893", "2203.07615"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_15682"} +{"question": "Can you provide me some papers about text-to-image diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["2112.10752", "2307.01952", "2204.06125", "2205.11487", "2211.01324", "2203.13131"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_15683"} +{"question": "What works deal with the diffusion models designed on Gaussian prior?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Generative Modeling by Estimating Gradients of the Data Distribution", "Improved Techniques for Training Score-Based Generative Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["1503.03585", "2006.11239", "1907.05600", "2006.09011", "2011.13456"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_15684"} +{"question": "Which study shows a contradiction to having a SGD error rate O​(t−ξ) with ξ≥1 in the MNIST's loss evolution?", "answer": ["Last iterate convergence of SGD for Least-Squares in the Interpolation regime"], "answer_arxiv_id": ["2102.03183"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_15685"} +{"question": "Which research covers robust algorithm design in the context of stochastic bandits?", "answer": ["The Intrinsic Robustness of Stochastic Bandits to Strategic Manipulation", "Robust Stochastic Bandit Algorithms under Probabilistic Unbounded Adversarial Attack", "On Optimal Robustness to Adversarial Corruption in Online Decision Problems"], "answer_arxiv_id": ["1906.01528", "2002.07214v1", "2109.10963"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15686"} +{"question": "Any works that perform CFPI by solving a linear approximation in BCPO paradigm?", "answer": ["Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "A Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["1906.00949", "2106.06860"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_15687"} +{"question": "Which model, important for directed graphs in GNNs, used the Magnetic Laplacian within its message passing?", "answer": ["MagNet: A Neural Network for Directed Graphs"], "answer_arxiv_id": ["2102.11391"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_15688"} +{"question": "What works showed a strong correlation between ImageNet accuracy of different models and downstream accuracy on various web-scraped object-centric computer vision benchmark tasks?", "answer": ["Do Better ImageNet Models Transfer Better?"], "answer_arxiv_id": ["1805.08974"], "source_meta": {"published_time": "20230111"}, "qid": "AutoScholarQuery_train_15689"} +{"question": "What are some of the works that involve constructing clusters using interpretable model classes?", "answer": ["Interpretable Clustering using Unsupervised Binary Trees"], "answer_arxiv_id": ["1103.5339"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_15690"} +{"question": "What studies utilize time-conditioned NeRF to represent a 4D scene?", "answer": ["Neural 3D Video Synthesis from Multi-view Video"], "answer_arxiv_id": ["2103.02597"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_15691"} +{"question": "What dataset features radar images from diverse automotive scenarios captured using a Texas Instruments (TI) automotive radar prototype?", "answer": ["RADDet: Range-Azimuth-Doppler based Radar Object Detection for Dynamic\n Road Users"], "answer_arxiv_id": ["2105.00363"], "source_meta": {"published_time": "20240428"}, "qid": "AutoScholarQuery_train_15692"} +{"question": "What paper proposed the popular method of learning with adversarial examples?", "answer": ["Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1412.6572"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_15693"} +{"question": "Any works about end-to-end vectorization methods using supervised learning for persistent homology?", "answer": ["Deep Learning with Topological Signatures", "PersLay: A Neural Network Layer for Persistence Diagrams and New Graph Topological Signatures"], "answer_arxiv_id": ["1707.04041", "1904.09378"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_15694"} +{"question": "Could you tell me about studies that venture into new methodologies, such as adding extra parameters alongside inputs?", "answer": ["Visual Prompt Tuning"], "answer_arxiv_id": ["2203.12119"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_15695"} +{"question": "Which papers introduce the encoder-only and encoder-decoder pretrained language models?", "answer": ["Attention Is All You Need", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "DeBERTa: Decoding-enhanced BERT with Disentangled Attention", "DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with\n Gradient-Disentangled Embedding Sharing", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension"], "answer_arxiv_id": ["1706.03762", "1810.04805", "1907.11692", "2006.03654", "2111.09543", "1910.10683", "1910.13461"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_15696"} +{"question": "Who proposes using additional layers in each transformer block for multi-task fine-tuning?", "answer": ["Parameter-efficient Multi-task Fine-tuning for Transformers via Shared\n Hypernetworks"], "answer_arxiv_id": ["2106.04489"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_15697"} +{"question": "Which papers developed methods utilizing structured matrices to improve model efficiency?", "answer": ["Monarch: Expressive Structured Matrices for Efficient and Accurate Training", "Structured Transforms for Small-Footprint Deep Learning"], "answer_arxiv_id": ["2204.00595", "1510.01722"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_15698"} +{"question": "What studies focus on capturing better spatial and context information in the semantic segmentation?", "answer": ["Rethinking Atrous Convolution for Semantic Image Segmentation", "High-Resolution Representations for Labeling Pixels and Regions"], "answer_arxiv_id": ["1706.05587", "1904.04514"], "source_meta": {"published_time": "20240416"}, "qid": "AutoScholarQuery_train_15699"} +{"question": "Are there any studies that compute the OT maps, over the plans, in generative modeling?", "answer": ["Large-Scale Optimal Transport and Mapping Estimation", "Generative Modeling with Optimal Transport Maps"], "answer_arxiv_id": ["1711.02283", "2110.02999"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_15700"} +{"question": "What papers are about datasets in the field of insect and common crop-weed?", "answer": ["CWD30: A Comprehensive and Holistic Dataset for Crop Weed Recognition in Precision Agriculture"], "answer_arxiv_id": ["2305.10084"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_15701"} +{"question": "Which works primarily discuss detecting and evaluating hallucinations in generated responses?", "answer": ["Siren's Song in the AI Ocean: A Survey on Hallucination in Large\n Language Models", "Evaluating Object Hallucination in Large Vision-Language Models"], "answer_arxiv_id": ["2309.01219", "2305.10355"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_15702"} +{"question": "What papers are related to the design of randomized exploration algorithms with action probabilities in closed forms for sub-Gaussian bandits?", "answer": ["Boltzmann Exploration Done Right"], "answer_arxiv_id": ["1705.10257"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_15703"} +{"question": "What works classify biases in visual datasets into selection bias, framing bias, and label bias?", "answer": ["A Survey on Bias in Visual Datasets"], "answer_arxiv_id": ["2107.07919"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_15704"} +{"question": "Which papers considered unpaired interventional data in the context of causal disentanglement?", "answer": ["Interventional Causal Representation Learning", "Linear Causal Disentanglement via Interventions"], "answer_arxiv_id": ["2209.11924", "2211.16467v3"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_15705"} +{"question": "Which research showcased the performance of ‘map-free’ navigation agents on a variety of tasks?", "answer": ["Habitat: A Platform for Embodied AI Research", "DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion Frames", "RobustNav: Towards Benchmarking Robustness in Embodied Navigation", "Simple but Effective: CLIP Embeddings for Embodied AI", "Is Mapping Necessary for Realistic PointGoal Navigation?", "A Generalist Agent"], "answer_arxiv_id": ["1904.01201", "1911.00357", "2106.04531", "2111.09888", "2206.00997", "2205.06175"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15706"} +{"question": "Which research papers are devoted to learning cyclic dependencies in the underlying graph?", "answer": ["Discovering Cyclic Causal Models by Independent Components Analysis", "Structure Learning for Cyclic Linear Causal Models"], "answer_arxiv_id": ["1206.3273v1", "2006.05978"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_15707"} +{"question": "Can you provide works that used few-shot learning for better training efficiency in the benchmark protocol of stochastic prediction?", "answer": ["Social-STGCNN: A Social Spatio-Temporal Graph Convolutional Neural\n Network for Human Trajectory Prediction"], "answer_arxiv_id": ["2002.11927"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_15708"} +{"question": "What papers demonstrated the benefits of contrastive learning in graph representations?", "answer": ["Graph Contrastive Learning with Adaptive Augmentation", "Graph-less Collaborative Filtering", "Prototypical Graph Contrastive Learning", "AutoGCL: Automated Graph Contrastive Learning via Learnable View Generators"], "answer_arxiv_id": ["2010.14945", "2303.08537v3", "2106.09645", "2109.10259"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_15709"} +{"question": "Which research presented a deep SSM-based model for long sequence modeling tasks?", "answer": ["Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layers"], "answer_arxiv_id": ["2110.13985"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_15710"} +{"question": "What works proposed keypoints detection and overlap attention in fully convolutional methods?", "answer": ["KPConv: Flexible and Deformable Convolution for Point Clouds", "D3Feat: Joint Learning of Dense Detection and Description of 3D Local\n Features", "PREDATOR: Registration of 3D Point Clouds with Low Overlap"], "answer_arxiv_id": ["1904.08889", "2003.03164", "2011.13005"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_15711"} +{"question": "Which papers proposed methods for finding local directions in image manipulation using text in GANs, Diffusion models, and Vision transformers?", "answer": ["StyleMC: Multi-Channel Based Fast Text-Guided Image Generation and Manipulation", "TediGAN: Text-Guided Diverse Face Image Generation and Manipulation", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Blended Diffusion for Text-driven Editing of Natural Images", "More Control for Free! Image Synthesis with Semantic Diffusion Guidance", "MaskGIT: Masked Generative Image Transformer"], "answer_arxiv_id": ["2112.08493", "2012.03308", "2103.17249", "2112.10741", "2111.14818", "2112.05744", "2202.04200"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_15712"} +{"question": "Which work shows a straightforward model could surpass complex ones by training in latent space in layout-to-image generation field?", "answer": ["High-Resolution Complex Scene Synthesis with Transformers"], "answer_arxiv_id": ["2105.06458"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_15713"} +{"question": "Which study suggest minimizing symmetric KL-divergence between the model output of an original input and that of the input perturbed by Gaussian noise?", "answer": ["Better Fine-Tuning by Reducing Representational Collapse"], "answer_arxiv_id": ["2008.03156"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_15714"} +{"question": "What works have used self-supervised mask language modeling training on large dialogues datasets to tackle few-shot intent detection?", "answer": ["Efficient Intent Detection with Dual Sentence Encoders", "DialoGLUE: A Natural Language Understanding Benchmark for Task-Oriented Dialogue", "Example-Driven Intent Prediction with Observers"], "answer_arxiv_id": ["2003.04807", "2009.13570", "2010.08684"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_15715"} +{"question": "Are there any recent studies on the design of stronger canaries for privacy auditing?", "answer": ["Adversary Instantiation: Lower Bounds for Differentially Private Machine Learning", "Debugging Differential Privacy: A Case Study for Privacy Auditing", "Tight Auditing of Differentially Private Machine Learning", "Canife: Crafting Canaries for Empirical Privacy Measurement in Federated Learning"], "answer_arxiv_id": ["2101.04535", "2202.12219", "2302.07956", "2210.02912"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_15716"} +{"question": "Which papers presented empirical results for discontinuities or sharp gradients in PDEs?", "answer": ["Learning Operators with Coupled Attention"], "answer_arxiv_id": ["2201.01032"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_15717"} +{"question": "Which studies estimate expected model changes or predicted loss as sample importance in generic active learning?", "answer": ["Learning Loss for Active Learning"], "answer_arxiv_id": ["1905.03677"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_15718"} +{"question": "Can you mention any paper that proposed generating outlier samples of a novel domain to improve the generalization of the classifier for OOD generalization?", "answer": ["Learning to Generate Novel Domains for Domain Generalization"], "answer_arxiv_id": ["2007.03304"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_15719"} +{"question": "What studies have analyzed regular Transformers as differentiable dictionaries applying powerful associative memory mechanisms?", "answer": ["Large Associative Memory Problem in Neurobiology and Machine Learning", "Hopfield Networks is All You Need"], "answer_arxiv_id": ["2008.06996", "2008.02217"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_15720"} +{"question": "Which works proposed CNN-based binary classifiers for deepfake detection?", "answer": ["MesoNet: a Compact Facial Video Forgery Detection Network", "FaceForensics++: Learning to Detect Manipulated Facial Images"], "answer_arxiv_id": ["1809.00888", "1901.08971"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_15721"} +{"question": "Which studies proposed algorithmic enhancements and optimizations to improve the efficiency of diffusion models?", "answer": ["Denoising Diffusion Implicit Models", "Progressive Distillation for Fast Sampling of Diffusion Models", "On Distillation of Guided Diffusion Models"], "answer_arxiv_id": ["2010.02502", "2202.00512", "2210.03142"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_15722"} +{"question": "What work proposed the Latent Diffusion Model (LDM) using a VAE?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_15723"} +{"question": "What is the line of work that deals with the max-margin analysis of neural networks?", "answer": ["The Implicit Bias of Gradient Descent on Separable Data", "Characterizing Implicit Bias in Terms of Optimization Geometry", "Convergence of Gradient Descent on Separable Data", "Gradient Descent Maximizes the Margin of Homogeneous Neural Networks", "Lexicographic and Depth-Sensitive Margins in Homogeneous and Non-Homogeneous Deep Models", "Implicit Bias of Gradient Descent for Wide Two-layer Neural Networks Trained with the Logistic Loss", "Implicit Bias in Deep Linear Classification: Initialization Scale vs Training Accuracy", "Directional convergence and alignment in deep learning", "Feature selection with gradient descent on two-layer networks in low-rotation regimes", "The Double-Edged Sword of Implicit Bias: Generalization vs. Robustness in ReLU Networks", "Benign Overfitting in Linear Classifiers and Leaky ReLU Networks from KKT Conditions for Margin Maximization", "Gradient Descent on Two-layer Nets: Margin Maximization and Simplicity Bias"], "answer_arxiv_id": ["1710.10345", "1802.08246", "1803.01905", "1906.05890", "1905.07325", "2002.04486", "2007.06738", "2006.06657", "2208.02789", "2303.01456", "2303.01462", "2110.13905"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_15724"} +{"question": "Are there any papers regarding the application of deep image segmentation?", "answer": ["Image Segmentation Using Deep Learning: A Survey"], "answer_arxiv_id": ["2001.05566"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_15725"} +{"question": "Which work used the implicit representation of convolution kernels to model long-range dependencies in sequential data?", "answer": ["CKConv: Continuous Kernel Convolution For Sequential Data"], "answer_arxiv_id": ["2102.02611"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_15726"} +{"question": "Which research papers discussed the use of Laplace approximation in Bayesian deep learning?", "answer": ["Overcoming catastrophic forgetting in neural networks"], "answer_arxiv_id": ["1612.00796"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_15727"} +{"question": "What papers discuss the implementation of the fill-in-the-middle (FIM) training objective and its evaluation on HumanEval infilling tasks?", "answer": ["Efficient Training of Language Models to Fill in the Middle"], "answer_arxiv_id": ["2207.14255"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_15728"} +{"question": "What are some research papers on heatmap-based methods for pose estimation?", "answer": ["Distribution-Aware Coordinate Representation for Human Pose Estimation", "Cascaded Pyramid Network for Multi-Person Pose Estimation", "HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human\n Pose Estimation"], "answer_arxiv_id": ["1910.06278", "1711.07319", "1908.10357"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_15729"} +{"question": "Could you tell me a work which studied graphs in which the local structure depends on the graph size?", "answer": ["From Local Structures to Size Generalization in Graph Neural Networks"], "answer_arxiv_id": ["2010.08853"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_15730"} +{"question": "Which papers have proposed SR benchmarks for 2D images?", "answer": ["Open High-Resolution Satellite Imagery: The WorldStrat Dataset – With Application to Super-Resolution"], "answer_arxiv_id": ["2207.06418"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_15731"} +{"question": "What papers have been following the formulation of decentralized optimization and employ the continuized framework?", "answer": ["Asynchrony and Acceleration in Gossip Algorithms", "A Continuized View on Nesterov Acceleration for Stochastic Gradient Descent and Randomized Gossip"], "answer_arxiv_id": ["2011.02379", "2106.07644"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_15732"} +{"question": "Which papers discuss the use of AI agents based on large pre-trained models?", "answer": ["The Rise and Potential of Large Language Model Based Agents: A Survey", "Foundation Models for Decision Making: Problems, Methods, and\n Opportunities"], "answer_arxiv_id": ["2309.07864v3", "2303.04129"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_15733"} +{"question": "Where could I find more details about neural fields?", "answer": ["Neural Fields in Visual Computing and Beyond"], "answer_arxiv_id": ["2111.11426"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_15734"} +{"question": "What studies describe the use of generative models to directly generate logical forms?", "answer": ["Case-based Reasoning for Natural Language Queries over Knowledge Bases", "RnG-KBQA: Generation Augmented Iterative Ranking for Knowledge Base\n Question Answering", "FC-KBQA: A Fine-to-Coarse Composition Framework for Knowledge Base\n Question Answering", "DecAF: Joint Decoding of Answers and Logical Forms for Question\n Answering over Knowledge Bases"], "answer_arxiv_id": ["2104.08762", "2109.08678", "2306.14722", "2210.00063"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_15735"} +{"question": "Could you provide me some studies about extending NeRFs to dynamic scenes?", "answer": ["Non-Rigid Neural Radiance Fields: Reconstruction and Novel View\n Synthesis of a Dynamic Scene From Monocular Video", "Nerfies: Deformable Neural Radiance Fields", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Space-time Neural Irradiance Fields for Free-Viewpoint Video"], "answer_arxiv_id": ["2012.12247", "2011.12948", "2011.13961", "2106.13228v2", "2011.13084", "2011.12950"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_15736"} +{"question": "What are some works that have extended RGCN's architecture?", "answer": ["Composition-based Multi-Relational Graph Convolutional Networks"], "answer_arxiv_id": ["1911.03082"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_15737"} +{"question": "What works train a NeRF representation of a scene, doing so by optimizing an individual NeRF model per-scene conditioned on an input text?", "answer": ["Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis", "Zero-Shot Text-Guided Object Generation with Dream Fields", "DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2104.00677", "2112.01455", "2209.14988"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_15738"} +{"question": "What research papers focused on the development of AI models aligned with human values?", "answer": ["RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models", "Challenges in Detoxifying Language Models", "Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasets", "On the Opportunities and Risks of Foundation Models", "Generative Language Models and Automated Influence Operations: Emerging Threats and Potential Mitigations"], "answer_arxiv_id": ["2009.11462", "2109.07445", "2106.10328", "2108.07258v3", "2301.04246"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_15739"} +{"question": "Are there any works that study instance-wise performance in the NLP setting?", "answer": ["Are Larger Pretrained Language Models Uniformly Better? Comparing Performance at the Instance Level"], "answer_arxiv_id": ["2105.06020"], "source_meta": {"published_time": "20220220"}, "qid": "AutoScholarQuery_train_15740"} +{"question": "Which works detail the use of score distillation technique in the context of 3D generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2305.16213", "2212.00774v1"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_15741"} +{"question": "Any works about MDP with general function approximation settings?", "answer": ["Online Sub-Sampling for Reinforcement Learning with General Function Approximation"], "answer_arxiv_id": ["2106.07203v2"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_15742"} +{"question": "Which works examined systematic generalization through sequence-to-sequence tasks in language domain?", "answer": ["RobustFill: Neural Program Learning under Noisy I/O", "COGS: A Compositional Generalization Challenge Based on Semantic Interpretation", "Compositional Generalization and Decomposition in Neural Program Synthesis", "Program Induction by Rationale Generation: Learning to Solve and Explain Algebraic Word Problems", "MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalisms", "Measuring Compositional Generalization: A Comprehensive Method on Realistic Data", "Improving Text-to-SQL Evaluation Methodology", "ReCOGS: How Incidental Details of a Logical Form Overshadow an Evaluation of Semantic Interpretation"], "answer_arxiv_id": ["1703.07469", "2010.05465", "2204.03758", "1705.04146", "1905.13319", "1912.09713", "1806.09029", "2303.13716"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_15743"} +{"question": "What studies include the integration of point cloud representation with 2D neural rendering?", "answer": ["Neural Point-Based Graphics", "NPBG++: Accelerating Neural Point-Based Graphics"], "answer_arxiv_id": ["1906.08240", "2203.13318"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15744"} +{"question": "To which works the introduction of Covariance Matrix Adaptation MAP-Annealing (CMA-MAE) is attributed?", "answer": ["Covariance Matrix Adaptation MAP-Annealing", "Covariance Matrix Adaptation for the Rapid Illumination of Behavior Space"], "answer_arxiv_id": ["2205.10752", "1912.02400"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_15745"} +{"question": "Could you provide examples of neural models that have achieved state-of-the-art performance on in-distribution test sets?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "MLP-Mixer: An all-MLP Architecture for Vision", "How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers", "When Vision Transformers Outperform ResNets without Pre-training or Strong Data Augmentations", "Surrogate Gap Minimization Improves Sharpness-Aware Training", "Deep Residual Learning for Image Recognition", "Very Deep Convolutional Networks for Large-Scale Image Recognition", "Going deeper with convolutions", "Densely Connected Convolutional Networks", "Wide Residual Networks"], "answer_arxiv_id": ["2010.11929", "2105.01601", "2106.10270", "2106.01548", "2203.08065", "1512.03385", "1409.1556", "1409.4842", "1608.06993", "1605.07146"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_15746"} +{"question": "Could you provide me works about the online computation of the infinite-dimensional Fisher market equilibrium?", "answer": ["Online Market Equilibrium with Application to Fair Division", "Nonstationary Dual Averaging and Online Fair Allocation"], "answer_arxiv_id": ["2103.12936", "2202.11614"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_15747"} +{"question": "Can you provide references that define curvature exclusively for pairs of nodes, rather than for hyperedges?", "answer": ["A new transport distance and its associated Ricci curvature of hypergraphs", "Coarse Ricci Curvature of Hypergraphs and Its Generalization"], "answer_arxiv_id": ["2105.06710", "2102.00698"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_15748"} +{"question": "Could you mention the papers that proposed the use of a balanced replay buffer and ensembled networks to aid the transition from offline to online in RL?", "answer": ["Offline-to-Online Reinforcement Learning via Balanced Replay and Pessimistic Q-Ensemble"], "answer_arxiv_id": ["2107.00591"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_15749"} +{"question": "Could you provide me some works that discuss the production of human motions that interact with 3D scenes while avoiding collisions?", "answer": ["Synthesizing Diverse Human Motions in 3D Indoor Scenes", "The Wanderings of Odysseus in 3D Scenes", "Diffusion-based Generation, Optimization, and Planning in 3D Scenes"], "answer_arxiv_id": ["2305.12411", "2112.09251", "2301.06015"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_15750"} +{"question": "Are there any studies where contrastive learning is performed at the observation-level?", "answer": ["Time-Series Representation Learning via Temporal and Contextual Contrasting", "TS2Vec: Towards Universal Representation of Time Series"], "answer_arxiv_id": ["2106.14112", "2106.10466"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_15751"} +{"question": "What is an example of a dataset that annotates camera poses and depths with SfM tool COLMAP?", "answer": ["Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction"], "answer_arxiv_id": ["2109.00512v1"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_15752"} +{"question": "Which work presents a detailed analysis on parameter-efficient tuning?", "answer": ["Towards a Unified View of Parameter-Efficient Transfer Learning"], "answer_arxiv_id": ["2110.04366"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_15753"} +{"question": "Who dealt with the challenge of modeling a variable number of primitives?", "answer": ["3D-PRNN: Generating Shape Primitives with Recurrent Neural Networks", "GRASS: Generative Recursive Autoencoders for Shape Structures", "CSGNet: Neural Shape Parser for Constructive Solid Geometry", "3D Point Capsule Networks", "Learning Shape Abstractions by Assembling Volumetric Primitives", "Superquadrics Revisited: Learning 3D Shape Parsing beyond Cuboids"], "answer_arxiv_id": ["1708.01648", "1705.02090", "1712.08290", "1812.10775", "1612.00404", "1904.09970"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_15754"} +{"question": "Which papers studied the use of Periodic functions and grid features in Implicit Neural Representations (INRs)?", "answer": ["Implicit Neural Representations with Periodic Activation Functions", "Modulated Periodic Activations for Generalizable Local Functional Representations", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2006.09661", "2104.03960", "2003.08934", "2111.11215", "2201.05989", "2203.09517"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_15755"} +{"question": "Could you provide me some studies on methods that use gradient ascent to optimize existing designs in offline model-based optimization?", "answer": ["Conservative Objective Models for Effective Offline Model-Based Optimization", "RoMA: Robust Model Adaptation for Offline Model-based Optimization", "Offline Model-Based Optimization via Normalized Maximum Likelihood Estimation", "Bidirectional Learning for Offline Infinite-width Model-based Optimization", "Bidirectional Learning for Offline Model-based Biological Sequence Design"], "answer_arxiv_id": ["2107.06882", "2110.14188", "2102.07970", "2209.07507", "2301.02931"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_15756"} +{"question": "Can you name the studies that add polynomials as features for developing equivariant neural networks?", "answer": ["Universal approximations of invariant maps by neural networks", "Scalars are universal: Equivariant machine learning, structured like classical physics"], "answer_arxiv_id": ["1804.10306", "2106.06610"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15757"} +{"question": "What work proposed the incorporation of peripheral position encoding into the multi-head self-attention layers ?", "answer": ["Peripheral Vision Transformer"], "answer_arxiv_id": ["2206.06801"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_15758"} +{"question": "Which works focus on applying GANs to class-conditioned image generation?", "answer": ["Large Scale GAN Training for High Fidelity Natural Image Synthesis", "A Style-Based Generator Architecture for Generative Adversarial Networks", "Multimodal Conditional Image Synthesis with Product-of-Experts GANs"], "answer_arxiv_id": ["1809.11096", "1812.04948", "2112.05130"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_15759"} +{"question": "Which studies have proposed various algorithms for achieving problem-dependent regret guarantees with benign properties?", "answer": ["Dynamic Regret of Convex and Smooth Functions", "Adaptivity and Non-stationarity: Problem-dependent Dynamic Regret for Online Convex Optimization"], "answer_arxiv_id": ["2007.03479", "2112.14368"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_15760"} +{"question": "Which studies focused on the effect of positional encoding on length generalization?", "answer": ["The Devil is in the Detail: Simple Tricks Improve Systematic Generalization of Transformers", "Making Transformers Solve Compositional Tasks", "Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation"], "answer_arxiv_id": ["2108.12284", "2108.04378v2", "2108.12409"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15761"} +{"question": "What previous works discussed the effectiveness of pre-training on self-supervised tasks for large language and vision models?", "answer": ["Language Models are Few-Shot Learners", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2005.14165", "2111.06377"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_15762"} +{"question": "Which works were inspired by DeepCoder for the integer manipulation domain?", "answer": ["Automatic Program Synthesis of Long Programs with a Learned Garbage Collector", "Learning to Combine Per-Example Solutions for Neural Program Synthesis"], "answer_arxiv_id": ["1809.04682", "2106.07175"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_15763"} +{"question": "Could you provide me some studies that focus on the detection of blending artifacts in deepfake detection?", "answer": ["Exposing DeepFake Videos By Detecting Face Warping Artifacts", "Face X-ray for More General Face Forgery Detection", "Self-supervised Learning of Adversarial Example: Towards Good Generalizations for Deepfake Detection"], "answer_arxiv_id": ["1811.00656", "1912.13458", "2203.12208"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_15764"} +{"question": "Which papers introduced Context-Optimization (CoOp) and feature adapters methods, such as Tip-Adapter, for vision-language model adaptation?", "answer": ["Learning to Prompt for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention"], "answer_arxiv_id": ["2109.01134", "2111.03930", "2304.15010", "2303.16199"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_15765"} +{"question": "What research works focus on creating object-agnostic methods for partitioning an image into coherent regions or their dual representation as contours?", "answer": ["Fast Edge Detection Using Structured Forests", "High-for-Low and Low-for-High: Efficient Boundary Detection from Deep\n Object Features and its Applications to High-Level Vision", "Pushing the Boundaries of Boundary Detection using Deep Learning"], "answer_arxiv_id": ["1406.5549", "1504.06201", "1511.07386"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_15766"} +{"question": "Are there any studies that introduced a global model momentum to stabilize FL training?", "answer": ["No Fear of Heterogeneity: Classifier Calibration for Federated Learning with Non-IID Data", "Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classification"], "answer_arxiv_id": ["2106.05001", "1909.06335"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15767"} +{"question": "Which works introduced additional components for training based on the foundation model - addition-based methods?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "BERT and PALs: Projected Attention Layers for Efficient Adaptation in\n Multi-Task Learning", "Compacter: Efficient Low-Rank Hypercomplex Adapter Layers", "AdapterDrop: On the Efficiency of Adapters in Transformers", "Language Models are Few-Shot Learners", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Multitask Prompt Tuning Enables Parameter-Efficient Transfer Learning"], "answer_arxiv_id": ["1902.00751", "1902.02671", "2106.04647", "2010.11918", "2005.14165", "2104.08691", "2101.00190", "2303.02861"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_15768"} +{"question": "What study introduces MORBO, a method that performs BO in parallel on multiple local regions of the candidate space?", "answer": ["Multi-Objective Bayesian Optimization over High-Dimensional Search Spaces"], "answer_arxiv_id": ["2109.10964"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_15769"} +{"question": "Which papers develop algorithms for online reinforcement learning with additional access to offline data?", "answer": ["Exploration-Enhanced Politex", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning"], "answer_arxiv_id": ["1908.10479", "2106.04895"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_15770"} +{"question": "Which works consider alternate regret notions or behavior models for repeated Stackelberg games?", "answer": ["Learning in Stackelberg Games with Non-myopic Agents"], "answer_arxiv_id": ["2208.09407"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15771"} +{"question": "Which papers proposed the use of paraphrasing and retokenization as defenses against optimization-based attacks?", "answer": ["Baseline Defenses for Adversarial Attacks Against Aligned Language\n Models"], "answer_arxiv_id": ["2309.00614"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_15772"} +{"question": "Which research papers discussed image-text feature alignment in LMMs?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2301.12597", "2304.08485", "2305.06500", "2304.10592"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_15773"} +{"question": "Could you provide studies that use a two-pass architecture for generating target text and target speech in S2ST?", "answer": ["UnitY: Two-pass Direct Speech-to-speech Translation with Discrete Units", "Translatotron 2: High-quality direct speech-to-speech translation with voice preservation", "Speech-to-Speech Translation For A Real-world Unwritten Language"], "answer_arxiv_id": ["2212.08055", "2107.08661", "2211.06474"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_15774"} +{"question": "Could you provide me some studies about the importance of large step sizes for generalization?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "A Bayesian Perspective on Generalization and Stochastic Gradient Descent", "On the Origin of Implicit Regularization in Stochastic Gradient Descent", "Stochastic gradient descent introduces an effective landscape-dependent regularization favoring flat solutions"], "answer_arxiv_id": ["1609.04836", "1710.06451", "2101.12176", "2206.01246"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15775"} +{"question": "Which works studied the least-square TD learning rule?", "answer": ["Accelerated Gradient Temporal Difference Learning"], "answer_arxiv_id": ["1611.09328"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_15776"} +{"question": "Which studies introduced AI-assistant agents using supervised fine-tuning and reinforcement learning from human feedback on user queries?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_15777"} +{"question": "Which studies discuss extending the coverage of transfer via suitable pre-trainings like meta-learning?", "answer": ["Meta-Learning with Differentiable Convex Optimization", "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Patching open-vocabulary modelsby interpolating weights"], "answer_arxiv_id": ["1904.03758", "1703.03400", "2208.05592"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_15778"} +{"question": "Who used Stein’s unbiased risk estimator for training unsupervised denoisers?", "answer": ["Unsupervised Learning with Stein’s Unbiased Risk Estimator", "Extending Stein’s unbiased risk estimator to train deep denoisers with correlated pairs of noisy images", "Adaptive Denoising via GainTuning"], "answer_arxiv_id": ["1805.10531", "1902.02452", "2107.12815"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15779"} +{"question": "Are there any papers that present parameter-efficient fine-tuning strategies to reduce the cost of fine-tuning?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["1902.00751", "2106.09685", "2303.16199", "2101.00190", "2104.08691"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_15780"} +{"question": "What works recently found that Transformers are more robust than CNNs on out-of-distribution samples?", "answer": ["Are Transformers More Robust Than CNNs?", "Delving Deep into the Generalization of Vision Transformers under Distribution Shifts", "Vision Transformers are Robust Learners"], "answer_arxiv_id": ["2111.05464", "2106.07617", "2105.07581"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_15781"} +{"question": "Which researchers performed an analysis of computational and memory complexity for computing the eNTK?", "answer": ["Fast Finite Width Neural Tangent Kernel"], "answer_arxiv_id": ["2206.08720"], "source_meta": {"published_time": "20220625"}, "qid": "AutoScholarQuery_train_15782"} +{"question": "Which papers have demonstrated the capabilities of Language Literacy Models in handling various tasks without further tuning?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "GPT-4 Technical Report"], "answer_arxiv_id": ["2005.14165", "2204.02311", "2303.08774"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_15783"} +{"question": "What works proposed ways to accelerate the training of diffusion models or reduce the number of sampling steps?", "answer": ["Denoising Diffusion Implicit Models", "Progressive Distillation for Fast Sampling of Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Autoregressive Diffusion Models", "WaveGrad: Estimating Gradients for Waveform Generation", "On Fast Sampling of Diffusion Probabilistic Models", "Tackling the Generative Learning Trilemma with Denoising Diffusion GANs", "Elucidating the Design Space of Diffusion-Based Generative Models", "Poisson Flow Generative Models"], "answer_arxiv_id": ["2010.02502", "2202.00512", "2112.10752", "2110.02037", "2009.00713v2", "2106.00132", "2112.07804", "2206.00364", "2209.11178"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_15784"} +{"question": "What are the research papers that involve measuring the uncertainty of the Q-functions for sampling curriculum goals?", "answer": ["Automatic Curriculum Learning through Value Disagreement"], "answer_arxiv_id": ["2006.09641"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15785"} +{"question": "What were the papers that addressed the phenomenon of Grokking in NLP tasks?", "answer": ["Grokking: Generalization Beyond Overfitting on Small Algorithmic\n Datasets"], "answer_arxiv_id": ["2201.02177"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_15786"} +{"question": "Are there any works that have made attempts to modify sampling using neural networks or occupancy caching in the context of NeRF?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction", "Improved Direct Voxel Grid Optimization for Radiance Fields Reconstruction", "Neural Sparse Voxel Fields"], "answer_arxiv_id": ["2102.13090", "2111.11215", "2206.05085", "2007.11571"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_15787"} +{"question": "Could you provide me some studies about prompt tuning method for long-tailed image classification?", "answer": ["Learning to Prompt for Vision-Language Models", "Visual Prompt Tuning", "Exploring Visual Prompts for Adapting Large-Scale Models"], "answer_arxiv_id": ["2109.01134", "2203.12119", "2203.17274"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_15788"} +{"question": "What researches illustrated how neural networks in the field of medical imaging interpret hospital-specific tokens or incidental cues instead of actual disease symptoms?", "answer": ["Hidden Stratification Causes Clinically Meaningful Failures in Machine\n Learning for Medical Imaging"], "answer_arxiv_id": ["1909.12475"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_15789"} +{"question": "Which works discuss challenges in using Bayesian Neural Networks (BNNs) for Quantifying predictive uncertainty in deep learning models?", "answer": ["Adapting the Linearised Laplace Model Evidence for Modern Deep Learning"], "answer_arxiv_id": ["2206.08900"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_15790"} +{"question": "What studies showed the connections between GNNs and Weisfeiler–Leman type algorithms?", "answer": ["Let’s Agree to Degree: Comparing Graph Convolutional Networks in the Message-Passing Framework", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["2004.02593", "1810.02244", "1810.00826"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_15791"} +{"question": "Which work presented the HRNet for semantic segmentation?", "answer": ["High-Resolution Representations for Labeling Pixels and Regions"], "answer_arxiv_id": ["1904.04514"], "source_meta": {"published_time": "20240416"}, "qid": "AutoScholarQuery_train_15792"} +{"question": "Are there any works showcasing emergent analogical structure in language models?", "answer": ["Efficient Estimation of Word Representations in Vector Space", "Emergent Analogical Reasoning in Large Language Models"], "answer_arxiv_id": ["1301.3781", "2212.09196v3"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_15793"} +{"question": "What are some of the many tensor decomposition methods that have been developed?", "answer": ["Bayesian Conditional Tensor Factorizations for High-Dimensional Classification", "Distributed Flexible Nonlinear Tensor Factorization", "Nonparametric Sparse Tensor Factorization with Hierarchical Gamma Processes", "Nonparametric Embeddings of Sparse High-Order Interaction Events"], "answer_arxiv_id": ["1301.4950", "1604.07928", "2110.10082", "2207.03639"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_15794"} +{"question": "What studies proposed methods on improving the text generation process, such as addressing issues of low-quality samples and exposure bias?", "answer": ["Text Generation by Learning from Demonstrations", "Step-unrolled Denoising Autoencoders for Text Generation", "RankGen: Improving Text Generation with Large Ranking Models"], "answer_arxiv_id": ["2009.07839", "2112.06749", "2205.09726"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_15795"} +{"question": "Could you provide me a study which shows that training provably enters the edge of stability with modified gradient descent or modified loss, and then its associated flow goes to flat regions?", "answer": ["Understanding Gradient Descent on Edge of Stability in Deep Learning"], "answer_arxiv_id": ["2205.09745"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_15796"} +{"question": "Which of the referenced works involve the use of quantization techniques for training acceleration?", "answer": ["FracTrain: Fractionally Squeezing Bit Savings Both Temporally and\n Spatially for Efficient DNN Training", "E2-Train: Training State-of-the-art CNNs with Over 80% Energy Savings"], "answer_arxiv_id": ["2012.13113", "1910.13349"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_15797"} +{"question": "Which works propose memory-based methods that approximate Bayes-optimal agents in the field of meta RL?", "answer": ["Learning to reinforcement learn", "Meta-learning of Sequential Strategies"], "answer_arxiv_id": ["1611.05763", "1905.03030"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_15798"} +{"question": "Which papers explored rule-based foils for language-based detection?", "answer": ["Discriminative Learning of Open-Vocabulary Object Retrieval and\n Localization by Negative Phrase Augmentation"], "answer_arxiv_id": ["1711.09509"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_15799"} +{"question": "Which papers discussed the creation of synthetic fog datasets by rendering fog onto clean images, using depth information?", "answer": ["Benchmarking Single Image Dehazing and Beyond", "Semantic Foggy Scene Understanding with Synthetic Data", "Semantic Understanding of Foggy Scenes with Purely Synthetic Data", "Realistic Large-Scale Fine-Depth Dehazing Dataset from 3D Videos"], "answer_arxiv_id": ["1712.04143", "1708.07819", "1910.03997", "2004.08554"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_15800"} +{"question": "Can you name studies that propose methods to generalize representations to unseen nodes?", "answer": ["LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation", "Graph Convolutional Matrix Completion", "Graph Convolutional Neural Networks for Web-Scale Recommender Systems", "Inductive Matrix Completion Based on Graph Neural Networks", "Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach"], "answer_arxiv_id": ["2002.02126", "1706.02263", "1806.01973", "1904.12058", "2007.04833"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_15801"} +{"question": "Which studies employed a Vision Transformer to the masked autoencoder?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "PeCo: Perceptual Codebook for BERT Pre-training of Vision Transformers", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: A Simple Framework for Masked Image Modeling"], "answer_arxiv_id": ["2106.08254", "2111.12710", "2112.09133", "2111.06377", "2111.09886"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_15802"} +{"question": "Could you list the works that offer a theoretical understanding of KD, primarily related to linear classifier/nets or neural tangent kernel (NTK) analysis?", "answer": ["Towards Understanding Knowledge Distillation", "Knowledge Distillation in Wide Neural Networks: Risk Bound, Data Efficiency and Imperfect Teacher", "Supervision Complexity and its Role in Knowledge Distillation"], "answer_arxiv_id": ["2105.13093", "2010.10090", "2301.12245"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_15803"} +{"question": "What are some studies that have discussed primitive fitting to point clouds in 3D shapes?", "answer": ["Supervised Fitting of Geometric Primitives to 3D Point Clouds", "CPFN: Cascaded Primitive Fitting Networks for High-Resolution Point Clouds"], "answer_arxiv_id": ["1811.08988", "2109.00113"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_15804"} +{"question": "Can you cite an example of research where the implicit representation of convolution kernels was utilized for signal representation?", "answer": ["Implicit Neural Representations with Periodic Activation Functions"], "answer_arxiv_id": ["2006.09661"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_15805"} +{"question": "Could you provide me with some works regarding the use of GANs in creating Deep Tabular Generators?", "answer": ["Modeling Tabular Data using Conditional GAN", "CTAB-GAN+: Enhancing Tabular Data Synthesis", "Conditional Wasserstein GAN-based Oversampling of Tabular Data for Imbalanced Learning"], "answer_arxiv_id": ["1907.00503", "2204.00401", "2008.09202"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_15806"} +{"question": "Which works show that Physics-informed neural networks (PINNs) can accurately solve PDEs?", "answer": ["Implicit Neural Representations with Periodic Activation Functions"], "answer_arxiv_id": ["2006.09661"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_15807"} +{"question": "What works are related to adopting a language model as an index for document storage and retrieval?", "answer": ["Rethinking Search: Making Domain Experts out of Dilettantes", "Transformer Memory as a Differentiable Search Index", "A Neural Corpus Indexer for Document Retrieval", "Scalable and Effective Generative Information Retrieval"], "answer_arxiv_id": ["2105.02274", "2202.06991", "2206.02743", "2311.09134"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_15808"} +{"question": "Which studies keep the model size small at the expense of higher computation and latency?", "answer": ["Lite Vision Transformer with Enhanced Self-Attention", "Separable Self-attention for Mobile Vision Transformers", "EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers"], "answer_arxiv_id": ["2112.10809", "2206.02680", "2205.03436"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15809"} +{"question": "What works focus on employing inter-task interpolation to mitigate task scarcity in few-task datasets?", "answer": ["Meta-Learning with Fewer Tasks through Task Interpolation"], "answer_arxiv_id": ["2106.02695"], "source_meta": {"published_time": "20220407"}, "qid": "AutoScholarQuery_train_15810"} +{"question": "What works proposed improvements to Adam optimizers based on the large variance in the early stage?", "answer": ["Attention Is All You Need", "Training Tips for the Transformer Model", "On the Variance of the Adaptive Learning Rate and Beyond"], "answer_arxiv_id": ["1706.03762", "1804.00247", "1908.03265"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_15811"} +{"question": "What researches utilize encoder-only masked language models in training large-scale neural language models on source code?", "answer": ["CodeBERT: A Pre-Trained Model for Programming and Natural Languages", "Learning and Evaluating Contextual Embedding of Source Code"], "answer_arxiv_id": ["2002.08155", "2001.00059"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_15812"} +{"question": "Which paper showed that the robust dataset may present a false sense of robustness under AutoAttack?", "answer": ["Adversarial Examples Are Not Real Features"], "answer_arxiv_id": ["2310.18936"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15813"} +{"question": "What papers on expectation-maximization considered non-Gaussian noise?", "answer": ["Efficient Algorithms for Estimating the Parameters of Mixed Linear Regression Models"], "answer_arxiv_id": ["2105.05953"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_15814"} +{"question": "What papers have focused on finding prototypical examples that represent a larger corpus of data?", "answer": ["Prototype selection for interpretable classification", "The Bayesian Case Model: A Generative Approach for Case-Based Reasoning and Prototype Classification"], "answer_arxiv_id": ["1202.5933", "1503.01161"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_15815"} +{"question": "Which studies are related to performing statistical inference on the value function to aid in OMS?", "answer": ["Statistical Inference of the Value Function for Reinforcement Learning in Infinite-Horizon Settings"], "answer_arxiv_id": ["2001.04515"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_15816"} +{"question": "Could you provide some studies about diffusion models being used in generating images from text prompts (T2I)?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2205.11487", "2111.14822", "2112.10752"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15817"} +{"question": "Which studies have focused on decision boundaries in deep classifiers through analysing small margins in adversarial directions?", "answer": ["Characterizing the Decision Boundary of Deep Neural Networks"], "answer_arxiv_id": ["1912.11460"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_15818"} +{"question": "What work produced a glossary of over 300 dog whistles used in settings?", "answer": ["From Dogwhistles to Bullhorns: Unveiling Coded Rhetoric with Language\n Models"], "answer_arxiv_id": ["2305.17174"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_15819"} +{"question": "Could you provide me some studies in video grounding where models decode moment boundaries from global features?", "answer": ["To Find Where You Talk: Temporal Sentence Localization in Video with\n Attention Based Location Regression", "Local-Global Video-Text Interactions for Temporal Grounding", "Embracing Uncertainty: Decoupling and De-bias for Robust Temporal\n Grounding", "Compositional Temporal Grounding with Structured Variational Cross-Graph\n Correspondence Learning"], "answer_arxiv_id": ["1804.07014", "2004.07514", "2103.16848", "2203.13049"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_15820"} +{"question": "Could you name the works that propose maintaining an agent's memory externally or integrating it directly into the feature extraction pipeline in RL?", "answer": ["Grounded Language Learning Fast and Slow", "Unsupervised Predictive Memory in a Goal-Directed Agent"], "answer_arxiv_id": ["2009.01719", "1803.10760"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_15821"} +{"question": "Which datasets incorporate LiDAR signal modality that is similar to vision?", "answer": ["LiDARCap: Long-range Marker-less 3D Human Motion Capture with LiDAR\n Point Clouds", "SLOPER4D: A Scene-Aware Dataset for Global 4D Human Pose Estimation in\n Urban Environments", "CIMI4D: A Large Multimodal Climbing Motion Dataset under Human-scene\n Interactions"], "answer_arxiv_id": ["2203.14698", "2303.09095", "2303.17948"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_15822"} +{"question": "Which works have performed editing of factual knowledge directly in the weights of pre-trained Transformers?", "answer": ["Mass-Editing Memory in a Transformer", "Fast Model Editing at Scale", "Editing Factual Knowledge in Language Models"], "answer_arxiv_id": ["2210.07229", "2110.11309", "2104.08164"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15823"} +{"question": "Could you provide me with studies that focus on learning Bayesian priors for large-scale machine learning models?", "answer": ["Learning Approximately Objective Priors", "Predictive Complexity Priors", "The Deep Weight Prior"], "answer_arxiv_id": ["1704.01168", "2006.10801v3", "1810.06943"], "source_meta": {"published_time": "20220717"}, "qid": "AutoScholarQuery_train_15824"} +{"question": "Which papers propose methods for training retrievers and readers in retrieval-augmented language models?", "answer": ["Dense Passage Retrieval for Open-Domain Question Answering"], "answer_arxiv_id": ["2004.04906"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15825"} +{"question": "Which works discuss key factors such as reduced inference overhead, memory efficiency, and storage optimization in parameter-efficient fine-tuning (PEFT)?", "answer": ["Learning multiple visual domains with residual adapters", "Parameter-Efficient Transfer Learning for NLP", "AdapterFusion: Non-Destructive Task Composition for Transfer Learning", "AdapterDrop: On the Efficiency of Adapters in Transformers", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["1705.08045", "1902.00751", "2005.00247", "2010.11918", "2101.00190", "2104.08691", "2106.09685"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_15826"} +{"question": "What research used no-regret learning to address the non-correspondence problem in two-player zero-sum games?", "answer": ["Combining Deep Reinforcement Learning and Search for Imperfect-Information Games"], "answer_arxiv_id": ["2007.13544"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_15827"} +{"question": "Which papers in the field of Generative Masked Representation Learning proposed to reconstruct online teacher tokens or HOG features?", "answer": ["iBOT : Image BERT Pre-Training with Online Tokenizer", "Masked Feature Prediction for Self-Supervised Visual Pre-Training"], "answer_arxiv_id": ["2111.07832", "2112.09133"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_15828"} +{"question": "Are there previous works that considered learning in repeated first-price auctions with binary feedback?", "answer": ["Contextual Bandits with Cross-learning"], "answer_arxiv_id": ["1809.09582v3"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_15829"} +{"question": "What studies proposed the LSVI-UCB and LSVI-UCB++ algorithms?", "answer": ["Provably Efficient Reinforcement Learning with Linear Function Approximation", "Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes"], "answer_arxiv_id": ["1907.05388", "2212.06132"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_15830"} +{"question": "Can you provide research works that connect prompting and CL?", "answer": ["DualPrompt: Complementary Prompting for Rehearsal-free Continual Learning", "Learning to Prompt for Continual Learning"], "answer_arxiv_id": ["2204.04799", "2112.08654"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_15831"} +{"question": "Are there any works that adapted ToM through Bayesian approaches?", "answer": ["Learning Triadic Belief Dynamics in Nonverbal Communication from Videos"], "answer_arxiv_id": ["2104.02841"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_15832"} +{"question": "What are some of the frameworks in which diffusion models have become the gold standard for controllable image generation?", "answer": ["Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2102.12092", "2205.11487", "2302.08453", "2302.05543"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_15833"} +{"question": "Which works deal with the study of context in learning representations for language models?", "answer": ["Distributed Representations of Words and Phrases and their Compositionality", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1310.4546", "1810.04805"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_15834"} +{"question": "Which works are related to public-data-assisted methods that leverage public data for saving privacy budgets?", "answer": ["Leveraging Public Data for Practical Private Query Release", "Iterative Methods for Private Synthetic Data: Unifying Framework and New Methods"], "answer_arxiv_id": ["2102.08598v2", "2106.07153"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15835"} +{"question": "Any works that modeled a calorimeter response for cosmic ray-induced air showers in the Earth’s atmosphere?", "answer": ["Generating and refining particle detector simulations using the Wasserstein distance in adversarial networks"], "answer_arxiv_id": ["1802.03325"], "source_meta": {"published_time": "20220210"}, "qid": "AutoScholarQuery_train_15836"} +{"question": "Could you provide me some studies that utilised dynamic sparse training methods to promote the prunability of the network?", "answer": ["Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science", "Sparse Networks from Scratch: Faster Training without Losing Performance", "Rigging the Lottery: Making All Tickets Winners"], "answer_arxiv_id": ["1707.04780v2", "1907.04840", "1911.11134"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_15837"} +{"question": "Could you provide me some studies about the collection of single-person hand-object interaction datasets?", "answer": ["FreiHAND: A Dataset for Markerless Capture of Hand Pose and Shape from Single RGB Images", "ContactPose: A Dataset of Grasps with Object Contact and Hand Pose", "GRAB: A Dataset of Whole-Body Human Grasping of Objects", "DexYCB: A Benchmark for Capturing Hand Grasping of Objects", "BEHAVE: Dataset and Method for Tracking Human Object Interactions", "InterCap: Joint Markerless 3D Tracking of Humans and Objects in Interaction", "HOnnotate: A method for 3D Annotation of Hand and Object Poses", "H2O: Two Hands Manipulating Objects for First Person Interaction Recognition", "Keypoint Transformer: Solving Joint Identification in Challenging Hands and Object Interactions for Accurate 3D Pose Estimation", "Assembly101: A Large-Scale Multi-View Video Dataset for Understanding Procedural Activities", "ARCTIC: A Dataset for Dexterous Bimanual Hand-Object Manipulation"], "answer_arxiv_id": ["1909.04349", "2007.09545", "2008.11200", "2104.04631", "2204.06950", "2209.12354", "1907.01481", "2104.11181", "2104.14639", "2203.14712v2", "2204.13662"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_15838"} +{"question": "What study introduced the MovieGraphs dataset?", "answer": ["MovieGraphs: Towards Understanding Human-Centric Situations from Videos"], "answer_arxiv_id": ["1712.06761"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_15839"} +{"question": "What research paper presented the TEACH method?", "answer": ["TEACH: Temporal Action Composition for 3D Humans"], "answer_arxiv_id": ["2209.04066"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_15840"} +{"question": "Which works show that REALM and DPR can be used for open-domain question answering?", "answer": ["REALM: Retrieval-Augmented Language Model Pre-Training", "Dense Passage Retrieval for Open-Domain Question Answering"], "answer_arxiv_id": ["2002.08909", "2004.04906"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_15841"} +{"question": "Which studies introduced efficient ways to collect gaze direction data from panoramic cameras?", "answer": ["Gaze360: Physically Unconstrained Gaze Estimation in the Wild"], "answer_arxiv_id": ["1910.10088"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15842"} +{"question": "What projects currently exist that integrates a memory network into autoencoder architecture?", "answer": ["Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detection"], "answer_arxiv_id": ["1904.02639"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_15843"} +{"question": "Any research papers tried to improve the TS method?", "answer": ["Post-hoc Uncertainty Calibration for Domain Drift Scenarios"], "answer_arxiv_id": ["2012.10988"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_15844"} +{"question": "What works had to solve large-scale optimization problems for synthetic data generation?", "answer": ["Differentially Private Query Release Through Adaptive Projection"], "answer_arxiv_id": ["2103.06641"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15845"} +{"question": "Could you provide me the works that conducted research on specific topological properties of deep networks that can indicate their test accuracy?", "answer": ["FLASH: Fast Neural Architecture Search with Hardware Optimization"], "answer_arxiv_id": ["2108.00568"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_15846"} +{"question": "Which work combined Dividemix with self-supervised pre-training to boost its performance by improving the quality of the extracted features?", "answer": ["Contrast to Divide: Self-Supervised Pre-Training for Learning with Noisy Labels"], "answer_arxiv_id": ["2103.13646"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_15847"} +{"question": "What work proposed a method where Graph Neural Prompt was used to extract valuable knowledge from KGs for integration into pre-trained LLMs?", "answer": ["Graph Neural Prompting with Large Language Models"], "answer_arxiv_id": ["2309.15427"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_15848"} +{"question": "Which benchmark consists of tests from major domains such as medicine, law, psychology and education?", "answer": ["Measuring Massive Multitask Chinese Understanding"], "answer_arxiv_id": ["2304.12986"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_15849"} +{"question": "Which papers use masked modeling to predict the masked words of a sentence?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_15850"} +{"question": "Which works proposed approaches to address the Open Catalyst challenge?", "answer": ["Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs", "Simple GNN Regularisation for 3D Molecular Property Prediction & Beyond", "GemNet: Universal Directional Graph Neural Networks for Molecules"], "answer_arxiv_id": ["2206.11990", "2106.07971", "2106.08903v10"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_15851"} +{"question": "Which studies define a pretext task useful for downstream tasks in the field of self-supervised learning?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction", "Unsupervised Learning of Visual Representations using Videos", "Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles", "Colorful Image Colorization", "Learning Features by Watching Objects Move", "Unsupervised Representation Learning by Predicting Image Rotations", "Masked Autoencoders Are Scalable Vision Learners", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["1505.05192", "1505.00687", "1603.09246", "1603.08511", "1612.06370", "1803.07728", "2111.06377", "2006.07733"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_15852"} +{"question": "Which studies observed a tight analysis in the context of federated learning?", "answer": ["Advances and Open Problems in Federated Learning", "On the Linear Speedup Analysis of Communication Efficient Momentum SGD for Distributed Non-Convex Optimization", "Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning", "Tighter Theory for Local SGD on Identical and Heterogeneous Data", "Minibatch vs Local SGD for Heterogeneous Distributed Learning", "Is Local SGD Better than Minibatch SGD?", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Local SGD with Periodic Averaging: Tighter Analysis and Adaptive Synchronization"], "answer_arxiv_id": ["1912.04977", "1905.03817", "1807.06629", "1909.04746v4", "2006.04735", "2002.07839", "1910.06378", "1910.13598"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_15853"} +{"question": "Which works have explored how natural language descriptions can be used to improve performance and generalization in few-shot or domain adaptation settings?", "answer": ["Using Language to Extend to Unseen Domains", "Is synthetic data from generative models ready for image recognition?", "On Guiding Visual Attention with Language Specification"], "answer_arxiv_id": ["2210.09520", "2210.07574", "2202.08926"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15854"} +{"question": "Could you name the papers that focus on aleatoric uncertainty in image segmentation?", "answer": ["What Uncertainties Do We Need in Bayesian Deep Learning for Computer\n Vision?"], "answer_arxiv_id": ["1703.04977"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_15855"} +{"question": "Which research posed that the self-attention matrix 'P' is close to a low rank matrix, inspiring Linformers and the fine-tuning algorithm LoRA?", "answer": ["A Survey of Transformers"], "answer_arxiv_id": ["2106.04554"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_15856"} +{"question": "What work proves that neural networks have simplicity bias and favor the linear function in early training stages?", "answer": ["The Pitfalls of Simplicity Bias in Neural Networks"], "answer_arxiv_id": ["2006.07710"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_15857"} +{"question": "Are there any benchmarks that include tasks in text, image, and math, designed for scalability?", "answer": ["Long Range Arena: A Benchmark for Efficient Transformers"], "answer_arxiv_id": ["2011.04006"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_15858"} +{"question": "What works have demonstrated LLMs' learning capability from the context samples in a M-ICL setting?", "answer": ["A Survey on In-context Learning", "MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action", "Chameleon: Plug-and-Play Compositional Reasoning with Large Language\n Models", "Visual Programming: Compositional visual reasoning without training"], "answer_arxiv_id": ["2301.00234", "2303.11381", "2304.09842", "2211.11559"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_15859"} +{"question": "What studies have focused on finding discriminative embeddings of measures to be compared?", "answer": ["Learning Generative Models with Sinkhorn Divergences", "Amortized Projection Optimization for Sliced Wasserstein Generative Models"], "answer_arxiv_id": ["1706.00292v3", "2203.13417"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_15860"} +{"question": "What research papers utilized attention and graph neural networks for ICD coding?", "answer": ["ICD Coding from Clinical Text Using Multi-Filter Residual Convolutional Neural Network", "Code Synonyms Do Matter: Multiple Synonyms Matching Network for Automatic ICD Coding"], "answer_arxiv_id": ["1912.00862", "2203.01515"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_15861"} +{"question": "Any works about balancing modalities by modulating gradients or adjusting modality losses in the context of multi-modal FAS?", "answer": ["Balanced Multimodal Learning via On-the-fly Gradient Modulation", "PMR: Prototypical Modal Rebalance for Multimodal Learning"], "answer_arxiv_id": ["2203.15332", "2211.07089"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_15862"} +{"question": "Can you mention some papers where LLMs have been applied as agents for vision tasks?", "answer": ["An Empirical Study of GPT-3 for Few-Shot Knowledge-Based VQA", "PromptCap: Prompt-Guided Task-Aware Image Captioning", "MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action", "AVIS: Autonomous Visual Information Seeking with Large Language Model\n Agent"], "answer_arxiv_id": ["2109.05014", "2211.09699", "2303.11381", "2306.08129"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_15863"} +{"question": "Are there any papers discussing the the use of LLMs for generating inferential rules?", "answer": ["Large Language Models can Learn Rules"], "answer_arxiv_id": ["2310.07064v2"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_15864"} +{"question": "Which works propose joint optimization of motions by back-propagating the gradient of the image reconstruction loss to the pose parameters in human reconstruction from monocular videos?", "answer": ["Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via\n Self-supervised Scene Decomposition", "InstantAvatar: Learning Avatars from Monocular Video in 60 Seconds"], "answer_arxiv_id": ["2302.11566", "2212.10550"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15865"} +{"question": "Which research suggests the importance of mixed-sign high-pass vs low-pass filters for the spectral GPRGNN?", "answer": ["Adaptive Universal Generalized PageRank Graph Neural Network"], "answer_arxiv_id": ["2006.07988"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_15866"} +{"question": "Which papers assert that current LLMs’ probabilistic predictions correspond well with actual token occurrence frequencies?", "answer": ["Scaling Language Models: Methods, Analysis & Insights from Training Gopher", "Language Models (Mostly) Know What They Know", "On Calibration of Modern Neural Networks", "How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answering", "Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generation"], "answer_arxiv_id": ["2112.11446", "2207.05221", "1706.04599", "2012.00955", "2302.09664"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_15867"} +{"question": "What papers propose single-modal methods for Semantic Scene Completion that take only TSDF?", "answer": ["Semantic Scene Completion from a Single Depth Image", "EdgeNet: Semantic Scene Completion from a Single RGB-D Image"], "answer_arxiv_id": ["1611.08974", "1908.02893"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_15868"} +{"question": "Which works focused on black-box certificates for multi-output classifiers?", "answer": ["Collective Robustness Certificates: Exploiting Interdependence in Graph Neural Networks", "Scalable Certified Segmentation via Randomized Smoothing", "On Collective Robustness of Bagging Against Data Poisoning", "Certified Defences Against Adversarial Patch Attacks on Semantic Segmentation"], "answer_arxiv_id": ["2302.02829", "2107.00228", "2205.13176", "2209.05980"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_15869"} +{"question": "Can you provide papers about the application and improvement of Diffusion Models?", "answer": ["Elucidating the Design Space of Diffusion-Based Generative Models", "Classifier-Free Diffusion Guidance", "Progressive Distillation for Fast Sampling of Diffusion Models", "Blended Diffusion for Text-driven Editing of Natural Images", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Denoising Diffusion Restoration Models", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic\n Models", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2206.00364v2", "2207.12598", "2202.00512", "2111.14818", "2211.09794", "2208.01618", "2201.11793", "2104.14951", "2201.09865", "2108.01073", "2208.12242"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_15870"} +{"question": "Which research paper acknowledged variations in the contents of a datasheet based on different research circumstances?", "answer": ["Datasheets for Datasets"], "answer_arxiv_id": ["1803.09010"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_15871"} +{"question": "Could you provide me some research papers that focus on a warm-started max flow algorithm?", "answer": ["Learning-Augmented Maximum Flow"], "answer_arxiv_id": ["2207.12911"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_15872"} +{"question": "Which works discussed deep mutual learning scheme as a regularization for the seq2seq model?", "answer": ["Deep Mutual Learning"], "answer_arxiv_id": ["1706.00384"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_15873"} +{"question": "Any studies about synthesizing data for human body pose estimation?", "answer": ["Synthesizing Training Images for Boosting Human 3D Pose Estimation", "Learning from Synthetic Humans", "Sim2real transfer learning for 3D human pose estimation: motion to the rescue"], "answer_arxiv_id": ["1604.02703", "1701.01370", "1907.02499"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_15874"} +{"question": "What are some networks used for estimating optical flow?", "answer": ["Optical Flow Estimation using a Spatial Pyramid Network"], "answer_arxiv_id": ["1611.00850"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_15875"} +{"question": "Could you provide me some studies about subject-aware text-to-image generation?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2212.04488"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_15876"} +{"question": "What research has been done on modality-agnostic Self-Supervised Learning?", "answer": ["Towards Domain-Agnostic Contrastive Learning", "i-Mix: A Domain-Agnostic Strategy for Contrastive Representation Learning", "DABS: A Domain-Agnostic Benchmark for Self-Supervised Learning", "Representation Learning with Contrastive Predictive Coding", "Viewmaker Networks: Learning Views for Unsupervised Representation Learning", "ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2011.04419", "2010.08887", "2111.12062", "1807.03748", "2010.07432", "2003.10555", "2111.06377"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_15877"} +{"question": "Which research uses CLIP to perform open vocabulary segmentation by only modifying its image encoder?", "answer": ["Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2112.01071"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_15878"} +{"question": "Can you name papers that tackle challenges in multiagent collaboration to get rid of the limitations in single-agent perception in autonomous driving?", "answer": ["Learning Distilled Collaboration Graph for Multi-Agent Perception", "Among Us: Adversarially Robust Collaborative Perception by Consensus", "Uncertainty Quantification of Collaborative Detection for Self-Driving", "Collaboration Helps Camera Overtake LiDAR in 3D Detection", "Collaborative Multi-Object Tracking with Conformal Uncertainty\n Propagation", "ActFormer: Scalable Collaborative Perception via Active Queries"], "answer_arxiv_id": ["2111.00643", "2303.09495", "2209.08162", "2303.13560", "2303.14346", "2403.04968"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_15879"} +{"question": "Could you provide me some studies focused on the semantic correspondence between video and audio?", "answer": ["Objects that Sound", "Look, Listen and Learn"], "answer_arxiv_id": ["1712.06651", "1705.08168"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_15880"} +{"question": "Which studies demonstrate that regularization is not necessary for obtaining linear convergence in NPG?", "answer": ["On the Linear convergence of Natural Policy Gradient Algorithm", "On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2105.01424v1", "2201.07443"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_15881"} +{"question": "Which paper designs estimators for the optimal stopping problem by recursively calling the rMLMC algorithm?", "answer": ["Unbiased Optimal Stopping via the MUSE"], "answer_arxiv_id": ["2106.02263"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_15882"} +{"question": "What research works aimed at consolidating various modalities and tasks by representing them as images?", "answer": ["Transframer: Arbitrary Frame Prediction with Generative Models", "Visual Prompting via Image Inpainting", "Images Speak in Images: A Generalist Painter for In-Context Visual Learning"], "answer_arxiv_id": ["2203.09494", "2209.00647", "2212.02499"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_15883"} +{"question": "Which work used INRs for anomaly detection in time series?", "answer": ["Time-Series Anomaly Detection with Implicit Neural Representation"], "answer_arxiv_id": ["2201.11950"], "source_meta": {"published_time": "20220713"}, "qid": "AutoScholarQuery_train_15884"} +{"question": "What researches use depth maps or LiDAR point clouds as guidance in 3D detection?", "answer": ["Learning Depth-Guided Convolutions for Monocular 3D Object Detection", "Depth-conditioned Dynamic Message Propagation for Monocular 3D Object\n Detection", "Is Pseudo-Lidar needed for Monocular 3D Object detection?", "MonoDistill: Learning Spatial Features for Monocular 3D Object Detection", "Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving", "DID-M3D: Decoupling Instance Depth for Monocular 3D Object Detection"], "answer_arxiv_id": ["1912.04799", "2103.16470", "2108.06417", "2201.10830", "2203.02112", "2207.08531"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15885"} +{"question": "Could you provide me with studies that offer theoretical optimization guarantees under related conditions to power laws?", "answer": ["Last iterate convergence of SGD for Least-Squares in the Interpolation regime", "Tight Nonparametric Convergence Rates for Stochastic Gradient Descent under the Noiseless Linear Model", "Benign Overfitting of Constant-Stepsize SGD for Linear Regression", "Learning curves for Gaussian process regression with power-law priors and targets", "Spectral Bias Outside the Training Set for Deep Networks in the Kernel Regime"], "answer_arxiv_id": ["2102.03183", "2006.08212", "2103.12692", "2110.12231", "2206.02927"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_15886"} +{"question": "What work presents a method for task selection in Natural Language Processing (NLP) by building a graph among common NLP tasks?", "answer": ["TaskWeb: Selecting Better Source Tasks for Multi-task NLP"], "answer_arxiv_id": ["2305.13256"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_15887"} +{"question": "What studies have proposed algorithmic approaches for fast clustering of points in metric spaces?", "answer": ["DBSCAN++: Towards fast and scalable density clustering", "Faster k-Medoids Clustering: Improving the PAM, CLARA, and CLARANS Algorithms"], "answer_arxiv_id": ["1810.13105", "1810.05691v4"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_15888"} +{"question": "Which works introduced metric-based methods for few-shot learning?", "answer": ["Prototypical Networks for Few-shot Learning", "Learning to Compare: Relation Network for Few-Shot Learning", "Matching Networks for One Shot Learning", "Feature Generating Networks for Zero-Shot Learning"], "answer_arxiv_id": ["1703.05175", "1711.06025", "1606.04080", "1712.00981"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_15889"} +{"question": "Could you mention papers that reported results in terms of four calibration metrics including ECE and Adaptive-ECE?", "answer": ["On Calibration of Modern Neural Networks", "Revisiting the Evaluation of Uncertainty Estimation and Its Application to Explore Model Complexity-Uncertainty Trade-Off", "Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration"], "answer_arxiv_id": ["1706.04599", "1903.02050", "1910.12656"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_15890"} +{"question": "Could you provide me some studies about global ε-optimal policy using sample complexity for soft-max policy parameterization?", "answer": ["On the Convergence and Sample Efficiency of Variance-Reduced Policy Gradient Method"], "answer_arxiv_id": ["2102.08607"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_15891"} +{"question": "What studies have addressed fairness under noisy labels?", "answer": ["Fair Classification with Group-Dependent Label Noise", "Fairness-Aware PAC Learning from Corrupted Data", "Fairness Improves Learning from Noisily Labeled Long-Tailed Data"], "answer_arxiv_id": ["2011.00379", "2102.06004", "2303.12291"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_15892"} +{"question": "What work discusses a four-stage mechanism related to Gradient Descent convergence?", "answer": ["Analyzing Sharpness along GD Trajectory: Progressive Sharpening and Edge of Stability"], "answer_arxiv_id": ["2207.12678"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_15893"} +{"question": "What are some studies about open-source efforts enabling LLaMA to use image inputs?", "answer": ["LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention"], "answer_arxiv_id": ["2303.16199"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_15894"} +{"question": "What papers propose Gaussian-based representations to improve the performance of oriented object detection?", "answer": ["Rethinking Rotated Object Detection with Gaussian Wasserstein Distance\n Loss", "Learning High-Precision Bounding Box for Rotated Object Detection via\n Kullback-Leibler Divergence"], "answer_arxiv_id": ["2101.11952", "2106.01883"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_15895"} +{"question": "Are there any publications that try to uncover the underlying working mechanisms of in-context learning?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_15896"} +{"question": "Can you provide me studies related to learning with deferral?", "answer": ["Consistent Estimators for Learning to Defer to an Expert"], "answer_arxiv_id": ["2006.01862v3"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_15897"} +{"question": "Which papers discusse the use of Transformer-based methods for Long Shot-Term Forecasting (LSTF)?", "answer": ["Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecasting", "Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting", "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting", "ETSformer: Exponential Smoothing Transformers for Time-series Forecasting", "FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting"], "answer_arxiv_id": ["1907.00235", "2012.07436", "2106.13008", "2202.01381", "2201.12740"], "source_meta": {"published_time": "20220713"}, "qid": "AutoScholarQuery_train_15898"} +{"question": "What research work mentions that the success of NeRF extensions depends on the dominance of camera motion over scene motion?", "answer": ["Monocular Dynamic View Synthesis: A Reality Check"], "answer_arxiv_id": ["2210.13445"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_15899"} +{"question": "Are there works that propose parameter-efficient fine-tuning?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "LoRA: Low-Rank Adaptation of Large Language Models", "Visual Prompt Tuning"], "answer_arxiv_id": ["1902.00751", "2101.00190", "2106.09685", "2203.12119"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_15900"} +{"question": "Which papers have developed patchifying operation in transformer encoder architectures, specifically in image encoder models?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_15901"} +{"question": "What research used variational graph autoencoder architecture (VGAE) and graph neural network (GNN) architecture like in VACA?", "answer": ["Relating Graph Neural Networks to Structural Causal Models"], "answer_arxiv_id": ["2109.04173v3"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_15902"} +{"question": "What works center around hyperparameter optimization and are formulated as BLO?", "answer": ["Forward and Reverse Gradient-Based Hyperparameter Optimization", "Optimizing Millions of Hyperparameters by Implicit Differentiation", "Gradient-based Hyperparameter Optimization through Reversible Learning"], "answer_arxiv_id": ["1703.01785", "1911.02590", "1502.03492"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_15903"} +{"question": "What papers follow the first main approach which focuses on the transformation from static data structures to dynamic structures in non-private algorithms in the streaming model?", "answer": ["Lower Bounds for Oblivious Near-Neighbor Search"], "answer_arxiv_id": ["1904.04828"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_15904"} +{"question": "Are there any works that examine self-supervised learning for task-specific fields?", "answer": ["DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2304.07193"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_15905"} +{"question": "Could you provide me some papers that proposed the use of VAEs and energy-based models to learn the latent distribution?", "answer": ["A Contrastive Learning Approach for Training Variational Autoencoder Priors"], "answer_arxiv_id": ["2010.02917"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_15906"} +{"question": "Which works are about reducing the sequence length for creating efficient architectures?", "answer": ["Efficient conformer: Progressive downsampling and grouped attention for\n automatic speech recognition", "Squeezeformer: An Efficient Transformer for Automatic Speech Recognition", "Efficient Transformers with Dynamic Token Pooling", "Fast Conformer with Linearly Scalable Attention for Efficient Speech\n Recognition"], "answer_arxiv_id": ["2109.01163", "2206.00888", "2211.09761", "2305.05084"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_15907"} +{"question": "What paper proposed a model that uses large patch size initially and decreases patch size adaptively for ViT?", "answer": ["Not All Images are Worth 16x16 Words: Dynamic Transformers for Efficient Image Recognition"], "answer_arxiv_id": ["2105.15075"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15908"} +{"question": "Which works consider adversarial attacks on single agent RL?", "answer": ["Vulnerability of Deep Reinforcement Learning to Policy Induction Attacks", "Deceptive Reinforcement Learning Under Adversarial Manipulations on Cost Signals", "Policy Poisoning in Batch Reinforcement Learning and Control", "Adaptive Reward-Poisoning Attacks against Reinforcement Learning", "Vulnerability-Aware Poisoning Mechanism for Online RL with Unknown Dynamics", "Policy Teaching via Environment Poisoning: Training-time Adversarial Attacks against Reinforcement Learning", "Reward Poisoning in Reinforcement Learning: Attacks Against Unknown Learners in Unknown Environments"], "answer_arxiv_id": ["1701.04143", "1906.10571", "1910.05821", "2003.12613", "2009.00774v5", "2003.12909", "2102.08492"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_15909"} +{"question": "Any works about proposed multi-hop QA datasets and methods?", "answer": ["HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering", "DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs", "Answering Complex Open-Domain Questions with Multi-Hop Dense Retrieval"], "answer_arxiv_id": ["1809.09600", "1903.00161", "2009.12756"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_15910"} +{"question": "What research proposed the use of the set encoder for task-adaptive accuracy prediction models?", "answer": ["Rapid Neural Architecture Search by Learning to Generate Graphs from Datasets", "Task-Adaptive Neural Network Search with Meta-Contrastive Learning"], "answer_arxiv_id": ["2107.00860", "2103.01495"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_15911"} +{"question": "Which work presented a meta-gradient algorithm for the discovery of GVF-based questions?", "answer": ["Discovery of Useful Questions as Auxiliary Tasks"], "answer_arxiv_id": ["1909.04607"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_15912"} +{"question": "Which studies introduce using colored boxes and circle mark for visual attention during zero-shot classification?", "answer": ["CPT: Colorful Prompt Tuning for Pre-trained Vision-Language Models"], "answer_arxiv_id": ["2109.11797"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_15913"} +{"question": "Which works suggest methods to approximate real-world degradation by complex degradation models?", "answer": ["Designing a Practical Degradation Model for Deep Blind Image\n Super-Resolution", "Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure\n Synthetic Data"], "answer_arxiv_id": ["2103.14006", "2107.10833"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_15914"} +{"question": "What paper heuristically demonstrates the universal approximation idea in the context of state-space models?", "answer": ["Resurrecting Recurrent Neural Networks for Long Sequences"], "answer_arxiv_id": ["2303.06349"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_15915"} +{"question": "Which paper studied transfer learning for solving the same PDEs at different resolutions?", "answer": ["Transfer learning based multi-fidelity physics informed deep neural network"], "answer_arxiv_id": ["2005.10614v2"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_15916"} +{"question": "What research studies causal matching, where an experimenter can transform the system to a desired state through interventions?", "answer": ["Matching a Desired Causal State via Shift Interventions"], "answer_arxiv_id": ["2107.01850"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_15917"} +{"question": "Which work introduced additional semantically similar supervision to contrastive representation learning?", "answer": ["Constrained Mean Shift Using Distant Yet Related Neighbors for Representation Learning"], "answer_arxiv_id": ["2112.04607"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_15918"} +{"question": "What work proposed a generative model of environment maps by applying probabilistic diffusion as a prior on the illumination?", "answer": ["Diffusion Posterior Illumination for Ambiguity-aware Inverse Rendering"], "answer_arxiv_id": ["2310.00362"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15919"} +{"question": "Could you provide me works that propose architectural modifications to enhance the performance of low-precision models?", "answer": ["XNOR-Net: ImageNet Classification Using Binary Convolutional Neural\n Networks", "VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision\n Neural Network Inference", "FracBNN: Accurate and FPGA-Efficient Binary Neural Networks with\n Fractional Activations", "PokeBNN: A Binary Pursuit of Lightweight Accuracy", "Delving Deep into Rectifiers: Surpassing Human-Level Performance on\n ImageNet Classification", "ReActNet: Towards Precise Binary Neural Network with Generalized\n Activation Functions", "How Do Adam and Training Strategies Help BNNs Optimization?"], "answer_arxiv_id": ["1603.05279", "2102.04503", "2012.12206", "2112.00133", "1502.01852", "2003.03488", "2106.11309"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_15920"} +{"question": "Which research introduced low-cost and efficient training methods, such as S2​-AttnsuperscriptS2-Attn or Activation Beacon, for reducing training costs when fine-tuning LLMs?", "answer": ["LongLoRA: Efficient Fine-tuning of Long-Context Large Language Models", "Soaring from 4K to 400K: Extending LLM's Context with Activation Beacon"], "answer_arxiv_id": ["2309.12307", "2401.03462"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_15921"} +{"question": "What papers highlight the universal approximation properties of universal operators particularly in the context of Fourier neural operators?", "answer": ["On universal approximation and error bounds for Fourier Neural Operators", "Neural Operator: Learning Maps Between Function Spaces"], "answer_arxiv_id": ["2107.07562", "2108.08481v6"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_15922"} +{"question": "Which studies used high-level self-supervised learning to reduce the sampling space of generative algorithms?", "answer": ["wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations", "HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2006.11477", "2106.07447", "2111.06377"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_15923"} +{"question": "What is an example of work that reduces memory usage for the processing of high-resolution images by identifying regions of interest in low resolution?", "answer": ["Processing Megapixel Images with Deep Attention-Sampling Models", "Differentiable Patch Selection for Image Recognition"], "answer_arxiv_id": ["1905.03711v2", "2104.03059"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_15924"} +{"question": "What papers mention obfuscating face images by adding crafted noise?", "answer": ["Privacy Preserving Face Recognition Utilizing Differential Privacy", "IdentityDP: Differential Private Identification Protection for Face\n Images"], "answer_arxiv_id": ["2005.10486", "2103.01745"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_15925"} +{"question": "Could you give me an example of a study that pre-trains the soft prompt of a target task with data formulated similarly with target data?", "answer": ["PPT: Pre-trained Prompt Tuning for Few-shot Learning"], "answer_arxiv_id": ["2109.04332"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_15926"} +{"question": "What works focus on hardware-accelerated environments for continuous control?", "answer": ["Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learning"], "answer_arxiv_id": ["2108.10470"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_15927"} +{"question": "What papers discuss the utilization of transformers for reinforcement learning tasks?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Offline Reinforcement Learning as One Big Sequence Modeling Problem", "TransDreamer: Reinforcement Learning with Transformer World Models", "Transformers are Sample-Efficient World Models"], "answer_arxiv_id": ["2106.01345", "2106.02039", "2202.09481", "2209.00588"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_15928"} +{"question": "Which papers have suggested the application of knowledge distillation techniques to expedite the sampling process of diffusion models?", "answer": ["Knowledge Distillation in Iterative Generative Models for Improved\n Sampling Speed"], "answer_arxiv_id": ["2101.02388"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_15929"} +{"question": "Are there any works that tested T2I-Adapters that are jointly trained, like CoAdapter?", "answer": ["T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.08453"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15930"} +{"question": "Which studies report on the development of HDR-VDP metric and its improvements?", "answer": ["HDR-VDP-3: A multi-metric for predicting image differences, quality and\n contrast distortions in high dynamic range and regular content", "Consolidated Dataset and Metrics for High-Dynamic-Range Image Quality"], "answer_arxiv_id": ["2304.13625", "2012.10758"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_15931"} +{"question": "What are some studies that achieved success in machine learning tasks using Graph Neural Networks (GNNs)?", "answer": ["Link Prediction Based on Graph Neural Networks"], "answer_arxiv_id": ["1802.09691"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_15932"} +{"question": "What research papers applied Subsampled Randomized Hadamard Transformation (SRHT) to solve machine learning tasks?", "answer": ["Optimal Randomized First-Order Methods for Least-Squares Problems"], "answer_arxiv_id": ["2002.09488"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_15933"} +{"question": "Which papers conducted research in expanding the utility of NeRF for different tasks?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Neuralangelo: High-Fidelity Neural Surface Reconstruction", "NeRD: Neural Reflectance Decomposition from Image Collections", "NeRFactor: Neural Factorization of Shape and Reflectance Under an\n Unknown Illumination", "NeRO: Neural Geometry and BRDF Reconstruction of Reflective Objects from\n Multiview Images", "BARF: Bundle-Adjusting Neural Radiance Fields", "NeRF--: Neural Radiance Fields Without Known Camera Parameters", "BAD-NeRF: Bundle Adjusted Deblur Neural Radiance Fields", "Depth-supervised NeRF: Fewer Views and Faster Training for Free", "FreeNeRF: Improving Few-shot Neural Rendering with Free Frequency\n Regularization"], "answer_arxiv_id": ["2106.10689", "2306.03092", "2012.03918", "2106.01970", "2305.17398", "2104.06405", "2102.07064", "2211.12853", "2107.02791", "2303.07418"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_15934"} +{"question": "What work introduced the concept of instruction tuning?", "answer": ["Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2109.01652"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_15935"} +{"question": "Which works have studied the sim-to-real transfer theoretically?", "answer": ["How Does an Approximate Model Help in Reinforcement Learning?", "PAC Reinforcement Learning without Real-World Feedback"], "answer_arxiv_id": ["1912.02986", "1909.10449"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_15936"} +{"question": "What papers studied the vulnerability of deep neural networks to adversarial attacks?", "answer": ["Intriguing properties of neural networks", "Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1312.6199", "1412.6572"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_15937"} +{"question": "What papers have proposed parameterizing the diffusion equation on graphs with a neural network?", "answer": ["GRAND: Graph Neural Diffusion"], "answer_arxiv_id": ["2106.10934"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_15938"} +{"question": "Any researches about subject-driven image generation that don't require test-time fintuning?", "answer": ["InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models", "Face0: Instantaneously Conditioning a Text-to-Image Model on a Face", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models", "Taming Encoder for Zero Fine-tuning Image Customization with\n Text-to-Image Diffusion Models", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning"], "answer_arxiv_id": ["2304.03411", "2302.12228", "2306.06638", "2302.13848", "2308.06721", "2304.02642", "2305.14720", "2307.11410"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_15939"} +{"question": "Which papers used pose estimation models to derive 2D/3D poses from online dance videos?", "answer": ["Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields", "AlphaPose: Whole-Body Regional Multi-Person Pose Estimation and Tracking\n in Real-Time", "mm-Pose: Real-Time Human Skeletal Posture Estimation using mmWave Radars\n and CNNs"], "answer_arxiv_id": ["1611.08050", "2211.03375", "1911.09592"], "source_meta": {"published_time": "20240506"}, "qid": "AutoScholarQuery_train_15940"} +{"question": "Which work addressed the limitations of traditional numerical solvers for dynamics forecasting by introducing a GNN approach to model adaptively sampled points in a graph architecture?", "answer": ["Graph Element Networks: adaptive, structured computation and memory"], "answer_arxiv_id": ["1904.09019"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_15941"} +{"question": "What works are there on the subject of OOD Detection?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks"], "answer_arxiv_id": ["1610.02136", "1706.02690"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_15942"} +{"question": "Could you provide some studies about compressing original data to a fixed dimension with neural autoencoders?", "answer": ["Neural Distributed Source Coding"], "answer_arxiv_id": ["2106.02797"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_15943"} +{"question": "What works suggested geometric-based approaches in neural rendering?", "answer": ["Self-Calibrating Neural Radiance Fields", "BARF : Bundle-Adjusting Neural Radiance Fields", "GARF: Gaussian Activated Radiance Fields for High Fidelity Reconstruction and Pose Estimation"], "answer_arxiv_id": ["2108.13826", "2104.06405", "2204.05735"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_15944"} +{"question": "Could you tell me a work that provides an overview of prior theoretical results for SSL?", "answer": ["Improvability through Semi-Supervised Learning: A Survey of Theoretical Results"], "answer_arxiv_id": ["1908.09574"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15945"} +{"question": "What studies proposed alternatives to Local SGDA for improving sample and communication complexities?", "answer": ["Federated Minimax Optimization: Improved Convergence Analyses and Algorithms", "FedNest: Federated Bilevel, Minimax, and Compositional Optimization"], "answer_arxiv_id": ["2203.04850", "2205.02215v3"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_15946"} +{"question": "Could you list some studies that implement the Class-Prototype (CP) accumulation strategy?", "answer": ["Continual Prototype Evolution: Learning Online from Non-Stationary Data Streams", "Don’t Stop Learning: Towards Continual Learning for the CLIP Model", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Memory Efficient Continual Learning with Transformers", "Lifelong Machine Learning with Deep Streaming Linear Discriminant Analysis", "Towards a Unified View of Parameter-Efficient Transfer Learning"], "answer_arxiv_id": ["2009.00919", "2207.09248", "2010.11929", "2203.04640", "1909.01520", "2110.04366"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_15947"} +{"question": "What are some recent attempts to model task difficulty in crowdsourcing?", "answer": ["Achieving Budget-optimality with Adaptive Schemes in Crowdsourcing", "A Permutation-based Model for Crowd Labeling: Optimal Estimation and Robustness", "Reducing Crowdsourcing to Graphon Estimation, Statistically"], "answer_arxiv_id": ["1602.03481", "1606.09632v3", "1703.08085v4"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_15948"} +{"question": "What studies introduced artificial transformations onto images such as style transfer, corruptions and perturbations to investigate the robustness of machine learning models?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness", "Benchmarking Neural Network Robustness to Common Corruptions and Perturbations", "On Interaction Between Augmentations and Corruptions in Natural Corruption Robustness"], "answer_arxiv_id": ["1811.12231", "1903.12261", "2102.11273"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_15949"} +{"question": "What papers have explored methods for LLM instruction-tuning?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_15950"} +{"question": "Which studies introduced polarizing optics to a conventional projector for facilitating 3D imaging of translucent objects and polarimetric light transport analysis?", "answer": ["Polarimetric Spatio-Temporal Light Transport Probing", "Polarimetric iToF: Measuring High-Fidelity Depth through Scattering\n Media"], "answer_arxiv_id": ["2105.11609", "2306.17618"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_15951"} +{"question": "What research has been done to improve the initialization in deep state space models?", "answer": ["How to Train Your HiPPO: State Space Models with Generalized Orthogonal Basis Projections"], "answer_arxiv_id": ["2206.12037"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_15952"} +{"question": "Can you point out research works that built on RUDDER by replacing it with expressive language models and the continuous modern Hopfield network?", "answer": ["Sequence Modeling of Temporal Credit Assignment for Episodic Reinforcement Learning"], "answer_arxiv_id": ["1905.13420"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_15953"} +{"question": "Which literature discusses Dataset Ownership Verification using backdoor attacks?", "answer": ["Black-box Dataset Ownership Verification via Backdoor Watermarking", "Untargeted Backdoor Watermark: Towards Harmless and Stealthy Dataset Copyright Protection", "Did You Train on My Dataset? Towards Public Dataset Protection with Clean-Label Backdoor Watermarking"], "answer_arxiv_id": ["2209.06015", "2210.00875", "2303.11470"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_15954"} +{"question": "Which studies in video grounding generate temporal segments as proposals, then score them and refine their boundaries?", "answer": ["Localizing Moments in Video with Natural Language", "TALL: Temporal Activity Localization via Language Query"], "answer_arxiv_id": ["1708.01641", "1705.02101"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_15955"} +{"question": "What studies are based on the GAN paradigm for generative image editing?", "answer": ["User-Controllable Latent Transformer for StyleGAN Image Layout Editing", "Drag Your GAN: Interactive Point-based Manipulation on the Generative\n Image Manifold", "StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated\n Images using Conditional Continuous Normalizing Flows", "FreeStyleGAN: Free-view Editable Portrait Rendering with the Camera\n Manifold", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "Interpreting the Latent Space of GANs for Semantic Face Editing", "Closed-Form Factorization of Latent Semantics in GANs", "StyleRig: Rigging StyleGAN for 3D Control over Portrait Images", "GANSpace: Discovering Interpretable GAN Controls", "Generative Visual Manipulation on the Natural Image Manifold", "LinkGAN: Linking GAN Latents to Pixels for Controllable Image Synthesis"], "answer_arxiv_id": ["2208.12408", "2305.10973", "2008.02401", "2109.09378", "2103.17249", "1907.10786", "2007.06600v4", "2004.00121", "2004.02546", "1609.03552", "2301.04604"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_15956"} +{"question": "What research papers suggest the use of sophisticated grouping methods for quantization of LLMs?", "answer": ["ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers"], "answer_arxiv_id": ["2206.01861"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_15957"} +{"question": "Can you name the studies that focused on the scalability of Dynamic Topic Models?", "answer": ["Scalable Generalized Dynamic Topic Models", "Scaling up Dynamic Topic Models"], "answer_arxiv_id": ["1803.07868", "1602.06049v1"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_15958"} +{"question": "What papers demonstrated limitations of the Message Passing Neural Network's expressive power?", "answer": ["How Powerful are Graph Neural Networks?", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks"], "answer_arxiv_id": ["1810.00826", "1810.02244"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_15959"} +{"question": "Which papers are about the risk of copying, style mimicry and copyright at inference due to the memorization in foundation models?", "answer": ["Extracting Training Data from Large Language Models", "Extracting Training Data from Diffusion Models", "Glaze: Protecting Artists from Style Mimicry by Text-to-Image Models"], "answer_arxiv_id": ["2012.07805", "2301.13188", "2302.04222"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_15960"} +{"question": "Could you provide me the references where they used diffusion models to construct continuous space planners?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis", "Is Conditional Generative Modeling all you need for Decision-Making?", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "Guided Conditional Diffusion for Controllable Traffic Simulation", "Learning Universal Policies via Text-Guided Video Generation"], "answer_arxiv_id": ["2205.09991", "2211.15657", "2208.06193", "2210.17366", "2302.00111"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_15961"} +{"question": "Which studies proposed a method for constraining the offline RL training by penalizing the Q function?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Fisher Divergence Critic Regularization", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["2006.04779", "2103.08050", "2110.06169"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_15962"} +{"question": "What studies attempted to integrate open-vocabulary object detection and dense captioning?", "answer": ["CapDet: Unifying Dense Captioning and Open-World Detection Pretraining", "GRiT: A Generative Region-to-text Transformer for Object Understanding"], "answer_arxiv_id": ["2303.02489", "2212.00280"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_15963"} +{"question": "Which studies have proposed different decoding strategies for language models?", "answer": ["Controlling Linguistic Style Aspects in Neural Language Generation", "Hierarchical Neural Story Generation", "Learning to Write with Cooperative Discriminators", "The Curious Case of Neural Text Degeneration", "Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural Machine Translation"], "answer_arxiv_id": ["1707.02633", "1805.04833", "1805.06087", "1904.09751", "2005.10283"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_15964"} +{"question": "What are the studies on individual fairness suggesting that similar individuals should receive similar treatment?", "answer": ["Fairness Through Awareness"], "answer_arxiv_id": ["1104.3913"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_15965"} +{"question": "Which study showed a strong linear correlation between OOD agreement and ID agreement in the context of neural networks?", "answer": ["Agreement-on-the-Line: Predicting the Performance of Neural Networks under Distribution Shift"], "answer_arxiv_id": ["2206.13089"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_15966"} +{"question": "What are some of the works that attempted to unify different contrastive learning techniques in a single theoretical framework?", "answer": ["Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods"], "answer_arxiv_id": ["2205.11508"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_15967"} +{"question": "What is the citation for the lemma that provides a lower bound on the expected sample complexity of a best arm identification algorithm?", "answer": ["Optimal Best Arm Identification with Fixed Confidence"], "answer_arxiv_id": ["1602.04589v2"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_15968"} +{"question": "What works have been used benchmarks spanning multiple modalities and applications?", "answer": ["Wilds: A Benchmark of in-the-Wild Distribution Shifts"], "answer_arxiv_id": ["2012.07421"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_15969"} +{"question": "What work proposed a hybrid implicit-explicit triplane representation as a compromise between rendering speed and memory consumption for 3D reconstruction?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["2112.07945"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_15970"} +{"question": "Can you provide examples of works that focused on the communication of intra-context object attributes in emergent language?", "answer": ["Emergent Linguistic Phenomena in Multi-Agent Communication Games", "Multi-Agent Cooperation and the Emergence of (Natural) Language"], "answer_arxiv_id": ["1901.08706", "1612.07182"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_15971"} +{"question": "Which papers proposed polynomial-time algorithms for learning one-hidden-layer neural networks under certain input distributions?", "answer": ["Beating the Perils of Non-Convexity: Guaranteed Training of Neural Networks using Tensor Methods", "Recovery Guarantees for One-hidden-layer Neural Networks", "Learning One-hidden-layer Neural Networks with Landscape Design", "Learning Two-layer Neural Networks with Symmetric Inputs", "Learning Two Layer Rectified Neural Networks in Polynomial Time", "Learning One-hidden-layer ReLU Networks via Gradient Descent", "Efficient Algorithms for Learning Depth-2 Neural Networks with General ReLU Activations"], "answer_arxiv_id": ["1506.08473", "1706.03175", "1711.00501", "1810.06793", "1811.01885", "1806.07808", "2107.10209"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_15972"} +{"question": "Which work addresses privacy concerns that arise from using generative models due to their data replication characteristic?", "answer": ["Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models"], "answer_arxiv_id": ["2212.03860"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_15973"} +{"question": "Which studies indicated that topologies with a fast consensus rate improve the accuracy of decentralized learning?", "answer": ["A Unified Theory of Decentralized SGD with Changing Topology and Local Updates", "Consensus Control for Decentralized Deep Learning", "Throughput-Optimal Topology Design for Cross-Silo Federated Learning", "Matcha: Speeding Up Decentralized SGD via Matching Decomposition Sampling"], "answer_arxiv_id": ["2003.10422", "2102.04828", "2010.12229", "1905.09435"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_15974"} +{"question": "What studies used convolutional neural networks for tabular data synthesis?", "answer": ["Data Synthesis based on Generative Adversarial Networks"], "answer_arxiv_id": ["1806.03384v5"], "source_meta": {"published_time": "20221008"}, "qid": "AutoScholarQuery_train_15975"} +{"question": "Which researches focused on sparser setups for encoding scenes in free space?", "answer": ["Learning-Based View Synthesis for Light Field Cameras", "Light Field Networks: Neural Scene Representations with\n Single-Evaluation Rendering", "Learning Neural Light Fields with Ray-Space Embedding Networks", "Light Field Neural Rendering", "Generalizable Patch-Based Neural Rendering"], "answer_arxiv_id": ["1609.02974", "2106.02634", "2112.01523", "2112.09687", "2207.10662"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_15976"} +{"question": "Which works have used language-conditioned imitation learning and reinforcement learning to enhance robotic behavior?", "answer": ["Language Conditioned Imitation Learning over Unstructured Data", "Language-Conditioned Imitation Learning for Robot Manipulation Tasks", "Mapping Instructions and Visual Observations to Actions with Reinforcement Learning", "Learning with Latent Language", "A Survey of Reinforcement Learning Informed by Natural Language"], "answer_arxiv_id": ["2005.07648", "2010.12083", "1704.08795", "1711.00482", "1906.03926"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_15977"} +{"question": "What papers propose the use of reinforcement learning for discovering better minimum energy structures in materials?", "answer": ["Reinforcement Learning for Molecular Design Guided by Quantum Mechanics"], "answer_arxiv_id": ["2002.07717"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_15978"} +{"question": "Can you mention some papers where they achieve implicit 3D scene reconstruction and novel view synthesis from multi-view images?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "S-NeRF: Neural Radiance Fields for Street Views", "SceneRF: Self-Supervised Monocular 3D Scene Reconstruction with Radiance\n Fields", "StreetSurf: Extending Multi-view Implicit Surface Reconstruction to\n Street Views"], "answer_arxiv_id": ["2003.08934", "2303.00749", "2212.02501", "2306.04988"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_15979"} +{"question": "What papers explored the use of a two-step approach to improve autoregressive generative models using fixed noise levels?", "answer": ["Improved Autoregressive Modeling with Distribution Smoothing"], "answer_arxiv_id": ["2103.15089"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_15980"} +{"question": "Are there any studies that have improved the realism of synthetic tumors in the liver and pancreas?", "answer": ["Acquiring Weak Annotations for Tumor Localization in Temporal and\n Volumetric Data"], "answer_arxiv_id": ["2310.15098"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_15981"} +{"question": "Where was the data for the repository-level code completion benchmark in Python (RepoEval) derived from?", "answer": ["The Stack: 3 TB of permissively licensed source code"], "answer_arxiv_id": ["2211.15533"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_15982"} +{"question": "Which studies are related to the integration of transformer blocks with a U-net architecture?", "answer": ["All are Worth Words: A ViT Backbone for Diffusion Models"], "answer_arxiv_id": ["2209.12152"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_15983"} +{"question": "Which papers introduced the concept of a world model which tries to completely imitate the environment?", "answer": ["Recurrent World Models Facilitate Policy Evolution"], "answer_arxiv_id": ["1809.01999"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_15984"} +{"question": "What papers attempted to combine the strength of score distillation and mesh texturization?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation"], "answer_arxiv_id": ["2211.10440"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_15985"} +{"question": "Can you name the video summarization models that use attention algorithms to determine the essential frames?", "answer": ["Video Summarization with Attention-Based Encoder-Decoder Networks", "Summarizing Videos with Attention", "Exploring global diverse attention via pairwise temporal relation for\n video summarization", "Video Joint Modelling Based on Hierarchical Transformer for\n Co-summarization", "CLIP-It! Language-Guided Video Summarization"], "answer_arxiv_id": ["1708.09545", "1812.01969", "2009.10942", "2112.13478", "2107.00650"], "source_meta": {"published_time": "20240520"}, "qid": "AutoScholarQuery_train_15986"} +{"question": "Could you name the methods that leveraged visual text embedding for synthesizing sketches?", "answer": ["CLIPDraw: Exploring Text-to-Drawing Synthesis through Language-Image Encoders", "StyleCLIPDraw: Coupling Content and Style in Text-to-Drawing Synthesis", "CLIP-CLOP: CLIP-Guided Collage and Photomontage", "CLIPascene: Scene Sketching with Different Types and Levels of Abstraction"], "answer_arxiv_id": ["2106.14843", "2111.03133", "2205.03146", "2211.17256"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_15987"} +{"question": "What research treated 3DVG as a matching problem and utilized object identifiers for generating candidate objects?", "answer": ["InstanceRefer: Cooperative Holistic Understanding for Visual Grounding\n on Point Clouds through Instance Multi-level Contextual Referring", "SAT: 2D Semantics Assisted Training for 3D Visual Grounding", "LanguageRefer: Spatial-Language Model for 3D Visual Grounding", "Language Conditioned Spatial Relation Reasoning for 3D Object Grounding"], "answer_arxiv_id": ["2103.01128", "2105.11450", "2107.03438", "2211.09646"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_15988"} +{"question": "In which studies have they found that lack of standardised splits leads to challenges in comparing different approaches and fair evaluation of models’ performance?", "answer": ["Automation of Citation Screening for Systematic Literature Reviews using Neural Networks: A Replicability Study"], "answer_arxiv_id": ["2201.07534"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_15989"} +{"question": "What paper discussed single pose estimation using an implicit neural representation on S3?", "answer": ["Implicit-PDF: Non-Parametric Representation of Probability Distributions on the Rotation Manifold"], "answer_arxiv_id": ["2106.05965"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_15990"} +{"question": "Which work is about extending convolutional operators to irregular grids to accelerate graph-based surrogate models?", "answer": ["Fourier Neural Operator with Learned Deformations for PDEs on General Geometries"], "answer_arxiv_id": ["2207.05209"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_15991"} +{"question": "Which papers examined the design choices in coordinate neural networks for implicitly representing images?", "answer": ["Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains", "Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Beyond Periodicity: Towards a Unifying Framework for Activations in Coordinate-MLPs", "Trading Positional Complexity vs. Deepness in Coordinate Networks"], "answer_arxiv_id": ["2006.10739", "1906.01618", "2003.08934", "2111.15135", "2205.08987"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_15992"} +{"question": "Which study suggested using semantic consistency loss to train on missing views in Neural Radiance Fields models?", "answer": ["Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis"], "answer_arxiv_id": ["2104.00677"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_15993"} +{"question": "What papers are about refining contexts or adapting generation models to handle extended contexts in RAG?", "answer": ["Large Language Models with Controllable Working Memory", "Active Retrieval Augmented Generation", "Retrieval meets Long Context Large Language Models", "LongLoRA: Efficient Fine-tuning of Long-Context Large Language Models"], "answer_arxiv_id": ["2211.05110", "2305.06983", "2310.03025", "2309.12307"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_15994"} +{"question": "What work developed a DP version of Hamiltonian Monte Carlo?", "answer": ["Differentially Private Hamiltonian Monte Carlo"], "answer_arxiv_id": ["2106.09376"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_15995"} +{"question": "What works are related to the recovery of a causal DAG from data where the causal representation is directly observed?", "answer": ["Causal Structure Learning"], "answer_arxiv_id": ["1706.09141"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_15996"} +{"question": "Which papers are about Bellman Residual Minimization (BRM) methods?", "answer": ["SBEED: Convergent Reinforcement Learning with Nonlinear Function Approximation"], "answer_arxiv_id": ["1712.10285"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_15997"} +{"question": "Who initiated the study on sparsity-dependent bound?", "answer": ["Gains and Losses are Fundamentally Different in Regret Minimization: The Sparse Case"], "answer_arxiv_id": ["1511.08405"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_15998"} +{"question": "Which works explored the impact of timestep and layer on pre-trained SD's capabilities in semantic matching?", "answer": ["Emergent Correspondence from Image Diffusion", "A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot\n Semantic Correspondence"], "answer_arxiv_id": ["2306.03881", "2305.15347"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_15999"} +{"question": "Which studies measured the OOD generalization of LLMs in the context of deductive reasoning?", "answer": ["RobustLR: A Diagnostic Benchmark for Evaluating Logical Robustness of Deductive Reasoners", "LAMBADA: Backward Chaining for Automated Reasoning in Natural Language"], "answer_arxiv_id": ["2205.12598", "2212.13894"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16000"} +{"question": "Can you list research papers that propose single-instance out-of-distribution (OOD) detection methods?", "answer": ["Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Deep Anomaly Detection with Outlier Exposure", "Deep Anomaly Detection Using Geometric Transformations", "Likelihood Ratios for Out-of-Distribution Detection", "Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicality", "Uncertainty Baselines: Benchmarks for Uncertainty & Robustness in Deep Learning", "Exploring the Limits of Out-of-Distribution Detection"], "answer_arxiv_id": ["1706.02690", "1610.02136", "1812.04606", "1805.10917", "1906.02845", "1906.02994", "2106.04015", "2106.03004"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_16001"} +{"question": "Which work combined Monte Carlo localization method with a pre-trained NeRF to build a real-time global localization method?", "answer": ["Loc-NeRF: Monte Carlo Localization using Neural Radiance Fields"], "answer_arxiv_id": ["2209.09050"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_16002"} +{"question": "What are the studies that have attempted to model hand-object interactions using skeletons or customised meshes?", "answer": ["HOnnotate: A method for 3D Annotation of Hand and Object Poses", "H2O: Two Hands Manipulating Objects for First Person Interaction\n Recognition"], "answer_arxiv_id": ["1907.01481", "2104.11181"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_16003"} +{"question": "Which trained models are capable of generating images and texts?", "answer": ["CM3: A Causal Masked Multimodal Model of the Internet"], "answer_arxiv_id": ["2201.07520"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_16004"} +{"question": "What are some works about Unsupervised Skill Discovery?", "answer": ["Variational Intrinsic Control", "Diversity is All You Need: Learning Skills without a Reward Function", "Fast Task Inference with Variational Intrinsic Successor Features", "Dynamics-Aware Unsupervised Discovery of Skills", "Explore, Discover and Learn: Unsupervised Discovery of State-Covering Skills", "APS: Active Pretraining with Successor Features", "Direct then Diffuse: Incremental Unsupervised Skill Discovery for State Covering and Goal Reaching", "Lipschitz-constrained Unsupervised Skill Discovery", "Variational Option Discovery Algorithms", "Hierarchical Reinforcement Learning by Discovering Intrinsic Options", "Controllability-Aware Unsupervised Skill Discovery"], "answer_arxiv_id": ["1611.07507", "1802.06070", "1906.05030", "1907.01657", "2002.03647", "2108.13956", "2110.14457", "2202.00914", "1807.10299", "2101.06521", "2302.05103"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_16005"} +{"question": "What work first utilized neighbor consistency to create pseudo-clean labels for diffusion training in the context of noisy labels?", "answer": ["Label-Retrieval-Augmented Diffusion Models for Learning from Noisy Labels"], "answer_arxiv_id": ["2305.19518"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_16006"} +{"question": "What works studied data-poisoning backdoor attacks, a category of backdoor attacks?", "answer": ["Dataset Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defenses", "Invisible Backdoor Attack with Sample-Specific Triggers", "Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning", "Poisoning and Backdooring Contrastive Learning", "Label-Consistent Backdoor Attacks"], "answer_arxiv_id": ["2012.10544", "2012.03816", "1712.05526v1", "2106.09667", "1912.02771"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_16007"} +{"question": "What studies have used the von Mises-Fisher distribution in tasks like facial recognition, out-of-distribution detection, and long-tailed learning?", "answer": ["How to Exploit Hyperspherical Embeddings for Out-of-Distribution\n Detection?", "Towards Calibrated Hyper-Sphere Representation via Distribution Overlap\n Coefficient for Long-tailed Learning"], "answer_arxiv_id": ["2203.04450", "2208.10043"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_16008"} +{"question": "Could you provide me studies about using self-supervised learning-based methods to alleviate the feature bias in computational pathology?", "answer": ["Scaling Vision Transformers to Gigapixel Images via Hierarchical\n Self-Supervised Learning", "Dual-stream Multiple Instance Learning Network for Whole Slide Image\n Classification with Self-supervised Contrastive Learning", "Self supervised learning improves dMMR/MSI detection from histology\n slides across multiple cancers"], "answer_arxiv_id": ["2206.02647", "2011.08939", "2109.05819"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16009"} +{"question": "Which papers first reported linear correlations in dataset reconstruction settings?", "answer": ["Do ImageNet Classifiers Generalize to ImageNet?", "Do CIFAR-10 Classifiers Generalize to CIFAR-10?"], "answer_arxiv_id": ["1902.10811", "1806.00451"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_16010"} +{"question": "What research deals with ReLU reduction for efficient PI?", "answer": ["CryptoNAS: Private Inference on a ReLU Budget", "Sphynx: ReLU-Efficient Network Design for Private Inference"], "answer_arxiv_id": ["2006.08733", "2106.11755v1"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_16011"} +{"question": "What work mentions that multicalibration does not give omniprediction for all non-convex losses?", "answer": ["Omnipredictors"], "answer_arxiv_id": ["2109.05389"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_16012"} +{"question": "Can you point to works that proposed GroupViT?", "answer": ["GroupViT: Semantic Segmentation Emerges from Text Supervision"], "answer_arxiv_id": ["2202.11094"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_16013"} +{"question": "Which papers have addressed the issue of LLMs generating toxic content?", "answer": ["Red Teaming Language Models with Language Models", "Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors,\n and Lessons Learned"], "answer_arxiv_id": ["2202.03286", "2209.07858"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_16014"} +{"question": "Which works discuss offline RL and incorporate safety constraints?", "answer": ["Batch Policy Learning under Constraints", "COptiDICE: Offline Constrained Reinforcement Learning via Stationary Distribution Correction Estimation", "Constraints Penalized Q-learning for Safe Offline Reinforcement Learning"], "answer_arxiv_id": ["1903.08738", "2204.08957", "2107.09003"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_16015"} +{"question": "Can you provide some studies where nested sequences of subspaces (flags) appear in principal directions?", "answer": ["Barycentric Subspace Analysis on Manifolds", "Optimization on flag manifolds"], "answer_arxiv_id": ["1607.02833", "1907.00949"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_16016"} +{"question": "Which papers propose deformation-based methods for dynamic 3D representations?", "answer": ["HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "Nerfies: Deformable Neural Radiance Fields", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Neural Surface Reconstruction of Dynamic Scenes with Monocular RGB-D Camera", "Fast Dynamic Radiance Fields with Time-Aware Neural Voxels", "Decoupling Dynamic Monocular Videos for Dynamic View Synthesis"], "answer_arxiv_id": ["2106.13228", "2011.12948", "2011.13961", "2206.15258", "2205.15285", "2304.01716"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_16017"} +{"question": "Are there any studies that proposed light-weight models for neural image codecs?", "answer": ["EVC: Towards Real-Time Neural Image Compression with Mask Decay", "Computationally-Efficient Neural Image Compression with Shallow Decoders"], "answer_arxiv_id": ["2302.05071", "2304.06244"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16018"} +{"question": "Which studies initially developed and adapted the Key Point Analysis (KPA) to summarize and quantify arguments?", "answer": ["From Arguments to Key Points: Towards Automatic Argument Summarization"], "answer_arxiv_id": ["2005.01619"], "source_meta": {"published_time": "20240719"}, "qid": "AutoScholarQuery_train_16019"} +{"question": "Could you tell me about studies that have introduced variants of the transformer architecture to allow editing of generated text?", "answer": ["PEER: A Collaborative Language Model", "Levenshtein Transformer", "Insertion Transformer: Flexible Sequence Generation via Insertion Operations", "Generating Information-Seeking Conversations from Unlabeled Documents"], "answer_arxiv_id": ["2208.11663", "1905.11006", "1902.03249", "2205.12609"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_16020"} +{"question": "Which work proposed the method of Deep Ensembles?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1612.01474"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_16021"} +{"question": "What research proposed improvements over MaskFormer by implementing various architectural enhancements?", "answer": ["Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2112.01527"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16022"} +{"question": "Can you mention references on diffusion models used in probabilistic generative models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_16023"} +{"question": "Which works are examples of LMMs demonstrating impressive performance on benchmarks?", "answer": ["Making the V in VQA Matter: Elevating the Role of Image Understanding in\n Visual Question Answering", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science\n Question Answering", "From Recognition to Cognition: Visual Commonsense Reasoning", "The Abduction of Sherlock Holmes: A Dataset for Visual Abductive\n Reasoning"], "answer_arxiv_id": ["1612.00837", "2209.09513", "1811.10830", "2202.04800"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_16024"} +{"question": "Can you list works where researchers manually created a set of inferential rules for inductive reasoning?", "answer": ["CLUTRR: A Diagnostic Benchmark for Inductive Reasoning from Text"], "answer_arxiv_id": ["1908.06177"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_16025"} +{"question": "Which papers deal with the problem of open-set SSL?", "answer": ["Multi-Task Curriculum Framework for Open-Set Semi-Supervised Learning"], "answer_arxiv_id": ["2007.11330"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_16026"} +{"question": "What work proposed a method for finding a minimal perturbation that achieves a different target class label?", "answer": ["Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1608.04644"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_16027"} +{"question": "What works exist on the topic of sound localization in videos?", "answer": ["Objects that Sound", "Audio-Visual Scene Analysis with Self-Supervised Multisensory Features", "Learning to Localize Sound Sources in Visual Scenes: Analysis and Applications", "Mix and Localize: Localizing Sound Sources in Mixtures", "Class-aware Sounding Objects Localization via Audiovisual Correspondence"], "answer_arxiv_id": ["1712.06651", "1804.03641", "1911.09649", "2211.15058", "2112.11749"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_16028"} +{"question": "Are there any works that extensively utilized and compared supervised pretrained ResNet for object detection and segmentation?", "answer": ["Benchmarking Detection Transfer Learning with Vision Transformers", "Exploring Plain Vision Transformer Backbones for Object Detection", "Rethinking ImageNet Pre-training", "MMDetection: Open MMLab Detection Toolbox and Benchmark"], "answer_arxiv_id": ["2111.11429", "2203.16527", "1811.08883", "1906.07155"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16029"} +{"question": "What work proposed the use of multiple modalities for both intermodal interaction and intramodal refinement in relation prediction?", "answer": ["Stacked Hybrid-Attention and Group Collaborative Learning for Unbiased\n Scene Graph Generation"], "answer_arxiv_id": ["2203.09811"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_16030"} +{"question": "Could you provide me with papers about the use of a root-finding algorithm for transformations in human avatar reconstruction?", "answer": ["SNARF: Differentiable Forward Skinning for Animating Non-Rigid Neural\n Implicit Shapes"], "answer_arxiv_id": ["2104.03953"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_16031"} +{"question": "Which works discuss tuning the embedding layer inputs as a PEFT method?", "answer": ["Input-Tuning: Adapting Unfamiliar Inputs to Frozen Pretrained Models"], "answer_arxiv_id": ["2203.03131"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_16032"} +{"question": "Which papers have refined the use of transformer architectures for superpoint matching?", "answer": ["REGTR: End-to-end Point Cloud Correspondences with Transformers", "Geometric Transformer for Fast and Robust Point Cloud Registration"], "answer_arxiv_id": ["2203.14517", "2202.06688"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_16033"} +{"question": "Could you provide me some studies about multi-policy approaches in MORL?", "answer": ["Multi-Objective Deep Reinforcement Learning", "Meta-Learning for Multi-objective Reinforcement Learning"], "answer_arxiv_id": ["1610.02707", "1811.03376"], "source_meta": {"published_time": "20220816"}, "qid": "AutoScholarQuery_train_16034"} +{"question": "Could you provide me some studies about high-quality correspondence matching in RefSR?", "answer": ["RRSR:Reciprocal Reference-based Image Super-Resolution with Progressive\n Feature Alignment and Selection", "Coarse-to-Fine Embedded PatchMatch and Multi-Scale Dynamic Aggregation\n for Reference-based Super-Resolution"], "answer_arxiv_id": ["2211.04203", "2201.04358"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_16035"} +{"question": "What are some examples of voxel-based methods in surface reconstruction from multi-view images?", "answer": ["SurfaceNet+: An End-to-end 3D Neural Network for Very Sparse Multi-view\n Stereopsis", "Learning a Multi-View Stereo Machine", "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video"], "answer_arxiv_id": ["2005.12690", "1708.05375", "2104.00681"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_16036"} +{"question": "What studies form decision boundaries in an unsupervised manner through neighborhood consensus ?", "answer": ["Unsupervised Deep Learning by Neighbourhood Discovery"], "answer_arxiv_id": ["1904.11567"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_16037"} +{"question": "Which studies use voxel or grid-based formats for 3D shape representation learning?", "answer": ["PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1912.13192", "1904.08755"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_16038"} +{"question": "Which studies introduced an almost annotation-free data creation pipeline for SFT tasks?", "answer": ["Self-Instruct: Aligning Language Models with Self-Generated Instructions"], "answer_arxiv_id": ["2212.10560"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_16039"} +{"question": "In the context of NLP, are there any concurrent studies that make use of trainable gates?", "answer": ["An Extensible Plug-and-Play Method for Multi-Aspect Controllable Text Generation"], "answer_arxiv_id": ["2212.09387"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16040"} +{"question": "What works have found hard-parameter sharing for multi-task learning to be effective on time series, language and graph data?", "answer": ["Multi-Task Deep Neural Networks for Natural Language Understanding", "Strategies for Pre-training Graph Neural Networks"], "answer_arxiv_id": ["1901.11504", "1905.12265"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_16041"} +{"question": "What are some references that provide insights into parameter sharing in neural networks?", "answer": ["ACDC: Weight Sharing in Atom-Coefficient Decomposed Convolution", "Rethinking Hard-Parameter Sharing in Multi-Domain Learning", "Task Adaptive Parameter Sharing for Multi-Task Learning"], "answer_arxiv_id": ["2009.02386", "2107.11359", "2203.16708"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_16042"} +{"question": "Could you provide me some studies about directly sampling image features with queries for 3D object detection?", "answer": ["DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries", "PETRv2: A Unified Framework for 3D Perception from Multi-Camera Images", "CAPE: Camera View Position Embedding for Multi-View 3D Object Detection", "Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D\n Object Detection"], "answer_arxiv_id": ["2110.06922", "2206.01256", "2303.10209", "2303.11926"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_16043"} +{"question": "Could you provide some references on deep learning-based multi-view representation learning?", "answer": ["Reconsidering Representation Alignment for Multi-view Clustering"], "answer_arxiv_id": ["2103.07738"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_16044"} +{"question": "Which studies have used non-parametric citation augmented strategies in LLMs?", "answer": ["Enabling Large Language Models to Generate Text with Citations", "Teaching language models to support answers with verified quotes", "Leveraging Passage Retrieval with Generative Models for Open Domain\n Question Answering"], "answer_arxiv_id": ["2305.14627", "2203.11147", "2007.01282"], "source_meta": {"published_time": "20240225"}, "qid": "AutoScholarQuery_train_16045"} +{"question": "What are some works that introduced scalable and memory-efficient map representations in MLP-based NeRF-based SLAM?", "answer": ["iMAP: Implicit Mapping and Positioning in Real-Time"], "answer_arxiv_id": ["2103.12352"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_16046"} +{"question": "Could you provide some state-of-the-art planning methods using diffusion models?", "answer": ["MetaDiffuser: Diffusion Model as Conditional Planner for Offline Meta-RL", "Motion Planning Diffusion: Learning and Planning of Robot Motions with\n Diffusion Models", "Learning Universal Policies via Text-Guided Video Generation", "AdaptDiffuser: Diffusion Models as Adaptive Self-evolving Planners", "Diffusion Policy: Visuomotor Policy Learning via Action Diffusion"], "answer_arxiv_id": ["2305.19923", "2308.01557", "2302.00111", "2302.01877", "2303.04137"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_16047"} +{"question": "Could you provide me some works about attacking and defending the federated learning system?", "answer": ["Gradient Disaggregation: Breaking Privacy in Federated Learning by Reconstructing the User Participant Matrix"], "answer_arxiv_id": ["2106.06089"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_16048"} +{"question": "Which works take inspiration from human mesh recovery and use the MANO parametric hand model?", "answer": ["3D Hand Shape and Pose from Images in the Wild", "Pushing the Envelope for RGB-based Dense 3D Hand Pose Estimation via\n Neural Rendering", "End-to-end Hand Mesh Recovery from a Monocular RGB Image"], "answer_arxiv_id": ["1902.03451", "1904.04196", "1902.09305"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_16049"} +{"question": "What papers are about implementing Adapter tuning as part of their PETL methods?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "AdaptFormer: Adapting Vision Transformers for Scalable Visual\n Recognition"], "answer_arxiv_id": ["1902.00751", "2205.13535"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_16050"} +{"question": "Who proposed SE(3)-equivariant neural network architectures using the atomic cluster expansion framework?", "answer": ["Atomic Cluster Expansion: Completeness, Efficiency and Stability", "Learning Local Equivariant Representations for Large-Scale Atomistic Dynamics"], "answer_arxiv_id": ["1911.03550", "2204.05249"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_16051"} +{"question": "Who focused on improving out-of-distribution generalization by training multiple models followed by weight-averaging?", "answer": ["DART: Diversify-Aggregate-Repeat Training Improves Generalization of\n Neural Networks", "Diverse Weight Averaging for Out-of-Distribution Generalization", "Ensemble of Averages: Improving Model Selection and Boosting Performance\n in Domain Generalization", "Model Ratatouille: Recycling Diverse Models for Out-of-Distribution\n Generalization"], "answer_arxiv_id": ["2302.14685", "2205.09739", "2110.10832", "2212.10445"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_16052"} +{"question": "Which papers discussed using transformer architectures for VideoQA methods?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_16053"} +{"question": "Which works are about applications of Shapley values to machine learning model explanations and feature importance?", "answer": ["A Unified Approach to Interpreting Model Predictions", "Consistent Individualized Feature Attribution for Tree Ensembles", "The Many Shapley Values for Model Explanation", "Shapley Flow: A Graph-based Approach to Interpreting Model Predictions", "Interpreting Multivariate Shapley Interactions in DNNs", "Asymmetric Shapley values: incorporating causal knowledge into model-agnostic explainability", "On Explainability of Graph Neural Networks via Subgraph Explorations", "Understanding Global Feature Contributions With Additive Importance Measures"], "answer_arxiv_id": ["1705.07874", "1802.03888", "1908.08474", "2010.14592", "2010.05045", "1910.06358", "2102.05152", "2004.00668"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_16054"} +{"question": "What works discuss ASC and SSC conditions in identifiability analysis of nonnegative matrix factorization?", "answer": ["On Identifiability of Nonnegative Matrix Factorization", "Self-Dictionary Sparse Regression for Hyperspectral Unmixing: Greedy Pursuit and Pure Pixel Search are Related", "Nonnegative Matrix Factorization for Signal and Data Analytics: [Identifiability, Algorithms, and Applications]", "Robust Volume Minimization-Based Matrix Factorization for Remote Sensing and Document Clustering", "The Why and How of Nonnegative Matrix Factorization", "Robust Near-Separable Nonnegative Matrix Factorization Using Linear Optimization", "Memory-Efficient Convex Optimization for Self-Dictionary Separable Nonnegative Matrix Factorization: A Frank-Wolfe Approach"], "answer_arxiv_id": ["1709.00614", "1409.4320v2", "1803.01257", "1608.04290v1", "1401.5226", "1302.4385", "2109.11135"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16055"} +{"question": "Who proposed the modification of the latent representation to generate bias-conflicting samples?", "answer": ["Latent Adversarial Debiasing: Mitigating Collider Bias in Deep Neural Networks"], "answer_arxiv_id": ["2011.11486"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_16056"} +{"question": "What studies proposed pre-processing approaches in fair learning techniques?", "answer": ["Obtaining fairness using optimal transport theory", "Identifying and Correcting Label Bias in Machine Learning"], "answer_arxiv_id": ["1806.03195", "1901.04966"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_16057"} +{"question": "Which works discuss the ideas behind approximate message passing algorithms being rooted in physics of spin glasses?", "answer": ["An iterative construction of solutions of the TAP equations for the Sherrington-Kirkpatrick model", "Statistical physics of inference: Thresholds and algorithms"], "answer_arxiv_id": ["1201.2891", "1511.02476"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_16058"} +{"question": "Which work used the recent Conffusion method in generative models for fast confidence interval prediction?", "answer": ["Conffusion: Confidence Intervals for Diffusion Models"], "answer_arxiv_id": ["2211.09795"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_16059"} +{"question": "Could you provide me some works about fine-tuning pre-trained LLMs with contrastive loss using positive query-corpus paired data?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than\n In-Context Learning", "Promptagator: Few-shot Dense Retrieval From 8 Examples"], "answer_arxiv_id": ["2106.09685", "2205.05638", "2209.11755"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_16060"} +{"question": "Could you provide me some works about prompt-based methods for eliciting reasoning behaviour in LLMs?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2201.11903", "2205.10625", "2203.11171"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_16061"} +{"question": "Which works focused on developing strategies for 3D object detection?", "answer": ["Deep Hough Voting for 3D Object Detection in Point Clouds", "MLCVNet: Multi-Level Context VoteNet for 3D Object Detection", "Learning Geometry-Guided Depth via Projective Modeling for Monocular 3D Object Detection", "Delving into Localization Errors for Monocular 3D Object Detection", "H3DNet: 3D Object Detection Using Hybrid Geometric Primitives", "Back-tracing Representative Points for Voting-based 3D Object Detection in Point Clouds"], "answer_arxiv_id": ["1904.09664", "2004.05679", "2107.13931", "2103.16237", "2006.05682", "2104.06114"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_16062"} +{"question": "What papers described the use of BERT for predicting randomly-masked parts of neural spiking to extract motor neuron representations?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Representation learning for neural population activity with Neural Data Transformers"], "answer_arxiv_id": ["1810.04805", "2108.01210"], "source_meta": {"published_time": "20230812"}, "qid": "AutoScholarQuery_train_16063"} +{"question": "In what papers does the researcher use graph neural networks in learning-based provers?", "answer": ["Learning to Prove Theorems via Interacting with Proof Assistants", "Graph Representations for Higher-Order Logic and Theorem Proving"], "answer_arxiv_id": ["1905.09381", "1905.10006"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_16064"} +{"question": "What papers focused on developing GNN-enhanced crime predictive models to encode crime dynamics?", "answer": ["Spatial-Temporal Sequential Hypergraph Network for Crime Prediction with Dynamic Multiplex Relation Learning", "HAGEN: Homophily-Aware Graph Convolutional Recurrent Network for Crime Forecasting"], "answer_arxiv_id": ["2201.02435", "2109.12846"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_16065"} +{"question": "Which papers introduced different propagation methods in decoupled GCN?", "answer": ["Predict then Propagate: Graph Neural Networks meet Personalized PageRank", "Simplifying Graph Convolutional Networks"], "answer_arxiv_id": ["1810.05997", "1902.07153"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_16066"} +{"question": "Could you name some UDA methods available for real-world SR?", "answer": ["Unsupervised Image Super-Resolution using Cycle-in-Cycle Generative\n Adversarial Networks", "Closed-loop Matters: Dual Regression Networks for Single Image\n Super-Resolution", "Unsupervised Real-world Image Super Resolution via Domain-distance Aware\n Training", "Dual Adversarial Adaptation for Cross-Device Real-World Image\n Super-Resolution"], "answer_arxiv_id": ["1809.00437", "2003.07018", "2004.01178", "2205.03524"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_16067"} +{"question": "Which works theoretically analyzed how different architectural parameters impact expressivity and implicit regularization using separation rank?", "answer": ["On the Expressive Power of Deep Learning: A Tensor Analysis", "Convolutional Rectifier Networks as Generalized Tensor Decompositions", "Inductive Bias of Deep Convolutional Networks through Pooling Geometry", "Boosting Dilated Convolutional Networks with Mixed Tensor Decompositions", "On the Expressive Power of Overlapping Architectures of Deep Learning", "Deep Learning and Quantum Entanglement: Fundamental Connections with Implications to Network Design", "Expressive power of recurrent neural networks", "Generalized Tensor Models for Recurrent Neural Networks", "The Inductive Bias of In-Context Learning: Rethinking Pretraining Example Design", "Implicit Regularization in Deep Learning May Not Be Explainable by Norms", "Implicit Regularization in Tensor Factorization", "Implicit Regularization in Hierarchical Tensor Factorization and Deep Convolutional Neural Networks"], "answer_arxiv_id": ["1509.05009", "1603.00162", "1605.06743", "1703.06846v3", "1703.02065", "1704.01552v2", "1711.00811", "1901.10801", "2110.04541", "2005.06398", "2102.09972", "2201.11729"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_16068"} +{"question": "Could you provide me some studies on safe RL by control-based approaches?", "answer": ["Safe Model-based Reinforcement Learning with Stability Guarantees", "A Lyapunov-based Approach to Safe Reinforcement Learning", "Safe Exploration in Continuous Action Spaces", "Enforcing Hard Constraints with Soft Barriers: Safe Reinforcement Learning in Unknown Stochastic Environments"], "answer_arxiv_id": ["1705.08551", "1805.07708", "1801.08757", "2209.15090"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_16069"} +{"question": "Which papers used test-time training for better adaptation to each specific instance?", "answer": ["Learning to Adapt for Stereo", "Real-time self-adaptive deep stereo", "Tent: Fully Test-time Adaptation by Entropy Minimization", "Improving robustness against common corruptions by covariate shift adaptation", "Test-Time Training with Self-Supervision for Generalization under\n Distribution Shifts", "Test-Time Training with Masked Autoencoders"], "answer_arxiv_id": ["1904.02957", "1810.05424", "2006.10726", "2006.16971v2", "1909.13231", "2209.07522"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_16070"} +{"question": "Which methods worked on reconstructing humans in 3D by leveraging parametric body models?", "answer": ["Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a\n Single Image", "End-to-end Recovery of Human Shape and Pose", "Learning to Estimate 3D Human Pose and Shape from a Single Color Image", "Neural Body Fitting: Unifying Deep Learning and Model-Based Human Pose\n and Shape Estimation", "Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the\n Loop", "Synthetic Training for Accurate 3D Human Pose and Shape Estimation in the Wild", "PARE: Part Attention Regressor for 3D Human Body Estimation", "Neural Descent for Visual 3D Human Pose and Shape", "THUNDR: Transformer-based 3D HUmaN Reconstruction with Markers"], "answer_arxiv_id": ["1607.08128", "1712.06584", "1805.04092", "1808.05942", "1909.12828", "2009.10013v2", "2104.08527", "2008.06910", "2106.09336"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_16071"} +{"question": "Could you provide me an example of a study that made improvements in the synthesis quality of vector fonts using a transformer architecture?", "answer": ["DeepVecFont-v2: Exploiting Transformers to Synthesize Vector Fonts with\n Higher Quality"], "answer_arxiv_id": ["2303.14585"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_16072"} +{"question": "Could you give an example of a work that introduced a method for debiasing VLMs using synthetically-constructed contrast sets?", "answer": ["Balancing the Picture: Debiasing Vision-Language Datasets with Synthetic\n Contrast Sets"], "answer_arxiv_id": ["2305.15407"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16073"} +{"question": "What are some studies that focused on human-object and human-scene interactions reconstruction from monocular input?", "answer": ["Perceiving 3D Human-Object Spatial Arrangements from a Single Image in\n the Wild", "Resolving 3D Human Pose Ambiguities with 3D Scene Constraints", "Generating 3D People in Scenes without People", "Long-term Human Motion Prediction with Scene Context", "Stochastic Scene-Aware Motion Prediction", "BEHAVE: Dataset and Method for Tracking Human Object Interactions", "InterCap: Joint Markerless 3D Tracking of Humans and Objects in\n Interaction", "Capturing and Inferring Dense Full-Body Human-Scene Contact", "HUMANISE: Language-conditioned Human Motion Generation in 3D Scenes", "ARCTIC: A Dataset for Dexterous Bimanual Hand-Object Manipulation", "Full-Body Articulated Human-Object Interaction", "CHORE: Contact, Human and Object REconstruction from a single RGB image", "StackFLOW: Monocular Human-Object Reconstruction by Stacked Normalizing Flow with Offset", "Visibility Aware Human-Object Interaction Tracking from Single RGB\n Camera"], "answer_arxiv_id": ["2007.15649", "1908.06963", "1912.02923", "2007.03672", "2108.08284", "2204.06950", "2209.12354", "2206.09553", "2210.09729", "2204.13662", "2212.10621", "2204.02445", "2407.20545v1", "2303.16479"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_16074"} +{"question": "Which papers discuss high-performing text-to-image diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["2112.10752", "2205.11487", "2204.06125", "2203.13131"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_16075"} +{"question": "Are there any works about data construction for IFT?", "answer": ["Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low\n Training Data Instruction Tuning", "Rethinking Data Selection for Supervised Fine-Tuning"], "answer_arxiv_id": ["2305.09246", "2402.06094"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_16076"} +{"question": "Any studies have proposed methods for HBox-supervised instance segmentation?", "answer": ["Simple Does It: Weakly Supervised Instance and Semantic Segmentation", "BoxInst: High-Performance Instance Segmentation with Box Annotations", "Box-supervised Instance Segmentation with Level Set Evolution", "Segment Anything"], "answer_arxiv_id": ["1603.07485", "2012.02310", "2207.09055", "2401.14159"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_16077"} +{"question": "Any works about generating high-resolution images based on the CLIP text encoder?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2204.06125"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_16078"} +{"question": "Which studies propose modular multi-step approaches for multi-view stereo (MVS)?", "answer": ["MVSNet: Depth Inference for Unstructured Multi-view Stereo", "Recurrent MVSNet for High-resolution Multi-view Stereo Depth Inference", "Visibility-aware Multi-view Stereo Network", "Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching", "DELTAS: Depth Estimation by Learning Triangulation And densification of Sparse points", "SurfaceNet: An End-to-end 3D Neural Network for Multiview Stereopsis", "Atlas: End-to-End 3D Scene Reconstruction from Posed Images"], "answer_arxiv_id": ["1804.02505", "1902.10556", "2008.07928", "1912.06378", "2003.08933", "1708.01749", "2003.10432"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_16079"} +{"question": "Could you name the works that leveraged claim decomposition strategy for fine-grained verification?", "answer": ["Revisiting the Gold Standard: Grounding Summarization Evaluation with\n Robust Human Evaluation", "WiCE: Real-World Entailment for Claims in Wikipedia", "Complex Claim Verification with Evidence Retrieved in the Wild", "FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long\n Form Text Generation"], "answer_arxiv_id": ["2212.07981", "2303.01432", "2305.11859", "2305.14251"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_16080"} +{"question": "What works have been done on improving the functionality of LLMs in generating summaries in different formats?", "answer": ["LLM is Like a Box of Chocolates: the Non-determinism of ChatGPT in Code\n Generation"], "answer_arxiv_id": ["2308.02828"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_16081"} +{"question": "Could you provide me with studies about Incremental Learning methods?", "answer": ["Re-evaluating Continual Learning Scenarios: A Categorization and Case\n for Strong Baselines", "Three scenarios for continual learning"], "answer_arxiv_id": ["1810.12488", "1904.07734"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_16082"} +{"question": "Which work improved language models by retrieving tokens?", "answer": ["Improving language models by retrieving from trillions of tokens"], "answer_arxiv_id": ["2112.04426"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_16083"} +{"question": "What papers provide the theoretical means to the evolution of the network parameters via a Wasserstein gradient flow?", "answer": ["A Mean Field View of the Landscape of Two-Layer Neural Networks", "Gradient Descent on Infinitely Wide Neural Networks: Global Convergence and Generalization", "On the Global Convergence of Gradient Descent for Over-parameterized Models using Optimal Transport"], "answer_arxiv_id": ["1804.06561", "2110.08084", "1805.09545"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_16084"} +{"question": "What studies proposed extensions to model global absolute encodings?", "answer": ["Encoding word order in complex embeddings"], "answer_arxiv_id": ["1912.12333"], "source_meta": {"published_time": "20210222"}, "qid": "AutoScholarQuery_train_16085"} +{"question": "Which studies begin their data curation pipelines using Common Crawl as the internet source?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "mT5: A massively multilingual pre-trained text-to-text transformer", "Unsupervised Cross-lingual Representation Learning at Scale", "The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora\n with Web Data, and Web Data Only"], "answer_arxiv_id": ["1910.10683", "2010.11934", "1911.02116", "2306.01116"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_16086"} +{"question": "Which works aim to reduce the latent separation between poison and clean samples in the context of adaptive backdoor attacks?", "answer": ["Bypassing Backdoor Detection Algorithms in Deep Learning", "Enhancing Backdoor Attacks with Multi-Level MMD Regularization", "Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification", "Imperceptible Backdoor Attack: From Input Space to Feature Representation"], "answer_arxiv_id": ["1905.13409", "2111.05077", "2012.11212", "2205.03190"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_16087"} +{"question": "Are there any works pointed out the overly optimistic behavior of DT in RCSL?", "answer": ["Addressing Optimism Bias in Sequence Modeling for Reinforcement Learning"], "answer_arxiv_id": ["2207.10295"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_16088"} +{"question": "Are there any research papers that proposed training approaches other than maximizing the log-likelihood for neural TPPs?", "answer": ["Learning Temporal Point Processes via Reinforcement Learning", "Exploring Generative Neural Temporal Point Process", "Wasserstein Learning of Deep Generative Point Process Models"], "answer_arxiv_id": ["1811.05016", "2208.01874", "1705.08051"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_16089"} +{"question": "Which work used the post-processing technique to improve the recognition result?", "answer": ["FastCorrect: Fast Error Correction with Edit Alignment for Automatic\n Speech Recognition"], "answer_arxiv_id": ["2105.03842"], "source_meta": {"published_time": "20240210"}, "qid": "AutoScholarQuery_train_16090"} +{"question": "What research has been done on data augmentation methods in the field of pseudo-labeling and consistency regularization?", "answer": ["Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning", "Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning", "MixMatch: A Holistic Approach to Semi-Supervised Learning", "Unsupervised Data Augmentation for Consistency Training", "RandAugment: Practical automated data augmentation with a reduced search space"], "answer_arxiv_id": ["1606.04586", "1704.03976", "1905.02249", "1904.12848", "1909.13719"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_16091"} +{"question": "Which study developed a KG-to-Text approach to create high quality prompts and ultimately improved LLM performance in KG-based question answering?", "answer": ["Retrieve-Rewrite-Answer: A KG-to-Text Enhanced LLMs Framework for\n Knowledge Graph Question Answering"], "answer_arxiv_id": ["2309.11206"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_16092"} +{"question": "Who proposed the use of memory-guided attention to incorporate category information into the domain adaptation process?", "answer": ["MeGA-CDA: Memory Guided Attention for Category-Aware Unsupervised Domain Adaptive Object Detection"], "answer_arxiv_id": ["2103.04224"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_16093"} +{"question": "Which research proposes the use of the (ridge) leverage score distribution for quadrature node sampling?", "answer": ["On the Equivalence between Kernel Quadrature Rules and Random Feature Expansions"], "answer_arxiv_id": ["1502.06800"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_16094"} +{"question": "Which papers discuss the concept of 'Bellman-completeness' in the context of OPE?", "answer": ["Information-Theoretic Considerations in Batch Reinforcement Learning"], "answer_arxiv_id": ["1905.00360"], "source_meta": {"published_time": "20230725"}, "qid": "AutoScholarQuery_train_16095"} +{"question": "Which study uses 4-bit quantization in LoRA?", "answer": ["QLoRA: Efficient Finetuning of Quantized LLMs"], "answer_arxiv_id": ["2305.14314"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16096"} +{"question": "Are there any work exploring CLIP for IQA?", "answer": ["Exploring CLIP for Assessing the Look and Feel of Images", "Blind Image Quality Assessment via Vision-Language Correspondence: A\n Multitask Learning Perspective", "Advancing Zero-Shot Digital Human Quality Assessment through\n Text-Prompted Evaluation", "VILA: Learning Image Aesthetics from User Comments with Vision-Language\n Pretraining"], "answer_arxiv_id": ["2207.12396", "2303.14968", "2307.02808", "2303.14302"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_16097"} +{"question": "Could you provide me some studies about price of anarchy in autobidding frameworks?", "answer": ["Towards Efficient Auctions in an Auto-bidding World", "Robust Auction Design in the Auto-bidding World", "Auction Design in an Auto-bidding Setting: Randomization Improves Efficiency Beyond VCG", "Efficiency of the First-Price Auction in the Autobidding World"], "answer_arxiv_id": ["2103.13356", "2111.02468", "2204.10956", "2208.10650"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_16098"} +{"question": "Which publications offer insights on using bootstrapping to improve standard CoT?", "answer": ["STaR: Bootstrapping Reasoning With Reasoning"], "answer_arxiv_id": ["2203.14465"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_16099"} +{"question": "Which papers explore the generation of probabilistic labels in weak supervision?", "answer": ["Snorkel: Rapid Training Data Creation with Weak Supervision", "Training Complex Models with Multi-Task Weak Supervision", "Fast and Three-rious: Speeding Up Weak Supervision with Triplet Methods", "Universalizing Weak Supervision", "Learning Hyper Label Model for Programmatic Weak Supervision", "Lifting Weak Supervision To Structured Prediction"], "answer_arxiv_id": ["1711.10160", "1810.02840v2", "2002.11955", "2112.03865", "2207.13545", "2211.13375"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_16100"} +{"question": "Which works have utilized pre-trained generator models and CLIP for text-driven image manipulation?", "answer": ["Paint by Word", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "FuseDream: Training-Free Text-to-Image Generation with Improved CLIP+GAN\n Space Optimization", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "CLIPstyler: Image Style Transfer with a Single Text Condition", "Text2LIVE: Text-Driven Layered Image and Video Editing"], "answer_arxiv_id": ["2103.10951", "2108.00946", "2112.01573", "2103.17249", "2112.00374", "2204.02491"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16101"} +{"question": "What is the largest flow physics dataset and who proposed it?", "answer": ["A public turbulence database cluster and applications to study Lagrangian evolution of velocity increments in turbulence"], "answer_arxiv_id": ["0804.1703v1"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_16102"} +{"question": "What research proposes mapping the value function constraining problem as constraining densities of state visitation?", "answer": ["Density Constrained Reinforcement Learning"], "answer_arxiv_id": ["2106.12764"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_16103"} +{"question": "What studies employed feature mimicking in knowledge distillation?", "answer": ["FitNets: Hints for Thin Deep Nets", "Distilling Object Detectors via Decoupled Features"], "answer_arxiv_id": ["1412.6550", "2103.14475"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16104"} +{"question": "Which works evaluate model performance on program synthesis tasks with Python as the target language?", "answer": ["Measuring Coding Challenge Competence With APPS", "DS-1000: A Natural and Reliable Benchmark for Data Science Code\n Generation", "Execution-based Evaluation for Data Science Code Generation Models", "CodeGen: An Open Large Language Model for Code with Multi-Turn Program\n Synthesis", "Training and Evaluating a Jupyter Notebook Data Science Assistant"], "answer_arxiv_id": ["2105.09938", "2211.11501", "2211.09374", "2203.13474", "2201.12901"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_16105"} +{"question": "What papers have made connections between generative models and contrastive learning?", "answer": ["Generative Models as a Data Source for Multiview Representation Learning"], "answer_arxiv_id": ["2106.05258"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_16106"} +{"question": "Which papers offer methods to handle Multivariate Irregularly Sampled Time Series (MISTS)?", "answer": ["A Transformer-based Framework for Multivariate Time Series Representation Learning", "Recurrent Neural Networks for Multivariate Time Series with Missing Values", "Set Functions for Time Series", "Graph-Guided Network for Irregularly Sampled Multivariate Time Series"], "answer_arxiv_id": ["2010.02803", "1606.01865", "1909.12064", "2110.05357"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_16107"} +{"question": "What research papers compare to notable works like the graph convolutional network (GCN), the graph attention network (GAT), the graph isomorphism network (GIN) and the principal neighbourhood aggregation (PNA)?", "answer": ["Graph Attention Networks", "How Powerful are Graph Neural Networks?", "Principal Neighbourhood Aggregation for Graph Nets"], "answer_arxiv_id": ["1710.10903", "1810.00826", "2004.05718"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_16108"} +{"question": "Which studies proposed speculative decoding as a strategy for boosting inference speed of LLMs?", "answer": ["Accelerating Large Language Model Decoding with Speculative Sampling", "Fast Inference from Transformers via Speculative Decoding"], "answer_arxiv_id": ["2302.01318", "2211.17192"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_16109"} +{"question": "Can you provide me with work that discusses strategies to extend the passage retriever in long-document retrieval?", "answer": ["Approximate Nearest Neighbor Negative Contrastive Learning for Dense\n Text Retrieval", "Simple Local Attentions Remain Competitive for Long-Context Tasks", "SeDR: Segment Representation Learning for Long Documents Dense Retrieval"], "answer_arxiv_id": ["2007.00808", "2112.07210", "2211.10841"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_16110"} +{"question": "Are there any works that show the difficulty of learning depth-333 Boolean circuits under a uniform distribution?", "answer": ["From Local Pseudorandom Generators to Hardness of Learning"], "answer_arxiv_id": ["2101.08303"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_16111"} +{"question": "Who provides a comprehensive survey of large language model-based research on in-context learning?", "answer": ["A Survey on In-context Learning"], "answer_arxiv_id": ["2301.00234"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_16112"} +{"question": "What works contribute to the development of deep learning based GIQA methods?", "answer": ["RAPIQUE: Rapid and Accurate Video Quality Prediction of User Generated\n Content", "Blind Image Quality Assessment Using A Deep Bilinear Convolutional\n Neural Network"], "answer_arxiv_id": ["2101.10955", "1907.02665"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_16113"} +{"question": "Which work showed how to obtain O(T) and O(log T) regret guarantees for general convex sets?", "answer": ["Efficient Projection-Free Online Convex Optimization with Membership Oracle"], "answer_arxiv_id": ["2111.05818v1"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16114"} +{"question": "Which papers propose various architectures or designs for YOLO?", "answer": ["YOLOv6: A Single-Stage Object Detection Framework for Industrial\n Applications", "YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for\n real-time object detectors", "YOLOX: Exceeding YOLO Series in 2021", "PP-YOLO: An Effective and Efficient Implementation of Object Detector", "PP-YOLOE: An evolved version of YOLO"], "answer_arxiv_id": ["2209.02976", "2207.02696", "2107.08430", "2007.12099", "2203.16250"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_16115"} +{"question": "Could you provide me some works which apply differential privacy in machine learning?", "answer": ["Evaluating Differentially Private Machine Learning in Practice", "Bayesian Differential Privacy for Machine Learning"], "answer_arxiv_id": ["1902.08874", "1901.09697"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_16116"} +{"question": "What studies proposed methods for Multi-modal FAS, such as cross-modality attention or multi-modal adapters, for pretrained ViT?", "answer": ["FM-ViT: Flexible Modal Vision Transformers for Face Anti-Spoofing", "MA-ViT: Modality-Agnostic Vision Transformers for Face Anti-Spoofing", "Rethinking Vision Transformer and Masked Autoencoder in Multimodal Face\n Anti-Spoofing", "Visual Prompt Flexible-Modal Face Anti-Spoofing"], "answer_arxiv_id": ["2305.03277", "2304.07549", "2302.05744", "2307.13958"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16117"} +{"question": "Are there any works that focus on studying single-neuron behavior for mechanistic interpretability?", "answer": ["Toy Models of Superposition", "Finding Neurons in a Haystack: Case Studies with Sparse Probing"], "answer_arxiv_id": ["2209.10652", "2305.01610"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_16118"} +{"question": "What papers have tackled learning robust or even causal explanations?", "answer": ["Robust and Stable Black Box Explanations", "Explanations can be manipulated and geometry is to blame", "Towards Robust Explanations for Deep Neural Networks", "Asymmetric Shapley values: incorporating causal knowledge into model-agnostic explainability", "Causal Shapley Values: Exploiting Causal Knowledge to Explain Individual Predictions of Complex Models", "Causal Mediation Analysis for Interpreting Neural NLP: The Case of Gender Bias"], "answer_arxiv_id": ["2011.06169", "1906.07983", "2012.10425", "1910.06358", "2011.01625", "2004.12265"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_16119"} +{"question": "What studies elaborated on concept-based methods in post-hoc explanations?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with\n Concept Activation Vectors (TCAV)", "CRAFT: Concept Recursive Activation FacTorization for Explainability"], "answer_arxiv_id": ["1711.11279", "2211.10154"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_16120"} +{"question": "What studies provide lower bounds on robust loss when the set of classifiers under consideration is all measurable functions?", "answer": ["Lower Bounds on Adversarial Robustness from Optimal Transport", "Lower Bounds on Cross-Entropy Loss in the Presence of Test-time Adversaries", "Adversarial Risk via Optimal Transport and Optimal Couplings"], "answer_arxiv_id": ["1909.12272", "2104.08382", "1912.02794v2"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_16121"} +{"question": "What is the work that adds sparsity to higher-order GNNs by proposing δ-k-LWL?", "answer": ["Weisfeiler and Leman go sparse: Towards scalable higher-order graph embeddings"], "answer_arxiv_id": ["1904.01543"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_train_16122"} +{"question": "Could you give me some papers that discuss examples of model editing?", "answer": ["Fast Model Editing at Scale", "Editing a classifier by rewriting its prediction rules", "Editable Neural Networks"], "answer_arxiv_id": ["2110.11309", "2112.01008", "2004.00345"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_16123"} +{"question": "Can you mention studies that work on the concepts of realistic neighborhoods by learning the data manifold for LIME?", "answer": ["P", "Fairwashing Explanations with Off-Manifold Detergent"], "answer_arxiv_id": ["0704.0320", "2007.09969"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_16124"} +{"question": "What works used generation for query expansion?", "answer": ["Generation-Augmented Retrieval for Open-Domain Question Answering"], "answer_arxiv_id": ["2009.08553"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_16125"} +{"question": "Which works introduce popular multilingual evaluation benchmarks?", "answer": ["XTREME: A Massively Multilingual Multi-task Benchmark for Evaluating\n Cross-lingual Generalization", "XGLUE: A New Benchmark Dataset for Cross-lingual Pre-training,\n Understanding and Generation"], "answer_arxiv_id": ["2003.11080", "2004.01401"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_16126"} +{"question": "What papers avoid texture ambiguity and huge computational costs in 3D reconstruction by relying on geometric details?", "answer": ["Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image", "MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2306.07881", "2309.03453", "2308.16512"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_16127"} +{"question": "Can you provide me with some works in which diffusion models were used in text-to-3D generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16128"} +{"question": "Which papers proposed personalized federated learning approach using local fine-tuning?", "answer": ["Federated Evaluation of On-device Personalization", "Salvaging Federated Learning by Local Adaptation"], "answer_arxiv_id": ["1910.10252", "2002.04758"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_16129"} +{"question": "Which work demonstrated a unified approach to solve these learning problems in optimal control?", "answer": ["Pontryagin Differentiable Programming: An End-to-End Learning and Control Framework"], "answer_arxiv_id": ["1912.12970v5"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_16130"} +{"question": "Can you provide any references where Gaussian process bandits have been extensively studied under a non-contextual setting?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design", "On Kernelized Multi-armed Bandits"], "answer_arxiv_id": ["0912.3995", "1704.00445"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_16131"} +{"question": "Which work achieved state-of-the-art performance on 3D instance segmentation using a transformer-based network?", "answer": ["Mask3D: Mask Transformer for 3D Semantic Instance Segmentation"], "answer_arxiv_id": ["2210.03105"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_16132"} +{"question": "Could you give me examples of research that applied diffusion models for tasks including face editing, animation, and body?", "answer": ["Collaborative Diffusion for Multi-Modal Face Generation and Editing", "Face Animation with an Attribute-Guided Diffusion Model", "Diffused Heads: Diffusion Models Beat GANs on Talking-Face Generation", "DINAR: Diffusion Inpainting of Neural Textures for One-Shot Human\n Avatars"], "answer_arxiv_id": ["2304.10530", "2304.03199", "2301.03396", "2303.09375"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16133"} +{"question": "What research uses graph neural networks to encode social influence into node embedding?", "answer": ["Influence maximization in unknown social networks: Learning Policies for Effective Graph Sampling", "Multi-task Learning for Influence Estimation and Maximization"], "answer_arxiv_id": ["1907.11625", "1904.08804"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_16134"} +{"question": "What researches further improve the relative positional encoding?", "answer": ["XLNet: Generalized Autoregressive Pretraining for Language Understanding", "DeBERTa: Decoding-enhanced BERT with Disentangled Attention"], "answer_arxiv_id": ["1906.08237", "2006.03654"], "source_meta": {"published_time": "20210222"}, "qid": "AutoScholarQuery_train_16135"} +{"question": "Any works about redesigning training objectives for improving likelihood estimation in diffusion models?", "answer": ["Variational Diffusion Models", "Maximum Likelihood Training of Score-Based Diffusion Models", "Score-based Generative Modeling in Latent Space"], "answer_arxiv_id": ["2107.00630", "2101.09258", "2106.05931"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16136"} +{"question": "What was the first work proposing Non-Autoregressive (NAR) method which improved generation speed and efficiency?", "answer": ["Non-Autoregressive Neural Machine Translation"], "answer_arxiv_id": ["1711.02281"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_16137"} +{"question": "What works have integrated the idea of GLOM with reversible networks in the learning of disentangled representations?", "answer": ["Reversible Column Networks"], "answer_arxiv_id": ["2212.11696"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_16138"} +{"question": "Which papers use models in model-based RL methods to search for optimal action sequences?", "answer": ["Local Search for Policy Iteration in Continuous Control", "Learning Latent Dynamics for Planning from Pixels", "Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models", "Dream to Control: Learning Behaviors by Latent Imagination", "Latent Skill Planning for Exploration and Transfer"], "answer_arxiv_id": ["2010.05545", "1811.04551", "1805.12114", "1912.01603", "2011.13897"], "source_meta": {"published_time": "20220918"}, "qid": "AutoScholarQuery_train_16139"} +{"question": "What works develop writing assistants allowing users to mark words that the system should omit?", "answer": ["QuickEdit: Editing Text & Translations by Crossing Words Out"], "answer_arxiv_id": ["1711.04805"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_16140"} +{"question": "Could you list some research which proposed variants for increasing the expressivity of PCFGs without relaxing the context-free assumptions?", "answer": ["Compound Probabilistic Context-Free Grammars for Grammar Induction"], "answer_arxiv_id": ["1906.10225"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_16141"} +{"question": "Could you give me examples of research on hyperparameter tuning and its federated counterparts in the context of bilevel models?", "answer": ["Hyperparameter Optimization: A Spectral Approach", "Advances and Open Problems in Federated Learning"], "answer_arxiv_id": ["1706.00764", "1912.04977"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_16142"} +{"question": "Which articles contributed to bounding Bayesian regret for Thompson Sampling (TS)?", "answer": ["An Information-Theoretic Analysis for Thompson Sampling with Many Actions", "First-Order Bayesian Regret Analysis of Thompson Sampling"], "answer_arxiv_id": ["1805.11845", "1902.00681"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16143"} +{"question": "Any research indicating that large LMs can self-verify their output in a few-shot setting for a range of tasks?", "answer": ["Language Models (Mostly) Know What They Know"], "answer_arxiv_id": ["2207.05221"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_16144"} +{"question": "Are there any works about quantum state tomography that can perform state tomography for an unknown quantum state?", "answer": ["Sample-optimal tomography of quantum states"], "answer_arxiv_id": ["1508.01797"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_16145"} +{"question": "Can you provide examples where modeling the forward dynamics of the system is considered as a pretraining objective?", "answer": ["Provable Representation Learning for Imitation with Contrastive Fourier Features", "Pretraining Representations for Data-Efficient Reinforcement Learning", "CURL: Contrastive Unsupervised Representations for Reinforcement Learning", "Playing hard exploration games by watching YouTube", "PI-ARS: Accelerating Evolution-Learned Visual-Locomotion with Predictive Information Representations", "Masked Trajectory Models for Prediction, Representation, and Control"], "answer_arxiv_id": ["2105.12272", "2106.04799", "2004.04136", "1805.11592", "2207.13224", "2305.02968"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_16146"} +{"question": "Which work has shown how sparse attention patterns can be achieved by self-attention units?", "answer": ["On the Expressive Power of Self-Attention Matrices"], "answer_arxiv_id": ["2106.03764"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_16147"} +{"question": "Can you refer me to some research on the use of self-training models in anomaly detection?", "answer": ["Self-trained Deep Ordinal Regression for End-to-End Video Anomaly Detection", "Latent Outlier Exposure for Anomaly Detection with Contaminated Data"], "answer_arxiv_id": ["2003.06780", "2202.08088"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16148"} +{"question": "What studies have been conducted on adaptive search algorithms?", "answer": ["Learning Causal Graphs with Small Interventions", "Active Structure Learning of Causal DAGs via Directed Clique Trees", "Verification and search algorithms for causal DAGs", "Subset verification and search algorithms for causal DAGs"], "answer_arxiv_id": ["1511.00041", "2011.00641", "2206.15374", "2301.03180"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_16149"} +{"question": "Could you provide examples of GAN-based methods for generating image variations?", "answer": ["Image-to-Image Translation with Conditional Adversarial Networks", "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["1611.07004", "1703.10593", "1812.04948"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16150"} +{"question": "What papers have worked on using auxiliary tasks to explore attentive knowledge on sub-trajectory in Offline RL?", "answer": ["Representation Matters: Offline Pretraining for Sequential Decision Making"], "answer_arxiv_id": ["2102.05815"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_16151"} +{"question": "What works have applied spectral analysis to kernels associated with standard convolutional architectures that include no skip connections?", "answer": ["On the Spectral Bias of Convolutional Neural Tangent and Gaussian Process Kernels", "Eigenspace Restructuring: a Principle of Space and Frequency in Neural Networks"], "answer_arxiv_id": ["2203.09255", "2112.05611v1"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_16152"} +{"question": "Which references discuss enforcing invariance conditions across training domains through a regularization term to learn invariant features?", "answer": ["Invariant Risk Minimization", "Out-of-Distribution Generalization via Risk Extrapolation (REx)", "Accounting for Unobserved Confounding in Domain Generalization", "On Calibration and Out-of-domain Generalization", "Invariant Causal Mechanisms through Distribution Matching"], "answer_arxiv_id": ["1907.02893", "2003.00688v5", "2007.10653", "2102.10395", "2206.11646"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_16153"} +{"question": "What research has explored the use of human feedback through annotation or selection of intended output for semantic parsing?", "answer": ["Learning a Neural Semantic Parser from User Feedback"], "answer_arxiv_id": ["1704.08760"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_16154"} +{"question": "Which work established a dataset named Known-Unknown Questions (KUQ) to evaluate the capability of Large Language Models (LLMs) to classify known and unknown questions?", "answer": ["Knowledge of Knowledge: Exploring Known-Unknowns Uncertainty with Large\n Language Models"], "answer_arxiv_id": ["2305.13712"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_16155"} +{"question": "What works propose learning the distribution with better mode coverage and video quality than GAN-based approaches?", "answer": ["Video Generation From Text", "VideoGPT: Video Generation using VQ-VAE and Transformers", "VideoFlow: A Conditional Flow-Based Model for Stochastic Video\n Generation", "Video Pixel Networks", "Scaling Autoregressive Video Models", "Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive\n Transformer"], "answer_arxiv_id": ["1710.00421", "2104.10157", "1903.01434", "1610.00527", "1906.02634", "2204.03638"], "source_meta": {"published_time": "20230826"}, "qid": "AutoScholarQuery_train_16156"} +{"question": "Which metric learning-based methods are used in few-shot learning?", "answer": ["Prototypical Networks for Few-shot Learning", "TADAM: Task dependent adaptive metric for improved few-shot learning", "Learning to Compare: Relation Network for Few-Shot Learning", "Cross Attention Network for Few-shot Classification", "Meta-Baseline: Exploring Simple Meta-Learning for Few-Shot Learning"], "answer_arxiv_id": ["1703.05175", "1805.10123", "1711.06025", "1910.07677", "2003.04390"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_16157"} +{"question": "Which studies aimed at identifying the OOD data in the testing set?", "answer": ["Energy-based Out-of-Distribution Detection for Graph Neural Networks"], "answer_arxiv_id": ["2302.02914"], "source_meta": {"published_time": "20221105"}, "qid": "AutoScholarQuery_train_16158"} +{"question": "Could you provide me some works about prompt learning in vision-language models?", "answer": ["Conditional Prompt Learning for Vision-Language Models", "Prompt Distribution Learning", "Learning to Prompt for Vision-Language Models", "Visual Prompt Tuning", "Unified Vision and Language Prompt Learning", "MaPLe: Multi-modal Prompt Learning", "Distribution-Aware Prompt Tuning for Vision-Language Models", "Prompt-aligned Gradient for Prompt Tuning", "Gradient-Regulated Meta-Prompt Learning for Generalizable\n Vision-Language Models", "Self-regulating Prompts: Foundational Model Adaptation without Forgetting"], "answer_arxiv_id": ["2203.05557", "2205.03340", "2109.01134", "2203.12119", "2210.07225", "2210.03117", "2309.03406", "2205.14865", "2303.06571", "2307.06948v2"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_16159"} +{"question": "What papers lack any generative component related to sequence design in the methods they propose?", "answer": ["DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking", "EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction"], "answer_arxiv_id": ["2210.01776", "2202.05146"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_16160"} +{"question": "Which resources provide real noisy data for the evaluation of unsupervised denoising techniques?", "answer": ["Benchmarking Denoising Algorithms with Real Photographs", "Real-world Noisy Image Denoising: A New Benchmark", "A Poisson-Gaussian Denoising Dataset with Real Fluorescence Microscopy Images"], "answer_arxiv_id": ["1707.01313v1", "1804.02603v1", "1812.10366"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16161"} +{"question": "What research work introduced LLaVA to leverage the capabilities of GPT3.5 and GPT4?", "answer": ["Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_16162"} +{"question": "Which works propose solutions on outlier exposure (OE) for out-of-distribution detection?", "answer": ["Deep Anomaly Detection with Outlier Exposure"], "answer_arxiv_id": ["1812.04606"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_16163"} +{"question": "Which studies are related to learning a mapping between aligned language and trajectories using a human-generated dataset?", "answer": ["Listen, Attend, and Walk: Neural Mapping of Navigational Instructions to Action Sequences", "Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments", "Learning to Map Natural Language Instructions to Physical Quadcopter Control using Simulated Flight", "Imitating Interactive Intelligence", "ALFRED A Benchmark for Interpreting Grounded Instructions for Everyday Tasks", "Speaker-Follower Models for Vision-and-Language Navigation"], "answer_arxiv_id": ["1506.04089", "1711.07280", "1910.09664", "2012.05672", "1912.01734", "1806.02724"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_16164"} +{"question": "What studies focus on obtaining well-calibrated confidence scores for model predictions?", "answer": ["On Calibration of Modern Neural Networks"], "answer_arxiv_id": ["1706.04599"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_16165"} +{"question": "What are the pioneering works in zero-shot depth estimation?", "answer": ["Single-Image Depth Perception in the Wild"], "answer_arxiv_id": ["1604.03901"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_16166"} +{"question": "Could you refer me to some researches that focus on self-supervised learning that learns view-invariant representations?", "answer": ["Time-Contrastive Networks: Self-Supervised Learning from Video"], "answer_arxiv_id": ["1704.06888"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_16167"} +{"question": "Could you list the works that focus on hard prompt tuning?", "answer": ["Universal Adversarial Triggers for Attacking and Analyzing NLP", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts", "ALLSH: Active Learning Guided by Local Sensitivity and Hardness"], "answer_arxiv_id": ["1908.07125v3", "2010.15980", "2205.04980"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_16168"} +{"question": "Which works have proposed the Gumbel-Softmax trick?", "answer": ["Categorical Reparameterization with Gumbel-Softmax", "The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables"], "answer_arxiv_id": ["1611.01144", "1611.00712"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_16169"} +{"question": "Are there any studies that extended IG to constrained strongly-convex lower problems?", "answer": ["Alternating Implicit Projected SGD and Its Efficient Variants for Equality-constrained Bilevel Optimization", "Inexact bilevel stochastic gradient methods for constrained and unconstrained lower-level problems"], "answer_arxiv_id": ["2211.07096v2", "2110.00604"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_16170"} +{"question": "What research has been conducted on online learning with delayed feedback?", "answer": ["Gradient-free Online Learning in Games with Delayed Rewards"], "answer_arxiv_id": ["2006.10911"], "source_meta": {"published_time": "20230709"}, "qid": "AutoScholarQuery_train_16171"} +{"question": "Can you provide some studies related to 'Masksembles or Dropout Ensembles'?", "answer": ["Masksembles for Uncertainty Estimation"], "answer_arxiv_id": ["2012.08334"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_16172"} +{"question": "Any studies about improving the calibration of deep models during the training phase?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning", "A Simple Baseline for Bayesian Uncertainty in Deep Learning", "Improving model calibration with accuracy versus uncertainty optimization"], "answer_arxiv_id": ["1612.01474", "2003.07329", "1902.02476", "2012.07923"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16173"} +{"question": "Which studies have discussed significantly improved image quality and sample diversity by diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Variational Diffusion Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239", "2105.05233", "2107.00630", "2102.09672"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_16174"} +{"question": "Which works used supervised fine-tuning of RoBERTa for the development of A.I text detection classifier?", "answer": ["RoBERTa: A Robustly Optimized BERT Pretraining Approach", "Release Strategies and the Social Impacts of Language Models"], "answer_arxiv_id": ["1907.11692", "1908.09203"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_16175"} +{"question": "Could you provide me some studies which showed the effectiveness of data augmentation in visual reinforcement learning?", "answer": ["Reinforcement Learning with Augmented Data", "Automatic Data Augmentation for Generalization in Deep Reinforcement Learning", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Mastering Visual Continuous Control: Improved Data-Augmented Reinforcement Learning", "Generalization in Reinforcement Learning by Soft Data Augmentation", "Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation", "A Comprehensive Survey of Data Augmentation in Visual Reinforcement Learning", "Temporal Difference Learning for Model Predictive Control"], "answer_arxiv_id": ["2004.14990", "2006.12862v2", "2004.13649", "2107.09645", "2011.13389", "2107.00644v2", "2210.04561", "2203.04955"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_16176"} +{"question": "What works have proposed approaches for domain generalization via domain randomization or adversarial data augmentation?", "answer": ["Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Generalizing to Unseen Domains via Adversarial Data Augmentation", "Deep Domain-Adversarial Image Generation for Domain Generalisation"], "answer_arxiv_id": ["1703.06907", "1805.12018", "2003.06054"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_16177"} +{"question": "Could you provide studies that combats label noise through sample selection?", "answer": ["MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels", "Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels", "How does Disagreement Help Generalization against Label Corruption?", "Confident Learning: Estimating Uncertainty in Dataset Labels", "Searching to Exploit Memorization Effect in Learning with Noisy Labels", "Combating Noisy Labels by Agreement: A Joint Training Method with Co-Regularization"], "answer_arxiv_id": ["1712.05055", "1804.06872", "1901.04215", "1911.00068", "1911.02377", "2003.02752"], "source_meta": {"published_time": "20211018"}, "qid": "AutoScholarQuery_train_16178"} +{"question": "What studies adopt differential privacy-inspired definitions of approximate unlearning based on the goal of indistinguishability from retrain-from-scratch?", "answer": ["Remember What You Want to Forget: Algorithms for Machine Unlearning", "Descent-to-Delete: Gradient-Based Methods for Machine Unlearning"], "answer_arxiv_id": ["2103.03279v2", "2007.02923v1"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_16179"} +{"question": "Which works shifted to feature-based knowledge transfer in knowledge distillation?", "answer": ["A Comprehensive Overhaul of Feature Distillation", "TinyBERT: Distilling BERT for Natural Language Understanding", "FitNets: Hints for Thin Deep Nets", "ViTKD: Practical Guidelines for ViT feature knowledge distillation"], "answer_arxiv_id": ["1904.01866", "1909.10351", "1412.6550", "2209.02432"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_16180"} +{"question": "Which works discuss that the choice of the best time-series anomaly detection method depends on the dataset?", "answer": ["A review on outlier/anomaly detection in time series data"], "answer_arxiv_id": ["2002.04236"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_16181"} +{"question": "Which papers exploit the NTK framework for the generalization analysis of GNNs?", "answer": ["Graph Neural Tangent Kernel: Fusing Graph Neural Networks with Graph Kernels", "Optimization of Graph Neural Networks: Implicit Acceleration by Skip Connections and More Depth"], "answer_arxiv_id": ["1905.13192", "2105.04550"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_16182"} +{"question": "Any developments in Encoder-Decoder architecture with bidirectional attention mechanisms for conditional context handling in infilling?", "answer": ["BART: Denoising Sequence-to-Sequence Pre-training for Natural Language\n Generation, Translation, and Comprehension", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer"], "answer_arxiv_id": ["1910.13461", "1910.10683"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_16183"} +{"question": "What work proposed a new parameterization for State Space Models?", "answer": ["Efficiently Modeling Long Sequences with Structured State Spaces"], "answer_arxiv_id": ["2111.00396"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_16184"} +{"question": "What methods use low-rank decomposition and approximating to hessian diagonal to reduce the computational cost of the second order information?", "answer": ["Apollo: An Adaptive Parameter-wise Diagonal Quasi-Newton Method for Nonconvex Stochastic Optimization", "AdaHessian: An Adaptive Second Order Optimizer for Machine Learning", "Scalable Second Order Optimization for Deep Learning"], "answer_arxiv_id": ["2009.13586", "2006.00719", "2002.09018"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_16185"} +{"question": "Could you refer me to papers that focus on proving recurrence/oscillating behavior of learning dynamics in time-varying periodic games?", "answer": ["Online Learning in Periodic Zero-Sum Games"], "answer_arxiv_id": ["2111.03377"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_16186"} +{"question": "Which works focus on sampling-based two-player zero-sum Markov games?", "answer": ["Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity", "A Sharp Analysis of Model-based Reinforcement Learning with Self-Play", "Near-Optimal Reinforcement Learning with Self-Play", "Provable Self-Play Algorithms for Competitive Reinforcement Learning", "Solving Discounted Stochastic Two-Player Games with Near-Optimal Time and Sample Complexity", "Online Learning in Unknown Markov Games", "Online Reinforcement Learning in Stochastic Games", "Learning Zero-Sum Simultaneous-Move Markov Games Using Function Approximation and Correlated Equilibrium", "Almost Optimal Algorithms for Two-player Zero-Sum Linear Mixture Markov Games"], "answer_arxiv_id": ["2007.07461", "2010.01604", "2006.12007", "2002.04017", "1908.11071", "2010.15020", "1712.00579", "2002.07066", "2102.07404"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_16187"} +{"question": "Which work presented a network which re-purposes transformers into language-conditioned regressors?", "answer": ["End-to-End Dense Video Grounding via Parallel Regression"], "answer_arxiv_id": ["2109.11265"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_16188"} +{"question": "Which paper proved exponential convergence by assuming the function is locally quadratic in a vicinity around the global optimum?", "answer": ["Exponential Regret Bounds for Gaussian Process Bandits with Deterministic Observations"], "answer_arxiv_id": ["1206.6457"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16189"} +{"question": "Can you mention a few studies that focused on language agents that interact directly with the environment via text?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Inner Monologue: Embodied Reasoning through Planning with Language\n Models"], "answer_arxiv_id": ["2204.01691", "2207.05608"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16190"} +{"question": "Which papers focused on direct optimization on 3D data using like voxels and point clouds for explicit 3D representation learning?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space"], "answer_arxiv_id": ["1612.00593", "1706.02413"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_16191"} +{"question": "What research provides an efficient algorithm with low-regret to solve LQR problems with known cost functions and stochastic perturbations?", "answer": ["Regret Bounds for Robust Adaptive Control of the Linear Quadratic Regulator"], "answer_arxiv_id": ["1805.09388"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16192"} +{"question": "Which research studies conducted analysis about optimal transport with arbitrary strongly convex regularization?", "answer": ["Smooth and Sparse Optimal Transport"], "answer_arxiv_id": ["1710.06276"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16193"} +{"question": "Which studies have used room impulse responses to localize inanimate acoustic reflectors?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_16194"} +{"question": "Could you give me examples of research that employs summarization techniques to design compressors?", "answer": ["RECOMP: Improving Retrieval-Augmented LMs with Compression and Selective\n Augmentation", "Learning to Filter Context for Retrieval-Augmented Generation"], "answer_arxiv_id": ["2310.04408", "2311.08377"], "source_meta": {"published_time": "20240603"}, "qid": "AutoScholarQuery_train_16195"} +{"question": "Which works highlight that CLIP is miscalibrated and has imbalanced predictions, causing challenges in pseudolabeling?", "answer": ["Enabling Calibration In The Zero-Shot Inference of Large Vision-Language Models", "Debiased Learning from Naturally Imbalanced Pseudo-Labels"], "answer_arxiv_id": ["2303.12748", "2201.01490v2"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_16196"} +{"question": "In what works learning rate schedules for SGD, under fixed distribution, and for the setting of least squares have been studied?", "answer": ["The Step Decay Schedule: A Near Optimal, Geometrically Decaying Learning Rate Procedure For Least Squares", "Making the Last Iterate of SGD Information Theoretically Optimal"], "answer_arxiv_id": ["1904.12838v2", "1904.12443"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_16197"} +{"question": "What studies investigated the limitations of these supervised approaches in generalizing to other text domains?", "answer": ["Real or Fake? Learning to Discriminate Machine from Human Generated Text", "Release Strategies and the Social Impacts of Language Models"], "answer_arxiv_id": ["1906.03351", "1908.09203"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_16198"} +{"question": "What works determined that the expressive power of randomized nets is low with specific settings?", "answer": ["On the Power and Limitations of Random Features for Understanding Neural Networks"], "answer_arxiv_id": ["1904.00687"], "source_meta": {"published_time": "20200819"}, "qid": "AutoScholarQuery_train_16199"} +{"question": "Are there any works that investigated overparameterization in low-rank matrix recovery from a spectral initialization?", "answer": ["Rank Overspecified Robust Matrix Recovery: Subgradient Method and Exact Recovery"], "answer_arxiv_id": ["2109.11154v2"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_16200"} +{"question": "What work adopted a set of learnable adaption prompts and prepend them to word tokens at higher transformer layers for LLaMA fine-tuning?", "answer": ["LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init\n Attention"], "answer_arxiv_id": ["2303.16199"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16201"} +{"question": "What works modified the cross-attention map by altering the textual description for image manipulation?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2208.01626", "2211.12572"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_16202"} +{"question": "Can you list some works about short-term tracking methods?", "answer": ["End-to-end representation learning for Correlation Filter based tracking", "Discriminative Correlation Filter with Channel and Spatial Reliability"], "answer_arxiv_id": ["1704.06036", "1611.08461"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_16203"} +{"question": "Could you provide me some studies about NeRF-based parametric models in the context of 3D generative models?", "answer": ["HeadNeRF: A Real-time NeRF-based Parametric Head Model", "MoFaNeRF: Morphable Facial Neural Radiance Field", "Preface: A Data-driven Volumetric Prior for Few-shot Ultra\n High-resolution Face Synthesis"], "answer_arxiv_id": ["2112.05637", "2112.02308", "2309.16859"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_16204"} +{"question": "Who investigates the sparsity effect of large depth on the global minima of L2-regularized networks?", "answer": ["Implicit Regularization Towards Rank Minimization in ReLU Networks"], "answer_arxiv_id": ["2201.12760"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16205"} +{"question": "What work suggests that receiving a trajectory feedback at the end of each trajectory can hinder learning efficiency in long-horizon problems?", "answer": ["Learning Long-Term Reward Redistribution via Randomized Return Decomposition"], "answer_arxiv_id": ["2111.13485"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_16206"} +{"question": "What studies demonstrate that LLMs exhibit enhanced performance when they produce reasoning process traces along with the final answer?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Large Language Models are Zero-Shot Reasoners", "Measuring and Narrowing the Compositionality Gap in Language Models"], "answer_arxiv_id": ["2201.11903", "2112.00114", "2205.11916", "2210.03350"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_16207"} +{"question": "What papers have studied Wasserstein gradient flows of f-divergences in deep generative modeling?", "answer": ["Refining Deep Generative Models via Discriminator Gradient Flow"], "answer_arxiv_id": ["2012.00780"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_16208"} +{"question": "What research introduces a method for achieving tighter bounds on certain QBRMs by focusing the statistical power of the Berk-Jones bound?", "answer": ["Quantile Risk Control: A Flexible Framework for Bounding the Probability of High-Loss Predictions"], "answer_arxiv_id": ["2212.13629"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_16209"} +{"question": "Which work presented SimCLR that uses contrastive loss for effective representation learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_16210"} +{"question": "What studies proposed learning architectures operated directly on B-rep data structures in CAD representations?", "answer": ["BRepNet: A topological message passing system for solid models", "UV-Net: Learning from Boundary Representations"], "answer_arxiv_id": ["2104.00706", "2006.10211"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16211"} +{"question": "Which papers focused on optimizing the prediction sets’ efficiency in addition to coverage?", "answer": ["High-Quality Prediction Intervals for Deep Learning: A Distribution-Free, Ensembled Approach", "PAC Confidence Sets for Deep Neural Networks via Calibrated Prediction", "Finite-sample Efficient Conformal Prediction", "Learning Optimal Conformal Classifiers", "Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control", "Uncertainty Sets for Image Classifiers using Conformal Prediction", "Efficient and Differentiable Conformal Prediction with General Function Classes"], "answer_arxiv_id": ["1802.07167v3", "2001.00106", "2104.13871", "2110.09192", "2110.01052", "2009.14193", "2202.11091"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_16212"} +{"question": "Could you provide me some works that propose parametric methods focused on different approximations to the weight space metric matrix?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Online Structured Laplace Approximations For Overcoming Catastrophic Forgetting"], "answer_arxiv_id": ["1612.00796", "1805.07810"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_16213"} +{"question": "What works proposed regularization-based methods to address the incremental learning problem?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Continual Learning Through Synaptic Intelligence", "Learning without Forgetting"], "answer_arxiv_id": ["1612.00796", "1703.04200", "1606.09282"], "source_meta": {"published_time": "20220211"}, "qid": "AutoScholarQuery_train_16214"} +{"question": "Which works discuss off-policy RL?", "answer": ["Playing Atari with Deep Reinforcement Learning", "Continuous control with deep reinforcement learning"], "answer_arxiv_id": ["1312.5602", "1509.02971"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_16215"} +{"question": "What studies leverage a pre-trained CLIP model for retrieval and segmentation of personalized objects?", "answer": ["“This is my unicorn, Fluffy”: Personalizing frozen vision-language representations"], "answer_arxiv_id": ["2204.01694"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_16216"} +{"question": "Which works have focused on extending GFlowNets to stochastic environments?", "answer": ["Stochastic Generative Flow Networks", "Distributional GFlowNets with Quantile Flows"], "answer_arxiv_id": ["2302.09465", "2302.05793"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_16217"} +{"question": "What research found that Curriculum Learning strategy can achieve better generalization performance when the given training data is noisy?", "answer": ["MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels", "Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels"], "answer_arxiv_id": ["1712.05055", "1804.06872"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_16218"} +{"question": "Which papers deal with the utilization of auxiliary networks specifically in DKS and FRSKD?", "answer": ["Deeply-supervised Knowledge Synergy", "Refine Myself by Teaching Myself: Feature Refinement via Self-Knowledge\n Distillation"], "answer_arxiv_id": ["1906.00675", "2103.08273"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16219"} +{"question": "Which works have studied the effects of reducing the size of language model evaluation sets?", "answer": ["Efficient Benchmarking of Language Models", "Benchmarking Large Language Model Capabilities for Conditional\n Generation"], "answer_arxiv_id": ["2308.11696", "2306.16793"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_16220"} +{"question": "Could you provide me some studies about using a specifically-designed formal language to represent schemas in universal information extraction?", "answer": ["Unified Structure Generation for Universal Information Extraction"], "answer_arxiv_id": ["2203.12277"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_16221"} +{"question": "Which recent studies provided Chinese safety benchmarks for more comprehensive safety evaluation of LLMs?", "answer": ["Safety Assessment of Chinese Large Language Models", "CValues: Measuring the Values of Chinese Large Language Models from\n Safety to Responsibility"], "answer_arxiv_id": ["2304.10436", "2307.09705"], "source_meta": {"published_time": "20230913"}, "qid": "AutoScholarQuery_train_16222"} +{"question": "What studies reformulated the DRO problem into a max-min problem as a duality approach?", "answer": ["Data-Driven Distributionally Robust Optimization Using the Wasserstein Metric: Performance Guarantees and Tractable Reformulations", "Quantifying Distributional Model Risk via Optimal Transport", "Computational aspects of robust optimized certainty equivalents and option pricing", "Distributionally Robust Stochastic Optimization with Wasserstein Distance", "Coresets for Wasserstein Distributionally Robust Optimization Problems"], "answer_arxiv_id": ["1505.05116", "1604.01446", "1706.10186", "1604.02199v3", "2210.04260"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_16223"} +{"question": "Could you provide a work that discusses the gap between poison-only backdoor attack and latent separation?", "answer": ["Demon in the Variant: Statistical Analysis of DNNs for Robust Backdoor Contamination Detection"], "answer_arxiv_id": ["1908.00686"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_16224"} +{"question": "Can you specify research that tackled the problem of adapting to newly specified tasks often from the perspective of meta-learning?", "answer": ["Matching Networks for One Shot Learning", "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"], "answer_arxiv_id": ["1606.04080", "1703.03400"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_16225"} +{"question": "What paper noted that neural message passing can be derived through certain embedded inference on probabilistic graphical models?", "answer": ["Discriminative Embeddings of Latent Variable Models for Structured Data"], "answer_arxiv_id": ["1603.05629"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_16226"} +{"question": "Which papers introduced the GLM-tron algorithm to solve the isotonic regression in ReLU regression problem?", "answer": ["Efficient Learning of Generalized Linear and Single Index Models with Isotonic Regression"], "answer_arxiv_id": ["1104.2018"], "source_meta": {"published_time": "20220804"}, "qid": "AutoScholarQuery_train_16227"} +{"question": "Which papers propose strategies for unified frameworks for multi-modal perception?", "answer": ["Perceiver: General Perception with Iterative Attention", "Perceiver IO: A General Architecture for Structured Inputs & Outputs", "data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language", "Uni-Perceiver: Pre-training Unified Architecture for Generic Perception for Zero-shot and Few-shot Tasks", "Uni-Perceiver v2: A Generalist Model for Large-Scale Vision and Vision-Language Tasks"], "answer_arxiv_id": ["2103.03206", "2107.14795", "2202.03555", "2112.01522", "2211.09808"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_16228"} +{"question": "Could you give me an example of work that explores view-conditioned diffusion to learn control mechanisms that manipulate the camera viewpoint in large-scale diffusion models?", "answer": ["Zero-1-to-3: Zero-shot One Image to 3D Object"], "answer_arxiv_id": ["2303.11328"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_16229"} +{"question": "Can you name the study that pioneered the use of deep functional map methods to autonomously learn features from training data?", "answer": ["Deep Functional Maps: Structured Prediction for Dense Shape\n Correspondence"], "answer_arxiv_id": ["1704.08686"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_16230"} +{"question": "What papers show the results achieved with VLMs in language instruction and visual reasoning ability?", "answer": ["Visual Instruction Tuning", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models"], "answer_arxiv_id": ["2304.08485", "2305.06500", "2305.03726", "2304.10592"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_16231"} +{"question": "Could you provide me some studies that applied weak supervision to alleviate the lack of training data and labels for ICD coding?", "answer": ["Rare Disease Identification from Clinical Notes with Ontologies and Weak Supervision", "Classifying Unstructured Clinical Notes via Automatic Weak Supervision"], "answer_arxiv_id": ["2105.01995", "2206.12088"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_16232"} +{"question": "Can you mention the study that proposed a method to adapt the step size for large-batch training by estimating the gradient diversity?", "answer": ["AdaScale SGD: A User-Friendly Algorithm for Distributed Training"], "answer_arxiv_id": ["2007.05105"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_16233"} +{"question": "Which study showed that distilling knowledge in deep neural networks is an effective way to obtain better generalizing models?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_16234"} +{"question": "Which studies made significant contributions in hard prompt tuning, also known as prompt engineering?", "answer": ["AutoPrompt: Eliciting Knowledge from Language Models with Automatically\n Generated Prompts", "Large Language Models Are Human-Level Prompt Engineers", "Large Language Models as Optimizers"], "answer_arxiv_id": ["2010.15980", "2211.01910", "2309.03409"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_16235"} +{"question": "Which works provide examples of neural-symbolic reasoning using a symbolic-driven neural fashion?", "answer": ["Knowledge Graph Embedding with Iterative Guidance from Soft Rules", "Logic Rules Powered Knowledge Graph Embedding"], "answer_arxiv_id": ["1711.11231", "1903.03772"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_16236"} +{"question": "Any works that utilized Lie group theory to achieve equivariance?", "answer": ["Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data"], "answer_arxiv_id": ["2002.12880"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_16237"} +{"question": "What works used user input for rendering based methods in shadow generation?", "answer": ["Automatic Scene Inference for 3D Object Compositing", "Controllable Shadow Generation Using Pixel Height Maps", "PixHt-Lab: Pixel Height Based Light Effect Generation for Image\n Compositing"], "answer_arxiv_id": ["1912.12297", "2207.05385", "2303.00137"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_16238"} +{"question": "Which work used variance reduction to obtain an intermediate runtime?", "answer": ["Variance Reduction for Matrix Games"], "answer_arxiv_id": ["1907.02056"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_16239"} +{"question": "What studies aim to estimate the density functions of the score under null and alternative hypotheses, primarily through training shadow models?", "answer": ["Membership Inference Attacks Against Machine Learning Models", "White-box vs Black-box: Bayes Optimal Strategies for Membership Inference", "Systematic Evaluation of Privacy Risks of Machine Learning Models", "Membership Inference Attacks From First Principles"], "answer_arxiv_id": ["1610.05820", "1908.11229", "2003.10595", "2112.03570"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_16240"} +{"question": "What paper uses a data driven strategy to train neural networks to predict depth maps with depth supervision?", "answer": ["MVSNet: Depth Inference for Unstructured Multi-view Stereo"], "answer_arxiv_id": ["1804.02505"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_16241"} +{"question": "Which research papers have contributed to Masked Image Modeling (MIM) in the field of computer vision?", "answer": ["BEiT: BERT Pre-Training of Image Transformers", "SimMIM: A Simple Framework for Masked Image Modeling", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2106.08254", "2111.09886", "2111.06377"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_16242"} +{"question": "Which paper introduces a method to reduce the number of trainable parameters by using a shared frozen random LoRA module?", "answer": ["VeRA: Vector-based Random Matrix Adaptation"], "answer_arxiv_id": ["2310.11454"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16243"} +{"question": "Could you provide me with the studies that discuss Bayesian Optimization (BO) for hyperparameter tuning of deep neural networks?", "answer": ["Practical Bayesian Optimization of Machine Learning Algorithms"], "answer_arxiv_id": ["1206.2944"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_16244"} +{"question": "Can you provide some references that exploited deep generative models for the development of synthetic data?", "answer": ["Beyond Privacy: Navigating the Opportunities and Challenges of Synthetic Data", "DECAF: Generating Fair Synthetic Data Using Causally-Aware Generative Networks", "Data Augmentation Generative Adversarial Networks", "Effect of Balancing Data Using Synthetic Data on the Performance of Machine Learning Classifiers for Intrusion Detection in Computer Networks", "Conditional Synthetic Data Generation for Robust Machine Learning Applications with Limited Pandemic Data", "Conditional Generation of Medical Time Series for Extrapolation to Underrepresented Populations"], "answer_arxiv_id": ["2304.03722v1", "2110.12884", "1711.04340", "2204.00144", "2109.06486", "2201.08186"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_16245"} +{"question": "Which work is closest to using an automatic method to identify challenging subsets by fitting a Gaussian mixture model in CLIP space?", "answer": ["Domino: Discovering Systematic Errors with Cross-Modal Embeddings"], "answer_arxiv_id": ["2203.14960"], "source_meta": {"published_time": "20220629"}, "qid": "AutoScholarQuery_train_16246"} +{"question": "Could you provide me some works that show the alternate minimization approach for overcomplete dictionary learning?", "answer": ["New Algorithms for Learning Incoherent and Overcomplete Dictionaries", "Simple, Efficient, and Neural Algorithms for Sparse Coding", "Learning Sparsely Used Overcomplete Dictionaries via Alternating Minimization"], "answer_arxiv_id": ["1308.6273", "1503.00778", "1310.7991"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16247"} +{"question": "What works talk about self-competition in learning environments?", "answer": ["Emergent Complexity via Multi-Agent Competition", "Intrinsic Motivation and Automatic Curricula via Asymmetric Self-Play", "Automated Synthesis of Steady-State Continuous Processes using Reinforcement Learning"], "answer_arxiv_id": ["1710.03748", "1703.05407", "2101.04422v2"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_16248"} +{"question": "Could you provide me some studies about the use of distributional distances or Optimal Transport in data valuation?", "answer": ["LAVA: Data Valuation without Pre-Specified Learning Algorithms"], "answer_arxiv_id": ["2305.00054"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_16249"} +{"question": "What research proposed using large language models for designing and regularizing reward functions in text-based games?", "answer": ["Reward Design with Language Models", "Language Instructed Reinforcement Learning for Human-AI Coordination"], "answer_arxiv_id": ["2303.00001", "2304.07297"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_16250"} +{"question": "What work suggested optimizing the policy over the policy distribution space rather than the action space?", "answer": ["Distributional Policy Optimization: An Alternative Approach for Continuous Control"], "answer_arxiv_id": ["1905.09855"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_16251"} +{"question": "What work proposed a paradigm in image understanding that maps non-linear augmented views projections into a unit sphere of K classes?", "answer": ["Deep Clustering for Unsupervised Learning of Visual Features", "Self-labelling via simultaneous clustering and representation learning", "Unsupervised Pre-Training of Image Features on Non-Curated Data", "Learning Representations by Predicting Bags of Visual Words", "CliqueCNN: Deep Unsupervised Exemplar Learning"], "answer_arxiv_id": ["1807.05520", "1911.05371", "1905.01278", "2002.12247", "1608.08792"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_16252"} +{"question": "Which works extend diffusion models to realize image editing?", "answer": ["DiffEdit: Diffusion-based semantic image editing with mask guidance", "Prompt-to-Prompt Image Editing with Cross Attention Control", "InstructPix2Pix: Learning to Follow Image Editing Instructions"], "answer_arxiv_id": ["2210.11427", "2208.01626", "2211.09800"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_16253"} +{"question": "Which papers applied a supervised learning approach to seedless versions of graph matching problems?", "answer": ["Learning Combinatorial Embedding Networks for Deep Graph Matching", "Neural Graph Matching Network: Learning Lawler's Quadratic Assignment Problem with Extension to Hypergraph and Multiple-graph Matching", "Deep Graph Matching Consensus", "Deep Graph Matching via Blackbox Differentiation of Combinatorial Solvers", "Deep Graph Matching under Quadratic Constraint"], "answer_arxiv_id": ["1904.00597", "1911.11308", "2001.09621", "2003.11657", "2103.06643"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_16254"} +{"question": "Give examples of research that explores the decomposition of model parameters into base layers and personalized layers in PFL?", "answer": ["Federated Learning with Personalization Layers", "Exploiting Shared Representations for Personalized Federated Learning", "FedBABU: Towards Enhanced Representation for Federated Image\n Classification", "FedTP: Federated Learning by Transformer Personalization"], "answer_arxiv_id": ["1912.00818v1", "2102.07078", "2106.06042", "2211.01572"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16255"} +{"question": "What researches introduce regularization-based approaches in continual learning?", "answer": ["Learning without Forgetting", "Memory Aware Synapses: Learning what (not) to forget", "Overcoming catastrophic forgetting in neural networks", "Continual Learning Through Synaptic Intelligence"], "answer_arxiv_id": ["1606.09282", "1711.09601", "1612.00796", "1703.04200"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_16256"} +{"question": "What research introduces a new 'pseudo-word' to invert concepts in textual embedding space?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion"], "answer_arxiv_id": ["2208.01618"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_16257"} +{"question": "Which papers discussed the use of symmetry as a design principle for Geometric Deep Learning?", "answer": ["Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["2104.13478"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_16258"} +{"question": "What research papers have extensively studied the effects of batch normalization in deep neural networks?", "answer": ["A Mean Field Theory of Batch Normalization", "Batch Normalization Orthogonalizes Representations in Deep Random Networks", "On Bridging the Gap between Mean Field and Finite Width in Deep Random Neural Networks with Batch Normalization"], "answer_arxiv_id": ["1902.08129", "2106.03970", "2205.13076"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_16259"} +{"question": "What researches are about improving robustness in retrieval-augmented translation?", "answer": ["Improving Robustness of Retrieval Augmented Translation via Shuffling of\n Suggestions", "Rethinking Translation Memory Augmented Neural Machine Translation", "Improving Retrieval Augmented Neural Machine Translation by Controlling\n Source and Fuzzy-Match Interactions"], "answer_arxiv_id": ["2210.05059", "2306.06948", "2210.05047"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_16260"} +{"question": "Which research suggests that large step sizes prevent the increase of local curvature during the early phase of training?", "answer": ["Catastrophic Fisher Explosion: Early Phase Fisher Matrix Impacts Generalization"], "answer_arxiv_id": ["2012.14193"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16261"} +{"question": "Can you provide me some papers where deep neural networks are applied to TCR-peptide binding prediction?", "answer": ["TITAN: T Cell Receptor Specificity Prediction with Bimodal Attention Networks"], "answer_arxiv_id": ["2105.03323"], "source_meta": {"published_time": "20221015"}, "qid": "AutoScholarQuery_train_16262"} +{"question": "What papers have used videos to predict regions of interactions for robot learning?", "answer": ["Human Hands as Probes for Interactive Object Understanding", "Grounded Human-Object Interaction Hotspots from Video"], "answer_arxiv_id": ["2112.09120", "1812.04558"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16263"} +{"question": "What studies integrated weak supervision with continual learning in realistic scenarios?", "answer": ["Incremental Learning in Semantic Segmentation from Image Labels"], "answer_arxiv_id": ["2112.01882"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_16264"} +{"question": "Could you provide me some works studying non-monotone suodular maximization with cardinality constraint k in offline and secretary settings?", "answer": ["Constrained Non-Monotone Submodular Maximization: Offline and Secretary Algorithms"], "answer_arxiv_id": ["1003.1517"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_16265"} +{"question": "Could you provide me some studies involving the re-parameterization technique to boost the efficiency of networks?", "answer": ["MobileOne: An Improved One millisecond Mobile Backbone"], "answer_arxiv_id": ["2206.04040"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_16266"} +{"question": "Are there any studies where premise selection was tackled in isolation without theorem proving or the premises were fed to a symbolic prover?", "answer": ["Machine-Learned Premise Selection for Lean", "DeepMath - Deep Sequence Models for Premise Selection", "Learning to Prove Theorems by Learning to Generate Theorems", "Premise Selection for Mathematics by Corpus Analysis and Kernel Methods", "Magnushammer: A Transformer-based Approach to Premise Selection"], "answer_arxiv_id": ["2304.00994v2", "1606.04442", "2002.07019", "1108.3446", "2303.04488"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_16267"} +{"question": "Can you provide any research papers discussing the weaknesses of using language models as soft reasoners?", "answer": ["Curriculum: A Broad-Coverage Benchmark for Linguistic Phenomena in Natural Language Understanding", "Do Neural Models Learn Systematicity of Monotonicity Inference in Natural Language?", "Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Instead"], "answer_arxiv_id": ["2204.06283v2", "2004.14839", "1811.10154"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_16268"} +{"question": "Which studies discuss the framework for sequential decision making in the field of multi-armed bandits?", "answer": ["Regret Analysis of Stochastic and Nonstochastic Multi-armed Bandit Problems", "Introduction to Multi-Armed Bandits", "Stochastic Contextual Bandits with Long Horizon Rewards"], "answer_arxiv_id": ["1204.5721v2", "1904.07272", "2302.00814"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_16269"} +{"question": "Which papers suggested the use of decoder-only models for generating code from left to right?", "answer": ["Evaluating Large Language Models Trained on Code", "Program Synthesis with Large Language Models", "GPT-NeoX-20B: An Open-Source Autoregressive Language Model", "A Systematic Evaluation of Large Language Models of Code", "InCoder: A Generative Model for Code Infilling and Synthesis"], "answer_arxiv_id": ["2107.03374", "2108.07732", "2204.06745", "2202.13169", "2204.05999"], "source_meta": {"published_time": "20220626"}, "qid": "AutoScholarQuery_train_16270"} +{"question": "Is there any work illustrating simple mechanisms falsely reporting privacy guarantees?", "answer": ["On the Privacy Properties of Variants on the Sparse Vector Technique", "Understanding the Sparse Vector Technique for Differential Privacy"], "answer_arxiv_id": ["1508.07306", "1603.01699"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_16271"} +{"question": "What papers discuss symbolic reasoning over KGs?", "answer": ["Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "Inductive Logic Programming via Differentiable Deep Neural Logic\n Networks", "Learning Algorithms via Neural Logic Networks", "Learn to Explain Efficiently via Neural Logic Inductive Learning", "RNNLogic: Learning Logic Rules for Reasoning on Knowledge Graphs"], "answer_arxiv_id": ["1702.08367", "1906.03523", "1904.01554", "1910.02481", "2010.04029"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_16272"} +{"question": "Which work introduced spatial boxes as input and trained the model on region-text pairs?", "answer": ["GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest"], "answer_arxiv_id": ["2307.03601"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_16273"} +{"question": "What works modify or optimize the forward diffusion process to make the reverse denoising process more efficient?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling with Critically-Damped Langevin Diffusion", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2102.09672", "2112.07068", "2206.00364"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_16274"} +{"question": "What research developed the Prototypical Part Network (ProtoPNet)?", "answer": ["This Looks Like That: Deep Learning for Interpretable Image Recognition"], "answer_arxiv_id": ["1806.10574"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_16275"} +{"question": "Which studies proposed a robust SBI method based on Bayesian nonparametric learning and the posterior bootstrap?", "answer": ["Robust Bayesian Inference for Simulator-based Models via the MMD Posterior Bootstrap"], "answer_arxiv_id": ["2202.04744v3"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16276"} +{"question": "Can you list studies that finetune semantic embeddings for the task of concept attribution?", "answer": ["Evaluating Data Attribution for Text-to-Image Models"], "answer_arxiv_id": ["2306.09345"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_16277"} +{"question": "What research paper helps with understanding the asymmetric nature of self-attention by leveraging the kernel tricks from RKBS?", "answer": ["Transformers are Deep Infinite-Dimensional Non-Mercer Binary Kernel Machines"], "answer_arxiv_id": ["2106.01506"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_16278"} +{"question": "Which works focus on inverse rendering on single RGB images?", "answer": ["Shape, Illumination, and Reflectance from Shading", "Neural Inverse Rendering of an Indoor Scene from a Single Image", "InverseRenderNet: Learning single image inverse rendering", "Object-based Illumination Estimation with Rendering-aware Neural\n Networks", "Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying\n Lighting and SVBRDF from a Single Image", "Shape and Material Capture at Home"], "answer_arxiv_id": ["2010.03592", "1901.02453", "1811.12328", "2008.02514", "1905.02722", "2104.06397"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_16279"} +{"question": "What is the literature on Label-Noise Representation Learning that focus on designing accurate estimators of the noise transition matrix?", "answer": ["Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach", "Learning with Bounded Instance- and Label-dependent Label Noise", "Classification with Noisy Labels by Importance Reweighting", "Are Anchor Points Really Indispensable in Label-Noise Learning?", "Estimating Instance-dependent Bayes-label Transition Matrix using a Deep Neural Network"], "answer_arxiv_id": ["1609.03683", "1709.03768", "1411.7718", "1906.00189", "2105.13001"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_16280"} +{"question": "Which works have developed datasets for compressed video quality enhancement?", "answer": ["BVI-DVC: A Training Database for Deep Video Compression", "NTIRE 2021 Challenge on Quality Enhancement of Compressed Video: Dataset\n and Study", "NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of\n Compressed Video: Dataset, Methods and Results", "MFQE 2.0: A New Approach for Multi-frame Quality Enhancement on\n Compressed Video"], "answer_arxiv_id": ["2003.13552", "2104.10782", "2204.09314", "1902.09707"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_16281"} +{"question": "Which research combine the attention mechanism with convolutional neuroscience?", "answer": ["Convolutional Sequence to Sequence Learning", "CoAtNet: Marrying Convolution and Attention for All Data Sizes"], "answer_arxiv_id": ["1705.03122v3", "2106.04803"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_16282"} +{"question": "What papers have investigated how sharp minimizers with concentrated posterior parameter distributions impact robustness to data perturbations?", "answer": ["Entropy-SGD: Biasing Gradient Descent Into Wide Valleys"], "answer_arxiv_id": ["1611.01838"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_16283"} +{"question": "Which studies employed GAN-based methods in text-to-image generation?", "answer": ["StackGAN: Text to Photo-realistic Image Synthesis with Stacked\n Generative Adversarial Networks", "StackGAN++: Realistic Image Synthesis with Stacked Generative\n Adversarial Networks", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks"], "answer_arxiv_id": ["1612.03242", "1710.10916", "1711.10485"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16284"} +{"question": "Which papers describe impulse control models applied to financial and economic environments?", "answer": ["Optimal Capital Injections with the Risk of Ruin: A Stochastic Differential Game of Impulse Control and Stopping Approach"], "answer_arxiv_id": ["1805.01578"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_16285"} +{"question": "Could you provide some studies that used State-of-the-art backbones like VAE-based formulations and Self-supervised pre-trained for Sketch-Based Image Retrieval?", "answer": ["StyleMeUp: Towards Style-Agnostic Sketch-Based Image Retrieval", "A Zero-Shot Framework for Sketch-based Image Retrieval"], "answer_arxiv_id": ["2103.15706", "1807.11724"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_16286"} +{"question": "What works proposed that SE2 equivariant layers can be implemented efficiently using regular 2D convolutions?", "answer": ["General E(2) - Equivariant Steerable CNNs"], "answer_arxiv_id": ["1911.08251"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16287"} +{"question": "Which study utilize the Subset-Sum approach for weights on dense networks resulting in a logarithmic overparametrization of the width of a layer?", "answer": ["Optimal Lottery Tickets via SubsetSum: Logarithmic Over-Parameterization is Sufficient"], "answer_arxiv_id": ["2006.07990"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_16288"} +{"question": "What researches consider perturbation-based methods for visual interpretation by observing changes in output scores?", "answer": ["RISE: Randomized Input Sampling for Explanation of Black-box Models", "“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "A Unified Approach to Interpreting Model Predictions", "Real Time Image Saliency for Black Box Classifiers", "Explaining Image Classifiers by Counterfactual Generation", "Interpretable Explanations of Black Boxes by Meaningful Perturbation", "Interpretable and Fine-Grained Visual Explanations for Convolutional Neural Networks", "BBAM: Bounding Box Attribution Map for Weakly Supervised Semantic and Instance Segmentation"], "answer_arxiv_id": ["1806.07421", "1602.04938", "1705.07874", "1705.07857", "1807.08024", "1704.03296", "1908.02686", "2103.08907"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_16289"} +{"question": "Are there any studies that generate bird's-eye view by generating a 3D point cloud?", "answer": ["Lift, Splat, Shoot: Encoding Images from Arbitrary Camera Rigs by Implicitly Unprojecting to 3D", "FIERY: Future Instance Prediction in Bird’s-Eye View from Surround Monocular Cameras"], "answer_arxiv_id": ["2008.05711", "2104.10490"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_16290"} +{"question": "Can you tell me what papers utilized the nearest neighbor method for domain adaption in computer vision?", "answer": ["Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation", "Nearest Neighborhood-Based Deep Clustering for Source Data-absent Unsupervised Domain Adaptation"], "answer_arxiv_id": ["2110.04202", "2107.12585"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_16291"} +{"question": "Which studies suggested the use of the Triplanes representation, an answer to the cubic scaling problem of voxel grids?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "3D Neural Field Generation using Triplane Diffusion", "3DGen: Triplane Latent Diffusion for Textured Mesh Generation", "Rodin: A Generative Model for Sculpting 3D Digital Avatars Using\n Diffusion"], "answer_arxiv_id": ["2112.07945", "2301.10241", "2211.16677", "2303.05371", "2212.06135"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_16292"} +{"question": "What work developed a metric to learn state representations by modeling the latent dynamic transition as Gaussian?", "answer": ["Learning Invariant Representations for Reinforcement Learning without Reconstruction"], "answer_arxiv_id": ["2006.10742"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_16293"} +{"question": "What works propose point-based methods to reconstruct generic clothes?", "answer": ["Point-Based Modeling of Human Clothing", "xCloth: Extracting Template-free Textured 3D Clothes from a Monocular\n Image"], "answer_arxiv_id": ["2104.08230", "2208.12934"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_16294"} +{"question": "What study combines gap-dependent regret with variances?", "answer": ["Non-Asymptotic Gap-Dependent Regret Bounds for Tabular MDPs"], "answer_arxiv_id": ["1905.03814"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_16295"} +{"question": "Which work introduced an edge branch in guiding stereo matching networks?", "answer": ["EdgeStereo: An Effective Multi-Task Learning Network for Stereo Matching\n and Edge Detection"], "answer_arxiv_id": ["1903.01700"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16296"} +{"question": "Any study that proposed probabilistic face embeddings to address feature ambiguity?", "answer": ["Probabilistic Face Embeddings"], "answer_arxiv_id": ["1904.09658"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_16297"} +{"question": "Who introduced Top Two Probability Sampling (TTPS) and Top Two Thompson Sampling (TTTS)?", "answer": ["Simple Bayesian Algorithms for Best-Arm Identification"], "answer_arxiv_id": ["1602.08448"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16298"} +{"question": "Could you provide me some studies about unsupervised object discovery?", "answer": ["Localizing Objects with Self-Supervised Transformers and no Labels", "Vision Transformers Need Registers", "Self-Supervised Transformers for Unsupervised Object Discovery using\n Normalized Cut", "Large-Scale Unsupervised Object Discovery", "Unsupervised Object Discovery and Localization in the Wild: Part-based\n Matching with Bottom-up Region Proposals", "Unsupervised Image Matching and Object Discovery as Optimization"], "answer_arxiv_id": ["2109.14279", "2309.16588", "2202.11539", "2106.06650", "1501.06170", "1904.03148"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16299"} +{"question": "Which research papers used language to augment the reward function in Reinforcement Learning?", "answer": ["Reinforced Cross-Modal Matching and Self-Supervised Imitation Learning for Vision-Language Navigation", "Learning to Understand Goal Specifications by Modelling Reward", "Using Natural Language for Reward Shaping in Reinforcement Learning", "EAGER: Asking and Answering Questions for Automatic Reward Shaping in Language-guided RL", "Reward Design with Language Models"], "answer_arxiv_id": ["1811.10092", "1806.01946", "1903.02020", "2206.09674", "2303.00001"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_16300"} +{"question": "Could you provide me some works that used post-hoc watermarking techniques in AI-text detection?", "answer": ["Frustratingly Easy Edit-based Linguistic Steganography with a Masked Language Model", "A Watermark for Large Language Models"], "answer_arxiv_id": ["2104.09833", "2301.10226v4"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_16301"} +{"question": "In what studies curiosity-driven exploration is explored in the context of increasing an agent's understanding of the task environment?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Curiosity-driven Exploration by Self-supervised Prediction", "Intrinsic Reward Driven Imitation Learning via Generative Model"], "answer_arxiv_id": ["1507.00814", "1705.05363", "2006.15061"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_16302"} +{"question": "What works attempt to learn semantic relationships at the pixel level for unsupervised semantic segmentation?", "answer": ["Invariant Information Clustering for Unsupervised Image Classification and Segmentation", "PiCIE: Unsupervised Semantic Segmentation using Invariance and Equivariance in Clustering", "Unsupervised Hierarchical Semantic Segmentation with Multiview Cosegmentation and Clustering Transformers"], "answer_arxiv_id": ["1807.06653", "2103.17070", "2204.11432"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_16303"} +{"question": "What research introduces the idea of approximating Wasserstein distances using multiple nonlinear projections?", "answer": ["Generalized Sliced Wasserstein Distances"], "answer_arxiv_id": ["1902.00434"], "source_meta": {"published_time": "20201028"}, "qid": "AutoScholarQuery_train_16304"} +{"question": "Which research studies contain datasets emphasizing the noun properties of the query text for visual grounding?", "answer": ["Microsoft COCO: Common Objects in Context", "Flickr30k Entities: Collecting Region-to-Phrase Correspondences for\n Richer Image-to-Sentence Models"], "answer_arxiv_id": ["1405.0312", "1505.04870"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16305"} +{"question": "Which studies are related to generating high-quality masks in the field of semantic segmentation?", "answer": ["Conditional Random Fields as Recurrent Neural Networks", "Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials", "Fully Convolutional Networks for Semantic Segmentation", "DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs", "CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement", "Semantic Segmentation Refinement by Monte Carlo Region Growing of High Confidence Detections"], "answer_arxiv_id": ["1502.03240v3", "1210.5644", "1411.4038", "1606.00915", "2005.02551", "1802.07789"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_16306"} +{"question": "Which papers studied a simplified unconstrained feature model, also known as layer-peeled model, and demonstrated that Neural Collapse occurs under various settings?", "answer": ["Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training", "Neural collapse with cross-entropy loss", "On the emergence of simplex symmetry in the final and penultimate layers of neural network classifiers", "Neural collapse with unconstrained features", "A Geometric Analysis of Neural Collapse with Unconstrained Features", "An unconstrained layer-peeled perspective on neural collapse", "Neural Collapse Under MSE Loss: Proximity to and Dynamics on the Central Path", "On the Optimization Landscape of Neural Collapse under MSE Loss: Global Optimality with Unconstrained Features", "Extended Unconstrained Features Model for Exploring Deep Neural Collapse", "Neural Collapse with Normalized Features: A Geometric Analysis over the Riemannian Manifold"], "answer_arxiv_id": ["2101.12699", "2012.08465", "2012.05420", "2011.11619", "2105.02375", "2110.02796", "2106.02073", "2203.01238", "2202.08087", "2209.09211"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_16307"} +{"question": "Are there any works showcasing the use of DPMs in semantic segmentation?", "answer": ["Label-Efficient Semantic Segmentation with Diffusion Models"], "answer_arxiv_id": ["2112.03126"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16308"} +{"question": "What works applied equivariant networks for updating coordinates in ℝ3 in the context of geometric deep learning on point clouds?", "answer": ["SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks"], "answer_arxiv_id": ["2006.10503"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_16309"} +{"question": "What research uses focal distillation to decrease the number of negative flips in the prediction setting?", "answer": ["Positive-Congruent Training: Towards Regression-Free Model Updates"], "answer_arxiv_id": ["2011.09161"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_16310"} +{"question": "Which works are concerned with the applications of the denoising diffusion models?", "answer": ["CARD: Classification and Regression Diffusion Models", "Denoising Diffusion Probabilistic Models", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic Models"], "answer_arxiv_id": ["2206.07275", "2006.11239", "2104.14951"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_16311"} +{"question": "What studies proposed the usage of control maps as guidance modalities in video generation models?", "answer": ["MagicAvatar: Multimodal Avatar Generation and Animation", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "Follow Your Pose: Pose-Guided Text-to-Video Generation using Pose-Free\n Videos", "ControlVideo: Training-free Controllable Text-to-Video Generation", "Control-A-Video: Controllable Text-to-Video Generation with Diffusion\n Models", "DreamPose: Fashion Image-to-Video Synthesis via Stable Diffusion"], "answer_arxiv_id": ["2308.14748", "2306.02018", "2303.13439", "2304.01186", "2305.13077", "2305.13840", "2304.06025"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_16312"} +{"question": "What works regarded category names as short and rough text descriptions in semantic segmentation?", "answer": ["Language-driven Semantic Segmentation", "Global Knowledge Calibration for Fast Open-Vocabulary Segmentation"], "answer_arxiv_id": ["2201.03546", "2303.09181"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_16313"} +{"question": "Could you provide me some works about learning invariant representations in a supervised setting?", "answer": ["Domain Generalization via Invariant Feature Representation", "Invariant Risk Minimization", "Out-of-Distribution Generalization via Risk Extrapolation (REx)", "On Calibration and Out-of-domain Generalization", "Towards efficient representation identification in supervised learning", "Desiderata for Representation Learning: A Causal Perspective", "Probable Domain Generalization via Quantile Risk Minimization", "Spuriosity Didn’t Kill the Classifier: Using Invariant Predictions to Harness Spurious Features"], "answer_arxiv_id": ["1301.2115", "1907.02893", "2003.00688v5", "2102.10395", "2204.04606", "2109.03795v2", "2207.09944", "2307.09933"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16314"} +{"question": "Which works have extended geometric feature extraction from local to global in object variation studies?", "answer": ["HS-Pose: Hybrid Scope Feature Extraction for Category-level Object Pose\n Estimation"], "answer_arxiv_id": ["2303.15743"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_16315"} +{"question": "Which studies proposed multi-modal methods that process depth as 2D images for Semantic Scene Completion?", "answer": ["View-volume Network for Semantic Scene Completion from a Single Depth\n Image", "RGBD Based Dimensional Decomposition Residual Network for 3D Semantic\n Scene Completion", "Anisotropic Convolutional Networks for 3D Semantic Scene Completion"], "answer_arxiv_id": ["1806.05361", "1903.00620", "2004.02122"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_16316"} +{"question": "Could you provide me with some studies about altering model parameters?", "answer": ["Embedding Watermarks into Deep Neural Networks", "Protect, Show, Attend and Tell: Empowering Image Captioning Models with\n Ownership Protection"], "answer_arxiv_id": ["1701.04082", "2008.11009"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_16317"} +{"question": "Could you provide the research that used a hybrid regularizer incorporating the log-barrier in ADA-BARRONS?", "answer": ["Efficient Online Portfolio with Logarithmic Regret"], "answer_arxiv_id": ["1805.07430"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_16318"} +{"question": "Which papers consider different notions of variation like 'path variation', 'functional variation', and 'path length' to capture changes in the comparator?", "answer": ["Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient", "Non-stationary Stochastic Optimization", "Tracking Moving Agents via Inexact Online Gradient Descent Algorithm"], "answer_arxiv_id": ["1605.04638", "1307.5449", "1710.05133v2"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_16319"} +{"question": "What are recent works on progressive refinement in image segmentation?", "answer": ["End-to-End Object Detection with Transformers", "MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "ReMaX: Relaxing for Better Training on Efficient Panoptic Segmentation"], "answer_arxiv_id": ["2005.12872", "2012.00759", "2107.06278", "2112.01527", "2306.17319"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_16320"} +{"question": "Which studies aimed to augmented text inputs to LMs with audio signals or visual perception?", "answer": ["Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding"], "answer_arxiv_id": ["1810.02338"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16321"} +{"question": "Which paper showed the problem of selecting the k coins out of d coins, in a DP algorithm, that land heads with the highest probability?", "answer": ["Tight Lower Bounds for Differentially Private Selection"], "answer_arxiv_id": ["1704.03024"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_16322"} +{"question": "Which studies have adapted scene graph techniques to video, focusing on temporal relationships?", "answer": ["Panoptic Scene Graph Generation", "Not All Relations are Equal: Mining Informative Labels for Scene Graph\n Generation", "The Devil is in the Labels: Noisy Label Correction for Robust Scene\n Graph Generation", "HL-Net: Heterophily Learning Network for Scene Graph Generation"], "answer_arxiv_id": ["2207.11247", "2111.13517", "2206.03014", "2205.01316"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_16323"} +{"question": "Which papers provided rates of convergence dependent on the ambient dimension of the instance space and the complexity of the critic family?", "answer": ["Nonparametric Density Estimation under Adversarial Losses", "How Well Generative Adversarial Networks Learn Distributions", "Nonparametric Density Estimation & Convergence Rates for GANs under Besov IPM Losses"], "answer_arxiv_id": ["1805.08836v2", "1811.03179", "1902.03511v4"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_16324"} +{"question": "Which work discussed the higher compute density of alternative algebras in relation to Clifford convolutions?", "answer": ["AlgebraNets"], "answer_arxiv_id": ["2006.07360"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_16325"} +{"question": "Could you suggest any prior work that improves the layout quality with auto-regressive decoders or GAN-based methods?", "answer": ["Aesthetic Text Logo Synthesis via Content-aware Layout Inferring", "Geometry Aligned Variational Transformer for Image-conditioned Layout\n Generation", "Composition-aware Graphic Layout GAN for Visual-textual Presentation\n Designs", "PosterLayout: A New Benchmark and Approach for Content-aware\n Visual-Textual Presentation Layout"], "answer_arxiv_id": ["2204.02701", "2209.00852", "2205.00303", "2303.15937"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_16326"} +{"question": "Can you name the studies that proposed different versions of MCMC for efficient sampling in the continuous latent space?", "answer": ["Learning Non-Convergent Non-Persistent Short-Run MCMC Toward Energy-Based Model"], "answer_arxiv_id": ["1904.09770"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16327"} +{"question": "Which work integrated a computational physics engine into the language modeling process?", "answer": ["Mind's Eye: Grounded Language Model Reasoning through Simulation"], "answer_arxiv_id": ["2210.05359"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_16328"} +{"question": "What works introduced vector quantization for compressing pre-trained hybrid NeRF representations?", "answer": ["Variable Bitrate Neural Fields", "Compressing Volumetric Radiance Fields to 1 MB"], "answer_arxiv_id": ["2206.07707", "2211.16386"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_16329"} +{"question": "What papers discuss examples of rearrangement such as picking and placing rigid objects or manipulating articulated objects?", "answer": ["robosuite: A Modular Simulation Framework and Benchmark for Robot Learning", "Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning", "DoorGym: A Scalable Door Opening Environment and Baseline Agent", "ManiSkill: Generalizable Manipulation Skill Benchmark with Large-Scale Demonstrations"], "answer_arxiv_id": ["2009.12293", "1910.10897", "1908.01887", "2107.14483"], "source_meta": {"published_time": "20220906"}, "qid": "AutoScholarQuery_train_16330"} +{"question": "What works have studied the challenge of catastrophic forgetting in continual learning?", "answer": ["A continual learning survey: Defying forgetting in classification tasks", "Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics", "Sequoia: A Software Framework to Unify Continual Learning Research"], "answer_arxiv_id": ["1909.08383", "2007.07400", "2108.01005"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_16331"} +{"question": "Can you name the methods that pushed the state-of-the-art in kepoint tracking in the last two years?", "answer": ["TAPIR: Tracking Any Point with per-frame Initialization and temporal\n Refinement", "TAP-Vid: A Benchmark for Tracking Any Point in a Video", "Particle Video Revisited: Tracking Through Occlusions Using Point\n Trajectories", "PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point\n Tracking"], "answer_arxiv_id": ["2306.08637", "2211.03726", "2204.04153", "2307.15055"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_16332"} +{"question": "What are some examples of Large Language Models with strong zero-shot transfer abilities?", "answer": ["Language Models are Few-Shot Learners", "GPT-4 Technical Report", "LLaMA: Open and Efficient Foundation Language Models", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2005.14165", "2303.08774", "2302.13971", "2204.02311"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_16333"} +{"question": "What research works focus on semi-supervised high-resolution disentanglement learning?", "answer": ["Semi-Supervised StyleGAN for Disentanglement Learning"], "answer_arxiv_id": ["2003.03461"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_16334"} +{"question": "Could you tell me what works explored Transformers for point clouds?", "answer": ["Transformers in 3D Point Clouds: A Survey", "Point Transformer", "PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Sampling", "PCT: Point Cloud Transformer", "3D Object Detection with Pointformer", "Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling"], "answer_arxiv_id": ["2205.07417", "2012.09164", "2003.00492", "2012.09688", "2012.11409", "2111.14819"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_16335"} +{"question": "Any works related to layout generation using transformer-based models?", "answer": ["Dolfin: Diffusion Layout Transformers without Autoencoder"], "answer_arxiv_id": ["2310.16305"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16336"} +{"question": "Which works proposed a model that accepts partially diacritized text during decoding and showed improved diacritization performance?", "answer": ["Deep Diacritization: Efficient Hierarchical Recurrence for Improved\n Arabic Diacritization"], "answer_arxiv_id": ["2011.00538"], "source_meta": {"published_time": "20240609"}, "qid": "AutoScholarQuery_train_16337"} +{"question": "Which papers discuss about prior-training sparsity techniques for SNN?", "answer": ["Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science", "Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization", "Sparse Networks from Scratch: Faster Training without Losing Performance", "Rigging the Lottery: Making All Tickets Winners", "Do We Actually Need Dense Over-Parameterization? In-Time Over-Parameterization in Sparse Training", "Powerpropagation: A sparsity inducing weight reparameterisation"], "answer_arxiv_id": ["1707.04780v2", "1902.05967", "1907.04840", "1911.11134", "2102.02887", "2110.00296"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_16338"} +{"question": "What studies train networks for Novel Views Synthesis purely from videos and use point clouds as intermediate 3D representations?", "answer": ["Infinite Nature: Perpetual View Generation of Natural Scenes from a\n Single Image", "PixelSynth: Generating a 3D-Consistent Experience from a Single Image", "SynSin: End-to-end View Synthesis from a Single Image"], "answer_arxiv_id": ["2012.09855", "2108.05892", "1912.08804"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_16339"} +{"question": "What work was used as a starting point for exploration and finetuning in Minecraft?", "answer": ["Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos"], "answer_arxiv_id": ["2206.11795"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_16340"} +{"question": "Can you name the studies that developed state-of-the-art DAG learning methods for linear Structural Equation Models (SEMs)?", "answer": ["DAGMA: Learning DAGs via M-matrices and a Log-Determinant Acyclicity Characterization", "On the Role of Sparsity and DAG Constraints for Learning Linear DAGs"], "answer_arxiv_id": ["2209.08037", "2006.10201"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16341"} +{"question": "Could you point out the studies that focus on RGB-D fusion in 2D based depth completion?", "answer": ["Confidence Propagation through CNNs for Guided Sparse Depth Regression", "PENet: Towards Precise and Efficient Image Guided Depth Completion", "Depth Completion with Twin Surface Extrapolation at Occlusion Boundaries", "LRRU: Long-short Range Recurrent Updating Networks for Depth Completion"], "answer_arxiv_id": ["1811.01791", "2103.00783", "2104.02253", "2310.08956v1"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_16342"} +{"question": "What studies propose methods of post-processing a predictor to enforce these fairness constraints while minimizing loss?", "answer": ["Equality of Opportunity in Supervised Learning"], "answer_arxiv_id": ["1610.02413"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_16343"} +{"question": "Are there any studies that focused on bounding reconstruction for language tasks?", "answer": ["Defending against Reconstruction Attacks with Rényi Differential Privacy"], "answer_arxiv_id": ["2202.07623"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_16344"} +{"question": "Which papers have utilized flow in video recognition?", "answer": ["Deep Feature Flow for Video Recognition"], "answer_arxiv_id": ["1611.07715"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_16345"} +{"question": "Could you provide me some research that decode multiple times from the same perspective and aggregate the results to enhance LLMs?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models", "Maieutic Prompting: Logically Consistent Reasoning with Recursive\n Explanations", "Recitation-Augmented Language Models"], "answer_arxiv_id": ["2203.11171", "2205.10625", "2205.11822", "2210.01296"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_16346"} +{"question": "Can you provide research that extended gaze target detection into videos?", "answer": ["Detecting Attended Visual Targets in Video"], "answer_arxiv_id": ["2003.02501"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16347"} +{"question": "What works have investigated entropy regularization in two-player zero-sum matrix games?", "answer": ["Faster Last-iterate Convergence of Policy Optimization in Zero-Sum Markov Games"], "answer_arxiv_id": ["2210.01050"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_16348"} +{"question": "Which work decomposed in-the-wild objects using cross-instance constraints?", "answer": ["Share With Thy Neighbors: Single-View Reconstruction by Cross-Instance\n Consistency"], "answer_arxiv_id": ["2204.10310"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16349"} +{"question": "Which researchers have modeled federated learning as a hedonic game?", "answer": ["Optimality and Stability in Federated Learning: A Game-theoretic Approach", "Model-sharing Games: Analyzing Federated Learning Under Voluntary Participation"], "answer_arxiv_id": ["2106.09580", "2010.00753"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_16350"} +{"question": "Could you provide me some works that integrate human feedback into training using reinforcement learning for user-tailored tasks?", "answer": ["Training language models to follow instructions with human feedback", "Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback"], "answer_arxiv_id": ["2203.02155", "2204.05862"], "source_meta": {"published_time": "20240718"}, "qid": "AutoScholarQuery_train_16351"} +{"question": "Which work suggests that the advancements to adversarial training, since PGD, can be matched with early stopping?", "answer": ["Overfitting in adversarially robust deep learning"], "answer_arxiv_id": ["2002.11569"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_16352"} +{"question": "Which paper studied optimal local DP algorithms under a given utility function and prior?", "answer": ["Extremal Mechanisms for Local Differential Privacy"], "answer_arxiv_id": ["1407.1338"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_16353"} +{"question": "What works have done architectural improvements in audio-visual video parsing (AVVP)?", "answer": ["Exploiting Audio-Visual Consistency with Partial Supervision for Spatial Audio Generation", "MM-Pyramid: Multimodal Pyramid Attentional Network for Audio-Visual Event Localization and Video Parsing"], "answer_arxiv_id": ["2105.00708", "2111.12374"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_16354"} +{"question": "Which papers have discussed the role of text encoder prompts in the field of V&L models?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "LASP: Text-to-Text Optimization for Language-Aware Soft Prompting of\n Vision & Language Models"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2210.01115"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_16355"} +{"question": "Which work explained the use of random feature matrices for defining U⋆ and V⋆ in the unnormalized softmax case?", "answer": ["Rethinking Attention with Performers"], "answer_arxiv_id": ["2009.14794"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_16356"} +{"question": "Could you provide me works about an unsupervised pretraining algorithm to learn representations based on an eigendecomposition of transition matrix Pπ?", "answer": ["Learning One Representation to Optimize All Rewards", "Learning Successor States and Goal-Dependent Values: A Mathematical Viewpoint"], "answer_arxiv_id": ["2103.07945", "2101.07123"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_16357"} +{"question": "Could you name the works that have proposed rule-learning and GNNs methods for KG completion?", "answer": ["Differentiable Learning of Logical Rules for Knowledge Base Reasoning", "Modeling Relational Data with Graph Convolutional Networks", "Neural Bellman-Ford Networks: A General Graph Neural Network Framework for Link Prediction"], "answer_arxiv_id": ["1702.08367", "1703.06103v4", "2106.06935"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_16358"} +{"question": "Which works are the backbone of the recent popular vision-language models (VLMs)?", "answer": ["Language Models are Few-Shot Learners", "Training language models to follow instructions with human feedback", "OpenChat: Advancing Open-source Language Models with Mixed-Quality Data"], "answer_arxiv_id": ["2005.14165", "2203.02155", "2309.11235"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_16359"} +{"question": "Which studies focus on optimal experiment design, a broad topic that includes Bayesian Optimization?", "answer": ["Bayesian Algorithm Execution: Estimating Computable Properties of Black-box Functions Using Mutual Information"], "answer_arxiv_id": ["2104.09460"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16360"} +{"question": "Could you cite examples of research about the utilization of latent variables for offline RL?", "answer": ["OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning", "PLAS: Latent Action Space for Offline Reinforcement Learning"], "answer_arxiv_id": ["2010.13611", "2011.07213"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_16361"} +{"question": "Any works that converted existing datasets of RC tasks into the instruction following format?", "answer": ["Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic"], "answer_arxiv_id": ["2306.15195"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_16362"} +{"question": "What studies in the field of explicit 3D shape representations used large-scale datasets for training?", "answer": ["ShapeNet: An Information-Rich 3D Model Repository"], "answer_arxiv_id": ["1512.03012"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16363"} +{"question": "Could you provide me some studies that explore pre-training end-to-end models for document understanding?", "answer": ["OCR-free Document Understanding Transformer", "End-to-end Document Recognition and Understanding with Dessurt", "Pix2Struct: Screenshot Parsing as Pretraining for Visual Language\n Understanding", "Attention Where It Matters: Rethinking Visual Document Understanding\n with Selective Region Concentration"], "answer_arxiv_id": ["2111.15664", "2203.16618", "2210.03347", "2309.01131"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_16364"} +{"question": "Which works progressively improve the Probabilistic Circuits (PC) structure during the optimization process?", "answer": ["Sparse Probabilistic Circuits via Pruning and Growing"], "answer_arxiv_id": ["2211.12551"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_16365"} +{"question": "What works evaluated various Protein Language Models (PLMs) in contact prediction?", "answer": ["Evaluating Protein Transfer Learning with TAPE"], "answer_arxiv_id": ["1906.08230"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16366"} +{"question": "Which studies have applied the replica method to non-convex losses?", "answer": ["Contrasting random and learned features in deep Bayesian linear regression", "Optimal Learning of Deep Random Networks of Extensive-width"], "answer_arxiv_id": ["2203.00573", "2302.00375"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_16367"} +{"question": "Which studies have been done on coordinate based MLPs for learning implicit functions to represent continuous 3D shapes?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "Neural Unsigned Distance Fields for Implicit Function Learning", "RangeUDF: Semantic Surface Reconstruction from 3D Point Clouds", "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction"], "answer_arxiv_id": ["1901.05103", "1812.03828", "2010.13938", "2204.09138", "2003.08934", "2106.10689"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_16368"} +{"question": "What works proposed to learn a generalized Lipschitz transition model with respect to the Wasserstein metric?", "answer": ["Lipschitz Continuity in Model-based Reinforcement Learning"], "answer_arxiv_id": ["1804.07193"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_16369"} +{"question": "Which work introduced adversarial training of a classifier?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_16370"} +{"question": "Which works propose the use of Knowledge Distillation (KD) to improve the performance of self-supervised pre-training for low-compute network architectures?", "answer": ["Distilling the Knowledge in a Neural Network", "SEED: Self-supervised Distillation For Visual Representation", "CompRess: Self-Supervised Learning by Compressing Representations", "SimReg: Regression as a Simple Yet Effective Tool for Self-supervised Knowledge Distillation", "DisCo: Remedy Self-supervised Learning on Lightweight Models with Distilled Contrastive Learning"], "answer_arxiv_id": ["1503.02531", "2101.04731", "2010.14713", "2201.05131", "2104.09124"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_16371"} +{"question": "Which paper introduced the Segment Anything (SA) model?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_16372"} +{"question": "What research connects reinforcement learning to game theory?", "answer": ["Actor-Critic Policy Optimization in Partially Observable Multiagent Environments", "Neural Replicator Dynamics: Multiagent Learning via Hedging Policy Gradients", "From Poincaré Recurrence to Convergence in Imperfect Information Games: Finding Equilibrium via Regularization"], "answer_arxiv_id": ["1810.09026", "1906.00190", "2002.08456"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_16373"} +{"question": "Are there any papers discussing the difficulty in providing semantically consistent augmentation for user behavior sequences because of diverse user interests and heavy noise, which hampers the effectiveness of existing contrastive learning-based user model pre-training methods?", "answer": ["Sampler Design for Implicit Feedback Data by Noisy-label Robust Learning", "Contrastive Learning for Representation Degeneration Problem in Sequential Recommendation"], "answer_arxiv_id": ["2007.07204", "2110.05730"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_16374"} +{"question": "Could you provide me some studies about metric-based meta learning methods for relational graphs, network alignment and generic graphs?", "answer": ["Meta Relational Learning for Few-Shot Link Prediction in Knowledge Graphs", "Few-Shot Knowledge Graph Completion"], "answer_arxiv_id": ["1909.01515", "1911.11298"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_16375"} +{"question": "Which papers explored architectural enhancements to improve the performance of deep learning methods specifically for single image super-resolution?", "answer": ["Accurate Image Super-Resolution Using Very Deep Convolutional Networks", "Enhanced Deep Residual Networks for Single Image Super-Resolution", "Residual Dense Network for Image Super-Resolution", "Image Super-Resolution Using Very Deep Residual Channel Attention\n Networks", "Single Image Super-Resolution via a Holistic Attention Network", "Image Super-Resolution with Cross-Scale Non-Local Attention and\n Exhaustive Self-Exemplars Mining", "Deeply-Recursive Convolutional Network for Image Super-Resolution"], "answer_arxiv_id": ["1511.04587", "1707.02921", "1802.08797", "1807.02758", "2008.08767", "2006.01424", "1511.04491"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_16376"} +{"question": "What research employed a conditional GAN within an autoencoder-based compression model to enhance perceptual quality?", "answer": ["High-Fidelity Generative Image Compression"], "answer_arxiv_id": ["2006.09965"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_16377"} +{"question": "Which studies discussed expectational-maximization (EM) based methods in Safe Reinforcement Learning?", "answer": ["Constrained Variational Policy Optimization for Safe Reinforcement Learning", "Conservative Distributional Reinforcement Learning with Safety Constraints"], "answer_arxiv_id": ["2201.11927", "2201.07286"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_16378"} +{"question": "What study uses predefined measurement functions to handle time series forecasting with changing temporal distribution?", "answer": ["Koopman Neural Forecaster for Time Series with Temporal Distribution Shifts"], "answer_arxiv_id": ["2210.03675"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16379"} +{"question": "Which work explores the cause of negative flips due to large logit margins and presents methods to reduce large logit displacement?", "answer": ["ELODI: Ensemble Logit Difference Inhibition for Positive-Congruent Training"], "answer_arxiv_id": ["2205.06265v3"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_16380"} +{"question": "What papers are included in the example of periodic learning that involves predicting weather and environmental changes?", "answer": ["MetNet: A Neural Weather Model for Precipitation Forecasting", "Skillful Twelve Hour Precipitation Forecasts using Large Context Neural Networks"], "answer_arxiv_id": ["2003.12140", "2111.07470v1"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_16381"} +{"question": "What studies focus on the investigation of internal mechanisms in GANs to identify and remove units causing artifacts in images?", "answer": ["GAN Dissection: Visualizing and Understanding Generative Adversarial Networks"], "answer_arxiv_id": ["1811.10597"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16382"} +{"question": "What are some works that have investigated or improved the calibration of language models?", "answer": ["Holistic Evaluation of Language Models", "A Benchmark Study on Calibration", "Investigating Uncertainty Calibration of Aligned Language Models under\n the Multiple-Choice Setting", "Calibration of Pre-trained Transformers", "Calibrated Language Model Fine-Tuning for In- and Out-of-Distribution\n Data", "Calibrating Structured Output Predictors for Natural Language Processing", "Selective Question Answering under Domain Shift", "How Can We Know When Language Models Know? On the Calibration of\n Language Models for Question Answering", "Reducing conversational agents' overconfidence through linguistic\n calibration", "Teaching Models to Express Their Uncertainty in Words", "Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence\n Scores from Language Models Fine-Tuned with Human Feedback", "Navigating the Grey Area: How Expressions of Uncertainty and\n Overconfidence Affect Language Models"], "answer_arxiv_id": ["2211.09110", "2308.11838", "2310.11732", "2003.07892", "2010.11506", "2004.04361", "2006.09462", "2012.00955", "2012.14983", "2205.14334", "2305.14975", "2302.13439"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_16383"} +{"question": "In what papers do Large Language Models (LLMs) act as decoders in vision-language applications by using a modality bridge module to align visual with language features?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning", "LLaVA-Med: Training a Large Language-and-Vision Assistant for\n Biomedicine in One Day", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Cheap and Quick: Efficient Vision-Language Instruction Tuning for Large\n Language Models", "mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality", "Otter: A Multi-Modal Model with In-Context Instruction Tuning", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "PMC-VQA: Visual Instruction Tuning for Medical Visual Question Answering"], "answer_arxiv_id": ["2204.14198", "2301.12597", "2304.10592", "2304.08485", "2306.00890", "2304.15010", "2305.15023", "2304.14178", "2305.03726", "2305.11175", "2305.10415"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_16384"} +{"question": "Which studies have used EvolveGraph to employ static and dynamic interaction graphs?", "answer": ["EvolveGraph: Multi-Agent Trajectory Prediction with Dynamic Relational Reasoning"], "answer_arxiv_id": ["2003.13924"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16385"} +{"question": "What research found that neural networks can be trained in lower-dimensional subspaces?", "answer": ["Intrinsic Dimensionality Explains the Effectiveness of Language Model\n Fine-Tuning"], "answer_arxiv_id": ["2012.13255"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_16386"} +{"question": "Could you provide me some studies that have been offered to improve the rendering quality with additional constraints?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "RealFusion: 360{\\deg} Reconstruction of Any Object from a Single Image", "NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with\n 360{\\deg} Views", "Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D\n Generation", "Zero-Shot Text-Guided Object Generation with Dream Fields", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "Re-imagine the Negative Prompt Algorithm: Transform 2D Diffusion into\n 3D, alleviate Janus problem and Beyond", "Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and\n Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2211.10440", "2302.10663", "2211.16431", "2303.07937", "2112.01455", "2303.13873", "2305.16213", "2304.04968", "2212.14704"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_16387"} +{"question": "Are there any research papers you can recommend on self-supervised pre-training in tabular deep learning?", "answer": ["SubTab: Subsetting Features of Tabular Data for Self-Supervised Representation Learning", "SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training", "Revisiting Pretraining Objectives for Tabular Deep Learning", "Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning"], "answer_arxiv_id": ["2110.04361", "2106.01342", "2207.03208", "2106.02584v2"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_16388"} +{"question": "Which works leveraged language models for recommendation tasks in the early implementations?", "answer": ["Recommender Systems in the Era of Large Language Models (LLMs)"], "answer_arxiv_id": ["2307.02046"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_16389"} +{"question": "Which research incorporates an atrous spatial pyramid pooling module in DeeplabV3+?", "answer": ["Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation"], "answer_arxiv_id": ["1802.02611"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_16390"} +{"question": "Could you give me studies that perform semi-supervised medical segmentation?", "answer": ["Uncertainty-aware Self-ensembling Model for Semi-supervised 3D Left Atrium Segmentation", "Semi-supervised Medical Image Segmentation through Dual-task Consistency", "Contrastive learning of global and local features for medical image segmentation with limited annotations", "SimCVD: Simple Contrastive Voxel-Wise Representation Distillation for Semi-Supervised Medical Image Segmentation", "Bootstrapping Semi-supervised Medical Image Segmentation with Anatomical-aware Contrastive Distillation", "Mine yOur owN Anatomy: Revisiting Medical Image Segmentation with Extremely Limited Labels", "Mutual Consistency Learning for Semi-supervised Medical Image Segmentation", "Exploring Smoothness and Class-Separation for Semi-supervised Medical Image Segmentation", "ACTION++: Improving Semi-supervised Medical Image Segmentation with Adaptive Anatomical Contrast"], "answer_arxiv_id": ["1907.07034", "2009.04448", "2006.10511", "2108.06227", "2206.02307", "2209.13476", "2109.09960", "2203.01324", "2304.02689"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_16391"} +{"question": "What studies have shown that their evaluation benchmarks have a high agreement to human evaluation?", "answer": ["Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena"], "answer_arxiv_id": ["2306.05685"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_16392"} +{"question": "What are the works that managed to generate human motion isolated from environmental context?", "answer": ["HuMoR: 3D Human Motion Model for Robust Pose Estimation", "MoGlow: Probabilistic and controllable motion synthesis using\n normalising flows", "Human Motion Diffusion Model", "The Wanderings of Odysseus in 3D Scenes", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE"], "answer_arxiv_id": ["2105.04668", "1905.06598", "2209.14916", "2112.09251", "2104.05670"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_16393"} +{"question": "What study compares the intuitive physics engine of bib.bib8 with the convolutional neural network of bib.bib25?", "answer": ["A Comparative Evaluation of Approximate Probabilistic Simulation and Deep Neural Networks as Accounts of Human Physical Scene Understanding"], "answer_arxiv_id": ["1605.01138"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16394"} +{"question": "Which papers' method of shift detection do we focus on that rely on deep neural representations of the data?", "answer": ["Detecting and Correcting for Label Shift with Black Box Predictors"], "answer_arxiv_id": ["1802.03916"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_16395"} +{"question": "Could you provide examples of research that significantly ventured away from prior-based motion representation in model-free methods?", "answer": ["One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing", "Implicit Warping for Animation with Image Sets"], "answer_arxiv_id": ["2011.15126", "2210.01794"], "source_meta": {"published_time": "20231021"}, "qid": "AutoScholarQuery_train_16396"} +{"question": "What research articles discuss synthetic object collections lacking environment context?", "answer": ["ShapeNet: An Information-Rich 3D Model Repository", "ABO: Dataset and Benchmarks for Real-World 3D Object Understanding"], "answer_arxiv_id": ["1512.03012", "2110.06199"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_16397"} +{"question": "Are there any works discussing self-supervised representations for association in object tracking?", "answer": ["Improved Baselines with Momentum Contrastive Learning", "Rethinking Self-supervised Correspondence Learning: A Video Frame-level\n Similarity Perspective"], "answer_arxiv_id": ["2003.04297", "2103.17263"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_16398"} +{"question": "Can you provide any works that proposed probabilistic formulation of MS for Gaussian process based dynamics?", "answer": ["Variational multiple shooting for Bayesian ODEs with Gaussian processes"], "answer_arxiv_id": ["2106.10905"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_16399"} +{"question": "What work separated the generation of temporal latent variables and spatial information in video generation?", "answer": ["Temporal Generative Adversarial Nets with Singular Value Clipping"], "answer_arxiv_id": ["1611.06624"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_16400"} +{"question": "What other prompting methods were inspired by the idea of conditioning the prompt on input images?", "answer": ["MaPLe: Multi-modal Prompt Learning", "Prompt-aligned Gradient for Prompt Tuning"], "answer_arxiv_id": ["2210.03117", "2205.14865"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_16401"} +{"question": "What studies address the issue of large intra-class variations by mining the class-specific representation from the query branch in few-shot segmentation?", "answer": ["Self-Support Few-Shot Semantic Segmentation", "Intermediate Prototype Mining Transformer for Few-Shot Semantic Segmentation"], "answer_arxiv_id": ["2207.11549", "2210.06780"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_16402"} +{"question": "What papers explore two-step methods that focus on sample-selection tasks and large margin principles for Positive Unlabeled Learning?", "answer": ["A Variational Approach for Learning from Positive and Unlabeled Data"], "answer_arxiv_id": ["1906.00642"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_16403"} +{"question": "Could you provide literature that applies Optimal Transport in dealing with noisy labels?", "answer": ["CSOT: Curriculum and Structure-Aware Optimal Transport for Learning with\n Noisy Labels"], "answer_arxiv_id": ["2312.06221"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16404"} +{"question": "What research discussed the effect of overparameterization in PGMs?", "answer": ["Empirical Study of the Benefits of Overparameterization in Learning Latent Variable Models"], "answer_arxiv_id": ["1907.00030"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_16405"} +{"question": "Which works are about offline meta-reinforcement learning?", "answer": ["Offline Meta-Reinforcement Learning with Online Self-Supervision", "Contextual Transformer for Offline Meta Reinforcement Learning"], "answer_arxiv_id": ["2107.03974", "2211.08016"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_16406"} +{"question": "Which paper investigates a range of simple synthetic tasks for pre-training?", "answer": ["Insights into Pre-training via Simpler Synthetic Tasks"], "answer_arxiv_id": ["2206.10139"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_16407"} +{"question": "Which paper deals with hypernetwork CFNN, where the weights of a model are first randomly generated?", "answer": ["Coin-Flipping Neural Networks"], "answer_arxiv_id": ["2206.09182"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_16408"} +{"question": "What are some studies discussing MHA-based tuning in the context of PETL?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "DualPrompt: Complementary Prompting for Rehearsal-free Continual Learning", "Visual Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally Across Scales and Tasks", "GPT Understands, Too"], "answer_arxiv_id": ["2106.09685", "2204.04799", "2203.12119", "2101.00190", "2110.07602", "2103.10385"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16409"} +{"question": "Which papers reviewed the generation of executable actions predicted by Language Learning Models (LLMs) in household environments?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Skill Induction and Planning with Latent Language", "Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents", "On Grounded Planning for Embodied Tasks with Language Models"], "answer_arxiv_id": ["2204.01691", "2110.01517", "2201.07207", "2209.00465v3"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16410"} +{"question": "Could you provide me some works about obtaining 'slimmable neural networks'?", "answer": ["Slimmable Neural Networks", "Universally Slimmable Networks and Improved Training Techniques", "AutoSlim: Towards One-Shot Architecture Search for Channel Numbers"], "answer_arxiv_id": ["1812.08928", "1903.05134", "1903.11728"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_16411"} +{"question": "What works suggest sample-efficient algorithms for learning POMDPs?", "answer": ["A Spectral Algorithm for Learning Hidden Markov Models", "Reinforcement Learning of POMDPs using Spectral Methods", "A PAC RL Algorithm for Episodic POMDPs", "Sublinear Regret for Learning POMDPs", "Online Learning for Unknown Partially Observable MDPs"], "answer_arxiv_id": ["0811.4413", "1602.07764", "1605.08062v2", "2107.03635v4", "2102.12661"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16412"} +{"question": "What research works discuss the use of projection-based methods in multimodal language models?", "answer": ["PMC-VQA: Visual Instruction Tuning for Medical Visual Question Answering", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "Cheap and Quick: Efficient Vision-Language Instruction Tuning for Large\n Language Models", "PandaGPT: One Model To Instruction-Follow Them All", "CogVLM: Visual Expert for Pretrained Language Models"], "answer_arxiv_id": ["2305.10415", "2304.15010", "2305.15023", "2305.16355", "2311.03079"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_16413"} +{"question": "Which work proposed the procedural generation pipeline for interactable scenes within the context of indoor scene synthesis?", "answer": ["ProcTHOR: Large-Scale Embodied AI Using Procedural Generation"], "answer_arxiv_id": ["2206.06994"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_16414"} +{"question": "What research focused on continuous-time dynamic graphs (CTDGs) through sequence-based methods?", "answer": ["Deep Coevolutionary Network: Embedding User and Item Features for Recommendation", "Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks"], "answer_arxiv_id": ["1609.03675", "1908.01207"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_16415"} +{"question": "Which works utilize point clouds at different timestamps as views for unsupervised contrastive learning?", "answer": ["Spatio-temporal Self-Supervised Representation Learning for 3D Point Clouds"], "answer_arxiv_id": ["2109.00179"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_16416"} +{"question": "What are the studies aimed at improving the descriptiveness and discriminability of image captioning using prompting strategies?", "answer": ["CapEnrich: Enriching Caption Semantics for Web Images via Cross-modal Pre-trained Knowledge", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2211.09371", "2107.13586v1"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_16417"} +{"question": "Could you name some works that have used graph partitioning techniques in Graph Neural Networks?", "answer": ["Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks", "PipeGCN: Efficient Full-Graph Training of Graph Convolutional Networks with Pipelined Feature Communication", "BNS-GCN: Efficient Full-Graph Training of Graph Convolutional Networks with Partition-Parallelism and Random Boundary Node Sampling"], "answer_arxiv_id": ["1905.07953", "2203.10428", "2203.10983"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_16418"} +{"question": "Which works have conducted feature interaction selection in DSNs by borrowing ideas from neural architecture search?", "answer": ["AutoFIS: Automatic Feature Interaction Selection in Factorization Models for Click-Through Rate Prediction"], "answer_arxiv_id": ["2003.11235"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_16419"} +{"question": "Which papers focus on improving the robustness of Vision Transformers (ViTs)?", "answer": ["Towards Robust Vision Transformer", "Understanding Robustness of Transformers for Image Classification", "Are Transformers More Robust Than CNNs?", "Robustness Verification for Transformers", "Vision Transformers are Robust Learners", "Robustifying Token Attention for Vision Transformers", "Understanding The Robustness in Vision Transformers"], "answer_arxiv_id": ["2105.07926", "2103.14586", "2111.05464", "2002.06622", "2105.07581", "2303.11126", "2204.12451"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_16420"} +{"question": "What works discuss the hPINN that treats PDE constraints as regularization terms?", "answer": ["Physics-informed neural networks with hard constraints for inverse design"], "answer_arxiv_id": ["2102.04626v1"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_16421"} +{"question": "Could you provide papers that researched on the mistake bound under uniform ball manipulations and regret under a pre-defined and known manipulation?", "answer": ["The Strategic Perceptron", "Fundamental Bounds on Online Strategic Classification"], "answer_arxiv_id": ["2008.01710", "2302.12355v2"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16422"} +{"question": "What works focus on resolving specific tasks in image restoration with unique model designs?", "answer": ["Implicit Neural Representation for Cooperative Low-light Image\n Enhancement", "Learning A Sparse Transformer Network for Effective Image Deraining", "Image Restoration with Mean-Reverting Stochastic Differential Equations", "Learning Enriched Features for Fast Image Restoration and Enhancement", "Vision Transformers for Single Image Dehazing", "RestoreFormer: High-Quality Blind Face Restoration from Undegraded\n Key-Value Pairs", "Restoring Vision in Adverse Weather Conditions with Patch-Based\n Denoising Diffusion Models", "Residual Denoising Diffusion Models"], "answer_arxiv_id": ["2303.11722", "2303.11950", "2301.11699", "2205.01649", "2204.03883", "2201.06374", "2207.14626", "2308.13712"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_16423"} +{"question": "Which research works have used the re-ranking approach to improve the quality of generation in language models?", "answer": ["Generate & Rank: A Multi-task Framework for Math Word Problems"], "answer_arxiv_id": ["2109.03034"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_16424"} +{"question": "Which work explains the concept of self-distillation and its positive impact on the generalization performance of a model?", "answer": ["Born-Again Neural Networks"], "answer_arxiv_id": ["1805.04770"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16425"} +{"question": "Which studies proposed corrections to the distortions and projections occurring in high-dimensional continuous spaces?", "answer": ["A warped kernel improving robustness in Bayesian optimization via random embeddings", "On the choice of the low-dimensional domain for global optimization via random embeddings"], "answer_arxiv_id": ["1411.3685", "1704.05318"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_16426"} +{"question": "Can you cite works that discuss few-shot class-incremental learning?", "answer": ["Few-Shot Class-Incremental Learning via Entropy-Regularized Data-Free Replay", "Few-Shot Class-Incremental Learning", "Matching Networks for One Shot Learning"], "answer_arxiv_id": ["2207.11213", "2004.10956", "1606.04080"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_16427"} +{"question": "Which benchmarks are used for physical reasoning in computer vision?", "answer": ["PHYRE: A New Benchmark for Physical Reasoning", "Rapid trial-and-error learning with simulation supports flexible tool use and physical reasoning", "Generalization to New Actions in Reinforcement Learning"], "answer_arxiv_id": ["1908.05656", "1907.09620", "2011.01928"], "source_meta": {"published_time": "20220202"}, "qid": "AutoScholarQuery_train_16428"} +{"question": "In which work is a sparse voxel octree used for modeling via NSVF and challenges with its flexibility discussed?", "answer": ["Neural Sparse Voxel Fields"], "answer_arxiv_id": ["2007.11571"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_16429"} +{"question": "Which paper introduced the trailblazing point-based method known as PCN for point cloud completion?", "answer": ["PCN: Point Completion Network"], "answer_arxiv_id": ["1808.00671"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_16430"} +{"question": "Could you provide papers that discuss combining approaches by using a multi-particle, multi-step AIS transdimensional parameter proposal?", "answer": ["Estimators of Entropy and Information via Inference in Probabilistic Models"], "answer_arxiv_id": ["2202.12363v4"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_16431"} +{"question": "What works are considered as 2D-to-2D projection methods in the context of open-vocabulary 3D instance segmentation?", "answer": ["OpenMask3D: Open-Vocabulary 3D Instance Segmentation", "OpenIns3D: Snap and Lookup for 3D Open-vocabulary Instance Segmentation", "OpenScene: 3D Scene Understanding with Open Vocabularies"], "answer_arxiv_id": ["2306.13631", "2309.00616", "2211.15654"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_16432"} +{"question": "Which papers talk about leveraging natural language for classification?", "answer": ["The Curious Layperson: Fine-Grained Image Recognition without Expert Labels", "Link the head to the “beak”: Zero Shot Learning from Noisy Text Description at Part Precision", "Visual Classification via Description from Large Language Models", "What does a platypus look like? Generating customized prompts for zero-shot image classification"], "answer_arxiv_id": ["2111.03651", "1709.01148", "2210.07183", "2209.03320"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16433"} +{"question": "Is there any research that trains a transformer-based architecture for generation of RGB point clouds conditioned on complex prompts?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts"], "answer_arxiv_id": ["2212.08751"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_16434"} +{"question": "Which papers introduced the idea of learning UDFs for handling open surfaces with neural networks?", "answer": ["Neural Unsigned Distance Fields for Implicit Function Learning"], "answer_arxiv_id": ["2010.13938"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_16435"} +{"question": "Can you name some papers that demonstrated remarkable success in understanding and reasoning about 3D objects using LLMs?", "answer": ["PointLLM: Empowering Large Language Models to Understand Point Clouds", "Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D\n Understanding, Generation, and Instruction Following", "Scalable 3D Captioning with Pretrained Models", "RegionBLIP: A Unified Multi-modal Pre-training Framework for Holistic\n and Regional Comprehension"], "answer_arxiv_id": ["2308.16911", "2309.00615", "2306.07279", "2308.02299"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16436"} +{"question": "What studies uses attention control to minimize changes to unrelated parts in image editing?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2208.01626", "2211.12572"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_16437"} +{"question": "Which works in text-to-image generation field propose the use of a transformer model to infer image tokens from text?", "answer": ["Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers"], "answer_arxiv_id": ["2102.12092", "2105.13290"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_16438"} +{"question": "Could you provide me the works where recurrent neural networks were introduced to predict pedestrian future trajectories?", "answer": ["Social GAN: Socially Acceptable Trajectories with Generative Adversarial\n Networks", "Trajectron++: Dynamically-Feasible Trajectory Forecasting With\n Heterogeneous Data", "Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory\n Prediction"], "answer_arxiv_id": ["1803.10892", "2001.03093", "2005.08514"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_16439"} +{"question": "In what papers the researchers guide the simulation towards a desired target state?", "answer": ["Contact Points Discovery for Soft-Body Manipulations with Differentiable Physics"], "answer_arxiv_id": ["2205.02835"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_16440"} +{"question": "What studies focused on certified backdoor defenses that modify the training process to prevent learning the backdoor?", "answer": ["BagFlip: A Certified Defense against Data Poisoning", "RAB: Provable Robustness Against Backdoor Attacks", "Certified Robustness of Nearest Neighbors against Data Poisoning and Backdoor Attacks", "Certified Robustness to Label-Flipping Attacks via Randomized Smoothing"], "answer_arxiv_id": ["2205.13634", "2003.08904", "2012.03765", "2002.03018"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_16441"} +{"question": "Which research papers have proposed using 3D LUTs for image enhancement?", "answer": ["Learning Image-adaptive 3D Lookup Tables for High Performance Photo\n Enhancement in Real-time", "Real-time Image Enhancer via Learnable Spatial-aware 3D Lookup Tables", "AdaInt: Learning Adaptive Intervals for 3D Lookup Tables on Real-time\n Image Enhancement", "SepLUT: Separable Image-adaptive Lookup Tables for Real-time Image\n Enhancement", "4D LUT: Learnable Context-Aware 4D Lookup Table for Image Enhancement"], "answer_arxiv_id": ["2009.14468", "2108.08697", "2204.13983", "2207.08351", "2209.01749"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_16442"} +{"question": "Could you give me examples of studies that formulated AFT as an auto ML task?", "answer": ["DIFER: Differentiable Automated Feature Engineering"], "answer_arxiv_id": ["2010.08784"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_16443"} +{"question": "Which studies evaluate the generalization of DNN based on the stability of the learning algorithm?", "answer": ["Train faster, generalize better: Stability of stochastic gradient descent"], "answer_arxiv_id": ["1509.01240"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_16444"} +{"question": "What is the work that introduced Neural Radiance Fields (NeRF)?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_16445"} +{"question": "What studies utilized linear probing to illustrate that BERT is inherently resilient to catastrophic forgetting, even without buffer data in task-incremental learning?", "answer": ["Can BERT Refrain from Forgetting on Sequential Tasks? A Probing Study"], "answer_arxiv_id": ["2303.01081"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_16446"} +{"question": "Could you provide me some studies about methods preserving parameters and storing edit instances explicitly?", "answer": ["Memory-Based Model Editing at Scale", "MQuAKE: Assessing Knowledge Editing in Language Models via Multi-Hop\n Questions"], "answer_arxiv_id": ["2206.06520", "2305.14795"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_16447"} +{"question": "What studies have proposed solutions for saliency-guided incremental learning?", "answer": ["Learning without Memorizing", "Remembering for the Right Reasons: Explanations Reduce Catastrophic\n Forgetting"], "answer_arxiv_id": ["1811.08051", "2010.01528"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_16448"} +{"question": "Could you specify some works that present a distributed robust learning procedure?", "answer": ["Robust Learning from Untrusted Sources", "Distributed Robust Learning"], "answer_arxiv_id": ["1901.10310", "1409.5937"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_16449"} +{"question": "Which works propose comparable individual CMI mentioned in this paper?", "answer": ["On Random Subset Generalization Error Bounds and the Stochastic Gradient Langevin Dynamics Algorithm", "Individually Conditional Individual Mutual Information Bound on Generalization Error"], "answer_arxiv_id": ["2010.10994", "2012.09922"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_16450"} +{"question": "Which study explored MAE in ISLR context and found it highly useful?", "answer": ["Self-Supervised Video Transformers for Isolated Sign Language\n Recognition"], "answer_arxiv_id": ["2309.02450"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_16451"} +{"question": "Can you name any studies that used deep learning for climate model downscaling?", "answer": ["DeepSD: Generating High Resolution Climate Change Projections through Single Image Super-Resolution"], "answer_arxiv_id": ["1703.03126"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_16452"} +{"question": "Which studies use logical reasoning and declarative programming in designing architectures for algorithm inference?", "answer": ["Learning Explanatory Rules from Noisy Data", "Neural Logic Machines"], "answer_arxiv_id": ["1711.04574", "1904.11694"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_16453"} +{"question": "Are there any works that have explored various parallelism strategies for accelerating the training and inference of large language models?", "answer": ["GPipe: Efficient Training of Giant Neural Networks using Pipeline\n Parallelism", "TeraPipe: Token-Level Pipeline Parallelism for Training Large-Scale\n Language Models", "Beyond Data and Model Parallelism for Deep Neural Networks", "Efficient Large-Scale Language Model Training on GPU Clusters Using\n Megatron-LM", "GSPMD: General and Scalable Parallelization for ML Computation Graphs", "OneFlow: Redesign the Distributed Deep Learning Framework from Scratch", "Alpa: Automating Inter- and Intra-Operator Parallelism for Distributed\n Deep Learning", "ZeRO: Memory Optimizations Toward Training Trillion Parameter Models", "ZeRO-Offload: Democratizing Billion-Scale Model Training", "PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel"], "answer_arxiv_id": ["1811.06965", "2102.07988", "1807.05358", "2104.04473", "2105.04663", "2110.15032", "2201.12023", "1910.02054", "2101.06840", "2304.11277"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16454"} +{"question": "Which research utilize importance sampling for correcting the distributional mismatch in offline policy evaluation?", "answer": ["Doubly Robust Policy Evaluation and Optimization"], "answer_arxiv_id": ["1503.02834"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_16455"} +{"question": "What studies propose training on audio-image pairs for learning audio-visual representations?", "answer": ["AudioCLIP: Extending CLIP to Image, Text and Audio", "Wav2CLIP: Learning Robust Audio Representations From CLIP"], "answer_arxiv_id": ["2106.13043v1", "2110.11499"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_16456"} +{"question": "Which papers explored defenses against targeted data poisoning attacks during the fine-tuning of CLIP?", "answer": ["Data Poisoning Attacks Against Multimodal Encoders"], "answer_arxiv_id": ["2209.15266"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_16457"} +{"question": "Could you tell me about the studies considering exploration in procedurally generated environments?", "answer": ["Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models", "Unifying Count-Based Exploration and Intrinsic Motivation", "#Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning", "Curiosity-driven Exploration by Self-supervised Prediction", "Large-Scale Study of Curiosity-Driven Learning", "Exploration by Random Network Distillation", "Never Give Up: Learning Directed Exploration Strategies", "First return, then explore"], "answer_arxiv_id": ["1507.00814", "1606.01868", "1611.04717v3", "1705.05363", "1808.04355", "1810.12894", "2002.06038", "2004.12919"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_16458"} +{"question": "Which studies introduce neural implicit mapping or large-scale environmental reconstruction in SLAM?", "answer": ["iMAP: Implicit Mapping and Positioning in Real-Time", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM"], "answer_arxiv_id": ["2103.12352", "2112.12130"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_16459"} +{"question": "What studies adapt online learning with delay in the BCO algorithm?", "answer": ["Online Learning with Optimism and Delay"], "answer_arxiv_id": ["2106.06885"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16460"} +{"question": "Which papers have used synthetic deployment by simulating feedback using supervised datasets for continual learning?", "answer": ["Reinforcement Learning for Bandit Neural Machine Translation with Simulated Human Feedback", "Improving a Neural Semantic Parser by Counterfactual Learning from Human Bandit Feedback"], "answer_arxiv_id": ["1707.07402", "1805.01252"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_16461"} +{"question": "Which work introduces the concept of 'outlier exposure'?", "answer": ["Deep Anomaly Detection with Outlier Exposure"], "answer_arxiv_id": ["1812.04606"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_16462"} +{"question": "What paper constructed a dataset with BabyAI oracle plans?", "answer": ["Think Before You Act: Unified Policy for Interleaving Language Reasoning with Actions"], "answer_arxiv_id": ["2304.11063"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16463"} +{"question": "Are there studies which apply learning-augmented methods to improve and accelerate heuristic solvers via operation selection?", "answer": ["Learning to Perform Local Rewriting for Combinatorial Optimization", "Learning 2-opt Heuristics for the Traveling Salesman Problem via Deep Reinforcement Learning"], "answer_arxiv_id": ["1810.00337", "2004.01608"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_16464"} +{"question": "Are there any studies exploring the issue of overfitting in the context of vision-language model adaptation?", "answer": ["Learning to Prompt for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling", "LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model", "LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention"], "answer_arxiv_id": ["2109.01134", "2111.03930", "2304.15010", "2303.16199"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_16465"} +{"question": "Which papers discuss selective weight loading in the context of large language models?", "answer": ["Sparse GPU Kernels for Deep Learning", "FlexGen: High-Throughput Generative Inference of Large Language Models\n with a Single GPU", "vDNN: Virtualized Deep Neural Networks for Scalable, Memory-Efficient\n Neural Network Design"], "answer_arxiv_id": ["2006.10901", "2303.06865", "1602.08124"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_16466"} +{"question": "Which papers have discussed the operations of typical diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models", "Improved Techniques for Training Score-Based Generative Models", "Improved Denoising Diffusion Probabilistic Models", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["1503.03585", "1907.05600", "2006.11239", "2006.09011", "2102.09672", "2206.00364"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_16467"} +{"question": "Could you name the studies that offered a solution to the issue of neural logic programming approaches being unable to handle large KGs?", "answer": ["RNNLogic: Learning Logic Rules for Reasoning on Knowledge Graphs"], "answer_arxiv_id": ["2010.04029"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_16468"} +{"question": "What are examples of datasets that have addressed two-hand interactions?", "answer": ["Capturing Hands in Action using Discriminative Salient Points and Physics Simulation"], "answer_arxiv_id": ["1506.02178"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_16469"} +{"question": "What is the paper behind the Progressive distillation strategy for online distillation?", "answer": ["Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2202.00512"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_16470"} +{"question": "What research papers are there about the causal interpretation of language models?", "answer": ["Localizing Model Behavior With Path Patching", "Interpretability at Scale: Identifying Causal Mechanisms in Alpaca"], "answer_arxiv_id": ["2304.05969", "2305.08809"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_16471"} +{"question": "Can you point out research about active in-context learning?", "answer": ["Active Prompting with Chain-of-Thought for Large Language Models", "Selective Annotation Makes Language Models Better Few-Shot Learners"], "answer_arxiv_id": ["2302.12246", "2209.01975"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_16472"} +{"question": "Could you provide me some research that have shown the impact of gradient sparsity on global convergence and communication complexity in Federated Learning?", "answer": ["Federated Learning over Wireless Networks: Convergence Analysis and Resource Allocation"], "answer_arxiv_id": ["1910.13067"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_16473"} +{"question": "Can you cite a work that regularizes the terminal state distribution of a skill to be close to the initial set of the following skill?", "answer": ["Adversarial Skill Chaining for Long-Horizon Robot Manipulation via Terminal State Regularization"], "answer_arxiv_id": ["2111.07999"], "source_meta": {"published_time": "20220906"}, "qid": "AutoScholarQuery_train_16474"} +{"question": "Could you name some studies that improved the performance of Knowledge Tracing by utilizing temporal feature embeddings?", "answer": ["SAINT+: Integrating Temporal Features for EdNet Correctness Prediction"], "answer_arxiv_id": ["2010.12042"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_16475"} +{"question": "Which works focus on analyzing the expressivity of networks based on an expected Gaussian complexity?", "answer": ["Zen-NAS: A Zero-Shot NAS for High-Performance Deep Image Recognition"], "answer_arxiv_id": ["2102.01063"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_16476"} +{"question": "What works focused on post-hoc interpretation methods of interpretable multimodal models?", "answer": ["MultiViz: Towards Visualizing and Understanding Multimodal Models", "M2Lens: Visualizing and Explaining Multimodal Models for Sentiment\n Analysis", "VL-InterpreT: An Interactive Visualization Tool for Interpreting\n Vision-Language Transformers"], "answer_arxiv_id": ["2207.00056", "2107.08264", "2203.17247"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_16477"} +{"question": "Could you provide me some works that use a guidance method for generating actions that maximize the learned Q-functions?", "answer": ["Is Conditional Generative Modeling all you need for Decision-Making?", "Offline Reinforcement Learning via High-Fidelity Generative Behavior Modeling", "Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2211.15657", "2209.14548", "2205.09991"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_16478"} +{"question": "Which work is TSDiff, the proposed model in this study, based on?", "answer": ["Diffusion-based Time Series Imputation and Forecasting with Structured State Space Models", "DiffWave: A Versatile Diffusion Model for Audio Synthesis"], "answer_arxiv_id": ["2208.09399", "2009.09761v3"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_16479"} +{"question": "Could you provide me some studies that use point clouds for 3D data representation?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "LION: Latent Point Diffusion Models for 3D Shape Generation", "Diffusion Probabilistic Models for 3D Point Cloud Generation"], "answer_arxiv_id": ["2212.08751", "2210.06978", "2103.01458"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16480"} +{"question": "Which works describe the method of region subdivision for extracting the exact polyhedral complex from a ReLU-network?", "answer": ["On the Expressive Power of Deep Neural Networks"], "answer_arxiv_id": ["1606.05336"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_16481"} +{"question": "Which papers contributed to the evaluation of the performance of supervised detectors?", "answer": ["M4: Multi-generator, Multi-domain, and Multi-lingual Black-Box\n Machine-Generated Text Detection"], "answer_arxiv_id": ["2305.14902"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_16482"} +{"question": "Which papers addressed molecular geometry pretraining?", "answer": ["Molecular Geometry Pretraining with SE(3)-Invariant Denoising Distance Matching"], "answer_arxiv_id": ["2206.13602"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_16483"} +{"question": "Could you provide me with examples of studies that focus on retrieval-based approaches for controllable text generation?", "answer": ["Megatron-Cntrl: Controllable Story Generation with External Knowledge Using Large-Scale Language Models"], "answer_arxiv_id": ["2010.00840"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_16484"} +{"question": "Which research works use deep learning methods for animating landscapes?", "answer": ["Animating Landscape: Self-Supervised Learning of Decoupled Motion and\n Appearance for Single-Image Video Synthesis"], "answer_arxiv_id": ["1910.07192"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_16485"} +{"question": "Which papers discussed analogical reasoning in the context of identifying a relational structure between two domains?", "answer": ["Abstraction and Analogy-Making in Artificial Intelligence"], "answer_arxiv_id": ["2102.10717"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_16486"} +{"question": "What work uses a 3D-UNet to denoise explicit voxel grids storing density and color for object shape and appearance synthesis?", "answer": ["DiffRF: Rendering-Guided 3D Radiance Field Diffusion"], "answer_arxiv_id": ["2212.01206"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_16487"} +{"question": "What work showed that a suitable discretization of the reverse SDE run for polynomially many steps generates samples that are close in statistical distance to the data distribution?", "answer": ["Sampling is as easy as learning the score: theory for diffusion models with minimal data assumptions", "Convergence of score-based generative modeling for general data distributions", "Improved Analysis of Score-based Generative Modeling: User-Friendly Bounds under Minimal Smoothness Assumptions"], "answer_arxiv_id": ["2209.11215", "2209.12381", "2211.01916"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_16488"} +{"question": "What studies propose neural improvement heuristics?", "answer": ["Learning to Perform Local Rewriting for Combinatorial Optimization", "Neural Large Neighborhood Search for the Capacitated Vehicle Routing Problem", "Learning Improvement Heuristics for Solving Routing Problems", "Learning to Iteratively Solve Routing Problems with Dual-Aspect Collaborative Transformer"], "answer_arxiv_id": ["1810.00337", "1911.09539", "1912.05784", "2110.02544"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_16489"} +{"question": "Which papers reported the use of DeepCAD, a dataset of 3D shapes corresponding to objects such as flanges, pipes and screws?", "answer": ["DeepCAD: A Deep Generative Network for Computer-Aided Design Models"], "answer_arxiv_id": ["2105.09492"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16490"} +{"question": "What papers provide model-free approaches in offline RL?", "answer": ["Is Pessimism Provably Efficient for Offline RL?", "A Minimalist Approach to Offline Reinforcement Learning", "Bellman-consistent Pessimism for Offline Reinforcement Learning", "Adversarially Trained Actor Critic for Offline Reinforcement Learning", "Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["2012.15085", "2106.06860", "2106.06926", "2202.02446", "2006.04779", "2110.06169"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_16491"} +{"question": "Could you provide me some research about cascaded models, which are a category of text-conditional diffusion models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2205.11487", "2112.10741"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_16492"} +{"question": "What papers focus on backward-warping-based Video Frame Interpolation?", "answer": ["Depth-Aware Video Frame Interpolation", "MEMC-Net: Motion Estimation and Motion Compensation Driven Neural\n Network for Video Interpolation and Enhancement", "IFRNet: Intermediate Feature Refine Network for Efficient Frame\n Interpolation", "AMT: All-Pairs Multi-Field Transforms for Efficient Frame Interpolation"], "answer_arxiv_id": ["1904.00830", "1810.08768", "2205.14620", "2304.09790"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_16493"} +{"question": "Which studies proposed a chart-based method for simulating Recursive Neural Networks through dynamic programming?", "answer": ["Jointly Learning Sentence Embeddings and Syntax with Unsupervised Tree-LSTMs"], "answer_arxiv_id": ["1705.09189"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_16494"} +{"question": "Which datasets have been translated to Chinese language for commonsense reasoning studies?", "answer": ["XNLI: Evaluating Cross-lingual Sentence Representations", "XCOPA: A Multilingual Dataset for Causal Commonsense Reasoning", "Few-shot Learning with Multilingual Language Models"], "answer_arxiv_id": ["1809.05053", "2005.00333", "2112.10668"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_16495"} +{"question": "Which work augmented a policy gradient update with a demonstration data-included weighted update?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations"], "answer_arxiv_id": ["1709.10087"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_16496"} +{"question": "Which works focus on the theoretical analysis of the representational capabilities of neural networks for solving PDEs?", "answer": ["Solving parametric PDE problems with artificial neural networks", "DGM: A deep learning algorithm for solving partial differential equations", "A proof that deep artificial neural networks overcome the curse of dimensionality in the numerical approximation of Kolmogorov partial differential equations with constant diffusion and nonlinear drift coefficients", "Deep neural network approximation for high-dimensional elliptic PDEs with boundary conditions", "A proof that rectified deep neural networks overcome the curse of dimensionality in the numerical approximation of semilinear heat equations"], "answer_arxiv_id": ["1707.03351", "1708.07469", "1809.07321", "2007.05384", "1901.10854"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_16497"} +{"question": "What works attempt to improve the generalization ability of graph neural networks?", "answer": ["Discovering Invariant Rationales for Graph Neural Networks", "Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanism", "Graph Invariant Learning with Subgraph Co-mixup for Out-Of-Distribution Generalization"], "answer_arxiv_id": ["2201.12872", "2201.12987", "2312.10988"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_16498"} +{"question": "What datasets consist of news articles for text summarization?", "answer": ["Abstractive Text Summarization Using Sequence-to-Sequence RNNs and\n Beyond", "Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional\n Neural Networks for Extreme Summarization"], "answer_arxiv_id": ["1602.06023", "1808.08745"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_16499"} +{"question": "Which research introduced the concept of a two-stage approach in the field of zero-shot open-domain models?", "answer": ["Zero-Shot Text-to-Image Generation", "Taming Transformers for High-Resolution Image Synthesis"], "answer_arxiv_id": ["2102.12092", "2012.09841"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_16500"} +{"question": "What research studies off-policy risk evaluation in bandit or RL setting?", "answer": ["Universal Off-Policy Evaluation", "Off-Policy Risk Assessment in Contextual Bandits"], "answer_arxiv_id": ["2104.12820", "2104.08977v2"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_16501"} +{"question": "What papers discussed anchor-free methods for action detection?", "answer": ["Learning Salient Boundary Feature for Anchor-free Temporal Action\n Localization", "An Efficient Spatio-Temporal Pyramid Transformer for Action Detection", "Proposal-Free Temporal Action Detection via Global Segmentation Mask\n Learning", "ActionFormer: Localizing Moments of Actions with Transformers", "DiffTAD: Temporal Action Detection with Proposal Denoising Diffusion", "Temporal Action Detection with Structured Segment Networks", "Actionness Estimation Using Hybrid Fully Convolutional Networks", "BSN: Boundary Sensitive Network for Temporal Action Proposal Generation"], "answer_arxiv_id": ["2103.13137", "2207.10448", "2207.06580", "2202.07925", "2303.14863", "1704.06228", "1604.07279", "1806.02964"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_16502"} +{"question": "Which studies focus on instruction following as a cross-task generalization across a pool of NLP tasks?", "answer": ["Cross-Task Generalization via Natural Language Crowdsourcing Instructions", "Finetuned Language Models Are Zero-Shot Learners", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "ExT5: Towards Extreme Multi-Task Scaling for Transfer Learning", "Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks"], "answer_arxiv_id": ["2104.08773v4", "2109.01652", "2110.08207", "2111.10952", "2204.07705v3"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_16503"} +{"question": "What paper tried and failed in adequately fine-tuning the CLIP for open-vocabulary segmentation?", "answer": ["Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2112.01071"], "source_meta": {"published_time": "20230930"}, "qid": "AutoScholarQuery_train_16504"} +{"question": "What work introduced a joint optimization problem with camera pose represented as a 6-degree-of-freedom matrix?", "answer": ["NeRF-⁣-: Neural Radiance Fields Without Known Camera Parameters"], "answer_arxiv_id": ["2102.07064"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_16505"} +{"question": "Which work developed the reference-free approach GPTScore for evaluating NLG tasks?", "answer": ["GPTScore: Evaluate as You Desire"], "answer_arxiv_id": ["2302.04166"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_16506"} +{"question": "Which publication presented the Point-Voxel Diffusion (PVD), a 3D diffusion model for generating shapes?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion"], "answer_arxiv_id": ["2104.03670"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_16507"} +{"question": "Which works applied contrastive loss objective to obtain low-dimension time series representations?", "answer": ["Learnable latent embeddings for joint behavioral and neural analysis"], "answer_arxiv_id": ["2204.00673"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_16508"} +{"question": "Can you name work that seeks to make datasets appear more optimal in the context of goal-conditioned reinforcement learning?", "answer": ["Hindsight Experience Replay", "Rewriting History with Inverse RL: Hindsight Inference for Policy Improvement", "Generalized Hindsight for Reinforcement Learning"], "answer_arxiv_id": ["1707.01495v3", "2002.11089", "2002.11708"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_16509"} +{"question": "What papers discuss the use of residual optical flow to account for dynamical objects?", "answer": ["GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose", "When the Sun Goes Down: Repairing Photometric Losses for All-Day Depth Estimation"], "answer_arxiv_id": ["1803.02276", "2206.13850"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_16510"} +{"question": "Are there any works that have taken the approach of learning an explainer model to directly predict the Shapley value?", "answer": ["FastSHAP: Real-Time Shapley Value Estimation"], "answer_arxiv_id": ["2107.07436"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_16511"} +{"question": "Which studies implement traditional PDE simulations in a differentiable manner for applying in a gradient-based optimization workflow?", "answer": ["DiffAqua: A Differentiable Computational Design Pipeline for Soft Underwater Swimmers with Shape Interpolation", "ChainQueen: A Real-Time Differentiable Physical Simulator for Soft Robotics", "DiffTaichi: Differentiable Programming for Physical Simulation", "PlasticineLab: A Soft-Body Manipulation Benchmark with Differentiable Physics", "DiffPD: Differentiable Projective Dynamics", "Differentiable Simulation of Soft Multi-body Systems", "Efficient Differentiable Simulation of Articulated Bodies", "ADD: Analytically Differentiable Dynamics for Multi-Body Systems with Frictional Contact", "FluidLab: A Differentiable Environment for Benchmarking Complex Fluid Manipulation", "A Differentiable Physics Engine for Deep Learning in Robotics"], "answer_arxiv_id": ["2104.00837", "1810.01054", "1910.00935", "2104.03311", "2101.05917", "2205.01758v1", "2109.07719v1", "2007.00987", "2303.02346", "1611.01652"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_16512"} +{"question": "What research introduced an image transform network for fast stylization to address the issue of traditional neural style transfer?", "answer": ["Perceptual Losses for Real-Time Style Transfer and Super-Resolution"], "answer_arxiv_id": ["1603.08155"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16513"} +{"question": "What is the pioneer work in point-based methods for point clouds semantic segmentation?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation"], "answer_arxiv_id": ["1612.00593"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_16514"} +{"question": "What research proposed techniques for improving the diffusion and training processes of diffusion probabilistic models?", "answer": ["Contextualized Diffusion Models for Text-Guided Image and Video\n Generation"], "answer_arxiv_id": ["2402.16627"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16515"} +{"question": "Which paper demonstrated the scalability of the video diffusion for conditional and unconditional video generation using the cascaded diffusion approach?", "answer": ["Video Diffusion Models"], "answer_arxiv_id": ["2204.03458"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_16516"} +{"question": "What papers have used Satisfiability Modulo Theories(SMT) or Mixed Integer Linear Programming(MILP) for neural networks verification?", "answer": ["Formal Verification of Piece-Wise Linear Feed-Forward Neural Networks"], "answer_arxiv_id": ["1705.01320"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_16517"} +{"question": "What are some of the recent source-free domain adaptation methods that use nearest neighbor information?", "answer": ["Nearest Neighborhood-Based Deep Clustering for Source Data-absent Unsupervised Domain Adaptation", "Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation"], "answer_arxiv_id": ["2107.12585", "2110.04202"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_16518"} +{"question": "In what papers a novel semantic preserving module was proposed to improve the performance of semantic image synthesis?", "answer": ["Semantic Image Synthesis with Spatially-Adaptive Normalization"], "answer_arxiv_id": ["1903.07291"], "source_meta": {"published_time": "20200331"}, "qid": "AutoScholarQuery_train_16519"} +{"question": "Are there studies that explored relevance of feature suppression in contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Intriguing Properties of Contrastive Losses"], "answer_arxiv_id": ["2002.05709", "2011.02803"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_16520"} +{"question": "Which are the existing models that primarily rely on CLIP or BLIP for multi-modal embedding?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2103.00020", "2201.12086"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_16521"} +{"question": "Which works explore the performance of Language Learning Models in performing in-context learning from natural language prompts?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "GPT-4 Technical Report"], "answer_arxiv_id": ["2005.14165", "2204.02311", "2211.05100", "2303.08774"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_16522"} +{"question": "What works discuss Competence-based IMs maximising the diversity of skills mastered by the agent?", "answer": ["Autotelic Agents with Intrinsically Motivated Goal-Conditioned Reinforcement Learning: A Short Survey"], "answer_arxiv_id": ["2012.09830"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_16523"} +{"question": "What researchers used an ensemble of one-step predictive models for exploration?", "answer": ["Planning to Explore via Self-Supervised World Models"], "answer_arxiv_id": ["2005.05960"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16524"} +{"question": "Which works initially proposed backdoor attacks to poison deep learning models?", "answer": ["Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning", "Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks", "Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networks", "Hidden Trigger Backdoor Attacks"], "answer_arxiv_id": ["1712.05526v1", "2007.02343", "1804.00792", "1910.00033"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_16525"} +{"question": "Are there recent developments that refine NeRF for unbounded environments?", "answer": ["Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields", "Nerfstudio: A Modular Framework for Neural Radiance Field Development"], "answer_arxiv_id": ["2304.06706", "2302.04264"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_16526"} +{"question": "Which papers provide information on expansion-based methods in continual learning?", "answer": ["Progressive Neural Networks", "Lifelong Learning with Dynamically Expandable Networks", "PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning", "Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgetting", "Compacting, Picking and Growing for Unforgetting Continual Learning", "Model Zoo: A Growing “Brain” That Learns Continually", "DyTox: Transformers for Continual Learning with DYnamic TOken eXpansion"], "answer_arxiv_id": ["1606.04671", "1708.01547", "1711.05769", "1904.00310", "1910.06562", "2106.03027", "2111.11326"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_16527"} +{"question": "Which papers have applied diffusion models in 2D image synthesis applications like image inpainting?", "answer": ["RePaint: Inpainting using Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2201.09865"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16528"} +{"question": "Which studies propose the use of natural language concept-based explanations that are better understood by the end users?", "answer": ["Automated Rationale Generation: A Technique for Explainable AI and its Effects on Human Perceptions", "Semantic-Based Explainable AI: Leveraging Semantic Scene Graphs and Pairwise Ranking to Explain Robot Failures", "Subgoal-Based Explanations for Unreliable Intelligent Decision Support Systems"], "answer_arxiv_id": ["1901.03729", "2108.03554", "2201.04204"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_16529"} +{"question": "What papers used contrastive learning as the pretraining objective and extended the SimCLR framework to tabular tasks?", "answer": ["Scarf: Self-Supervised Contrastive Learning using Random Feature Corruption"], "answer_arxiv_id": ["2106.15147"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_16530"} +{"question": "Which works use batch optimal solutions and their gradients for training generative models?", "answer": ["MMD GAN: Towards Deeper Understanding of Moment Matching Network", "Learning Generative Models with Sinkhorn Divergences", "Learning with minibatch Wasserstein : asymptotic and gradient properties"], "answer_arxiv_id": ["1705.08584", "1706.00292", "1910.04091"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_16531"} +{"question": "Are there any studies focused on controllable generation or sequence-to-sequence tasks with diffusion language models?", "answer": ["Diffusion-LM Improves Controllable Text Generation", "Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning", "DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models", "SeqDiffuSeq: Text Diffusion with Encoder-Decoder Transformers", "DiNoiSer: Diffused Conditional Sequence Learning by Manipulating Noises"], "answer_arxiv_id": ["2205.14217", "2208.04202", "2210.08933", "2212.10325", "2302.10025"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16532"} +{"question": "What are some studies about pre-training methods for molecules that use both 2D and 3D information?", "answer": ["3D Infomax improves GNNs for Molecular Property Prediction", "Pre-training Molecular Graph Representation with 3D Geometry", "Unified 2D and 3D Pre-Training of Molecular Representations"], "answer_arxiv_id": ["2110.04126", "2110.07728", "2207.08806"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_16533"} +{"question": "Which papers provide discussions about the limitations of the functional model TransE?", "answer": ["Quaternion Knowledge Graph Embeddings", "BoxE: A Box Embedding Model for Knowledge Base Completion"], "answer_arxiv_id": ["1904.10281", "2007.06267v2"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_16534"} +{"question": "Which papers use a few-shot technique in person-agnostic methods for talking-head synthesis?", "answer": ["Few-Shot Adversarial Learning of Realistic Neural Talking Head Models", "Neural Head Reenactment with Latent Pose Descriptors"], "answer_arxiv_id": ["1905.08233", "2004.12000"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_16535"} +{"question": "Which studies proposed diffusion-based text-to-image methods?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242", "2212.04488", "2302.13848"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_16536"} +{"question": "Could you provide me some studies that used cascaded diffusion models in video diffusion?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models"], "answer_arxiv_id": ["2210.02303"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_16537"} +{"question": "Could you provide me some studies on efficient design of MoE models?", "answer": ["GShard: Scaling Giant Models with Conditional Computation and Automatic\n Sharding", "Beyond Distillation: Task-level Mixture-of-Experts for Efficient\n Inference", "Task-Specific Expert Pruning for Sparse Mixture-of-Experts", "Parameter-Efficient Mixture-of-Experts Architecture for Pre-trained\n Language Models"], "answer_arxiv_id": ["2006.16668", "2110.03742", "2206.00277", "2203.01104"], "source_meta": {"published_time": "20230829"}, "qid": "AutoScholarQuery_train_16538"} +{"question": "Which papers have proposed domain generalization algorithms to mitigate the accuracy degradation caused by distribution shifts?", "answer": ["Invariant Risk Minimization", "Out-of-Distribution Generalization via Risk Extrapolation (REx)", "Learning to Generalize: Meta-Learning for Domain Generalization", "DecAug: Out-of-Distribution Generalization via Decomposed Feature Representation and Semantic Augmentation", "Stable Prediction across Unknown Environments", "NAS-OoD: Neural Architecture Search for Out-of-Distribution Generalization", "SWAD: Domain Generalization by Seeking Flat Minima", "Out-of-distribution Generalization with Causal Invariant Transformations", "Breaking Correlation Shift via Conditional Invariant Regularizer"], "answer_arxiv_id": ["1907.02893", "2003.00688v5", "1710.03463", "2012.09382", "1806.06270v2", "2109.02038", "2102.08604", "2203.11528", "2207.06687"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_16539"} +{"question": "What paper argued that the general architecture of transformers is key for performance, introducing a version with simple spatial pooling operator?", "answer": ["MetaFormer Is Actually What You Need for Vision"], "answer_arxiv_id": ["2111.11418"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_16540"} +{"question": "What are the neural network-driven approaches being used in surface completion from point clouds?", "answer": ["Unsupervised 3D Shape Completion through GAN Inversion", "Multimodal Shape Completion via Conditional Generative Adversarial Networks", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "Convolutional Occupancy Networks"], "answer_arxiv_id": ["2104.13366", "2003.07717", "1812.03828", "2003.04618"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_16541"} +{"question": "What are some papers that explore grounding language in environment states for action plan generation using LLMs?", "answer": ["Inner Monologue: Embodied Reasoning through Planning with Language Models", "LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Models"], "answer_arxiv_id": ["2207.05608", "2212.04088"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_16542"} +{"question": "What works utilize inducing point methods to improve the scalability of gps?", "answer": ["Scalable Gaussian Process Variational Autoencoders", "Sparse Gaussian Process Variational Autoencoders", "Longitudinal Variational Autoencoder"], "answer_arxiv_id": ["2010.13472v3", "2010.10177", "2006.09763"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_16543"} +{"question": "What studies introduce a discrete graph diffusion process by defining a Markov transition matrix for different node and edge types?", "answer": ["DiGress: Discrete Denoising diffusion for graph generation", "Diffusion Models for Graphs Benefit From Discrete State Spaces"], "answer_arxiv_id": ["2209.14734", "2210.01549"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_16544"} +{"question": "Which work developed a method to deal with inputs that exceed the context size of a large language model (LLM) by forming a tree of hierarchical summary nodes?", "answer": ["Walking Down the Memory Maze: Beyond Context Limit through Interactive\n Reading"], "answer_arxiv_id": ["2310.05029"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_16545"} +{"question": "Which research proposed a primal-dual stochastic approximation algorithm to solve a min-max reformulation employing kernel embedding techniques?", "answer": ["Learning from Conditional Distributions via Dual Embeddings"], "answer_arxiv_id": ["1607.04579"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_16546"} +{"question": "What works have been done in the field of cross-modal video recognition?", "answer": ["Disentangling Spatial and Temporal Learning for Efficient Image-to-Video\n Transfer Learning", "Implicit Temporal Modeling with Learnable Alignment for Video\n Recognition", "ActionCLIP: A New Paradigm for Video Action Recognition", "CLIP-ViP: Adapting Pre-trained Image-Text Model to Video-Language\n Representation Alignment", "Prompting Visual-Language Models for Efficient Video Understanding", "Expanding Language-Image Pretrained Models for General Video Recognition", "Bidirectional Cross-Modal Knowledge Exploration for Video Recognition\n with Pre-trained Vision-Language Models", "Revisiting Classifier: Transferring Vision-Language Models for Video\n Recognition"], "answer_arxiv_id": ["2309.07911", "2304.10465", "2109.08472", "2209.06430", "2112.04478", "2208.02816", "2301.00182", "2207.01297"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16547"} +{"question": "What papers contributed to improving recognition tasks through keypoint detection?", "answer": ["Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action\n Recognition"], "answer_arxiv_id": ["1801.07455"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_16548"} +{"question": "Which work proposed a unified framework known as stochastic interpolants and discussed the optimal choice between the probability flow and diffusion models?", "answer": ["Stochastic Interpolants: A Unifying Framework for Flows and Diffusions"], "answer_arxiv_id": ["2303.08797"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_16549"} +{"question": "Which works focus on diversity as a sampling criteria in Active Learning?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Variational Adversarial Active Learning"], "answer_arxiv_id": ["1708.00489", "1904.00370"], "source_meta": {"published_time": "20230917"}, "qid": "AutoScholarQuery_train_16550"} +{"question": "Could you provide some studies that adopted model distillation to eliminate hidden backdoors?", "answer": ["Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks", "Eliminating Backdoor Triggers for Deep Neural Networks Using Attention Relation Graph Distillation"], "answer_arxiv_id": ["2101.05930", "2204.09975"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_16551"} +{"question": "What papers discuss the use of inductive logic programming (ILP) for logical rule generation?", "answer": ["Learn to Explain Efficiently via Neural Logic Inductive Learning", "RNNLogic: Learning Logic Rules for Reasoning on Knowledge Graphs", "Neuro-Symbolic Inductive Logic Programming with Logical Neural Networks"], "answer_arxiv_id": ["1910.02481", "2010.04029", "2112.03324"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_16552"} +{"question": "Which paper proposed the concept of transformers in natural language processing?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20220607"}, "qid": "AutoScholarQuery_train_16553"} +{"question": "What researches have implemented Hamiltonian dynamics in the context of neural ODE?", "answer": ["Symplectic ODE-Net: Learning Hamiltonian Dynamics with Control"], "answer_arxiv_id": ["1909.12077"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_16554"} +{"question": "Can you list some works that concentrate on learning the inference machine in black-box VI?", "answer": ["Black Box Variational Inference"], "answer_arxiv_id": ["1401.0118v1"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_16555"} +{"question": "Could you provide me studies about using multi-task learning to develop customized models in the FL environment?", "answer": ["Federated Multi-Task Learning", "Decentralized Collaborative Learning of Personalized Models over Networks"], "answer_arxiv_id": ["1705.10467", "1610.05202"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_16556"} +{"question": "Which paper proposes a biologically-inspired mechanism to improve robustness of neural networks to adversarial perturbations?", "answer": ["Biologically Inspired Mechanisms for Adversarial Robustness"], "answer_arxiv_id": ["2006.16427"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_16557"} +{"question": "What works provide more details on obtaining the required Jacobians in an implicit layer using the implicit function theorem?", "answer": ["OptNet: Differentiable Optimization as a Layer in Neural Networks", "On the Differentiability of the Solution to Convex Optimization Problems", "Efficient and Modular Implicit Differentiation", "Differentiable Convex Optimization Layers", "Implicit Deep Learning"], "answer_arxiv_id": ["1703.00443", "1804.05098", "2105.15183", "1910.12430", "1908.06315"], "source_meta": {"published_time": "20220718"}, "qid": "AutoScholarQuery_train_16558"} +{"question": "Which papers demonstrated the efficacy of goal-conditional prediction in long-range prediction tasks?", "answer": ["TNT: Target-driveN Trajectory Prediction", "DenseTNT: End-to-end Trajectory Prediction from Dense Goal Sets", "CoverNet: Multimodal Behavior Prediction using Trajectory Sets", "MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction", "Map-Adaptive Goal-Based Trajectory Prediction"], "answer_arxiv_id": ["2008.08294", "2108.09640", "1911.10298", "1910.05449", "2009.04450"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16559"} +{"question": "What research poses theories for reducing sharpness-aware minimization's computational demands?", "answer": ["Towards Efficient and Scalable Sharpness-Aware Minimization", "Efficient Sharpness-aware Minimization for Improved Training of Neural\n Networks"], "answer_arxiv_id": ["2203.02714", "2110.03141"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16560"} +{"question": "Which works study the out-of-domain generalization?", "answer": ["On Guiding Visual Attention with Language Specification", "Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution", "Editing a classifier by rewriting its prediction rules", "Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["2202.08926", "2202.10054", "2112.01008", "1903.12261"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_16561"} +{"question": "Which study proposed a smaller encoder from the original SAM encoder for faster mask generation?", "answer": ["Faster Segment Anything: Towards Lightweight SAM for Mobile Applications"], "answer_arxiv_id": ["2306.14289"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_16562"} +{"question": "What research work developed the ODinW (Object Detection in the Wild) dataset?", "answer": ["Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2112.03857"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_16563"} +{"question": "Which work relaxed the bounded domain assumption for the extragradient method in the monotone case?", "answer": ["Revisiting Stochastic Extragradient"], "answer_arxiv_id": ["1905.11373v2"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_16564"} +{"question": "What research proposed controller-based search algorithm in neural architecture search?", "answer": ["Neural Architecture Search with Reinforcement Learning"], "answer_arxiv_id": ["1611.01578"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_16565"} +{"question": "Which papers have been published on the topic of Vision-Language Representational Learning based on contrastive loss?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Learning Transferable Visual Models From Natural Language Supervision", "Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm", "Robust Cross-Modal Representation Learning with Progressive Self-Distillation"], "answer_arxiv_id": ["1807.03748", "2103.00020", "2110.05208", "2204.04588"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_16566"} +{"question": "What works proposed a distance ratio in the latent space of a classifier in the context of failure detection?", "answer": ["To Trust Or Not To Trust A Classifier"], "answer_arxiv_id": ["1805.11783"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_16567"} +{"question": "Can you provide me some works on proactive motion capture with a single mobile camera?", "answer": ["Human Motion Capture Using a Drone"], "answer_arxiv_id": ["1804.06112"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_16568"} +{"question": "Which references extended the proposed smoothed GDA to the stochastic setting and obtained a specific iteration complexity?", "answer": ["Faster Single-loop Algorithms for Minimax Optimization without Strong Concavity"], "answer_arxiv_id": ["2112.05604"], "source_meta": {"published_time": "20221226"}, "qid": "AutoScholarQuery_train_16569"} +{"question": "What work describes an algorithm with minimax optimal sample complexity for infinite-horizon MDPs, depending on a particular MDP structure called anchor state-action pairs?", "answer": ["Sample-Optimal Parametric Q-Learning Using Linearly Additive Features"], "answer_arxiv_id": ["1902.04779"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_16570"} +{"question": "What works use graph neural networks, specifically GATs and Graph Convolutional Networks, to simulate interactions among different agents?", "answer": ["DAG-Net: Double Attentive Graph Neural Network for Trajectory\n Forecasting", "AVGCN: Trajectory Prediction using Graph Convolutional Networks Guided\n by Human Attention", "Social-BiGAT: Multimodal Trajectory Forecasting using Bicycle-GAN and\n Graph Attention Networks", "SGCN:Sparse Graph Convolution Network for Pedestrian Trajectory\n Prediction", "MSR-GCN: Multi-Scale Residual Graph Convolution Networks for Human\n Motion Prediction", "AVGCN: Trajectory Prediction using Graph Convolutional Networks Guided\n by Human Attention", "Social-STGCNN: A Social Spatio-Temporal Graph Convolutional Neural\n Network for Human Trajectory Prediction"], "answer_arxiv_id": ["2005.12661", "2101.05682", "1907.03395", "2104.01528", "2108.07152", "2101.05682", "2002.11927"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_16571"} +{"question": "Which studies discuss adaptive control for specific system classes such as fully-actuated systems?", "answer": ["Adaptive-Control-Oriented Meta-Learning for Nonlinear Systems"], "answer_arxiv_id": ["2103.04490"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_16572"} +{"question": "Which papers explored Diffusion Model for mel-spectrogram generation and waveform generation?", "answer": ["Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech", "ResGrad: Residual Denoising Diffusion Probabilistic Models for Text to\n Speech", "Bilateral Denoising Diffusion Models", "PriorGrad: Improving Conditional Denoising Diffusion Models with\n Data-Dependent Adaptive Prior"], "answer_arxiv_id": ["2105.06337", "2212.14518", "2108.11514", "2106.06406"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_16573"} +{"question": "Which works are about the use of dynamic network mechanisms to drop blocks in the field of model automation designing?", "answer": ["BlockDrop: Dynamic Inference Paths in Residual Networks", "Multi-Scale Dense Networks for Resource Efficient Image Classification", "SkipNet: Learning Dynamic Routing in Convolutional Networks"], "answer_arxiv_id": ["1711.08393", "1703.09844", "1711.09485"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_16574"} +{"question": "Could you provide me some studies about handcrafted feature-based methods for rapid development in ReID?", "answer": ["Person Re-identification by Local Maximal Occurrence Representation and\n Metric Learning"], "answer_arxiv_id": ["1406.4216"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_16575"} +{"question": "Could you provide me some papers that explore the use of large language models for social science questions?", "answer": ["ChatGPT outperforms crowd-workers for text-annotation tasks", "Large Language Models Can Be Used to Estimate the Latent Positions of Politicians", "Can Large Language Models Transform Computational Social Science?"], "answer_arxiv_id": ["2303.15056", "2303.12057", "2305.03514"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_16576"} +{"question": "What studies address the overconfidence issue in the inference of density estimations?", "answer": ["Confident Multiple Choice Learning"], "answer_arxiv_id": ["1706.03475"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_16577"} +{"question": "Could you provide some examples of research papers that use synthetic images for robustness evaluation?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias\n improves accuracy and robustness", "Benchmarking Neural Network Robustness to Common Corruptions and\n Perturbations", "Noise or Signal: The Role of Image Backgrounds in Object Recognition"], "answer_arxiv_id": ["1811.12231", "1903.12261", "2006.09994"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_16578"} +{"question": "What works proposed the general Double Oracle framework in the intersection of Game Theory and MARL", "answer": ["Online Double Oracle", "XDO: A Double Oracle Algorithm for Extensive-Form Games"], "answer_arxiv_id": ["2103.07780", "2103.06426"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_16579"} +{"question": "Which studies use historical trajectories from offline datasets to infer task representations during meta-training in OMRL?", "answer": ["FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization", "Multi-task Batch Reinforcement Learning with Metric Learning", "Robust Task Representations for Offline Meta-Reinforcement Learning via Contrastive Learning", "Offline Meta Reinforcement Learning with In-Distribution Online Adaptation"], "answer_arxiv_id": ["2010.01112", "1909.11373", "2206.10442", "2305.19529"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_16580"} +{"question": "Could you provide me some works that discuss meta-learning-based methods to simulate domain shifts during training in order to learn a robust representative feature space?", "answer": ["Meta-Teacher For Face Anti-Spoofing", "Learning Meta Pattern for Face Anti-Spoofing", "Adaptive Mixture of Experts Learning for Generalizable Face\n Anti-Spoofing", "Learning Meta Pattern for Face Anti-Spoofing"], "answer_arxiv_id": ["2111.06638", "2110.06753", "2207.09868", "2110.06753"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16581"} +{"question": "What works explored the concept of having a readable and writable tape in turing machines as differentiable versions of computer parts?", "answer": ["Neural Turing Machines", "Reinforcement Learning Neural Turing Machines - Revised", "Neural Stored-program Memory"], "answer_arxiv_id": ["1410.5401", "1505.00521", "1906.08862"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_16582"} +{"question": "What works investigate the generation of curiosity-driven and inquisitive questions in the field of QG?", "answer": ["Inquisitive Question Generation for High Level Text Comprehension", "Ask to Learn: A Study on Curiosity-driven Question Generation", "FOLLOWUPQG: Towards Information-Seeking Follow-up Question Generation"], "answer_arxiv_id": ["2010.01657", "1911.03350", "2309.05007"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_16583"} +{"question": "Could you provide me some studies about non-contrastive methods for self-supervised representation learning?", "answer": ["The Mechanism of Prediction Head in Non-contrastive Self-supervised Learning", "Understanding Self-Supervised Learning Dynamics without Contrastive Pairs", "On the duality between contrastive and non-contrastive self-supervised learning", "Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods"], "answer_arxiv_id": ["2205.06226", "2102.06810", "2206.02574", "2205.11508"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_16584"} +{"question": "What works did the researcher cite about learning a sequence of MDPs with side information?", "answer": ["Markov Decision Processes with Continuous Side Information"], "answer_arxiv_id": ["1711.05726"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16585"} +{"question": "Which studies successfully detected dynamic tracks with additional depth information?", "answer": ["Self-supervised Learning with Geometric Constraints in Monocular Video:\n Connecting Flow, Depth, and Camera"], "answer_arxiv_id": ["1907.05820"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_16586"} +{"question": "Which studies have successfully applied mitigating pessimistic lower bounds of classic worst-case analysis via untrusted predictions to scheduling problems?", "answer": ["Flow Time Scheduling with Uncertain Processing Time", "Distortion-Oblivious Algorithms for Minimizing Flow Time", "Permutation Predictions for Non-Clairvoyant Scheduling", "A Novel Prediction Setup for Online Speed-Scaling", "Algorithms with Prediction Portfolios"], "answer_arxiv_id": ["2103.05604", "2109.08424", "2202.10199", "2112.03082", "2210.12438"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_16587"} +{"question": "What papers discuss the topic of test-time training with self-supervised learning in computer vision, robotics, and visual RL?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "Test-Time Training with Masked Autoencoders", "Self-Supervised Policy Adaptation during Deployment"], "answer_arxiv_id": ["1909.13231", "1909.13231", "2209.07522", "2007.04309"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_16588"} +{"question": "Can you provide papers that focus on research related to diffusion models in the context of image-text data?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_16589"} +{"question": "Could you provide me some studies about the similarities between Contrastive and Non-contrastive Self-Supervised Learning?", "answer": ["On the duality between contrastive and non-contrastive self-supervised\n learning"], "answer_arxiv_id": ["2206.02574"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_16590"} +{"question": "What papers worked on mitigating bias by applying strategies such as re-sampling and weighting, distributional robust optimization, and so on?", "answer": ["REPAIR: Removing Representation Bias by Dataset Resampling", "Bias Mimicking: A Simple Sampling Approach for Bias Mitigation", "Mitigating Unwanted Biases with Adversarial Learning"], "answer_arxiv_id": ["1904.07911", "2209.15605", "1801.07593"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_16591"} +{"question": "Which works focused on research in Human-Scene Interaction (HSI) pertaining to populating static human figures into 3D scenes?", "answer": ["Holistic++ Scene Understanding: Single-view 3D Holistic Scene Parsing\n and Human Pose Estimation with Human-Object Interaction and Physical\n Commonsense", "Generating 3D People in Scenes without People", "Populating 3D Scenes by Learning Human-Scene Interaction", "Detecting Human-Object Contact in Images"], "answer_arxiv_id": ["1909.01507", "1912.02923", "2012.11581", "2303.03373"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_16592"} +{"question": "Which paper proposed a solution to MVPS using neural radiance fields (NeRFs)?", "answer": ["Neural Radiance Fields Approach to Deep Multi-View Photometric Stereo"], "answer_arxiv_id": ["2110.05594"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_16593"} +{"question": "Which studies deal with the efforts to equip machines with the ability of abductive reasoning?", "answer": ["Interactive Visual Reasoning under Uncertainty"], "answer_arxiv_id": ["2206.09203"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_16594"} +{"question": "What studies have replaced hand-crafted prompts with learnable soft prompts?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16595"} +{"question": "Could you provide me with studies about the use of data augmentation methods to deal with image corruptions in deep neural networks?", "answer": ["A simple way to make neural networks robust against diverse image\n corruptions", "ImageNet-trained CNNs are biased towards texture; increasing shape bias\n improves accuracy and robustness", "AutoAugment: Learning Augmentation Policies from Data", "AugMix: A Simple Data Processing Method to Improve Robustness and\n Uncertainty", "The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution\n Generalization"], "answer_arxiv_id": ["2001.06057", "1811.12231", "1805.09501", "1912.02781", "2006.16241"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_16596"} +{"question": "Could you provide me a study that discovers a larger language model improves the quality of text-to-image generation?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2205.11487"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_16597"} +{"question": "Which study proposed CLIP as a representative Visual Language Model?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_16598"} +{"question": "Could you name a study that proposed splitting the tokens based on informativeness and then fusing the tokens considering their diversity?", "answer": ["Beyond Attentive Tokens: Incorporating Token Importance and Diversity\n for Efficient Vision Transformers"], "answer_arxiv_id": ["2211.11315"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_16599"} +{"question": "Which works provide an empirical comparison of Experience Replay (ER) with learned models under a variety of search control strategies?", "answer": ["Organizing Experience: A Deeper Look at Replay Mechanisms for Sample-based Planning in Continuous State Domains"], "answer_arxiv_id": ["1806.04624"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_16600"} +{"question": "Which papers study Pandora's Box when a box can be selected without paying for it?", "answer": ["Pandora’s Problem with Nonobligatory Inspection"], "answer_arxiv_id": ["1905.01428"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_16601"} +{"question": "What works adopted the same methodology and showcase noticeable capabilities in various domains, like Atari games?", "answer": ["Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models", "Mastering Diverse Domains through World Models"], "answer_arxiv_id": ["1912.01603", "2010.02193", "2301.04104"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_16602"} +{"question": "Are there any works that guide LLMs to produce hallucinatory responses artificially?", "answer": ["HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large\n Language Models", "Generating Benchmarks for Factuality Evaluation of Language Models"], "answer_arxiv_id": ["2305.11747", "2307.06908"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_16603"} +{"question": "What studies made use of optimization-based methods for single-agent visual SLAM?", "answer": ["ORB-SLAM: a Versatile and Accurate Monocular SLAM System", "ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D Cameras", "VINS-Mono: A Robust and Versatile Monocular Visual-Inertial State Estimator", "ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAM"], "answer_arxiv_id": ["1502.00956", "1610.06475", "1708.03852", "2007.11898"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_16604"} +{"question": "Which works represent new benchmark datasets for visual object tracking tasks?", "answer": ["LaSOT: A High-quality Benchmark for Large-scale Single Object Tracking", "TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild", "GOT-10k: A Large High-Diversity Benchmark for Generic Object Tracking in the Wild", "Towards More Flexible and Accurate Object Tracking with Natural Language: Algorithms and Benchmark"], "answer_arxiv_id": ["1809.07845", "1803.10794", "1810.11981", "2103.16746"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16605"} +{"question": "Which methods based on federated learning constraing local updates by adding penalty terms to address data heterogeneity?", "answer": ["Federated Optimization in Heterogeneous Networks", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Federated Learning Based on Dynamic Regularization"], "answer_arxiv_id": ["1812.06127", "1910.06378", "2111.04263"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_16606"} +{"question": "What papers presented works related to reconstructing dynamic NeRFs from monocular videos by learning a deformation field that maps sampled points from the deformed space to the canonical space?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View\n Synthesis of a Dynamic Scene From Monocular Video"], "answer_arxiv_id": ["2011.13961", "2011.12948", "2106.13228v2", "2012.12247"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_16607"} +{"question": "Which works introduced counterfactual explanations that provide information on why a specific decision was made?", "answer": ["Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR"], "answer_arxiv_id": ["1711.00399v3"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_16608"} +{"question": "What researches explored scene flow estimation in relation with the task of predicting depth from a single frame?", "answer": ["EffiScene: Efficient Per-Pixel Rigidity Inference for Unsupervised Joint Learning of Optical Flow, Depth, Camera Pose and Motion Segmentation", "Self-Supervised Monocular Scene Flow Estimation", "Self-Supervised Multi-Frame Monocular Scene Flow", "Learning Optical Flow, Depth, and Scene Flow without Real-World Labels"], "answer_arxiv_id": ["2011.08332", "2004.04143", "2105.02216", "2203.15089"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_16609"} +{"question": "Which research papers have proposed the use of learnable attention weights in aggregating neighbor information for GNNs?", "answer": ["Graph Attention Networks"], "answer_arxiv_id": ["1710.10903"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_16610"} +{"question": "Can you tell me which research discusses the gradient starvation phenomenon that is observed when a network already possesses spurious features?", "answer": ["Gradient Starvation: A Learning Proclivity in Neural Networks"], "answer_arxiv_id": ["2011.09468"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_16611"} +{"question": "What study processed neural networks using another neural network that operates on a combination of their high-order spatial derivatives?", "answer": ["Signal Processing for Implicit Neural Representations"], "answer_arxiv_id": ["2210.08772"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_16612"} +{"question": "Which works discuss training the online network to predict the representation of another view of the same image in self-supervised learning?", "answer": ["Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2006.07733", "2104.14294"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_16613"} +{"question": "Could you point out some studies about Meta-classification-based detectors (MCDs) for backdoor model detection?", "answer": ["Detecting AI Trojans Using Meta Neural Analysis", "Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs"], "answer_arxiv_id": ["1910.03137", "1906.10842"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_16614"} +{"question": "What researches focus on development of Concept Bottleneck Models (CBMs)?", "answer": ["Concept Bottleneck Models"], "answer_arxiv_id": ["2007.04612"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_16615"} +{"question": "What are some studies that have used behavioral testing in their counterfactual benchmarks to test the faithfulness of interpretability method?", "answer": ["Faithfulness Tests for Natural Language Explanations", "FIND: A Function Description Benchmark for Evaluating Interpretability Methods", "ALMANACS: A Simulatability Benchmark for Language Model Explainability"], "answer_arxiv_id": ["2305.18029", "2309.03886v3", "2312.12747"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16616"} +{"question": "Which studies work on different training recipes for VLMs where they freeze the LLM and train auxiliary components?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "CogVLM: Visual Expert for Pretrained Language Models"], "answer_arxiv_id": ["2204.14198", "2201.12086", "2301.12597", "2311.03079"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_16617"} +{"question": "Which papers have been published about black-box adversarial attacks?", "answer": ["Black-Box Adversarial Attack with Transferable Model-based Embedding"], "answer_arxiv_id": ["1911.07140"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16618"} +{"question": "What studies introduce approaches to enable or enhance the serving of dynamic Deep Learning models?", "answer": ["Nimble: Efficiently Compiling Dynamic Neural Networks for Model\n Inference", "NestDNN: Resource-Aware Multi-Tenant On-Device Deep Learning for\n Continuous Mobile Vision", "LegoDNN: Block-grained Scaling of Deep Neural Networks for Mobile Vision", "AdaptiveNet: Post-deployment Neural Architecture Adaptation for Diverse\n Edge Environments", "Dynamic Slicing for Deep Neural Networks"], "answer_arxiv_id": ["2006.03031", "1810.10090", "2112.09852", "2303.07129", "2009.13747"], "source_meta": {"published_time": "20230829"}, "qid": "AutoScholarQuery_train_16619"} +{"question": "Could you provide me some studies that improved frame-level detector in VQL?", "answer": ["Negative Frames Matter in Egocentric Visual Query 2D Localization", "Where is my Wallet? Modeling Object Proposal Sets for Egocentric Visual Query Localization"], "answer_arxiv_id": ["2208.01949", "2211.10528"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_16620"} +{"question": "Could you tell me about the works that introduce a threshold in BNNs optimization to determine whether to flip a binary weight?", "answer": ["Latent Weights Do Not Exist: Rethinking Binarized Neural Network Optimization", "A Bop and Beyond: A Second Order Optimizer for Binarized Neural Networks"], "answer_arxiv_id": ["1906.02107", "2104.05124v1"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_16621"} +{"question": "Which work popularized the phenomenon of adversarial attacks?", "answer": ["Intriguing properties of neural networks"], "answer_arxiv_id": ["1312.6199"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_16622"} +{"question": "Which research papers incorporate DFKD methods in Federated Learning?", "answer": ["Data-Free Knowledge Distillation for Heterogeneous Federated Learning"], "answer_arxiv_id": ["2105.10056"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_16623"} +{"question": "Are there any works that employ a classifier-free manner in text-to-motion by conditioning on CLIP?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2204.06125", "2205.11487"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16624"} +{"question": "What papers utilized text input for style condition or determining the content in neural style transfer?", "answer": ["Zero-Shot Contrastive Loss for Text-Guided Diffusion Image Style\n Transfer"], "answer_arxiv_id": ["2303.08622"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_16625"} +{"question": "Can you provide me a research that improves the quality of reconstruction by fixing each timestep drifting in diffusion-based image editing?", "answer": ["Null-text Inversion for Editing Real Images using Guided Diffusion\n Models"], "answer_arxiv_id": ["2211.09794"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_16626"} +{"question": "Which works focused on estimating poses of individual parts without using shape databases or known semantic information?", "answer": ["Generative 3D Part Assembly via Dynamic Graph Learning", "RGL-NET: A Recurrent Graph Learning framework for Progressive Part\n Assembly", "3D Part Assembly Generation with Instance Encoded Transformer", "Neural Shape Mating: Self-Supervised Object Assembly with Adversarial\n Shape Priors"], "answer_arxiv_id": ["2006.07793", "2107.12859", "2207.01779", "2205.14886"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16627"} +{"question": "What works follow a similar approach with DCLS by re-parameterize the kernel weights?", "answer": ["CKConv: Continuous Kernel Convolution For Sequential Data", "FlexConv: Continuous Kernel Convolutions with Differentiable Kernel Sizes"], "answer_arxiv_id": ["2102.02611", "2110.08059"], "source_meta": {"published_time": "20211207"}, "qid": "AutoScholarQuery_train_16628"} +{"question": "Any works that perform point cloud completion using diffusion models over a compact latent representation of neural SDFs?", "answer": ["Diffusion-SDF: Conditional Generative Modeling of Signed Distance Functions", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation"], "answer_arxiv_id": ["2211.13757", "2212.04493"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_16629"} +{"question": "Which papers proposed using personalized prototypes in non-IID Federated Learning?", "answer": ["FedProto: Federated Prototype Learning across Heterogeneous Clients", "Federated Learning from Pre-Trained Models: A Contrastive Learning Approach"], "answer_arxiv_id": ["2105.00243", "2209.10083"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_16630"} +{"question": "What work empowers the trend of using LLMs to create targeted, high-quality, and diverse annotations?", "answer": ["Verbs in Action: Improving verb understanding in video-language models"], "answer_arxiv_id": ["2304.06708"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_16631"} +{"question": "Which research proposed the Arcade Learning Environment (ALE) for reinforcement learning?", "answer": ["The Arcade Learning Environment: An Evaluation Platform for General Agents"], "answer_arxiv_id": ["1207.4708"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_16632"} +{"question": "Which studies show the growing interest in using machine learning for the problem of generating de novo sequences that improve upon natural sequences?", "answer": ["Machine learning-guided directed evolution for protein engineering", "Population-Based Black-Box Optimization for Biological Sequence Design"], "answer_arxiv_id": ["1811.10775", "2006.03227v2"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_16633"} +{"question": "Can you list the works which allocate tensors and computation operations to either GPUs or CPUs according to the data-flow graph?", "answer": ["ZeRO-Offload: Democratizing Billion-Scale Model Training"], "answer_arxiv_id": ["2101.06840"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_16634"} +{"question": "Are there any studies employing a universal large multimodal model for text detection, recognition, spotting, and understanding?", "answer": ["UniDoc: A Universal Large Multimodal Model for Simultaneous Text\n Detection, Recognition, Spotting and Understanding"], "answer_arxiv_id": ["2308.11592"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16635"} +{"question": "Which work described the data programming (DP) framework in the context of programmatic weak supervision?", "answer": ["Data Programming: Creating Large Training Sets, Quickly"], "answer_arxiv_id": ["1605.07723"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_16636"} +{"question": "What statistical methods are used in AI-text detection?", "answer": ["GLTR: Statistical Detection and Visualization of Generated Text", "DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1906.04043", "2301.11305", "1910.10683"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_16637"} +{"question": "Any examples of INRs applications in various domains?", "answer": ["​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​", "NeRV: Neural Representations for Videos", "COIN: COmpression with Implicit Neural representations", "Light Field Networks: Neural Scene Representations with Single-Evaluation Rendering", "Block-NeRF: Scalable Large Scene Neural View Synthesis"], "answer_arxiv_id": ["2103.13415", "2110.13903", "2103.03123", "2106.02634", "2202.05263"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_16638"} +{"question": "What are some of the benchmarks for evaluating language progress in NLP?", "answer": ["GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language\n Understanding", "SuperGLUE: A Stickier Benchmark for General-Purpose Language\n Understanding Systems"], "answer_arxiv_id": ["1804.07461", "1905.00537"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_16639"} +{"question": "Could you provide studies that highlighted the development of language models towards solving math benchmarks?", "answer": ["Large Language Models for Mathematical Reasoning: Progresses and\n Challenges", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2402.00157", "2302.13971"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16640"} +{"question": "Which works discuss the convergence guarantee of policy gradient in Cooperative MGs using direct parameterization?", "answer": ["Global Convergence of Multi-Agent Policy Gradient in Markov Potential Games"], "answer_arxiv_id": ["2106.01969"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_16641"} +{"question": "Which research works focus on devising procedures to evaluate the harmfulness and toxicity in Large Language Models' outputs?", "answer": ["Reducing Sentiment Bias in Language Models via Counterfactual Evaluation", "Language Models are Few-Shot Learners", "Toxicity in ChatGPT: Analyzing Persona-assigned Language Models"], "answer_arxiv_id": ["1911.03064", "2005.14165", "2304.05335"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_16642"} +{"question": "Which research proposed Sliced Score Matching loss as an approximation method for computational efficiency?", "answer": ["Sliced Score Matching: A Scalable Approach to Density and Score Estimation"], "answer_arxiv_id": ["1905.07088"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_16643"} +{"question": "Could you give me some examples of studies that have used randomized algorithms to improve the competitive ratio?", "answer": ["Online Vertex-Weighted Bipartite Matching and Single-bid Budgeted Allocations"], "answer_arxiv_id": ["1007.1271"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_16644"} +{"question": "Which papers utilized locality-sensitive-hashing (LSH) to retrieve the closest key vectors in terms of attention matrix?", "answer": ["Reformer: The Efficient Transformer"], "answer_arxiv_id": ["2001.04451"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16645"} +{"question": "Which study proposed the optimization of training time and learned details by adding a trainable embedding layer?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​"], "answer_arxiv_id": ["2201.05989", "2103.13415"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_16646"} +{"question": "Which studies utilize domain adaptation to solve out-of-distribution generalization problems?", "answer": ["Domain-Adversarial Training of Neural Networks", "Deep CORAL: Correlation Alignment for Deep Domain Adaptation"], "answer_arxiv_id": ["1505.07818", "1607.01719"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_16647"} +{"question": "Could you provide me some papers discussing knowledge distillation for compressing models?", "answer": ["Patch Slimming for Efficient Vision Transformers", "Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillation", "MSD: Multi-Self-Distillation Learning via Multi-classifiers within Deep Neural Networks", "Similarity-Preserving Knowledge Distillation", "Knowledge Distillation by On-the-Fly Native Ensemble"], "answer_arxiv_id": ["2106.02852", "1905.08094", "1911.09418", "1907.09682", "1806.04606"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_16648"} +{"question": "Which works used Neural Theorem Provers in the ILP?", "answer": ["End-to-End Differentiable Proving", "Logical Rule Induction and Theory Learning Using Neural Theorem Proving", "Neuro-Symbolic Hierarchical Rule Induction"], "answer_arxiv_id": ["1705.11040", "1809.02193", "2112.13418"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_16649"} +{"question": "What paper demonstrated that in two-layer leaky ReLU networks, SGD on the hinge loss for linearly separable data converges to zero loss?", "answer": ["SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data"], "answer_arxiv_id": ["1710.10174"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_16650"} +{"question": "Which papers proposed methods that only compress up-link messages in non-private compression in Federated Learning (FL)?", "answer": ["Qsparse-local-SGD: Distributed SGD with Quantization, Sparsification, and Local Computations", "FedPAQ: A Communication-Efficient Federated Learning Method with Periodic Averaging and Quantization", "Federated Learning with Compression: Unified Analysis and Sharp Guarantees", "Linear Convergence in Federated Learning: Tackling Client Heterogeneity and Sparse Gradients", "Federated Optimization Algorithms with Random Reshuffling and Gradient Compression"], "answer_arxiv_id": ["1906.02367", "1909.13014", "2007.01154v2", "2102.07053", "2206.07021"], "source_meta": {"published_time": "20221108"}, "qid": "AutoScholarQuery_train_16651"} +{"question": "What papers offer ways to accelerate the DDIM sampling process?", "answer": ["On Fast Sampling of Diffusion Probabilistic Models", "Learning to Efficiently Sample from Diffusion Probabilistic Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2106.00132", "2106.03802", "2202.09778"], "source_meta": {"published_time": "20220611"}, "qid": "AutoScholarQuery_train_16652"} +{"question": "Could you tell me what researches introduced Temperature Scaling (TS) and Ensemble Temperature Scaling (ETS)?", "answer": ["On Calibration of Modern Neural Networks", "Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning"], "answer_arxiv_id": ["1706.04599", "2003.07329"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_16653"} +{"question": "Which works discussed the use of goal relabeling to handle the sparse reward problem in Goal-conditioned Reinforcement Learning (GCRL)?", "answer": ["Hindsight Experience Replay", "Generalized Hindsight for Reinforcement Learning", "Rewriting History with Inverse RL: Hindsight Inference for Policy Improvement", "MHER: Model-based Hindsight Experience Replay"], "answer_arxiv_id": ["1707.01495v3", "2002.11708", "2002.11089", "2107.00306"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16654"} +{"question": "What publications describe the use of factorized space-time U-Net compatible with joint image-video training in text-guided video editing?", "answer": ["Video Diffusion Models"], "answer_arxiv_id": ["2204.03458"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16655"} +{"question": "Could you point me to a study that shows the relationship between the singular values of the Jacobian matrix of a DNN layer and the Lipshitz Directed Initialization?", "answer": ["A Signal Propagation Perspective for Pruning Neural Networks at Initialization"], "answer_arxiv_id": ["1906.06307"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_16656"} +{"question": "Could you provide me some studies about masked image modeling in the context of self-supervised learning?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers", "iBOT : Image BERT Pre-Training with Online Tokenizer", "SimMIM: a Simple Framework for Masked Image Modeling"], "answer_arxiv_id": ["2111.06377", "2106.08254", "2111.07832", "2111.09886"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16657"} +{"question": "Which works use surface features to rank and extract salient sentences into summaries as part of extractive review summarization approaches?", "answer": ["Summarizing Opinions: Aspect Extraction Meets Sentiment Prediction and\n They Are Both Weakly Supervised", "Weakly-Supervised Opinion Summarization by Leveraging External\n Information"], "answer_arxiv_id": ["1808.08858", "1911.09844"], "source_meta": {"published_time": "20240719"}, "qid": "AutoScholarQuery_train_16658"} +{"question": "What approach did Wang et al. utilize to deal with noisy samples?", "answer": ["PiCO+: Contrastive Label Disambiguation for Robust Partial Label Learning"], "answer_arxiv_id": ["2201.08984"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_16659"} +{"question": "Are there any papers about test-time adaptation methods that use Masked autoencoding loss?", "answer": ["Test-Time Training with Masked Autoencoders", "Context Encoders: Feature Learning by Inpainting", "Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2209.07522", "1604.07379", "2111.06377"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_16660"} +{"question": "Which works focus on building general group equivariant neural networks?", "answer": ["Group Equivariant Convolutional Networks", "On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "A General Theory of Equivariant CNNs on Homogeneous Spaces", "Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data", "A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups"], "answer_arxiv_id": ["1602.07576", "1802.03690", "1811.02017", "2002.12880", "2104.09459"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_16661"} +{"question": "Which works explored environments progressively with an incremental 3DSG methodology?", "answer": ["SceneGraphFusion: Incremental 3D Scene Graph Prediction from RGB-D Sequences"], "answer_arxiv_id": ["2103.14898"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_16662"} +{"question": "Could you provide me a work that generalizes NLF by enabling cross-view communication through the attention mechanism?", "answer": ["Generalizable Patch-Based Neural Rendering"], "answer_arxiv_id": ["2207.10662v2"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_16663"} +{"question": "Which paper is about unusual paraphrases in a diagnostic set?", "answer": ["BERTs of a feather do not generalize together: Large variability in generalization across models with similar test set performance"], "answer_arxiv_id": ["1911.02969"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_16664"} +{"question": "Which works proposed visual grounding datasets for indoor 3D scene understanding?", "answer": ["REVERIE: Remote Embodied Visual Referring Expression in Real Indoor Environments", "Refer-it-in-RGBD: A Bottom-up Approach for 3D Visual Grounding in RGBD Images", "ScanRefer: 3D Object Localization in RGB-D Scans using Natural Language"], "answer_arxiv_id": ["1904.10151", "2103.07894", "1912.08830"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_16665"} +{"question": "Which paper suggested employing two distinct batch normalization layers for different types of augmented training samples?", "answer": ["Adversarial Examples Improve Image Recognition"], "answer_arxiv_id": ["1911.09665"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_16666"} +{"question": "What research papers discuss the application of code prompts to story understanding?", "answer": ["CoRRPUS: Code-based Structured Prompting for Neurosymbolic Story\n Understanding"], "answer_arxiv_id": ["2212.10754"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_16667"} +{"question": "Which paper first formalized Novel Category Discovery (NCD) as deep transfer clustering for discovering unlabeled new classes?", "answer": ["Learning to Discover Novel Visual Categories via Deep Transfer\n Clustering"], "answer_arxiv_id": ["1908.09884"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_16668"} +{"question": "Can you name some works that introduced a spatial structure on the latent codes using potentially adaptive grids?", "answer": ["3D Shape Generation with Grid-based Implicit Functions", "Local Implicit Grid Representations for 3D Scenes", "Convolutional Occupancy Networks", "Implicit Functions in Feature Space for 3D Shape Reconstruction and Completion"], "answer_arxiv_id": ["2107.10607", "2003.08981", "2003.04618", "2003.01456"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_16669"} +{"question": "Do any algorithms apart from ProtoNet design methods principally or exclusively for classification?", "answer": ["Prototypical Networks for Few-shot Learning"], "answer_arxiv_id": ["1703.05175"], "source_meta": {"published_time": "20220505"}, "qid": "AutoScholarQuery_train_16670"} +{"question": "Which works focus on pedestrian attribute recognition in human-centric perception?", "answer": ["Rethinking of Pedestrian Attribute Recognition: A Reliable Evaluation under Zero-Shot Pedestrian Identity Setting"], "answer_arxiv_id": ["2107.03576"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_16671"} +{"question": "Which studies focus on designing Word Sense Disambiguation systems in supervised settings using pre-trained language models?", "answer": ["Moving Down the Long Tail of Word Sense Disambiguation with\n Gloss-Informed Biencoders"], "answer_arxiv_id": ["2005.02590"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_16672"} +{"question": "Which studies used kernels for uncertainty quantification, measure calibration quality, or as regularization during training?", "answer": ["To Trust Or Not To Trust A Classifier", "Calibration tests in multi-class classification: A unifying framework"], "answer_arxiv_id": ["1805.11783", "1910.11385"], "source_meta": {"published_time": "20220215"}, "qid": "AutoScholarQuery_train_16673"} +{"question": "What works utilized conjugate gradient to accelerate the classical EM algorithm for sparse Bayesian learning?", "answer": ["Covariance-Free Sparse Bayesian Learning", "High-Dimensional Sparse Bayesian Learning without Covariance Matrices"], "answer_arxiv_id": ["2105.10439", "2202.12808"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_16674"} +{"question": "Who hypothesized that the sharpness tends to grow but cannot exceed a stability criterion in stochastic gradient descent dynamics?", "answer": ["The break-even point on optimization trajectories of deep neural networks"], "answer_arxiv_id": ["2002.09572"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16675"} +{"question": "Which works used multimodal vision-large language model to directly answer questions about images?", "answer": ["X-IQE: eXplainable Image Quality Evaluation for Text-to-Image Generation with Visual Large Language Models"], "answer_arxiv_id": ["2305.10843"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_16676"} +{"question": "Any papers about fast biased approximation of the sliced Wasserstein distance?", "answer": ["Fast Approximation of the Sliced-Wasserstein Distance Using Concentration of Random Projections"], "answer_arxiv_id": ["2106.15427"], "source_meta": {"published_time": "20220927"}, "qid": "AutoScholarQuery_train_16677"} +{"question": "What works were made to propose black-box defense methods?", "answer": ["DeepSweep: An Evaluation Framework for Mitigating DNN Backdoor Attacks using Data Augmentation", "Februus: Input Purification Defense Against Trojan Attacks on Deep Neural Network Systems"], "answer_arxiv_id": ["2012.07006", "1908.03369"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_16678"} +{"question": "What studies have proposed to improve NN generalization in scientific problems by incorporating domain constraints into the ML framework?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations", "Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers"], "answer_arxiv_id": ["2010.08895", "2007.00016"], "source_meta": {"published_time": "20220718"}, "qid": "AutoScholarQuery_train_16679"} +{"question": "Could you provide me the studies where code prompts have been used for generative outputs such as knowledge graph construction?", "answer": ["CodeKGC: Code Language Model for Generative Knowledge Graph Construction"], "answer_arxiv_id": ["2304.09048"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_16680"} +{"question": "Are there any studies that have integrated the concept of assigning roles with complex tasks such as code generation?", "answer": ["ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate", "Enhancing Large Language Models in Coding Through Multi-Perspective\n Self-Consistency", "AutoAgents: A Framework for Automatic Agent Generation"], "answer_arxiv_id": ["2308.07201", "2309.17272", "2309.17288"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_16681"} +{"question": "Which works focus on breaking the symmetries between slots in slot-based object-centric representations by enforcing an order on the slots?", "answer": ["Attend, Infer, Repeat: Fast Scene Understanding with Generative Models", "MONet: Unsupervised Scene Decomposition and Representation", "Genesis: Generative Scene Inference and Sampling with Object-Centric Latent Representations"], "answer_arxiv_id": ["1603.08575", "1901.11390", "1907.13052"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16682"} +{"question": "Which research papers have dealt with the generation of proof sequences at once by Language Models (LMs)?", "answer": ["PRover: Proof Generation for Interpretable Reasoning over Rules", "Measuring Systematic Generalization in Neural Proof Generation with Transformers", "Explaining Answers with Entailment Trees", "Probabilistic Graph Reasoning for Natural Proof Generation"], "answer_arxiv_id": ["2010.02830", "2009.14786", "2104.08661", "2107.02418v1"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_16683"} +{"question": "What research proposed rationale generation methods for problem-solving tasks in neural sequence models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Large Language Models are Zero-Shot Reasoners", "Unsupervised Commonsense Question Answering with Self-Talk", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning"], "answer_arxiv_id": ["2201.11903", "2112.00114", "2205.11916", "2004.05483", "2203.11171", "2205.10625", "2205.09712"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16684"} +{"question": "What studies focus on Evidential Deep Learning (EDL)?", "answer": ["Evidential Deep Learning to Quantify Classification Uncertainty", "Deep Evidential Regression", "The Unreasonable Effectiveness of Deep Evidential Regression", "Uncertainty Aware Semi-Supervised Learning on Graph Data", "OpenTAL: Towards Open Set Temporal Action Localization", "Pitfalls of Epistemic Uncertainty Quantification through Loss Minimisation"], "answer_arxiv_id": ["1806.01768", "1910.02600", "2205.10060", "2010.12783", "2203.05114", "2203.06102"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_16685"} +{"question": "Could you give me some studies about traditional Eulerian methods that used hand-crafted filters?", "answer": ["Video Acceleration Magnification"], "answer_arxiv_id": ["1704.04186"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_16686"} +{"question": "Could you provide me with some studies which used generative probabilistic models to reconstruct full-body motion from three 6D trackers?", "answer": ["Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking\n Inputs with Diffusion Model"], "answer_arxiv_id": ["2304.08577"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_16687"} +{"question": "Which works have demonstrated significant success in diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2006.11239", "1503.03585", "1907.05600", "2011.13456"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_16688"} +{"question": "Which works propose using consensus motifs or simple structural patterns like k-mers as predictive features for TCR-peptide binding prediction?", "answer": ["TITAN: T Cell Receptor Specificity Prediction with Bimodal Attention Networks"], "answer_arxiv_id": ["2105.03323"], "source_meta": {"published_time": "20221015"}, "qid": "AutoScholarQuery_train_16689"} +{"question": "Which studies were motivated by trying to identify bugs in the implementations of differentially private data analysis algorithms?", "answer": ["Detecting Violations of Differential Privacy"], "answer_arxiv_id": ["1805.10277"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_16690"} +{"question": "What is the work that introduces Concept Activation Vectors?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with\n Concept Activation Vectors (TCAV)"], "answer_arxiv_id": ["1711.11279"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_16691"} +{"question": "What studies proposed an algorithm for decentralized linear bandits with a finite-capacity uplink channel and an infinite-capacity downlink channel?", "answer": ["Linear Stochastic Bandits over a Bit-Constrained Channel"], "answer_arxiv_id": ["2203.01198"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_16692"} +{"question": "What papers have discussed the application of Fourier Neural Operators for efficient PDE solver surrogates?", "answer": ["Fourier Neural Operator for Parametric Partial Differential Equations"], "answer_arxiv_id": ["2010.08895"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_16693"} +{"question": "What papers revision symbolic and neuro-symbolic techniques for code generation?", "answer": ["DeepCoder: Learning to Write Programs", "Learning to Represent Programs with Property Signatures", "DreamCoder: Growing generalizable, interpretable knowledge with wake-sleep Bayesian program learning", "RobustFill: Neural Program Learning under Noisy I/O"], "answer_arxiv_id": ["1611.01989", "2002.09030", "2006.08381", "1703.07469"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_16694"} +{"question": "What papers utilized Viterbi for solving alignment objectives in weakly-supervised action segmentation models?", "answer": ["Weakly Supervised Action Learning with RNN based Fine-to-coarse Modeling", "NeuralNetwork-Viterbi: A Framework for Weakly Supervised Video Learning", "Weakly Supervised Energy-Based Learning for Action Segmentation", "Weakly supervised learning of actions from transcripts", "A Hybrid RNN-HMM Approach for Weakly Supervised Temporal Action\n Segmentation"], "answer_arxiv_id": ["1703.08132", "1805.06875", "1909.13155", "1610.02237", "1906.01028"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_16695"} +{"question": "What research suggested adding Gumbel noise to the Sinkhorn operator to learn latent permutations?", "answer": ["Learning Latent Permutations with Gumbel-Sinkhorn Networks"], "answer_arxiv_id": ["1802.08665"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_16696"} +{"question": "What research leveraged ChatGPT to automatically summarize experience from annotated NLP datasets?", "answer": ["ExpeL: LLM Agents Are Experiential Learners", "Grimoire is All You Need for Enhancing Large Language Models"], "answer_arxiv_id": ["2308.10144", "2401.03385"], "source_meta": {"published_time": "20240712"}, "qid": "AutoScholarQuery_train_16697"} +{"question": "What paper discusses the sample complexity for robust learning?", "answer": ["Adversarially Robust Generalization Requires More Data"], "answer_arxiv_id": ["1804.11285"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_16698"} +{"question": "Which papers conducted research on inferring constraints based on discrete state-action space?", "answer": ["Learning Constraints from Demonstrations", "Maximum Likelihood Constraint Inference for Inverse Reinforcement Learning", "Maximum Likelihood Constraint Inference from Stochastic Demonstrations"], "answer_arxiv_id": ["1812.07084", "1909.05477", "2102.12554"], "source_meta": {"published_time": "20220620"}, "qid": "AutoScholarQuery_train_16699"} +{"question": "Could you provide me the references where depth priors have been used to supervise training of neural implicit representations?", "answer": ["Depth-supervised NeRF: Fewer Views and Faster Training for Free", "MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction", "NerfingMVS: Guided Optimization of Neural Radiance Fields for Indoor Multi-view Stereo", "Dense Depth Priors for Neural Radiance Fields from Sparse Input Views"], "answer_arxiv_id": ["2107.02791", "2206.00665", "2109.01129", "2112.03288"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_16700"} +{"question": "What research discusses the challenges of inducing neural collapse in imbalanced datasets due to minority collapse?", "answer": ["Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse\n in Imbalanced Training"], "answer_arxiv_id": ["2101.12699"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_16701"} +{"question": "What works have utilized scaling operations, like the proposed method in this paper, to achieve efficient fine-tuning of Vision Transformers?", "answer": ["AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition"], "answer_arxiv_id": ["2205.13535"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16702"} +{"question": "What papers studied the distribution shift robustness of ViTs?", "answer": ["Delving Deep into the Generalization of Vision Transformers under Distribution Shifts"], "answer_arxiv_id": ["2106.07617"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_16703"} +{"question": "What studies have proposed fair estimation criteria to decrease algorithmic biases?", "answer": ["Equality of Opportunity in Supervised Learning", "Algorithmic decision making and the cost of fairness"], "answer_arxiv_id": ["1610.02413", "1701.08230"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_16704"} +{"question": "Could you provide me some works about efficient processing of high-resolution images through sparse activations?", "answer": ["SparseViT: Revisiting Activation Sparsity for Efficient High-Resolution\n Vision Transformer"], "answer_arxiv_id": ["2303.17605"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_16705"} +{"question": "Are there any specific studies on Inductive Logic Programming?", "answer": ["Inductive Logic Programming At 30: A New Introduction"], "answer_arxiv_id": ["2008.07912"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16706"} +{"question": "Which study reports reward scaling as a power law with data frames when training a data efficient variant of MuZero?", "answer": ["Online and Offline Reinforcement Learning by Planning with a Learned Model"], "answer_arxiv_id": ["2104.06294"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16707"} +{"question": "Which research papers explored text-to-video retrieval?", "answer": ["VATEX: A Large-Scale, High-Quality Multilingual Dataset for\n Video-and-Language Research", "The \"something something\" video database for learning and evaluating\n visual common sense"], "answer_arxiv_id": ["1904.03493", "1706.04261"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_16708"} +{"question": "What research works have studied regularization inspired by dynamical optimal transport in the context of continuous normalizing flows?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_16709"} +{"question": "Which study explores the use of EBM to replace hand-crafted permutation-invariance loss functions?", "answer": ["Set Prediction without Imposing Structure as Conditional Density Estimation"], "answer_arxiv_id": ["2010.04109"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16710"} +{"question": "Which papers propose trajectory predicting models based on conditions of trajectory prototypes?", "answer": ["MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction"], "answer_arxiv_id": ["1910.05449"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_16711"} +{"question": "Could you provide me some works about Formal methods in model-based safe RL?", "answer": ["Reachability-based Trajectory Safeguard (RTS): A Safe and Fast Reinforcement Learning Safety Layer for Continuous Control", "Reinforcement Learning for Safety-Critical Control under Model Uncertainty, using Control Lyapunov Functions and Control Barrier Functions", "End-to-End Safe Reinforcement Learning through Barrier Functions for Safety-Critical Continuous Control Tasks", "Joint Differentiable Optimization and Verification for Certified Reinforcement Learning", "Model-based Constrained Reinforcement Learning using Generalized Control Barrier Function", "Learning Barrier Certificates: Towards Safe Reinforcement Learning with Zero Training-time Violations", "Safe Model-based Reinforcement Learning with Stability Guarantees", "Learning for Safety-Critical Control with Control Barrier Functions", "Reachability Constrained Reinforcement Learning"], "answer_arxiv_id": ["2011.08421", "2004.07584", "1903.08792", "2201.12243v2", "2103.01556", "2108.01846", "1705.08551", "1912.10099", "2205.07536"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16712"} +{"question": "What researches are about approaches that try to fine-tune LLMs to produce the final answer directly, keeping reasoning implicit?", "answer": ["Critical Thinking for Language Models", "Transformers as Soft Reasoners over Language"], "answer_arxiv_id": ["2009.07185", "2002.05867"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_16713"} +{"question": "What studies are related to AI agents in the context of foundation language models?", "answer": ["A Survey on Large Language Model based Autonomous Agents"], "answer_arxiv_id": ["2308.11432"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_16714"} +{"question": "Could you provide me with works related to adapting models on OOD data?", "answer": ["Domain Generalization with MixStyle", "Domain Generalization via Model-Agnostic Learning of Semantic Features", "Learning to Generalize: Meta-Learning for Domain Generalization", "Deep Domain-Adversarial Image Generation for Domain Generalisation", "MEMO: Test Time Robustness via Adaptation and Augmentation", "Generative Interventions for Causal Learning", "Discrete Representations Strengthen Vision Transformer Robustness", "Tent: Fully Test-Time Adaptation by Entropy Minimization"], "answer_arxiv_id": ["2104.02008", "1910.13580", "1710.03463", "2003.06054", "2110.09506", "2012.12265", "2111.10493v2", "2006.10726"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_16715"} +{"question": "Could you provide some studies that incorporate inter-object relationships and spatial context in 3D object recognition and segmentation?", "answer": ["Putting visual object recognition in context", "Structure Inference Net: Object Detection Using Scene-Level Context and\n Instance-Level Relationships", "Graph-Structured Representations for Visual Question Answering"], "answer_arxiv_id": ["1911.07349", "1807.00119", "1609.05600"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_16716"} +{"question": "Which works presented exemplar-based CIL methods for model rehearsal during updating?", "answer": ["Selective Experience Replay for Lifelong Learning"], "answer_arxiv_id": ["1802.10269"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_16717"} +{"question": "Can you list the studies that focus on the application of pseudolabeling in transductive zero-shot learning?", "answer": ["Transductive Zero-Shot Learning with a Self-training dictionary approach", "Hardness Sampling for Self-Training Based Transductive Zero-Shot Learning", "An Iterative Co-Training Transductive Framework for Zero Shot Learning"], "answer_arxiv_id": ["1703.08893", "2106.00264", "2203.16041"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_16718"} +{"question": "Could you give me some examples of works that utilized NeRF for novel view image synthesis and 3D surface reconstruction?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "INeRF: Inverting Neural Radiance Fields for Pose Estimation", "NeRF--: Neural Radiance Fields Without Known Camera Parameters", "NoPe-NeRF: Optimising Neural Radiance Field with No Pose Prior", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "iMAP: Implicit Mapping and Positioning in Real-Time", "Direct-PoseNet: Absolute Pose Regression with Photometric Consistency"], "answer_arxiv_id": ["2003.08934", "2012.05877", "2102.07064", "2212.07388", "2112.12130", "2103.12352", "2104.04073"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_16719"} +{"question": "Could you provide me with studies giving comprehensive literature surveys on semi-supervised learning?", "answer": ["An Overview of Deep Semi-Supervised Learning"], "answer_arxiv_id": ["2006.05278"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_16720"} +{"question": "What papers pertains to Shapley value-based methods?", "answer": ["A Unified Approach to Interpreting Model Predictions", "The Many Shapley Values for Model Explanation", "WeightedSHAP: analyzing and improving Shapley based feature attributions"], "answer_arxiv_id": ["1705.07874", "1908.08474", "2209.13429"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_16721"} +{"question": "Are there any studies that improve the 3D generation models through the use of textual codes or depth maps?", "answer": ["Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D\n Generation", "NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as\n General Image Priors", "RealFusion: 360{\\deg} Reconstruction of Any Object from a Single Image"], "answer_arxiv_id": ["2303.07937", "2212.03267", "2302.10663"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16722"} +{"question": "Which works discuss the objective optimization function in age estimation?", "answer": ["PML: Progressive Margin Loss for Long-tailed Age Classification"], "answer_arxiv_id": ["2103.02140"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16723"} +{"question": "Which studies have derived the non-asymptotic rate of the actor-critic under i.i.d. assumptions?", "answer": ["On the Sample Complexity of Actor-Critic Method for Reinforcement Learning with Function Approximation", "Neural Policy Gradient Methods: Global Optimality and Rates of Convergence"], "answer_arxiv_id": ["1910.08412", "1909.01150"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_16724"} +{"question": "Could you mention some studies that applied GAN inversion to image restoration?", "answer": ["Towards Real-World Blind Face Restoration with Generative Facial Prior", "PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of\n Generative Models", "GAN Prior Embedded Network for Blind Face Restoration in the Wild"], "answer_arxiv_id": ["2101.04061", "2003.03808", "2105.06070"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_16725"} +{"question": "Which works provided insights about the statistical limits of deep learning and Fourier basis techniques for solving elliptic PDEs?", "answer": ["Machine Learning For Elliptic PDEs: Fast Rate Generalization Bound, Neural Scaling Law and Minimax Optimality"], "answer_arxiv_id": ["2110.06897"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_16726"} +{"question": "Could you tell me which papers proposed to progressively train the generator and the discriminator in GANs to stabilize the training process?", "answer": ["Progressive Growing of GANs for Improved Quality, Stability, and Variation"], "answer_arxiv_id": ["1710.10196"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_16727"} +{"question": "Could you mention some studies that focus on 2D multimodal datasets training?", "answer": ["LAION-5B: An open large-scale dataset for training next generation image-text models", "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs"], "answer_arxiv_id": ["2210.08402", "2111.02114"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_16728"} +{"question": "What work is closely related to the study in this paper that tried to handle simultaneity via a single model?", "answer": ["Environment-agnostic Multitask Learning for Natural Language Grounded\n Navigation"], "answer_arxiv_id": ["2003.00443"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_16729"} +{"question": "Which research works study different metrics for evaluating feature attributions?", "answer": ["Sanity Checks for Saliency Maps", "A Benchmark for Interpretability Methods in Deep Neural Networks", "ERASER: A Benchmark to Evaluate Rationalized NLP Models", "“Will You Find These Shortcuts?” A Protocol for Evaluating the Faithfulness of Input Salience Methods for Text Classification", "A Consistent and Efficient Evaluation Strategy for Attribution Methods", "The Solvability of Interpretability Evaluation Metrics", "From Anecdotal Evidence to Quantitative Evaluation Methods: A Systematic Review on Evaluating Explainable AI", "Post hoc Explanations may be Ineffective for Detecting Unknown Spurious Correlation", "Do Feature Attribution Methods Correctly Attribute Features?"], "answer_arxiv_id": ["1810.03292v3", "1806.10758", "1911.03429v2", "2111.07367", "2202.00449", "2205.08696", "2201.08164", "2212.04629", "2104.14403"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_16730"} +{"question": "Which works discuss the use of the R-Precision metric for measuring text recall scores from a reference dataset of captions?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks"], "answer_arxiv_id": ["1711.10485"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_16731"} +{"question": "Which studies have shown improvements using an explicit causally-motivated invariance loss in conjunction with the contrastive objective?", "answer": ["Representation Learning via Invariant Causal Mechanisms", "Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?"], "answer_arxiv_id": ["2010.07922", "2201.05119"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_16732"} +{"question": "What are some classical approaches to learning visual embeddings that align with pre-defined text embeddings?", "answer": ["Open Vocabulary Scene Parsing", "Zero-Shot Semantic Segmentation"], "answer_arxiv_id": ["1703.08769", "1906.00817"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_16733"} +{"question": "Could you provide some articles that utilize penalization on the estimated model in model-based offline RL?", "answer": ["MOPO: Model-based Offline Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning", "Offline Reinforcement Learning from Images with Latent Space Models"], "answer_arxiv_id": ["2005.13239", "2005.05951", "2012.11547"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_16734"} +{"question": "Can you give examples of researches that use logits interpolation from separate forward passes to incorporate retrieval information into LMs?", "answer": ["REPLUG: Retrieval-Augmented Black-Box Language Models", "RA-DIT: Retrieval-Augmented Dual Instruction Tuning"], "answer_arxiv_id": ["2301.12652", "2310.01352"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_16735"} +{"question": "Any studies discussing the various properties of information bottlenecks?", "answer": ["Emergence of Invariance and Disentanglement in Deep Representations"], "answer_arxiv_id": ["1706.01350"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16736"} +{"question": "What study accelerates the generative process of diffusion models by discarding Markovian assumption?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_16737"} +{"question": "Which works are pioneer of point-based methods in 3D point cloud analysis that adopt multi-layer perceptron?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space", "PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies"], "answer_arxiv_id": ["1612.00593", "1706.02413", "2206.04670"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_16738"} +{"question": "Can you mention some papers that criticized the CKA method by revealing its insensitivity and by sharing their observations about its usage?", "answer": ["On the Origins of the Block Structure Phenomenon in Neural Network Representations", "Generalized Shape Metrics on Neural Representations"], "answer_arxiv_id": ["2202.07184", "2110.14739"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_16739"} +{"question": "Do we have any works on visual object tracking in video localization tasks?", "answer": ["Transformer Tracking", "MixFormer: End-to-End Tracking with Iterative Mixed Attention"], "answer_arxiv_id": ["2103.15436", "2203.11082"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_16740"} +{"question": "What works leverage the idea of contrastive learning to connect images with their corresponding captions in vision-language pre-training?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16741"} +{"question": "Which research papers have explored ALADIN for style attribution?", "answer": ["EKILA: Synthetic Media Provenance and Attribution for Generative Art", "Evaluating Data Attribution for Text-to-Image Models", "ALADIN: All Layer Adaptive Instance Normalization for Fine-grained Style\n Similarity"], "answer_arxiv_id": ["2304.04639", "2306.09345", "2103.09776"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_16742"} +{"question": "What research uses methods like HMMR, VIBE and MEVA for video-based 3D HPS estimation", "answer": ["Learning 3D Human Dynamics from Video", "VIBE: Video Inference for Human Body Pose and Shape Estimation", "3D Human Motion Estimation via Motion Compression and Refinement"], "answer_arxiv_id": ["1812.01601", "1912.05656", "2008.03789"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_16743"} +{"question": "What work guides the Natural Evolution Strategies (NES) framework?", "answer": ["Evolution Strategies as a Scalable Alternative to Reinforcement Learning"], "answer_arxiv_id": ["1703.03864v2"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_16744"} +{"question": "Could you provide me some studies that decode shapes as raw point clouds or implicit representations?", "answer": ["Learning Representations and Generative Models for 3D Point Clouds", "PointDAN: A Multi-Scale 3D Domain Adaption Network for Point Cloud Representation", "3D Point Capsule Networks", "Implicit Geometric Regularization for Learning Shapes"], "answer_arxiv_id": ["1707.02392", "1911.02744", "1812.10775", "2002.10099"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_16745"} +{"question": "Which research works implemented regularization techniques for calibrating DNNs?", "answer": ["On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networks", "AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning", "Uncertainty Quantification and Deep Ensembles", "BatchEnsemble: An alternative approach to Efficient Ensemble and Lifelong Learning", "When Does Label Smoothing Help?", "On Calibration of Modern Neural Networks"], "answer_arxiv_id": ["1905.11001", "1912.02781", "1612.01474", "2003.07329", "2007.08792", "2002.06715", "1906.02629", "1706.04599"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_16746"} +{"question": "Who offers an accessible overview of distribution-free uncertainty quantification?", "answer": ["A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification"], "answer_arxiv_id": ["2107.07511"], "source_meta": {"published_time": "20221227"}, "qid": "AutoScholarQuery_train_16747"} +{"question": "What is the reference for the used algorithm in RLHF method?", "answer": ["Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1707.06347"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_16748"} +{"question": "Could you mention some studies that assume interventional data and linear mixing functions in a non-temporal setting?", "answer": ["Linear Causal Disentanglement via Interventions", "Score-based Causal Representation Learning with Interventions"], "answer_arxiv_id": ["2211.16467v3", "2301.08230"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_16749"} +{"question": "What research discussions have been done about gradient coherence in the training dynamics of mini-batch SGD in centralized learning?", "answer": ["Coherent Gradients: An Approach to Understanding Generalization in Gradient Descent-based Optimization", "Weak and Strong Gradient Directions: Explaining Memorization, Generalization, and Hardness of Examples at Scale", "Making Coherence Out of Nothing At All: Measuring the Evolution of Gradient Alignment", "Stiffness: A New Perspective on Generalization in Neural Networks"], "answer_arxiv_id": ["2002.10657", "2003.07422", "2008.01217", "1901.09491"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_16750"} +{"question": "Any works provided empirical evidence of ‘hidden progress’ during training?", "answer": ["Hidden Progress in Deep Learning: SGD Learns Parities Near the Computational Limit"], "answer_arxiv_id": ["2207.08799"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_16751"} +{"question": "Which papers discussed an evaluaton-based search algorithm for neural architecture search?", "answer": ["Large-Scale Evolution of Image Classifiers"], "answer_arxiv_id": ["1703.01041"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_16752"} +{"question": "Which works focused on the infinite-horizon linear quadratic regulator (LQR) study?", "answer": ["Naive Exploration is Optimal for Online LQR", "Black-Box Control for Linear Dynamical Systems", "Reinforcement Learning with Fast Stabilization in Linear Dynamical Systems"], "answer_arxiv_id": ["2001.09576", "2007.06650", "2007.12291v2"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_16753"} +{"question": "In which papers do researchers learn a joint embedding model between subgoal language commands and agent states?", "answer": ["A Narration-based Reward Shaping Approach using Grounded Natural Language Commands"], "answer_arxiv_id": ["1911.00497"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_16754"} +{"question": "Which papers aimed to search for the optimal Transformer-based architecture?", "answer": ["AutoFormer: Searching Transformers for Visual Recognition", "ViTAS: Vision Transformer Architecture Search", "Training-free Transformer Architecture Search"], "answer_arxiv_id": ["2107.00651", "2106.13700", "2203.12217"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_16755"} +{"question": "Any research demonstrating empirical DRO with the χ2-divergence has the effect of a variance regularization?", "answer": ["Variance-based regularization with convex objectives"], "answer_arxiv_id": ["1610.02581"], "source_meta": {"published_time": "20230218"}, "qid": "AutoScholarQuery_train_16756"} +{"question": "What is the reference for the paper that put prominence on the geometric structure of the last-layer feature and classifier in a well-trained model- a concept referred to as Neural Collapse?", "answer": ["Prevalence of Neural Collapse during the terminal phase of deep learning training"], "answer_arxiv_id": ["2008.08186"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_16757"} +{"question": "Could you list some papers that discuss about the settings under which benign overfitting occurs?", "answer": ["Uniform convergence may be unable to explain generalization in deep learning", "Benign Overfitting in Linear Regression", "In Defense of Uniform Convergence: Generalization via derandomization with an application to interpolating predictors", "Distributional Generalization: A New Kind of Generalization", "Exact Gap between Generalization Error and Uniform Convergence in Random Feature Models", "Uniform Convergence of Interpolators: Gaussian Width, Norm Bounds and Benign Overfitting", "Failures of model-dependent generalization bounds for least-norm interpolation", "Classification vs regression in overparameterized regimes: Does the loss function matter?", "Uniform Convergence, Adversarial Spheres and a Simple Remedy", "Optimistic Rates: A Unifying Theory for Interpolation Learning and Regularization in Linear Regression", "The Implicit Bias of Benign Overfitting"], "answer_arxiv_id": ["1902.04742", "1906.11300", "1912.04265", "2009.08092", "2103.04554", "2106.09276", "2010.08479", "2005.08054", "2105.03491", "2112.04470", "2201.11489"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_16758"} +{"question": "Which research demonstrated the success of MAE as a self-supervised learning paradigm in image domain?", "answer": ["Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2111.06377"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_16759"} +{"question": "What works investigated the use of Graph Neural Networks for object detection?", "answer": ["Pushing the Limits of Asynchronous Graph-based Object Detection with\n Event Cameras"], "answer_arxiv_id": ["2211.12324"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_16760"} +{"question": "What studies have explored the vulnerability of GNN models to adversarial attacks on graph structure?", "answer": ["Adversarial Attack on Graph Structured Data"], "answer_arxiv_id": ["1806.02371"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_16761"} +{"question": "Which work refers to similar questions they have previously encountered to judge retrieval necessity?", "answer": ["Self-Knowledge Guided Retrieval Augmentation for Large Language Models"], "answer_arxiv_id": ["2310.05002"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_16762"} +{"question": "What research has been conducted to improve GPT responses using human feedback from online data?", "answer": ["WebGPT: Browser-assisted question-answering with human feedback"], "answer_arxiv_id": ["2112.09332"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_16763"} +{"question": "What papers propose use of human template mesh in 3D human generation?", "answer": ["AvatarGen: A 3D Generative Model for Animatable Human Avatars", "AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "CLIP-Actor: Text-Driven Recommendation and Stylization for Animating\n Human Meshes", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "Chupa: Carving 3D Clothed Humans from Skinned Shape Priors using 2D\n Diffusion Probabilistic Models"], "answer_arxiv_id": ["2211.14589", "2205.08535", "2206.04382", "2304.00916", "2305.11870"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_16764"} +{"question": "Which works are related to the deep learning methods for optical flow estimation?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks", "LiteFlowNet: A Lightweight Convolutional Neural Network for Optical Flow\n Estimation", "PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume"], "answer_arxiv_id": ["1504.06852", "1612.01925", "1805.07036", "1709.02371"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_16765"} +{"question": "Can you provide resources about using skip connection to build multi-head self-attention models?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16766"} +{"question": "In the domain of compositional video synthesis, what work integrates multiple conditioning signals within a unified framework?", "answer": ["VideoComposer: Compositional Video Synthesis with Motion Controllability"], "answer_arxiv_id": ["2306.02018"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_16767"} +{"question": "Which papers discuss image-text contrastive learning as a possible loss function while training VLMs?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal\n Skip-connections", "Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual\n Concepts", "FLAVA: A Foundational Language And Vision Alignment Model"], "answer_arxiv_id": ["2103.00020", "2107.07651", "2201.12086", "2205.12005", "2111.08276", "2112.04482"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16768"} +{"question": "Could you give me examples of research that used importance weighting type approaches for the selection of examples within an SGD minibatch?", "answer": ["Not All Samples Are Created Equal: Deep Learning with Importance Sampling"], "answer_arxiv_id": ["1803.00942v3"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_16769"} +{"question": "Which paper proposed the discretization of latent variables for the improvement of generative modeling capacity?", "answer": ["Generating Diverse High-Fidelity Images with VQ-VAE-2"], "answer_arxiv_id": ["1906.00446"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_16770"} +{"question": "What works focused on modifying the diffusion sampling process in diffusion-based human motion composition?", "answer": ["DiffCollage: Parallel Generation of Large Content with Diffusion Models", "MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation", "Human Motion Diffusion as a Generative Prior"], "answer_arxiv_id": ["2303.17076", "2302.08113", "2303.01418"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_16771"} +{"question": "What works store episodic information in the time-varying neural activities of a recurrent network for meta-learning?", "answer": ["Learning to reinforcement learn"], "answer_arxiv_id": ["1611.05763"], "source_meta": {"published_time": "20211216"}, "qid": "AutoScholarQuery_train_16772"} +{"question": "Which papers studied the application of reinforcement learning to image classification?", "answer": ["Recurrent Models of Visual Attention", "Gaussian RAM: Lightweight Image Classification via Stochastic\n Retina-Inspired Glimpse and Reinforcement Learning"], "answer_arxiv_id": ["1406.6247", "2011.06190"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_16773"} +{"question": "Which studies utilized the Learning To Prompt (L2P) and DualPrompt methods in the prompting of transformer networks?", "answer": ["Improved Few-Shot Visual Classification", "Dark Experience for General Continual Learning: a Strong, Simple Baseline"], "answer_arxiv_id": ["1912.03432", "2004.07211"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_16774"} +{"question": "Can you mention any work that applies LSTMs to encode ancestors and descendants?", "answer": ["TaxoEnrich: Self-Supervised Taxonomy Completion via Structure-Semantic\n Representations"], "answer_arxiv_id": ["2202.04887"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_16775"} +{"question": "Who improved the approximation algorithm for maximizing a non-monotone suodular function subject to a cardinality constraint to a 3.6+ε approximation guarantee?", "answer": ["Optimal Streaming Algorithms for Submodular Maximization with Cardinality Constraints"], "answer_arxiv_id": ["1911.12959v3"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_16776"} +{"question": "Which works provided an average-case complexity analysis and implemented constraints on the data-generation process to prevent attacks?", "answer": ["Adversarially Robust Learning Could Leverage Computational Hardness"], "answer_arxiv_id": ["1905.11564"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_16777"} +{"question": "Which paper introduced GLNs and allowed a study of its convergence and generalization in the lazy kernel limit?", "answer": ["Decoupling Gating from Linearity"], "answer_arxiv_id": ["1906.05032"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_16778"} +{"question": "Which studies have introduced metrics to measure the alignment between the generated image and the text prompt?", "answer": ["Mutual Information Divergence: A Unified Metric for Multimodal\n Generative Models", "Improved Precision and Recall Metric for Assessing Generative Models", "Assessing Generative Models via Precision and Recall"], "answer_arxiv_id": ["2205.13445", "1904.06991", "1806.00035"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_16779"} +{"question": "Which study used a similar architecture to TubeDETR and extended it for object tracking?", "answer": ["Learning Spatio-Temporal Transformer for Visual Tracking"], "answer_arxiv_id": ["2103.17154"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_16780"} +{"question": "Which studies proposed methods to improve the speed and quality of sampling in diffusion-based models?", "answer": ["Denoising Diffusion Implicit Models", "Progressive Distillation for Fast Sampling of Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Autoregressive Diffusion Models", "WaveGrad: Estimating Gradients for Waveform Generation", "On Fast Sampling of Diffusion Probabilistic Models", "Tackling the Generative Learning Trilemma with Denoising Diffusion GANs", "Elucidating the Design Space of Diffusion-Based Generative Models", "Poisson Flow Generative Models"], "answer_arxiv_id": ["2010.02502", "2202.00512", "2112.10752", "2110.02037", "2009.00713v2", "2106.00132", "2112.07804", "2206.00364", "2209.11178"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_16781"} +{"question": "Could you provide me some literature in which supervised contrastive learning is discussed?", "answer": ["Supervised Contrastive Learning", "Supervised Contrastive Learning for Pre-trained Language Model\n Fine-tuning"], "answer_arxiv_id": ["2004.11362", "2011.01403"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_16782"} +{"question": "What studies collected diverse, large-value databases for text-to-SQL?", "answer": ["EHRSQL: A Practical Text-to-SQL Benchmark for Electronic Health Records", "Text-to-SQL in the Wild: A Naturally-Occurring Dataset Based on Stack Exchange Data", "Text-to-SQL Generation for Question Answering on Electronic Medical Records"], "answer_arxiv_id": ["2301.07695", "2106.05006", "1908.01839"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_16783"} +{"question": "What works propose ways to capture relations between multi-view data or keypoints?", "answer": ["Learning Actionable Representations from Visual Observations", "Multi-View Dreaming: Multi-View World Model with Contrastive Learning", "Time-Contrastive Networks: Self-Supervised Learning from Video", "Keypoints into the Future: Self-Supervised Correspondence in Model-Based Reinforcement Learning"], "answer_arxiv_id": ["1808.00928", "2203.11024", "1704.06888", "2009.05085"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_16784"} +{"question": "Which research paper establishes RGCN as a prototypical example of early Graph Neural Networks for knowledge graphs?", "answer": ["Modeling Relational Data with Graph Convolutional Networks"], "answer_arxiv_id": ["1703.06103v4"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_16785"} +{"question": "Could you provide me some works on image scaling attacks?", "answer": ["Rethinking Image-Scaling Attacks: The Interplay Between Vulnerabilities in Machine Learning Systems"], "answer_arxiv_id": ["2104.08690v3"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_16786"} +{"question": "What studies notified the lack of theoretical work characterizing when and why distillation is effective for compression?", "answer": ["Knowledge Distillation: A Survey"], "answer_arxiv_id": ["2006.05525"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_16787"} +{"question": "Which studies initiated the focus on computationally-efficient robust estimation of multivariate distributions?", "answer": ["Robust Estimators in High Dimensions without the Computational Intractability", "Agnostic Estimation of Mean and Covariance"], "answer_arxiv_id": ["1604.06443", "1604.06968"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_16788"} +{"question": "Which work first proposed the concept of test-time adaptation (TTA) under transductive learning setting?", "answer": ["Tent: Fully Test-Time Adaptation by Entropy Minimization"], "answer_arxiv_id": ["2006.10726"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_16789"} +{"question": "Which studies conducted in a semi-supervised setting for temporal action segmentation?", "answer": ["Iterative Contrast-Classify For Semi-supervised Temporal Action\n Segmentation", "Leveraging Action Affinity and Continuity for Semi-supervised Temporal\n Action Segmentation"], "answer_arxiv_id": ["2112.01402", "2207.08653"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_16790"} +{"question": "Can you provide some works on regularization-based methods for incremental learning?", "answer": ["Less-forgetting Learning in Deep Neural Networks", "Learning without Forgetting", "Overcoming catastrophic forgetting in neural networks", "Orthogonal Gradient Descent for Continual Learning", "Gradient Projection Memory for Continual Learning"], "answer_arxiv_id": ["1607.00122", "1606.09282", "1612.00796", "1910.07104v1", "2103.09762"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_16791"} +{"question": "Present some studies that train neural networks with the help of concepts.", "answer": ["Concept Bottleneck Model with Additional Unsupervised Concepts", "C-SENN: Contrastive Self-Explaining Neural Network"], "answer_arxiv_id": ["2202.01459", "2206.09575"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_16792"} +{"question": "Which paper presented the use of a GAN to generate radar received signals and derived radar images through signal processing?", "answer": ["Generation of Realistic Synthetic Raw Radar Data for Automated Driving\n Applications using Generative Adversarial Networks"], "answer_arxiv_id": ["2308.02632"], "source_meta": {"published_time": "20240428"}, "qid": "AutoScholarQuery_train_16793"} +{"question": "Which work introduced Sampled MuZero to address arbitrarily complex action spaces?", "answer": ["Learning and Planning in Complex Action Spaces"], "answer_arxiv_id": ["2104.06303"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_16794"} +{"question": "In which work is the T5 framework introduced?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1910.10683"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_16795"} +{"question": "What studies dealt with continual learning in the context of online data poisoning?", "answer": ["Continual Learning in the Teacher-Student Setup: Impact of Task Similarity"], "answer_arxiv_id": ["2107.04384"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_16796"} +{"question": "Which papers have used decoder-only architectures for the infilling task?", "answer": ["Insertion Transformer: Flexible Sequence Generation via Insertion Operations", "GLM: General Language Model Pretraining with Autoregressive Blank\n Infilling", "CM3: A Causal Masked Multimodal Model of the Internet", "InCoder: A Generative Model for Code Infilling and Synthesis"], "answer_arxiv_id": ["1902.03249v1", "2103.10360", "2201.07520", "2204.05999"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_16797"} +{"question": "Which studies explored 3D object detection with tasks like depth estimation or semantic segmentation?", "answer": ["PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation"], "answer_arxiv_id": ["1711.10871"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_16798"} +{"question": "What studies explore neural processes in meta-learning?", "answer": ["Fast and Flexible Multi-Task Classification Using Conditional Neural Adaptive Processes", "Neural Processes", "Conditional Neural Processes", "Attentive Neural Processes", "Doubly Stochastic Variational Inference for Neural Processes with Hierarchical Latent Variables", "Practical Conditional Neural Processes Via Tractable Dependent Predictions", "Contrastive Conditional Neural Processes", "Neural Processes with Stochastic Attention: Paying more attention to the context dataset"], "answer_arxiv_id": ["1906.07697", "1807.01622", "1807.01613", "1901.05761", "2008.09469", "2203.08775", "2203.03978", "2204.05449"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_16799"} +{"question": "What works presented feature extraction techniques that focus on extracting features suitable for objects of varying orientations?", "answer": ["ReDet: A Rotation-equivariant Detector for Aerial Object Detection", "Adaptive Rotated Convolution for Rotated Object Detection"], "answer_arxiv_id": ["2103.07733", "2303.07820"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_16800"} +{"question": "What are some research papers that incorporated deep learning techniques in non-rigid shape registration?", "answer": ["DeepDeform: Learning Non-rigid RGB-D Reconstruction with Semi-supervised Data", "Neural Non-Rigid Tracking", "Multiway Non-rigid Point Cloud Registration via Learned Functional Map Synchronization"], "answer_arxiv_id": ["1912.04302", "2006.13240", "2111.12878"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_16801"} +{"question": "What research has been conducted about interactive image editing using region masks as user input?", "answer": ["RePaint: Inpainting using Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2201.09865"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_16802"} +{"question": "What works discussed different variants of knowledge distillation and soft labels in image recognition training?", "answer": ["Distilling the Knowledge in a Neural Network", "TinyBERT: Distilling BERT for Natural Language Understanding", "DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter", "Training data-efficient image transformers & distillation through attention", "Contrastive Representation Distillation", "Re-labeling ImageNet: from Single to Multi-Labels, from Global to Localized Labels", "A Fast Knowledge Distillation Framework for Visual Recognition"], "answer_arxiv_id": ["1503.02531", "1909.10351", "1910.01108", "2012.12877", "1910.10699", "2101.05022", "2112.01528"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_16803"} +{"question": "Which works extend Deep RL algorithms for multi-objective settings?", "answer": ["A Generalized Algorithm for Multi-Objective Reinforcement Learning and Policy Adaptation", "A Distributional View on Multi-Objective Policy Optimization", "Pareto Set Learning for Neural Multi-objective Combinatorial Optimization"], "answer_arxiv_id": ["1908.08342", "2005.07513", "2203.15386"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_16804"} +{"question": "What papers discuss the convergence of various gradient-based algorithms to the globally optimal policy by using the gradient domination condition?", "answer": ["On the Global Convergence Rates of Softmax Policy Gradient Methods", "A Decentralized Policy Gradient Approach to Multi-Task Reinforcement Learning", "On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2005.06392", "2006.04338", "2201.07443"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_16805"} +{"question": "Which studies focus on meta-learning methods over graphs for few-shot learning?", "answer": ["Task-Adaptive Few-shot Node Classification", "Meta Relational Learning for Few-Shot Link Prediction in Knowledge Graphs", "Few-Shot Knowledge Graph Completion"], "answer_arxiv_id": ["2206.11972", "1909.01515", "1911.11298"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_16806"} +{"question": "What research observed that masked background regions affect the recognition ability of CLIP due to the distribution difference?", "answer": ["Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP"], "answer_arxiv_id": ["2210.04150"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16807"} +{"question": "Which research papers explore the approach of extracting high-resolution meshes from a learned radiance field?", "answer": ["MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures", "Re-ReND: Real-time Rendering of NeRFs across Devices", "Learning Neural Duplex Radiance Fields for Real-Time View Synthesis"], "answer_arxiv_id": ["2208.00277", "2303.08717", "2304.10537"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_16808"} +{"question": "Which studies introduced Federated Learning?", "answer": ["Federated Optimization: Distributed Optimization Beyond the Datacenter", "Agnostic Federated Learning"], "answer_arxiv_id": ["1511.03575", "1902.00146"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_16809"} +{"question": "Could you list some works dealing with the stochastic oracle model with bounded variance?", "answer": ["On the Convergence of SGD with Biased Gradients", "Better Theory for SGD in the Nonconvex World"], "answer_arxiv_id": ["2008.00051", "2002.03329"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_16810"} +{"question": "Which paper showed that LLMs perform poorly in comparison to fine-tuned smaller LMs when given few-shot examples?", "answer": ["Generating Natural Language Proofs with Verifier-Guided Search"], "answer_arxiv_id": ["2205.12443"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_16811"} +{"question": "What is the researching work about Entity-Message Passing (EMP) in the context of MARL?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_16812"} +{"question": "Which studies have applied CLIP for text-free image generation?", "answer": ["Lafite : Towards Language-Free Training for Text-to-Image Generation", "CLIP-GEN: Language-Free Training of a Text-to-Image Generator with CLIP"], "answer_arxiv_id": ["2111.13792", "2203.00386"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_16813"} +{"question": "What are some publications that explore learning to follow instructions?", "answer": ["Improving Passage Retrieval with Zero-Shot Question Generation"], "answer_arxiv_id": ["2204.07496"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_16814"} +{"question": "Which studies introduced the decoupled Graph Convolutional Network (decoupled GCN)?", "answer": ["On the Equivalence of Decoupled Graph Convolution Network and Label Propagation"], "answer_arxiv_id": ["2010.12408"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_16815"} +{"question": "Which work proposed a novel approach to model dynamic scenes at a large scale using a scene flow field representation?", "answer": ["Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes"], "answer_arxiv_id": ["2011.13084"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_16816"} +{"question": "Which studies have successfully applied diffusion models in various domains?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations", "Denoising Diffusion Probabilistic Models", "Diffusion Models: A Comprehensive Survey of Methods and Applications"], "answer_arxiv_id": ["2011.13456", "2006.11239", "2209.00796"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_16817"} +{"question": "Can you point out studies that propagate matrices as messages instead of vectors in GNN for achieving a permutation equivariant unique identification scheme?", "answer": ["Building powerful and equivariant graph neural networks with structural message-passing"], "answer_arxiv_id": ["2006.15107"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_16818"} +{"question": "What works have made contributions to established correspondence for various vision tasks such as image matching and retrieval, and representation learning?", "answer": ["LoFTR: Detector-Free Local Feature Matching with Transformers", "SuperGlue: Learning Feature Matching with Graph Neural Networks", "Unsupervised Learning of Visual Representations using Videos", "Matching Networks for One Shot Learning"], "answer_arxiv_id": ["2104.00680", "1911.11763", "1505.00687", "1606.04080"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_16819"} +{"question": "Are there any works that contributed to the incorporation of physics-based deformations into the NeRF framework?", "answer": ["NeRF-Editing: Geometry Editing of Neural Radiance Fields", "NeuPhysics: Editable Neural Geometry and Physics from Monocular Videos", "Neural Impostor: Editing Neural Radiance Fields with Explicit Shape\n Manipulation"], "answer_arxiv_id": ["2205.04978", "2210.12352", "2310.05391"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_16820"} +{"question": "Which works generalize symmetries on graphs to sets and point clouds in the study of equivariant networks?", "answer": ["Deep Sets", "Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks", "e3nn: Euclidean Neural Networks", "Scalars are universal: Equivariant machine learning, structured like classical physics", "A simple and universal rotation equivariant point-cloud network"], "answer_arxiv_id": ["1703.06114", "1802.08219", "2006.10503", "2207.09453", "2106.06610", "2203.01216"], "source_meta": {"published_time": "20230821"}, "qid": "AutoScholarQuery_train_16821"} +{"question": "Are there any works on unsupervised continual learning with a generative model adopting task-specific inference?", "answer": ["Continual Unsupervised Representation Learning"], "answer_arxiv_id": ["1910.14481"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_16822"} +{"question": "Which studies involve uncertainty-based methods for stereo matching?", "answer": ["Deep Stereo using Adaptive Thin Volume Representation with Uncertainty\n Awareness"], "answer_arxiv_id": ["1911.12012"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16823"} +{"question": "Any works about applying large language models to open-ended generation tasks like dialog?", "answer": ["Conversational AI: The Science Behind the Alexa Prize", "LaMDA: Language Models for Dialog Applications"], "answer_arxiv_id": ["1801.03604v1", "2201.08239"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_16824"} +{"question": "Which works produced robust word embeddings through metric learning with adversarial training?", "answer": ["Robust Textual Embedding against Word-level Adversarial Attacks"], "answer_arxiv_id": ["2202.13817"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_16825"} +{"question": "What works have modified the cross-attention in the original transformer to leverage the information from the encoder, allowing a fine spatial recovery in the decoder?", "answer": ["U-Net Transformer: Self and Cross Attention for Medical Image\n Segmentation"], "answer_arxiv_id": ["2103.06104"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_16826"} +{"question": "Which study proposed a topology-agnostic autoencoder framework for facial expression retargeting across different mesh topologies?", "answer": ["Neural Face Rigging for Animating and Retargeting Facial Meshes in the\n Wild"], "answer_arxiv_id": ["2305.08296"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_16827"} +{"question": "Which study established the technique where secrets are embedded into a cover image or extracted from a stego-image using an end-to-end learnable DNN?", "answer": ["Generating Steganographic Images via Adversarial Training"], "answer_arxiv_id": ["1703.00371"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16828"} +{"question": "What studies addressed applications of subgraph methods for link prediction in recommender systems?", "answer": ["Inductive Matrix Completion Based on Graph Neural Networks"], "answer_arxiv_id": ["1904.12058"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16829"} +{"question": "Can you refer to works on using self-consistency and chain of thought prompting for selecting best generations for supervised fine-tuning?", "answer": ["Large Language Models Can Self-Improve", "Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2210.11610", "2203.11171"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_16830"} +{"question": "Which studies verify the positive role of entropy regularization in game-theoretic settings?", "answer": ["Fast Policy Extragradient Methods for Competitive Games with Entropy Regularization", "Exploration-Exploitation in Multi-Agent Competition: Convergence with Bounded Rationality", "Independent Natural Policy Gradient Methods for Potential Games: Finite-time Global Convergence with Entropy Regularization"], "answer_arxiv_id": ["2105.15186", "2106.12928", "2204.05466"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_16831"} +{"question": "What study presented a generic biologically plausible neural network framework based on weighted similarity matching (WSM) and maximization of the output correlation determinant criterion?", "answer": ["Biologically-Plausible Determinant Maximization Neural Networks for Blind Separation of Correlated Sources"], "answer_arxiv_id": ["2209.12894"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_16832"} +{"question": "Can you identify any studies that look into textual backdoor attacks in the NLP tasks?", "answer": ["BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain", "Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning", "A Tale of Evil Twins: Adversarial Inputs versus Poisoned Models", "Weight Poisoning Attacks on Pre-trained Models", "Be Careful about Poisoned Word Embeddings: Exploring the Vulnerability of the Embedding Layers in NLP Models", "Trojaning Language Models for Fun and Profit", "A backdoor attack against LSTM-based text classification systems", "Hidden Killer: Invisible Textual Backdoor Attacks with Syntactic Trigger"], "answer_arxiv_id": ["1708.06733", "1712.05526", "1911.01559", "2004.06660", "2103.15543", "2008.00312", "1905.12457", "2105.12400"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_16833"} +{"question": "Please name some research papers that discussed the use of multi-modal large language models.", "answer": ["Sparks of Artificial General Intelligence: Early experiments with GPT-4", "Language Models are Few-Shot Learners", "GPT-4 Technical Report", "LLaMA: Open and Efficient Foundation Language Models", "MIMIC-IT: Multi-Modal In-Context Instruction Tuning", "Planting a SEED of Vision in Large Language Model"], "answer_arxiv_id": ["2303.12712v5", "2005.14165", "2303.08774", "2302.13971", "2306.05425", "2307.08041"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_16834"} +{"question": "What research works proposed the use of T5 in LLMs?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1910.10683"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_16835"} +{"question": "Which paper created GLUCOSE to model implicit commonsense knowledge in narrative contexts?", "answer": ["GLUCOSE: GeneraLized and COntextualized Story Explanations"], "answer_arxiv_id": ["2009.07758"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_16836"} +{"question": "Are there any works about temporal action segmentation in the unsupervised setup?", "answer": ["Temporally-Weighted Hierarchical Clustering for Unsupervised Action\n Segmentation", "Unsupervised Learning and Segmentation of Complex Activities from Video", "Unsupervised learning of action classes with continuous temporal\n embedding"], "answer_arxiv_id": ["2103.11264", "1803.09490", "1904.04189"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_16837"} +{"question": "What works originated the concept of DDPMs?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_16838"} +{"question": "Any works about learning a single periodic neuron over noisy isotropic Gaussian distributions challenging the Shortest Vector Problem?", "answer": ["On the Cryptographic Hardness of Learning Single Periodic Neurons"], "answer_arxiv_id": ["2106.10744"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_16839"} +{"question": "In what works are layout control signals such as bounding boxes, segmentation maps, and key points used in diffusion models?", "answer": ["LayoutLLM-T2I: Eliciting Layout Guidance from LLM for Text-to-Image\n Generation", "BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained\n Diffusion", "Freestyle Layout-to-Image Synthesis", "Zero-shot spatial layout conditioning for text-to-image diffusion models", "Continuous Layout Editing of Single Images with Diffusion Models"], "answer_arxiv_id": ["2308.05095", "2307.10816", "2303.14412", "2306.13754", "2306.13078"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_16840"} +{"question": "Could you provide some studies about using neural networks for program synthesis, debugging, code analysis and reverse engineering?", "answer": ["Program Synthesis with Large Language Models", "Static Prediction of Runtime Errors by Learning to Execute Programs with\n External Resource Descriptions"], "answer_arxiv_id": ["2108.07732", "2203.03771"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_16841"} +{"question": "What works are about image composition?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion", "Kosmos-G: Generating Images in Context with Multimodal Large Language\n Models"], "answer_arxiv_id": ["2208.12242", "2212.04488", "2310.02992"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16842"} +{"question": "What researches presented a Transformer-based RL framework that conducts autoregressive report modeling and study-report matching?", "answer": ["Generalized Radiograph Representation Learning via Cross-supervision between Images and Free-text Radiology Reports"], "answer_arxiv_id": ["2111.03452"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_16843"} +{"question": "Which papers have explored techniques to improve the generalization of CAD-based object pose estimation?", "answer": ["3D Object Detection and Pose Estimation of Unseen Objects in Color\n Images with Local Surface Embeddings", "ZS6D: Zero-shot 6D Object Pose Estimation using Vision Transformers", "ZePHyR: Zero-shot Pose Hypothesis Rating", "Templates for 3D Object Pose Estimation Revisited: Generalization to New Objects and Robustness to Occlusions", "Fusing Local Similarities for Retrieval-based 3D Orientation Estimation\n of Unseen Objects", "OSOP: A Multi-Stage One Shot Object Pose Estimation Framework", "Diff-DOPE: Differentiable Deep Object Pose Estimation", "MegaPose: 6D Pose Estimation of Novel Objects via Render & Compare"], "answer_arxiv_id": ["2010.04075", "2309.11986", "2104.13526", "2203.17234v1", "2203.08472", "2203.15533", "2310.00463", "2212.06870"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_16844"} +{"question": "What papers introduce the concept of adversarial examples?", "answer": ["Intriguing properties of neural networks"], "answer_arxiv_id": ["1312.6199"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_16845"} +{"question": "Could you provide me some studies about ordered clustering?", "answer": ["Coresets for Ordered Weighted Clustering"], "answer_arxiv_id": ["1903.04351"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_16846"} +{"question": "Which papers propose the usage of human preference datasets for training reward models?", "answer": ["ImageReward: Learning and Evaluating Human Preferences for Text-to-Image\n Generation", "Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image\n Generation", "Human Preference Score: Better Aligning Text-to-Image Models with Human\n Preference", "Human Preference Score v2: A Solid Benchmark for Evaluating Human\n Preferences of Text-to-Image Synthesis"], "answer_arxiv_id": ["2304.05977", "2305.01569", "2303.14420", "2306.09341"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_16847"} +{"question": "Which studies focused on addressing hallucination issue by limiting the length of instruction data?", "answer": ["VIGC: Visual Instruction Generation and Correction"], "answer_arxiv_id": ["2308.12714"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_16848"} +{"question": "What publications are related to data boosting for instance segmentation?", "answer": ["On Pre-Trained Image Features and Synthetic Images for Deep Learning", "Render for CNN: Viewpoint Estimation in Images Using CNNs Trained with Rendered 3D Model Views", "Photorealistic Image Synthesis for Object Instance Detection", "Playing for Data: Ground Truth from Computer Games", "Playing for Benchmarks", "Cut, Paste and Learn: Surprisingly Easy Synthesis for Instance Detection", "Modeling Visual Context is Key to Augmenting Object Detection Datasets", "InstaBoost: Boosting Instance Segmentation via Probability Map Guided Copy-Pasting", "Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentation", "LAION-5B: An open large-scale dataset for training next generation image-text models", "Detecting Twenty-thousand Classes using Image-level Supervision", "Weakly Supervised Semantic Segmentation using Web-Crawled Videos", "STC: A Simple to Complex Framework for Weakly-supervised Semantic Segmentation", "RegionCLIP: Region-based Language-Image Pretraining"], "answer_arxiv_id": ["1710.10710", "1505.05641", "1902.03334", "1608.02192v1", "1709.07322v1", "1708.01642", "1807.07428", "1908.07801", "2012.07177", "2210.08402", "2201.02605", "1701.00352", "1509.03150", "2112.09106"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_16849"} +{"question": "Who proposed a novel positional embedding schema that supports variable aspect ratios?", "answer": ["Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding"], "answer_arxiv_id": ["2210.03347"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_16850"} +{"question": "Which work introduced Variational Causal Autoencoders (VACA)?", "answer": ["VACA: Design of Variational Graph Autoencoders for Interventional and Counterfactual Queries"], "answer_arxiv_id": ["2110.14690"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16851"} +{"question": "Any research papers that obtained small-loss bounds in contextual bandits?", "answer": ["Make the Minority Great Again: First-Order Regret Bound for Contextual Bandits", "Efficient First-Order Contextual Bandits: Prediction, Allocation, and Triangular Discrimination"], "answer_arxiv_id": ["1802.03386", "2107.02237"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16852"} +{"question": "What papers propose the use of latent-conditioned NeRF?", "answer": ["NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections", "pixelNeRF: Neural Radiance Fields from One or Few Images", "IBRNet: Learning Multi-View Image-Based Rendering"], "answer_arxiv_id": ["2008.02268", "2012.02190", "2102.13090"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_16853"} +{"question": "Which work introduced FedEx for optimizing client optimizer hyperparameters in Federated Learning?", "answer": ["Federated Hyperparameter Tuning: Challenges, Baselines, and Connections to Weight-Sharing"], "answer_arxiv_id": ["2106.04502"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_16854"} +{"question": "Can you provide me with some examples of recently published large text-to-image models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2205.11487", "2112.10741", "2204.06125"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16855"} +{"question": "Who introduced the SelfAware, a dataset of unanswerable questions and their answerable counterparts, to assess the uncertainty in the LLM’s responses?", "answer": ["Do Large Language Models Know What They Don't Know?"], "answer_arxiv_id": ["2305.18153"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_16856"} +{"question": "What paper explored and addressed the expert utilization problem of MoEs in a mixture of dataset setting?", "answer": ["On the Representation Collapse of Sparse Mixture of Experts"], "answer_arxiv_id": ["2204.09179"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_16857"} +{"question": "Could you provide me some studies that used deep learning-based stimulus encoders?", "answer": ["Hybrid Neural Autoencoders for Stimulus Encoding in Visual and Other Sensory Neuroprostheses"], "answer_arxiv_id": ["2205.13623"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_16858"} +{"question": "What works have used programming languages for information extraction?", "answer": ["Code4Struct: Code Generation for Few-Shot Event Structure Prediction"], "answer_arxiv_id": ["2210.12810"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_16859"} +{"question": "What works used a continuous token embedding space for diffusion models?", "answer": ["Diffusion-LM Improves Controllable Text Generation", "DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models"], "answer_arxiv_id": ["2205.14217", "2210.08933"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_16860"} +{"question": "Which studies use infants’ egocentric videos in SAYCam and deep learning methods to mimic children’s word learning experience?", "answer": ["Self-supervised learning through the eyes of a child", "Predicting Word Learning in Children from the Performance of Computer Vision Systems", "A Computational Acquisition Model for Multimodal Word Categorization", "Learning word-referent mappings and concepts from raw inputs"], "answer_arxiv_id": ["2007.16189", "2207.09847", "2205.05974", "2003.05573"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16861"} +{"question": "Any studies have explored alignment-based objectives for self-supervised video features?", "answer": ["Temporal Cycle-Consistency Learning", "Aligning Videos in Space and Time", "Representation Learning via Global Temporal Alignment and Cycle-Consistency", "Learning by Aligning Videos in Time", "Learning to Align Sequential Actions in the Wild", "Context-Aware Sequence Alignment using 4D Skeletal Augmentation"], "answer_arxiv_id": ["1904.07846", "2007.04515", "2105.05217", "2103.17260", "2111.09301", "2204.12223"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16862"} +{"question": "What works used diffusion models in exemplar-based image generation?", "answer": ["MIDMs: Matching Interleaved Diffusion Models for Exemplar-based Image\n Translation", "Person Image Synthesis via Denoising Diffusion Model", "Adding Conditional Control to Text-to-Image Diffusion Models", "Composer: Creative and Controllable Image Synthesis with Composable\n Conditions", "Paint by Example: Exemplar-based Image Editing with Diffusion Models", "In-Context Learning Unlocked for Diffusion Models"], "answer_arxiv_id": ["2209.11047", "2211.12500", "2302.05543", "2302.09778", "2211.13227", "2305.01115"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16863"} +{"question": "Could you provide me some works that introduced large language models on code?", "answer": ["Evaluating Large Language Models Trained on Code", "Competition-Level Code Generation with AlphaCode"], "answer_arxiv_id": ["2107.03374", "2203.07814"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_16864"} +{"question": "What paper is related to the development of a single-shot pipeline to regress object pose, shape, and appearance?", "answer": ["ShAPO: Implicit Representations for Multi-Object Shape, Appearance, and\n Pose Optimization"], "answer_arxiv_id": ["2207.13691"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_16865"} +{"question": "Could you provide me studies that adopted Kullback-Leibler divergence loss in inner maximization to improve the quality of adversarial examples?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy", "Revisiting Adversarial Robustness Distillation: Robust Soft Labels Make\n Student Better"], "answer_arxiv_id": ["1901.08573", "2108.07969"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_16866"} +{"question": "In which papers the researchers discuss difficulties in locating referents caused by ambiguities in referring expressions in robotic systems and how to address them?", "answer": ["INVIGORATE: Interactive Visual Grounding and Grasping in Clutter", "Talk-to-Resolve: Combining scene understanding and spatial dialogue to resolve granular task ambiguity for a collocated robot"], "answer_arxiv_id": ["2108.11092", "2111.11099"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_16867"} +{"question": "Could you provide me some works on selective methods in Parameter-efficient finetuning?", "answer": ["BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models", "Three things everyone should know about Vision Transformers", "Towards a Unified View of Parameter-Efficient Transfer Learning"], "answer_arxiv_id": ["2106.10199", "2203.09795", "2110.04366"], "source_meta": {"published_time": "20240205"}, "qid": "AutoScholarQuery_train_16868"} +{"question": "What studies reported examples of overoptimization as a form of specification gaming in different settings?", "answer": ["The Surprising Creativity of Digital Evolution: A Collection of Anecdotes from the Evolutionary Computation and Artificial Life Research Communities"], "answer_arxiv_id": ["1803.03453v4"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_16869"} +{"question": "Can you list some papers on the exploration of LLM capabilities in open-ended generative tasks?", "answer": ["Bounding the Capabilities of Large Language Models in Open Text\n Generation with Prompt Constraints", "Art or Artifice? Large Language Models and the False Promise of\n Creativity"], "answer_arxiv_id": ["2302.09185", "2309.14556"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_16870"} +{"question": "In which citations, did the researchers notice that the position of key information in long contexts greatly impacts the capability of large language models to correctly understand and process text?", "answer": ["Lost in the Middle: How Language Models Use Long Contexts"], "answer_arxiv_id": ["2307.03172"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_16871"} +{"question": "What papers suggest using pessimism to correct overestamation in value function?", "answer": ["Addressing Function Approximation Error in Actor-Critic Methods", "Reinforcement Learning with Augmented Data", "Predictive Information Accelerates Learning in RL", "Tactical Optimism and Pessimism for Deep Reinforcement Learning"], "answer_arxiv_id": ["1802.09477", "2004.14990", "2007.12401", "2102.03765"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_16872"} +{"question": "Can you provide some examples of research that studied multi-modal Masked Auto-Encoder?", "answer": ["Multimodal Masked Autoencoders Learn Transferable Representations", "Masked Vision and Language Modeling for Multi-modal Representation Learning"], "answer_arxiv_id": ["2205.14204", "2208.02131"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_16873"} +{"question": "Which research papers are based on datasets from instructional videos for research in procedure planning?", "answer": ["Procedure Planning in Instructional Videos", "Procedure Planning in Instructional Videos via Contextual Modeling and\n Model-based Policy Learning", "Learning Procedure-aware Video Representation from Instructional Videos\n and Their Narrations"], "answer_arxiv_id": ["1907.01172", "2110.01770", "2303.17839"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_16874"} +{"question": "Which work considered reusing data from other tasks but assumes an oracle reward function for the new tasks?", "answer": ["Conservative Data Sharing for Multi-Task Offline Reinforcement Learning"], "answer_arxiv_id": ["2109.08128"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_16875"} +{"question": "What works have validated the effectiveness of curriculum learning in computer vision?", "answer": ["FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo\n Labeling", "Multi-Task Curriculum Framework for Open-Set Semi-Supervised Learning"], "answer_arxiv_id": ["2110.08263", "2007.11330"], "source_meta": {"published_time": "20240418"}, "qid": "AutoScholarQuery_train_16876"} +{"question": "What papers highlight the trend of large-scale pre-training in vision-language (VL) research?", "answer": ["UNITER: UNiversal Image-TExt Representation Learning", "An Empirical Study of Training End-to-End Vision-and-Language Transformers", "Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal Transformers", "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision", "UNIMO: Towards Unified-Modal Understanding and Generation via Cross-Modal Contrastive Learning", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["1909.11740", "2111.02387", "2004.00849", "2102.03334", "2012.15409", "1908.08530", "2203.05557"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_16877"} +{"question": "Are there studies that focus on improving in-context learning through task instructions?", "answer": ["The Turking Test: Can Language Models Understand Instructions?", "Cross-Task Generalization via Natural Language Crowdsourcing Instructions"], "answer_arxiv_id": ["2010.11982", "2104.08773"], "source_meta": {"published_time": "20220905"}, "qid": "AutoScholarQuery_train_16878"} +{"question": "What studies utilized Implicit Environment Functions (IEF) and cost-to-go (c2g) functions in their methodologies?", "answer": ["Learning Continuous Environment Fields via Implicit Functions", "Cost-to-Go Function Generating Networks for High Dimensional Motion Planning"], "answer_arxiv_id": ["2111.13997", "2012.06023"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16879"} +{"question": "What work is there on exploring strategies to grow the representation of collocation points with high residuals in PINNs?", "answer": ["A comprehensive study of non-adaptive and residual-based adaptive sampling for physics-informed neural networks", "DeepXDE: A deep learning library for solving differential equations", "Efficient Training of Physics-Informed Neural Networks via Importance Sampling"], "answer_arxiv_id": ["2207.10289", "1907.04502", "2104.12325"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_16880"} +{"question": "Which papers propose generative models over the outcome trajectory for continuous-time treatments and outcomes?", "answer": ["A Framework for Individualizing Predictions of Disease Trajectories by Exploiting Multi-Resolution Structure", "A Bayesian Nonparametric Approach for Estimating Individualized Treatment-Response Curves", "Treatment-Response Models for Counterfactual Reasoning with Continuous-time, Continuous-valued Interventions", "Errors-in-variables Modeling of Personalized Treatment-Response Trajectories", "Patient-Specific Effects of Medication Using Latent Force Models with Gaussian Processes"], "answer_arxiv_id": ["1601.04674", "1608.05182v2", "1704.02038", "1906.03989v1", "1906.00226"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_16881"} +{"question": "Can you name some studies that have investigated path-length bounds and sparsity-dependent bounds?", "answer": ["Gains and Losses are Fundamentally Different in Regret Minimization: The Sparse Case", "Improved Path-length Regret Bounds for Bandits", "Sparsity, variance and curvature in multi-armed bandits", "More Adaptive Algorithms for Adversarial Bandits", "Equipping Experts/Bandits with Long-term Memory"], "answer_arxiv_id": ["1511.08405", "1901.10604", "1711.01037", "1801.03265", "1905.12950"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_16882"} +{"question": "What studies employed Exponential Moving Average in semi-supervised learning algorithms for tasks such as classification and object detection?", "answer": ["Mean teachers are better role models: Weight-averaged consistency\n targets improve semi-supervised deep learning results", "Unbiased Teacher for Semi-Supervised Object Detection"], "answer_arxiv_id": ["1703.01780", "2102.09480"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_16883"} +{"question": "Could you name some papers on bi-level optimization recently attracting attention?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-Learning Priors for Efficient Online Bayesian Regression", "Learning Convex Optimization Control Policies", "OptNet: Differentiable Optimization as a Layer in Neural Networks", "Bilevel Optimization for Planning through Contact: A Semidirect Method", "Understanding and correcting pathologies in the training of learned optimizers"], "answer_arxiv_id": ["1703.03400", "1807.08912", "1912.09529", "1703.00443", "1906.04292", "1810.10180"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_16884"} +{"question": "Which works have been done on representation learning on dynamic graphs?", "answer": ["Representation Learning for Dynamic Graphs: A Survey", "Dynamic Network Embedding Survey"], "answer_arxiv_id": ["1905.11485", "2103.15447"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_16885"} +{"question": "Which paper introduced the concept of probabilities of causation into representation learning?", "answer": ["Desiderata for Representation Learning: A Causal Perspective"], "answer_arxiv_id": ["2109.03795v2"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_16886"} +{"question": "Which research papers proposed two-stream architectures that mutually leverage appearance and motion cues in the field of unsupervised VOS?", "answer": ["Motion-Attentive Transition for Zero-Shot Video Object Segmentation", "Full-Duplex Strategy for Video Object Segmentation", "Hierarchical Feature Alignment Network for Unsupervised Video Object\n Segmentation", "Unsupervised Video Object Segmentation via Prototype Memory Network", "Treating Motion as Option to Reduce Motion Dependency in Unsupervised\n Video Object Segmentation"], "answer_arxiv_id": ["2003.04253", "2108.03151v3", "2207.08485", "2209.03712", "2209.03138"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_16887"} +{"question": "Can you provide some references that model the adjacency matrices using score matching at different noise scales?", "answer": ["Permutation Invariant Graph Generation via Score-Based Generative Modeling"], "answer_arxiv_id": ["2003.00638"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_16888"} +{"question": "What papers analyzied the convergence of linear student networks trained with the cross-entropy loss?", "answer": ["Towards Understanding Knowledge Distillation", "Knowledge Distillation in Wide Neural Networks: Risk Bound, Data Efficiency and Imperfect Teacher"], "answer_arxiv_id": ["2105.13093", "2010.10090"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_16889"} +{"question": "What papers propose the use of Fourier transform and MLPs for univariate forecasting?", "answer": ["DEPTS: Deep Expansion Learning for Periodic Time Series Forecasting"], "answer_arxiv_id": ["2203.07681"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_16890"} +{"question": "Which papers discuss on region-based methods in object detection research?", "answer": ["Rich feature hierarchies for accurate object detection and semantic\n segmentation", "Fast R-CNN", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks", "Feature Pyramid Networks for Object Detection", "Mask R-CNN"], "answer_arxiv_id": ["1311.2524", "1504.08083", "1506.01497", "1612.03144", "1703.06870"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_16891"} +{"question": "Which works showed that similar results can be shown in the Ising model setting by using the screening estimator?", "answer": ["Interaction Screening: Efficient and Sample-Optimal Learning of Ising Models"], "answer_arxiv_id": ["1605.07252"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_16892"} +{"question": "Could you tell me what papers discussed the approximation power of sparse Deep Neural Networks (DNNs) across different classes of functions?", "answer": ["Nonparametric regression using deep neural networks with ReLU activation function", "Optimal Approximation with Sparsely Connected Deep Neural Networks"], "answer_arxiv_id": ["1708.06633", "1705.01714"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_16893"} +{"question": "Which studies proposed models fine-tuned on pseudo-labels for zero-shot segmentation task?", "answer": ["Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2112.01071"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16894"} +{"question": "Which research papers discussed the combination of various building blocks in ANNS?", "answer": ["Billion-scale similarity search with GPUs", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1702.08734", "2103.00020"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_16895"} +{"question": "Could you provide me papers about Variational Inference schemes, an approximation for deep learning?", "answer": ["Global inducing point variational posteriors for Bayesian neural networks and deep Gaussian processes"], "answer_arxiv_id": ["2005.08140"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_16896"} +{"question": "What research has been done on using VQVAE as generative prior?", "answer": ["Neural Discrete Representation Learning", "Towards Robust Blind Face Restoration with Codebook Lookup Transformer", "VQFR: Blind Face Restoration with Vector-Quantized Dictionary and Parallel Decoder", "RestoreFormer: High-Quality Blind Face Restoration from Undegraded Key-Value Pairs"], "answer_arxiv_id": ["1711.00937", "2206.11253", "2205.06803", "2201.06374"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_16897"} +{"question": "What studies discuss the possibility of catastrophic safety risks as LLMs approach or reach human-level capabilities?", "answer": ["Frontier AI Regulation: Managing Emerging Risks to Public Safety"], "answer_arxiv_id": ["2307.03718"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_16898"} +{"question": "Any studies about super-resolving 'fixel' or fibre segmentation in diffusion MRI?", "answer": ["Enhancing Fiber Orientation Distributions using Convolutional Neural Networks"], "answer_arxiv_id": ["2008.05409"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16899"} +{"question": "Which papers presented the methods focusing on image-to-video transfer learning?", "answer": ["VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding", "Revealing Single Frame Bias for Video-and-Language Learning", "Prompting Visual-Language Models for Efficient Video Understanding", "X-CLIP: End-to-End Multi-grained Contrastive Learning for Video-Text Retrieval", "CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "Less is More: ClipBERT for Video-and-Language Learning via Sparse Sampling", "LocVTP: Video-Text Pre-training for Temporal Localization", "CLIP-ViP: Adapting Pre-trained Image-Text Model to Video-Language Alignment", "Object-aware Video-language Pre-training for Retrieval"], "answer_arxiv_id": ["2109.14084", "2206.03428", "2112.04478", "2207.07285", "2106.11097", "2102.06183", "2207.10362", "2209.06430", "2112.00656"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_16900"} +{"question": "Which study presents a version-space based active learning method with performance guarantees for cost-sensitive multiclass classification?", "answer": ["Active Learning for Cost-Sensitive Classification"], "answer_arxiv_id": ["1703.01014"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_16901"} +{"question": "Can you provide me studies where Fisher market equilibrium is used as a model for online ad auction platforms?", "answer": ["Multiplicative Pacing Equilibria in Auction Markets"], "answer_arxiv_id": ["1706.07151v5"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_16902"} +{"question": "Could you provide me some works on scale-free learning that consider full-information feedback?", "answer": ["Scale-Free Online Learning"], "answer_arxiv_id": ["1601.01974"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_16903"} +{"question": "What research papers are associated with creation of significant benchmarks like FaceForensics++ and Deepfake Detection Challenge Dataset?", "answer": ["FaceForensics++: Learning to Detect Manipulated Facial Images"], "answer_arxiv_id": ["1901.08971"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_16904"} +{"question": "Which works illustrate that pre-trained LLMs are susceptible to backdoor attacks and can be transferred to downstream tasks?", "answer": ["Red Alarm for Pre-trained Models: Universal Vulnerability to\n Neuron-Level Backdoor Attacks", "BadPre: Task-agnostic Backdoor Attacks to Pre-trained NLP Foundation\n Models"], "answer_arxiv_id": ["2101.06969", "2110.02467"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_16905"} +{"question": "Could you give me some works that utilized convolutions with a projection for the rendering of novel views?", "answer": ["Equivariant Neural Rendering"], "answer_arxiv_id": ["2006.07630"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_16906"} +{"question": "Which studies propose graph-based models for neural architecture encoding?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "NASGEM: Neural Architecture Search via Graph Embedding Method", "Graph-based Neural Architecture Search with Operation Embeddings"], "answer_arxiv_id": ["1609.02907", "2007.04452", "2105.04885"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_16907"} +{"question": "Which paper established a standard method for training robust neural networks against adversarial examples?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_16908"} +{"question": "Which work generalized the world model of Neural Radiance Fields to handle anti-aliasing, different cameras, and lighting?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo\n Collections", "Block-NeRF: Scalable Large Scene Neural View Synthesis"], "answer_arxiv_id": ["2103.13415", "2008.02268", "2202.05263"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_16909"} +{"question": "Could you mention some works on imputing where to insert calls to tools within language models?", "answer": ["TALM: Tool Augmented Language Models", "Toolformer: Language Models Can Teach Themselves to Use Tools"], "answer_arxiv_id": ["2205.12255", "2302.04761"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_16910"} +{"question": "Which studies investigated utilizing word embeddings to generalize representation of linguistic instructions?", "answer": ["Ask Your Humans: Using Human Instructions to Improve Generalization in Reinforcement Learning", "Robust Navigation with Language Pretraining and Stochastic Sampling"], "answer_arxiv_id": ["2011.00517", "1909.02244v1"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_16911"} +{"question": "Which works discuss the role of human input in natural language processing tasks?", "answer": ["Is Reinforcement Learning (Not) for Natural Language Processing: Benchmarks, Baselines, and Building Blocks for Natural Language Policy Optimization", "Boosting Offline Reinforcement Learning with Action Preference Query"], "answer_arxiv_id": ["2210.01241", "2306.03362"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_16912"} +{"question": "Could you provide me some studies that focus on giving privacy guarantees that are tailored to the dataset at hand?", "answer": ["Selling Privacy at Auction", "Per-instance Differential Privacy"], "answer_arxiv_id": ["1011.1375", "1707.07708"], "source_meta": {"published_time": "20210620"}, "qid": "AutoScholarQuery_train_16913"} +{"question": "What works translated viseme sequences to animations by morphing templates or 3D rigged models?", "answer": ["wav2vec 2.0: A Framework for Self-Supervised Learning of Speech\n Representations"], "answer_arxiv_id": ["2006.11477"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_16914"} +{"question": "What studies introduce high-order non-stationarity learning in designing LSTM modules?", "answer": ["Memory In Memory: A Predictive Neural Network for Learning Higher-Order Non-Stationarity from Spatiotemporal Dynamics"], "answer_arxiv_id": ["1811.07490"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_16915"} +{"question": "Could you provide me a study which trains the diffusion backbone and segmentation head with a consistency loss?", "answer": ["Prompting Diffusion Representations for Cross-Domain Semantic\n Segmentation"], "answer_arxiv_id": ["2307.02138"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_16916"} +{"question": "Any research works that focus on parameter-efficient transfer learning, enabling a pretrained ViT to efficiently adapt to unseen domains with a lower risk of over-fitting?", "answer": ["Rehearsal-Free Domain Continual Face Anti-Spoofing: Generalize More and\n Forget Less", "Adaptive Transformers for Robust Few-shot Cross-domain Face\n Anti-spoofing", "FLIP: Cross-domain Face Anti-spoofing with Language Guidance", "S-Adapter: Generalizing Vision Transformer for Face Anti-Spoofing with\n Statistical Tokens"], "answer_arxiv_id": ["2303.09914", "2203.12175", "2309.16649", "2309.04038"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16917"} +{"question": "Any works about using generative models, like GANs and AutoEncoders, for shape completion?", "answer": ["SDF-StyleGAN: Implicit SDF-Based StyleGAN for 3D Shape Generation", "Unsupervised 3D Shape Completion through GAN Inversion", "Unpaired Point Cloud Completion on Real Scans using Adversarial Training", "Multimodal Shape Completion via Conditional Generative Adversarial Networks", "Improved Adversarial Systems for 3D Object Generation and Reconstruction", "AutoSDF: Shape Priors for 3D Completion, Reconstruction, and Generation", "Learning Representations and Generative Models for 3D Point Clouds"], "answer_arxiv_id": ["2206.12055", "2104.13366", "1904.00069", "2003.07717", "1707.09557", "2203.09516", "1707.02392"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_16918"} +{"question": "What study advances the concept to a fully automatic system, also achieving accelerated rendering speeds by distilling knowledge to a student Light Field Network (LFN)?", "answer": ["CoNFies: Controllable Neural Face Avatars"], "answer_arxiv_id": ["2211.08610"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_16919"} +{"question": "Which works propose utilizing pseudo-count as a useful surrogate of novelty in count-based exploration method?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "Count-Based Exploration with Neural Density Models"], "answer_arxiv_id": ["1606.01868", "1703.01310"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_16920"} +{"question": "Could you provide me some papers that develop algorithms for automatically generating jailbreak prompts?", "answer": ["GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts"], "answer_arxiv_id": ["2309.10253v4"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_16921"} +{"question": "What paper used the back substitution approach leveraging minimal area relaxations and Box relaxation to bound the worst case loss?", "answer": ["Towards Stable and Efficient Training of Verifiably Robust Neural Networks"], "answer_arxiv_id": ["1906.06316"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_16922"} +{"question": "Which works proposed shape-based attack methods in the context of generating sufficiently smooth and imperceptible adversarial examples?", "answer": ["3D Adversarial Attacks Beyond Point Cloud", "Isometric 3D Adversarial Examples in the Physical World", "Adversarial shape perturbations on 3D point clouds"], "answer_arxiv_id": ["2104.12146", "2210.15291", "1908.06062"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_16923"} +{"question": "What studies are about enhancing retrieval systems with artificially created data?", "answer": ["Document Expansion by Query Prediction", "Promptagator: Few-shot Dense Retrieval From 8 Examples", "GPL: Generative Pseudo Labeling for Unsupervised Domain Adaptation of\n Dense Retrieval", "Query2doc: Query Expansion with Large Language Models"], "answer_arxiv_id": ["1904.08375", "2209.11755", "2112.07577", "2303.07678"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_16924"} +{"question": "Any works about learning to select cutting planes in the context of improving BnB using Machine Learning?", "answer": ["Reinforcement Learning for Integer Programming: Learning to Cut", "Learning to Cut by Looking Ahead: Cutting Plane Selection via Imitation Learning", "Learning to Select Cuts for Efficient Mixed-Integer Programming"], "answer_arxiv_id": ["1906.04859", "2206.13414", "2105.13645"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_16925"} +{"question": "Which research works are most related to recovering 3D shapes and textures from sparse collections of animal images?", "answer": ["LASSIE: Learning Articulated Shapes from Sparse Image Ensemble via 3D Part Discovery", "Hi-LASSIE: High-Fidelity Articulated Shape and Skeleton Discovery from Sparse Image Ensemble", "MagicPony: Learning Articulated 3D Animals in the Wild"], "answer_arxiv_id": ["2207.03434", "2212.11042", "2211.12497"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_16926"} +{"question": "Which study designed a novel codebook transfer network with part-of-speech?", "answer": ["SPAE: Semantic Pyramid AutoEncoder for Multimodal Generation with Frozen\n LLMs"], "answer_arxiv_id": ["2306.17842"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_16927"} +{"question": "What works attempt to enhance the Rico dataset?", "answer": ["VINS: Visual Search for Mobile User Interface Design", "Learning to Denoise Raw Mobile UI Layouts for Improving Datasets at Scale"], "answer_arxiv_id": ["2102.05216", "2201.04100"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_16928"} +{"question": "Which works have applied non-Euclidean convolutional neural networks and point cloud-based learning models to encode the molecular surface?", "answer": ["Geometric deep learning on graphs and manifolds using mixture model CNNs"], "answer_arxiv_id": ["1611.08402"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_16929"} +{"question": "What are some key works that employ large-scale video-language pretraining in the context of egocentric vision?", "answer": ["EgoVLPv2: Egocentric Video-Language Pre-training with Fusion in the Backbone", "HierVL: Learning Hierarchical Video-Language Embeddings", "Learning Video Representations from Large Language Models"], "answer_arxiv_id": ["2307.05463v2", "2301.02311", "2212.04501v1"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_16930"} +{"question": "Could you provide me some studies that focus on understanding human pose from videos using pose and motion priors?", "answer": ["VIBE: Video Inference for Human Body Pose and Shape Estimation"], "answer_arxiv_id": ["1912.05656"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_16931"} +{"question": "What works proposed white-box attacks for models exposed to the attacker?", "answer": ["Explaining and Harnessing Adversarial Examples", "Adversarial examples in the physical world", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1412.6572", "1607.02533", "1706.06083"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_16932"} +{"question": "What works develop computer vision architectures using standard convolution operation?", "answer": ["Very Deep Convolutional Networks for Large-Scale Image Recognition", "Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1409.1556", "1512.03385"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_16933"} +{"question": "Which papers explored the generalization of minimum-norm interpolators and their role in overparametrized models?", "answer": ["Understanding deep learning requires rethinking generalization"], "answer_arxiv_id": ["1611.03530"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_16934"} +{"question": "Which studies encode spatial information in the network weights of an MLP through neural fields?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "Occupancy Networks: Learning 3D Reconstruction in Function Space", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation", "Fourier Features Let Networks Learn High Frequency Functions in Low\n Dimensional Domains"], "answer_arxiv_id": ["1812.02822", "1812.03828", "1901.05103", "2006.10739"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_16935"} +{"question": "Which papers have handled combinations of different modalities using a single model for visual object tracking?", "answer": ["Prompting for Multi-Modal Tracking", "Visual Prompt Multi-Modal Tracking"], "answer_arxiv_id": ["2207.14571", "2303.10826"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_16936"} +{"question": "Which works focus on probabilistic models that separate between static and dynamic factors in sequential disentanglement?", "answer": ["Unsupervised Learning of Disentangled and Interpretable Representations from Sequential Data", "Disentangled Sequential Autoencoder"], "answer_arxiv_id": ["1709.07902", "1803.02991"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_16937"} +{"question": "What work they refer for replacing its 2D backbone with lightweight DLA34 in their architecture design?", "answer": ["Deep Layer Aggregation"], "answer_arxiv_id": ["1707.06484"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_16938"} +{"question": "Which works propose changing the loss function to handle the encoding discontinuity in discontinuous representation of OBBs?", "answer": ["Optimization for Arbitrary-Oriented Object Detection via Representation\n Invariance Loss", "Learning Modulated Loss for Rotated Object Detection", "PIoU Loss: Towards Accurate Oriented Object Detection in Complex\n Environments", "SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated\n Objects", "Rethinking Rotated Object Detection with Gaussian Wasserstein Distance\n Loss", "Learning High-Precision Bounding Box for Rotated Object Detection via\n Kullback-Leibler Divergence", "The KFIoU Loss for Rotated Object Detection"], "answer_arxiv_id": ["2103.11636", "1911.08299", "2007.09584", "1811.07126", "2101.11952", "2106.01883", "2201.12558"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16939"} +{"question": "Which works have proposed methods for unsupervised optical flow estimation?", "answer": ["UNSUPERVISED CONVOLUTIONAL NEURAL NETWORKS FOR MOTION ESTIMATION", "UPFlow: Upsampling Pyramid for Unsupervised Optical Flow Learning"], "answer_arxiv_id": ["1601.06087", "2012.00212"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_16940"} +{"question": "Any works about employing neural networks that output a distance given a query coordinate?", "answer": ["Neural Fields in Visual Computing and Beyond", "Grasping Field: Learning Implicit Representations for Human Grasps", "Pose-NDF: Modeling Human Pose Manifolds with Neural Distance Fields", "Neural Grasp Distance Fields for Robot Manipulation"], "answer_arxiv_id": ["2111.11426", "2008.04451", "2207.13807", "2211.02647"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_16941"} +{"question": "Which papers have discussed the usage of GPT-like models in domains of automated scoring and low-resource language generation?", "answer": ["Using GPT-4 to Augment Unbalanced Data for Automatic Scoring", "DALE: Generative Data Augmentation for Low-Resource Legal NLP"], "answer_arxiv_id": ["2310.18365", "2310.15799"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_16942"} +{"question": "Which researches probe LLMs like Delphi and GPT-3, using ethics questionnaires such as the Moral Foundation Questionnaire or Shweder’s 'Big Three' Ethics?", "answer": ["Does Moral Code Have a Moral Code? Probing Delphi’s Moral Philosophy", "Moral Foundations of Large Language Models"], "answer_arxiv_id": ["2205.12771", "2310.15337"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_16943"} +{"question": "What works explore identifiability of latent processes under stationary environments and distribution shifts?", "answer": ["Temporally Disentangled Representation Learning"], "answer_arxiv_id": ["2210.13647"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_16944"} +{"question": "Who conducted research on tracking-based semi-supervised learning in Lidar?", "answer": ["Motion Inspired Unsupervised Perception and Prediction in Autonomous\n Driving", "Towards Unsupervised Object Detection From LiDAR Point Clouds"], "answer_arxiv_id": ["2210.08061", "2311.02007"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_16945"} +{"question": "Which works focused on volumetric representation for geometrical information input in neural networks?", "answer": ["Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose", "Ordinal Depth Supervision for 3D Human Pose Estimation", "VoxelPose: Towards Multi-Camera 3D Human Pose Estimation in Wild\n Environment", "Faster VoxelPose: Real-time 3D Human Pose Estimation by Orthographic\n Projection", "VoxelTrack: Multi-Person 3D Human Pose Estimation and Tracking in the\n Wild", "VTP: Volumetric Transformer for Multi-view Multi-person 3D Pose\n Estimation"], "answer_arxiv_id": ["1611.07828", "1805.04095", "2004.06239", "2207.10955", "2108.02452", "2205.12602"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_16946"} +{"question": "What papers report on the use of task-specific workarounds to implement restes in real-world robotics applications?", "answer": ["Collective Robot Reinforcement Learning with Distributed Asynchronous Guided Policy Search", "TossingBot: Learning to Throw Arbitrary Objects with Residual Physics", "Emergent Real-World Robotic Skills via Unsupervised Off-Policy Reinforcement Learning", "Recovery RL: Safe Reinforcement Learning with Learned Recovery Zones"], "answer_arxiv_id": ["1610.00673", "1903.11239", "2004.12974", "2010.15920"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_16947"} +{"question": "Which studies proposed the use of adapters in parameter-efficient fine-tuning?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Simple, Scalable Adaptation for Neural Machine Translation", "Learning multiple visual domains with residual adapters", "AdapterHub: A Framework for Adapting Transformers"], "answer_arxiv_id": ["1902.00751", "1909.08478", "1705.08045", "2007.07779"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_16948"} +{"question": "Can you provide some examples of research introducing spoken TOD datasets?", "answer": ["“How Robust r u?”: Evaluating Task-Oriented Dialogue Systems on Spoken Conversations", "EVI: Multilingual Spoken Dialogue Tasks and Dataset for Knowledge-Based Enrolment, Verification, and Identification"], "answer_arxiv_id": ["2109.13489", "2204.13496"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_16949"} +{"question": "Which works proposed methods to accelerate post-training by defining the reconstruction error as a linear least squares problem?", "answer": ["A Fast Post-Training Pruning Framework for Transformers"], "answer_arxiv_id": ["2204.09656"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_16950"} +{"question": "Which recent diffusion models achieve state-of-the-art performance on text-to-image generation?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_16951"} +{"question": "Are there any previous studies about database lineage tracking?", "answer": ["Efficient Approximate Search for Sets of Vectors"], "answer_arxiv_id": ["2107.06817"], "source_meta": {"published_time": "20221027"}, "qid": "AutoScholarQuery_train_16952"} +{"question": "What works proposed self-supervised methods for image denoising?", "answer": ["Noise2Noise: Learning Image Restoration without Clean Data", "Noise2Void - Learning Denoising from Single Noisy Images", "Noise2Self: Blind Denoising by Self-Supervision", "High-Quality Self-Supervised Deep Image Denoising", "Unpaired Learning of Deep Image Denoising", "Noisier2Noise: Learning to Denoise from Unpaired Noisy Data", "IDR: Self-Supervised Image Denoising via Iterative Data Refinement", "Neighbor2Neighbor: Self-Supervised Denoising from Single Noisy Images", "Blind2Unblind: Self-Supervised Image Denoising with Visible Blind Spots"], "answer_arxiv_id": ["1803.04189v3", "1811.10980", "1901.11365", "1901.10277v3", "2008.13711", "1910.11908", "2111.14358", "2101.02824", "2203.06967"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_16953"} +{"question": "What references talk about the optimization of domain weights for downstream tasks?", "answer": ["Practical Bayesian Optimization of Machine Learning Algorithms", "Neural Architecture Search with Reinforcement Learning"], "answer_arxiv_id": ["1206.2944", "1611.01578"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_16954"} +{"question": "Could you provide me some studies that uses data-centric method, specifically through fine-tuning the NMT model using domain-specific corpora?", "answer": ["Translation Transformers Rediscover Inherent Data Domains", "Joint Training for Neural Machine Translation Models with Monolingual Data", "Selecting Artificially-Generated Sentences for Fine-Tuning Neural Machine Translation", "Iterative Domain-Repaired Back-Translation", "Domain Adaptation of Neural Machine Translation by Lexicon Induction", "Dictionary-based Data Augmentation for Cross-Domain Neural Machine Translation"], "answer_arxiv_id": ["2109.07864", "1803.00353", "1909.12016", "2010.02473", "1906.00376", "2004.02577"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_16955"} +{"question": "Which works used Large Language Models with visual features as evaluators on images?", "answer": ["LLMScore: Unveiling the Power of Large Language Models in Text-to-Image\n Synthesis Evaluation", "T2I-CompBench: A Comprehensive Benchmark for Open-world Compositional\n Text-to-image Generation"], "answer_arxiv_id": ["2305.11116", "2307.06350"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_16956"} +{"question": "What research showed that for Transformer-based models, most recent tokens play a greater role than older tokens for next-token prediction?", "answer": ["Do Long-Range Language Models Actually Use Long-Range Context?"], "answer_arxiv_id": ["2109.09115"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_16957"} +{"question": "Which studies represent the class of generative models consisting of forward and reverse processes, known as Diffusion Probabilistic Models (DPMs)?", "answer": ["Denoising Diffusion Probabilistic Models", "Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2006.11239", "1907.05600", "2011.13456"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_16958"} +{"question": "What are some studies that have explored the use of GAN models in generating images for segmentation model training?", "answer": ["BigDatasetGAN: Synthesizing ImageNet with Pixel-wise Annotations", "DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort"], "answer_arxiv_id": ["2201.04684", "2104.06490"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_16959"} +{"question": "What works introduced the batch bandit methods?", "answer": ["Batched bandit problems", "Batched Multi-armed Bandits Problem"], "answer_arxiv_id": ["1505.00369", "1904.01763"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_16960"} +{"question": "Which research proposes to modify Adam optimizer for fine-tuning a pre-trained model?", "answer": ["Recall and Learn: Fine-tuning Deep Pretrained Language Models with Less Forgetting"], "answer_arxiv_id": ["2004.12651"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_16961"} +{"question": "Any studies that extended RotatE for dealing with emerging entities?", "answer": ["Inductively Representing Out-of-Knowledge-Graph Entities by Optimal Estimation Under Translational Assumptions"], "answer_arxiv_id": ["2009.12765"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_16962"} +{"question": "Can you provide research work which proposed the use of overlapped losses to reduce the impact of greedy learning?", "answer": ["LoCo: Local Contrastive Representation Learning", "Interlocking Backpropagation: Improving depthwise model-parallelism"], "answer_arxiv_id": ["2008.01342", "2010.04116"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_16963"} +{"question": "What works introduced the row-wise attention to capture inter-sample interactions using transformer?", "answer": ["SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training"], "answer_arxiv_id": ["2106.01342"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_16964"} +{"question": "Which studies have applied normalizing flows and Boltzmann generators to molecular sampling and free energy estimation?", "answer": ["Boltzmann Generators – Sampling Equilibrium States of Many-Body Systems with Deep Learning", "Temperature Steerable Flows and Boltzmann Generators", "Smooth Normalizing Flows", "Targeted free energy estimation via learned mappings", "Flow Annealed Importance Sampling Bootstrap"], "answer_arxiv_id": ["1812.01729", "2108.01590", "2110.00351", "2002.04913", "2208.01893"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_16965"} +{"question": "Which studies employ counterfactual reasoning for learning explainable machine learning models?", "answer": ["Explainable Reinforcement Learning Through a Causal Lens", "Learning \"What-if\" Explanations for Sequential Decision-Making", "Algorithmic recourse under imperfect causal knowledge: a probabilistic approach", "Decisions, Counterfactual Explanations and Strategic Behavior"], "answer_arxiv_id": ["1905.10958", "2007.13531", "2006.06831", "2002.04333"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_16966"} +{"question": "Which papers discussed reinforcement learning that improves a policy through continuous online interactions with the environment?", "answer": ["Value-Consistent Representation Learning for Data-Efficient Reinforcement Learning"], "answer_arxiv_id": ["2206.12542"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_16967"} +{"question": "Which works proposed other accounting tools to obtain tighter bounds in analyzing DP-SGD?", "answer": ["Numerical Composition of Differential Privacy"], "answer_arxiv_id": ["2106.02848"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_16968"} +{"question": "Which works are about latent diffusion models, and among them - Stable Diffusion model?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis"], "answer_arxiv_id": ["2112.10752", "2307.01952"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_16969"} +{"question": "Which papers about the Vision Transformer (ViT) model in computer vision?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Swin Transformer V2: Scaling Up Capacity and Resolution"], "answer_arxiv_id": ["2010.11929", "2103.14030", "2111.09883"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_16970"} +{"question": "Which research contributions discuss the challenges in assessing empathy due to its subjective nature?", "answer": ["E-CORE: Emotion Correlation Enhanced Empathetic Dialogue Generation"], "answer_arxiv_id": ["2311.15016"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_16971"} +{"question": "What studies have used perturbation-based approaches for instance-level GNN explanation?", "answer": ["GNNExplainer: Generating Explanations for Graph Neural Networks", "Parameterized Explainer for Graph Neural Network", "Interpreting Graph Neural Networks for NLP With Differentiable Edge Masking"], "answer_arxiv_id": ["1903.03894", "2011.04573", "2010.00577v3"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_16972"} +{"question": "Which works first extended the guided diffusion models for training a student model?", "answer": ["On Distillation of Guided Diffusion Models", "SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two\n Seconds"], "answer_arxiv_id": ["2210.03142v3", "2306.00980"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_16973"} +{"question": "Which paper originally proposed the use of noisy image pairs for unsupervised learning of image denoising?", "answer": ["Noise2Noise: Learning Image Restoration without Clean Data"], "answer_arxiv_id": ["1803.04189"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_16974"} +{"question": "What paper empirically investigates factors contributing to the formation of an effective vision encoder in a multimodal language model from the perspective of pretraining?", "answer": ["What Makes for Good Visual Tokenizers for Large Language Models?"], "answer_arxiv_id": ["2305.12223"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_16975"} +{"question": "Which works developed generic frameworks for Combinatorial Optimization?", "answer": ["Learning Combinatorial Optimization Algorithms over Graphs"], "answer_arxiv_id": ["1704.01665"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_16976"} +{"question": "Can you provide papers that proposed post-hoc prompting strategies and mentioned their strategies such as 'Reflexion', 'Self-Refine', and 'Self-Contrast'?", "answer": ["Automatically Correcting Large Language Models: Surveying the landscape\n of diverse self-correction strategies", "Self-Refine: Iterative Refinement with Self-Feedback", "REFINER: Reasoning Feedback on Intermediate Representations", "Large Language Models Can Self-Improve", "Self-Contrast: Better Reflection Through Inconsistent Solving\n Perspectives"], "answer_arxiv_id": ["2308.03188", "2303.17651", "2304.01904", "2210.11610", "2401.02009"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16977"} +{"question": "What works use the NOTEARS DAG penalty in the field of variational inference?", "answer": ["DAGs with NO TEARS: Continuous Optimization for Structure Learning"], "answer_arxiv_id": ["1803.01422"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_16978"} +{"question": "Are there any works that studied instantaneous hard constraints with unsafe states under deterministic state transitions?", "answer": ["Safe Exploration in Finite Markov Decision Processes with Gaussian Processes"], "answer_arxiv_id": ["1606.04753"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_16979"} +{"question": "Could you provide me the references that provide a survey on efficient methods of entropic OT for discrete setting?", "answer": ["Computational Optimal Transport"], "answer_arxiv_id": ["1803.00567"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_16980"} +{"question": "What research introduced the Program-of-Thought-Prompting that generates code and executes it using an interpreter?", "answer": ["Program of Thoughts Prompting: Disentangling Computation from Reasoning\n for Numerical Reasoning Tasks", "PAL: Program-aided Language Models"], "answer_arxiv_id": ["2211.12588", "2211.10435"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_16981"} +{"question": "Which papers are about efficient feature transfer method like SSF in pre-trained models?", "answer": ["Scaling & Shifting Your Features: A New Baseline for Efficient Model\n Tuning"], "answer_arxiv_id": ["2210.08823"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_16982"} +{"question": "What are some recent contrastive methods in self-supervised embedding prediction?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance Discrimination", "Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning"], "answer_arxiv_id": ["1805.01978", "1807.03748", "2002.05709", "1911.05722", "2003.04297"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_16983"} +{"question": "Which studies have combined neural network encoders with cost volumes to match features across views?", "answer": ["MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "Neural Rays for Occlusion-aware Image-based Rendering", "GeoNeRF: Generalizing NeRF with Geometry Priors"], "answer_arxiv_id": ["2103.15595", "2107.13421", "2111.13539"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_16984"} +{"question": "What studies focus on solving the problem of self-supervised pretraining being less significant for denser downstream tasks?", "answer": ["Rethinking ImageNet Pre-training", "Self-Supervised Pretraining Improves Self-Supervised Pretraining"], "answer_arxiv_id": ["1811.08883", "2103.12718"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_16985"} +{"question": "Could you provide me some research about personalized federated learning using local/global model interpolation?", "answer": ["Adaptive Personalized Federated Learning", "Variational Federated Multi-Task Learning", "Three Approaches for Personalization with Applications to Federated Learning"], "answer_arxiv_id": ["2003.13461", "1906.06268", "2002.10619"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_16986"} +{"question": "Which studies are related to the technique of cost aggregation in the process of establishing correspondence between visually or semantically similar images?", "answer": ["End-to-End Learning of Geometry and Context for Deep Stereo Regression", "Group-wise Correlation Stereo Network", "Hierarchical Deep Stereo Matching on High-resolution Images", "CATs: Cost Aggregation Transformers for Visual Correspondence", "Cost Aggregation with 4D Convolutional Swin Transformer for Few-Shot\n Segmentation"], "answer_arxiv_id": ["1703.04309", "1903.04025", "1912.06704", "2106.02520", "2207.10866"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_16987"} +{"question": "Could you name the studies that focused on improving feature description in Structure-from-Motion?", "answer": ["LIFT: Learned Invariant Feature Transform", "SuperPoint: Self-Supervised Interest Point Detection and Description", "D2-Net: A Trainable CNN for Joint Detection and Description of Local\n Features", "DISK: Learning local features with policy gradient"], "answer_arxiv_id": ["1603.09114", "1712.07629", "1905.03561", "2006.13566"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_16988"} +{"question": "Who introduced an approach to resize the source domain labels based on the mean object size (MOS) of the target domain for Lidar 3D-OD?", "answer": ["Train in Germany, Test in The USA: Making 3D Object Detectors Generalize"], "answer_arxiv_id": ["2005.08139"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_16989"} +{"question": "What is the paper that showed a parallelized gradient algorithm achieving the optimal rate in the context of realizable linear regression?", "answer": ["Least Squares Regression with Markovian Data: Fundamental Limits and Algorithms"], "answer_arxiv_id": ["2006.08916"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_16990"} +{"question": "Could you provide me some works inspired by Video Object Segmentation, Multi-Object Tracking, and Multi-Object Tracking and Segmentation and applied in VIS?", "answer": ["Video Object Segmentation using Space-Time Memory Networks", "MOTChallenge: A Benchmark for Single-Camera Multiple Target Tracking", "MOTS: Multi-Object Tracking and Segmentation", "CompFeat: Comprehensive Feature Aggregation for Video Instance Segmentation", "Video Instance Segmentation with a Propose-Reduce Paradigm", "VISOLO: Grid-Based Space-Time Aggregation for Efficient Online Video Instance Segmentation"], "answer_arxiv_id": ["1904.00607", "2010.07548", "1902.03604", "2012.03400", "2103.13746", "2112.04177"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_16991"} +{"question": "Which works adopted edge perturbation as topology augmentation and assumed each edge has equal importance to the property of the input graph?", "answer": ["Deep Graph Contrastive Representation Learning", "Graph Barlow Twins: A self-supervised representation learning framework for graphs", "Graph Contrastive Learning with Augmentations"], "answer_arxiv_id": ["2006.04131", "2106.02466", "2010.13902"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_16992"} +{"question": "What works utilized two collaborative latent variables to model the distributions of knowledge and response or enhance knowledge selection?", "answer": ["There Are a Thousand Hamlets in a Thousand People’s Eyes: Enhancing Knowledge-grounded Dialogue with Personal Memory"], "answer_arxiv_id": ["2204.02624"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_16993"} +{"question": "Which works utilized RNN and CNN-based architectures for time series forecasting tasks?", "answer": ["Recurrent Neural Filters: Learning Independent Bayesian Filtering Steps for Time Series Prediction", "Deep Factors for Forecasting", "DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks", "WaveNet: A Generative Model for Raw Audio", "An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling"], "answer_arxiv_id": ["1901.08096", "1905.12417", "1704.04110", "1609.03499", "1803.01271"], "source_meta": {"published_time": "20220719"}, "qid": "AutoScholarQuery_train_16994"} +{"question": "Could you provide me studies where reinforcement learning was used to generate diverse molecules for drug development?", "answer": ["Hit and Lead Discovery with Explorative RL and Fragment-based Molecule Generation"], "answer_arxiv_id": ["2110.01219"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_16995"} +{"question": "Which work simplified the T5’s Relative encoding and introduced a more efficient variant called ALiBi?", "answer": ["Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation"], "answer_arxiv_id": ["2108.12409"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_16996"} +{"question": "Any works about utilizing diffusion models for solving inverse problems?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "AVID: Any-Length Video Inpainting with Diffusion Model"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456", "2112.10752", "2205.11487", "2208.12242", "2303.11305", "2312.03816"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_16997"} +{"question": "Can you give me examples of studies that proposed methods for recovering material properties from a few images using synthetic data for training?", "answer": ["Flexible SVBRDF Capture with a Multi-Image Deep Network", "Guided Fine-Tuning for Large-Scale Material Transfer", "MaterialGAN: Reflectance Capture using a Generative SVBRDF Model", "Metappearance: Meta-Learning for Visual Appearance Reproduction"], "answer_arxiv_id": ["1906.11557", "2007.03059", "2010.00114", "2204.08993"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_16998"} +{"question": "Could you provide me some studies about molecule topology and geometry joint pretraining?", "answer": ["Pre-training Molecular Graph Representation with 3D Geometry", "3D Infomax improves GNNs for Molecular Property Prediction"], "answer_arxiv_id": ["2110.07728", "2110.04126"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_16999"} +{"question": "Could you provide me studies about spectral message passing?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks"], "answer_arxiv_id": ["1609.02907"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_17000"} +{"question": "What research achieved the same rate as bib.bib13 using OMD, which is more commonly referred to as OMWU when entropy regularization is in use for the mirror descent update rule?", "answer": ["Optimization, Learning, and Games with Predictable Sequences"], "answer_arxiv_id": ["1311.1869"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_17001"} +{"question": "Who used this analysis to bound Bayesian regret in partial monitoring settings?", "answer": ["An Information-Theoretic Approach to Minimax Regret in Partial Monitoring", "Mirror Descent and the Information Ratio"], "answer_arxiv_id": ["1902.00470", "2009.12228"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_17002"} +{"question": "Which paper proposed injecting learnable low-rank decomposition matrices into pretrained model parameters, known as Low-Rank Adaptation?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_17003"} +{"question": "Can you name the research papers that tried to combine cost-sensitive methods with techniques such as denoise, self-paced curriculum, and heuristic mix up?", "answer": ["Self-PU: Self Boosted and Calibrated Positive-Unlabeled Training"], "answer_arxiv_id": ["2006.11280"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_17004"} +{"question": "Could you mention papers that presented the use of Score distillation Sampling (SDS) method and its extended methods?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion"], "answer_arxiv_id": ["2209.14988"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_17005"} +{"question": "Which studies attempted to transform the structured reasoning into a linearized sequence?", "answer": ["Explaining Answers with Entailment Trees"], "answer_arxiv_id": ["2104.08661"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_17006"} +{"question": "Who studied interactive clustering where there is an oracle providing guidance on cluster merging?", "answer": ["Local algorithms for interactive clustering", "Interactive Bayesian Hierarchical Clustering"], "answer_arxiv_id": ["1312.6724", "1602.03258"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_17007"} +{"question": "Could you provide me some research on models based on sparse communication for scalability?", "answer": ["Distributed SLIDE: Enabling Training Large Neural Networks on Low Bandwidth and Simple CPU-Clusters via Model Parallelism and Sparsity"], "answer_arxiv_id": ["2201.12667"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_17008"} +{"question": "Can you list the studies that focused on the network pruning to compress the language model?", "answer": ["The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models", "Rethinking Network Pruning— under the Pre-train and Fine-tune Paradigm"], "answer_arxiv_id": ["2203.07259", "2104.08682"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_17009"} +{"question": "What researches studied intermediate chain-of-thought reasoning in in-context learning?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Automatic Chain of Thought Prompting in Large Language Models"], "answer_arxiv_id": ["2201.11903", "2210.03493"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_17010"} +{"question": "What paper achieves linear convergence of a variational PG method by considering a general strongly-concave utility function of the state-action occupancy measure and exploiting the hidden convexity of the problem?", "answer": ["Variational Policy Gradient Method for Reinforcement Learning with General Utilities"], "answer_arxiv_id": ["2007.02151"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_17011"} +{"question": "Can you mention some studies that used unpaired data to learn the unknown degradation process?", "answer": ["To learn image super-resolution, use a GAN to learn how to do image\n degradation first", "Unsupervised Learning for Real-World Super-Resolution", "Frequency Separation for Real-World Super-Resolution", "Unsupervised Real-world Image Super Resolution via Domain-distance Aware\n Training", "DeFlow: Learning Complex Image Degradations from Unpaired Data with\n Conditional Flows", "Learn from Unpaired Data for Image Restoration: A Variational Bayes\n Approach"], "answer_arxiv_id": ["1807.11458", "1909.09629", "1911.07850", "2004.01178", "2101.05796", "2204.10090"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17012"} +{"question": "Which paper noted that DP SGLD may have unbounded private loss even if the exact posterior was differentially private as desired?", "answer": ["Can Stochastic Gradient Langevin Dynamics Provide Differential Privacy for Deep Learning?"], "answer_arxiv_id": ["2110.05057"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_17013"} +{"question": "What papers focus on two-layer fully connected networks and the 'benign overfitting' phenomenon when trained on high-dimensional mixture model data?", "answer": ["Benign Overfitting without Linearity: Neural Network Classifiers Trained by Gradient Descent for Noisy Linear Data"], "answer_arxiv_id": ["2202.05928"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_17014"} +{"question": "Which studies have applied VQ-based generative models in various generation tasks such as image generation, video generation, text-to-image generation, and face restoration?", "answer": ["Vector-quantized Image Modeling with Improved VQGAN", "MaskGIT: Masked Generative Image Transformer", "Taming Transformers for High-Resolution Image Synthesis", "Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive\n Transformer", "VideoGPT: Video Generation using VQ-VAE and Transformers", "CogVideo: Large-scale Pretraining for Text-to-Video Generation via\n Transformers", "Zero-Shot Text-to-Image Generation", "CogView: Mastering Text-to-Image Generation via Transformers", "High-Resolution Image Synthesis with Latent Diffusion Models", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "VQFR: Blind Face Restoration with Vector-Quantized Dictionary and\n Parallel Decoder", "RestoreFormer: High-Quality Blind Face Restoration from Undegraded\n Key-Value Pairs", "Towards Robust Blind Face Restoration with Codebook Lookup Transformer"], "answer_arxiv_id": ["2110.04627", "2202.04200", "2012.09841", "2204.03638", "2104.10157", "2205.15868", "2102.12092", "2105.13290", "2112.10752", "2206.10789", "2205.06803", "2201.06374", "2206.11253"], "source_meta": {"published_time": "20221206"}, "qid": "AutoScholarQuery_train_17015"} +{"question": "Can you cite an example of a real-world dataset that is used for person tracking?", "answer": ["Human in Events: A Large-Scale Benchmark for Human-centric Video\n Analysis in Complex Events", "MOT16: A Benchmark for Multi-Object Tracking", "MOT20: A benchmark for multi object tracking in crowded scenes", "DanceTrack: Multi-Object Tracking in Uniform Appearance and Diverse\n Motion"], "answer_arxiv_id": ["2005.04490", "1603.00831", "2003.09003", "2111.14690"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_17016"} +{"question": "Does any work focus on trainability, optimization, connections to existing forms of calibration, and applications to decision-making in the context of MMD-based calibration metrics?", "answer": ["Calibration tests in multi-class classification: A unifying framework"], "answer_arxiv_id": ["1910.11385"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_17017"} +{"question": "What works effectively fix less-complex program bugs but rely on ground truth test cases?", "answer": ["Competition-Level Code Generation with AlphaCode", "Teaching Large Language Models to Self-Debug", "CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning"], "answer_arxiv_id": ["2203.07814", "2304.05128v2", "2207.01780"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_17018"} +{"question": "What studies investigated ways to decrease runtime and focus on a single input image for personalization?", "answer": ["Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models", "Subject-driven Text-to-Image Generation via Apprenticeship Learning", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "Taming Encoder for Zero Fine-tuning Image Customization with\n Text-to-Image Diffusion Models", "HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image\n Models"], "answer_arxiv_id": ["2302.12228", "2304.00186", "2304.03411", "2304.02642", "2307.06949"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_17019"} +{"question": "Are there any recent works that have proven the effectiveness of model expansion in Class-Incremental Learning?", "answer": ["Adaptive Aggregation Networks for Class-Incremental Learning", "DER: Dynamically Expandable Representation for Class Incremental Learning", "FOSTER: Feature Boosting and Compression for Class-Incremental Learning", "Overcoming Catastrophic Forgetting with Hard Attention to the Task", "Learning without Forgetting for Vision-Language Models"], "answer_arxiv_id": ["2010.05063", "2103.16788", "2204.04662", "1801.01423", "2305.19270"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_17020"} +{"question": "Could you provide me some examples of the studies that developed algorithms for learning POMDPs?", "answer": ["Policy Gradient in Partially Observable Environments: Approximation and Convergence"], "answer_arxiv_id": ["1810.07900"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_17021"} +{"question": "What works use an underlying triangle mesh, and how do they benefit from the traditional triangle rasterization pipeline?", "answer": ["MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient\n Neural Field Rendering on Mobile Architectures"], "answer_arxiv_id": ["2208.00277"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_17022"} +{"question": "What works are related to the robustness of the revenue objective for BIC mechanisms under various notions of statistical distance?", "answer": ["Multi-Item Mechanisms without Item-Independence: Learnability via Robustness"], "answer_arxiv_id": ["1911.02146v2"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_17023"} +{"question": "What studies train a posterior estimator in amortized SBI methods?", "answer": ["Automatic Posterior Transformation for Likelihood-free Inference"], "answer_arxiv_id": ["1905.07488"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_17024"} +{"question": "Which papers discuss machine learning with differentiable physics?", "answer": ["Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids", "Learning Mesh-Based Simulation with Graph Networks", "HyperDynamics: Meta-Learning Object and Agent Dynamics with Hypernetworks", "Learning to Simulate Complex Physics with Graph Networks", "DiffTaichi: Differentiable Programming for Physical Simulation", "Accelerated Policy Learning with Parallel Differentiable Simulation", "PlasticineLab: A Soft-Body Manipulation Benchmark with Differentiable Physics", "An End-to-End Differentiable Framework for Contact-Aware Robot Design", "DiffSkill: Skill Abstraction from Differentiable Physics for Deformable Object Manipulations with Tools", "SoftZoo: A Soft Robot Co-design Benchmark For Locomotion In Diverse Environments"], "answer_arxiv_id": ["1810.01566", "2010.03409", "2103.09439v1", "2002.09405", "1910.00935", "2204.07137", "2104.03311", "2107.07501", "2203.17275", "2303.09555"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_17025"} +{"question": "What research has been conducted on purely convolutional architectures for vision tasks on resource constraint devices?", "answer": ["Deep Residual Learning for Image Recognition", "A ConvNet for the 2020s", "Designing Network Design Spaces", "RepVGG: Making VGG-style ConvNets Great Again", "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision\n Applications", "MobileNetV2: Inverted Residuals and Linear Bottlenecks", "Searching for MobileNetV3", "MobileOne: An Improved One millisecond Mobile Backbone"], "answer_arxiv_id": ["1512.03385", "2201.03545", "2003.13678", "2101.03697", "1704.04861", "1801.04381", "1905.02244", "2206.04040"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_17026"} +{"question": "Can you cite some studies that proposed improving the global aggregation stages in Federated Learning?", "answer": ["Federated learning with matched averaging"], "answer_arxiv_id": ["2002.06440"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_17027"} +{"question": "What works used expert algorithms for online bipartite matching?", "answer": ["Online Vertex-Weighted Bipartite Matching: Beating 1-1/e with Random Arrivals"], "answer_arxiv_id": ["1804.07458v2"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17028"} +{"question": "Which work proposed and tested an entropy approximation for an implicit density used as a variational approximation?", "answer": ["Kernel Implicit Variational Inference", "Hierarchical Implicit Models and Likelihood-Free Variational Inference", "Implicit Weight Uncertainty in Neural Networks"], "answer_arxiv_id": ["1705.10119", "1702.08896", "1711.01297"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_17029"} +{"question": "Which papers discuss the use of domain knowledge to enhance the safety of a reinforcement learning agent?", "answer": ["Safe Exploration in Continuous Action Spaces", "Trial without Error: Towards Safe Reinforcement Learning via Human Intervention", "Safe Reinforcement Learning via Shielding", "Towards Safe Reinforcement Learning with a Safety Editor Policy", "Learning Barrier Certificates: Towards Safe Reinforcement Learning with Zero Training-time Violations", "Sauté RL: Almost Surely Safe Reinforcement Learning Using State Augmentation", "Context-Aware Safe Reinforcement Learning for Non-Stationary Environments"], "answer_arxiv_id": ["1801.08757", "1707.05173", "1708.08611", "2201.12427", "2108.01846", "2202.06558", "2101.00531"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_17030"} +{"question": "Can you identify the papers focused on designing task-aware functions to eliminate improper predictions in motion prediction?", "answer": ["Injecting Planning-Awareness into Prediction and Detection Evaluation", "Task-Relevant Failure Detection for Trajectory Predictors in Autonomous Vehicles"], "answer_arxiv_id": ["2110.03270", "2207.12380"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_17031"} +{"question": "What are the papers that discuss the adaptation of the QUD for discourse coherence?", "answer": ["Inquisitive Question Generation for High Level Text Comprehension", "Discourse Comprehension: A Question Answering Framework to Represent\n Sentence Connections", "Discourse Analysis via Questions and Answers: Parsing Dependency\n Structures of Questions Under Discussion"], "answer_arxiv_id": ["2010.01657", "2111.00701", "2210.05905"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_17032"} +{"question": "What research papers focus on classical semantic search methods in text similarity, classification, or information retrieval?", "answer": ["SimCSE: Simple Contrastive Learning of Sentence Embeddings", "Unsupervised Dense Information Retrieval with Contrastive Learning"], "answer_arxiv_id": ["2104.08821", "2112.09118"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_17033"} +{"question": "What works are focused on MindCraft environment where two agents converse to achieve a shared goal?", "answer": ["MindCraft: Theory of Mind Modeling for Situated Dialogue in\n Collaborative Tasks", "Towards Collaborative Plan Acquisition through Theory of Mind Modeling\n in Situated Dialogue"], "answer_arxiv_id": ["2109.06275", "2305.11271"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_17034"} +{"question": "What research works are there on Prompt Learning in Continual Learning?", "answer": ["When Prompt-based Incremental Learning Does Not Meet Strong Pretraining", "Learning to Prompt for Continual Learning", "CODA-Prompt: COntinual Decomposed Attention-based Prompting for\n Rehearsal-Free Continual Learning", "DualPrompt: Complementary Prompting for Rehearsal-free Continual\n Learning", "POP: Prompt Of Prompts for Continual Learning", "PromptFusion: Decoupling Stability and Plasticity for Continual Learning", "Introducing Language Guidance in Prompt-based Continual Learning", "Online Class Incremental Learning on Stochastic Blurry Task Boundary via\n Mask and Visual Prompt Tuning", "Multimodal Parameter-Efficient Few-Shot Class Incremental Learning"], "answer_arxiv_id": ["2308.10445", "2112.08654", "2211.13218", "2204.04799", "2306.08200", "2303.07223", "2308.15827", "2308.09303", "2303.04751"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_17035"} +{"question": "Could you provide me some works where Shapley value was used for data evaluation?", "answer": ["Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms", "Scalability vs. Utility: Do We Have to Sacrifice One for the Other in Data Importance Quantification?"], "answer_arxiv_id": ["1908.08619", "1911.07128"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_17036"} +{"question": "What papers included research on self-supervised methods in monocular depth estimation?", "answer": ["Learning to Fuse Monocular and Multi-view Cues for Multi-frame Depth\n Estimation in Dynamic Scenes", "Digging Into Self-Supervised Monocular Depth Estimation", "Disentangling Object Motion and Occlusion for Unsupervised Multi-frame\n Monocular Depth", "Kick Back & Relax: Learning to Reconstruct the World by Watching SlowTV", "Adaptive Fusion of Single-View and Multi-View Depth for Autonomous\n Driving", "Self-Supervised Monocular Depth Estimation with Internal Feature Fusion", "MonoViT: Self-Supervised Monocular Depth Estimation with a Vision\n Transformer", "HR-Depth: High Resolution Self-Supervised Monocular Depth Estimation", "Attention Concatenation Volume for Accurate and Efficient Stereo\n Matching", "Semantically-Guided Representation Learning for Self-Supervised\n Monocular Depth", "Learning Depth via Leveraging Semantics: Self-supervised Monocular Depth\n Estimation with Both Implicit and Explicit Semantic Guidance", "SC-DepthV3: Robust Self-supervised Monocular Depth Estimation for\n Dynamic Scenes", "R3D3: Dense 3D Reconstruction of Dynamic Scenes from Multiple Cameras"], "answer_arxiv_id": ["2304.08993", "1806.01260", "2203.15174", "2307.10713", "2403.07535", "2110.09482", "2208.03543", "2012.07356", "2203.02146", "2002.12319", "2102.06685", "2211.03660", "2308.14713"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_17037"} +{"question": "Which works offer studies on the abstraction of language models?", "answer": ["On the Measure of Intelligence", "Abstraction and Analogy-Making in Artificial Intelligence", "Abstraction, Reasoning and Deep Learning: A Study of the “Look and Say” Sequence"], "answer_arxiv_id": ["1911.01547", "2102.10717", "2109.12755"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_17038"} +{"question": "What research explored factual knowledge stored in Transformer MLP layers?", "answer": ["Transformer Feed-Forward Layers Are Key-Value Memories", "Knowledge Neurons in Pretrained Transformers", "Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2012.14913", "2104.08696", "2202.05262"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17039"} +{"question": "Are there any studies that show there's no REC algorithm whose expected runtime scales polynomially without making additional assumptions on Q and P?", "answer": ["Universally Quantized Neural Compression"], "answer_arxiv_id": ["2006.09952"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_17040"} +{"question": "Which research papers focus on the use of Bayesian Optimal Experiment Design in selecting interventions for causal discovery?", "answer": ["Learning Neural Causal Models with Active Interventions", "Interventions, Where and How? Experimental Design for Causal Models at Scale", "Bayesian Active Learning for Classification and Preference Learning"], "answer_arxiv_id": ["2109.02429", "2203.02016", "1112.5745"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_17041"} +{"question": "Are there any works that considered sparse 3D convolutional network in voxel-based methods?", "answer": ["4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1904.08755"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_17042"} +{"question": "Could you provide the works that come with guarantees that match the lower bound for the general monotone VI methods?", "answer": ["Last Iterate is Slower than Averaged Iterate in Smooth Convex-Concave Saddle Point Problems"], "answer_arxiv_id": ["2002.00057v2"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_17043"} +{"question": "What work showed empirical results on temporal discretization playing a role in determining the effectiveness of fitted Q-iteration?", "answer": ["Control Frequency Adaptation via Action Persistence in Batch Reinforcement Learning"], "answer_arxiv_id": ["2002.06836v2"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_17044"} +{"question": "What studies adopt similar approaches to DPM-Solver in modeling the denoising process?", "answer": ["Fast Sampling of Diffusion Models with Exponential Integrator", "Improved Order Analysis and Design of Exponential Integrator for\n Diffusion Models Sampling", "Elucidating the Design Space of Diffusion-Based Generative Models", "UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds"], "answer_arxiv_id": ["2204.13902", "2308.02157", "2206.00364v2", "2302.04867v4", "2202.09778"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_17045"} +{"question": "What papers are focused on exploring shared knowledge across various languages in multilingual models?", "answer": ["Same Neurons, Different Languages: Probing Morphosyntax in Multilingual\n Pre-trained Models", "Journey to the Center of the Knowledge Neurons: Discoveries of\n Language-Independent Knowledge Neurons and Degenerate Knowledge Neurons", "Unveiling A Core Linguistic Region in Large Language Models"], "answer_arxiv_id": ["2205.02023", "2308.13198", "2310.14928"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_17046"} +{"question": "Could you provide me some works about Large Language Models for Video Understanding?", "answer": ["MM-VID: Advancing Video Understanding with GPT-4V(ision)", "Valley: Video Assistant with Large Language model Enhanced abilitY", "MovieChat: From Dense Token to Sparse Memory for Long Video\n Understanding", "VideoChat: Chat-Centric Video Understanding"], "answer_arxiv_id": ["2310.19773", "2306.07207", "2307.16449", "2305.06355"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17047"} +{"question": "What studies have focused on developing an A.I text detection model that could generalized to unseen text generation models and domains?", "answer": ["Defending Against Neural Fake News", "DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature", "Real or Fake? Learning to Discriminate Machine from Human Generated Text", "Release Strategies and the Social Impacts of Language Models"], "answer_arxiv_id": ["1905.12616", "2301.11305", "1906.03351", "1908.09203"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_17048"} +{"question": "Which works are about pre-training LLMs on vast multilingual text corpora?", "answer": ["Few-shot Learning with Multilingual Language Models", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "PolyLM: An Open Source Polyglot Large Language Model"], "answer_arxiv_id": ["2112.10668", "2211.05100", "2307.06018"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_17049"} +{"question": "Which works looked into the evaluation of image captions in a reference-free context?", "answer": ["CLIPScore: A Reference-free Evaluation Metric for Image Captioning", "Positive-Augmented Contrastive Learning for Image and Video Captioning\n Evaluation"], "answer_arxiv_id": ["2104.08718", "2303.12112"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_17050"} +{"question": "What are the studies that have significantly improved the field of text-to-image generation?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Diffusion Models Beat GANs on Image Synthesis", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers"], "answer_arxiv_id": ["2103.00020", "2006.11239", "2010.02502", "2105.05233", "2204.06125", "2205.11487", "2112.10752", "2211.01324"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_17051"} +{"question": "Could you provide the works that focus on minimizing the total completion time where jobs either arrive online and/or are clairvoyant?", "answer": ["Permutation Predictions for Non-Clairvoyant Scheduling", "Algorithms with Prediction Portfolios"], "answer_arxiv_id": ["2202.10199", "2210.12438"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_17052"} +{"question": "Which papers have focused on the CoT prompting method and its variants?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Automatic Chain of Thought Prompting in Large Language Models", "Recitation-Augmented Language Models", "Measuring and Narrowing the Compositionality Gap in Language Models"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2210.03493", "2210.01296", "2210.03350"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17053"} +{"question": "Which works formulated the design of expressive equivariant GNNs that match the power of k-FWL test for k>1?", "answer": ["Provably Powerful Graph Networks", "On the Universality of Invariant Networks", "Expressive Power of Invariant and Equivariant Graph Neural Networks", "Expressiveness and Approximation Properties of Graph Neural Networks"], "answer_arxiv_id": ["1905.11136", "1901.09342", "2006.15646", "2204.04661"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_17054"} +{"question": "Which works analyze deep networks with a single non-linear layer?", "answer": ["The Role of Linear Layers in Nonlinear Interpolating Networks"], "answer_arxiv_id": ["2202.00856"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_17055"} +{"question": "What papers have focused on graph signal augmentation?", "answer": ["Graph Contrastive Learning with Augmentations"], "answer_arxiv_id": ["2010.13902"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_17056"} +{"question": "What research considers generative-based methods for synthesizing hand-object interaction images?", "answer": ["Hand-Object Interaction Image Generation", "Affordance Diffusion: Synthesizing Hand-Object Interactions"], "answer_arxiv_id": ["2211.15663", "2303.12538"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_17057"} +{"question": "What work evaluated the generalization ability of LLMs on the tasks like semantic parsing?", "answer": ["Evaluating the Impact of Model Scale for Compositional Generalization in Semantic Parsing"], "answer_arxiv_id": ["2205.12253"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_17058"} +{"question": "Are there any works that use ChatGPT with other models to improve image captioning?", "answer": ["ChatGPT Asks, BLIP-2 Answers: Automatic Questioning Towards Enriched Visual Descriptions"], "answer_arxiv_id": ["2303.06594"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17059"} +{"question": "Which works use techniques like DINOv2 in decoder-focused methods for task-specific feature extraction?", "answer": ["DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2304.07193"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17060"} +{"question": "Which works are related to zero-shot recognition?", "answer": ["KG-SP: Knowledge Guided Simple Primitives for Open World Compositional\n Zero-Shot Learning", "K-LITE: Learning Transferable Visual Models with External Knowledge", "CLIP-Event: Connecting Text and Images with Event Structures", "ASIF: Coupled Data Turns Unimodal Models to Multimodal Without Training", "Retrieval Augmented Classification for Long-Tail Visual Recognition", "Hybrid Routing Transformer for Zero-Shot Learning"], "answer_arxiv_id": ["2205.06784", "2204.09222", "2201.05078", "2210.01738", "2202.11233", "2203.15310"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17061"} +{"question": "Which works are dedicated to class-conditional video generation?", "answer": ["G3AN: Disentangling Appearance and Motion for Video Generation", "InMoDeGAN: Interpretable Motion Decomposition Generative Adversarial\n Network for Video Generation", "Adversarial Video Generation on Complex Datasets", "Generating Long Videos of Dynamic Scenes", "Generating Videos with Dynamics-aware Implicit Generative Adversarial\n Networks", "StyleGAN-V: A Continuous Video Generator with the Price, Image Quality\n and Perks of StyleGAN2", "A Good Image Generator Is What You Need for High-Resolution Video\n Synthesis", "Disentangling Multiple Features in Video Sequences using Gaussian\n Processes in Variational Autoencoders", "Motion-Based Generator Model: Unsupervised Disentanglement of\n Appearance, Trackable and Intrackable Motions in Dynamic Patterns", "VideoGPT: Video Generation using VQ-VAE and Transformers", "Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive\n Transformer"], "answer_arxiv_id": ["1912.05523", "2101.03049", "1907.06571", "2206.03429", "2202.10571", "2112.14683", "2104.15069", "2001.02408", "1911.11294", "2104.10157", "2204.03638"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_17062"} +{"question": "Which papers focus on learning dense representations for each pixel in the image and then clustering them?", "answer": ["Unsupervised Semantic Segmentation by Contrasting Object Mask Proposals", "Masked Autoencoders Are Scalable Vision Learners", "PiCIE: Unsupervised Semantic Segmentation using Invariance and Equivariance in Clustering", "Invariant Information Clustering for Unsupervised Image Classification and Segmentation", "SegSort: Segmentation by Discriminative Sorting of Segments", "Autoregressive Unsupervised Image Segmentation", "Unsupervised Hierarchical Semantic Segmentation with Multiview Cosegmentation and Clustering Transformers"], "answer_arxiv_id": ["2102.06191", "2111.06377", "2103.17070", "1807.06653", "1910.06962", "2007.08247", "2204.11432"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_17063"} +{"question": "Which papers talk about the usage of LLMs to generate candidates or supporting evidence for particular captions for KVQA?", "answer": ["KAT: A Knowledge Augmented Transformer for Vision-and-Language", "REVIVE: Regional Visual Representation Matters in Knowledge-Based Visual\n Question Answering"], "answer_arxiv_id": ["2112.08614", "2206.01201"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_17064"} +{"question": "What works present average-case analyses of the maximal initial learning rate and how it relates to the architecture and expected sharpness?", "answer": ["Deep ReLU Networks Preserve Expected Length", "Deep ReLU Networks Have Surprisingly Few Activation Patterns", "Complexity of Linear Regions in Deep Networks"], "answer_arxiv_id": ["2102.10492", "1906.00904", "1901.09021"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_17065"} +{"question": "Could you provide me studies about discrete-time generative models?", "answer": ["Conditional Sig-Wasserstein GANs for Time Series Generation"], "answer_arxiv_id": ["2006.05421v2"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_17066"} +{"question": "Could you provide some papers relevant to detector-free Structure-from-Motion framework that eliminates feature detection by performing coarse grid-level matching first and then refining 2D points for sub-pixel accuracy?", "answer": ["Structure-from-Motion using Dense CNN Features with Keypoint\n Relocalization", "OnePose++: Keypoint-Free One-Shot Object Pose Estimation without CAD\n Models"], "answer_arxiv_id": ["1805.03879", "2301.07673"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_17067"} +{"question": "What studies developed methods to fit the pre-defined maximum input length of transformer-based models for ICD coding?", "answer": ["PLM-ICD: Automatic ICD Coding with Pretrained Language Models", "Knowledge Injected Prompt Based Fine-tuning for Multi-label Few-shot ICD Coding"], "answer_arxiv_id": ["2207.05289", "2210.03304"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_17068"} +{"question": "What research works on using the signature kernel within gp models?", "answer": ["Bayesian Learning from Sequential Data using Gaussian Processes with Signature Covariances", "The Signature Kernel is the solution of a Goursat PDE"], "answer_arxiv_id": ["1906.08215", "2006.14794"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_17069"} +{"question": "Which study introduced averaged class-distance normalized variance (CDNV) as a metric to analyze feature-variability collapse?", "answer": ["On the Role of Neural Collapse in Transfer Learning"], "answer_arxiv_id": ["2112.15121"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_17070"} +{"question": "In what research papers have the authors applied contrastive methods to cross-modal data views?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Attention Bottlenecks for Multimodal Fusion", "Multimodal Self-Supervised Learning of General Audio Representations", "Large Scale Audiovisual Learning of Sounds with Weakly Labeled Data", "Telling Left from Right: Learning Spatial Correspondence of Sight and\n Sound", "Self-Supervised Learning by Cross-Modal Audio-Video Clustering", "Ambient Sound Provides Supervision for Visual Learning", "Cooperative Learning of Audio and Video Models from Self-Supervised\n Synchronization", "Audio-Visual Scene Analysis with Self-Supervised Multisensory Features", "Look, Listen and Learn", "SoundNet: Learning Sound Representations from Unlabeled Video", "Deep Multimodal Clustering for Unsupervised Audiovisual Learning", "Speech2Action: Cross-modal Supervision for Action Recognition", "Audiovisual SlowFast Networks for Video Recognition", "On Compositions of Transformations in Contrastive Self-Supervised\n Learning", "Active Contrastive Learning of Audio-Visual Video Representations"], "answer_arxiv_id": ["2103.00020", "2107.00135", "2104.12807", "2006.01595", "2006.06175", "1911.12667", "1608.07017", "1807.00230", "1804.03641", "1705.08168", "1610.09001", "1807.03094", "2003.13594", "2001.08740", "2003.04298", "2009.09805"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_17071"} +{"question": "Are there studiees that propose open-world semi-supervised learning?", "answer": ["Open-World Semi-Supervised Learning", "Dynamic Conceptional Contrastive Learning for Generalized Category Discovery", "PromptCAL: Contrastive Affinity Learning via Auxiliary Prompts for Generalized Novel Category Discovery", "OpenLDN: Learning to Discover Novel Classes for Open-World Semi-Supervised Learning", "Generalized Category Discovery", "OpenCon: Open-world Contrastive Learning"], "answer_arxiv_id": ["2102.03526", "2303.17393", "2212.05590", "2207.02261", "2201.02609", "2208.02764"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_17072"} +{"question": "What paper used the UCF101 dataset, originally created for human action recognition, for evaluating VFI methods?", "answer": ["UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild"], "answer_arxiv_id": ["1212.0402"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_17073"} +{"question": "Which work does the researcher build upon for learning techniques based on sentence embedding methods, like SentenceBERT?", "answer": ["Imagination is All You Need! Curved Contrastive Learning for Abstract\n Sequence Modeling Utilized on Long Short-Term Dialogue Planning"], "answer_arxiv_id": ["2211.07591"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_17074"} +{"question": "What works have demonstrated that shorter code lengths can be achieved by using the prequential approach with standard neural network architectures and optimizers?", "answer": ["The Description Length of Deep Learning Models"], "answer_arxiv_id": ["1802.07044"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_17075"} +{"question": "Are there any studies about box-supervised approaches for both semantic and instance segmentation?", "answer": ["DiscoBox: Weakly Supervised Instance Segmentation and Semantic Correspondence from Box Supervision", "Simple Does It: Weakly Supervised Instance and Semantic Segmentation"], "answer_arxiv_id": ["2105.06464", "1603.07485"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_17076"} +{"question": "Could you provide me some studies about segmentation approaches in Neural Radiance Fields (NeRFs)?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene\n Representation", "Panoptic Lifting for 3D Scene Understanding with Neural Fields", "NeSF: Neural Semantic Fields for Generalizable Semantic Segmentation of\n 3D Scenes", "Neural Volumetric Object Selection", "Segment Anything in 3D with Radiance Fields", "Interactive Segment Anything NeRF with Feature Imitation"], "answer_arxiv_id": ["2103.15875", "2212.09802", "2111.13260", "2205.14929", "2304.12308", "2305.16233"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_17077"} +{"question": "Are there any studies about using neural implicit representations, specifically the Signed Distance Function (SDF), for recovering 3D geometry from multi-view images?", "answer": ["Atlas: End-to-End 3D Scene Reconstruction from Posed Images", "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video", "VolumeFusion: Deep Depth Fusion for 3D Scene Reconstruction"], "answer_arxiv_id": ["2003.10432", "2104.00681", "2108.08623"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_17078"} +{"question": "Are there any studies of stylized image generation based on pre-trained deep convolutional or transformer-based neural networks?", "answer": ["ArtFlow: Unbiased Image Style Transfer via Reversible Neural Flows", "StyTr$^2$: Image Style Transfer with Transformers", "Style Transfer by Relaxed Optimal Transport and Self-Similarity", "Arbitrary Style Transfer with Style-Attentional Networks"], "answer_arxiv_id": ["2103.16877", "2105.14576", "1904.12785", "1812.02342"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_17079"} +{"question": "What are some works that discuss using data augmentation methods like jittering, rotation, and multi-cropping in the construction of consistent positive samples for contrastive learning in images?", "answer": ["Self-Supervised Learning of Pretext-Invariant Representations", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1912.01991", "2006.09882", "1911.05722"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_17080"} +{"question": "Please mention some researches about Text to 3D Scene Generation", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_17081"} +{"question": "What studies have focused on the importance of predicting the height to the ground in the field of camera-based roadside perception?", "answer": ["BEVHeight: A Robust Framework for Vision-based Roadside 3D Object\n Detection"], "answer_arxiv_id": ["2303.08498"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17082"} +{"question": "What studies have been proposed for unsupervised disentanglement metrics like UDR and MC?", "answer": ["Unsupervised Model Selection for Variational Disentangled Representation Learning", "InfoGAN-CR and ModelCentrality: Self-supervised Model Training and Selection for Disentangling GANs"], "answer_arxiv_id": ["1905.12614", "1906.06034"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_17083"} +{"question": "Which papers performed diffusion model on the contextualized BART embeddings?", "answer": ["Latent Diffusion for Language Generation"], "answer_arxiv_id": ["2212.09462"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_17084"} +{"question": "Who proposed a GCN-based method to handle emerging entities more flexibly?", "answer": ["Inductive Knowledge Graph Reasoning for Multi-batch Emerging Entities"], "answer_arxiv_id": ["2208.10378"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17085"} +{"question": "What works investigate possible mechanisms that enable in-context learning?", "answer": ["In-context Learning and Induction Heads", "What learning algorithm is in-context learning? Investigations with linear models", "Transformers Learn In-Context by Gradient Descent"], "answer_arxiv_id": ["2209.11895v1", "2211.15661", "2212.07677"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_17086"} +{"question": "Can you name the research that found quasi-polynomial time planning is attainable under the weak-observability assumption in computational challenge in POMDPs?", "answer": ["Planning in Observable POMDPs in Quasipolynomial Time"], "answer_arxiv_id": ["2201.04735"], "source_meta": {"published_time": "20220624"}, "qid": "AutoScholarQuery_train_17087"} +{"question": "What studies have used meshes in learning-based 3D reconstruction?", "answer": ["Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images", "GeoSim: Realistic Video Simulation via Geometry-Aware Composition for Self-Driving"], "answer_arxiv_id": ["1804.01654v2", "2101.06543"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_17088"} +{"question": "Are there any papers that proposed alternatives to the costly eigendecomposition operation and approximated the eigenfunctions?", "answer": ["Neural Eigenfunctions Are Structured Representation Learners"], "answer_arxiv_id": ["2210.12637"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17089"} +{"question": "Which paper presented the ACRE dataset to systematically evaluate current vision systems’ capability in causal induction?", "answer": ["ACRE: Abstract Causal REasoning Beyond Covariation"], "answer_arxiv_id": ["2103.14232"], "source_meta": {"published_time": "20220618"}, "qid": "AutoScholarQuery_train_17090"} +{"question": "Could you list me some studies that apply diffusion models to novel tasks in a zero-shot manner?", "answer": ["SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2108.01073", "2201.09865"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_17091"} +{"question": "Which publication considers data augmentation as a type of data-dependent regularization term?", "answer": ["A Kernel Theory of Modern Data Augmentation"], "answer_arxiv_id": ["1803.06084"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_17092"} +{"question": "Which studies are centered on the design of a model to learn to select different parameters with the same architecture for each of the input elements?", "answer": ["Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity", "GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding", "Mixture-of-Experts with Expert Choice Routing"], "answer_arxiv_id": ["2101.03961", "2006.16668", "2202.09368"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_17093"} +{"question": "Are there any works that integrated transformer models into semantic segmentation tasks?", "answer": ["Segmenter: Transformer for Semantic Segmentation", "SOTR: Segmenting Objects with Transformers"], "answer_arxiv_id": ["2105.05633", "2108.06747"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_17094"} +{"question": "What studies were cited as examples of LLMs becoming universal assistants?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_17095"} +{"question": "Are there any works that focus on a heterogeneous reward setting in the context of collision models in MPMAB model?", "answer": ["Distributed learning in congested environments with partial information"], "answer_arxiv_id": ["2103.15901"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17096"} +{"question": "Could you provide me some of the early methods proposed for single-image reflection removal?", "answer": ["Robust Reflection Removal with Reflection-free Flash-only Cues"], "answer_arxiv_id": ["2103.04273"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17097"} +{"question": "Which works discussed the use of backward adaptation in entropy estimation?", "answer": ["Joint Autoregressive and Hierarchical Priors for Learned Image\n Compression", "Checkerboard Context Model for Efficient Learned Image Compression", "Channel-wise Autoregressive Entropy Models for Learned Image Compression", "ELIC: Efficient Learned Image Compression with Unevenly Grouped\n Space-Channel Contextual Adaptive Coding", "Learning Accurate Entropy Model with Global Reference for Image\n Compression", "Entroformer: A Transformer-based Entropy Model for Learned Image\n Compression", "Contextformer: A Transformer with Spatio-Channel Attention for Context\n Modeling in Learned Image Compression", "MLIC: Multi-Reference Entropy Model for Learned Image Compression"], "answer_arxiv_id": ["1809.02736", "2103.15306", "2007.08739", "2203.10886", "2010.08321", "2202.05492", "2203.02452", "2211.07273"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17098"} +{"question": "Which work decomposes a video to atlases for video editing and why was this method limiting?", "answer": ["Text2LIVE: Text-Driven Layered Image and Video Editing"], "answer_arxiv_id": ["2204.02491"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_17099"} +{"question": "Which papers has the researcher referenced regarding the integration of the planning process into the generative model?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_17100"} +{"question": "What works discuss the development of universal speech and audio understanding models?", "answer": ["AudioGPT: Understanding and Generating Speech, Music, Sound, and Talking\n Head", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face", "Qwen-Audio: Advancing Universal Audio Understanding via Unified\n Large-Scale Audio-Language Models", "SALMONN: Towards Generic Hearing Abilities for Large Language Models", "Listen, Think, and Understand", "Joint Audio and Speech Understanding", "Pengi: An Audio Language Model for Audio Tasks"], "answer_arxiv_id": ["2304.12995", "2303.17580", "2311.07919", "2310.13289", "2305.10790", "2309.14405", "2305.11834"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_17101"} +{"question": "What studies have researched methods to improve LLMs’ math capabilities by prompting with carefully designed inputs, such as chain-of-thought prompting and program-of-thought prompting?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Neuro-Symbolic Integration Brings Causal and Reliable Reasoning Proofs", "Solving Challenging Math Word Problems Using GPT-4 Code Interpreter with\n Code-based Self-Verification", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "PAL: Program-aided Language Models", "Program of Thoughts Prompting: Disentangling Computation from Reasoning\n for Numerical Reasoning Tasks"], "answer_arxiv_id": ["2305.10601", "2311.09802", "2308.07921", "2201.11903", "2205.11916", "2211.10435", "2211.12588"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_17102"} +{"question": "What are the studies that focus on improving the computational efficiency of the diffusion model?", "answer": ["eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2211.01324", "2211.07600", "2205.11487"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_17103"} +{"question": "What works explored attention and feature injection for image editing?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2208.01626", "2211.12572"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_17104"} +{"question": "What research about LVLMs in the medical field can you share?", "answer": ["Med-Flamingo: a Multimodal Medical Few-shot Learner", "LLaVA-Med: Training a Large Language-and-Vision Assistant for\n Biomedicine in One Day", "PMC-VQA: Visual Instruction Tuning for Medical Visual Question Answering"], "answer_arxiv_id": ["2307.15189", "2306.00890", "2305.10415"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_17105"} +{"question": "Which papers provide a background on diffusion models used for generative models?", "answer": ["Denoising Diffusion Probabilistic Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2006.11239", "2112.10741"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17106"} +{"question": "What are some works that focus on accelerating the training of transformers by reducing the number of updated parameters?", "answer": ["bert2BERT: Towards Reusable Pretrained Language Models", "Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Automated Progressive Learning for Efficient Training of Vision\n Transformers"], "answer_arxiv_id": ["2110.07143", "2106.04533", "2203.14509"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_17107"} +{"question": "What works thoroughly investigated the importance of data augmentation techniques in creating optimal or hard samples without labels?", "answer": ["Unsupervised Visual Representation Learning by Context Prediction"], "answer_arxiv_id": ["1505.05192"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_17108"} +{"question": "What studies focused on maximizing performance on the training task distribution in memory-based Meta-RL?", "answer": ["RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "Learning to reinforcement learn", "A Survey of Meta-Reinforcement Learning"], "answer_arxiv_id": ["1611.02779", "1611.05763", "2301.08028"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_17109"} +{"question": "Could you tell me about works that infer per-timestep 3D voxel grids and use a CNN to regress frame-to-frame poses?", "answer": ["Video Autoencoder: self-supervised disentanglement of static 3D structure and motion", "Learning Spatial Common Sense with Geometry-Aware Recurrent Networks"], "answer_arxiv_id": ["2110.02951", "1901.00003"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17110"} +{"question": "What papers focused on improving the diffusion process by conducting it in the semantic latent space obtained with a pre-trained VAE?", "answer": ["Score-based Generative Modeling in Latent Space", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2106.05931", "2112.10752"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17111"} +{"question": "Which studies focused on single object synthesis in the context of Text-to-3D?", "answer": ["DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation", "DiffRF: Rendering-Guided 3D Radiance Field Diffusion", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation", "DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Text2NeRF: Text-Driven 3D Scene Generation with Neural Radiance Fields"], "answer_arxiv_id": ["2309.16653", "2212.01206", "2305.16213", "2209.14988", "2211.10440", "2305.11588"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_17112"} +{"question": "Could you provide me some works that improved the efficiency of search through temporal abstraction?", "answer": ["Efficient Black-Box Planning Using Macro-Actions with Focused Effects"], "answer_arxiv_id": ["2004.13242"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_17113"} +{"question": "What research has done a generalization analysis of MPNNs using a pre-defined finite set of graphons?", "answer": ["Generalization Analysis of Message Passing Neural Networks on Large Random Graphs"], "answer_arxiv_id": ["2202.00645"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17114"} +{"question": "Are there any works about the fast escape of saddle points by adaptive methods?", "answer": ["The Power of Normalization: Faster Evasion of Saddle Points", "Revisiting Normalized Gradient Descent: Fast Evasion of Saddle Points", "Adaptive Inertia: Disentangling the Effects of Adaptive Learning Rate and Momentum"], "answer_arxiv_id": ["1611.04831", "1711.05224", "2006.15815"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_17115"} +{"question": "Any works about BLOOM which is trained on a large multilingual dataset?", "answer": ["BLOOM: A 176B-Parameter Open-Access Multilingual Language Model"], "answer_arxiv_id": ["2211.05100"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_17116"} +{"question": "Which earlier studies generally formulate deepfake detection as a binary classification?", "answer": ["FaceForensics++: Learning to Detect Manipulated Facial Images", "Combining EfficientNet and Vision Transformers for Video Deepfake\n Detection"], "answer_arxiv_id": ["1901.08971", "2107.02612"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_17117"} +{"question": "Are there studies that explored pairwise similarity relations to regularize feature distributions of labeled and unlabeled data in novel class discovery?", "answer": ["OpenLDN: Learning to Discover Novel Classes for Open-World\n Semi-Supervised Learning"], "answer_arxiv_id": ["2207.02261"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_17118"} +{"question": "What papers presented execution-based benchmarks for code generation tasks?", "answer": ["Program Synthesis with Large Language Models", "Measuring Coding Challenge Competence With APPS", "Evaluating Large Language Models Trained on Code", "DS-1000: A Natural and Reliable Benchmark for Data Science Code Generation"], "answer_arxiv_id": ["2108.07732", "2105.09938", "2107.03374", "2211.11501"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_17119"} +{"question": "Could you tell me which study introduced the Transformer-based entity segmentation method, CropFormer?", "answer": ["High-Quality Entity Segmentation"], "answer_arxiv_id": ["2211.05776"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17120"} +{"question": "What papers proposed methods to look at diffusion models from a deep equilibrium (DEQ) perspective?", "answer": ["Deep Equilibrium Approaches to Diffusion Models"], "answer_arxiv_id": ["2210.12867"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_17121"} +{"question": "Which works explain black-box multimodal models primarily by relying on gradient-based visualizations?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "How to Explain Individual Classification Decisions"], "answer_arxiv_id": ["1312.6034", "0912.1128"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_17122"} +{"question": "Can you list some papers that used more images to enhance VLM's performance?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs", "LAION-5B: An open large-scale dataset for training next generation\n image-text models"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2111.02114", "2210.08402"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_17123"} +{"question": "What study used zero-convolutions for conditioning on text and image data?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_17124"} +{"question": "Could you mention some studies that propose techniques to combine the predictions of models, for instance by averaging, voting, or interpolation?", "answer": ["Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time", "Evaluating Deep Neural Network Ensembles by Majority Voting cum Meta-Learning scheme", "Patching open-vocabulary modelsby interpolating weights"], "answer_arxiv_id": ["2203.05482", "2105.03819v1", "2208.05592"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_17125"} +{"question": "What paper first introduced an online variant of the Frank-Wolfe algorithm (OFW)?", "answer": ["Projection-free Online Learning"], "answer_arxiv_id": ["1206.4657"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17126"} +{"question": "Could you provide some papers that focus on Knowledge-based IMs on maximising the diversity of states?", "answer": ["A survey on intrinsic motivation in reinforcement learning"], "answer_arxiv_id": ["1908.06976v2"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_17127"} +{"question": "Which works in metric learning consider using paired comparisons?", "answer": ["Guaranteed Classification via Regularized Similarity Learning", "Robustness and Generalization for Metric Learning"], "answer_arxiv_id": ["1306.3108", "1209.1086"], "source_meta": {"published_time": "20230908"}, "qid": "AutoScholarQuery_train_17128"} +{"question": "What papers indicate that free-form texts are susceptible to hallucinations?", "answer": ["Explain Yourself! Leveraging Language Models for Commonsense Reasoning", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["1906.02361", "2201.11903"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_17129"} +{"question": "Any works about creating a generative framework with part-level control for shape manipulation?", "answer": ["SPAGHETTI: Editing Implicit Shapes Through Part Aware Generation"], "answer_arxiv_id": ["2201.13168"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_17130"} +{"question": "Any works that used the transition matrix to preserve the discrete natures of the graph structure in diffusion models?", "answer": ["DiGress: Discrete Denoising diffusion for graph generation"], "answer_arxiv_id": ["2209.14734"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_17131"} +{"question": "What papers propose to leverage parameterized curves such as Bezier curves and Fourier contours to adaptively fit highly-curved text regions?", "answer": ["Fourier Contour Embedding for Arbitrary-Shaped Text Detection", "ABCNet: Real-time Scene Text Spotting with Adaptive Bezier-Curve Network", "DeepSolo++: Let Transformer Decoder with Explicit Points Solo for\n Multilingual Text Spotting", "Text Spotting Transformers"], "answer_arxiv_id": ["2104.10442", "2002.10200", "2305.19957", "2204.01918"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_17132"} +{"question": "What works discuss proposed SSL frameworks?", "answer": ["Emerging Properties in Self-Supervised Vision Transformers", "A Simple Framework for Contrastive Learning of Visual Representations", "Exploring Simple Siamese Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Masked Autoencoders Are Scalable Vision Learners", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction"], "answer_arxiv_id": ["2104.14294", "2002.05709", "2011.10566", "2006.07733", "2111.06377", "2103.03230"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_17133"} +{"question": "Could you give me examples of research that utilizes the neural tangent kernel to estimate generalization without the marginal likelihood, especially in the neural architecture search context?", "answer": ["Towards NNGP-guided Neural Architecture Search", "Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective"], "answer_arxiv_id": ["2011.06006", "2102.11535"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17134"} +{"question": "Which papers discuss the approach of empirical risk minimization with IPS-based estimators in off-policy learning?", "answer": ["Unbiased Learning-to-Rank with Biased Feedback", "Counterfactual Risk Minimization: Learning from Logged Bandit Feedback", "Off-policy evaluation for slate recommendation"], "answer_arxiv_id": ["1608.04468", "1502.02362", "1605.04812"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_17135"} +{"question": "Which works studied the generalization of visual prompts across models in the area of visual prompting?", "answer": ["Boosting Adversarial Attacks with Momentum"], "answer_arxiv_id": ["1710.06081"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_17136"} +{"question": "Could you provide me some studies that applying asynchronous distributed methods?", "answer": ["Asynchronous Online Federated Learning for Edge Devices with Non-IID Data", "Decentralized Consensus Optimization with Asynchrony and Delays"], "answer_arxiv_id": ["1911.02134", "1612.00150"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_17137"} +{"question": "What studies use a stochastic block model to analyze spectral contrastive learning for unsupervised domain adaption?", "answer": ["Connect, Not Collapse: Explaining Contrastive Learning for Unsupervised Domain Adaptation"], "answer_arxiv_id": ["2204.00570v4"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_17138"} +{"question": "What research is mentioned that utilizes policy optimization on surrogate policy spaces within safe RL?", "answer": ["Constrained Policy Optimization", "Projection-Based Constrained Policy Optimization", "Constrained Variational Policy Optimization for Safe Reinforcement Learning"], "answer_arxiv_id": ["1705.10528", "2010.03152", "2201.11927"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_17139"} +{"question": "Which research papers proposed techniques to balance between MAE and CE in robust loss functions?", "answer": ["Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels"], "answer_arxiv_id": ["1805.07836"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_17140"} +{"question": "Which papers discussed the use of classification-based methods in Event Argument Extraction (EAE)?", "answer": ["Entity, Relation, and Event Extraction with Contextualized Span\n Representations", "CLEVE: Contrastive Pre-training for Event Extraction"], "answer_arxiv_id": ["1909.03546", "2105.14485"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_17141"} +{"question": "Which works have trained large convolutional neural networks to predict the stability of wooden block towers?", "answer": ["Learning Physical Intuition of Block Towers by Example"], "answer_arxiv_id": ["1603.01312"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_17142"} +{"question": "Could you provide some studies on the assessment of language models in the reading comprehension tasks?", "answer": ["Know What You Don’t Know: Unanswerable Questions for SQuAD"], "answer_arxiv_id": ["1806.03822"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_17143"} +{"question": "Do we have any theoretical developments in preserving the characteristics of the input data while learning continuous latent representations in Wasserstein Auto-Encoder?", "answer": ["Wasserstein Auto-Encoders"], "answer_arxiv_id": ["1711.01558"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_17144"} +{"question": "Which work imroves inversion by conditioning the denoising process on noised low-pass filter data from the target image?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2108.02938"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_17145"} +{"question": "What research describes providing an instruction of the task in natural language without updating any parameters of LLM?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20220923"}, "qid": "AutoScholarQuery_train_17146"} +{"question": "Is there a specific study that addressed the need for non-rigid transformations in image editing using pre-trained diffusion models?", "answer": ["Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2210.09276"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_17147"} +{"question": "Which papers highlight the limitations of RS in terms of computational cost?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing"], "answer_arxiv_id": ["1902.02918"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_17148"} +{"question": "Which studies use direct regression to estimate 6D pose?", "answer": ["FS-Net: Fast Shape-based Network for Category-Level 6D Object Pose\n Estimation with Decoupled Rotation Mechanism", "DualPoseNet: Category-level 6D Object Pose and Size Estimation Using\n Dual Pose Network with Refined Learning of Pose Consistency"], "answer_arxiv_id": ["2103.07054", "2103.06526"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_17149"} +{"question": "Can you provide a work that proposes techniques for adapting large-scale models for zero-shot adversarial robustness?", "answer": ["Understanding Zero-Shot Adversarial Robustness for Large-Scale Models"], "answer_arxiv_id": ["2212.07016"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_17150"} +{"question": "Which studies suggest that the use of visual descriptors could lead to possible privacy breaches?", "answer": ["Privacy-Preserving Image Features via Adversarial Affine Subspace\n Embeddings", "Analysis and Mitigations of Reverse Engineering Attacks on Local Feature\n Descriptors", "Privacy Preserving Image-Based Localization"], "answer_arxiv_id": ["2006.06634", "2105.03812", "1903.05572"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_17151"} +{"question": "What works focus on improving the training approach of dense retrieval models?", "answer": ["Dense Passage Retrieval for Open-Domain Question Answering", "Approximate Nearest Neighbor Negative Contrastive Learning for Dense\n Text Retrieval", "RocketQA: An Optimized Training Approach to Dense Passage Retrieval for\n Open-Domain Question Answering"], "answer_arxiv_id": ["2004.04906", "2007.00808", "2010.08191"], "source_meta": {"published_time": "20230512"}, "qid": "AutoScholarQuery_train_17152"} +{"question": "Which papers discuss optimization and greedy approaches that depend on hand-crafted features to solve 2D jigsaw puzzles?", "answer": ["Solving Jigsaw Puzzles with Linear Programming", "Solving Jigsaw Puzzles By the Graph Connection Laplacian"], "answer_arxiv_id": ["1511.04472", "1811.03188v5"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_17153"} +{"question": "Are there any studies that proposed and analyzed the noisy gradient descent ascent flow for finding the MNE in minmax optimization?", "answer": ["A mean-field analysis of two-player zero-sum games"], "answer_arxiv_id": ["2002.06277"], "source_meta": {"published_time": "20221217"}, "qid": "AutoScholarQuery_train_17154"} +{"question": "Can you give me an example of a research work that has explored the potential of the traditional GAN framework?", "answer": ["Scaling up GANs for Text-to-Image Synthesis"], "answer_arxiv_id": ["2303.05511"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_17155"} +{"question": "What research developed frameworks to analyze the low-norm interpolator?", "answer": ["On Uniform Convergence and Low-Norm Interpolation Learning", "In Defense of Uniform Convergence: Generalization via derandomization with an application to interpolating predictors", "Uniform Convergence of Interpolators: Gaussian Width, Norm Bounds and Benign Overfitting"], "answer_arxiv_id": ["2006.05942", "1912.04265", "2106.09276"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_17156"} +{"question": "What works detail approximations for the bisimulation metric?", "answer": ["MICo: Improved representations via sampling-based state similarity for Markov decision processes"], "answer_arxiv_id": ["2106.08229"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_17157"} +{"question": "What are the studies that have used diffusion models for unconditional image inpainting?", "answer": ["Palette: Image-to-Image Diffusion Models", "RePaint: Inpainting using Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Come-Closer-Diffuse-Faster: Accelerating Conditional Diffusion Models for Inverse Problems through Stochastic Contraction", "Towards Coherent Image Inpainting Using Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2111.05826", "2201.09865", "2112.10752", "2112.05146", "2304.03322v1"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17158"} +{"question": "Which research papers have worked on the private multi-armed bandit problem?", "answer": ["Achieving Privacy in the Adversarial Multi-Armed Bandit", "Multi-Armed Bandits with Local Differential Privacy", "(Locally) Differentially Private Combinatorial Semi-Bandits", "Local Differential Privacy for Bayesian Optimization", "No-Regret Algorithms for Private Gaussian Process Bandit Optimization"], "answer_arxiv_id": ["1701.04222", "2007.03121", "2006.00706", "2010.06709", "2102.12467v1"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_17159"} +{"question": "In what studies do the algorithms estimate the inverse probability of selecting actions under the current policy using previously collected contexts?", "answer": ["Efficient Optimal Learning for Contextual Bandits", "Contextual Bandit Learning with Predictable Rewards", "Taming the Monster: A Fast and Simple Algorithm for Contextual Bandits"], "answer_arxiv_id": ["1106.2369", "1202.1334v2", "1402.0555"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_17160"} +{"question": "Can you find any studies that proposed to pre-train on distantly supervised one-hop paragraph-level commonsense inferences?", "answer": ["Paragraph-level Commonsense Transformers with Recurrent Memory"], "answer_arxiv_id": ["2010.01486"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_17161"} +{"question": "What works proposed optimization-based, filter-based, or learning-based solutions for motion estimation with event cameras?", "answer": ["IDOL: A Framework for IMU-DVS Odometry using Lines", "Continuous-Time Visual-Inertial Odometry for Event Cameras", "Event-based Vision meets Deep Learning on Steering Prediction for\n Self-driving Cars", "Event-Based Angular Velocity Regression with Spiking Networks"], "answer_arxiv_id": ["2008.05749", "1702.07389v2", "1804.01310", "2003.02790"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_17162"} +{"question": "Which references introduce a novel approach to use an LLM to assess the overall solution path?", "answer": ["ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world\n APIs"], "answer_arxiv_id": ["2307.16789"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_17163"} +{"question": "What studies on image style transfer in exemplar-guided image editing exist in the literature?", "answer": ["Cross-domain Correspondence Learning for Exemplar-based Image Translation"], "answer_arxiv_id": ["2004.05571"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_17164"} +{"question": "What research proposed a grouping loss to evaluate subgroup calibration error?", "answer": ["Beyond calibration: estimating the grouping loss of modern neural networks"], "answer_arxiv_id": ["2210.16315v3"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_17165"} +{"question": "Any works about improving geometry representation and proposing a more accurate material estimating pipeline using Image-Based Lighting?", "answer": ["Extracting Triangular 3D Models, Materials, and Lighting From Images", "Shape, Light, and Material Decomposition from Images using Monte Carlo\n Rendering and Denoising", "NeILF: Neural Incident Light Field for Physically-based Material\n Estimation"], "answer_arxiv_id": ["2111.12503", "2206.03380", "2203.07182"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17166"} +{"question": "Which work introduced a differentiable PnP layer for 6D pose estimation?", "answer": ["Coupled Iterative Refinement for 6D Multi-Object Pose Estimation"], "answer_arxiv_id": ["2204.12516"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_17167"} +{"question": "What works explored the idea of shared representation in the context of continual learning?", "answer": ["Provable and Efficient Continual Representation Learning", "Provable Lifelong Learning of Representations"], "answer_arxiv_id": ["2203.02026", "2110.14098"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_17168"} +{"question": "Which research improved Performance Fairness by incorporating a penalty term?", "answer": ["Ditto: Fair and Robust Federated Learning Through Personalization"], "answer_arxiv_id": ["2012.04221"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_17169"} +{"question": "Which paper proposed the use of mixture densities and introduced depth estimation as a proxy objective in Neural Radiance Fields models?", "answer": ["MixNeRF: Modeling a Ray with Mixture Density for Novel View Synthesis\n from Sparse Inputs"], "answer_arxiv_id": ["2302.08788"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_17170"} +{"question": "Any works discussing different triggers of backdoor attacks used to activate the backdoor effect, such as single pixel, reflection background, invisible patterns?", "answer": ["Spectral Signatures in Backdoor Attacks", "Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks", "Invisible Backdoor Attacks on Deep Neural Networks via Steganography and\n Regularization", "Hidden Trigger Backdoor Attacks"], "answer_arxiv_id": ["1811.00636", "2007.02343", "1909.02742", "1910.00033"], "source_meta": {"published_time": "20240430"}, "qid": "AutoScholarQuery_train_17171"} +{"question": "What works employed end-to-end models in chart-to-summary generation?", "answer": ["Chart-to-Text: Generating Natural Language Descriptions for Charts by\n Adapting the Transformer Model"], "answer_arxiv_id": ["2010.09142"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_17172"} +{"question": "What are the works that initiated the study of predictions in online algorithms?", "answer": ["Competitive caching with machine learned advice"], "answer_arxiv_id": ["1802.05399"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17173"} +{"question": "Which papers report on non-meta-learning methods that utilize supervised or unsupervised representation learning methods for few-shot image classification?", "answer": ["A Closer Look at Few-shot Classification", "Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?", "A Baseline for Few-Shot Image Classification", "Learning a Universal Template for Few-shot Dataset Generalization", "Universal Representation Learning from Multiple Domains for Few-shot Classification", "Exploring Complementary Strengths of Invariant and Equivariant Representations for Few-Shot Learning", "CrossTransformers: spatially-aware few-shot transfer", "Pushing the Limits of Simple Pipelines for Few-Shot Learning: External Data and Fine-Tuning Make a Difference", "Exploring Efficient Few-shot Adaptation for Vision Transformers"], "answer_arxiv_id": ["1904.04232", "2003.11539", "1909.02729", "2105.07029v2", "2103.13841", "2103.01315", "2007.11498", "2204.07305v1", "2301.02419"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_17174"} +{"question": "Which papers provide comprehensive reviews on the use of language in robotics planning tasks?", "answer": ["A Survey of Reinforcement Learning Informed by Natural Language"], "answer_arxiv_id": ["1906.03926"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17175"} +{"question": "Could you mention studies that propose neural methods for unbalanced OT?", "answer": ["Scalable Unbalanced Optimal Transport using Generative Adversarial Networks", "Neural Unbalanced Optimal Transport via Cycle-Consistent Semi-Couplings"], "answer_arxiv_id": ["1810.11447", "2209.15621"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_17176"} +{"question": "In which research papers is the concept of pseudo-counts used for exploration?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "Count-Based Exploration with the Successor Representation"], "answer_arxiv_id": ["1606.01868", "1807.11622"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_17177"} +{"question": "Which papers proposed Mask-reconstruction pretraining as an SSL approach?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language"], "answer_arxiv_id": ["2111.06377", "2202.03555"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_17178"} +{"question": "What studies involve deep learning approaches in VEA?", "answer": ["Learning Multi-level Deep Representations for Image Emotion\n Classification", "Stimuli-Aware Visual Emotion Analysis", "SOLVER: Scene-Object Interrelated Visual Emotion Reasoning Network"], "answer_arxiv_id": ["1611.07145", "2109.01812", "2110.12334"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_17179"} +{"question": "Which studies focused on developing the Dynamic Embedded Topic Model?", "answer": ["The Dynamic Embedded Topic Model"], "answer_arxiv_id": ["1907.05545"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_17180"} +{"question": "Which methods use supervision from natural language inference and labeled query-document pairs to train text embeddings?", "answer": ["A large annotated corpus for learning natural language inference", "MS MARCO: A Human Generated MAchine Reading COmprehension Dataset", "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", "Supervised Learning of Universal Sentence Representations from Natural\n Language Inference Data", "SimCSE: Simple Contrastive Learning of Sentence Embeddings"], "answer_arxiv_id": ["1508.05326", "1611.09268", "1908.10084", "1705.02364", "2104.08821"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_17181"} +{"question": "Can you give some references that introduced new activation functions and empirically studied their performance?", "answer": ["Improved Adversarial Robustness by Reducing Open Space Risk via Tent Activations", "Soft-Root-Sign Activation Function", "PLU: The Piecewise Linear Unit Activation Function"], "answer_arxiv_id": ["1908.02435", "2003.00547", "1809.09534"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_17182"} +{"question": "Could you show me some studies that develop contextualized language models?", "answer": ["Deep contextualized word representations", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "XLNet: Generalized Autoregressive Pretraining for Language Understanding", "ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators"], "answer_arxiv_id": ["1802.05365", "1810.04805", "1907.11692", "1906.08237", "2003.10555"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_17183"} +{"question": "Which works aim to match the sentiment of articles about the economy to both consumer sentiment and economic indicators?", "answer": ["News-based Business Sentiment and its Properties as an Economic Index"], "answer_arxiv_id": ["2110.10340"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_17184"} +{"question": "Which papers focus on domain adaptation in the field of egocentric vision?", "answer": ["Multi-Modal Domain Adaptation for Fine-Grained Action Recognition", "Temporal Attentive Alignment for Large-Scale Video Domain Adaptation", "What can a cook in Italy teach a mechanic in India? Action Recognition\n Generalisation Over Scenarios and Locations"], "answer_arxiv_id": ["2001.09691", "1907.12743", "2306.08713"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_17185"} +{"question": "What researches conducted one-step analysis in the framework of dynamic fairness?", "answer": ["Delayed Impact of Fair Machine Learning", "Downstream Effects of Affirmative Action", "From Fair Decision Making to Social Equality", "Group Retention when Using Machine Learning in Sequential Decision Making: the Interplay between User Dynamics and Fairness"], "answer_arxiv_id": ["1803.04383", "1808.09004", "1812.02952", "1905.00569"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_17186"} +{"question": "What studies propose methods for 3D SSL via multi-modal and multi-view settings?", "answer": ["DetMatch: Two Teachers are Better Than One for Joint 2D and 3D Semi-Supervised Object Detection"], "answer_arxiv_id": ["2203.09510"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_17187"} +{"question": "Which studies involve the use of behavior cloning frequently in real-world applications?", "answer": ["Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperation", "Reinforcement and Imitation Learning for Diverse Visuomotor Skills", "Vision-Based Multi-Task Manipulation for Inexpensive Robots Using End-To-End Learning from Demonstration", "Self-Supervised Correspondence in Visuomotor Policy Learning", "Transporter Networks: Rearranging the Visual World for Robotic Manipulation"], "answer_arxiv_id": ["1710.04615v2", "1802.09564", "1707.02920", "1909.06933", "2010.14406"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17188"} +{"question": "Could you provide studies on the usage of the Diffusion Models for image generation in video creation?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding"], "answer_arxiv_id": ["2204.06125", "2112.10752", "2112.10741", "2205.11487"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_17189"} +{"question": "What studies are about using large language models to build generic text quality scorers?", "answer": ["GPTScore: Evaluate as You Desire", "Large Language Models are Effective Text Rankers with Pairwise Ranking\n Prompting", "AlpaGasus: Training A Better Alpaca with Fewer Data"], "answer_arxiv_id": ["2302.04166", "2306.17563", "2307.08701"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_17190"} +{"question": "Can you list some papers that investigated the ability of language models on arithmetic tasks?", "answer": ["Measuring Vision-Language STEM Skills of Neural Models", "Goat: Fine-tuned LLaMA Outperforms GPT-4 on Arithmetic Tasks", "Show Your Work: Scratchpads for Intermediate Computation with Language\n Models", "Investigating the Limitations of Transformers with Simple Arithmetic\n Tasks", "From Interpolation to Extrapolation: Complete Length Generalization for\n Arithmetic Transformers"], "answer_arxiv_id": ["2402.17205", "2305.14201", "2112.00114", "2102.13019", "2310.11984"], "source_meta": {"published_time": "20240705"}, "qid": "AutoScholarQuery_train_17191"} +{"question": "Which study examined the lipschitzness of deep neural networks and developed an algorithm to upper-bound their Lipschitz constant?", "answer": ["Lipschitz regularity of deep neural networks: analysis and efficient estimation"], "answer_arxiv_id": ["1805.10965"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_17192"} +{"question": "Which work is considered the root of the field of distribution-free uncertainty quantification?", "answer": ["A Tutorial on Conformal Prediction"], "answer_arxiv_id": ["0706.3188"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_17193"} +{"question": "Which works propose asynchronous algorithms that require a shared memory/oracle from which clients grab the most up-to-date global parameters?", "answer": ["Hogwild!: A Lock-Free Approach to Parallelizing Stochastic Gradient Descent", "Asynchronous Parallel Stochastic Gradient for Nonconvex Optimization"], "answer_arxiv_id": ["1106.5730", "1506.08272v5"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_17194"} +{"question": "What studies concern about the application of DRO in the field of domain generalization?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the\n Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17195"} +{"question": "Could you give me examples of research that focused on communication uncertainty in federated learning?", "answer": ["Decentralized Federated Learning with Unreliable Communications"], "answer_arxiv_id": ["2108.02397"], "source_meta": {"published_time": "20221025"}, "qid": "AutoScholarQuery_train_17196"} +{"question": "What papers propose a variant of the traditional sequence model, namely pointer network, for combinatorial optimization?", "answer": ["Pointer Networks"], "answer_arxiv_id": ["1506.03134"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_17197"} +{"question": "Which works first studied neural architectures for extrapolation?", "answer": ["Neural GPUs Learn Algorithms"], "answer_arxiv_id": ["1511.08228"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_17198"} +{"question": "What works illustrate Maximum-Entropy RL as an example of CPO with a uniform policy distribution?", "answer": ["Reinforcement Learning with Deep Energy-Based Policies", "Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor"], "answer_arxiv_id": ["1702.08165", "1801.01290"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17199"} +{"question": "Which works studied or implemented IWERM with a misspecified model class or deep neural network?", "answer": ["Rethinking Importance Weighting for Deep Learning under Distribution Shift"], "answer_arxiv_id": ["2006.04662"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_17200"} +{"question": "What research papers have explored the use of more complex sub-shapes for shape assembly?", "answer": ["Learning elementary structures for 3D shape generation and matching", "CvxNet: Learnable Convex Decomposition", "Neural Star Domain as Primitive Representation", "ParSeNet: A Parametric Surface Fitting Network for 3D Point Clouds", "Physical Primitive Decomposition", "3D-PRNN: Generating Shape Primitives with Recurrent Neural Networks", "Discovering 3D Parts from Image Collections", "Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networks", "Learning 3D Part Assembly from a Single Image", "Composite Shape Modeling via Latent Space Factorization"], "answer_arxiv_id": ["1908.04725", "1909.05736", "2010.11248", "2003.12181", "1809.05070", "1708.01648", "2107.13629", "2103.10429", "2003.09754", "1901.02968"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_17201"} +{"question": "Which paper introduced the concept of Backwards Compatible Training (BCT)?", "answer": ["Towards Backward-Compatible Representation Learning"], "answer_arxiv_id": ["2003.11942"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_17202"} +{"question": "Can you provide a paper that addresses the issue of identifying independent latent factors in unsupervised learning?", "answer": ["Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations"], "answer_arxiv_id": ["1811.12359"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_17203"} +{"question": "Which works utilized a domain-based self-attention solver to reduce transformer complexity in image deblurring?", "answer": ["Efficient Frequency Domain-based Transformers for High-Quality Image\n Deblurring"], "answer_arxiv_id": ["2211.12250"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_17204"} +{"question": "Could you provide me with some studies on Conditional GANs incorporating segmentation maps as conditions?", "answer": ["Contrastive Learning for Unpaired Image-to-Image Translation", "Exploring Patch-wise Semantic Relation for Contrastive Learning in\n Image-to-Image Translation Tasks"], "answer_arxiv_id": ["2007.15651", "2203.01532"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_17205"} +{"question": "Could you provide me some studies about semantic augmentation in relation to Interpolation Regularization and FixMatch?", "answer": ["DistractFlow: Improving Optical Flow Estimation via Realistic\n Distractions and Pseudo-Labeling", "mixup: Beyond Empirical Risk Minimization", "Interpolation Consistency Training for Semi-Supervised Learning", "Interpolation-based semi-supervised learning for object detection", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["2303.14078", "1710.09412", "1903.03825v5", "2006.02158", "2001.07685v2"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17206"} +{"question": "Which work introduced a way to improve the Stable Diffusion model by using consistency trees or scene graphs to enhance the embedding learning of the prompt?", "answer": ["Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesis"], "answer_arxiv_id": ["2212.05032"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_17207"} +{"question": "What studies identified the causes and provided solutions for the issue of negative transfer in multi-task learning?", "answer": ["Gradient Surgery for Multi-Task Learning", "Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout", "Multi-Task Learning as Multi-Objective Optimization", "Multi-Task Learning as a Bargaining Game", "Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics", "GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks"], "answer_arxiv_id": ["2001.06782", "2010.06808", "1810.04650", "2202.01017", "1705.07115", "1711.02257"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17208"} +{"question": "Have there been any works that used binary cross-entropy for link prediction in graph learning?", "answer": ["Hyperbolic Graph Convolutional Neural Networks", "Learning Continuous Hierarchies in the Lorentz Model of Hyperbolic Geometry", "Hyperbolic Neural Networks++"], "answer_arxiv_id": ["1910.12933", "1806.03417", "2006.08210"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_17209"} +{"question": "Which works focused on speech voice conversion and aim to modify the timbre without changing the speech content?", "answer": ["Diffusion-Based Voice Conversion with Fast Maximum Likelihood Sampling Scheme", "Unsupervised Cross-Domain Singing Voice Conversion"], "answer_arxiv_id": ["2109.13821", "2008.02830"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_17210"} +{"question": "What research introduce the masked language modeling objective from BERT to non-autoregressively predict and refine translations?", "answer": ["Mask-Predict: Parallel Decoding of Conditional Masked Language Models"], "answer_arxiv_id": ["1904.09324"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_17211"} +{"question": "What are the works that expanded the network over time in continual learning for maintaining past performance?", "answer": ["Progressive Neural Networks"], "answer_arxiv_id": ["1606.04671"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_17212"} +{"question": "What are the studies that rely on heterogeneous data and distribution shifts in causal representation learning?", "answer": ["Learning Temporally Causal Latent Processes from General Temporal Data", "Temporally Disentangled Representation Learning", "Identifying Weight-Variant Latent Causal Models"], "answer_arxiv_id": ["2110.05428", "2210.13647", "2208.14153"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17213"} +{"question": "What papers provide solutions to alleviate catastrophic forgetting through regularization-based approaches?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Memory Aware Synapses: Learning what (not) to forget", "Variational Continual Learning", "Continual Learning Through Synaptic Intelligence"], "answer_arxiv_id": ["1612.00796", "1711.09601", "1710.10628", "1703.04200"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_17214"} +{"question": "In Inverse RL, what paper discusses learning a stationary reward function?", "answer": ["f-IRL: Inverse Reinforcement Learning via State Marginal Matching"], "answer_arxiv_id": ["2011.04709v2"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17215"} +{"question": "Which studies investigate embodied artificial intelligence enabling agents to learn through interactions?", "answer": ["Object Goal Navigation using Goal-Oriented Semantic Exploration", "Continuous Scene Representations for Embodied AI", "Interactron: Embodied Adaptive Object Detection", "Embodied Visual Active Learning for Semantic Segmentation"], "answer_arxiv_id": ["2007.00643", "2203.17251", "2202.00660", "2012.09503"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_17216"} +{"question": "Which paper first motivated the 'adaptive rounding' proxy objective?", "answer": ["A Systematic Classification of Knowledge, Reasoning, and Context within the ARC Dataset"], "answer_arxiv_id": ["1806.00358"], "source_meta": {"published_time": "20230725"}, "qid": "AutoScholarQuery_train_17217"} +{"question": "Can you mention some works that focused on consistency regularization?", "answer": ["Temporal Ensembling for Semi-Supervised Learning", "Regularization With Stochastic Transformations and Perturbations for\n Deep Semi-Supervised Learning", "Mean teachers are better role models: Weight-averaged consistency\n targets improve semi-supervised deep learning results", "Progressive Semantic-Visual Mutual Adaption for Generalized Zero-Shot\n Learning"], "answer_arxiv_id": ["1610.02242v3", "1606.04586", "1703.01780", "2303.15322"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_17218"} +{"question": "Which papers were included in the research about PCA-Net in the infinite-rank setting?", "answer": ["Model Reduction And Neural Networks For Parametric PDEs", "The Cost-Accuracy Trade-Off In Operator Learning With Neural Networks"], "answer_arxiv_id": ["2005.03180", "2203.13181"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17219"} +{"question": "What research has proposed methods for training an RL policy to be robust against an adversarial environment?", "answer": ["Robust Adversarial Reinforcement Learning"], "answer_arxiv_id": ["1703.02702"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_17220"} +{"question": "Could you provide me examples of studies that have used logographic data that has been transcribed into Latin or annotated with extra semantic information?", "answer": ["Finding structure in logographic writing with library learning"], "answer_arxiv_id": ["2405.06906v1"], "source_meta": {"published_time": "20240808"}, "qid": "AutoScholarQuery_train_17221"} +{"question": "What papers expanded the application of generative adversarial models to 3D content generation using neural rendering techniques?", "answer": ["GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "HoloGAN: Unsupervised Learning of 3D Representations From Natural Images", "GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields", "BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled Images", "GIRAFFE HD: A High-Resolution 3D-aware Generative Model"], "answer_arxiv_id": ["2007.02442", "1904.01326", "2011.12100", "2002.08988", "2203.14954"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_17222"} +{"question": "Which works introduced the notation of load in the Hopfield model of associative memory?", "answer": ["About the ergodic regime in the analogical Hopfield neural networks. Moments of the partition function", "On the equivalence of Hopfield Networks and Boltzmann Machines"], "answer_arxiv_id": ["0911.3515", "1105.2790"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_17223"} +{"question": "What are some of the novel techniques used for face recognition tasks?", "answer": ["CosFace: Large Margin Cosine Loss for Deep Face Recognition", "ArcFace: Additive Angular Margin Loss for Deep Face Recognition", "MagFace: A Universal Representation for Face Recognition and Quality Assessment"], "answer_arxiv_id": ["1801.09414", "1801.07698", "2103.06627"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_17224"} +{"question": "What works focus on integrating ViTs and ConvNets?", "answer": ["Rethinking Spatial Dimensions of Vision Transformers", "MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer", "Mobile-Former: Bridging MobileNet and Transformer"], "answer_arxiv_id": ["2103.16302", "2110.02178", "2108.05895"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_17225"} +{"question": "What works indicate that large language models (LLMs) can generate toxic outputs?", "answer": ["RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models"], "answer_arxiv_id": ["2009.11462v2"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_17226"} +{"question": "What papers underscore the importance of introducing diffusion noise?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2011.13456", "2006.11239"], "source_meta": {"published_time": "20220907"}, "qid": "AutoScholarQuery_train_17227"} +{"question": "What study used the model’s initially trained capabilities to generate additional training data?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2201.12086"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_17228"} +{"question": "Which works have achieved state-of-the-art performance in image or video captioning with large-scale multimodal models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Microsoft COCO: Common Objects in Context"], "answer_arxiv_id": ["2204.14198", "1405.0312"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_17229"} +{"question": "What papers deal with the minimization of the discrepancy between the sample distributions of S and T in domain alignment?", "answer": ["Deep Transfer Learning with Joint Adaptation Networks", "Normalized Wasserstein for Mixture Distributions with Applications in Adversarial Learning and Domain Adaptation", "Domain Conditioned Adaptation Network"], "answer_arxiv_id": ["1605.06636", "1902.00415", "2005.06717"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_17230"} +{"question": "Any works exploring multi-task learning in NeRF?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene Representation"], "answer_arxiv_id": ["2103.15875"], "source_meta": {"published_time": "20220919"}, "qid": "AutoScholarQuery_train_17231"} +{"question": "What are the preceding studies that explored text-driven image synthesis, on which AYG builds?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "ERNIE-ViLG 2.0: Improving Text-to-Image Diffusion Model with\n Knowledge-Enhanced Mixture-of-Denoising-Experts", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "Emu: Enhancing Image Generation Models Using Photogenic Needles in a\n Haystack", "RAPHAEL: Text-to-Image Generation via Large Mixture of Diffusion Paths"], "answer_arxiv_id": ["2205.11487", "2112.10752", "2204.06125", "2210.15257", "2307.01952", "2309.15807", "2305.18295"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_17232"} +{"question": "What is the work that utilizes the joint representation space of visual and textual features in the pre-trained multimodal CLIP model to learn the feature shift between the visual and textual descriptions of the target domain?", "answer": ["CLIP the Gap: A Single Domain Generalization Approach for Object\n Detection"], "answer_arxiv_id": ["2301.05499"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_17233"} +{"question": "What works use GAN inversion for style transfer?", "answer": ["VToonify: Controllable High-Resolution Portrait Video Style Transfer", "Pastiche Master: Exemplar-Based High-Resolution Portrait Style Transfer"], "answer_arxiv_id": ["2209.11224", "2203.13248"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_17234"} +{"question": "Can you list some papers regarding contrastive learning or architectural variants as methods to avoid representational collapse?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "BYOL works even without batch statistics", "Exploring Simple Siamese Representation Learning"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2006.07733", "2010.10241", "2011.10566"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_17235"} +{"question": "What works use iterative differentiation in gradient-based meta-learning algorithms?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Forward and Reverse Gradient-Based Hyperparameter Optimization", "Bilevel Programming for Hyperparameter Optimization and Meta-Learning", "Gradient-based Hyperparameter Optimization through Reversible Learning"], "answer_arxiv_id": ["1703.03400", "1703.01785", "1806.04910", "1502.03492"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_17236"} +{"question": "Could you provide me the papers which focused on providing guarantees for adversarial robustness?", "answer": ["Intriguing properties of neural networks", "Evasion attacks against machine learning at test time", "Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1312.6199", "1708.06131", "1412.6572", "1706.06083", "1608.04644"], "source_meta": {"published_time": "20230414"}, "qid": "AutoScholarQuery_train_17237"} +{"question": "Which early methods for text to image (T2I) generation are based on GANs?", "answer": ["Generative Adversarial Text to Image Synthesis", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "StackGAN: Text to Photo-realistic Image Synthesis with Stacked\n Generative Adversarial Networks", "Cross-Modal Contrastive Learning for Text-to-Image Generation", "StackGAN++: Realistic Image Synthesis with Stacked Generative\n Adversarial Networks"], "answer_arxiv_id": ["1605.05396", "1711.10485", "1612.03242", "2101.04702", "1710.10916"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_17238"} +{"question": "What research papers include the study of generative method and the progressive approach to image inpainting?", "answer": ["Progressive Image Inpainting with Full-Resolution Residual Network", "Recurrent Feature Reasoning for Image Inpainting", "High-Resolution Image Inpainting with Iterative Confidence Feedback and\n Guided Upsampling"], "answer_arxiv_id": ["1907.10478", "2008.03737", "2005.11742"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_17239"} +{"question": "What works utilize shared or regularized underlying representation of repetitive primitives or objects to constrain and reconstruct their 3D geometry?", "answer": ["Reconstructing 3D Human Pose by Watching Humans in the Mirror"], "answer_arxiv_id": ["2104.00340"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_17240"} +{"question": "Could you provide me a work that proposed emulating disalignment through sampling?", "answer": ["An Emulator for Fine-Tuning Large Language Models using Small Language\n Models"], "answer_arxiv_id": ["2310.12962"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_17241"} +{"question": "Which researches on T2I involve encoding spatial constraints via an extra condition module?", "answer": ["Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "High-Resolution Image Synthesis with Latent Diffusion Models", "Modeling Image Composition for Complex Scene Generation", "Frido: Feature Pyramid Diffusion for Complex Scene Image Synthesis", "ReCo: Region-Controlled Text-to-Image Generation", "GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2203.13131", "2112.10752", "2206.00923", "2208.13753v2", "2211.15518", "2301.07093"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_17242"} +{"question": "Which research papers discussed the use of Lp Norm and entropy regularization in handling overfitting and overconfidence issues in deep neural networks?", "answer": ["Revisiting Explicit Regularization in Neural Networks for Well-Calibrated Predictive Uncertainty", "Regularizing Neural Networks by Penalizing Confident Output Distributions"], "answer_arxiv_id": ["2006.06399", "1701.06548"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_17243"} +{"question": "What studies discuss the use of Direct Method and how it suffers from high bias when the reward model is misspecified?", "answer": ["Empirical Study of Off-Policy Policy Evaluation for Reinforcement Learning"], "answer_arxiv_id": ["1911.06854"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_17244"} +{"question": "Which work utilizes an explainable subgraph and attention mechanism for predictions?", "answer": ["xERTE: Explainable Reasoning on Temporal Knowledge Graphs for\n Forecasting Future Links"], "answer_arxiv_id": ["2012.15537"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_17245"} +{"question": "Which research papers focused on diversifying in-prompt examples to enhance generalization in transformers?", "answer": ["Diverse Demonstrations Improve In-context Compositional Generalization"], "answer_arxiv_id": ["2212.06800"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_17246"} +{"question": "Could you provide me some studies that leverage existing word alignment datasets to train high performing word aligners?", "answer": ["A Supervised Word Alignment Method based on Cross-Language Span\n Prediction using Multilingual BERT", "WSPAlign: Word Alignment Pre-training via Large-Scale Weakly Supervised\n Span Prediction"], "answer_arxiv_id": ["2004.14516", "2306.05644"], "source_meta": {"published_time": "20240716"}, "qid": "AutoScholarQuery_train_17247"} +{"question": "What research papers are about applying GANs in text-to-image generative models?", "answer": ["VQGAN-CLIP: Open Domain Image Generation and Editing with Natural Language Guidance"], "answer_arxiv_id": ["2204.08583"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_17248"} +{"question": "Which works introduced the concept of causal Bandit Optimization focusing on hard interventions?", "answer": ["Causal Bayesian Optimization"], "answer_arxiv_id": ["2005.11741v2"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_17249"} +{"question": "Which papers aimed at boosting the vanilla Vision Transformer by scaling?", "answer": ["Going deeper with Image Transformers", "Scaling Vision Transformers"], "answer_arxiv_id": ["2103.17239", "2106.04560"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17250"} +{"question": "Which works use compatibility between model features and task labels for transferability estimation?", "answer": ["LogME: Practical Assessment of Pre-trained Models for Transfer Learning", "Frustratingly Easy Transferability Estimation", "PACTran: PAC-Bayesian Metrics for Estimating the Transferability of Pretrained Models to Classification Tasks"], "answer_arxiv_id": ["2102.11005", "2106.09362", "2203.05126"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_17251"} +{"question": "Any works about the concept of using a communication game with a discrete bottleneck?", "answer": ["Interpretable agent communication from scratch (with a generic visual processor emerging on the side)"], "answer_arxiv_id": ["2106.04258"], "source_meta": {"published_time": "20220401"}, "qid": "AutoScholarQuery_train_17252"} +{"question": "What studies focus on addressing the encoding discontinuity by handling the Periodicity of Angular (PoA)?", "answer": ["Dense Label Encoding for Boundary Discontinuity Free Rotation Detection", "Multi-Grained Angle Representation for Remote Sensing Object Detection", "ARS-DETR: Aspect Ratio-Sensitive Detection Transformer for Aerial\n Oriented Object Detection", "Phase-Shifting Coder: Predicting Accurate Orientation in Oriented Object\n Detection", "Rethinking Boundary Discontinuity Problem for Oriented Object Detection"], "answer_arxiv_id": ["2011.09670", "2209.02884", "2303.04989", "2211.06368", "2305.10061"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_17253"} +{"question": "Could you provide me the research that examines the performance of cvHM on Markovian GP baselines?", "answer": ["State Space Expectation Propagation: Efficient Inference Schemes for Temporal Gaussian Processes"], "answer_arxiv_id": ["2007.05994v1"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17254"} +{"question": "Which previous works studied risk-sensitive reinforcement learning in single-agent settings?", "answer": ["Algorithms for CVaR Optimization in MDPs", "Risk-Sensitive and Robust Decision-Making: a CVaR Optimization Approach", "Risk-Sensitive Reinforcement Learning: Near-Optimal Risk-Sample Tradeoff in Regret", "Parametric Return Density Estimation for Reinforcement Learning", "Learning Robust Options by Conditional Value at Risk Optimization"], "answer_arxiv_id": ["1406.3339", "1506.02188", "2006.13827", "1203.3497v1", "1905.09191"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_17255"} +{"question": "Could you give me studies applying diffusion models to the domain of molecules?", "answer": ["GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation", "Equivariant Diffusion for Molecule Generation in 3D"], "answer_arxiv_id": ["2203.02923", "2203.17003"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_17256"} +{"question": "Which papers introduced a CNN-based method for enhancing the quality of JPEG-compressed images?", "answer": ["Compression Artifacts Reduction by a Deep Convolutional Network"], "answer_arxiv_id": ["1504.06993"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17257"} +{"question": "What studies have utilized Reinforcement Learning (RL) to solve tasks with weak supervision?", "answer": ["Reinforcement Learning for Relation Classification from Noisy Data"], "answer_arxiv_id": ["1808.08013"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_17258"} +{"question": "In which study the PixelNeRF is extended by sampling colours from epipolar lines and processing densities with an MLP head for each point in the target ray?", "answer": ["Behind the Scenes: Density Fields for Single View Reconstruction"], "answer_arxiv_id": ["2301.07668"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_17259"} +{"question": "What papers proposed baseline approaches such as Maximum Softmax Probability and MaxLogit to handle overconfidence problem in deep neural networks?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples\n in Neural Networks", "Scaling Out-of-Distribution Detection for Real-World Settings"], "answer_arxiv_id": ["1610.02136", "1911.11132"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_17260"} +{"question": "What papers proposed methods for assigning hard/soft pseudo-labels to unlabeled images?", "answer": ["FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["2001.07685v2"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_17261"} +{"question": "Which work indicates that GPRS requires numerically solving a certain ODE?", "answer": ["Greedy Poisson Rejection Sampling"], "answer_arxiv_id": ["2305.15313"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_17262"} +{"question": "What papers discuss a low-rank effect in the weight matrices due to a small weight norm?", "answer": ["Implicit Regularization Towards Rank Minimization in ReLU Networks"], "answer_arxiv_id": ["2201.12760"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_17263"} +{"question": "Which work first discovered the barren plateau phenomenon in quantum neural networks?", "answer": ["Barren plateaus in quantum neural network training landscapes"], "answer_arxiv_id": ["1803.11173v1"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_17264"} +{"question": "Which papers are about unconstrained motion generation in the early research phase?", "answer": ["Structure-Aware Human-Action Generation", "Learning Diverse Stochastic Human-Action Generators by Learning Smooth\n Latent Transitions", "Character Controllers Using Motion VAEs"], "answer_arxiv_id": ["2007.01971", "1912.10150", "2103.14274"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_17265"} +{"question": "Could you provide me some studies that investigated the piecewise smooth functions in the context of Transformer networks?", "answer": ["Optimal approximation of piecewise smooth functions using deep ReLU neural networks", "Deep Neural Networks Learn Non-Smooth Functions Effectively"], "answer_arxiv_id": ["1709.05289", "1802.04474"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17266"} +{"question": "Which works have utilized StyleGAN for image manipulation with a text condition?", "answer": ["StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "HairCLIP: Design Your Hair by Text and Reference Image"], "answer_arxiv_id": ["2103.17249", "2108.00946", "2112.05142"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_17267"} +{"question": "Could you provide me some studies related to model compression techniques like distillation?", "answer": ["Distilling the Knowledge in a Neural Network", "Well-Read Students Learn Better: On the Importance of Pre-training Compact Models", "Autoregressive Knowledge Distillation through Imitation Learning"], "answer_arxiv_id": ["1503.02531", "1908.08962", "2009.07253"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_17268"} +{"question": "Which works developed text-based Large Language Models (LLMs)?", "answer": ["GPT-4 Technical Report", "PaLM: Scaling Language Modeling with Pathways", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2303.08774", "2204.02311", "2307.09288"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_17269"} +{"question": "What works pointed out the limitations of graph neural network (GNN) models when addressing network heterophily?", "answer": ["Convolutional Networks on Graphs for Learning Molecular Fingerprints", "Semi-Supervised Classification with Graph Convolutional Networks", "Inductive Representation Learning on Large Graphs"], "answer_arxiv_id": ["1509.09292", "1609.02907", "1706.02216"], "source_meta": {"published_time": "20211104"}, "qid": "AutoScholarQuery_train_17270"} +{"question": "Which works leverages reinforcement learning to pursue optimal bit-widths in mixed-precision quantization?", "answer": ["HAQ: Hardware-Aware Automated Quantization with Mixed Precision", "AutoQ: Automated Kernel-Wise Neural Network Quantization"], "answer_arxiv_id": ["1811.08886", "1902.05690"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17271"} +{"question": "Which papers propose the pseudo-fake generation approach to improve deepfake detection?", "answer": ["Face X-ray for More General Face Forgery Detection", "Detecting Deepfakes with Self-Blended Images", "AltFreezing for More General Video Face Forgery Detection"], "answer_arxiv_id": ["1912.13458", "2204.08376", "2307.08317"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_17272"} +{"question": "What studies use diffusion for image purification from out-of-domain data?", "answer": ["Diffusion Models for Adversarial Purification", "Back to the Source: Diffusion-Driven Test-Time Adaptation"], "answer_arxiv_id": ["2205.07460v1", "2207.03442"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_17273"} +{"question": "Can you provide me studies about semantic-aware learning methods used in part assembly tasks?", "answer": ["AutoMate: A Dataset and Learning Approach for Automatic Mating of CAD Assemblies", "JoinABLe: Learning Bottom-up Assembly of Parametric CAD Joints", "CompoNet: Learning to Generate the Unseen by Part Synthesis and Composition", "Learning 3D Part Assembly from a Single Image", "PQ-NET: A Generative Part Seq2Seq Network for 3D Shapes", "COALESCE: Component Assembly by Learning to Synthesize Connections", "Composite Shape Modeling via Latent Space Factorization", "Generative 3D Part Assembly via Dynamic Graph Learning", "RGL-NET: A Recurrent Graph Learning framework for Progressive Part Assembly"], "answer_arxiv_id": ["2105.12238", "2111.12772", "1811.07441", "2003.09754", "1911.10949", "2008.01936", "1901.02968", "2006.07793", "2107.12859"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_17274"} +{"question": "Can you provide any studies about the use of human feedback in semantic parsing?", "answer": ["Model-based Interactive Semantic Parsing: A Unified Framework and A Text-to-SQL Case Study"], "answer_arxiv_id": ["1910.05389"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_17275"} +{"question": "Which studies have created 'task embeddings'?", "answer": ["Task2Vec: Task Embedding for Meta-Learning"], "answer_arxiv_id": ["1902.03545"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_17276"} +{"question": "Which studies have attempted to leverage the prototypes to serve as a constraint on the feature alignment?", "answer": ["Deep Clustering for Unsupervised Learning of Visual Features", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Emerging Properties in Self-Supervised Vision Transformers", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "Combating Representation Learning Disparity with Geometric Harmonization"], "answer_arxiv_id": ["1807.05520", "2006.09882", "2104.14294", "2103.03230", "2310.17622"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_17277"} +{"question": "Could you provide me some studies about applying reinforcement learning in NLP tasks?", "answer": ["Tackling Error Propagation through Reinforcement Learning: A Case of Greedy Dependency Parsing", "Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation", "Reinforcement Learning based Curriculum Optimization for Neural Machine Translation", "A Deep Reinforced Model for Abstractive Summarization", "Deep Reinforcement Learning for Dialogue Generation", "Fine-Tuning Language Models from Human Preferences", "Recursively Summarizing Books with Human Feedback", "Quark: Controllable Text Generation with Reinforced [Un]learning", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["1702.06794", "1609.08144", "1903.00041", "1705.04304", "1606.01541", "1909.08593", "2109.10862", "2205.13636", "2203.02155"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_17278"} +{"question": "Which studies have combined machine learning with analytic physics simulators to correct analytic models for various real-world domains?", "answer": ["Augmenting Physical Simulators with Stochastic Neural Networks: Case Study of Planar Pushing and Bouncing", "TossingBot: Learning to Throw Arbitrary Objects with Residual Physics", "Learning Agile and Dynamic Motor Skills for Legged Robots", "NeuralSim: Augmenting Differentiable Simulators with Neural Networks"], "answer_arxiv_id": ["1808.03246", "1903.11239", "1901.08652", "2011.04217"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_17279"} +{"question": "What publication proposes a frequency-based fingerprint feature?", "answer": ["Watch your Up-Convolution: CNN Based Generative Deep Neural Networks are\n Failing to Reproduce Spectral Distributions"], "answer_arxiv_id": ["2003.01826"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_17280"} +{"question": "Which works contributed to the open-world generation application in subject-driven image generation?", "answer": ["GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2301.07093"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17281"} +{"question": "What are the examples of research papers related to Coupled Graph Neural Networks (GNNs)?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Inductive Representation Learning on Large Graphs", "Graph Attention Networks"], "answer_arxiv_id": ["1609.02907", "1706.02216", "1710.10903"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_17282"} +{"question": "Which work handled high-stakes privacy-critical applications of mixed-type tabular data?", "answer": ["Deep Neural Networks and Tabular Data: A Survey"], "answer_arxiv_id": ["2110.01889"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_17283"} +{"question": "Which research introduced 3D Gaussian splatting as an alternative to NeRF for 3D tasks?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_17284"} +{"question": "Are there any studies on solving phase retrieval via gradient descent?", "answer": ["Gradient Descent with Random Initialization: Fast Global Convergence for Nonconvex Phase Retrieval", "A Geometric Analysis of Phase Retrieval"], "answer_arxiv_id": ["1803.07726", "1602.06664"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17285"} +{"question": "What research papers are relating to the recently introduced class of score-based models termed diffusion Schrödinger bridges?", "answer": ["On the relation between optimal transport and Schrödinger bridges: A stochastic control viewpoint", "Diffusion Schrödinger Bridge with Applications to Score-Based Generative Modeling", "Entropic Neural Optimal Transport via Diffusion Processes"], "answer_arxiv_id": ["1412.4430", "2106.01357", "2211.01156"], "source_meta": {"published_time": "20220214"}, "qid": "AutoScholarQuery_train_17286"} +{"question": "Which works focused on studying commonsense reasoning aspect of reasoning ability in machine learning models?", "answer": ["SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense\n Inference", "CommonsenseQA: A Question Answering Challenge Targeting Commonsense\n Knowledge", "Abductive Commonsense Reasoning"], "answer_arxiv_id": ["1808.05326", "1811.00937", "1908.05739"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_17287"} +{"question": "What papers showed that there exist certain cryptography-inspired classification tasks such that learning a classifier requires solving the NP-hard Learning Parity with Noise problem?", "answer": ["Computational Limitations in Robust Classification and Win-Win Results", "Adversarial Examples from Cryptographic Pseudo-Random Generators", "Adversarial examples from computational constraints"], "answer_arxiv_id": ["1902.01086", "1811.06418", "1805.10204"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_17288"} +{"question": "What works propose harnessing unique characteristics inherent to satellite imagery such as time-series observations or geolocation meta-data?", "answer": ["DynamicEarthNet: Daily Multi-Spectral Satellite Dataset for Semantic\n Change Segmentation", "Geography-Aware Self-Supervised Learning"], "answer_arxiv_id": ["2203.12560", "2011.09980"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_17289"} +{"question": "Could you provide me some studies utilizing the mask image modeling for remote sensing data?", "answer": ["SatMAE: Pre-training Transformers for Temporal and Multi-Spectral\n Satellite Imagery"], "answer_arxiv_id": ["2207.08051"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_17290"} +{"question": "Which works proposed the representation of source code as a sequence of tokens?", "answer": ["A deep language model for software code", "Deep API Learning"], "answer_arxiv_id": ["1608.02715", "1605.08535"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17291"} +{"question": "Which research proposes to move the optimization to the smaller latent space of a generative model to satisfy the gradient matching objective?", "answer": ["Gradient Inversion with Generative Image Prior"], "answer_arxiv_id": ["2110.14962"], "source_meta": {"published_time": "20220912"}, "qid": "AutoScholarQuery_train_17292"} +{"question": "Which works have been undertaken to alleviate the overestimation bias in value function?", "answer": ["Addressing Function Approximation Error in Actor-Critic Methods", "Deep Reinforcement Learning with Double Q-learning", "Maxmin Q-learning: Controlling the Estimation Bias of Q-learning"], "answer_arxiv_id": ["1802.09477", "1509.06461", "2002.06487"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_17293"} +{"question": "Can you name the work which applies the principles of GPT-f to Lean and studies the benefits from co-training on self-supervised objectives?", "answer": ["Proof Artifact Co-training for Theorem Proving with Language Models"], "answer_arxiv_id": ["2102.06203"], "source_meta": {"published_time": "20220203"}, "qid": "AutoScholarQuery_train_17294"} +{"question": "Which studies presented visual and textual features fused via cross-modality fusion for Visual Question Answering Systems?", "answer": ["Beyond Bilinear: Generalized Multimodal Factorized High-order Pooling for Visual Question Answering", "Towards VQA Models That Can Read", "Deep Modular Co-Attention Networks for Visual Question Answering", "In Defense of Grid Features for Visual Question Answering", "Bilinear Graph Networks for Visual Question Answering"], "answer_arxiv_id": ["1708.03619", "1904.08920", "1906.10770", "2001.03615", "1907.09815"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_17295"} +{"question": "Which works discussed the incorporation of reinforcement learning tasks into the sequence-to-sequence framework?", "answer": ["A Generalist Agent"], "answer_arxiv_id": ["2205.06175"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17296"} +{"question": "What studies have applied diffusion methods in the field of motion generation?", "answer": ["Human Motion Diffusion Model", "FLAME: Free-form Language-based Motion Synthesis & Editing", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "EDGE: Editable Dance Generation From Music", "Executing your Commands via Motion Diffusion in Latent Space", "Priority-Centric Human Motion Generation in Discrete Latent Space", "DiverseMotion: Towards Diverse Human Motion Generation via Discrete\n Diffusion"], "answer_arxiv_id": ["2209.14916", "2209.00349", "2208.15001", "2211.10658", "2212.04048", "2308.14480", "2309.01372"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17297"} +{"question": "What work have been done in the field of Visual Question Answering that prompts textual answers to questions about images?", "answer": ["VQA: Visual Question Answering"], "answer_arxiv_id": ["1505.00468"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_17298"} +{"question": "Which research works adopted generative methods for alignment?", "answer": ["More Grounded Image Captioning by Distilling Image-Text Matching Model", "DenseCap: Fully Convolutional Localization Networks for Dense Captioning", "RegionCLIP: Region-based Language-Image Pretraining"], "answer_arxiv_id": ["2004.00390", "1511.07571", "2112.09106"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_17299"} +{"question": "Which papers proposed methods to enforce robustness by re-weighting losses on individual data points?", "answer": ["Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach", "Large-Scale Methods for Distributionally Robust Optimization", "Detecting and Correcting for Label Shift with Black Box Predictors", "Distributionally Robust Language Modeling"], "answer_arxiv_id": ["1610.03425", "2010.05893", "1802.03916", "1909.02060"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_17300"} +{"question": "Which papers proposed data synthesis methods based on GANs to generate tabular data?", "answer": ["Correlated discrete data generation using adversarial training", "Boosting Deep Learning Risk Prediction with Generative Adversarial Networks for Electronic Health Records", "Data Synthesis based on Generative Adversarial Networks", "Modeling Tabular Data using Conditional GAN", "OCT-GAN: Neural ODE-based Conditional Tabular GANs", "SOS: Score-based Oversampling for Tabular Data"], "answer_arxiv_id": ["1804.00925", "1709.01648", "1806.03384v5", "1907.00503", "2105.14969", "2206.08555"], "source_meta": {"published_time": "20221008"}, "qid": "AutoScholarQuery_train_17301"} +{"question": "Could you provide me with any studies on explicit and hybrid scene representations?", "answer": ["TensoRF: Tensorial Radiance Fields", "Plenoxels: Radiance Fields without Neural Networks", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Point-NeRF: Point-based Neural Radiance Fields", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "Neural Sparse Voxel Fields"], "answer_arxiv_id": ["2203.09517", "2112.05131", "2201.05989", "2201.08845", "2304.06706", "2111.11215", "2007.11571"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_17302"} +{"question": "What studies support prompts in both text and image for instance-level, i. e. , foreground objects, perception tasks?", "answer": ["Universal Instance Perception as Object Discovery and Retrieval"], "answer_arxiv_id": ["2303.06674"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_17303"} +{"question": "In the context of adversarial purification, which works propose the usage of the score-based match model to eliminate the influence of perturbation and recover clean images?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations", "Adversarial purification with Score-based generative models"], "answer_arxiv_id": ["2011.13456", "2106.06041"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_17304"} +{"question": "Is there any work that proposes a unifying framework under spectral manifold learning, which closely relates to this paper's focus?", "answer": ["Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods"], "answer_arxiv_id": ["2205.11508"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_17305"} +{"question": "Is there any research that embedded a 3D-aware inductive bias into a unified diffusion model architecture?", "answer": ["HoloDiffusion: Training a 3D Diffusion Model using 2D Images"], "answer_arxiv_id": ["2303.16509"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_17306"} +{"question": "Could you provide me some studies about different formulations of individual fairness in clustering?", "answer": ["A Center in Your Neighborhood: Fairness in Facility Location", "Distributional Individual Fairness in Clustering", "A Pairwise Fair and Community-preserving Approach to k-Center Clustering", "Feature-based Individual Fairness in k-clustering"], "answer_arxiv_id": ["1908.09041", "2006.12589", "2007.07384v1", "2109.04554"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_17307"} +{"question": "What are some studies that used sketching for the purpose of training neural networks?", "answer": ["Generalized Leverage Score Sampling for Neural Networks", "Training (Overparametrized) Neural Networks in Near-Linear Time", "Training Multi-Layer Over-Parametrized Neural Network in Subquadratic Time", "Training Overparametrized Neural Networks in Sublinear Time"], "answer_arxiv_id": ["2009.09829v1", "2006.11648", "2112.07628v2", "2208.04508"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_17308"} +{"question": "Which paper proposes the Segment Anything Model (SAM) that introduces spatial prompts for segmenting arbitrary objects?", "answer": ["Segment Anything"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_17309"} +{"question": "Which research papers have shifted their concentration towards training the model with deterministic logical constraints?", "answer": ["On the relation between Loss Functions and T-Norms", "Analyzing Differentiable Fuzzy Logic Operators", "Deep Learning with Logical Constraints"], "answer_arxiv_id": ["1907.07904v1", "2002.06100", "2205.00523"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_17310"} +{"question": "Which paper conducted a detailed survey on the research in data influence estimation?", "answer": ["Training Data Influence Analysis and Estimation: A Survey"], "answer_arxiv_id": ["2212.04612"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_17311"} +{"question": "Which works used AI algorithms trained on egocentric images from head-mounted cameras of human babies and children?", "answer": ["Self-supervised learning through the eyes of a child"], "answer_arxiv_id": ["2007.16189"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_17312"} +{"question": "Which studies have been made on unimodal reasoning tasks such as question answering?", "answer": ["NewsQA: A Machine Comprehension Dataset", "Coarse-to-Fine Question Answering for Long Documents", "Conversational Question Answering: A Survey"], "answer_arxiv_id": ["1611.09830", "1611.01839", "2106.00874v2"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_17313"} +{"question": "What studies discussed the distribution shift problem in multi-view Supervised Fine Tuning (SFT)?", "answer": ["Instant3D: Fast Text-to-3D with Sparse-View Generation and Large\n Reconstruction Model"], "answer_arxiv_id": ["2311.06214"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_17314"} +{"question": "Which works develop methods for the incorporation of additional conditions into existing text-to-image diffusion models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_17315"} +{"question": "Could you provide me studies about state entropy maximization?", "answer": ["Efficient Exploration via State Marginal Matching", "Reward-Free Exploration for Reinforcement Learning", "On Reward-Free Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1906.05274", "2002.02794", "2006.11274"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_17316"} +{"question": "Which papers have explored the theory of out-of-distribution (OOD) generalization in structured causal models (SCM)?", "answer": ["Invariant Risk Minimization", "Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization"], "answer_arxiv_id": ["1907.02893", "2106.06607"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_17317"} +{"question": "Could you provide some works about constructing a contrastive language-audio pre-training model?", "answer": ["CLAP : Learning Audio Concepts From Natural Language Supervision", "Contrastive Learning of Medical Visual Representations from Paired Images and Text"], "answer_arxiv_id": ["2206.04769", "2010.00747"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_17318"} +{"question": "Which research led to negative results, like a bad approximation ratio, in robust forecast aggregation?", "answer": ["Robust Forecast Aggregation", "Are You Smarter Than a Random Expert? The Robust Aggregation of Substitutable Signals"], "answer_arxiv_id": ["1710.02838v3", "2111.03153"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_17319"} +{"question": "Which works contributed to the development of large language models (LLMs)?", "answer": ["BLOOM: A 176B-Parameter Open-Access Multilingual Language Model", "PaLM: Scaling Language Modeling with Pathways", "Scaling Instruction-Finetuned Language Models", "LLaMA: Open and Efficient Foundation Language Models", "Language Models are Few-Shot Learners", "GPT-4 Technical Report"], "answer_arxiv_id": ["2211.05100", "2204.02311", "2210.11416", "2302.13971", "2005.14165", "2303.08774"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_17320"} +{"question": "What previous work has studied the generalization gap in deep learning?", "answer": ["Uniform convergence may be unable to explain generalization in deep learning", "How Does Mixup Help With Robustness and Generalization?", "Robustness Implies Generalization via Data-Dependent Generalization Bounds", "Generalization in Deep Learning"], "answer_arxiv_id": ["1902.04742", "2010.04819", "2206.13497", "1710.05468"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17321"} +{"question": "Which works contributed to traditional emotion recognition by inferring emotional states from facial expressions?", "answer": ["Deep Facial Expression Recognition: A Survey"], "answer_arxiv_id": ["1804.08348"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_17322"} +{"question": "Which works utilized text embeddings to estimate user satisfaction in conversational systems?", "answer": ["Self-Supervised Contrastive Learning for Efficient User Satisfaction\n Prediction in Conversational Agents", "Simulating User Satisfaction for the Evaluation of Task-oriented\n Dialogue Systems"], "answer_arxiv_id": ["2010.11230", "2105.03748"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_17323"} +{"question": "Which work used an Auto-Encoder to generate new samples based on the informative training data selected by the acquisition function?", "answer": ["Bayesian Generative Active Deep Learning"], "answer_arxiv_id": ["1904.11643"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_17324"} +{"question": "What studies provided indoor 3D datasets which are used to create 3D visual grounding datasets?", "answer": ["Joint 2D-3D-Semantic Data for Indoor Scene Understanding", "ARKitScenes: A Diverse Real-World Dataset For 3D Indoor Scene Understanding Using Mobile RGB-D Data", "Matterport3D: Learning from RGB-D Data in Indoor Environments", "ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "RIO: 3D Object Instance Re-Localization in Changing Indoor Environments", "Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments for Embodied AI", "Gibson Env: Real-World Perception for Embodied Agents"], "answer_arxiv_id": ["1702.01105", "2111.08897", "1709.06158", "1702.04405v2", "1908.06109", "2109.08238", "1808.10654"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_17325"} +{"question": "In what papers was the concept of selective rationalization first introduced?", "answer": ["Rationalizing Neural Predictions"], "answer_arxiv_id": ["1606.04155"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_17326"} +{"question": "What studies provide methods to reduce hallucinations?", "answer": ["Woodpecker: Hallucination Correction for Multimodal Large Language\n Models", "VIGC: Visual Instruction Generation and Correction"], "answer_arxiv_id": ["2310.16045", "2308.12714"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_17327"} +{"question": "What papers propose the Linear Scaling Rule for SGD aimed at improving generalization in large-batch training?", "answer": ["One weird trick for parallelizing convolutional neural networks", "Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour"], "answer_arxiv_id": ["1404.5997", "1706.02677"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_17328"} +{"question": "Which studies aim to segment an image into semantic regions indicated by text descriptions through open-vocabulary semantic segmentation?", "answer": ["Zero-Shot Semantic Segmentation", "Language-driven Semantic Segmentation", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "GroupViT: Semantic Segmentation Emerges from Text Supervision", "Learning Open-vocabulary Semantic Segmentation Models From Natural Language Supervision", "Unsupervised Domain Generalization by Learning a Bridge Across Domains"], "answer_arxiv_id": ["1906.00817", "2201.03546", "2112.12143", "2112.01518", "2210.04150", "2202.11094", "2301.09121", "2112.02300"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_17329"} +{"question": "Which paper uses a binning approach to estimate mutual information?", "answer": ["Opening the black box of Deep Neural Networks via Information"], "answer_arxiv_id": ["1703.00810"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_17330"} +{"question": "Who introduces a generalized noise operator that incorporates noise in non-Gaussian forward processes?", "answer": ["Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise"], "answer_arxiv_id": ["2208.09392"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_17331"} +{"question": "What papers studied the influence of Transformers on abstractive summarization?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_17332"} +{"question": "Which work established a shared latent space between visual and tactile modalities using the Gelsight touch sensor, aiming at precise fabric classification?", "answer": ["Connecting Look and Feel: Associating the visual and tactile properties\n of physical materials"], "answer_arxiv_id": ["1704.03822"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_17333"} +{"question": "Could you provide me some papers that focused on enhancing language models performance using high-quality instruction examples?", "answer": ["LIMA: Less Is More for Alignment"], "answer_arxiv_id": ["2305.11206"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_17334"} +{"question": "What papers introduced variants of attention-based models for irregular time series?", "answer": ["Multi-Time Attention Networks for Irregularly Sampled Time Series", "Self-Attentive Hawkes Process"], "answer_arxiv_id": ["2101.10318", "1907.07561"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_17335"} +{"question": "Could you provide me some studies about adapting pretrained models to novel domains using parameter-efficient fine-tuning methods?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["1902.00751", "2106.09685"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_17336"} +{"question": "Which papers used the similar factorization of conditional probability for sampling in controllable generation?", "answer": ["FUDGE: Controlled Text Generation With Future Discriminators"], "answer_arxiv_id": ["2104.05218"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_17337"} +{"question": "Can you list research that uses domain-invariant predictors as a proxy for unknown target labels?", "answer": ["Estimating Generalization under Distribution Shifts via Domain-Invariant Representations"], "answer_arxiv_id": ["2007.03511"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17338"} +{"question": "Which studies discuss the use of multiple cameras in robotics?", "answer": ["Asynchronous Multi-View SLAM"], "answer_arxiv_id": ["2101.06562"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_17339"} +{"question": "Could you provide me some works where RL has been used to improve models in machine translation?", "answer": ["Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation", "Reinforcement Learning for Bandit Neural Machine Translation with Simulated Human Feedback", "Revisiting the Weaknesses of Reinforcement Learning for Neural Machine Translation"], "answer_arxiv_id": ["1609.08144", "1707.07402", "2106.08942"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_17340"} +{"question": "Which works categorize cooperative perception into early, intermediate, and late fusion?", "answer": ["A Survey and Framework of Cooperative Perception: From Heterogeneous\n Singleton to Hierarchical Cooperation", "Collaborative Perception in Autonomous Driving: Methods, Datasets and\n Challenges"], "answer_arxiv_id": ["2208.10590", "2301.06262"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17341"} +{"question": "Any recent methods that apply vision transformer based architectures to tackle the perception task?", "answer": ["OneFormer: One Transformer to Rule Universal Image Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers", "Vision Transformers for Dense Prediction", "Omnivore: A Single Model for Many Visual Modalities", "Deformable DETR: Deformable Transformers for End-to-End Object Detection"], "answer_arxiv_id": ["2211.06220", "2112.01527", "2105.15203", "2103.13413", "2201.08377", "2010.04159"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_17342"} +{"question": "What papers focused on the utilization of policy distillation from single-task policies?", "answer": ["Distral: Robust Multitask Reinforcement Learning", "Policy Distillation"], "answer_arxiv_id": ["1707.04175", "1511.06295"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_17343"} +{"question": "Are there any studies providing methods to reduce the annotation requirement in scene flow estimation?", "answer": ["Neural Scene Flow Prior", "SCOOP: Self-Supervised Correspondence and Optimization-Based Scene Flow", "Re-Evaluating LiDAR Scene Flow for Autonomous Driving"], "answer_arxiv_id": ["2111.01253", "2211.14020", "2304.02150v2"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_17344"} +{"question": "What papers have showcased learning dynamics models using low-dimensional latent space?", "answer": ["Embed to Control: A Locally Linear Latent Dynamics Model for Control from Raw Images", "Learning to Poke by Poking: Experiential Learning of Intuitive Physics", "Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model", "DayDreamer: World Models for Physical Robot Learning"], "answer_arxiv_id": ["1506.07365", "1606.07419", "1811.04551", "1912.01603", "1911.08265", "2206.14176"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_17345"} +{"question": "Which papers discuss that decoupling subsequent steps of optimization procedures can lead to speed-ups?", "answer": ["An Optimal Algorithm for Decentralized Finite Sum Optimization", "A principled framework for the design and analysis of token algorithms", "Decoupled Greedy Learning of CNNs", "Decoupled Greedy Learning of CNNs for Synchronous and Asynchronous Distributed Learning"], "answer_arxiv_id": ["2005.10675", "2205.15015", "1901.08164", "2106.06401"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_17346"} +{"question": "Could you provide me with some studies about recent methods using contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Supporting Clustering with Contrastive Learning"], "answer_arxiv_id": ["2002.05709", "2103.12953"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_17347"} +{"question": "What research developed an extension of GFlowNets for DAGs with continuous state-action spaces?", "answer": ["A Theory of Continuous Generative Flow Networks"], "answer_arxiv_id": ["2301.12594"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_17348"} +{"question": "What papers used relevant gradient computation techniques to alleviate the difficulties of training PINNs?", "answer": ["Universal Differential Equations for Scientific Machine Learning"], "answer_arxiv_id": ["2001.04385"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_17349"} +{"question": "Which works proposed text-based local image editing by manipulating cross-attention maps?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control"], "answer_arxiv_id": ["2208.01626"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_17350"} +{"question": "In which papers were depth separations between depth 2 and 3 networks first carried out?", "answer": ["The Power of Depth for Feedforward Neural Networks"], "answer_arxiv_id": ["1512.03965"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_17351"} +{"question": "Which papers proposed and studied the optimization variant of a notion of fairness for centroid-based clustering?", "answer": ["Improved Approximation Algorithms for Individually Fair Clustering"], "answer_arxiv_id": ["2106.14043"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_17352"} +{"question": "What works have conducted detailed comparison with existing work on reward-free RL?", "answer": ["On Reward-Free Reinforcement Learning with Linear Function Approximation", "Provably Efficient Reward-Agnostic Navigation with Linear Value Iteration", "Reward-Free RL is No Harder Than Reward-Aware RL in Linear Markov Decision Processes", "Towards Deployment-Efficient Reinforcement Learning: Lower Bound and Optimality", "Sample-Efficient Reinforcement Learning with loglog(T) Switching Cost"], "answer_arxiv_id": ["2006.11274", "2008.07737", "2201.11206", "2202.06450", "2202.06385"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_17353"} +{"question": "Which studies demonstrated that human-authored counterfactuals pose a significant challenge for existing models?", "answer": ["Learning the Difference that Makes a Difference with Counterfactually-Augmented Data"], "answer_arxiv_id": ["1909.12434"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_17354"} +{"question": "What work leveraged DAG for dynamically generating evaluation data in reasoning tasks as a means to mitigate data contamination?", "answer": ["DyVal: Dynamic Evaluation of Large Language Models for Reasoning Tasks"], "answer_arxiv_id": ["2309.17167"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_17355"} +{"question": "Which studies have considered different variants of zero-sum games and are notable for developing quantum algorithms?", "answer": ["Quantum algorithms for zero-sum games", "Sublinear quantum algorithms for training linear and kernel-based classifiers", "Sublinear Classical and Quantum Algorithms for General Matrix Games"], "answer_arxiv_id": ["1904.03180", "1904.02276", "2012.06519"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_17356"} +{"question": "What studies proposed active learning algorithms for detecting rare classes and data that is out-of-distribution?", "answer": ["SIMILAR: Submodular Information Measures Based Active Learning In Realistic Scenarios"], "answer_arxiv_id": ["2107.00717"], "source_meta": {"published_time": "20211227"}, "qid": "AutoScholarQuery_train_17357"} +{"question": "Which works indicate that the model pre-trained with simple contrastive objectives on noisy image-text pairs can generate powerful vision-language representation?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_17358"} +{"question": "What studies have explored 'dual supervision' in the field of NLP?", "answer": ["Want To Reduce Labeling Cost? GPT-3 Can Help"], "answer_arxiv_id": ["2108.13487"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_17359"} +{"question": "Which research works extend knowledge distillation to feature distillation in intermediate feature learning and pairwise relations?", "answer": ["FitNets: Hints for Thin Deep Nets"], "answer_arxiv_id": ["1412.6550"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_17360"} +{"question": "What papers applied diffusion model to video generation?", "answer": ["Video-P2P: Video Editing with Cross-attention Control", "Swap Attention in Spatiotemporal Diffusions for Text-to-Video Generation", "Structure and Content-Guided Video Synthesis with Diffusion Models", "Latent-Shift: Latent Diffusion with Temporal Shift for Efficient\n Text-to-Video Generation", "ControlVideo: Training-free Controllable Text-to-Video Generation", "Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models", "Free-Bloom: Zero-Shot Text-to-Video Generator with LLM Director and LDM Animator", "LAMP: Learn A Motion Pattern for Few-Shot-Based Video Generation", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing", "LAVIE: High-Quality Video Generation with Cascaded Latent Diffusion\n Models", "Show-1: Marrying Pixel and Latent Diffusion Models for Text-to-Video\n Generation", "I2VGen-XL: High-Quality Image-to-Video Synthesis via Cascaded Diffusion\n Models"], "answer_arxiv_id": ["2303.04761", "2305.10874", "2302.03011", "2304.08477", "2305.13077", "2305.10474v3", "2309.14494v1", "2310.10769", "2303.09535", "2309.15103", "2309.15818", "2311.04145"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17361"} +{"question": "Could you tell me some works about using models to help humans write useful tests or formally verify the correctness of the predicted solutions?", "answer": ["Unit Test Case Generation with Transformers and Focal Context"], "answer_arxiv_id": ["2009.05617"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_17362"} +{"question": "Which works address the localization problem through a multi-class classification and use of a dense multimodal space?", "answer": ["Visual Cross-View Metric Localization with Dense Uncertainty Estimates"], "answer_arxiv_id": ["2208.08519"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_17363"} +{"question": "Can you mention some studies that emphasize the necessity of conducting explicit behavioral cloning in offline RL?", "answer": ["Know Your Boundaries: The Necessity of Explicit Behavioral Cloning in Offline RL", "Is Conditional Generative Modeling all you need for Decision-Making?"], "answer_arxiv_id": ["2206.00695", "2211.15657"], "source_meta": {"published_time": "20220812"}, "qid": "AutoScholarQuery_train_17364"} +{"question": "Which works combine value-based methods following the CTDE-based training paradigm with coordination graphs in order to introduce correlations among policies in Multi-Agent Reinforcement Learning?", "answer": ["QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning", "Deep Coordination Graphs"], "answer_arxiv_id": ["1803.11485", "1910.00091"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_17365"} +{"question": "In which study does an LLM serve as an optimizer and generate new prompts with the objective of maximizing task accuracy?", "answer": ["Large Language Models as Optimizers"], "answer_arxiv_id": ["2309.03409"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_17366"} +{"question": "In what research paper is the Chain-of-Thought method proposed to enhance in-context learning by incorporating additional signals into the prompt?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2201.11903"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_17367"} +{"question": "Could you give me examples of datasets that provide insights for study and analysis of human-motion through self-contact?", "answer": ["Learning Complex 3D Human Self-Contact", "On Self-Contact and Human Pose"], "answer_arxiv_id": ["2012.10366", "2104.03176"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17368"} +{"question": "Could you provide some references about multi-modal systems applied in complex tasks like OKVQA?", "answer": ["A-OKVQA: A Benchmark for Visual Question Answering using World Knowledge"], "answer_arxiv_id": ["2206.01718"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_17369"} +{"question": "What research subdivides and adds offsets to FLAME to enhance its geometry and enable a dynamic texture via an expression-dependent texture field?", "answer": ["Neural Head Avatars from Monocular RGB Videos"], "answer_arxiv_id": ["2112.01554"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_17370"} +{"question": "Are there any studies using spatial self-correlation in images for revealing geometric layout of objects?", "answer": ["Relational Embedding for Few-Shot Classification", "FCSS: Fully Convolutional Self-Similarity for Dense Semantic\n Correspondence"], "answer_arxiv_id": ["2108.09666", "1702.00926"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_17371"} +{"question": "What papers focused on the trend of flattening feature subspaces and the inclusion of graph structure matrices in GNNs?", "answer": ["Simple and Deep Graph Convolutional Networks", "Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods", "Benchmarking Graph Neural Networks", "Simplifying approach to Node Classification in Graph Neural Networks"], "answer_arxiv_id": ["2007.02133", "2110.14446", "2003.00982", "2111.06748"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_17372"} +{"question": "Which works explored the use of large language models for program synthesis?", "answer": ["Evaluating Large Language Models Trained on Code", "Program Synthesis with Large Language Models", "InCoder: A Generative Model for Code Infilling and Synthesis"], "answer_arxiv_id": ["2107.03374", "2108.07732", "2204.05999"], "source_meta": {"published_time": "20220325"}, "qid": "AutoScholarQuery_train_17373"} +{"question": "Which works developed advanced networks, ranging from convolution-based architectures to transformer-based approaches, in Monocular Metric Depth Estimation?", "answer": ["Deep Ordinal Regression Network for Monocular Depth Estimation", "Deeper Depth Prediction with Fully Convolutional Residual Networks", "Learning Depth from Single Monocular Images Using Deep Convolutional\n Neural Fields", "P3Depth: Monocular Depth Estimation with a Piecewise Planarity Prior", "Transformer-Based Attention Networks for Continuous Pixel-Wise\n Prediction", "AdaBins: Depth Estimation using Adaptive Bins", "iDisc: Internal Discretization for Monocular Depth Estimation"], "answer_arxiv_id": ["1806.02446", "1606.00373", "1502.07411", "2204.02091", "2103.12091", "2011.14141", "2304.06334"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_17374"} +{"question": "What studies adopted a generative paradigm, redefining tracking as a sequence generation task?", "answer": ["Unified Sequence-to-Sequence Learning for Single- and Multi-Modal Visual\n Object Tracking"], "answer_arxiv_id": ["2304.14394"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_17375"} +{"question": "Could you provide me some works about online data poisoning attacks?", "answer": ["Accumulative Poisoning Attacks on Real-time Data", "Iterative Machine Teaching", "An Optimal Control Approach to Sequential Machine Teaching", "Online Data Poisoning Attacks"], "answer_arxiv_id": ["2106.09993", "1705.10470", "1810.06175", "1903.01666"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17376"} +{"question": "Which papers outlined approaches to training large-scale text-to-video models using autoregressive Transformers or Diffusion Models?", "answer": ["Phenaki: Variable Length Video Generation From Open Domain Textual\n Description", "N\\\"UWA: Visual Synthesis Pre-training for Neural visUal World creAtion", "CogVideo: Large-scale Pretraining for Text-to-Video Generation via\n Transformers", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models", "State of the Art on Diffusion Models for Visual Computing"], "answer_arxiv_id": ["2210.02399", "2111.12417", "2205.15868", "2304.08818", "2210.02303", "2209.14792", "2304.08818", "2305.10474v3", "2310.07204"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_17377"} +{"question": "Could you provide the work which proposes the DepEdit framework to assess ME methods?", "answer": ["Evaluating Dependencies in Fact Editing for Language Models: Specificity\n and Implication Awareness"], "answer_arxiv_id": ["2312.01858"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_17378"} +{"question": "Can you provide some studies on Virtual OOD synthesis methods for OOD detection?", "answer": ["VOS: Learning What You Don’t Know by Virtual Outlier Synthesis", "Non-parametric Outlier Synthesis"], "answer_arxiv_id": ["2202.01197", "2303.02966"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_17379"} +{"question": "Which work showed that biased parities are learnable by SGD on a differentiable model consisting of a linear predictor and fixed module implementing the parity?", "answer": ["Quantifying the Benefit of Using Differentiable Learning over Tangent Kernels"], "answer_arxiv_id": ["2103.01210v1"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_17380"} +{"question": "Are there any examples of researches that utilized SAM's segmentation results for image editing and restoration?", "answer": ["Matte Anything: Interactive Natural Image Matting with Segment Anything\n Models", "Personalize Segment Anything Model with One Shot"], "answer_arxiv_id": ["2306.04121", "2305.03048"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17381"} +{"question": "Are there any preparatory works about creating a chemical agent using Large Language Models?", "answer": ["ChemCrow: Augmenting large-language models with chemistry tools", "Emergent autonomous scientific research capabilities of large language models"], "answer_arxiv_id": ["2304.05376v5", "2304.05332v1"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_17382"} +{"question": "What papers propose using the heavy-tailed phenomena to construct metrics for evaluating the generalization of Neural Networks?", "answer": ["Predicting trends in the quality of state-of-the-art neural networks without access to training or testing data", "Evaluating natural language processing models with generalization metrics that do not need access to any training or testing data"], "answer_arxiv_id": ["2002.06716v2", "2202.02842"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_17383"} +{"question": "What studies use semantic maps to optimize the feature space for low-light enhancement?", "answer": ["Learning Semantic-Aware Knowledge Guidance for Low-Light Image\n Enhancement"], "answer_arxiv_id": ["2304.07039"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_17384"} +{"question": "What studies were conducted about bio-inspired Spiking Neural Networks (SNNs) that can benefit from advanced deep learning and neuroscience knowledge?", "answer": ["Probabilistic Modeling: Proving the Lottery Ticket Hypothesis in Spiking Neural Network"], "answer_arxiv_id": ["2305.12148"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_17385"} +{"question": "Are there any studies that employed binary-tree mechanism for private online learning?", "answer": ["Differentially Private Online Learning", "The Price of Differential Privacy for Online Learning"], "answer_arxiv_id": ["1109.0105", "1701.07953"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_17386"} +{"question": "Which references elaborate on the application of CLIP for image-text alignment?", "answer": ["Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models", "MaskCLIP: Masked Self-Distillation Advances Contrastive Language-Image\n Pretraining", "Prompting Diffusion Representations for Cross-Domain Semantic\n Segmentation", "PromptStyler: Prompt-driven Style Generation for Source-free Domain\n Generalization", "Convolutions Die Hard: Open-Vocabulary Segmentation with Single Frozen\n Convolutional CLIP", "CAT-Seg: Cost Aggregation for Open-Vocabulary Semantic Segmentation"], "answer_arxiv_id": ["2303.04803", "2208.12262", "2307.02138", "2307.15199", "2308.02487", "2303.11797"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_17387"} +{"question": "Which works introduced PETL that refers to updating only a small number of pre-trained or additional parameters during fine-tuning?", "answer": ["Towards a Unified View of Parameter-Efficient Transfer Learning", "Delta Tuning: A Comprehensive Study of Parameter Efficient Methods for Pre-trained Language Models"], "answer_arxiv_id": ["2110.04366", "2203.06904v2"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_17388"} +{"question": "Which studies collected natural images with specific factor changes for the study of model robustness?", "answer": ["Noise or Signal: The Role of Image Backgrounds in Object Recognition"], "answer_arxiv_id": ["2006.09994"], "source_meta": {"published_time": "20230808"}, "qid": "AutoScholarQuery_train_17389"} +{"question": "Which works have estimated the attribution/importance/saliency of each input variable in DNNs?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "Axiomatic Attribution for Deep Networks", "A Unified Approach to Interpreting Model Predictions", "Interpretable Explanations of Black Boxes by Meaningful Perturbation", "Learning Deep Features for Discriminative Localization", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization"], "answer_arxiv_id": ["1602.04938", "1703.01365", "1705.07874", "1704.03296", "1512.04150", "1610.02391"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_17390"} +{"question": "What works explore learning invariant or diverse features for improved robustness?", "answer": ["Invariant Risk Minimization", "Model Patching: Closing the Subgroup Performance Gap with Data\n Augmentation", "Rich Feature Construction for the Optimization-Generalization Dilemma", "Controlling Directions Orthogonal to a Classifier"], "answer_arxiv_id": ["1907.02893", "2008.06775", "2203.15516", "2201.11259"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_17391"} +{"question": "What work pioneered unsupervised denoising using a technique where a neural network is trained on pairs of noisy images?", "answer": ["Noise2Noise: Learning Image Restoration without Clean Data"], "answer_arxiv_id": ["1803.04189"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_17392"} +{"question": "Which work introduced a versatile new input encoding for neural primitives that decreases the cost and enables the use of a smaller network without affecting quality?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_17393"} +{"question": "What studies used a min-max optimization to boost the performance of the worst-performing clients?", "answer": ["Agnostic Federated Learning"], "answer_arxiv_id": ["1902.00146"], "source_meta": {"published_time": "20240526"}, "qid": "AutoScholarQuery_train_17394"} +{"question": "Could you provide me some studies investigating the effect of memorization of difficult samples on the generalization performance of adversarial training?", "answer": ["How benign is benign overfitting?", "Exploring Memorization in Adversarial Training"], "answer_arxiv_id": ["2007.04028", "2106.01606"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_17395"} +{"question": "What literature discusses the idea of Panoptic Segmentation?", "answer": ["Panoptic Segmentation", "Panoptic Feature Pyramid Networks", "UPSNet: A Unified Panoptic Segmentation Network", "Attention-guided Unified Network for Panoptic Segmentation", "An End-to-End Network for Panoptic Segmentation", "Learning Instance Occlusion for Panoptic Segmentation", "Unifying Training and Inference for Panoptic Segmentation", "Pixel Consensus Voting for Panoptic Segmentation", "End-to-End Object Detection with Transformers", "Fully Convolutional Networks for Panoptic Segmentation", "Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation", "Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation", "MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation"], "answer_arxiv_id": ["1801.00868", "1901.02446", "1901.03784", "1812.03904", "1903.05027", "1906.05896", "2001.04982", "2004.01849", "2005.12872", "2012.00720", "1911.10194", "2003.07853", "2012.00759", "2206.08948"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_17396"} +{"question": "What studies provide innovative modifications to the architecture of ResNet?", "answer": ["A ConvNet for the 2020s"], "answer_arxiv_id": ["2201.03545"], "source_meta": {"published_time": "20220620"}, "qid": "AutoScholarQuery_train_17397"} +{"question": "What research introduced the RefineMask method that incorporates a semantic head to Mask R-CNN?", "answer": ["RefineMask: Towards High-Quality Instance Segmentation with Fine-Grained Features"], "answer_arxiv_id": ["2104.08569"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_17398"} +{"question": "Which papers are about the impact of deduplication on language models?", "answer": ["Quantifying Memorization Across Neural Language Models", "Deduplicating Training Data Makes Language Models Better", "Scaling Laws and Interpretability of Learning from Repeated Data"], "answer_arxiv_id": ["2202.07646", "2107.06499", "2205.10487"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_17399"} +{"question": "Which papers proved that Transformers with learnable positional encodings are universal approximators of continuous sequence-to-sequence functions?", "answer": ["Are Transformers universal approximators of sequence-to-sequence functions?", "Big Bird: Transformers for Longer Sequences"], "answer_arxiv_id": ["1912.10077", "2007.14062"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17400"} +{"question": "Could you provide me some works about no-reference IQA methods that compute the overall quality score of an input image without requiring a reference?", "answer": ["MANIQA: Multi-dimension Attention Network for No-Reference Image Quality\n Assessment", "Exploring CLIP for Assessing the Look and Feel of Images"], "answer_arxiv_id": ["2204.08958", "2207.12396"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_17401"} +{"question": "Could you provide some works that have attempted to convert problems into formal logic to bridge the gap between deep learning and symbolic AI?", "answer": ["Improving Coherence and Consistency in Neural Sequence Models with Dual-System, Neuro-Symbolic Reasoning"], "answer_arxiv_id": ["2107.02794"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_17402"} +{"question": "What is the most multilingual language model, as per the previous studies?", "answer": ["BLOOM: A 176B-Parameter Open-Access Multilingual Language Model"], "answer_arxiv_id": ["2211.05100"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_17403"} +{"question": "Who adapted adversarial attacks originally developed for classification tasks to SR?", "answer": ["Evaluating Robustness of Deep Image Super-Resolution against Adversarial\n Attacks"], "answer_arxiv_id": ["1904.06097"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_17404"} +{"question": "Which papers describe Text VAEs being useful for learning a smooth and interpretable representation space and generating diverse text?", "answer": ["Generating Sentences from a Continuous Space", "A Tutorial on Deep Latent Variable Models of Natural Language", "Optimus: Organizing Sentences via Pre-trained Modeling of a Latent Space"], "answer_arxiv_id": ["1511.06349", "1812.06834", "2004.04092"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_17405"} +{"question": "Which methods are used by the probabilistic approach to estimate model uncertainty in image segmentation?", "answer": ["Bayesian Deep Learning and a Probabilistic Perspective of Generalization"], "answer_arxiv_id": ["2002.08791"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_17406"} +{"question": "Can you give an example of a user attribution technique that incorporates user-specific fingerprints?", "answer": ["Decentralized Attribution of Generative Models"], "answer_arxiv_id": ["2010.13974"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_17407"} +{"question": "Which papers initiated the OOD detection task in pre-trained vision-language models?", "answer": ["Exploring the Limits of Out-of-Distribution Detection"], "answer_arxiv_id": ["2106.03004"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_17408"} +{"question": "What studies have focused on the development of neural audio codec?", "answer": ["SoundStream: An End-to-End Neural Audio Codec", "High Fidelity Neural Audio Compression", "AudioDec: An Open-source Streaming High-fidelity Neural Audio Codec", "HiFi-Codec: Group-residual Vector quantization for High Fidelity Audio\n Codec", "High-Fidelity Audio Compression with Improved RVQGAN"], "answer_arxiv_id": ["2107.03312", "2210.13438", "2305.16608", "2305.02765", "2306.06546"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_17409"} +{"question": "Which works utilize geometric objects or probabilistic distributions in embedding-based methods for complex queries?", "answer": ["Embedding Logical Queries on Knowledge Graphs", "Beta Embeddings for Multi-Hop Logical Reasoning in Knowledge Graphs", "ConE: Cone Embeddings for Multi-Hop Reasoning over Knowledge Graphs", "Self-Supervised Hyperboloid Representations from Logical Queries over Knowledge Graphs"], "answer_arxiv_id": ["1806.01445", "2010.11465", "2110.13715", "2012.13023"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_17410"} +{"question": "Which work proposes using a memory module for storing and retrieving prototypical object representations for segmentation?", "answer": ["Unsupervised Video Object Segmentation via Prototype Memory Network"], "answer_arxiv_id": ["2209.03712"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_17411"} +{"question": "Could you give me examples of papers that developed algorithms to find permutations in SGD solutions?", "answer": ["Git Re-Basin: Merging Models Modulo Permutation Symmetries"], "answer_arxiv_id": ["2209.04836"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_17412"} +{"question": "What papers focus on the standard zero-reward collision model in MPMAB?", "answer": ["Selfish Robustness and Equilibria in Multi-Player Bandits"], "answer_arxiv_id": ["2002.01197"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17413"} +{"question": "Could you tell me about the works that have studied the effect of machine unlearning on the generalization loss?", "answer": ["Adaptive Machine Unlearning", "Remember What You Want to Forget: Algorithms for Machine Unlearning"], "answer_arxiv_id": ["2106.04378", "2103.03279v2"], "source_meta": {"published_time": "20221221"}, "qid": "AutoScholarQuery_train_17414"} +{"question": "Which papers present research on indoor 3D object detection methods?", "answer": ["Deep Hough Voting for 3D Object Detection in Point Clouds", "H3DNet: 3D Object Detection Using Hybrid Geometric Primitives", "Back-tracing Representative Points for Voting-based 3D Object Detection in Point Clouds", "CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Clouds"], "answer_arxiv_id": ["1904.09664", "2006.05682", "2104.06114", "2210.04264"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_17415"} +{"question": "Which works are the examples of prediction methods for intrinsic behavioral learning without a reward?", "answer": ["Large-Scale Study of Curiosity-Driven Learning", "Curiosity-driven Exploration by Self-supervised Prediction", "Self-Supervised Exploration via Disagreement"], "answer_arxiv_id": ["1808.04355", "1705.05363", "1906.04161"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_17416"} +{"question": "Can you provide some works where a cluster-based regularization was used?", "answer": ["NodeTrans: A Graph Transfer Learning Approach for Traffic Prediction"], "answer_arxiv_id": ["2207.01301"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_17417"} +{"question": "Could you mention some papers that set an objective to best approximate the information being exchanged via techniques such as quantization and sparsification?", "answer": ["Linear Stochastic Bandits over a Bit-Constrained Channel"], "answer_arxiv_id": ["2203.01198"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_17418"} +{"question": "Which works utilized Proximal Policy Optimization (PPO) to fine-tune GPT-2 based on human preferences?", "answer": ["Reliability and Learnability of Human Bandit Feedback for\n Sequence-to-Sequence Reinforcement Learning", "Proximal Policy Optimization Algorithms", "Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["1805.10627", "1707.06347", "1909.08593"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_17419"} +{"question": "What was the initial paper that proposed the framework of Total Variation (TV) stability?", "answer": ["Machine Unlearning via Algorithmic Stability"], "answer_arxiv_id": ["2102.13179"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_17420"} +{"question": "Can you name some works that used radar point clouds for object detection?", "answer": ["2D Car Detection in Radar Data with PointNets", "Radar-based Road User Classification and Novelty Detection with\n Recurrent Neural Network Ensembles", "Using Machine Learning to Detect Ghost Images in Automotive Radar"], "answer_arxiv_id": ["1904.08414", "1905.11703", "2007.05280"], "source_meta": {"published_time": "20240428"}, "qid": "AutoScholarQuery_train_17421"} +{"question": "Are there any research papers that examined how to build visual correspondence models with weak supervision?", "answer": ["Learning Feature Descriptors using Camera Pose Supervision"], "answer_arxiv_id": ["2004.13324"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17422"} +{"question": "Are there any works that showed the potential of kNN-MT for unsupervised domain adaptation and online learning?", "answer": ["Non-Parametric Unsupervised Domain Adaptation for Neural Machine Translation", "Non-Parametric Domain Adaptation for End-to-End Speech Translation", "Non-Parametric Online Learning from Human Feedback for Neural Machine Translation"], "answer_arxiv_id": ["2109.06604", "2205.11211", "2109.11136"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_17423"} +{"question": "Could you provide some references on the enhancement of quality and speed of sampling in diffusion models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Implicit Models", "Progressive Distillation for Fast Sampling of Diffusion Models"], "answer_arxiv_id": ["2102.09672", "2105.05233", "2010.02502", "2202.00512"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_17424"} +{"question": "Could you give examples of the works where co-training is used to handle confirmation bias?", "answer": ["Deep Co-Training for Semi-Supervised Image Recognition", "Semi-Supervised Semantic Segmentation with Cross-Consistency Training", "Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision"], "answer_arxiv_id": ["1803.05984", "2003.09005", "2106.01226"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_17425"} +{"question": "Are there any studies that have investigated cross-modality adaptation by modeling the vision input using pretrained language models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "Multimodal Few-Shot Learning with Frozen Language Models", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2204.14198", "2106.13884", "2301.12597"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17426"} +{"question": "How has metric learning utilized contrastive learning to cluster similar examples and separate the dis-similar ones?", "answer": ["Smart Mining for Deep Metric Learning"], "answer_arxiv_id": ["1704.01285"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_17427"} +{"question": "Which paper introduces a decomposed, attention-based prompting method for continual learning?", "answer": ["CODA-Prompt: COntinual Decomposed Attention-based Prompting for\n Rehearsal-Free Continual Learning"], "answer_arxiv_id": ["2211.13218"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_17428"} +{"question": "Which paper proposed a multi-agent policy gradient algorithm in the context of CTDE?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments"], "answer_arxiv_id": ["1706.02275"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_17429"} +{"question": "Which references used an MLP in vanilla RCRL settings?", "answer": ["RvS: What is Essential for Offline RL via Supervised Learning?"], "answer_arxiv_id": ["2112.10751"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17430"} +{"question": "Which papers present voxel-based methods for 3D detection?", "answer": ["VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection", "PointPillars: Fast Encoders for Object Detection from Point Clouds"], "answer_arxiv_id": ["1711.06396", "1812.05784"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_17431"} +{"question": "Who proposed the 'BBQ' and what does it do?", "answer": ["BBQ: A Hand-Built Bias Benchmark for Question Answering"], "answer_arxiv_id": ["2110.08193"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_17432"} +{"question": "Can you inform me about the recent researches that focus on advanced planning methods for improving the multi-step problem-solving abilities of LLMs?", "answer": ["Self-Refine: Iterative Refinement with Self-Feedback", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "ReAct: Synergizing Reasoning and Acting in Language Models", "Language Agent Tree Search Unifies Reasoning Acting and Planning in\n Language Models"], "answer_arxiv_id": ["2303.17651", "2203.11171", "2305.10601", "2210.03629", "2310.04406"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_17433"} +{"question": "Any studies on ensemble methods for enhancing the performance of single-model defenses?", "answer": ["Improving Adversarial Robustness of Ensembles with Diversity Training", "Improving Adversarial Robustness via Promoting Ensemble Diversity", "EMPIR: Ensembles of Mixed Precision Deep Networks for Increased Robustness against Adversarial Attacks", "DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensembles", "TRS: Transferability Reduced Ensemble via Promoting Gradient Diversity and Model Smoothness", "A Multiclass Boosting Framework for Achieving Fast and Provable Adversarial Robustness", "Building Robust Ensembles via Margin Boosting"], "answer_arxiv_id": ["1901.09981", "1901.08846", "2004.10162", "2009.14720", "2104.00671", "2103.01276", "2206.03362"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_17434"} +{"question": "Which works have recently focused on making graphical neural networks more powerful than Weisfeiler-Lehman test?", "answer": ["Building powerful and equivariant graph neural networks with structural message-passing", "Relational Pooling for Graph Representations", "Coloring graph neural networks for node disambiguation", "Random Features Strengthen Graph Neural Networks", "The Surprising Power of Graph Neural Networks with Random Node Initialization", "Identity-aware Graph Neural Networks", "Breaking the Limits of Message Passing Graph Neural Networks", "Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks", "Weisfeiler and Lehman Go Cellular: CW Networks", "Directional Graph Networks"], "answer_arxiv_id": ["2006.15107", "1903.02541", "1912.06058", "2002.03155", "2010.01179", "2101.10320", "2106.04319", "2103.03212", "2106.12575", "2010.02863"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_17435"} +{"question": "What works discuss that humans naively trust explanations regardless of their 'correctness'?", "answer": ["“How do I fool you?”: Manipulating User Trust via Misleading Black Box Explanations"], "answer_arxiv_id": ["1911.06473"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_17436"} +{"question": "Which study represents an adversarial attack that can drastically reduce the accuracy of co-salient object detection?", "answer": ["Can You Spot the Chameleon? Adversarially Camouflaging Images from\n Co-Salient Object Detection"], "answer_arxiv_id": ["2009.09258"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_17437"} +{"question": "What are the research works that utilize trained post-editing models to edit the outputs of the LLM or MT system?", "answer": ["APE at Scale and its Implications on MT Evaluation Biases", "Context-Aware Monolingual Repair for Neural Machine Translation", "Generating Sequences by Learning to Self-Correct"], "answer_arxiv_id": ["1904.04790", "1909.01383", "2211.00053"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_17438"} +{"question": "What works introduced the Decision Transformers(DT) in the context of offline RL?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["2106.01345"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_17439"} +{"question": "What research synthesizes and edits videos with cross-frame attention, initial frame integration, and background smoothing?", "answer": ["Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators"], "answer_arxiv_id": ["2303.13439"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17440"} +{"question": "What works develop different parameter efficient tuning techniques?", "answer": ["FedTP: Federated Learning by Transformer Personalization", "Learning Federated Visual Prompt in Null Space for MRI Reconstruction", "Efficient Model Personalization in Federated Learning via\n Client-Specific Prompt Generation"], "answer_arxiv_id": ["2211.01572", "2303.16181", "2308.15367"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_17441"} +{"question": "Which studies are closer in our spirit that involved the design of CTR evaluation metrics or suitable loss functions?", "answer": ["Cost-sensitive Learning for Utility Optimization in Online Advertising Auctions"], "answer_arxiv_id": ["1603.03713"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17442"} +{"question": "What are some research papers discussing scene text detection methods?", "answer": ["Arbitrary-Oriented Scene Text Detection via Rotation Proposals", "TextBoxes++: A Single-Shot Oriented Scene Text Detector", "Rotation-Sensitive Regression for Oriented Scene Text Detection"], "answer_arxiv_id": ["1703.01086", "1801.02765", "1803.05265"], "source_meta": {"published_time": "20220129"}, "qid": "AutoScholarQuery_train_17443"} +{"question": "What are some examples of research using pre-trained network descriptors for anomaly detection?", "answer": ["Uninformed Students: Student-Teacher Anomaly Detection with\n Discriminative Latent Embeddings", "Sub-Image Anomaly Detection with Deep Pyramid Correspondences", "PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and\n Localization", "CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via\n Conditional Normalizing Flows", "PANDA: Adapting Pretrained Features for Anomaly Detection and\n Segmentation", "Transfer Learning Gaussian Anomaly Detection by Fine-tuning\n Representations", "Towards Total Recall in Industrial Anomaly Detection", "PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and\n Localization", "CutPaste: Self-Supervised Learning for Anomaly Detection and\n Localization", "Self-Supervised Predictive Convolutional Attentive Block for Anomaly\n Detection", "SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection\n and Segmentation", "Collaborative Discrepancy Optimization for Reliable Image Anomaly Localization", "Pushing the Limits of Fewshot Anomaly Detection in Industry Vision:\n Graphcore", "Asymmetric Student-Teacher Networks for Industrial Anomaly Detection"], "answer_arxiv_id": ["1911.02357", "2005.02357", "2011.08785", "2107.12571", "2010.05903", "2108.04116", "2106.08265", "2011.08785", "2104.04015", "2111.09099", "2207.14315", "2302.08769v1", "2301.12082", "2210.07829"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_17444"} +{"question": "Can you point out the research papers that explored point-based representations with differential point splatting?", "answer": ["PointAvatar: Deformable Point-based Head Avatars from Videos"], "answer_arxiv_id": ["2212.08377"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_17445"} +{"question": "What studies have extended the PSRs to the non-tabular setting using conditional mean embeddings?", "answer": ["Hilbert Space Embeddings of Predictive State Representations"], "answer_arxiv_id": ["1309.6819v1"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_17446"} +{"question": "What research proposed using a vision abstractor and finetuning the vision encoder?", "answer": ["mPLUG-Owl: Modularization Empowers Large Language Models with\n Multimodality"], "answer_arxiv_id": ["2304.14178"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_17447"} +{"question": "Any works about SGD with (biased) compression?", "answer": ["Sparsified SGD with Memory", "Linearly Converging Error Compensated SGD"], "answer_arxiv_id": ["1809.07599", "2010.12292"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_17448"} +{"question": "Could you provide names of the works that extended the generation process to the temporal domain in the context of 2D video generation?", "answer": ["MoCoGAN: Decomposing Motion and Content for Video Generation", "A Good Image Generator Is What You Need for High-Resolution Video Synthesis", "StyleGAN-V: A Continuous Video Generator with the Price, Image Quality and Perks of StyleGAN2", "Generating Videos with Dynamics-aware Implicit Generative Adversarial Networks"], "answer_arxiv_id": ["1707.04993", "2104.15069", "2112.14683", "2202.10571"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_17449"} +{"question": "Which papers quantified memorization using data extraction attacks?", "answer": ["Extracting Training Data from Large Language Models", "Training Data Extraction From Pre-trained Language Models: A Survey"], "answer_arxiv_id": ["2012.07805", "2305.16157"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_17450"} +{"question": "Which study shows that handcrafted features perform well when training from scratch, although fine-tuning deep models outperforms them?", "answer": ["Differentially Private Learning Needs Better Features (or Much More Data)"], "answer_arxiv_id": ["2011.11660"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_17451"} +{"question": "Could you mention works that incorporate textual information into the protein encoder through multimodal pre-training?", "answer": ["ProtST: Multi-Modality Learning of Protein Sequences and Biomedical\n Texts"], "answer_arxiv_id": ["2301.12040"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_17452"} +{"question": "Which work showed that Gaussian mixtures describe the deep learning representation of GAN data?", "answer": ["Random Matrix Theory Proves that Deep Learning Representations of GAN-data Behave as Gaussian Mixtures"], "answer_arxiv_id": ["2001.08370v1"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_17453"} +{"question": "Which works focused on compressing local model updates via gradient quantization in federated learning?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding", "Distributed Mean Estimation with Limited Communication", "TernGrad: Ternary Gradients to Reduce Communication in Distributed Deep Learning", "Gradient Sparsification for Communication-Efficient Distributed Optimization", "Communication Lower Bounds for Statistical Estimation Problems via a Distributed Data Processing Inequality"], "answer_arxiv_id": ["1602.05629", "1610.02132", "1611.00429", "1705.07878", "1710.09854v1", "1506.07216"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_17454"} +{"question": "What studies have utilized Detection Transformer (DETR) for advances in segmentation tasks?", "answer": ["Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers", "Masked-attention Mask Transformer for Universal Image Segmentation", "Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentation", "OneFormer: One Transformer to Rule Universal Image Segmentation", "MP-Former: Mask-Piloted Transformer for Image Segmentation"], "answer_arxiv_id": ["2109.03814", "2112.01527", "2206.02777", "2211.06220", "2303.07336"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_17455"} +{"question": "Could you provide me some research that has similar goal of adding new dimensions to existing image encoders?", "answer": ["All in One: Exploring Unified Video-Language Pre-training", "HNeRV: A Hybrid Neural Representation for Videos", "PointNetLK: Robust & Efficient Point Cloud Registration using PointNet"], "answer_arxiv_id": ["2203.07303", "2304.02633", "1903.05711"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_17456"} +{"question": "What papers introduce models generating text descriptions for localizing risk objects in autonomous driving?", "answer": ["DRAMA: Joint Risk Localization and Captioning in Driving"], "answer_arxiv_id": ["2209.10767"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_17457"} +{"question": "Could you name some studies that explored the diffusion-based methods in the context of dance generation?", "answer": ["EDGE: Editable Dance Generation From Music", "FineDance: A Fine-grained Choreography Dataset for 3D Full Body Dance\n Generation"], "answer_arxiv_id": ["2211.10658", "2212.03741"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17458"} +{"question": "Who proposed question-answer decomposition model for fact-checking?", "answer": ["EXPLAINABLE FACT-CHECKING THROUGH QUESTION ANSWERING"], "answer_arxiv_id": ["2110.05369"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_17459"} +{"question": "In which papers are the improved versions of UED, POET and Enhanced POET discussed?", "answer": ["Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions", "Enhanced POET: Open-Ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions"], "answer_arxiv_id": ["1901.01753", "2003.08536"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_17460"} +{"question": "Could you mention some recent advancements in multi-view diffusion models that focus on generating a fixed number of target views with fixed camera poses?", "answer": ["MVDream: Multi-view Diffusion for 3D Generation", "Consistent-1-to-3: Consistent Image to 3D View Synthesis via\n Geometry-aware Diffusion Models", "Zero123++: a Single Image to Consistent Multi-view Diffusion Base Model", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image", "Wonder3D: Single Image to 3D using Cross-Domain Diffusion", "One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization"], "answer_arxiv_id": ["2308.16512", "2310.03020", "2310.15110", "2309.03453", "2310.15008", "2306.16928"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_17461"} +{"question": "What research papers discuss prompt tuning method for task adaptation?", "answer": ["Learning to Decompose Visual Features with Latent Textual Prompts", "Prompt-aligned Gradient for Prompt Tuning", "DETA: Denoised Task Adaptation for Few-Shot Learning", "Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2210.04287", "2205.14865", "2303.06315", "2109.01134", "2203.05557"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_17462"} +{"question": "Which research papers introduced method to generate the intermediate motion in Motion completion task?", "answer": ["Human Motion Diffusion Model"], "answer_arxiv_id": ["2209.14916"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_17463"} +{"question": "What studies first introduced the task of pose-guided person image synthesis?", "answer": ["Pose Guided Person Image Generation"], "answer_arxiv_id": ["1705.09368"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_17464"} +{"question": "Which papers discuss the utilization of large language models for scoring in subjective evaluations?", "answer": ["INSTRUCTEVAL: Towards Holistic Evaluation of Instruction-Tuned Large\n Language Models", "G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment", "GPTScore: Evaluate as You Desire", "Phoenix: Democratizing ChatGPT across Languages"], "answer_arxiv_id": ["2306.04757", "2303.16634", "2302.04166", "2304.10453"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_17465"} +{"question": "Could you point me to research works that studied the fairness Pareto frontier and the fair Bayes optimal classifier?", "answer": ["Equality of Opportunity in Supervised Learning", "Algorithmic decision making and the cost of fairness", "Leveraging Labeled and Unlabeled Data for Consistent Fair Binary Classification", "Bayes-Optimal Classifiers under Group Fairness", "Fair Bayes-Optimal Classifiers Under Predictive Parity"], "answer_arxiv_id": ["1610.02413", "1701.08230", "1906.05082", "2202.09724", "2205.07182"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_17466"} +{"question": "Can you tell me the studies that focus on multi-modal generation tasks, using a conditional diffusion procedure to gradually transform noise and source images into the target image?", "answer": ["Unsupervised Medical Image Translation with Adversarial Diffusion Models", "Conversion Between CT and MRI Images Using Diffusion and Score-Matching\n Models", "Adaptive Diffusion Priors for Accelerated MRI Reconstruction"], "answer_arxiv_id": ["2207.08208", "2209.12104", "2207.05876"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_17467"} +{"question": "Which works show the success of transformer-based pre-trained LLMs in code, both in understanding and generation?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer"], "answer_arxiv_id": ["1810.04805", "1910.10683"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_17468"} +{"question": "Could you provide some works that illustrate deep neural networks' vulnerability to adversarial perturbations?", "answer": ["Intriguing properties of neural networks", "Evasion attacks against machine learning at test time"], "answer_arxiv_id": ["1312.6199", "1708.06131"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_17469"} +{"question": "What study proposed FedTHE, a test-time adaptation algorithm for federated learning?", "answer": ["Test-Time Robust Personalization for Federated Learning"], "answer_arxiv_id": ["2205.10920"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_17470"} +{"question": "Which study developed a cross-resolution attention that enables the gathering of comprehensive contextual information for high-resolution features?", "answer": ["RTFormer: Efficient Design for Real-Time Semantic Segmentation with\n Transformer"], "answer_arxiv_id": ["2210.07124"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_17471"} +{"question": "What papers are about learning-based metrics for perceptual similarity that use deep features from pre-trained networks?", "answer": ["A Neural Algorithm of Artistic Style", "Perceptual Losses for Real-Time Style Transfer and Super-Resolution", "Generating Images with Perceptual Similarity Metrics based on Deep Networks"], "answer_arxiv_id": ["1508.06576", "1603.08155", "1602.02644"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_17472"} +{"question": "What studies evaluate humanoid traits of LLMs used as human proxies?", "answer": ["Evaluating Psychological Safety of Large Language Models", "Machine Psychology"], "answer_arxiv_id": ["2212.10529", "2303.13988v6"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_17473"} +{"question": "Could you provide the research paper where the authors proposed a Contrastive Language-Image Pre-Training (CLIP) to learn a robust representation for each text and image?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17474"} +{"question": "Which work developed the state of the art model DINO v2, leveraging both contrastive learning and masked image modelling?", "answer": ["DINOv2: Learning Robust Visual Features without Supervision"], "answer_arxiv_id": ["2304.07193"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_17475"} +{"question": "What works have analyzed the degree of learned permutation symmetry in networks that process weights?", "answer": ["Predicting Neural Network Accuracy from Weights"], "answer_arxiv_id": ["2002.11448"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_17476"} +{"question": "Which papers mention the problem of Shampoo memory costs in large models like BERT-Large?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_17477"} +{"question": "Can you identify some studies that utilized Monte-Carlo Tree Search to reduce the complexity of continuous space?", "answer": ["Learning Search Space Partition for Black-box Optimization using Monte Carlo Tree Search", "Monte Carlo Tree Search based Variable Selection for High Dimensional Bayesian Optimization"], "answer_arxiv_id": ["2007.00708", "2210.01628"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_17478"} +{"question": "What papers proposed masking out detected moving objects from photometric loss computation?", "answer": ["Self-Supervised Monocular Depth Estimation: Solving the Dynamic Object\n Problem by Semantic Guidance"], "answer_arxiv_id": ["2007.06936"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_17479"} +{"question": "Which works use the encoder-decoder framework for the link prediction task using GNNs?", "answer": ["Variational Graph Auto-Encoders", "Graph Convolutional Matrix Completion", "Modeling Relational Data with Graph Convolutional Networks", "Graph Convolutional Neural Networks for Web-Scale Recommender Systems", "Hyperspherical Variational Auto-Encoders", "Neural Bellman-Ford Networks: A General Graph Neural Network Framework for Link Prediction", "Neo-GNNs: Neighborhood Overlap-aware Graph Neural Networks for Link Prediction", "Learning from Counterfactual Links for Link Prediction"], "answer_arxiv_id": ["1611.07308", "1706.02263", "1703.06103v4", "1806.01973", "1804.00891", "2106.06935", "2206.04216", "2106.02172"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_17480"} +{"question": "What research exists concerning the use of LLMs in the improvement of software development procedures?", "answer": ["ChatDev: Communicative Agents for Software Development"], "answer_arxiv_id": ["2307.07924"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_17481"} +{"question": "Can you name a human-annotated dataset used in multimodal instruction tuning?", "answer": ["MultiInstruct: Improving Multi-Modal Zero-Shot Learning via Instruction\n Tuning"], "answer_arxiv_id": ["2212.10773"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_17482"} +{"question": "Which papers demonstrate the promising direction of employing visual grounding datasets to endow MLLMs with region-level visual understanding abilities?", "answer": ["Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "Kosmos-2: Grounding Multimodal Large Language Models to the World", "Ferret: Refer and Ground Anything Anywhere at Any Granularity", "Pink: Unveiling the Power of Referential Comprehension for Multi-modal\n LLMs"], "answer_arxiv_id": ["2306.15195", "2306.14824", "2310.07704", "2310.00582"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_17483"} +{"question": "Which study uses reinforcement learning at the meta level to improve the performance of the low-level black box optimizers?", "answer": ["Discovering Attention-Based Genetic Algorithms via Meta-Black-Box Optimization"], "answer_arxiv_id": ["2304.03995"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_17484"} +{"question": "Can you provide the names of the papers that proposed datasets inspired by the Raven’s Progressive Matrices (RPM)?", "answer": ["Measuring abstract reasoning in neural networks", "RAVEN: A Dataset for Relational and Analogical Visual rEasoNing"], "answer_arxiv_id": ["1807.04225", "1903.02741"], "source_meta": {"published_time": "20220618"}, "qid": "AutoScholarQuery_train_17485"} +{"question": "What research works are there on unsupervised RL that actively collect new datasets for a yet to be specified task?", "answer": ["Don’t Change the Algorithm, Change the Data: Exploratory Data for Offline Reinforcement Learning", "Visual Backtracking Teleoperation: A Data Collection Protocol for Offline Image-Based Reinforcement Learning"], "answer_arxiv_id": ["2201.13425", "2210.02343"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_17486"} +{"question": "Do we have instances where language models were trained on a mix of abstracts from PubMed and full text articles from PubMed Central?", "answer": ["BioBERT: a pre-trained biomedical language representation model for biomedical text mining", "SciBert: A Pretrained Language Model for Scientific Text", "Domain-Specific Language Model Pretraining for Biomedical Natural Language Processing"], "answer_arxiv_id": ["1901.08746", "1903.10676", "2007.15779"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_17487"} +{"question": "Where can I find the process of using VIGC to generate concise answers and reduce hallucinations?", "answer": ["VIGC: Visual Instruction Generation and Correction"], "answer_arxiv_id": ["2308.12714"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_17488"} +{"question": "Which work proposed a method for VLM agents to execute actions on the web?", "answer": ["GPT-4V(ision) is a Generalist Web Agent, if Grounded"], "answer_arxiv_id": ["2401.01614"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_17489"} +{"question": "Which work introduced a cyclic inference scheme for generating panoramic images?", "answer": ["Diverse Plausible 360-Degree Image Outpainting for Efficient 3DCG\n Background Creation"], "answer_arxiv_id": ["2203.14668"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_17490"} +{"question": "Could you provide me some works that apply model fine-tuning methods in white-box source-free UDA?", "answer": ["Domain-Specific Batch Normalization for Unsupervised Domain Adaptation"], "answer_arxiv_id": ["1906.03950"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_17491"} +{"question": "Which works applied stochastic approximation to the graph drawing objective?", "answer": ["The Laplacian in RL: Learning Representations with Efficient Approximations", "Discovering Options for Exploration by Minimizing Cover Time", "Towards Better Laplacian Representation in Reinforcement Learning with Generalized Graph Drawing"], "answer_arxiv_id": ["1810.04586", "1903.00606v2", "2107.05545"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17492"} +{"question": "Are there any papers proposing a multi-candidate representation function to capture the diversity in different modalities?", "answer": ["Polysemous Visual-Semantic Embedding for Cross-Modal Retrieval"], "answer_arxiv_id": ["1906.04402v2"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_17493"} +{"question": "Which research investigates problems in classical methods of dealing with data imbalance?", "answer": ["Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss", "Decoupling Representation and Classifier for Long-Tailed Recognition"], "answer_arxiv_id": ["1906.07413", "1910.09217"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_17494"} +{"question": "What works are about integrating cross-modal information to leverage knowledge transferred from language or image models for 3D learning?", "answer": ["Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining", "Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?"], "answer_arxiv_id": ["2302.02318", "2212.08320"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_17495"} +{"question": "What research claims that gradient-based explanations are unfaithful?", "answer": ["P"], "answer_arxiv_id": ["0704.0320"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_17496"} +{"question": "Which study proposes to use time-invariant variables in the context of discrete-time stochastic video prediction?", "answer": ["Stochastic Latent Residual Video Prediction"], "answer_arxiv_id": ["2002.09219"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_17497"} +{"question": "Which papers propose solutions to unstable convergence at training time in GANs?", "answer": ["Improved Training of Wasserstein GANs"], "answer_arxiv_id": ["1704.00028"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_17498"} +{"question": "Could you provide the papers where GD attains tighter generalization error and excess risk bounds than those of SGD for convex losses?", "answer": ["Fine-Grained Analysis of Stability and Generalization for Stochastic Gradient Descent"], "answer_arxiv_id": ["2006.08157"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_17499"} +{"question": "What are the works where cocoercivity was applied to the extensions for monotone problems?", "answer": ["Stochastic Gradient Descent-Ascent: Unified Theory and New Efficient Methods"], "answer_arxiv_id": ["2202.07262v3"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_17500"} +{"question": "Which studies demonstrated consistency-training methods which maintain the consistency among the segmentation results of different perturbations of the same unlabeled samples?", "answer": ["Semi-supervised semantic segmentation needs strong, varied perturbations", "Semi-Supervised Semantic Segmentation with Cross-Consistency Training", "Pixel Contrastive-Consistent Semi-Supervised Semantic Segmentation", "Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation", "Dual Student: Breaking the Limits of the Teacher in Semi-supervised Learning"], "answer_arxiv_id": ["1906.01916", "2003.09005", "2108.09025", "2111.12903", "1909.01804"], "source_meta": {"published_time": "20230805"}, "qid": "AutoScholarQuery_train_17501"} +{"question": "Which studies discussed connections between robustness and privacy in distribution learning under the constraint of differential privacy?", "answer": ["Private Hypothesis Selection", "Private Mean Estimation of Heavy-Tailed Distributions", "Covariance-Aware Private Mean Estimation Without Private Covariance Estimation", "Robust and differentially private mean estimation", "Differential privacy and robust statistics in high dimensions", "Private Robust Estimation by Stabilizing Convex Relaxations", "Concentration of the exponential mechanism and differentially private multivariate medians", "Perturbed M-Estimation: A Further Investigation of Robust Statistics for Differential Privacy", "Efficient Mean Estimation with Pure Differential Privacy via a Sum-of-Squares Exponential Mechanism", "Privacy Induces Robustness: Information-Computation Gaps and Sparse Mean Estimation", "Robustness Implies Privacy in Statistical Estimation", "Privately Estimating a Gaussian: Efficient, Robust and Optimal"], "answer_arxiv_id": ["1905.13229", "2002.09464", "2106.13329v3", "2102.09159", "2111.06578", "2112.03548", "2210.06459", "2108.08266", "2111.12981", "2211.00724", "2212.05015v3", "2212.08018v2"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_17502"} +{"question": "Can you name the studies that scrutinize the trade-offs for regular rounding in lossy quantization approaches?", "answer": ["The case for 4-bit precision: k-bit Inference Scaling Laws", "GLM-130B: An Open Bilingual Pre-trained Model", "Efficiently Scaling Transformer Inference"], "answer_arxiv_id": ["2212.09720", "2210.02414", "2211.05102v1"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_17503"} +{"question": "Can you name some studies on deduplication and memorization analysis of large datasets?", "answer": ["Deduplicating Training Data Makes Language Models Better", "The ROOTS Search Tool: Data Transparency for LLMs", "Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpus"], "answer_arxiv_id": ["2107.06499", "2302.14035", "2104.08758"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_17504"} +{"question": "What research developed the forward diffusion process for xt?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_17505"} +{"question": "What are some works that have utilized self-supervised learning in underwater imaging?", "answer": ["Self-Supervised Monocular Depth Underwater"], "answer_arxiv_id": ["2210.03206v1"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_17506"} +{"question": "What papers used deep generative models to recover detailed 3D facial appearance materials?", "answer": ["Photorealistic Facial Texture Inference Using Deep Neural Networks", "UV-GAN: Adversarial Facial UV Map Completion for Pose-invariant Face\n Recognition", "GANFIT: Generative Adversarial Network Fitting for High Fidelity 3D Face\n Reconstruction", "AvatarMe: Realistically Renderable 3D Facial Reconstruction\n \"in-the-wild\"", "Learning Formation of Physically-Based Face Attributes", "AvatarMe++: Facial Shape and BRDF Inference with Photorealistic\n Rendering-Aware GANs", "High-Fidelity 3D Digital Human Head Creation from RGB-D Selfies", "FitMe: Deep Photorealistic 3D Morphable Model Avatars", "Unsupervised High-Fidelity Facial Texture Generation and Reconstruction", "Normalized Avatar Synthesis Using StyleGAN and Perceptual Refinement", "BareSkinNet: De-makeup and De-lighting via 3D Face Reconstruction", "Makeup Extraction of 3D Representation via Illumination-Aware Image\n Decomposition", "FFHQ-UV: Normalized Facial UV-Texture Dataset for 3D Face Reconstruction", "Single-Shot Implicit Morphable Faces with Consistent Texture\n Parameterization", "ClipFace: Text-guided Editing of Textured 3D Morphable Models"], "answer_arxiv_id": ["1612.00523", "1712.04695", "1902.05978", "2003.13845", "2004.03458", "2112.05957", "2010.05562", "2305.09641", "2110.04760", "2106.11423", "2209.09029", "2302.13279", "2211.13874", "2305.03043", "2212.01406"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17507"} +{"question": "What studies have explored the scaling laws in language or vision systems?", "answer": ["Scaling Laws for Neural Language Models", "Revisiting Neural Scaling Laws in Language and Vision"], "answer_arxiv_id": ["2001.08361", "2209.06640"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_17508"} +{"question": "Which papers employed a differentiable point cloud renderer and an inpainter for the more complex scenes?", "answer": ["SynSin: End-to-end View Synthesis from a Single Image"], "answer_arxiv_id": ["1912.08804"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_17509"} +{"question": "Can you list some studies that curate data sets from videos in the natural domain?", "answer": ["MERLOT: Multimodal Neural Script Knowledge Models", "MERLOT Reserve: Neural Script Knowledge through Vision and Language and\n Sound", "Connecting Vision and Language with Video Localized Narratives", "MIMIC-IT: Multi-Modal In-Context Instruction Tuning"], "answer_arxiv_id": ["2106.02636", "2201.02639", "2302.11217", "2306.05425"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17510"} +{"question": "Which works are related to docking in the context of protein structure prediction?", "answer": ["Independent SE(3)-Equivariant Models for End-to-End Rigid Protein Docking", "EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction"], "answer_arxiv_id": ["2111.07786", "2202.05146"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17511"} +{"question": "What are some examples of work on robust RL in Robust Markov Decision Process (RMDP) in the generative model?", "answer": ["Towards Theoretical Understandings of Robust Markov Decision Processes: Sample Complexity and Asymptotics", "Sample Complexity of Robust Reinforcement Learning with a Generative Model", "Distributionally Robust Batch Contextual Bandits", "A Finite Sample Complexity Bound for Distributionally Robust Q-learning", "Improved Sample Complexity Bounds for Distributionally Robust Reinforcement Learning", "Towards Minimax Optimality of Model-based Robust Reinforcement Learning"], "answer_arxiv_id": ["2105.03863", "2112.01506v3", "2006.05630", "2302.13203v2", "2303.02783v2", "2302.05372"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_17512"} +{"question": "What papers have been published focused on efficiently generating larger graphs with diffusion models?", "answer": ["Efficient and Degree-Guided Graph Generation via Discrete Diffusion Modeling"], "answer_arxiv_id": ["2305.04111"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_17513"} +{"question": "What research work exists on the topic of explaining the capability of in-context learning of Transformer?", "answer": ["What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "Transformers Learn In-Context by Gradient Descent", "Transformers as Statisticians: Provable In-Context Learning with In-Context Algorithm Selection", "In-context Learning and Induction Heads", "What learning algorithm is in-context learning? Investigations with linear models", "The Closeness of In-Context Learning and Weight Shifting for Softmax Regression"], "answer_arxiv_id": ["2208.01066", "2212.07677", "2306.04637", "2209.11895v1", "2211.15661", "2304.13276"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_17514"} +{"question": "Could you mention studies that showed the introduction of biases in predictions due to the subject’s class?", "answer": ["E-BERT: Efficient-Yet-Effective Entity Embeddings for BERT"], "answer_arxiv_id": ["1911.03681"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_17515"} +{"question": "Which works were the first to propose a Bernstein-type concentration inequality for self-normalized martingales in linear mixture MDPs?", "answer": ["Nearly Minimax Optimal Reinforcement Learning for Linear Mixture Markov Decision Processes"], "answer_arxiv_id": ["2012.08507"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_17516"} +{"question": "Which paper proposed aggregate backward gradients and LRP throughout all layers and heads in the attention modules in order to derive explanation relevancy?", "answer": ["Transformer Interpretability Beyond Attention Visualization"], "answer_arxiv_id": ["2012.09838"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_17517"} +{"question": "What studies make base and incremental sessions compatible via pseudo-feature, augmentation or finding a flat minima?", "answer": ["Forward Compatible Few-Shot Class-Incremental Learning", "Few-Shot Class-Incremental Learning from an Open-Set Perspective", "Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by Finding Flat Minima"], "answer_arxiv_id": ["2203.06953", "2208.00147", "2111.01549"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_17518"} +{"question": "Could you provide me with papers that proposed re-formulations of self-attention to reduce memory requirements?", "answer": ["Longformer: The Long-Document Transformer", "Nystr\\\"omformer: A Nystr\\\"om-Based Algorithm for Approximating\n Self-Attention", "Perceiver: General Perception with Iterative Attention", "Rethinking Attention with Performers", "Flowformer: Linearizing Transformers with Conservation Flows", "Transformer Quality in Linear Time"], "answer_arxiv_id": ["2004.05150", "2102.03902", "2103.03206", "2009.14794", "2202.06258", "2202.10447"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_17519"} +{"question": "Which works detect Out-of-distribution (OOD) examples based on the Mahalanobis distance?", "answer": ["A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks"], "answer_arxiv_id": ["1807.03888"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_17520"} +{"question": "Which works have developed purification methods based on GANs for adversarial defense?", "answer": ["Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models"], "answer_arxiv_id": ["1805.06605"], "source_meta": {"published_time": "20221101"}, "qid": "AutoScholarQuery_train_17521"} +{"question": "Which paper discussed the need to evaluate both the quality and diversity of the generated samples?", "answer": ["Assessing Generative Models via Precision and Recall"], "answer_arxiv_id": ["1806.00035"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_17522"} +{"question": "Which research papers developed adversarial MDPs with linear-function approximations?", "answer": ["Provably Efficient Exploration in Policy Optimization", "Near-optimal Policy Optimization Algorithms for Learning Adversarial Linear Mixture MDPs", "Online Learning in MDPs with Linear Function Approximation and Bandit Feedback", "Policy Optimization in Adversarial MDPs: Improved Exploration via Dilated Bonuses", "Policy Optimization in Adversarial MDPs: Improved Exploration via Dilated Bonuses"], "answer_arxiv_id": ["1912.05830", "2102.08940v2", "2007.01612", "2107.08346", "2107.08346"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_17523"} +{"question": "Could you provide some works that highlight the continuous need for convolution in order to help the transformer achieve better visual representation?", "answer": ["UniFormer: Unifying Convolution and Self-attention for Visual\n Recognition", "CvT: Introducing Convolutions to Vision Transformers", "Early Convolutions Help Transformers See Better"], "answer_arxiv_id": ["2201.09450", "2103.15808", "2106.14881"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17524"} +{"question": "Which works focused on the classification of all PM games based on their minimax regrets?", "answer": ["Cleaning up the neighborhood: A full classification for adversarial partial monitoring"], "answer_arxiv_id": ["1805.09247"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17525"} +{"question": "Which studies use few-shot learning to adapt existing models to new domains?", "answer": ["Transferable Multi-Domain State Generator for Task-Oriented Dialogue\n Systems", "XQA-DST: Multi-Domain and Multi-Lingual Dialogue State Tracking"], "answer_arxiv_id": ["1905.08743", "2204.05895"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_17526"} +{"question": "Could you provide some studies about temporal reasoning in dialogues?", "answer": ["TimeDial: Temporal Commonsense Reasoning in Dialog"], "answer_arxiv_id": ["2106.04571"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_17527"} +{"question": "What are some studies that enforce 3D detectors to learn 3D cues from diverse geometric constraints?", "answer": ["Total3DUnderstanding: Joint Layout, Object Pose and Mesh Reconstruction\n for Indoor Scenes from a Single Image", "MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty\n Propagation", "Cooperative Holistic Scene Understanding: Unifying 3D Object, Layout,\n and Camera Pose Estimation", "Factoring Shape, Pose, and Layout from the 2D Image of a 3D Scene", "3D-RelNet: Joint Object and Relational Network for 3D Prediction", "Probabilistic and Geometric Depth: Detecting Objects in Perspective"], "answer_arxiv_id": ["2002.12212", "2103.12605", "1810.13049", "1712.01812", "1906.02729", "2107.14160"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_17528"} +{"question": "What studies focus on leveraging NeRF as an external module within a self-content SLAM system?", "answer": ["ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial\n and Multi-Map SLAM", "ORB-SLAM: a Versatile and Accurate Monocular SLAM System", "ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D\n Cameras"], "answer_arxiv_id": ["2007.11898", "1502.00956", "1610.06475"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_17529"} +{"question": "What studies tried to enforce conservation by adding the continuity equation as a soft regularizer using PINNs approach?", "answer": ["USING CONSERVATION LAWS TO INFER DEEP LEARNING MODEL ACCURACY OF RICHTMYER-MESHKOV INSTABILITIES"], "answer_arxiv_id": ["2208.11477"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_17530"} +{"question": "Could you tell me about any research papers that experiment with combining Vision Transformers (ViTs) and convolutions?", "answer": ["MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer"], "answer_arxiv_id": ["2110.02178v2"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_17531"} +{"question": "Which paper initiated the theme of Semantic Scene Completion (SSC)?", "answer": ["Semantic Scene Completion from a Single Depth Image"], "answer_arxiv_id": ["1611.08974"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_17532"} +{"question": "Can you provide studies where only the observation model is non-linear in system identification?", "answer": ["Certainty Equivalent Perception-Based Control", "Learning the Linear Quadratic Regulator from Nonlinear Observations"], "answer_arxiv_id": ["2008.12332", "2010.03799v1"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17533"} +{"question": "Which research works have been done on point cloud diffusion models in 3D generation?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "LION: Latent Point Diffusion Models for 3D Shape Generation", "GECCO: Geometrically-Conditioned Point Diffusion Models", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts"], "answer_arxiv_id": ["2104.03670", "2103.01458", "2210.06978", "2303.05916", "2212.08751"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_17534"} +{"question": "Could you name the studies that introduced methods utilizing self-supervised photometric consistency to NeRF?", "answer": ["Improving neural implicit surfaces geometry with patch warping", "StructNeRF: Neural Radiance Fields for Indoor Scenes with Structural Hints", "Geo-Neus: Geometry-Consistent Neural Implicit Surfaces Learning for Multi-view Reconstruction"], "answer_arxiv_id": ["2112.09648", "2209.05277", "2205.15848"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17535"} +{"question": "Can you name the studies that proposed scaling up the parameters of language models to the billion-scale?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2302.13971", "2307.09288"], "source_meta": {"published_time": "20240605"}, "qid": "AutoScholarQuery_train_17536"} +{"question": "Could you list some works on CAM-based explanations?", "answer": ["Learning Deep Features for Discriminative Localization", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks", "SS-CAM: Smoothed Score-CAM for Sharper Visual Feature Localization"], "answer_arxiv_id": ["1512.04150", "1610.02391", "1910.01279", "2006.14255"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_17537"} +{"question": "Could you list studies that tackle the challenge of learning from low-density regions?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Statistical Efficiency of Score Matching: The View from Isoperimetry", "Generating High Fidelity Data from Low-density Regions using Diffusion Models"], "answer_arxiv_id": ["1907.05600", "2210.00726", "2203.17260"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_17538"} +{"question": "Are there studies that employ spatial cross-correlations between consecutive frames for estimating optical flow or learning motion features for action recognition in the video domain?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "MotionSqueeze: Neural Motion Feature Learning for Video Understanding", "Video Modeling with Correlation Networks"], "answer_arxiv_id": ["1504.06852", "2007.09933", "1906.03349"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_17539"} +{"question": "Which works aim to minimize the distance between samples from the same class while maximizing the distance between samples from different classes?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Exploring Simple Siamese Representation Learning", "Supervised Contrastive Learning"], "answer_arxiv_id": ["2002.05709", "2011.10566", "2004.11362"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_17540"} +{"question": "What papers suggest a linkage between instability across settings and the usage of a single CATE estimator?", "answer": ["Doing Great at Estimating CATE? On the Neglected Assumptions in Benchmark Comparisons of Treatment Effect Estimators"], "answer_arxiv_id": ["2107.13346"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_17541"} +{"question": "What research articles are about token pruning to lessen self-attention complexity in transformers?", "answer": ["TORE: Token Reduction for Efficient Human Mesh Recovery with Transformer", "ClusTR: Exploring Efficient Self-attention via Clustering for Vision\n Transformers", "Beyond Attentive Tokens: Incorporating Token Importance and Diversity\n for Efficient Vision Transformers", "SPViT: Enabling Faster Vision Transformers via Soft Token Pruning", "Making Vision Transformers Efficient from A Token Sparsification View"], "answer_arxiv_id": ["2211.10705", "2208.13138", "2211.11315", "2112.13890", "2303.08685"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_17542"} +{"question": "Can you name studies highlighting the influence of large-scale pre-training on computer vision?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_17543"} +{"question": "Which work mention the use of the CLIP text encoder in Stable Diffusion?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_17544"} +{"question": "What works have applied machine learning techniques, such as kernel-based methods and deep neural networks, to estimate bounds under the marginal sensitivity model?", "answer": ["Interval Estimation of Individual-Level Causal Effects Under Unobserved Confounding", "Quantifying Ignorance in Individual-Level Causal-Effect Estimates under Hidden Confounding"], "answer_arxiv_id": ["1810.02894", "2103.04850"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17545"} +{"question": "What works fall under the in-processing method in algorithmic fairness?", "answer": ["A Reductions Approach to Fair Classification", "Optimization with Non-Differentiable Constraints with Applications to Fairness, Recall, Churn, and Other Goals", "A Confidence-Based Approach for Balancing Fairness and Accuracy", "AdaFair: Cumulative Fairness Adaptive Boosting", "FairXGBoost: Fairness-aware Classification in XGBoost", "Fair Adversarial Gradient Tree Boosting"], "answer_arxiv_id": ["1803.02453", "1809.04198", "1601.05764", "1909.08982", "2009.01442", "1911.05369"], "source_meta": {"published_time": "20220916"}, "qid": "AutoScholarQuery_train_17546"} +{"question": "Which works considered delegation through labeling of data points?", "answer": ["Optimum Statistical Estimation with Strategic Data Sources"], "answer_arxiv_id": ["1408.2539"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_17547"} +{"question": "Could you provide me some deep learning-based studies for co-speech gesture synthesis?", "answer": ["Learning Hierarchical Cross-Modal Association for Co-Speech Gesture\n Generation", "Taming Diffusion Models for Audio-Driven Co-Speech Gesture Generation", "Speech Gesture Generation from the Trimodal Context of Text, Audio, and\n Speaker Identity", "Generating Holistic 3D Human Motion from Speech", "BEAT: A Large-Scale Semantic and Emotional Multi-Modal Dataset for\n Conversational Gestures Synthesis", "EmotionGesture: Audio-Driven Diverse Emotional Co-Speech 3D Gesture\n Generation", "LivelySpeaker: Towards Semantic-Aware Co-Speech Gesture Generation", "GestureDiffuCLIP: Gesture Diffusion Model with CLIP Latents"], "answer_arxiv_id": ["2203.13161", "2303.09119", "2009.02119", "2212.04420", "2203.05297", "2305.18891", "2309.09294", "2303.14613"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17548"} +{"question": "Which study focuses on risk-sensitive reinforcement learning with dynamic risk measures and leverages the Lipschitz property?", "answer": ["Regret Bounds for Risk-sensitive Reinforcement Learning with Lipschitz Dynamic Risk Measures"], "answer_arxiv_id": ["2306.02399"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_17549"} +{"question": "In which paper the risk of SGD with replacement achieving a risk of O​(1∕n) was discussed?", "answer": ["Fine-Grained Analysis of Stability and Generalization for Stochastic Gradient Descent"], "answer_arxiv_id": ["2006.08157"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_17550"} +{"question": "What research extend the sequential sampling procedure to a reparameterizable estimator using REINFORCE?", "answer": ["Estimating Gradients for Discrete Random Variables by Sampling without Replacement"], "answer_arxiv_id": ["2002.06043"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_17551"} +{"question": "Which works are about utilizing the idea of Video Quantized Variational AutoEncoder (VQVAE) in early explorations on text-to-video models?", "answer": ["Phenaki: Variable Length Video Generation From Open Domain Textual\n Description", "CogVideo: Large-scale Pretraining for Text-to-Video Generation via\n Transformers"], "answer_arxiv_id": ["2210.02399", "2205.15868"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_17552"} +{"question": "Which studies focus on designing algorithms and analyses for adversarial learning in queueing systems?", "answer": ["Queue Scheduling with Adversarial Bandit Learning"], "answer_arxiv_id": ["2303.01745"], "source_meta": {"published_time": "20230815"}, "qid": "AutoScholarQuery_train_17553"} +{"question": "What works have focused on generating coherent scenes using autoregressive prior?", "answer": ["Genesis: Generative Scene Inference and Sampling with Object-Centric Latent Representations", "GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinement"], "answer_arxiv_id": ["1907.13052", "2104.09958"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_17554"} +{"question": "What papers studied the global convergences of the learning dynamics of gradient-based methods in an overparameterized regime?", "answer": ["Gradient Descent Provably Optimizes Over-parameterized Neural Networks", "Gradient Descent Finds Global Minima of Deep Neural Networks", "Generalization Guarantees for Neural Networks via Harnessing the Low-rank Structure of the Jacobian", "Overparameterized Nonlinear Learning: Gradient Descent Takes the Shortest Path?", "On the Proof of Global Convergence of Gradient Descent for Deep ReLU Networks with Linear Widths", "Loss landscapes and optimization in over-parameterized non-linear systems and neural networks", "Improved Overparametrization Bounds for Global Convergence of SGD for Shallow Neural Networks", "Convergence of gradient descent for deep neural networks"], "answer_arxiv_id": ["1810.02054", "1811.03804", "1906.05392v2", "1812.10004", "2101.09612", "2003.00307", "2201.12052", "2203.16462"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_17555"} +{"question": "Which studies are about the runtime adaptation of Anytime Neural Networks based on available hardware resources?", "answer": ["S2DNAS: Transforming Static CNN Model for Dynamic Inference via Neural Architecture Search"], "answer_arxiv_id": ["1911.07033"], "source_meta": {"published_time": "20230513"}, "qid": "AutoScholarQuery_train_17556"} +{"question": "Are there any RGB-D methods that treat the depth information as an extra channel of RGB images or convert it into a bird-eye-view image?", "answer": ["Joint 3D Proposal Generation and Object Detection from View Aggregation", "A Unified Framework for Multi-View Multi-Class Object Pose Estimation", "Deep Continuous Fusion for Multi-Sensor 3D Object Detection"], "answer_arxiv_id": ["1712.02294", "1803.08103", "2012.10992"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_17557"} +{"question": "What works focus on backdoor attacks with more requirements (like control model training)?", "answer": ["Input-Aware Dynamic Backdoor Attack", "Few-Shot Backdoor Attacks on Visual Object Tracking", "One-bit Flip is All You Need: When Bit-flip Attack Meets Model Training"], "answer_arxiv_id": ["2010.08138", "2201.13178", "2308.07934"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_17558"} +{"question": "What works adopted graph neural networks in supervised IC?", "answer": ["Auto-Encoding Scene Graphs for Image Captioning", "Exploring Visual Relationship for Image Captioning"], "answer_arxiv_id": ["1812.02378", "1809.07041"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_17559"} +{"question": "Could you list some papers that used deep encoder-decoder networks or flow-based models in image watermarking?", "answer": ["HiDDeN: Hiding Data With Deep Networks", "ReDMark: Framework for Residual Diffusion Watermarking on Deep Networks", "FaceSigns: Semi-Fragile Neural Watermarks for Media Authentication and\n Countering Deepfakes", "SepMark: Deep Separable Watermarking for Unified Source Tracing and\n Deepfake Detection", "Towards Blind Watermarking: Combining Invertible and Non-invertible\n Mechanisms"], "answer_arxiv_id": ["1807.09937", "1810.07248", "2204.01960", "2305.06321", "2212.12678"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_17560"} +{"question": "What studies bring out impressive generation results based on open-vocabulary text descriptions using diffusion models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2205.11487", "2204.06125", "2112.10741", "2112.10752"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_17561"} +{"question": "Which works address conflicting gradients and other optimization challenges in MTL?", "answer": ["Gradient Vaccine: Investigating and Improving Multi-task Optimization in Massively Multilingual Models", "Gradient Surgery for Multi-Task Learning", "Ray Interference: a Source of Plateaus in Deep Reinforcement Learning"], "answer_arxiv_id": ["2010.05874", "2001.06782", "1904.11455"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_17562"} +{"question": "Could you provide me some works that followed the topic of neural ODEs in the GNN field?", "answer": ["Discrete and Continuous Deep Residual Learning Over Graphs", "Graph Neural Ordinary Differential Equations", "Hamiltonian Graph Networks with ODE Integrators"], "answer_arxiv_id": ["1911.09554", "1911.07532", "1909.12790"], "source_meta": {"published_time": "20220611"}, "qid": "AutoScholarQuery_train_17563"} +{"question": "Which works identified the relationship between training amount and model size in language modeling?", "answer": ["Training Compute-Optimal Large Language Models"], "answer_arxiv_id": ["2203.15556"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_17564"} +{"question": "Which paper proposed using an attention-based framework in displaying light fields with view consistency?", "answer": ["Light Field Neural Rendering"], "answer_arxiv_id": ["2112.09687"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_17565"} +{"question": "Which works are related to the personalization of text-to-image diffusion models?", "answer": ["Subject-driven Text-to-Image Generation via Apprenticeship Learning", "Multi-Concept Customization of Text-to-Image Diffusion", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Break-A-Scene: Extracting Multiple Concepts from a Single Image", "Taming Encoder for Zero Fine-tuning Image Customization with\n Text-to-Image Diffusion Models", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Encoder-based Domain Tuning for Fast Personalization of Text-to-Image\n Models", "BLIP-Diffusion: Pre-trained Subject Representation for Controllable\n Text-to-Image Generation and Editing", "Key-Locked Rank One Editing for Text-to-Image Personalization"], "answer_arxiv_id": ["2304.00186", "2212.04488", "2304.03411", "2208.12242", "2305.16311", "2304.02642", "2302.13848", "2208.01618", "2302.12228", "2305.14720", "2305.01644"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_17566"} +{"question": "Which studies have proposed the use of Transformer-based approach in part-aware panoptic segmentation?", "answer": ["Panoptic-PartFormer: Learning a Unified Model for Panoptic Part\n Segmentation", "PanopticPartFormer++: A Unified and Decoupled View for Panoptic Part\n Segmentation"], "answer_arxiv_id": ["2204.04655", "2301.00954"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_17567"} +{"question": "In which research were eigenvectors of adjacency and Laplacian matrices used as extra information to improve the effectiveness of global attention in graph transformers?", "answer": ["Benchmarking Graphormer on Large-Scale Molecular Modeling Datasets", "Rethinking Graph Transformers with Spectral Attention"], "answer_arxiv_id": ["2203.04810", "2106.03893"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_17568"} +{"question": "Could you provide me some works about generating molecules in 3D space?", "answer": ["Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules", "Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models", "Reinforcement Learning for Molecular Design Guided by Quantum Mechanics", "Symmetry-Aware Actor-Critic for 3D Molecular Design"], "answer_arxiv_id": ["1906.00957", "2010.08687", "2002.07717", "2011.12747"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_17569"} +{"question": "What model proposes Discrete Fourier Transform-based attention mechanism with low-rank approximation in frequency?", "answer": ["FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting"], "answer_arxiv_id": ["2201.12740"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_17570"} +{"question": "What studies propose to enhance caption density and quality in vision-language models?", "answer": ["Dense and Aligned Captions (DAC) Promote Compositional Reasoning in VL\n Models"], "answer_arxiv_id": ["2305.19595"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_17571"} +{"question": "Are there any works discussing the role of prompt learning techniques for Vision Language Models?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "MaPLe: Multi-modal Prompt Learning", "Prompt-aligned Gradient for Prompt Tuning"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2210.03117", "2205.14865"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_17572"} +{"question": "Which large-scale video datasets were mentioned for pre-training in the study?", "answer": ["YouTube-8M: A Large-Scale Video Classification Benchmark", "The Kinetics Human Action Video Dataset", "Moments in Time Dataset: one million videos for event understanding"], "answer_arxiv_id": ["1609.08675", "1705.06950", "1801.03150"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_17573"} +{"question": "Which studies introduced the techniques SpIN and NeuralEF to learn spectral decompositions of kernels?", "answer": ["Spectral Inference Networks: Unifying Deep and Spectral Learning", "NeuralEF: Deconstructing Kernels by Deep Neural Networks"], "answer_arxiv_id": ["1806.02215", "2205.00165"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_17574"} +{"question": "What studies demonstrated the effectiveness of learning only a small subset of parameters and keeping the remaining parameters frozen?", "answer": ["Compacter: Efficient Low-Rank Hypercomplex Adapter Layers", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Polyhistor: Parameter-Efficient Multi-Task Adaptation for Dense Vision Tasks", "Learning to Prompt for Continual Learning", "LST: Ladder Side-Tuning for Parameter and Memory Efficient Transfer Learning", "Learning multiple visual domains with residual adapters", "Efficient parametrization of multi-domain deep neural networks"], "answer_arxiv_id": ["2106.04647", "2104.08691", "2210.03265", "2112.08654", "2206.06522", "1705.08045", "1803.10082"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_17575"} +{"question": "Who developed the stepwise methods that surprisingly produced longer proofs?", "answer": ["ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language", "Natural Language Deduction through Search over Statement Compositions", "FaiRR: Faithful and Robust Deductive Reasoning over Natural Language", "Generating Natural Language Proofs with Verifier-Guided Search"], "answer_arxiv_id": ["2012.13048", "2201.06028", "2203.10261", "2205.12443"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_17576"} +{"question": "Could you provide me some references about the Transformer architecture?", "answer": ["Attention Is All You Need", "An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["1706.03762", "2010.11929"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_17577"} +{"question": "Could you give me examples of research that leveraged pre-trained Generative Adversarial Networks (GANs) to improve the super-resolution process?", "answer": ["GLEAN: Generative Latent Bank for Large-Factor Image Super-Resolution", "Image Processing Using Multi-Code GAN Prior", "PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of\n Generative Models", "Best-Buddy GANs for Highly Detailed Image Super-Resolution", "Exploiting Deep Generative Prior for Versatile Image Restoration and\n Manipulation", "Towards Real-World Blind Face Restoration with Generative Facial Prior", "GAN Prior Embedded Network for Blind Face Restoration in the Wild"], "answer_arxiv_id": ["2012.00739", "1912.07116", "2003.03808", "2103.15295", "2003.13659", "2101.04061", "2105.06070"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_17578"} +{"question": "What papers proposed Evol-Instruct for evolving instructions to enhance difficulty and diversity?", "answer": ["WizardLM: Empowering Large Language Models to Follow Complex\n Instructions", "WizardCoder: Empowering Code Large Language Models with Evol-Instruct"], "answer_arxiv_id": ["2304.12244", "2306.08568"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_17579"} +{"question": "What is the study that described a smoothly broken power law functional form in their paper?", "answer": ["Scaling Laws for Transfer"], "answer_arxiv_id": ["2102.01293"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_17580"} +{"question": "Which works developed video unsupervised domain adaptation techniques using adversarial methods?", "answer": ["Video Unsupervised Domain Adaptation with Deep Learning: A Comprehensive\n Survey", "Domain-Adversarial Training of Neural Networks", "Learning Transferable Features with Deep Adaptation Networks", "Temporal Attentive Alignment for Large-Scale Video Domain Adaptation"], "answer_arxiv_id": ["2211.10412", "1505.07818", "1502.02791", "1907.12743"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_17581"} +{"question": "Can you name the studies that take advantage of the semi-linear feature of diffusion processes for an accurate approximation?", "answer": ["UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models", "Fast Sampling of Diffusion Models with Exponential Integrator", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models"], "answer_arxiv_id": ["2302.04867", "2204.13902", "2206.00927", "2211.01095"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_17582"} +{"question": "Which works model the text regions as semantic masks for text detection?", "answer": ["Multi-Oriented Text Detection with Fully Convolutional Networks", "Real-time Scene Text Detection with Differentiable Binarization", "Real-Time Scene Text Detection with Differentiable Binarization and\n Adaptive Scale Fusion", "Shape Robust Text Detection with Progressive Scale Expansion Network", "TextSnake: A Flexible Representation for Detecting Text of Arbitrary\n Shapes"], "answer_arxiv_id": ["1604.04018", "1911.08947", "2202.10304", "1903.12473", "1807.01544"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_17583"} +{"question": "Could you provide me some works classifying distribution shift problems?", "answer": ["Wilds: A Benchmark of in-the-Wild Distribution Shifts"], "answer_arxiv_id": ["2012.07421"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_17584"} +{"question": "Which works made research about the hardness of learning depth-222 ReLU networks without assumptions on the input distribution or weights?", "answer": ["Complexity theoretic limitations on learning DNF’s"], "answer_arxiv_id": ["1404.3378"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_17585"} +{"question": "What articles propose techniques for Dynamic Sparse Training (DST)?", "answer": ["Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science", "Deep Rewiring: Training very sparse deep networks", "Sparse Networks from Scratch: Faster Training without Losing Performance", "Rigging the Lottery: Making All Tickets Winners", "MEST: Accurate and Fast Memory-Economic Sparse Training Framework on the Edge", "Discovering Neural Wirings", "A Brain-inspired Algorithm for Training Highly Sparse Neural Networks", "NeST: A Neural Network Synthesis Tool Based on a Grow-and-Prune Paradigm", "AutoSparse: Towards Automated Sparse Training of Deep neural Networks"], "answer_arxiv_id": ["1707.04780", "1711.05136", "1907.04840", "1911.11134", "2110.14032", "1906.00586", "1903.07138v3", "1711.02017", "2304.06941"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_17586"} +{"question": "Which paper presents an application of Transformers in RL?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_17587"} +{"question": "Can you list down some studies about the fusion of different types of sensors at input level in radar?", "answer": ["Distant Vehicle Detection Using Radar and Vision", "A Deep Learning-based Radar and Camera Sensor Fusion Architecture for Object Detection"], "answer_arxiv_id": ["1901.10951", "2005.07431"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_17588"} +{"question": "What papers show the importance of prior information during the Model Inversion process and propose that proper prior of a generative model could contribute to the inversion of a classification model?", "answer": ["The Secret Revealer: Generative Model-Inversion Attacks Against Deep\n Neural Networks", "Do Gradient Inversion Attacks Make Federated Learning Unsafe?", "Knowledge-Enriched Distributional Model Inversion Attacks"], "answer_arxiv_id": ["1911.07135", "2202.06924", "2010.04092"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_17589"} +{"question": "Could you cite works that have avoided knowledge selection but focused on enhancing knowledge usage during response generation?", "answer": ["Knowledge-Grounded Dialogue Generation with Pre-trained Language Models", "A Three-Stage Learning Framework for Low-Resource Knowledge-Grounded Dialogue Generation"], "answer_arxiv_id": ["2010.08824", "2109.04096"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17590"} +{"question": "Which research aims at maximizing the mutual information shared by all modalities in the context of semi-supervised multimodal learning?", "answer": ["TCGM: An Information-Theoretic Framework for Semi-Supervised Multi-Modality Learning"], "answer_arxiv_id": ["2007.06793"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_17591"} +{"question": "What papers discuss the challenges faced by automated metrics in handling the diversity of potential outputs?", "answer": ["Evaluation of Text Generation: A Survey", "Hurdles to Progress in Long-form Question Answering"], "answer_arxiv_id": ["2006.14799", "2103.06332"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_17592"} +{"question": "Could you provide me some studies stating that fine-tuning the self-supervised model performs well for semi-supervised learning?", "answer": ["S4L: Self-Supervised Semi-Supervised Learning", "Semi-Supervised Learning with Scarce Annotations", "Big Self-Supervised Models are Strong Semi-Supervised Learners", "Semi-supervised Vision Transformers at Scale"], "answer_arxiv_id": ["1905.03670", "1905.08845", "2006.10029", "2208.05688"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_17593"} +{"question": "What works discuss free-form explanations in interpreting NLP models?", "answer": ["Explain Yourself! Leveraging Language Models for Commonsense Reasoning", "Investigating the Benefits of Free-Form Rationales", "Measuring Association Between Labels and Free-Text Rationales", "Reframing Human-AI Collaboration for Generating Free-Text Explanations", "PINTO: Faithful Language Reasoning Using Prompt-Generated Rationales", "Explanation-based Finetuning Makes Models More Robust to Spurious Cues"], "answer_arxiv_id": ["1906.02361", "2206.11083", "2010.12762", "2112.08674", "2211.01562", "2305.04990"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_17594"} +{"question": "Which works developed innovative model architectures for automatic multi-organ segmentation?", "answer": ["C2FNAS: Coarse-to-Fine Neural Architecture Search for 3D Medical Image\n Segmentation", "Multi-organ Segmentation over Partially Labeled Datasets with\n Multi-scale Feature Abstraction", "DoDNet: Learning to segment multi-organ and tumors from multiple\n partially labeled datasets"], "answer_arxiv_id": ["1912.09628", "2001.00208", "2011.10217"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17595"} +{"question": "Can you mention some studies that focus on open-world scene understanding, identifying novel object categories?", "answer": ["Zero-shot point cloud segmentation by transferring geometric primitives", "Generative Zero-Shot Learning for Semantic Segmentation of 3D Point Clouds", "See More and Know More: Zero-shot Point Cloud Segmentation via Multi-modal Visual Data", "OpenScene: 3D Scene Understanding with Open Vocabularies", "Novel Class Discovery for 3D Point Cloud Semantic Segmentation", "Language-driven Semantic Segmentation", "Zero-Shot Semantic Segmentation"], "answer_arxiv_id": ["2210.09923v3", "2108.06230", "2307.10782", "2211.15654", "2303.11610", "2201.03546", "1906.00817"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17596"} +{"question": "What paper presents a generative method for estimating pose and global translation?", "answer": ["MOVIN: Real-time Motion Capture using a Single LiDAR"], "answer_arxiv_id": ["2309.09314"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17597"} +{"question": "Which works approach to compute the OT plan through the primal formulation, which involves the usage of generative models?", "answer": ["Large-Scale Optimal Transport via Adversarial Training with Cycle-Consistency", "On Scalable and Efficient Computation of Large Scale Optimal Transport"], "answer_arxiv_id": ["2003.06635", "1905.00158"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_17598"} +{"question": "What papers have contributed to developing and benchmarking machine learning models to improve and analyze compositional generalization abilities?", "answer": ["Unsupervised Learning of Compositional Energy Concepts", "Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMC", "Compositional Visual Generation with Composable Diffusion Models", "Prompting Large Pre-trained Vision-Language Models For Compositional Concept Learning", "The Role of Syntactic Planning in Compositional Image Captioning", "Compositional Generalization in Grounded Language Learning via Induced Model Sparsity", "Ablating Concepts in Text-to-Image Diffusion Models", "Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality", "Measuring Compositionality in Representation Learning", "Does CLIP Bind Concepts? Probing Compositionality in Large Image Models", "Generalization without Systematicity: On the Compositional Skills of Sequence-to-Sequence Recurrent Networks", "Break It Down: Evidence for Structural Compositionality in Neural Networks", "CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning", "Testing Relational Understanding in Text-Guided Image Generation", "Visual Representation Learning Does Not Generalize Strongly Within the Same Domain", "Benchmarking Spatial Relationships in Text-to-Image Generation", "Benchmarking Compositionality with Formal Languages"], "answer_arxiv_id": ["2111.03042", "2302.11552", "2206.01714", "2211.05077", "2101.11911", "2207.02518", "2303.13516", "2204.03162", "1902.07181", "2212.10537", "1711.00350", "2301.10884", "1612.06890", "2208.00005", "2107.08221", "2212.10015v3", "2208.08195"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_17599"} +{"question": "What works leverage Data Flow Graphs and Control Flow Graphs for grounding program understanding?", "answer": ["funcGNN: A Graph Neural Network Approach to Program Similarity"], "answer_arxiv_id": ["2007.13239"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_17600"} +{"question": "What papers have developed robustness verification methods for transformers?", "answer": ["Robustness Verification for Transformers"], "answer_arxiv_id": ["2002.06622"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_17601"} +{"question": "What researches mentioned prevalent realizations of anytime-valid confidence intervals based on specific estimators such as the least-squares estimator and noise sub-families?", "answer": ["Improved Optimistic Algorithms for Logistic Bandits"], "answer_arxiv_id": ["2002.07530"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_17602"} +{"question": "Are there any studies which proposed methods to generate reflection images more consistent with real-world scenarios?", "answer": ["Single Image Reflection Removal with Physically-Based Training Images"], "answer_arxiv_id": ["1904.11934"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17603"} +{"question": "Any works about the use of diffusion models for human motion priors?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Human Motion Diffusion Model", "Human Motion Diffusion as a Generative Prior", "PhysDiff: Physics-Guided Human Motion Diffusion Model"], "answer_arxiv_id": ["2006.11239", "2010.02502", "2209.14916", "2303.01418", "2212.02500"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_17604"} +{"question": "In which papers the knowledge of LLMs is hard to identify or control, sensitive to specific words or phrases, and only emitted when prompted appropriately?", "answer": ["Quantifying Memorization Across Neural Language Models", "Analyzing Commonsense Emergence in Few-shot Knowledge Models", "Measuring and Improving Consistency in Pretrained Language Models"], "answer_arxiv_id": ["2202.07646", "2101.00297", "2102.01017"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_17605"} +{"question": "Could you provide me some works about dynamic bit-width neural networks?", "answer": ["AdaBits: Neural Network Quantization with Adaptive Bit-Widths", "Arbitrary Bit-width Network: A Joint Layer-Wise Quantization and\n Adaptive Inference Approach", "Bit-Mixer: Mixed-precision networks with runtime bit-width selection", "EQ-Net: Elastic Quantization Neural Networks"], "answer_arxiv_id": ["1912.09666", "2204.09992", "2103.17267", "2308.07650"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_17606"} +{"question": "What papers discuss local editing approaches in text-to-image models?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images", "Blended Latent Diffusion", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2111.14818", "2206.02779", "2112.10741"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_17607"} +{"question": "Which studies proposed methods that perform a projection of 3D input to a unit sphere?", "answer": ["Spherical CNNs", "Learning SO(3) Equivariant Representations with Spherical CNNs"], "answer_arxiv_id": ["1801.10130", "1711.06721"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_17608"} +{"question": "What studies provide an insight into the expressiveness of Graphical Neural Networks (GNNs) with respect to the Weisfeiler-Leman (WL) graph isomorphism test?", "answer": ["How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["1810.00826"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_17609"} +{"question": "What works differ from the current research but also utilize transformers?", "answer": ["Vision Transformer for NeRF-Based View Synthesis from a Single Input Image", "Scene Representation Transformer: Geometry-Free Novel View Synthesis Through Set-Latent Scene Representations"], "answer_arxiv_id": ["2207.05736", "2111.13152"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_17610"} +{"question": "What work is done on creating spatially grounded instruction-tuning datasets in the medical domain using videos?", "answer": ["Quilt-1M: One Million Image-Text Pairs for Histopathology"], "answer_arxiv_id": ["2306.11207"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17611"} +{"question": "Which works have used diffusion models in the field of image generation?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2204.06125", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17612"} +{"question": "What studies have improved the performance of two-stage action localisation models by incorporating more contextual information using feature banks extracted from additional frames in the video?", "answer": ["MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient\n Long-Term Video Recognition", "Actor-Context-Actor Relation Network for Spatio-Temporal Action\n Localization", "Asynchronous Interaction Aggregation for Action Detection", "Long-Term Feature Banks for Detailed Video Understanding"], "answer_arxiv_id": ["2201.08383", "2006.07976", "2004.07485", "1812.05038"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_17613"} +{"question": "Could you provide me examples of studies that introduced Uncertainty estimation to overcome overestimation in offline RL?", "answer": ["Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble", "Pessimistic Bootstrapping for Uncertainty-Driven Offline Reinforcement Learning"], "answer_arxiv_id": ["2110.01548", "2202.11566"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_17614"} +{"question": "Could you provide me with some works that have focused on contrastive learning in the domain of deep graph clustering?", "answer": ["Unifying Graph Contrastive Learning with Flexible Contextual Scopes"], "answer_arxiv_id": ["2210.08792"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_17615"} +{"question": "Could you provide the work that proposes a direct adversarial attack on Stable Diffusion by maximizing the Mean Squared Error loss during the optimization process?", "answer": ["Adversarial Example Does Good: Preventing Painting Imitation from\n Diffusion Models via Adversarial Examples"], "answer_arxiv_id": ["2302.04578"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_17616"} +{"question": "Is there any research that leverages representations of writing styles for authorship attribution?", "answer": ["Few-Shot Detection of Machine-Generated Text using Style Representations"], "answer_arxiv_id": ["2401.06712"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_17617"} +{"question": "What papers discuss the use of out-of-distribution data during the training of evidential models?", "answer": ["Reverse KL-Divergence Training of Prior Networks: Improved Uncertainty and Adversarial Robustness", "Predictive Uncertainty Estimation via Prior Networks"], "answer_arxiv_id": ["1905.13472", "1802.10501"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_17618"} +{"question": "What papers discussed invariant learning by separating semantic factors of variations in data using disentangle-based methods?", "answer": ["Representation Learning: A Review and New Perspectives", "Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations"], "answer_arxiv_id": ["1206.5538", "1811.12359"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_17619"} +{"question": "Who described methods of using LiDAR for human body segmentation?", "answer": ["STCrowd: A Multimodal Dataset for Pedestrian Perception in Crowded\n Scenes"], "answer_arxiv_id": ["2204.01026"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_17620"} +{"question": "Which work integrates graphical models with GANs for structured data?", "answer": ["Graphical Generative Adversarial Networks"], "answer_arxiv_id": ["1804.03429"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_17621"} +{"question": "Are there any works that investigate the relationship between causal models at different levels of coarse-graining or abstraction?", "answer": ["Visual Causal Feature Learning", "Multi-Level Cause-Effect Systems", "Causal Consistency of Structural Equation Models", "Abstracting Causal Models", "Abstraction between Structural Causal Models: A Review of Definitions and Properties", "Jointly Learning Consistent Causal Abstractions Over Multiple Interventional Distributions"], "answer_arxiv_id": ["1412.2309", "1512.07942v1", "1707.00819", "1812.03789", "2207.08603", "2301.05893"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17622"} +{"question": "Which researches worked on introducing semi-autoregressive decoding?", "answer": ["Semi-Autoregressive Neural Machine Translation", "Learning to Recover from Multi-Modality Errors for Non-Autoregressive Neural Machine Translation", "Semi-Autoregressive Training Improves Mask-Predict Decoding", "Hybrid-Regressive Neural Machine Translation"], "answer_arxiv_id": ["1808.08583", "2006.05165", "2001.08785", "2210.10416"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_17623"} +{"question": "Could you mention the work which demonstrated the potential of graph hypernetworks for training with RL to estimate variable architecture policies for locomotion and manipulation?", "answer": ["Efficiently Learning Small Policies for Locomotion and Manipulation"], "answer_arxiv_id": ["2210.00140"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17624"} +{"question": "What studies reported the importance of extrapolation in the context of language models?", "answer": ["Generalization without Systematicity: On the Compositional Skills of Sequence-to-Sequence Recurrent Networks", "On Evaluating the Generalization of LSTM Models in Formal Languages"], "answer_arxiv_id": ["1711.00350", "1811.01001"], "source_meta": {"published_time": "20200415"}, "qid": "AutoScholarQuery_train_17625"} +{"question": "What studies have proposed complex architectures for the alignment module in LVLMs?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "OpenFlamingo: An Open-Source Framework for Training Large Autoregressive\n Vision-Language Models"], "answer_arxiv_id": ["2204.14198", "2308.01390"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_17626"} +{"question": "What's the paper that proposed the test-time prompt tuning (TPT) in prompt engineering?", "answer": ["Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models"], "answer_arxiv_id": ["2209.07511"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_17627"} +{"question": "What works have been done on achieving uncertainty estimates through confidence calibration?", "answer": ["On Calibration of Modern Neural Networks", "Taking a Step Back with KCal: Multi-Class Kernel-Based Calibration for Deep Neural Networks", "Revisiting the Calibration of Modern Neural Networks"], "answer_arxiv_id": ["1706.04599", "2202.07679", "2106.07998"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_17628"} +{"question": "Which works are related to Multi-input multi-output (MIMO) settings in multi-task learning?", "answer": ["Variational Multi-Task Learning with Gumbel-Softmax Priors", "Learning Multiple Tasks with Multilinear Relationship Networks", "Association Graph Learning for Multi-Task Classification with Category Shifts", "Scheduled Multi-task Learning for Neural Chat Translation"], "answer_arxiv_id": ["2111.05323", "1506.02117", "2210.04637", "2205.03766"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_17629"} +{"question": "What research focuses on the application of short, localized convolutions?", "answer": ["Patches Are All You Need?"], "answer_arxiv_id": ["2201.09792"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_17630"} +{"question": "What works specifically evaluated the object hallucinations of LVLMs?", "answer": ["Evaluating Object Hallucination in Large Vision-Language Models"], "answer_arxiv_id": ["2305.10355"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_17631"} +{"question": "Which papers explored early approaches for visual representation learning such as inpainting?", "answer": ["Context Encoders: Feature Learning by Inpainting"], "answer_arxiv_id": ["1604.07379"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17632"} +{"question": "What is the best-performing model in the survey on dynamical VAEs for time series data?", "answer": ["Sequential Neural Models with Stochastic Layers"], "answer_arxiv_id": ["1605.07571"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_17633"} +{"question": "What studies demonstrated learning of an animatable avatar of a human face from only monocular video?", "answer": ["I M Avatar: Implicit Morphable Head Avatars from Videos", "PointAvatar: Deformable Point-based Head Avatars from Videos"], "answer_arxiv_id": ["2112.07471", "2212.08377"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_17634"} +{"question": "Which work introduces the PointOE module for generating point-wise local descriptors?", "answer": ["SOE-Net: A Self-Attention and Orientation Encoding Network for Point\n Cloud based Place Recognition"], "answer_arxiv_id": ["2011.12430"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_17635"} +{"question": "Which works delve into the implications of using non-convergent short-run MCMC for sampling in EBMs?", "answer": ["Learning Non-Convergent Non-Persistent Short-Run MCMC Toward Energy-Based Model", "Latent Diffusion Energy-Based Model for Interpretable Text Modeling", "Learning Energy-Based Models by Diffusion Recovery Likelihood"], "answer_arxiv_id": ["1904.09770", "2206.05895", "2012.08125"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_17636"} +{"question": "What work studies L2-regularized optimal transport problems?", "answer": ["Smooth and Sparse Optimal Transport"], "answer_arxiv_id": ["1710.06276"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_17637"} +{"question": "Could you provide me studies that implemented regularization term based on Mixup to enhance incomplete labels?", "answer": ["A Variational Approach for Learning from Positive and Unlabeled Data", "Positive Label Is All You Need for Multi-Label Classification"], "answer_arxiv_id": ["1906.00642", "2306.16016"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_17638"} +{"question": "Which study shows convergence in the two-time-scaled GDA with non-adaptive stepsize or Adam, given an asymptotically stable attractor exists?", "answer": ["GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium"], "answer_arxiv_id": ["1706.08500"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_17639"} +{"question": "Can you provide works that focus on the computation of the Euclidean distance degree for the purpose of 3D reconstruction?", "answer": ["Euclidean distance degree of the multiview variety", "Theoretical and Numerical Analysis of 3D Reconstruction Using Point and\n Line Incidences"], "answer_arxiv_id": ["1812.05648", "2303.13593"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_17640"} +{"question": "Are there any works providing evidence that performance of downstream models correlates with training set diversity?", "answer": ["An Investigation of the (In)effectiveness of Counterfactually Augmented\n Data", "Toward Learning Human-aligned Cross-domain Robust Models by Countering\n Misaligned Features"], "answer_arxiv_id": ["2107.00753", "2111.03740"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_17641"} +{"question": "Which papers proposed advanced variance-reduction techniques in improving the sample complexity of solving finite-sum coupled compositional optimization problems?", "answer": ["Multi-block-Single-probe Variance Reduced Estimator for Coupled Compositional Optimization"], "answer_arxiv_id": ["2207.08540"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_17642"} +{"question": "What works utilize pre-trained text-to-video diffusion model in motion transfer?", "answer": ["ModelScope Text-to-Video Technical Report"], "answer_arxiv_id": ["2308.06571"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_17643"} +{"question": "What studies utilized transformers for video-based classification or recognition?", "answer": ["ViViT: A Video Vision Transformer", "Video Swin Transformer", "VidTr: Video Transformer Without Convolutions", "Is Space-Time Attention All You Need for Video Understanding?"], "answer_arxiv_id": ["2103.15691", "2106.13230", "2104.11746", "2102.05095"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_17644"} +{"question": "What research proposes the approach that selects top K most confident examples per class to improve performance in pseudolabeling?", "answer": ["Unsupervised Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2204.03649"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_17645"} +{"question": "Which papers propose combining type checking with inference-time search for type prediction?", "answer": ["TypeWriter: Neural Type Prediction with Search-based Validation", "Static Inference Meets Deep Learning: A Hybrid Type Inference Approach for Python"], "answer_arxiv_id": ["1912.03768", "2105.03595v2"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_17646"} +{"question": "Which works suggest that promoting orthogonality on each layer is a more effective way to reduce the Lipschitz constant and increase robustness?", "answer": ["Sorting Out Lipschitz Function Approximation", "Parseval Networks: Improving Robustness to Adversarial Examples"], "answer_arxiv_id": ["1811.05381", "1704.08847"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_17647"} +{"question": "Which work investigates the under-explored text encoder in T2I pipelines?", "answer": ["The Hidden Language of Diffusion Models"], "answer_arxiv_id": ["2306.00966"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_17648"} +{"question": "Are there any works that utilize the hyperlinks between Wikipedia pages in pre-training tailored for dense retrieval?", "answer": ["Pre-training for Ad-hoc Retrieval: Hyperlink is Also You Need", "Hyperlink-induced Pre-training for Passage Retrieval in Open-domain Question Answering"], "answer_arxiv_id": ["2108.09346", "2203.06942"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_17649"} +{"question": "What research extends AMP to produce motions facilitating interactions with the scene?", "answer": ["Synthesizing Physical Character-Scene Interactions"], "answer_arxiv_id": ["2302.00883"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_17650"} +{"question": "Are there any other gradient-based methods proposed besides TracIn?", "answer": ["Input Similarity from the Neural Network Perspective", "Evaluation of Similarity-based Explanations"], "answer_arxiv_id": ["2102.05262", "2006.04528"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17651"} +{"question": "In what paper does a masked-autoencoder trained on diverse video data serve as an effective visual embedding for online RL?", "answer": ["Masked Visual Pre-training for Motor Control", "Masked Autoencoders Are Scalable Vision Learners", "The “something something” video database for learning and evaluating visual common sense", "Understanding Human Hands in Contact at Internet Scale"], "answer_arxiv_id": ["2203.06173", "2111.06377", "1706.04261", "2006.06669"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_17652"} +{"question": "Which works explore how to optimize the reasoning pipeline for chain-of-thought (CoT) methods in LLMs using approaches such as programming, problem decomposition, and calibration coordination?", "answer": ["Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Decomposed Prompting: A Modular Approach for Solving Complex Tasks", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Complexity-Based Prompting for Multi-step Reasoning"], "answer_arxiv_id": ["2211.12588", "2205.10625", "2210.02406", "2203.11171", "2210.00720"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_17653"} +{"question": "Could you provide me some works related to mask image modeling for self-supervised training?", "answer": ["Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2111.06377"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17654"} +{"question": "Which works analyzed the convergence of decentralized temporal difference (TD) learning?", "answer": ["On the Convergence of Consensus Algorithms with Markovian Noise and Gradient Bias"], "answer_arxiv_id": ["2008.07841"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17655"} +{"question": "What works address context-based meta-RL based on a PODMP model?", "answer": ["Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables", "VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning", "Context-aware Dynamics Model for Generalization in Model-Based Reinforcement Learning", "AdaRL: What, Where, and How to Adapt in Transfer Reinforcement Learning"], "answer_arxiv_id": ["1903.08254", "1910.08348", "2005.06800", "2107.02729"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_17656"} +{"question": "Could you provide me some studies that employ the sparse-voxel data structure to scale to large scenes?", "answer": ["4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "Cylinder3D: An Effective 3D Framework for Driving-scene LiDAR Semantic\n Segmentation", "PVNAS: 3D Neural Architecture Search with Point-Voxel Convolution", "RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR\n Point Cloud Segmentation", "(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection\n for Sparse Semantic Segmentation Network", "RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR\n Point Cloud Segmentation", "Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning\n Contextual Shape Priors from Scene Completion", "Point-to-Voxel Knowledge Distillation for LiDAR Semantic Segmentation", "2DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Clouds"], "answer_arxiv_id": ["1904.08755", "1812.05784", "2008.01550", "2204.11797", "2103.12978", "2102.04530", "2103.12978", "2012.03762", "2206.02099", "2207.04397"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_17657"} +{"question": "What studies achieved optimal convergence rate for negatively comonotone operators?", "answer": ["Fast Extra Gradient Methods for Smooth Structured Nonconvex-Nonconcave Minimax Problems"], "answer_arxiv_id": ["2106.02326"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_17658"} +{"question": "Which work refines the prompts for the text encoder of CLIP model by formulating them as learnable variables?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_17659"} +{"question": "Which paper introduced Sparse Tensor Cores in the NVIDIA Ampere GPU architecture?", "answer": ["Accelerating Sparse Deep Neural Networks"], "answer_arxiv_id": ["2104.08378"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_17660"} +{"question": "Could you provide me some research that discusses using normalizing flows for negative samples generation and a latent space in EBM training?", "answer": ["Flow Contrastive Estimation of Energy-Based Models", "MCMC Should Mix: Learning Energy-Based Model with Neural Transport Latent Space MCMC", "A Tale of Two Flows: Cooperative Learning of Langevin Flow and Normalizing Flow Toward Energy-Based Model"], "answer_arxiv_id": ["1912.00589", "2006.06897", "2205.06924"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_17661"} +{"question": "Which works proposed methods for face anonymization?", "answer": ["Natural and Effective Obfuscation by Head Inpainting", "DeepPrivacy: A Generative Adversarial Network for Face Anonymization", "AnonymousNet: Natural Face De-Identification with Measurable Privacy", "Password-conditioned Anonymization and Deanonymization with Face\n Identity Transformers", "CIAGAN: Conditional Identity Anonymization Generative Adversarial\n Networks", "PrivacyNet: Semi-Adversarial Networks for Multi-attribute Face Privacy", "Differentially Private Imaging via Latent Space Manipulation", "Perceptual Indistinguishability-Net (PI-Net): Facial Image Obfuscation\n with Manipulable Semantics", "The UU-Net: Reversible Face De-Identification for Visual Surveillance\n Video Footage", "CFA-Net: Controllable Face Anonymization Network with Identity\n Representation Manipulation", "IdentityDP: Differential Private Identification Protection for Face\n Images", "On Generating Identifiable Virtual Faces"], "answer_arxiv_id": ["1711.09001", "1909.04538", "1904.12620", "1911.11759", "2005.09544", "2001.00561", "2103.05472", "2104.01753", "2007.04316", "2105.11137", "2103.01745", "2110.07986"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_17662"} +{"question": "Which works produce pseudo-labels to perform supervised training in test-time adaptation?", "answer": ["Test-Time Adaptation via Conjugate Pseudo-labels", "If your data distribution shifts, use self-learning"], "answer_arxiv_id": ["2207.09640", "2104.12928"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_17663"} +{"question": "What studies have utilized DPMs for generating protein backbones?", "answer": ["SE⁢(3) diffusion model with application to protein backbone generation", "Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem", "Two for One: Diffusion Models and Force Fields for Coarse-Grained Molecular Dynamics", "DiffMD: A Geometric Diffusion Model for Molecular Dynamics Simulations"], "answer_arxiv_id": ["2302.02277", "2206.04119", "2302.00600", "2204.08672"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_17664"} +{"question": "Which studies use 3D measurements for pose estimation methods?", "answer": ["PointNetLK: Robust & Efficient Point Cloud Registration using PointNet", "Deep Closest Point: Learning Representations for Point Cloud\n Registration", "PRNet: Self-Supervised Learning for Partial-to-Partial Registration", "RPM-Net: Robust Point Matching using Learned Features", "DeepGMR: Learning Latent Gaussian Mixture Models for Registration", "MatchNorm: Learning-based Point Cloud Registration for 6D Object Pose\n Estimation in the Real World"], "answer_arxiv_id": ["1903.05711", "1905.03304", "1910.12240", "2003.13479", "2008.09088v1", "2203.15309"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_17665"} +{"question": "Is there any work that showed Transformers attention is a special case of the MHN association?", "answer": ["Hopfield Networks is All You Need"], "answer_arxiv_id": ["2008.02217"], "source_meta": {"published_time": "20230322"}, "qid": "AutoScholarQuery_train_17666"} +{"question": "Could you mention studies that proposed models for body gesture co-speech animation?", "answer": ["Style Transfer for Co-Speech Gesture Animation: A Multi-Speaker\n Conditional-Mixture Approach", "Moving fast and slow: Analysis of representations and post-processing in\n speech-driven automatic gesture generation", "Audio2Gestures: Generating Diverse Gestures from Speech Audio with\n Conditional Variational Autoencoders", "Speech Drives Templates: Co-Speech Gesture Synthesis with Learned\n Templates", "MoGlow: Probabilistic and controllable motion synthesis using\n normalising flows"], "answer_arxiv_id": ["2007.12553", "2007.09170", "2108.06720", "2108.08020", "1905.06598"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_17667"} +{"question": "Which studies show that adversarial training requires more data, compared to its standard training counterpart, to achieve the same level of generalization?", "answer": ["Adversarially Robust Generalization Requires More Data"], "answer_arxiv_id": ["1804.11285"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_17668"} +{"question": "What studies proposed new tasks such as Subject-driven image generation and editing and Multi-concept image composition?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "DreamEdit: Subject-driven Image Editing", "Multi-Concept Customization of Text-to-Image Diffusion", "Cones 2: Customizable Image Synthesis with Multiple Subjects"], "answer_arxiv_id": ["2208.01618", "2208.12242", "2306.12624", "2212.04488", "2305.19327"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_17669"} +{"question": "Which studies combined DM and IS for robust OPE?", "answer": ["Doubly Robust Off-policy Value Evaluation for Reinforcement Learning", "Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning", "More Robust Doubly Robust Off-policy Evaluation", "Doubly Robust Bias Reduction in Infinite Horizon Off-Policy Estimation", "Minimax Weight and Q-Function Learning for Off-Policy Evaluation", "Deeply-Debiased Off-Policy Interval Estimation", "Batch Policy Learning in Average Reward Markov Decision Processes", "Efficiently Breaking the Curse of Horizon in Off-Policy Evaluation with Double Reinforcement Learning"], "answer_arxiv_id": ["1511.03722", "1604.00923", "1802.03493", "1910.07186", "1910.12809", "2105.04646", "2007.11771", "1909.05850v6"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_17670"} +{"question": "What works propose selection algorithms to yield the smallest conformal prediction intervals given a family of learning algorithms?", "answer": ["Finite-sample Efficient Conformal Prediction"], "answer_arxiv_id": ["2104.13871"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17671"} +{"question": "Which studies proposed self-training in semi-supervised learning?", "answer": ["Self-training with Noisy Student improves ImageNet classification", "Rethinking Pre-training and Self-training", "Meta Pseudo Labels"], "answer_arxiv_id": ["1911.04252", "2006.06882", "2003.10580"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_17672"} +{"question": "Could you provide some examples of works that applied reinforcement learning for solving inverse problems?", "answer": ["Model-based Reinforcement Learning: A Survey.", "Soft Actor-Critic Algorithms and Applications", "A Deep Reinforcement Learning Approach for Dynamically Stable Inverse Kinematics of Humanoid Robots"], "answer_arxiv_id": ["2006.16712", "1812.05905", "1801.10425v1"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_17673"} +{"question": "What papers have been proposed to improve computational efficiency in vision transformers?", "answer": ["Quadtree Attention for Vision Transformers", "Axial Attention In Multidimensional Transformers", "DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification"], "answer_arxiv_id": ["2201.02767", "1912.12180", "2106.02034"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_17674"} +{"question": "Can you specify which work did the causal policy in this research most resemble to?", "answer": ["Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos"], "answer_arxiv_id": ["2206.11795"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_17675"} +{"question": "What research have been done related to SSC in path planning?", "answer": ["Object Goal Navigation using Goal-Oriented Semantic Exploration"], "answer_arxiv_id": ["2007.00643"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_17676"} +{"question": "Which work introduced Neural radiance fields (NeRF), a technique to encode scene geometry in a compact neural network?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_17677"} +{"question": "Which papers discussed memory compression using quantization in the context of model parameters storage?", "answer": ["Understanding and Overcoming the Challenges of Efficient Transformer Quantization", "Post training 4-bit quantization of convolutional networks for rapid-deployment", "Quantizing deep convolutional networks for efficient inference: A whitepaper"], "answer_arxiv_id": ["2109.12948", "1810.05723", "1806.08342"], "source_meta": {"published_time": "20220201"}, "qid": "AutoScholarQuery_train_17678"} +{"question": "What studies developed the detection-based approach for handling dynamic scenes in NeRF?", "answer": ["Neural Scene Graphs for Dynamic Scenes", "Panoptic Neural Fields: A Semantic Object-Aware Neural Scene\n Representation"], "answer_arxiv_id": ["2011.10379", "2205.04334"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17679"} +{"question": "Are there any works that use sentence embedding for classifying input at test time to the semantically closest class seen during training?", "answer": ["Matching Networks for One Shot Learning", "Prototypical Networks for Few-shot Learning", "SentEval: An Evaluation Toolkit for Universal Sentence Representations"], "answer_arxiv_id": ["1606.04080", "1703.05175", "1803.05449"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_17680"} +{"question": "What sources provide detailed overviews of stochastic convex optimization methods?", "answer": ["4"], "answer_arxiv_id": ["2304.09215"], "source_meta": {"published_time": "20230803"}, "qid": "AutoScholarQuery_train_17681"} +{"question": "What studies in the field of object-centric learning use the approach of Vision Transformers (ViT)?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_17682"} +{"question": "Which papers have explored the use of part-based models in pose estimation?", "answer": ["Unsupervised Part-Based Disentangling of Object Shape and Appearance", "Learning Local RGB-to-CAD Correspondences for Object Pose Estimation"], "answer_arxiv_id": ["1903.06946", "1811.07249"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_17683"} +{"question": "Could you tell me about the studies that added a layer based on functional principal components to transform the scalar-valued output back to functional data or proposed fully functional neurons?", "answer": ["Remaining Useful Life Estimation Using Functional Data Analysis", "A Non-linear Function-on-Function Model for Regression with Time Series Data", "Non-linear Functional Modeling using Neural Networks", "Modern Non-Linear Function-on-Function Regression"], "answer_arxiv_id": ["1904.06442", "2011.12378", "2104.09371", "2107.14151"], "source_meta": {"published_time": "20230114"}, "qid": "AutoScholarQuery_train_17684"} +{"question": "Which works have investigated the presence of noisy, hard-to-learn, and/or negatively influential samples in popular vision benchmarks?", "answer": ["Deconstructing Distributions: A Pointwise Framework of Learning", "Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamics", "Metadata Archaeology: Unearthing Data Subsets by Leveraging Training Dynamics", "A Data-Based Perspective on Transfer Learning", "Characterizing Datapoints via Second-Split Forgetting"], "answer_arxiv_id": ["2202.09931", "2009.10795", "2209.10015", "2207.05739", "2210.15031"], "source_meta": {"published_time": "20220823"}, "qid": "AutoScholarQuery_train_17685"} +{"question": "Could you provide me with some recent works that use the self-supervised method?", "answer": ["DeCap: Decoding CLIP Latents for Zero-Shot Captioning via Text-Only\n Training", "Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image\n Retrieval", "Zero-Shot Composed Image Retrieval with Textual Inversion", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Key-Locked Rank One Editing for Text-to-Image Personalization"], "answer_arxiv_id": ["2303.03032", "2302.03084", "2303.15247", "2208.01618", "2305.01644"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17686"} +{"question": "Which papers proposed model-based methods that estimate the uncertainty via ensembles to address DS problem?", "answer": ["When to Trust Your Model: Model-Based Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["1906.08253", "2005.05951"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_17687"} +{"question": "What papers shed lights on the limitations of LLMs in understanding prompts like humans?", "answer": ["Do Prompt-Based Models Really Understand the Meaning of Their Prompts?", "Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2109.01247", "2104.08786"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_17688"} +{"question": "What studies have proposed methods for debiasing classifiers?", "answer": ["Data-Centric Debugging: mitigating model failures via targeted data\n collection", "ModelDiff: A Framework for Comparing Learning Algorithms", "Editing a classifier by rewriting its prediction rules", "Fast Model Editing at Scale", "Learning Robust Global Representations by Penalizing Local Predictive\n Power", "Invariant Risk Minimization"], "answer_arxiv_id": ["2211.09859", "2211.12491", "2112.01008", "2110.11309", "1905.13549", "1907.02893"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17689"} +{"question": "Can you name studies that generated instructions and responses from image annotations by prompting GPT-4 using self-instruction methods?", "answer": ["Visual Instruction Tuning", "SVIT: Scaling up Visual Instruction Tuning"], "answer_arxiv_id": ["2304.08485", "2307.04087"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_17690"} +{"question": "Are there any works with indirect substantiation for the assumption that reversal of diffusion process has the same functional form as the forward process by introducing a reverse SDE?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_17691"} +{"question": "Could you name a stereotype repository that includes stereotypes for state-level identities within the US and India?", "answer": ["SeeGULL: A Stereotype Benchmark with Broad Geo-Cultural Coverage\n Leveraging Generative Models"], "answer_arxiv_id": ["2305.11840"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_17692"} +{"question": "What studies optimize over both discrete instruction and example sets in the prompts?", "answer": ["TEMPERA: Test-Time Prompting via Reinforcement Learning"], "answer_arxiv_id": ["2211.11890"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_17693"} +{"question": "Which studies first introduced prompt tuning to adapt pretrained language models to downstream tasks in natural language processing?", "answer": ["AutoPrompt: Eliciting Knowledge from Language Models with Automatically\n Generated Prompts", "How Can We Know What Language Models Know?"], "answer_arxiv_id": ["2010.15980", "1911.12543"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_17694"} +{"question": "Can you mention any research that applied graph-based methods for sarcasm detection?", "answer": ["Towards Multi-Modal Sarcasm Detection via Hierarchical Congruity\n Modeling with Knowledge Enhancement"], "answer_arxiv_id": ["2210.03501"], "source_meta": {"published_time": "20240501"}, "qid": "AutoScholarQuery_train_17695"} +{"question": "What works introduced popular deep learning approaches such as Physics inspired neural networks (PINN) and Deep Ritz method for PDEs?", "answer": ["The Deep Ritz method: A deep learning-based numerical algorithm for solving variational problems"], "answer_arxiv_id": ["1710.00211"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_17696"} +{"question": "Which papers introduced the concept of Vector Quantization (VQ)?", "answer": ["Neural Discrete Representation Learning", "Generating Diverse High-Fidelity Images with VQ-VAE-2"], "answer_arxiv_id": ["1711.00937", "1906.00446"], "source_meta": {"published_time": "20231022"}, "qid": "AutoScholarQuery_train_17697"} +{"question": "What are some examples of pre-trained models applied to the VLN task?,", "answer": ["A Recurrent Vision-and-Language BERT for Navigation", "History Aware Multimodal Transformer for Vision-and-Language Navigation", "Think Global, Act Local: Dual-scale Graph Transformer for Vision-and-Language Navigation", "HOP: History-and-Order Aware Pre-training for Vision-and-Language\n Navigation", "Airbert: In-domain Pretraining for Vision-and-Language Navigation", "March in Chat: Interactive Prompting for Remote Embodied Referring\n Expression", "Scaling Data Generation in Vision-and-Language Navigation", "LangNav: Language as a Perceptual Representation for Navigation"], "answer_arxiv_id": ["2011.13922", "2110.13309", "2202.11742v1", "2203.11591", "2108.09105", "2308.10141", "2307.15644", "2310.07889"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_17698"} +{"question": "Could you provide examples of studies in the field of Human Image Synthesis?", "answer": ["Deep Image Spatial Transformation for Person Image Generation", "Towards Fine-grained Human Pose Transfer with Detail Replenishing\n Network", "PISE: Person Image Synthesis and Editing with Decoupled GAN", "Neural Texture Extraction and Distribution for Controllable Person Image\n Synthesis", "Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "HumanSD: A Native Skeleton-Guided Diffusion Model for Human Image\n Generation", "HyperHuman: Hyper-Realistic Human Generation with Latent Structural\n Diffusion", "Exploring Dual-task Correlation for Pose Guided Person Image Generation", "Beyond Appearance: a Semantic Controllable Self-Supervised Learning\n Framework for Human-Centric Visual Tasks"], "answer_arxiv_id": ["2003.00696", "2005.12494", "2103.04023", "2204.06160", "2302.05543", "2302.08453", "2304.04269", "2310.08579", "2203.02910", "2303.17602"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_17699"} +{"question": "Which paper considered a binary classification problem and provided bonds on number of queries and bound on excess risk in the active learning framework?", "answer": ["Efficient Active Learning with Abstention"], "answer_arxiv_id": ["2204.00043"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_17700"} +{"question": "Who considered the Bayesian tabular MDP case while also introducing additional Dirichlet prior assumptions?", "answer": ["Why is Posterior Sampling Better than Optimism for Reinforcement Learning?"], "answer_arxiv_id": ["1607.00215"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_17701"} +{"question": "What papers employed bilinear pooling or developed transformer-based multimodal architectures such as MMBT, ViLBERT, and Visual-BERT for hateful meme detection?", "answer": ["\"Subverting the Jewtocracy\": Online Antisemitism Detection Using\n Multimodal Deep Learning", "Supervised Multimodal Bitransformers for Classifying Images and Text", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "VisualBERT: A Simple and Performant Baseline for Vision and Language"], "answer_arxiv_id": ["2104.05947", "1909.02950", "1908.02265", "1908.03557"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_17702"} +{"question": "Which work proposed a simple and effective attempt to roadside perception utilizing camera specifications and the ground knowledge?", "answer": ["Rope3D: TheRoadside Perception Dataset for Autonomous Driving and\n Monocular 3D Object Detection Task"], "answer_arxiv_id": ["2203.13608"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17703"} +{"question": "What study improved on CaloGAN using a normalizing flow architecture called CaloFlow for generating shower images?", "answer": ["CaloFlow: Fast and Accurate Generation of Calorimeter Showers with Normalizing Flows"], "answer_arxiv_id": ["2106.05285"], "source_meta": {"published_time": "20220210"}, "qid": "AutoScholarQuery_train_17704"} +{"question": "What studies proposed the 'once-for-all' approach, which trains a large network containing multiple specialized sub-nets?", "answer": ["Once-for-All: Train One Network and Specialize it for Efficient Deployment"], "answer_arxiv_id": ["1908.09791"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_17705"} +{"question": "Which works utilize camera positions from head-mounted cameras for motion capture with IMU sensors?", "answer": ["Human POSEitioning System (HPS): 3D Human Pose Estimation and\n Self-localization in Large Scenes from Body-Mounted Sensors", "EgoLocate: Real-time Motion Capture, Localization, and Mapping with\n Sparse Body-mounted Sensors"], "answer_arxiv_id": ["2103.17265", "2305.01599"], "source_meta": {"published_time": "20240101"}, "qid": "AutoScholarQuery_train_17706"} +{"question": "Could you provide works studying the challenging problem of learning adversarial linear MDPs with only bandit feedback?", "answer": ["Online Learning in MDPs with Linear Function Approximation and Bandit Feedback", "Policy Optimization in Adversarial MDPs: Improved Exploration via Dilated Bonuses", "Improved Regret Bounds for Linear Adversarial MDPs via Linear Optimization", "Refined Regret for Adversarial MDPs with Linear Function Approximation", "Improved Regret for Efficient Online Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["2007.01612", "2107.08346", "2302.06834", "2301.12942", "2301.13087"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_17707"} +{"question": "Could you provide me some studies about open-vocabulary or zero-shot image semantic segmentation?", "answer": ["Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "Language-driven Semantic Segmentation", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Extract Free Dense Labels from CLIP", "Semantic Segmentation In-the-Wild Without Seeing Any Segmentation Examples", "GroupViT: Semantic Segmentation Emerges from Text Supervision", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "Open-vocabulary Semantic Segmentation with Frozen Vision-Language Models", "Open-Vocabulary Multi-Label Classification via Multi-Modal Knowledge Transfer", "Open-vocabulary Object Detection via Vision and Language Knowledge Distillation", "ConceptFusion: Open-set Multimodal 3D Mapping", "Decoupling Zero-Shot Semantic Segmentation"], "answer_arxiv_id": ["2112.12143", "2201.03546", "2210.04150", "2112.01071", "2112.03185", "2202.11094", "2112.01518", "2210.15138", "2207.01887", "2104.13921", "2302.07241", "2112.07910v2"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_17708"} +{"question": "Which work interprets the method as a one-shot architecture search using Laplace approximations?", "answer": ["BayesNAS: A Bayesian Approach for Neural Architecture Search"], "answer_arxiv_id": ["1905.04919"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_17709"} +{"question": "Are there studies where depth and camera ego-motion are learned simultaneously from monocular videos?", "answer": ["Unsupervised Learning of Depth and Ego-Motion from Video"], "answer_arxiv_id": ["1704.07813"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_17710"} +{"question": "Can you tell me some of the studies where RL is applied in multi-agent games?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments", "Multi-Agent Reinforcement Learning: A Selective Overview of Theories and Algorithms"], "answer_arxiv_id": ["1706.02275", "1911.10635"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_17711"} +{"question": "What studies used a population to achieve better coverage of the policy space for improving skills in RL?", "answer": ["Diversity-Driven Exploration Strategy for Deep Reinforcement Learning", "Attraction-Repulsion Actor-Critic for Continuous Control Reinforcement Learning", "Population-Guided Parallel Policy Search for Reinforcement Learning", "Effective Diversity in Population Based Reinforcement Learning"], "answer_arxiv_id": ["1802.04564", "1909.07543", "2001.02907", "2002.00632"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_17712"} +{"question": "Could you provide me some studies about semi-discrete approaches aiming at computing Wasserstein barycenter?", "answer": ["Stochastic Wasserstein Barycenters", "Parallel Streaming Wasserstein Barycenters", "Variational Wasserstein Barycenters for Geometric Clustering"], "answer_arxiv_id": ["1802.05757", "1705.07443", "2002.10543"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17713"} +{"question": "Could you name the work that implemented soft top-k operator for differentiable beam search using iterated softmax?", "answer": ["A Continuous Relaxation of Beam Search for End-to-end Training of Neural Sequence Models"], "answer_arxiv_id": ["1708.00111"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17714"} +{"question": "Could you provide me with studies that discuss the concept of temporal grounding?", "answer": ["LocVTP: Video-Text Pre-training for Temporal Localization", "End-to-end Multi-modal Video Temporal Grounding", "Local-Global Video-Text Interactions for Temporal Grounding", "Can Shuffling Video Benefit Temporal Bias Problem: A Novel Training\n Framework for Temporal Grounding", "Text-Visual Prompting for Efficient 2D Temporal Video Grounding", "Localizing Moments in Long Video Via Multimodal Guidance"], "answer_arxiv_id": ["2207.10362", "2107.05624", "2004.07514", "2207.14698", "2303.04995", "2302.13372"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_17715"} +{"question": "What works are associated with coarse-grained amino-acid-level representations in protein structure encoding methods?", "answer": ["Protein Representation Learning by Geometric Structure Pretraining"], "answer_arxiv_id": ["2203.06125"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_17716"} +{"question": "Can you list some works that studied diffusion on graphs?", "answer": ["The Emerging Field of Signal Processing on Graphs Extending High-Dimensional Data Analysis to Networks and Other Irregular Domains"], "answer_arxiv_id": ["1211.0053"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17717"} +{"question": "What papers describe the use of graph neural networks for encoding the context information and sample atoms in the context of structure-based drug design?", "answer": ["Generating 3D Molecules for Target Protein Binding", "Pocket2Mol: Efficient Molecular Sampling Based on 3D Protein Pockets"], "answer_arxiv_id": ["2204.09410", "2205.07249"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_17718"} +{"question": "What works have been done on localizing the objects with a self-supervised transformer?", "answer": ["Localizing Objects with Self-Supervised Transformers and no Labels"], "answer_arxiv_id": ["2109.14279"], "source_meta": {"published_time": "20220919"}, "qid": "AutoScholarQuery_train_17719"} +{"question": "Which paper first proposed the concept of knowledge transfer across modalities?", "answer": ["Cross Modal Distillation for Supervision Transfer"], "answer_arxiv_id": ["1507.00448"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_17720"} +{"question": "What works proposed the concept of order learning for ordinal classification by comparison between instances?", "answer": ["Moving Window Regression: A Novel Approach to Ordinal Regression"], "answer_arxiv_id": ["2203.13122"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_17721"} +{"question": "What works have pretrained vision-language models from scratch using image-caption data?", "answer": ["Florence: A New Foundation Model for Computer Vision"], "answer_arxiv_id": ["2111.11432"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_17722"} +{"question": "Which papers have researched non-smooth normalizing flows on Riemannian submanifolds?", "answer": ["Normalizing Flows on Riemannian Manifolds", "Density estimation on smooth manifolds with normalizing flows"], "answer_arxiv_id": ["1611.02304", "2106.03500"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17723"} +{"question": "In which papers did researchers use language models to break down complex instructions?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents", "Inner Monologue: Embodied Reasoning through Planning with Language Models", "Collaborating with language models for embodied reasoning", "ProgPrompt: Generating Situated Robot Task Plans using Large Language Models"], "answer_arxiv_id": ["2204.01691", "2201.07207", "2207.05608", "2302.00763", "2209.11302"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_17724"} +{"question": "Which works in deepfake detection make use of ID information in detecting image forgery?", "answer": ["Implicit Identity Leakage: The Stumbling Block to Improving Deepfake\n Detection Generalization"], "answer_arxiv_id": ["2210.14457"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_17725"} +{"question": "What works have used ConvLSTM for vegetation prediction in Africa?", "answer": ["Learning to forecast vegetation greenness at fine resolution over Africa\n with ConvLSTMs"], "answer_arxiv_id": ["2210.13648"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_17726"} +{"question": "Could you provide examples of work which have explored the use of learnable tokens in place of the original queries or keys to generate dynamic affinity matrices?", "answer": ["Synthesizer: Rethinking Self-Attention in Transformer Models", "Involution: Inverting the Inherence of Convolution for Visual\n Recognition", "VOLO: Vision Outlooker for Visual Recognition", "Learned Queries for Efficient Local Attention"], "answer_arxiv_id": ["2005.00743", "2103.06255", "2106.13112", "2112.11435"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_17727"} +{"question": "What paper has addressed the issue of output discrepancy between teacher and student models in network compression?", "answer": ["Mind the Gap in Distilling StyleGANs"], "answer_arxiv_id": ["2208.08840"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_17728"} +{"question": "In which studies was fairness considered in the context of linear bandits and general anytime constraints?", "answer": ["An Efficient Pessimistic-Optimistic Algorithm for Stochastic Linear Bandits with General Constraints"], "answer_arxiv_id": ["2102.05295"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_17729"} +{"question": "What references cover empirical experiments that show LLM performance decrease with less effective information in a prompt?", "answer": ["LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding", "Large Language Models Can Be Easily Distracted by Irrelevant Context"], "answer_arxiv_id": ["2308.14508v2", "2302.00093"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_17730"} +{"question": "In the realm of Random Matrix Theory, what works have been completed in the analysis of multi-task learning?", "answer": ["PCA-based Multi Task Learning: a Random Matrix Approach"], "answer_arxiv_id": ["2111.00924v1"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_17731"} +{"question": "Which studies have worked on using a two-stage approach for dense video captioning?", "answer": ["Dense-Captioning Events in Videos", "A Better Use of Audio-Visual Cues: Dense Video Captioning with Bi-modal\n Transformer", "Multi-modal Dense Video Captioning", "Bidirectional Attentive Fusion with Context Gating for Dense Video\n Captioning"], "answer_arxiv_id": ["1705.00754", "2005.08271", "2003.07758", "1804.00100"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_17732"} +{"question": "The post-hoc text detection research that tackles detection problems without additional training procedures?", "answer": ["GLTR: Statistical Detection and Visualization of Generated Text", "DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability\n Curvature", "Dynamic Prompting: A Unified Framework for Prompt Tuning"], "answer_arxiv_id": ["1906.04043", "2301.11305", "2303.02909"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_17733"} +{"question": "Which studies proposed to use total variation distance in bidirectional language models?", "answer": ["Regularizing Neural Machine Translation by Target-bidirectional Agreement"], "answer_arxiv_id": ["1808.04064"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_17734"} +{"question": "Which studies utilized contrastive learning in the early stage of adversarial SSL methods?", "answer": ["Adversarial Self-Supervised Contrastive Learning", "Robust Pre-Training by Adversarial Contrastive Learning"], "answer_arxiv_id": ["2006.07589", "2010.13337"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_17735"} +{"question": "What research have generated counterfactual explanations to understand model behavior?", "answer": ["DISSECT: Disentangled Simultaneous Explanations via Concept Traversals", "Meaningfully Debugging Model Mistakes using Conceptual Counterfactual Explanations"], "answer_arxiv_id": ["2105.15164", "2106.12723"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_17736"} +{"question": "What is the first Semi-Supervised Attentional Visual Sound Localization work?", "answer": ["Learning to Localize Sound Source in Visual Scenes", "Learning to Localize Sound Sources in Visual Scenes: Analysis and Applications"], "answer_arxiv_id": ["1803.03849v1", "1911.09649"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_17737"} +{"question": "Which studies first introduced the concept of entity segmentation?", "answer": ["Open-World Entity Segmentation", "High Quality Segmentation for Ultra High-resolution Images", "High-Quality Entity Segmentation"], "answer_arxiv_id": ["2107.14228", "2111.14482", "2211.05776"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17738"} +{"question": "Any works about using spectral properties of the transition matrix to learn options for aiding exploration?", "answer": ["A Laplacian Framework for Option Discovery in Reinforcement Learning", "Eigenoption Discovery through the Deep Successor Representation", "The Eigenoption-Critic Framework"], "answer_arxiv_id": ["1703.00956", "1710.11089", "1712.04065"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_17739"} +{"question": "What works proposed flaggers and scorers for Automatic Error Detection (AED)?", "answer": ["Annotation Error Detection: Analyzing the Past and Present for a More\n Coherent Future"], "answer_arxiv_id": ["2206.02280"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_17740"} +{"question": "Which studies utilized the Wasserstein GAN which enforced the Lipschitz continuity on the discriminator via gradient penalties?", "answer": ["Improved Training of Wasserstein GANs"], "answer_arxiv_id": ["1704.00028"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_17741"} +{"question": "Which research used automated evaluation metrics to rank model generations and then select the top few best and worst?", "answer": ["PairReranker: Pairwise Reranking for Natural Language Generation"], "answer_arxiv_id": ["2212.10555"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_17742"} +{"question": "What works describe common data augmentation techniques employed in machine learning?", "answer": ["mixup: Beyond Empirical Risk Minimization", "CutMix: Regularization Strategy to Train Strong Classifiers with\n Localizable Features", "RandAugment: Practical automated data augmentation with a reduced search\n space", "Unsupervised Data Augmentation for Consistency Training", "EDA: Easy Data Augmentation Techniques for Boosting Performance on Text\n Classification Tasks"], "answer_arxiv_id": ["1710.09412", "1905.04899", "1909.13719", "1904.12848", "1901.11196"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_17743"} +{"question": "What studies pertain to pairwise AUC optimization, a formulation of objective functions in AUC?", "answer": ["On the Consistency of AUC Pairwise Optimization"], "answer_arxiv_id": ["1208.0645"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_17744"} +{"question": "What research work indicates the presence of adversarial examples in various neural networks?", "answer": ["Intriguing properties of neural networks", "Tactics of Adversarial Attack on Deep Reinforcement Learning Agents", "On Adversarial Examples for Character-Level Neural Machine Translation", "Adversarial Attack on Graph Structured Data"], "answer_arxiv_id": ["1312.6199", "1703.06748", "1806.09030", "1806.02371"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_17745"} +{"question": "Can you provide studies that utilized instruction-tuning and RLHF for language learning models?", "answer": ["ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detection", "The Turking Test: Can Language Models Understand Instructions?", "Red Teaming Language Models with Language Models", "Generating Datasets with Pretrained Language Models"], "answer_arxiv_id": ["2203.09509v4", "2010.11982", "2202.03286v1", "2104.07540"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_17746"} +{"question": "What studies discuss additional regularization in training as a method to enhance the performance of MLP?", "answer": ["OrthoReg: Improving Graph-regularized MLPs via Orthogonality Regularization"], "answer_arxiv_id": ["2302.00109"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_17747"} +{"question": "What research focuses on fine-tuning the diffusion model in text-to-image models?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "Multi-Concept Customization of Text-to-Image Diffusion", "Key-Locked Rank One Editing for Text-to-Image Personalization"], "answer_arxiv_id": ["2208.12242", "2212.04488", "2305.01644"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_17748"} +{"question": "Which works have considered an extension where the state model is a pretrained text2image model?", "answer": ["Learning Universal Policies via Text-Guided Video Generation"], "answer_arxiv_id": ["2302.00111"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_17749"} +{"question": "What are the studies that have extended code synthesis tasks to general-purpose programming languages?", "answer": ["A Syntactic Neural Model for General-Purpose Code Generation", "Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["1704.01696", "2107.03374"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17750"} +{"question": "What are the studies related to object-centric approaches, which aim to abstract representations for objects in a scene?", "answer": ["Illiterate DALL-E Learns to Compose", "Object-Centric Learning with Slot Attention", "Generalization and Robustness Implications in Object-Centric Learning"], "answer_arxiv_id": ["2110.11405", "2006.15055", "2107.00637"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_17751"} +{"question": "What studies discarded the smooth or Lipschitz assumptions for stability analysis of SGD?", "answer": ["Stability of Stochastic Gradient Descent on Nonsmooth Convex Losses", "Beyond Lipschitz: Sharp Generalization and Excess Risk Bounds for Full-Batch GD"], "answer_arxiv_id": ["2006.06914", "2204.12446v5"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_17752"} +{"question": "What research introduced methods that randomly remove input parts as a form of dropout in the input layer?", "answer": ["Hide-and-Seek: Forcing a Network to be Meticulous for Weakly-supervised Object and Action Localization", "Improved Regularization of Convolutional Neural Networks with Cutout", "Random Erasing Data Augmentation"], "answer_arxiv_id": ["1704.04232", "1708.04552", "1708.04896"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_17753"} +{"question": "What papers proposed adversarial methods for information removal?", "answer": ["Censoring Representations with an Adversary", "Controllable Invariance through Adversarial Feature Learning", "Adversarial Deep Averaging Networks for Cross-Lingual Sentiment Classification", "Adversarial Removal of Demographic Attributes from Text Data", "Mitigating Unwanted Biases with Adversarial Learning"], "answer_arxiv_id": ["1511.05897", "1705.11122", "1606.01614", "1808.06640", "1801.07593"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17754"} +{"question": "Which studies provide examples of representing stochastic policies for continuous control by Gaussians with diagonal covariance matrix?", "answer": ["Proximal Policy Optimization Algorithms", "Soft Actor-Critic Algorithms and Applications", "Maximum a Posteriori Policy Optimisation"], "answer_arxiv_id": ["1707.06347v2", "1812.05905", "1806.06920"], "source_meta": {"published_time": "20221022"}, "qid": "AutoScholarQuery_train_17755"} +{"question": "Could you mention some research related to extracting features from RGB and depth images individually and then fusing them for pose estimation?", "answer": ["PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation", "Frustum PointNets for 3D Object Detection from RGB-D Data", "DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion", "MoreFusion: Multi-object Reasoning for 6D Pose Estimation from\n Volumetric Fusion", "PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose\n Estimation", "FFB6D: A Full Flow Bidirectional Fusion Network for 6D Pose Estimation", "Deep Fusion Transformer Network with Weighted Vector-Wise Keypoints\n Voting for Robust 6D Object Pose Estimation"], "answer_arxiv_id": ["1711.10871", "1711.08488", "1901.04780", "2004.04336", "1911.04231", "2103.02242", "2308.05438"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_17756"} +{"question": "What papers modeled diverse tasks as sequence generation tasks inspired by sequence-to-sequence models?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning", "GIT: A Generative Image-to-text Transformer for Vision and Language"], "answer_arxiv_id": ["2204.14198", "2205.14100"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17757"} +{"question": "Can you mention the studies that deal with text-guided image generation using diffusion models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2205.11487", "2112.10741", "2204.06125"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_17758"} +{"question": "What are some studies that explored in-context learning for LLMs?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "PaLM 2 Technical Report"], "answer_arxiv_id": ["2005.14165", "2204.02311", "2305.10403"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_17759"} +{"question": "What papers have investigated or used data augmentation in the graph domain?", "answer": ["ForceNet: A Graph Neural Network for Large-Scale Quantum Calculations", "Equivariance versus Augmentation for Spherical Images"], "answer_arxiv_id": ["2103.01436", "2202.03990"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_17760"} +{"question": "Can you provide details on the studies which found interpretable directions in the latent space of GANs?", "answer": ["On the \"steerability\" of generative adversarial networks", "Understanding the Role of Individual Units in a Deep Neural Network"], "answer_arxiv_id": ["1907.07171", "2009.05041"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_17761"} +{"question": "Which works proposed parameter-efficient finetuning (PEFT) methods due to computational costs associated with fully finetuning a large language model (LLM)?", "answer": ["GPT Understands, Too", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Progressive Neural Networks", "Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than\n In-Context Learning", "Towards a Unified View of Parameter-Efficient Transfer Learning", "AutoPEFT: Automatic Configuration Search for Parameter-Efficient\n Fine-Tuning", "Neural Architecture Search with Reinforcement Learning", "UniPELT: A Unified Framework for Parameter-Efficient Language Model\n Tuning"], "answer_arxiv_id": ["2103.10385", "2101.00190", "2104.08691", "1606.04671", "1902.00751", "2106.09685", "2205.05638", "2110.04366", "2301.12132", "1611.01578", "2110.07577"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_17762"} +{"question": "Any studies that model multiple dog breeds and employ intra-breed similarities using a triplet loss?", "answer": ["BARC: Learning to Regress 3D Dog Shape from Images by Exploiting Breed\n Information"], "answer_arxiv_id": ["2203.15536"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_17763"} +{"question": "What works used GAN prior for neural network visualization?", "answer": ["Synthesizing the preferred inputs for neurons in neural networks via deep generator networks", "Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space"], "answer_arxiv_id": ["1605.09304", "1612.00005"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_17764"} +{"question": "Which paper proposed the emrQA dataset for medical QA systems based on clinical texts?", "answer": ["emrQA: A Large Corpus for Question Answering on Electronic Medical Records"], "answer_arxiv_id": ["1809.00732v1"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_17765"} +{"question": "Which papers focus on benchmarking classical machine learning AD (Anomaly Detection) methods?", "answer": ["Benchmarking Unsupervised Outlier Detection with Realistic Synthetic Data"], "answer_arxiv_id": ["2004.06947"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_17766"} +{"question": "What studies involve reducing the scope of optimization in modeling?", "answer": ["Local Bayesian optimization via maximizing probability of descent"], "answer_arxiv_id": ["2210.11662v2"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_17767"} +{"question": "What research studies investigated general bias correction techniques?", "answer": ["Minimax Estimation of Divergences between Discrete Distributions", "Bias Correction with Jackknife, Bootstrap, and Taylor Series"], "answer_arxiv_id": ["1605.09124", "1709.06183"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_17768"} +{"question": "Which works discussed image generation based on GANs?", "answer": ["Generative Adversarial Networks", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["2203.00667", "1812.04948"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_17769"} +{"question": "What works proposed a full transformer-based pipeline to decode a NeRF representation from triplane features?", "answer": ["GINA-3D: Learning to Generate Implicit Neural Assets in the Wild", "LRM: Large Reconstruction Model for Single Image to 3D"], "answer_arxiv_id": ["2304.02163", "2311.04400"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_17770"} +{"question": "What papers consider solutions for nonconvex-concave saddle point problems through projection-free based algorithms?", "answer": ["The Complexity of Large-scale Convex Programming under a Linear Optimization Oracle", "Faster Rates for the Frank-Wolfe Method over Strongly-Convex Sets", "Conditional Accelerated Lazy Stochastic Gradient Descent"], "answer_arxiv_id": ["1309.5550", "1406.1305", "1703.05840v5"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_17771"} +{"question": "Can you mention a study that proposed using a low temperature for softmax activation in training to improve transferability?", "answer": ["Why Do Better Loss Functions Lead to Less Transferable Features?"], "answer_arxiv_id": ["2010.16402"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17772"} +{"question": "Which works incorporate MIM within Siamese frameworks?", "answer": ["MST: Masked Self-Supervised Transformer for Visual Representation", "iBOT : Image BERT Pre-Training with Online Tokenizer", "Are Large-scale Datasets Necessary for Self-Supervised Pre-training?"], "answer_arxiv_id": ["2106.05656", "2111.07832", "2112.10740"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_17773"} +{"question": "What works propose to estimate importance scores using coreset methods based on k-means?", "answer": ["Beyond neural scaling laws: beating power law scaling via data pruning"], "answer_arxiv_id": ["2206.14486v6"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_17774"} +{"question": "Who proposed slot attention as an approach for learning object-centric representations in an unsupervised manner?", "answer": ["Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["2006.15055"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17775"} +{"question": "What research employs the nearest-neighbor search in the type embedding space for user-defined type prediction?", "answer": ["Typilus: Neural Type Hints"], "answer_arxiv_id": ["2004.10657v1"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_17776"} +{"question": "Any works that explained brain-to-network mapping at the ROI-level?", "answer": ["The Algonauts Project 2021 Challenge: How the Human Brain Makes Sense of a World in Motion", "Teaching Matters: Investigating the Role of Supervision in Vision\n Transformers"], "answer_arxiv_id": ["2104.13714v1", "2212.03862"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_17777"} +{"question": "What papers explore applications of Federated Learning in next word prediction?", "answer": ["Federated learning for mobile keyboard prediction", "Federated Learning of N-gram Language Models"], "answer_arxiv_id": ["1811.03604", "1910.03432"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_17778"} +{"question": "What researches have been made in generating relation triplets using LLMs?", "answer": ["RelationPrompt: Leveraging Prompts to Generate Synthetic Data for Zero-Shot Relation Triplet Extraction"], "answer_arxiv_id": ["2203.09101"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_17779"} +{"question": "What paper proposes AdaMixup which trains an extra network to dynamically determine interpolation coefficient parameter?", "answer": ["MixUp as Locally Linear Out-Of-Manifold Regularization"], "answer_arxiv_id": ["1809.02499"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_17780"} +{"question": "Any studies that consider the notion of computational efficiency in sequential probability assignment and portfolio selection?", "answer": ["Efficient Online Portfolio with Logarithmic Regret", "Pushing the Efficiency-Regret Pareto Frontier for Online Learning of Portfolios and Quantum States", "Efficient and Near-Optimal Online Portfolio Selection"], "answer_arxiv_id": ["1805.07430", "2202.02765", "2209.13932"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_17781"} +{"question": "What works have suggested deep-learning models to improve temporal consistency and to make use of spatial-temporal information for low-light video enhancement and denoising?", "answer": ["Recurrent Self-Supervised Video Denoising with Denser Receptive Field"], "answer_arxiv_id": ["2308.03608"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_17782"} +{"question": "Which studies focus on the program-based prompting strategy?", "answer": ["Program of Thoughts Prompting: Disentangling Computation from Reasoning\n for Numerical Reasoning Tasks", "PAL: Program-aided Language Models", "Binding Language Models in Symbolic Languages"], "answer_arxiv_id": ["2211.12588", "2211.10435", "2210.02875"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_17783"} +{"question": "Could you provide me some works about model-based methods for recovering 3D HPS from images?", "answer": ["End-to-end Recovery of Human Shape and Pose", "Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the\n Loop", "Humans in 4D: Reconstructing and Tracking Humans with Transformers", "NeuralAnnot: Neural Annotator for 3D Human Mesh Training Sets", "CLIFF: Carrying Location Information in Full Frames into Human Pose and\n Shape Estimation", "Three Recipes for Better 3D Pseudo-GTs of 3D Human Mesh Estimation in\n the Wild"], "answer_arxiv_id": ["1712.06584", "1909.12828", "2305.20091", "2011.11232", "2208.00571", "2304.04875"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_17784"} +{"question": "What are some studies that have proposed task-specific taxonomies to identify potential sources of disagreement in NLP tasks?", "answer": ["Investigating Reasons for Disagreement in Natural Language Inference"], "answer_arxiv_id": ["2209.03392"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_17785"} +{"question": "Which works focus on the text-to-3D or image-to-3D generation methodologies?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "DreamBooth3D: Subject-Driven Text-to-3D Generation", "Zero-1-to-3: Zero-shot One Image to 3D Object", "Magic3D: High-Resolution Text-to-3D Content Creation", "3DDesigner: Towards Photorealistic 3D Object Generation and Editing with\n Text-guided Diffusion Models", "SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation"], "answer_arxiv_id": ["2209.14988", "2303.13508", "2303.11328", "2211.10440", "2211.14108", "2212.04493"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_17786"} +{"question": "What studies are about sequence labeling methods for EAE?", "answer": ["Leveraging Knowledge Bases in LSTMs for Improving Machine Reading", "The Devil is in the Details: On the Pitfalls of Event Extraction\n Evaluation"], "answer_arxiv_id": ["1902.09091", "2306.06918"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_17787"} +{"question": "Can you name papers that discuss enhancing compositionality in T2I models with the help of layout or scene graph?", "answer": ["Image Generation from Scene Graphs", "Inferring Semantic Layout for Hierarchical Text-to-Image Synthesis", "Modeling Image Composition for Complex Scene Generation", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors"], "answer_arxiv_id": ["1804.01622", "1801.05091", "2206.00923", "2203.13131"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_17788"} +{"question": "Which papers are focused on the optimization of KV cache?", "answer": ["H$_2$O: Heavy-Hitter Oracle for Efficient Generative Inference of Large\n Language Models", "Transformers are Multi-State RNNs", "Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs"], "answer_arxiv_id": ["2306.14048", "2401.06104", "2310.01801"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_17789"} +{"question": "Which researches discuss continuing to fine-tune the entire model after training only a subset of parameters?", "answer": ["Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution", "Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2202.10054v1", "2202.06856"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_17790"} +{"question": "Which papers focused on BayesPCN model for memorization and recall of a stream of patterns?", "answer": ["BayesPCN: A Continually Learnable Predictive Coding Associative Memory"], "answer_arxiv_id": ["2205.09930"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_17791"} +{"question": "Can you name any research papers that propose methods for document ranking or permutation learning?", "answer": ["Ranking via Sinkhorn Propagation", "DeepPermNet: Visual Permutation Learning", "Learning Latent Permutations with Gumbel-Sinkhorn Networks"], "answer_arxiv_id": ["1106.1925", "1704.02729", "1802.08665"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17792"} +{"question": "Which paper proposed the IIC method for semantic segmentation, aiming to maximize the mutual information between different augmented versions of an image?", "answer": ["Invariant Information Clustering for Unsupervised Image Classification\n and Segmentation"], "answer_arxiv_id": ["1807.06653"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_17793"} +{"question": "Which works implement an ICL-based system for few-shot KBQA?", "answer": ["Code-Style In-Context Learning for Knowledge-Based Question Answering"], "answer_arxiv_id": ["2309.04695"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_17794"} +{"question": "Which works follow the approach of loss reweighting for resolving class imbalance?", "answer": ["Class-Balanced Loss Based on Effective Number of Samples", "Deep Imbalanced Learning for Face Recognition and Attribute Prediction", "Class-Balanced Loss Based on Effective Number of Samples", "Focal Loss for Dense Object Detection", "Anchor Loss: Modulating Loss Scale based on Prediction Difficulty", "Data-Driven Robust Optimization"], "answer_arxiv_id": ["1901.05555", "1806.00194", "1901.05555", "1708.02002", "1909.11155", "1401.0212"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_17795"} +{"question": "Where can I find works about the reasoning abilities of Large Language Models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Graph of Thoughts: Solving Elaborate Problems with Large Language Models"], "answer_arxiv_id": ["2201.11903", "2305.10601", "2203.11171", "2308.09687"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_17796"} +{"question": "Could you list some studies that incorporate instance modeling into NeRF?", "answer": ["Panoptic NeRF: 3D-to-2D Label Transfer for Panoptic Urban Scene Segmentation", "Panoptic Neural Fields: A Semantic Object-Aware Neural Scene Representation", "Decomposing NeRF for Editing via Feature Field Distillation"], "answer_arxiv_id": ["2203.15224", "2205.04334", "2205.15585"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_17797"} +{"question": "What works explored the use of Diffusion Models for global changes and perform local manipulations using user-provided masks?", "answer": ["Blended Diffusion for Text-driven Editing of Natural Images", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation"], "answer_arxiv_id": ["2111.14818", "2110.02711"], "source_meta": {"published_time": "20220802"}, "qid": "AutoScholarQuery_train_17798"} +{"question": "Which research introduced neural ODEs, where the vector fields of an ODE are parameterized by a neural network?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_17799"} +{"question": "What studies suggest using scheduled sampling as an advanced training scheme?", "answer": ["Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks"], "answer_arxiv_id": ["1506.03099"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_17800"} +{"question": "Which works discuss methods to allocate compute conditioned on the input?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer", "Efficient Transformers with Dynamic Token Pooling", "CoLT5: Faster Long-Range Transformers with Conditional Computation"], "answer_arxiv_id": ["1701.06538", "2211.09761", "2303.09752v3"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_17801"} +{"question": "Which work designs a semi-automatic framework to extend a 2D scene graph into 3D space?", "answer": ["3D Scene Graph: A Structure for Unified Semantics, 3D Space, and Camera"], "answer_arxiv_id": ["1910.02527"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_17802"} +{"question": "Are there any sources that demonstrate the need for further research in character-level infilling?", "answer": ["Efficient Training of Language Models to Fill in the Middle"], "answer_arxiv_id": ["2207.14255"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_17803"} +{"question": "What studies focus on neighborhood consistency in relation to self-supervised learning?", "answer": ["Mean Shift for Self-Supervised Learning", "With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations", "Constrained Mean Shift Using Distant Yet Related Neighbors for Representation Learning"], "answer_arxiv_id": ["2105.07269", "2104.14548", "2112.04607"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_17804"} +{"question": "Can you provide references for works that use text commands to learn multi-task and generalist control policies?", "answer": ["Inner Monologue: Embodied Reasoning through Planning with Language Models", "Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "RT-1: Robotics Transformer for Real-World Control at Scale", "Language Conditioned Imitation Learning over Unstructured Data"], "answer_arxiv_id": ["2207.05608", "2204.01691", "2212.06817", "2005.07648"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_17805"} +{"question": "Which work needs a Lipschitz loss while analyzing convex objectives?", "answer": ["Fine-Grained Analysis of Stability and Generalization for Stochastic Gradient Descent"], "answer_arxiv_id": ["2006.08157"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_17806"} +{"question": "Who explored an ample data space for memorizing the category-level information?", "answer": ["Exploring Cross-Image Pixel Contrast for Semantic Segmentation"], "answer_arxiv_id": ["2101.11939"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_17807"} +{"question": "Could you provide me some studies on principal component regression (PCR)?", "answer": ["On Robustness of Principal Component Regression"], "answer_arxiv_id": ["1902.10920"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_17808"} +{"question": "What studies utilize reward shaping to deal with the delayed rewards in reinforcement learning?", "answer": ["Learning to Utilize Shaping Rewards: A New Approach of Reward Shaping", "Controllable Neural Story Plot Generation via Reward Shaping"], "answer_arxiv_id": ["2011.02669", "1809.10736"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_17809"} +{"question": "Which studies suggest the performance gain of combining online RL with offline datasets?", "answer": ["Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations", "Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewards", "Deep Q-learning from Demonstrations", "Overcoming Exploration in Reinforcement Learning with Demonstrations", "AWAC: Accelerating Online Reinforcement Learning with Offline Datasets"], "answer_arxiv_id": ["1709.10087", "1707.08817", "1704.03732", "1709.10089", "2006.09359"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_17810"} +{"question": "Which research leverages attention mechanism to estimate facial landmarks?", "answer": ["Sparse Local Patch Transformer for Robust Face Alignment and Landmarks\n Inherent Relation Learning"], "answer_arxiv_id": ["2203.06541"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_17811"} +{"question": "What works used graph convolution with linear messages in equivariant neural networks?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds"], "answer_arxiv_id": ["1802.08219"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_17812"} +{"question": "What concurrent study proposed to use LLM to combine information from raw and synthetic captions?", "answer": ["VeCLIP: Improving CLIP Training via Visual-enriched Captions"], "answer_arxiv_id": ["2310.07699"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_17813"} +{"question": "Can you provide me studies that focused on improving the ability of GNNs to learn joint dependencies?", "answer": ["Structured Prediction Energy Networks", "Deep Structured Prediction with Nonlinear Output Transformations", "Combining Generative and Discriminative Models for Hybrid Inference", "GMNN: Graph Markov Neural Networks", "Learning Iterative Reasoning through Energy Minimization", "Neural Structured Prediction for Inductive Node Classification"], "answer_arxiv_id": ["1511.06350", "1811.00539", "1906.02547", "1905.06214", "2206.15448", "2204.07524"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_17814"} +{"question": "What research papers discuss the effective use of federated learning with decentralized private data?", "answer": ["Federated Optimization: Distributed Machine Learning for On-Device Intelligence", "Communication-Efficient Learning of Deep Networks from Decentralized Data", "Federated Optimization in Heterogeneous Networks", "Advances and Open Problems in Federated Learning", "A Field Guide to Federated Optimization"], "answer_arxiv_id": ["1610.02527", "1602.05629", "1812.06127", "1912.04977", "2107.06917"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_17815"} +{"question": "What work emphasized the importance of having a computationally verifiable form of documentation across various scenarios?", "answer": ["Datasheets for Datasets"], "answer_arxiv_id": ["1803.09010"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_17816"} +{"question": "Could you provide some papers that covered the actions of LLMs within social science and micro-societies experiments?", "answer": ["Generative Agents: Interactive Simulacra of Human Behavior", "Can Large Language Models Transform Computational Social Science?"], "answer_arxiv_id": ["2304.03442", "2305.03514"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_17817"} +{"question": "What studies proposed different data augmentation strategies to deal with overfitting in RL?", "answer": ["Reinforcement Learning with Augmented Data", "Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels", "Quantifying Generalization in Reinforcement Learning"], "answer_arxiv_id": ["2004.14990", "2004.13649", "1812.02341"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_17818"} +{"question": "What papers proposed directly working with higher-order graph representations?", "answer": ["Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "Provably Powerful Graph Networks"], "answer_arxiv_id": ["1810.02244", "1905.11136"], "source_meta": {"published_time": "20220622"}, "qid": "AutoScholarQuery_train_17819"} +{"question": "Which works extend the study of conceptualization by constructing abstraction benchmarks based on eventualities?", "answer": ["AbsPyramid: Benchmarking the Abstraction Ability of Language Models with\n a Unified Entailment Graph"], "answer_arxiv_id": ["2311.09174"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_17820"} +{"question": "Can you name the researches which use LSTM, BERT, or graph neural networks for learning representations on the encoder side in text-to-SQL tasks?", "answer": ["RYANSQL: Recursively Applying Sketch-based Slot Fillings for Complex Text-to-SQL in Cross-Domain Databases", "A Comprehensive Exploration on WikiSQL with Table-Aware Word Contextualization", "RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers", "SADGA: Structure-Aware Dual Graph Aggregation Network for Text-to-SQL", "LGESQL: Line Graph Enhanced Text-to-SQL Model with Mixed Local and Non-Local Relations"], "answer_arxiv_id": ["2004.03125", "1902.01069", "1911.04942", "2111.00653", "2106.01093"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_17821"} +{"question": "What papers focus on improving the computational efficiency of Gromov-Wasserstein algorithms?", "answer": ["Semi-relaxed Gromov-Wasserstein divergence with applications on graphs", "Linear-Time Gromov Wasserstein Distances using Low Rank Couplings and Costs", "A Convergent Single-Loop Algorithm for Relaxation of Gromov-Wasserstein in Graph Data"], "answer_arxiv_id": ["2110.02753", "2106.01128", "2303.06595"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_17822"} +{"question": "Could you provide me a study that used the machine learning framework to model physical domains?", "answer": ["Learning to Simulate Complex Physics with Graph Networks"], "answer_arxiv_id": ["2002.09405"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_17823"} +{"question": "Which studies propose acceleration methods for Diffusion Models (DMs)?", "answer": ["Pseudo Numerical Methods for Diffusion Models on Manifolds", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2202.09778", "2206.00364"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_17824"} +{"question": "Can you provide me an example of a work that introduced event-based semantic segmentation through DSEC-Semantic dataset?", "answer": ["ESS: Learning Event-based Semantic Segmentation from Still Images"], "answer_arxiv_id": ["2203.10016"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_17825"} +{"question": "Which study is considered a concurrent work on linear-Q or linear MDPs with adversarial losses, bandit feedback, and unknown transitions?", "answer": ["Improved Regret for Efficient Online Reinforcement Learning with Linear Function Approximation", "Improved Regret Bounds for Linear Adversarial MDPs via Linear Optimization", "Delay-Adapted Policy Optimization and Improved Regret for Adversarial MDP with Delayed Bandit Feedback"], "answer_arxiv_id": ["2301.13087", "2302.06834", "2305.07911"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_17826"} +{"question": "Which works introduced the concept of activation editing?", "answer": ["Emergent world representations: Exploring a sequence model trained on a synthetic task"], "answer_arxiv_id": ["2210.13382"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_17827"} +{"question": "What works use task loss as a weighting term in the MSE loss for model-based RL?", "answer": ["Value Gradient weighted Model-Based Reinforcement Learning"], "answer_arxiv_id": ["2204.01464"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_17828"} +{"question": "Could you list some papers discussing regression-based methods for 3D face reconstruction?", "answer": ["Regressing Robust and Discriminative 3D Morphable Models with a very\n Deep Neural Network", "MoFA: Model-based Deep Convolutional Face Autoencoder for Unsupervised\n Monocular Reconstruction", "CNN-based Real-time Dense Face Reconstruction with Inverse-rendered\n Photo-realistic Face Images", "Unsupervised Training for 3D Morphable Model Regression", "Self-supervised Multi-level Face Model Learning for Monocular\n Reconstruction at over 250 Hz", "Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From\n Single Image to Image Set", "Learning an Animatable Detailed 3D Face Model from In-The-Wild Images", "Riggable 3D Face Reconstruction via In-Network Optimization", "Self-Supervised 3D Face Reconstruction via Conditional Estimation", "EMOCA: Emotion Driven Monocular Face Capture and Animation", "Towards Metrical Reconstruction of Human Faces", "Towards Racially Unbiased Skin Tone Estimation via Scene Disambiguation"], "answer_arxiv_id": ["1612.04904", "1703.10580", "1708.00980", "1806.06098", "1712.02859", "1903.08527", "2012.04012", "2104.03493", "2110.04800", "2204.11312", "2204.06607", "2205.03962"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17829"} +{"question": "Which research introduced the Mean Field IRL approach for discounted finite-horizon mean field games?", "answer": ["Individual-Level Inverse Reinforcement Learning for Mean Field Games"], "answer_arxiv_id": ["2202.06401"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_17830"} +{"question": "Which studies applied the min-max based adversarial training in the context of NLP and found it less effective?", "answer": ["Searching for an Effective Defender: Benchmarking Defense against Adversarial Word Substitution", "FreeLB: Enhanced Adversarial Training for Natural Language Understanding", "Towards Robustness Against Natural Language Word Substitutions", "InfoBERT: Improving Robustness of Language Models from An Information Theoretic Perspective"], "answer_arxiv_id": ["2108.12777", "1909.11764", "2107.13541", "2010.02329"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_17831"} +{"question": "What studies discuss self-supervised learning (SSL) approaches for speech that are based on masked span prediction?", "answer": ["wav2vec 2.0: A Framework for Self-Supervised Learning of Speech\n Representations", "HuBERT: Self-Supervised Speech Representation Learning by Masked\n Prediction of Hidden Units", "data2vec: A General Framework for Self-supervised Learning in Speech,\n Vision and Language", "Self-supervised Learning with Random-projection Quantizer for Speech\n Recognition"], "answer_arxiv_id": ["2006.11477", "2106.07447", "2202.03555", "2202.01855"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_17832"} +{"question": "Could you provide me some studies that developed knowledge distillation-based methods for class-incremental learning?", "answer": ["Learning without Forgetting", "iCaRL: Incremental Classifier and Representation Learning", "PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning", "Class-incremental Learning via Deep Model Consolidation", "On Learning the Geodesic Path for Incremental Learning", "Learning without Memorizing"], "answer_arxiv_id": ["1606.09282", "1611.07725", "2004.13513", "1903.07864", "2104.08572", "1811.08051"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_17833"} +{"question": "Can you list papers that explored action-free imitation by learning a dynamics model or a reward function?", "answer": ["Behavioral Cloning from Observation", "State-Only Imitation Learning for Dexterous Manipulation", "State Alignment-based Imitation Learning", "Imitating Latent Policies from Observation", "Playing hard exploration games by watching YouTube", "XIRL: Cross-embodiment Inverse Reinforcement Learning", "Unsupervised Perceptual Rewards for Imitation Learning"], "answer_arxiv_id": ["1805.01954", "2004.04650", "1911.10947", "1805.07914", "1805.11592", "2106.03911", "1612.06699v3"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_17834"} +{"question": "What studies proposed methods like RCI to enhance LLM’s reasoning performance?", "answer": ["Language Models can Solve Computer Tasks"], "answer_arxiv_id": ["2303.17491"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_17835"} +{"question": "Can you identify the study that addressed modeling temporal correlations through patch shifting rather than channel shifting?", "answer": ["Spatiotemporal Self-attention Modeling with Temporal Patch Shift for\n Action Recognition"], "answer_arxiv_id": ["2207.13259"], "source_meta": {"published_time": "20230818"}, "qid": "AutoScholarQuery_train_17836"} +{"question": "What papers have explored spatial hash tables and function decompositions for view angle dependence as world representations in Neural Radiance Fields?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "PlenOctrees for Real-time Rendering of Neural Radiance Fields"], "answer_arxiv_id": ["2201.05989", "2112.05131", "2103.14024"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_17837"} +{"question": "Can you provide a study that considered the biological implausibility of CNNs and showed a neuro-inspired approach to improve traditional CNNs and locally-connected networks?", "answer": ["Towards Biologically Plausible Convolutional Networks"], "answer_arxiv_id": ["2106.13031"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17838"} +{"question": "What work has developed a strategy similar to DBARF, which estimates camera poses alongside a generalizable NeRF?", "answer": ["DBARF: Deep Bundle-Adjusting Generalizable Neural Radiance Fields"], "answer_arxiv_id": ["2303.14478"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17839"} +{"question": "What papers applied VLMs to visual question answering?", "answer": ["PaLI: A Jointly-Scaled Multilingual Language-Image Model", "VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and Dataset", "PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2209.06794", "2304.08345", "2303.03378v1"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_17840"} +{"question": "Who defined α-separator and α-clique separator in their work?", "answer": ["Verification and search algorithms for causal DAGs"], "answer_arxiv_id": ["2206.15374"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_17841"} +{"question": "Which works focus on augmenting or synthetically-generating additional training images for domain generalization?", "answer": ["Adversarial Domain Adaptation with Domain Mixup", "Improving Out-of-Distribution Robustness via Selective Augmentation", "Domain Randomization and Pyramid Consistency: Simulation-to-Real Generalization without Accessing Target Domain Data", "Deep Domain-Adversarial Image Generation for Domain Generalisation"], "answer_arxiv_id": ["1912.01805", "2201.00299", "1909.00889", "2003.06054"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_17842"} +{"question": "Could you provide me some research that utilized neural adjoint methods to explore the solution space with different initializations?", "answer": ["Benchmarking Deep Inverse Models over time, and the Neural-Adjoint method", "Neural-adjoint method for the inverse design of all-dielectric metasurfaces", "Nanophotonic Particle Simulation and Inverse Design Using Artificial Neural Networks"], "answer_arxiv_id": ["2009.12919", "2012.05020", "1712.03222"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17843"} +{"question": "What works propose end-to-end methods for temporal action detection which use raw frames as input?", "answer": ["Learning Salient Boundary Feature for Anchor-free Temporal Action\n Localization", "RGB Stream Is Enough for Temporal Action Detection", "An Empirical Study of End-to-End Temporal Action Detection"], "answer_arxiv_id": ["2103.13137", "2107.04362", "2204.02932"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_17844"} +{"question": "What studies use the layer-to-layer strategy where only necessary tensors for a particular layer are transferred to the GPU memory?", "answer": ["Training Large Neural Networks with Constant Memory using a New\n Execution Algorithm"], "answer_arxiv_id": ["2002.05645"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_17845"} +{"question": "Which works describe methodologies to check whether a treatment effect estimate is robust to changes by varying the adjustment set?", "answer": ["A Robustness Test for Estimating Total Effects with Covariate Adjustment"], "answer_arxiv_id": ["2206.07533"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_17846"} +{"question": "Which works assume a threshold or a capacity for each arm in handling collisions in MPMAB model?", "answer": ["Multi-User Multi-Armed Bandits for Uncoordinated Spectrum Access", "Resource Allocation in NOMA-based Self-Organizing Networks using Stochastic Multi-Armed Bandits", "Dynamic Spectrum Access using Stochastic Multi-User Bandits", "Decentralized Heterogeneous Multi-Player Multi-Armed Bandits with Non-Zero Rewards on Collisions", "Multi-Player Multi-Armed Bandits with Finite Shareable Resources Arms: Learning Algorithms & Applications", "Multiple-Play Stochastic Bandits with Shareable Finite-Capacity Arms"], "answer_arxiv_id": ["1807.00867v5", "2101.06340", "2101.04388", "1910.09089", "2204.13502v1", "2206.08776"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_17847"} +{"question": "Could you specify the work that focuses on medical image-to-text generation tasks using diffusion-based methods?", "answer": ["ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on\n Weakly-Supervised Classification and Localization of Common Thorax Diseases"], "answer_arxiv_id": ["1705.02315"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_17848"} +{"question": "Which paper constructed a dataset to study the robustness of detectors?", "answer": ["MAGE: Machine-generated Text Detection in the Wild"], "answer_arxiv_id": ["2305.13242"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_17849"} +{"question": "Could you let me know works that are proposed for explicit augmentation methods?", "answer": ["Contrastive Learning for Sequential Recommendation", "Contrastive Self-supervised Sequential Recommendation with Robust Augmentation"], "answer_arxiv_id": ["2010.14395", "2108.06479"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_17850"} +{"question": "Which papers focused on face detectors designed to detect human faces?", "answer": ["Finding Tiny Faces", "S$^3$FD: Single Shot Scale-invariant Face Detector", "Feature Pyramid Networks for Object Detection", "SSH: Single Stage Headless Face Detector", "PyramidBox: A Context-assisted Single Shot Face Detector", "DSFD: Dual Shot Face Detector", "Sample and Computation Redistribution for Efficient Face Detection", "SSH: Single Stage Headless Face Detector"], "answer_arxiv_id": ["1612.04402", "1708.05237", "1612.03144", "1708.03979", "1803.07737", "1810.10220", "2105.04714", "1708.03979"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_17851"} +{"question": "What research papers generate inputs that maximally or minimally activate a neuron either using DNNs or iterative algorithms?", "answer": ["Understanding Deep Image Representations by Inverting Them", "Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space"], "answer_arxiv_id": ["1412.0035", "1612.00005"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_17852"} +{"question": "What papers are about constructing efficient attacks on deep learning architectures?", "answer": ["Explaining and Harnessing Adversarial Examples", "Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models", "Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1412.6572", "1712.04248", "1608.04644"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17853"} +{"question": "Which research proposed a simple contrastive learning framework for inter- and intra-modal dense feature contrast?", "answer": ["SimIPU: Simple 2D Image and 3D Point Cloud Unsupervised Pre-Training for Spatial-Aware Visual Representations"], "answer_arxiv_id": ["2112.04680"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_17854"} +{"question": "Can you tell me about any works that sequentially train the weights corresponding to a single iteration locally for training scheduling?", "answer": ["Learned Belief-Propagation Decoding with Simple Scaling and SNR Adaptation"], "answer_arxiv_id": ["1901.08621"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_17855"} +{"question": "What is the research about adding self-attention layers and scaling up to very large architectures?", "answer": ["Self-Attention Generative Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Scaling up GANs for Text-to-Image Synthesis"], "answer_arxiv_id": ["1805.08318", "1809.11096", "2303.05511"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_17856"} +{"question": "Which papers focused on GANs that concentrate on network architectures or latent space?", "answer": ["Unsupervised Representation Learning with Deep Convolutional Generative\n Adversarial Networks", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Conditional Generative Adversarial Nets", "InfoGAN: Interpretable Representation Learning by Information Maximizing\n Generative Adversarial Nets"], "answer_arxiv_id": ["1511.06434", "1809.11096", "1411.1784", "1606.03657"], "source_meta": {"published_time": "20220521"}, "qid": "AutoScholarQuery_train_17857"} +{"question": "Which works involve translation-based EA models?", "answer": ["Multilingual Knowledge Graph Embeddings for Cross-lingual Knowledge Alignment", "Cross-lingual Entity Alignment via Joint Attribute-Preserving Embedding", "TransEdge: Translating Relation-contextualized Embeddings for Knowledge Graphs"], "answer_arxiv_id": ["1611.03954", "1708.05045", "2004.13579"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_17858"} +{"question": "What studies extended question answering (QA) tasks by incorporating common sense?", "answer": ["From Recognition to Cognition: Visual Commonsense Reasoning"], "answer_arxiv_id": ["1811.10830"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_17859"} +{"question": "Which research papers focus on using 'Deep Ensembles' for uncertainty quantification in deep learning?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift"], "answer_arxiv_id": ["1612.01474", "1906.02530"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_17860"} +{"question": "What papers have explored the learnability of certain boolean concept classes using transformer-based hypothesis classes?", "answer": ["Inductive Biases and Variable Creation in Self-Attention Mechanisms", "Simplicity Bias in Transformers and their Ability to Learn Sparse Boolean Functions"], "answer_arxiv_id": ["2110.10090", "2211.12316"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_17861"} +{"question": "Could you point me towards some works that use diffusion or transformer models to transform 2D images into 3D representations for novel-view synthesis?", "answer": ["Generative Novel View Synthesis with 3D-Aware Diffusion Models", "Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion\n Prior", "NerfDiff: Single-image View Synthesis with NeRF-guided Distillation from\n 3D-aware Diffusion", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image", "Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision", "Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors"], "answer_arxiv_id": ["2304.02602", "2303.14184", "2302.10109", "2309.03453", "2306.11719", "2306.17843"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_17862"} +{"question": "Which studies proposed attention guidances using external annotations like bounding boxes?", "answer": ["BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained\n Diffusion", "Directed Diffusion: Direct Control of Object Placement through Attention\n Guidance", "Training-Free Location-Aware Text-to-Image Synthesis"], "answer_arxiv_id": ["2307.10816", "2302.13153", "2304.13427"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_17863"} +{"question": "Which works focus on text-to-image generation?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Muse: Text-To-Image Generation via Masked Generative Transformers"], "answer_arxiv_id": ["2112.10741", "2112.10752", "2301.00704"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_17864"} +{"question": "Could you provide me some studies that considered different approaches for conditional sampling in infinite dimensional diffusion models?", "answer": ["Neural Diffusion Processes"], "answer_arxiv_id": ["2206.03992"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_17865"} +{"question": "What sources provide the existing lower complexity bounds that were used to conclude about improvable dependency of GDA and GDmax on epsilon?", "answer": ["The Complexity of Nonconvex-Strongly-Concave Minimax Optimization", "Complexity Lower Bounds for Nonconvex-Strongly-Concave Min-Max Optimization"], "answer_arxiv_id": ["2103.15888", "2104.08708"], "source_meta": {"published_time": "20221031"}, "qid": "AutoScholarQuery_train_17866"} +{"question": "Which work introduced Residual Log-likelihood Estimation (RLE) into the DETR-based top-down framework?", "answer": ["Poseur: Direct Human Pose Regression with Transformers"], "answer_arxiv_id": ["2201.07412v2"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_17867"} +{"question": "Which works establish the concept of diffusion models estimating a target data distribution?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_17868"} +{"question": "Which hybrid methodologies let neural learners make decisions within algorithmic loops?", "answer": ["Learning to Iteratively Solve Routing Problems with Dual-Aspect Collaborative Transformer", "Learning Improvement Heuristics for Solving Routing Problems", "Learning 2-opt Heuristics for the Traveling Salesman Problem via Deep Reinforcement Learning", "Learning to Perform Local Rewriting for Combinatorial Optimization", "Learning Large Neighborhood Search Policy for Integer Programming", "Learning to Delegate for Large-scale Vehicle Routing"], "answer_arxiv_id": ["2110.02544", "1912.05784", "2004.01608", "1810.00337", "2111.03466", "2107.04139"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_17869"} +{"question": "Which works focus on objective evaluations of large language models, particularly through multiple-choice benchmarks?", "answer": ["Measuring Massive Multitask Language Understanding", "C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for\n Foundation Models", "CMMLU: Measuring massive multitask language understanding in Chinese"], "answer_arxiv_id": ["2009.03300", "2305.08322", "2306.09212"], "source_meta": {"published_time": "20240126"}, "qid": "AutoScholarQuery_train_17870"} +{"question": "Which research used BERT and BART for evaluation and prediction in the context of text generation?", "answer": ["BERTScore: Evaluating Text Generation with BERT", "MoverScore: Text Generation Evaluating with Contextualized Embeddings and Earth Mover Distance", "BLEURT: Learning Robust Metrics for Text Generation", "BARTScore: Evaluating Generated Text as Text Generation"], "answer_arxiv_id": ["1904.09675", "1909.02622", "2004.04696", "2106.11520"], "source_meta": {"published_time": "20220801"}, "qid": "AutoScholarQuery_train_17871"} +{"question": "Which papers involved the use of weakly-supervised information for supervised anomaly detection?", "answer": ["Real-world Anomaly Detection in Surveillance Videos", "Weakly-supervised Video Anomaly Detection with Robust Temporal Feature\n Magnitude Learning", "Not only Look, but also Listen: Learning Multimodal Violence Detection\n under Weak Supervision", "VadCLIP: Adapting Vision-Language Models for Weakly Supervised Video\n Anomaly Detection", "Unbiased Multiple Instance Learning for Weakly Supervised Video Anomaly\n Detection", "MGFN: Magnitude-Contrastive Glance-and-Focus Network for\n Weakly-Supervised Video Anomaly Detection"], "answer_arxiv_id": ["1801.04264", "2101.10030", "2007.04687", "2308.11681", "2303.12369", "2211.15098"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_17872"} +{"question": "What papers employed correlated Brownian motion to enhance SGD's capability to explore loss landscape?", "answer": ["On the Theoretical Properties of Noise Correlation in Stochastic Optimization"], "answer_arxiv_id": ["2209.09162"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_17873"} +{"question": "Which works cover Out-of-Distribution (OOD) generalization problem?", "answer": ["Domain Generalization: A Survey", "Towards Out-Of-Distribution Generalization: A Survey", "Improving Out-of-Distribution Robustness via Selective Augmentation"], "answer_arxiv_id": ["2103.02503", "2108.13624v2", "2201.00299"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_17874"} +{"question": "What studies are involved in image blending methods that target at combining the foreground and background seamlessly?", "answer": ["GP-GAN: Towards Realistic High-Resolution Image Blending", "Deep Image Blending", "Deep Image Compositing"], "answer_arxiv_id": ["1703.07195", "1910.11495", "2011.02146"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_17875"} +{"question": "What are the pieces of literature that focus on evaluating generative models, particularly using Fréchet Inception Distance?", "answer": ["GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium"], "answer_arxiv_id": ["1706.08500"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_17876"} +{"question": "What works have applied sampling-based decoding strategies to neural machine translation?", "answer": ["Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural Machine Translation", "Energy-Based Reranking: Improving Neural Machine Translation Using Energy-Based Models", "Understanding the Properties of Minimum Bayes Risk Decoding in Neural Machine Translation", "High Quality Rather than High Model Probability: Minimum Bayes Risk Decoding with Neural Metrics"], "answer_arxiv_id": ["2005.10283", "2009.13267", "2105.08504", "2111.09388"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_17877"} +{"question": "What papers reported toxic or harmful behavior exhibited by language models?", "answer": ["The Woman Worked as a Babysitter: On Biases in Language Generation", "Towards Understanding and Mitigating Social Biases in Language Models", "Universal Adversarial Triggers for Attacking and Analyzing NLP"], "answer_arxiv_id": ["1909.01326", "2106.13219", "1908.07125"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_17878"} +{"question": "What papers directly estimate either densities or density ratios in the context of state-action occupancy measure?", "answer": ["DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections", "GenDICE: Generalized Offline Estimation of Stationary Values"], "answer_arxiv_id": ["1906.04733", "2002.09072"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_17879"} +{"question": "What papers deal with adversarial training methods used for domain adaptation?", "answer": ["Domain-Adversarial Training of Neural Networks", "Adversarial Discriminative Domain Adaptation", "Bridging Theory and Algorithm for Domain Adaptation", "Unsupervised Domain Adaptation Based on Source-guided Discrepancy", "Blending-target Domain Adaptation by Adversarial Meta-Adaptation Networks", "Adaptation Across Extreme Variations using Unlabeled Domain Bridges"], "answer_arxiv_id": ["1505.07818", "1702.05464", "1904.05801", "1809.03839", "1907.03389", "1906.02238v2"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_17880"} +{"question": "Which works have proposed using fusion-encoder transformers in generative VLMs?", "answer": ["LXMERT: Learning Cross-Modality Encoder Representations from\n Transformers", "UNITER: UNiversal Image-TExt Representation Learning", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks"], "answer_arxiv_id": ["1908.07490", "1909.11740", "2004.06165"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_17881"} +{"question": "Can you cite the works that are about retrieval-augmented methods for knowledge-intensive NLP?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "REALM: Retrieval-Augmented Language Model Pre-Training"], "answer_arxiv_id": ["2005.11401", "2002.08909"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_17882"} +{"question": "Any works about using NeRF-based per-scene optimization for 3D reconstruction?", "answer": ["Wonder3D: Single Image to 3D using Cross-Domain Diffusion", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image", "MVDream: Multi-view Diffusion for 3D Generation"], "answer_arxiv_id": ["2310.15008", "2309.03453", "2308.16512"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_17883"} +{"question": "Could you cite studies that proposed to use DNN explainers based on exact or approximated Shapley values?", "answer": ["Shapley Explanation Networks", "FastSHAP: Real-Time Shapley Value Estimation", "Learning to Estimate Shapley Values with Vision Transformers", "CXPlain: Causal Explanations for Model Interpretation under Uncertainty"], "answer_arxiv_id": ["2104.02297", "2107.07436", "2206.05282", "1910.12336"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_17884"} +{"question": "Which works focus on enhancing the object localization capabilities of multimodal models?", "answer": ["Kosmos-2: Grounding Multimodal Large Language Models to the World", "Universal Instance Perception as Object Discovery and Retrieval", "DetGPT: Detect What You Need via Reasoning"], "answer_arxiv_id": ["2306.14824", "2303.06674", "2305.14167"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_17885"} +{"question": "Which works belong to the field of large-scale pre-trained text-to-image diffusion models?", "answer": ["eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Improved Denoising Diffusion Probabilistic Models", "Distribution Knowledge Embedding for Graph Pooling", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2211.01324", "2112.10752", "2204.06125", "2112.10741", "2102.09672", "2109.14333", "2105.05233"], "source_meta": {"published_time": "20240522"}, "qid": "AutoScholarQuery_train_17886"} +{"question": "Could you mention some research papers that have studied binary classification in overparameterized linear models?", "answer": ["Classification vs regression in overparameterized regimes: Does the loss function matter?", "Finite-sample Analysis of Interpolating Linear Classifiers in the Overparameterized Regime"], "answer_arxiv_id": ["2005.08054", "2004.12019"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_17887"} +{"question": "Could you provide me with some research papers that proposed variants of the SPC assumption to account for partial data coverage when general function approximation is used?", "answer": ["Bellman-consistent Pessimism for Offline Reinforcement Learning", "Adversarially Trained Actor Critic for Offline Reinforcement Learning"], "answer_arxiv_id": ["2106.06926", "2202.02446"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_17888"} +{"question": "Which dataset is reported to have domain weights that are tuned using downstream data?", "answer": ["GLaM: Efficient Scaling of Language Models with Mixture-of-Experts"], "answer_arxiv_id": ["2112.06905"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_17889"} +{"question": "Could you mention the studies that first define the Iterated CVaR measure and consider iterated coherent risk measures in MDPs?", "answer": ["Iterated risk measures for risk-sensitive Markov decision processes with discounted cost", "Markov Decision Processes with Recursive Risk Measures"], "answer_arxiv_id": ["1202.3755", "2010.07220"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_17890"} +{"question": "What papers focused on deriving global minimizers of optimization problems related to DNNs within the context of MSE-NC in unconstrained/layer-peeled models?", "answer": ["Revealing the Structure of Deep Neural Networks via Convex Duality", "Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training", "Extended Unconstrained Features Model for Exploring Deep Neural Collapse"], "answer_arxiv_id": ["2002.09773", "2101.12699", "2202.08087"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_17891"} +{"question": "Which studies have adopted the idea of masked modeling in the context of self-supervised learning for 3D domain?", "answer": ["Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point\n Modeling", "Masked Autoencoders for Point Cloud Self-supervised Learning"], "answer_arxiv_id": ["2111.14819", "2203.06604"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_17892"} +{"question": "What other papers extended the work of the Denoising Diffusion Probabilistic Model?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2211.01324", "2112.10752", "2205.11487", "2307.01952"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_17893"} +{"question": "Could you provide me studies that introduced learning a parameterized reward function with inverse reinforcement learning?", "answer": ["Learning Robust Rewards with Adversarial Inverse Reinforcement Learning"], "answer_arxiv_id": ["1710.11248"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_17894"} +{"question": "Which works pertain to optimization methods in multi-task learning for dense prediction tasks in computer vision?", "answer": ["Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics", "GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep\n Multitask Networks", "A Modulation Module for Multi-task Learning with Applications in Image\n Retrieval", "Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign\n Dropout"], "answer_arxiv_id": ["1705.07115v3", "1711.02257", "1807.06708", "2010.06808"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17895"} +{"question": "Could you provide me some studies about the methods sample goals that maximize Learning Progress in exploratory goal methods?", "answer": ["CURIOUS: Intrinsically Motivated Modular Multi-Goal Reinforcement Learning", "Teacher algorithms for curriculum learning of Deep RL in continuously parameterized environments"], "answer_arxiv_id": ["1810.06284", "1910.07224"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_17896"} +{"question": "What works propose to relax the nonnegative lower bound to a negative one?", "answer": ["Extragradient method with variance reduction for stochastic variational inequalities", "Fast Extra Gradient Methods for Smooth Structured Nonconvex-Nonconcave Minimax Problems", "Accelerated Single-Call Methods for Constrained Min-Max Optimization", "Efficient Methods for Structured Nonconvex-Nonconcave Min-Max Optimization"], "answer_arxiv_id": ["1703.00260v1", "2106.02326", "2210.03096", "2011.00364"], "source_meta": {"published_time": "20221226"}, "qid": "AutoScholarQuery_train_17897"} +{"question": "Which papers have defined rules on top of annotations by surrogate models for image data?", "answer": ["Slice-based Learning: A Programming Model for Residual Learning in Critical Data Slices"], "answer_arxiv_id": ["1909.06349"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_17898"} +{"question": "Are there any works in which masked autoencoding has been applied in NLP?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1810.04805", "2005.14165"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_17899"} +{"question": "Which paper reformulated the complex OCDA problem as multiple UDA problems?", "answer": ["Discover, Hallucinate, and Adapt: Open Compound Domain Adaptation for Semantic Segmentation"], "answer_arxiv_id": ["2110.04111"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_17900"} +{"question": "Any works examining models trained on GPT model outputs?", "answer": ["The False Promise of Imitating Proprietary LLMs"], "answer_arxiv_id": ["2305.15717"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_17901"} +{"question": "What works proposed adversarial inimax objectives to generate robust recourses and what were their results?", "answer": ["Towards Robust and Reliable Algorithmic Recourse", "On the Adversarial Robustness of Causal Algorithmic Recourse"], "answer_arxiv_id": ["2102.13620", "2112.11313"], "source_meta": {"published_time": "20220313"}, "qid": "AutoScholarQuery_train_17902"} +{"question": "Could you list the studies that worked on non-autoregressive models by allowing multiple additional iterations to refine any inaccurate predictions?", "answer": ["Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative Refinement", "Mask-Predict: Parallel Decoding of Conditional Masked Language Models"], "answer_arxiv_id": ["1802.06901", "1904.09324"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_17903"} +{"question": "What are the research papers proposing methods for generating sound and music using language models?", "answer": ["AudioGen: Textually Guided Audio Generation", "MusicLM: Generating Music From Text", "SingSong: Generating musical accompaniments from singing"], "answer_arxiv_id": ["2209.15352", "2301.11325", "2301.12662"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_17904"} +{"question": "Could you provide me the papers that introduce a metric IER to indicate triangle presence in reconstructed meshes?", "answer": ["Meshing Point Clouds with Predicted Intrinsic-Extrinsic Ratio Guidance"], "answer_arxiv_id": ["2007.09267"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_17905"} +{"question": "Could you mention some works that use additional annotated data for training in HBox-supervised oriented object detection?", "answer": ["Leveraging Orientation for Weakly Supervised Object Detection with Application to Firearm Localization", "Knowledge Combination to Learn Rotated Detection Without Rotated Annotation"], "answer_arxiv_id": ["1904.10032", "2304.02199"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_17906"} +{"question": "What research papers have focused on fair methods for supervised learning tasks?", "answer": ["The Frontiers of Fairness in Machine Learning", "Empirical Risk Minimization Under Fairness Constraints", "Fairness Through Awareness", "Fairness Constraints: Mechanisms for Fair Classification"], "answer_arxiv_id": ["1810.08810", "1802.08626", "1104.3913", "1507.05259"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_17907"} +{"question": "What studies simply modify the quantization operator as part of the non-uniform quantization approach?", "answer": ["Convolutional Neural Networks using Logarithmic Data Representation", "Incremental Network Quantization: Towards Lossless CNNs with Low-Precision Weights"], "answer_arxiv_id": ["1603.01025v2", "1702.03044"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_17908"} +{"question": "What papers discussed about neural processes in context of deep probabilistic regression tools?", "answer": ["Neural Processes", "Conditional Neural Processes"], "answer_arxiv_id": ["1807.01622", "1807.01613"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_17909"} +{"question": "What studies enhanced the realism of reconstructed images by optimizing neural codecs using a triple rate-distortion-perception loss?", "answer": ["High-Fidelity Generative Image Compression", "Perceptual Learned Video Compression with Recurrent Conditional GAN", "Multi-Realism Image Compression with a Conditional Generator"], "answer_arxiv_id": ["2006.09965", "2109.03082", "2212.13824"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_17910"} +{"question": "Could you provide some works that shown BNNs to be robust to gradient-based attacks?", "answer": ["Robustness of Bayesian Neural Networks to Gradient-Based Attacks"], "answer_arxiv_id": ["2002.04359"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_17911"} +{"question": "Can you provide some research papers where pseudo-labeling technique was used for UDA?", "answer": ["Prototypical Pseudo Label Denoising and Target Structure Learning for\n Domain Adaptive Semantic Segmentation", "Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain\n Adaptive Semantic Segmentation", "Confidence Regularized Self-Training", "HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic\n Segmentation", "DAFormer: Improving Network Architectures and Training Strategies for\n Domain-Adaptive Semantic Segmentation", "MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation"], "answer_arxiv_id": ["2101.10979", "2003.03773", "1908.09822v3", "2204.13132", "2111.14887", "2212.01322"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_17912"} +{"question": "What papers proposed the symmetric losses and their combinations with the CE loss or its variants?", "answer": ["Robust Loss Functions under Label Noise for Deep Neural Networks", "Symmetric Cross Entropy for Robust Learning with Noisy Labels", "Normalized Loss Functions for Deep Learning with Noisy Labels", "Generalized Jensen-Shannon Divergence Loss for Learning with Noisy Labels"], "answer_arxiv_id": ["1712.09482", "1908.06112", "2006.13554", "2105.04522"], "source_meta": {"published_time": "20211230"}, "qid": "AutoScholarQuery_train_17913"} +{"question": "Could you provide me with some works that analysed Lyapunov stability of neural networks?", "answer": ["Step Size Matters in Deep Learning"], "answer_arxiv_id": ["1805.08890"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_17914"} +{"question": "What studies agreed on the effectiveness of BERT models trained with domain-specific data?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1810.04805"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_17915"} +{"question": "Which studies have presented backdoor defense methods specifically designed for white-box models?", "answer": ["Backdoor Learning: A Survey", "BackdoorBench: A Comprehensive Benchmark of Backdoor Learning"], "answer_arxiv_id": ["2007.08745", "2206.12654"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_17916"} +{"question": "What are some of the applications of iterated learning in existing literature?", "answer": ["Compositional Languages Emerge in a Neural Iterated Learning Model", "Iterated learning for emergent systematicity in VQA", "Countering Language Drift with Seeded Iterated Learning", "Multi-label Iterated Learning for Image Classification with Label Ambiguity", "The Primacy Bias in Deep Reinforcement Learning"], "answer_arxiv_id": ["2002.01365", "2105.01119", "2003.12694", "2111.12172", "2205.07802"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_17917"} +{"question": "Could you provide me some research that incorporated long-tail object categories in object detection methods?", "answer": ["LVIS: A Dataset for Large Vocabulary Instance Segmentation"], "answer_arxiv_id": ["1908.03195"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_17918"} +{"question": "Which research works discuss low bounds in optimization?", "answer": ["A Lower Bound for the Optimization of Finite Sums", "Lower Bounds for Parallel and Randomized Convex Optimization", "Communication Complexity of Distributed Convex Learning and Optimization", "Parallelization does not Accelerate Convex Optimization: Adaptivity Lower Bounds for Non-smooth Convex Minimization", "Lower Bounds and Nearly Optimal Algorithms in Distributed Learning with Communication Compression", "Revisiting Optimal Convergence Rate for Smooth and Non-convex Stochastic Decentralized Optimization", "The Complexity of Making the Gradient Small in Stochastic Convex Optimization", "Bidirectional compression in heterogeneous settings for distributed or federated learning with partial participation: tight convergence guarantees", "Lower Bounds and Accelerated Algorithms in Distributed Stochastic Optimization with Communication Compression"], "answer_arxiv_id": ["1410.0723", "1811.01903", "1506.01900", "1808.03880", "2206.03665", "2210.07863", "1902.04686v2", "2006.14591", "2305.07612"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_17919"} +{"question": "Which studies are about enriching the training degradation space through data distribution learning in blind image super-resolution?", "answer": ["Better \"CMOS\" Produces Clearer Images: Learning Space-Variant Blur\n Estimation for Blind Image Super-Resolution", "From Face to Natural Image: Learning Real Degradation for Blind Image\n Super-Resolution", "To learn image super-resolution, use a GAN to learn how to do image\n degradation first"], "answer_arxiv_id": ["2304.03542", "2210.00752", "1807.11458"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_17920"} +{"question": "Which work relaxed the condition to the star-variant referred to as the weak Minty variational inequality (MVI)?", "answer": ["Efficient Methods for Structured Nonconvex-Nonconcave Min-Max Optimization"], "answer_arxiv_id": ["2011.00364"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_17921"} +{"question": "Which works propose directly correcting the training labels by estimating the noise transition matrix?", "answer": ["Are Anchor Points Really Indispensable in Label-Noise Learning?", "Dual T: Reducing Estimation Error for Transition Matrix in Label-noise\n Learning", "Making Deep Neural Networks Robust to Label Noise: a Loss Correction\n Approach"], "answer_arxiv_id": ["1906.00189", "2006.07805", "1609.03683"], "source_meta": {"published_time": "20220209"}, "qid": "AutoScholarQuery_train_17922"} +{"question": "What papers proposed to augment documents at indexing time with a number of generated queries?", "answer": ["Bridging the Gap Between Indexing and Retrieval for Differentiable Search Index with Query Generation", "A Neural Corpus Indexer for Document Retrieval", "Ultron: An Ultimate Retriever on Corpus with a Model-based Indexer"], "answer_arxiv_id": ["2206.10128", "2206.02743", "2208.09257"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_17923"} +{"question": "Which works focused on improving dialogue systems by collecting human feedback?", "answer": ["Deploying Lifelong Open-Domain Dialogue Learning", "Learning New Skills after Deployment: Improving open-domain internet-driven dialogue with human feedback"], "answer_arxiv_id": ["2008.08076", "2208.03270"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_17924"} +{"question": "Can you list the studies that propose different self-supervised formulations for gradient-based causal structure learning?", "answer": ["Differentiable Causal Discovery from Interventional Data", "DiBS: Differentiable Bayesian Structure Learning", "BCD Nets: Scalable Variational Approaches for Bayesian Causal Discovery", "Variational Causal Networks: Approximate Bayesian Inference over Causal Structures", "Deep End-to-end Causal Inference", "Bayesian Structure Learning with Generative Flow Networks", "Learning Neural Causal Models from Unknown Interventions", "Efficient Neural Causal Discovery without Acyclicity Constraints"], "answer_arxiv_id": ["2007.01754", "2105.11839", "2112.02761", "2106.07635", "2202.02195", "2202.13903", "1910.01075", "2107.10483"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_17925"} +{"question": "Could you mention some works that utilized disagreements between SGD runs for predicting in-distribution generalization?", "answer": ["Assessing Generalization of SGD via Disagreement"], "answer_arxiv_id": ["2106.13799"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_17926"} +{"question": "What researches propose techniques for bridging the domain gap and the data gap in unsupervised machine translation?", "answer": ["Bridging the Data Gap between Training and Inference for Unsupervised Neural Machine Translation"], "answer_arxiv_id": ["2203.08394"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_17927"} +{"question": "Are there any known works inspired by the cross-attention module of Perceiver?", "answer": ["Perceiver: General Perception with Iterative Attention"], "answer_arxiv_id": ["2103.03206"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_17928"} +{"question": "Which studies have discussed the utilization of Large Language Models (LLMs) to generate explicit textual knowledge or training data for smaller models?", "answer": ["COMET: Commonsense Transformers for Automatic Knowledge Graph\n Construction", "Symbolic Knowledge Distillation: from General Language Models to\n Commonsense Models", "Self-Instruct: Aligning Language Models with Self-Generated Instructions", "Distilling Reasoning Capabilities into Smaller Language Models", "Symbolic Chain-of-Thought Distillation: Small Models Can Also \"Think\"\n Step-by-Step"], "answer_arxiv_id": ["1906.05317", "2110.07178", "2212.10560", "2212.00193", "2306.14050"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_17929"} +{"question": "Which papers discuss fine-tuning LLMs with annotated instructional data to follow general language instructions?", "answer": ["Learning from Task Descriptions", "Multitask Prompted Training Enables Zero-Shot Task Generalization", "Cross-Task Generalization via Natural Language Crowdsourcing\n Instructions"], "answer_arxiv_id": ["2011.08115", "2110.08207", "2104.08773"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_17930"} +{"question": "Which method ensures exact representation of local cost without learning optimal values?", "answer": ["Plan2Vec: Unsupervised Representation Learning by Latent Plans"], "answer_arxiv_id": ["2005.03648v1"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_17931"} +{"question": "Which studies discussed deep learning methods used for time-series forecasting?", "answer": ["FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting", "Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting", "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting"], "answer_arxiv_id": ["2201.12740", "2012.07436", "2106.13008"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_17932"} +{"question": "What work proposed mapping classification tasks to a question-answering format?", "answer": ["Adapting Language Models for Zero-shot Learning by Meta-tuning on Dataset and Prompt Collections"], "answer_arxiv_id": ["2104.04670"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_17933"} +{"question": "What studies have explored probabilistic models for anomaly detection in videos?", "answer": ["Joint Detection and Recounting of Abnormal Events by Learning Deep\n Generic Knowledge"], "answer_arxiv_id": ["1709.09121"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_17934"} +{"question": "What are the works that develop approaches for multi-scene representation using a single network to tackle the computational issue related to neural radiance fields?", "answer": ["R"], "answer_arxiv_id": ["1210.6589"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_17935"} +{"question": "Which work obtains Ĥ-consistency bounds of the multiclass logistic loss under a little stronger assumption ?", "answer": ["Cross-Entropy Loss Functions: Theoretical Analysis and Applications"], "answer_arxiv_id": ["2304.07288"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_17936"} +{"question": "Which studies proposed end-to-end deep learning frameworks with surface normal as intermediate representations to guide depth enhancement?", "answer": ["DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene\n from Sparse LiDAR Data and Single Color Image", "Depth Completion from Sparse LiDAR Data with Depth-Normal Constraints"], "answer_arxiv_id": ["1812.00488", "1910.06727"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_17937"} +{"question": "Which works discussed the design of Diffusion Models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["1503.03585", "2011.13456"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_17938"} +{"question": "What studies discussed the limitations of discrete VAE models in using continuous relaxation for backpropagation?", "answer": ["Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_17939"} +{"question": "What research proposed using proxy tasks like masked language modeling and masked frame modeling during pre-training of VLMs?", "answer": ["Less is More: ClipBERT for Video-and-Language Learning via Sparse Sampling", "All in One: Exploring Unified Video-Language Pre-training"], "answer_arxiv_id": ["2102.06183", "2203.07303"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_17940"} +{"question": "What studies improved the efficiency of ViTs by applying token-level sparsification?", "answer": ["Adaptive Sparse ViT: Towards Learnable Adaptive Token Pruning by Fully Exploiting Self-Attention", "AdaViT: Adaptive Tokens for Efficient Vision Transformer", "SPViT: Enabling Faster Vision Transformers via Soft Token Pruning", "Learning A Sparse Transformer Network for Effective Image Deraining", "Chasing Sparsity in Vision Transformers: An End-to-End Exploration", "Learned Token Pruning for Transformers", "DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification"], "answer_arxiv_id": ["2209.13802v2", "2112.07658", "2112.13890", "2303.11950", "2106.04533", "2107.00910", "2106.02034v2"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_17941"} +{"question": "Which papers discuss modifications of Gradient Descent in relation to gradient flow that minimizes the GFS sharpness?", "answer": ["Understanding the Generalization Benefit of Normalization Layers: Sharpness Reduction"], "answer_arxiv_id": ["2206.07085"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_17942"} +{"question": "Which works utilized Euler angles as rotation representations?", "answer": ["Viewpoints and Keypoints"], "answer_arxiv_id": ["1411.6067"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_17943"} +{"question": "Can you tell me about the research paper which introduced Shap·E that targets training diffusion model on the 3D implicit function?", "answer": ["Shap-E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2305.02463"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_17944"} +{"question": "Which work provides a comprehensive comparison of Meta-Continual Learning (MCL) with other fields such as continual meta-learning?", "answer": ["When Meta-Learning Meets Online and Continual Learning: A Survey"], "answer_arxiv_id": ["2311.05241"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_17945"} +{"question": "Which work integrates contrastive learning into EEG-to-Text decoding?", "answer": ["BELT:Bootstrapping Electroencephalography-to-Language Decoding and Zero-Shot Sentiment Classification by Natural Language Supervision"], "answer_arxiv_id": ["2309.12056v2"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_17946"} +{"question": "Do you know any work like ProcGen which supports infinite procedurally generated environments?", "answer": ["Leveraging Procedural Generation to Benchmark Reinforcement Learning"], "answer_arxiv_id": ["1912.01588"], "source_meta": {"published_time": "20221123"}, "qid": "AutoScholarQuery_train_17947"} +{"question": "Which works introduced the concept of vision language foundation models?", "answer": ["Contrastive Learning of Medical Visual Representations from Paired\n Images and Text", "Learning Deep Structure-Preserving Image-Text Embeddings", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "VL-BERT: Pre-training of Generic Visual-Linguistic Representations"], "answer_arxiv_id": ["2010.00747", "1511.06078", "2107.07651", "2004.06165", "1908.02265", "1908.08530"], "source_meta": {"published_time": "20230907"}, "qid": "AutoScholarQuery_train_17948"} +{"question": "Could you provide an example of works that proposed nearly linear time approximations for clustering problems with provable approximation guarantees?", "answer": ["Local Search Yields a PTAS for k-Means in Doubling Metrics", "Local search yields approximation schemes for k-means and k-median in Euclidean and minor-free metrics", "Near-Linear Time Approximation Schemes for Clustering in Doubling Metrics"], "answer_arxiv_id": ["1603.08976v2", "1603.09535", "1812.08664"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_17949"} +{"question": "What papers are associated with the application of image-conditioned diffusion to enhance the generation of unseen viewpoints?", "answer": ["NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as\n General Image Priors"], "answer_arxiv_id": ["2212.03267"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_17950"} +{"question": "What are the studies about VCOD that segmented the moving camouflaged objects via dense optical flow?", "answer": ["It's Moving! A Probabilistic Model for Causal Motion Segmentation in\n Moving Camera Videos"], "answer_arxiv_id": ["1604.00136"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_17951"} +{"question": "Which studies have been referenced as making efforts to reduce the variance of REINFORCE?", "answer": ["MuProp: Unbiased Backpropagation for Stochastic Neural Networks", "REBAR: Low-variance, unbiased gradient estimates for discrete latent variable models", "Backpropagation through the Void: Optimizing control variates for black-box gradient estimation", "Gradient Estimation with Discrete Stein Operators"], "answer_arxiv_id": ["1511.05176", "1703.07370", "1711.00123", "2202.09497"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_17952"} +{"question": "What studies are related to the statistical aspects of time-series prediction?", "answer": ["Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification"], "answer_arxiv_id": ["1802.08334"], "source_meta": {"published_time": "20230117"}, "qid": "AutoScholarQuery_train_17953"} +{"question": "What studies proposed solutions to limitations such as slow inference speed and performance degradation in sequence-to-sequence models?", "answer": ["Uni-Perceiver: Pre-training Unified Architecture for Generic Perception for Zero-shot and Few-shot Tasks", "Uni-Perceiver-MoE: Learning Sparse Generalist Models with Conditional MoEs", "Uni-Perceiver v2: A Generalist Model for Large-Scale Vision and Vision-Language Tasks"], "answer_arxiv_id": ["2112.01522", "2206.04674", "2211.09808"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_17954"} +{"question": "Are there any studies that directly manipulate NeRF fields or include physical simulators to achieve more dynamic behaviors?", "answer": ["NeuralEditor: Editing Neural Radiance Fields via Manipulating Point\n Clouds", "Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis", "Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene\n Reconstruction", "4D Gaussian Splatting for Real-Time Dynamic Scene Rendering", "Real-time Photorealistic Dynamic Scene Representation and Rendering with\n 4D Gaussian Splatting", "PAC-NeRF: Physics Augmented Continuum Neural Radiance Fields for\n Geometry-Agnostic System Identification"], "answer_arxiv_id": ["2305.03049", "2308.09713", "2309.13101", "2310.08528", "2310.10642", "2303.05512"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_17955"} +{"question": "Do you know any papers that implemented approximations in Transformer models through kernelization?", "answer": ["Rethinking Attention with Performers", "Random Feature Attention"], "answer_arxiv_id": ["2009.14794", "2103.02143"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_17956"} +{"question": "Any works about synthesizing plausible 3D scene arrangements?", "answer": ["Human-centric Indoor Scene Synthesis Using Stochastic Grammar"], "answer_arxiv_id": ["1808.08473"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_17957"} +{"question": "What are the studies that focus on estimating and measuring the consistency of LLMs given diverse prompts?", "answer": ["Measuring and Improving Consistency in Pretrained Language Models", "Statistical Knowledge Assessment for Large Language Models"], "answer_arxiv_id": ["2102.01017", "2305.10519"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_17958"} +{"question": "What sources discuss how uniformity and alignment in feature space are keys to a good representation in unsupervised learning?", "answer": ["Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere"], "answer_arxiv_id": ["2005.10242"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_17959"} +{"question": "What studies employed equivariant probability distributions for modeling symmetrical density functions?", "answer": ["Equivariant Flows: Exact Likelihood Generative Learning for Symmetric Densities", "E(n) Equivariant Normalizing Flows", "GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation", "Exchangeable Neural ODE for Set Modeling", "Scalable Normalizing Flows for Permutation Invariant Densities"], "answer_arxiv_id": ["2006.02425", "2105.09016", "2203.02923", "2008.02676", "2010.03242"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_17960"} +{"question": "What studies discuss model stealing from classifiers?", "answer": ["SSLGuard: A Watermarking Scheme for Self-supervised Learning Pre-trained\n Encoders", "StolenEncoder: Stealing Pre-trained Encoders in Self-supervised Learning"], "answer_arxiv_id": ["2201.11692", "2201.05889"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_17961"} +{"question": "Which studies discuss physically based inverse rendering of humans?", "answer": ["Neural Light Transport for Relighting and View Synthesis", "Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing", "Learning Visibility Field for Detailed 3D Human Reconstruction and\n Relighting", "RANA: Relightable Articulated Neural Avatars", "Relighting4D: Neural Relightable Human from Videos"], "answer_arxiv_id": ["2008.03806", "2204.08906", "2304.11900", "2212.03237", "2207.07104"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_17962"} +{"question": "Which research papers presented pioneering approaches for distilling semantic features from pretrained models into NeRF for decomposition?", "answer": ["Decomposing NeRF for Editing via Feature Field Distillation", "Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D\n Image Representations"], "answer_arxiv_id": ["2205.15585", "2209.03494"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_17963"} +{"question": "Could you provide me some studies that propose variance reduction techniques for training diffusion models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Maximum Likelihood Training of Score-Based Diffusion Models", "Variational Diffusion Models"], "answer_arxiv_id": ["2102.09672", "2101.09258", "2107.00630"], "source_meta": {"published_time": "20230506"}, "qid": "AutoScholarQuery_train_17964"} +{"question": "Which papers discuss text-to-image generation using GANs?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image\n Synthesis", "Improving Text-to-Image Synthesis Using Contrastive Learning", "DF-GAN: A Simple and Effective Baseline for Text-to-Image Synthesis"], "answer_arxiv_id": ["1711.10485", "1904.01310", "2107.02423", "2008.05865"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_17965"} +{"question": "What are the benchmarks that evaluate both the knowledge memorization and reasoning capabilities of LLMs within specific domains?", "answer": ["KoLA: Carefully Benchmarking World Knowledge of Large Language Models", "SeaEval for Multilingual Foundation Models: From Cross-Lingual Alignment\n to Cultural Reasoning", "LawBench: Benchmarking Legal Knowledge of Large Language Models"], "answer_arxiv_id": ["2306.09296v3", "2309.04766", "2309.16289"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_17966"} +{"question": "What studies estimated fairness under the scenario where one develops a predictor for an unobserved covariate?", "answer": ["Measuring Model Fairness under Noisy Covariates: A Theoretical Perspective"], "answer_arxiv_id": ["2105.09985"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_17967"} +{"question": "Are there any works that investigate the principle of multi-sampling for image denoising?", "answer": ["Noise2Noise: Learning Image Restoration without Clean Data", "Noise2NoiseFlow: Realistic Camera Noise Modeling without Clean Images", "Noise2Void - Learning Denoising from Single Noisy Images", "Neighbor2Neighbor: Self-Supervised Denoising from Single Noisy Images"], "answer_arxiv_id": ["1803.04189v3", "2206.01103", "1811.10980", "2101.02824"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_17968"} +{"question": "Which work developed a deep learning-based matcher that can simultaneously match interest points and reject outliers?", "answer": ["SuperGlue: Learning Feature Matching with Graph Neural Networks"], "answer_arxiv_id": ["1911.11763"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_17969"} +{"question": "Any works about quantizing SSL representations using k-means and later perform language modeling?", "answer": ["On Generative Spoken Language Modeling from Raw Audio", "textless-lib: a Library for Textless Spoken Language Processing", "AudioLM: a Language Modeling Approach to Audio Generation"], "answer_arxiv_id": ["2102.01192", "2202.07359", "2209.03143"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_17970"} +{"question": "Could you provide me some works involving the application of low-rank structure in earlier MTL methods?", "answer": ["Trace Norm Regularised Deep Multi-Task Learning"], "answer_arxiv_id": ["1606.04038"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_17971"} +{"question": "Which paper provides a regret bound that scales with the variance under the MDP setting?", "answer": ["Variance-Aware Regret Bounds for Undiscounted Reinforcement Learning in MDPs"], "answer_arxiv_id": ["1803.01626"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_17972"} +{"question": "Which work introduced an approach in synthesizing realistic clean-noisy image pairs by considering in-camera Image Signal Processing (ISP) pipeline for noise synthesis?", "answer": ["Toward Convolutional Blind Denoising of Real Photographs"], "answer_arxiv_id": ["1807.04686"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_17973"} +{"question": "Which study parameterizes a graph using a particle variational inference method?", "answer": ["DiBS: Differentiable Bayesian Structure Learning"], "answer_arxiv_id": ["2105.11839"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_17974"} +{"question": "What are some studies that explore the connection between model generalization and loss landscape geometry?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data", "Fantastic Generalization Measures and Where to Find Them"], "answer_arxiv_id": ["1609.04836", "1703.11008", "1912.02178"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_17975"} +{"question": "Who evaluated the GPT-3.5 model on WebNLG?", "answer": ["Using Large Language Models for Zero-Shot Natural Language Generation\n from Knowledge Graphs"], "answer_arxiv_id": ["2307.07312"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_17976"} +{"question": "Which works have explored the use of sports data in natural language tasks?", "answer": ["Challenges in Data-to-Document Generation", "Data-to-text Generation with Entity Modeling", "Investigating Sports Commentator Bias within a Large Corpus of American\n Football Broadcasts"], "answer_arxiv_id": ["1707.08052", "1906.03221", "1909.03343"], "source_meta": {"published_time": "20240215"}, "qid": "AutoScholarQuery_train_17977"} +{"question": "What works are involved with identifying scenario-specific causal relations and outcomes?", "answer": ["SemEval-2010 Task 8: Multi-Way Classification of Semantic Relations Between Pairs of Nominals"], "answer_arxiv_id": ["1911.10422"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_17978"} +{"question": "What works conducted numerical experiments to validate their hypothesis about the relationship between optimal transport and probability flow?", "answer": ["Understanding DDPM Latent Codes Through Optimal Transport"], "answer_arxiv_id": ["2202.07477"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_17979"} +{"question": "What works are related to the use of Lp,q norm with q>1 associated with homogeneous measures of comparator smoothness in universal dynamic regret?", "answer": ["Online Forecasting of Total-Variation-bounded Sequences", "Optimal Dynamic Regret in Exp-Concave Online Learning"], "answer_arxiv_id": ["1906.03364", "2104.11824"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_17980"} +{"question": "Which studies proposed various methods to empower VLMs with multi-modal in-context learning (M-ICL) capabilities?", "answer": ["Visual Programming: Compositional visual reasoning without training", "ICL-D3IE: In-Context Learning with Diverse Demonstrations Updating for\n Document Information Extraction", "ICL-D3IE: In-Context Learning with Diverse Demonstrations Updating for\n Document Information Extraction"], "answer_arxiv_id": ["2211.11559", "2303.05063", "2303.05063"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_17981"} +{"question": "What are some classical works that studied the effect of data augmentation methods?", "answer": ["A Kernel Theory of Modern Data Augmentation", "On the Generalization Effects of Linear Transformations in Data Augmentation", "How Data Augmentation affects Optimization for Linear Regression", "Data Augmentation as Feature Manipulation", "The Benefits of Mixup for Feature Learning"], "answer_arxiv_id": ["1803.06084", "2005.00695", "2010.11171", "2203.01572", "2303.08433v1"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_17982"} +{"question": "Which studies have focused on instance-aware part segmentation in the same context?", "answer": ["Self-Correction for Human Parsing", "Devil in the Details: Towards Accurate Single and Multiple Human Parsing", "Renovating Parsing R-CNN for Accurate Multiple Human Parsing", "Parsing R-CNN for Instance-Level Human Analysis", "AIParsing: Anchor-free Instance-level Human Parsing", "Towards Open-World Segmentation of Parts"], "answer_arxiv_id": ["1910.09777", "1809.05996", "2009.09447", "1811.12596", "2207.06854", "2305.16804"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_17983"} +{"question": "What studies proposed communication methods to lessen the slow down effect from compression?", "answer": ["Decentralized Stochastic Optimization and Gossip Algorithms with Compressed Communication"], "answer_arxiv_id": ["1902.00340"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_17984"} +{"question": "What studies have been done on the use of simple motion models in visual navigation training policies?", "answer": ["Object Goal Navigation using Goal-Oriented Semantic Exploration", "Learning to Explore using Active Neural SLAM", "Navigating to Objects in the Real World", "Combining Optimal Control and Learning for Visual Navigation in Novel\n Environments", "ViNT: A Foundation Model for Visual Navigation", "Success Weighted by Completion Time: A Dynamics-Aware Evaluation\n Criteria for Embodied Navigation"], "answer_arxiv_id": ["2007.00643", "2004.05155", "2212.00922", "1903.02531", "2306.14846", "2103.08022"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_17985"} +{"question": "Which work proposed to use off-the-shelf language models like Roberta and BERT to align images to language models without text supervision?", "answer": ["RoBERTa: A Robustly Optimized BERT Pretraining Approach", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["1907.11692", "1810.04805"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_17986"} +{"question": "Could you provide me with studies that explored how transformers learn topics by studying co-occurrences of words in the training data?", "answer": ["How Do Transformers Learn Topic Structure: Towards a Mechanistic Understanding"], "answer_arxiv_id": ["2303.04245"], "source_meta": {"published_time": "20230726"}, "qid": "AutoScholarQuery_train_17987"} +{"question": "What studies utilized novel architectural innovations for fast adaptation?", "answer": ["Visual Prompt Tuning", "Exploring Visual Prompts for Adapting Large-Scale Models", "What Makes Good Examples for Visual In-Context Learning?", "Training BatchNorm and Only BatchNorm: On the Expressive Power of Random Features in CNNs", "Learning multiple visual domains with residual adapters"], "answer_arxiv_id": ["2203.12119", "2203.17274", "2301.13670", "2003.00152", "1705.08045"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_17988"} +{"question": "Which research works focus on Interpretable ML straddling machine learning, language, vision, and HCI?", "answer": ["Interpretable Machine Learning: Moving From Mythos to Diagnostics", "Explaining Explanations: An Overview of Interpretability of Machine Learning", "The Language Interpretability Tool: Extensible, Interactive Visualizations and Analysis for NLP Models", "Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps"], "answer_arxiv_id": ["2103.06254", "1806.00069", "2008.05122", "1312.6034"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_17989"} +{"question": "Any research that addressed the issue of LLMs struggling to provide accurate critiques by incorporating external information?", "answer": ["Re3: Generating Longer Stories With Recursive Reprompting and Revision", "Self-RAG: Learning to Retrieve, Generate, and Critique through\n Self-Reflection"], "answer_arxiv_id": ["2210.06774", "2310.11511"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_17990"} +{"question": "Any works aimed at improving cross-attention by adding additional spatial control?", "answer": ["GLIGEN: Open-Set Grounded Text-to-Image Generation"], "answer_arxiv_id": ["2301.07093"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_17991"} +{"question": "What papers propose the use of domain-specific knowledge for aligning texts and images?", "answer": ["ERNIE-ViL: Knowledge Enhanced Vision-Language Representations Through\n Scene Graph", "Align, Reason and Learn: Enhancing Medical Vision-and-Language\n Pre-training with Knowledge", "Medical Image Understanding with Pretrained Vision Language Models: A\n Comprehensive Study"], "answer_arxiv_id": ["2006.16934", "2209.07118", "2209.15517"], "source_meta": {"published_time": "20240203"}, "qid": "AutoScholarQuery_train_17992"} +{"question": "Which research studies have been conducted on protein structures in the context of protein representation learning?", "answer": ["Intrinsic-Extrinsic Convolution and Pooling for Learning on 3D Protein Structures", "Multi-Scale Representation Learning on Proteins"], "answer_arxiv_id": ["2007.06252", "2204.02337"], "source_meta": {"published_time": "20220311"}, "qid": "AutoScholarQuery_train_17993"} +{"question": "What is the paper that the A coding algorithm was generalised from?", "answer": ["Strong Functional Representation Lemma and Applications to Coding Theorems"], "answer_arxiv_id": ["1701.02827"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_17994"} +{"question": "Any studies extending the method of extracting memorized images from Stable Diffusion?", "answer": ["Extracting Training Data from Diffusion Models"], "answer_arxiv_id": ["2301.13188"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_17995"} +{"question": "What papers proposed the usage of powerful cross-modal algorithms for text-to-video generation?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240117"}, "qid": "AutoScholarQuery_train_17996"} +{"question": "What papers have made contributions to deep learning-based image registration?", "answer": ["Learn2Reg: comprehensive multi-task medical image registration\n challenge, dataset and evaluation in the era of deep learning", "Biomedical image analysis competitions: The state of current\n participation practice", "CoMIR: Contrastive Multimodal Image Representation for Registration"], "answer_arxiv_id": ["2112.04489", "2212.08568", "2006.06325"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_17997"} +{"question": "What studies cover the analysis of the generalization of deep linear GNNs with skip connections?", "answer": ["Optimization of Graph Neural Networks: Implicit Acceleration by Skip Connections and More Depth"], "answer_arxiv_id": ["2105.04550"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_17998"} +{"question": "Could you provide me some works about voxel-based methods for point clouds semantic segmentation?", "answer": ["RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR\n Point Cloud Segmentation", "Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR\n Segmentation", "Spherical Transformer for LiDAR-based 3D Recognition", "Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["2103.12978", "2011.10033", "2303.12766", "2007.16100", "1904.08755"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_17999"} +{"question": "What papers have studied the relationship between sharpness and generalization in DNNs?", "answer": ["On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima", "Three Factors Influencing Minima in SGD", "Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data", "Exploring Generalization in Deep Learning", "On the Generalization of Models Trained with SGD: Information-Theoretic Bounds and Implications"], "answer_arxiv_id": ["1609.04836", "1711.04623", "1703.11008", "1706.08947", "2110.03128"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_18000"} +{"question": "Which works are associated with the development of Q-learning algorithms that HER is designed for?", "answer": ["Deep Reinforcement Learning with Double Q-learning", "Playing Atari with Deep Reinforcement Learning"], "answer_arxiv_id": ["1509.06461", "1312.5602"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18001"} +{"question": "Which papers mentioned the effect of large models pre-trained on self-supervised tasks in NLP?", "answer": ["Deep contextualized word representations", "Universal Language Model Fine-tuning for Text Classification", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "RoBERTa: A Robustly Optimized BERT Pretraining Approach", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"], "answer_arxiv_id": ["1802.05365", "1801.06146", "1810.04805", "1907.11692", "1910.10683"], "source_meta": {"published_time": "20211216"}, "qid": "AutoScholarQuery_train_18002"} +{"question": "Which papers utilize saliency-based methods in their black-box explanations?", "answer": ["Learning Deep Features for Discriminative Localization", "Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization", "SmoothGrad: removing noise by adding noise", "Axiomatic Attribution for Deep Networks", "XRAI: Better Attributions Through Regions", "RISE: Randomized Input Sampling for Explanation of Black-box Models"], "answer_arxiv_id": ["1512.04150", "1610.02391", "1706.03825", "1703.01365", "1906.02825", "1806.07421"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_18003"} +{"question": "Could you tell me about some transformer-based models that capture global-level perspectives in neural architecture encoding?", "answer": ["CATE: Computation-aware Neural Architecture Encoding with Transformers", "NAR-Former: Neural Architecture Representation Learning towards Holistic\n Attributes Prediction"], "answer_arxiv_id": ["2102.07108", "2211.08024"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_18004"} +{"question": "What papers examined the finite-sample complexity bounds for RMDPs in finite-horizon setting?", "answer": ["Improved Sample Complexity Bounds for Distributionally Robust Reinforcement Learning", "Online Policy Optimization for Robust MDP"], "answer_arxiv_id": ["2303.02783v2", "2209.13841"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18005"} +{"question": "Are there any papers featuring an approach to learning data augmentations as an invariance learning problem?", "answer": ["Learning Invariances using the Marginal Likelihood", "Last Layer Marginal Likelihood for Invariance Learning", "Data augmentation in Bayesian neural networks and the cold posterior effect", "Invariance Learning in Deep Neural Networks with Differentiable Laplace Approximations"], "answer_arxiv_id": ["1808.05563", "2106.07512", "2106.05586v2", "2202.10638"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_18006"} +{"question": "Which work introduces LogDet divergence measure in SONew for efficient preconditioning process?", "answer": ["Efficient Second Order Online Learning by Sketching"], "answer_arxiv_id": ["1602.02202"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_18007"} +{"question": "Which works propose different methods to deal with the challenge of the sparse reward mechanism and generalization performance in RL-based navigation?", "answer": ["Curious Representation Learning for Embodied Intelligence", "Zero Experience Required: Plug & Play Modular Transfer Learning for Semantic Visual Navigation", "ZSON: Zero-Shot Object-Goal Navigation using Multimodal Goal Embeddings", "Learning Vision-and-Language Navigation from YouTube Videos"], "answer_arxiv_id": ["2105.01060", "2202.02440", "2206.12403", "2307.11984"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_18008"} +{"question": "Which papers discuss methods to downsample the K𝐾K and V𝑉V in self-attention while preserving the global receptive field?", "answer": ["Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions", "CvT: Introducing Convolutions to Vision Transformers", "EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers", "Shunted Self-Attention via Multi-Scale Token Aggregation", "Focal Self-attention for Local-Global Interactions in Vision Transformers", "Twins: Revisiting the Design of Spatial Attention in Vision Transformers"], "answer_arxiv_id": ["2102.12122", "2103.15808", "2205.03436", "2111.15193", "2107.00641", "2104.13840"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_18009"} +{"question": "Could you provide me with research where LLMs are prompted to generate both an answer and a level of confidence?", "answer": ["Teaching Models to Express Their Uncertainty in Words"], "answer_arxiv_id": ["2205.14334"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_18010"} +{"question": "Could you provide me some works that aim to make Bayesian Optimization robust?", "answer": ["Unscented Bayesian Optimization for Safe Robot Grasping", "Practical Bayesian optimization in the presence of outliers", "Bayesian optimisation under uncertain inputs", "No-Regret Learning in Unknown Games with Correlated Payoffs"], "answer_arxiv_id": ["1603.02038", "1712.04567v1", "1902.07908v1", "1909.08540"], "source_meta": {"published_time": "20220304"}, "qid": "AutoScholarQuery_train_18011"} +{"question": "What papers have proposed approaches for training dedicated retrievers in the context of HQA?", "answer": ["MuGER$^2$: Multi-Granularity Evidence Retrieval and Reasoning for Hybrid\n Question Answering", "S$^3$HQA: A Three-Stage Approach for Multi-hop Text-Table Hybrid\n Question Answering"], "answer_arxiv_id": ["2210.10350", "2305.11725"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_18012"} +{"question": "Which studies proposed traditional vision-based models for surgical instrument segmentation?", "answer": ["Automatic Instrument Segmentation in Robot-Assisted Surgery Using Deep Learning", "U-Net: Convolutional Networks for Biomedical Image Segmentation", "ISINet: An Instance-Based Approach for Surgical Instrument Segmentation", "Incorporating Temporal Prior from Motion Flow for Instrument Segmentation in Minimally Invasive Surgery Video"], "answer_arxiv_id": ["1803.01207", "1505.04597", "2007.05533", "1907.07899"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_18013"} +{"question": "Are there any works specialized in neural fields for different types of data such as climate data, audio, and medical images?", "answer": ["Compressing multidimensional weather and climate data into neural\n networks", "Siamese SIREN: Audio Compression with Implicit Neural Representations", "Lossy compression of multidimensional medical images using sinusoidal\n activation networks: an evaluation study", "COIN++: Neural Compression Across Modalities", "A Novel Implicit Neural Representation for Volume Data", "SINCO: A Novel structural regularizer for image compression using\n implicit neural representations"], "answer_arxiv_id": ["2210.12538", "2306.12957", "2208.01602", "2201.12904", "2403.08566v1", "2210.14974"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18014"} +{"question": "What works used dropout inference in Bayesian deep learning?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "Concrete Dropout"], "answer_arxiv_id": ["1506.02142", "1703.04977", "1705.07832"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_18015"} +{"question": "Can you mention research papers that discuss the development and utility of various datasets designed for egocentric video understanding?", "answer": ["Scaling Egocentric Vision: The EPIC-KITCHENS Dataset", "Actor and Observer: Joint Modeling of First and Third-Person Videos", "Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["1804.02748", "1804.09627", "2110.07058"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_18016"} +{"question": "What research work includes methods for injecting entity knowledge into pre-training by introducing named entity masking, hypernym linking, or using entity knowledge triples to construct knowledge-rich sentence trees?", "answer": ["ERNIE 2.0: A Continual Pre-Training Framework for Language Understanding", "Knowledge Enhanced Contextual Word Representations", "K-BERT: Enabling Language Representation with Knowledge Graph"], "answer_arxiv_id": ["1907.12412", "1909.04164", "1909.07606"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_18017"} +{"question": "Any works about the usage of stolen encoders for transfer attacks on the target encoder?", "answer": ["Stealing the Decoding Algorithms of Language Models", "Thieves on Sesame Street! Model Extraction of BERT-based APIs", "Model Extraction and Adversarial Transferability, Your BERT is\n Vulnerable!"], "answer_arxiv_id": ["2303.04729", "1910.12366", "2103.10013"], "source_meta": {"published_time": "20240612"}, "qid": "AutoScholarQuery_train_18018"} +{"question": "What studies discuss the exponential growth in the number of class label spaces as a challenge in multi-label learning?", "answer": ["The Emerging Trends of Multi-Label Learning"], "answer_arxiv_id": ["2011.11197"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_18019"} +{"question": "Which studies prove positive results for adaptive spiked kernels in low-dimensional kernel learning regime?", "answer": ["Does data interpolation contradict statistical optimality?"], "answer_arxiv_id": ["1806.09471"], "source_meta": {"published_time": "20230118"}, "qid": "AutoScholarQuery_train_18020"} +{"question": "What work conducted empirical evaluations on how the configuration of LLMs affect uncertainty?", "answer": ["Uncertainty Quantification with Pre-trained Language Models: A\n Large-Scale Empirical Analysis"], "answer_arxiv_id": ["2210.04714"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_18021"} +{"question": "What works talked about using variational auto-encoders (VAEs) or generative adversarial networks (GANs) to generate 3D structures of periodic materials?", "answer": ["Auto-Encoding Variational Bayes", "Generative Adversarial Nets"], "answer_arxiv_id": ["1312.6114", "1406.2661"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_18022"} +{"question": "Can you cite studies that discuss the equivariance assumption in neural network architectures?", "answer": ["Group Equivariant Convolutional Networks", "General $E(2)$-Equivariant Steerable CNNs", "Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["1602.07576", "1911.08251", "2104.13478v2"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18023"} +{"question": "Which studies focused on the use of Mask Transformers for image segmentation?", "answer": ["MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers", "Attention Is All You Need", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["2012.00759", "1706.03762", "2005.12872"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_18024"} +{"question": "Could you tell me about the studies that focus on the computational and memory requirements related to selective execution in large language models?", "answer": ["Deja Vu: Contextual Sparsity for Efficient LLMs at Inference Time", "ReLU Strikes Back: Exploiting Activation Sparsity in Large Language\n Models", "Adaptive Computation Time for Recurrent Neural Networks", "Alternating Updates for Efficient Transformers"], "answer_arxiv_id": ["2310.17157", "2310.04564", "1603.08983", "2301.13310"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_18025"} +{"question": "What studies discussed ensemble methods combined with Pretraining Models (PTMs)?", "answer": ["Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization", "Domain Generalization using Pretrained Models without Fine-tuning", "ZooD: Exploiting Model Zoo for Out-of-Distribution Generalization", "Diverse Weight Averaging for Out-of-Distribution Generalization", "Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time"], "answer_arxiv_id": ["2110.10832", "2203.04600", "2210.09236", "2205.09739", "2203.05482"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_18026"} +{"question": "Could you provide me the reference that develops the augmentation graph framework this paper builds upon?", "answer": ["Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss"], "answer_arxiv_id": ["2106.04156"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_18027"} +{"question": "What works discussed the blob method for reducing KL-divergence over the Wasserstein space?", "answer": ["A Blob Method For Diffusion"], "answer_arxiv_id": ["1709.09195"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_18028"} +{"question": "Could you mention some works on physically based face models which inherently have the ability to model anatomy constraints through simulation?", "answer": ["Implicit Neural Representation for Physics-driven Actuated Soft Bodies"], "answer_arxiv_id": ["2401.14861"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_18029"} +{"question": "Any works about prompt tuning as a parameter freezing strategy?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "WARP: Word-level Adversarial ReProgramming", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2101.00190", "2101.00121", "2104.08691", "2109.01652"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_18030"} +{"question": "What are some works related to the use of diffusion models in voice synthesis?", "answer": ["DiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism", "WaveGrad: Estimating Gradients for Waveform Generation", "WaveGrad 2: Iterative Refinement for Text-to-Speech Synthesis"], "answer_arxiv_id": ["2105.02446", "2009.00713v2", "2106.09660"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_18031"} +{"question": "What research proves that the Fourier Neural Operator can approximate any continuous operator?", "answer": ["On universal approximation and error bounds for Fourier Neural Operators"], "answer_arxiv_id": ["2107.07562"], "source_meta": {"published_time": "20211127"}, "qid": "AutoScholarQuery_train_18032"} +{"question": "Could you provide me some studies about video segmentation using the Segment Anything Model (SAM)?", "answer": ["Segment Anything Meets Point Tracking", "UVOSAM: A Mask-free Paradigm for Unsupervised Video Object Segmentation\n via Segment Anything Model"], "answer_arxiv_id": ["2307.01197", "2305.12659"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_18033"} +{"question": "Which papers have studied the social impacts of RLHF and Preference Tuning?", "answer": ["Perspectives on the Social Impacts of Reinforcement Learning with Human\n Feedback"], "answer_arxiv_id": ["2303.02891"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_18034"} +{"question": "Can you point me to any studies that tackle the challenge of capturing inter-token dependency in non-autoregressive sequence generation?", "answer": ["Fully Non-autoregressive Neural Machine Translation: Tricks of the Trade"], "answer_arxiv_id": ["2012.15833"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18035"} +{"question": "Which works address the semi-supervised learning problem through empirical efforts?", "answer": ["Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning", "Temporal Ensembling for Semi-Supervised Learning", "S4L: Self-Supervised Semi-Supervised Learning", "Semi-Supervised Learning with Scarce Annotations", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Multi-Task Curriculum Framework for Open-Set Semi-Supervised Learning", "OpenCoS: Contrastive Semi-supervised Learning for Handling Open-set Unlabeled Data", "OpenMatch: Open-set Consistency Regularization for Semi-supervised Learning with Outliers", "Trash to Treasure: Harvesting OOD Data with Cross-Modal Matching for Open-Set Semi-Supervised Learning", "Class-Aware Contrastive Semi-Supervised Learning"], "answer_arxiv_id": ["1606.04586", "1610.02242v3", "1905.03670", "1905.08845", "2001.07685v2", "2007.11330", "2107.08943", "2105.14148", "2108.05617", "2203.02261"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_18036"} +{"question": "What paper has applied DPPs for in-context learning in compositional tasks?", "answer": ["Diverse Demonstrations Improve In-context Compositional Generalization"], "answer_arxiv_id": ["2212.06800"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_18037"} +{"question": "Could you provide some examples of research papers that discuss the use of pre-trained models from other modalities for the purpose of multimodal video understanding?", "answer": ["ActionCLIP: A New Paradigm for Video Action Recognition", "MERLOT: Multimodal Neural Script Knowledge Models", "VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding"], "answer_arxiv_id": ["2109.08472", "2106.02636", "2109.14084"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_18038"} +{"question": "Any studies about augmenting Transformer-based models to have inductive biases for nested, hierarchical structures?", "answer": ["Context-Free Transductions with Neural Stacks", "Pushdown Layers: Encoding Recursive Structure in Transformer Language\n Models", "Injecting structural hints: Using language models to study inductive\n biases in language learning"], "answer_arxiv_id": ["1809.02836", "2310.19089", "2304.13060"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_18039"} +{"question": "What studies unify the label mapping of detection datasets manually?", "answer": ["Object Detection with a Unified Label Space from Multiple Datasets"], "answer_arxiv_id": ["2008.06614"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_18040"} +{"question": "Which work proposes a method for data augmentation in visual question answering tasks by applying back-translation on text and adversarial noise on images?", "answer": ["Semantic Equivalent Adversarial Data Augmentation for Visual Question Answering"], "answer_arxiv_id": ["2007.09592"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_18041"} +{"question": "What works introduced instruction tuning to enhance the zero-shot capabilities of Large Language Models?", "answer": ["Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2109.01652"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_18042"} +{"question": "Any works showed that the 'SUBG' method can be an effective approach to learning worst-group-robust predictors?", "answer": ["Simple data balancing achieves competitive worst-group-accuracy"], "answer_arxiv_id": ["2110.14503"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_18043"} +{"question": "What works analyzed algorithms for convex-concave problems in the Federated Learning (FL) setting?", "answer": ["Distributionally Robust Federated Averaging", "Efficient Algorithms for Federated Saddle Point Optimization", "A Communication-efficient Algorithm with Linear Convergence for Federated Minimax Learning"], "answer_arxiv_id": ["2102.12660", "2102.06333", "2206.01132"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_18044"} +{"question": "What are the papers about the sampling methods for diffusion models?", "answer": ["Denoising Diffusion Implicit Models", "Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models", "Fast Sampling of Diffusion Models with Exponential Integrator", "gDDIM: Generalized denoising diffusion implicit models", "Progressive Distillation for Fast Sampling of Diffusion Models", "On Distillation of Guided Diffusion Models"], "answer_arxiv_id": ["2010.02502", "2201.06503", "2206.00927", "2211.01095", "2204.13902", "2206.05564", "2202.00512", "2210.03142"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_18045"} +{"question": "Can you provide reference about the role of positive and negative pairs' selection in the success of contrastive learning?", "answer": ["Context Encoders: Feature Learning by Inpainting"], "answer_arxiv_id": ["1604.07379"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_18046"} +{"question": "Could you provide me some references conducted on unstructured pruning in network pruning research?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding"], "answer_arxiv_id": ["1510.00149"], "source_meta": {"published_time": "20220725"}, "qid": "AutoScholarQuery_train_18047"} +{"question": "What works introduced new larger datasets or made larger datasets more usable in the field of machine learning fairness?", "answer": ["Retiring Adult: New Datasets for Fair Machine Learning"], "answer_arxiv_id": ["2108.04884"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_18048"} +{"question": "What are some previous studies on techniques for learning search indices?", "answer": ["The Case for Learned Index Structures"], "answer_arxiv_id": ["1712.01208"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_18049"} +{"question": "Which studies benefit from a pretrained text model in sign language translation?", "answer": ["YouTube-ASL: A Large-Scale, Open-Domain American Sign Language-English\n Parallel Corpus"], "answer_arxiv_id": ["2306.15162"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_18050"} +{"question": "Which works enhance the memory of the detectors by introducing recurrent modules?", "answer": ["Convolutional LSTM Network: A Machine Learning Approach for\n Precipitation Nowcasting"], "answer_arxiv_id": ["1506.04214"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18051"} +{"question": "What papers detail the replacement of certain modules such as motion compensation, transform coding, and entropy coding with powerful learning-based models for video compression?", "answer": ["Learning for Video Compression with Hierarchical Quality and Recurrent Enhancement", "Variational image compression with a scale hyperprior", "VCT: A Video Compression Transformer"], "answer_arxiv_id": ["2003.01966", "1802.01436v2", "2206.07307"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_18052"} +{"question": "Are there any works on the impact of demographic differences in raters on the detection of hate speech?", "answer": ["Is Your Toxicity My Toxicity? Exploring the Impact of Rater Identity on Toxicity Annotation"], "answer_arxiv_id": ["2205.00501"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_18053"} +{"question": "Which prior works use methods based on approximate dynamic programming in the context of offline RL for NLP?", "answer": ["Human-centric dialog training via offline reinforcement learning", "Sequence Tutor: Conservative Fine-Tuning of Sequence Generation Models with KL-control", "CHAI: A CHatbot AI for Task-Oriented Dialogue with Offline Reinforcement Learning"], "answer_arxiv_id": ["2010.05848", "1611.02796v9", "2204.08426"], "source_meta": {"published_time": "20220605"}, "qid": "AutoScholarQuery_train_18054"} +{"question": "What works proposed neural radiance fields derived from event streams?", "answer": ["Ev-NeRF: Event Based Neural Radiance Field", "EventNeRF: Neural Radiance Fields from a Single Colour Event Camera"], "answer_arxiv_id": ["2206.12455", "2206.11896"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_18055"} +{"question": "Which publications detail efforts to speed up training in NeRFs?", "answer": ["Depth-supervised NeRF: Fewer Views and Faster Training for Free", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks", "TensoRF: Tensorial Radiance Fields", "BakedSDF: Meshing Neural SDFs for Real-Time View Synthesis"], "answer_arxiv_id": ["2107.02791", "2111.11215", "2201.05989", "2112.05131", "2203.09517", "2302.14859"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_18056"} +{"question": "What paper surveyed recent efforts in data-centric benchmarking?", "answer": ["Data-centric Artificial Intelligence: A Survey"], "answer_arxiv_id": ["2303.10158"], "source_meta": {"published_time": "20220720"}, "qid": "AutoScholarQuery_train_18057"} +{"question": "What studies provided methods on subpopulation shift setting?", "answer": ["Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "What is the Effect of Importance Weighting in Deep Learning?", "An investigation of why overparameterization exacerbates spurious correlations", "Simple data balancing achieves competitive worst-group-accuracy"], "answer_arxiv_id": ["1610.03425", "1911.08731", "1812.03372", "2005.04345", "2110.14503"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_18058"} +{"question": "What studies have proposed storing view-direction-independent information on voxel grids to accelerate the training speed of NeRF?", "answer": ["Neural Sparse Voxel Fields", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Baking Neural Radiance Fields for Real-Time View Synthesis"], "answer_arxiv_id": ["2007.11571", "2103.14024", "2103.14645"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_18059"} +{"question": "Any works about the use of RLHF in guiding RL policy optimization?", "answer": ["Deep Reinforcement Learning from Human Preferences", "Fine-Tuning Language Models from Human Preferences", "Interactive Learning from Policy-Dependent Human Feedback"], "answer_arxiv_id": ["1706.03741", "1909.08593", "1701.06049"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_18060"} +{"question": "Who developed methods on transferring style based on temporal consistency in video sequences?", "answer": ["Interactive Video Stylization Using Few-Shot Patch-Based Training", "Coherent Online Video Style Transfer"], "answer_arxiv_id": ["2004.14489", "1703.09211"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_18061"} +{"question": "Which research papers focus on the issue that LLMs struggle to rectify their initial mistakes and worsen their performance after self-correction?", "answer": ["Large Language Models Cannot Self-Correct Reasoning Yet", "LLMs cannot find reasoning errors, but can correct them given the error\n location", "CritiqueLLM: Towards an Informative Critique Generation Model for\n Evaluation of Large Language Model Generation"], "answer_arxiv_id": ["2310.01798", "2311.08516", "2311.18702"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_18062"} +{"question": "What followup works focused on improving the speed or the generalization ability of NeRF?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Nerfies: Deformable Neural Radiance Fields"], "answer_arxiv_id": ["2012.02190", "2103.14024", "2011.12948"], "source_meta": {"published_time": "20220512"}, "qid": "AutoScholarQuery_train_18063"} +{"question": "Which work proposed a series of methods that has been inspired by FixMatch?", "answer": ["CoMatch: Semi-supervised Learning with Contrastive Graph Regularization", "In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "All Labels Are Not Created Equal: Enhancing Semi-supervision via Label Grouping and Co-training", "Meta Pseudo Labels", "SimPLE: Similar Pseudo Label Exploitation for Semi-Supervised Classification"], "answer_arxiv_id": ["2011.11183", "2101.06329", "2110.08263", "2104.05248", "2003.10580", "2103.16725"], "source_meta": {"published_time": "20230805"}, "qid": "AutoScholarQuery_train_18064"} +{"question": "Which works exemplify the use of edge-independent models in probabilistic models for large networks?", "answer": ["Variational Graph Auto-Encoders", "Stochastic Blockmodels meet Graph Neural Networks", "Dirichlet Graph Variational Autoencoder", "Interpretable Node Representation with Attribute Decoding", "NetGAN: Generating Graphs via Random Walks"], "answer_arxiv_id": ["1611.07308", "1905.05738", "2010.04408", "2212.01682", "1803.00816"], "source_meta": {"published_time": "20230506"}, "qid": "AutoScholarQuery_train_18065"} +{"question": "What works are about processing 3D motion through recurrent network?", "answer": ["RAFT: Recurrent All-Pairs Field Transforms for Optical Flow"], "answer_arxiv_id": ["2003.12039"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_18066"} +{"question": "Which works specifically involve training neural fields with coarse-to-fine LODs through adaptive data structures like octrees?", "answer": ["Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D\n Shapes"], "answer_arxiv_id": ["2101.10994"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_18067"} +{"question": "Are there any studies about crowdsourced labeling?", "answer": ["Budget-Optimal Task Allocation for Reliable Crowdsourcing Systems"], "answer_arxiv_id": ["1110.3564"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_18068"} +{"question": "Can you mention some studies that address the inflexibility of diffusion models conditioned on sensor measurements to structural changes?", "answer": ["EDGE: Editable Dance Generation From Music", "Human Motion Diffusion Model"], "answer_arxiv_id": ["2211.10658", "2209.14916"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_18069"} +{"question": "Which studies applied the self-attention computation pattern, normally used in natural language processing, to point clouds?", "answer": ["Point Transformer"], "answer_arxiv_id": ["2012.09164"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_18070"} +{"question": "Which works propose the use of PEFT in order to reduce the number of trainable parameters when fine-tuning the entire network?", "answer": ["Visual Prompt Tuning", "LoRA: Low-Rank Adaptation of Large Language Models", "Efficient Adaptation of Large Vision Transformer via Adapter\n Re-Composing", "Scaling & Shifting Your Features: A New Baseline for Efficient Model\n Tuning"], "answer_arxiv_id": ["2203.12119", "2106.09685", "2310.06234", "2210.08823"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_18071"} +{"question": "What studies on attention mechanisms have been successfully adopted in computer vision tasks?", "answer": ["End-to-End Object Detection with Transformers", "An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Segmenter: Transformer for Semantic Segmentation", "PETR: Position Embedding Transformation for Multi-View 3D Object\n Detection"], "answer_arxiv_id": ["2005.12872", "2010.11929", "2105.05633", "2203.05625"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_18072"} +{"question": "Any works about parallel tempering?", "answer": ["Parallel tempering on optimized paths"], "answer_arxiv_id": ["2102.07720"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_18073"} +{"question": "What was the key idea behind the work of bib.bib50 in the context of iterative learning?", "answer": ["Learning to Score Behaviors for Guided Policy Optimization"], "answer_arxiv_id": ["1906.04349"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_18074"} +{"question": "Could you provide me some studies that provided theoretical understanding on contrastive learning?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere", "Contrastive estimation reveals topic posterior information to linear models", "Understanding Self-supervised Learning with Dual Deep Networks", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Toward Understanding the Feature Learning Process of Self-supervised Contrastive Learning", "Contrastive Learning Inverts the Data Generating Process"], "answer_arxiv_id": ["1902.09229v1", "2005.10242", "2003.02234", "2010.00578", "2106.04156", "2105.15134", "2102.08850v4"], "source_meta": {"published_time": "20220202"}, "qid": "AutoScholarQuery_train_18075"} +{"question": "Which work designed a method called ECON specifically for clothed human recovery from images?", "answer": ["ECON: Explicit Clothed humans Optimized via Normal integration"], "answer_arxiv_id": ["2212.07422"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_18076"} +{"question": "Which works focused on training models on multiple tasks?", "answer": ["Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-Scale Convolutional Architecture", "UberNet : Training a ‘Universal’ Convolutional Neural Network for Low-, Mid-, and High-Level Vision using Diverse Datasets and Limited Memory", "Multi-Task Self-Training for Learning General Representations", "A Generalist Agent", "MulT: An End-to-End Multitask Learning Transformer"], "answer_arxiv_id": ["1411.4734", "1609.02132", "2108.11353", "2205.06175", "2205.08303"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_18077"} +{"question": "Which works have used concepts to build interpretable models?", "answer": ["Concept Bottleneck Models", "Post-hoc Concept Bottleneck Models"], "answer_arxiv_id": ["2007.04612", "2205.15480"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_18078"} +{"question": "Could you provide me some works about handling complex logical structures in reasoning?", "answer": ["Query2box: Reasoning over Knowledge Graphs in Vector Space using Box\n Embeddings", "Benchmarking the Combinatorial Generalizability of Complex Query\n Answering on Knowledge Graphs", "Logical Message Passing Networks with One-hop Inference on Atomic\n Formulas", "Complex Query Answering on Eventuality Knowledge Graph with Implicit\n Logical Constraints"], "answer_arxiv_id": ["2002.05969", "2109.08925", "2301.08859", "2305.19068"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_18079"} +{"question": "Could you provide me some studies about side observations settings in stochastic, adversarial, graph feedback, and cascading feedback approaches?", "answer": ["Leveraging Side Observations in Stochastic Bandits", "From Bandits to Experts: On the Value of Side-Observations", "Online Learning with Feedback Graphs: Beyond Bandits", "Online Learning with Gaussian Payoffs and Side Observations", "Nonstochastic Multi-Armed Bandits with Graph-Structured Feedback", "Online Algorithm for Unsupervised Sensor Selection", "Thompson Sampling for Unsupervised Sequential Selection", "Online Algorithm for Unsupervised Sequential Selection with Contextual Information"], "answer_arxiv_id": ["1210.4839v1", "1106.2436", "1502.07617", "1510.08108", "1409.8428v1", "1901.04676v2", "2009.07554", "2010.12353"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_18080"} +{"question": "Could you provide me any studies that focuses on determining when and why 'grokking' arises?", "answer": ["Towards Understanding Grokking: An Effective Theory of Representation Learning", "A Tale of Two Circuits: Grokking as Competition of Sparse and Dense Subnetworks", "Explaining grokking through circuit efficiency"], "answer_arxiv_id": ["2205.10343", "2303.11873", "2309.02390"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18081"} +{"question": "Could you provide me some works that implement rationales for facilitating commonsense and social reasoning?", "answer": ["STaR: Bootstrapping Reasoning With Reasoning", "Knowledge-Grounded Self-Rationalization via Extractive and Natural\n Language Explanations"], "answer_arxiv_id": ["2203.14465", "2106.13876"], "source_meta": {"published_time": "20240627"}, "qid": "AutoScholarQuery_train_18082"} +{"question": "Could you provide examples of the work done on the generative methods in pre-training user models?", "answer": ["Parameter-Efficient Transfer from Sequential Behaviors for User Modeling and Recommendation", "PTUM: Pre-training User Model from Unlabeled User Behaviors via Self-supervision"], "answer_arxiv_id": ["2001.04253", "2010.01494"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_18083"} +{"question": "What studies discuss advanced techniques for long-tailed recognition, such as class-conditional sharpness-aware minimization, feature clusters compression, and global-local mixture consistency cumulative learning?", "answer": ["Global and Local Mixture Consistency Cumulative Learning for Long-tailed\n Visual Recognitions"], "answer_arxiv_id": ["2305.08661"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_18084"} +{"question": "Could you give examples of papers that use Mini-batch SGD with small batches for distributed optimization?", "answer": ["Minibatch vs Local SGD for Heterogeneous Distributed Learning", "Is Local SGD Better than Minibatch SGD?"], "answer_arxiv_id": ["2006.04735", "2002.07839"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_18085"} +{"question": "Could you provide me with information about research on DiffusionCLIP, a method used in image-to-image translation?", "answer": ["DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation", "StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators", "Learning Transferable Visual Models From Natural Language Supervision", "ArcFace: Additive Angular Margin Loss for Deep Face Recognition"], "answer_arxiv_id": ["2110.02711", "2108.00946", "2103.00020", "1801.07698"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18086"} +{"question": "What studies employ a Q-Former to link the frozen LLM and vision encoder?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models"], "answer_arxiv_id": ["2301.12597"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_18087"} +{"question": "Which papers are related to the field of continual learning, which deals with mitigating catastrophic forgetting?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Gradient Episodic Memory for Continual Learning"], "answer_arxiv_id": ["1612.00796", "1706.08840"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_18088"} +{"question": "Which work is the most similar to our research about the continuous-time discrete diffusion?", "answer": ["A Continuous Time Framework for Discrete Denoising Models"], "answer_arxiv_id": ["2205.14987"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_18089"} +{"question": "Which works employed transformer-based architectures for novel view synthesis?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering", "Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction"], "answer_arxiv_id": ["2102.13090", "2109.00512v1"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_18090"} +{"question": "Can you provide an example of studies that focus on prompt-based learning in the field of few-shot learning?", "answer": ["Language Models as Knowledge Bases?", "Language Models are Few-Shot Learners", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "Making Pre-trained Language Models Better Few-shot Learners", "Differentiable Prompt Makes Pre-trained Language Models Better Few-shot Learners", "The Power of Scale for Parameter-Efficient Prompt Tuning", "GPT Understands, Too"], "answer_arxiv_id": ["1909.01066", "2005.14165", "2107.13586", "2012.15723", "2108.13161", "2104.08691", "2103.10385"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_18091"} +{"question": "What research discusses the effect of distribution shifts on pre-processing methods?", "answer": ["Impossibility results for fair representations", "Diagnosing failures of fairness transfer across distribution shift in real-world medical settings"], "answer_arxiv_id": ["2107.03483", "2202.01034"], "source_meta": {"published_time": "20201201"}, "qid": "AutoScholarQuery_train_18092"} +{"question": "Can you provide research papers where they proposed efficient attention operations?", "answer": ["Scaling Laws vs Model Architectures: How does Inductive Bias Influence\n Scaling?", "A Practical Survey on Faster and Lighter Transformers"], "answer_arxiv_id": ["2207.10551", "2103.14636"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_18093"} +{"question": "Could you provide me some research papers about self-supervised methods that rely on contrastive augmentations?", "answer": ["Self-Supervised Representation Learning: Introduction, Advances and Challenges"], "answer_arxiv_id": ["2110.09327"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_18094"} +{"question": "Which work implemented Neural Implicit Functions for modeling and reconstructing 3D shapes?", "answer": ["Neural Fields in Visual Computing and Beyond"], "answer_arxiv_id": ["2111.11426"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_18095"} +{"question": "Which researches used large language models (LLMs) like the closed source GPT series and the opensource LLaMA?", "answer": ["GPT-4 Technical Report", "LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2303.08774", "2302.13971", "2307.09288"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_18096"} +{"question": "What studies replace the convolutions in the high-resolution ConvNet backbone with self-attention blocks?", "answer": ["Multi-Scale High-Resolution Vision Transformer for Semantic Segmentation"], "answer_arxiv_id": ["2111.01236"], "source_meta": {"published_time": "20221213"}, "qid": "AutoScholarQuery_train_18097"} +{"question": "Which studies approximate a symbolic policy using genetic programming?", "answer": ["Interpretable Policies for Reinforcement Learning by Genetic Programming"], "answer_arxiv_id": ["1712.04170"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_18098"} +{"question": "Which research papers proposed the use of multi-layer perceptrons to learn occupancy probability or signed distance functions in neural fields?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "Learning Implicit Fields for Generative Shape Modeling", "DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation", "SAL: Sign Agnostic Learning of Shapes from Raw Data"], "answer_arxiv_id": ["1812.03828", "1812.02822", "1901.05103", "1911.10414"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_18099"} +{"question": "Which papers proposed the use of large kernels in convolutional networks?", "answer": ["Going Deeper with Convolutions"], "answer_arxiv_id": ["1409.4842"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_18100"} +{"question": "Could you specify some studies that showed common hybrid models underperform compared to RNNs on POMDPs?", "answer": ["POPGym: Benchmarking Partially Observable Reinforcement Learning"], "answer_arxiv_id": ["2303.01859"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_18101"} +{"question": "Which works explored the concept of joint motion prediction?", "answer": ["THOMAS: Trajectory Heatmap Output with learned Multi-Agent Sampling", "Latent Variable Sequential Set Transformers for Joint Multi-Agent Motion Prediction", "Heterogeneous Edge-Enhanced Graph Attention Network For Multi-Agent Trajectory Prediction", "M2I: From Factored Marginal Trajectory Prediction to Interactive Prediction"], "answer_arxiv_id": ["2110.06607", "2104.00563", "2106.07161", "2202.11884"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_18102"} +{"question": "What works proposed methods for vertical federated learning to mitigate the straggler problem, specifically those that allow asynchronous data collection and model updating?", "answer": ["VAFL: a Method of Vertical Asynchronous Federated Learning", "Secure Bilevel Asynchronous Vertical Federated Learning with Backward Updating", "FDML: A Collaborative Machine Learning Framework for Distributed Features"], "answer_arxiv_id": ["2007.06081", "2103.00958", "1812.06415"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_18103"} +{"question": "What studies discuss the issue of training SNN models over simulation time steps by using backpropagation?", "answer": ["Training Deep Spiking Neural Networks using Backpropagation", "SLAYER: Spike Layer Error Reassignment in Time"], "answer_arxiv_id": ["1608.08782", "1810.08646"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_18104"} +{"question": "Which research introduced pre-trained code language models that showed high performance in translating natural language to code?", "answer": ["Evaluating Large Language Models Trained on Code", "InCoder: A Generative Model for Code Infilling and Synthesis"], "answer_arxiv_id": ["2107.03374", "2204.05999"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_18105"} +{"question": "What research introduced Sharpness-Aware Minimization (SAM) that improves model generalization by simultaneously minimizing the loss value and sharpness?", "answer": ["Sharpness-Aware Minimization for Efficiently Improving Generalization"], "answer_arxiv_id": ["2010.01412"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18106"} +{"question": "What studies applied foundational language models into vision-language tasks and mathematical problem-solving?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Learning to Prompt for Vision-Language Models", "CogView: Mastering Text-to-Image Generation via Transformers", "Automatic Generation of Socratic Subquestions for Teaching Math Word\n Problems"], "answer_arxiv_id": ["2103.00020", "2109.01134", "2105.13290", "2211.12835"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_18107"} +{"question": "Which work introduced generated queries and a novel auto-regressive decoder architecture to improve the DSI performance?", "answer": ["A Neural Corpus Indexer for Document Retrieval"], "answer_arxiv_id": ["2206.02743"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_18108"} +{"question": "Which research presents the EGNN model that learns equivariant features by integrating directional distance vectors between atom pairs?", "answer": ["E(n) Equivariant Graph Neural Networks"], "answer_arxiv_id": ["2102.09844"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_18109"} +{"question": "What works proposed one-stage methods that directly predict the target bounding box in referring image detection?", "answer": ["X-DETR: A Versatile Architecture for Instance-wise Vision-Language Tasks", "Referring Transformer: A One-step Approach to Multi-task Visual\n Grounding", "A Real-Time Cross-modality Correlation Filtering Method for Referring\n Expression Comprehension", "A Fast and Accurate One-Stage Approach to Visual Grounding", "SeqTR: A Simple yet Universal Network for Visual Grounding"], "answer_arxiv_id": ["2204.05626", "2106.03089", "1909.07072", "1908.06354", "2203.16265"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_18110"} +{"question": "What works focused on maximizing the state visitation distribution’s entropy?", "answer": ["Reinforcement Learning with Prototypical Representations", "APS: Active Pretraining with Successor Features", "Behavior From the Void: Unsupervised Active Pre-Training"], "answer_arxiv_id": ["2102.11271v2", "2108.13956", "2103.04551"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18111"} +{"question": "Any works about extending the traditional Marching Cubes, Poisson surface reconstruction, and Dual Contouring using neural networks?", "answer": ["Neural Marching Cubes", "Shape As Points: A Differentiable Poisson Solver", "Neural Dual Contouring"], "answer_arxiv_id": ["2106.11272", "2106.03452", "2202.01999"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_18112"} +{"question": "Which paper transferred each reasoning program into several sub-questions and QDG to evaluate the compositional consistency of existing VidQA methods?", "answer": ["Measuring Compositional Consistency for Video Question Answering"], "answer_arxiv_id": ["2204.07190"], "source_meta": {"published_time": "20240703"}, "qid": "AutoScholarQuery_train_18113"} +{"question": "What paper proposes the concept of Onestep RL for policy evaluation on in-distribution data?", "answer": ["Offline RL Without Off-Policy Evaluation"], "answer_arxiv_id": ["2106.08909"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_18114"} +{"question": "Which papers have proposed CNN-based and transformer-based methods for View-homogeneous ReID?", "answer": ["Deep Learning for Person Re-identification: A Survey and Outlook", "Beyond Part Models: Person Retrieval with Refined Part Pooling (and a\n Strong Convolutional Baseline)", "TransReID: Transformer-based Object Re-Identification"], "answer_arxiv_id": ["2001.04193", "1711.09349", "2102.04378"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_18115"} +{"question": "Which research papers proposed generating synthetic data to address data availability issues in anti-money laundering research?", "answer": ["Learning Deep Object Detectors from 3D Models", "The ParallelEye Dataset: Constructing Large-Scale Artificial Scenes for Traffic Vision Research"], "answer_arxiv_id": ["1412.7122", "1712.08394"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_18116"} +{"question": "What works discuss the drawback of Masked Image Modeling (MIM) in terms of its poor linear separability and data-efficiency in few-shot scenarios?", "answer": ["Masked Siamese Networks for Label-Efficient Learning"], "answer_arxiv_id": ["2204.07141"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18117"} +{"question": "Could you give me an example of a study that employed a specific tool to generate reasoning chains for each question in NLP tasks?", "answer": ["ExpeL: LLM Agents Are Experiential Learners"], "answer_arxiv_id": ["2308.10144"], "source_meta": {"published_time": "20240712"}, "qid": "AutoScholarQuery_train_18118"} +{"question": "What works show the effectiveness of using the attention mechanism for target-specific context representations?", "answer": ["Neural Processes with Stochastic Attention: Paying more attention to the context dataset", "Latent Bottlenecked Attentive Neural Processes", "Transformer Neural Processes: Uncertainty-Aware Meta Learning Via Sequence Modeling", "Robustifying Sequential Neural Processes", "NPCL: Neural Processes for Uncertainty-Aware Continual Learning"], "answer_arxiv_id": ["2204.05449", "2211.08458", "2207.04179", "2006.15987v1", "2310.19272"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_18119"} +{"question": "What works research the vulnerability of models to over-reliance on texture?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness"], "answer_arxiv_id": ["1811.12231"], "source_meta": {"published_time": "20220629"}, "qid": "AutoScholarQuery_train_18120"} +{"question": "What papers have studied norm concentration for subgamma and subgaussian random vectors?", "answer": ["A tail inequality for quadratic forms of subgaussian random vectors"], "answer_arxiv_id": ["1110.2842"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_18121"} +{"question": "Any studies have suggested a locally discriminative learning framework?", "answer": ["Details or Artifacts: A Locally Discriminative Learning Approach to\n Realistic Image Super-Resolution"], "answer_arxiv_id": ["2203.09195"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18122"} +{"question": "What are the works that have contributed to generating audio-description from the video?", "answer": ["AutoAD: Movie Description in Context", "AutoAD II: The Sequel -- Who, When, and What in Movie Audio Description"], "answer_arxiv_id": ["2303.16899", "2310.06838"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_18123"} +{"question": "What research compared gradient flow and gradient descent trajectories?", "answer": ["Implicit Gradient Regularization"], "answer_arxiv_id": ["2009.11162"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_18124"} +{"question": "Which work uses preference learning to align a student model to user intent by using preferences derived from different candidate models from a large teacher, concurrent to your work?", "answer": ["Zephyr: Direct Distillation of LM Alignment"], "answer_arxiv_id": ["2310.16944"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_18125"} +{"question": "Which works use the linear programming (LP) formulation of Reinforcement Learning (RL) to derive DICE algorithms for policy evaluation?", "answer": ["DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections", "AlgaeDICE: Policy Gradient from Arbitrary Experience", "Reinforcement Learning via Fenchel-Rockafellar Duality"], "answer_arxiv_id": ["1906.04733", "1912.02074", "2001.01866"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_18126"} +{"question": "Which papers use a process resembling gradient optimization for prompt optimization?", "answer": ["Automatic Prompt Optimization with \"Gradient Descent\" and Beam Search"], "answer_arxiv_id": ["2305.03495"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_18127"} +{"question": "Can you give some works that incorporate distances measured by LiDAR as auxiliary information in their methods?", "answer": ["Neural Fields meet Explicit Geometric Representation for Inverse\n Rendering of Urban Scenes"], "answer_arxiv_id": ["2304.03266"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_18128"} +{"question": "Which papers have shown that gradient-based explanations are insensitive to the randomization of model parameters?", "answer": ["Sanity Checks for Saliency Maps"], "answer_arxiv_id": ["1810.03292v3"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_18129"} +{"question": "Which work utilized Graph Convolution Networks in predicting polygonal boundaries?", "answer": ["Fast Interactive Object Annotation with Curve-GCN"], "answer_arxiv_id": ["1903.06874v1"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_18130"} +{"question": "Could you tell me which papers demonstrated that LLMs like GPT-4 can process long sequence length of up to 32,768 tokens?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18131"} +{"question": "What papers mention the use of cross-lingual transfer, where the skills from one language are transferred to another, to enhance capabilities in languages with scarce resources?", "answer": ["Do Multilingual Language Models Think Better in English?", "Not All Languages Are Created Equal in LLMs: Improving Multilingual\n Capability by Cross-Lingual-Thought Prompting", "Empowering Multi-step Reasoning across Languages via Tree-of-Thoughts"], "answer_arxiv_id": ["2308.01223", "2305.07004", "2311.08097"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18132"} +{"question": "Can you name studies where KD has been applied in action recognition?", "answer": ["Modality Distillation with Multiple Stream Networks for Action Recognition", "Graph Distillation for Action Detection with Privileged Modalities", "CROSS-MODAL KNOWLEDGE DISTILLATION FOR ACTION RECOGNITION"], "answer_arxiv_id": ["1806.07110", "1712.00108", "1910.04641"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_18133"} +{"question": "Which papers studied reward poisoning on vanilla bandit algorithms like UCB and epsilon-greedy?", "answer": ["Adversarial Attacks on Stochastic Bandits", "Near Optimal Adversarial Attack on UCB Bandits", "Data Poisoning Attacks in Contextual Bandits", "Data Poisoning Attacks on Stochastic Bandits", "When Are Linear Stochastic Bandits Attackable?"], "answer_arxiv_id": ["1810.12188", "2008.09312", "1808.05760", "1905.06494", "2110.09008"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_18134"} +{"question": "Any works about the cross-domain few-shot classification method involving the use of linear layers to align the features from different domains?", "answer": ["Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation"], "answer_arxiv_id": ["2001.08735"], "source_meta": {"published_time": "20231104"}, "qid": "AutoScholarQuery_train_18135"} +{"question": "What is the study that leverages the low-rank decomposition concept for PEFT methods?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20240113"}, "qid": "AutoScholarQuery_train_18136"} +{"question": "What are some examples of studies that apply DMs in text-guided image editing?", "answer": ["Custom-Edit: Text-Guided Image Editing with Customized Diffusion Models", "SINE: SINgle Image Editing with Text-to-Image Diffusion Models", "Paint by Example: Exemplar-based Image Editing with Diffusion Models", "Null-text Inversion for Editing Real Images using Guided Diffusion\n Models"], "answer_arxiv_id": ["2305.15779", "2212.04489", "2211.13227", "2211.09794"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_18137"} +{"question": "What papers applied the Maximum Mutual Information objective to code generating LLMs for reranking?", "answer": ["Coder Reviewer Reranking for Code Generation"], "answer_arxiv_id": ["2211.16490"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_18138"} +{"question": "Which papers are foundational works for the diffusion models?", "answer": ["Diffusion Models: A Comprehensive Survey of Methods and Applications", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2209.00796", "1503.03585", "1907.05600", "2006.11239"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_18139"} +{"question": "Which research papers have addressed the issue of symmetric objects in seen object pose estimation?", "answer": ["Explaining the Ambiguity of Object Detection and 6D Pose From Visual\n Data", "DPOD: 6D Pose Object Detector and Refiner", "GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D\n Object Pose Estimation"], "answer_arxiv_id": ["1812.00287", "1902.11020", "2102.12145"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_18140"} +{"question": "What benchmarks are used for comprehensive evaluation of self-supervised learning models?", "answer": ["SUPERB: Speech processing Universal PERformance Benchmark", "HEAR: Holistic Evaluation of Audio Representations"], "answer_arxiv_id": ["2105.01051", "2203.03022"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_18141"} +{"question": "Which studies have focused on point cloud completion?", "answer": ["Point Cloud Completion by Skip-attention Network with Hierarchical\n Folding", "PF-Net: Point Fractal Network for 3D Point Cloud Completion", "Morphing and Sampling Network for Dense Point Cloud Completion"], "answer_arxiv_id": ["2005.03871", "2003.00410", "1912.00280"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18142"} +{"question": "Can you point me to any recent works on representation learning in Markov games?", "answer": ["Representation Learning for General-sum Low-rank Markov Games"], "answer_arxiv_id": ["2210.16976"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_18143"} +{"question": "Which papers investigated the effect of monetary incentives on the predictions in crowdsourcing?", "answer": ["Incentivizing High Quality Crowdwork", "Adaptive Contract Design for Crowdsourcing Markets: Bandit Algorithms for Repeated Principal-Agent Problems", "Double or Nothing: Multiplicative Incentive Mechanisms for Crowdsourcing"], "answer_arxiv_id": ["1503.05897", "1405.2875v2", "1408.1387"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_18144"} +{"question": "Which works attempted to study text-to-image synthesis using GANs?", "answer": ["Generative Adversarial Text to Image Synthesis", "Learning What and Where to Draw", "DF-GAN: A Simple and Effective Baseline for Text-to-Image Synthesis"], "answer_arxiv_id": ["1605.05396", "1610.02454", "2008.05865"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_18145"} +{"question": "Could you mention some representative corpora for studying speaking styles in spoken conversations?", "answer": ["Towards Multimodal Sarcasm Detection (An _Obviously_ Perfect Paper)", "MELD: A Multimodal Multi-Party Dataset for Emotion Recognition in\n Conversations"], "answer_arxiv_id": ["1906.01815", "1810.02508"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_18146"} +{"question": "What papers have proposed the training of neural SDEs as GANs?", "answer": ["Neural SDEs as Infinite-Dimensional GANs"], "answer_arxiv_id": ["2102.03657"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_18147"} +{"question": "What works proposed to learn an intrinsic reward that is consistent with the expert intention?", "answer": ["CLUE: Calibrated Latent Guidance for Offline Reinforcement Learning"], "answer_arxiv_id": ["2306.13412"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_18148"} +{"question": "What work conducted contrastive learning to align local and global feature spaces using local and globally shared pseudo-data to reduce local learning bias?", "answer": ["FedBR: Improving Federated Learning on Heterogeneous Data via Local Learning Bias Reduction"], "answer_arxiv_id": ["2205.13462v4"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_18149"} +{"question": "What papers provide early work on quantification of uncertainty in deep models using Bayesian modeling?", "answer": ["Subspace Inference for Bayesian Deep Learning"], "answer_arxiv_id": ["1907.07504"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_18150"} +{"question": "Could you provide works that discuss VLMs with alignment objectives?", "answer": ["FLAVA: A Foundational Language And Vision Alignment Model", "Coarse-to-Fine Vision-Language Pre-training with Fusion in the Backbone"], "answer_arxiv_id": ["2112.04482", "2206.07643"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_18151"} +{"question": "What is the pioneering work in editing both the shape and color of neural fields by conditioning them on latent codes?", "answer": ["Editing Conditional Radiance Fields"], "answer_arxiv_id": ["2105.06466"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_18152"} +{"question": "What are the works that use a low-rank approximation of the attention matrix?", "answer": ["Linformer: Self-Attention with Linear Complexity"], "answer_arxiv_id": ["2006.04768"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18153"} +{"question": "What works are about off-policy evaluation in the area of contextual bandits?", "answer": ["Doubly Robust Policy Evaluation and Optimization", "Optimal and Adaptive Off-policy Evaluation in Contextual Bandits", "Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation", "More Robust Doubly Robust Off-policy Evaluation", "CAB: Continuous Adaptive Blending for Policy Evaluation and Learning", "Doubly robust off-policy evaluation with shrinkage", "Optimal Off-Policy Evaluation from Multiple Logging Policies"], "answer_arxiv_id": ["1503.02834", "1612.01205", "1810.12429", "1802.03493", "1811.02672", "1907.09623", "2010.11002"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_18154"} +{"question": "Could you provide me some studies employing robust loss functions with provable noise tolerance?", "answer": ["Normalized Loss Functions for Deep Learning with Noisy Labels"], "answer_arxiv_id": ["2006.13554"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_18155"} +{"question": "What works utilized large language models for sequence of tokens representations?", "answer": ["CodeBERT: A Pre-Trained Model for Programming and Natural Languages", "GraphCodeBERT: Pre-training Code Representations with Data Flow", "UniXcoder: Unified Cross-Modal Pre-training for Code Representation"], "answer_arxiv_id": ["2002.08155", "2009.08366", "2203.03850"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_18156"} +{"question": "What researches argue that ensuring convergence of the algorithm provides better transferability to new imaging tasks?", "answer": ["Image reconstruction algorithms in radio interferometry: from\n handcrafted to learned regularization denoisers"], "answer_arxiv_id": ["2202.12959"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18157"} +{"question": "Could you provide me some studies about the use of expert models in the domain of natural language processing?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated\n Mixture-of-Experts Layer", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts", "MoEBERT: from BERT to Mixture-of-Experts via Importance-Guided\n Adaptation"], "answer_arxiv_id": ["1701.06538", "2112.06905", "2204.07675"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_18158"} +{"question": "Which studies mention that equivariances can enforce hard constraints on functions and can even form a linear subspace that can be explicitly computed?", "answer": ["MDP Homomorphic Networks: Group Symmetries in Reinforcement Learning", "A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups"], "answer_arxiv_id": ["2006.16908", "2104.09459"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_18159"} +{"question": "Which study expands the method of merging individually fine-tuned models by introducing gradient fusion?", "answer": ["Mix-of-Show: Decentralized Low-Rank Adaptation for Multi-Concept\n Customization of Diffusion Models"], "answer_arxiv_id": ["2305.18292"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18160"} +{"question": "Which works assums that coarse GPS location and gravity direction are known?", "answer": ["Large-scale, real-time visual-inertial localization revisited"], "answer_arxiv_id": ["1907.00338"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18161"} +{"question": "Could you provide me some studies using box-supervised approaches for semantic segmentation?", "answer": ["Box-driven Class-wise Region Masking and Filling Rate Guided Loss for Weakly Supervised Semantic Segmentation"], "answer_arxiv_id": ["1904.11693"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_18162"} +{"question": "What papers discussed the use of fully convolutional networks (FCNs) and subsequent improvements in the field of Semantic Segmentation?", "answer": ["Fully Convolutional Networks for Semantic Segmentation", "U-Net: Convolutional Networks for Biomedical Image Segmentation", "Conditional Random Fields as Recurrent Neural Networks", "Multi-Scale Context Aggregation by Dilated Convolutions", "Segmentation-Aware Convolutional Networks Using Local Attention Masks", "Non-local Neural Networks", "Squeeze-and-Excitation Networks", "Dual Attention Network for Scene Segmentation", "Exploring Cross-Image Pixel Contrast for Semantic Segmentation", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers", "Visual Recognition with Deep Nearest Centroids", "GMMSeg: Gaussian Mixture based Generative Semantic Segmentation Models", "Attention Is All You Need", "SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers", "Segmenter: Transformer for Semantic Segmentation", "Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers", "A Unified Efficient Pyramid Transformer for Semantic Segmentation", "Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "Multi-Scale High-Resolution Vision Transformer for Semantic Segmentation"], "answer_arxiv_id": ["1411.4038", "1505.04597", "1502.03240v3", "1511.07122", "1708.04607", "1711.07971", "1709.01507", "1809.02983", "2101.11939", "2012.15840", "2209.07383", "2210.02025", "1706.03762", "2105.15203", "2105.05633", "2012.15840", "2107.14209", "2107.06278", "2112.01527", "2111.01236"], "source_meta": {"published_time": "20230503"}, "qid": "AutoScholarQuery_train_18163"} +{"question": "What papers used VQGAN-CLIP for image generation from text?", "answer": ["VQGAN-CLIP: Open Domain Image Generation and Editing with Natural\n Language Guidance", "Taming Transformers for High-Resolution Image Synthesis", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2204.08583", "2012.09841", "2103.00020"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_18164"} +{"question": "Which papers focus on applying DUN for super-resolution and deblurring?", "answer": ["Deep Unfolding Network for Image Super-Resolution", "Unfolded Deep Kernel Estimation for Blind Image Super-resolution", "Deep Generalized Unfolding Networks for Image Restoration"], "answer_arxiv_id": ["2003.10428", "2203.05568", "2204.13348"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_18165"} +{"question": "Which papers demonstrate the importance of human preference data for RLHF training?", "answer": ["RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models", "Training a Helpful and Harmless Assistant with Reinforcement Learning\n from Human Feedback", "RRHF: Rank Responses to Align Language Models with Human Feedback\n without tears", "BeaverTails: Towards Improved Safety Alignment of LLM via a\n Human-Preference Dataset", "Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language\n Models that Follow Instructions"], "answer_arxiv_id": ["2009.11462v2", "2204.05862", "2304.05302", "2307.04657", "2309.07875"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_18166"} +{"question": "Could you provide examples of recent studies related to MPDocVQA?", "answer": ["Hierarchical multimodal transformers for Multi-Page DocVQA", "Document Understanding Dataset and Evaluation (DUDE)"], "answer_arxiv_id": ["2212.05935", "2305.08455"], "source_meta": {"published_time": "20240107"}, "qid": "AutoScholarQuery_train_18167"} +{"question": "Do any researches discuss the use of prediction error of a model as the basis for the intrinsic reward?", "answer": ["VIME: Variational Information Maximizing Exploration", "Curiosity-driven Exploration by Self-supervised Prediction", "Self-Supervised Exploration via Disagreement", "Exploration by Random Network Distillation", "Novelty Search in Representational Space for Sample Efficient Exploration"], "answer_arxiv_id": ["1605.09674", "1705.05363", "1906.04161", "1810.12894", "2009.13579"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_18168"} +{"question": "What studies provide proof for the assumption of reversal of diffusion process, where they show the density of the reverse process as a mixture of Gaussian distributions?", "answer": ["On the Mathematics of Diffusion Models"], "answer_arxiv_id": ["2301.11108"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_18169"} +{"question": "What research proposes the reconstruction of images in the spectral domain for obtaining robust image representations?", "answer": ["Focal Frequency Loss for Image Reconstruction and Synthesis"], "answer_arxiv_id": ["2012.12821"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_18170"} +{"question": "What research work proposed a deformation and matching strategy to handle the intra-category shape variation in category-level 6D object pose estimation ?", "answer": ["Shape Prior Deformation for Categorical 6D Object Pose and Size\n Estimation"], "answer_arxiv_id": ["2007.08454"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_18171"} +{"question": "What research utilized cycle consistency in time to learn visual temporal correspondence?", "answer": ["Learning Correspondence from the Cycle-consistency of Time"], "answer_arxiv_id": ["1903.07593"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_18172"} +{"question": "Could you provide me some studies that have been developed for adversarial training or randomized smoothing as defences?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Theoretically Principled Trade-off between Robustness and Accuracy", "Adversarial Training for Free!", "Certified Adversarial Robustness via Randomized Smoothing", "(Certified!!) Adversarial Robustness for Free!"], "answer_arxiv_id": ["1706.06083", "1901.08573", "1904.12843", "1902.02918", "2206.10550"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_18173"} +{"question": "Could you provide me some works that utilized KPA to summarize and quantify the prevalence of opinions in business reviews?", "answer": ["Quantitative Argument Summarization and Beyond: Cross-Domain Key Point\n Analysis", "Every Bite Is an Experience: Key Point Analysis of Business Reviews"], "answer_arxiv_id": ["2010.05369", "2106.06758"], "source_meta": {"published_time": "20240719"}, "qid": "AutoScholarQuery_train_18174"} +{"question": "Which papers pioneered models that can quickly adapt to new tasks in machine learning?", "answer": ["A Model of Inductive Bias Learning"], "answer_arxiv_id": ["1106.0245v1"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_18175"} +{"question": "What works claim to outperform the latest video compression standards?", "answer": ["AlphaVC: High-Performance and Efficient Learned Video Compression", "Neural Video Compression with Diverse Contexts"], "answer_arxiv_id": ["2207.14678", "2302.14402"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_18176"} +{"question": "Can you mention some works about pure transformer architectures in vision tasks?", "answer": ["Segmenter: Transformer for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation", "SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers"], "answer_arxiv_id": ["2105.05633", "2112.01527", "2105.15203"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_18177"} +{"question": "Which works introduced Inverse Reinforcement Learning to deal with temporal drifting in trajectories?", "answer": ["A Connection Between Generative Adversarial Networks, Inverse Reinforcement Learning, and Energy-Based Models", "Generative Adversarial Imitation Learning"], "answer_arxiv_id": ["1611.03852", "1606.03476"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_18178"} +{"question": "Are there any studies that propose to reuse the off-policy algorithm with modifications for offline-to-online RL?", "answer": ["Efficient Online Reinforcement Learning with Offline Data"], "answer_arxiv_id": ["2302.02948"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_18179"} +{"question": "What study suggests the use of a pure test-time optimization approach for inferring the point trajectory and occlusion directly from videos?", "answer": ["Tracking Everything Everywhere All at Once"], "answer_arxiv_id": ["2306.05422"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_18180"} +{"question": "Could you provide me some studies that utilized prompting with special prefixes in multilingual NMT?", "answer": ["Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation", "Massively Multilingual Neural Machine Translation in the Wild: Findings and Challenges", "Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation"], "answer_arxiv_id": ["1611.04558", "1907.05019", "2004.11867"], "source_meta": {"published_time": "20230117"}, "qid": "AutoScholarQuery_train_18181"} +{"question": "Are there any methods that remove the poison effect from the image in order to defend against attacks?", "answer": ["Mask and Restore: Blind Backdoor Defense at Test Time with Masked Autoencoder"], "answer_arxiv_id": ["2303.15564"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_18182"} +{"question": "Which papers have proposed procedures to mitigate the trade-off phenomenon in the study of robust self training?", "answer": ["Are Labels Required for Improving Adversarial Robustness?", "Unlabeled Data Improves Adversarial Robustness", "Adversarially Robust Generalization Just Requires More Unlabeled Data", "Understanding and Mitigating the Tradeoff Between Robustness and Accuracy"], "answer_arxiv_id": ["1905.13725", "1905.13736", "1906.00555", "2002.10716"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_18183"} +{"question": "Can you point out works that explored fairness in education using machine learning models?", "answer": ["Algorithmic Fairness in Education"], "answer_arxiv_id": ["2007.05443v3"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_18184"} +{"question": "Could you provide some research on robust reinforcement learning where they tried to learn without the true reward models?", "answer": ["Robust Reinforcement Learning using Offline Data"], "answer_arxiv_id": ["2208.05129"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_18185"} +{"question": "Can you provide me some studies that construct a family of PI neural architectures for learning set representations, for instance, Deep Set?", "answer": ["Deep Sets"], "answer_arxiv_id": ["1703.06114"], "source_meta": {"published_time": "20220310"}, "qid": "AutoScholarQuery_train_18186"} +{"question": "What papers use 7B for training in LLaVA-Med?", "answer": ["LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2302.13971"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_18187"} +{"question": "Which works have achieved good performance in generating simple code pieces using language models?", "answer": ["Evaluating Large Language Models Trained on Code", "Program Synthesis with Large Language Models"], "answer_arxiv_id": ["2107.03374", "2108.07732"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_18188"} +{"question": "Can you name works where assumptions about lighting were made for multi-view approaches in inverse rendering of materials and lighting?", "answer": ["Neural Reflectance Fields for Appearance Acquisition", "Deep Reflectance Volumes: Relightable Reconstructions from Multi-View\n Photometric Images", "NeRV: Neural Reflectance and Visibility Fields for Relighting and View\n Synthesis"], "answer_arxiv_id": ["2008.03824", "2007.09892", "2012.03927"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18189"} +{"question": "Which works use contrastive regularization to improve the reliability of pseudo-labeling in semi-supervised learning?", "answer": ["Contrastive Regularization for Semi-Supervised Learning", "Class-Aware Contrastive Semi-Supervised Learning"], "answer_arxiv_id": ["2201.06247", "2203.02261"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_18190"} +{"question": "Which publications demonstrate the effectiveness of the pretrained CLIP model in the context of out-of-domain representation learning for control?", "answer": ["Simple but Effective: CLIP Embeddings for Embodied AI", "CLIPort: What and Where Pathways for Robotic Manipulation"], "answer_arxiv_id": ["2111.09888", "2109.12098"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_18191"} +{"question": "What work investigates the effect of duplicating text samples in the training set on memorization by LMs?", "answer": ["Deduplicating Training Data Mitigates Privacy Risks in Language Models"], "answer_arxiv_id": ["2202.06539"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_18192"} +{"question": "Which papers have proposed to verify reasoning chains for LM training?", "answer": ["Let's Verify Step by Step"], "answer_arxiv_id": ["2305.20050"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_18193"} +{"question": "What are the alternative reinforcement learning benchmarks to ALE?", "answer": ["Leveraging Procedural Generation to Benchmark Reinforcement Learning", "ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning", "DeepMind Lab", "Gotta Learn Fast: A New Benchmark for Generalization in RL"], "answer_arxiv_id": ["1912.01588", "1605.02097", "1612.03801", "1804.03720"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_18194"} +{"question": "Which papers have explored the use of the number of linear regions as a measure to approximate the expressivity of a deep neural network?", "answer": ["Neural Architecture Search without Training", "Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective", "Restructurable Activation Networks"], "answer_arxiv_id": ["2006.04647", "2102.11535", "2208.08562"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_18195"} +{"question": "In what papers the researchers design structured convolutional filters?", "answer": ["Exploiting Cyclic Symmetry in Convolutional Neural Networks", "Inception-v4, Inception-ResNet and the Impact of Residual Connections on\n Learning", "SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural\n Networks for Real-Time Object Detection for Autonomous Driving"], "answer_arxiv_id": ["1602.02660", "1602.07261", "1612.01051"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_18196"} +{"question": "What studies are there on developing audio segmentation and transcription system to collect high-quality video datasets with temporally aligned ADs?", "answer": ["A Dataset for Movie Description", "Movie Description", "Using Descriptive Video Services to Create a Large Data Source for Video\n Annotation Research", "MAD: A Scalable Dataset for Language Grounding in Videos from Movie\n Audio Descriptions"], "answer_arxiv_id": ["1501.02530", "1605.03705", "1503.01070", "2112.00431"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18197"} +{"question": "Could you provide me some research using sensitivity-based methods for neural network pruning?", "answer": ["SNIP: Single-shot Network Pruning based on Connection Sensitivity", "A Signal Propagation Perspective for Pruning Neural Networks at Initialization", "NISP: Pruning Networks using Neuron Importance Score Propagation", "Optimal Brain Compression: A Framework for Accurate Post-Training Quantization and Pruning"], "answer_arxiv_id": ["1810.02340", "1906.06307", "1711.05908", "2208.11580"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_18198"} +{"question": "Which papers discuss the extension of pre-trained T2I diffusion models and their finetuning on video data?", "answer": ["Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Video Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models"], "answer_arxiv_id": ["2305.10474v3", "2304.08818", "2204.03458", "2210.02303"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_18199"} +{"question": "What works have studied adversarial robustness for specific problems in robust streaming frameworks?", "answer": ["The Adversarial Robustness of Sampling", "Adversarially Robust Streaming via Dense–Sparse Trade-offs", "On the Robustness of CountSketch to Adaptive Inputs", "Adversarial Robustness of Streaming Algorithms through Importance Sampling", "Adversarial Laws of Large Numbers and Optimal Regret in Online Classification", "Adversarially Robust Coloring for Graph Streams"], "answer_arxiv_id": ["1906.11327", "2109.03785", "2202.13736", "2106.14952", "2101.09054", "2109.11130v1"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_18200"} +{"question": "Are there any studies that have looked into the discovery of concepts that are used by a model?", "answer": ["On Completeness-aware Concept-Based Explanations in Deep Neural Networks", "Towards Automatic Concept-based Explanations", "Explaining in Style: Training a GAN to explain a classifier in StyleSpace"], "answer_arxiv_id": ["1910.07969", "1902.03129", "2104.13369"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_18201"} +{"question": "What research concerns with few-shot in-context learning, zero-shot problem solving, and other abilities resulted from scaling LLMs?", "answer": ["Emergent Abilities of Large Language Models"], "answer_arxiv_id": ["2206.07682"], "source_meta": {"published_time": "20221103"}, "qid": "AutoScholarQuery_train_18202"} +{"question": "Which works used the method of knowledge distillation in continual learning?", "answer": ["Overcoming Catastrophic Forgetting in Incremental Object Detection via\n Elastic Response Distillation", "Large Scale Incremental Learning", "Dark Experience for General Continual Learning: a Strong, Simple\n Baseline"], "answer_arxiv_id": ["2204.02136", "1905.13260", "2004.07211"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_18203"} +{"question": "Any works that propose instruction back-translation that trains an LLM to generate instructions?", "answer": ["Self-Alignment with Instruction Backtranslation"], "answer_arxiv_id": ["2308.06259"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_18204"} +{"question": "What references discussed the challenges of federated learning including how to utilize the distributed data to learn a machine learning model with light communication cost without harming the data privacy?", "answer": ["Federated Optimization: Distributed Machine Learning for On-Device Intelligence", "Communication-Efficient Learning of Deep Networks from Decentralized Data"], "answer_arxiv_id": ["1610.02527", "1602.05629"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_18205"} +{"question": "Which works focus on pre-training in VideoQA techniques?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_18206"} +{"question": "Which paper discuss the combination of NeRF with intuitive fluid dynamics leveraging neural simulators?", "answer": ["NeuroFluid: Fluid Dynamics Grounding with Particle-Driven Neural Radiance Fields"], "answer_arxiv_id": ["2203.01762"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_18207"} +{"question": "What studies demonstrated that many methods exhibit inadequate calibration in uncertainty estimation?", "answer": ["On Calibration of Modern Neural Networks", "Revisiting the Calibration of Modern Neural Networks"], "answer_arxiv_id": ["1706.04599", "2106.07998"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_18208"} +{"question": "What works focused on localizing distinct feed-forward layers that are responsible for factual recall for model editing purposes?", "answer": ["Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2202.05262"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_18209"} +{"question": "Any research that uses image captions from PubMed articles for instruction-tuning within the Medical domain?", "answer": ["LLaVA-Med: Training a Large Language-and-Vision Assistant for\n Biomedicine in One Day"], "answer_arxiv_id": ["2306.00890"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_18210"} +{"question": "Can you mention some studies that explored applications of Graph Neural Networks?", "answer": ["Structure-aware Interactive Graph Neural Networks for the Prediction of Protein-Ligand Binding Affinity", "Discovering Symbolic Models from Deep Learning with Inductive Biases", "Learning Mesh-Based Simulation with Graph Networks", "Graph Neural Networks for Social Recommendation", "LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation"], "answer_arxiv_id": ["2107.10670", "2006.11287", "2010.03409", "1902.07243", "2002.02126"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_18211"} +{"question": "What are some use cases of these reduction techniques, like game solving and decision-time planning?", "answer": ["Improving Policies via Search in Cooperative Partially Observable Games", "Scalable Online Planning via Reinforcement Learning Fine-Tuning"], "answer_arxiv_id": ["1912.02318", "2109.15316"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_18212"} +{"question": "What works are there on supervised approaches that rely on the gradients estimated by a differentiable approximate function for training SNNs?", "answer": ["SLAYER: Spike Layer Error Reassignment in Time", "Spiking Deep Networks with LIF Neurons", "Long short-term memory and learning-to-learn in networks of spiking neurons", "Gradient Descent for Spiking Neural Networks"], "answer_arxiv_id": ["1810.08646", "1510.08829", "1803.09574", "1706.04698"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_18213"} +{"question": "What work has explored the training of GANs on neural representations for 3D shape generation?", "answer": ["Learning Implicit Fields for Generative Shape Modeling", "Adversarial Generation of Continuous Implicit Shape Representations", "3D Shape Generation with Grid-based Implicit Functions", "SDF-StyleGAN: Implicit SDF-Based StyleGAN for 3D Shape Generation"], "answer_arxiv_id": ["1812.02822", "2002.00349", "2107.10607", "2206.12055"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_18214"} +{"question": "What work introduced the concept of panoptic segmentation?", "answer": ["Panoptic Segmentation"], "answer_arxiv_id": ["1801.00868"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_18215"} +{"question": "What works tried to leverage k-nearest neighbor predictions for clean identification and label correction?", "answer": ["Multi-Objective Interpolation Training for Robustness to Label Noise", "Neighborhood Collective Estimation for Noisy Label Identification and Correction"], "answer_arxiv_id": ["2012.04462", "2208.03207"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_18216"} +{"question": "Are there any studies that made use of hyperbolic space for embedding words in natural language processing tasks?", "answer": ["Representation Tradeoffs for Hyperbolic Embeddings"], "answer_arxiv_id": ["1804.03329"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_18217"} +{"question": "What prior studies deal with large-scale image-text pretraining in image-language models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Grounded Language-Image Pre-training", "GLIPv2: Unifying Localization and Vision-Language Understanding", "Flamingo: a Visual Language Model for Few-Shot Learning", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2103.00020", "2102.05918", "2112.03857", "2206.05836", "2204.14198", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18218"} +{"question": "Could you name the work that introduced the concept of Instance Pattern Composers and their role in achieving high performance of generalizable INRs?", "answer": ["Generalizable Implicit Neural Representations via Instance Pattern Composers"], "answer_arxiv_id": ["2211.13223"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_18219"} +{"question": "Which paper discusses the assignment of weight to training samples based on the direction of their gradients?", "answer": ["Learning to Reweight Examples for Robust Deep Learning"], "answer_arxiv_id": ["1803.09050"], "source_meta": {"published_time": "20240709"}, "qid": "AutoScholarQuery_train_18220"} +{"question": "Can you list studies using Graphs and GNNs for spatial applications?", "answer": ["3D Dynamic Scene Graphs: Actionable Spatial Perception with Places,\n Objects, and Humans", "Graph-to-3D: End-to-End Generation and Manipulation of 3D Scenes Using\n Scene Graphs", "Spatial Commonsense Graph for Object Localisation in Partial Scenes", "PoserNet: Refining Relative Camera Poses Exploiting Object Detections", "Hierarchical Representations and Explicit Memory: Learning Effective\n Navigation Policies on 3D Scene Graphs using Graph Neural Networks"], "answer_arxiv_id": ["2002.06289", "2108.08841", "2203.05380", "2207.09445v2", "2108.01176"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18221"} +{"question": "In what papers was the conditional latent diffusion model, designed to improve inference speed, proposed?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_18222"} +{"question": "Among the transformer model architectures for establishing correspondence, where is TubeDETR employed?", "answer": ["TubeDETR: Spatio-Temporal Video Grounding with Transformers"], "answer_arxiv_id": ["2203.16434"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_18223"} +{"question": "Who define a notion of optimal representations for batch Monte Carlo optimization?", "answer": ["A Geometric Perspective on Optimal Representations for Reinforcement Learning"], "answer_arxiv_id": ["1901.11530"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_18224"} +{"question": "Could you provide examples of research that employ Data coverage maximization methods in RL?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation", "Provably Efficient Maximum Entropy Exploration", "Efficient Exploration via State Marginal Matching", "Behavior From the Void: Unsupervised Active Pre-Training", "Reinforcement Learning with Prototypical Representations", "State Entropy Maximization with Random Encoders for Efficient Exploration"], "answer_arxiv_id": ["1606.01868", "1812.02690", "1906.05274", "2103.04551", "2102.11271v2", "2102.09430"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_18225"} +{"question": "What works proposed data augmentation strategies to manage the overfitting problem of the discriminator in generative adversarial networks?", "answer": ["Differentiable Augmentation for Data-Efficient GAN Training", "Training Generative Adversarial Networks with Limited Data", "Deceive D: Adaptive Pseudo Augmentation for GAN Training with Limited Data"], "answer_arxiv_id": ["2006.10738", "2006.06676", "2111.06849"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_18226"} +{"question": "Which works discuss the use of list decoding and cyclic redundancy checks (CRC) to address the sub-optimality of successive cancellation (SC)?", "answer": ["List Decoding of Polar Codes", "LLR-based Successive Cancellation List Decoding of Polar Codes", "An Adaptive Successive Cancellation List Decoder for Polar Codes with Cyclic Redundancy Check"], "answer_arxiv_id": ["1206.0050", "1401.3753v4", "1208.3091v1"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_18227"} +{"question": "What works have proposed techniques for knowledge transfer?", "answer": ["Continual Learning of a Mixed Sequence of Similar and Dissimilar Tasks", "Achieving Forgetting Prevention and Knowledge Transfer in Continual Learning", "Beyond Not-Forgetting: Continual Learning with Backward Knowledge Transfer"], "answer_arxiv_id": ["2112.10017", "2112.02706", "2211.00789"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_18228"} +{"question": "Which research unified preceding techniques into a labeling trick?", "answer": ["Labeling Trick: A Theory of Using Graph Neural Networks for Multi-Node Representation Learning"], "answer_arxiv_id": ["2010.16103"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18229"} +{"question": "Can you suggest some studies that discuss the creation of adversarial examples in a white-box setting?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_18230"} +{"question": "Could you provide me some studies that have used additional single-view data or custom training strategies in material capture?", "answer": ["Modeling Surface Appearance from a Single Photograph using\n Self-augmented Convolutional Neural Networks", "SurfaceNet: Adversarial SVBRDF Estimation from a Single Image", "Guided Fine-Tuning for Large-Scale Material Transfer"], "answer_arxiv_id": ["1809.00886", "2107.11298", "2007.03059"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_18231"} +{"question": "Which paper proposed Multiexpert Gaussian Processes (MGP) that combines the predictive posterior distributions of multiple Gaussian Processes?", "answer": ["Uncertainty Estimation for Multi-view Data: The Power of Seeing the Whole Picture"], "answer_arxiv_id": ["2210.02676"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_18232"} +{"question": "Where have blueprint only methods achieved state-of-the-art performance?", "answer": ["DouZero: Mastering DouDizhu with Self-Play Deep Reinforcement Learning", "Suphx: Mastering Mahjong with Deep Reinforcement Learning", "Pipeline PSRO: A Scalable Approach for Finding Approximate Nash Equilibria in Large Games", "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning"], "answer_arxiv_id": ["2106.06135", "2003.13590", "2006.08555", "2206.15378"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_18233"} +{"question": "Which researcher proposed a fast deep learning-based solution to MVPS?", "answer": ["MVPSNet: Fast Generalizable Multi-view Photometric Stereo"], "answer_arxiv_id": ["2305.11167"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_18234"} +{"question": "Can you mention any work that utilizes similar delay scheme?", "answer": ["Non-Stochastic Control with Bandit Feedback"], "answer_arxiv_id": ["2008.05523"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_18235"} +{"question": "Which papers have derived algorithmic fairness with respect to statistics on certain factual/counterfactual groups?", "answer": ["Principal Fairness for Human and Algorithmic Decision-Making", "Counterfactual Risk Assessments, Evaluation, and Fairness", "Fairness in Risk Assessment Instruments: Post-Processing to Achieve Counterfactual Equalized Odds"], "answer_arxiv_id": ["2005.10400", "1909.00066", "2009.02841"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_18236"} +{"question": "Which research has worked on deriving bounds in the non-asymptotic regime for TD learning?", "answer": ["A Finite Time Analysis of Temporal Difference Learning With Linear Function Approximation", "Finite Sample Analyses for TD(0) with Function Approximation", "Finite time analysis of temporal difference learning with linear function approximation: Tail averaging and regularisation", "Finite-Time Error Bounds For Linear Stochastic Approximation and TD Learning", "Concentration bounds for temporal difference learning with linear function approximation: The case of batch data and uniform sampling", "A Non-asymptotic Analysis of Non-parametric Temporal-Difference Learning"], "answer_arxiv_id": ["1806.02450", "1704.01161", "2210.05918v2", "1902.00923", "1306.2557", "2205.11831"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_18237"} +{"question": "What work proposed an extension of the findings of earlier works regarding Lipschitz continuity and adversarial examples in DNNs with ReLU activations?", "answer": ["Evaluating the Robustness of Neural Networks: An Extreme Value Theory Approach"], "answer_arxiv_id": ["1801.10578"], "source_meta": {"published_time": "20230309"}, "qid": "AutoScholarQuery_train_18238"} +{"question": "What are the studies involved in solving SINDy problems with greater success, ability to verify optimality, but scale issues?", "answer": ["Learning Sparse Nonlinear Dynamics via Mixed-Integer Optimization"], "answer_arxiv_id": ["2206.00176"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_18239"} +{"question": "What studies have used contrastive learning in image-text alignment models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "Sigmoid Loss for Language Image Pre-Training"], "answer_arxiv_id": ["2103.00020", "2111.07991", "2303.15343v4"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_18240"} +{"question": "Are there any works that use conditional generative adversarial networks to generate a BEV occupancy image?", "answer": ["SimNet: Learning Reactive Self-driving Simulations from Real-world Observations"], "answer_arxiv_id": ["2105.12332"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_18241"} +{"question": "What works focus on efficient fine-tuning VLMs on downstream tasks with BLIP-base?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2201.12086"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_18242"} +{"question": "Which research highlighted the problem with relative positional encoding in Transformers?", "answer": ["A Length-Extrapolatable Transformer"], "answer_arxiv_id": ["2212.10554"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18243"} +{"question": "Which research works extended the idea of modeling diverse tasks as sequence generation tasks by using discrete coordinate tokens?", "answer": ["Unified-IO: A unified model for vision, language, and multi-modal tasks", "A Unified Sequence Interface for Vision Tasks", "UniTAB: Unifying Text and Box Outputs for Grounded Vision-Language Modeling"], "answer_arxiv_id": ["2206.08916", "2206.07669", "2111.12085"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18244"} +{"question": "Which studies used conditional generative models to address ill-posedness of super-resolution?", "answer": ["Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network", "ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks", "SRFlow: Learning the Super-Resolution Space with Normalizing Flow", "Conditional Injective Flows for Bayesian Imaging"], "answer_arxiv_id": ["1609.04802", "1809.00219", "2006.14200", "2204.07664"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_18245"} +{"question": "Could you name the studies where the MCTS approach was proposed for sequential decision-making problems?", "answer": ["Monte Carlo Tree Search: A Review of Recent Modifications and Applications"], "answer_arxiv_id": ["2103.04931"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_18246"} +{"question": "Which studies on backdoor attacks imply that the prompts of LLMs present a surface for Trojan attacks?", "answer": ["TrojText: Test-time Invisible Textual Trojan Insertion", "SSL-Cleanse: Trojan Detection and Mitigation in Self-Supervised Learning", "Audit and Improve Robustness of Private Neural Networks on Encrypted Data", "ESTAS: Effective and Stable Trojan Attacks in Self-supervised Encoders with One Target Unlabelled Sample", "TrojViT: Trojan Insertion in Vision Transformers"], "answer_arxiv_id": ["2303.02242", "2303.09079", "2209.09996", "2211.10908v2", "2208.13049"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_18247"} +{"question": "Which papers propose LLMs specifically trained for evaluation tasks?", "answer": ["PandaLM: An Automatic Evaluation Benchmark for LLM Instruction Tuning\n Optimization", "Generative Judge for Evaluating Alignment", "Prometheus: Inducing Fine-grained Evaluation Capability in Language\n Models"], "answer_arxiv_id": ["2306.05087", "2310.05470", "2310.08491"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_18248"} +{"question": "What works have explored attributed prompts to reduce the issue of low informativeness and redundancy in training data generation?", "answer": ["Mixture of Soft Prompts for Controllable Data Generation"], "answer_arxiv_id": ["2303.01580"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_18249"} +{"question": "Which research paper delivers a comprehensive survey on rearrangement?", "answer": ["Rearrangement: A Challenge for Embodied AI"], "answer_arxiv_id": ["2011.01975"], "source_meta": {"published_time": "20220906"}, "qid": "AutoScholarQuery_train_18250"} +{"question": "What studies have applied LLMs to video comprehension?", "answer": ["VideoChat: Chat-Centric Video Understanding", "Video-LLaVA: Learning United Visual Representation by Alignment Before\n Projection"], "answer_arxiv_id": ["2305.06355", "2311.10122"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_18251"} +{"question": "Which papers introduced ways to inject knowledge without changing the model architecture?", "answer": ["Fine-Tuning or Retrieval? Comparing Knowledge Injection in LLMs", "RAG vs Fine-tuning: Pipelines, Tradeoffs, and a Case Study on\n Agriculture"], "answer_arxiv_id": ["2312.05934", "2401.08406"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_18252"} +{"question": "Which works leverage meta-learning for model selection in unsupervised AD, based on the similarity between the new task and historical datasets?", "answer": ["Toward Unsupervised Outlier Model Selection"], "answer_arxiv_id": ["2211.01834"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_18253"} +{"question": "Which studies proposed the concept of selectively tuning some parameters of a pre-trained network?", "answer": ["DeCAF: A Deep Convolutional Activation Feature for Generic Visual\n Recognition", "BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models"], "answer_arxiv_id": ["1310.1531", "2106.10199"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_18254"} +{"question": "Which research first mentioned that a fixed model can result in poor performance due to the possible shift between training and test data distribution?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["1903.12261"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_18255"} +{"question": "Which studies address optimization schemes in Federated Learning?", "answer": ["Adaptive Federated Optimization", "FedPD: A Federated Learning Framework with Optimal Rates and Adaptivity to Non-IID Data"], "answer_arxiv_id": ["2003.00295", "2005.11418"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_18256"} +{"question": "Which works discussed the necessity of explicit multi-scale supervision through downsampling images?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale\n Scene Rendering"], "answer_arxiv_id": ["2103.13415", "2112.05504"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_18257"} +{"question": "Which studies have used graph neural networks to reason spatial constraints of human joints for 3D human pose estimation?", "answer": ["Graph Stacked Hourglass Networks for 3D Human Pose Estimation", "Learning Skeletal Graph Neural Networks for Hard 3D Pose Estimation", "GraphMLP: A Graph MLP-Like Architecture for 3D Human Pose Estimation"], "answer_arxiv_id": ["2103.16385", "2108.07181", "2206.06420"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_18258"} +{"question": "What studies have extended the application of CLIP into the 3D domain?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP", "PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning", "Can Language Understand Depth?", "Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D\n Understanding, Generation, and Instruction Following"], "answer_arxiv_id": ["2112.02413", "2211.11682", "2207.01077", "2309.00615"], "source_meta": {"published_time": "20240131"}, "qid": "AutoScholarQuery_train_18259"} +{"question": "What papers dealt with aliasing issue in Neural Radiance Fields?", "answer": ["​​Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields​", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Exact-NeRF: An Exploration of a Precise Volumetric Parameterization for Neural Radiance Fields", "BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale Scene Rendering", "Local Implicit Ray Function for Generalizable Radiance Field Representation"], "answer_arxiv_id": ["2103.13415", "2111.12077", "2211.12285", "2112.05504", "2304.12746"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_18260"} +{"question": "Can you provide the references where transformers-based language models have been used to solve hypothesis classification?", "answer": ["Transformers as Soft Reasoners over Language", "RuleBERT: Teaching Soft Rules to Pre-Trained Language Models", "Logically Consistent Adversarial Attacks for Soft Theorem Provers", "Pushing the Limits of Rule Reasoning in Transformers through Natural Language Satisfiability", "RobustLR: A Diagnostic Benchmark for Evaluating Logical Robustness of Deductive Reasoners"], "answer_arxiv_id": ["2002.05867", "2109.13006", "2205.00047", "2112.09054", "2205.12598"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_18261"} +{"question": "Which works used SDE approximation to analyze the factors influencing the minima found by SGD?", "answer": ["Three Factors Influencing Minima in SGD"], "answer_arxiv_id": ["1711.04623"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_18262"} +{"question": "In which research has contextual structure been preserved to guarantee the local feature/pixel continuity for image inpainting?", "answer": ["Coherent Semantic Attention for Image Inpainting"], "answer_arxiv_id": ["1905.12384"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_18263"} +{"question": "Could you name the works which focused on extending the SPO framework for combinatorial problems?", "answer": ["CombOptNet: Fit the Right NP-Hard Problem by Learning Integer Programming Constraints", "Melding the Data-Decisions Pipeline: Decision-Focused Learning for Combinatorial Optimization", "Differentiation of Blackbox Combinatorial Solvers", "Implicit MLE: Backpropagating Through Discrete Exponential Family Distributions", "SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver", "MIPaaL: Mixed Integer Program as a Layer", "Smart Predict-and-Optimize for Hard Combinatorial Optimization Problems"], "answer_arxiv_id": ["2105.02343", "1809.05504", "1912.02175", "2106.01798", "1905.12149", "1907.05912", "1911.10092"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_18264"} +{"question": "Which research papers focus on the surface form of textual input to VLMs?", "answer": ["Disentangling visual and written concepts in CLIP"], "answer_arxiv_id": ["2206.07835"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_18265"} +{"question": "What research papers tend to tackle structural distribution shifts on static graphs?", "answer": ["Out-Of-Distribution Generalization on Graphs: A Survey", "Discovering Invariant Rationales for Graph Neural Networks", "Handling Distribution Shifts on Graphs: An Invariance Perspective", "Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training Data", "OOD-GNN: Out-of-Distribution Generalized Graph Neural Network", "Learning to Solve Travelling Salesman Problem with Hardness-Adaptive Curriculum", "Revisiting Transformation Invariant Geometric Deep Learning: Are Initial Representations All You Need?", "Generalizing Graph Neural Networks on Out-Of-Distribution Graphs"], "answer_arxiv_id": ["2202.07987", "2201.12872", "2202.02466", "2108.01099", "2112.03806", "2204.03236", "2112.12345", "2111.10657"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_18266"} +{"question": "What papers have been cited for efficient transformers architecture design that could improve inference speed and reduce memory usage?", "answer": ["Reformer: The Efficient Transformer", "SqueezeBERT: What can computer vision teach NLP about efficient neural networks?", "MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices", "Linformer: Self-Attention with Linear Complexity", "Lite Transformer with Long-Short Range Attention", "TraDE: Transformers for Density Estimation", "ALBERT: A Lite BERT for Self-supervised Learning of Language Representations", "LittleBird: Efficient Faster & Longer Transformer for Question Answering"], "answer_arxiv_id": ["2001.04451", "2006.11316", "2004.02984", "2006.04768", "2004.11886", "2004.02441", "1909.11942", "2210.11870"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_18267"} +{"question": "What is the research that proposed NeuralUDF for learning UDF in reconstructing open models?", "answer": ["NeuralUDF: Learning Unsigned Distance Fields for Multi-view\n Reconstruction of Surfaces with Arbitrary Topologies"], "answer_arxiv_id": ["2211.14173"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_18268"} +{"question": "Could you provide some works that explored bias in question answering?", "answer": ["Compositional Questions Do Not Necessitate Multi-hop Reasoning"], "answer_arxiv_id": ["1906.02900"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18269"} +{"question": "Are there any papers that generalized the 111-WL limitation to the k-WL?", "answer": ["The expressive power of kth-order invariant graph networks", "Provably Powerful Graph Networks", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "SpeqNets: Sparsity-aware Permutation-equivariant Graph Networks"], "answer_arxiv_id": ["2007.12035v1", "1905.11136", "1810.02244", "2203.13913"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_18270"} +{"question": "What studies utilized perturbations in non-generative methods for explainability in Graph Neural Networks?", "answer": ["GraphLIME:Local Interpretable Model Explanations for Graph Neural Networks", "Interpreting Graph Neural Networks for NLP With Differentiable Edge Masking"], "answer_arxiv_id": ["2001.06216", "2010.00577"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18271"} +{"question": "What studies inspired Yang et al. to introduced a pioneering method for 3D facial makeup acquisition?", "answer": ["PSGAN: Pose and Expression Robust Spatial-Aware GAN for Customizable\n Makeup Transfer", "LADN: Local Adversarial Disentangling Network for Facial Makeup and\n De-Makeup", "Lipstick ain't enough: Beyond Color Matching for In-the-Wild Makeup\n Transfer"], "answer_arxiv_id": ["1909.06956", "1904.11272", "2104.01867"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_18272"} +{"question": "What studies attempts to reduce the number of communication in decentralized optimization with fixed topology?", "answer": ["ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!†"], "answer_arxiv_id": ["2202.09357"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_18273"} +{"question": "In which works are topological graphs introduced for environment modeling in VLN?", "answer": ["Structured Scene Memory for Vision-Language Navigation", "Evolving Graphical Planner: Contextual Global Planning for\n Vision-and-Language Navigation", "Language and Visual Entity Relationship Graph for Agent Navigation", "Topological Planning with Transformers for Vision-and-Language\n Navigation", "Think Global, Act Local: Dual-scale Graph Transformer for Vision-and-Language Navigation", "ETPNav: Evolving Topological Planning for Vision-Language Navigation in Continuous Environments"], "answer_arxiv_id": ["2103.03454", "2007.05655", "2010.09304", "2012.05292", "2202.11742v1", "2304.03047v3"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_18274"} +{"question": "Are there any papers that have generalized the horizon-free results to other MDP problems?", "answer": ["Improved Variance-Aware Confidence Sets for Linear Bandits and Linear Mixture MDP", "Improved Regret Analysis for Variance-Adaptive Linear Bandits and Horizon-Free Linear Mixture MDPs", "Stochastic Shortest Path: Minimax, Parameter-Free and Towards Horizon-Free Regret", "Computationally Efficient Horizon-Free Reinforcement Learning for Linear Mixture MDPs"], "answer_arxiv_id": ["2101.12745", "2111.03289", "2104.11186", "2205.11507"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_18275"} +{"question": "Can you name the works specifically discussing pre-trained text-to-image latent diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_18276"} +{"question": "What work extended the study of ridge ensembles for RMT features?", "answer": ["Bagging in overparameterized learning: Risk characterization and risk monotonization"], "answer_arxiv_id": ["2210.11445"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18277"} +{"question": "What studies demonstrate the ability of diffusion models to generate training data in zero or few shot settings?", "answer": ["Is synthetic data from generative models ready for image recognition?", "Effective Data Augmentation With Diffusion Models"], "answer_arxiv_id": ["2210.07574", "2302.07944"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18278"} +{"question": "Could you provide works that accelerate per-prediction of online learning on a graph by working on trees and paths of the graph?", "answer": ["Random Spanning Trees and the Prediction of Weighted Graphs"], "answer_arxiv_id": ["1212.5637"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18279"} +{"question": "Which papers treat the complex queries as operator trees?", "answer": ["Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings", "Beta Embeddings for Multi-Hop Logical Reasoning in Knowledge Graphs", "Benchmarking the Combinatorial Generalizability of Complex Query Answering on Knowledge Graphs"], "answer_arxiv_id": ["2002.05969", "2010.11465", "2109.08925"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_18280"} +{"question": "Who developed stereo matching networks that leverage a multi-scale cost volume?", "answer": ["Hierarchical Deep Stereo Matching on High-resolution Images", "AANet: Adaptive Aggregation Network for Efficient Stereo Matching", "PCW-Net: Pyramid Combination and Warping Cost Volume for Stereo Matching"], "answer_arxiv_id": ["1912.06704", "2004.09548", "2006.12797"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18281"} +{"question": "Can you name some papers that discuss dividing complex tasks into smaller sub-tasks within the concept of hierarchical reinforcement learning?", "answer": ["The Option-Critic Architecture", "FeUdal Networks for Hierarchical Reinforcement Learning", "Data-Efficient Hierarchical Reinforcement Learning"], "answer_arxiv_id": ["1609.05140", "1703.01161", "1805.08296"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_18282"} +{"question": "What works derived deformable attention from deformable CNN?", "answer": ["Deformable Convolutional Networks", "Deformable ConvNets v2: More Deformable, Better Results"], "answer_arxiv_id": ["1703.06211", "1811.11168"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_18283"} +{"question": "Which papers focus on sample construction approaches such as diversity and uncertainty sampling in the field of active learning?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Deep Active Learning over the Long Tail", "Discriminative Active Learning"], "answer_arxiv_id": ["1708.00489", "1711.00941", "1907.06347"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_18284"} +{"question": "What papers have divided the matting task into a trimap generation and a trimap-based matting subtasks?", "answer": ["Bridging Composite and Real: Towards End-to-end Deep Image Matting"], "answer_arxiv_id": ["2010.16188v3"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_18285"} +{"question": "What work is the most related to the Tiered RL framework and considered the case when MHi=MLo?", "answer": ["Tiered Reinforcement Learning: Pessimism in the Face of Uncertainty and Constant Regret"], "answer_arxiv_id": ["2205.12418"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_18286"} +{"question": "Which works have dealt with stochastic variational inequalities with i. i. d. noise?", "answer": ["Reducing Noise in GAN Training with Variance Reduced Extragradient", "Stochastic Variance Reduction Methods for Saddle-Point Problems", "Stochastic Variance Reduction for Variational Inequality Methods", "Stochastic Gradient Descent-Ascent: Unified Theory and New Efficient Methods"], "answer_arxiv_id": ["1904.08598", "1605.06398", "2102.08352", "2202.07262v3"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18287"} +{"question": "Which studies discuss the problem of measuring the transferability from the source data pre-training the PTM to the target downstream task?", "answer": ["Scalable Diverse Model Selection for Accessible Transfer Learning", "Newer is not always better: Rethinking transferability metrics, their peculiarities, stability and performance", "How stable are Transferability Metrics evaluations?", "Which Model to Transfer? Finding the Needle in the Growing Haystack"], "answer_arxiv_id": ["2111.06977", "2110.06893", "2204.01403", "2010.06402"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_18288"} +{"question": "Which studies applied deep neural networks into the task of open-set recognition?", "answer": ["Towards Open Set Deep Networks"], "answer_arxiv_id": ["1511.06233"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_18289"} +{"question": "What paper introduces a method for estimating confidence levels by using linguistic invariances to calculate semantic entropy?", "answer": ["Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation\n in Natural Language Generation"], "answer_arxiv_id": ["2302.09664"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_18290"} +{"question": "What studies have proposed using probabilistic modeling approaches for unsupervised accuracy estimation in classifier models?", "answer": ["Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach"], "answer_arxiv_id": ["1705.07086"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_18291"} +{"question": "What works used projections to enforce a close distance to the pre-trained model?", "answer": ["Distance-Based Regularisation of Deep Networks for Fine-Tuning", "Trainable Projected Gradient Method for Robust Fine-tuning"], "answer_arxiv_id": ["2002.08253", "2303.10720"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_18292"} +{"question": "What were the studies that relax the assumption of known poses in the context of Neural Radiance Fields?", "answer": ["NeRF-⁣-: Neural Radiance Fields Without Known Camera Parameters", "GNeRF: GAN-based Neural Radiance Field without Posed Camera", "BARF : Bundle-Adjusting Neural Radiance Fields", "NeRF-Pose: A First-Reconstruct-Then-Regress Approach for Weakly-supervised 6D Object Pose Estimation", "GARF: Gaussian Activated Radiance Fields for High Fidelity Reconstruction and Pose Estimation", "SparsePose: Sparse-View Camera Pose Regression and Refinement"], "answer_arxiv_id": ["2102.07064", "2103.15606", "2104.06405", "2203.04802", "2204.05735v1", "2211.16991"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_18293"} +{"question": "Could you provide me some work that advanced temporal action detection by merging the advantages of both anchor-based and anchor-free methods?", "answer": ["Revisiting Anchor Mechanisms for Temporal Action Localization"], "answer_arxiv_id": ["2008.09837"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_18294"} +{"question": "Which work proposed non-local approximated gradients that utilize gradients backpropagated from a loss function in the field of differentiable rendering?", "answer": ["Neural 3D Mesh Renderer"], "answer_arxiv_id": ["1711.07566"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_18295"} +{"question": "Any works about applying self-play in complex games?", "answer": ["Emergent Tool Use From Multi-Agent Autocurricula"], "answer_arxiv_id": ["1909.07528v2"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_18296"} +{"question": "Could you provide me some studies about GAN-based approaches for editing facial images?", "answer": ["Generative Adversarial Networks", "TransEditor: Transformer-Based Dual-Space GAN for Highly Controllable\n Facial Editing", "SemanticStyleGAN: Learning Compositional Generative Priors for\n Controllable Image Synthesis and Editing", "StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via\n Pre-trained StyleGAN", "Training and Tuning Generative Neural Radiance Fields for\n Attribute-Conditional 3D-Aware Face Generation", "Barbershop: GAN-based Image Compositing using Segmentation Masks", "Style Your Hair: Latent Optimization for Pose-Invariant Hairstyle\n Transfer via Local-Style-Aware Hair Alignment", "In-Domain GAN Inversion for Real Image Editing", "Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661", "2203.17266", "2112.02236", "2203.04036", "2208.12550", "2106.01505", "2208.07765", "2004.00049", "2112.07945"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_18297"} +{"question": "What works are there on estimating maximal performance for regression problems?", "answer": ["Performance of regression models as a function of experiment noise"], "answer_arxiv_id": ["1912.08141v3"], "source_meta": {"published_time": "20220201"}, "qid": "AutoScholarQuery_train_18298"} +{"question": "What works improved the approximation results of deep belief networks by reducing the number of layers?", "answer": ["Refinements of Universal Approximation Results for Deep Belief Networks and Restricted Boltzmann Machines"], "answer_arxiv_id": ["1005.1593v2"], "source_meta": {"published_time": "20220818"}, "qid": "AutoScholarQuery_train_18299"} +{"question": "Which works are about safe reinforcement learning?", "answer": ["Reinforcement Learning with Almost Sure Constraints", "Safe Reinforcement Learning with Linear Function Approximation", "Safe Adaptive Learning-based Control for Constrained Linear Quadratic Regulators with Regret Guarantees"], "answer_arxiv_id": ["2112.05198v3", "2106.06239", "2111.00411"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_18300"} +{"question": "Which work proposed multi-modal prompt tuning (MaPLe)?", "answer": ["MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2210.03117"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_18301"} +{"question": "What studies explored CS for efficient standard training in supervised settings?", "answer": ["Coresets for Data-efficient Training of Machine Learning Models", "Glister: Generalization based Data Subset Selection for Efficient and Robust Learning", "Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training"], "answer_arxiv_id": ["1906.01827", "2012.10630", "2103.00123"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_18302"} +{"question": "Are there any researches identified that specifically address convolutional neural networks and hierarchically connected networks within MP-NAS methods?", "answer": ["DeepMAD: Mathematical Architecture Design for Deep Convolutional Neural Network", "LayerNAS: Neural Architecture Search in Polynomial Complexity"], "answer_arxiv_id": ["2303.02165v3", "2304.11517"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_18303"} +{"question": "Which papers propose methods for model editing by retaining requirements while extending to pretrain hypernetworks that predict edits?", "answer": ["Editing Factual Knowledge in Language Models", "Memory-Based Model Editing at Scale", "Fast Model Editing at Scale"], "answer_arxiv_id": ["2104.08164", "2206.06520", "2110.11309"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_18304"} +{"question": "Which paper conducted an analysis in the spectral domain on the convolution degree of graphon neural networks?", "answer": ["Graphon Neural Networks and the Transferability of Graph Neural Networks"], "answer_arxiv_id": ["2006.03548"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_18305"} +{"question": "Which papers focused on how the attention maps of Transformers encode the syntactic structure of a sentence?", "answer": ["Do Attention Heads in BERT Track Syntactic Dependencies?", "What Does BERT Look At? An Analysis of BERT's Attention", "Universal Dependencies according to BERT: both more specific and more\n general"], "answer_arxiv_id": ["1911.12246", "1906.04341", "2004.14620"], "source_meta": {"published_time": "20240209"}, "qid": "AutoScholarQuery_train_18306"} +{"question": "Could you tell me about the studies that extended the prompt tuning technique?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "AutoPrompt: Eliciting Knowledge from Language Models with Automatically\n Generated Prompts", "Language Models as Knowledge Bases?", "E-BERT: Efficient-Yet-Effective Entity Embeddings for BERT", "Learning to Prompt for Vision-Language Models", "Visual-Language Prompt Tuning with Knowledge-guided Context Optimization", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2107.13586v1", "2010.15980", "1909.01066", "1911.03681", "2109.01134", "2303.13283", "2203.05557"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_18307"} +{"question": "Which papers have developed conditional GANs incorporating degraded images as conditions?", "answer": ["Unsupervised Night Image Enhancement: When Layer Decomposition Meets\n Light-Effects Suppression", "Unsupervised Deraining: Where Contrastive Learning Meets Self-similarity", "FD-GAN: Generative Adversarial Networks with Fusion-discriminator for\n Single Image Dehazing"], "answer_arxiv_id": ["2207.10564", "2203.11509", "2001.06968"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_18308"} +{"question": "Which methods construct a generalizable 3D representation by making use of features from various observed viewpoints?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "pixelNeRF: Neural Radiance Fields from One or Few Images"], "answer_arxiv_id": ["2102.13090", "2103.15595", "2012.02190"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_18309"} +{"question": "What papers used the strategy of providing intermediate reasoning steps, also known as Chain-of-Thought, for multi-step reasoning tasks?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2201.11903", "2204.02311"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_18310"} +{"question": "Can you provide examples of studies that address intra-class shape variations through the incorporation of additional shape priors?", "answer": ["Shape Prior Deformation for Categorical 6D Object Pose and Size\n Estimation", "SAR-Net: Shape Alignment and Recovery Network for Category-level 6D\n Object Pose and Size Estimation", "ACR-Pose: Adversarial Canonical Representation Reconstruction Network\n for Category Level 6D Object Pose Estimation"], "answer_arxiv_id": ["2007.08454", "2106.14193", "2111.10524"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_18311"} +{"question": "Can you point me to the research about large-scale fundamental detection models?", "answer": ["DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via Word-Region Alignment", "Grounded Language-Image Pre-training", "GLIPv2: Unifying Localization and Vision-Language Understanding", "MDETR - Modulated Detection for End-to-End Multi-Modal Understanding", "OmDet: Language-Aware Object Detection with Large-scale Vision-Language Multi-dataset Pre-training"], "answer_arxiv_id": ["2304.04514", "2112.03857", "2206.05836", "2104.12763", "2209.05946"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_18312"} +{"question": "Any studies propose the sequential versions of the Modern Hopfield Network and the Modern Continuous Hopfield Network?", "answer": ["Dense Associative Memory for Pattern Recognition", "Hopfield Networks is All You Need", "Attention Is All You Need"], "answer_arxiv_id": ["1606.01164", "2008.02217", "1706.03762"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_18313"} +{"question": "Could you provide me some researches proposing the combination of Bayesian Optimization and Multi-fidelity methods?", "answer": ["BOHB: Robust and Efficient Hyperparameter Optimization at Scale"], "answer_arxiv_id": ["1807.01774"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_18314"} +{"question": "What work initially introduced SR-STE that uses sparse refinement when evaluating gradients via masked weights?", "answer": ["Learning N:M Fine-grained Structured Sparse Neural Networks From Scratch"], "answer_arxiv_id": ["2102.04010"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_18315"} +{"question": "What papers have been presented to tackle vision tasks using the GS-Event method?", "answer": ["Event-Based Fusion for Motion Deblurring with Cross-modal Attention", "Motion Deblurring with Real Events"], "answer_arxiv_id": ["2112.00167", "2109.13695"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_18316"} +{"question": "What are some examples of research that estimated transition matrix by relying on anchor points?", "answer": ["Classification with Noisy Labels by Importance Reweighting", "Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach", "Are Anchor Points Really Indispensable in Label-Noise Learning?", "Confident Learning: Estimating Uncertainty in Dataset Labels"], "answer_arxiv_id": ["1411.7718", "1609.03683", "1906.00189", "1911.00068"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_18317"} +{"question": "What paper proposed the use of an already trained NeRF in image registration?", "answer": ["INeRF: Inverting Neural Radiance Fields for Pose Estimation"], "answer_arxiv_id": ["2012.05877"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_18318"} +{"question": "Which works focused on the relation between observation x and latent z?", "answer": ["Info3D: Representation Learning on 3D Objects using Mutual Information Maximization and Contrastive Learning", "Learning Disentangled Representations and Group Structure of Dynamical Environments", "Equivariant Neural Rendering", "Spatio-temporal Self-Supervised Representation Learning for 3D Point Clouds"], "answer_arxiv_id": ["2006.02598", "2002.06991", "2006.07630", "2109.00179"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_18319"} +{"question": "What works have tried to achieve improvements in sign language translation by treating different cues in the input video differently?", "answer": ["Spatial-Temporal Multi-Cue Network for Continuous Sign Language Recognition", "Better Sign Language Translation with STMC-Transformer"], "answer_arxiv_id": ["2002.03187", "2004.00588"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_18320"} +{"question": "Which studies focused on the problem of catastrophic forgetting in Deep Neural Networks (DNNs)?", "answer": ["A continual learning survey: Defying forgetting in classification tasks"], "answer_arxiv_id": ["1909.08383"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_18321"} +{"question": "Which paper introduces the neural ordinary differential equations (NODEs)?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_18322"} +{"question": "What research used pretrained text encoder, such as CLIP, to encode the text prompt into a feature vector to condition the diffusion model?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10741", "2204.06125"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18323"} +{"question": "Which papers addressed the need to reduce the size of parameters and latency in language models?", "answer": ["ALBERT: A Lite BERT for Self-supervised Learning of Language Representations", "MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices"], "answer_arxiv_id": ["1909.11942", "2004.02984"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_18324"} +{"question": "Which papers attempt to generalize operations to symmetric spaces?", "answer": ["Fully-Connected Network on Noncompact Symmetric Space and Ridgelet Transform based on Helgason-Fourier Analysis"], "answer_arxiv_id": ["2203.01631"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_18325"} +{"question": "Which research papers focus on the one-stage paradigm regarding spatio-temporal video grounding without relying on pre-trained object detectors?", "answer": ["TubeDETR: Spatio-Temporal Video Grounding with Transformers", "Embracing Consistency: A One-Stage Approach for Spatio-Temporal Video\n Grounding"], "answer_arxiv_id": ["2203.16434", "2209.13306"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_18326"} +{"question": "What paper utilizes cringe loss for optimization?", "answer": ["Some things are more CRINGE than others: Iterative Preference\n Optimization with the Pairwise Cringe Loss"], "answer_arxiv_id": ["2312.16682"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_18327"} +{"question": "What works unify object detection and phrase grounding through a language-image pre-training model?", "answer": ["Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2112.03857"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_18328"} +{"question": "What works are about pose-based methods in ISLR?", "answer": ["Word-level Deep Sign Language Recognition from Video: A New Large-scale Dataset and Methods Comparison", "OpenHands: Making Sign Language Recognition Accessible with Pose-based Pretrained Models across Languages"], "answer_arxiv_id": ["1910.11006", "2110.05877v1"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_18329"} +{"question": "What references proposed fusion strategies to find the consensus representation among views?", "answer": ["Fast Multi-view Clustering via Ensembles: Towards Scalability,\n Superiority, and Simplicity"], "answer_arxiv_id": ["2203.11572"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_18330"} +{"question": "Which research papers have explored fingerprinting techniques in the Stable Diffusion model?", "answer": ["A Recipe for Watermarking Diffusion Models", "The Stable Signature: Rooting Watermarks in Latent Diffusion Models", "Tree-Ring Watermarks: Fingerprints for Diffusion Images that are\n Invisible and Robust"], "answer_arxiv_id": ["2303.10137", "2303.15435", "2305.20030"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_18331"} +{"question": "Which research proposed the mimic-exp algorithm?", "answer": ["Toward the Fundamental Limits of Imitation Learning"], "answer_arxiv_id": ["2009.05990"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18332"} +{"question": "Which paper presented an evolution in CBMs by removing the impact of confounding information to secure causality?", "answer": ["Debiasing Concept-based Explanations with Causal Analysis"], "answer_arxiv_id": ["2007.11500"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_18333"} +{"question": "Could you provide references that are based on discrete-time and continuous-time models such as VRNN, KVAE, and latentODE?", "answer": ["A Recurrent Latent Variable Model for Sequential Data", "A Disentangled Recognition and Nonlinear Dynamics Model for Unsupervised Learning"], "answer_arxiv_id": ["1506.02216", "1710.05741"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_18334"} +{"question": "What works have proposed aggregation schemes in Federated Learning?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization", "FedBN: Federated Learning on Non-IID Features via Local Batch Normalization"], "answer_arxiv_id": ["1602.05629", "2007.07481", "2102.07623"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_18335"} +{"question": "Could you provide me with some research papers that improve the sampling strategy for differentiable volume rendering?", "answer": ["DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields\n using Depth Oracle Networks", "TermiNeRF: Ray Termination Prediction for Efficient Neural Rendering", "HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling"], "answer_arxiv_id": ["2103.03231", "2111.03643", "2301.02238"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_18336"} +{"question": "Which papers discuss adversarial perturbations that can fool Deep Neural Networks (DNNs)?", "answer": ["Intriguing properties of neural networks", "Towards Evaluating the Robustness of Neural Networks", "Universal Adversarial Attack on Attention and the Resulting Dataset DAmageNet", "Understanding Black-box Predictions via Influence Functions", "Adversarial Examples Make Strong Poisons"], "answer_arxiv_id": ["1312.6199", "1608.04644", "2001.06325", "1703.04730", "2106.10807"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_18337"} +{"question": "What works propose the proportionally fair clustering based on points forming coalitions?", "answer": ["Proportionally Fair Clustering"], "answer_arxiv_id": ["1905.03674"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_18338"} +{"question": "Which papers demonstrate the effectiveness of the two-stage training strategy for transformer models?", "answer": ["RoBERTa: A Robustly Optimized BERT Pretraining Approach", "Well-Read Students Learn Better: On the Importance of Pre-training Compact Models"], "answer_arxiv_id": ["1907.11692", "1908.08962"], "source_meta": {"published_time": "20221102"}, "qid": "AutoScholarQuery_train_18339"} +{"question": "Could you refer me to a work that provides a theoretical explanation of the stability improvements offered by clipped SGD for (stochastic) first-order methods?", "answer": ["Why gradient clipping accelerates training: A theoretical justification for adaptivity"], "answer_arxiv_id": ["1905.11881"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_18340"} +{"question": "Which works have developed various Bayesian Optimization methods?", "answer": ["Entropy Search for Information-Efficient Global Optimization", "Predictive Entropy Search for Efficient Global Optimization of Black-box Functions"], "answer_arxiv_id": ["1112.1217v1", "1406.2541"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_18341"} +{"question": "Could you provide me some studies that have adopted model-based approaches for AED?", "answer": ["Unsupervised Label Noise Modeling and Loss Correction", "Dataset Cartography: Mapping and Diagnosing Datasets with Training\n Dynamics"], "answer_arxiv_id": ["1904.11238", "2009.10795"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18342"} +{"question": "What studies address training costs in the context of NeRF rendering?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2201.05989", "2112.05131"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_18343"} +{"question": "What works propose methods for optimizing architectures using gradient-based optimization?", "answer": ["DARTS: Differentiable Architecture Search", "Neural Architecture Search: A Survey"], "answer_arxiv_id": ["1806.09055", "1808.05377"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_18344"} +{"question": "What studies contribute to boosting diffusion-based T2I models to a larger scale?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2112.10752", "2211.01324"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_18345"} +{"question": "Are there any works that focus on fair representation learning as a pre-processing method?", "answer": ["Learning Adversarially Fair and Transferable Representations", "Flexibly Fair Representation Learning by Disentanglement", "Inherent Tradeoffs in Learning Fair Representations", "Towards Fairness in Visual Recognition: Effective Strategies for Bias\n Mitigation", "Contrastive Learning for Fair Representations", "Fairness via Representation Neutralization"], "answer_arxiv_id": ["1802.06309", "1906.02589", "1906.08386", "1911.11834", "2109.10645", "2106.12674"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_18346"} +{"question": "Could you provide me studies that recast the iterative slot refinement problem for improving the stability and computational cost of the slot-attention technique?", "answer": ["Object Representations as Fixed Points: Training Iterative Refinement Algorithms with Implicit Differentiation"], "answer_arxiv_id": ["2207.00787"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18347"} +{"question": "Could you provide me some studies that developed the optimistic maximum likelihood estimation (OMLE) algorithm?", "answer": ["When Is Partially Observable Reinforcement Learning Not Scary?", "Sample-Efficient Reinforcement Learning of Partially Observable Markov Games"], "answer_arxiv_id": ["2204.08967", "2206.01315"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_18348"} +{"question": "Could you provide me some works about the study of representations learned under class imbalance?", "answer": ["Decoupling Representation and Classifier for Long-Tailed Recognition", "Rethinking the Value of Labels for Improving Class-Imbalanced Learning", "Contrasting Contrastive Self-Supervised Representation Learning Pipelines", "Rethinking the Value of Labels for Improving Class-Imbalanced Learning"], "answer_arxiv_id": ["1910.09217", "2006.07529", "2103.14005", "2006.07529"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18349"} +{"question": "What works have been completed on the simultaneous update methods in game theory?", "answer": ["Training GANs with Optimism", "Optimistic Mirror Descent in Saddle-Point Problems: Going the Extra (Gradient) Mile", "The Numerics of GANs", "Competitive Gradient Descent", "Learning with Opponent-Learning Awareness", "Stabilizing Adversarial Nets With Prediction Methods", "The Mechanics of n-Player Differentiable Games"], "answer_arxiv_id": ["1711.00141", "1807.02629", "1705.10461", "1905.12103", "1709.04326", "1705.07364", "1802.05642v2"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_18350"} +{"question": "Any works conducted theoretical analysis and performance evaluation to identify the limitations of existing editing methods?", "answer": ["Does Localization Inform Editing? Surprising Differences in\n Causality-Based Localization vs. Knowledge Editing in Language Models", "Can LMs Learn New Entities from Descriptions? Challenges in Propagating\n Injected Knowledge"], "answer_arxiv_id": ["2301.04213", "2305.01651"], "source_meta": {"published_time": "20231223"}, "qid": "AutoScholarQuery_train_18351"} +{"question": "Which works describe image-based fashion editing methods that use text descriptions of styles?", "answer": ["Text2Human: Text-Driven Controllable Human Image Generation", "Multimodal Garment Designer: Human-Centric Latent Diffusion Models for\n Fashion Image Editing", "StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery", "TediGAN: Text-Guided Diverse Face Image Generation and Manipulation", "HyperStyle: StyleGAN Inversion with HyperNetworks for Real Image Editing"], "answer_arxiv_id": ["2205.15996", "2304.02051", "2103.17249", "2012.03308", "2111.15666"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_18352"} +{"question": "What studies illustrate the advantages of hybrid representations over singular MLP structures in the context of NeRF-centric approaches?", "answer": ["NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM", "Vox-Fusion: Dense Tracking and Mapping with Voxel-based Neural Implicit\n Representation", "ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of\n Signed Distance Fields", "Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural\n Real-Time SLAM"], "answer_arxiv_id": ["2302.03594", "2210.15858", "2211.11704", "2304.14377"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_18353"} +{"question": "Which works replace the simple TDNNs in the encoder with powerful ECAPA-TDNN or ResNet series models when designing speaker recognition methods?", "answer": ["ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verification", "MFA: TDNN with Multi-scale Frequency-channel Attention for Text-independent Speaker Verification with Short Utterances", "Integrating Frequency Translational Invariance in TDNNs and Frequency Positional Information in 2D ResNets to Enhance Speaker Verification", "Frequency and Multi-Scale Selective Kernel Attention for Speaker Verification", "PL-EESR: PERCEPTUAL LOSS BASED END-TO-END ROBUST SPEAKER REPRESENTATION EXTRACTION", "ResNeXt and Res2Net Structures for Speaker Verification", "Attention Back-end for Automatic Speaker Verification with Multiple Enrollment Utterances", "Utterance-level aggregation for speaker recognition in the wild"], "answer_arxiv_id": ["2005.07143", "2202.01624", "2104.02370", "2204.01005", "2110.00940", "2007.02480", "2104.01541", "1902.10107"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_18354"} +{"question": "Which paper first proposed to represent crystals with radius graphs and used a graph convolutional network for property prediction?", "answer": ["Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties"], "answer_arxiv_id": ["1710.10324"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_18355"} +{"question": "Which works proposed automatic program repair to alleviate the effort of debugging?", "answer": ["Learning to Represent Programs with Graphs", "Neural Program Repair by Jointly Learning to Localize and Repair", "HEAT: Hyperedge Attention Networks", "SequenceR: Sequence-to-Sequence Learning for End-to-End Program Repair", "Graph-based, Self-Supervised Program Repair from Diagnostic Feedback", "Break-It-Fix-It: Unsupervised Learning for Program Repair"], "answer_arxiv_id": ["1711.00740", "1904.01720", "2201.12113", "1901.01808", "2005.10636", "2106.06600"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_18356"} +{"question": "Could you provide me some studies about finding dense correspondences employing transformer architectures in scene flow estimation?", "answer": ["CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registration", "Geometric Transformer for Fast and Robust Point Cloud Registration"], "answer_arxiv_id": ["2110.14076", "2202.06688"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_18357"} +{"question": "What are the studies that have attempted molecule generation tasks by generating sequential representations of molecules - SMILES?", "answer": ["Grammar Variational Autoencoder", "Syntax-Directed Variational Autoencoder for Structured Data"], "answer_arxiv_id": ["1703.01925", "1802.08786"], "source_meta": {"published_time": "20230505"}, "qid": "AutoScholarQuery_train_18358"} +{"question": "What research works have studied experimental design and analysis for individual treatment rules (ITR)?", "answer": ["Experimental Evaluation of Individualized Treatment Rules", "Policy Learning with Observational Data"], "answer_arxiv_id": ["1905.05389v6", "1702.02896"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_18359"} +{"question": "Could you list the papers addressing the challenge of discrimination learning under label scarcity in unsupervised domain adaptation?", "answer": ["Dynamic Weighted Learning for Unsupervised Domain Adaptation", "Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation"], "answer_arxiv_id": ["2103.13814", "2002.08546"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18360"} +{"question": "Are there any works making use of model-generated instruction-augmented data?", "answer": ["Unnatural Instructions: Tuning Language Models with (Almost) No Human Labor", "WizardLM: Empowering Large Language Models to Follow Complex Instructions"], "answer_arxiv_id": ["2212.09689", "2304.12244"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_18361"} +{"question": "Which paper compared zeroth- and first-order methods and discussed their pros and cons?", "answer": ["Do Differentiable Simulators Give Better Policy Gradients?"], "answer_arxiv_id": ["2202.00817"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_18362"} +{"question": "Which studies focus on using regularizations or constraints in learning the latent space for different tasks?", "answer": ["Improving black-box optimization in VAE latent space using decoder uncertainty", "Multi-Objective Latent Space Optimization of Generative Molecular Design Models", "Contrastive Adaptation Network for Unsupervised Domain Adaptation", "Domain Adaptation by Class Centroid Matching and Local Manifold Self-Learning", "Latent Space Regularization for Unsupervised Domain Adaptation in Semantic Segmentation", "Continual Semantic Segmentation via Repulsion-Attraction of Sparse and Disentangled Latent Representations", "Consistency Regularization for Variational Auto-Encoders", "Learning Smooth Neural Functions via Lipschitz Regularization", "Reduced Representation of Deformation Fields for Effective Non-rigid Shape Matching", "Towards Robust Bisimulation Metric Learning"], "answer_arxiv_id": ["2107.00096", "2203.00526", "1901.00976", "2003.09391", "2104.02633", "2103.06342", "2105.14859", "2202.08345", "2211.14604", "2110.14096"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_18363"} +{"question": "Which research introduced the Bayesian framework for ICL?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference"], "answer_arxiv_id": ["2111.02080"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_18364"} +{"question": "Which papers use the same technique of re-parameterization of a convolution layer like LiResNet block?", "answer": ["DiracNets: Training Very Deep Neural Networks Without Skip-Connections", "ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks", "RepVGG: Making VGG-style ConvNets Great Again", "RepMLPNet: Hierarchical Vision MLP with Re-parameterized Locality"], "answer_arxiv_id": ["1706.00388", "1908.03930", "2101.03697", "2112.11081"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_18365"} +{"question": "Which papers leverage RNN or adopt auto-regression for camera pose estimation in sparse camera view settings?", "answer": ["DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras", "DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent\n Convolutional Neural Networks", "D3VO: Deep Depth, Deep Pose and Deep Uncertainty for Monocular Visual\n Odometry"], "answer_arxiv_id": ["2108.10869", "1709.08429", "2003.01060"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_18366"} +{"question": "Could you indicate some studies that dealt with the forward uncertainty quantification case where the model performance is characterized when the inputs are perturbed?", "answer": ["The Principles of Deep Learning Theory"], "answer_arxiv_id": ["2106.10165"], "source_meta": {"published_time": "20230219"}, "qid": "AutoScholarQuery_train_18367"} +{"question": "Could you mention some studies about a simple contrastive learning framework through dropout augmentation?", "answer": ["SimCSE: Simple Contrastive Learning of Sentence Embeddings"], "answer_arxiv_id": ["2104.08821"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_18368"} +{"question": "Which works dive into prevalent classes of algorithms for non-convex constrained optimization?", "answer": ["ADMM for Multiaffine Constrained Optimization", "A Proximal Alternating Direction Method of Multiplier for Linearly Constrained Nonconvex Minimization", "A global dual error bound and its application to the analysis of linearly constrained nonconvex optimization", "On the Iteration Complexity of Smoothed Proximal ALM for Nonconvex Optimization Problem with Convex Constraints", "Convergence Analysis of Alternating Direction Method of Multipliers for a Family of Nonconvex Problems", "Complexity of a quadratic penalty accelerated inexact proximal point method for solving linearly constrained nonconvex composite programs", "Iteration-complexity of an inexact proximal accelerated augmented Lagrangian method for solving linearly constrained smooth nonconvex composite optimization problems", "Iteration-complexity of an inner accelerated inexact proximal augmented Lagrangian method based on the classical Lagrangian function", "An Accelerated Inexact Dampened Augmented Lagrangian Method for Linearly-Constrained Nonconvex Composite Optimization Problems", "An adaptive superfast inexact proximal augmented Lagrangian method for smooth nonconvex composite optimization problems", "Augmented Lagrangian based first-order methods for convex-constrained programs with weakly-convex objective", "An Inexact Augmented Lagrangian Framework for Nonconvex Optimization with Nonlinear Constraints", "Rate-improved Inexact Augmented Lagrangian Method for Constrained Nonconvex Optimization", "Zeroth-order Optimization for Composite Problems with Functional Constraints", "Majorization-minimization procedures and convergence of SQP methods for semi-algebraic and tame programs", "Ghost Penalties in Nonconvex Constrained Optimization: Diminishing Stepsizes and Iteration Complexity", "Level Constrained First Order Methods for Function Constrained Optimization", "Inexact Sequential Quadratic Optimization for Minimizing a Stochastic Objective Function Subject to Deterministic Nonlinear Equality Constraints", "A Sequential Quadratic Programming Method with High Probability Complexity Bounds for Nonlinear Equality Constrained Stochastic Optimization", "Sequential Quadratic Optimization for Nonlinear Equality Constrained Stochastic Optimization", "A Stochastic Sequential Quadratic Optimization Algorithm for Nonlinear Equality Constrained Optimization with Rank-Deficient Jacobians", "Worst-Case Complexity of an SQP Method for Nonlinear Equality Constrained Stochastic Optimization"], "answer_arxiv_id": ["1802.09592", "1812.10229v4", "2006.16440v1", "2207.06304v3", "1410.1390", "1802.03504", "2006.08048", "2008.00562", "2110.11151v3", "2207.11905", "2003.08880", "1906.11357", "2007.01284v2", "2112.11420", "1409.8147", "1709.03384v3", "2205.08011v2", "2107.03512", "2301.00477", "2007.10525", "2106.13015", "2112.14799"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_18369"} +{"question": "Which works proposed to analyze the layout of document images?", "answer": ["PubLayNet: largest dataset ever for document layout analysis"], "answer_arxiv_id": ["1908.07836"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_18370"} +{"question": "What papers discussed about explore-then-commit algorithms?", "answer": ["On Explore-Then-Commit Strategies", "Double Explore-then-Commit: Asymptotic Optimality and Beyond", "Risk-Averse Explore-Then-Commit Algorithms for Finite-Time Bandits"], "answer_arxiv_id": ["1605.08988", "2002.09174", "1904.13387"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_18371"} +{"question": "What studies applied entropy minimization to improve the video INR compression?", "answer": ["NIRVANA: Neural Implicit Representations of Videos with Adaptive\n Networks and Autoregressive Patch-wise Modeling"], "answer_arxiv_id": ["2212.14593"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_18372"} +{"question": "Which works discuss the use of orthogonality constraints in recurrent neural networks to avoid vanishing and exploding gradients problem?", "answer": ["Unitary Evolution Recurrent Neural Networks", "Cheap Orthogonal Constraints in Neural Networks: A Simple Parametrization of the Orthogonal and Unitary Group", "Fast and accurate optimization on the orthogonal manifold without retraction", "projUNN: efficient method for training deep networks with unitary matrices"], "answer_arxiv_id": ["1511.06464", "1901.08428", "2102.07432v2", "2203.05483"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_18373"} +{"question": "Which works concern scientific sentiment summarization in the field of meta-review generation?", "answer": ["Summarizing Multiple Documents with Conversational Structure for\n Meta-Review Generation"], "answer_arxiv_id": ["2305.01498"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_18374"} +{"question": "Which studies approach the development of a single model trained across diverse datasets and tasks in universal image/video segmentation?", "answer": ["OneFormer: One Transformer to Rule Universal Image Segmentation", "RAP-SAM: Towards Real-Time All-Purpose Segment Anything", "Universal Instance Perception as Object Discovery and Retrieval", "DaTaSeg: Taming a Universal Multi-Dataset Multi-Task Segmentation Model"], "answer_arxiv_id": ["2211.06220", "2401.10228", "2303.06674", "2306.01736"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_18375"} +{"question": "What papers extend the vanilla Decision-Estimation Coefficient (DEC) by incorporating an optimistic modification?", "answer": ["Unified Algorithms for RL with Decision-Estimation Coefficients: No-Regret, PAC, and Reward-Free Learning"], "answer_arxiv_id": ["2209.11745"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18376"} +{"question": "What paper discusses differences and similarities with SAR in terms of entropy?", "answer": ["Efficient Test-Time Model Adaptation without Forgetting"], "answer_arxiv_id": ["2204.02610"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_18377"} +{"question": "What literature can provide an analysis of the tension between group fairness measures and model performance metrics?", "answer": ["Inherent Trade-Offs in the Fair Determination of Risk Scores", "Fair prediction with disparate impact: A study of bias in recidivism prediction instruments", "Algorithmic decision making and the cost of fairness", "Why Is My Classifier Discriminatory?", "Is There a Trade-Off Between Fairness and Accuracy? A Perspective Using Mismatched Hypothesis Testing", "To Split or Not to Split: The Impact of Disparate Treatment in Classification"], "answer_arxiv_id": ["1609.05807", "1610.07524", "1701.08230", "1805.12002", "1910.07870", "2002.04788"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_18378"} +{"question": "What works architect benchmarks for tool utilization, response comparison, in particular?", "answer": ["ToolQA: A Dataset for LLM Question Answering with External Tools"], "answer_arxiv_id": ["2306.13304"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18379"} +{"question": "Which study introduces a natural data-dependent parameter that can limit the complexity of data for logistic and probit regression?", "answer": ["On Coresets for Logistic Regression"], "answer_arxiv_id": ["1805.08571"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_18380"} +{"question": "Can you tell me some prior works focusing on the point cloud understanding?", "answer": ["Volumetric and Multi-View CNNs for Object Classification on 3D Data", "4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks", "3D ShapeNets: A Deep Representation for Volumetric Shapes", "OctNet: Learning Deep 3D Representations at High Resolutions", "Octree guided CNN with Spherical Kernels for 3D Point Clouds", "3D Semantic Segmentation with Submanifold Sparse Convolutional Networks", "Multi-View 3D Object Detection Network for Autonomous Driving", "Multi-view Convolutional Neural Networks for 3D Shape Recognition", "PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space", "KPConv: Flexible and Deformable Convolution for Point Clouds", "PointConv: Deep Convolutional Networks on 3D Point Clouds", "Dynamic Graph CNN for Learning on Point Clouds", "PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Clouds", "RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds"], "answer_arxiv_id": ["1604.03265", "1904.08755", "1406.5670v3", "1611.05009", "1903.00343", "1711.10275", "1611.07759", "1505.00880", "1612.00593", "1706.02413", "1904.08889", "1811.07246", "1801.07829", "2103.14635", "1911.11236"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_18381"} +{"question": "Could you provide me some studies that contributed to proposing methods for feature aggregation in multi-modal vehicle ReID?", "answer": ["Multi-spectral Vehicle Re-identification with Cross-directional Consistency Network and a High-quality Benchmark"], "answer_arxiv_id": ["2208.00632v2"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_18382"} +{"question": "Did any research study the causes for the lack of generalization performance?", "answer": ["Evaluating the Generalization Ability of Super-Resolution Networks"], "answer_arxiv_id": ["2205.07019"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_18383"} +{"question": "Which studies introduce open-source LLMs, including the LLaMA series, Vicuna, InternLM, MOSS, ChatGLM, Qwen, Baichuan, and Falcon?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "Llama 2: Open Foundation and Fine-Tuned Chat Models", "Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena", "GLM: General Language Model Pretraining with Autoregressive Blank\n Infilling", "Qwen Technical Report", "Baichuan 2: Open Large-scale Language Models", "The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora\n with Web Data, and Web Data Only", "Skywork: A More Open Bilingual Foundation Model", "Efficient and Effective Text Encoding for Chinese LLaMA and Alpaca"], "answer_arxiv_id": ["2302.13971", "2307.09288", "2306.05685", "2103.10360", "2309.16609v1", "2309.10305", "2306.01116", "2310.19341", "2304.08177"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18384"} +{"question": "What papers have contributed to the datasets for ground-only camera networks in View-homogeneous ReID?", "answer": ["Person Transfer GAN to Bridge Domain Gap for Person Re-Identification"], "answer_arxiv_id": ["1711.08565"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_18385"} +{"question": "Could you cite some studies on activation quantization?", "answer": ["ActNN: Reducing Training Memory Footprint via 2-Bit Activation Compressed Training", "GACT: Activation Compressed Training for Generic Network Architectures", "BED: A Real-Time Object Detection System for Edge Devices"], "answer_arxiv_id": ["2104.14129", "2206.11357", "2202.07503"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_18386"} +{"question": "Which works investigated alternative labeling methods for existing benchmarks?", "answer": ["Agreeing to Disagree: Annotating Offensive Language Datasets with Annotators’ Disagreement", "Dealing with Disagreements: Looking Beyond the Majority Vote in Subjective Annotations"], "answer_arxiv_id": ["2109.13563", "2110.05719"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18387"} +{"question": "What works introduce benchmark datasets to examine aspects of the robustness of ME?", "answer": ["Unveiling the Pitfalls of Knowledge Editing for Large Language Models", "Evaluating the Ripple Effects of Knowledge Editing in Language Models"], "answer_arxiv_id": ["2310.02129", "2307.12976"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_18388"} +{"question": "What studies proposed methods to mitigate the consequences of over-smoothing?", "answer": ["DeepViT: Towards Deeper Vision Transformer", "Anti-Oversmoothing in Deep Vision Transformers via the Fourier Domain Analysis: From Theory to Practice"], "answer_arxiv_id": ["2103.11886", "2203.05962"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_18389"} +{"question": "In which paper is LoRA, a parameter-efficient fine-tuning (PEFT) method with competitive trade-offs between performance and parameter efficiency, introduced?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2106.09685"], "source_meta": {"published_time": "20240224"}, "qid": "AutoScholarQuery_train_18390"} +{"question": "What are the studies using Transformer in sequence-to-sequence modeling for land-cover mapping?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "TransUNet: Transformers Make Strong Encoders for Medical Image\n Segmentation", "UNetFormer: A UNet-like Transformer for Efficient Semantic Segmentation of Remote Sensing Urban Scene Imagery"], "answer_arxiv_id": ["2103.14030", "2102.04306", "2109.08937v4"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_18391"} +{"question": "Which models have been tested for adversarial robustness and hallucination according to the all-round benchmarks?", "answer": ["On the Hidden Mystery of OCR in Large Multimodal Models", "On Evaluating Adversarial Robustness of Large Vision-Language Models", "Holistic Analysis of Hallucination in GPT-4V(ision): Bias and\n Interference Challenges", "Evaluating Object Hallucination in Large Vision-Language Models", "Evaluation and Analysis of Hallucination in Large Vision-Language Models"], "answer_arxiv_id": ["2305.07895", "2305.16934", "2311.03287", "2305.10355", "2308.15126"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_18392"} +{"question": "Which paper proposed an external classifier to identify regions of low visual fidelity in individual generations and to detect internal units associated with those regions?", "answer": ["Automatic Correction of Internal Units in Generative Neural Networks"], "answer_arxiv_id": ["2104.06118"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18393"} +{"question": "Which research papers have demonstrated improved image classification due to the siamese network approach?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1911.05722", "2002.05709"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_18394"} +{"question": "What works are about the label distribution method and its relevance between labels?", "answer": ["Deep Differentiable Random Forests for Age Estimation", "Deep Differentiable Random Forests for Age Estimation"], "answer_arxiv_id": ["1907.10665", "1907.10665"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_18395"} +{"question": "Could you list some works focusing on robust semi-supervised learning where the distributions of unlabeled and labeled data shift?", "answer": ["Realistic Evaluation of Deep Semi-Supervised Learning Algorithms", "OpenCoS: Contrastive Semi-supervised Learning for Handling Open-set Unlabeled Data", "RoPAWS: Robust Semi-supervised Representation Learning from Uncurated Data"], "answer_arxiv_id": ["1804.09170", "2107.08943", "2302.14483"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_18396"} +{"question": "What are some works that incorporate hand-crafted or learnable latent variables to improve inter-token dependency in non-autoregressive sequence generation?", "answer": ["Guiding Non-Autoregressive Neural Machine Translation Decoding with\n Reordering Information", "Syntactically Supervised Transformers for Faster Neural Machine\n Translation", "Latent-Variable Non-Autoregressive Neural Machine Translation with\n Deterministic Inference Using a Delta Posterior", "Fast Decoding in Sequence Models using Discrete Latent Variables", "$\\textit{latent}$-GLAT: Glancing at Latent Variables for Parallel Text\n Generation"], "answer_arxiv_id": ["1911.02215", "1906.02780", "1908.07181", "1803.03382", "2204.02030"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18397"} +{"question": "What works on the integration mechanism of visual features into the LLM used a cross-attention-based transformer model?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning"], "answer_arxiv_id": ["2301.12597", "2305.06500"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_18398"} +{"question": "Could you list some studies about approximating the posterior of the weights in Bayesian modeling?", "answer": ["Markov Chain Monte Carlo and Variational Inference: Bridging the Gap", "Multiplicative Normalizing Flows for Variational Bayesian Neural Networks", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1410.6460", "1703.01961v2", "1506.02142"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_18399"} +{"question": "What are some research papers that focus on finding static sparse networks before training?", "answer": ["SNIP: Single-shot Network Pruning based on Connection Sensitivity", "Picking Winning Tickets Before Training by Preserving Gradient Flow"], "answer_arxiv_id": ["1810.02340", "2002.07376"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_18400"} +{"question": "Which research showed that an ensemble of teachers trained with noisy labels can be used to label a new unlabeled dataset for training a student?", "answer": ["Knowledge Distillation: Bad Models Can Be Good Role Models"], "answer_arxiv_id": ["2203.14649v1"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_18401"} +{"question": "Which studies focus on how views of different samples are connected by the augmentation process in contrastive learning?", "answer": ["Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss"], "answer_arxiv_id": ["2106.04156"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_18402"} +{"question": "Can you mention studies that propose a method for distribution shift explanation based on transport mappings?", "answer": ["Towards Explaining Distribution Shifts"], "answer_arxiv_id": ["2210.10275"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_18403"} +{"question": "Which works developed an algorithm that does not rely on the prior knowledge of T∗ without assuming cmin>0?", "answer": ["Stochastic Shortest Path: Minimax, Parameter-Free and Towards Horizon-Free Regret"], "answer_arxiv_id": ["2104.11186"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_18404"} +{"question": "What works emphasize pretraining on synthetic and real data for enhanced depth estimation?", "answer": ["The Surprising Effectiveness of Diffusion Models for Optical Flow and\n Monocular Depth Estimation"], "answer_arxiv_id": ["2306.01923"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18405"} +{"question": "Could you provide me some works that focused on the intersection of algorithmic fairness and policy learning?", "answer": ["Learning Optimal Fair Policies", "Fair Policy Targeting"], "answer_arxiv_id": ["1809.02244", "2005.12395v3"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_18406"} +{"question": "What are the papers that detail how error-minimization, gradient alignment and influence functions can help protect data?", "answer": ["Unlearnable Examples: Making Personal Data Unexploitable", "Preventing Unauthorized Use of Proprietary Data: Poisoning for Secure Dataset Release", "Influence Function based Data Poisoning Attacks to Top-N Recommender Systems"], "answer_arxiv_id": ["2101.04898", "2103.02683", "2002.08025"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_18407"} +{"question": "Which works implement generative neural fields with the concept of hyper-network?", "answer": ["Learning Signal-Agnostic Manifolds of Neural Fields", "Generative Models as Distributions of Functions"], "answer_arxiv_id": ["2111.06387", "2102.04776"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18408"} +{"question": "What papers mention the synthetic scenario databases that are handcrafted?", "answer": ["SafeBench: A Benchmarking Platform for Safety Evaluation of Autonomous Vehicles", "Scenic: A Language for Scenario Specification and Scene Generation", "OpenCDA: An Open Cooperative Driving Automation Framework Integrated with Co-Simulation"], "answer_arxiv_id": ["2206.09682", "1809.09310", "2107.06260"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_18409"} +{"question": "Can you show me a work that introduces qualitative assessment of natural language explanations?", "answer": ["e-SNLI: Natural Language Inference with Natural Language Explanations"], "answer_arxiv_id": ["1812.01193"], "source_meta": {"published_time": "20221215"}, "qid": "AutoScholarQuery_train_18410"} +{"question": "Which study established the first sublinear cumulative regret bounds of a global BO algorithm, GP-UCB, in the noisy setting?", "answer": ["Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design"], "answer_arxiv_id": ["0912.3995v4"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_18411"} +{"question": "Could you provide me some works that leveraged the sparse nature of 3D scenes in high-resolution 3D-aware GANs?", "answer": ["GRAM: Generative Radiance Manifolds for 3D-Aware Image Generation", "GRAM-HD: 3D-Consistent Image Generation at High Resolution with\n Generative Radiance Manifolds", "Generative Multiplane Images: Making a 2D GAN 3D-Aware", "VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids"], "answer_arxiv_id": ["2112.08867", "2206.07255", "2207.10642", "2206.07695"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_18412"} +{"question": "What papers were significant in the literature discussing the training of abstaining classifiers?", "answer": ["Selective Classification for Deep Neural Networks", "SelectiveNet: A Deep Neural Network with an Integrated Reject Option", "Selective Classification via One-Sided Prediction"], "answer_arxiv_id": ["1705.08500", "1901.09192", "2010.07853"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_18413"} +{"question": "What papers have studied set generation in the context of n-gram generation?", "answer": ["Order Matters: Sequence to sequence for sets"], "answer_arxiv_id": ["1511.06391"], "source_meta": {"published_time": "20230311"}, "qid": "AutoScholarQuery_train_18414"} +{"question": "Are there any works that provide robust results for datasets lacking exploration without relying on dataset coverage assumptions?", "answer": ["Information-Theoretic Considerations in Batch Reinforcement Learning"], "answer_arxiv_id": ["1905.00360"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_18415"} +{"question": "Could you provide me some studies about supervised methods in monocular depth estimation?", "answer": ["Depth Map Prediction from a Single Image using a Multi-Scale Deep\n Network", "AdaBins: Depth Estimation using Adaptive Bins", "Deep Ordinal Regression Network for Monocular Depth Estimation", "LocalBins: Improving Depth Estimation by Learning Local Distributions", "NeW CRFs: Neural Window Fully-connected CRFs for Monocular Depth\n Estimation", "Vision Transformers for Dense Prediction", "EdgeConv with Attention Module for Monocular Depth Estimation", "DeepV2D: Video to Depth with Differentiable Structure from Motion"], "answer_arxiv_id": ["1406.2283", "2011.14141", "1806.02446", "2203.15132", "2203.01502", "2103.13413", "2106.08615", "1812.04605"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_18416"} +{"question": "What papers use density-based methods in traditional object counting?", "answer": ["Deep Rank-Consistent Pyramid Model for Enhanced Crowd Counting"], "answer_arxiv_id": ["2201.04819"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_18417"} +{"question": "Are there any works that extend multi-view diffusion to produce non-square panorama images?", "answer": ["MVDiffusion: Enabling Holistic Multi-view Image Generation with Correspondence-Aware Diffusion"], "answer_arxiv_id": ["2307.01097"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18418"} +{"question": "Which works have utilized video data for self-supervised learning?", "answer": ["MERLOT: Multimodal Neural Script Knowledge Models", "MERLOT Reserve: Neural Script Knowledge through Vision and Language and Sound", "Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos", "Connecting Vision and Language with Video Localized Narratives", "Connecting Vision and Language with Localized Narratives"], "answer_arxiv_id": ["2106.02636", "2201.02639", "2206.11795", "2302.11217", "1912.03098v4"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_18419"} +{"question": "Which studies mentioned conceptual problems by remembering training data not only unique to adversarial training?", "answer": ["Graph Robustness Benchmark: Benchmarking the Adversarial Robustness of Graph Machine Learning", "Randomized Message-Interception Smoothing: Gray-box Certificates for Graph Neural Networks"], "answer_arxiv_id": ["2111.04314", "2301.02039"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_18420"} +{"question": "What studies have expounded on the use of PGD attack as an adversarial attack method?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks"], "answer_arxiv_id": ["1706.06083", "2003.01690"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_18421"} +{"question": "What studies used fixed routing strategies for more stable routing and training in MoE models?", "answer": ["Hash Layers For Large Sparse Models", "StableMoE: Stable Routing Strategy for Mixture of Experts"], "answer_arxiv_id": ["2106.04426", "2204.08396"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_18422"} +{"question": "Which work expands the single-flow diffusion pipeline into a multi-task multimodal network?", "answer": ["Versatile Diffusion: Text, Images and Variations All in One Diffusion Model"], "answer_arxiv_id": ["2211.08332"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_18423"} +{"question": "Could you provide me some studies about gradient-based methods for generating class-aware saliency maps for CNNs?", "answer": ["Striving for Simplicity: The All Convolutional Net", "SmoothGrad: removing noise by adding noise", "Axiomatic Attribution for Deep Networks"], "answer_arxiv_id": ["1412.6806", "1706.03825", "1703.01365"], "source_meta": {"published_time": "20231015"}, "qid": "AutoScholarQuery_train_18424"} +{"question": "Could you provide me some studies about 3D segmentation using grid space such as voxels?", "answer": ["Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution", "Point-Voxel CNN for Efficient 3D Deep Learning"], "answer_arxiv_id": ["2007.16100", "1907.03739"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_18425"} +{"question": "Which studies discuss the emergence of in-context learning through implicit gradient descent?", "answer": ["Transformers Learn In-Context by Gradient Descent", "What learning algorithm is in-context learning? Investigations with linear models"], "answer_arxiv_id": ["2212.07677", "2211.15661"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18426"} +{"question": "Could you provide me some studies that concentrate on attacking language models within the input space by identifying adversarial prompts?", "answer": ["Universal and Transferable Adversarial Attacks on Aligned Language\n Models", "\"Do Anything Now\": Characterizing and Evaluating In-The-Wild Jailbreak\n Prompts on Large Language Models", "AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language\n Models", "DeepInception: Hypnotize Large Language Model to Be Jailbreaker", "Jailbreaking Black Box Large Language Models in Twenty Queries"], "answer_arxiv_id": ["2307.15043", "2308.03825", "2310.04451", "2311.03191", "2310.08419"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18427"} +{"question": "What work uses MLP to process sampled points on the ray and then apply a transformer to predict density?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering"], "answer_arxiv_id": ["2102.13090"], "source_meta": {"published_time": "20220727"}, "qid": "AutoScholarQuery_train_18428"} +{"question": "Are there any other techniques proposed to tackle the problem of unpaired image-to-image translation?", "answer": ["Learning Cycle-Consistent Cooperative Networks via Alternating MCMC Teaching for Unsupervised Cross-Domain Translation", "Cooperative Training of Descriptor and Generator Networks", "A Theory of Generative ConvNet"], "answer_arxiv_id": ["2103.04285", "1609.09408v3", "1602.03264v3"], "source_meta": {"published_time": "20230804"}, "qid": "AutoScholarQuery_train_18429"} +{"question": "Are there any research papers that have predicted the error of a classifier under distribution shift with unlabeled test data?", "answer": ["Are Labels Always Necessary for Classifier Accuracy Evaluation?", "What Does Rotation Prediction Tell Us about Classifier Accuracy under Varying Testing Environments?", "Detecting Errors and Estimating Accuracy on Unlabeled Data with Self-training Ensembles", "Mandoline: Model Evaluation under Distribution Shift", "Predicting with Confidence on Unseen Distributions", "Estimating Generalization under Distribution Shifts via Domain-Invariant Representations", "Leveraging Unlabeled Data to Predict Out-of-Distribution Performance", "Agreement-on-the-Line: Predicting the Performance of Neural Networks under Distribution Shift"], "answer_arxiv_id": ["2007.02915", "2106.05961", "2106.15728", "2107.00643", "2107.03315", "2007.03511", "2201.04234", "2206.13089"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_18430"} +{"question": "Could you provide me some works that considered the robustness to the uncertainty of the action in reinforcement learning?", "answer": ["Action Robust Reinforcement Learning and Applications in Continuous Control", "Robustifying Reinforcement Learning Agents via Action Space Adversarial Training"], "answer_arxiv_id": ["1901.09184", "2007.07176"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18431"} +{"question": "Which works introduced diffusion models as a class of generative models in image synthesis?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_18432"} +{"question": "Can you give examples of research that developed methods for precise editing areas localization?", "answer": ["Text2LIVE: Text-Driven Layered Image and Video Editing", "FEAT: Face Editing with Attention", "CoralStyleCLIP: Co-optimized Region and Layer Selection for Image\n Editing", "DiffEdit: Diffusion-based semantic image editing with mask guidance", "Watch Your Steps: Local Image and Scene Editing by Text Instructions", "InstructEdit: Improving Automatic Masks for Diffusion-based Image\n Editing With User Instructions", "Object-aware Inversion and Reassembly for Image Editing", "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set\n Object Detection", "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "Localizing Object-level Shape Variations with Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2204.02491", "2202.02713", "2303.05031", "2210.11427", "2308.08947", "2305.18047", "2310.12149", "2303.05499", "2401.14159", "2303.11306"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_18433"} +{"question": "Which works extended the Probably Approximately Correct learning framework to distribution-dependent bounds?", "answer": ["Distribution-Dependent Sample Complexity of Large Margin Learning"], "answer_arxiv_id": ["1204.1276"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_18434"} +{"question": "What works discuss the output type of multi-trajectory in motion prediction?", "answer": ["Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks"], "answer_arxiv_id": ["1803.10892"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_18435"} +{"question": "Which papers address the task of automatically finding a program satisfying user intent?", "answer": ["Sparks of Artificial General Intelligence: Early experiments with GPT-4", "InCoder: A Generative Model for Code Infilling and Synthesis"], "answer_arxiv_id": ["2303.12712", "2204.05999"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18436"} +{"question": "What researches propose the idea that most of the knowledge in large language models is learned during pre-training?", "answer": ["An Explanation of In-context Learning as Implicit Bayesian Inference", "Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "LIMA: Less Is More for Alignment"], "answer_arxiv_id": ["2111.02080", "2202.12837", "2305.11206"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_18437"} +{"question": "Which works have utilized RNN for generating intensities in the context of NPP models?", "answer": ["The Neural Hawkes Process: A Neurally Self-Modulating Multivariate Point Process"], "answer_arxiv_id": ["1612.09328"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_18438"} +{"question": "Which works have been conducted on the explainability in QA systems?", "answer": ["A Survey on Explainability in Machine Reading Comprehension", "Teach Me to Explain: A Review of Datasets for Explainable Natural\n Language Processing", "QED: A Framework and Dataset for Explanations in Question Answering"], "answer_arxiv_id": ["2010.00389", "2102.12060", "2009.06354"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_18439"} +{"question": "Which papers present model-based imitation learning techniques that can plan to go back in distribution when the agent visits out of distribution states?", "answer": ["Imitating, Fast and Slow: Robust learning from demonstrations via decision-time planning"], "answer_arxiv_id": ["2204.03597"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_18440"} +{"question": "What are the works that focus on the gradient conflict problem by proposing specific approaches?", "answer": ["Gradient Surgery for Multi-Task Learning", "Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout", "RotoGrad: Gradient Homogenization in Multitask Learning", "Conflict-Averse Gradient Descent for Multi-task Learning"], "answer_arxiv_id": ["2001.06782", "2010.06808", "2103.02631", "2110.14048"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_18441"} +{"question": "Could you provide me some studies about mixup, a method that uses a convex sum of images and their labels?", "answer": ["mixup: Beyond Empirical Risk Minimization"], "answer_arxiv_id": ["1710.09412"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_18442"} +{"question": "Could you provide me the works which explained that a min-max loss improves generalization in distributionally robust learning (DRL)?", "answer": ["Learning Models with Uniform Performance via Distributionally Robust Optimization", "Minimizing the Maximal Loss: How and Why"], "answer_arxiv_id": ["1810.08750", "1602.01690"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_18443"} +{"question": "Are there studies that have used LayerNorm to enhance performance in online and offline reinforcement learning?", "answer": ["Q-Ensemble for Offline RL: Don’t Scale the Ensemble, Scale the Batch Size", "Efficient Online Reinforcement Learning with Offline Data", "Improving and Benchmarking Offline Reinforcement Learning Algorithms", "Offline Q-Learning on Diverse Multi-Task Data Both Scales And Generalizes"], "answer_arxiv_id": ["2211.11092", "2302.02948", "2306.00972", "2211.15144"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_18444"} +{"question": "Could you provide me some papers about the goal of meta-learning to adapt quickly to unseen tasks with a few examples?", "answer": ["Matching Networks for One Shot Learning", "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-Learning in Neural Networks: A Survey"], "answer_arxiv_id": ["1606.04080", "1703.03400", "2004.05439"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_18445"} +{"question": "What research introduced the red teaming dataset?", "answer": ["Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned"], "answer_arxiv_id": ["2209.07858"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_18446"} +{"question": "Could you tell me about papers that discussed reinforcement learning applied to interact with discrete memory interfaces in memory-augmented networks?", "answer": ["Learning Simple Algorithms from Examples"], "answer_arxiv_id": ["1511.07275"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_18447"} +{"question": "What studies have explored the traditional message-passing GNN architectures like GCN and GAT using the binary contextual stochastic block model?", "answer": ["Graph Convolution for Semi-Supervised Classification: Improved Linear Separability and Out-of-Distribution Generalization", "Node Feature Extraction by Self-Supervised Multi-scale Neighborhood Prediction", "On Classification Thresholds for Graph Attention with Edge Features", "Learnable Graph Convolutional Attention Networks", "Effects of Graph Convolutions in Multi-layer Networks"], "answer_arxiv_id": ["2102.06966", "2111.00064", "2210.10014", "2211.11853v2", "2204.09297"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_18448"} +{"question": "What work has shown connections of GFlowNets to other generative models and hierarchical variational inference?", "answer": ["GFlowNets and variational inference"], "answer_arxiv_id": ["2210.00580"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_18449"} +{"question": "Which papers discuss the concept of canonicalization in symmetrization?", "answer": ["Equivariance with Learned Canonicalization Functions"], "answer_arxiv_id": ["2211.06489"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_18450"} +{"question": "Which work introduced Separated Attribute Predictability (SAP)?", "answer": ["Variational Inference of Disentangled Latent Concepts from Unlabeled Observations"], "answer_arxiv_id": ["1711.00848"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_18451"} +{"question": "What papers propose methods for gradient manipulations in domain generalization?", "answer": ["Domain Generalization via Gradient Surgery", "Learning to Balance Specificity and Invariance for In and Out of Domain Generalization", "SAND-mask: An Enhanced Gradient Masking Strategy for the Discovery of Invariances in Domain Generalization"], "answer_arxiv_id": ["2108.01621", "2008.12839", "2106.02266"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_18452"} +{"question": "Any works about use of Mixup in Split learning?", "answer": ["Differentially Private CutMix for Split Learning with Vision Transformer"], "answer_arxiv_id": ["2210.15986"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_18453"} +{"question": "Which studies have explored advancements in Natural Language Processing (NLP) driven by language models?", "answer": ["GPT-4 Technical Report", "Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2303.08774", "2307.09288"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_18454"} +{"question": "Could you mention works on video editing based on diffusion models?", "answer": ["Pix2Video: Video Editing using Image Diffusion", "Structure and Content-Guided Video Synthesis with Diffusion Models", "Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation", "Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video\n Generators", "FateZero: Fusing Attentions for Zero-shot Text-based Video Editing"], "answer_arxiv_id": ["2303.12688", "2302.03011", "2212.11565", "2303.13439", "2303.09535"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_18455"} +{"question": "Any works that focus on message passing formulations in hypergraph learning?", "answer": ["Wasserstein Soft Label Propagation on Hypergraphs: Algorithm and Generalization Error Bounds"], "answer_arxiv_id": ["1809.01833"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_18456"} +{"question": "Can you name the studies that focus on synthesizing natural lip motion for talking heads based on audio signals?", "answer": ["DiffTalk: Crafting Diffusion Models for Generalized Audio-Driven\n Portraits Animation", "Learning Dynamic Facial Radiance Fields for Few-Shot Talking Head\n Synthesis"], "answer_arxiv_id": ["2301.03786", "2207.11770"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18457"} +{"question": "Could you provide me studies about the GANs?", "answer": ["Large Scale GAN Training for High Fidelity Natural Image Synthesis"], "answer_arxiv_id": ["1809.11096"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_18458"} +{"question": "Which research proposed an asymmetric design to allow the large encoder to operate only on unmasked patches in Masked image modeling?", "answer": ["Masked Autoencoders Are Scalable Vision Learners"], "answer_arxiv_id": ["2111.06377"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_18459"} +{"question": "Are there any works presenting hybrid approaches in recalibration algorithms?", "answer": ["Verified Uncertainty Calibration"], "answer_arxiv_id": ["1909.10155"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18460"} +{"question": "Could you provide me some studies about semantic image synthesis using diffusion models?", "answer": ["Semantic Image Synthesis via Diffusion Models", "Pretraining is All You Need for Image-to-Image Translation", "Zero-shot spatial layout conditioning for text-to-image diffusion models", "SpaText: Spatio-Textual Representation for Controllable Image Generation"], "answer_arxiv_id": ["2207.00050", "2205.12952", "2306.13754", "2211.14305"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_18461"} +{"question": "Which research casts the distributionally robust optimization problem over contexts in Bayesian Optimization?", "answer": ["Distributionally Robust Bayesian Optimization"], "answer_arxiv_id": ["2002.09038"], "source_meta": {"published_time": "20220304"}, "qid": "AutoScholarQuery_train_18462"} +{"question": "What works analyze the statistical limits of phase retrieval in a high-dimensional limit for certain types of sensing designs?", "answer": ["Phase retrieval in high dimensions: Statistical and computational phase transitions"], "answer_arxiv_id": ["2006.05228"], "source_meta": {"published_time": "20230622"}, "qid": "AutoScholarQuery_train_18463"} +{"question": "Which methods learn conditionally invariant features across domains?", "answer": ["Domain Generalization via Conditional Invariant Representation", "Domain Generalization via Multidomain Discriminant Analysis"], "answer_arxiv_id": ["1807.08479", "1907.11216"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18464"} +{"question": "Which papers developed models that use 4D deformation fields as an additional layer over the canonical scene representation?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "STaR: Self-supervised Tracking and Reconstruction of Rigid Objects in\n Motion with Neural Rendering"], "answer_arxiv_id": ["2011.13961", "2011.12948", "2106.13228v2", "2101.01602"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_18465"} +{"question": "What studies have used the principle of minimizing the worst-case loss over different groups of data?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Just Train Twice: Improving Group Robustness without Training Group Information", "Model Agnostic Sample Reweighting for Out-of-Distribution Learning"], "answer_arxiv_id": ["1911.08731", "2107.09044", "2301.09819"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_18466"} +{"question": "Any references about agents asking clarification questions when faced with ambiguous or conflicting evidence?", "answer": ["Learning to Ask Good Questions: Ranking Clarification Questions using\n Neural Expected Value of Perfect Information"], "answer_arxiv_id": ["1805.04655"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18467"} +{"question": "Any research about applying large language models for compiler optimization?", "answer": ["Large Language Models for Compiler Optimization"], "answer_arxiv_id": ["2309.07062"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_18468"} +{"question": "Are there any works that used Deep Ensembles and Test-time augmentations to estimate uncertainty?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks"], "answer_arxiv_id": ["1612.01474", "1807.07356"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_18469"} +{"question": "Could you list line-based localization methods that leverage the photometric information from the images used for line extraction?", "answer": ["SOLD2: Self-supervised Occlusion-aware Line Description and Detection", "Line as a Visual Sentence: Context-aware Line Descriptor for Visual\n Localization", "Robust Line Segments Matching via Graph Convolution Networks", "VLASE: Vehicle Localization by Aggregating Semantic Edges", "GlueStick: Robust Image Matching by Sticking Points and Lines Together", "Pose Refinement with Joint Optimization of Visual Points and Lines", "3D Line Mapping Revisited", "LDL: Line Distance Functions for Panoramic Localization"], "answer_arxiv_id": ["2104.03362", "2109.04753", "2004.04993", "1807.02536", "2304.02008", "2110.03940", "2303.17504", "2308.13989"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_18470"} +{"question": "In what papers entity-centric models also referred to as slots or object files are described?", "answer": ["On the Binding Problem in Artificial Neural Networks", "Dynamic Routing Between Capsules", "Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects", "Genesis: Generative Scene Inference and Sampling with Object-Centric Latent Representations", "Object Files and Schemata: Factorizing Declarative and Procedural Knowledge in Dynamical Systems", "Systematic Evaluation of Causal Discovery in Visual Model Based Reinforcement Learning", "MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational Inference", "Unsupervised Video Decomposition using Spatio-temporal Iterative Inference", "S2RMs: Spatially Structured Recurrent Modules"], "answer_arxiv_id": ["2012.05208", "1710.09829", "1806.01794", "1907.13052", "2006.16225", "2107.00848", "1901.11390", "1903.00450", "2006.14727", "2007.06533v1"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_18471"} +{"question": "What works utilize pre-trained models as an oracle for better general representations across various domains?", "answer": ["Exploiting Domain-Specific Features to Enhance Domain Generalization", "Learning to Balance Specificity and Invariance for In and Out of Domain\n Generalization", "Domain Generalization by Mutual-Information Regularization with\n Pre-trained Models"], "answer_arxiv_id": ["2110.09410", "2008.12839", "2203.10789"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18472"} +{"question": "Which studies are related to the order k Tensor PCA problem?", "answer": ["A statistical model for tensor PCA", "Homotopy Analysis for Tensor PCA", "How to iron out rough landscapes and get optimal performances: Averaged Gradient Descent and its application to tensor PCA"], "answer_arxiv_id": ["1411.1076", "1610.09322", "1905.12294"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18473"} +{"question": "Which work extends the 2D heatmap to 3D by incorporating a bird-eye-view in HPS?", "answer": ["Putting People in their Place: Monocular Regression of 3D People in\n Depth"], "answer_arxiv_id": ["2112.08274"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_18474"} +{"question": "What studies developed FL algorithms designed to mitigate discrepancies between local and global model parameters?", "answer": ["FedDC: Federated Learning with Non-IID Data via Local Drift Decoupling\n and Correction"], "answer_arxiv_id": ["2203.11751"], "source_meta": {"published_time": "20240429"}, "qid": "AutoScholarQuery_train_18475"} +{"question": "Which studies have used the Proxy Tuning method?", "answer": ["Tuning Language Models by Proxy", "DExperts: Decoding-Time Controlled Text Generation with Experts and\n Anti-Experts"], "answer_arxiv_id": ["2401.08565", "2105.03023"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_18476"} +{"question": "Could you provide me some works that incorporated tensor anchor graphs into their methods?", "answer": ["Multi-view MERA Subspace Clustering"], "answer_arxiv_id": ["2305.09095"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_18477"} +{"question": "Can you list some works that focus on 6D object pose estimation using 3D registration techniques?", "answer": ["6-DoF Object Pose from Semantic Keypoints", "Real-Time Seamless Single Shot 6D Object Pose Prediction", "Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects", "Implicit 3D Orientation Learning for 6D Object Detection from RGB Images", "Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects"], "answer_arxiv_id": ["1703.04670", "1711.08848", "1809.10790", "1902.01275", "1809.10790"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_18478"} +{"question": "Could you provide me some studies about audio tokenization model used in generative audio models?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation", "MusicLM: Generating Music From Text", "Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers"], "answer_arxiv_id": ["2209.03143", "2301.11325", "2301.02111"], "source_meta": {"published_time": "20230611"}, "qid": "AutoScholarQuery_train_18479"} +{"question": "What papers have built on SAM for task application and achievement enhancements?", "answer": ["Track Anything: Segment Anything Meets Videos", "Semantic-SAM: Segment and Recognize Anything at Any Granularity", "Recognize Anything: A Strong Image Tagging Model"], "answer_arxiv_id": ["2304.11968", "2307.04767", "2306.03514"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18480"} +{"question": "Which works first fully trains the original model or a smaller proxy when finding coresets for training machine learning models?", "answer": ["Semantic Redundancies in Image-Classification Datasets: The 10% You Don’t Need", "Selection via Proxy: Efficient Data Selection for Deep Learning"], "answer_arxiv_id": ["1901.11409", "1906.11829"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_18481"} +{"question": "Which work revisited Noise2Noise method and generated noisy image pairs from a single noisy image via spatial subsampling?", "answer": ["Neighbor2Neighbor: Self-Supervised Denoising from Single Noisy Images"], "answer_arxiv_id": ["2101.02824"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_18482"} +{"question": "What studies explore the concept of targeted alteration in the model?", "answer": ["Editable Neural Networks"], "answer_arxiv_id": ["2004.00345"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_18483"} +{"question": "Which paper introduced Decoupled-Mixup that distinguishes discriminative and noise-prone parts of images and fuses these parts separately?", "answer": ["Decoupled Mixup for Generalized Visual Recognition"], "answer_arxiv_id": ["2210.14783"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18484"} +{"question": "Which papers focused on representation learning using contrastive learning algorithms?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2006.07733"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_18485"} +{"question": "Which papers proposed a decentralized shadow reward actor-critic (DSAC) method in multi-agent systems?", "answer": ["MARL with General Utilities via Decentralized Shadow Reward Actor-Critic"], "answer_arxiv_id": ["2106.00543"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_18486"} +{"question": "Can you name the works that employed reinforcement learning-based methods to search for optimal device parameters to fulfill desired circuit specifications, particularly in the graph learning for analog circuit design automation context?", "answer": ["GCN-RL Circuit Designer: Transferable Transistor Sizing with Graph Neural Networks and Reinforcement Learning", "Domain Knowledge-Infused Deep Learning for Automated Analog/Radio-Frequency Circuit Parameter Optimization"], "answer_arxiv_id": ["2005.00406", "2204.12948"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_18487"} +{"question": "Could you tell me some studies that used the supervision of Contrastive Text-Image Pre-Training for inpainting neural radiance fields?", "answer": ["CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields", "LaTeRF: Label and Text Driven Object Radiance Fields"], "answer_arxiv_id": ["2112.05139", "2207.01583"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_18488"} +{"question": "What papers study keyphrase generation for long documents?", "answer": ["Select, Extract and Generate: Neural Keyphrase Generation with\n Layer-wise Coverage Attention", "Keyphrase Generation Beyond the Boundaries of Title and Abstract"], "answer_arxiv_id": ["2008.01739", "2112.06776"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_18489"} +{"question": "What studies involve minimizing some distribution-distance measures for learning invariant representations?", "answer": ["Learning Transferable Features with Deep Adaptation Networks"], "answer_arxiv_id": ["1502.02791"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_18490"} +{"question": "In what works is ViT utilized for impressive results on ImageNet?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "ImageNet Large Scale Visual Recognition Challenge"], "answer_arxiv_id": ["2010.11929", "1409.0575"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_18491"} +{"question": "Which paper reported a recent advancement in learning specific relations, including some interactions, from exemplar images?", "answer": ["ReVersion: Diffusion-Based Relation Inversion from Images"], "answer_arxiv_id": ["2303.13495"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_18492"} +{"question": "Can you cite works that proposed representing scenes via neural radiance fields?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "HumanNeRF: Efficiently Generated Human Radiance Field from Sparse Inputs", "HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular\n Video", "Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via\n Self-supervised Scene Decomposition"], "answer_arxiv_id": ["2003.08934", "2012.15838", "2112.02789", "2201.04127", "2302.11566"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18493"} +{"question": "Which works applied Convolutional networks (CNNs) on video inputs before the rise of Transformers?", "answer": ["Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset", "A Closer Look at Spatiotemporal Convolutions for Action Recognition", "Non-local Neural Networks"], "answer_arxiv_id": ["1705.07750", "1711.11248", "1711.07971"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_18494"} +{"question": "Could you provide me some studies about building realistic tactile simulation system in the context of visual-tactile learning?", "answer": ["TACTO: A Fast, Flexible, and Open-source Simulator for High-Resolution\n Vision-based Tactile Sensors", "Sim-to-Real for Robotic Tactile Sensing via Physics-Based Simulation and\n Learned Latent Projections"], "answer_arxiv_id": ["2012.08456", "2103.16747"], "source_meta": {"published_time": "20240116"}, "qid": "AutoScholarQuery_train_18495"} +{"question": "Which works cover point-based approaches in LiDAR semantic segmentation?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "SPLATNet: Sparse Lattice Networks for Point Cloud Processing", "Point-Voxel CNN for Efficient 3D Deep Learning", "KPConv: Flexible and Deformable Convolution for Point Clouds", "PCT: Point cloud transformer", "ConDaFormer: Disassembled Transformer with Local Structure Enhancement\n for 3D Point Cloud Understanding"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1802.08275", "1907.03739", "1904.08889", "2012.09688", "2312.11112"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_18496"} +{"question": "Which papers use an extra special token for vision transformer-based object detection?", "answer": ["ViDT: An Efficient and Effective Fully Transformer-based Object Detector"], "answer_arxiv_id": ["2110.03921"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_18497"} +{"question": "What researches used meta-learning for targets for TD learning?", "answer": ["Meta-Gradient Reinforcement Learning with an Objective Discovered Online"], "answer_arxiv_id": ["2007.08433"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_18498"} +{"question": "Which researches proposed adaptable personalization policies for local fine-tuning?", "answer": ["Personalized Federated Learning using Hypernetworks", "Layer-wised Model Aggregation for Personalized Federated Learning", "Adaptive Personalized Federated Learning"], "answer_arxiv_id": ["2103.04628", "2205.03993", "2003.13461"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_18499"} +{"question": "Which studies provided insight into the theoretical aspects of contrastive learning?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Contrastive estimation reveals topic posterior information to linear models", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere", "Towards the Generalization of Contrastive Self-Supervised Learning"], "answer_arxiv_id": ["1902.09229v1", "1902.09229", "2003.02234", "2106.04156", "2005.10242", "2111.00743"], "source_meta": {"published_time": "20230218"}, "qid": "AutoScholarQuery_train_18500"} +{"question": "Which work proposed a protocol for reference-free token-level annotation of complex D2T generation output?", "answer": ["A Gold Standard Methodology for Evaluating Accuracy in Data-To-Text\n Systems"], "answer_arxiv_id": ["2011.03992"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_18501"} +{"question": "Which works discuss the concept of combining noise estimates from multiple conditional distributions in the context of diffusion models?", "answer": ["Compositional Visual Generation with Composable Diffusion Models", "Reduce, Reuse, Recycle: Compositional Generation with Energy-Based\n Diffusion Models and MCMC"], "answer_arxiv_id": ["2206.01714", "2302.11552"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18502"} +{"question": "Can you provide some representative approaches of NAS for ASR?", "answer": ["LightSpeech: Lightweight and Fast Text to Speech with Neural Architecture Search", "Darts-Conformer: Towards Efficient Gradient-Based Neural Architecture Search For End-to-End ASR"], "answer_arxiv_id": ["2102.04040", "2104.02868"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_18503"} +{"question": "Which work proposes a diversity-based selection method in active learning that selects a subset by performing k-means++ on the gradient embedding based on the latest model?", "answer": ["Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds"], "answer_arxiv_id": ["1906.03671"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_18504"} +{"question": "Which studies have utilized volume rendering to model geometry and color together?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Nerfies: Deformable Neural Radiance Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "ADOP: Approximate Differentiable One-Pixel Point Rendering", "Plenoxels: Radiance Fields without Neural Networks", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "Neural RGB-D Surface Reconstruction", "GO-Surf: Neural Feature Grid Optimization for Fast, High-Fidelity RGB-D Surface Reconstruction", "TransformerFusion: Monocular RGB Scene Reconstruction using Transformers", "MonoNeuralFusion: Online Monocular Neural 3D Reconstruction with Geometric Priors", "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video"], "answer_arxiv_id": ["2003.08934", "2011.12948", "2201.05989", "2110.06635", "2112.05131", "2112.12130", "2104.04532", "2206.14735v2", "2107.02191", "2209.15153", "2104.00681"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_18505"} +{"question": "Which research works define the concept of staircase mechanisms optimal for objective function minimization in local DP algorithms?", "answer": ["Extremal Mechanisms for Local Differential Privacy"], "answer_arxiv_id": ["1407.1338"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_18506"} +{"question": "Which study has proposed the PDS strategy for constructing a pessimistic reward function using an ensemble of neural networks?", "answer": ["The Provable Benefits of Unsupervised Data Sharing for Offline Reinforcement Learning"], "answer_arxiv_id": ["2302.13493"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_18507"} +{"question": "What works study the convergence of Policy Gradient methods?", "answer": ["Global Convergence of Policy Gradient Methods for the Linear Quadratic Regulator", "On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shift", "PC-PG: Policy Cover Directed Exploration for Provable Policy Gradient Learning", "Global Optimality Guarantees For Policy Gradient Methods", "On the Global Convergence Rates of Softmax Policy Gradient Methods", "Global Convergence of Policy Gradient Methods to (Almost) Locally Optimal Policies", "PC-PG: Policy Cover Directed Exploration for Provable Policy Gradient Learning", "Sample Efficient Reinforcement Learning with REINFORCE", "Softmax Policy Gradient Methods Can Take Exponential Time to Converge", "Fast Global Convergence of Natural Policy Gradient Methods with Entropy Regularization", "Beyond Exact Gradients: Convergence of Stochastic Soft-Max Policy Gradient Methods with Entropy Regularization", "A general sample complexity analysis of vanilla policy gradient", "Towards an Understanding of Default Policies in Multitask Policy Optimization"], "answer_arxiv_id": ["1801.05039", "1908.00261", "2007.08459", "1906.01786", "2005.06392", "1906.08383", "2007.08459", "2010.11364", "2102.11270v3", "2007.06558", "2110.10117v3", "2107.11433v5", "2111.02994v4"], "source_meta": {"published_time": "20221224"}, "qid": "AutoScholarQuery_train_18508"} +{"question": "Which studies utilized causal inference for off-policy evaluation?", "answer": ["Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search", "Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models"], "answer_arxiv_id": ["1811.06272", "1905.05824"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_18509"} +{"question": "Can you provide examples of studies that exploit domain adversarial learning in their augmentation approaches to domain generalization?", "answer": ["Generalizing Across Domains via Cross-Gradient Training", "Deep Domain-Adversarial Image Generation for Domain Generalisation"], "answer_arxiv_id": ["1804.10745", "2003.06054"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_18510"} +{"question": "Which work recommends a data-constraining encoder?", "answer": ["Regularized target encoding outperforms traditional methods in supervised machine learning with high cardinality features"], "answer_arxiv_id": ["2104.00629"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_18511"} +{"question": "Which were the early works on external memory that can be retrieved and written?", "answer": ["Neural Turing Machines", "Memory Networks"], "answer_arxiv_id": ["1410.5401", "1410.3916"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18512"} +{"question": "What research includes regularization terms in the loss function to achieve Personalized federated learning (PFL)?", "answer": ["Federated Optimization in Heterogeneous Networks", "Ditto: Fair and Robust Federated Learning Through Personalization", "Personalized Federated Learning with Moreau Envelopes"], "answer_arxiv_id": ["1812.06127", "2012.04221", "2006.08848"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18513"} +{"question": "What studies apply amodal completion approach on toy datasets?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational\n Inference", "GENESIS: Generative Scene Inference and Sampling with Object-Centric\n Latent Representations"], "answer_arxiv_id": ["1901.11390", "1903.00450", "1907.13052"], "source_meta": {"published_time": "20231224"}, "qid": "AutoScholarQuery_train_18514"} +{"question": "What are some works that equip AI models with Theory of Mind (ToM) based on deep learning approaches?", "answer": ["Machine Theory of Mind", "MindCraft: Theory of Mind Modeling for Situated Dialogue in\n Collaborative Tasks", "Baby Intuitions Benchmark (BIB): Discerning the goals, preferences, and\n actions of others", "I Cast Detect Thoughts: Learning to Converse and Guide with Intents and\n Theory-of-Mind in Dungeons and Dragons", "Computational Language Acquisition with Theory of Mind", "Neural Reasoning About Agents' Goals, Preferences, and Actions"], "answer_arxiv_id": ["1802.07740", "2109.06275", "2102.11938", "2212.10060", "2303.01502", "2312.07122"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_18515"} +{"question": "Can you provide sources of traditional GAN-based models used in text-to-image generation?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image Synthesis"], "answer_arxiv_id": ["1711.10485", "1904.01310"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_18516"} +{"question": "Which studies are related to vision and language understanding for tasks such as image captioning, visual question answering and visual grounding?", "answer": ["Language Models are Few-Shot Learners", "Flamingo: a Visual Language Model for Few-Shot Learning", "PaLM-E: An Embodied Multimodal Language Model"], "answer_arxiv_id": ["2005.14165", "2204.14198", "2303.03378"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_18517"} +{"question": "What papers address memory models in the area of action understanding?", "answer": ["Memory-and-Anticipation Transformer for Online Action Understanding"], "answer_arxiv_id": ["2308.07893"], "source_meta": {"published_time": "20230731"}, "qid": "AutoScholarQuery_train_18518"} +{"question": "Could you give me examples of research that use graph-based neural networks for tabular data representation learning?", "answer": ["Answering Conversational Questions on Structured Data without Logical Forms", "Retrieving Complex Tables with Multi-Granular Graph Representation Learning", "TCN: Table Convolutional Network for Web Table Interpretation"], "answer_arxiv_id": ["1908.11787", "2105.01736", "2102.09460"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_18519"} +{"question": "In which papers does the research involve leveraging the memorization effect of DNNs to address the noisy data problem?", "answer": ["A Closer Look at Memorization in Deep Networks"], "answer_arxiv_id": ["1706.05394"], "source_meta": {"published_time": "20230819"}, "qid": "AutoScholarQuery_train_18520"} +{"question": "Can you name the papers that used Minecraft to demonstrate how modular policies can improve exploration?", "answer": ["Align-RUDDER: Learning From Few Demonstrations by Reward Redistribution"], "answer_arxiv_id": ["2009.14108"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_18521"} +{"question": "What research works have explored the idea of self-conditioning for diffusion models?", "answer": ["Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning", "Self-conditioned Embedding Diffusion for Text Generation", "Continuous diffusion for categorical data", "A Generalist Framework for Panoptic Segmentation of Images and Videos"], "answer_arxiv_id": ["2208.04202", "2211.04236", "2211.15089", "2210.06366"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_18522"} +{"question": "Are there any resources that detail the use of GPT-4 for chatbot evaluation instead of costly human annotation?", "answer": ["Instruction Tuning with GPT-4"], "answer_arxiv_id": ["2304.03277"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_18523"} +{"question": "Which works address the issue of data contamination in the context of pre-training mixture?", "answer": ["NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination\n for each Benchmark", "Time Travel in LLMs: Tracing Data Contamination in Large Language Models"], "answer_arxiv_id": ["2310.18018", "2308.08493"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_18524"} +{"question": "Could you provide me studies done on the empirical study of the label memorization concept in computer vision?", "answer": ["What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation"], "answer_arxiv_id": ["2008.03703"], "source_meta": {"published_time": "20211224"}, "qid": "AutoScholarQuery_train_18525"} +{"question": "Are there any image editing studies that use text-guided image-to-image translation method?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2211.12572"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_18526"} +{"question": "Which papers have employed curriculum learning in L2L systems?", "answer": ["Training Stronger Baselines for Learning to Optimize"], "answer_arxiv_id": ["2010.09089"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_18527"} +{"question": "Are there any research papers that adopt the CLIP guidance as part of their loss function during the learning process?", "answer": ["DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation", "Diffusion Models already have a Semantic Latent Space"], "answer_arxiv_id": ["2110.02711", "2210.10960v2"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_18528"} +{"question": "What papers discussed the aim of unsupervised RL towards discovering a diverse range of goals and learning corresponding goal-reaching policies?", "answer": ["Unsupervised Control through Non-Parametric Discriminative Rewards", "Skew-Fit: State-Covering Self-Supervised Reinforcement Learning", "Maximum Entropy Gain Exploration for Long Horizon Multi-goal Reinforcement Learning", "Discovering and Achieving Goals via World Models"], "answer_arxiv_id": ["1811.11359", "1903.03698", "2007.02832v1", "2110.09514"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_18529"} +{"question": "Which papers present Large Language Models that handle inputs from images in the field of autonomous driving?", "answer": ["DetGPT: Detect What You Need via Reasoning", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "PerceptionGPT: Effectively Fusing Visual Perception into LLM", "G-LLaVA: Solving Geometric Problem with Multi-Modal Large Language Model", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic"], "answer_arxiv_id": ["2305.14167", "2201.12086", "2301.12597", "2311.06612", "2312.11370", "2304.10592", "2204.14198", "2306.15195"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_18530"} +{"question": "What papers introduced improvements to the Sinkhorn algorithm, providing enhanced numerical stability or memory-efficient versions?", "answer": ["Stabilized Sparse Scaling Algorithms for Entropy Regularized Transport Problems", "Interpolating between Optimal Transport and MMD using Sinkhorn Divergences"], "answer_arxiv_id": ["1610.06519", "1810.08278"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18531"} +{"question": "Which papers follow LSS and predict a distribution over depth bins for BEV detection in the context of multi-camera operation?", "answer": ["Lift, Splat, Shoot: Encoding Images from Arbitrary Camera Rigs by Implicitly Unprojecting to 3D", "Categorical Depth Distribution Network for Monocular 3D Object Detection", "BEVDet: High-Performance Multi-Camera 3D Object Detection in Bird-Eye-View", "Unifying Voxel-based Representation with Transformer for 3D Object Detection"], "answer_arxiv_id": ["2008.05711", "2103.01100", "2112.11790", "2206.00630"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_18532"} +{"question": "Which research involves the use of 'Attention rollout' for calculating the contributions using attention maps?", "answer": ["Quantifying Attention Flow in Transformers"], "answer_arxiv_id": ["2005.00928"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_18533"} +{"question": "What studies have proposed transformer architecture for natural language processing tasks?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18534"} +{"question": "Are there studies that utilize self-attention with downsampling technique to improve computational efficiency?", "answer": ["Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions", "PVT v2: Improved Baselines with Pyramid Vision Transformer", "Conformer: Local Features Coupling Global Representations for Visual Recognition", "Fast Vision Transformers with HiLo Attention", "Inception Transformer", "Vision Transformer with Deformable Attention", "Twins: Revisiting the Design of Spatial Attention in Vision Transformers"], "answer_arxiv_id": ["2102.12122", "2106.13797", "2105.03889", "2205.13213", "2205.12956", "2201.00520", "2104.13840"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_18535"} +{"question": "What studies are focused on active fine-tuning in Active Learning?", "answer": ["Active Learning Helps Pretrained Models Learn the Intended Task"], "answer_arxiv_id": ["2204.08491"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_18536"} +{"question": "Which papers analyze the adoption of data-augmentation techniques to address the sample inefficiency of image-based RL?", "answer": ["Reinforcement Learning with Augmented Data", "Mastering Visual Continuous Control: Improved Data-Augmented Reinforcement Learning"], "answer_arxiv_id": ["2004.14990", "2107.09645"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_18537"} +{"question": "What studies have been done to address the problem of sampling bias in contrastive learning?", "answer": ["Hard Negative Mixing for Contrastive Learning", "Contrastive Learning with Hard Negative Samples"], "answer_arxiv_id": ["2010.01028v2", "2010.04592"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18538"} +{"question": "Which papers have focused on acquiring large number of scenes especially for reflection separation?", "answer": ["Polarized Reflection Removal with Perfect Alignment in the Wild"], "answer_arxiv_id": ["2003.12789"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18539"} +{"question": "What source provides an overview of LLM attacks?", "answer": ["Attacks, Defenses and Evaluations for LLM Conversation Safety: A Survey"], "answer_arxiv_id": ["2402.09283"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18540"} +{"question": "Any studies on modifications of TD more conducive for convergence analysis?", "answer": ["An Emphatic Approach to the Problem of Off-policy Temporal-Difference Learning", "Breaking the Deadly Triad with a Target Network", "A Convergent Off-Policy Temporal Difference Algorithm", "Regularized Q-learning", "Gradient Temporal-Difference Learning with Regularized Corrections"], "answer_arxiv_id": ["1503.04269", "2101.08862", "1911.05697", "2202.05404", "2007.00611"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_18541"} +{"question": "Are there any works proposing learning-based methods for fisheye cameras?", "answer": ["Rethinking Generic Camera Models for Deep Single Image Camera\n Calibration to Recover Rotation and Fisheye Distortion"], "answer_arxiv_id": ["2111.12927"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_18542"} +{"question": "Are there any benchmarks that focus on evaluating the stealthiness of adversarial examples during transfer-based attacks?", "answer": ["Towards Good Practices in Evaluating Transfer Adversarial Attacks"], "answer_arxiv_id": ["2211.09565"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_18543"} +{"question": "Can you mention studies that advanced SSL through a threshold strategy for pseudo labels?", "answer": ["Dash: Semi-Supervised Learning with Dynamic Thresholding", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning"], "answer_arxiv_id": ["2109.00650", "2110.08263", "2205.07246"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_18544"} +{"question": "Could you provide me the reference where Gumbel-Sinkhorn is used in the context of differentiable DAG learning?", "answer": ["Learning Latent Permutations with Gumbel-Sinkhorn Networks"], "answer_arxiv_id": ["1802.08665"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_18545"} +{"question": "Could you provide me some studies about the application of statistical methods in unsupervised machine translation?", "answer": ["Phrase-Based & Neural Unsupervised Machine Translation", "A robust self-learning method for fully unsupervised cross-lingual mappings of word embeddings"], "answer_arxiv_id": ["1804.07755", "1805.06297"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_18546"} +{"question": "What studies have attempted to curate a set of text prompts to evaluate key aspects of text-conditioned generative tasks?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "T$^3$Bench: Benchmarking Current Progress in Text-to-3D Generation"], "answer_arxiv_id": ["2209.14988", "2310.02977v2"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_18547"} +{"question": "Which papers discussed about block-floating point schemes?", "answer": ["Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks", "Training DNNs with Hybrid Block Floating Point"], "answer_arxiv_id": ["1711.02213", "1804.01526"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18548"} +{"question": "What work claimed that denoising a noisy measurement towards another measurement is statistically equivalent to the supervised training on clean images up to a constant?", "answer": ["Noise2Noise: Learning Image Restoration without Clean Data"], "answer_arxiv_id": ["1803.04189"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_18549"} +{"question": "Could you point out the research that has been conducted on using deep learning for indoor lighting estimation?", "answer": ["Deep Parametric Indoor Lighting Estimation"], "answer_arxiv_id": ["1910.08812"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_18550"} +{"question": "Which papers discuss the most common approximations used in Bayesian Gaussian linear models?", "answer": ["Gaussian Processes for Big Data"], "answer_arxiv_id": ["1309.6835"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_18551"} +{"question": "Which paper utilized cross-national surveys to assess how well LLMs capture subjective opinions from various countries?", "answer": ["Towards Measuring the Representation of Subjective Global Opinions in\n Language Models"], "answer_arxiv_id": ["2306.16388"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_18552"} +{"question": "Which works improved the training scheme for balanced learning in long-tailed object detection?", "answer": ["LVIS: A Dataset for Large Vocabulary Instance Segmentation", "Equalization Loss for Long-Tailed Object Recognition", "Equalization Loss v2: A New Gradient Balance Approach for Long-tailed Object Detection", "Seesaw Loss for Long-Tailed Instance Segmentation", "Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentation", "Scaling Object Detection by Transferring Classification Weights", "Decoupling Representation and Classifier for Long-Tailed Recognition"], "answer_arxiv_id": ["1908.03195", "2003.05176", "2012.08548", "2008.10032", "2012.07177", "1909.06804", "1910.09217"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_18553"} +{"question": "What works represent the 'source-free' domain adaptation approach?", "answer": ["Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation", "Class-Incremental Domain Adaptation", "Universal Source-Free Domain Adaptation", "Domain Impression: A Source Data Free Domain Adaptation Method"], "answer_arxiv_id": ["2002.08546", "2008.01389", "2004.04393", "2102.09003"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_18554"} +{"question": "What are the research papers that considered ϕ−d​i​v​e​r​g​e​n​c​e and the Wasserstein distance as distance metrics in Distributionally Robust Optimization (DRO)?", "answer": ["Does Distributionally Robust Supervised Learning Give Robust Classifiers?", "Variance-based regularization with convex objectives", "Certifying Some Distributional Robustness with Principled Adversarial Training", "Data-Driven Distributionally Robust Optimization Using the Wasserstein Metric: Performance Guarantees and Tractable Reformulations", "Wasserstein Distributionally Robust Optimization: Theory and Applications in Machine Learning", "Robust Wasserstein Profile Inference and Applications to Machine Learning", "Distributionally Robust Stochastic Optimization with Wasserstein Distance"], "answer_arxiv_id": ["1611.02041", "1610.02581", "1710.10571v5", "1505.05116", "1908.08729v1", "1610.05627", "1604.02199"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_18555"} +{"question": "Can you mention some datasets that use internet videos sourced from YouTube?", "answer": ["UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild", "The Kinetics Human Action Video Dataset", "Moments in Time Dataset: one million videos for event understanding"], "answer_arxiv_id": ["1212.0402", "1705.06950", "1801.03150"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18556"} +{"question": "What work utilizes the idea of upper confidence bound by estimating the uncertainty via ensembles to improve the sample efficiency?", "answer": ["UCB Exploration via Q-Ensembles"], "answer_arxiv_id": ["1706.01502v3"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18557"} +{"question": "What papers cite that local DP models usually incur much larger estimation errors compared to central DP?", "answer": ["Inference under Information Constraints II: Communication Constraints and Shared Randomness", "Fisher information under local differential privacy"], "answer_arxiv_id": ["1905.08302v2", "2005.10783"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_18558"} +{"question": "Are there any existing works that offer formal guarantees in the setting of online RL with access to logged data?", "answer": ["Agnostic System Identification for Model-Based Reinforcement Learning", "Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning", "Hybrid RL: Using Both Offline and Online Data Can Make RL Efficient", "Bandits with Partially Observable Confounded Data"], "answer_arxiv_id": ["1203.1007", "2106.04895", "2210.06718", "2006.06731"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_18559"} +{"question": "Which works propose domain generalization solutions based on aligning between multiple source domains?", "answer": ["Deep CORAL: Correlation Alignment for Deep Domain Adaptation", "Domain-Adversarial Training of Neural Networks", "Domain Generalization via Conditional Invariant Representations"], "answer_arxiv_id": ["1607.01719", "1505.07818", "1807.08479"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_18560"} +{"question": "Which work showed that LLMs are good zero-shot reasoners?", "answer": ["Large Language Models are Zero-Shot Reasoners"], "answer_arxiv_id": ["2205.11916"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_18561"} +{"question": "Which studies represent NAS approaches in NLP?", "answer": ["NAS-BERT: Task-Agnostic and Adaptive-Size BERT Compression with Neural Architecture Search", "HAT: Hardware-Aware Transformers for Efficient Natural Language Processing"], "answer_arxiv_id": ["2105.14444", "2005.14187"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_18562"} +{"question": "Which works have made significant advancements in multi-modal capabilities of Large Language Models (LLMs)?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "OFA: Unifying Architectures, Tasks, and Modalities Through a Simple\n Sequence-to-Sequence Learning Framework", "Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2301.12597", "2202.03052", "2204.14198"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_18563"} +{"question": "Which works introduce the 'EmpCov' prior?", "answer": ["Dangers of Bayesian Model Averaging under Covariate Shift"], "answer_arxiv_id": ["2106.11905"], "source_meta": {"published_time": "20220624"}, "qid": "AutoScholarQuery_train_18564"} +{"question": "Could you provide me some works discussing the recent trend in instruction following methods?", "answer": ["Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback"], "answer_arxiv_id": ["2204.05862"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_18565"} +{"question": "Which research works have applied policy gradient methods to solve zero-sum stochastic games and established their finite-sample complexity?", "answer": ["Independent Policy Gradient Methods for Competitive Reinforcement Learning", "Provably Efficient Policy Optimization for Two-Player Zero-Sum Markov Games", "Derivative-Free Policy Optimization for Linear Risk-Sensitive and Robust Control Design: Implicit Regularization and Sample Complexity"], "answer_arxiv_id": ["2101.04233v1", "2102.08903", "2101.01041"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_18566"} +{"question": "Which works introduced HLoc pipeline for scaling up to large scenes in camera pose prediction?", "answer": ["From Coarse to Fine: Robust Hierarchical Localization at Large Scale", "SuperGlue: Learning Feature Matching with Graph Neural Networks"], "answer_arxiv_id": ["1812.03506", "1911.11763"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18567"} +{"question": "What works established last-iterate performance for several policy-based primal-dual algorithms in the Lagrangian-based framework?", "answer": ["Risk-Constrained Reinforcement Learning with Percentile Risk Criteria", "Reward Constrained Policy Optimization"], "answer_arxiv_id": ["1512.01629", "1805.11074"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_18568"} +{"question": "What work explores addressing downstream tasks in a data-efficient semi-supervised way?", "answer": ["Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts"], "answer_arxiv_id": ["2012.09165"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_18569"} +{"question": "What studies have used diffusion models for text-driven image manipulation?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Dual Diffusion Implicit Bridges for Image-to-Image Translation", "Blended Diffusion for Text-driven Editing of Natural Images", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential\n Equations", "More Control for Free! Image Synthesis with Semantic Diffusion Guidance", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image\n Manipulation", "Imagic: Text-Based Real Image Editing with Diffusion Models", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2112.10741", "2203.08382", "2111.14818", "2108.01073", "2112.05744", "2110.02711", "2210.09276", "2208.01618", "2208.12242"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_18570"} +{"question": "Which papers have proposed methods to quantify and mitigate representation collapse in neural networks?", "answer": ["Grafit: Learning fine-grained image representations with coarse labels", "Hidden State Variability of Pretrained Language Models Can Guide Computation Reduction for Transfer Learning"], "answer_arxiv_id": ["2011.12982", "2210.10041"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_18571"} +{"question": "Could you tell me about the work where a point-dragging formulation for fine-grained image editing was presented?", "answer": ["Drag Your GAN: Interactive Point-based Manipulation on the Generative\n Image Manifold"], "answer_arxiv_id": ["2305.10973"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_18572"} +{"question": "What works discuss leveraging a pretrained teacher policy to improve the learning efficiency of the student policy?", "answer": ["DisCoRL: Continual Reinforcement Learning via Policy Distillation"], "answer_arxiv_id": ["1907.05855"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_18573"} +{"question": "Any works about the use of neural networks for neural decompilation?", "answer": ["Towards Neural Decompilation", "Semantics-Recovering Decompilation through Neural Machine Translation"], "answer_arxiv_id": ["1905.08325", "2112.15491"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_18574"} +{"question": "Which works focus on viewing transformers as function approximators?", "answer": ["Are Transformers universal approximators of sequence-to-sequence functions?"], "answer_arxiv_id": ["1912.10077"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_18575"} +{"question": "Could you name some studies that used autoregressive generative models inspired from LLMs in RAG context?", "answer": ["Re-Imagen: Retrieval-Augmented Text-to-Image Generator", "Retrieval-Augmented Multimodal Language Modeling"], "answer_arxiv_id": ["2209.14491", "2211.12561"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_18576"} +{"question": "What works incorporate attention mechanisms for point cloud segmentation?", "answer": ["Point Transformer", "Stratified Transformer for 3D Point Cloud Segmentation", "Learning Inner-Group Relations on Point Clouds", "Fast Point Transformer", "PatchFormer: An Efficient Point Transformer with Patch Attention"], "answer_arxiv_id": ["2012.09164", "2203.14508", "2108.12468", "2112.04702", "2111.00207"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_18577"} +{"question": "Which research papers used VQ-VAE for the construction of world model with lower number of parameters?", "answer": ["Smaller World Models for Reinforcement Learning"], "answer_arxiv_id": ["2010.05767"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_18578"} +{"question": "Which works showcased the socio-demographic disparities in commercial face recognition systems?", "answer": ["Demographic Bias in Biometrics: A Survey on an Emerging Challenge", "Robustness Disparities in Face Detection", "Two-Face: Adversarial Audit of Commercial Face Recognition Systems", "Comparing Human and Machine Bias in Face Recognition"], "answer_arxiv_id": ["2003.02488", "2211.15937", "2111.09137", "2110.08396"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_18579"} +{"question": "Could you provide me some works about using reinforcement learning (RL) for pruning?", "answer": ["AMC: AutoML for Model Compression and Acceleration on Mobile Devices", "N2N Learning: Network to Network Compression via Policy Gradient\n Reinforcement Learning", "BlockDrop: Dynamic Inference Paths in Residual Networks", "Learning to Prune Filters in Convolutional Neural Networks", "Topology-Aware Network Pruning using Multi-stage Graph Embedding and\n Reinforcement Learning"], "answer_arxiv_id": ["1802.03494", "1709.06030", "1711.08393", "1801.07365", "2102.03214"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_18580"} +{"question": "Could you provide me some works that explored the domain generalization in terms of Out-of-Distribution (OOD) Generalization?", "answer": ["Domain Generalization via Invariant Feature Representation"], "answer_arxiv_id": ["1301.2115"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_18581"} +{"question": "Which works discuss the universal approximation property of Transformers when treated as discrete-time systems?", "answer": ["Are Transformers universal approximators of sequence-to-sequence functions?"], "answer_arxiv_id": ["1912.10077"], "source_meta": {"published_time": "20230509"}, "qid": "AutoScholarQuery_train_18582"} +{"question": "Which work introduces the prompt-based segmentation approach on Segment Anything Model (SAM)?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_18583"} +{"question": "Could you tell me some works about denoising diffusion implicit models (DDIM)?", "answer": ["Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2010.02502"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_18584"} +{"question": "Which work proposed the SketchGraphs dataset, a collection of sketches extracted from parametric CAD models?", "answer": ["SketchGraphs: A Large-Scale Dataset for Modeling Relational Geometry in Computer-Aided Design"], "answer_arxiv_id": ["2007.08506"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18585"} +{"question": "Who studied font generation, a subfield of OCR?", "answer": ["Word-As-Image for Semantic Typography", "Diff-Font: Diffusion Model for Robust One-Shot Font Generation", "Look Closer to Supervise Better: One-Shot Font Generation via Component-Based Discriminator", "Multiple Heads are Better than One: Few-shot Font Generation with Multiple Localized Experts"], "answer_arxiv_id": ["2303.01818", "2212.05895", "2205.00146", "2104.00887"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18586"} +{"question": "What papers studied the disentanglement of animal behavior in videos by learning distinct behavioral embeddings?", "answer": ["Disentangled Sequential Autoencoder", "Decomposing Motion and Content for Natural Video Sequence Prediction"], "answer_arxiv_id": ["1803.02991", "1706.08033"], "source_meta": {"published_time": "20230315"}, "qid": "AutoScholarQuery_train_18587"} +{"question": "Any works about hierarchical clustering under different objective functions?", "answer": ["Hierarchical Clustering: Objective Functions and Algorithms", "Hierarchical Clustering: a 0.585 Revenue Approximation", "Hierarchical Clustering better than Average-Linkage", "Bisect and Conquer: Hierarchical Clustering via Max-Uncut Bisection"], "answer_arxiv_id": ["1704.02147", "2006.01933", "1808.02227", "1912.06983"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_18588"} +{"question": "Which research papers propose methods utilizing memory network for anomaly detection in computer vision?", "answer": ["Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detection", "Learning Memory-guided Normality for Anomaly Detection", "A Hybrid Video Anomaly Detection Framework via Memory-Augmented Flow Reconstruction and Flow-Guided Frame Prediction"], "answer_arxiv_id": ["1904.02639", "2003.13228", "2108.06852"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18589"} +{"question": "Can you list some research papers that are about the emergence of large language models?", "answer": ["Language Models are Few-Shot Learners", "Training language models to follow instructions with human feedback", "PaLM: Scaling Language Modeling with Pathways", "OPT: Open Pre-trained Transformer Language Models", "GLM-130B: An Open Bilingual Pre-trained Model", "Emergent Abilities of Large Language Models", "LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2005.14165", "2203.02155", "2204.02311", "2205.01068", "2210.02414", "2206.07682", "2302.13971"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_18590"} +{"question": "Are there any researches that attempt to break the symmetries between slots using an iterative procedure?", "answer": ["Tagger: Deep Unsupervised Perceptual Grouping", "Neural Expectation Maximization", "Multi-Object Representation Learning with Iterative Variational Inference", "Conditional Set Generation with Transformers", "Unsupervised Learning of Compositional Energy Concepts", "Recurrent Independent Mechanisms", "Robust and Controllable Object-Centric Learning through Energy-based Models"], "answer_arxiv_id": ["1606.06724", "1708.03498", "1903.00450", "2006.16841", "2111.03042", "1909.10893v6", "2210.05519"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_18591"} +{"question": "Which works report on the development of closed-source LLMs, such as GPT-4, Claude, and PaLM?", "answer": ["GPT-4 Technical Report", "PaLM: Scaling Language Modeling with Pathways"], "answer_arxiv_id": ["2303.08774", "2204.02311"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_18592"} +{"question": "Which researches employ parameter-efficient fine-tuning methods that aim to improve OOD generalization performance?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Composed Fine-Tuning: Freezing Pre-Trained Denoising Autoencoders for Improved Generalization", "The Power of Scale for Parameter-Efficient Prompt Tuning", "Avoiding Inference Heuristics in Few-shot Prompt-based Finetuning", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["1902.00751", "2101.00190", "2006.16205", "2104.08691", "2109.04144", "2109.01134"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_18593"} +{"question": "Which studies propose methods of inferring object-centric representations as scene mixture components?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Multi-Object Representation Learning with Iterative Variational Inference", "Tagger: Deep Unsupervised Perceptual Grouping", "Neural Expectation Maximization", "Unsupervised Learning of Compositional Energy Concepts"], "answer_arxiv_id": ["1901.11390", "1903.00450", "1606.06724", "1708.03498", "2111.03042"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_18594"} +{"question": "What research works have been focused on learning the region-based visual features to enhance holistic visual features in Embedding-based Zero-Shot Learning?", "answer": ["Semantic-Guided Multi-Attention Localization for Zero-Shot Learning", "Attribute Prototype Network for Zero-Shot Learning", "TransZero++: Cross Attribute-Guided Transformer for Zero-Shot Learning"], "answer_arxiv_id": ["1903.00502", "2008.08290", "2112.08643"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_18595"} +{"question": "What works focus on using low precision for the gradients of the weights in distributed training?", "answer": ["signSGD: Compressed Optimisation for Non-Convex Problems"], "answer_arxiv_id": ["1802.04434"], "source_meta": {"published_time": "20211219"}, "qid": "AutoScholarQuery_train_18596"} +{"question": "Any works on solving the grouping problem by optimal transport in self-supervised learning?", "answer": ["Self-labelling via simultaneous clustering and representation learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments"], "answer_arxiv_id": ["1911.05371", "2006.09882"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_18597"} +{"question": "Which papers have identified a connection to the super-resolution problem where tensor methods are employed?", "answer": ["Towards a Mathematical Theory of Super-Resolution", "Super-resolution, Extremal Functions and the Condition Number of Vandermonde Matrices", "Algorithmic Foundations for the Diffraction Limit", "Super-Resolution Off the Grid"], "answer_arxiv_id": ["1203.5871", "1408.1681", "2004.07659", "1509.07943"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_18598"} +{"question": "What works are considered as quantization-aware training methods applying straight-through estimator for computing the gradient of rounding operations?", "answer": ["Deep Learning with Limited Numerical Precision", "Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference", "DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients", "PACT: Parameterized Clipping Activation for Quantized Neural Networks", "Relaxed Quantization for Discretized Neural Networks", "Learned Step Size Quantization", "LSQ+: Improving low-bit quantization through learnable offsets and better initialization", "Mixed Precision DNNs: All you need is a good parametrization", "Overcoming Oscillations in Quantization-Aware Training"], "answer_arxiv_id": ["1502.02551", "1712.05877", "1606.06160", "1805.06085", "1810.01875", "1902.08153", "2004.09576", "1905.11452v3", "2203.11086"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_18599"} +{"question": "What paper discusses generating multiple geometry-consistent descriptions using GPT-3?", "answer": ["Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2005.14165"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_18600"} +{"question": "Which works have been conducted on model pruning techniques that selectively prune weights based on their importance or magnitude?", "answer": ["A Comprehensive Study on Large-Scale Graph Training: Benchmarking and Rethinking"], "answer_arxiv_id": ["2210.07494"], "source_meta": {"published_time": "20220805"}, "qid": "AutoScholarQuery_train_18601"} +{"question": "What works discuss the necessity of precise subgoal generation in complex, discrete reasoning domains?", "answer": ["Fast and Precise: Adjusting Planning Horizon with Adaptive Subgoal Search"], "answer_arxiv_id": ["2206.00702"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_18602"} +{"question": "What papers have discussed GANs for time series synthesis?", "answer": ["Generative adversarial networks in time series: A survey and taxonomy", "Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs", "TTS-GAN: A Transformer-based Time-Series Generative Adversarial Network"], "answer_arxiv_id": ["2107.11098", "1706.02633", "2202.02691"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_18603"} +{"question": "Are there any papers where adversarial attacks are proposed to aggressively degrade the achievable accuracy of DNNs?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Towards Evaluating the Robustness of Neural Networks", "Square Attack: a query-efficient black-box adversarial attack via random search"], "answer_arxiv_id": ["1706.06083", "1608.04644", "1912.00049"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_18604"} +{"question": "Which works aimed at allowing the user to generate precise masks using SmartMask for object inpainting?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2112.10752"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_18605"} +{"question": "Could you provide examples of work studying embedding of symmetries in domains of fluid dynamics and turbulence modeling?", "answer": ["Incorporating Symmetry into Deep Dynamics Models for Improved Generalization", "Approximately Equivariant Networks for Imperfectly Symmetric Dynamics"], "answer_arxiv_id": ["2002.03061", "2201.11969"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_18606"} +{"question": "Which works utilized TCN-based models in time series forecasting?", "answer": ["An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling", "TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis"], "answer_arxiv_id": ["1803.01271", "2210.02186"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_18607"} +{"question": "What research papers propose improvement-based methods to iteratively improve a feasible initial solution?", "answer": ["Exploratory Combinatorial Optimization with Reinforcement Learning", "Learning Improvement Heuristics for Solving Routing Problems", "Learning to Iteratively Solve Routing Problems with Dual-Aspect Collaborative Transformer"], "answer_arxiv_id": ["1909.04063", "1912.05784", "2110.02544"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_18608"} +{"question": "What is the recent work that considered a setup where a small number of labelled actions are available along with a large unlabelled dataset?", "answer": ["Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos"], "answer_arxiv_id": ["2206.11795"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_18609"} +{"question": "Could you provide me some works demonstrating the ability of robotic systems to generalize to new tasks through language-based planning and control?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances", "BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning"], "answer_arxiv_id": ["2204.01691", "2202.02005"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_18610"} +{"question": "What research discusses the critical issue of how to build a scalable, expressive Transformer for learning node-pair interactions given the large number of node instances?", "answer": ["NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification"], "answer_arxiv_id": ["2306.08385"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_18611"} +{"question": "Could you provide me some works using models in model-based RL methods to generate synthetic data?", "answer": ["Algorithmic Framework for Model-based Deep Reinforcement Learning with Theoretical Guarantees", "Dream to Control: Learning Behaviors by Latent Imagination", "When to Trust Your Model: Model-Based Policy Optimization", "Model-based Policy Optimization with Unsupervised Model Adaptation"], "answer_arxiv_id": ["1807.03858", "1912.01603", "1906.08253", "2010.09546"], "source_meta": {"published_time": "20220918"}, "qid": "AutoScholarQuery_train_18612"} +{"question": "What studies proposed multi-resolution hash encoding for improving the learning of neural implicit field?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_18613"} +{"question": "Who constructed a robust model using an adversarial attack based on pixel-wise and perceptual losses?", "answer": ["Generalized Real-World Super-Resolution through Adversarial Robustness"], "answer_arxiv_id": ["2108.11505"], "source_meta": {"published_time": "20240523"}, "qid": "AutoScholarQuery_train_18614"} +{"question": "Can you give examples of studies that proposed Multiplicative filters networks?", "answer": ["FlexConv: Continuous Kernel Convolutions with Differentiable Kernel\n Sizes"], "answer_arxiv_id": ["2110.08059"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_18615"} +{"question": "Can you provide some studies that aimed at creating policies for continuous motion synthesis in virtual spaces?", "answer": ["Character Controllers Using Motion VAEs", "AMP: Adversarial Motion Priors for Stylized Physics-Based Character\n Control", "ASE: Large-Scale Reusable Adversarial Skill Embeddings for Physically\n Simulated Characters", "The Wanderings of Odysseus in 3D Scenes", "Synthesizing Diverse Human Motions in 3D Indoor Scenes", "AnySkill: Learning Open-Vocabulary Physical Skill for Interactive Agents"], "answer_arxiv_id": ["2103.14274", "2104.02180", "2205.01906", "2112.09251", "2305.12411", "2403.12835"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_18616"} +{"question": "Which works discuss the concept of optimizing the representation space with positive and negative pairs in contrastive learning?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance Discrimination", "A Survey on Contrastive Self-supervised Learning"], "answer_arxiv_id": ["1805.01978", "2011.00362"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_18617"} +{"question": "Could you name some studies using user input in the form of trimap in auxiliary input-based matting?", "answer": ["Deep Image Matting", "Natural Image Matting via Guided Contextual Attention", "Context-Aware Image Matting for Simultaneous Foreground and Alpha\n Estimation", "Long-Range Feature Propagating for Natural Image Matting"], "answer_arxiv_id": ["1703.03872", "2001.04069", "1909.09725", "2109.12252v1"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_18618"} +{"question": "Could you provide me some works that discuss the addition-based approaches of PEFT methods?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation", "Parameter-Efficient Transfer Learning for NLP", "GPT Understands, Too"], "answer_arxiv_id": ["2101.00190", "1902.00751", "2103.10385"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_18619"} +{"question": "What works have attempted to combine vision-language models and prompts?", "answer": ["Conditional Prompt Learning for Vision-Language Models", "CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "MaPLe: Multi-modal Prompt Learning"], "answer_arxiv_id": ["2203.05557", "2110.04544", "2210.03117"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_18620"} +{"question": "Which studies have applied deep feature fusion in a shared representation space for multi-modality 3D object detectors?", "answer": ["3D-CVF: Generating Joint Camera and LiDAR Features Using Cross-View\n Spatial Feature Fusion for 3D Object Detection", "Unifying Voxel-based Representation with Transformer for 3D Object\n Detection", "Voxel Field Fusion for 3D Object Detection"], "answer_arxiv_id": ["2004.12636", "2206.00630", "2205.15938"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_18621"} +{"question": "What researches introduced solutions to enhance the generalization ability for constructive NCO models in the inference stage?", "answer": ["Efficient Active Search for Combinatorial Optimization Problems", "Simulation-guided Beam Search for Neural Combinatorial Optimization"], "answer_arxiv_id": ["2106.05126", "2207.06190"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_18622"} +{"question": "Which studies employed the VAE or conditional GAN frameworks in the area of text-to-video models?", "answer": ["Sync-DRAW: Automatic Video Generation using Deep Recurrent Attentive\n Architectures", "To Create What You Tell: Generating Videos from Captions", "Video Generation From Text"], "answer_arxiv_id": ["1611.10314", "1804.08264", "1710.00421"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_18623"} +{"question": "What research does not satisfy learning to plan, using a learned model, or being evaluated in a complex environment?", "answer": ["Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model", "Value Prediction Network", "TreeQN and ATreeC: Differentiable Tree-Structured Models for Deep Reinforcement Learning", "Imagination-Augmented Agents for Deep Reinforcement Learning", "Value Iteration Networks", "An Investigation of Model-Free Planning", "Policy Gradient Search: Online Planning and Expert Iteration without Search Trees", "Scalable Online Planning via Reinforcement Learning Fine-Tuning", "Learning model-based planning from scratch", "Model-Based Planning with Discrete and Continuous Actions"], "answer_arxiv_id": ["1911.08265", "1707.03497", "1710.11417", "1707.06203", "1602.02867", "1901.03559", "1904.03646", "2109.15316", "1707.06170", "1705.07177"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_18624"} +{"question": "Could you provide me some works that explore the impacts of the KL regularization term on the representation in VAEs?", "answer": ["Emergence of Invariance and Disentanglement in Deep Representations", "Fixing a Broken ELBO"], "answer_arxiv_id": ["1706.01350", "1711.00464"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_18625"} +{"question": "What papers can be referred to as breakthroughs in Natural Language Processing (NLP)?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Language Models are Few-Shot Learners", "GPT-4 Technical Report", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "Universal Language Model Fine-tuning for Text Classification", "Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["1810.04805", "2005.14165", "2303.08774", "1910.10683", "1801.06146", "2109.01652"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_18626"} +{"question": "Which work supports merging of individually fine-tuned networks through solving a constrained customization problem?", "answer": ["Multi-Concept Customization of Text-to-Image Diffusion"], "answer_arxiv_id": ["2212.04488"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18627"} +{"question": "Could you name some studies that use complex transformations in backdoor attacks?", "answer": ["Input-Aware Dynamic Backdoor Attack", "Deep Feature Space Trojan Attack of Neural Networks by Controlled\n Detoxification", "Dynamic Backdoor Attacks Against Machine Learning Models", "Invisible Backdoor Attack with Sample-Specific Triggers"], "answer_arxiv_id": ["2010.08138v1", "2012.11212", "2003.03675", "2012.03816"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_18628"} +{"question": "Any works about the application of conformal prediction to image segmentation tasks?", "answer": ["Distribution-Free, Risk-Controlling Prediction Sets", "Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control"], "answer_arxiv_id": ["2101.02703", "2110.01052"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_18629"} +{"question": "Which studies are responsible for the creation of the kPam dataset?", "answer": ["kPAM: KeyPoint Affordances for Category-Level Robotic Manipulation", "LabelFusion: A Pipeline for Generating Ground Truth Labels for Real RGBD\n Data of Cluttered Scenes"], "answer_arxiv_id": ["1903.06684", "1707.04796"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_18630"} +{"question": "Could you provide some studies that demonstrate promising results in training neural constitutive laws directly from motion observations?", "answer": ["Learning Elastic Constitutive Material and Damping Models", "Learning Constitutive Relations from Indirect Observations Using Deep Neural Networks"], "answer_arxiv_id": ["1909.01875", "1905.12530v4"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_18631"} +{"question": "Can you mention studies about best ways to aggregate feature maps in image retrieval?", "answer": ["Visual Instance Retrieval with Deep Convolutional Networks", "Particular object retrieval with integral max-pooling of CNN activations", "Fine-tuning CNN Image Retrieval with No Human Annotation"], "answer_arxiv_id": ["1412.6574", "1511.05879", "1711.02512"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_18632"} +{"question": "Could you tell me about the studies using individual MLPs to represent the deformation field and a canonical field?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "PREF: Predictability Regularized Neural Motion Fields", "DeformToon3D: Deformable 3D Toonification from Neural Radiance Fields", "NeRF-DS: Neural Radiance Fields for Dynamic Specular Objects", "Unbiased 4D: Monocular 4D Reconstruction with a Neural Deformation Model"], "answer_arxiv_id": ["2011.13961", "2209.10691", "2309.04410", "2303.14435", "2206.08368"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18633"} +{"question": "Which papers discuss the successes of autoregressive models as constructive heuristic solvers for combinatorial optimization (CO) problems?", "answer": ["Attention Is All You Need", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1706.03762", "2005.14165"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_18634"} +{"question": "Could you provide me studies about the definition of functions for one-hidden layer ReLU networks in the multivariate case?", "answer": ["A Function Space View of Bounded Norm Infinite Width ReLU Nets: The Multivariate Case"], "answer_arxiv_id": ["1910.01635"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_18635"} +{"question": "Which research works point out that ties in document scoring can affect the optimization procedure and lead to an overestimated ranking quality?", "answer": ["StochasticRank: Global Optimization of Scale-Free Discrete Functions"], "answer_arxiv_id": ["2003.02122"], "source_meta": {"published_time": "20220404"}, "qid": "AutoScholarQuery_train_18636"} +{"question": "What work did not impose the constant weight assumption and is slightly more general to construct a few scalar functions that characterize the dynamics as compared to the existing works?", "answer": ["A Comparative Analysis of Expected and Distributional Reinforcement Learning"], "answer_arxiv_id": ["1901.11084"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18637"} +{"question": "What work using offline RL mimics the classifier-guidance method but lacks a detailed convergence discussion?", "answer": ["Planning with Diffusion for Flexible Behavior Synthesis"], "answer_arxiv_id": ["2205.09991"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_18638"} +{"question": "Which work proposed a video demoiréing method with a relation-based temporal consistency loss?", "answer": ["Video Demoireing with Relation-Based Temporal Consistency"], "answer_arxiv_id": ["2204.02957v1"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_18639"} +{"question": "Could you provide me some works that studied the properties of creator-side equilibrium in the C3 game, under given creator incentives?", "answer": ["Modeling Content Creator Incentives on Algorithm-Curated Platforms", "Supply-Side Equilibria in Recommender Systems"], "answer_arxiv_id": ["2206.13102", "2206.13489"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_18640"} +{"question": "Could you provide some works that leverage CAD models in single-view methods?", "answer": ["AutoShape: Real-Time Shape-Aware Monocular 3D Object Detection", "ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape", "Monocular 3D Object Detection via Geometric Reasoning on Keypoints"], "answer_arxiv_id": ["2108.11127", "1812.02781", "1905.05618"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_18641"} +{"question": "What papers introduced, investigated, and explored the visual prompt tuning?", "answer": ["Learning to Prompt for Vision-Language Models", "Exploring Visual Prompts for Adapting Large-Scale Models", "PromptCAL: Contrastive Affinity Learning via Auxiliary Prompts for\n Generalized Novel Category Discovery", "MaPLe: Multi-modal Prompt Learning", "Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2109.01134", "2203.17274", "2212.05590", "2210.03117", "2203.05557"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_18642"} +{"question": "Which research studies have proposed techniques to alleviate communication overhead in Federated Learning?", "answer": ["Federated Learning: Strategies for Improving Communication Efficiency", "Sparse Binary Compression: Towards Distributed Deep Learning with minimal Communication", "Ternary Compression for Communication-Efficient Federated Learning", "Communication-Efficient On-Device Machine Learning: Federated Distillation and Augmentation under Non-IID Private Data", "Distillation-Based Semi-Supervised Federated Learning for Communication-Efficient Collaborative Training with Non-IID Private Data", "Communication-Efficient Federated Distillation", "Ensemble Distillation for Robust Model Fusion in Federated Learning", "Client Selection for Federated Learning with Heterogeneous Resources in Mobile Edge"], "answer_arxiv_id": ["1610.05492", "1805.08768", "2003.03564", "1811.11479", "2008.06180", "2012.00632", "2006.07242", "1804.08333"], "source_meta": {"published_time": "20231113"}, "qid": "AutoScholarQuery_train_18643"} +{"question": "Which papers studied the concept of a covered edge in a directed acyclic graph?", "answer": ["A Transformational Characterization of Equivalent Bayesian Network Structures"], "answer_arxiv_id": ["1302.4938v1"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18644"} +{"question": "Which researches provided frameworks or discussions about the trade-off between efficacy and efficiency on non-amortized methods and RTX?", "answer": ["Efficient XAI Techniques: A Taxonomic Survey"], "answer_arxiv_id": ["2302.03225"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_18645"} +{"question": "Which papers discuss methods to improve contrastive learning by mining nearest neighbour?", "answer": ["With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations", "Mean Shift for Self-Supervised Learning", "Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction"], "answer_arxiv_id": ["2104.14548", "2105.07269", "2102.10106"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_18646"} +{"question": "Which works utilized unlabeled samples predicted as minority classes more frequently for iterative self-training?", "answer": ["CReST: A Class-Rebalancing Self-Training Framework for Imbalanced\n Semi-Supervised Learning"], "answer_arxiv_id": ["2102.09559"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_18647"} +{"question": "Which studies pointed out that enforcing symmetry where it is unclear can be harmful to performance?", "answer": ["An intriguing failing of convolutional neural networks and the CoordConv solution", "Approximately Equivariant Networks for Imperfectly Symmetric Dynamics"], "answer_arxiv_id": ["1807.03247", "2201.11969"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_18648"} +{"question": "What papers utilize pure image-text data as supervision or reformulate labeled data into image-text data?", "answer": ["GroupViT: Semantic Segmentation Emerges from Text Supervision", "Language-driven Semantic Segmentation"], "answer_arxiv_id": ["2202.11094", "2201.03546"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_18649"} +{"question": "Which study proposed DMASNet for better cross-domain transferability in shadow generation?", "answer": ["Shadow Generation with Decomposed Mask Prediction and Attentive Shadow\n Filling"], "answer_arxiv_id": ["2306.17358"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_18650"} +{"question": "What research proposed a setting similar to ours, called 'Hybrid RL'?", "answer": ["Hybrid RL: Using Both Offline and Online Data Can Make RL Efficient"], "answer_arxiv_id": ["2210.06718"], "source_meta": {"published_time": "20221109"}, "qid": "AutoScholarQuery_train_18651"} +{"question": "Which works discussed model-based reinforcement learning?", "answer": ["When to Trust Your Model: Model-Based Policy Optimization", "World Models", "Model-based Reinforcement Learning and the Eluder Dimension"], "answer_arxiv_id": ["1906.08253", "1803.10122", "1406.1853"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_18652"} +{"question": "Could you give me examples of works that proposed to use efficient representations like voxel grids and triplanes to improve the spatial resolution of the scene representation?", "answer": ["VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids", "Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["2206.07695", "2112.07945"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_18653"} +{"question": "What are some of the synthetic benchmarks used for evaluating long-range motion estimation?", "answer": ["FlowNet: Learning Optical Flow with Convolutional Networks", "A Large Dataset to Train Convolutional Networks for Disparity, Optical\n Flow, and Scene Flow Estimation", "AutoFlow: Learning a Better Training Set for Optical Flow", "Kubric: A scalable dataset generator", "PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point\n Tracking", "Particle Video Revisited: Tracking Through Occlusions Using Point\n Trajectories"], "answer_arxiv_id": ["1504.06852", "1512.02134", "2104.14544", "2203.03570", "2307.15055", "2204.04153"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_18654"} +{"question": "Could you provide me the reference that used causal inference to measure the effect of pre-training data statistics on QA performance?", "answer": ["Measuring Causal Effects of Data Statistics on Language Model’s ‘Factual’ Predictions"], "answer_arxiv_id": ["2207.14251"], "source_meta": {"published_time": "20221115"}, "qid": "AutoScholarQuery_train_18655"} +{"question": "Which work proposed an architecture to encode the image but decode a depth map under DDP approach?", "answer": ["DDP: Diffusion Model for Dense Visual Prediction"], "answer_arxiv_id": ["2303.17559"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18656"} +{"question": "What are the papers that analysed and proposed minibatch SGD, FedAvg/Local SGD or their extensions?", "answer": ["A Unified Theory of Decentralized SGD with Changing Topology and Local Updates", "On the Convergence of FedAvg on Non-IID Data", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "Is Local SGD Better than Minibatch SGD?", "Minibatch vs Local SGD for Heterogeneous Distributed Learning", "Federated Accelerated Stochastic Gradient Descent", "Communication-Efficient Learning of Deep Networks from Decentralized Data", "SCAFFOLD: Stochastic Controlled Averaging for Federated Learning"], "answer_arxiv_id": ["2003.10422", "1907.02189", "1910.06378", "2002.07839", "2006.04735", "2006.08950", "1602.05629", "1910.06378"], "source_meta": {"published_time": "20210617"}, "qid": "AutoScholarQuery_train_18657"} +{"question": "What is the pioneering work in the field of 3D representation learning using point-based methods?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation"], "answer_arxiv_id": ["1612.00593"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_18658"} +{"question": "What studies proposed a universal data-driven optimal likelihood hypothesis test with linear spectral statistics for a spiked random model?", "answer": ["Detection of Signal in the Spiked Rectangular Models"], "answer_arxiv_id": ["2104.13517"], "source_meta": {"published_time": "20240106"}, "qid": "AutoScholarQuery_train_18659"} +{"question": "Which works focus on the application of DRO in handling data distribution shift?", "answer": ["Certifying Some Distributional Robustness with Principled Adversarial Training", "Distributionally Robust Learning with Stable Adversarial Training"], "answer_arxiv_id": ["1710.10571v5", "2106.15791"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_18660"} +{"question": "Which works mention the information loss and disregarding inter-observation dynamics when discretizing the timeline into time bins or padding missing values?", "answer": ["MiME: Multilevel Medical Embedding of Electronic Health Records for Predictive Healthcare", "Discrete Event, Continuous Time RNNs"], "answer_arxiv_id": ["1810.09593", "1710.04110v1"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_18661"} +{"question": "What papers have proposed various methods to generate shapes from parts in assembly-based 3D modeling?", "answer": ["ShapeAssembly: Learning to Generate Programs for 3D Shape Structure\n Synthesis", "ABC: A Big CAD Model Dataset For Geometric Deep Learning", "Compositionally Generalizable 3D Structure Prediction", "SDM-NET: Deep Generative Network for Structured Deformable Mesh", "Learning Generative Models of Shape Handles"], "answer_arxiv_id": ["2009.08026", "1812.06216", "2012.02493", "1908.04520", "2004.03028"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_18662"} +{"question": "Which references talked about the prevalence of vision models that are limited to around 1 billion parameters?", "answer": ["EVA: Exploring the Limits of Masked Visual Representation Learning at\n Scale"], "answer_arxiv_id": ["2211.07636"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18663"} +{"question": "Which research papers implemented 3D Gaussian Splatting for synthesizing static 3D scenes?", "answer": ["DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content\n Creation", "Text-to-3D using Gaussian Splatting"], "answer_arxiv_id": ["2309.16653", "2309.16585"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18664"} +{"question": "Could you mention the work that tackles extrinsic transformation between a 360° camera and a LiDAR?", "answer": ["INF: Implicit Neural Fusion for LiDAR and Camera"], "answer_arxiv_id": ["2308.14414"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_18665"} +{"question": "What studies use graph augmentation to generate multiple views in graph contrastive learning?", "answer": ["Graph Barlow Twins: A self-supervised representation learning framework for graphs", "Adversarial Graph Augmentation to Improve Graph Contrastive Learning", "Graph Contrastive Learning Automated", "Adversarial Graph Contrastive Learning with Information Regularization", "Data Augmentation for Deep Graph Learning: A Survey"], "answer_arxiv_id": ["2106.02466", "2106.05819", "2106.07594", "2202.06491", "2202.08235"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_18666"} +{"question": "Which works are associated with handcrafted optimizers?", "answer": ["signSGD: Compressed Optimisation for Non-Convex Problems", "Adafactor: Adaptive Learning Rates with Sublinear Memory Cost", "AdaBelief Optimizer: Adapting Stepsizes by the Belief in Observed Gradients", "Scalable Second Order Optimization for Deep Learning", "On the Variance of the Adaptive Learning Rate and Beyond", "On the convergence of Adam and Beyond", "Shampoo: Preconditioned Stochastic Tensor Optimization", "Quasi-hyperbolic momentum and Adam for deep learning"], "answer_arxiv_id": ["1802.04434", "1804.04235", "2010.07468", "2002.09018", "1908.03265", "1904.09237", "1802.09568", "1810.06801"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_18667"} +{"question": "Can you name some author studies that addressed the imbalance of the gradient contributions in PINNs by developing different weighting strategies for the individual components of the loss?", "answer": ["Optimally weighted loss functions for solving PDEs with Neural Networks", "When and why PINNs fail to train: A neural tangent kernel perspective"], "answer_arxiv_id": ["2002.06269", "2007.14527"], "source_meta": {"published_time": "20230225"}, "qid": "AutoScholarQuery_train_18668"} +{"question": "Could you tell me about existing works on cooperative multi-agent RL designed for environments that are non-stationary and/or heterogeneous?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments", "The Surprising Effectiveness of PPO in Cooperative Multi-Agent Games", "Trust Region Policy Optimisation in Multi-Agent Reinforcement Learning", "Stateful active facilitator: Coordination and Environmental Heterogeneity in Cooperative Multi-Agent Reinforcement Learning", "Federated Reinforcement Learning with Environment Heterogeneity"], "answer_arxiv_id": ["1706.02275", "2103.01955", "2109.11251", "2210.03022", "2204.02634v1"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_18669"} +{"question": "Which works utilized normalizing flow or GAN to generate realistic noisy images with paired training data?", "answer": ["Noise Flow: Noise Modeling with Conditional Normalizing Flows", "Modeling sRGB Camera Noise with Normalizing Flows", "Dual Adversarial Network: Toward Real-world Noise Removal and Noise\n Generation"], "answer_arxiv_id": ["1908.08453", "2206.00812", "2007.05946"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_18670"} +{"question": "In what studies do we find the robust optimization and multi source domain adaptation methods?", "answer": ["Domain-Adversarial Training of Neural Networks", "Domain Generalization via Invariant Feature Representation", "A Theory of Multiple-Source Adaptation with Limited Target Labeled Data", "Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach", "Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "In Search of Lost Domain Generalization"], "answer_arxiv_id": ["1505.07818", "1301.2115", "2007.09762", "1610.03425", "1911.08731", "2007.01434"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_18671"} +{"question": "Could you provide me the studies that adopted temporal convolution network and transformer model to estimate per-frame action probability?", "answer": ["Skeleton-Based Action Segmentation with Multi-Stage Spatial-Temporal Graph Convolutional Neural Networks", "LocATe: End-to-end Localization of Actions in 3D with Transformers"], "answer_arxiv_id": ["2202.01727", "2203.10719"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_18672"} +{"question": "Could you mention some papers that utilize recurrent networks for fast adaption in the meta-RL setting?", "answer": ["RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning"], "answer_arxiv_id": ["1611.02779"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_18673"} +{"question": "Are there any studies that have used large teacher models to guide MAE pretraining?", "answer": ["MILAN: Masked Image Pretraining on Language Assisted Representation", "Masked Autoencoders Enable Efficient Knowledge Distillers"], "answer_arxiv_id": ["2208.06049", "2208.12256"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_18674"} +{"question": "What studies discuss the rapid scale-up of vision datasets from thousands to billion examples?", "answer": ["Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "Scaling Vision Transformers"], "answer_arxiv_id": ["2102.05918", "2106.04560"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_18675"} +{"question": "Are there any research on visual autonomous driving that were based on Image-LiDAR pairs?", "answer": ["Is Pseudo-Lidar needed for Monocular 3D Object detection?", "UniPAD: A Universal Pre-training Paradigm for Autonomous Driving"], "answer_arxiv_id": ["2108.06417", "2310.08370"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_18676"} +{"question": "What works proposed methods for certification of Bayesian neural networks (BNNs)?", "answer": ["Probabilistic Safety for Bayesian Neural Networks", "Make Sure You’re Unsure: A Framework for Verifying Probabilistic Specifications", "Infinite Time Horizon Safety of Bayesian Neural Networks"], "answer_arxiv_id": ["2004.10281", "2102.09479", "2111.03165"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_18677"} +{"question": "Who are the authors of works on adversarial reprogramming?", "answer": ["Adversarial Reprogramming of Neural Networks", "Transfer Learning without Knowing: Reprogramming Black-box Machine Learning Models with Scarce Data and Limited Resources", "Voice2Series: Reprogramming Acoustic Models for Time Series Classification"], "answer_arxiv_id": ["1806.11146", "2007.08714", "2106.09296"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_18678"} +{"question": "What study argued that the problem is as hard as online learning a one-dimensional threshold function in the linear case?", "answer": ["Efficient and robust algorithms for adversarial linear contextual bandits"], "answer_arxiv_id": ["2002.00287"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_18679"} +{"question": "What are the studies considering 3D geometries of compounds for drug design?", "answer": ["Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules"], "answer_arxiv_id": ["1906.00957"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18680"} +{"question": "Which research papers provided a Riemannian perspective on the latent space of generative models?", "answer": ["Latent Space Oddity: on the Curvature of Deep Generative Models", "Metrics for Deep Generative Models"], "answer_arxiv_id": ["1710.11379", "1711.01204v2"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_18681"} +{"question": "What papers describe the categories of self-supervised learning (SSL)?", "answer": ["Graph Self-Supervised Learning: A Survey", "Self-Supervised Learning of Graph Neural Networks: A Unified Review", "Self-supervised Learning: Generative or Contrastive"], "answer_arxiv_id": ["2103.00111", "2102.10757", "2006.08218"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_18682"} +{"question": "Which research provided empirical evidence that decision boundaries of classifiers cling onto the data manifold?", "answer": ["The Dimpled Manifold Model of Adversarial Examples in Machine Learning"], "answer_arxiv_id": ["2106.10151"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_18683"} +{"question": "What works involve tasks such as inverse rendering and intrinsic image decomposition?", "answer": ["Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying\n Lighting and SVBRDF from a Single Image", "IRISformer: Dense Vision Transformers for Single-Image Inverse Rendering\n in Indoor Scenes"], "answer_arxiv_id": ["1905.02722", "2206.08423"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_18684"} +{"question": "What works have engaged methods of estimating flow in real-world videos using pretrained models?", "answer": ["RAFT: Recurrent All-Pairs Field Transforms for Optical Flow", "GMFlow: Learning Optical Flow via Global Matching", "Unifying Flow, Stereo and Depth Estimation"], "answer_arxiv_id": ["2003.12039", "2111.13680", "2211.05783"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_18685"} +{"question": "What works have critiqued standard metrics for generative evaluation?", "answer": ["The Glass Ceiling of Automatic Evaluation in Natural Language Generation", "Of Human Criteria and Automatic Metrics: A Benchmark of the Evaluation\n of Story Generation", "Holistic Evaluation of Language Models"], "answer_arxiv_id": ["2208.14585", "2208.11646", "2211.09110"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_18686"} +{"question": "What are the original papers that conformal prediction originated from?", "answer": ["A Tutorial on Conformal Prediction"], "answer_arxiv_id": ["0706.3188"], "source_meta": {"published_time": "20221227"}, "qid": "AutoScholarQuery_train_18687"} +{"question": "Which papers provide theoretical frameworks for real-time reinforcement learning in asynchronous environments?", "answer": ["Real-Time Reinforcement Learning", "Reactive Reinforcement Learning in Asynchronous Environments"], "answer_arxiv_id": ["1911.04448", "1802.06139"], "source_meta": {"published_time": "20210610"}, "qid": "AutoScholarQuery_train_18688"} +{"question": "Could you provide me some works that discuss open-vocabulary object detection?", "answer": ["Open-Vocabulary Object Detection Using Captions", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "Florence: A New Foundation Model for Computer Vision", "LiT: Zero-Shot Transfer with Locked-image text Tuning", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision", "Learning Transferable Visual Models From Natural Language Supervision", "RegionCLIP: Region-based Language-Image Pretraining", "Simple Open-Vocabulary Object Detection with Vision Transformers", "Exploiting Unlabeled Data with Vision and Language Models for Object Detection", "Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "Detecting Twenty-thousand Classes using Image-level Supervision"], "answer_arxiv_id": ["2011.10678", "2205.01917", "2111.11432", "2111.07991", "2102.05918", "2103.00020", "2112.09106", "2205.06230", "2207.08954", "2203.14940", "2112.12143", "2201.02605"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_18689"} +{"question": "What studies include transferring pre-trained AlexNet features to downstream tasks?", "answer": ["DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition"], "answer_arxiv_id": ["1310.1531v1"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_18690"} +{"question": "Could you provide me the study where over-robustness was discussed under the name of invariance-based adversarial examples for the image domain?", "answer": ["Fundamental Tradeoffs between Invariance and Sensitivity to Adversarial Perturbations"], "answer_arxiv_id": ["2002.04599"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_18691"} +{"question": "Could you mention papers which discussed obtaining a texture atlas using human shape models that unwrap to a unified UV space?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "REALY: Rethinking the Evaluation of 3D Face Reconstruction"], "answer_arxiv_id": ["1904.05866", "2203.09729"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_18692"} +{"question": "Which papers discuss the improvement of global 3D consistency and visual quality in perpetual view generation models?", "answer": ["Persistent Nature: A Generative Model of Unbounded 3D Worlds", "DiffDreamer: Towards Consistent Unsupervised Single-view Scene\n Extrapolation with Conditional Diffusion Models"], "answer_arxiv_id": ["2303.13515", "2211.12131"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_18693"} +{"question": "Which papers study binary labeling of graph nodes in the online learning setting?", "answer": ["Online Prediction of Switching Graph Labelings with Cluster Specialists"], "answer_arxiv_id": ["1806.06439"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18694"} +{"question": "What research works have improved accuracy of implicit functions using additional priors or losses related to depth, normals and multi-view consistency?", "answer": ["MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction", "Neural RGB-D Surface Reconstruction", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM", "NeuRIS: Neural Reconstruction of Indoor Scenes Using Normal Priors", "Neural 3D Scene Reconstruction with the Manhattan-world Assumption", "Geo-Neus: Geometry-Consistent Neural Implicit Surfaces Learning for Multi-view Reconstruction"], "answer_arxiv_id": ["2206.00665", "2104.04532", "2112.12130", "2206.13597", "2205.02836", "2205.15848"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_18695"} +{"question": "Could you provide me some examples of research about continual learning methods like Elastic Weight Consolidation, Experience Replay, and Maximally Interfered Replay?", "answer": ["On Continual Model Refinement in Out-of-Distribution Data Streams", "Overcoming catastrophic forgetting in neural networks", "Experience Replay for Continual Learning", "Online Continual Learning with Maximally Interfered Retrieval"], "answer_arxiv_id": ["2205.02014v1", "1612.00796", "1811.11682", "1908.04742"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_18696"} +{"question": "What works have proposed different measures for topic diversity, redundancy, or overlap?", "answer": ["Topic Modeling in Embedding Spaces"], "answer_arxiv_id": ["1907.04907"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_18697"} +{"question": "Which works considered contrastive estimation in the context of disentanglement of latent factors?", "answer": ["InfoGAN-CR and ModelCentrality: Self-supervised Model Training and Selection for Disentangling GANs", "Self-Supervised Learning Disentangled Group Representation as Feature"], "answer_arxiv_id": ["1906.06034", "2110.15255"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18698"} +{"question": "Could you mention some known streaming algorithms that work in the insertion-only streaming model?", "answer": ["Streaming Non-monotone Submodular Maximization: Personalized Video Summarization on the Fly", "Do Less, Get More: Streaming Submodular Maximization with Subsampling"], "answer_arxiv_id": ["1706.03583", "1802.07098"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_18699"} +{"question": "What research papers propose segmenting large city-scale scenes into blocks and applying NeRF within each for 3D scene generation?", "answer": ["Block-NeRF: Scalable Large Scene Neural View Synthesis", "Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual\n Fly-Throughs", "SUDS: Scalable Urban Dynamic Scenes"], "answer_arxiv_id": ["2202.05263", "2112.10703", "2303.14536"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_18700"} +{"question": "What works provide a comprehensive review on learning algorithms for bandits?", "answer": ["A Tutorial on Thompson Sampling"], "answer_arxiv_id": ["1707.02038"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_18701"} +{"question": "Which papers discuss the capabilities of Diffusion models in the field of image synthesis?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2006.11239", "2011.13456", "2105.05233"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_18702"} +{"question": "What research papers extend DDPG and TRPO in the context of multi-agent policy gradient?", "answer": ["Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments", "Trust Region Policy Optimisation in Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1706.02275", "2109.11251"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_18703"} +{"question": "What works introduced reparameterization method to YOLO-series models and proposed EfficientRep Backbone and Rep-PAN Neck?", "answer": ["YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications", "YOLOv6 v3.0: A Full-Scale Reloading"], "answer_arxiv_id": ["2209.02976", "2301.05586"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_18704"} +{"question": "Could you mention some works that discussed attribution in the context of providing external evidence supporting the claims made by the model?", "answer": ["Measuring Attribution in Natural Language Generation Models", "A Survey of Large Language Models Attribution"], "answer_arxiv_id": ["2112.12870", "2311.03731"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18705"} +{"question": "Which study does the sketch generator in the SQA game base on?", "answer": ["Learning to generate line drawings that convey geometry and semantics"], "answer_arxiv_id": ["2203.12691"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_18706"} +{"question": "Which works feature the application of consistency regularization and pseudo-labeling in semi-supervised learning?", "answer": ["MixMatch: A Holistic Approach to Semi-Supervised Learning", "AdaMatch: A Unified Approach to Semi-Supervised Learning and Domain Adaptation", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["1905.02249", "2106.04732", "2001.07685v2"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_18707"} +{"question": "Which papers have studies about ℓ2 mean estimation problem under communication constraints?", "answer": ["rTop-k: A Statistical Estimation Approach to Distributed SGD", "vqSGD: Vector Quantized Stochastic Gradient Descent", "Distributed Mean Estimation with Limited Communication", "DRIVE: One-bit Distributed Mean Estimation", "Communication-Efficient Algorithms for Statistical Optimization"], "answer_arxiv_id": ["2005.10761", "1911.07971v4", "1611.00429", "2105.08339", "1209.4129"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18708"} +{"question": "Could you provide some works that introduced novel approaches in Neural Implicit Functions like implicit moving least-squares surfaces, a differentiable Poisson solver, a complex Gabor wavelet, and a level set alignment loss?", "answer": ["Deep Implicit Moving Least-Squares Functions for 3D Reconstruction", "Shape As Points: A Differentiable Poisson Solver", "WIRE: Wavelet Implicit Neural Representations", "Towards Better Gradient Consistency for Neural Signed Distance Functions\n via Level Set Alignment"], "answer_arxiv_id": ["2103.12266", "2106.03452", "2301.05187", "2305.11601"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_18709"} +{"question": "Which works modify the model architecture during training to train it with self-supervised loss for test time training (TTT)?", "answer": ["Test-Time Training with Self-Supervision for Generalization under\n Distribution Shifts", "TTAPS: Test-Time Adaption by Aligning Prototypes using Self-Supervision", "MT3: Meta Test-Time Training for Self-Supervised Test-Time Adaption"], "answer_arxiv_id": ["1909.13231", "2205.08731", "2103.16201"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_18710"} +{"question": "Could you name some works on unsupervised methods for image segmentation?", "answer": ["Unsupervised Semantic Segmentation with Self-supervised Object-centric\n Representations", "PiCIE: Unsupervised Semantic Segmentation using Invariance and\n Equivariance in Clustering", "Unsupervised Semantic Segmentation by Distilling Feature Correspondences", "SegSort: Segmentation by Discriminative Sorting of Segments", "Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised\n Semantic Segmentation and Localization"], "answer_arxiv_id": ["2207.05027", "2103.17070", "2203.08414", "1910.06962", "2205.07839"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_18711"} +{"question": "Which paper introduces a benchmark focused on detecting GAN-generated images using continual learning?", "answer": ["A Continual Deepfake Detection Benchmark: Dataset, Methods, and Essentials"], "answer_arxiv_id": ["2205.05467"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_18712"} +{"question": "Which papers classify the three categories of model adaptation methods?", "answer": ["VL-Adapter: Parameter-Efficient Transfer Learning for Vision-and-Language Tasks"], "answer_arxiv_id": ["2112.06825"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_18713"} +{"question": "What research explores utilizing a frozen text-to-image diffusion model for multimodal generation capabilities in LLMs?", "answer": ["Any-to-Any Generation via Composable Diffusion"], "answer_arxiv_id": ["2305.11846"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18714"} +{"question": "What studies utilized the outlier-robust Wasserstein distance for their analysis?", "answer": ["Outlier-Robust Optimal Transport: Duality, Structure, and Statistical Analysis", "Robust Estimation under the Wasserstein Distance"], "answer_arxiv_id": ["2111.01361", "2302.01237"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_18715"} +{"question": "What is the foundational study behind the DCLS method's consideration of the effective receptive field (ERF)?", "answer": ["Understanding the Effective Receptive Field in Deep Convolutional Neural Networks"], "answer_arxiv_id": ["1701.04128"], "source_meta": {"published_time": "20211207"}, "qid": "AutoScholarQuery_train_18716"} +{"question": "Could you specify the paper that focused on low-frequency relations?", "answer": ["Learning of Visual Relations: The Devil is in the Tails"], "answer_arxiv_id": ["2108.09668"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_18717"} +{"question": "Could you provide me some studies that explore self-supervised learning with language, vision, and sound?", "answer": ["VATT: Transformers for Multimodal Self-Supervised Learning from Raw\n Video, Audio and Text"], "answer_arxiv_id": ["2104.11178"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_18718"} +{"question": "Which papers introduced approaches that use low-rank tensor to model time-varying AR processes?", "answer": ["Time-varying Autoregression with Low Rank Tensors", "Bayesian Time-Varying Tensor Vector Autoregressive Models for Dynamic Effective Connectivity"], "answer_arxiv_id": ["1905.08389", "2106.14083"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_18719"} +{"question": "In what papers the researcher used Transformer-based architectures to learn multimodal representation for language and image modalities?", "answer": ["Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language\n Representation Learning", "ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual\n Concepts"], "answer_arxiv_id": ["2104.03135", "2102.03334", "2107.07651", "2111.08276"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_18720"} +{"question": "Which works are about using Implicit Neural Spatial Representation in strictly spatially dependent PDEs?", "answer": ["NTopo: Mesh-free Topology Optimization using Implicit Neural Representations"], "answer_arxiv_id": ["2102.10782"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_18721"} +{"question": "Are there any studies of leveraging LLMs to extract object affordances to build agents for tidying up an indoor environment?", "answer": ["Housekeep: Tidying Virtual Households using Commonsense Reasoning", "TIDEE: Tidying Up Novel Rooms using Visuo-Semantic Commonsense Priors", "TidyBot: Personalized Robot Assistance with Large Language Models"], "answer_arxiv_id": ["2205.10712", "2207.10761", "2305.05658"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_18722"} +{"question": "Which works improved the training objectives, architecture, and sampling process for Denoising diffusion probabilistic models (DDPMs)?", "answer": ["Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Tackling the Generative Learning Trilemma with Denoising Diffusion GANs", "Elucidating the Design Space of Diffusion-Based Generative Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Improved Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2105.05233", "2112.07804", "2206.00364", "2112.10752", "2102.09672", "2010.02502"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_18723"} +{"question": "Which research introduced a hyperbolic neural network for modeling tree-structured data?", "answer": ["Hyperbolic Neural Networks++"], "answer_arxiv_id": ["2006.08210"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_18724"} +{"question": "Could you provide me studies about 'word order' and whether it really affects multilingual ability?", "answer": ["How multilingual is Multilingual BERT?", "When is BERT Multilingual? Isolating Crucial Ingredients for\n Cross-lingual Transfer"], "answer_arxiv_id": ["1906.01502", "2110.14782"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_18725"} +{"question": "Could you provide me some studies that approximate variational Bayesian inference using dropout or ensemble?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1506.02142"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_18726"} +{"question": "What papers are related to the concept of BRIO for summarization, which reuses the generation model as the evaluation model?", "answer": ["BRIO: Bringing Order to Abstractive Summarization", "On Learning to Summarize with Large Language Models as References"], "answer_arxiv_id": ["2203.16804", "2305.14239"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_18727"} +{"question": "Could you provide works that employed diffusion guidance for text-guided image generation and editing?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Blended Diffusion for Text-driven Editing of Natural Images"], "answer_arxiv_id": ["2112.10741", "2111.14818"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_18728"} +{"question": "What works have used special training objectives to limit artifacts and reliably assess models?", "answer": ["Don't Take the Premise for Granted: Mitigating Artifacts in Natural\n Language Inference"], "answer_arxiv_id": ["1907.04380"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18729"} +{"question": "Is there any study about open-source AI assistant following a similar approach to ChatGPT by collecting its own data?", "answer": ["OpenAssistant Conversations - Democratizing Large Language Model Alignment"], "answer_arxiv_id": ["2304.07327"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_18730"} +{"question": "What works used low-rank adaption for a model using both text instruction data and vision-language instruction data?", "answer": ["mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality"], "answer_arxiv_id": ["2304.14178"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_18731"} +{"question": "Which studies utilised an RL approach for prompt optimization?", "answer": ["RLPrompt: Optimizing Discrete Text Prompts with Reinforcement Learning"], "answer_arxiv_id": ["2205.12548"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_18732"} +{"question": "Which papers fall under the category of Offline Model-based Optimization methods that generate novel designs through generative models?", "answer": ["Model Inversion Networks for Model-Based Optimization", "Conditioning by adaptive sampling for robust design", "Autofocused oracles for model-based design", "Bootstrapped Training of Score-Conditioned Generator for Offline Design of Biological Sequences"], "answer_arxiv_id": ["1912.13464", "1901.10060", "2006.08052", "2306.03111"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_18733"} +{"question": "Which papers provide knowledge on poisoning attacks in the context of training data manipulation?", "answer": ["Poisoning Attacks against Support Vector Machines", "Data Poisoning attack against Unsupervised Node Embedding Methods", "Data Poisoning Attacks against Online Learning", "Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learning", "Transferable Clean-Label Poisoning Attacks on Deep Neural Nets"], "answer_arxiv_id": ["1206.6389", "1810.12881", "1808.08994", "1804.00308", "1905.05897"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_18734"} +{"question": "Which works address computational difficulties in planning in POMDPs?", "answer": ["Nonapproximability Results for Partially Observable Markov Decision Processes"], "answer_arxiv_id": ["1106.0242v1"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_18735"} +{"question": "Which works pose the first-known lower bound on sample complexity for FC-BAI with ε-global DP?", "answer": ["Differentially Private Contextual Linear Bandits", "When Privacy Meets Partial Information: A Refined Analysis of Differentially Private Bandits"], "answer_arxiv_id": ["1810.00068", "2209.02570"], "source_meta": {"published_time": "20230905"}, "qid": "AutoScholarQuery_train_18736"} +{"question": "Which research work contributed to the generalization of the k-WL test to graphons?", "answer": ["Weisfeiler-Leman Indistinguishability of Graphons"], "answer_arxiv_id": ["2112.09001"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_18737"} +{"question": "Could you provide me some studies that aimed on the non-Bayesian approach to uncertainty estimation?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Evidential Deep Learning to Quantify Classification Uncertainty", "Trusted Multi-View Classification"], "answer_arxiv_id": ["1612.01474", "1806.01768", "2102.02051"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_18738"} +{"question": "What papers proposed score distillation that learns 3D representation directly from pre-trained 2D diffusion models?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation", "Magic3D: High-Resolution Text-to-3D Content Creation", "SparseFusion: Distilling View-conditioned Diffusion for 3D Reconstruction", "NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as General Image Priors"], "answer_arxiv_id": ["2209.14988", "2212.00774v1", "2211.10440", "2212.00792", "2212.03267"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_18739"} +{"question": "What work proposes the benchmark for Scene Graph Generation?", "answer": ["Visual Genome: Connecting Language and Vision Using Crowdsourced Dense\n Image Annotations"], "answer_arxiv_id": ["1602.07332"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_18740"} +{"question": "Is there any research demonstrating improved performance in multilingual tasks by aligning a multilingual language model with another language model?", "answer": ["LLM Augmented LLMs: Expanding Capabilities through Composition"], "answer_arxiv_id": ["2401.02412"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_18741"} +{"question": "Can you name any works that enhanced Flownet3D with the inclusion of point-to-plane distance and angular distance as additional geometric constraints?", "answer": ["FlowNet3D++: Geometric Losses For Deep Scene Flow Estimation"], "answer_arxiv_id": ["1912.01438"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_18742"} +{"question": "Which paper performs conformal prediction in the Federated Learning (FL) setting?", "answer": ["Distribution-Free Federated Learning with Conformal Predictions"], "answer_arxiv_id": ["2110.07661"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_18743"} +{"question": "What works performed feature matching on complete images, enabling end-to-end joint learning?", "answer": ["End-to-End Training of Hybrid CNN-CRF Models for Stereo"], "answer_arxiv_id": ["1611.10229"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_18744"} +{"question": "What research also uses priors to improve DP-SGD image classification?", "answer": ["Differentially Private Learning Needs Better Features (or Much More Data)"], "answer_arxiv_id": ["2011.11660"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_18745"} +{"question": "Which works have introduced directly-trained spiking vision transformers with a pure SNN architecture?", "answer": ["Spikformer: When Spiking Neural Network Meets Transformer", "Spikingformer: Spike-driven Residual Learning for Transformer-based\n Spiking Neural Network", "Spike-driven Transformer"], "answer_arxiv_id": ["2209.15425", "2304.11954", "2307.01694"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_18746"} +{"question": "Can you list some papers that focus on calibration studies in NLP?", "answer": ["On Calibration of Modern Neural Networks", "Calibration of Pre-trained Transformers"], "answer_arxiv_id": ["1706.04599", "2003.07892"], "source_meta": {"published_time": "20240211"}, "qid": "AutoScholarQuery_train_18747"} +{"question": "What studies have attempted the pruning of heads or features in creating efficient transformers?", "answer": ["AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned", "Are Sixteen Heads Really Better than One?"], "answer_arxiv_id": ["2111.15668", "1905.09418", "1905.10650"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_18748"} +{"question": "Which papers proposed efficient training strategies for models without knowledge priors?", "answer": ["Adam: A Method for Stochastic Optimization", "Large Batch Optimization for Deep Learning: Training BERT in 76 minutes", "SuperNeurons: Dynamic GPU Memory Management for Training Deep Neural Networks", "Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism", "Exploring Low Rank Training of Deep Neural Networks", "Taking Notes on the Fly Helps Language Pre-Training", "Accelerating Training of Transformer-Based Language Models with Progressive Layer Dropping", "Knowledge Inheritance for Pre-trained Language Models", "Token Merging: Your ViT But Faster"], "answer_arxiv_id": ["1412.6980v9", "1904.00962", "1801.04380", "1909.08053", "2209.13569v1", "2008.01466", "2010.13369", "2105.13880", "2210.09461"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_18749"} +{"question": "What papers propose usage of similarity encoders?", "answer": ["Encoding high-cardinality string categorical variables"], "answer_arxiv_id": ["1907.01860"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_18750"} +{"question": "Are there any research papers that use coordinate matrices or voxel images of materials as the direct generation targets?", "answer": ["Generative Adversarial Networks for Crystal Structure Prediction", "High-throughput discovery of novel cubic crystal materials using deep generative neural networks", "Data-Driven Approach to Encoding and Decoding 3-D Crystal Structures"], "answer_arxiv_id": ["2004.01396v4", "2102.01880", "1909.00949"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_18751"} +{"question": "Which papers proposed rationale generation to enhance the interpretability of the reasoning problem solving model?", "answer": ["Towards Interpretable Natural Language Understanding with Explanations as Latent Variables", "Teach Me to Explain: A Review of Datasets for Explainable Natural Language Processing", "Measuring Association Between Labels and Free-Text Rationales"], "answer_arxiv_id": ["2011.05268", "2102.12060", "2010.12762"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_18752"} +{"question": "Which work proposed the Pointer Network as a neural solver to directly construct solutions for combinatorial optimization problems?", "answer": ["Pointer Networks"], "answer_arxiv_id": ["1506.03134"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_18753"} +{"question": "What works have studied the applications of diffusion models in generation tasks?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "High-Resolution Image Synthesis with Latent Diffusion Models", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "D2C: Diffusion-Denoising Models for Few-shot Conditional Generation", "Score-Based Generative Modeling with Critically-Damped Langevin\n Diffusion"], "answer_arxiv_id": ["2105.05233", "2112.10752", "2112.10741", "2106.06819", "2112.07068"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18754"} +{"question": "Can you provide references that translated monolingual corpora into pseudo parallel corpora for cross-lingual representation pre-training?", "answer": ["Unsupervised Multilingual Sentence Embeddings for Parallel Corpus Mining", "Ernie-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora"], "answer_arxiv_id": ["2105.10419", "2012.15674"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_18755"} +{"question": "Which works formulated detection as grounding and learns instance-level visual representation with language-aware deep fusion?", "answer": ["Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2112.03857"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18756"} +{"question": "What works proposed resampling for long-tail recognition?", "answer": ["A systematic study of the class imbalance problem in convolutional\n neural networks"], "answer_arxiv_id": ["1710.05381"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_18757"} +{"question": "Which works have implemented the joint embedding space concept for text-driven motion generation?", "answer": ["Language2Pose: Natural Language Grounded Pose Forecasting", "Synthesis of Compositional Animations from Textual Descriptions", "MotionCLIP: Exposing Human Motion Generation to CLIP Space"], "answer_arxiv_id": ["1907.01108", "2103.14675", "2203.08063"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_18758"} +{"question": "Which work is considered as the first proposition of Multi-fidelity multi-armed bandits (MF-MAB)?", "answer": ["The Multi-fidelity Multi-armed Bandit"], "answer_arxiv_id": ["1610.09726"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_18759"} +{"question": "Which works discussed the technique of token pruning in context of reducing model parameters and computational complexity?", "answer": ["Token Merging: Your ViT But Faster", "Shunted Self-Attention via Multi-Scale Token Aggregation", "Making Vision Transformers Efficient from A Token Sparsification View", "Revisiting Token Pruning for Object Detection and Instance Segmentation"], "answer_arxiv_id": ["2210.09461", "2111.15193", "2303.08685", "2306.07050"], "source_meta": {"published_time": "20240410"}, "qid": "AutoScholarQuery_train_18760"} +{"question": "Which studies conducted a theoretical analysis of federated learning in the context of various data corruption models?", "answer": ["Mitigating Bias in Federated Learning"], "answer_arxiv_id": ["2012.02447"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18761"} +{"question": "Which work introduces different types of adapter blocks in a Vision Transformer pre-trained model?", "answer": ["Lossless Adaptation of Pretrained Vision Models For Robotic Manipulation"], "answer_arxiv_id": ["2304.06600"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_18762"} +{"question": "Which papers involved in designing ways to alleviate the over-smoothing problem in GCNs?", "answer": ["Break the Ceiling: Stronger Multi-scale Deep Graph Convolutional Networks"], "answer_arxiv_id": ["1906.02174"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_18763"} +{"question": "Which studies use explicit regularization in joint-embedding approaches?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Emerging Properties in Self-Supervised Vision Transformers", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samples"], "answer_arxiv_id": ["2002.05709", "2104.14294", "2105.04906", "2104.13963"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_18764"} +{"question": "What works have been done on trigger synthesis as a backdoor defense?", "answer": ["AEVA: Black-box Backdoor Detection Using Adversarial Extreme Value Analysis"], "answer_arxiv_id": ["2110.14880"], "source_meta": {"published_time": "20220526"}, "qid": "AutoScholarQuery_train_18765"} +{"question": "Can you mention the research that found a multi-lingual translation system with pixel inputs outperforms its textual counterpart?", "answer": ["Multilingual Pixel Representations for Translation and Effective\n Cross-lingual Transfer"], "answer_arxiv_id": ["2305.14280"], "source_meta": {"published_time": "20240808"}, "qid": "AutoScholarQuery_train_18766"} +{"question": "Which works focus on zero-cost NAS proxies?", "answer": ["How Powerful are Performance Predictors in Neural Architecture Search?", "ProxyBO: Accelerating Neural Architecture Search via Bayesian Optimization with Zero-Cost Proxies", "Neural Architecture Search without Training"], "answer_arxiv_id": ["2104.01177", "2110.10423", "2006.04647"], "source_meta": {"published_time": "20230113"}, "qid": "AutoScholarQuery_train_18767"} +{"question": "Can you provide studies that explored model auditing with individual fairness?", "answer": ["Statistical inference for individual fairness", "Verifying Individual Fairness in Machine Learning Models"], "answer_arxiv_id": ["2103.16714", "2006.11737"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_18768"} +{"question": "What studies constructed tractable cost functions that can be sampled with MCMC in GBI?", "answer": ["Robust Generalised Bayesian Inference for Intractable Likelihoods"], "answer_arxiv_id": ["2104.07359"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_18769"} +{"question": "What researches have been conducted on the topic of HBox-supervised weakly-supervised oriented object detection?", "answer": ["BoxInst: High-Performance Instance Segmentation with Box Annotations", "Box-supervised Instance Segmentation with Level Set Evolution", "H2RBox: Horizontal Box Annotation is All You Need for Oriented Object\n Detection", "Leveraging Orientation for Weakly Supervised Object Detection with\n Application to Firearm Localization", "Knowledge Combination to Learn Rotated Detection Without Rotated\n Annotation"], "answer_arxiv_id": ["2012.02310", "2207.09055", "2210.06742", "1904.10032", "2304.02199"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_18770"} +{"question": "Which papers focus on training models for selective attention to salient objects in a scene?", "answer": ["Recurrent Models of Visual Attention", "Attention for Fine-Grained Categorization"], "answer_arxiv_id": ["1406.6247", "1412.7054"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_18771"} +{"question": "Could you name the studies that used prompt tuning in other visual recognition tasks like object detection, semantic segmentation, and video recognition?", "answer": ["Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model", "DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting", "Expanding Language-Image Pretrained Models for General Video Recognition"], "answer_arxiv_id": ["2203.14940", "2112.01518", "2208.02816"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_18772"} +{"question": "Which studies have replaced the naive model update averaging strategy of FedAvg with more efficient aggregation schemes?", "answer": ["Bayesian Nonparametric Federated Learning of Neural Networks", "Federated learning with matched averaging", "Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization"], "answer_arxiv_id": ["1905.12022", "2002.06440", "2007.07481"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_18773"} +{"question": "Which previous works are related to ensuring worst-case performance to invariance in ML models?", "answer": ["Invariance, Causality and Robustness"], "answer_arxiv_id": ["1812.08233v1"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_18774"} +{"question": "Could you provide me some works about using vision Transformers as the DNN backbone instead of conventional CNNs?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Training data-efficient image transformers & distillation through attention"], "answer_arxiv_id": ["2010.11929", "2012.12877"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_18775"} +{"question": "What work proposed the distillation of a referential SFT policy by polarizing the preference, known as direct preference optimization (DPO)?", "answer": ["Direct Preference Optimization: Your Language Model is Secretly a Reward\n Model"], "answer_arxiv_id": ["2305.18290"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_18776"} +{"question": "Can you provide me the study which attributes the instability of language models to different biases of prompts and proposes a contextual calibration approach?", "answer": ["Calibrate Before Use: Improving Few-Shot Performance of Language Models"], "answer_arxiv_id": ["2102.09690"], "source_meta": {"published_time": "20220520"}, "qid": "AutoScholarQuery_train_18777"} +{"question": "Which studies centered on the application of gradient-based meta-learning in generalizable INRs?", "answer": ["Spatial Functa: Scaling Functa to ImageNet Classification and Generation", "From data to functa: Your data point is a function and you can treat it like one"], "answer_arxiv_id": ["2302.03130", "2201.12204"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_18778"} +{"question": "What studies have expanded the method of predicting functional map in shape matching?", "answer": ["DPFM: Deep Partial Functional Maps", "Deep Geometric Functional Maps: Robust Feature Learning for Shape Correspondence", "SRFeat: Learning Locally Accurate and Globally Consistent Non-Rigid Shape Correspondence"], "answer_arxiv_id": ["2110.09994v1", "2003.14286", "2209.07806"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_18779"} +{"question": "Which studies demonstrate the integration of artificial intelligence via LLMs across various psychological research domains?", "answer": ["PsyQA: A Chinese Dataset for Generating Long Counseling Text for Mental\n Health Support", "Exploring the Frontiers of LLMs in Psychological Applications: A Comprehensive Review", "Towards a Psychological Generalist AI: A Survey of Current Applications\n of Large Language Models and Future Prospects", "Turning large language models into cognitive models", "Can Large Language Models Transform Computational Social Science?"], "answer_arxiv_id": ["2106.01702", "2401.01519v3", "2312.04578", "2306.03917", "2305.03514"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18780"} +{"question": "Which papers studied the performance of ERM models when they rely on spurious attributes such as background in vision?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization", "Noise or Signal: The Role of Image Backgrounds in Object Recognition"], "answer_arxiv_id": ["1911.08731", "2006.09994"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_18781"} +{"question": "What studies explored inpainting and conditional sampling within diffusion models?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion", "Score-Based Generative Modeling through Stochastic Differential Equations", "Palette: Image-to-Image Diffusion Models"], "answer_arxiv_id": ["2104.03670", "2011.13456", "2111.05826"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_18782"} +{"question": "Which works use the concept of reward machines to improve performance in structured reinforcement learning?", "answer": ["Reward Machines: Exploiting Reward Function Structure in Reinforcement Learning"], "answer_arxiv_id": ["2010.03950"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_18783"} +{"question": "Which references are about text-to-3D asset generation methods which adopted neural radiance fields or its variants?", "answer": ["Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance\n Fields", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2103.13415", "2201.05989"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_18784"} +{"question": "Can you name studies that utilize an Ordinary Differential Equation (ODE) termed ‘probability flow’ to establish the diffusion model in continuous time?", "answer": ["Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2011.13456"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_18785"} +{"question": "Which work conditioned the prompt on input images to enhance generalizability?", "answer": ["Conditional Prompt Learning for Vision-Language Models"], "answer_arxiv_id": ["2203.05557"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_18786"} +{"question": "What are some examples of research that utilize multimodal representation learning techniques in robot learning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230919"}, "qid": "AutoScholarQuery_train_18787"} +{"question": "What studies discuss image generation methods based on diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models", "Improved Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2010.02502", "2102.09672", "2105.05233"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_18788"} +{"question": "Which papers talked about using a vision transformer with an image Q-Former to encode individual frames?", "answer": ["Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding"], "answer_arxiv_id": ["2306.02858"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18789"} +{"question": "What studies utilize Vision Transformers as a diverse model to CNNs in CPS-based framework?", "answer": ["Diverse Cotraining Makes Strong Semi-Supervised Segmentor"], "answer_arxiv_id": ["2308.09281"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18790"} +{"question": "Could you list studies that have explored weight quantization for neural networks?", "answer": ["XNOR-Net: ImageNet Classification Using Binary Convolutional Neural\n Networks", "On Quantizing Implicit Neural Representations", "Quantization and Training of Neural Networks for Efficient\n Integer-Arithmetic-Only Inference"], "answer_arxiv_id": ["1603.05279", "2209.01019", "1712.05877"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_18791"} +{"question": "In which research papers the nearest neighbor method was used in contrastive learning framework?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2006.09882", "2104.14548"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_18792"} +{"question": "Which work was used in creating fixed pipelines in the perception model?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_18793"} +{"question": "What research paper proposed augmentation methods based on the frequency domain to prevent spectral bias during the training?", "answer": ["Frequency Domain Model Augmentation for Adversarial Attack"], "answer_arxiv_id": ["2207.05382v1"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_18794"} +{"question": "Are there any works about identifying sparse initializations produced by static approaches that are invariant to parameter reshuffling and reinitialization?", "answer": ["Pruning Neural Networks at Initialization: Why Are We Missing the Mark?"], "answer_arxiv_id": ["2009.08576"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_18795"} +{"question": "Can you mention a work that attempts to optimize hard prompts for CLIP with image-text similarity matching?", "answer": ["Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt\n Tuning and Discovery"], "answer_arxiv_id": ["2302.03668"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_18796"} +{"question": "What is the study that incorporates DDIM inversion features into the text-to-image generation process alongside SD?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2211.12572"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_18797"} +{"question": "What papers discuss the voxel methods in LiDAR based 3D object detection?", "answer": ["Center-based 3D Object Detection and Tracking", "3D-MAN: 3D Multi-frame Attention Network for Object Detection", "PointPillars: Fast Encoders for Object Detection from Point Clouds", "PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection"], "answer_arxiv_id": ["2006.11275", "2103.16054", "1812.05784", "1912.13192"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_18798"} +{"question": "Can you provide some recent advances in privacy accounting that use more statistical information from specific mechanisms to be composed?", "answer": ["Deep Learning with Differential Privacy", "Rényi Differential Privacy", "Computing Tight Differential Privacy Guarantees Using FFT", "Deep Learning with Gaussian Differential Privacy", "Computing Differential Privacy Guarantees for Heterogeneous Compositions Using FFT", "Tight Differential Privacy for Discrete-Valued Mechanisms and for the Subsampled Gaussian Mechanism Using FFT", "Numerical Composition of Differential Privacy", "Optimal Accounting of Differential Privacy via Characteristic Function", "Faster Privacy Accounting via Evolving Discretization", "Connect the Dots: Tighter Discrete Approximations of Privacy Loss Distributions", "Analytical Composition of Differential Privacy via the Edgeworth Accountant"], "answer_arxiv_id": ["1607.00133", "1702.07476", "1906.03049", "1911.11607", "2102.12412", "2006.07134", "2106.02848", "2106.08567", "2207.04381", "2207.04380", "2206.04236"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_18799"} +{"question": "Which works focus on the generalization ability of SSCL on downstream tasks?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning", "Contrastive learning, multi-view redundancy, and linear models", "Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss", "Beyond Separability: Analyzing the Linear Transferability of Contrastive Representations to Related Subpopulations", "Toward Understanding the Feature Learning Process of Self-supervised Contrastive Learning", "Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data", "Towards the Generalization of Contrastive Self-Supervised Learning", "The Power of Contrast for Feature Learning: A Theoretical Analysis", "Deciphering the Projection Head: Representation Evaluation Self-supervised Learning"], "answer_arxiv_id": ["1902.09229v1", "2008.10150", "2106.04156", "2204.02683v2", "2105.15134", "2010.03622", "2111.00743", "2110.02473", "2301.12189"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_18800"} +{"question": "Can you mention papers that incorporated MoE paradigm in visual models or within multi-modal transformers?", "answer": ["Scaling Vision with Sparse Mixture of Experts", "Scaling Vision-Language Models with Sparse Mixture of Experts"], "answer_arxiv_id": ["2106.05974", "2303.07226"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18801"} +{"question": "What studies have been done on the use of deep learning in software design?", "answer": ["Feature Maps: A Comprehensible Software Representation for Design\n Pattern Detection", "GUIGAN: Learning to Generate GUI Designs Using Generative Adversarial\n Networks"], "answer_arxiv_id": ["1812.09873", "2101.09978"], "source_meta": {"published_time": "20230716"}, "qid": "AutoScholarQuery_train_18802"} +{"question": "In what study is a contrastive loss employed to align local features with the global prototypes to reduce the representation gap?", "answer": ["FedProc: Prototypical Contrastive Federated Learning on Non-IID data"], "answer_arxiv_id": ["2109.12273"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_18803"} +{"question": "Which publications propose differentiable data selection?", "answer": ["Optimizing Data Usage via Differentiable Rewards", "Balancing Training for Multilingual Neural Machine Translation"], "answer_arxiv_id": ["1911.10088", "2004.06748"], "source_meta": {"published_time": "20230418"}, "qid": "AutoScholarQuery_train_18804"} +{"question": "Could you provide me some works about the kernel limit of transformers?", "answer": ["Infinite attention: NNGP and NTK for deep attention networks", "Tensor Programs II: Neural Tangent Kernel for Any Architecture"], "answer_arxiv_id": ["2006.10540", "2006.14548v4"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_18805"} +{"question": "Could you provide me a work that considered rule and network-based attention weights for modeling emerging entities?", "answer": ["Logic Attention Based Neighborhood Aggregation for Inductive Knowledge Graph Embedding"], "answer_arxiv_id": ["1811.01399"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_18806"} +{"question": "What is the concurrent work that also augments TPP with CL abilities?", "answer": ["HyperHawkes: Hypernetwork based Neural Temporal Point Process"], "answer_arxiv_id": ["2210.00213"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_18807"} +{"question": "What studies have improved the efficiency of rendering by using voxel hashing or tensor decomposition?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2201.05989", "2203.09517"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18808"} +{"question": "Which papers established Denoising Diffusion Probabilistic Modeling (DDPM)?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20230717"}, "qid": "AutoScholarQuery_train_18809"} +{"question": "Could you provide me the study that introduces a factuality benchmark HaluEval?", "answer": ["HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large Language Models"], "answer_arxiv_id": ["2305.11747"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_18810"} +{"question": "Which work extended the EPL by allowing for the use of the norm notation 𝑿t−psuperscriptsubscript?", "answer": ["The Elliptical Potential Lemma Revisited"], "answer_arxiv_id": ["2010.10182"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_18811"} +{"question": "Which studies proposed the concept of neural module networks (NMN) to decompose complex reasoning tasks into subtasks?", "answer": ["Learning to Compose Neural Networks for Question Answering", "Neural Module Networks"], "answer_arxiv_id": ["1601.01705", "1511.02799v4"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_18812"} +{"question": "Which work proposes to perform depth position encoding to inject global depth information into Transformer in object detection?", "answer": ["MonoDTR: Monocular 3D Object Detection with Depth-Aware Transformer"], "answer_arxiv_id": ["2203.10981"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_18813"} +{"question": "What papers propose methods for 2D-3D joint depth completion using 3D representations such as surface normals, graphs, point clouds, and voxels?", "answer": ["Depth Completion from Sparse LiDAR Data with Depth-Normal Constraints", "DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene\n from Sparse LiDAR Data and Single Color Image", "Adaptive Context-Aware Multi-Modal Network for Depth Completion", "GraphCSPN: Geometry-Aware Depth Completion via Dynamic GCNs", "Learning Joint 2D-3D Representations for Depth Completion", "Boosting Monocular Depth Estimation with Lightweight 3D Point Fusion"], "answer_arxiv_id": ["1910.06727", "1812.00488", "2008.10833", "2210.10758", "2012.12402", "2012.10296"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_18814"} +{"question": "Are there any studies that proposed extending adversarial training by using additional unlabeled data?", "answer": ["Unlabeled Data Improves Adversarial Robustness", "Improving Robustness using Generated Data"], "answer_arxiv_id": ["1905.13736", "2110.09468"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_18815"} +{"question": "What literature proposes a restarting procedure in online convex optimization?", "answer": ["Non-stationary Stochastic Optimization"], "answer_arxiv_id": ["1307.5449"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_18816"} +{"question": "Which papers define the concept of network motifs?", "answer": ["Higher-order organization of complex networks", "Motifs in Temporal Networks"], "answer_arxiv_id": ["1612.08447", "1612.09259"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18817"} +{"question": "Which works proposed the contextual explanation networks?", "answer": ["Contextual Explanation Networks"], "answer_arxiv_id": ["1705.10301"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_18818"} +{"question": "Are there any works about the design of neural decoders for existing linear codes?", "answer": ["Learning to Decode Linear Codes Using Deep Learning", "An Introduction to Deep Learning for the Physical Layer", "Neural Offset Min-Sum Decoding", "An Iterative BP-CNN Architecture for Channel Decoding", "Deep Learning for Decoding of Linear Codes - A Syndrome-Based Approach", "DeepTurbo : Deep Turbo Decoder", "Hyper-Graph-Network Decoders for Block Codes", "Pruning Neural Belief Propagation Decoders", "Model-Driven DNN Decoder for Turbo Codes: Design, Simulation and Experimental Results"], "answer_arxiv_id": ["1607.04793", "1702.00832", "1701.05931", "1707.05697", "1802.04741", "1903.02295", "1909.09036", "2001.07464v2", "2006.08896"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_18819"} +{"question": "Are there any studies that modeled the adversarial interaction as a zero sum game?", "answer": ["Robust Adversarial Reinforcement Learning", "Adversarial Policies: Attacking Deep Reinforcement Learning"], "answer_arxiv_id": ["1703.02702", "1905.10615"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_18820"} +{"question": "What studies are there about creating separate global descriptors for the ground image and the satellite image in cross-view retrieval tasks?", "answer": ["Coming Down to Earth: Satellite-to-Street View Synthesis for Geo-Localization", "Localizing and Orienting Street Views Using Overhead Imagery"], "answer_arxiv_id": ["2103.06818", "1608.00161v2"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_18821"} +{"question": "What papers cover the use of consistency regularization in Semi-Supervised Learning?", "answer": ["Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results", "Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning", "Temporal Ensembling for Semi-Supervised Learning"], "answer_arxiv_id": ["1703.01780", "1606.04586", "1610.02242"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_18822"} +{"question": "Could you provide me with the references about source-free domain adaptation methods?", "answer": ["Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation", "Source-free Domain Adaptation via Distributional Alignment by Matching Batch Normalization Statistics", "Source-Free Adaptation to Measurement Shift via Bottom-Up Feature Restoration"], "answer_arxiv_id": ["2002.08546", "2101.10842", "2107.05446v3"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_18823"} +{"question": "Any works that improve performance by replacing clustering with candidate localization?", "answer": ["3D Instances as 1D Kernels"], "answer_arxiv_id": ["2207.07372"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_18824"} +{"question": "What research papers analyze the generalization of neural networks that are trained with Stochastic Gradient Descent (SGD) or its noisy variant Stochastic Gradient Langevin Descent (SGLD)?", "answer": ["Information-theoretic analysis of generalization capability of learning algorithms", "How much does your data exploration overfit? Controlling bias via information usage.", "Reasoning About Generalization via Conditional Mutual Information", "Non-Convex Learning via Stochastic Gradient Langevin Dynamics: A Nonasymptotic Analysis", "Sharpened Generalization Bounds based on Conditional Mutual Information and an Application to Noisy, Iterative Algorithms", "Information-Theoretic Generalization Bounds for Stochastic Gradient Descent"], "answer_arxiv_id": ["1705.07809", "1511.05219", "2001.09122", "1702.03849", "2004.12983v2", "2102.00931"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_18825"} +{"question": "What works have studied how to apply neural rendering techniques to automotive data?", "answer": ["Neural Scene Graphs for Dynamic Scenes", "Panoptic Neural Fields: A Semantic Object-Aware Neural Scene\n Representation", "Panoptic NeRF: 3D-to-2D Label Transfer for Panoptic Urban Scene\n Segmentation", "S-NeRF: Neural Radiance Fields for Street Views", "Block-NeRF: Scalable Large Scene Neural View Synthesis", "SUDS: Scalable Urban Dynamic Scenes"], "answer_arxiv_id": ["2011.10379", "2205.04334", "2203.15224", "2303.00749", "2202.05263", "2303.14536"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_18826"} +{"question": "Which papers proposed methods for single sound source localization?", "answer": ["Learning to Localize Sound Source in Visual Scenes", "Localizing Visual Sounds the Hard Way", "Learning Sound Localization Better From Semantically Similar Samples", "Unsupervised Sound Localization via Iterative Contrastive Learning", "Audio-Visual Spatial Integration and Recursive Attention for Robust\n Sound Source Localization", "Learning Audio-Visual Source Localization via False Negative Aware\n Contrastive Learning", "Hear The Flow: Optical Flow-Based Self-Supervised Visual Sound Source\n Localization", "Exploring Simple Siamese Representation Learning", "A Closer Look at Weakly-Supervised Audio-Visual Source Localization", "FlowGrad: Using Motion for Visual Sound Source Localization", "Exploiting Transformation Invariance and Equivariance for\n Self-supervised Sound Localisation", "Sound Source Localization is All about Cross-Modal Alignment", "Self-Supervised Predictive Learning: A Negative-Free Method for Sound Source Localization in Visual Scenes"], "answer_arxiv_id": ["1803.03849", "2104.02691", "2202.03007", "2104.00315", "2308.06087", "2303.11302", "2211.03019", "2011.10566", "2209.09634", "2211.08367", "2206.12772", "2309.10724", "2203.13412v1"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_18827"} +{"question": "Which works focused on formulating comprehensive conditional generation frameworks?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2112.10752", "2302.05543"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_18828"} +{"question": "Which works used Randomized Smoothing as a defense for offering provable robustness guarantees in machine learning?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing", "Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers"], "answer_arxiv_id": ["1902.02918", "1906.04584"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_18829"} +{"question": "Which papers describe the semi-template-based methods for retrosynthesis prediction?", "answer": ["A Graph to Graphs Framework for Retrosynthesis Prediction", "RetroXpert: Decompose Retrosynthesis Prediction Like A Chemist", "Learning Graph Models for Retrosynthesis Prediction"], "answer_arxiv_id": ["2003.12725", "2011.02893", "2006.07038"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_18830"} +{"question": "Which work designed a context-tree switching density model to perform state pseudo-count?", "answer": ["Unifying Count-Based Exploration and Intrinsic Motivation"], "answer_arxiv_id": ["1606.01868"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_18831"} +{"question": "Who introduced the idea of running Model-X Knockoffs multiple times and computing for each the proportion of runs for which it was selected?", "answer": ["Derandomizing Knockoffs"], "answer_arxiv_id": ["2012.02717"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_18832"} +{"question": "What alternative algorithms could be employed to increase retrieval efficiency in Open-Domain Question Answering systems?", "answer": ["Billion-scale similarity search with GPUs", "Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs"], "answer_arxiv_id": ["1702.08734", "1603.09320v4"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_18833"} +{"question": "Who introduced the setting that aims to maximize the XP returns of independently trained agents using the same algorithm which is called intra-algorithm cross play (intra-AXP)?", "answer": ["“Other-Play” for Zero-Shot Coordination"], "answer_arxiv_id": ["2003.02979"], "source_meta": {"published_time": "20220129"}, "qid": "AutoScholarQuery_train_18834"} +{"question": "What paper addresses the challenge of representing and rendering detailed 3D meshes of large-scale scenes efficiently?", "answer": ["Neural 3D Mesh Renderer"], "answer_arxiv_id": ["1711.07566"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_18835"} +{"question": "Any works studied learning a single neuron in agnostic and noisy settings?", "answer": ["Agnostic Learning of a Single Neuron with Gradient Descent"], "answer_arxiv_id": ["2005.14426"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_18836"} +{"question": "Could you provide some works that applied neural-based volumetric representations to head and hair reconstruction?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Neural Volumes: Learning Dynamic Renderable Volumes from Images", "Mixture of Volumetric Primitives for Efficient Neural Rendering"], "answer_arxiv_id": ["2003.08934", "1906.07751", "2103.01954"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_18837"} +{"question": "Could you point out the works that have investigated fair and local objectives?", "answer": ["Correlation Clustering and Biclustering with Locally Bounded Errors", "Fair Clustering Through Fairlets", "Local Guarantees in Graph Cuts and Clustering", "Min-Max Correlation Clustering via MultiCut", "Fair Correlation Clustering", "Local Correlation Clustering with Asymmetric Classification Errors", "Improved Approximation for Fair Correlation Clustering", "Fast Combinatorial Algorithms for Min Max Correlation Clustering"], "answer_arxiv_id": ["1506.08189", "1802.05733", "1704.00355", "1907.00117v1", "2002.03508", "2108.05697v1", "2206.05050", "2301.13079"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_18838"} +{"question": "Could you provide me works about methods to address the limitations of optimization-based methods by introducing encoder-based methods?", "answer": ["Prompt-Free Diffusion: Taking \"Text\" out of Text-to-Image Diffusion\n Models", "Enhancing Detail Preservation for Customized Text-to-Image Generation: A\n Regularization-Free Approach", "Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models"], "answer_arxiv_id": ["2305.16223", "2305.13579", "2307.11410", "2308.06721"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_18839"} +{"question": "What works involve the use of secondary neural models in translation quality estimation?", "answer": ["Improving Back-Translation with Uncertainty-based Confidence Estimation", "Uncertainty-Aware Machine Translation Evaluation"], "answer_arxiv_id": ["1909.00157", "2109.06352"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_18840"} +{"question": "Which studies aim to train an object detector using image-level labels without any bounding box annotation?", "answer": ["Weakly Supervised Deep Detection Networks", "Weakly Supervised Object Localization with Multi-fold Multiple Instance Learning", "Multiple Instance Detection Network with Online Instance Classifier Refinement", "Cap2Det: Learning to Amplify Weak Caption Supervision for Object Detection", "Omni-DETR: Omni-Supervised Object Detection with Transformers"], "answer_arxiv_id": ["1511.02853", "1503.00949", "1704.00138", "1907.10164", "2203.16089"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_18841"} +{"question": "What works describe the extraction of coarse object masks in the context of unsupervised instance segmentation?", "answer": ["FreeSOLO: Learning to Segment Objects without Annotations"], "answer_arxiv_id": ["2202.12181"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_18842"} +{"question": "Could you tell me which research showed that the ELBO is invariant to the choice of noise schedule?", "answer": ["Variational Diffusion Models"], "answer_arxiv_id": ["2107.00630"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_18843"} +{"question": "What are some primary works on transformer-based Protein Language Models (PLMs)?", "answer": ["ProGen2: Exploring the Boundaries of Protein Language Models", "xTrimoPGLM: Unified 100B-Scale Pre-trained Transformer for Deciphering\n the Language of Protein"], "answer_arxiv_id": ["2206.13517", "2401.06199"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_18844"} +{"question": "Are there any studies showing that a deep BNN is equivalent to a GP with a compositional kernel?", "answer": ["Deep Neural Networks as Gaussian Processes"], "answer_arxiv_id": ["1711.00165"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_18845"} +{"question": "Which works discuss the fast development of novel behaviors in neural networks as they scale up or trained longer?", "answer": ["Predictability and Surprise in Large Generative Models", "Emergent Abilities of Large Language Models"], "answer_arxiv_id": ["2202.07785", "2206.07682"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_18846"} +{"question": "In what papers have matrix factorization been used to learn propensity scores?", "answer": ["Modeling User Exposure in Recommendation", "CausCF: Causal Collaborative Filtering for Recommendation Effect Estimation"], "answer_arxiv_id": ["1510.07025", "2105.13881"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_18847"} +{"question": "What research papers discuss LLMs’ potential in using a myriad of tools to augment their capacity?", "answer": ["Augmented Language Models: a Survey", "Tool Learning with Foundation Models"], "answer_arxiv_id": ["2302.07842", "2304.08354"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_18848"} +{"question": "Which research papers have implemented algorithms that respect the symplectic structure or conservation laws of the Hamiltonian for Hamiltonian systems?", "answer": ["Simplifying Hamiltonian and Lagrangian Neural Networks via Explicit Constraints"], "answer_arxiv_id": ["2010.13581"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_18849"} +{"question": "Can you give examples of works on WSI analysis where instance-based methods were used in Multiple Instance Learning'", "answer": ["Weakly supervised multiple instance learning histopathological tumor\n segmentation", "CAMEL: A Weakly Supervised Learning Framework for Histopathology Image\n Segmentation"], "answer_arxiv_id": ["2004.05024", "1908.10555"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18850"} +{"question": "Are there any references about studies on how individuals may take strategic actions to improve their outcomes given a classifier?", "answer": ["Linear Classifiers that Encourage Constructive Adaptation"], "answer_arxiv_id": ["2011.00355"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_18851"} +{"question": "What studies have given nearly minimax optimal bounds?", "answer": ["REGAL: A Regularization based Algorithm for Reinforcement Learning in Weakly Communicating MDPs", "(More) Efficient Reinforcement Learning via Posterior Sampling", "Why is Posterior Sampling Better than Optimism for Reinforcement Learning?", "Near Optimal Exploration-Exploitation in Non-Communicating Markov Decision Processes", "Variance-Aware Regret Bounds for Undiscounted Reinforcement Learning in MDPs", "Non-Asymptotic Gap-Dependent Regret Bounds for Tabular MDPs", "Worst-Case Regret Bounds for Exploration via Randomized Value Functions", "Regret Minimization for Reinforcement Learning by Evaluating the Optimal Bias Function", "A Unifying View of Optimism in Episodic Reinforcement Learning"], "answer_arxiv_id": ["1205.2661v1", "1306.0940", "1607.00215", "1807.02373", "1803.01626", "1905.03814", "1906.02870", "1906.05110", "2007.01891"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_18852"} +{"question": "What works use a left-to-right autoregressive generation approach in the development of Large Language Models?", "answer": ["Language Models are Few-Shot Learners", "Training language models to follow instructions with human feedback", "GPT-4 Technical Report"], "answer_arxiv_id": ["2005.14165", "2203.02155", "2303.08774"], "source_meta": {"published_time": "20240527"}, "qid": "AutoScholarQuery_train_18853"} +{"question": "In what research is Stochastic Gradient Descent discussed to provide highly accurate results?", "answer": ["Practical Recommendations for Gradient-Based Training of Deep Architectures"], "answer_arxiv_id": ["1206.5533"], "source_meta": {"published_time": "20221114"}, "qid": "AutoScholarQuery_train_18854"} +{"question": "What works proposed the specialized solvers for efficient and accurate sampling in Probabilistic-flow Ordinal Differential Equation (PF-ODE)?", "answer": ["DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling\n in Around 10 Steps", "UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models", "DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic\n Models", "Denoising Diffusion Implicit Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Fast Sampling of Diffusion Models with Exponential Integrator", "Gotta Go Fast When Generating Data with Score-Based Models"], "answer_arxiv_id": ["2206.00927", "2302.04867v4", "2211.01095", "2010.02502", "2206.00364v2", "2204.13902", "2105.14080"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_18855"} +{"question": "Which works showcased the reasoning abilities of large language models (LLM)?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Visual Instruction Tuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation"], "answer_arxiv_id": ["2103.00020", "2304.08485", "2304.10592", "2201.12086"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_18856"} +{"question": "What papers apply training-free NAS methods to automate designing novel network architectures?", "answer": ["Neural Architecture Search on ImageNet in Four GPU Hours: A\n Theoretically Inspired Perspective", "Neural Architecture Search without Training", "Zen-NAS: A Zero-Shot NAS for High-Performance Deep Image Recognition", "Zero-Cost Proxies for Lightweight NAS", "Neural Architecture Search on ImageNet in Four GPU Hours: A\n Theoretically Inspired Perspective", "GradSign: Model Performance Inference with Theoretical Insights", "ZiCo: Zero-shot NAS via Inverse Coefficient of Variation on Gradients"], "answer_arxiv_id": ["2102.11535", "2006.04647", "2102.01063", "2101.08134", "2102.11535", "2110.08616", "2301.11300"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_18857"} +{"question": "Which works initiated the work on Visual Question Answering (VQA)?", "answer": ["VQA: Visual Question Answering", "Visual7W: Grounded Question Answering in Images"], "answer_arxiv_id": ["1505.00468", "1511.03416"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_18858"} +{"question": "Which paper developed an interpretable temporal convolutional network for action recognition?", "answer": ["Interpretable 3D Human Action Analysis with Temporal Convolutional\n Networks"], "answer_arxiv_id": ["1704.04516"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_18859"} +{"question": "Which paper implements state abstraction in reinforcement learning by encoding the observations directly as discrete state variables?", "answer": ["Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["2010.02193"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_18860"} +{"question": "Any studies that underscores the connection of SDE-based analysis and Linear Scaling Rule for SGD?", "answer": ["Three Factors Influencing Minima in SGD"], "answer_arxiv_id": ["1711.04623"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_18861"} +{"question": "Could you list the studies that pre-train the point cloud backbone by generating the 2D projections of the point cloud?", "answer": ["Take-A-Photo: 3D-to-2D Generative Pre-training of Point Cloud Models", "Ponder: Point Cloud Pre-training via Neural Rendering"], "answer_arxiv_id": ["2307.14971", "2301.00157"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_18862"} +{"question": "Any notable work dealing with post-training sparsity using zero-order criteria?", "answer": ["Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding"], "answer_arxiv_id": ["1510.00149"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_18863"} +{"question": "Which works propose policies to directly cancel out the matched uncertainty in adaptive control?", "answer": ["Neural-Fly Enables Rapid Learning for Agile Flight in Strong Winds", "Regret Bounds for Adaptive Nonlinear Control"], "answer_arxiv_id": ["2205.06908", "2011.13101"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_18864"} +{"question": "Which works are involved in using replay methods for continual learning where replaying includes storing a small portion of data from previous tasks in episodic memory?", "answer": ["REMIND Your Neural Network to Prevent Catastrophic Forgetting", "Task-Free Continual Learning", "Online Continual Learning on Class Incremental Blurry Task Configuration\n with Anytime Inference", "Rainbow Memory: Continual Learning with a Memory of Diverse Samples", "Online Coreset Selection for Rehearsal-based Continual Learning"], "answer_arxiv_id": ["1910.02509", "1812.03596", "2110.10031", "2103.17230", "2106.01085"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_18865"} +{"question": "What research introduced the local gradient tracking to reduce local inconsistency in the local SGD method?", "answer": ["Variance Reduced Local SGD with Lower Communication Complexity"], "answer_arxiv_id": ["1912.12844"], "source_meta": {"published_time": "20230221"}, "qid": "AutoScholarQuery_train_18866"} +{"question": "What research papers focus on the transition from rule-based systems and expert knowledge to statistical models and deep neural networks in coreference resolution?", "answer": ["Learning Global Features for Coreference Resolution", "Deep Reinforcement Learning for Mention-Ranking Coreference Models", "Improving Coreference Resolution by Learning Entity-Level Distributed Representations", "End-to-end Neural Coreference Resolution"], "answer_arxiv_id": ["1604.03035", "1609.08667v3", "1606.01323", "1707.07045"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_18867"} +{"question": "Which work first used 3D CNN for structure-based drug design to encode the protein-ligand structures and generate ligands by atom fitting and bond inference from the predicted atom densities?", "answer": ["Generating 3D Molecules Conditional on Receptor Binding Sites with Deep Generative Models"], "answer_arxiv_id": ["2110.15200v2"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_18868"} +{"question": "What works proposed instance-aware, modular, and realistic simulators for monocular dynamic scenes?", "answer": ["MARS: An Instance-aware, Modular and Realistic Simulator for Autonomous\n Driving", "UniSim: A Neural Closed-Loop Sensor Simulator"], "answer_arxiv_id": ["2307.15058", "2308.01898"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_18869"} +{"question": "Which papers discuss exemplar-based image translation tasks in instances of exemplar-guided image editing?", "answer": ["Example-Guided Style-Consistent Image Synthesis from Semantic Labeling", "Multimodal Unsupervised Image-to-Image Translation", "CoCosNet v2: Full-Resolution Correspondence Learning for Image Translation"], "answer_arxiv_id": ["1906.01314", "1804.04732", "2012.02047"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18870"} +{"question": "Which research papers study the Masked Autoencoders to self-supervised representation learning from audio spectrograms?", "answer": ["Masked Autoencoders that Listen"], "answer_arxiv_id": ["2207.06405"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_18871"} +{"question": "Which works mention the use of a pre-trained text encoder to extract text embeddings for training diffusion models?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2205.11487", "2206.10789"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_18872"} +{"question": "What works have been conducted on algorithm selection and parameters configurations?", "answer": ["How much data is sufficient to learn high-performing algorithms? Generalization guarantees for data-driven algorithm design", "Generalization in portfolio-based algorithm selection", "ParamILS: An Automatic Algorithm Configuration Framework"], "answer_arxiv_id": ["1908.02894", "2012.13315", "1401.3492v1"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_18873"} +{"question": "Which works attempted to estimate a reflectance map or an appearance map?", "answer": ["Deep Reflectance Maps", "Deep Appearance Maps"], "answer_arxiv_id": ["1511.04384", "1804.00863"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_18874"} +{"question": "What work is dedicated to Out-Of-Distribution prediction in the field of drug discovery?", "answer": ["DrugOOD: Out-of-Distribution (OOD) Dataset Curator and Benchmark for AI-aided Drug Discovery – A Focus on Affinity Prediction Problems with Noise Annotations"], "answer_arxiv_id": ["2201.09637"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_18875"} +{"question": "Which papers studied the mode connectivity property in deep neural networks’ loss surfaces?", "answer": ["Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Essentially No Barriers in Neural Network Energy Landscape"], "answer_arxiv_id": ["1802.10026", "1803.00885"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_18876"} +{"question": "Which researches expanded the capabilities of text-to-image models with the help of 'plugins'?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "GLIGEN: Open-Set Grounded Text-to-Image Generation", "LoRA: Low-Rank Adaptation of Large Language Models", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Inserting Anybody in Diffusion Models via Celeb Basis", "IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image\n Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning"], "answer_arxiv_id": ["2112.10752", "2301.07093", "2106.09685", "2208.12242", "2208.01618", "2306.00926", "2308.06721", "2302.05543", "2302.08453", "2307.04725"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18877"} +{"question": "Could you provide me some studies about the adaptation of these models to other domains?", "answer": ["Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models", "LLaVA-Med: Training a Large Language-and-Vision Assistant for\n Biomedicine in One Day", "XrayGPT: Chest Radiographs Summarization using Medical Vision-Language\n Models", "RSGPT: A Remote Sensing Vision Language Model and Benchmark"], "answer_arxiv_id": ["2306.05424", "2306.00890", "2306.07971", "2307.15266"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_18878"} +{"question": "Can you list the papers that used Transformers and supervised learning approaches to train on purely offline data for learning a generalist policy?", "answer": ["Attention Is All You Need", "A Generalist Agent", "Multi-Game Decision Transformers"], "answer_arxiv_id": ["1706.03762", "2205.06175", "2205.15241"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_18879"} +{"question": "Which research papers describe methods that improve upon vanilla zero-shot prompting?", "answer": ["Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing", "Calibrate Before Use: Improving Few-Shot Performance of Language Models", "Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "MetaICL: Learning to Learn In Context"], "answer_arxiv_id": ["2107.13586v1", "2102.09690", "2104.08786", "2201.11903", "2110.15943"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_18880"} +{"question": "Any research work on scene text spotting which use separate detection-recognition heads according to different annotation formats?", "answer": ["SwinTextSpotter: Scene Text Spotting via Better Synergy between Text\n Detection and Text Recognition", "ESTextSpotter: Towards Better Scene Text Spotting with Explicit Synergy\n in Transformer"], "answer_arxiv_id": ["2203.10209", "2308.10147"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_18881"} +{"question": "What are the most recent promising approaches in disentanglement learning?", "answer": ["Orthogonal Jacobian Regularization for Unsupervised Disentanglement in Image Generation"], "answer_arxiv_id": ["2108.07668"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_18882"} +{"question": "In what studies was the use of thermal cameras explored?", "answer": ["MFGNet: Dynamic Modality-Aware Filter Generation for RGB-T Tracking", "Learning Deep Multi-Level Similarity for Thermal Infrared Object\n Tracking", "Multi-Adapter RGBT Tracking"], "answer_arxiv_id": ["2107.10433", "1906.03568", "1907.07485"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_18883"} +{"question": "What studies proposed the Task-Informed method for motion prediction?", "answer": ["TIP: Task-Informed Motion Prediction for Intelligent Vehicles"], "answer_arxiv_id": ["2110.08750"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_18884"} +{"question": "Any prior work that used the concept of composing the gradients of multiple DMs for controllable image generation?", "answer": ["Compositional Visual Generation with Composable Diffusion Models", "Reduce, Reuse, Recycle: Compositional Generation with Energy-Based\n Diffusion Models and MCMC"], "answer_arxiv_id": ["2206.01714", "2302.11552"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18885"} +{"question": "What are some notable papers on multi-modal learning with multi-modal transformer?", "answer": ["Multimodal Machine Learning: A Survey and Taxonomy", "Multimodal Learning with Transformers: A Survey", "Multimodal Transformer for Unaligned Multimodal Language Sequences"], "answer_arxiv_id": ["1705.09406", "2206.06488", "1906.00295"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_18886"} +{"question": "Are there any studies that focus not on optimizing downstream performances during feature selection?", "answer": ["DeepPINK: reproducible feature selection in deep neural networks"], "answer_arxiv_id": ["1809.01185"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_18887"} +{"question": "What research works reduced training time by leveraging more computational resources through distributed training systems?", "answer": ["ZeRO: Memory Optimizations Toward Training Trillion Parameter Models", "GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism"], "answer_arxiv_id": ["1910.02054", "1811.06965v5"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_18888"} +{"question": "Which papers originally formulated GFlowNets as a reinforcement learning algorithm?", "answer": ["Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation", "GFlowNet Foundations"], "answer_arxiv_id": ["2106.04399", "2111.09266"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_18889"} +{"question": "Which studies explore the use of multiple normalization module to improve model generalization in adversarial training and insufficient data scenarios?", "answer": ["Adversarial Examples Improve Image Recognition", "Pure Noise to the Rescue of Insufficient Data: Improving Imbalanced Classification by Training on Random Noise Images"], "answer_arxiv_id": ["1911.09665", "2112.08810"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_18890"} +{"question": "Could you provide me some studies employing reinforcement learning or synthetic/auxiliary data to address data scarcity in learning-based theorem proving?", "answer": ["Learning to Reason in Large Theories without Imitation", "TacticZero: Learning to Prove Theorems from Scratch with Deep Reinforcement Learning", "HyperTree Proof Search for Neural Theorem Proving", "Formal Mathematics Statement Curriculum Learning", "Learning to Prove Theorems by Learning to Generate Theorems", "Mathematical Reasoning via Self-supervised Skip-tree Training", "LIME: Learning Inductive Bias for Primitives of Mathematical Reasoning", "Proof Artifact Co-training for Theorem Proving with Language Models"], "answer_arxiv_id": ["1905.10501", "2102.09756v2", "2205.11491", "2202.01344", "2002.07019", "2006.04757", "2101.06223", "2102.06203"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_18891"} +{"question": "In the context of transfer learning, what papers focus on prompting and in-context learning?", "answer": ["Language Models are Few-Shot Learners", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2005.14165", "2107.13586v1"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_18892"} +{"question": "Which studies were conducted where viewport is uniformly extracted over the sphere?", "answer": ["No-Reference Quality Assessment for 360-degree Images by Analysis of Multi-frequency Information and Local-global Naturalness", "Perceptual Quality Assessment of Omnidirectional Images"], "answer_arxiv_id": ["2102.11393", "2207.02674"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18893"} +{"question": "Could you provide me some studies about the application of diffusion models in text-to-image and text-to-video generation tasks?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Imagen Video: High Definition Video Generation with Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2205.11487", "2210.02303"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_18894"} +{"question": "Which studies focus on medical VLP methods?", "answer": ["Learning to Exploit Temporal Structure for Biomedical Vision-Language\n Processing", "Making the Most of Text Semantics to Improve Biomedical Vision--Language\n Processing", "MedCLIP: Contrastive Learning from Unpaired Medical Images and Text", "Contrastive Learning of Medical Visual Representations from Paired\n Images and Text", "Multi-Granularity Cross-modal Alignment for Generalized Medical Visual\n Representation Learning"], "answer_arxiv_id": ["2301.04558", "2204.09817", "2210.10163", "2010.00747", "2210.06044"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_18895"} +{"question": "Which work discusses local elasticity as a possible mechanism behind the emergence of Negative Curvature (NC)?", "answer": ["Neural Collapse: A Review on Modelling Principles and Generalization"], "answer_arxiv_id": ["2206.04041"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18896"} +{"question": "What work obtained an offline 6.5-approximation for non-monotone suodular maximization with cardinality constraint?", "answer": ["Constrained Non-Monotone Submodular Maximization: Offline and Secretary Algorithms"], "answer_arxiv_id": ["1003.1517"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_18897"} +{"question": "What work decomposes the eigenfunctions of the Laplace–Beltrami operator of a manifold to find a product structure?", "answer": ["Product Manifold Learning"], "answer_arxiv_id": ["2010.09908v1"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_18898"} +{"question": "Any studies that have considered adversarial delays in reinforcement learning?", "answer": ["Learning Adversarial Markov Decision Processes with Delayed Feedback", "Near-Optimal Regret for Adversarial MDP with Delayed Bandit Feedback"], "answer_arxiv_id": ["2012.14843", "2201.13172"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_18899"} +{"question": "Which papers focused on whole-body 3D pose estimation by leveraging the synchronized egocentric camera and external cameras?", "answer": ["Estimating Egocentric 3D Human Pose in the Wild with External Weak\n Supervision"], "answer_arxiv_id": ["2201.07929"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_18900"} +{"question": "Could you provide me some works that proposed diffusion-based T2I models for generating photorealistic images?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "More Control for Free! Image Synthesis with Semantic Diffusion Guidance", "Vector Quantized Diffusion Model for Text-to-Image Synthesis"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2205.11487", "2112.10752", "2112.05744", "2111.14822"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_18901"} +{"question": "Which work validated the SDE model experimentally?", "answer": ["On the Validity of Modeling SGD with Stochastic Differential Equations (SDEs)"], "answer_arxiv_id": ["2102.12470"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_18902"} +{"question": "Which researches propose methods for integrating labels with features in graph learning tasks?", "answer": ["Masked Label Prediction: Unified Message Passing Model for Semi-Supervised Classification"], "answer_arxiv_id": ["2009.03509"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_18903"} +{"question": "What studies discuss methods for solving the update and backward locking problems with an eye towards parallelization?", "answer": ["Decoupled Neural Interfaces using Synthetic Gradients", "Understanding Synthetic Gradients and Decoupled Neural Interfaces", "Decoupled Parallel Backpropagation with Convergence Guarantee", "Training Neural Networks Using Features Replay"], "answer_arxiv_id": ["1608.05343", "1703.00522", "1804.10574", "1807.04511"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_18904"} +{"question": "What works extended masked autoencoder from image to video?", "answer": ["VideoMAE: Masked Autoencoders are Data-Efficient Learners for\n Self-Supervised Video Pre-Training", "Masked Autoencoders As Spatiotemporal Learners", "VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking", "MGMAE: Motion Guided Masking for Video Masked Autoencoding"], "answer_arxiv_id": ["2203.12602", "2205.09113", "2303.16727v2", "2308.10794"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_18905"} +{"question": "What are some papers that explore multilingual instruction tuning?", "answer": ["Phoenix: Democratizing ChatGPT across Languages", "PolyLM: An Open Source Polyglot Large Language Model"], "answer_arxiv_id": ["2304.10453", "2307.06018"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_18906"} +{"question": "What studies claim that GNN models are overly reliant on homophilic patterns and unsuited to capturing heterophilic patterns?", "answer": ["Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs", "Adaptive Universal Generalized PageRank Graph Neural Network", "Finding Global Homophily in Graph Neural Networks When Meeting Heterophily", "Grale: Designing Networks for Graph Learning"], "answer_arxiv_id": ["2006.11468v2", "2006.07988", "2205.07308", "2007.12002"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_18907"} +{"question": "What studies focus on achieving fast test-time customization in the field of concept customization?", "answer": ["InstantBooth: Personalized Text-to-Image Generation without Test-Time Finetuning", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized Text-to-Image Generation", "Taming Encoder for Zero Fine-tuning Image Customization with Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2304.03411", "2302.13848", "2304.02642"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_18908"} +{"question": "What papers attempted to enhance model compatibility by learning maps between old and new model embedding spaces?", "answer": ["Memory-Efficient Incremental Learning Through Feature Adaptation", "Unified Representation Learning for Cross Model Compatibility", "Privacy-Preserving Model Upgrades with Bidirectional Compatible Training in Image Retrieval"], "answer_arxiv_id": ["2004.00713v2", "2008.04821", "2204.13919"], "source_meta": {"published_time": "20230308"}, "qid": "AutoScholarQuery_train_18909"} +{"question": "Can you provide the works about single-stage detectors that increase the speed and simplicity of the framework compared to the two-stage detector?", "answer": ["You Only Look Once: Unified, Real-Time Object Detection", "YOLOv3: An Incremental Improvement", "YOLOv4: Optimal Speed and Accuracy of Object Detection", "YOLO9000: Better, Faster, Stronger", "YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors", "SSD: Single Shot MultiBox Detector", "Deep High-Resolution Representation Learning for Visual Recognition", "Focal Loss for Dense Object Detection"], "answer_arxiv_id": ["1506.02640", "1804.02767", "2004.10934", "1612.08242", "2207.02696", "1512.02325", "1908.07919", "1708.02002"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_18910"} +{"question": "Which studies maximizes mutual information between node representations and corresponding high-level summaries of graphs?", "answer": ["Deep Graph Infomax"], "answer_arxiv_id": ["1809.10341"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_18911"} +{"question": "Could you give me examples of research that train the recurrent state space model (RSSM) using a growing experience dataset?", "answer": ["Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1811.04551", "1912.01603", "2010.02193"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_18912"} +{"question": "Could you provide me with some resources that utilized iterative instance-specific methods to generate adversarial perturbations?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Evaluating the Robustness of Neural Networks", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1412.6572", "1608.04644", "1706.06083"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_18913"} +{"question": "Will you list down some works where a mean-field game modeling approach is used for statistical learning and inference with equilibrium effects?", "answer": ["Treatment Effects in Market Equilibrium", "Policy Learning with Competing Agents"], "answer_arxiv_id": ["2109.11647v3", "2204.01884"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_18914"} +{"question": "What studies did the researcher refer to elucidating the connection between offline bandit/RL theory and off-policy regret analysis of generated samples?", "answer": ["Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation", "Information-Theoretic Considerations in Batch Reinforcement Learning", "A Theoretical Analysis of Deep Q-Learning", "Is Pessimism Provably Efficient for Offline RL?", "Offline Neural Contextual Bandits: Pessimism, Optimization and Generalization", "Offline Contextual Bandits with Overparameterized Models"], "answer_arxiv_id": ["1810.12429", "1905.00360", "1901.00137", "2012.15085", "2111.13807", "2006.15368"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_18915"} +{"question": "Which papers improved pre-training by encouraging perceptual similarity?", "answer": ["PeCo: Perceptual Codebook for BERT Pre-training of Vision Transformers"], "answer_arxiv_id": ["2111.12710"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_18916"} +{"question": "In what paper does the researcher propose an unforgeable publicly verifiable watermarking algorithm?", "answer": ["An Unforgeable Publicly Verifiable Watermark for Large Language Models"], "answer_arxiv_id": ["2307.16230"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_18917"} +{"question": "Are there any studies questioning the correlation between localization and editing efficacy?", "answer": ["Does Localization Inform Editing? Surprising Differences in\n Causality-Based Localization vs. Knowledge Editing in Language Models"], "answer_arxiv_id": ["2301.04213"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18918"} +{"question": "Which work introduced the concept of generating noisy versions of the data through data augmentation in contrastive learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709"], "source_meta": {"published_time": "20230812"}, "qid": "AutoScholarQuery_train_18919"} +{"question": "What research proposed the retrieval of relevant examples that are semantically similar to the test sample to address the issue of in-context example selection?", "answer": ["Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity"], "answer_arxiv_id": ["2104.08786"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_18920"} +{"question": "What papers come up with models for generating torsion angles in the context of small molecular design?", "answer": ["Torsional Diffusion for Molecular Conformer Generation", "DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking"], "answer_arxiv_id": ["2206.01729", "2210.01776"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_18921"} +{"question": "What studies have shown that language models tend to reflect biases and stereotypes based on societal attributes such as gender, race, religion and nationality?", "answer": ["Persistent Anti-Muslim Bias in Large Language Models", "The Woman Worked as a Babysitter: On Biases in Language Generation", "Social Biases in NLP Models as Barriers for Persons with Disabilities", "Societal Biases in Language Generation: Progress and Challenges", "Nationality Bias in Text Generation"], "answer_arxiv_id": ["2101.05783", "1909.01326v2", "2005.00813", "2105.04054", "2302.02463"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_18922"} +{"question": "What studies are there on the establishment of correspondence between multi-modal and spatio-temporal inputs in computer vision?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2103.00020", "2104.14294"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_18923"} +{"question": "Can you list the works that explored tuning-free methods to tackle challenges posed by resource-intensive backpropagation?", "answer": ["X&Fuse: Fusing Visual Information in Text-to-Image Generation", "ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation", "InstantBooth: Personalized Text-to-Image Generation without Test-Time\n Finetuning"], "answer_arxiv_id": ["2303.01000", "2302.13848", "2304.03411"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_18924"} +{"question": "What research has proven that the pessimistic model-based offline algorithm is capable of achieving optimality in offline RL?", "answer": ["Settling the Sample Complexity of Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["2204.05275"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_18925"} +{"question": "Can you cite studies using the teacher-student framework in semi-supervised learning?", "answer": ["FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results"], "answer_arxiv_id": ["2001.07685", "1703.01780"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18926"} +{"question": "What works feature more in-depth presentations of CP?", "answer": ["Conformal Prediction: a Unified Review of Theory and New Challenges"], "answer_arxiv_id": ["2005.07972"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_18927"} +{"question": "What research has worked on recovering human trajectories in global coordinates from the per-frame local human poses?", "answer": ["D&D: Learning Human Dynamics from Dynamic Camera", "GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Cameras"], "answer_arxiv_id": ["2209.08790", "2112.01524"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18928"} +{"question": "Any studies done to apply Decoder-only architecture in simultaneous machine translation?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2307.09288"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_18929"} +{"question": "In which papers do agents interact with a code-grounded environment?", "answer": ["Lemur: Harmonizing Natural Language and Code for Language Agents"], "answer_arxiv_id": ["2310.06830"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_18930"} +{"question": "Which research found that specialized architectures do not transfer well to tasks beyond the context in which they were designed?", "answer": ["Compositional Generalization in Semantic Parsing: Pre-training vs.\n Specialized Architectures"], "answer_arxiv_id": ["2007.08970"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_18931"} +{"question": "What paper introduced the SlIsotron algorithm for learning a model with unknown link function?", "answer": ["Efficient Learning of Generalized Linear and Single Index Models with Isotonic Regression"], "answer_arxiv_id": ["1104.2018"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_18932"} +{"question": "What are some studies that used reinforcement learning approaches like POMO and PPO for variable length input combinatorial optimization problems?", "answer": ["POMO: Policy Optimization with Multiple Optima for Reinforcement Learning", "Learning to Iteratively Solve Routing Problems with Dual-Aspect Collaborative Transformer"], "answer_arxiv_id": ["2010.16011", "2110.02544"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_18933"} +{"question": "What work extended this result to constrained and regularized settings?", "answer": ["Escaping limit cycles: Global convergence for constrained nonconvex-nonconcave minimax problems"], "answer_arxiv_id": ["2302.09831"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_18934"} +{"question": "What are the papers that use semi-supervised training in 6D pose estimation?", "answer": ["Category-Level 6D Object Pose Estimation in the Wild: A Semi-Supervised\n Learning Approach and A New Dataset", "CPS++: Improving Class-level 6D Pose and Shape Estimation From Monocular\n Images With Self-Supervised Learning"], "answer_arxiv_id": ["2206.15436", "2003.05848"], "source_meta": {"published_time": "20240123"}, "qid": "AutoScholarQuery_train_18935"} +{"question": "Could you provide me with the paper studying the generalization abilities of 111-layer GNNs in a transductive setting based on algorithmic stability?", "answer": ["Stability and Generalization of Graph Convolutional Neural Networks"], "answer_arxiv_id": ["1905.01004"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_18936"} +{"question": "Which paper discussed float16 mixed precision training to prevent underflows?", "answer": ["Mixed Precision Training"], "answer_arxiv_id": ["1710.03740"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_18937"} +{"question": "Do you know any research papers that describe fragment-by-fragment generation in graph-based models?", "answer": ["Learning to Extend Molecular Scaffolds with Structural Motifs", "Motif-based Graph Self-Supervised Learning for Molecular Property Prediction", "Data-Efficient Graph Grammar Learning for Molecular Generation"], "answer_arxiv_id": ["2103.03864", "2110.00987", "2203.08031"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_18938"} +{"question": "Which works use neural radiance fields to build digital twins?", "answer": ["Neural Scene Graphs for Dynamic Scenes", "Panoptic Neural Fields: A Semantic Object-Aware Neural Scene Representation", "SUDS: Scalable Urban Dynamic Scenes", "UniSim: A Neural Closed-Loop Sensor Simulator"], "answer_arxiv_id": ["2011.10379", "2205.04334", "2303.14536", "2308.01898"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_18939"} +{"question": "Which works focus on the machine unlearning in supervised learning tasks?", "answer": ["Adaptive Machine Unlearning", "Fast Yet Effective Machine Unlearning", "Machine Unlearning: Linear Filtration for Logit-based Classifiers", "Making AI Forget You: Data Deletion in Machine Learning", "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep\n Networks", "Mixed-Privacy Forgetting in Deep Networks", "Zero-Shot Machine Unlearning", "Forgetting Outside the Box: Scrubbing Deep Networks of Information\n Accessible from Input-Output Observations"], "answer_arxiv_id": ["2106.04378", "2111.08947", "2002.02730", "1907.05012", "1911.04933", "2012.13431", "2201.05629", "2003.02960"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_18940"} +{"question": "Which papers constructed manually-crafted prompts for jailbreaking LLMs?", "answer": ["GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts"], "answer_arxiv_id": ["2309.10253v4"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_18941"} +{"question": "Which papers proposed backdoor defense methods by identifying and eliminating backdoor neurons?", "answer": ["Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks", "Adversarial Neuron Pruning Purifies Backdoored Deep Models", "Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks"], "answer_arxiv_id": ["1805.12185", "2110.14430", "2101.05930"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_18942"} +{"question": "Which works achieved horizon-free property in model-based methods?", "answer": ["Is Reinforcement Learning More Difficult Than Bandits? A Near-optimal Algorithm Escaping the Curse of Horizon", "Horizon-Free Reinforcement Learning in Polynomial Time: the Power of Stationary Policies", "Is Long Horizon Reinforcement Learning More Difficult Than Short Horizon Reinforcement Learning?"], "answer_arxiv_id": ["2009.13503", "2203.12922v2", "2005.00527"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_18943"} +{"question": "Could you provide me with a study analyzing the adversarial co-authorship of human and machine?", "answer": ["Authorship Obfuscation in Multilingual Machine-Generated Text Detection"], "answer_arxiv_id": ["2401.07867"], "source_meta": {"published_time": "20240217"}, "qid": "AutoScholarQuery_train_18944"} +{"question": "What research takes advantage of 3D Gaussians as a point-based representation for generating high-quality renderings in human avatar reconstruction?", "answer": ["3D Gaussian Splatting for Real-Time Radiance Field Rendering"], "answer_arxiv_id": ["2308.04079"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18945"} +{"question": "Which works discuss fair distribution of the benefits resulting from collaboration as data sharing incentives?", "answer": ["Collaborative Fairness in Federated Learning", "Towards Fair and Privacy-Preserving Federated Deep Models", "One for One, or All for All: Equilibria and Optimality of Collaboration in Federated Learning"], "answer_arxiv_id": ["2008.12161", "1906.01167", "2103.03228"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_18946"} +{"question": "Any works about diffusion models for generative modeling on large-scale datasets?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_18947"} +{"question": "Could you provide me some papers where noise-robust frameworks are developed with robust loss functions?", "answer": ["Robust Loss Functions under Label Noise for Deep Neural Networks", "Generalized Cross Entropy Loss for Training Deep Neural Networks with\n Noisy Labels", "Symmetric Cross Entropy for Robust Learning with Noisy Labels"], "answer_arxiv_id": ["1712.09482", "1805.07836", "1908.06112"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_18948"} +{"question": "Which works exploited pre-trained language models in language tasks for domain-specific attacks?", "answer": ["Recovering Private Text in Federated Learning of Language Models"], "answer_arxiv_id": ["2205.08514"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_18949"} +{"question": "What papers discuss representation learning benchmarks?", "answer": ["Revisiting Training Strategies and Generalization Performance in Deep Metric Learning", "Supervised Contrastive Learning", "Representation Learning: A Review and New Perspectives"], "answer_arxiv_id": ["2002.08473", "2004.11362", "1206.5538"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_18950"} +{"question": "What are the recent works in video-language pretraining?", "answer": ["Expanding Language-Image Pretrained Models for General Video Recognition", "CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "X-Pool: Cross-Modal Language-Video Attention for Text-Video Retrieval", "CLIP-ViP: Adapting Pre-trained Image-Text Model to Video-Language\n Representation Alignment"], "answer_arxiv_id": ["2208.02816", "2106.11097", "2203.15086", "2209.06430"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_18951"} +{"question": "What work presented a matured version of the PFNs idea and displayed excellent performance with small tabular data?", "answer": ["TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second"], "answer_arxiv_id": ["2207.01848"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_18952"} +{"question": "Could you provide me some studies that leveraged text-conditioned inpainting models for 3D scene generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_18953"} +{"question": "What studies have used the method of linearizing knowledge bases for question answering?", "answer": ["UniK-QA: Unified Representations of Structured and Unstructured Knowledge for Open-Domain Question Answering"], "answer_arxiv_id": ["2012.14610"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_18954"} +{"question": "What works try to optimize poses with differentiable rendering in Structure-from-Motion?", "answer": ["BARF: Bundle-Adjusting Neural Radiance Fields", "Self-Calibrating Neural Radiance Fields"], "answer_arxiv_id": ["2104.06405", "2108.13826"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_18955"} +{"question": "Which works have formalized the concept of disentangled representations in capsule networks?", "answer": ["Learning the Irreducible Representations of Commutative Lie Groups", "Intertwiners between Induced Representations (with Applications to the Theory of Equivariant Neural Networks)"], "answer_arxiv_id": ["1402.4437", "1803.10743"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_18956"} +{"question": "Are there any specific studies where the partial information decomposition definition is proposed?", "answer": ["Quantifying unique information"], "answer_arxiv_id": ["1311.2852"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_18957"} +{"question": "What research proposed learning-free methods for in-context example selection?", "answer": ["What Makes Good In-Context Examples for GPT-3?", "Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity", "Complementary Explanations for Effective In-Context Learning", "Selective Annotation Makes Language Models Better Few-Shot Learners", "Diverse Demonstrations Improve In-context Compositional Generalization", "In-context Examples Selection for Machine Translation"], "answer_arxiv_id": ["2101.06804", "2104.08786", "2211.13892", "2209.01975", "2212.06800", "2212.02437"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_18958"} +{"question": "Which studies are about prompt tuning and prefix tuning for prompt engineering in vision transformer?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2104.08691", "2101.00190"], "source_meta": {"published_time": "20240402"}, "qid": "AutoScholarQuery_train_18959"} +{"question": "What papers defined the task of generating the image of the whole object as amodal completion?", "answer": ["SeGAN: Segmenting and Generating the Invisible", "Self-Supervised Scene De-occlusion"], "answer_arxiv_id": ["1703.10239", "2004.02788"], "source_meta": {"published_time": "20240125"}, "qid": "AutoScholarQuery_train_18960"} +{"question": "Which research considers data acquisition from parties with varying privacy requirements but focuses on the mean estimation problem and designing payment and privacy loss functions to get parties to report their true unit cost of privacy loss?", "answer": ["Optimal and Differentially Private Data Acquisition: Central and Local Mechanisms"], "answer_arxiv_id": ["2201.03968v2"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_18961"} +{"question": "Which works used off-policy distribution matching approach in IL from observations?", "answer": ["Off-Policy Imitation Learning from Observations", "Versatile Offline Imitation from Observations and Examples via Regularized State-Occupancy Matching"], "answer_arxiv_id": ["2102.13185", "2202.02433"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_18962"} +{"question": "Are there any papers about integrating PEFT and neural network quantization?", "answer": ["Quadapter: Adapter for GPT-2 Quantization", "AlphaTuning: Quantization-Aware Parameter-Efficient Adaptation of Large-Scale Pre-Trained Language Models"], "answer_arxiv_id": ["2211.16912", "2210.03858"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_18963"} +{"question": "What paper studied clipping for the FedAvg algorithm?", "answer": ["Understanding Clipping for Federated Learning: Convergence and Client-Level Differential Privacy"], "answer_arxiv_id": ["2106.13673"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_18964"} +{"question": "Can you name some works that proposed improving performance of CL methods by generating better views?", "answer": ["Viewmaker Networks: Learning Views for Unsupervised Representation Learning", "What Makes for Good Views for Contrastive Learning?", "Robust Contrastive Learning Using Negative Samples with Diminished Semantics", "Contrastive Learning with Stronger Augmentations"], "answer_arxiv_id": ["2010.07432", "2005.10243", "2110.14189", "2104.07713"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_18965"} +{"question": "Could you provide some works that adapted the techniques of matrix estimation beyond the independent noise assumption?", "answer": ["Entrywise Estimation of Singular Vectors of Low-Rank Matrices with Heteroskedasticity and Dependence"], "answer_arxiv_id": ["2105.13346"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_18966"} +{"question": "Could you provide the references which involved in the simulation of deformable objects?", "answer": ["Learning to Manipulate Deformable Objects without Demonstrations", "SoftGym: Benchmarking Deep Reinforcement Learning for Deformable Object Manipulation", "Learning Visible Connectivity Dynamics for Cloth Smoothing", "Graph-based Task-specific Prediction Models for Interactions between Deformable and Rigid Objects", "DiffCloud: Real-to-Sim from Point Clouds with Differentiable Simulation and Rendering of Deformable Objects"], "answer_arxiv_id": ["1910.13439", "2011.07215v2", "2105.10389", "2103.02932", "2204.03139"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_18967"} +{"question": "Are there any papers exploring the concept of model “stitching”?", "answer": ["Understanding image representations by measuring their equivariance and equivalence"], "answer_arxiv_id": ["1411.5908"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_18968"} +{"question": "Could you provide examples of studies tha only update the crucial part of the generator?", "answer": ["Image Generation From Small Datasets via Batch Statistics Adaptation"], "answer_arxiv_id": ["1904.01774"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_18969"} +{"question": "What studies showed that Boolean Multilayer Perceptrons (MLPs) are biased towards low-entropy functions?", "answer": ["Random deep neural networks are biased towards simple functions", "Neural networks are a priori biased towards Boolean functions with low\n entropy"], "answer_arxiv_id": ["1812.10156", "1909.11522"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_18970"} +{"question": "Can you mention some sources of synthetic images used in learning?", "answer": ["Playing for Data: Ground Truth from Computer Games", "Learning to See by Looking at Noise", "Replacing Labeled Real-image Datasets with Auto-generated Contours", "Procedural Image Programs for Representation Learning", "Visual Atoms: Pre-training Vision Transformers with Sinusoidal Waves"], "answer_arxiv_id": ["1608.02192v1", "2106.05963", "2206.09132", "2211.16412", "2303.01112"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_18971"} +{"question": "What works suggest resetting agent's last layer could prevent overfitting in RL?", "answer": ["The Primacy Bias in Deep Reinforcement Learning"], "answer_arxiv_id": ["2205.07802"], "source_meta": {"published_time": "20230703"}, "qid": "AutoScholarQuery_train_18972"} +{"question": "Could you point me to some research works that extend the IRM framework by considering game theory, variance penalization, information theory, nonlinear prediction functions?", "answer": ["Invariant Risk Minimization Games", "Out-of-Distribution Generalization via Risk Extrapolation (REx)", "Nonlinear Invariant Risk Minimization: A Causal Approach", "Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization", "ZIN: When and How to Learn Invariance Without Environment Partition?"], "answer_arxiv_id": ["2002.04692", "2003.00688v5", "2102.12353", "2106.06607", "2203.05818"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_18973"} +{"question": "Which works proposed text-to-video generation models that synthesize natural videos, particularly related to VLDM?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning"], "answer_arxiv_id": ["2304.08818", "2307.04725"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18974"} +{"question": "What are the works that use data augmentation as regularization by simple transformations or advanced techniques?", "answer": ["Image Data Augmentation for Deep Learning: A Survey", "mixup: Beyond Empirical Risk Minimization", "AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "AutoAugment: Learning Augmentation Strategies from Data", "Adversarial AutoAugment", "MaxUp: A Simple Way to Improve Generalization of Neural Network Training", "Pyramid Adversarial Training Improves ViT Performance", "Enhance the Visual Representation via Discrete Adversarial Training"], "answer_arxiv_id": ["2204.08610", "1710.09412", "1912.02781", "1805.09501", "1912.11188", "2002.09024", "2111.15121", "2209.07735"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_18975"} +{"question": "Could you provide me some works about mapping different expressions into the latent distribution of a mixture model through normalizing flow?", "answer": ["Variational Inference with Normalizing Flows"], "answer_arxiv_id": ["1505.05770"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_18976"} +{"question": "What paper discusses the (correlational) statistical dimension which governs the complexity of sparse linear regression in correlational SQ model?", "answer": ["Superpolynomial Lower Bounds for Learning One-Layer Neural Networks using Gradient Descent"], "answer_arxiv_id": ["2006.12011"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_18977"} +{"question": "Could you reference a paper that proposes classifier guidance for text-driven image generation?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_18978"} +{"question": "What studies focus on aligning protein with human language using LLMs?", "answer": ["Mol-Instructions: A Large-Scale Biomolecular Instruction Dataset for Large Language Models", "InstructProtein: Aligning Human and Protein Language via Knowledge\n Instruction"], "answer_arxiv_id": ["2306.08018v5", "2310.03269"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_18979"} +{"question": "Which paper introduced other distances on graphs as relative positional encodings in the Generalized-Distance Transformers?", "answer": ["Rethinking the Expressive Power of GNNs via Graph Biconnectivity"], "answer_arxiv_id": ["2301.09505"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_18980"} +{"question": "Which study introduced TracIn to estimate the influence of each training data?", "answer": ["Estimating Training Data Influence by Tracing Gradient Descent"], "answer_arxiv_id": ["2002.08484"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_18981"} +{"question": "What are the works that discuss two-point estimator is optimal for convex Lipschitz functions in online optimization?", "answer": ["An Optimal Algorithm for Bandit and Zero-Order Convex Optimization with Two-Point Feedback"], "answer_arxiv_id": ["1507.08752"], "source_meta": {"published_time": "20220927"}, "qid": "AutoScholarQuery_train_18982"} +{"question": "What works show deep linear networks optimization exhibits similar properties to those of the optimization of deep nonlinear models?", "answer": ["Exact solutions to the nonlinear dynamics of learning in deep linear neural networks", "Deep Learning without Poor Local Minima", "Deep Linear Networks with Arbitrary Loss: All Local Minima Are Global"], "answer_arxiv_id": ["1312.6120", "1605.07110", "1712.01473"], "source_meta": {"published_time": "20230101"}, "qid": "AutoScholarQuery_train_18983"} +{"question": "Are there any papers on using data manifolds obtained from unlabeled data?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "Revisiting Semi-Supervised Learning with Graph Embeddings", "Semi-supervised learning with GANs: revisiting manifold regularization", "An Overview of Deep Semi-Supervised Learning", "Overcoming the curse of dimensionality with Laplacian regularization in semi-supervised learning"], "answer_arxiv_id": ["1609.02907", "1603.08861", "1805.08957", "2006.05278", "2009.04324"], "source_meta": {"published_time": "20221001"}, "qid": "AutoScholarQuery_train_18984"} +{"question": "Which studies developed NLP models based on the Transformer?", "answer": ["Language Models are Few-Shot Learners", "Training language models to follow instructions with human feedback", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["2005.14165", "2203.02155", "1810.04805", "1910.10683", "2205.01068"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_18985"} +{"question": "In the area of contrastive learning in NLP, what studies constructed positives and negatives through parallel corpora or other labeled data?", "answer": ["Supervised Contrastive Learning for Pre-trained Language Model Fine-tuning", "Sentence-T5: Scalable Sentence Encoders from Pre-trained Text-to-Text Models"], "answer_arxiv_id": ["2011.01403", "2108.08877"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_18986"} +{"question": "What are the papers where progress in monocular depth estimation through large-scale datasets is discussed?", "answer": ["ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "The Cityscapes Dataset for Semantic Urban Scene Understanding"], "answer_arxiv_id": ["1702.04405", "1604.01685v2"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_18987"} +{"question": "What papers discussed ways of modifying the network structure, particularly in the development of Spike-Element-Wise ResNet and Spikformer?", "answer": ["Deep Residual Learning in Spiking Neural Networks", "Spikformer: When Spiking Neural Network Meets Transformer"], "answer_arxiv_id": ["2102.04159", "2209.15425"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_18988"} +{"question": "Could you name the works which used Contrastive VLMs to drive exploration, construct semantic representation, or act as agent’s vision and text encoders?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Semantic Exploration from Language Abstractions and Pretrained Representations", "Open-vocabulary Queryable Scene Representations for Real World Planning", "ProgPrompt: Generating Situated Robot Task Plans using Large Language Models", "ZSON: Zero-Shot Object-Goal Navigation using Multimodal Goal Embeddings", "LM-Nav: Robotic Navigation with Large Pre-Trained Models of Language, Vision, and Action", "CLIPort: What and Where Pathways for Robotic Manipulation", "LATTE: LAnguage Trajectory TransformEr"], "answer_arxiv_id": ["2103.00020", "2204.05080", "2209.09874", "2209.11302", "2206.12403", "2207.04429", "2109.12098", "2208.02918"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_18989"} +{"question": "What works integrate the Transformer into point cloud processing?", "answer": ["Voxel Transformer for 3D Object Detection", "Embracing Single Stride 3D Object Detector with Sparse Transformer", "SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds", "DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets", "UniTR: A Unified and Efficient Multi-Modal Transformer for Bird’s-Eye-View Representation"], "answer_arxiv_id": ["2109.02497", "2112.06375", "2210.07372", "2301.06051", "2308.07732"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_18990"} +{"question": "Which work proposed a pipeline called 'AnimateDiff' for traning a plug-and-play motion module?", "answer": ["AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning"], "answer_arxiv_id": ["2307.04725"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_18991"} +{"question": "What works have been done on improving the effectiveness of in-context learning through prompt and exemplars engineering?", "answer": ["What Makes Good In-Context Examples for GPT-$3$?", "Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing"], "answer_arxiv_id": ["2101.06804", "2107.13586v1"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_18992"} +{"question": "What papers propose to enhance visual sign representations and text decoding capabilities of SLT methods?", "answer": ["Neural Sign Language Translation based on Human Keypoint Estimation", "Sign Language Translation with Hierarchical Spatio-TemporalGraph Neural\n Network", "Sign Language Transformers: Joint End-to-end Sign Language Recognition\n and Translation", "TSPNet: Hierarchical Feature Learning via Temporal Semantic Pyramid for\n Sign Language Translation", "Better Sign Language Translation with STMC-Transformer", "A Token-level Contrastive Framework for Sign Language Translation", "Two-Stream Network for Sign Language Recognition and Translation", "SLTUNET: A Simple Unified Model for Sign Language Translation", "Gloss-free Sign Language Translation: Improving from Visual-Language\n Pretraining"], "answer_arxiv_id": ["1811.11436", "2111.07258", "2003.13830", "2010.05468", "2004.00588", "2204.04916", "2211.01367", "2305.01778", "2307.14768"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_18993"} +{"question": "What are the studies that utilized diffusion models for face morphing?", "answer": ["Diffusion Autoencoders: Toward a Meaningful and Decodable Representation"], "answer_arxiv_id": ["2111.15640"], "source_meta": {"published_time": "20230826"}, "qid": "AutoScholarQuery_train_18994"} +{"question": "Can you name the study that trained a powerful decoder model to decode the entire sequence by representing the embedding of sentences as the initial token?", "answer": ["Sentence Embedding Leaks More Information than You Expect: Generative\n Embedding Inversion Attack to Recover the Whole Sentence"], "answer_arxiv_id": ["2305.03010"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_18995"} +{"question": "Could you provide me the studies about generative adversarial networks (GANs) replacing the PCA-basis in mesh-based 3DMMs?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_18996"} +{"question": "What are the recent iterative optimization-based methods that have shown impressive performance in stereo matching?", "answer": ["Practical Stereo Matching via Cascaded Recurrent Network with Adaptive\n Correlation", "RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching"], "answer_arxiv_id": ["2203.11483", "2109.07547"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_18997"} +{"question": "Which studies propose intrinsically motivated RL algorithms that choose to explore outcomes rather than actions?", "answer": ["Exploration in Deep Reinforcement Learning: A Survey"], "answer_arxiv_id": ["2205.00824v1"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_18998"} +{"question": "Which works suggest using parameter-shared networks, called supernets, in one-/few-shot NAS methods?", "answer": ["DARTS: Differentiable Architecture Search", "ProxylessNAS: Direct Neural Architecture Search on Target Task and\n Hardware", "Once-for-All: Train One Network and Specialize it for Efficient\n Deployment", "Few-shot Neural Architecture Search", "Generalizing Few-Shot NAS with Gradient Matching"], "answer_arxiv_id": ["1806.09055", "1812.00332", "1908.09791", "2006.06863", "2203.15207"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_18999"} +{"question": "Which works adopted a global perspective for video comprehension in the VidSitu benchmark?", "answer": ["Grounded Video Situation Recognition"], "answer_arxiv_id": ["2210.10828"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19000"} +{"question": "Could you provide me some studies about segmentation from language supervision?", "answer": ["GroupViT: Semantic Segmentation Emerges from Text Supervision", "Open-world Semantic Segmentation via Contrasting and Clustering Vision-Language Embedding", "Learning to Generate Text-grounded Mask for Open-world Semantic Segmentation from Only Image-Text Pairs", "Open Vocabulary Semantic Segmentation with Patch Aligned Contrastive Learning", "ViewCo: Discovering Text-Supervised Segmentation Masks via Multi-View Semantic Consistency", "Learning Open-vocabulary Semantic Segmentation Models From Natural Language Supervision"], "answer_arxiv_id": ["2202.11094", "2207.08455", "2212.00785", "2212.04994", "2302.10307", "2301.09121"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_19001"} +{"question": "Which works use decentralized SGD algorithms for large-scale deep training?", "answer": ["On the Convergence of Decentralized Gradient Descent", "Decentralized Stochastic Optimization and Gossip Algorithms with Compressed Communication"], "answer_arxiv_id": ["1310.7063", "1902.00340"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_19002"} +{"question": "Which papers introduced diffusion probabilistic models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_19003"} +{"question": "Which research papers have combined fuzzy logic to tackle complex query answering?", "answer": ["Fuzzy Logic Based Logical Query Answering on Knowledge Graphs", "Neural-Symbolic Models for Logical Queries on Knowledge Graphs"], "answer_arxiv_id": ["2108.02390", "2205.10128"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_19004"} +{"question": "What works proposed advanced loss based on optimal transportation and feed-forward paint transformer to improve the painting process?", "answer": ["Stylized Neural Painting", "Paint Transformer: Feed Forward Neural Painting with Stroke Prediction"], "answer_arxiv_id": ["2011.08114", "2108.03798"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19005"} +{"question": "What are some of the studies that extended pairwise flow by incorporating multi-frame contexts?", "answer": ["A Fusion Approach for Multi-Frame Optical Flow Estimation"], "answer_arxiv_id": ["1810.10066"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_19006"} +{"question": "Can you list research works that utilize reinforcement learning for pruning in DNNs?", "answer": ["AMC: AutoML for Model Compression and Acceleration on Mobile Devices"], "answer_arxiv_id": ["1802.03494"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_19007"} +{"question": "Are there any works on distilling large-scale pre-trained language models?", "answer": ["DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter", "TinyBERT: Distilling BERT for Natural Language Understanding", "MiniLMv2: Multi-Head Self-Attention Relation Distillation for Compressing Pretrained Transformers", "MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices"], "answer_arxiv_id": ["1910.01108", "1909.10351", "2012.15828", "2004.02984"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_19008"} +{"question": "Which research paper discusses the study of population guarantees for stationarity in constrained settings?", "answer": ["Differentially Private Stochastic Optimization: New Results in Convex and Non-Convex Settings"], "answer_arxiv_id": ["2107.05585"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_19009"} +{"question": "Which studies used a neural network for volume density mapping in 3D spatial locations?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20220512"}, "qid": "AutoScholarQuery_train_19010"} +{"question": "Which papers focus on the utilization of volume rendering techniques in the context of neural radiance fields?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "RenderDiffusion: Image Diffusion for 3D Reconstruction, Inpainting and\n Generation", "Neuralangelo: High-Fidelity Neural Surface Reconstruction", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Shape, Pose, and Appearance from a Single Image via Bootstrapped\n Radiance Field Inversion"], "answer_arxiv_id": ["2003.08934", "2211.09869", "2306.03092", "2106.10689", "2211.11674"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_19011"} +{"question": "Could you tell me about the works that proposed to measure the cross-correlation matrix of distorted views in contrastive learning?", "answer": ["Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "Whitening for Self-Supervised Representation Learning", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning"], "answer_arxiv_id": ["2103.03230", "2007.06346", "2105.04906"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_19012"} +{"question": "Which works presented impressive results in the field of combining planning algorithms with deep learning and their applications?", "answer": ["Generative Language Modeling for Automated Theorem Proving", "Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm", "Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model", "MuZero with Self-competition for Rate Control in VP9 Video Compression"], "answer_arxiv_id": ["2009.03393", "1712.01815", "1911.08265", "2202.06626"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_19013"} +{"question": "Which works improved NeRFs by making it dynamic and conditioning the network on both space and time?", "answer": ["NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections", "Neural 3D Video Synthesis from Multi-view Video"], "answer_arxiv_id": ["2008.02268", "2103.02597"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_19014"} +{"question": "Which papers discuss sliding window-based methods in the context of VSR?", "answer": ["Real-Time Video Super-Resolution with Spatio-Temporal Networks and\n Motion Compensation", "Video Super-resolution with Temporal Group Attention", "MuCAN: Multi-Correspondence Aggregation Network for Video\n Super-Resolution", "TDAN: Temporally Deformable Alignment Network for Video Super-Resolution", "EDVR: Video Restoration with Enhanced Deformable Convolutional Networks"], "answer_arxiv_id": ["1611.05250", "2007.10595", "2007.11803", "1812.02898", "1905.02716"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_19015"} +{"question": "Can you point me to works that used attention mechanism in neural network building blocks?", "answer": ["Choose a Transformer: Fourier or Galerkin"], "answer_arxiv_id": ["2105.14995"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_19016"} +{"question": "Who proposed language-aware methods for keyframe localization?", "answer": ["LGDN: Language-Guided Denoising Network for Video-Language Modeling", "Revisiting the “Video” in Video-Language Understanding", "MIST: Multi-modal Iterative Spatial-Temporal Transformer for Long-form Video Question Answering", "Locate before Answering: Answer Guided Question Localization for Video Question Answering", "Semi-Parametric Video-Grounded Text Generation"], "answer_arxiv_id": ["2209.11388", "2206.01720", "2212.09522", "2210.02081", "2301.11507"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_19017"} +{"question": "Which papers deal with the intersection of nonstochastic control and learning, specifically for adversarially chosen cost functions and perturbations?", "answer": ["Online Control with Adversarial Disturbances"], "answer_arxiv_id": ["1902.08721"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19018"} +{"question": "Which research works employed fine-grained action datasets annotated with sub-action boundaries for fine-grained action recognition?", "answer": ["Recognizing Fine-Grained and Composite Activities using Hand-Centric Features and Script Data", "FineGym: A Hierarchical Video Dataset for Fine-grained Action Understanding"], "answer_arxiv_id": ["1502.06648", "2004.06704"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_19019"} +{"question": "What works involve utilizing the FPN-series in encoder-decoder structures?", "answer": ["Feature Pyramid Networks for Object Detection", "EfficientDet: Scalable and Efficient Object Detection", "NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detection"], "answer_arxiv_id": ["1612.03144", "1911.09070", "1904.07392"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_19020"} +{"question": "What work apply active learning in the context of simulation and deep learning?", "answer": ["A Survey of Deep Active Learning"], "answer_arxiv_id": ["2009.00236"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_19021"} +{"question": "Which study combines Adversarial Training (AT) with a Zero-Shot Learning (ZSL)?", "answer": ["ATZSL: Defensive Zero-Shot Recognition in the Presence of Adversaries"], "answer_arxiv_id": ["1910.10994"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_19022"} +{"question": "Which works have popularized knowledge distillation for classification?", "answer": ["Distilling the Knowledge in a Neural Network"], "answer_arxiv_id": ["1503.02531"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_19023"} +{"question": "What are some of the works that utilized transformer architectures for continual learning?", "answer": ["Improving Vision Transformers for Incremental Learning", "DyTox: Transformers for Continual Learning with DYnamic TOken eXpansion", "Towards Exemplar-Free Continual Learning in Vision Transformers: an Account of Attention, Functional and Weight Regularization", "Continual Learning with Transformers for Image Classification", "Learning to Prompt for Continual Learning", "DualPrompt: Complementary Prompting for Rehearsal-free Continual Learning"], "answer_arxiv_id": ["2112.06103", "2111.11326", "2203.13167", "2206.14085", "2112.08654", "2204.04799"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_19024"} +{"question": "What papers have made progress in refined camera pose during SLAM or Visual Odometry when multiple views are available?", "answer": ["DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras", "DeepFactors: Real-Time Probabilistic Dense Monocular SLAM", "XVO: Generalized Visual Odometry via Cross-Modal Self-Training", "TartanVO: A Generalizable Learning-based VO", "Deep Patch Visual Odometry"], "answer_arxiv_id": ["2108.10869", "2001.05049", "2309.16772", "2011.00359", "2208.04726"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_19025"} +{"question": "Which papers criticise the shallow approach of evaluating NLP progress using benchmarks?", "answer": ["Memorization vs. Generalization: Quantifying Data Leakage in NLP\n Performance Evaluation"], "answer_arxiv_id": ["2102.01818"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_19026"} +{"question": "What studies employ parameter isolation-based approaches in continual learning?", "answer": ["Efficient Feature Transformations for Discriminative and Generative Continual Learning", "Efficient Architecture Search for Continual Learning", "Overcoming Catastrophic Forgetting with Hard Attention to the Task", "Achieving Forgetting Prevention and Knowledge Transfer in Continual Learning"], "answer_arxiv_id": ["2103.13558", "2006.04027", "1801.01423", "2112.02706"], "source_meta": {"published_time": "20220722"}, "qid": "AutoScholarQuery_train_19027"} +{"question": "Could you provide me some papers about generative-based approaches in unpaired image denoising?", "answer": ["Generative Adversarial Networks"], "answer_arxiv_id": ["1406.2661"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_19028"} +{"question": "What research papers mention the relationship of equivariance to disentangled representations in capsule networks?", "answer": ["Dynamic Routing Between Capsules", "Stacked Capsule Autoencoders"], "answer_arxiv_id": ["1710.09829", "1906.06818"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_19029"} +{"question": "Which benchmarks are used for assessing LLMs across a diverse range of tasks?", "answer": ["Holistic Evaluation of Language Models", "Beyond the Imitation Game: Quantifying and extrapolating the\n capabilities of language models", "SuperGlue: Learning Feature Matching with Graph Neural Networks", "Language Models as Knowledge Bases?", "Chain-of-Thought Hub: A Continuous Effort to Measure Large Language\n Models' Reasoning Performance"], "answer_arxiv_id": ["2211.09110", "2206.04615", "1911.11763", "1909.01066", "2305.17306"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_19030"} +{"question": "What studies have focused on creating tasks for evaluating forms of inferential reasoning common to law?", "answer": ["LexGLUE: A Benchmark Dataset for Legal Language Understanding in English", "Holistic Evaluation of Language Models", "When Does Pretraining Help? Assessing Self-Supervised Learning for Law and the CaseHOLD Dataset of 53,000+ Legal Holdings", "A Dataset for Statutory Reasoning in Tax Law Entailment and Question Answering"], "answer_arxiv_id": ["2110.00976", "2211.09110", "2104.08671", "2005.05257"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_19031"} +{"question": "What are some works focusing on rule mining in abductive knowledge graph reasoning?", "answer": ["Neural Compositional Rule Learning for Knowledge Graph Reasoning"], "answer_arxiv_id": ["2303.03581"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_19032"} +{"question": "Which study utilises gradients to guide the optimization of features in InstantNGP?", "answer": ["Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"], "answer_arxiv_id": ["2201.05989"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_19033"} +{"question": "Which studies are proposed in the field of controllable image generation?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2302.05543"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19034"} +{"question": "What study improves the spiking neurons by adding input gates, forget gates, recurrent connections and multi-bit outputs?", "answer": ["Spiking Neural Networks with Improved Inherent Recurrence Dynamics for Sequential Learning"], "answer_arxiv_id": ["2109.01905"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_19035"} +{"question": "Could you provide me some papers where freezing parameters has been exhibited to preserve learned information?", "answer": ["Unsupervised Domain Adaptation with Residual Transfer Networks", "Early Stopping without a Validation Set", "Freeze the Discriminator: a Simple Baseline for Fine-Tuning GANs", "Fast Context Adaptation via Meta-Learning", "Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML", "Learning a Universal Template for Few-shot Dataset Generalization"], "answer_arxiv_id": ["1602.04433", "1703.09580", "2002.10964", "1810.03642", "1909.09157", "2105.07029v2"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_19036"} +{"question": "Which papers study joint training on labelled and unlabelled data in semi-supervised learning?", "answer": ["Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning", "S4L: Self-Supervised Semi-Supervised Learning", "Interpolation Consistency Training for Semi-Supervised Learning", "MultiGrain: a unified image embedding for classes and instances", "Unsupervised Data Augmentation for Consistency Training"], "answer_arxiv_id": ["1704.03976", "1905.03670", "1903.03825", "1902.05509", "1904.12848"], "source_meta": {"published_time": "20230112"}, "qid": "AutoScholarQuery_train_19037"} +{"question": "What are some papers that have made attempts to treat semantic/instance/panoptic segmentation tasks in a unified manner?", "answer": ["Per-Pixel Classification is Not All You Need for Semantic Segmentation", "Masked-attention Mask Transformer for Universal Image Segmentation"], "answer_arxiv_id": ["2107.06278", "2112.01527"], "source_meta": {"published_time": "20220818"}, "qid": "AutoScholarQuery_train_19038"} +{"question": "What research introduced an end-to-end approach that reprograms pre-trained acoustic models for time series classification by input transformation learning and output label mapping?", "answer": ["Voice2Series: Reprogramming Acoustic Models for Time Series Classification"], "answer_arxiv_id": ["2106.09296"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_19039"} +{"question": "In which papers the authors haven't considered the symmetries when designing architectures for processing weight space objects, but managing it through data augmentation?", "answer": ["Learning to learn by gradient descent by gradient descent", "Learning to Optimize", "HyperNetworks", "Bayesian Hypernetworks", "Graph HyperNetworks for Neural Architecture Search", "A Generative Model for Sampling High-Performance and Diverse Weights for Neural Networks", "Parameter Prediction for Unseen Deep Architectures", "Learning to Learn with Generative Models of Neural Network Checkpoints", "VeLO: Training Versatile Learned Optimizers by Scaling Up"], "answer_arxiv_id": ["1606.04474", "1606.01885", "1609.09106v4", "1710.04759", "1810.05749", "1905.02898", "2110.13100", "2209.12892", "2211.09760"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_19040"} +{"question": "Which research works have used optimizing the model with unlabeled test data in machine learning tasks?", "answer": ["Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models", "Neural Networks with Recurrent Generative Feedback", "“Zero-Shot” Super-Resolution using Deep Internal Learning", "Efficient Test-Time Model Adaptation without Forgetting", "Continual Test-Time Domain Adaptation"], "answer_arxiv_id": ["2209.07511", "2007.09200", "1712.06087", "2204.02610", "2203.13591"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_19041"} +{"question": "Which research introduced a suite of transformers pre-trained at scale in the area of Large Language Models?", "answer": ["OPT: Open Pre-trained Transformer Language Models"], "answer_arxiv_id": ["2205.01068"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19042"} +{"question": "What studies have focused on legal analysis in the context of NLP models?", "answer": ["Learning to Predict Charges for Criminal Cases with Legal Basis", "ILDC for CJPE: Indian Legal Documents Corpus for Court Judgment Prediction and Explanation", "LeSICiN: A Heterogeneous Graph-based Approach for Automatic Legal Statute Identification from Indian Legal Documents", "Legal-Tech Open Diaries: Lesson learned on how to develop and deploy light-weight models in the era of humongous Language Models"], "answer_arxiv_id": ["1707.09168", "2105.13562", "2112.14731", "2210.13086"], "source_meta": {"published_time": "20230820"}, "qid": "AutoScholarQuery_train_19043"} +{"question": "Which papers provide study about sound generation from visual data?", "answer": ["Foley Music: Learning to Generate Music from Videos", "Collaborative Learning to Generate Audio-Video Jointly", "Audeo: Audio Generation for a Silent Performance Video", "2.5D Visual Sound", "Beyond Mono to Binaural: Generating Binaural Audio from Mono Audio with Depth and Cross Modal Attention", "Sep-Stereo: Visually Guided Stereophonic Audio Generation by Associating Source Separation", "Visually Informed Binaural Audio Generation without Binaural Audios", "Exploiting Audio-Visual Consistency with Partial Supervision for Spatial Audio Generation"], "answer_arxiv_id": ["2007.10984", "2104.02656", "2006.14348", "1812.04204", "2111.08046", "2007.09902", "2104.06162", "2105.00708"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_19044"} +{"question": "What research works provide conditions under which complete dictionary learning (DL) can be solved?", "answer": ["Exact Recovery of Sparsely-Used Dictionaries", "Simple Alternating Minimization Provably Solves Complete Dictionary Learning"], "answer_arxiv_id": ["1206.5882", "2210.12816v1"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19045"} +{"question": "Which papers are there on optimizing FSDP?", "answer": ["OSDP: Optimal Sharded Data Parallel for Distributed Deep Learning", "Galvatron: Efficient Transformer Training over Multiple GPUs Using Automatic Parallelism"], "answer_arxiv_id": ["2209.13258", "2211.13878"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_19046"} +{"question": "Which studies utilize diffusion models for face restoration?", "answer": ["Denoising Diffusion Probabilistic Models", "DifFace: Blind Face Restoration with Diffused Error Contraction", "DR2: Diffusion-based Robust Degradation Remover for Blind Face\n Restoration"], "answer_arxiv_id": ["2006.11239", "2212.06512", "2303.06885"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_19047"} +{"question": "Which works proposed learning a combined state-object semantic representation with a transformation function?", "answer": ["Task-Driven Modular Networks for Zero-Shot Compositional Learning", "Learning Graph Embeddings for Compositional Zero-shot Learning", "Learning Graph Embeddings for Open World Compositional Zero-Shot\n Learning", "On Leveraging Variational Graph Embeddings for Open World Compositional\n Zero-Shot Learning"], "answer_arxiv_id": ["1905.05908", "2102.01987", "2105.01017", "2204.11848"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_19048"} +{"question": "What studies have used embeddings based on the Hodge-Laplacian?", "answer": ["Random Walks on Simplicial Complexes and the normalized Hodge 1-Laplacian", "A Notion of Harmonic Clustering in Simplicial Complexes"], "answer_arxiv_id": ["1807.05044v5", "1910.07247"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_19049"} +{"question": "Are there any studies that introduce a trainable object-level augmentor to enrich the variance of foreground objects and improve data augmentation?", "answer": ["Learning Object-level Point Augmentor for Semi-supervised 3D Object Detection"], "answer_arxiv_id": ["2212.09273"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19050"} +{"question": "What are the works that have continued piano audios unconditionally?", "answer": ["AudioLM: a Language Modeling Approach to Audio Generation"], "answer_arxiv_id": ["2209.03143"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19051"} +{"question": "What research used SVD to decompose degradation operators while dealing with the image inpainting problem?", "answer": ["Denoising Diffusion Restoration Models"], "answer_arxiv_id": ["2201.11793"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_19052"} +{"question": "What research uses reinforcement learning to learn an optimal dialog policy for identifying movie entries from a movie database?", "answer": ["Towards End-to-End Reinforcement Learning of Dialogue Agents for\n Information Access"], "answer_arxiv_id": ["1609.00777"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_19053"} +{"question": "Which works establish the use of the Polyak stepsize in stochastic non-convex optimization?", "answer": ["Stochastic Polyak Step-size for SGD: An Adaptive Learning Rate for Fast Convergence"], "answer_arxiv_id": ["2002.10542"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19054"} +{"question": "What works investigated the in-context reasoning ability of Large Language Models?", "answer": ["Language Models are Few-Shot Learners", "Emergent Abilities of Large Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Least-to-Most Prompting Enables Complex Reasoning in Large Language\n Models"], "answer_arxiv_id": ["2005.14165", "2206.07682", "2201.11903", "2205.10625"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_19055"} +{"question": "Could you provide me some studies about Transformer architectures that can model periodic finite-state languages or Dyck languages?", "answer": ["Theoretical Limitations of Self-Attention in Neural Sequence Models", "How Can Self-Attention Networks Recognize Dyck-n Languages?"], "answer_arxiv_id": ["1906.06755", "2010.04303"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_19056"} +{"question": "What studies have employed Rank-One Model Editing (ROME) in T2I personalization?", "answer": ["Key-Locked Rank One Editing for Text-to-Image Personalization", "Locating and Editing Factual Associations in GPT"], "answer_arxiv_id": ["2305.01644", "2202.05262"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_19057"} +{"question": "What paper first proposed using line-search to set the stepsize for AdaGrad to enhance its practical performance?", "answer": ["Adaptive Gradient Methods Converge Faster with Over-Parameterization (but you should do a line-search)"], "answer_arxiv_id": ["2006.06835v3"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_19058"} +{"question": "Which papers utilize Depthwise separable convolution for developing lightweight and efficient models?", "answer": ["A ConvNet for the 2020s", "ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture\n Design"], "answer_arxiv_id": ["2201.03545", "1807.11164"], "source_meta": {"published_time": "20240111"}, "qid": "AutoScholarQuery_train_19059"} +{"question": "In what research papers are different GNNs proposed to improve performance in low-homophily settings?", "answer": ["Geom-GCN: Geometric Graph Convolutional Networks", "MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing", "Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs", "Adaptive Universal Generalized PageRank Graph Neural Network", "BernNet: Learning Arbitrary Graph Spectral Filters via Bernstein Approximation", "Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods", "Revisiting Heterophily For Graph Neural Networks", "Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs", "Finding Global Homophily in Graph Neural Networks When Meeting Heterophily"], "answer_arxiv_id": ["2002.05287", "1905.00067", "2006.11468v2", "2006.07988", "2106.10994", "2110.14446", "2210.07606", "2202.04579", "2205.07308"], "source_meta": {"published_time": "20221125"}, "qid": "AutoScholarQuery_train_19060"} +{"question": "Could you provide me some works that applied graph coarsening techniques to scale up GNNs?", "answer": ["Scaling Up Graph Neural Networks Via Graph Coarsening"], "answer_arxiv_id": ["2106.05150"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_19061"} +{"question": "Which research focused on the implementation of adaptation via aligned representations method in domain adaptation?", "answer": ["Domain Adaptation In Reinforcement Learning Via Latent Unified State Representation", "DARLA: Improving Zero-Shot Transfer in Reinforcement Learning"], "answer_arxiv_id": ["2102.05714", "1707.08475"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_19062"} +{"question": "Any works attempted to address feature suppression in contrastive learning via auxiliary losses or modifying representations in the latent space?", "answer": ["Addressing Feature Suppression in Unsupervised Visual Representations", "Can contrastive learning avoid shortcut solutions?"], "answer_arxiv_id": ["2012.09962", "2106.11230"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_19063"} +{"question": "Which papers have contributed to compiling multi-choice questions for evaluating diverse reasoning capabilities of LLMs?", "answer": ["LogiQA: A Challenge Dataset for Machine Reading Comprehension with\n Logical Reasoning", "ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning"], "answer_arxiv_id": ["2007.08124", "2002.04326"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_19064"} +{"question": "In which works was the robustness of the ViTs over CNNs in terms of adversarial noise demonstrated?", "answer": ["Understanding and Improving Robustness of Vision Transformers through Patch-based Negative Augmentation"], "answer_arxiv_id": ["2110.07858"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_19065"} +{"question": "What datasets were categorized according to CFL learners in CSC?", "answer": ["YACLC: A Chinese Learner Corpus with Multidimensional Annotation"], "answer_arxiv_id": ["2112.15043"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_19066"} +{"question": "Which research papers have used variational autoencoders for learning 3D shapes?", "answer": ["SetVAE: Learning Hierarchical Composition for Generative Modeling of Set-Structured Data", "Adversarial Autoencoders", "Multiresolution Tree Networks for 3D Point Cloud Processing"], "answer_arxiv_id": ["2103.15619", "1511.05644", "1807.03520v2"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_19067"} +{"question": "Could you mention some works that use model stitching to measure the similarities for the inner representations learned by deep neural networks?", "answer": ["Revisiting Model Stitching to Compare Neural Representations", "Similarity and Matching of Neural Network Representations", "Model Stitching: Looking For Functional Similarity Between\n Representations"], "answer_arxiv_id": ["2106.07682", "2110.14633", "2303.11277"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19068"} +{"question": "What research addressed data poisoning in a streaming-data setting?", "answer": ["Data Poisoning Attacks against Online Learning"], "answer_arxiv_id": ["1808.08994"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19069"} +{"question": "Any works implementing the strategy of iterative trimming for sample reweighting?", "answer": ["Ordered SGD: A New Stochastic Optimization Framework for Empirical Risk Minimization", "Learning with Bad Training Data via Iterative Trimmed Loss Minimization"], "answer_arxiv_id": ["1907.04371", "1810.11874"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_19070"} +{"question": "Could you provide me some works related to classic matrix factorization problems?", "answer": ["Implicit Regularization in Matrix Factorization", "Algorithmic Regularization in Over-parameterized Matrix Sensing and Neural Networks with Quadratic Activations", "Implicit Regularization in Deep Matrix Factorization", "Nonconvex Optimization Meets Low-Rank Matrix Factorization: An Overview"], "answer_arxiv_id": ["1705.09280", "1712.09203", "1905.13655", "1809.09573"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_19071"} +{"question": "Could you provide some references regarding active learning approaches that prioritize data points to be labeled based on their informativeness?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "GeneDisco: A Benchmark for Experimental Design in Drug Discovery", "Bayesian Active Learning for Classification and Preference Learning", "BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning", "Test Distribution–Aware Active Learning: A Principled Approach Against Distribution Shift and Outliers"], "answer_arxiv_id": ["1708.00489", "2110.11875v1", "1112.5745v1", "1906.08158", "2106.11719"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19072"} +{"question": "What studies used importance sampling techniques employing the gradient norm or the loss to speed up learning?", "answer": ["Variance Reduction in SGD by Distributed Importance Sampling", "Not All Samples Are Created Equal: Deep Learning with Importance Sampling", "Online Batch Selection for Faster Training of Neural Networks", "Prioritized Experience Replay"], "answer_arxiv_id": ["1511.06481", "1803.00942v3", "1511.06343", "1511.05952"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_19073"} +{"question": "What paper proposed differentiable weighted finite state transducers?", "answer": ["Differentiable Weighted Finite-State Transducers"], "answer_arxiv_id": ["2010.01003"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_19074"} +{"question": "Which papers discuss how knowledge distillation, a method for knowledge transfer, has been used in the field of text generation?", "answer": ["Distilling the Knowledge in a Neural Network", "Knowledge Distillation: A Survey"], "answer_arxiv_id": ["1503.02531", "2006.05525"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_19075"} +{"question": "What are some studies that recommend generating 3D Point clouds as a solution for representating 3D shapes?", "answer": ["Diffusion Probabilistic Models for 3D Point Cloud Generation", "LION: Latent Point Diffusion Models for 3D Shape Generation", "3D Shape Generation and Completion through Point-Voxel Diffusion", "Point-E: A System for Generating 3D Point Clouds from Complex Prompts"], "answer_arxiv_id": ["2103.01458", "2210.06978", "2104.03670", "2212.08751"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_19076"} +{"question": "Could you mention some studies that use graph neural networks for graph classification?", "answer": ["Semi-Supervised Classification with Graph Convolutional Networks", "How Powerful are Graph Neural Networks?", "Weisfeiler and Lehman Go Cellular: CW Networks", "Accurate Learning of Graph Representations with Graph Multiset Pooling", "TGNN: A Joint Semi-Supervised Framework for Graph-Level Classification"], "answer_arxiv_id": ["1609.02907", "1810.00826", "2106.12575", "2102.11533", "2304.11688"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_19077"} +{"question": "Can you identify the study where a text-conditional diffusion model is fine-tuned for the inpainting task?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2112.10741"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_19078"} +{"question": "What are some recent studies that have gained popularity in the text-to-3D task?", "answer": ["Denoising Diffusion Probabilistic Models", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2006.11239", "2103.00020"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19079"} +{"question": "Are there any methods for music generation using diffusion models?", "answer": ["Noise2Music: Text-conditioned Music Generation with Diffusion Models", "Efficient Neural Music Generation"], "answer_arxiv_id": ["2302.03917", "2305.15719"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_19080"} +{"question": "Which works have been instrumental in building state-of-the-art object detectors for localizing objects in images?", "answer": ["Focal Loss for Dense Object Detection", "Fast R-CNN", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", "End-to-End Object Detection with Transformers"], "answer_arxiv_id": ["1708.02002", "1504.08083", "1506.01497", "2005.12872"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_19081"} +{"question": "Which research examines the Florence detector, which was pre-trained with large volumes of privately curated image-text pairs?", "answer": ["Florence: A New Foundation Model for Computer Vision"], "answer_arxiv_id": ["2111.11432"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_19082"} +{"question": "Could you please provide papers validating poor performance of the value decomposition structure and IGM assumption in coordination tasks such as the XOR game?", "answer": ["Revisiting Some Common Practices in Cooperative Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["2206.07505"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_19083"} +{"question": "Which paper proposed a debiased contrastive loss?", "answer": ["Debiased Contrastive Learning"], "answer_arxiv_id": ["2007.00224"], "source_meta": {"published_time": "20221023"}, "qid": "AutoScholarQuery_train_19084"} +{"question": "Where can I find a comprehensive survey of relational (relation-based) KD?", "answer": ["Knowledge Distillation: A Survey"], "answer_arxiv_id": ["2006.05525"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_19085"} +{"question": "What are some papers that have created full-body images considering additional input signals?", "answer": ["InsetGAN for Full-Body Image Generation", "Putting People in Their Place: Affordance-Aware Human Insertion into\n Scenes", "Collage Diffusion", "Pose with Style: Detail-Preserving Pose-Guided Image Synthesis with\n Conditional StyleGAN", "Reposing Humans by Warping 3D Features", "Pose Guided Person Image Generation", "Controllable Person Image Synthesis with Attribute-Decomposed GAN", "Learning Realistic Human Reposing using Cyclic Self-Supervision with 3D\n Shape, Pose, and Appearance Consistency", "Appearance and Pose-Conditioned Human Image Generation using Deformable\n GANs", "DreamPose: Fashion Image-to-Video Synthesis via Stable Diffusion", "Person Image Synthesis via Denoising Diffusion Model", "Thin-Plate Spline Motion Model for Image Animation", "Neural Texture Extraction and Distribution for Controllable Person Image\n Synthesis"], "answer_arxiv_id": ["2203.07293", "2304.14406", "2303.00262", "2109.06166", "2006.04898v1", "1705.09368", "2003.12267", "2110.05458", "1905.00007", "2304.06025", "2211.12500", "2203.14367", "2204.06160"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_19086"} +{"question": "Which works explore the combination of neural networks with clustering tasks?", "answer": ["Deep Learning with Nonparametric Clustering", "Invariant Information Clustering for Unsupervised Image Classification and Segmentation"], "answer_arxiv_id": ["1501.03084", "1807.06653"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_19087"} +{"question": "Which research combines the strengths of Head Probing (HP) and Fine Tuning (FT) and states that HP-FT provides the best performance on both in-distribution and out-of-distribution cases?", "answer": ["Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution"], "answer_arxiv_id": ["2202.10054v1"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_19088"} +{"question": "Which models in computational pathology are introduced for zero-shot classification and segmentation tasks?", "answer": ["Towards a Visual-Language Foundation Model for Computational Pathology", "Visual Language Pretrained Multiple Instance Zero-Shot Transfer for\n Histopathology Images"], "answer_arxiv_id": ["2307.12914", "2306.07831"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_19089"} +{"question": "Could you provide me some studies which incorporated cross-view learning approaches in semantic map learning?", "answer": ["Lift, Splat, Shoot: Encoding Images from Arbitrary Camera Rigs by Implicitly Unprojecting to 3D", "Cross-view Semantic Segmentation for Sensing Surroundings", "Cross-view Transformers for real-time Map-view Semantic Segmentation", "DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries", "FUTR3D: A Unified Sensor Fusion Framework for 3D Detection"], "answer_arxiv_id": ["2008.05711", "1906.03560", "2205.02833", "2110.06922", "2203.10642"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_19090"} +{"question": "Which papers discuss the use of large language models to generate necessary context for commonsense questions?", "answer": ["Generated Knowledge Prompting for Commonsense Reasoning"], "answer_arxiv_id": ["2110.08387"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_19091"} +{"question": "What researches focus on recognition, compositionality, and localization in videos?", "answer": ["Look for the Change: Learning Object States and State-Modifying Actions\n from Untrimmed Web Videos", "Action Modifiers: Learning from Adverbs in Instructional Videos", "Learning Graph Embeddings for Compositional Zero-shot Learning", "Learning Temporal Dynamics from Cycles in Narrated Video", "StepFormer: Self-supervised Step Discovery and Localization in\n Instructional Videos"], "answer_arxiv_id": ["2203.11637", "1912.06617", "2102.01987", "2101.02337", "2304.13265"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_19092"} +{"question": "What studies use additional images to mitigate the modality gap in zero-shot transfer?", "answer": ["SuS-X: Training-Free Name-Only Transfer of Vision-Language Models"], "answer_arxiv_id": ["2211.16198"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_19093"} +{"question": "Which studies attempted to improve the performance of variational autoencoders by employing flexible priors such as mixture distributions?", "answer": ["VAE with a VampPrior", "Resampled Priors for Variational Autoencoders"], "answer_arxiv_id": ["1705.07120v5", "1810.11428v2"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_19094"} +{"question": "Which research papers discuss the use of generative models for tabular problems?", "answer": ["Modeling Tabular Data using Conditional GAN", "Conditional Wasserstein GAN-based Oversampling of Tabular Data for Imbalanced Learning", "Relational Data Synthesis using Generative Adversarial Networks: A Design Space Exploration", "Differentially Private Synthetic Medical Data Generation using Convolutional GANs", "CTAB-GAN: Effective Table Data Synthesizing", "OCT-GAN: Neural ODE-based Conditional Tabular GANs", "Generative Trees: Adversarial and Copycat"], "answer_arxiv_id": ["1907.00503", "2008.09202", "2008.12763", "2012.11774", "2102.08369", "2105.14969", "2201.11205"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19095"} +{"question": "Are there any works about general-purpose methods in classification scoring?", "answer": ["Estimating Example Difficulty using Variance of Gradients", "Deep Learning on a Data Diet: Finding Important Examples Early in Training", "Metadata Archaeology: Unearthing Data Subsets by Leveraging Training Dynamics"], "answer_arxiv_id": ["2008.11600", "2107.07075", "2209.10015"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_19096"} +{"question": "Any works about using specific data corpora in training of VLMs?", "answer": ["PaLI-X: On Scaling up a Multilingual Vision and Language Model", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "PaLM-E: An Embodied Multimodal Language Model", "Visual Instruction Tuning", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models", "Kosmos-2: Grounding Multimodal Large Language Models to the World", "Improved Baselines with Visual Instruction Tuning"], "answer_arxiv_id": ["2305.18565", "2201.12086", "2303.03378", "2304.08485", "2306.05424", "2306.14824", "2310.03744"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_19097"} +{"question": "Can you provide me a paper that discussed using an ODE approach to model hazard and cumulative hazard functions?", "answer": ["SODEN: A Scalable Continuous-Time Survival Model through Ordinary Differential Equation Networks"], "answer_arxiv_id": ["2008.08637"], "source_meta": {"published_time": "20230318"}, "qid": "AutoScholarQuery_train_19098"} +{"question": "What are some early optimizers in the Adam family, based on the available literature?", "answer": ["ADADELTA: An Adaptive Learning Rate Method"], "answer_arxiv_id": ["1212.5701"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_19099"} +{"question": "Which work discusses the application of contrastive learning and manifold learning in self-supervised learning?", "answer": ["Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods"], "answer_arxiv_id": ["2205.11508"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_19100"} +{"question": "Are there any studies that combined Local SGD with gradient compression?", "answer": ["Qsparse-local-SGD: Distributed SGD with Quantization, Sparsification, and Local Computations"], "answer_arxiv_id": ["1906.02367"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19101"} +{"question": "What papers proposed methods for mitigating bias where the model exploits the identified bias?", "answer": ["Environment Inference for Invariant Learning"], "answer_arxiv_id": ["2010.07249"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_19102"} +{"question": "Which study uses a similar approach of using models to turn plain text into a specific sequence type(i.e., dialogue sequences)?", "answer": ["Dialog Inpainting: Turning Documents into Dialogs"], "answer_arxiv_id": ["2205.09073"], "source_meta": {"published_time": "20220824"}, "qid": "AutoScholarQuery_train_19103"} +{"question": "Could you provide some work in mathematical reasoning domain that developed an interpretable solver?", "answer": ["Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoning"], "answer_arxiv_id": ["2105.04165"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_19104"} +{"question": "In recent developments, which studies present the prompt-based semantic segmentation techniques? For example, systems like CLIPSeg and Segment Anything Model (SAM).", "answer": ["Image Segmentation Using Text and Image Prompts", "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2112.10003", "2401.14159"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19105"} +{"question": "Can you provide examples of studies that used memory network in computer vision, specifically in video representation learning, one-shot learning, and text-to-image synthesis?", "answer": ["Memory-augmented Dense Predictive Coding for Video Representation Learning", "Learning to Remember Rare Events", "Memory Matching Networks for One-Shot Image Recognition", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image Synthesis"], "answer_arxiv_id": ["2008.01065", "1703.03129", "1804.08281v1", "1904.01310"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19106"} +{"question": "What works re-popularised the linearisation step by showing that it improves the quality of uncertainty estimates?", "answer": ["Approximate Inference Turns Deep Networks into Gaussian Processes", "Improving predictions of Bayesian neural nets via local linearization"], "answer_arxiv_id": ["1906.01930", "2008.08400"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_19107"} +{"question": "In knowledge distillation, which works demonstrate the student model mimicking the teacher's intermediate features?", "answer": ["FitNets: Hints for Thin Deep Nets"], "answer_arxiv_id": ["1412.6550"], "source_meta": {"published_time": "20240308"}, "qid": "AutoScholarQuery_train_19108"} +{"question": "Who proposed continuous prompts to reduce prompt engineering in prompt-based learning?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning", "Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2104.08691", "2101.00190"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_19109"} +{"question": "Any works about the issues in existing evaluation methods for these agents?", "answer": ["Is GPT-4 a Good Data Analyst?", "AlpacaFarm: A Simulation Framework for Methods that Learn from Human\n Feedback", "LIDA: A Tool for Automatic Generation of Grammar-Agnostic Visualizations\n and Infographics using Large Language Models"], "answer_arxiv_id": ["2305.15038", "2305.14387", "2303.02927"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19110"} +{"question": "What research used an explicit voxel grid to model temporal information for dynamic scenes?", "answer": ["Fast Dynamic Radiance Fields with Time-Aware Neural Voxels"], "answer_arxiv_id": ["2205.15285"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_19111"} +{"question": "Which work introduced Softmax Splatting for forward-warping-based VFI?", "answer": ["Softmax Splatting for Video Frame Interpolation"], "answer_arxiv_id": ["2003.05534v1"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_19112"} +{"question": "What works presented the concept of retrieval-augmented generation (RAG)?", "answer": ["Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["2005.11401"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_19113"} +{"question": "Any research papers suggesting the application of gating in addition to the improvements in nODEs?", "answer": ["Learning differential equations that are easy to solve", "How to Train Your Neural ODE: the World of Jacobian and Kinetic Regularization", "Opening the Blackbox: Accelerating Neural Differential Equations by Regularizing Internal Solver Heuristics"], "answer_arxiv_id": ["2007.04504", "2002.02798", "2105.03918"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_19114"} +{"question": "In what research was the gyrovector space framework proposed to handle addition and multiplication issue in hyperbolic spaces? ", "answer": ["Hyperbolic Neural Networks++", "Hyperbolic Neural Networks++", "Continuous Hierarchical Representations with Poincaré Variational Auto-Encoders", "Hyperbolic Graph Attention Network", "Graph Geometry Interaction Learning"], "answer_arxiv_id": ["2006.08210", "2006.08210", "1901.06033", "1912.03046", "2010.12135"], "source_meta": {"published_time": "20220214"}, "qid": "AutoScholarQuery_train_19115"} +{"question": "What research papers pioneered in diffusion-based motion generation?", "answer": ["Human Motion Diffusion Model"], "answer_arxiv_id": ["2209.14916"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_19116"} +{"question": "What research was the proposition of CSRO based on?", "answer": ["Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables"], "answer_arxiv_id": ["1903.08254"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_19117"} +{"question": "Which studies argue the rise of contrastive learning as a mainstream framework for pretraining vision models?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning", "An Empirical Study of Training Self-Supervised Vision Transformers"], "answer_arxiv_id": ["2002.05709", "1911.05722", "2104.02057"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19118"} +{"question": "What works can be prominent examples of slot-centric generative models?", "answer": ["MONet: Unsupervised Scene Decomposition and Representation", "Genesis: Generative Scene Inference and Sampling with Object-Centric Latent Representations", "Multi-Object Representation Learning with Iterative Variational Inference", "Object-Centric Learning with Slot Attention"], "answer_arxiv_id": ["1901.11390", "1907.13052", "1903.00450", "2006.15055"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_19119"} +{"question": "What studies made significant steps in estimating the 6-DoF part poses of articulated objects?", "answer": ["Category-Level Articulated Object Pose Estimation", "Nothing But Geometric Constraints: A Model-Free Method for Articulated Object Pose Estimation", "CAPTRA: CAtegory-level Pose Tracking for Rigid and Articulated Objects from Point Clouds", "Self-Supervised Category-Level Articulated Object Pose Estimation with Part-Level SE(3) Equivariance", "GAPartNet: Cross-Category Domain-Generalizable Object Perception and Manipulation via Generalizable and Actionable Parts"], "answer_arxiv_id": ["1912.11913", "2012.00088", "2104.03437", "2302.14268", "2211.05272"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_19120"} +{"question": "Which literature reviews model-based IRL methods?", "answer": ["Continuous Inverse Optimal Control with Locally Optimal Examples"], "answer_arxiv_id": ["1206.4617v1"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_19121"} +{"question": "Which study talks about the inherent nature of the reverse diffusion process?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_19122"} +{"question": "Which studies tackled unsupervised learning to localize sound sources in video?", "answer": ["Localizing Visual Sounds the Hard Way"], "answer_arxiv_id": ["2104.02691"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_19123"} +{"question": "Which papers delve into the potential of Large Language Models (LLMs) for complex reasoning tasks?", "answer": ["MAP-Neo: Highly Capable and Transparent Bilingual Large Language Model Series", "E^2-LLM: Efficient and Extreme Length Extension of Large Language Models", "Chinese Tiny LLM: Pretraining a Chinese-Centric Large Language Model"], "answer_arxiv_id": ["2405.19327v4", "2401.06951", "2404.04167v5"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_19124"} +{"question": "Which papers focus on using deep reinforcement learning for solving dexterous manipulation tasks?", "answer": ["Data-efficient Deep Reinforcement Learning for Dexterous Manipulation", "Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations", "RRL: Resnet as representation for Reinforcement Learning", "VRL3: A Data-Driven Framework for Visual Deep Reinforcement Learning", "MoDem: Accelerating Visual Model-Based Reinforcement Learning with Demonstrations"], "answer_arxiv_id": ["1704.03073", "1709.10087", "2107.03380", "2202.10324", "2212.05698"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_19125"} +{"question": "What works gather annotations in the form of spatial heat maps for VG-Methods?", "answer": ["Human Attention in Visual Question Answering: Do Humans and Deep\n Networks Look at the Same Regions?"], "answer_arxiv_id": ["1606.03556"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_19126"} +{"question": "What works have extended the pre-trained T2I model to integrate layout information into generation and control instances’ position?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Training-Free Layout Control with Cross-Attention Guidance", "BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained\n Diffusion", "LayoutDiffusion: Controllable Diffusion Model for Layout-to-image\n Generation", "Guided Image Synthesis via Initial Image Editing in Diffusion Model"], "answer_arxiv_id": ["2112.10741", "2304.03373", "2307.10816", "2303.17189", "2305.03382"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_19127"} +{"question": "Which research papers address the problem of skeleton-based human action recognition?", "answer": ["Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action\n Recognition", "Two-Stream Adaptive Graph Convolutional Networks for Skeleton-Based\n Action Recognition", "Disentangling and Unifying Graph Convolutions for Skeleton-Based Action\n Recognition", "Semantics-Guided Neural Networks for Efficient Skeleton-Based Human\n Action Recognition", "Channel-wise Topology Refinement Graph Convolution for Skeleton-Based\n Action Recognition", "Learning Discriminative Representations for Skeleton Based Action\n Recognition", "Hierarchically Decomposed Graph Convolutional Networks for\n Skeleton-Based Action Recognition", "Hierarchically Decomposed Graph Convolutional Networks for\n Skeleton-Based Action Recognition"], "answer_arxiv_id": ["1801.07455", "1805.07694", "2003.14111", "1904.01189", "2107.12213", "2303.03729", "2208.10741", "2208.10741"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_19128"} +{"question": "What paper used ICL as a framework for Automated Machine Learning (AutoML)?", "answer": ["TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second"], "answer_arxiv_id": ["2207.01848"], "source_meta": {"published_time": "20230117"}, "qid": "AutoScholarQuery_train_19129"} +{"question": "Which research works create human body avatars with forward deformation and cage-based deformation?", "answer": ["Animatable and Relightable Gaussians for High-fidelity Human Avatar\n Modeling", "GART: Gaussian Articulated Template Models", "Animatable 3D Gaussians for High-fidelity Synthesis of Human Motions", "SplatArmor: Articulated Gaussian splatting for animatable humans from\n monocular RGB videos", "HUGS: Human Gaussian Splats", "Drivable 3D Gaussian Avatars"], "answer_arxiv_id": ["2311.16096", "2311.16099", "2311.13404v2", "2311.10812", "2311.17910", "2311.08581"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_19130"} +{"question": "What is the research proposing that downstream tasks can be jointly learned by using the pretrained LM head to generate answers in natural language?", "answer": ["Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension"], "answer_arxiv_id": ["1910.10683", "1910.13461"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19131"} +{"question": "Which works discuss the benefits of machine learning models in numerical simulations?", "answer": ["Deep Dynamical Modeling and Control of Unsteady Fluid Flows", "Learning Mesh-Based Simulation with Graph Networks"], "answer_arxiv_id": ["1805.07472", "2010.03409"], "source_meta": {"published_time": "20230124"}, "qid": "AutoScholarQuery_train_19132"} +{"question": "Which datasets perform language-grounding tasks in the field of autonomous driving?", "answer": ["Object Referring in Videos with Language and Human Gaze", "Talk2Car: Taking Control of Your Self-Driving Car"], "answer_arxiv_id": ["1801.01582", "1909.10838"], "source_meta": {"published_time": "20240102"}, "qid": "AutoScholarQuery_train_19133"} +{"question": "Which datasets are comprised of images and bounding box annotations of invasive species?", "answer": ["Pink-Eggs Dataset V1: A Step Toward Invasive Species Management Using Deep Learning Embedded Solutions", "The Ciona17 Dataset for Semantic Segmentation of Invasive Species in a Marine Aquaculture Environment"], "answer_arxiv_id": ["2305.09302", "1702.05564"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_19134"} +{"question": "What studies can be classified under static correction methods in variance reduction?", "answer": ["SCAFFOLD: Stochastic Controlled Averaging for Federated Learning", "From Deterioration to Acceleration: A Calibration Approach to Rehabilitating Step Asynchronism in Federated Optimization", "Variance Reduced Local SGD with Lower Communication Complexity", "Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning", "Linear Convergence in Federated Learning: Tackling Client Heterogeneity and Sparse Gradients"], "answer_arxiv_id": ["1910.06378", "2112.09355", "1912.12844", "2008.03606", "2102.07053"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_19135"} +{"question": "Which studies showcase the application of diffusion models in text-to-image generation?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2112.10741", "2204.06125", "2112.10752", "2205.11487"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_19136"} +{"question": "What works implemented instruction tuning across a wide range of human-crafted instructions and task categories to enhance the understanding of LLMs?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_19137"} +{"question": "Which works are about architectural improvements beyond the conventional feed-forward style, such as FiLM or Ada-In layers?", "answer": ["FiLM: Visual Reasoning with a General Conditioning Layer", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["1709.07871", "1812.04948"], "source_meta": {"published_time": "20200614"}, "qid": "AutoScholarQuery_train_19138"} +{"question": "Which works utilized IoU-like loss functions for training oriented object detectors?", "answer": ["SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated\n Objects", "Rethinking Rotated Object Detection with Gaussian Wasserstein Distance\n Loss", "Learning High-Precision Bounding Box for Rotated Object Detection via\n Kullback-Leibler Divergence", "The KFIoU Loss for Rotated Object Detection", "IoU Loss for 2D/3D Object Detection"], "answer_arxiv_id": ["1811.07126", "2101.11952", "2106.01883", "2201.12558", "1908.03851"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19139"} +{"question": "Could you provide me some studies that improved the estimation of 3D scene graphs for large-scale environments?", "answer": ["3D Dynamic Scene Graphs: Actionable Spatial Perception with Places,\n Objects, and Humans", "Hydra: A Real-time Spatial Perception System for 3D Scene Graph\n Construction and Optimization", "Kimera: from SLAM to Spatial Perception with 3D Dynamic Scene Graphs"], "answer_arxiv_id": ["2002.06289", "2201.13360", "2101.06894"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_19140"} +{"question": "Which works discuss the unsuccessful application of RL algorithms like DDPG or PPO in agent training?", "answer": ["Continuous control with deep reinforcement learning", "Proximal Policy Optimization Algorithms", "Rethinking Action Spaces for Reinforcement Learning in End-to-end Dialog Agents with Latent Variable Models", "A Survey on Recent Advances and Challenges in Reinforcement Learning Methods for Task-Oriented Dialogue Policy Learning"], "answer_arxiv_id": ["1509.02971", "1707.06347v2", "1902.08858", "2202.13675"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_19141"} +{"question": "What recent work extended the approach of learning classes of polynomials to utilize three-layer networks?", "answer": ["Towards Understanding Hierarchical Learning: Benefits of Neural Representations", "Learning and Generalization in Overparameterized Neural Networks, Going Beyond Two Layers"], "answer_arxiv_id": ["2006.13436", "1811.04918"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_19142"} +{"question": "What papers utilize SGD with momentum and adaptive SGD to design the new global optimizer for FL?", "answer": ["On the Linear Speedup Analysis of Communication Efficient Momentum SGD for Distributed Non-Convex Optimization", "Distributed Stochastic Optimization via Adaptive SGD"], "answer_arxiv_id": ["1905.03817", "1802.05811"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_19143"} +{"question": "Which papers proposed the method to maximize the dimensionality of the latent manifold by standardizing the covariance matrix of the representations in the mini-batch?", "answer": ["Decorrelated Batch Normalization", "Optimal Whitening and Decorrelation"], "answer_arxiv_id": ["1804.08450", "1512.00809"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_19144"} +{"question": "What works demonstrate that LLMs have reasoning abilities and exhibit emergent behaviours when they are large enough?", "answer": ["Emergent Abilities of Large Language Models", "Are Emergent Abilities of Large Language Models a Mirage?"], "answer_arxiv_id": ["2206.07682", "2304.15004"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_19145"} +{"question": "In what papers do they use LLMs for code generation in robotic planning?", "answer": ["Voyager: An Open-Ended Embodied Agent with Large Language Models", "Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language", "Code as Policies: Language Model Programs for Embodied Control"], "answer_arxiv_id": ["2305.16291", "2204.00598", "2209.07753"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19146"} +{"question": "Which works provide convergence analyses for stochastic Self-accelerated Method (SAM) and its variants under smooth nonconvex functions?", "answer": ["Surrogate Gap Minimization Improves Sharpness-Aware Training", "Make Sharpness-Aware Minimization Stronger: A Sparsified Perturbation Approach", "An Adaptive Policy to Employ Sharpness-Aware Minimization", "AdaSAM: Boosting Sharpness-Aware Minimization with Adaptive Learning Rate and Momentum for Training Deep Neural Networks"], "answer_arxiv_id": ["2203.08065", "2210.05177", "2304.14647", "2303.00565"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_19147"} +{"question": "Could you provide me some studies about different definitions of memorization in deep neural networks?", "answer": ["The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks", "Extracting Training Data from Large Language Models", "What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation", "Counterfactual Memorization in Neural Language Models"], "answer_arxiv_id": ["1802.08232", "2012.07805", "2008.03703", "2112.12938"], "source_meta": {"published_time": "20220215"}, "qid": "AutoScholarQuery_train_19148"} +{"question": "What works propose the conversion of the aggregation process of GNNs as a learnable function?", "answer": ["How Powerful are Graph Neural Networks?"], "answer_arxiv_id": ["1810.00826"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_19149"} +{"question": "What papers worked on task-specific approaches for identifying reasoning paths by constructing semantic graphs?", "answer": ["Exploiting Reasoning Chains for Multi-hop Science Question Answering"], "answer_arxiv_id": ["2109.02905v1"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_19150"} +{"question": "Which papers proposed representation-based anomaly detection methods in text-based anomaly detection?", "answer": ["FastFlow: Unsupervised Anomaly Detection and Localization via 2D\n Normalizing Flows", "CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via\n Conditional Normalizing Flows"], "answer_arxiv_id": ["2111.07677", "2107.12571"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_19151"} +{"question": "Which studies in an unsupervised setting extend the positive sampling procedure to other views of different instances that are close to the anchor in the latent space?", "answer": ["With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations", "Weakly Supervised Contrastive Learning", "ReSSL: Relational Self-Supervised Learning with Weak Augmentation", "Prototypical Contrastive Learning of Unsupervised Representations"], "answer_arxiv_id": ["2104.14548", "2110.04770", "2107.09282", "2005.04966"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_19152"} +{"question": "Who explored the linear contextual bandits with adversarial corruptions?", "answer": ["Nearly Optimal Algorithms for Linear Contextual Bandits with Adversarial Corruptions"], "answer_arxiv_id": ["2205.06811"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_19153"} +{"question": "Which papers established and described MAP-Elites?", "answer": ["Robots that can adapt like animals", "Illuminating search spaces by mapping elites"], "answer_arxiv_id": ["1407.3501", "1504.04909"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_19154"} +{"question": "Are there any works that use the angular point spread function to reduce the glow effect in nighttime scenes for NiSID?", "answer": ["Enhancing Visibility in Nighttime Haze Images Using Guided APSF and\n Gradient Adaptive Convolution"], "answer_arxiv_id": ["2308.01738"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19155"} +{"question": "Could you give me a list of studies that proposed learning improvement heuristics for VRPs?", "answer": ["Learning to Perform Local Rewriting for Combinatorial Optimization", "Neural Large Neighborhood Search for the Capacitated Vehicle Routing Problem", "Learning 2-opt Heuristics for the Traveling Salesman Problem via Deep Reinforcement Learning", "Learning Improvement Heuristics for Solving Routing Problems", "A Bi-Level Framework for Learning to Solve Combinatorial Optimization on Graphs", "Learning to Iteratively Solve Routing Problems with Dual-Aspect Collaborative Transformer", "NeuroLKH: Combining Deep Learning Model with Lin-Kernighan-Helsgaun Heuristic for Solving the Traveling Salesman Problem", "Learning Collaborative Policies to Solve NP-hard Routing Problems", "Graph Neural Network Guided Local Search for the Traveling Salesperson Problem"], "answer_arxiv_id": ["1810.00337", "1911.09539", "2004.01608", "1912.05784", "2106.04927", "2110.02544", "2110.07983", "2110.13987", "2110.05291"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19156"} +{"question": "What research aims to find shared and private representations using VAE but doesn't provide a way to quantify the information content of the private and shared components?", "answer": ["Deep Variational Canonical Correlation Analysis"], "answer_arxiv_id": ["1610.03454v3"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_19157"} +{"question": "What papers introduced regularization methods based on the fact that most of the space in a scene is empty?", "answer": ["Baking Neural Radiance Fields for Real-Time View Synthesis", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Plenoxels: Radiance Fields without Neural Networks"], "answer_arxiv_id": ["2103.14645", "2103.14024", "2112.05131"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_19158"} +{"question": "What studies have had progress with fine-tuning text-to-image diffusion models for customization on certain objects?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.01618", "2208.12242"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19159"} +{"question": "What papers are there on logical-rule-based methods for link prediction on tKGs which are mainly based on random walks?", "answer": ["dynnode2vec: Scalable Dynamic Network Embedding", "TLogic: Temporal Logical Rules for Explainable Link Forecasting on Temporal Knowledge Graphs"], "answer_arxiv_id": ["1812.02356", "2112.08025"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_19160"} +{"question": "Any works about training self-supervised representation learning models by predicting unseen speech signals?", "answer": ["An Unsupervised Autoregressive Model for Speech Representation Learning"], "answer_arxiv_id": ["1904.03240"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_19161"} +{"question": "Which paper demonstrated that image artifacts could be detected by leveraging the degree of partial decay at high frequencies?", "answer": ["Fourier Spectrum Discrepancies in Deep Network Generated Images"], "answer_arxiv_id": ["1911.06465"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_19162"} +{"question": "Which paper was the first npf model and uses simple mlp layers as its encoder and decoder?", "answer": ["Conditional Neural Processes"], "answer_arxiv_id": ["1807.01613"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_19163"} +{"question": "Which works discuss video recognition methods that learn from RGB coupled with other modalities but comparatively fewer use IMU?", "answer": ["Vision and Inertial Sensing Fusion for Human Action Recognition : A Review"], "answer_arxiv_id": ["2008.00380v1"], "source_meta": {"published_time": "20230105"}, "qid": "AutoScholarQuery_train_19164"} +{"question": "Any works about the use of hierarchical Transformer or cross-attention Transformer?", "answer": ["InterBERT: Vision-and-Language Interaction for Multi-modal Pretraining", "Improved Fusion of Visual and Language Representations by Dense\n Symmetric Co-Attention for Visual Question Answering", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks", "Large-scale Pretraining for Visual Dialog: A Simple State-of-the-Art\n Baseline"], "answer_arxiv_id": ["2003.13198", "1804.00775", "1908.02265", "1912.02379"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_19165"} +{"question": "What studies implemented recursive least-squares applied to deep networks in a layer-wise manner?", "answer": ["Continual Learning of Context-dependent Processing in Neural Networks"], "answer_arxiv_id": ["1810.01256"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_19166"} +{"question": "What previous works provided 2D laneline annotations for the lane detection task?", "answer": ["Real time Detection of Lane Markers in Urban Streets", "VPGNet: Vanishing Point Guided Network for Lane and Road Marking Detection and Recognition", "CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending"], "answer_arxiv_id": ["1411.7113v1", "1710.06288", "2007.12147"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_19167"} +{"question": "Are there any research studies that examine emergent outliers in large language model quantization?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "GLM-130B: An Open Bilingual Pre-trained Model", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models"], "answer_arxiv_id": ["2208.07339", "2210.02414", "2211.10438"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_19168"} +{"question": "Could you name a benchmark that was designed to separate covariate and concept shifts?", "answer": ["GOOD: A Graph Out-of-Distribution Benchmark"], "answer_arxiv_id": ["2206.08452"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_19169"} +{"question": "Which research papers discussed the use of geometric tensors as node embeddings in equivariant models?", "answer": ["Clebsch–Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks"], "answer_arxiv_id": ["1806.09231", "2006.10503"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_19170"} +{"question": "Which work introduced the concept of Test-time Batch Normalization?", "answer": ["Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shift"], "answer_arxiv_id": ["2006.10963"], "source_meta": {"published_time": "20230129"}, "qid": "AutoScholarQuery_train_19171"} +{"question": "Which research proposed a unified ranking model for solving information alignment style tasks?", "answer": ["Text Alignment Is An Efficient Unified Model for Massive NLP Tasks"], "answer_arxiv_id": ["2307.02729"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_19172"} +{"question": "Who introduced Boltzmann generators in Cartesian coordinates for molecules, potentially enabling transferable training?", "answer": ["Equivariant flow matching", "SE(3) Equivariant Augmented Coupling Flows"], "answer_arxiv_id": ["2306.15030", "2308.10364"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19173"} +{"question": "Which work extracts K-nearest neighbor patches from the 3D point cloud and parameterizes them into 2D space using neural networks?", "answer": ["Learning Delaunay Surface Elements for Mesh Reconstruction"], "answer_arxiv_id": ["2012.01203"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_19174"} +{"question": "Which work proves the effectiveness of Optimal Transport (OT) empirically?", "answer": ["Joint Distribution Optimal Transportation for Domain Adaptation"], "answer_arxiv_id": ["1705.08848"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_19175"} +{"question": "What papers utilize MLPs in their research?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Block-NeRF: Scalable Large Scene Neural View Synthesis"], "answer_arxiv_id": ["2003.08934", "2202.05263"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_19176"} +{"question": "Which work leverages the radiance field representation and lift the estimated 2D labels to 3D through per-scene optimization?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_19177"} +{"question": "Could you provide me some research on MLP-based models for time series forecasting?", "answer": ["N-BEATS: Neural basis expansion analysis for interpretable time series forecasting", "Are Transformers Effective for Time Series Forecasting?", "Less Is More: Fast Multivariate Time Series Forecasting with Light Sampling-oriented MLP Structures"], "answer_arxiv_id": ["1905.10437", "2205.13504", "2207.01186"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19178"} +{"question": "Could you give me some examples of research attempts to fine-tune the CLIP model efficiently?", "answer": ["Learning to Prompt for Vision-Language Models", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling", "MaPLe: Multi-modal Prompt Learning", "Black Box Few-Shot Adaptation for Vision-Language models"], "answer_arxiv_id": ["2109.01134", "2111.03930", "2210.03117", "2304.01752"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_19179"} +{"question": "What studies have used entropy minimization objectives for Unsupervised Learning, Semi-Supervised Learning, or Domain Adaptation?", "answer": ["Unsupervised Domain Adaptation with Residual Transfer Networks", "ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation", "Entropy Minimization vs. Diversity Maximization for Domain Adaptation"], "answer_arxiv_id": ["1602.04433", "1811.12833", "2002.01690"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_19180"} +{"question": "What research works involve fine-tuning of the generative models for image editing?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation", "Imagic: Text-Based Real Image Editing with Diffusion Models"], "answer_arxiv_id": ["2208.12242", "2210.09276"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_19181"} +{"question": "Can you provide some studies that have used encoder-decoder methods to solve the scene flow by an hourglass architecture neural network?", "answer": ["HPLFlowNet: Hierarchical Permutohedral Lattice FlowNet for Scene Flow Estimation on Large-scale Point Clouds", "FlowNet3D: Learning Scene Flow in 3D Point Clouds"], "answer_arxiv_id": ["1906.05332", "1806.01411"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_19182"} +{"question": "Could you provide me some works studying parallel transfer learning in multi-agent system?", "answer": ["Parallel Knowledge Transfer in Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["2003.13085"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_19183"} +{"question": "Can you list some studies that dealt with the transfer of the implicit category structure to infer unknown categories?", "answer": ["A Unified Objective for Novel Class Discovery", "Unsupervised Deep Embedding for Clustering Analysis", "Automatically Discovering and Learning New Visual Categories with Ranking Statistics", "Novel Visual Category Discovery with Dual Ranking Statistics and Mutual Knowledge Distillation", "OpenMix: Reviving Known Knowledge for Discovering Novel Visual Categories in An Open World", "Novel Class Discovery without Forgetting", "Class-incremental Novel Class Discovery", "Towards Realistic Semi-Supervised Learning", "Novel Visual Category Discovery with Dual Ranking Statistics and Mutual Knowledge Distillation", "Modeling Inter-Class and Intra-Class Constraints in Novel Class Discovery", "Novel Class Discovery for 3D Point Cloud Semantic Segmentation", "Class-relation Knowledge Distillation for Novel Class Discovery", "When and How Does Known Class Help Discover Unknown Ones? Provable Understanding Through Spectral Analysis", "Novel Class Discovery for Long-tailed Recognition", "An Interactive Interface for Novel Class Discovery in Tabular Data", "Supervised Knowledge May Hurt Novel Class Discovery Performance", "Open-world Semi-supervised Novel Class Discovery", "Novel Class Discovery in Semantic Segmentation", "Meta Discovery: Learning to Discover Novel Classes given Very Limited Data", "Grow and Merge: A Unified Framework for Continuous Categories Discovery", "Spacing Loss for Discovering Novel Categories", "A Closer Look at Novel Class Discovery from the Labeled Set"], "answer_arxiv_id": ["2108.08536", "1511.06335", "2002.05714", "2107.03358v2", "2004.05551", "2207.10659", "2207.08605", "2207.02269", "2107.03358v2", "2210.03591", "2303.11610", "2307.09158", "2308.05017", "2308.02989", "2306.12919", "2306.03648v1", "2305.13095", "2112.01900", "2102.04002", "2210.04174", "2204.10595", "2209.09120v4"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_19184"} +{"question": "Which studies have estimated mesh-free camera pose by 'inverting' a NeRF?", "answer": ["INeRF: Inverting Neural Radiance Fields for Pose Estimation"], "answer_arxiv_id": ["2012.05877"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19185"} +{"question": "Which study demonstrates the heavy reliance of Convolutional Neural Networks (CNNs) on texture information for object recognition?", "answer": ["ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness"], "answer_arxiv_id": ["1811.12231"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_19186"} +{"question": "What studies used universal approximation theorems to understand the expressivity of MPNNs?", "answer": ["Expressiveness and Approximation Properties of Graph Neural Networks"], "answer_arxiv_id": ["2204.04661"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_19187"} +{"question": "Can you provide some examples of research utilizing meta-gradient methods for intrinsic reward learning?", "answer": ["On Learning Intrinsic Rewards for Policy Gradient Methods"], "answer_arxiv_id": ["1804.06459"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_19188"} +{"question": "Which works have proposed prototype-based metric learning method for image segmentation?", "answer": ["Weakly Supervised Semantic Segmentation by Pixel-to-Prototype Contrast", "PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment", "Part-aware Prototype Network for Few-shot Semantic Segmentation", "Semi-supervised Semantic Segmentation with Prototype-based Consistency\n Regularization", "Rethinking Semantic Segmentation: A Prototype View", "Soft Neighbors are Positive Supporters in Contrastive Visual\n Representation Learning"], "answer_arxiv_id": ["2110.07110", "1908.06391", "2007.06309", "2210.04388", "2203.15102", "2303.17142"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_19189"} +{"question": "What papers apply attention mechanisms in natural language processing?", "answer": ["Effective Approaches to Attention-based Neural Machine Translation", "Attention Is All You Need", "BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding"], "answer_arxiv_id": ["1508.04025", "1706.03762", "1810.04805"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19190"} +{"question": "Which papers are related to the use of diffusion models in text-driven image generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2010.02502"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19191"} +{"question": "What works propose techniques that significantly boost 'few-shot prompting'?", "answer": ["Language Models are Few-Shot Learners", "PaLM: Scaling Language Modeling with Pathways", "Program Induction by Rationale Generation: Learning to Solve and Explain Algebraic Word Problems", "Training Verifiers to Solve Math Word Problems", "Show Your Work: Scratchpads for Intermediate Computation with Language Models", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them", "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models", "Compositional Semantic Parsing with Large Language Models", "Language Model Cascades", "Decomposed Prompting: A Modular Approach for Solving Complex Tasks", "Measuring and Narrowing the Compositionality Gap in Language Models", "Program Synthesis with Large Language Models", "PAL: Program-aided Language Models", "Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks", "Self-Consistency Improves Chain of Thought Reasoning in Language Models", "Natural Language to Code Translation with Execution", "Rationale-Augmented Ensembles in Language Models"], "answer_arxiv_id": ["2005.14165", "2204.02311", "1705.04146", "2110.14168", "2112.00114", "2201.11903", "2210.09261v1", "2205.10625", "2209.15003", "2207.10342", "2210.02406", "2210.03350", "2108.07732", "2211.10435", "2211.12588", "2203.11171", "2204.11454", "2207.00747"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19192"} +{"question": "What papers use learning strategies like meta-learning, gradient operation, and self-supervised learning to improve generalization performance?", "answer": ["Learning to Generalize: Meta-Learning for Domain Generalization", "Self-Challenging Improves Cross-Domain Generalization", "Domain Generalization by Solving Jigsaw Puzzles"], "answer_arxiv_id": ["1710.03463", "2007.02454", "1903.06864"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_19193"} +{"question": "What researches adopted gradient-based methods for explaining deep learning models?", "answer": ["Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps", "Learning Important Features Through Propagating Activation Differences", "Visualizing and Understanding Convolutional Networks", "Axiomatic Attribution for Deep Networks", "Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned"], "answer_arxiv_id": ["1312.6034", "1704.02685", "1311.2901", "1703.01365", "1905.09418"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_19194"} +{"question": "Could you mention some studies that proposed datasets covering rooms with simple cuboid layouts?", "answer": ["LayoutNet: Reconstructing the 3D Room Layout from a Single RGB Image", "LSUN: Construction of a Large-Scale Image Dataset using Deep Learning with Humans in the Loop"], "answer_arxiv_id": ["1803.08999", "1506.03365"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_19195"} +{"question": "Which papers provide examples of using regularization to turn the bilinear objective in normal-form games into a strongly-convex-strongly-concave one?", "answer": ["Fast Policy Extragradient Methods for Competitive Games with Entropy Regularization"], "answer_arxiv_id": ["2105.15186"], "source_meta": {"published_time": "20220619"}, "qid": "AutoScholarQuery_train_19196"} +{"question": "Which works have studied the reward poisoning in the general context of reinforcement learning?", "answer": ["Data Poisoning Attacks in Contextual Bandits", "Policy Poisoning in Batch Reinforcement Learning and Control", "Adaptive Reward-Poisoning Attacks against Reinforcement Learning"], "answer_arxiv_id": ["1808.05760", "1910.05821", "2003.12613"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_19197"} +{"question": "What studies can be pointed to that use visual self-supervision?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers", "Momentum Contrast for Unsupervised Visual Representation Learning", "Emerging Properties in Self-Supervised Vision Transformers", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2111.06377", "2106.08254", "1911.05722", "2104.14294", "2002.05709"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19198"} +{"question": "Which papers studied the implicit bias of specific optimization algorithms such as gradient descent?", "answer": ["Gradient Descent Finds Global Minima of Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization"], "answer_arxiv_id": ["1811.03804", "1811.03962"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_19199"} +{"question": "Could you list some studies that use the diffusion model approach for generating language-conditioned human motion?", "answer": ["Human Motion Diffusion Model", "MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model"], "answer_arxiv_id": ["2209.14916", "2208.15001"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_19200"} +{"question": "Which papers have demonstrated the ability of gradient-based methods to generate maximum-excited images?", "answer": ["NeuroGen: activation optimized image synthesis for discovery\n neuroscience", "Second Sight: Using brain-optimized encoding models to align image\n distributions with human brain activity"], "answer_arxiv_id": ["2105.07140", "2306.00927"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_19201"} +{"question": "Which studies used image editing techniques to explain modelvfailures?", "answer": ["LANCE: Stress-testing Visual Models by Generating Language-guided\n Counterfactual Images", "ImageNet-E: Benchmarking Neural Network Robustness via Attribute Editing"], "answer_arxiv_id": ["2305.19164", "2303.17096"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_19202"} +{"question": "What are some studies on cross-modal retrieval in multi-modal learning?", "answer": ["Fine-grained Video-Text Retrieval with Hierarchical Graph Reasoning", "Fisher Vectors Derived from Hybrid Gaussian-Laplacian Mixture Models for Image Annotation", "VSE++: Improving Visual-Semantic Embeddings with Hard Negatives", "Polysemous Visual-Semantic Embedding for Cross-Modal Retrieval", "Learning Transferable Visual Models From Natural Language Supervision", "CLIP2Video: Mastering Video-Text Retrieval via Image CLIP", "CenterCLIP: Token Clustering for Efficient Text-Video Retrieval", "X-Pool: Cross-Modal Language-Video Attention for Text-Video Retrieval", "Improving Cross-Modal Retrieval with Set of Diverse Embeddings"], "answer_arxiv_id": ["2003.00392", "1411.7399", "1707.05612", "1906.04402v2", "2103.00020", "2106.11097", "2205.00823", "2203.15086", "2211.16761"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_19203"} +{"question": "Which study extends the technique of generating similar heatmaps for vision transformers?", "answer": ["Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformers"], "answer_arxiv_id": ["2103.15679"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_19204"} +{"question": "Could you provide me some works on 3D scene reconstruction using representations such as (truncated) signed distance fields to combine multiple observations from RGB-D sensors?", "answer": ["iMAP: Implicit Mapping and Positioning in Real-Time"], "answer_arxiv_id": ["2103.12352"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19205"} +{"question": "What works discuss homogenization via confounding within recommendation systems?", "answer": ["How Algorithmic Confounding in Recommendation Systems Increases Homogeneity and Decreases Utility"], "answer_arxiv_id": ["1710.11214"], "source_meta": {"published_time": "20221124"}, "qid": "AutoScholarQuery_train_19206"} +{"question": "Could you provide me some works that move the focus from proprioceptive states to images or other more complex entities?", "answer": ["NeoRL: A Near Real-World Benchmark for Offline Reinforcement Learning", "An Optimistic Perspective on Offline Reinforcement Learning"], "answer_arxiv_id": ["2102.00714", "1907.04543v4"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_19207"} +{"question": "Which studies depict the usage of patches or stickers for carrying out adversarial attacks?", "answer": ["Certified Defenses for Adversarial Patches", "Dirty Road Can Attack: Security of Deep Learning based Automated Lane\n Centering under Physical-World Attack", "Robust Physical-World Attacks on Deep Learning Models"], "answer_arxiv_id": ["2003.06693", "2009.06701", "1707.08945"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_19208"} +{"question": "Which works have attempted to quantify the structure of data via quantum entanglement and mutual information?", "answer": ["Entanglement and tensor networks for supervised image classification", "Mutual Information Scaling for Tensor Network Machine Learning", "Tensor networks and efficient descriptions of classical data", "Information Perspective to Probabilistic Modeling: Boltzmann Machines versus Born Machines", "Entanglement Entropy of Target Functions for Image Classification and Convolutional Neural Network", "Entanglement Area Law for Shallow and Deep Quantum Neural Network States", "Entangled q-Convolutional Neural Nets", "Tensor network to learn the wavefunction of data."], "answer_arxiv_id": ["2007.06082", "2103.00105", "2103.06872", "1712.04144", "1710.05520", "1907.11333", "2103.11785", "2111.08014"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_19209"} +{"question": "Could you provide me works about constructing NeRFs from moving event cameras under various real-world conditions?", "answer": ["Robust e-NeRF: NeRF from Sparse & Noisy Events under Non-Uniform Motion"], "answer_arxiv_id": ["2309.08596"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_19210"} +{"question": "Which studies have shown remarkable performance on zero-shot image classification using VLP models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19211"} +{"question": "Which research studies compute the hypergradient based on implicit function theorem?", "answer": ["Optimizing Millions of Hyperparameters by Implicit Differentiation", "Scalable One-Pass Optimisation of High-Dimensional Weight-Update Hyperparameters by Implicit Differentiation", "Meta-Learning with Implicit Gradients"], "answer_arxiv_id": ["1911.02590", "2110.10461v3", "1909.04630"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_19212"} +{"question": "Which researches are about Private LSVI-UCB and Privacy-Preserving LSVI-UCB?", "answer": ["Improved Regret for Differentially Private Exploration in Linear MDP", "Differentially Private Exploration in Reinforcement Learning with Linear Representation"], "answer_arxiv_id": ["2202.01292", "2112.01585"], "source_meta": {"published_time": "20220602"}, "qid": "AutoScholarQuery_train_19213"} +{"question": "Which papers explore the mechanisms to produce more explicit equivariant representations of natural movies including changes in architecture, training diet and objectives?", "answer": ["Deep Predictive Coding Networks for Video Prediction and Unsupervised Learning", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["1605.08104", "1807.03748"], "source_meta": {"published_time": "20230826"}, "qid": "AutoScholarQuery_train_19214"} +{"question": "Could you provide me some works about learning numerical simulators according to data supervision?", "answer": ["Latent Space Physics: Towards Learning the Temporal Evolution of Fluid Flow", "Learning Mesh-Based Simulation with Graph Networks", "Learning to Simulate Complex Physics with Graph Networks", "Accelerating Eulerian Fluid Simulation With Convolutional Networks", "Data-Driven Synthesis of Smoke Flows with CNN-based Feature Descriptors", "tempoGAN: A Temporally Coherent, Volumetric GAN for Super-resolution Fluid Flow", "Lagrangian Neural Style Transfer for Fluids", "A Compositional Object-Based Approach to Learning Physical Dynamics", "Interaction Networks for Learning about Objects, Relations and Physics", "Graph Networks as Learnable Physics Engines for Inference and Control", "Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids"], "answer_arxiv_id": ["1802.10123", "2010.03409", "2002.09405", "1607.03597", "1705.01425", "1801.09710", "2005.00803", "1612.00341", "1612.00222", "1806.01242", "1810.01566"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_19215"} +{"question": "Any work introduced general un-projection model by using B-spline interpolation with nearest points?", "answer": ["Why Having 10,000 Parameters in Your Camera Model is Better Than Twelve"], "answer_arxiv_id": ["1912.02908"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_19216"} +{"question": "What are some works on subject-driven image generation that require test-time fintuning?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion", "SVDiff: Compact Parameter Space for Diffusion Fine-Tuning", "LoRA: Low-Rank Adaptation of Large Language Models", "Break-A-Scene: Extracting Multiple Concepts from a Single Image", "DisenBooth: Identity-Preserving Disentangled Tuning for Subject-Driven\n Text-to-Image Generation"], "answer_arxiv_id": ["2208.12242", "2208.01618", "2212.04488", "2303.11305", "2106.09685", "2305.16311", "2305.03374"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_19217"} +{"question": "Which papers laid the foundation for dynamic batching in dynamic neural networks?", "answer": ["DyNet: The Dynamic Neural Network Toolkit", "Deep Learning with Dynamic Computation Graphs"], "answer_arxiv_id": ["1701.03980", "1702.02181"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_19218"} +{"question": "What research uses orthogonal matrices to build directions?", "answer": ["Structured Evolution with Compact Architectures for Scalable Policy Optimization"], "answer_arxiv_id": ["1804.02395v2"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19219"} +{"question": "What papers demonstrated the use of gradient-based finetuning for adaptation in RL?", "answer": ["Experience-Embedded Visual Foresight", "RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning", "Self-Supervised Policy Adaptation during Deployment", "A Geometric Perspective on Self-Supervised Policy Adaptation", "VRL3: A Data-Driven Framework for Visual Deep Reinforcement Learning", "Visual Reinforcement Learning with Self-Supervised 3D Representations", "Lifelong Robotic Reinforcement Learning by Retaining Experiences", "Don’t Start From Scratch: Leveraging Prior Data to Automate Robotic Reinforcement Learning"], "answer_arxiv_id": ["1911.05071", "1611.02779", "2007.04309", "2011.07318", "2202.10324", "2210.07241", "2109.09180", "2207.04703"], "source_meta": {"published_time": "20221019"}, "qid": "AutoScholarQuery_train_19220"} +{"question": "Any studies about conducting Euclidean addition on graphons for structural mixup?", "answer": ["G-Mixup: Graph Data Augmentation for Graph Classification"], "answer_arxiv_id": ["2202.07179v2"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_19221"} +{"question": "Any studies about clean-label attacks that increase the stealthiness of attacks?", "answer": ["Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networks", "A New backdoor attack in CNNs by training set corruption without label poisoning"], "answer_arxiv_id": ["1804.00792", "1902.11237"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_19222"} +{"question": "What works have discussed exploiting generative models in self-supervised learning?", "answer": ["Generative Adversarial Nets", "Auto-Encoding Variational Bayes", "Large Scale Adversarial Representation Learning"], "answer_arxiv_id": ["1406.2661", "1312.6114", "1907.02544"], "source_meta": {"published_time": "20220601"}, "qid": "AutoScholarQuery_train_19223"} +{"question": "Which works represent video as 2D images?", "answer": ["Layered Neural Atlases for Consistent Video Editing", "Text2LIVE: Text-Driven Layered Image and Video Editing", "Shape-aware Text-driven Layered Video Editing", "CoDeF: Content Deformation Fields for Temporally Consistent Video\n Processing"], "answer_arxiv_id": ["2109.11418", "2204.02491", "2301.13173", "2308.07926"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_19224"} +{"question": "Which studies show that vision transformers have performed very well in Continual Learning?", "answer": ["Meta-attention for ViT-backed Continual Learning", "DyTox: Transformers for Continual Learning with DYnamic TOken eXpansion", "BiRT: Bio-inspired Replay in Vision Transformers for Continual Learning", "When Prompt-based Incremental Learning Does Not Meet Strong Pretraining", "Exemplar-Free Continual Transformer with Convolutions"], "answer_arxiv_id": ["2203.11684", "2111.11326", "2305.04769", "2308.10445", "2308.11357"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_19225"} +{"question": "Which work designed a selective watermark generation and detection approach for low-entropy code generation?", "answer": ["Who Wrote this Code? Watermarking for Code Generation"], "answer_arxiv_id": ["2305.15060"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_19226"} +{"question": "What works have investigated the capacity of transformers to learn target functions?", "answer": ["What Can Transformers Learn In-Context? A Case Study of Simple Function Classes"], "answer_arxiv_id": ["2208.01066"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19227"} +{"question": "Could you provide me with studies that used 2D lifting methods in Text-to-3D generation?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with\n Variational Score Distillation"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2211.07600", "2303.13873", "2305.16213"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_19228"} +{"question": "Which work improved performance by training the generation and the reranking models separately on non-overlapping halves of the fine-tuning data?", "answer": ["SummaReranker: A Multi-Task Mixture-of-Experts Re-ranking Framework for Abstractive Summarization"], "answer_arxiv_id": ["2203.06569"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19229"} +{"question": "Which studies proposed dual-encoder approach in image-text matching?", "answer": ["Unifying Visual-Semantic Embeddings with Multimodal Neural Language\n Models", "Learning Deep Structure-Preserving Image-Text Embeddings", "VSE++: Improving Visual-Semantic Embeddings with Hard Negatives", "Visual Semantic Reasoning for Image-Text Matching"], "answer_arxiv_id": ["1411.2539", "1511.06078", "1707.05612", "1909.02701"], "source_meta": {"published_time": "20240617"}, "qid": "AutoScholarQuery_train_19230"} +{"question": "Which studies have combined the functional maps framework with extrinsic features in the area of shape understanding?", "answer": ["Continuous and Orientation-preserving Correspondences via Functional Maps", "Complex Functional Maps : a Conformal Link Between Tangent Bundles", "Spatially and Spectrally Consistent Deep Functional Maps"], "answer_arxiv_id": ["1806.04455v3", "2112.09546", "2308.08871"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_19231"} +{"question": "Which studies focus on Deep Kernel Processes that optimize a variational approximate posterior over Gram matrices?", "answer": ["Deep kernel processes", "A variational approximate posterior for the deep Wishart process", "An Improved Variational Approximate Posterior for the Deep Wishart Process"], "answer_arxiv_id": ["2010.01590", "2107.10125", "2305.14454"], "source_meta": {"published_time": "20210830"}, "qid": "AutoScholarQuery_train_19232"} +{"question": "What research paper introduced an adaptive weighting scheme from the optimization perspective by approximately minimizing the variance of the MSBE loss?", "answer": ["Sample-Efficient Reinforcement Learning with Stochastic Ensemble Value Expansion"], "answer_arxiv_id": ["1807.01675"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_19233"} +{"question": "What work modified the MIWAE bound for the MNAR scenario?", "answer": ["not-MIWAE: Deep Generative Modelling with Missing not at Random Data"], "answer_arxiv_id": ["2006.12871"], "source_meta": {"published_time": "20230813"}, "qid": "AutoScholarQuery_train_19234"} +{"question": "What research work has made contributions towards the empirical understanding of how robustness is transferred from pretraining to the downstream task?", "answer": ["Adversarially robust transfer learning", "Does Robustness on ImageNet Transfer to Downstream Tasks?", "ImageNet Pre-training also Transfers Non-robustness"], "answer_arxiv_id": ["1905.08232", "2204.03934", "2106.10989"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_19235"} +{"question": "What papers introduced new methods to benchmark a fine-tuned dialogue system?", "answer": ["Bootstrapping LLM-based Task-Oriented Dialogue Agents via Self-Talk"], "answer_arxiv_id": ["2401.05033"], "source_meta": {"published_time": "20240517"}, "qid": "AutoScholarQuery_train_19236"} +{"question": "Any works that discuss the challenges of meta-reinforcement learning methods in scenarios requiring complex exploration or when lacking task distribution structure?", "answer": ["Distributionally Adaptive Meta Reinforcement Learning", "On the Effectiveness of Fine-tuning Versus Meta-reinforcement Learning"], "answer_arxiv_id": ["2210.03104", "2206.03271"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_19237"} +{"question": "Which studies proposed using the GOAL method and the CIRCLE dataset for 3D motion sequences?", "answer": ["GOAL: Generating 4D Whole-Body Motion for Hand-Object Grasping", "CIRCLE: Capture In Rich Contextual Environments"], "answer_arxiv_id": ["2112.11454", "2303.17912"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_19238"} +{"question": "Could you provide me some works where specialized architectures based on pretrained language models are proposed for data from big e-commerce platforms?", "answer": ["Controllable and Diverse Text Generation in E-commerce"], "answer_arxiv_id": ["2102.11497"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_19239"} +{"question": "What study inspired the focus on enhancing LLMs’s performance on search-related tasks?", "answer": ["Finetuned Language Models Are Zero-Shot Learners"], "answer_arxiv_id": ["2109.01652"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_19240"} +{"question": "In which works, the exploration noise is parametrized by the current state in the unstructured exploration?", "answer": ["Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor"], "answer_arxiv_id": ["1801.01290"], "source_meta": {"published_time": "20220831"}, "qid": "AutoScholarQuery_train_19241"} +{"question": "What are some examples of recent model editing benchmarks that provide counterfactuals?", "answer": ["Locating and Editing Factual Associations in GPT", "MQuAKE: Assessing Knowledge Editing in Language Models via Multi-Hop\n Questions", "Evaluating the Ripple Effects of Knowledge Editing in Language Models"], "answer_arxiv_id": ["2202.05262", "2305.14795", "2307.12976"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19242"} +{"question": "What research provides a description of the application of a GAN to generate synthetic samples from past tasks as an alternative to storing actual data?", "answer": ["Data-Free Learning of Student Networks"], "answer_arxiv_id": ["1904.01186"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_19243"} +{"question": "Which works introduced an iterative training scheme for joint training of CT and X-ray data in unpaired multi-modal SSL?", "answer": ["UniMiSS: Universal Medical Self-Supervised Learning via Breaking\n Dimensionality Barrier"], "answer_arxiv_id": ["2112.09356"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19244"} +{"question": "What is the research that develops the understanding of synthetic images and texts relation?", "answer": ["Benchmarking Spatial Relationships in Text-to-Image Generation", "Do DALL-E and Flamingo Understand Each Other?", "Breaking Common Sense: WHOOPS! A Vision-and-Language Benchmark of Synthetic and Compositional Images", "Generated Faces in the Wild: Quantitative Comparison of Stable Diffusion, Midjourney and DALL-E 2", "Evaluating a Synthetic Image Dataset Generated with Stable Diffusion"], "answer_arxiv_id": ["2212.10015v3", "2212.12249", "2303.07274", "2210.00586", "2211.01777"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19245"} +{"question": "Any work about using recent states for planning in the context of adaptation?", "answer": ["Towards Evaluating Adaptivity of Model-Based Reinforcement Learning Methods"], "answer_arxiv_id": ["2204.11464"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_19246"} +{"question": "Could you provide me some studies about text-planning for improving output quality and reducing hallucinations?", "answer": ["Bottom-Up Abstractive Summarization", "Planning with Learned Entity Prompts for Abstractive Summarization", "Conditional Generation with a Question-Answering Blueprint"], "answer_arxiv_id": ["1808.10792", "2104.07606", "2207.00397"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_19247"} +{"question": "What papers employ gating mechanisms to dynamically accelerate both training and inference?", "answer": ["DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification"], "answer_arxiv_id": ["2106.02034v2"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_19248"} +{"question": "Could you provide me some works about variational clustering that implicitly define RPMs to perform clustering?", "answer": ["Variational Deep Embedding: An Unsupervised and Generative Approach to Clustering", "Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoders", "Deep Conditional Gaussian Mixture Model for Constrained Clustering"], "answer_arxiv_id": ["1611.05148v3", "1611.02648", "2106.06385"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19249"} +{"question": "Which work introduced the concept of Forward Gradient and its ability to remove the backward lock in backpropagation algorithm?", "answer": ["Decoupled Neural Interfaces using Synthetic Gradients"], "answer_arxiv_id": ["1608.05343"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_19250"} +{"question": "What works contributed to the advancements in the field of text-conditioned image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2205.11487", "2204.06125"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_19251"} +{"question": "In what papers did the researchers find that varying temperature-like parameters during dense model training significantly affect pruning performance?", "answer": ["When to Prune? A Policy towards Early Structural Pruning", "Sparse Training via Boosting Pruning Plasticity with Neuroregeneration"], "answer_arxiv_id": ["2110.12007", "2106.10404"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_19252"} +{"question": "Which studies have characterized the sample complexity of pure offline RL and determined it with the quality of the offline dataset?", "answer": ["Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Offline Reinforcement Learning with Realizability and Single-policy Concentrability"], "answer_arxiv_id": ["2103.12021v2", "2202.04634"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19253"} +{"question": "Could you provide me some studies about commonsense reasoning datasets for large language models?", "answer": ["CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge", "Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategies"], "answer_arxiv_id": ["1811.00937", "2101.02235"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19254"} +{"question": "What research work explores the idea of model- and image-level fingerprints?", "answer": ["CNN-generated images are surprisingly easy to spot... for now", "Attributing Fake Images to GANs: Learning and Analyzing GAN Fingerprints"], "answer_arxiv_id": ["1912.11035", "1811.08180"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_19255"} +{"question": "Which works primarily focus on learning the non-rigid deformation of dense points on SMPL UV maps to represent dynamic garment wrinkles?", "answer": ["SCALE: Modeling Clothed Humans with a Surface Codec of Articulated Local\n Elements", "The Power of Points for Modeling Humans in Clothing"], "answer_arxiv_id": ["2104.07660", "2109.01137"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19256"} +{"question": "Can you provide some references where authors examined the outcomes of using alternative degradation methods instead of mask prediction as the pretext task?", "answer": ["Beyond Masking: Demystifying Token-Based Pre-Training for Vision Transformers", "Masked Frequency Modeling for Self-Supervised Visual Pre-Training", "Corrupted Image Modeling for Self-Supervised Visual Pre-Training"], "answer_arxiv_id": ["2203.14313", "2206.07706", "2202.03382"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19257"} +{"question": "Which paper extended the continuous-time framework to model node-edge joint distribution?", "answer": ["Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations"], "answer_arxiv_id": ["2202.02514"], "source_meta": {"published_time": "20230317"}, "qid": "AutoScholarQuery_train_19258"} +{"question": "Which works are about predicting the listener's fine-grain 2D or 3D gestural motion from the speaker's motion and audio?", "answer": ["Learning to Listen: Modeling Non-Deterministic Dyadic Facial Motion"], "answer_arxiv_id": ["2204.08451"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_19259"} +{"question": "Which research has utilized unCLIP model for generating a backward flow using retrieved images?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2204.06125"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19260"} +{"question": "Which work proposed the alignment of model predictions with trustworthy sources via prompting as a way to incorporate extraneous information about source trustworthiness?", "answer": ["Trusted Source Alignment in Large Language Models"], "answer_arxiv_id": ["2311.06697"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_19261"} +{"question": "Which papers discuss the usage of deep ensembles for DNN uncertainty quantification in i.i.d. inputs?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1612.01474"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_19262"} +{"question": "Which research articles discuss the importance of activation functions in neural networks?", "answer": ["Gaussian Error Linear Units (GELUs)"], "answer_arxiv_id": ["1606.08415"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_19263"} +{"question": "Which research extends LSS and query frameworks to process multiple frames?", "answer": ["BEVDet4D: Exploit Temporal Cues in Multi-camera 3D Object Detection", "BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detection", "Unifying Voxel-based Representation with Transformer for 3D Object Detection", "Polar Parametrization for Vision-based Surround-View 3D Detection", "PETRv2: A Unified Framework for 3D Perception from Multi-Camera Images", "BEVFormer: Learning Bird’s-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers"], "answer_arxiv_id": ["2203.17054", "2206.10092", "2206.00630", "2206.10965", "2206.01256", "2203.17270"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_19264"} +{"question": "Which studies discuss speaker recognition learned from audiovisual correspondences?", "answer": ["VoxCeleb2: Deep Speaker Recognition"], "answer_arxiv_id": ["1806.05622"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_19265"} +{"question": "Any recent work involves incorporating Gaussian mixture or Gamma noise into the forward process?", "answer": ["Non Gaussian Denoising Diffusion Models"], "answer_arxiv_id": ["2106.07582"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_19266"} +{"question": "Which paper analyzed a deformation that aligns a source surface with a target surface?", "answer": ["A Survey of Non-Rigid 3D Registration"], "answer_arxiv_id": ["2203.07858v2"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_19267"} +{"question": "Can you name the research that uses learning-based methods to predict aspects of a conversation such as turn-taking?", "answer": ["Towards Social Artificial Intelligence: Nonverbal Social Signal\n Prediction in A Triadic Interaction", "To React or not to React: End-to-End Visual Pose Forecasting for\n Personalized Avatar during Dyadic Conversations"], "answer_arxiv_id": ["1906.04158", "1910.02181"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_19268"} +{"question": "What research papers delve into safety-critical scenario generation?", "answer": ["SafeBench: A Benchmarking Platform for Safety Evaluation of Autonomous\n Vehicles", "Generating Useful Accident-Prone Driving Scenarios via a Learned Traffic\n Prior", "AdvDO: Realistic Adversarial Attacks for Trajectory Prediction", "Trace and Pace: Controllable Pedestrian Animation via Guided Trajectory\n Diffusion"], "answer_arxiv_id": ["2206.09682", "2112.05077", "2209.08744", "2304.01893"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19269"} +{"question": "Could you provide me works that generalize Sample Variance Penalization using the distributional robustness framework?", "answer": ["Distributionally Robust Counterfactual Risk Minimization", "Improving Offline Contextual Bandits with Distributional Robustness"], "answer_arxiv_id": ["1906.06211", "2011.06835"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_19270"} +{"question": "What papers have been published on the topic of sparsely-activated Mixture-of-Experts (MoE)?", "answer": ["GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding", "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity", "Scaling Vision with Sparse Mixture of Experts", "Multimodal Contrastive Learning with LIMoE: the Language-Image Mixture of Experts"], "answer_arxiv_id": ["2006.16668", "2101.03961", "2106.05974", "2206.02770"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_19271"} +{"question": "Which research constructs a dictionary of deep features from important cropped regions and then align the features of the input and reference images?", "answer": ["Learning Dual Memory Dictionaries for Blind Face Restoration"], "answer_arxiv_id": ["2210.08160"], "source_meta": {"published_time": "20240313"}, "qid": "AutoScholarQuery_train_19272"} +{"question": "What study builds a composition space by simulating all the visual changes of attributes performed on objects in CZSL?", "answer": ["Attributes as Operators: Factorizing Unseen Attribute-Object\n Compositions"], "answer_arxiv_id": ["1803.09851"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19273"} +{"question": "Could you provide me with some works about predicting speech given context in infilling?", "answer": ["On Generative Spoken Language Modeling from Raw Audio", "AudioLM: a Language Modeling Approach to Audio Generation"], "answer_arxiv_id": ["2102.01192", "2209.03143"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_19274"} +{"question": "Which works conduct timeline modeling for textual data in natural language processing?", "answer": ["Examining the State-of-the-Art in News Timeline Summarization"], "answer_arxiv_id": ["2005.10107"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_19275"} +{"question": "Which works ease CAD authoring by using differentiable execution?", "answer": ["Differentiable 3D CAD Programs for Bidirectional Editing"], "answer_arxiv_id": ["2110.01182"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_19276"} +{"question": "What papers interleave equivariant linear maps and equivariant pointwise nonlinearities for the design of equivariant neural networks?", "answer": ["Group Equivariant Convolutional Networks", "Steerable CNNs", "Equivariance Through Parameter-Sharing", "Invariant and Equivariant Graph Networks", "On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups", "A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups", "Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges"], "answer_arxiv_id": ["1602.07576", "1612.08498", "1702.08389", "1812.09902", "1802.03690", "2104.09459", "2104.13478"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_19277"} +{"question": "What papers have proposed the use of text prompts for more natural and general speech synthesis in TTS models?", "answer": ["PromptTTS: Controllable Text-to-Speech with Text Descriptions", "PromptTTS 2: Describing and Generating Voices with Text Prompt", "PromptTTS++: Controlling Speaker Identity in Prompt-Based Text-to-Speech\n Using Natural Language Descriptions", "TextrolSpeech: A Text Style Control Speech Corpus With Codec Language\n Text-to-Speech Models"], "answer_arxiv_id": ["2211.12171", "2309.02285", "2309.08140", "2308.14430"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_19278"} +{"question": "Can you name the researches that focused on datasets addressing the multi-step and structured aspect reasoning with explanations?", "answer": ["Transformers as Soft Reasoners over Language"], "answer_arxiv_id": ["2002.05867"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_19279"} +{"question": "What are some references discussing competition between bandit algorithms where data directly comes from users?", "answer": ["Competition, Alignment, and Equilibria in Digital Marketplaces"], "answer_arxiv_id": ["2208.14423"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_19280"} +{"question": "Which studies focused on explaining the representation capacity of a DNN in the time domain?", "answer": ["Deep Learning and the Information Bottleneck Principle", "Opening the black box of Deep Neural Networks via Information", "Learning Representations for Neural Network-Based Classification Using the Information Bottleneck Principle", "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks", "Deep Double Descent: Where Bigger Models and More Data Hurt", "Early Stopping in Deep Networks: Double Descent and How to Eliminate it", "L2 Regularization versus Batch and Weight Normalization", "Discovering and Explaining the Representation Bottleneck of DNNs"], "answer_arxiv_id": ["1503.02406", "1703.00810", "1802.09766", "1803.03635", "1912.02292", "2007.10099v2", "1706.05350v1", "2111.06236"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_19281"} +{"question": "What researches focus on learning the branching structure of multi-task networks?", "answer": ["A Tree-Structured Multi-Task Model Recommender", "Controllable Dynamic Multi-Task Architectures", "Learning to Branch for Multi-Task Learning", "Automated Search for Resource-Efficient Branched Multi-Task Networks", "Fully-adaptive Feature Sharing in Multi-Task Networks with Applications in Person Attribute Classification", "AdaShare: Learning What To Share For Efficient Deep Multi-Task Learning"], "answer_arxiv_id": ["2203.05092", "2203.14949", "2006.01895", "2008.10292", "1611.05377", "1911.12423"], "source_meta": {"published_time": "20230430"}, "qid": "AutoScholarQuery_train_19282"} +{"question": "Could you list some papers that build music-dance datasets using MoCap to record 3D skeletons?", "answer": ["Music2Dance: DanceNet for Music-driven Dance Generation", "Transflower: probabilistic autoregressive dance generation with\n multimodal attention", "FineDance: A Fine-grained Choreography Dataset for 3D Full Body Dance\n Generation"], "answer_arxiv_id": ["2002.03761", "2106.13871", "2212.03741"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_19283"} +{"question": "Could you name any researches that were outperformed by the Bonsai model in experiments?", "answer": ["LLM-Pruner: On the Structural Pruning of Large Language Models"], "answer_arxiv_id": ["2305.11627"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_19284"} +{"question": "In what papers did researchers identify low-level features like color, texture and style to predict visual emotions?", "answer": ["Learning Multi-level Deep Representations for Image Emotion\n Classification"], "answer_arxiv_id": ["1611.07145"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_19285"} +{"question": "Could you give me references on studies that demonstrated gradient descent with early stopping can still achieve near-optimal test loss?", "answer": ["Implicit Regularization for Optimal Sparse Recovery", "Implicit Sparse Regularization: The Impact of Depth and Early Stopping"], "answer_arxiv_id": ["1909.05122", "2108.05574"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_19286"} +{"question": "Could you tell me any works that proposed Multi-view CNN for 3D object recognition?", "answer": ["Multi-view Convolutional Neural Networks for 3D Shape Recognition"], "answer_arxiv_id": ["1505.00880"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_19287"} +{"question": "What research has been conducted on Multimodal Large Language Models (MLLMs) focusing on audio modality?", "answer": ["AudioGPT: Understanding and Generating Speech, Music, Sound, and Talking\n Head", "Text-to-Audio Generation using Instruction-Tuned LLM and Latent\n Diffusion Model"], "answer_arxiv_id": ["2304.12995", "2304.13731"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19288"} +{"question": "Which papers propose a method to estimate the model’s uncertainty and penalize the actions with highly uncertain consequences in offline RL?", "answer": ["When to Trust Your Model: Model-Based Policy Optimization", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["1906.08253", "2005.05951"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_19289"} +{"question": "What papers proposed guaranteed monotonicity for standard ReLU networks?", "answer": ["Counterexample-Guided Learning of Monotonic Neural Networks"], "answer_arxiv_id": ["2006.08852"], "source_meta": {"published_time": "20230714"}, "qid": "AutoScholarQuery_train_19290"} +{"question": "What papers describe Vision-Language Models that excel in image-text generation?", "answer": ["Zero-Shot Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2102.12092", "2204.06125"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_19291"} +{"question": "Have any contrastive pre-training methodologies been applied for the boolean satisfiability problem?", "answer": ["Augment with Care: Contrastive Learning for Combinatorial Problems"], "answer_arxiv_id": ["2202.08396"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_19292"} +{"question": "What studies developed novel approaches to parameter-efficient fine-tuning for transformer architecture such as LoRA and BitFit?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models"], "answer_arxiv_id": ["2106.09685", "2106.10199"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_19293"} +{"question": "Which strategies have been demonstrated to be effective for both SGG and the PSG tasks?", "answer": ["Neural Motifs: Scene Graph Parsing with Global Context", "Scene Graph Generation by Iterative Message Passing", "Unbiased Scene Graph Generation from Biased Training", "Panoptic Scene Graph Generation", "HiLo: Exploiting High Low Frequency Relations for Unbiased Panoptic\n Scene Graph Generation"], "answer_arxiv_id": ["1711.06640", "1701.02426", "2002.11949", "2207.11247", "2303.15994"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_19294"} +{"question": "What are some research papers in the category of methods that employ specially designed attention masks?", "answer": ["A Survey of Transformers"], "answer_arxiv_id": ["2106.04554"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19295"} +{"question": "Can you mention the works that detail single-call stochastic methods for monotone problems?", "answer": ["On the Convergence of Single-Call Stochastic Extra-Gradient Methods"], "answer_arxiv_id": ["1908.08465"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_19296"} +{"question": "What works have highlighted the issues faced by direct multi-concept generation using text prompts alone, such as missing objects and attribute binding?", "answer": ["Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesis", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "Aligning Text-to-Image Models using Human Feedback", "Directed Diffusion: Direct Control of Object Placement through Attention Guidance"], "answer_arxiv_id": ["2212.05032", "2206.10789", "2302.12192", "2302.13153"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_19297"} +{"question": "What papers describe a technique of gloss supervision with supervised SLT fine-tuning when predicting glosses from continuous signing?", "answer": ["Sign Language Transformers: Joint End-to-end Sign Language Recognition\n and Translation"], "answer_arxiv_id": ["2003.13830"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_19298"} +{"question": "What are the papers discussing time series analysis using autoregressive modeling?", "answer": ["Autoregressive Times Series Methods for Time Domain Astronomy"], "answer_arxiv_id": ["1901.08003v1"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_19299"} +{"question": "Are there any works that concentrated on entity-based conflicts in language models?", "answer": ["Entity-Based Knowledge Conflicts in Question Answering"], "answer_arxiv_id": ["2109.05052"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_19300"} +{"question": "Which research provides a non-convex inverse problem for recovering the ground metric and interaction kernel?", "answer": ["A mean field game inverse problem"], "answer_arxiv_id": ["2007.11551"], "source_meta": {"published_time": "20220518"}, "qid": "AutoScholarQuery_train_19301"} +{"question": "Could you provide me some works that have developed stronger adversarial attack methods recently?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Towards Evaluating the Robustness of Neural Networks", "Square Attack: a query-efficient black-box adversarial attack via random search", "Reliable Evaluation of Adversarial Robustness with an Ensemble of Diverse Parameter-free Attacks", "Generating Adversarial Examples with Adversarial Networks", "Spatially Transformed Adversarial Examples", "MeshAdv: Adversarial Meshes for Visual Recognition", "Characterizing Attacks on Deep Reinforcement Learning", "DensePure: Understanding Diffusion Models towards Adversarial Robustness", "Adversarial Objects Against LiDAR-Based Autonomous Driving Systems", "Adversarial Sensor Attack on LiDAR-based Perception in Autonomous Driving", "AdvDO: Realistic Adversarial Attacks for Trajectory Prediction"], "answer_arxiv_id": ["1706.06083", "1608.04644", "1912.00049", "2003.01690", "1801.02610", "1801.02612", "1810.05206", "1907.09470", "2211.00322", "1907.05418", "1907.06826", "2209.08744"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19302"} +{"question": "Which papers have proposed advanced techniques such as adversarial training and dynamic distillation in the context of classical models?", "answer": ["Adversarial Retriever-Ranker for dense text retrieval", "RocketQAv2: A Joint Training Method for Dense Passage Retrieval and\n Passage Re-ranking"], "answer_arxiv_id": ["2110.03611", "2110.07367"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_19303"} +{"question": "Which papers discuss the use of deep neural networks in keypoint detection and description for image matching?", "answer": ["Quad-networks: unsupervised learning to rank for interest point\n detection", "Key.Net: Keypoint Detection by Handcrafted and Learned CNN Filters", "Working hard to know your neighbor's margins: Local descriptor learning\n loss", "SOSNet: Second Order Similarity Regularization for Local Descriptor\n Learning", "Beyond Cartesian Representations for Local Descriptors"], "answer_arxiv_id": ["1611.07571", "1904.00889", "1705.10872", "1904.05019", "1908.05547"], "source_meta": {"published_time": "20240307"}, "qid": "AutoScholarQuery_train_19304"} +{"question": "Which work interpreted forward passing in neural networks as an integration of ODEs?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_19305"} +{"question": "What papers initiated the concept of diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_19306"} +{"question": "What works enriched the representation power of NAT by using a longer decoder?", "answer": ["Directed Acyclic Transformer for Non-Autoregressive Machine Translation", "Viterbi Decoding of Directed Acyclic Transformer for Non-Autoregressive Machine Translation"], "answer_arxiv_id": ["2205.07459", "2210.05193"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_19307"} +{"question": "What works zero in on MDP ambiguity for task inference in offline meta-RL with task-dependent behavior policies?", "answer": ["FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization"], "answer_arxiv_id": ["2010.01112"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19308"} +{"question": "Which studies discuss the difficulties in acquiring distributional labels?", "answer": ["Deep Differentiable Random Forests for Age Estimation"], "answer_arxiv_id": ["1907.10665"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19309"} +{"question": "What papers explore causal Bandit Optimization in the setting of unknown Direct Acyclic Graphs?", "answer": ["Causal Entropy Optimization", "BoGraph: Structured Bayesian Optimization From Logs for Expensive Systems with Many Parameters"], "answer_arxiv_id": ["2208.10981", "2112.08774"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_19310"} +{"question": "Could you provide me a study that uses multi-resolution tri-planes in training neural fields?", "answer": ["Anti-Aliased Neural Implicit Surfaces with Encoding Level of Detail"], "answer_arxiv_id": ["2309.10336"], "source_meta": {"published_time": "20240419"}, "qid": "AutoScholarQuery_train_19311"} +{"question": "What was the sole dataset used for cross-system data in learning-based theorem proving?", "answer": ["miniF2F: a cross-system benchmark for formal Olympiad-level mathematics"], "answer_arxiv_id": ["2109.00110"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_19312"} +{"question": "What papers have studied self-critiquing as a method of assistance in LLMs?", "answer": ["Self-critiquing models for assisting human evaluators"], "answer_arxiv_id": ["2206.05802v2"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_19313"} +{"question": "Which papers describe distilling knowledge from a pre-trained teacher GNN to a shallower student GNN?", "answer": ["Distilling Knowledge from Graph Convolutional Networks"], "answer_arxiv_id": ["2003.10477"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_19314"} +{"question": "Are there any papers about generating 3D molecules by voxelizing molecules in atomic density grids?", "answer": ["Generating 3D Molecules Conditional on Receptor Binding Sites with Deep Generative Models"], "answer_arxiv_id": ["2110.15200"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_19315"} +{"question": "Which works contribute towards understanding engineering advancements in unsupervised methods?", "answer": ["Deep Roto-Translation Scattering for Object Classification", "Deep Scattering Spectrum", "Enhanced Convolutional Neural Tangent Kernels", "On Exact Computation with an Infinitely Wide Neural Net", "Neural Kernels Without Tangents"], "answer_arxiv_id": ["1412.8659", "1304.6763", "1911.00809v1", "1904.11955", "2003.02237"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19316"} +{"question": "What research introduced reinforcement learning and physical simulation environments to enhance the physical realism of generated movements?", "answer": ["CALM: Conditional Adversarial Latent Models for Directable Virtual\n Characters", "ASE: Large-Scale Reusable Adversarial Skill Embeddings for Physically\n Simulated Characters"], "answer_arxiv_id": ["2305.02195", "2205.01906"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_19317"} +{"question": "Which studies developed weakly supervised 3D object detection in a monocular setting?", "answer": ["WeakM3D: Towards Weakly Supervised Monocular 3D Object Detection", "Weakly Supervised Monocular 3D Object Detection using Multi-View\n Projection and Direction Consistency"], "answer_arxiv_id": ["2203.08332", "2303.08686"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19318"} +{"question": "Are there any studies that employ language models to generate answers directly by finetuning and few-shot prompting?", "answer": ["How Much Knowledge Can You Pack Into the Parameters of a Language Model?", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2002.08910", "2005.14165"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_19319"} +{"question": "Which studies have focused on the extension of NeRF techniques to dynamic deformable domains?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View\n Synthesis of a Dynamic Scene From Monocular Video", "Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar\n Reconstruction", "Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2012.12247", "2012.03065", "2011.12948", "2106.13228v2"], "source_meta": {"published_time": "20231209"}, "qid": "AutoScholarQuery_train_19320"} +{"question": "Which works have used human priors for robust reconstruction of face and body?", "answer": ["Portrait Neural Radiance Fields from a Single Image", "Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar Reconstruction", "Neural Body: Implicit Neural Representations with Structured Latent Codes for Novel View Synthesis of Dynamic Humans"], "answer_arxiv_id": ["2012.05903", "2012.03065", "2012.15838"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_19321"} +{"question": "Are there any works about using pixel-level segmentation technique combined with post-processing in lane detection?", "answer": ["Spatial As Deep: Spatial CNN for Traffic Scene Understanding", "Towards End-to-End Lane Detection: an Instance Segmentation Approach"], "answer_arxiv_id": ["1712.06080", "1802.05591"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_19322"} +{"question": "Which works discuss activity cliffs?", "answer": ["Activity Cliff Prediction: Dataset and Benchmark"], "answer_arxiv_id": ["2302.07541"], "source_meta": {"published_time": "20231010"}, "qid": "AutoScholarQuery_train_19323"} +{"question": "Could you provide some works that have proposed policy gradient-based RL algorithms to solve text degeneration?", "answer": ["Reward Augmented Maximum Likelihood for Neural Structured Prediction", "Multi-Reward Reinforced Summarization with Saliency and Entailment", "A Study of Reinforcement Learning for Neural Machine Translation"], "answer_arxiv_id": ["1609.00150", "1804.06451", "1808.08866"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_19324"} +{"question": "Can you mention some studies where OT was formulated for semi-supervised learning?", "answer": ["Sinkhorn Label Allocation: Semi-Supervised Classification via Annealed Self-Training", "Confident Sinkhorn Allocation for Pseudo-Labeling"], "answer_arxiv_id": ["2102.08622", "2206.05880"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_19325"} +{"question": "What are the research papers that used NeRFs for talking head synthesis?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Nerfies: Deformable Neural Radiance Fields", "Efficient Geometry-aware 3D Generative Adversarial Networks"], "answer_arxiv_id": ["2003.08934", "2011.12948", "2112.07945"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19326"} +{"question": "What are the works that have parameterised vector sequences of sketches with parametric curves like Béziers and B-Splines?", "answer": ["B\\'ezierSketch: A generative model for scalable vector sketches", "CLIPasso: Semantically-Aware Object Sketching", "Fast B-spline Curve Fitting by L-BFGS"], "answer_arxiv_id": ["2007.02190", "2202.05822", "1201.0070"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_19327"} +{"question": "Could you provide me some works about in-processing backdoor defense measures used during training?", "answer": ["Anti-Backdoor Learning: Training Clean Models on Poisoned Data", "Backdoor Defense via Decoupling the Training Process"], "answer_arxiv_id": ["2110.11571", "2202.03423"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_19328"} +{"question": "Which work reported the degradation of image reconstruction quality when the LQAE tokens were used?", "answer": ["Language Quantized AutoEncoders: Towards Unsupervised Text-Image Alignment", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding"], "answer_arxiv_id": ["2302.00902", "1810.04805"], "source_meta": {"published_time": "20230630"}, "qid": "AutoScholarQuery_train_19329"} +{"question": "Which papers directly optimize the LMs against the native sequence-level feedback in language model training?", "answer": ["Self-critical Sequence Training for Image Captioning", "Sequence Level Training with Recurrent Neural Networks", "A Deep Reinforced Model for Abstractive Summarization", "Reward Optimization for Neural Machine Translation with Learned Metrics", "Quark: Controllable Text Generation with Reinforced [Un]learning", "Offline RL for Natural Language Generation with Implicit Language Q Learning"], "answer_arxiv_id": ["1612.00563", "1511.06732", "1705.04304", "2104.07541", "2205.13636", "2206.11871"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_19330"} +{"question": "Which papers introduced Poisson processes to channel simulation literature and proposed the Poisson functional representation (PFR)?", "answer": ["Strong Functional Representation Lemma and Applications to Coding Theorems"], "answer_arxiv_id": ["1701.02827"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19331"} +{"question": "What works adopted learning from human feedback for developing human-friendly general language models?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_19332"} +{"question": "What papers introduce positional encodings to make style-GAN generator shift and scale invariant?", "answer": ["Positional Encoding as Spatial Inductive Bias in GANs", "Arbitrary-Scale Image Synthesis"], "answer_arxiv_id": ["2012.05217", "2204.02273"], "source_meta": {"published_time": "20221220"}, "qid": "AutoScholarQuery_train_19333"} +{"question": "Which works first achieved impressive zero-shot text-to-image generation with good fidelity and image-text alignment?", "answer": ["Zero-Shot Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19334"} +{"question": "What studies explored the concept of storing past representations to improve memory and long-range sequence modeling in Transformers?", "answer": ["Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Generating Wikipedia by Summarizing Long Sequences", "Compressive Transformers for Long-Range Sequence Modelling", "Memorizing Transformers", "Poolingformer: Long Document Modeling with Pooling Attention"], "answer_arxiv_id": ["1901.02860", "1801.10198", "1911.05507", "2203.08913", "2105.04371"], "source_meta": {"published_time": "20230417"}, "qid": "AutoScholarQuery_train_19335"} +{"question": "Could you provide me some studies that use iterative prompts to decide if extra information is required for retrieval necessity?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "What Makes Good In-Context Examples for GPT-$3$?", "Learning To Retrieve Prompts for In-Context Learning"], "answer_arxiv_id": ["2201.11903", "2101.06804", "2112.08633"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_19336"} +{"question": "What works discuss improving the generalization in long-tailed learning?", "answer": ["Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss", "Long-Tail Learning via Logit Adjustment", "Identifying and Compensating for Feature Deviation in Imbalanced Deep Learning", "Adjusting Decision Boundary for Class Imbalanced Learning", "Label-Imbalanced and Group-Sensitive Classification under Overparameterization"], "answer_arxiv_id": ["1906.07413", "2007.07314", "2001.01385", "1912.01857", "2103.01550"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_19337"} +{"question": "Which works investigate relighting and view synthesis in a unified framework, but rely on a multi-view setup?", "answer": ["SunStage: Portrait Reconstruction and Relighting using the Sun as a\n Light Stage"], "answer_arxiv_id": ["2204.03648"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_19338"} +{"question": "Which dataset is used to dissect user intention distribution, co-occurrence, and flow patterns?", "answer": ["Analyzing and Characterizing User Intent in Information-seeking\n Conversations"], "answer_arxiv_id": ["1804.08759"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_19339"} +{"question": "What are works that argued about feature-level perspective of GBML?", "answer": ["Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML", "BOIL: Towards Representation Change for Few-shot Learning"], "answer_arxiv_id": ["1909.09157", "2008.08882"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_19340"} +{"question": "Could you cite works that discussed the crucial role of linearity of the final layer in establishing the approximate constancy of the tangent kernel for wide networks?", "answer": ["On the linearity of large non-linear models: when and why the tangent kernel is constant"], "answer_arxiv_id": ["2010.01092"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_19341"} +{"question": "Could you provide me some studies about single-stage methods for oriented detection?", "answer": ["Dynamic Refinement Network for Oriented and Densely Packed Object\n Detection", "R3Det: Refined Single-Stage Detector with Feature Refinement for\n Rotating Object", "Align Deep Features for Oriented Object Detection"], "answer_arxiv_id": ["2005.09973", "1908.05612", "2008.09397"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19342"} +{"question": "In which papers does the effectiveness of text augmentation for NLP tasks been discussed?", "answer": ["How Transferable are Neural Networks in NLP Applications?", "EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks", "EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks", "Data Augmentation Using Pre-trained Transformer Models", "Improving Neural Machine Translation Models with Monolingual Data"], "answer_arxiv_id": ["1603.06111", "1901.11196", "1901.11196", "2003.02245", "1511.06709v4"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19343"} +{"question": "Could you provide some research papers that focus on annotation efficiency in LiDAR segmentation?", "answer": ["LESS: Label-Efficient Semantic Segmentation for LiDAR Point Clouds", "Segment Any Point Cloud Sequences by Distilling Vision Foundation Models", "Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Data", "Scribble-Supervised LiDAR Semantic Segmentation", "COARSE3D: Class-Prototypes for Contrastive Learning in Weakly-Supervised\n 3D Point Cloud Segmentation"], "answer_arxiv_id": ["2210.08064", "2306.09347", "2203.16258", "2203.08537", "2210.01784"], "source_meta": {"published_time": "20240502"}, "qid": "AutoScholarQuery_train_19344"} +{"question": "What works implement the time-inhomogeneous variant of Langevin dynamics using equal-sized steps?", "answer": ["Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2006.11239", "2011.13456"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_19345"} +{"question": "Are there any researches that designed a multi-round communication game for the sender and receiver?", "answer": ["Emergent Graphical Conventions in a Visual Communication Game"], "answer_arxiv_id": ["2111.14210"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_19346"} +{"question": "Which papers discuss automatic differentiation techniques to estimate gradients in UCGs?", "answer": ["Automatic Differentiation in Machine Learning: a Survey"], "answer_arxiv_id": ["1502.05767"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_19347"} +{"question": "Could you provide me some works about enabling LLMs to comprehend 3D point-cloud?", "answer": ["Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D\n Understanding, Generation, and Instruction Following"], "answer_arxiv_id": ["2309.00615"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_19348"} +{"question": "What papers developed amodal completion methods?", "answer": ["Self-Supervised Scene De-occlusion", "A Weakly Supervised Amodal Segmenter with Boundary Uncertainty\n Estimation"], "answer_arxiv_id": ["2004.02788", "2108.09897"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_19349"} +{"question": "What research works are in the field of proxy methods to scale-up expensive ab-initio solvers?", "answer": ["MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields", "SE(3)-equivariant prediction of molecular wavefunctions and electronic densities"], "answer_arxiv_id": ["2206.07697", "2106.02347"], "source_meta": {"published_time": "20230715"}, "qid": "AutoScholarQuery_train_19350"} +{"question": "Can you tell me about research conducting finetuning open-source models to improve the mathematical reasoning capabilities of LLMs?", "answer": ["Scaling Relationship on Learning Mathematical Reasoning with Large\n Language Models", "WizardMath: Empowering Mathematical Reasoning for Large Language Models\n via Reinforced Evol-Instruct", "MAmmoTH: Building Math Generalist Models through Hybrid Instruction\n Tuning", "Making Large Language Models Better Reasoners with Alignment", "MetaMath: Bootstrap Your Own Mathematical Questions for Large Language\n Models", "ToRA: A Tool-Integrated Reasoning Agent for Mathematical Problem Solving"], "answer_arxiv_id": ["2308.01825", "2308.09583", "2309.05653", "2309.02144", "2309.12284", "2309.17452"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_19351"} +{"question": "What works introduced the concept of using historical embeddings in Graph Neural Networks?", "answer": ["Stochastic Training of Graph Convolutional Networks with Variance Reduction", "GNNAutoScale: Scalable and Expressive Graph Neural Networks via Historical Embeddings", "GraphFM: Improving Large-Scale GNN Training via Feature Momentum"], "answer_arxiv_id": ["1710.10568v3", "2106.05609", "2206.07161"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_19352"} +{"question": "Which studies improved the sample complexity for reward-free RL for the tabular case?", "answer": ["Adaptive Reward-Free Exploration", "Fast active learning for pure exploration in reinforcement learning"], "answer_arxiv_id": ["2006.06294", "2007.13442"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_19353"} +{"question": "Which papers come up with approximate unlearning by calculating influence functions?", "answer": ["Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep\n Networks", "Certified Data Removal from Machine Learning Models", "Remember What You Want to Forget: Algorithms for Machine Unlearning", "Knowledge Unlearning for Mitigating Privacy Risks in Language Models"], "answer_arxiv_id": ["1911.04933", "1911.03030", "2103.03279", "2210.01504"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_19354"} +{"question": "Could you provide me some studies about the trajectory-first methods in variational option discovery?", "answer": ["Explore, Discover and Learn: Unsupervised Discovery of State-Covering Skills", "OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning"], "answer_arxiv_id": ["2002.03647", "2010.13611"], "source_meta": {"published_time": "20221201"}, "qid": "AutoScholarQuery_train_19355"} +{"question": "Which study noticed instability when directly minimizing the worst-group loss?", "answer": ["Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization"], "answer_arxiv_id": ["1911.08731"], "source_meta": {"published_time": "20220714"}, "qid": "AutoScholarQuery_train_19356"} +{"question": "What papers proposed using Truncated Neumann series and efficient vector-Jacobian products for BLO?", "answer": ["Reviving and Improving Recurrent Back-Propagation", "Optimizing Millions of Hyperparameters by Implicit Differentiation", "Teaching with Commentaries"], "answer_arxiv_id": ["1803.06396v4", "1911.02590", "2011.03037"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19357"} +{"question": "Could you provide me some studies that investigate converting conversational intents into programmable representations, such as SQL or dataflow graph?", "answer": ["CoSQL: A Conversational Text-to-SQL Challenge Towards Cross-Domain Natural Language Interfaces to Databases", "SParC: Cross-Domain Semantic Parsing in Context", "Task-Oriented Dialogue as Dataflow Synthesis"], "answer_arxiv_id": ["1909.05378", "1906.02285", "2009.11423"], "source_meta": {"published_time": "20220325"}, "qid": "AutoScholarQuery_train_19358"} +{"question": "Are there any papers that achieved nonasymptotic variance reduction guarantees for the case of nonconvex optimization?", "answer": ["Cyclic Block Coordinate Descent With Variance Reduction for Composite Nonconvex Optimization", "Block Stochastic Gradient Iteration for Convex and Nonconvex Optimization"], "answer_arxiv_id": ["2212.05088", "1408.2597"], "source_meta": {"published_time": "20230328"}, "qid": "AutoScholarQuery_train_19359"} +{"question": "What papers proposed modeling moiré patterns in the frequency domain for image demoiréing?", "answer": ["Wavelet-Based Dual-Branch Network for Image Demoiréing"], "answer_arxiv_id": ["2007.07173"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_19360"} +{"question": "Which works explored image editing by inserting and harmonizing objects using generative models?", "answer": ["Image Harmonization with Diffusion Model"], "answer_arxiv_id": ["2306.10441"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_19361"} +{"question": "What papers consider finite actions and continuous state space by discretizing states or state features via kNN?", "answer": ["DeepAveragers: Offline Reinforcement Learning by Solving Derived Non-Parametric MDPs"], "answer_arxiv_id": ["2010.08891"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_19362"} +{"question": "Can you name some of the research that conducted a mean-field analysis of neural networks?", "answer": ["A Mean Field View of the Landscape of Two-Layer Neural Networks", "On the Global Convergence of Gradient Descent for Over-parameterized Models using Optimal Transport", "Mean-field theory of two-layers neural networks: dimension-free bounds and kernel limit", "Mean Field Analysis of Neural Networks: A Central Limit Theorem", "On Feature Learning in Neural Networks with Global Convergence Guarantees", "Depth Separation with Multilayer Mean-Field Networks", "Overparameterization of deep ResNet: zero loss and mean-field analysis"], "answer_arxiv_id": ["1804.06561", "1805.09545", "1902.06015", "1808.09372", "2204.10782", "2304.01063", "2105.14417"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_19363"} +{"question": "Which works have proposed to rely purely on implicit representation or combine it with statistical models for reconstructing expressive 3D human bodies?", "answer": ["PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization", "ICON: Implicit Clothed humans Obtained from Normals", "Neural Body: Implicit Neural Representations with Structured Latent Codes for Novel View Synthesis of Dynamic Humans", "Geometry-Guided Progressive NeRF for Generalizable and Efficient Neural Human Rendering"], "answer_arxiv_id": ["1905.05172v3", "2112.09127", "2012.15838", "2112.04312"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_19364"} +{"question": "Are there any research papers showing deep learning systems can infer causal relations from data where interventions are present but latent?", "answer": ["Differentiable Causal Discovery Under Latent Interventions"], "answer_arxiv_id": ["2203.02336v1"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19365"} +{"question": "What works utilized transformers and pre-trained language models for encoding contextual semantics in sentences for text representation learning on TAGs?", "answer": ["Attention Is All You Need", "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "XLNet: Generalized Autoregressive Pretraining for Language Understanding"], "answer_arxiv_id": ["1706.03762", "1810.04805", "1906.08237"], "source_meta": {"published_time": "20221026"}, "qid": "AutoScholarQuery_train_19366"} +{"question": "Could you provide me some tools and datasets for the Lean proof assistant?", "answer": ["Proof Artifact Co-training for Theorem Proving with Language Models", "Formal Mathematics Statement Curriculum Learning"], "answer_arxiv_id": ["2102.06203", "2202.01344"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_19367"} +{"question": "Which work introduced a 4D Convolutional Swin Transformer?", "answer": ["Cost Aggregation with 4D Convolutional Swin Transformer for Few-Shot Segmentation"], "answer_arxiv_id": ["2207.10866"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_19368"} +{"question": "What research work investigates locking and finetuning from JFT-pretrained models?", "answer": ["LiT: Zero-Shot Transfer with Locked-image text Tuning", "Combined Scaling for Zero-shot Transfer Learning"], "answer_arxiv_id": ["2111.07991", "2111.10050"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19369"} +{"question": "Which studies tackled the issue of weak performance in heterophilic setups in graph structure processing?", "answer": ["Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs"], "answer_arxiv_id": ["2202.04579"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_19370"} +{"question": "Which work optimizes LLMs with PPO algorithm for model optimization?", "answer": ["Fine-Tuning Language Models from Human Preferences", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["1909.08593", "2203.02155"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_19371"} +{"question": "Which works are related to model-level explanation methods for GNNs?", "answer": ["XGNN: Towards Model-Level Explanations of Graph Neural Networks", "Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps"], "answer_arxiv_id": ["2006.02587", "1312.6034"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_19372"} +{"question": "Who proposed modifications on message passing rule in GNNs with the goal of enhancing graph structure utilization?", "answer": ["Directional Graph Networks", "Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks", "Hierarchical Inter-Message Passing for Learning on Molecular Graphs", "Weisfeiler and Lehman Go Cellular: CW Networks"], "answer_arxiv_id": ["2010.02863", "1810.02244", "2006.12179", "2106.12575"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_19373"} +{"question": "What works proposed DeepONets, an alternative to the Fourier Neural Operator?", "answer": ["Learning the solution operator of parametric partial differential equations with physics-informed DeepOnets", "MIONet: Learning multiple-input operators via tensor product"], "answer_arxiv_id": ["2103.10974", "2202.06137"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_19374"} +{"question": "Which paper drew a relation between '5+1' phase transitions and degrees of regularization of Neural Networks?", "answer": ["Predicting trends in the quality of state-of-the-art neural networks without access to training or testing data"], "answer_arxiv_id": ["2002.06716v2"], "source_meta": {"published_time": "20221111"}, "qid": "AutoScholarQuery_train_19375"} +{"question": "Is there a comparable model to SAM with impressive open-vocabulary segmentation capabilities?", "answer": ["Segment Everything Everywhere All at Once"], "answer_arxiv_id": ["2304.06718"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_19376"} +{"question": "What studies included research on data poisoning in attacks of multi-armed bandit models?", "answer": ["Data Poisoning against Differentially-Private Learners: Attacks and Defenses"], "answer_arxiv_id": ["1903.09860"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_19377"} +{"question": "Which works made advancements in text-conditioned diffusion for image quality and image-text alignment?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2112.10752", "2205.11487"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19378"} +{"question": "What works have integrated this idea with view-conditioned diffusion methods?", "answer": ["One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization", "Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors", "Wonder3D: Single Image to 3D using Cross-Domain Diffusion"], "answer_arxiv_id": ["2306.16928", "2306.17843", "2310.15008"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_19379"} +{"question": "Which papers have discussed nonparametric score estimation?", "answer": ["A Kernelized Stein Discrepancy for Goodness-of-fit Tests", "A Spectral Approach to Gradient Estimation for Implicit Distributions", "Nonparametric Score Estimators"], "answer_arxiv_id": ["1602.03253", "1806.02925", "2005.10099"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_19380"} +{"question": "Which papers suggest the global loss landscape metrics such as mode connectivity and model similarity to indicate the phase transition of model training?", "answer": ["Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs", "Similarity of Neural Network Representations Revisited"], "answer_arxiv_id": ["1802.10026", "1905.00414"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_19381"} +{"question": "Could you provide me some studies that attempted to freeze DNN layers adaptively using gradient-norm or SVCCA score?", "answer": ["AutoFreeze: Automatically Freezing Model Blocks to Accelerate Fine-tuning", "PipeTransformer: Automated Elastic Pipelining for Distributed Training of Transformers"], "answer_arxiv_id": ["2102.01386v2", "2102.03161"], "source_meta": {"published_time": "20240130"}, "qid": "AutoScholarQuery_train_19382"} +{"question": "What studies extended the image-to-image translation paradigm to synthesize videos and panoptic masks?", "answer": ["A Generalist Framework for Panoptic Segmentation of Images and Videos"], "answer_arxiv_id": ["2210.06366"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_19383"} +{"question": "Are there any contributions to universality in the context of Transformer model variants?", "answer": ["Your Transformer May Not be as Powerful as You Expect"], "answer_arxiv_id": ["2205.13401"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19384"} +{"question": "What studies explored aligning visual features with pre-trained text embeddings in open-vocabulary semantic segmentation?", "answer": ["Open Vocabulary Scene Parsing", "Zero-Shot Semantic Segmentation"], "answer_arxiv_id": ["1703.08769", "1906.00817"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19385"} +{"question": "What works developed a private follow-the-regularized-leader algorithm for online convex optimization?", "answer": ["Practical and Private (Deep) Learning Without Sampling or Shuffling"], "answer_arxiv_id": ["2103.00039"], "source_meta": {"published_time": "20230227"}, "qid": "AutoScholarQuery_train_19386"} +{"question": "What research introduced depth prediction conditioned on latent codes for geometric priors?", "answer": ["CodeSLAM - Learning a Compact, Optimisable Representation for Dense\n Visual SLAM"], "answer_arxiv_id": ["1804.00874"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_19387"} +{"question": "Which work refines adversarial perturbations iteratively to increase its efficacy?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1706.06083"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_19388"} +{"question": "Which studies attempted to reversely infer camera extrinsics from the built neural implicit field?", "answer": ["iNeRF: Inverting Neural Radiance Fields for Pose Estimation", "NeRF-⁣-: Neural Radiance Fields Without Known Camera Parameters"], "answer_arxiv_id": ["2012.05877", "2102.07064"], "source_meta": {"published_time": "20231114"}, "qid": "AutoScholarQuery_train_19389"} +{"question": "Could you show me the works developed methods for estimating transferability through empirical estimation of the joint distribution of pseudo-source labels and target labels?", "answer": ["LEEP: A New Measure to Evaluate Transferability of Learned Representations"], "answer_arxiv_id": ["2002.12462"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_19390"} +{"question": "What are the studies carried out in multi-modal generative modeling for improving training and sampling techniques of diffusion models?", "answer": ["Improved Denoising Diffusion Probabilistic Models", "Maximum Likelihood Training of Score-Based Diffusion Models", "Variational Diffusion Models", "Score-based Generative Modeling in Latent Space", "EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations", "Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models", "Maximum Likelihood Training for Score-Based Diffusion ODEs by High-Order Denoising Score Matching", "Elucidating the Design Space of Diffusion-Based Generative Models"], "answer_arxiv_id": ["2102.09672", "2101.09258", "2107.00630", "2106.05931", "2207.06635", "2206.07309", "2206.08265", "2206.00364"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_19391"} +{"question": "What papers about developing algorithms for the approximate constrained policy search?", "answer": ["Constrained Policy Optimization", "Convergent Policy Optimization for Safe Reinforcement Learning", "Projection-Based Constrained Policy Optimization", "First Order Constrained Optimization in Policy Space", "CUP: A Conservative Update Policy Algorithm for Safe Reinforcement Learning"], "answer_arxiv_id": ["1705.10528", "1910.12156", "2010.03152", "2002.06506", "2202.07565"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_19392"} +{"question": "Which researches have focused on supervised pretraining for learning useful representations of molecules?", "answer": ["Strategies for Pre-training Graph Neural Networks", "Does GNN Pretraining Help Molecular Representation?"], "answer_arxiv_id": ["1905.12265", "2207.06010"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_19393"} +{"question": "Which papers focus on developing advanced prompting strategies for language models?", "answer": ["Making Pre-trained Language Models Better Few-shot Learners", "What Makes Good In-Context Examples for GPT-3?", "An Information-theoretic Approach to Prompt Engineering Without Ground Truth Labels", "Active Example Selection for In-Context Learning", "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models"], "answer_arxiv_id": ["2012.15723", "2101.06804", "2203.11364", "2211.04486", "2201.11903"], "source_meta": {"published_time": "20230117"}, "qid": "AutoScholarQuery_train_19394"} +{"question": "Which prior real-world datasets suffer from context dependence and lack sufficient annotations for intermediate steps in fact-checking?", "answer": ["Explainable Automated Fact-Checking: A Survey"], "answer_arxiv_id": ["2011.03870"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_19395"} +{"question": "Which work discussed the sample complexity of achieving an ϵ-stationary point in double-loop algorithms?", "answer": ["Approximation Methods for Bilevel Programming"], "answer_arxiv_id": ["1802.02246"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19396"} +{"question": "Which papers introduce methods on reducing computational cost and improve the segmentation efficiency for Efficient VSS?", "answer": ["Accel: A Corrective Fusion Network for Efficient Semantic Segmentation\n on Video", "Clockwork Convnets for Video Semantic Segmentation", "GSVNet: Guided Spatially-Varying Convolution for Fast Semantic\n Segmentation on Video", "Temporally Distributed Networks for Fast Video Semantic Segmentation", "Low-Latency Video Semantic Segmentation", "Dynamic Video Segmentation Network", "Deep Feature Flow for Video Recognition"], "answer_arxiv_id": ["1807.06667", "1608.03609", "2103.08834", "2004.01800", "1804.00389", "1804.00931", "1611.07715"], "source_meta": {"published_time": "20240127"}, "qid": "AutoScholarQuery_train_19397"} +{"question": "Any works showcasing the use of temporal structure of data available in reinforcement learning to learn a disentangled representation online with the reinforcement learning policy?", "answer": ["DARLA: Improving Zero-Shot Transfer in Reinforcement Learning"], "answer_arxiv_id": ["1707.08475"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_19398"} +{"question": "Which studies show a shift towards joint training of captioning and localization modules in dense video captioning?", "answer": ["Bidirectional Attentive Fusion with Context Gating for Dense Video\n Captioning", "End-to-End Dense Video Captioning with Masked Transformer", "Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense\n Video Captioning"], "answer_arxiv_id": ["1804.00100", "1804.00819", "2302.14115"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19399"} +{"question": "Which research proposed to learn segmentation masks from cheaper forms of supervision?", "answer": ["Weakly- and Semi-Supervised Learning of a Deep Convolutional Network for Semantic Image Segmentation", "BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation", "ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation"], "answer_arxiv_id": ["1502.02734", "1503.01640", "1604.05144v1"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_19400"} +{"question": "Can you name the studies that extend text-to-video capabilities to image-to-video?", "answer": ["VideoComposer: Compositional Video Synthesis with Motion Controllability", "AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "VideoCrafter1: Open Diffusion Models for High-Quality Video Generation"], "answer_arxiv_id": ["2306.02018", "2307.04725", "2310.19512"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_19401"} +{"question": "Which papers developed generative model-based methods to address the ill-posed nature of the SR problem?", "answer": ["SRFlow: Learning the Super-Resolution Space with Normalizing Flow", "ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks", "RankSRGAN: Generative Adversarial Networks with Ranker for Image\n Super-Resolution", "SRDiff: Single Image Super-Resolution with Diffusion Probabilistic\n Models", "Hierarchical Conditional Flow: A Unified Framework for Image\n Super-Resolution and Image Rescaling", "Photo-Realistic Single Image Super-Resolution Using a Generative\n Adversarial Network", "Normalizing Flow as a Flexible Fidelity Objective for Photo-Realistic\n Super-resolution", "To learn image super-resolution, use a GAN to learn how to do image\n degradation first", "EDFace-Celeb-1M: Benchmarking Face Hallucination with a Million-scale\n Dataset", "A Survey of Deep Face Restoration: Denoise, Super-Resolution, Deblur,\n Artifact Removal"], "answer_arxiv_id": ["2006.14200", "1809.00219", "1908.06382", "2104.14951", "2108.05301", "1609.04802", "2111.03649", "1807.11458", "2110.05031", "2211.02831"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_19402"} +{"question": "What paper proposed a general and effective recipe on the topic of video VLMs?", "answer": ["VindLU: A Recipe for Effective Video-and-Language Pretraining"], "answer_arxiv_id": ["2212.05051"], "source_meta": {"published_time": "20230405"}, "qid": "AutoScholarQuery_train_19403"} +{"question": "What research papers involve methods for graph signal sampling and matrix completion together?", "answer": ["Kernel-based Reconstruction of Graph Signals", "Song recommendation with Non-Negative Matrix factorization and graph total variation", "Temporal Graph Signal Decomposition"], "answer_arxiv_id": ["1605.07174", "1601.01892", "2106.13517"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_19404"} +{"question": "What papers presented methods for learning CSG representations in geometric deep learning?", "answer": ["JoinABLe: Learning Bottom-up Assembly of Parametric CAD Joints", "SketchGen: Generating Constrained CAD Sketches"], "answer_arxiv_id": ["2111.12772", "2106.02711"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_19405"} +{"question": "What papers focus on the use of LLMs in predicting mental health conditions from existing textual data?", "answer": ["Empowering Psychotherapy with Large Language Models: Cognitive\n Distortion Detection through Diagnosis of Thought Prompting"], "answer_arxiv_id": ["2310.07146"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_19406"} +{"question": "What works extended the approaches to general function approximations for Markov games?", "answer": ["The Power of Exploiter: Provable Multi-Agent RL in Large State Spaces", "Towards General Function Approximation in Zero-Sum Markov Games", "A Self-Play Posterior Sampling Algorithm for Zero-Sum Markov Games"], "answer_arxiv_id": ["2106.03352", "2107.14702", "2210.01907"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_19407"} +{"question": "What research papers established federated learning as a decentralized machine learning framework that provides data confidentiality by retaining clients’ raw data on edge devices?", "answer": ["Federated Learning: Strategies for Improving Communication Efficiency", "Communication-Efficient Learning of Deep Networks from Decentralized Data", "Advances and Open Problems in Federated Learning"], "answer_arxiv_id": ["1610.05492", "1602.05629", "1912.04977"], "source_meta": {"published_time": "20230404"}, "qid": "AutoScholarQuery_train_19408"} +{"question": "What works about two-stage technique used in 3D Human Pose Estimation?", "answer": ["Semantic Graph Convolutional Networks for 3D Human Pose Regression", "3D human pose estimation in video with temporal convolutions and\n semi-supervised training", "3D Human Pose Estimation with Spatial and Temporal Transformers", "MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose\n Estimation in Video", "PoseFormerV2: Exploring Frequency Domain for Efficient and Robust 3D\n Human Pose Estimation", "MotionBERT: A Unified Perspective on Learning Human Motion\n Representations"], "answer_arxiv_id": ["1904.03345", "1811.11742", "2103.10455", "2203.00859", "2303.17472", "2210.06551"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_19409"} +{"question": "What model synthesizes an image conditioned on a text embedding provided by a trained language encoder?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2205.11487"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_19410"} +{"question": "Are there any publications that extend the basic paradigm of CLIP to tasks like semantic segmentation and object detection?", "answer": ["Language-driven Semantic Segmentation", "ActionCLIP: A New Paradigm for Video Action Recognition"], "answer_arxiv_id": ["2201.03546", "2109.08472"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_19411"} +{"question": "Which works are about non-parametric methods of generating new tileable images?", "answer": ["Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying\n Lighting and SVBRDF from a Single Image"], "answer_arxiv_id": ["1905.02722"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_19412"} +{"question": "What works tackled class-imbalanced learning through the method of transfer learning?", "answer": ["Large-Scale Long-Tailed Recognition in an Open World", "Feature Transfer Learning for Face Recognition with Under-Represented Data", "M2m: Imbalanced Classification via Major-to-minor Translation", "Feature Space Augmentation for Long-Tailed Data", "Deep Representation Learning on Long-tailed Data: A Learnable Embedding Augmentation Perspective", "Meta Feature Modulator for Long-tailed Recognition"], "answer_arxiv_id": ["1904.05160", "1803.09014", "2004.00431", "2008.03673", "2002.10826", "2008.03428"], "source_meta": {"published_time": "20220728"}, "qid": "AutoScholarQuery_train_19413"} +{"question": "Which papers propose recent contributions such as weight mirrors and stochastic approximation to tackle the weight transport problem?", "answer": ["Deep Learning without Weight Transport", "Towards Scaling Difference Target Propagation by Learning Backprop Targets"], "answer_arxiv_id": ["1904.05391", "2201.13415"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19414"} +{"question": "Which paper constructs an aggregated independence test using Gaussian kernels?", "answer": ["Adaptive test of independence based on HSIC measures"], "answer_arxiv_id": ["1902.06441v5"], "source_meta": {"published_time": "20211028"}, "qid": "AutoScholarQuery_train_19415"} +{"question": "Which works have introduced classifier guidance for improving the quality of images and retrieving specific class of images?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Generating High Fidelity Data from Low-density Regions using Diffusion Models", "Blended Diffusion for Text-driven Editing of Natural Images", "More Control for Free! Image Synthesis with Semantic Diffusion Guidance", "GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models"], "answer_arxiv_id": ["2105.05233", "2203.17260", "2111.14818", "2112.05744", "2112.10741"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_19416"} +{"question": "Could you provide me the study that presented competitive results in a wide range of tasks by emerging LLMs?", "answer": ["A Survey of Large Language Models"], "answer_arxiv_id": ["2303.18223"], "source_meta": {"published_time": "20240807"}, "qid": "AutoScholarQuery_train_19417"} +{"question": "In what works the equivalence of PFR construction to a certain variant of A* sampling was noted?", "answer": ["Fast Relative Entropy Coding with A* coding"], "answer_arxiv_id": ["2201.12857"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19418"} +{"question": "Which works are about the Self-Training method in semi-supervised learning?", "answer": ["MixMatch: A Holistic Approach to Semi-Supervised Learning", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["1905.02249", "2001.07685v2"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_19419"} +{"question": "Any works incorporated prior knowledge in a diffusion model to render object geometries for object reconstruction?", "answer": ["Diffusion-Guided Reconstruction of Everyday Hand-Object Interaction\n Clips"], "answer_arxiv_id": ["2309.05663"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19420"} +{"question": "Could you name some studies that started exploring visual large language models?", "answer": ["BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Visual Instruction Tuning", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "PandaGPT: One Model To Instruction-Follow Them All"], "answer_arxiv_id": ["2301.12597", "2304.08485", "2305.11175", "2305.16355"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_19421"} +{"question": "Which research papers presented strategies for differentiation of persistent homology-based functions?", "answer": ["Optimizing persistent homology based functions", "A Framework for Differential Calculus on Persistence Barcodes"], "answer_arxiv_id": ["2010.08356", "1910.00960"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19422"} +{"question": "Are there any works using spectral filters in the form of polynomials that revealed insights about the learned robust representation?", "answer": ["Convolutional Neural Networks on Graphs with Chebyshev Approximation, Revisited", "Adaptive Universal Generalized PageRank Graph Neural Network"], "answer_arxiv_id": ["2202.03580", "2006.07988"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_19423"} +{"question": "What works considers Symbolic Regression as a natural language processing task?", "answer": ["Neural Symbolic Regression that Scales", "SymbolicGPT: A Generative Transformer Model for Symbolic Regression", "Deep Symbolic Regression for Recurrent Sequences", "End-to-end symbolic regression with transformers", "SymFormer: End-to-end symbolic regression using transformer-based architecture", "Symbolic Expression Transformer: A Computer Vision Approach for Symbolic Regression", "Discovering ordinary differential equations that govern time-series"], "answer_arxiv_id": ["2106.06427", "2106.14131", "2201.04600", "2204.10532", "2205.15764", "2205.11798", "2211.02830"], "source_meta": {"published_time": "20230420"}, "qid": "AutoScholarQuery_train_19424"} +{"question": "What papers considered local context and prioritized various aspects in point-based methods?", "answer": ["PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on\n Point Clouds", "Adaptive Graph Convolution for Point Cloud Analysis", "Walk in the Cloud: Learning Curves for Point Clouds Shape Analysis"], "answer_arxiv_id": ["1706.02413", "2103.14635", "2108.08035", "2105.01288"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_19425"} +{"question": "Which works study the model-based setting where a learned model is employed to train a control policy?", "answer": ["Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models", "When to Trust Your Model: Model-Based Policy Optimization", "Model Based Reinforcement Learning for Atari", "Conservative Dual Policy Optimization for Efficient Model-Based Reinforcement Learning", "Mastering Diverse Domains through World Models"], "answer_arxiv_id": ["1805.12114", "1906.08253", "1903.00374", "2209.07676", "2301.04104"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_19426"} +{"question": "Which paper provided evidence that directly generating annotations for training planning and grounding modules with Large Language Models (LLMs) could lead to numerous errors?", "answer": ["AgentBench: Evaluating LLMs as Agents"], "answer_arxiv_id": ["2308.03688"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_19427"} +{"question": "What is the key difference mentioned between implicit kernels and applying G𝐺G-MLPs?", "answer": ["A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups"], "answer_arxiv_id": ["2104.09459"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_19428"} +{"question": "Which works proposed a planar factorization to decompose 4D spatiotemporal volumes into six feature planes?", "answer": ["HexPlane: A Fast Representation for Dynamic Scenes", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance"], "answer_arxiv_id": ["2301.09632", "2301.10241"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_19429"} +{"question": "Which works applied graph neural networks in an Encode-Process-Decode pipeline for learning interactions between particles in fluid mechanics?", "answer": ["Learning to Simulate Complex Physics with Graph Networks"], "answer_arxiv_id": ["2002.09405"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_19430"} +{"question": "What papers discuss the exposure bias or covariate shift problem in sequential prediction tasks?", "answer": ["Sequence Level Training with Recurrent Neural Networks", "Feedback in Imitation Learning: The Three Regimes of Covariate Shift"], "answer_arxiv_id": ["1511.06732", "2102.02872"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_19431"} +{"question": "Which study introduced a Transformer architecture to graph domain with Laplacian Positional Encoding?", "answer": ["A Generalization of Transformer Networks to Graphs"], "answer_arxiv_id": ["2012.09699"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_19432"} +{"question": "What paper introduced the KaggleDBQA dataset?", "answer": ["KaggleDBQA: Realistic Evaluation of Text-to-SQL Parsers"], "answer_arxiv_id": ["2106.11455"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_19433"} +{"question": "Any studies on incorporating multi-domain generation capabilities into a single generator similar to DoRM?", "answer": ["DynaGAN: Dynamic Few-shot Adaptation of GANs to Multiple Domains", "HyperDomainNet: Universal Domain Adaptation for Generative Adversarial Networks"], "answer_arxiv_id": ["2211.14554", "2210.08884"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_19434"} +{"question": "In pretraining methods for RL, which studies have proposed online pretraining systems?", "answer": ["Pretraining in Deep Reinforcement Learning: A Survey"], "answer_arxiv_id": ["2211.03959"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_19435"} +{"question": "What works propose approaches that build a large-scale concept dictionary for vision-language models?", "answer": ["DetCLIP: Dictionary-Enriched Visual-Concept Paralleled Pre-training for Open-world Detection"], "answer_arxiv_id": ["2209.09407"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_19436"} +{"question": "Could you provide me some studies about the difficulties of neural machine translation for code?", "answer": ["Tree-to-tree Neural Networks for Program Translation"], "answer_arxiv_id": ["1802.03691"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_19437"} +{"question": "Which work presents a method to generate test cases for dynamically typed languages like Python?", "answer": ["Pynguin: Automated Unit Test Generation for Python"], "answer_arxiv_id": ["2202.05218"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_19438"} +{"question": "Can you provide some references in which label correction methods are employed that use a pseudo-labeling strategy?", "answer": ["Probabilistic End-to-end Noise Correction for Learning with Noisy Labels", "Early-Learning Regularization Prevents Memorization of Noisy Labels", "DivideMix: Learning with Noisy Labels as Semi-supervised Learning"], "answer_arxiv_id": ["1903.07788", "2007.00151", "2002.07394"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_19439"} +{"question": "Which papers studied CLIP-like models with language image pre-training that showed exceptional effective robustness?", "answer": ["Learning Visual Representations with Caption Annotations", "Contrastive Learning of Medical Visual Representations from Paired Images and Text", "VirTex: Learning Visual Representations from Textual Annotations", "Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision"], "answer_arxiv_id": ["2008.01392", "2010.00747", "2006.06666", "2103.00020", "2102.05918"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19440"} +{"question": "Which works focus on Text-to-image generation with the help of pre-trained GAN-based methods?", "answer": ["Generative Adversarial Text to Image Synthesis", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image\n Synthesis", "DF-GAN: A Simple and Effective Baseline for Text-to-Image Synthesis", "AttnGAN: Fine-Grained Text to Image Generation with Attentional\n Generative Adversarial Networks", "VQGAN-CLIP: Open Domain Image Generation and Editing with Natural\n Language Guidance"], "answer_arxiv_id": ["1605.05396", "1904.01310", "2008.05865", "1711.10485", "2204.08583"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_19441"} +{"question": "Which works propose the approach signaling the durable learning as independent branches in object and relation feature learning?", "answer": ["RelTR: Relation Transformer for Scene Graph Generation"], "answer_arxiv_id": ["2201.11460"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_19442"} +{"question": "Which papers aimed to use Graph Neural Networks for propagating information among neighboring geographical regions for traffic pattern representations?", "answer": ["Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting", "Traffic Flow Forecasting with Spatial-Temporal Graph Diffusion Network", "Spatial-Temporal Fusion Graph Neural Networks for Traffic Flow Forecasting"], "answer_arxiv_id": ["1707.01926v3", "2110.04038", "2012.09641"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_19443"} +{"question": "Any works about heuristic approaches to discover common regions in images?", "answer": ["ReCo: Retrieve and Co-segment for Zero-shot Transfer"], "answer_arxiv_id": ["2206.07045"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_19444"} +{"question": "Could you provide studies that introduced techniques that improve GNN's structure-awareness and alleviate the oversmoothing problem?", "answer": ["Representation Learning on Graphs with Jumping Knowledge Networks", "Simple and Deep Graph Convolutional Networks", "On the Universality of Graph Neural Networks on Large Random Graphs"], "answer_arxiv_id": ["1806.03536", "2007.02133", "2105.13099"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19445"} +{"question": "What studies utilized a hierarchy of differential equations to compute small feature learning corrections to the kernel through training?", "answer": ["Dynamics of Deep Neural Networks and Neural Tangent Hierarchy", "Asymptotics of Wide Networks from Feynman Diagrams", "Asymptotics of Wide Convolutional Neural Networks", "The Principles of Deep Learning Theory"], "answer_arxiv_id": ["1909.08156", "1909.11304", "2008.08675", "2106.10165"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_19446"} +{"question": "What papers studied the impact aspects of in-context learning, such as input-output mapping and template format?", "answer": ["What In-Context Learning \"Learns\" In-Context: Disentangling Task\n Recognition and Task Learning", "Rethinking the Role of Demonstrations: What Makes In-Context Learning\n Work?", "Ground-Truth Labels Matter: A Deeper Look into Input-Label\n Demonstrations"], "answer_arxiv_id": ["2305.09731", "2202.12837", "2205.12685"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_19447"} +{"question": "What studies have explored the usage of free-form explanations and rationales as a viable approach in examining reasoning capabilities of Language Models?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Models are Zero-Shot Reasoners", "Unsupervised Commonsense Question Answering with Self-Talk", "DREAM: Improving Situational QA by First Elaborating the Situation"], "answer_arxiv_id": ["2201.11903", "2205.11916", "2004.05483", "2112.08656"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_19448"} +{"question": "In what studies do researchers parallelize the model updates by imposing coordinate update rules?", "answer": ["An Asynchronous Parallel Stochastic Coordinate Descent Algorithm", "Asynchronous Stochastic Coordinate Descent: Parallelism and Convergence Properties"], "answer_arxiv_id": ["1311.1873", "1403.3862"], "source_meta": {"published_time": "20221116"}, "qid": "AutoScholarQuery_train_19449"} +{"question": "What works propose to use multi-scale feature grids for rendering at different resolution or distance similar to the mipmap algorithm in mesh texture rendering?", "answer": ["Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural\n Radiance Fields", "Mip-Grid: Anti-aliased Grid Representations for Neural Radiance Fields"], "answer_arxiv_id": ["2307.11335", "2402.14196"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_19450"} +{"question": "What studies have applied influence functions in natural language processing?", "answer": ["Explaining Black Box Predictions and Unveiling Data Artifacts through Influence Functions"], "answer_arxiv_id": ["2005.06676"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_19451"} +{"question": "Could you provide me a work about treatment effect estimation using an ensemble of deep network-based instrumental variable estimators?", "answer": ["Valid Causal Inference with (Some) Invalid Instruments"], "answer_arxiv_id": ["2006.11386"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_19452"} +{"question": "Who proposed Rotary, a new method for encoding positional information by rotating the hidden representations in Transformer models?", "answer": ["RoFormer: Enhanced Transformer with Rotary Position Embedding"], "answer_arxiv_id": ["2104.09864"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19453"} +{"question": "Could you mention some works that are related to the text-to-image generation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "CLIP-Event: Connecting Text and Images with Event Structures"], "answer_arxiv_id": ["2103.00020", "2201.05078"], "source_meta": {"published_time": "20240310"}, "qid": "AutoScholarQuery_train_19454"} +{"question": "Could you provide me some works on prototype-based methods in 2D and 3D detection?", "answer": ["Universal-Prototype Enhancing for Few-Shot Object Detection", "Breaking Immutable: Information-Coupled Prototype Elaboration for\n Few-Shot Object Detection", "Prototypical VoteNet for Few-Shot 3D Point Cloud Object Detection", "GPA-3D: Geometry-aware Prototype Alignment for Unsupervised Domain\n Adaptive 3D Object Detection from Point Clouds", "CL3D: Unsupervised Domain Adaptation for Cross-LiDAR 3D Detection"], "answer_arxiv_id": ["2103.01077", "2211.14782", "2210.05593", "2308.08140", "2212.00244"], "source_meta": {"published_time": "20240425"}, "qid": "AutoScholarQuery_train_19455"} +{"question": "Which studies have applied inverse reinforcement learning to infer and model human preferences?", "answer": ["Deep Reinforcement Learning from Human Preferences", "Recursively Summarizing Books with Human Feedback", "Improving a Neural Semantic Parser by Counterfactual Learning from Human Bandit Feedback", "Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["1706.03741", "2109.10862", "1805.01252", "1909.08593"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_19456"} +{"question": "What works shed light on optimizing causal effect in recommendation systems?", "answer": ["A Survey on Causal Inference for Recommendation"], "answer_arxiv_id": ["2303.11666"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_19457"} +{"question": "Could you list some studies where low-rank matrix factorization has been used in Multi-View Clustering?", "answer": ["A Closed Form Solution to Multi-View Low-Rank Regression"], "answer_arxiv_id": ["1610.04668v1"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_19458"} +{"question": "Can you point out the studies that have proposed the use of polyhedrons, e.g., octahedron and icosahedron, to represent panoramas?", "answer": ["SpherePHD: Applying CNNs on a Spherical PolyHeDron Representation of 360\n degree Images", "Equivariant Networks for Pixelized Spheres"], "answer_arxiv_id": ["1811.08196", "2106.06662"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_19459"} +{"question": "Could you provide me some works that have incorporated the decomposition of training loss trajectory in gradient-based methods?", "answer": ["Estimating Training Data Influence by Tracing Gradient Descent", "TracInAD: Measuring Influence for Anomaly Detection", "Data Cleansing for Models Trained with SGD", "Gradient-Based Automated Iterative Recovery for Parameter-Efficient Tuning"], "answer_arxiv_id": ["2002.08484", "2205.01362v4", "1906.08473", "2302.06598"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_19460"} +{"question": "Which papers show that end-to-end training can lead to better feature optimization?", "answer": ["Is Someone Speaking? Exploring Long-term Temporal Features for\n Audio-visual Active Speaker Detection", "End-to-End Active Speaker Detection"], "answer_arxiv_id": ["2107.06592", "2203.14250"], "source_meta": {"published_time": "20230119"}, "qid": "AutoScholarQuery_train_19461"} +{"question": "In what works the researcher studied the decentralized stochastic bilevel optimization problem, considering both local and global lower problems?", "answer": ["Decentralized Gossip-Based Stochastic Bilevel Optimization over Communication Networks", "Decentralized Bilevel Optimization", "Decentralized Stochastic Bilevel Optimization with Improved per-Iteration Complexity"], "answer_arxiv_id": ["2206.10870", "2206.05670", "2210.12839"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_19462"} +{"question": "Can you mention some works that have investigated the domain generalization problem?", "answer": ["Generalizing to Unseen Domains: A Survey on Domain Generalization", "Domain Generalization: A Survey"], "answer_arxiv_id": ["2103.03097", "2103.02503"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_19463"} +{"question": "Which papers have revealed the limitation of contrastive pretraining on fine-grained compositional understanding, verb understanding, and temporal reasoning?", "answer": ["Probing Image-Language Transformers for Verb Understanding", "When and Why Vision-Language Models Behave like Bags-Of-Words, and What to Do About It?", "Revealing Single Frame Bias for Video-and-Language Learning", "Test of Time: Instilling Video-Language Models with a Sense of Time"], "answer_arxiv_id": ["2106.09141", "2210.01936", "2206.03428", "2301.02074"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19464"} +{"question": "Could you mention works for long-sequence processing in transformers that uses chunking methods?", "answer": ["Efficient Long-Text Understanding with Short-Text Models", "Leveraging Locality in Abstractive Text Summarization", "Unlimiformer: Long-Range Transformers with Unlimited Length Input"], "answer_arxiv_id": ["2208.00748", "2205.12476", "2305.01625"], "source_meta": {"published_time": "20230825"}, "qid": "AutoScholarQuery_train_19465"} +{"question": "Could you list some studies focusing on enhancing NeRF with low-quality input images?", "answer": ["NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo\n Collections", "NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw\n Images", "HDR-NeRF: High Dynamic Range Neural Radiance Fields"], "answer_arxiv_id": ["2008.02268", "2111.13679", "2111.14451"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_19466"} +{"question": "Which studies employ techniques like temperature sampling, top-k sampling and top-p sampling to enhance diversity?", "answer": ["Hierarchical Neural Story Generation", "The Curious Case of Neural Text Degeneration"], "answer_arxiv_id": ["1805.04833", "1904.09751"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_19467"} +{"question": "Could you provide me some works on hard prompt tuning methodology, particularly the one involves manually designing?", "answer": ["CoTDet: Affordance Knowledge Prompting for Task Driven Object Detection", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2309.01093", "2103.00020"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_19468"} +{"question": "Can you provide me with some works that incorporate clusters or prototypes into contrastive learning (CL)?", "answer": ["Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Unsupervised Deep Learning by Neighbourhood Discovery", "With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations", "Prototypical Contrastive Learning of Unsupervised Representations"], "answer_arxiv_id": ["2006.09882", "1904.11567", "2104.14548", "2005.04966"], "source_meta": {"published_time": "20220716"}, "qid": "AutoScholarQuery_train_19469"} +{"question": "What papers have used VLMs for video tasks primarily focusing on zero-shot verb recognition?", "answer": ["ActionCLIP: A New Paradigm for Video Action Recognition", "Prompting Visual-Language Models for Efficient Video Understanding", "Vita-CLIP: Video and text adaptive CLIP via Multimodal Prompting"], "answer_arxiv_id": ["2109.08472", "2112.04478", "2304.03307"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_19470"} +{"question": "Which papers have investigated the depth of cognitive processing in LLMs?", "answer": ["Mind meets machine: Unravelling GPT-4's cognitive psychology", "Cognitive Overload: Jailbreaking Large Language Models with Overloaded\n Logical Thinking"], "answer_arxiv_id": ["2303.11436", "2311.09827"], "source_meta": {"published_time": "20240122"}, "qid": "AutoScholarQuery_train_19471"} +{"question": "Which studies propose to evaluate fairness by generating counterfactual samples?", "answer": ["Image Counterfactual Sensitivity Analysis for Detecting Unintended Bias", "Gender Slopes: Counterfactual Fairness for Computer Vision Models by\n Attribute Manipulation", "Evaluating and Mitigating Bias in Image Classifiers: A Causal\n Perspective Using Counterfactuals"], "answer_arxiv_id": ["1906.06439", "2005.10430", "2009.08270"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_19472"} +{"question": "What works focus on transferring knowledge of large pre-trained networks based on CL to lightweight ConvNets in computer vision?", "answer": ["SEED: Self-supervised Distillation For Visual Representation", "CompRess: Self-Supervised Learning by Compressing Representations"], "answer_arxiv_id": ["2101.04731", "2010.14713"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_19473"} +{"question": "What researches are related to the usage of language models in machine translation?", "answer": ["Unsupervised neural machine translation with generative language models only"], "answer_arxiv_id": ["2110.05448"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_19474"} +{"question": "Which studies have performed minimax optimization in the nonconvex–(strongly)-concave case?", "answer": ["On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems", "Efficient Algorithms for Smooth Minimax Optimization"], "answer_arxiv_id": ["1906.00331", "1907.01543"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_19475"} +{"question": "What researches have used generative deep learning techniques like GANs for protein design?", "answer": ["Generative Adversarial Networks", "Computational Protein Design with Deep Learning Neural Networks", "Conditional Antibody Design as 3D Equivariant Graph Translation", "Protein Structure and Sequence Generation with Equivariant Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2203.00667", "1801.07130v2", "2208.06073", "2205.15019"], "source_meta": {"published_time": "20221005"}, "qid": "AutoScholarQuery_train_19476"} +{"question": "What research has been done on improving the reasoning and robustness of LLMs in the context of hate speech detection?", "answer": ["Probing LLMs for hate speech detection: strengths and vulnerabilities"], "answer_arxiv_id": ["2310.12860"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_19477"} +{"question": "What research has been done on partial interventional identification of continuous outcomes?", "answer": ["Quantifying Ignorance in Individual-Level Causal-Effect Estimates under Hidden Confounding", "Scalable Sensitivity and Uncertainty Analyses for Causal-Effect Estimates of Continuous-Valued Interventions"], "answer_arxiv_id": ["2103.04850", "2204.10022"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_19478"} +{"question": "Any works about Offline RL?", "answer": ["When Should We Prefer Offline Reinforcement Learning Over Behavioral Cloning?", "Conservative Q-Learning for Offline Reinforcement Learning", "Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["2204.05618", "2006.04779", "2106.01345"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_19479"} +{"question": "Which paper presented the prompt-based editor know as IKE?", "answer": ["Can We Edit Factual Knowledge by In-Context Learning?"], "answer_arxiv_id": ["2305.12740"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_19480"} +{"question": "What studies confirmed that explicit GR enhanced the generalization performance?", "answer": ["Information-Theoretic Local Minima Characterization and Regularization"], "answer_arxiv_id": ["1911.08192"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_19481"} +{"question": "Who has proposed to estimate the energy-based prior model based on the prior of a pre-trained VAE by noise contrastive estimation?", "answer": ["A Contrastive Learning Approach for Training Variational Autoencoder Priors"], "answer_arxiv_id": ["2010.02917"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_19482"} +{"question": "Which works are based on instance discrimination in self-supervised pre-training?", "answer": ["Unsupervised Feature Learning via Non-Parametric Instance-level\n Discrimination", "A Simple Framework for Contrastive Learning of Visual Representations", "Bootstrap your own latent: A new approach to self-supervised Learning", "Unsupervised Learning of Visual Features by Contrasting Cluster\n Assignments", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["1805.01978", "2002.05709", "2006.07733", "2006.09882", "2104.14294"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_19483"} +{"question": "Which previous research has concentrated on sentence or document level for FSA dataset?", "answer": ["Good Debt or Bad Debt: Detecting Semantic Orientations in Economic Texts", "SEntFiN 1.0: Entity-Aware Sentiment Analysis for Financial News"], "answer_arxiv_id": ["1307.5336", "2305.12257"], "source_meta": {"published_time": "20240408"}, "qid": "AutoScholarQuery_train_19484"} +{"question": "Could you provide me examples of research that theoretically characterize the expressive power of subgraph GNNs?", "answer": ["Understanding and Extending Subgraph GNNs by Rethinking Their Symmetries", "A Complete Expressiveness Hierarchy for Subgraph GNNs via Subgraph Weisfeiler-Lehman Tests"], "answer_arxiv_id": ["2206.11140", "2302.07090"], "source_meta": {"published_time": "20230910"}, "qid": "AutoScholarQuery_train_19485"} +{"question": "Which works are the closest to this setting in terms of employing Wasserstein penalization and constraint?", "answer": ["Certifying Some Distributional Robustness with Principled Adversarial Training", "On the regularized risk of distributionally robust learning over deep neural networks"], "answer_arxiv_id": ["1710.10571v5", "2109.06294"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_19486"} +{"question": "What studies have utilized the JFT-300M dataset for pre-training in large-scale models?", "answer": ["Revisiting Unreasonable Effectiveness of Data in Deep Learning Era"], "answer_arxiv_id": ["1707.02968"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_19487"} +{"question": "Could you provide me some works related to 3D shape understanding using CLIP?", "answer": ["PointCLIP: Point Cloud Understanding by CLIP", "ULIP: Learning a Unified Representation of Language, Images, and Point\n Clouds for 3D Understanding", "ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding", "Contrast with Reconstruct: Contrastive 3D Representation Learning Guided\n by Generative Pretraining", "CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D\n Recognition", "ViT-Lens: Towards Omni-modal Representations", "Uni3D: Exploring Unified 3D Representation at Scale"], "answer_arxiv_id": ["2112.02413", "2212.05171", "2305.08275", "2302.02318", "2303.11313", "2311.16081", "2310.06773"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_19488"} +{"question": "What is the study that presented PanoStretch that randomly adjusts layout aspect ratio?", "answer": ["HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch\n Data Augmentation"], "answer_arxiv_id": ["1901.03861"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19489"} +{"question": "Could you provide me some works that provide classic board game suits in RL environment libraries?", "answer": ["OpenSpiel: A Framework for Reinforcement Learning in Games", "PettingZoo: Gym for Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["1908.09453", "2009.14471v7"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_19490"} +{"question": "Which works used node embeddings in structure-learning modules?", "answer": ["Graph WaveNet for Deep Spatial-Temporal Graph Modeling", "Discrete Graph Structure Learning for Forecasting Multiple Time Series", "Graph Neural Network-Based Anomaly Detection in Multivariate Time Series"], "answer_arxiv_id": ["1906.00121", "2101.06861", "2106.06947"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_19491"} +{"question": "Could you name studies that worked on retrieval-augmented image captioning?", "answer": ["SmallCap: Lightweight Image Captioning Prompted with Retrieval\n Augmentation", "Retrieval-Augmented Multimodal Language Modeling", "Retrieval-augmented Image Captioning", "Re-ViLM: Retrieval-Augmented Visual Language Model for Zero and Few-Shot\n Image Captioning"], "answer_arxiv_id": ["2209.15323", "2211.12561", "2302.08268", "2302.04858"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19492"} +{"question": "Which paper proposed WAT framework considering the worst-class robust risk?", "answer": ["WAT: Improve the Worst-class Robustness in Adversarial Training"], "answer_arxiv_id": ["2302.04025"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19493"} +{"question": "What studies proposed 3D texture-based attacks to improve the robustness?", "answer": ["Dual Attention Suppression Attack: Generate Adversarial Camouflage in\n Physical World", "FCA: Learning a 3D Full-coverage Vehicle Camouflage for Multi-view\n Physical Adversarial Attack", "DTA: Physical Camouflage Attacks using Differentiable Transformation\n Network", "ACTIVE: Towards Highly Transferable 3D Physical Camouflage for Universal\n and Robust Vehicle Evasion"], "answer_arxiv_id": ["2103.01050", "2109.07193", "2203.09831", "2308.07009"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_19494"} +{"question": "Which research works achieved impressive results in 2D human generation by scaling up the dataset?", "answer": ["StyleGAN-Human: A Data-Centric Odyssey of Human Generation", "InsetGAN for Full-Body Image Generation"], "answer_arxiv_id": ["2204.11823", "2203.07293"], "source_meta": {"published_time": "20221010"}, "qid": "AutoScholarQuery_train_19495"} +{"question": "What works highlight the significance of the early phase of learning with SGD?", "answer": ["On the relation between the sharpest directions of DNN loss and the SGD step length", "The break-even point on optimization trajectories of deep neural networks"], "answer_arxiv_id": ["1807.05031", "2002.09572"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19496"} +{"question": "What work used Marionette for learning codebooks in object-centric learning?", "answer": ["MarioNette: Self-Supervised Sprite Learning"], "answer_arxiv_id": ["2104.14553"], "source_meta": {"published_time": "20230110"}, "qid": "AutoScholarQuery_train_19497"} +{"question": "Could you provide me some studies about compressing instructions into concise key-value attention prefixes?", "answer": ["Learning to Compress Prompts with Gist Tokens"], "answer_arxiv_id": ["2304.08467"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_19498"} +{"question": "In which studies the information-theoretic quantities are considered as notions of distributional stability?", "answer": ["Information-theoretic analysis of generalization capability of learning algorithms", "On Generalization Error Bounds of Noisy Gradient Methods for Non-Convex Learning", "Reasoning About Generalization via Conditional Mutual Information", "Stability Based Generalization Bounds for Exponential Family Langevin Dynamics"], "answer_arxiv_id": ["1705.07809", "1902.00621", "2001.09122", "2201.03064"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_19499"} +{"question": "Which studies proposed neural architecture search for automating identification of optimal architectures for specific tasks and datasets?", "answer": ["Neural Architecture Search with Reinforcement Learning", "DARTS: Differentiable Architecture Search", "Neural Architecture Optimization", "Large-Scale Evolution of Image Classifiers"], "answer_arxiv_id": ["1611.01578", "1806.09055", "1808.07233", "1703.01041"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_19500"} +{"question": "Which works looked into performance fairness aware federated learning, emphasizing a uniform distribution of accuracy among clients?", "answer": ["Agnostic Federated Learning", "Fair Resource Allocation in Federated Learning"], "answer_arxiv_id": ["1902.00146", "1905.10497"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_19501"} +{"question": "Which studies highlighted the role of LMs in proposing reasonable plans based on context and environment?", "answer": ["Inner Monologue: Embodied Reasoning through Planning with Language Models", "ReAct: Synergizing Reasoning and Acting in Language Models"], "answer_arxiv_id": ["2207.05608", "2210.03629"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19502"} +{"question": "What papers apply CLIP to object detection?", "answer": ["Open-Vocabulary Object Detection Using Captions", "Open-vocabulary Object Detection via Vision and Language Knowledge\n Distillation", "Detecting Twenty-thousand Classes using Image-level Supervision", "Open-Vocabulary DETR with Conditional Matching", "RegionCLIP: Region-based Language-Image Pretraining", "A Simple Framework for Open-Vocabulary Segmentation and Detection"], "answer_arxiv_id": ["2011.10678", "2104.13921", "2201.02605", "2203.11876", "2112.09106", "2303.08131"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_19503"} +{"question": "Can you list some works with a focus on using prompt engineering for synthetic image generation?", "answer": ["Fake it till you make it: Learning transferable representations from synthetic ImageNet clones", "Diversify Your Vision Datasets with Automatic Diffusion-Based\n Augmentation", "Is synthetic data from generative models ready for image recognition?"], "answer_arxiv_id": ["2212.08420v2", "2305.16289", "2210.07574"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19504"} +{"question": "What work demonstrated that the Sobolev-Slobodetskii norm is inadequate from a learning-theoretic perspective?", "answer": ["Functions with average smoothness: structure, algorithms, and learning"], "answer_arxiv_id": ["2007.06283"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_19505"} +{"question": "Which works proposed Moore-Lewis selection that selects examples with high cross-entropy difference?", "answer": ["Cynical Selection of Language Model Training Data", "Automatic Document Selection for Efficient Encoder Pretraining"], "answer_arxiv_id": ["1709.02279", "2210.10951"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19506"} +{"question": "What studies are about LLaMA?", "answer": ["LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2302.13971"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_19507"} +{"question": "Which research suggested the term oversmoothing in the context of GNNs?", "answer": ["Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning"], "answer_arxiv_id": ["1801.07606"], "source_meta": {"published_time": "20230312"}, "qid": "AutoScholarQuery_train_19508"} +{"question": "What is the research involved in using a transformer-based model with a graph-guided masked attention for training in graph structured data?", "answer": ["GraphCodeBERT: Pre-training Code Representations with Data Flow"], "answer_arxiv_id": ["2009.08366"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_19509"} +{"question": "Which works applied GAN-based frameworks for text-to-3D human generation?", "answer": ["EVA3D: Compositional 3D Human Generation from 2D Image Collections", "Efficient 3D Articulated Human Generation with Layered Surface Volumes", "GETAvatar: Generative Textured Meshes for Animatable Human Avatars", "Get3DHuman: Lifting StyleGAN-Human into a 3D Generative Model using\n Pixel-aligned Reconstruction Priors"], "answer_arxiv_id": ["2210.04888", "2307.05462", "2310.02714", "2302.01162"], "source_meta": {"published_time": "20231002"}, "qid": "AutoScholarQuery_train_19510"} +{"question": "Which papers display advances in diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Implicit Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["1503.03585", "2010.02502", "2011.13456"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_19511"} +{"question": "Could you provide some studies about reinforcement learning with human or model feedback (RLHF/RLAIF)?", "answer": ["Deep Reinforcement Learning from Human Preferences", "Scaling Laws for Reward Model Overoptimization"], "answer_arxiv_id": ["1706.03741", "2210.10760"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_19512"} +{"question": "Can you provide examples on existing adaptation methods that improve model robustness on out-of-distribution datasets?", "answer": ["Domain Generalization with MixStyle", "Domain Generalization via Model-Agnostic Learning of Semantic Features", "Learning to Generalize: Meta-Learning for Domain Generalization", "Deep Domain-Adversarial Image Generation for Domain Generalisation", "MEMO: Test Time Robustness via Adaptation and Augmentation", "Generative Interventions for Causal Learning", "Discrete Representations Strengthen Vision Transformer Robustness", "Tent: Fully Test-time Adaptation by Entropy Minimization"], "answer_arxiv_id": ["2104.02008", "1910.13580", "1710.03463", "2003.06054", "2110.09506", "2012.12265", "2111.10493", "2006.10726"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_19513"} +{"question": "Could you provide me some works focused on generating hand-object interactions?", "answer": ["Grasping Field: Learning Implicit Representations for Human Grasps", "ContactOpt: Optimizing Contact to Improve Grasps"], "answer_arxiv_id": ["2008.04451", "2104.07267"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_19514"} +{"question": "Which works have highlighted the emergence of diffusion models for generating synthetic samples?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Improved Techniques for Training Score-Based Generative Models", "Improved Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2112.10752", "2006.09011", "2102.09672"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_19515"} +{"question": "Can you tell me about the work that introduced covariance amplitude and lengthscale variation in spatial non-stationary GPs?", "answer": ["Non-Stationary Gaussian Process Regression with Hamiltonian Monte Carlo"], "answer_arxiv_id": ["1508.04319"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_19516"} +{"question": "What are some foundational papers on the subject of human feedback in reinforcement learning?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "HG-DAgger: Interactive Imitation Learning with Human Experts", "Reward learning from human preferences and demonstrations in Atari"], "answer_arxiv_id": ["1011.0686v3", "1810.02890", "1811.06521"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_19517"} +{"question": "Which work proposed a solution to domain shift problem in federated learning through the transmission of domain information across clients using continuous frequency space interpolation?", "answer": ["FedDG: Federated Domain Generalization on Medical Image Segmentation via\n Episodic Learning in Continuous Frequency Space"], "answer_arxiv_id": ["2103.06030"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19518"} +{"question": "What are the studies that incorporate 2D diffusion and CLIP priors into per-shape optimization via various score distillation sampling strategy?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Magic3D: High-Resolution Text-to-3D Content Creation", "Fantasia3D: Disentangling Geometry and Appearance for High-quality\n Text-to-3D Content Creation", "RealFusion: 360{\\deg} Reconstruction of Any Object from a Single Image"], "answer_arxiv_id": ["2209.14988", "2211.10440", "2303.13873", "2302.10663"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_19519"} +{"question": "What studies are there about composing pre-trained models through a common interface or other methods?", "answer": ["VisualBERT: A Simple and Performant Baseline for Vision and Language", "SimVLM: Simple Visual Language Model Pretraining with Weak Supervision", "Flamingo: a Visual Language Model for Few-Shot Learning", "ClipCap: CLIP Prefix for Image Captioning", "Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language", "Compositional Visual Generation with Composable Diffusion Models", "Learning to Compose Visual Relations", "Compositional Visual Generation with Energy Based Models"], "answer_arxiv_id": ["1908.03557", "2108.10904", "2204.14198", "2111.09734", "2204.00598", "2206.01714", "2111.09297", "2004.06030"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_19520"} +{"question": "Which research incorporated a 2D CNN to encode scene priors in the field of Neural Rendering?", "answer": ["Efficient Neural Radiance Fields for Interactive Free-viewpoint Video", "Representing Volumetric Videos as Dynamic MLP Maps"], "answer_arxiv_id": ["2112.01517", "2304.06717"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_19521"} +{"question": "What are some works that examines the error of hypergradient approximation methods for certain non-smooth bilevel problems?", "answer": ["Implicit differentiation of Lasso-type models for hyperparameter optimization", "Implicit Differentiation for Fast Hyperparameter Selection in Non-Smooth Convex Learning"], "answer_arxiv_id": ["2002.08943", "2105.01637"], "source_meta": {"published_time": "20220207"}, "qid": "AutoScholarQuery_train_19522"} +{"question": "Which papers developed a teleoperation system for dexterous manipulation?", "answer": ["DexPilot: Vision Based Teleoperation of Dexterous Robotic Hand-Arm System", "Dexterous Imitation Made Easy: A Learning-Based Framework for Efficient Dexterous Manipulation", "Robotic Telekinesis: Learning a Robotic Hand Imitator by Watching Humans on YouTube"], "answer_arxiv_id": ["1910.03135", "2203.13251", "2202.10448"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_19523"} +{"question": "What are some works about using Generative Adversarial Networks (GANs) for image editing and attribute manipulation?", "answer": ["Generative Adversarial Networks", "GAN Inversion: A Survey", "Unsupervised Domain Adaptation GAN Inversion for Image Editing", "In-Domain GAN Inversion for Real Image Editing", "A Style-Based Generator Architecture for Generative Adversarial Networks"], "answer_arxiv_id": ["2203.00667", "2101.05278", "2211.12123", "2004.00049", "1812.04948"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19524"} +{"question": "Could you provide me with studies of the initial works on diffusion-based image generation?", "answer": ["Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["2006.11239", "2105.05233", "2010.02502"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_19525"} +{"question": "Which research demonstrates that large language models can infer personal attributes from text?", "answer": ["Beyond Memorization: Violating Privacy Via Inference with Large Language Models"], "answer_arxiv_id": ["2310.07298v2"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_19526"} +{"question": "Which research work introduced a continuous-time version of Dynamic Topic Models?", "answer": ["Continuous Time Dynamic Topic Models"], "answer_arxiv_id": ["1206.3298v2"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_19527"} +{"question": "Could you provide me some work that develop algorithms for the setting of contextual bandits with linearly-structured actions and general function approximation?", "answer": ["Adapting to Misspecification in Contextual Bandits", "Upper Counterfactual Confidence Bounds: a New Optimism Principle for Contextual Bandits", "Contextual Bandits with Large Action Spaces: Made Practical"], "answer_arxiv_id": ["2107.05745v1", "2007.07876", "2207.05836"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_19528"} +{"question": "What research found that the ordering of data within a task affects catastrophic forgetting?", "answer": ["Overcoming Catastrophic Forgetting beyond Continual Learning: Balanced Training for Neural Machine Translation"], "answer_arxiv_id": ["2203.03910"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_19529"} +{"question": "Which paper showed that acceleration is achievable for an extragradient-type scheme even for cohypomonotone problems?", "answer": ["Fast Extra Gradient Methods for Smooth Structured Nonconvex-Nonconcave Minimax Problems"], "answer_arxiv_id": ["2106.02326"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_19530"} +{"question": "Which studies have pioneered the construction of an event conceptualization benchmark?", "answer": ["Acquiring and Modelling Abstract Commonsense Knowledge via\n Conceptualization"], "answer_arxiv_id": ["2206.01532"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_19531"} +{"question": "Which studies have introduced entropy to attention maps for interpretability analysis?", "answer": ["Entropy- and Distance-Based Predictors From GPT-2 Attention Patterns Predict Reading Times Over and Above GPT-2 Surprisal"], "answer_arxiv_id": ["2212.11185"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_19532"} +{"question": "What research studies have successfully applied reinforcement learning to dexterous manipulation on real hardware?", "answer": ["Learning Dexterous In-Hand Manipulation", "Solving Rubik’s Cube with a Robot Hand", "Transferring Dexterous Manipulation from GPU Simulation to a Remote Real-World TriFinger", "Dexterous Robotic Manipulation using Deep Reinforcement Learning and Knowledge Transfer for Complex Sparse Reward-based Tasks"], "answer_arxiv_id": ["1808.00177", "1910.07113", "2108.09779", "2205.09683"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_19533"} +{"question": "Which studies made significant contribution in the field of video-centric dialogue models?", "answer": ["VideoChat: Chat-Centric Video Understanding", "Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding", "Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and\n Language Models", "Valley: Video Assistant with Large Language model Enhanced abilitY"], "answer_arxiv_id": ["2305.06355", "2306.02858", "2306.05424", "2306.07207"], "source_meta": {"published_time": "20230927"}, "qid": "AutoScholarQuery_train_19534"} +{"question": "Who proposed leveraging deep reinforcement learning to compute the Bayesian Optimal Experimental Design (BOED)?", "answer": ["Deep Adaptive Design: Amortizing Sequential Bayesian Experimental Design", "Policy-Based Bayesian Experimental Design for Non-Differentiable Implicit Models", "Optimizing Sequential Experimental Design with Deep Reinforcement Learning"], "answer_arxiv_id": ["2103.02438", "2203.04272v1", "2202.00821"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_19535"} +{"question": "Which studies discussed the trade-off between adversarial robustness and clean accuracy caused by Adversarial Training (AT)?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy", "Robustness May Be at Odds with Accuracy"], "answer_arxiv_id": ["1901.08573", "1805.12152"], "source_meta": {"published_time": "20240501"}, "qid": "AutoScholarQuery_train_19536"} +{"question": "What research looked into stochastic training under the partial likelihood objective?", "answer": ["DeepSurv: Personalized Treatment Recommender System Using A Cox Proportional Hazards Deep Neural Network", "SODEN: A Scalable Continuous-Time Survival Model through Ordinary Differential Equation Networks"], "answer_arxiv_id": ["1606.00931", "2008.08637"], "source_meta": {"published_time": "20230318"}, "qid": "AutoScholarQuery_train_19537"} +{"question": "Which papers focus on the problem of low-rank matrix estimation from trace measurements?", "answer": ["Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization", "Estimation of (near) low-rank matrices with noise and high-dimensional scaling", "Estimation of high-dimensional low-rank matrices", "Sparse Representation of a Polytope and Recovery of Sparse Signals and Low-rank Matrices"], "answer_arxiv_id": ["0706.4138", "0912.5100v1", "0912.5338", "1306.1154v2"], "source_meta": {"published_time": "20230908"}, "qid": "AutoScholarQuery_train_19538"} +{"question": "What work generated formal specifications from unstructured natural language?", "answer": ["Formal Specifications from Natural Language"], "answer_arxiv_id": ["2206.01962"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19539"} +{"question": "Which papers are related to data-oriented methods that aim to resample the training set to achieve a more balanced data distribution?", "answer": ["SMOTE: Synthetic Minority Over-sampling Technique"], "answer_arxiv_id": ["1106.1813"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_19540"} +{"question": "Which works are about automating the generation of prompt templates?", "answer": ["AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts", "Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2010.15980", "2109.01134"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_19541"} +{"question": "What work assumed affine variance noise and gave a desired rate in the noisy regime without assuming uniformly bounded gradients?", "answer": ["The Power of Adaptivity in SGD: Self-Tuning Step Sizes with Unbounded Gradients and Affine Variance"], "answer_arxiv_id": ["2202.05791"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_19542"} +{"question": "What works use convolutional neural networks with Siamese network architectures for tracking?", "answer": ["Fast Online Object Tracking and Segmentation: A Unifying Approach", "SiamRPN++: Evolution of Siamese Visual Tracking with Very Deep Networks"], "answer_arxiv_id": ["1812.05050", "1812.11703"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_19543"} +{"question": "Which paper models the joint distribution of graph structures and node attributes through a structure-attribute transformer (SAT) for addressing missing node features?", "answer": ["Learning on Attribute-Missing Graphs"], "answer_arxiv_id": ["2011.01623"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19544"} +{"question": "Which works are there on instance-level prediction in 3D scene understanding using either a bottom-up strategy or a top-down solution?", "answer": ["PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation", "GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in\n Point Cloud", "Learning Object Bounding Boxes for 3D Instance Segmentation on Point\n Clouds", "Top-Down Beats Bottom-Up in 3D Instance Segmentation"], "answer_arxiv_id": ["2004.01658", "1812.03320", "1906.01140", "2302.02871"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_19545"} +{"question": "In what works do they align LLMs to frozen vision decoders through an encoder-decoder transformer?", "answer": ["MiniGPT-5: Interleaved Vision-and-Language Generation via Generative\n Vokens"], "answer_arxiv_id": ["2310.02239"], "source_meta": {"published_time": "20240120"}, "qid": "AutoScholarQuery_train_19546"} +{"question": "Which papers demonstrate the successful use of BERT and GPT in large language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1810.04805", "2005.14165"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_19547"} +{"question": "Which research used data augmentation to design a sample-efficient RL method, DrQ?", "answer": ["Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels"], "answer_arxiv_id": ["2004.13649"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19548"} +{"question": "Which papers discuss about the high variance of reparameterization gradient estimator?", "answer": ["Auto-Encoding Variational Bayes", "PIPPS: Flexible Model-Based Policy Search Robust to the Curse of Chaos", "A unified view of likelihood ratio and reparameterization gradients and an optimal importance sampling scheme", "Understanding and correcting pathologies in the training of learned optimizers", "Natural Evolution Strategies"], "answer_arxiv_id": ["1312.6114", "1902.01240", "1910.06419", "1810.10180", "1106.4487"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_19549"} +{"question": "Can you name some works that provided NL statements with parallel FOL annotations?", "answer": ["FOLIO: Natural Language Reasoning with First-Order Logic"], "answer_arxiv_id": ["2209.00840"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19550"} +{"question": "Which paper aims at calibrating a pair of camera and depth sensors considering temporal miscalibration?", "answer": ["Self-Aligning Depth-regularized Radiance Fields for Asynchronous RGB-D\n Sequences"], "answer_arxiv_id": ["2211.07459"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19551"} +{"question": "Which paper is considered the de facto method for multiview 3D reconstruction?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19552"} +{"question": "What works improve upon NeRF's anatomy by adding a coarse-to-fine component?", "answer": ["BARF: Bundle-Adjusting Neural Radiance Fields"], "answer_arxiv_id": ["2104.06405"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19553"} +{"question": "What study established equivalences between replicability and other notions of algorithmic stability when the domain is finite?", "answer": ["Stability is Stable: Connections between Replicability, Privacy, and Adaptive Generalization"], "answer_arxiv_id": ["2303.12921"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19554"} +{"question": "What works proved the minimax-optimal regret bound to be 𝑂~(𝐻2​​|𝒮|​​|𝒜|​​𝑇) for Tabular RL?", "answer": ["Episodic Reinforcement Learning in Finite MDPs: Minimax Lower Bounds Revisited"], "answer_arxiv_id": ["2010.03531"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19555"} +{"question": "Which papers discuss the formulation of diffusion model as a score-based generative model?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["1907.05600", "2011.13456"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19556"} +{"question": "What previous work has the same algorithm as AUTO-S?", "answer": ["Normalized/Clipped SGD with Perturbation for Differentially Private Non-Convex Optimization"], "answer_arxiv_id": ["2206.13033"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_19557"} +{"question": "What are some works that require training a NeRF from precise camera poses first before being able to localize new query images?", "answer": ["INeRF: Inverting Neural Radiance Fields for Pose Estimation", "Loc-NeRF: Monte Carlo Localization using Neural Radiance Fields", "CROSSFIRE: Camera Relocalization On Self-Supervised Features from an\n Implicit Representation"], "answer_arxiv_id": ["2012.05877", "2209.09050", "2303.04869"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19558"} +{"question": "Which studies do not account for uncertainty estimation while handling Semantic Scene Completion (SSC) or LiDAR Panoptic Segmentation?", "answer": ["SCPNet: Semantic Scene Completion on Point Cloud", "S3CNet: A Sparse Semantic Scene Completion Network for LiDAR Point\n Clouds", "Efficient Semantic Scene Completion Network with Spatial Group\n Convolution", "Panoptic Segmentation", "SMAC-Seg: LiDAR Panoptic Segmentation via Sparse Multi-directional\n Attention Clustering", "EfficientLPS: Efficient LiDAR Panoptic Segmentation", "MOPT: Multi-Object Panoptic Tracking"], "answer_arxiv_id": ["2303.06884", "2012.09242", "1907.05091", "1801.00868", "2108.13588", "2102.08009", "2004.08189"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_19559"} +{"question": "Which work applies CG in areas such as vision and language?", "answer": ["When and Why Vision-Language Models Behave like Bags-Of-Words, and What to Do About It?", "Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality"], "answer_arxiv_id": ["2210.01936", "2204.03162"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_19560"} +{"question": "Which papers propose that mixup as a regularizer can further improve out-of-distribution robustness?", "answer": ["RegMixup: Mixup as a Regularizer Can Surprisingly Improve Accuracy and Out Distribution Robustness"], "answer_arxiv_id": ["2206.14502"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_19561"} +{"question": "What research first introduced Variational Auto-Encoder (VAE) for learning continuous representations?", "answer": ["Auto-Encoding Variational Bayes"], "answer_arxiv_id": ["1312.6114"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_19562"} +{"question": "Could you provide some papers that discussed unsupervised skill discovery aiming for a compact set of distinguishable skills embedded in a possibly discrete skill space?", "answer": ["Discovering and Achieving Goals via World Models"], "answer_arxiv_id": ["2110.09514"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_19563"} +{"question": "What studies explored the learning of affordance information for manipulating tasks?", "answer": ["Where2Act: From Pixels to Actions for Articulated 3D Objects", "O2O-Afford: Annotation-Free Large-Scale Object-Object Affordance Learning", "GIFT: Generalizable Interaction-aware Functional Tool Affordances without Labels", "3D AffordanceNet: A Benchmark for Visual Object Affordance Understanding"], "answer_arxiv_id": ["2101.02692", "2106.15087", "2106.14973", "2103.16397"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_19564"} +{"question": "Which work discusses the slow mixing issue of Langevin dynamics?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_19565"} +{"question": "What research bypassed the need for direct vector supervision when doing generative modeling of vector graphics?", "answer": ["Im2Vec: Synthesizing Vector Graphics without Vector Supervision"], "answer_arxiv_id": ["2102.02798"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_19566"} +{"question": "In which study a provably efficient algorithm that achieves a sublinear regret against the best fixed policy has been developed?", "answer": ["Learning Markov Games with Adversarial Opponents: Efficient Algorithms and Fundamental Limits"], "answer_arxiv_id": ["2203.06803"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_19567"} +{"question": "Any works about the theoretical analysis of multi-modal learning?", "answer": ["TCGM: An Information-Theoretic Framework for Semi-Supervised Multi-Modality Learning", "What Makes Multi-modal Learning Better than Single (Provably)"], "answer_arxiv_id": ["2007.06793", "2106.04538"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_19568"} +{"question": "Who worked on learning structured representations among objects and their interactions?", "answer": ["COBRA: Data-Efficient Model-Based RL through Unsupervised Object Discovery and Curiosity-Driven Exploration", "Self-supervised Visual Reinforcement Learning with Object-centric Representations", "Self-supervised Reinforcement Learning with Independently Controllable Subgoals", "Structured Object-Aware Physics Prediction for Video Modeling and Planning", "Factored World Models for Zero-Shot Generalization in Robotic Manipulation", "Contrastive Learning of Structured World Models"], "answer_arxiv_id": ["1905.09275", "2011.14381", "2109.04150", "1910.02425", "2202.05333", "1911.12247"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_19569"} +{"question": "Which work discusses the bounding of Rademacher complexity in terms of covering number of the function class?", "answer": ["Spectrally-normalized margin bounds for neural networks"], "answer_arxiv_id": ["1706.08498"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_19570"} +{"question": "Can you provide studies that proposed generating a sparse feature space in language models?", "answer": ["Linear Algebraic Structure of Word Senses, with Applications to Polysemy", "Sparse Autoencoders Find Highly Interpretable Features in Language\n Models", "Codebook Features: Sparse and Discrete Interpretability for Neural\n Networks"], "answer_arxiv_id": ["1601.03764", "2309.08600", "2310.17230"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19571"} +{"question": "Could you provide me some studies where generative priors derived from diffusion models were utilized?", "answer": ["Denoising Diffusion Probabilistic Models", "Generative Modeling by Estimating Gradients of the Data Distribution", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Zero-Shot Text-to-Image Generation", "Score-Based Generative Modeling through Stochastic Differential\n Equations"], "answer_arxiv_id": ["2006.11239", "1907.05600", "2112.10741", "2112.10752", "2102.12092", "2011.13456"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_19572"} +{"question": "Which works use neural networks in unsupervised machine translation?", "answer": ["Towards cross-lingual distributed representations without parallel text trained with adversarial autoencoders", "Unsupervised Machine Translation Using Monolingual Corpora Only", "Phrase-Based & Neural Unsupervised Machine Translation", "MASS: Masked Sequence to Sequence Pre-training for Language Generation"], "answer_arxiv_id": ["1608.02996", "1711.00043", "1804.07755", "1905.02450"], "source_meta": {"published_time": "20221120"}, "qid": "AutoScholarQuery_train_19573"} +{"question": "What paper proposed a modified version of RLHF using adversarial probing?", "answer": ["Improving alignment of dialogue agents via targeted human judgements"], "answer_arxiv_id": ["2209.14375"], "source_meta": {"published_time": "20240607"}, "qid": "AutoScholarQuery_train_19574"} +{"question": "What documents describe hybrid-based methods in implicit video representation that utilise content-relevant embeddings as inputs?", "answer": ["CNeRV: Content-adaptive Neural Representation for Visual Data", "HNeRV: A Hybrid Neural Representation for Videos", "DNeRV: Modeling Inherent Dynamics via Difference Neural Representation\n for Videos", "FFNeRV: Flow-Guided Frame-Wise Neural Representations for Videos", "Scene Matters: Model-based Deep Video Compression"], "answer_arxiv_id": ["2211.10421", "2304.02633", "2304.06544", "2212.12294", "2303.04557"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_19575"} +{"question": "In the context of LNL, what works design a curriculum by ranking the complexity of the data?", "answer": ["CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images"], "answer_arxiv_id": ["1808.01097"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_19576"} +{"question": "What works refer to the concept of prompting that is designing textual instructions to steer the VLM towards a certain output?", "answer": ["How Can We Know What Language Models Know?"], "answer_arxiv_id": ["1911.12543"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_19577"} +{"question": "What papers present a categorization of these models into string-based, graph-based, and 3D geometry-based methods?", "answer": ["Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules", "Junction Tree Variational Autoencoder for Molecular Graph Generation", "MARS: Markov Molecular Sampling for Multi-objective Drug Discovery", "Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules"], "answer_arxiv_id": ["1610.02415", "1802.04364", "2103.10432", "1906.00957"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_19578"} +{"question": "What works have explored accelerating the sampling of autoregressive models using various approaches?", "answer": ["Blockwise Parallel Decoding for Deep Autoregressive Models", "Accelerating Feedforward Computation via Parallel Nonlinear Equation Solving", "TinyBERT: Distilling BERT for Natural Language Understanding", "LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "Accelerating Large Language Model Decoding with Speculative Sampling", "Fast Inference from Transformers via Speculative Decoding"], "answer_arxiv_id": ["1811.03115", "2002.03629", "1909.10351", "2208.07339", "2302.01318", "2211.17192"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19579"} +{"question": "Which studies first proposed the idea of exploiting class names in the setting of zero and few-shot classification?", "answer": ["Text Classification Using Label Names Only: A Language Model Self-Training Approach", "Benchmarking Zero-shot Text Classification: Datasets, Evaluation and Entailment Approach"], "answer_arxiv_id": ["2010.07245", "1909.00161"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19580"} +{"question": "Which papers addressed dynamic fairness using variants of Markov Decision Processes (MDPs)?", "answer": ["Fairness in Reinforcement Learning", "Learning Fair Policies in Multiobjective (Deep) Reinforcement Learning with Average and Discounted Rewards", "How Do Fair Decisions Fare in Long-term Qualification?", "Algorithms for Fairness in Sequential Decision Making", "Learning Fair Policies in Decentralized Cooperative Multi-Agent Reinforcement Learning", "Towards Long-term Fairness in Recommendation"], "answer_arxiv_id": ["1611.03071", "2008.07773", "2010.11300", "1901.08568", "2012.09421", "2101.03584"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_19581"} +{"question": "What tools in theorem proving are available for the Isabelle proof assistant?", "answer": ["IsarStep: a Benchmark for High-level Mathematical Reasoning"], "answer_arxiv_id": ["2006.09265"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_19582"} +{"question": "Which papers highlighted that the sampling procedure of DDPM with ϵ-prediction parametrization resembles LD of an EBM?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_19583"} +{"question": "Can you name the works proposing smaller, specialized bidirectional encoders?", "answer": ["Towards General Text Embeddings with Multi-stage Contrastive Learning", "C-Pack: Packaged Resources To Advance General Chinese Embedding"], "answer_arxiv_id": ["2308.03281v1", "2309.07597"], "source_meta": {"published_time": "20240625"}, "qid": "AutoScholarQuery_train_19584"} +{"question": "Which works focus on batch normalization statistics for updating the model in test-time adaptation?", "answer": ["Revisiting Batch Normalization For Practical Domain Adaptation"], "answer_arxiv_id": ["1603.04779"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_19585"} +{"question": "Can you mention works in which vision transformers models outperform convolutional neural networks?", "answer": ["Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1512.03385"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_19586"} +{"question": "Which work demonstrated that training on synthetic data can be beneficial to FL model performance?", "answer": ["On the Importance and Applicability of Pre-Training for Federated Learning"], "answer_arxiv_id": ["2206.11488"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_19587"} +{"question": "Any works discussing online test methods in OMRL?", "answer": ["Offline Meta-Reinforcement Learning with Online Self-Supervision"], "answer_arxiv_id": ["2107.03974"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_19588"} +{"question": "What researches have been completed on non-iterative methods in offline RL, specifically looking at weighted or conditional imitation learning?", "answer": ["A Theory of Regularized Markov Decision Processes", "Leverage the Average: an Analysis of KL Regularization in Reinforcement Learning", "Maximum a Posteriori Policy Optimisation", "BAIL: Best-Action Imitation Learning for Batch Deep Reinforcement Learning", "Critic Regularized Regression", "Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning"], "answer_arxiv_id": ["1901.11275", "2003.14089", "1806.06920", "1910.12179", "2006.15134", "1910.00177"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_19589"} +{"question": "What studies have incorporated depth for semantic segmentation in different settings?", "answer": ["Domain Adaptive Semantic Segmentation with Self-Supervised Depth\n Estimation", "Three Ways to Improve Semantic Segmentation with Self-Supervised Depth\n Estimation", "Plugging Self-Supervised Monocular Depth into Unsupervised Domain\n Adaptation for Semantic Segmentation"], "answer_arxiv_id": ["2104.13613", "2012.10782", "2110.06685"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_19590"} +{"question": "What work first introduced the use of log barrier in graph learning?", "answer": ["How to learn a graph from smooth signals"], "answer_arxiv_id": ["1601.02513v1"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_19591"} +{"question": "Which works have utilized self-interpretability methods, that use models grounded in concepts?", "answer": ["Concept Bottleneck Models", "Concept Embedding Models: Beyond the Accuracy-Explainability Trade-Off", "Label-Free Concept Bottleneck Models", "Post-hoc Concept Bottleneck Models", "Language in a Bottle: Language Model Guided Concept Bottlenecks for\n Interpretable Image Classification"], "answer_arxiv_id": ["2007.04612", "2209.09056", "2304.06129", "2205.15480", "2211.11158"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_19592"} +{"question": "Which datasets are used to study animal social behavior?", "answer": ["The Multi-Agent Behavior Dataset: Mouse Dyadic Social Interactions", "Animal Kingdom: A Large and Diverse Dataset for Animal Behavior Understanding"], "answer_arxiv_id": ["2104.02710", "2204.08129"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_19593"} +{"question": "What works have been referenced with respect to the construction of human-annotated CommonSense Knowledge Graphs (CSKG)?", "answer": ["ConceptNet 5.5: An Open Multilingual Graph of General Knowledge", "ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning", "COMET-ATOMIC 2020: On Symbolic and Neural Commonsense Knowledge Graphs", "\"I'm Not Mad\": Commonsense Implications of Negation and Contradiction", "GLUCOSE: GeneraLized and COntextualized Story Explanations"], "answer_arxiv_id": ["1612.03975", "1811.00146", "2010.05953", "2104.06511", "2009.07758"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_19594"} +{"question": "Which works originally investigated Contrastive Sentence Representation Learning?", "answer": ["SimCSE: Simple Contrastive Learning of Sentence Embeddings", "ConSERT: A Contrastive Framework for Self-Supervised Sentence\n Representation Transfer"], "answer_arxiv_id": ["2104.08821", "2105.11741"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_19595"} +{"question": "Could you provide any works that employ transformer-based flow estimators in optical flow estimation?", "answer": ["FlowFormer: A Transformer Architecture for Optical Flow", "FlowFormer++: Masked Cost Volume Autoencoding for Pretraining Optical\n Flow Estimation", "Global Matching with Overlapping Attention for Optical Flow Estimation", "Perceiver IO: A General Architecture for Structured Inputs & Outputs"], "answer_arxiv_id": ["2203.16194", "2303.01237", "2203.11335", "2107.14795"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_19596"} +{"question": "What study is closest to the researcher's work in consideration of GN optimization?", "answer": ["Gram-Gauss-Newton Method: Learning Overparameterized Neural Networks for Regression Problems"], "answer_arxiv_id": ["1905.11675"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_19597"} +{"question": "What works derive regret bounds for arbitrary choices of q in q-Tsallis regularizer for an analysis of bandits?", "answer": ["Tsallis-INF: An Optimal Algorithm for Stochastic and Adversarial Bandits", "Improved Best-of-Both-Worlds Guarantees for Multi-Armed Bandits: FTRL with General Regularizers and Multiple Optimal Arms"], "answer_arxiv_id": ["1807.07623v6", "2302.13534"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19598"} +{"question": "What is the research that provides a cascaded network architecture for pose and shape estimation from point clouds?", "answer": ["PointHPS: Cascaded 3D Human Pose and Shape Estimation from Point Clouds"], "answer_arxiv_id": ["2308.14492"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19599"} +{"question": "What papers introduce the concept of diffusion models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential\n Equations", "Diffusion Models Beat GANs on Image Synthesis", "Diffusion models as plug-and-play priors", "Zero-Shot Text-to-Image Generation", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Denoising Diffusion Implicit Models"], "answer_arxiv_id": ["1503.03585", "2006.11239", "2011.13456", "2105.05233", "2206.09012", "2102.12092", "2112.10752", "2205.11487", "2010.02502"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19600"} +{"question": "Which papers investigate predicting model performance related to a specific dataset without actual training?", "answer": ["Datamodels: Predicting Predictions from Training Data", "TRAK: Attributing Model Behavior at Scale", "Agreement-on-the-Line: Predicting the Performance of Neural Networks under Distribution Shift"], "answer_arxiv_id": ["2202.00622", "2303.14186", "2206.13089"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_19601"} +{"question": "What research has shown that deep RL algorithms are vulnerable to adversarial perturbations and developed robust policies?", "answer": ["Adversarial Attacks on Neural Network Policies", "Robust Deep Reinforcement Learning against Adversarial Perturbations on State Observations", "Robust Reinforcement Learning on State Observations with Learned Optimal Adversary", "Who Is the Strongest Enemy? Towards Optimal and Efficient Evasion Attacks in Deep RL"], "answer_arxiv_id": ["1702.02284", "2003.08938", "2101.08452", "2106.05087"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_19602"} +{"question": "What work is there on using LLMs to ground image generation?", "answer": ["Generating Images with Multimodal Language Models", "Emu: Generative Pretraining in Multimodality"], "answer_arxiv_id": ["2305.17216", "2307.05222"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19603"} +{"question": "What studies have further developed research on norms to M-estimators?", "answer": ["Dimensionality Reduction for Tukey Regression"], "answer_arxiv_id": ["1905.05376v1"], "source_meta": {"published_time": "20230331"}, "qid": "AutoScholarQuery_train_19604"} +{"question": "Which work proved a local PL property for matrix sensing with exact parameterization?", "answer": ["Low-rank Solutions of Linear Matrix Equations via Procrustes Flow"], "answer_arxiv_id": ["1507.03566"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_19605"} +{"question": "Which work utilized spatial transformer network modules to restore a video sequence and its underlying motion?", "answer": ["Restoration of Video Frames from a Single Blurred Image with Motion\n Understanding"], "answer_arxiv_id": ["2104.09134"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_19606"} +{"question": "Which studies investigated obtaining language model preferences through the interaction between inputs and in-context examples?", "answer": ["Compositional Exemplars for In-context Learning"], "answer_arxiv_id": ["2302.05698"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_19607"} +{"question": "What works have discussed the efficacy of leveraging data samples from different sources in enhancing robustness and generalizability?", "answer": ["Training of Convolutional Networks on Multiple Heterogeneous Datasets\n for Street Scene Semantic Segmentation"], "answer_arxiv_id": ["1803.05675"], "source_meta": {"published_time": "20240502"}, "qid": "AutoScholarQuery_train_19608"} +{"question": "Can you name the recent studies that have applied summarization prompts on Large Language Models (LLMs) to generate review summaries?", "answer": ["Prompted Opinion Summarization with GPT-3.5", "From Sparse to Dense: GPT-4 Summarization with Chain of Density\n Prompting"], "answer_arxiv_id": ["2211.15914", "2309.04269"], "source_meta": {"published_time": "20240719"}, "qid": "AutoScholarQuery_train_19609"} +{"question": "What works facilitated specifically engineered systems for Transformer-based Large Language Model (LLM) inference?", "answer": ["LightSeq: A High Performance Inference Library for Transformers", "Efficiently Scaling Transformer Inference", "TurboTransformers: An Efficient GPU Serving System For Transformer\n Models", "DeepSpeed Inference: Enabling Efficient Inference of Transformer Models\n at Unprecedented Scale"], "answer_arxiv_id": ["2010.13887", "2211.05102", "2010.05680", "2207.00032"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_19610"} +{"question": "Which study attempted to generalize neural surface reconstruction?", "answer": ["SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse Views"], "answer_arxiv_id": ["2206.05737"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_19611"} +{"question": "Which study discovered that repeating a small portion of data during LLM pre-training can significantly harm model performance?", "answer": ["Scaling Laws and Interpretability of Learning from Repeated Data"], "answer_arxiv_id": ["2205.10487"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_19612"} +{"question": "Could you provide me some studies that apply chain-of-thought prompting in robotics scenarios?", "answer": ["Do As I Can, Not As I Say: Grounding Language in Robotic Affordances"], "answer_arxiv_id": ["2204.01691"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_19613"} +{"question": "Which works presented advancement in processing point clouds through Graph Neural Networks (GNNs)?", "answer": ["4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks"], "answer_arxiv_id": ["1904.08755"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_19614"} +{"question": "In which publications do the approaches make agents maximize state entropy along with task rewards?", "answer": ["Novelty Search in Representational Space for Sample Efficient Exploration", "State Entropy Maximization with Random Encoders for Efficient Exploration", "k-Means Maximum Entropy Exploration", "Rényi State Entropy Maximization for Exploration Acceleration in Reinforcement Learning"], "answer_arxiv_id": ["2009.13579", "2102.09430", "2205.15623", "2203.04297"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19615"} +{"question": "What works have predominantly used supervised learning-based approaches to learn adjustments aligned with human perception for image enhancement?", "answer": ["Deep Bilateral Learning for Real-Time Image Enhancement", "DeepLPF: Deep Local Parametric Filters for Image Enhancement", "Conditional Sequential Modulation for Efficient Global Image Retouching", "Real-time Image Enhancer via Learnable Spatial-aware 3D Lookup Tables"], "answer_arxiv_id": ["1707.02880", "2003.13985", "2009.10390", "2108.08697"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_19616"} +{"question": "What are some examples of two-stage object detectors?", "answer": ["Mask R-CNN", "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks"], "answer_arxiv_id": ["1703.06870", "1506.01497"], "source_meta": {"published_time": "20221205"}, "qid": "AutoScholarQuery_train_19617"} +{"question": "What references discuss the approach of manual construction for knowledge bases?", "answer": ["ConceptNet 5.5: An Open Multilingual Graph of General Knowledge", "ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning"], "answer_arxiv_id": ["1612.03975", "1811.00146"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_19618"} +{"question": "What are the research papers that focused on multi-entity subject-driven generation?", "answer": ["Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning", "FastComposer: Tuning-Free Multi-Subject Image Generation with Localized\n Attention"], "answer_arxiv_id": ["2307.11410", "2305.10431"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_19619"} +{"question": "What works have been done on 3D Dense Captioning (3D-DC) where a model translates an input 3D scene into instance coordinates and natural language descriptions?", "answer": ["Scan2Cap: Context-aware Dense Captioning in RGB-D Scans", "Spatiality-guided Transformer for 3D Dense Captioning on Point Clouds", "End-to-End 3D Dense Captioning with Vote2Cap-DETR"], "answer_arxiv_id": ["2012.02206", "2204.10688", "2301.02508"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19620"} +{"question": "What papers discussed black-box attacks?", "answer": ["ZOO: Zeroth Order Optimization Based Black-box Attacks to Deep Neural Networks without Training Substitute Models", "Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models", "Black-box Adversarial Attacks with Limited Queries and Information"], "answer_arxiv_id": ["1708.03999", "1712.04248", "1804.08598"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_19621"} +{"question": "What research work employed flow-based models as the latent prior", "answer": ["FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow"], "answer_arxiv_id": ["1909.02480"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_19622"} +{"question": "Which paper proposes a counterfactual data augmentation method to enhance Q-learning?", "answer": ["Sample-Efficient Reinforcement Learning via Counterfactual-Based Data Augmentation"], "answer_arxiv_id": ["2012.09092"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_19623"} +{"question": "Can you provide some examples of work that use physics-based refinement in hand-object pose estimation?", "answer": ["CPF: Learning a Contact Potential Field to Model the Hand-Object Interaction", "Physical Interaction: Reconstructing Hand-object Interactions with Physics"], "answer_arxiv_id": ["2012.00924", "2209.10833"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_19624"} +{"question": "What is the research that trained a network to acquire discriminative features of objects for template-based methods?", "answer": ["Learning Descriptors for Object Recognition and 3D Pose Estimation"], "answer_arxiv_id": ["1502.05908"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19625"} +{"question": "Could you cite some studies that used differentiable renderers for 3D face reconstruction?", "answer": ["Accelerating 3D Deep Learning with PyTorch3D", "Kaolin: A PyTorch Library for Accelerating 3D Deep Learning Research"], "answer_arxiv_id": ["2007.08501", "1911.05063"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_19626"} +{"question": "What paper proposed the first computationally efficient algorithm LSVI-UCB that achieves worst-case regret in online learning settings?", "answer": ["Provably Efficient Reinforcement Learning with Linear Function Approximation"], "answer_arxiv_id": ["1907.05388"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_19627"} +{"question": "Which references discuss the challenges in the practical implementation of RLHF due to its dependence on high-quality feedback from adept labelers?", "answer": ["Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned", "Red Teaming Language Models with Language Models"], "answer_arxiv_id": ["2209.07858", "2202.03286"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_19628"} +{"question": "Which works used CLIP knowledge in supervised human-object interaction (HOI) detection?", "answer": ["Category-Aware Transformer Network for Better Human-Object Interaction Detection", "GEN-VLKT: Simplify Association and Enhance Interaction Understanding for HOI Detection"], "answer_arxiv_id": ["2204.04911", "2203.13954"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19629"} +{"question": "Which works introduced benchmarks for estimating scene flow from stereo videos?", "answer": ["A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimation"], "answer_arxiv_id": ["1512.02134"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_19630"} +{"question": "Could you provide me with some studies about fusion-encoder models in the context of VL pre-training?", "answer": ["Align before Fuse: Vision and Language Representation Learning with Momentum Distillation", "VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts", "CoCa: Contrastive Captioners are Image-Text Foundation Models"], "answer_arxiv_id": ["2107.07651", "2111.02358", "2205.01917"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_19631"} +{"question": "Could you provide some examples of recent research that extended multimodal representation learning to 3D modality?", "answer": ["Learning 3D Representations from 2D Pre-trained Models via\n Image-to-Point Masked Autoencoders", "Contrast with Reconstruct: Contrastive 3D Representation Learning Guided\n by Generative Pretraining"], "answer_arxiv_id": ["2212.06785", "2302.02318"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_19632"} +{"question": "Which work introduced contrastive pre-trained vision-language models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_19633"} +{"question": "Who proposed AdapterDrop for analyzing the role of adapters during adaptation to enhance the robustness and speed up the inference?", "answer": ["AdapterDrop: On the Efficiency of Adapters in Transformers"], "answer_arxiv_id": ["2010.11918"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_19634"} +{"question": "Which work is the first to bring the human feedback reward fine-tuning idea to diffusion models?", "answer": ["Aligning Text-to-Image Models using Human Feedback"], "answer_arxiv_id": ["2302.12192"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_19635"} +{"question": "Which studies have proposed low-rank approximation for attention matrices?", "answer": ["Linformer: Self-Attention with Linear Complexity", "Rethinking Attention with Performers"], "answer_arxiv_id": ["2006.04768", "2009.14794"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19636"} +{"question": "Which works tackled the blind docking task by directly predicting pocket keypoints?", "answer": ["EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction"], "answer_arxiv_id": ["2202.05146"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_19637"} +{"question": "What papers set the latent space of VAE to be the elliptic space?", "answer": ["Spherical Latent Spaces for Stable Variational Autoencoders", "Hyperspherical Variational Auto-Encoders"], "answer_arxiv_id": ["1808.10805", "1804.00891"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19638"} +{"question": "Could you refer me some works investigating the existence of shortcuts in different fields of deep learning?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20220524"}, "qid": "AutoScholarQuery_train_19639"} +{"question": "What papers have utilized the framework of conformal prediction?", "answer": ["A Gentle Introduction to Conformal Prediction and Distribution-Free\n Uncertainty Quantification"], "answer_arxiv_id": ["2107.07511"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_19640"} +{"question": "Which papers extend the grid or plane-based strategy to the acceleration or representation of time-conditioned 4D feature in dynamic scene modeling?", "answer": ["Fast Dynamic Radiance Fields with Time-Aware Neural Voxels", "Tensor4D : Efficient Neural 4D Decomposition for High-fidelity Dynamic\n Reconstruction and Rendering", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance", "HexPlane: A Fast Representation for Dynamic Scenes"], "answer_arxiv_id": ["2205.15285", "2211.11610", "2301.10241", "2301.09632"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_19641"} +{"question": "Could you provide me with studies that showed successful usage of cross-attention between different views of the same video?", "answer": ["Multiview Transformers for Video Recognition", "Dual Cross-Attention Learning for Fine-Grained Visual Categorization and Object Re-Identification", "CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification", "Cross-Attention Transformer for Video Interpolation"], "answer_arxiv_id": ["2201.04288", "2205.02151", "2103.14899", "2207.04132"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19642"} +{"question": "Which works illustrate the learning of lower-level representations through view synthesis from few images?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_19643"} +{"question": "What studies explored OOD detection in semantic parsing, speech recognition, and machine translation?", "answer": ["Uncertainty Estimation in Autoregressive Structured Prediction", "Shifts: A Dataset of Real Distributional Shift Across Multiple Large-Scale Tasks", "Wat zei je? Detecting Out-of-Distribution Translations with Variational Transformers"], "answer_arxiv_id": ["2002.07650", "2107.07455", "2006.08344"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19644"} +{"question": "Could you provide me some studies about generating audio conditioned on images?", "answer": ["I Hear Your True Colors: Image Guided Audio Generation"], "answer_arxiv_id": ["2211.03089"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19645"} +{"question": "What research includes latent variables in the Non-Autoregressive (NAR) model to exploit latent representation more sufficiently?", "answer": ["Non-Autoregressive Translation by Learning Target Categorical Codes"], "answer_arxiv_id": ["2103.11405"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_19646"} +{"question": "Could you provide me some works about online learning with dynamic comparators?", "answer": ["Online Optimization : Competing with Dynamic Comparators", "Efficient Contextual Bandits in Non-stationary Worlds", "Bandit Convex Optimization in Non-stationary Environments"], "answer_arxiv_id": ["1501.06225", "1708.01799v4", "1907.12340"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_19647"} +{"question": "What works introduced and developed Cross-modal foundation models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision"], "answer_arxiv_id": ["2103.00020", "2102.05918"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_19648"} +{"question": "Are there any works about the optimization of only task instructions in Automatic Prompt Engineer?", "answer": ["Large Language Models Are Human-Level Prompt Engineers"], "answer_arxiv_id": ["2211.01910"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19649"} +{"question": "Can you list the papers that investigated the role of self-attention and MLP layers in ViT models?", "answer": ["Do Vision Transformers See Like Convolutional Neural Networks?", "Understanding Robustness of Transformers for Image Classification", "How Do Vision Transformers Work?"], "answer_arxiv_id": ["2108.08810", "2103.14586", "2202.06709"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_19650"} +{"question": "Could you provide me the studies improving DGM-regularized compressive sensing problems?", "answer": ["Prior Image-Constrained Reconstruction using Style-Based Generative Models"], "answer_arxiv_id": ["2102.12525"], "source_meta": {"published_time": "20211207"}, "qid": "AutoScholarQuery_train_19651"} +{"question": "What are the studies that appied generic active learning strategies and used hand-crafted heuristics for 3D detection learning?", "answer": ["Deep Active Learning for Efficient Training of a LiDAR 3D Object Detector", "Localization-Aware Active Learning for Object Detection", "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"], "answer_arxiv_id": ["1901.10609v2", "1801.05124", "1506.02142"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_19652"} +{"question": "Who used a Variational Autoencoder conditioned on an object list and raster grid of the ground roadways to validate synthetic image similarity to real images?", "answer": ["Deep Stochastic Radar Models"], "answer_arxiv_id": ["1701.09180"], "source_meta": {"published_time": "20240428"}, "qid": "AutoScholarQuery_train_19653"} +{"question": "Which studies optimized a continuous array of prompt vectors within the language branch?", "answer": ["Learning to Prompt for Vision-Language Models"], "answer_arxiv_id": ["2109.01134"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19654"} +{"question": "Which studies conducted uniform-in-time propagation of chaos estimates in the context of mean-field Langevin dynamics?", "answer": ["Uniform-in-time propagation of chaos for mean field Langevin dynamics", "Convergence of mean-field Langevin dynamics: Time and space discretization, stochastic gradient, and variance reduction"], "answer_arxiv_id": ["2212.03050", "2306.07221"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_19655"} +{"question": "What papers continued the research on univariate Gaussian learning in the multivariate setting?", "answer": ["CoinPress: Practical Private Mean and Covariance Estimation"], "answer_arxiv_id": ["2006.06618"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_19656"} +{"question": "What papers trained a continuous map by learning the underlying velocity?", "answer": ["How to Train Your Neural ODE: the World of Jacobian and Kinetic Regularization", "OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport"], "answer_arxiv_id": ["2002.02798", "2006.00104"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19657"} +{"question": "Which papers involve the successful application of BERT and GPT in pre-training large-scale language models?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1810.04805", "2005.14165"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_19658"} +{"question": "Any works talking about 'Offloading' which reduces GPU memory consumption?", "answer": ["ZeRO-Infinity: Breaking the GPU Memory Wall for Extreme Scale Deep Learning", "ZeRO-Offload: Democratizing Billion-Scale Model Training"], "answer_arxiv_id": ["2104.07857", "2101.06840"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_19659"} +{"question": "Could you provide me some works that used Neural Architecture Search (NAS) to find optimal architectures?", "answer": ["CryptoNAS: Private Inference on a ReLU Budget", "Sphynx: ReLU-Efficient Network Design for Private Inference"], "answer_arxiv_id": ["2006.08733", "2106.11755v1"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_19660"} +{"question": "Any works about contrastive learning methods for self-supervised feature learning in correspondence?", "answer": ["Learning Correspondence from the Cycle-consistency of Time", "Space-Time Correspondence as a Contrastive Random Walk", "Rethinking Self-supervised Correspondence Learning: A Video Frame-level Similarity Perspective", "Emerging Properties in Self-Supervised Vision Transformers"], "answer_arxiv_id": ["1903.07593", "2006.14613", "2103.17263", "2104.14294"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_19661"} +{"question": "In which works have regularized encoders been used to maintain equivariance with respect to the group action?", "answer": ["Equivariant Neural Rendering", "Quantifying and Learning Linear Symmetry-Based Disentanglement"], "answer_arxiv_id": ["2006.07630", "2011.06070"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_19662"} +{"question": "What papers are about using Lagrangian dynamics in neural ODE models?", "answer": ["Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning"], "answer_arxiv_id": ["1907.04490"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_19663"} +{"question": "What works explored the use of the discriminator of a single-image generative model to find artifacts in the borders of generated textures?", "answer": ["SeamlessGAN: Self-Supervised Synthesis of Tileable Texture Maps"], "answer_arxiv_id": ["2201.05120"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_19664"} +{"question": "Which works are about training vision adaptors with language models?", "answer": ["MAGMA -- Multimodal Augmentation of Generative Models through\n Adapter-based Finetuning", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning", "Linearly Mapping from Image to Text Space"], "answer_arxiv_id": ["2112.05253", "2304.10592", "2304.08485", "2209.15162"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_19665"} +{"question": "Which research worked on single and multi-index models using the replica method?", "answer": ["Optimal Errors and Phase Transitions in High-Dimensional Generalized Linear Models", "Generalization error in high-dimensional perceptrons: Approaching Bayes error with convex optimization", "The committee machine: Computational to statistical gaps in learning a two-layers neural network"], "answer_arxiv_id": ["1708.03395", "2006.06560", "1806.05451"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19666"} +{"question": "Which research works involved solving PDEs in the reduced space using a data-driven approach?", "answer": ["Continuous PDE Dynamics Forecasting with Implicit Neural Representations"], "answer_arxiv_id": ["2209.14855v2"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_19667"} +{"question": "Which works use fixed routing strategies for more stable routing and training in MoE models?", "answer": ["Hash Layers For Large Sparse Models", "StableMoE: Stable Routing Strategy for Mixture of Experts"], "answer_arxiv_id": ["2106.04426", "2204.08396"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_19668"} +{"question": "Any works about self-talk models for commonsense reasoning?", "answer": ["Unsupervised Commonsense Question Answering with Self-Talk", "Generated Knowledge Prompting for Commonsense Reasoning"], "answer_arxiv_id": ["2004.05483", "2110.08387"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_19669"} +{"question": "Any examples of works in which bias was found in CATE estimators?", "answer": ["Meta-learners for Estimating Heterogeneous Treatment Effects using Machine Learning", "Towards Optimal Doubly Robust Estimation of Heterogeneous Causal Effects", "Double/Debiased Machine Learning for Treatment and Structural Parameters", "Nonparametric Estimation of Heterogeneous Treatment Effects: From Theory to Learning Algorithms"], "answer_arxiv_id": ["1706.03461", "2004.14497", "1608.00060", "2101.10943v2"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_19670"} +{"question": "Could you provide me some studies that discussed reward learning techniques using simpler forms of feedback like scalar scores or rankings?", "answer": ["Learning Multimodal Rewards from Rankings", "Extrapolating Beyond Suboptimal Demonstrations via Inverse Reinforcement Learning from Observations", "The Green Choice: Learning and Influencing Human Decisions on Shared Roads"], "answer_arxiv_id": ["2109.12750", "1904.06387", "1904.02209"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19671"} +{"question": "Which research papers describe the implementation of diffusion models in MRI reconstruction tasks?", "answer": ["Measurement-conditioned Denoising Diffusion Probabilistic Model for\n Under-sampled Medical Image Reconstruction", "Towards performant and reliable undersampled MR reconstruction via\n diffusion model sampling", "Score-based diffusion models for accelerated MRI", "Adaptive Diffusion Priors for Accelerated MRI Reconstruction", "DisC-Diff: Disentangled Conditional Diffusion Model for Multi-Contrast\n MRI Super-Resolution"], "answer_arxiv_id": ["2203.03623", "2203.04292", "2110.05243", "2207.05876", "2303.13933"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_19672"} +{"question": "Which studies suggest interpolating nodes to enrich minority classes in data augmentation for graph representation learning?", "answer": ["GraphSMOTE: Imbalanced Node Classification on Graphs with Graph Neural Networks"], "answer_arxiv_id": ["2103.08826"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_19673"} +{"question": "Who proposed avoiding the evaluation of out-of-distribution actions using the upper expectile value function or policy optimization within a latent action space?", "answer": ["Offline Reinforcement Learning with Implicit Q-Learning", "Latent-Variable Advantage-Weighted Policy Optimization for Offline RL", "PLAS: Latent Action Space for Offline Reinforcement Learning", "Let Offline RL Flow: Training Conservative Agents in the Latent Space of Normalizing Flows"], "answer_arxiv_id": ["2110.06169", "2203.08949v1", "2011.07213", "2211.11096"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19674"} +{"question": "What papers expound on the methods that always require extra information for fast adaptation in offline meta-RL?", "answer": ["FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization", "Offline Meta-Reinforcement Learning with Advantage Weighting", "Robust Task Representations for Offline Meta-Reinforcement Learning via Contrastive Learning", "Offline Meta-Reinforcement Learning with Online Self-Supervision"], "answer_arxiv_id": ["2010.01112", "2008.06043", "2206.10442", "2107.03974"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19675"} +{"question": "Which papers identified limited visual representations as a cause of multimodal hallucination in LMMs?", "answer": ["From CLIP to DINO: Visual Encoders Shout in Multi-modal Large Language Models", "Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMs", "Evaluation and Analysis of Hallucination in Large Vision-Language Models"], "answer_arxiv_id": ["2310.08825v3", "2401.06209", "2308.15126"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_19676"} +{"question": "Which works proposed methods based on Lagrangian relaxation in soft-constrainted RL?", "answer": ["Risk-Constrained Reinforcement Learning with Percentile Risk Criteria", "Responsive Safety in Reinforcement Learning by PID Lagrangian Methods", "Accelerated Primal-Dual Policy Optimization for Safe Reinforcement Learning"], "answer_arxiv_id": ["1512.01629", "2007.03964", "1802.06480"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_19677"} +{"question": "What research works have been done on minimizing the visibility of locating patterns in QR codes?", "answer": ["OAcode: Overall Aesthetic 2D Barcode on Screen"], "answer_arxiv_id": ["2302.02396"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_19678"} +{"question": "Could you provide some research that learn initialization and adapt the parameters with policy gradient methods?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "ProMP: Proximal Meta-Policy Search"], "answer_arxiv_id": ["1703.03400", "1810.06784"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_19679"} +{"question": "Any works that discuss early post-hoc explanation methods based on representation spaces?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)"], "answer_arxiv_id": ["1711.11279"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19680"} +{"question": "Which paper proposes to increase dropout rate as training progresses to address late-stage overfitting?", "answer": ["Curriculum Dropout"], "answer_arxiv_id": ["1703.06229"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19681"} +{"question": "Can you provide research papers on other model architectures like 3D convolution approaches, transformers, and standard RNNs?", "answer": ["U-Net: Convolutional Networks for Biomedical Image Segmentation", "3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation", "Diffusion Models for Video Prediction and Infilling", "VideoGPT: Video Generation using VQ-VAE and Transformers", "Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive Transformer", "Attention Is All You Need", "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "ViViT: A Video Vision Transformer", "MaskViT: Masked Visual Pre-Training for Video Prediction", "Clockwork Variational Autoencoders", "FitVid: Overfitting in Pixel-Level Video Prediction"], "answer_arxiv_id": ["1505.04597", "1606.06650", "2206.07696", "2104.10157", "2204.03638", "1706.03762", "2010.11929", "2103.15691", "2206.11894", "2102.09532", "2106.13195"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_19682"} +{"question": "What research papers have adopted Federated Learning for various medical image analysis tasks?", "answer": ["Federated Learning for Medical Image Analysis: A Survey"], "answer_arxiv_id": ["2306.05980"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_19683"} +{"question": "Which papers are on reconstructing geometry at every step of the denoising process?", "answer": ["Viewset Diffusion: (0-)Image-Conditioned 3D Generative Models from 2D\n Data", "Diffusion with Forward Models: Solving Stochastic Inverse Problems\n Without Direct Supervision"], "answer_arxiv_id": ["2306.07881", "2306.11719"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_19684"} +{"question": "Which works provide insight into 3D hand shape reconstruction employing a discretized representation of shape?", "answer": ["Robust 3D Hand Pose Estimation in Single Depth Images: from Single-View CNN to Multi-View CNNs", "Pushing the Envelope for RGB-based Dense 3D Hand Pose Estimation via Neural Rendering", "3D Hand Shape and Pose from Images in the Wild", "Leveraging Photometric Consistency over Time for Sparsely Supervised Hand-Object Reconstruction", "Learning joint reconstruction of hands and manipulated objects", "End-to-end Hand Mesh Recovery from a Monocular RGB Image", "ACR: Attention Collaboration-based Regressor for Arbitrary Two-Hand Reconstruction", "Interacting Attention Graph for Single Image Two-Hand Reconstruction", "Weakly-Supervised Mesh-Convolutional Hand Reconstruction in the Wild", "End-to-End Human Pose and Mesh Reconstruction with Transformers"], "answer_arxiv_id": ["1606.07253", "1904.04196", "1902.03451", "2004.13449", "1904.05767", "1902.09305", "2303.05938", "2203.09364", "2004.01946", "2012.09760"], "source_meta": {"published_time": "20230716"}, "qid": "AutoScholarQuery_train_19685"} +{"question": "What research exploited supervision of negative pairs from memory queues in order to relax the demand of large batch size during vision model pretraining?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning"], "answer_arxiv_id": ["1911.05722", "2003.04297"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19686"} +{"question": "Could you provide me some works studying sources of randomness in neural model training such as initialization or dropout?", "answer": ["Practical Recommendations for Gradient-Based Training of Deep Architectures", "Reporting Score Distributions Makes a Difference: Performance Study of LSTM-networks for Sequence Tagging", "On Model Stability as a Function of Random Seed"], "answer_arxiv_id": ["1206.5533", "1707.09861", "1909.10447"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_19687"} +{"question": "What papers have proposed approaches to search for shared features to solve issues in Visible-Infrared Person ReID?", "answer": ["Deep Learning for Person Re-identification: A Survey and Outlook", "Dynamic Dual-Attentive Aggregation Learning for Visible-Infrared Person\n Re-Identification"], "answer_arxiv_id": ["2001.04193", "2007.09314"], "source_meta": {"published_time": "20240318"}, "qid": "AutoScholarQuery_train_19688"} +{"question": "Which work studies a problem similar to the one mentioned in the equation?", "answer": ["Multi-block Min-max Bilevel Optimization with Applications in Multi-task Deep AUC Maximization"], "answer_arxiv_id": ["2206.00260"], "source_meta": {"published_time": "20220303"}, "qid": "AutoScholarQuery_train_19689"} +{"question": "Which paper introduced the concept of end-to-end learning for stochastic optimization?", "answer": ["Task-based End-to-end Model Learning in Stochastic Optimization"], "answer_arxiv_id": ["1703.04529"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_19690"} +{"question": "Any works focused on using a diffusion model conditioned on low resolution images for image restoration?", "answer": ["ResShift: Efficient Diffusion Model for Image Super-resolution by\n Residual Shifting", "DiffIR: Efficient Diffusion Model for Image Restoration", "High-Resolution Image Synthesis with Latent Diffusion Models", "Image Super-Resolution via Iterative Refinement", "Denoising Diffusion Probabilistic Models for Robust Image\n Super-Resolution in the Wild"], "answer_arxiv_id": ["2307.12348", "2303.09472", "2112.10752", "2104.07636", "2302.07864"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_19691"} +{"question": "Which work proposed an alternative to Bayesian Neural Networks for uncertainty estimates?", "answer": ["Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1612.01474"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_19692"} +{"question": "What papers discuss the benefits of large datasets in deep learning?", "answer": ["ImageNet Large Scale Visual Recognition Challenge", "Pointer Sentinel Mixture Models"], "answer_arxiv_id": ["1409.0575", "1609.07843"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_19693"} +{"question": "Which work addresses hallucination problems during the training of large language models by using high-fidelity data sources?", "answer": ["Llama 2: Open Foundation and Fine-Tuned Chat Models"], "answer_arxiv_id": ["2307.09288"], "source_meta": {"published_time": "20240225"}, "qid": "AutoScholarQuery_train_19694"} +{"question": "Which work originally proposed the one-shot NAS concept?", "answer": ["SMASH: One-Shot Model Architecture Search through HyperNetworks"], "answer_arxiv_id": ["1708.05344"], "source_meta": {"published_time": "20230428"}, "qid": "AutoScholarQuery_train_19695"} +{"question": "What studies describe the discriminative approach in SSL, specifically contrastive learning?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1911.05722", "2002.05709"], "source_meta": {"published_time": "20240129"}, "qid": "AutoScholarQuery_train_19696"} +{"question": "Could you tell me about the research that used unified encoder to represent features from both images and texts?", "answer": ["Unifying Vision, Text, and Layout for Universal Document Processing"], "answer_arxiv_id": ["2212.02623"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19697"} +{"question": "What work established the concept of Structural Re-parameterization?", "answer": ["RepVGG: Making VGG-style ConvNets Great Again"], "answer_arxiv_id": ["2101.03697"], "source_meta": {"published_time": "20220530"}, "qid": "AutoScholarQuery_train_19698"} +{"question": "What research works focused on understanding scenes and object affordances through 2D observations?", "answer": ["Binge Watching: Scaling Affordance Learning from Sitcoms", "Putting Humans in a Scene: Learning Affordance in 3D Indoor Environments", "Grounded Human-Object Interaction Hotspots from Video", "Putting People in Their Place: Affordance-Aware Human Insertion into\n Scenes"], "answer_arxiv_id": ["1804.03080", "1903.05690", "1812.04558", "2304.14406"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_19699"} +{"question": "Which research papers discuss the use of GANs in image generation?", "answer": ["AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks", "DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image Synthesis", "Cross-Modal Contrastive Learning for Text-to-Image Generation"], "answer_arxiv_id": ["1711.10485", "1904.01310", "2101.04702"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_19700"} +{"question": "What are some of the datasets proposed for the Vision-and-Language Navigation task?", "answer": ["Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments", "Stay on the Path: Instruction Fidelity in Vision-and-Language Navigation", "Vision-and-Dialog Navigation", "Touchdown: Natural Language Navigation and Spatial Reasoning in Visual Street Environments", "ALFRED A Benchmark for Interpreting Grounded Instructions for Everyday Tasks", "TEACh: Task-driven Embodied Agents that Chat", "Room-Across-Room: Multilingual Vision-and-Language Navigation with Dense Spatiotemporal Grounding", "REVERIE: Remote Embodied Visual Referring Expression in Real Indoor Environments"], "answer_arxiv_id": ["1711.07280", "1905.12255", "1907.04957", "1811.12354", "1912.01734", "2110.00534", "2010.07954", "1904.10151"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19701"} +{"question": "Which works employed teacher-student approach for unsupervised AD analysis of RGB images?", "answer": ["Uninformed Students: Student-Teacher Anomaly Detection with\n Discriminative Latent Embeddings", "Student-Teacher Feature Pyramid Matching for Anomaly Detection", "Multiresolution Knowledge Distillation for Anomaly Detection", "Anomaly Detection via Reverse Distillation from One-Class Embedding"], "answer_arxiv_id": ["1911.02357", "2103.04257", "2011.11108", "2201.10703"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19702"} +{"question": "Could you list the research that has described the effects of severe occlusions on model fitting?", "answer": ["3D Shape Estimation from 2D Landmarks: A Convex Relaxation Approach", "Towards Scene Understanding with Detailed 3D Object Representations", "ROCA: Robust CAD Model Retrieval and Alignment from a Single Image"], "answer_arxiv_id": ["1411.2942", "1411.5935", "2112.01988"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19703"} +{"question": "What studies presented learning-augmented algorithms for graph or metric problems?", "answer": ["Online metric algorithms with untrusted predictions", "Online Graph Algorithms with Predictions", "Online Facility Location with Predictions"], "answer_arxiv_id": ["2003.02144v3", "2112.11831", "2110.08840"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_19704"} +{"question": "Which works study error curves for Gaussian Processes under posterior predictive losses?", "answer": ["The Shape of Learning Curves: a Review"], "answer_arxiv_id": ["2103.10948"], "source_meta": {"published_time": "20221014"}, "qid": "AutoScholarQuery_train_19705"} +{"question": "Which works improved program synthesis abilities of Large language models (LLMs) by generating programming puzzles?", "answer": ["Language Models Can Teach Themselves to Program Better"], "answer_arxiv_id": ["2207.14502"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_19706"} +{"question": "What works have been done on scoring head design to select the representative template?", "answer": ["MixFormer: End-to-End Tracking with Iterative Mixed Attention", "Learning Spatio-Temporal Transformer for Visual Tracking"], "answer_arxiv_id": ["2203.11082", "2103.17154"], "source_meta": {"published_time": "20240222"}, "qid": "AutoScholarQuery_train_19707"} +{"question": "Could you mention any studies that presented techniques for pruning LLMs to high degrees of sparsity without modifying the remaining weights?", "answer": ["A Simple and Effective Pruning Approach for Large Language Models"], "answer_arxiv_id": ["2306.11695"], "source_meta": {"published_time": "20240519"}, "qid": "AutoScholarQuery_train_19708"} +{"question": "Could you provide me some studies about generalizable Neural Combinatorial Optimization?", "answer": ["Reinforcement Learning for Combinatorial Optimization: A Survey", "Generalize a Small Pre-trained Model to Arbitrarily Large TSP Instances", "Generalization of Neural Combinatorial Solvers Through the Lens of Adversarial Robustness", "On the Generalization of Neural Combinatorial Optimization Heuristics", "DIMES: A Differentiable Meta Solver for Combinatorial Optimization Problems", "Sym-NCO: Leveraging Symmetricity for Neural Combinatorial Optimization", "POMO: Policy Optimization with Multiple Optima for Reinforcement Learning", "Learning to Delegate for Large-scale Vehicle Routing"], "answer_arxiv_id": ["2003.03600", "2012.10658", "2110.10942", "2206.00787", "2210.04123", "2205.13209", "2010.16011", "2107.04139"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_19709"} +{"question": "Could you provide me some related works in the contrastive learning literature that inspired ConSpec?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Representation Learning with Contrastive Predictive Coding", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "1807.03748", "1911.05722"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_19710"} +{"question": "Could you provide me with some studies that focus on logical reasoning ability?", "answer": ["LogiQA: A Challenge Dataset for Machine Reading Comprehension with\n Logical Reasoning", "Transformers as Soft Reasoners over Language", "ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning"], "answer_arxiv_id": ["2007.08124", "2002.05867", "2002.04326"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19711"} +{"question": "Any works that introduced the notion of 'relative uncertainty'?", "answer": ["Pessimistic Minimax Value Iteration: Provably Efficient Equilibrium Learning from Offline Datasets"], "answer_arxiv_id": ["2202.07511"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_19712"} +{"question": "What papers are part of the extensive literature in the field of generic segmentation?", "answer": ["Object Detection in 20 Years: A Survey", "Image Segmentation Using Deep Learning: A Survey"], "answer_arxiv_id": ["1905.05055", "2001.05566"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_19713"} +{"question": "Could you provide me some works that explored the usage of GANs for data-free federated knowledge distillation?", "answer": ["Data-Free Knowledge Distillation for Heterogeneous Federated Learning", "Fine-tuning Global Model via Data-Free Knowledge Distillation for Non-IID Federated Learning"], "answer_arxiv_id": ["2105.10056", "2203.09249"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_19714"} +{"question": "What are some examples of studies in OOD generalization that focus on fine-tuning CLIP models?", "answer": ["Robust fine-tuning of zero-shot models", "Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution", "Finetune like you pretrain: Improved finetuning of zero-shot vision models"], "answer_arxiv_id": ["2109.01903", "2202.10054v1", "2212.00638"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19715"} +{"question": "Could you provide me some studies about the approximations of Taylor expansion of deep neural networks?", "answer": ["Zero-Cost Proxies for Lightweight NAS", "SNIP: Single-shot Network Pruning based on Connection Sensitivity", "Picking Winning Tickets Before Training by Preserving Gradient Flow"], "answer_arxiv_id": ["2101.08134", "1810.02340", "2002.07376"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_19716"} +{"question": "What researches focus on the case of shortcut learning where subpopulations are defined as the product of attributes and labels?", "answer": ["Shortcut Learning in Deep Neural Networks"], "answer_arxiv_id": ["2004.07780"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_19717"} +{"question": "Can you find any works that evaluate the Latent Diffusion Models (LDMs) in object-centric learning?", "answer": ["Object-Centric Slot Diffusion"], "answer_arxiv_id": ["2303.10834"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19718"} +{"question": "What papers have attempted to solve the 3D hand-mesh reconstruction problem by regressing voxels?", "answer": ["Hand Pose Estimation via Latent 2.5D Heatmap Regression", "I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human\n Pose and Mesh Estimation from a Single RGB Image", "InterHand2.6M: A Dataset and Baseline for 3D Interacting Hand Pose\n Estimation from a Single RGB Image"], "answer_arxiv_id": ["1804.09534v1", "2008.03713", "2008.09309"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_19719"} +{"question": "What are examples of MAML-based methods that are typically considered as a synonym of GBML?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-Learning with Implicit Gradients", "Meta-Learning with Latent Embedding Optimization", "Meta-Learning with Adaptive Hyperparameters", "Meta-learning with negative learning rates"], "answer_arxiv_id": ["1703.03400", "1909.04630", "1807.05960", "2011.00209", "2102.00940"], "source_meta": {"published_time": "20231004"}, "qid": "AutoScholarQuery_train_19720"} +{"question": "Provide some works about generative memory-based methods in incremental learning.", "answer": ["Continual Learning with Deep Generative Replay", "Learning to Remember: A Synaptic Plasticity Driven Framework for Continual Learning", "FearNet: Brain-Inspired Model for Incremental Learning", "EEC: Learning to Encode and Regenerate Images for Continual Learning"], "answer_arxiv_id": ["1705.08690", "1904.03137", "1711.10563", "2101.04904"], "source_meta": {"published_time": "20220211"}, "qid": "AutoScholarQuery_train_19721"} +{"question": "Could you mention researches that introduce auxiliary variables in order to adjust the (misspecified) observation to be well-specified?", "answer": ["Misspecification-robust Sequential Neural Likelihood"], "answer_arxiv_id": ["2301.13368"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19722"} +{"question": "Are there any works that select examples that are non-noisy, non-redundant, task-relevant, and reduce the loss on a holdout set the most?", "answer": ["Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt"], "answer_arxiv_id": ["2206.07137"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_19723"} +{"question": "Which works investigate the effectiveness of diffusion models in correspondence estimation tasks?", "answer": ["A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot\n Semantic Correspondence", "Diffusion Hyperfeatures: Searching Through Time and Space for Semantic\n Correspondence", "Unsupervised Semantic Correspondence Using Stable Diffusion", "Emergent Correspondence from Image Diffusion"], "answer_arxiv_id": ["2305.15347", "2305.14334", "2305.15581", "2306.03881"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19724"} +{"question": "Which research learned embeddings of high dimensional pixel data using an unsupervised model for downstream reinforcement learning tasks?", "answer": ["CURL: Contrastive Unsupervised Representations for Reinforcement Learning"], "answer_arxiv_id": ["2004.04136"], "source_meta": {"published_time": "20230109"}, "qid": "AutoScholarQuery_train_19725"} +{"question": "Which studies made great progress in diffusion models?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Score-Based Generative Modeling through Stochastic Differential Equations", "Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation"], "answer_arxiv_id": ["2105.05233", "2011.13456", "2106.05527"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_19726"} +{"question": "Which paper proposed the Visually-Guided Acoustic Synthesis (ViGAS) network?", "answer": ["Novel-View Acoustic Synthesis"], "answer_arxiv_id": ["2301.08730"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_19727"} +{"question": "Could you name some studies about data-driven methods for IMU-based motion capture?", "answer": ["Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse\n Inertial Measurements in Real Time", "TransPose: Real-time 3D Human Translation and Pose Estimation with Six\n Inertial Sensors", "Physical Inertial Poser (PIP): Physics-aware Real-time Human Motion\n Tracking from Sparse Inertial Sensors", "Transformer Inertial Poser: Real-time Human Motion Reconstruction from\n Sparse IMUs with Simultaneous Terrain Generation", "DiffusionPoser: Real-time Human Motion Reconstruction From Arbitrary\n Sparse Sensors Using Autoregressive Diffusion"], "answer_arxiv_id": ["1810.04703", "2105.04605", "2203.08528", "2203.15720", "2308.16682"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_19728"} +{"question": "Which papers discuss the use of conditional generation for class-conditional image generation, language-to-image generation, and reinforcement learning?", "answer": ["Conditional Generative Adversarial Nets", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Decision Transformer: Reinforcement Learning via Sequence Modeling"], "answer_arxiv_id": ["1411.1784", "2204.06125", "2106.01345"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_19729"} +{"question": "Which studies propose frameworks to regularize neural models using symbolic constraints?", "answer": ["Augmenting Neural Networks with First-order Logic"], "answer_arxiv_id": ["1906.06298"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_19730"} +{"question": "Which papers propose the use of generative models to convert discrete space into continuous space in the context of Bayesian optimization?", "answer": ["Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules", "Combining Latent Space and Structured Kernels for Bayesian Optimization over Combinatorial Spaces", "Local Latent Space Bayesian Optimization over Structured Inputs"], "answer_arxiv_id": ["1610.02415", "2111.01186", "2201.11872"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_19731"} +{"question": "Which research utilized genetic algorithms-based methods in optimization-based attacks?", "answer": ["AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language\n Models"], "answer_arxiv_id": ["2310.04451"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_19732"} +{"question": "Can you provide works where diffusion probabilistic models were utilized in image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Diffusion Models Beat GANs on Image Synthesis", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Hierarchical Text-Conditional Image Generation with CLIP Latents"], "answer_arxiv_id": ["2112.10752", "2105.05233", "2205.11487", "2204.06125"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_19733"} +{"question": "What studies increased research on text-to-video models by introducing diffusion models?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2112.10752", "2006.11239"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_19734"} +{"question": "What is the closest study with the researcher's work which uses the ERM to train models that focus on parts of the image with high spurious correlation to the label?", "answer": ["MaskTune: Mitigating Spurious Correlations by Forcing to Explore"], "answer_arxiv_id": ["2210.00055"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19735"} +{"question": "Could you provide me the papers about a benchmark called MiniWoB++?", "answer": ["Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration", "Learning to Navigate the Web", "DOM-Q-NET: Grounded RL on Structured Language", "Environment Generation for Zero-Shot Compositional Reinforcement Learning"], "answer_arxiv_id": ["1802.08802", "1812.09195", "1902.07257", "2201.08896"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_19736"} +{"question": "Could you provide me some works that studied in-context few-shot learning in large language models?", "answer": ["Language Models are Few-Shot Learners", "Larger language models do in-context learning differently"], "answer_arxiv_id": ["2005.14165", "2303.03846v2"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_19737"} +{"question": "Which studies proposed 2D models using popular and often pre-trained backbones for instance segmentation?", "answer": ["Deep Residual Learning for Image Recognition", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["1512.03385", "2103.14030"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_19738"} +{"question": "What papers use the Domain Randomization method to enhance policy robustness?", "answer": ["Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Solving Rubik’s Cube with a Robot Hand", "Understanding Domain Randomization for Sim-to-real Transfer"], "answer_arxiv_id": ["1703.06907", "1910.07113", "2110.03239"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_19739"} +{"question": "What early works demonstrated that an individual's image representation could be inverted from gradients of a deep neural network?", "answer": ["Beyond Inferring Class Representatives: User-Level Privacy Leakage From Federated Learning", "Deep Leakage from Gradients"], "answer_arxiv_id": ["1812.00535", "1906.08935"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_19740"} +{"question": "What papers have proposed codecs applicable to multiple modalities?", "answer": ["COIN++: Neural Compression Across Modalities", "Meta-Learning Sparse Compression Networks", "Modality-Agnostic Variational Compression of Implicit Neural\n Representations", "SHACIRA: Scalable HAsh-grid Compression for Implicit Neural\n Representations"], "answer_arxiv_id": ["2201.12904", "2205.08957", "2301.09479", "2309.15848"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19741"} +{"question": "Which paper explore multiplier bootstrap as a computationally efficient approximation to non-parametric bootstrap in bandit algorithms?", "answer": ["Thompson sampling with the online bootstrap"], "answer_arxiv_id": ["1410.4009"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_19742"} +{"question": "What studies have been conducted for the meta-evaluation of factual consistency metrics in summarization in the general domain?", "answer": ["Understanding Factuality in Abstractive Summarization with FRANK: A\n Benchmark for Factuality Metrics", "TRUE: Re-evaluating Factual Consistency Evaluation", "SummaC: Re-Visiting NLI-based Models for Inconsistency Detection in\n Summarization", "Understanding Factual Errors in Summarization: Errors, Summarizers,\n Datasets, Error Detectors", "FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long\n Form Text Generation"], "answer_arxiv_id": ["2104.13346", "2204.04991", "2111.09525", "2205.12854", "2305.14251"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_19743"} +{"question": "Which works refined unit matching techniques aiming to improve model merging?", "answer": ["Git Re-Basin: Merging Models modulo Permutation Symmetries", "Bayesian Nonparametric Federated Learning of Neural Networks", "Federated Learning with Matched Averaging", "Model Fusion via Optimal Transport", "REPAIR: REnormalizing Permuted Activations for Interpolation Repair", "Deep Model Fusion: A Survey", "ZipIt! Merging Models from Different Tasks without Training", "Optimizing Mode Connectivity via Neuron Alignment", "ModelGiF: Gradient Fields for Model Functional Distance", "On Cross-Layer Alignment for Model Fusion of Heterogeneous Neural\n Networks", "Random initialisations performing above chance and how to find them", "Convergent Learning: Do different neural networks learn the same\n representations?", "Re-basin via implicit Sinkhorn differentiation"], "answer_arxiv_id": ["2209.04836", "1905.12022", "2002.06440", "1910.05653", "2211.08403", "2309.15698", "2305.03053", "2009.02439", "2309.11013", "2110.15538", "2209.07509v2", "1511.07543", "2212.12042"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_19744"} +{"question": "Which studies fall into the category of semantic parsing in the field of controlling robots from natural language?", "answer": ["LM-Nav: Robotic Navigation with Large Pre-Trained Models of Language, Vision, and Action"], "answer_arxiv_id": ["2207.04429"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19745"} +{"question": "What study demonstrated a high probability risk bound for Batch Mirror Decent in linear models?", "answer": ["Stochastic linear optimization never overfits with quadratically-bounded losses on general data"], "answer_arxiv_id": ["2202.06915"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19746"} +{"question": "What studies used in-the-wild datasets for general-purpose visual pre-training?", "answer": ["RRL: Resnet as representation for Reinforcement Learning", "Simple but Effective: CLIP Embeddings for Embodied AI", "Masked Visual Pre-training for Motor Control", "The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "Pre-Trained Image Encoder for Generalizable Visual Reinforcement Learning", "R3M: A Universal Visual Representation for Robot Manipulation", "Where are we in the search for an Artificial Visual Cortex for Embodied Intelligence?"], "answer_arxiv_id": ["2107.03380", "2111.09888", "2203.06173", "2203.03580", "2212.08860", "2203.12601", "2303.18240"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_19747"} +{"question": "Can you provide references on how video latent diffusion models are implemented to avoid excessive computing demands?", "answer": ["Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "ModelScope Text-to-Video Technical Report", "VideoCrafter1: Open Diffusion Models for High-Quality Video Generation"], "answer_arxiv_id": ["2304.08818", "2211.11018", "2308.06571", "2310.19512"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19748"} +{"question": "What research is there on the struggles of VAEs to generate faithful samples?", "answer": ["Tackling the Generative Learning Trilemma with Denoising Diffusion GANs"], "answer_arxiv_id": ["2112.07804"], "source_meta": {"published_time": "20230324"}, "qid": "AutoScholarQuery_train_19749"} +{"question": "Could you provide the work that extract relevant triples from Knowledge Graphs (KGs) and used linear verbalization techniques?", "answer": ["Knowledge-Augmented Language Model Prompting for Zero-Shot Knowledge Graph Question Answering"], "answer_arxiv_id": ["2306.04136v1"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_19750"} +{"question": "Which works developed model-based methods for two-player zero-sum Markov games?", "answer": ["A Sharp Analysis of Model-based Reinforcement Learning with Self-Play", "Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity"], "answer_arxiv_id": ["2010.01604", "2007.07461"], "source_meta": {"published_time": "20221003"}, "qid": "AutoScholarQuery_train_19751"} +{"question": "Could you provide me some works that describe the pair-wise predictive gaps in individual fairness?", "answer": ["Fairness Through Awareness"], "answer_arxiv_id": ["1104.3913"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_19752"} +{"question": "What works have experimented the use of graph curvature in GNNs?", "answer": ["Rewiring Networks for Graph Neural Network Training Using Discrete Geometry"], "answer_arxiv_id": ["2207.08026v1"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_19753"} +{"question": "Could you provide me some studies about image editing methods that manipulate the attentions in the diffusion process according to target prompts?", "answer": ["Prompt-to-Prompt Image Editing with Cross Attention Control", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation"], "answer_arxiv_id": ["2208.01626", "2211.12572"], "source_meta": {"published_time": "20231229"}, "qid": "AutoScholarQuery_train_19754"} +{"question": "Which papers provide methods for bag prediction in computational pathology?", "answer": ["CAMEL: A Weakly Supervised Learning Framework for Histopathology Image\n Segmentation", "Dual-stream Multiple Instance Learning Network for Whole Slide Image\n Classification with Self-supervised Contrastive Learning", "Data Efficient and Weakly Supervised Computational Pathology on Whole\n Slide Images", "TransMIL: Transformer based Correlated Multiple Instance Learning for\n Whole Slide Image Classification", "DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning\n for Histopathology Whole Slide Image Classification", "Cluster-to-Conquer: A Framework for End-to-End Multi-Instance Learning\n for Whole Slide Image Classification"], "answer_arxiv_id": ["1908.10555", "2011.08939", "2004.09666", "2106.00908", "2203.12081", "2103.10626"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19755"} +{"question": "What papers discuss employing data-driven approaches for modeling constraints in contact for hand-object pose estimation?", "answer": ["ArtiBoost: Boosting Articulated 3D Hand-Object Pose Estimation via Online Exploration and Synthesis", "Keypoint Transformer: Solving Joint Identification in Challenging Hands and Object Interactions for Accurate 3D Pose Estimation", "HOPE-Net: A Graph-based Model for Hand-Object Pose Estimation", "Interacting Hand-Object Pose Estimation via Dense Mutual Attention"], "answer_arxiv_id": ["2109.05488", "2104.14639", "2004.00060", "2211.08805"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_19756"} +{"question": "Could you tell me the papers that provide non-asymptotic convergence rate for mirror descent and also discussed optimistic mirror descent?", "answer": ["On Last-Iterate Convergence Beyond Zero-Sum Games"], "answer_arxiv_id": ["2203.12056"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_19757"} +{"question": "What are some examples of research papers that have contributed to region of interest (ROI) methods in video crowd counting?", "answer": ["Locality-constrained Spatial Transformer Network for Video Crowd\n Counting", "Fast Video Crowd Counting with a Temporal Aware Network", "Spatiotemporal Modeling for Crowd Counting in Videos"], "answer_arxiv_id": ["1907.07911", "1907.02198", "1707.07890"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_19758"} +{"question": "In what studies do diffusion models get applied to 3D shapes?", "answer": ["3D Shape Generation and Completion through Point-Voxel Diffusion", "Diffusion-SDF: Text-to-Shape via Voxelized Diffusion", "Diffusion Probabilistic Models for 3D Point Cloud Generation", "LION: Latent Point Diffusion Models for 3D Shape Generation", "Texture Generation on 3D Meshes with Point-UV Diffusion", "MeshDiffusion: Score-based Generative 3D Mesh Modeling", "Shap-E: Generating Conditional 3D Implicit Functions", "Sin3DM: Learning a Diffusion Model from a Single 3D Textured Shape", "3D Neural Field Generation using Triplane Diffusion"], "answer_arxiv_id": ["2104.03670", "2212.03293", "2103.01458", "2210.06978", "2308.10490", "2303.08133", "2305.02463", "2305.15399", "2211.16677"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_19759"} +{"question": "Could you tell me about some works that propose methods for neural operator learning and predict the solution function for a set of PDEs?", "answer": ["Neural Operator: Learning Maps Between Function Spaces With Applications to PDEs"], "answer_arxiv_id": ["2108.08481"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_19760"} +{"question": "Are there any papers where DDPM has been used for semantic segmentation?", "answer": ["Label-Efficient Semantic Segmentation with Diffusion Models", "SegDiff: Image Segmentation with Diffusion Probabilistic Models"], "answer_arxiv_id": ["2112.03126", "2112.00390"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_19761"} +{"question": "Which works argue that experience replay strategy in continual learning could deteriorate the imbalance and overfitting of normalization statistics?", "answer": ["A Comprehensive Survey of Continual Learning: Theory, Method and Application", "Large Scale Incremental Learning", "Memory Replay with Data Compression for Continual Learning"], "answer_arxiv_id": ["2302.00487", "1905.13260", "2202.06592"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_19762"} +{"question": "Which works utilize Hessian information in their pruning methods?", "answer": ["Importance Estimation for Neural Network Pruning"], "answer_arxiv_id": ["1906.10771"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_19763"} +{"question": "What papers are discussing contrastive methods in Graph SSL?", "answer": ["InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization", "Strategies for Pre-training Graph Neural Networks", "Graph Contrastive Learning with Augmentations"], "answer_arxiv_id": ["1908.01000", "1905.12265", "2010.13902"], "source_meta": {"published_time": "20220616"}, "qid": "AutoScholarQuery_train_19764"} +{"question": "Which model used additional temporal Q-former layers in its architecture?", "answer": ["Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video\n Understanding"], "answer_arxiv_id": ["2306.02858"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_19765"} +{"question": "Which work first connected code tasks with pre-trained models?", "answer": ["CodeBERT: A Pre-Trained Model for Programming and Natural Languages"], "answer_arxiv_id": ["2002.08155"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_19766"} +{"question": "Can you provide me with the studies proposing weakly supervised methods based on text annotations only?", "answer": ["GroupViT: Semantic Segmentation Emerges from Text Supervision", "SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary\n Semantic Segmentation", "MaskCLIP: Masked Self-Distillation Advances Contrastive Language-Image\n Pretraining", "Decoupling Zero-Shot Semantic Segmentation", "Image Segmentation Using Text and Image Prompts", "Exploring Open-Vocabulary Semantic Segmentation without Human Labels"], "answer_arxiv_id": ["2202.11094", "2211.14813", "2208.12262", "2112.07910", "2112.10003", "2306.00450"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_19767"} +{"question": "Which studies have proposed confidence intervals for policy evaluation?", "answer": ["CoinDICE: Off-Policy Confidence Interval Estimation"], "answer_arxiv_id": ["2010.11652"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_19768"} +{"question": "What papers discuss the learning and improvement of dynamics model from NetHack text messages?", "answer": ["Improving Policy Learning via Language Dynamics Distillation"], "answer_arxiv_id": ["2210.00066v1"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19769"} +{"question": "What work uses a Lie group of transformations to disentangle data manifolds?", "answer": ["Disentangling by Subspace Diffusion"], "answer_arxiv_id": ["2006.12982"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_19770"} +{"question": "What studies use methods from physics to understand representation learning in neural networks?", "answer": ["Finite size corrections for neural network Gaussian processes", "Asymptotics of Wide Networks from Feynman Diagrams", "Finite Depth and Width Corrections to the Neural Tangent Kernel", "Why bigger is not always better: on finite and infinite neural networks", "Non-Gaussian processes and neural networks at finite widths", "Asymptotics of representation learning in finite Bayesian neural networks", "Exact marginal prior distributions of finite Bayesian neural networks", "The Principles of Deep Learning Theory", "A self consistent theory of Gaussian Processes captures feature learning effects in finite CNNs", "Neural Networks and Quantum Field Theory"], "answer_arxiv_id": ["1908.10030v1", "1909.11304", "1909.05989", "1910.08013", "1910.00019", "2106.00651", "2104.11734", "2106.10165", "2106.04110", "2008.08601v2"], "source_meta": {"published_time": "20210830"}, "qid": "AutoScholarQuery_train_19771"} +{"question": "What study pre-trained a seq2seq VLM with a simple prefix language modeling objective on text generation?", "answer": ["SimVLM: Simple Visual Language Model Pretraining with Weak Supervision"], "answer_arxiv_id": ["2108.10904"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_19772"} +{"question": "Which studies focus on imbalanced classification through techniques such as sample weighting or upsampling?", "answer": ["What is the Effect of Importance Weighting in Deep Learning?", "The Majority Can Help The Minority: Context-rich Minority Oversampling\n for Long-tailed Classification", "Class-Wise Difficulty-Balanced Loss for Solving Class-Imbalance"], "answer_arxiv_id": ["1812.03372", "2112.00412", "2010.01824"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_19773"} +{"question": "Could you provide me the recent monocular methods that incorporate more geometry constraints and extra priors?", "answer": ["M3D-RPN: Monocular 3D Region Proposal Network for Object Detection", "ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape", "RTM3D: Real-time Monocular 3D Detection from Object Keypoints for\n Autonomous Driving", "AutoShape: Real-Time Shape-Aware Monocular 3D Object Detection", "Objects are Different: Flexible Monocular 3D Object Detection", "Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular\n 3D Object Detection", "MonoJSG: Joint Semantic and Geometric Cost Volume for Monocular 3D\n Object Detection"], "answer_arxiv_id": ["1907.06038", "1812.02781", "2001.03343", "2108.11127", "2104.02323", "2205.09373", "2203.08563"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19774"} +{"question": "What studies used models trained on multiple tasks to achieve high performance on a certain task of interest?", "answer": ["Editing Models with Task Arithmetic"], "answer_arxiv_id": ["2212.04089"], "source_meta": {"published_time": "20221214"}, "qid": "AutoScholarQuery_train_19775"} +{"question": "Which research works have used long-form movie content as a data source?", "answer": ["MovieQA: Understanding Stories in Movies through Question-Answering", "MovieNet: A Holistic Dataset for Movie Understanding", "On Negative Sampling for Audio-Visual Contrastive Learning from Movies"], "answer_arxiv_id": ["1512.02902", "2007.10937", "2205.00073"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_19776"} +{"question": "Which work first used GRU based RNNs in sequential recommendations for recommender systems?", "answer": ["Session-based Recommendations with Recurrent Neural Networks"], "answer_arxiv_id": ["1511.06939"], "source_meta": {"published_time": "20230508"}, "qid": "AutoScholarQuery_train_19777"} +{"question": "Could you provide me some studies about instruction tuning in LLMs?", "answer": ["Finetuned Language Models Are Zero-Shot Learners", "Instruction Tuning for Large Language Models: A Survey", "Training language models to follow instructions with human feedback", "Self-Instruct: Aligning Language Models with Self-Generated Instructions", "WizardLM: Empowering Large Language Models to Follow Complex\n Instructions"], "answer_arxiv_id": ["2109.01652", "2308.10792", "2203.02155", "2212.10560", "2304.12244"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_19778"} +{"question": "What papers enhanced the practical application of pretrained models like CLIP and GPT through prompt tuning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["2103.00020", "2005.14165"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_19779"} +{"question": "Which research distinguished between common and private class samples using a one-vs-all classifier for each class?", "answer": ["OVANet: One-vs-All Network for Universal Domain Adaptation"], "answer_arxiv_id": ["2104.03344"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_19780"} +{"question": "Can you name the representative methods which extend Switching Linear Dynamical Systems to the non-linear case?", "answer": ["Collapsed amortized variational inference for switching nonlinear dynamical systems", "Deep Explicit Duration Switching Models for Time Series"], "answer_arxiv_id": ["1910.09588", "2110.13878"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_19781"} +{"question": "What researches involved the use of neural networks to learn depth from stereo images?", "answer": ["Learning monocular depth estimation with unsupervised trinocular\n assumptions", "Refine and Distill: Exploiting Cycle-Inconsistency and Knowledge\n Distillation for Unsupervised Monocular Depth Estimation", "Self-Supervised Monocular Depth Hints", "PLADE-Net: Towards Pixel-Level Accuracy for Self-Supervised Single-View\n Depth Estimation with Neural Positional Encoding and Distilled Matting Loss", "Self-distilled Feature Aggregation for Self-supervised Monocular Depth\n Estimation"], "answer_arxiv_id": ["1808.01606", "1903.04202", "1909.09051", "2103.07362", "2209.07088"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_19782"} +{"question": "What studies are about addressing the IL problem with morphological mismatch?", "answer": ["Domain Adaptive Imitation Learning"], "answer_arxiv_id": ["1910.00105v2"], "source_meta": {"published_time": "20210617"}, "qid": "AutoScholarQuery_train_19783"} +{"question": "Which work proposed a new vision-language foundation model with a masked multi-modal modeling objective?", "answer": ["FLAVA: A Foundational Language And Vision Alignment Model"], "answer_arxiv_id": ["2112.04482"], "source_meta": {"published_time": "20220615"}, "qid": "AutoScholarQuery_train_19784"} +{"question": "Can you tell me about the studies where they use a hierarchical manner to aggregate local neighborhood information in point-based networks?", "answer": ["Rethinking Network Design and Local Geometry in Point Cloud: A Simple\n Residual MLP Framework", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "SO-Net: Self-Organizing Network for Point Cloud Analysis", "Point Transformer", "PointNeXt: Revisiting PointNet++ with Improved Training and Scaling\n Strategies", "RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds", "Not All Points Are Equal: Learning Highly Efficient Point-based\n Detectors for 3D LiDAR Point Clouds", "3DSSD: Point-based 3D Single Stage Object Detector", "PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation"], "answer_arxiv_id": ["2202.07123", "1706.02413", "1803.04249", "2012.09164", "2206.04670", "1911.11236", "2203.11139", "2002.10187", "1612.00593"], "source_meta": {"published_time": "20230321"}, "qid": "AutoScholarQuery_train_19785"} +{"question": "Which works explored input-space backdoor attacks using simple trigger patterns such as a single pixel or random noise?", "answer": ["Spectral Signatures in Backdoor Attacks", "BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain", "Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning"], "answer_arxiv_id": ["1811.00636", "1708.06733", "1712.05526v1"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19786"} +{"question": "Which works are about differentially private gradient-based MCMC?", "answer": ["On Connecting Stochastic Gradient MCMC and Differential Privacy", "Differentially Private Bayesian Neural Networks on Accuracy, Privacy and Reliability", "Differentially Private Hamiltonian Monte Carlo"], "answer_arxiv_id": ["1712.09097v1", "2107.08461", "2106.09376"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_19787"} +{"question": "Could you provide me some examples where the attention mechanism is used in the field of 3D object detection?", "answer": ["BEVFormer: Learning Bird’s-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers"], "answer_arxiv_id": ["2203.17270"], "source_meta": {"published_time": "20221117"}, "qid": "AutoScholarQuery_train_19788"} +{"question": "Which works proposed biologically plausible ICA networks for solving the BSS problem?", "answer": ["A Normative and Biologically Plausible Algorithm for Independent Component Analysis", "Biologically plausible single-layer networks for nonnegative independent component analysis"], "answer_arxiv_id": ["2111.08858", "2010.12632"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_19789"} +{"question": "What paper introduced denoising score matching to enhance the score approximation accuracy?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["1907.05600"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_19790"} +{"question": "Could you provide me with studies that focus on conditional diffusion generation, specifically ones that include text conditions?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10741", "2112.10752"], "source_meta": {"published_time": "20240204"}, "qid": "AutoScholarQuery_train_19791"} +{"question": "Could you provide me some studies that used the attention mechanism to approximate the complexity of Vision Transformer to linear?", "answer": ["Transformers are RNNs: Fast Autoregressive Transformers with Linear\n Attention", "Fast Transformers with Clustered Attention", "Linformer: Self-Attention with Linear Complexity", "Reformer: The Efficient Transformer", "Big Bird: Transformers for Longer Sequences"], "answer_arxiv_id": ["2006.16236", "2007.04825", "2006.04768", "2001.04451", "2007.14062"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_19792"} +{"question": "What paper discusses the application of diffusion models in the real-number vector space and the simplex space?", "answer": ["Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning", "Ssd-LM: Semi-autoregressive Simplex-based Diffusion Language Model for Text Generation and Modular Control"], "answer_arxiv_id": ["2208.04202", "2210.17432"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_19793"} +{"question": "Could you provide me some studies that have used watermarking for protecting diffusion models?", "answer": ["The Stable Signature: Rooting Watermarks in Latent Diffusion Models", "A Watermark for Large Language Models", "A Recipe for Watermarking Diffusion Models", "Watermarking Diffusion Model", "DiffusionShield: A Watermark for Copyright Protection against Generative\n Diffusion Models", "Reverse Engineering of Generative Models: Inferring Model\n Hyperparameters from Generated Images", "Invisible Watermarking for Audio Generation Diffusion Models"], "answer_arxiv_id": ["2303.15435", "2301.10226", "2303.10137", "2305.12502", "2306.04642", "2106.07873", "2309.13166"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_19794"} +{"question": "What paper discusses ART-BP that uses randomly sampled truncation lengths?", "answer": ["Unbiasing Truncated Backpropagation Through Time"], "answer_arxiv_id": ["1705.08209"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_19795"} +{"question": "Can you name some papers where diffusion models are used for image translation?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation", "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations", "EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations"], "answer_arxiv_id": ["2108.02938", "2110.02711", "2108.01073", "2207.06635"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19796"} +{"question": "Could you provide some research papers about the implementation of Diffusion-based methods to obtain state-of-the-art performances in Text-to-image generation?", "answer": ["CogView2: Faster and Better Text-to-Image Generation via Hierarchical\n Transformers", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers"], "answer_arxiv_id": ["2204.14217", "2112.10741", "2204.06125", "2112.10752", "2205.11487", "2211.01324"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_19797"} +{"question": "What paper introduced the approach of matching an updated policy to the visit count distribution, a procedure used in Gumbel AlphaZero?", "answer": ["Monte-Carlo tree search as regularized policy optimization"], "answer_arxiv_id": ["2007.12509"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_19798"} +{"question": "Could you mention studies that have focused on learning lower-resolution synthetic images and upsampling?", "answer": ["Remember the Past: Distilling Datasets into Addressable Memories for Neural Networks", "Dataset Condensation with Contrastive Signals"], "answer_arxiv_id": ["2206.02916", "2202.02916"], "source_meta": {"published_time": "20221119"}, "qid": "AutoScholarQuery_train_19799"} +{"question": "What works are in line with the first type of methodology to solve inverse problems which is iterative or optimization-based methods?", "answer": ["A Fixed Mesh Method With Immersed Finite Elements for Solving Interface Inverse Problems", "Mathematical and numerical study of a three-dimensional inverse eddy current problem", "Numerical Solution of Inverse Problems by Weak Adversarial Networks"], "answer_arxiv_id": ["1805.03255v1", "1908.08683", "2002.11340"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_19800"} +{"question": "What works used Lookahead algorithm in the context of Federated Averaging and the Reptile algorithm?", "answer": ["Communication-Efficient Learning of Deep Networks from Decentralized Data", "On First-Order Meta-Learning Algorithms"], "answer_arxiv_id": ["1602.05629", "1803.02999"], "source_meta": {"published_time": "20231020"}, "qid": "AutoScholarQuery_train_19801"} +{"question": "Can you mention the works that commonly employ Bird’s-eye-view (BEV) representations?", "answer": ["Orthographic Feature Transform for Monocular 3D Object Detection", "Categorical Depth Distribution Network for Monocular 3D Object Detection", "BEVDet: High-performance Multi-camera 3D Object Detection in\n Bird-Eye-View", "PETR: Position Embedding Transformation for Multi-View 3D Object\n Detection", "BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers", "BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object\n Detection"], "answer_arxiv_id": ["1811.08188", "2103.01100", "2112.11790", "2203.05625", "2203.17270", "2206.10092"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19802"} +{"question": "Which works focused on incorporating aspects, sentiments, topics, and contrasting opinions in abstractive summarization of reviews?", "answer": ["Unsupervised Opinion Summarization with Content Planning", "OpinionDigest: A Simple Framework for Opinion Summarization", "Comparative Opinion Summarization via Collaborative Decoding"], "answer_arxiv_id": ["2012.07808v1", "2005.01901", "2110.07520"], "source_meta": {"published_time": "20240611"}, "qid": "AutoScholarQuery_train_19803"} +{"question": "What works popularized the supervised setting of pretraining deep models?", "answer": ["Rich feature hierarchies for accurate object detection and semantic segmentation", "DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition", "Visualizing and Understanding Convolutional Networks", "OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks"], "answer_arxiv_id": ["1311.2524", "1310.1531v1", "1311.2901", "1312.6229"], "source_meta": {"published_time": "20220401"}, "qid": "AutoScholarQuery_train_19804"} +{"question": "Which studies are related to the development of intelligent driving systems such as scene analysis?", "answer": ["Toward Driving Scene Understanding: A Dataset for Learning Driver\n Behavior and Causal Reasoning", "Spatiotemporal Relationship Reasoning for Pedestrian Intent Prediction", "Learning Road Scene-level Representations via Semantic Region Prediction"], "answer_arxiv_id": ["1811.02307", "2002.08945", "2301.00714"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19805"} +{"question": "Which papers highlight the issue of increased computational complexity in calculating KL divergence due to usage of isotropic or diagonal covariance in distributions?", "answer": ["A Rotated Hyperbolic Wrapped Normal Distribution for Hierarchical Representation Learning"], "answer_arxiv_id": ["2205.13371"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19806"} +{"question": "Which studies have introduced models that can generate high-quality arbitrary-resolution images?", "answer": ["Unlimited-Size Diffusion Restoration", "MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generation"], "answer_arxiv_id": ["2303.00354", "2302.08113"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_19807"} +{"question": "Could you provide studies that explain the advantage of using a large learning rate in SGD?", "answer": ["Implicit Gradient Regularization", "On the Origin of Implicit Regularization in Stochastic Gradient Descent"], "answer_arxiv_id": ["2009.11162", "2101.12176"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_19808"} +{"question": "In the field of HOI detection, can you name studies that deal with pairwise relatedness in two-stage methods?", "answer": ["Cascaded Human-Object Interaction Recognition"], "answer_arxiv_id": ["2003.04262"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_19809"} +{"question": "Are there any works that used language to provide inter-class correlations in visual recognition?", "answer": ["What does a platypus look like? Generating customized prompts for\n zero-shot image classification", "Visual Classification via Description from Large Language Models", "Multi-Modal Classifiers for Open-Vocabulary Object Detection", "MAtch, eXpand and Improve: Unsupervised Finetuning for Zero-Shot Action\n Recognition with Language Knowledge"], "answer_arxiv_id": ["2209.03320", "2210.07183", "2306.05493", "2303.08914"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19810"} +{"question": "Can you specify some works that introduce new methods for CRL?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Continual Learning Through Synaptic Intelligence", "Memory Aware Synapses: Learning what (not) to forget", "Gradient Episodic Memory for Continual Learning", "Efficient Lifelong Learning with A-GEM"], "answer_arxiv_id": ["1612.00796", "1703.04200", "1711.09601", "1706.08840", "1812.00420"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_19811"} +{"question": "Any works about data-dependent approaches that can achieve non-vacuouys bounds in neural networks?", "answer": ["Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data", "Non-vacuous Generalization Bounds at the ImageNet Scale: a PAC-Bayesian Compression Approach", "Information-Theoretic Generalization Bounds for SGLD via Data-Dependent Estimates"], "answer_arxiv_id": ["1703.11008", "1804.05862", "1911.02151v3"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_19812"} +{"question": "Which studies proposed generating a random simplex ETF matrix to replace the original classifier in imbalanced scenarios?", "answer": ["Inducing Neural Collapse in Imbalanced Learning: Do We Really Need a\n Learnable Classifier at the End of Deep Neural Network?", "Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class\n Incremental Learning"], "answer_arxiv_id": ["2203.09081", "2302.03004"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_19813"} +{"question": "Which studies used the TensoRF representation for efficient computation of visibility and indirect lighting?", "answer": ["TensoIR: Tensorial Inverse Rendering", "TensoRF: Tensorial Radiance Fields"], "answer_arxiv_id": ["2304.12461", "2203.09517"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_19814"} +{"question": "Which works discussed the creation of image-based distribution shift benchmarks?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["1903.12261"], "source_meta": {"published_time": "20230619"}, "qid": "AutoScholarQuery_train_19815"} +{"question": "What works have proposed algorithms for machine unlearning?", "answer": ["Making AI Forget You: Data Deletion in Machine Learning", "Certified Data Removal from Machine Learning Models", "Approximate Data Deletion from Machine Learning Models", "Descent-to-Delete: Gradient-Based Methods for Machine Unlearning", "Machine Unlearning via Algorithmic Stability", "Remember What You Want to Forget: Algorithms for Machine Unlearning", "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks", "Mixed-Privacy Forgetting in Deep Networks", "Variational Bayesian Unlearning", "Machine Unlearning"], "answer_arxiv_id": ["1907.05012", "1911.03030", "2002.10077", "2007.02923v1", "2102.13179", "2103.03279v2", "1911.04933", "2012.13431", "2010.12883", "1912.03817v3"], "source_meta": {"published_time": "20220630"}, "qid": "AutoScholarQuery_train_19816"} +{"question": "Which studies discussed hyperbolic graph neural networks?", "answer": ["Hyperbolic Graph Neural Networks", "Hyperbolic Graph Convolutional Neural Networks"], "answer_arxiv_id": ["1910.12892", "1910.12933"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_19817"} +{"question": "What works employ imputation-based techniques to tackle irregular sampling in time series analysis?", "answer": ["Multi-Time Attention Networks for Irregularly Sampled Time Series", "Interpolation-Prediction Networks for Irregularly Sampled Time Series", "Dynamic Gaussian Mixture based Deep Generative Model For Robust Forecasting on Sparse Multivariate Time Series"], "answer_arxiv_id": ["2101.10318", "1909.07782", "2103.02164"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_19818"} +{"question": "What work relies on the Gaussian equivalence conjecture studied and extensively used in the random features model?", "answer": ["The Gaussian equivalence of generative models for learning with shallow neural networks", "Universality Laws for High-Dimensional Learning with Random Features", "Universality of empirical risk minimization", "The curse of overparametrization in adversarial training: Precise analysis of robust generalization for random features regression", "Learning curves of generic features maps for realistic datasets with a teacher-student model", "On the interplay between data structure and loss function in classification problems"], "answer_arxiv_id": ["2006.14709", "2009.07669", "2202.08832", "2201.05149", "2102.08127", "2103.05524"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19819"} +{"question": "What are the studies that address class imbalance using two-stage fine-tuning and meta-learning methods?", "answer": ["Large-Scale Long-Tailed Recognition in an Open World", "Factors in Finetuning Deep Model for Object Detection with Long-tail Distribution", "Decoupling Representation and Classifier for Long-Tailed Recognition", "MetaBalance: High-Performance Neural Networks for Class-Imbalanced Data"], "answer_arxiv_id": ["1904.05160", "1601.05150", "1910.09217", "2106.09643"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19820"} +{"question": "What papers have proposed different memory architectures in the field of AI?", "answer": ["Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation", "Neural Turing Machines", "Memory Networks", "End-To-End Memory Networks"], "answer_arxiv_id": ["1406.1078", "1410.5401", "1410.3916", "1503.08895"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_19821"} +{"question": "Which papers deal with convergence analysis on over-parameterized neural networks that involve data separability?", "answer": ["Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Data", "On the Convergence Rate of Training Recurrent Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Learning and Generalization in Overparameterized Neural Networks, Going Beyond Two Layers"], "answer_arxiv_id": ["1808.01204", "1810.12065", "1811.03962", "1811.04918"], "source_meta": {"published_time": "20220807"}, "qid": "AutoScholarQuery_train_19822"} +{"question": "What papers have worked on generating plausible motion from textual descriptions before diffusion models gained popularity?", "answer": ["Executing your Commands via Motion Diffusion in Latent Space", "T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete\n Representations", "Make-An-Animation: Large-Scale Text-conditional 3D Human Motion\n Generation", "TEMOS: Generating diverse human motions from textual descriptions", "FLAME: Free-form Language-based Motion Synthesis & Editing"], "answer_arxiv_id": ["2212.04048", "2301.06052", "2305.09662", "2204.14109", "2209.00349"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19823"} +{"question": "Could you provide me some studies about KD for model compression?", "answer": ["On the Efficacy of Knowledge Distillation", "Improved Knowledge Distillation via Teacher Assistant"], "answer_arxiv_id": ["1910.01348", "1902.03393"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_19824"} +{"question": "Which research implicitly explains that model smoothing is the reason why DreamerV2 can achieve good performance without any complicated exploration mechanism?", "answer": ["Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["2010.02193"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_19825"} +{"question": "What paper conducted experiments to generate audio-driven talking faces with the utilization of an adversarial process?", "answer": ["A Lip Sync Expert Is All You Need for Speech to Lip Generation In The\n Wild"], "answer_arxiv_id": ["2008.10010"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_19826"} +{"question": "Could you provide works on using the spectral method for initialization in low-rank matrix estimation?", "answer": ["Spectral Methods for Data Science: A Statistical Perspective"], "answer_arxiv_id": ["2012.08496"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19827"} +{"question": "Could you tell me the papers that studied deep learning-based image reconstruction methods for spike cameras outperforming traditional methods by a significant margin?", "answer": ["Recurrent Spike-based Image Restoration under General Illumination"], "answer_arxiv_id": ["2308.03018"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_19828"} +{"question": "Which papers made use of generative models like Generative Adversarial Networks (GANs) in forecasting future trajectories?", "answer": ["Social GAN: Socially Acceptable Trajectories with Generative Adversarial\n Networks", "SoPhie: An Attentive GAN for Predicting Paths Compliant to Social and\n Physical Constraints", "Conditional Generative Neural System for Probabilistic Trajectory\n Prediction", "Social Ways: Learning Multi-Modal Distributions of Pedestrian\n Trajectories with GANs", "Social-BiGAT: Multimodal Trajectory Forecasting using Bicycle-GAN and\n Graph Attention Networks"], "answer_arxiv_id": ["1803.10892", "1806.01482", "1905.01631", "1904.09507", "1907.03395"], "source_meta": {"published_time": "20240331"}, "qid": "AutoScholarQuery_train_19829"} +{"question": "Which works discuss the techniques of strongly adaptive regret minimization that are used in online learning?", "answer": ["Strongly Adaptive Online Learning", "Improved Strongly Adaptive Online Learning using Coin Betting", "Dynamic Regret of Strongly Adaptive Methods"], "answer_arxiv_id": ["1502.07073", "1610.04578", "1701.07570"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_19830"} +{"question": "Can you name the study that identified activation outliers as the major quantization bottleneck for LLMs?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale"], "answer_arxiv_id": ["2208.07339"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_19831"} +{"question": "Which papers developed models via Distributionally Robust Optimization (DRO) that optimizes samples that most violate fairness?", "answer": ["Training individually fair ML models with Sensitive Subspace Robustness", "SenSeI: Sensitive Set Invariance for Enforcing Individual Fairness", "Learning Certified Individually Fair Representations", "Individual Fairness Revisited: Transferring Techniques from Adversarial Robustness"], "answer_arxiv_id": ["1907.00020", "2006.14168", "2002.10312", "2002.07738"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_19832"} +{"question": "What works have introduced motion modeling modules to animate T2I models once for all?", "answer": ["AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning"], "answer_arxiv_id": ["2307.04725"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_19833"} +{"question": "What studies focus on action poisoning attacks?", "answer": ["Efficient Action Poisoning Attacks on Linear Contextual Bandits"], "answer_arxiv_id": ["2112.05367"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_19834"} +{"question": "What papers cover the studies on model-based methods for robust discounted MDPs?", "answer": ["Distributionally Robust Counterpart in Markov Decision Processes"], "answer_arxiv_id": ["1501.07418"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_19835"} +{"question": "Could you provide studies related to the perturbed history exploration algorithm?", "answer": ["Perturbed-History Exploration in Stochastic Multi-Armed Bandits", "Perturbed-History Exploration in Stochastic Linear Bandits", "Randomized Exploration in Generalized Linear Bandits"], "answer_arxiv_id": ["1902.10089", "1903.09132", "1906.08947"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_19836"} +{"question": "What studies have showcased recent unified models that offer support for image locations, enabling tasks like object detection or region captioning?", "answer": ["Unifying Vision-and-Language Tasks via Text Generation", "Webly Supervised Concept Expansion for General Purpose Vision Models", "Towards General Purpose Vision Systems: An End-to-End Task-Agnostic Vision-Language Architecture", "Coarse-to-Fine Vision-Language Pre-training with Fusion in the Backbone", "A Unified Sequence Interface for Vision Tasks", "MDETR - Modulated Detection for End-to-End Multi-Modal Understanding", "Grounded Language-Image Pre-training"], "answer_arxiv_id": ["2102.02779", "2202.02317", "2104.00743", "2206.07643", "2206.07669", "2104.12763", "2112.03857"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_19837"} +{"question": "What studies have used selfies for reposing applications?", "answer": ["Unselfie: Translating Selfies to Neutral-pose Portraits in the Wild"], "answer_arxiv_id": ["2007.15068"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_19838"} +{"question": "What papers have detailed the connection between attention mechanism and random feature in machine learning?", "answer": ["Random Feature Attention"], "answer_arxiv_id": ["2103.02143"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_19839"} +{"question": "Could you list the papers that discuss the use of pseudo-3D convolutions to reduce computational costs?", "answer": ["AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models\n without Specific Tuning", "Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models", "Structure and Content-Guided Video Synthesis with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Structure and Content-Guided Video Synthesis with Diffusion Models"], "answer_arxiv_id": ["2307.04725", "2305.10474v3", "2302.03011", "2209.14792", "2302.03011"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_19840"} +{"question": "Can you list some research that have made attempts to decompose raw videos into temporally aligned slots?", "answer": ["Conditional Object-Centric Learning from Video", "SIMONe: View-Invariant, Temporally-Abstracted Object Representations via Unsupervised Video Decomposition", "Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos"], "answer_arxiv_id": ["2111.12594", "2106.03849", "2205.14065"], "source_meta": {"published_time": "20221012"}, "qid": "AutoScholarQuery_train_19841"} +{"question": "What are some works that introduced video inpainting techniques?", "answer": ["FuseFormer: Fusing Fine-Grained Information in Transformers for Video\n Inpainting", "Free-form Video Inpainting with 3D Gated Convolution and Temporal\n PatchGAN", "Video Inpainting by Jointly Learning Temporal Structure and Spatial\n Details", "Towards An End-to-End Framework for Flow-Guided Video Inpainting", "Flow-Guided Transformer for Video Inpainting", "Short-Term and Long-Term Context Aggregation Network for Video\n Inpainting", "ProPainter: Improving Propagation and Transformer for Video Inpainting"], "answer_arxiv_id": ["2109.02974", "1904.10247", "1806.08482", "2204.02663", "2208.06768", "2009.05721", "2309.03897"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_19842"} +{"question": "Are there any research aimed at zero-shot object goal navigation?", "answer": ["CoWs on Pasture: Baselines and Benchmarks for Language-Driven Zero-Shot Object Navigation", "Zero Experience Required: Plug & Play Modular Transfer Learning for Semantic Visual Navigation", "ProcTHOR: Large-Scale Embodied AI Using Procedural Generation"], "answer_arxiv_id": ["2203.10421", "2202.02440", "2206.06994"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_19843"} +{"question": "What are some studies demonstrating the promising results of LVMs in uncovering the low-dimensional structure underlying complex neural activities?", "answer": ["Learning identifiable and interpretable latent models of high-dimensional neural activity using pi-VAE", "Drop, Swap, and Generate: A Self-Supervised Approach for Generating Neural Activity"], "answer_arxiv_id": ["2011.04798", "2111.02338"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_19844"} +{"question": "Which studies propose dimensionality reduction techniques via PCA methods for k-means clustering?", "answer": ["Dimensionality Reduction for k-Means Clustering and Low Rank Approximation", "Strong Coresets for k-Median and Subspace Approximation: Goodbye Dimension"], "answer_arxiv_id": ["1410.6801v3", "1809.02961v2"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_19845"} +{"question": "Which studies explore other types of guidance in guided diffusions?", "answer": ["GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models", "Diffusion models as plug-and-play priors"], "answer_arxiv_id": ["2112.10741", "2206.09012"], "source_meta": {"published_time": "20230713"}, "qid": "AutoScholarQuery_train_19846"} +{"question": "Which research works are involved in the field of image-to-3D reconstruction?", "answer": ["NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as\n General Image Priors", "Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion\n Prior", "Magic123: One Image to High-Quality 3D Object Generation Using Both 2D\n and 3D Diffusion Priors"], "answer_arxiv_id": ["2212.03267", "2303.14184", "2306.17843"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_19847"} +{"question": "Which studies have employed deep learning methods for direct design in the engineering field?", "answer": ["Topology optimization of 2D structures with nonlinearities using deep learning", "Real-Time Topology Optimization in 3D via Deep Transfer Learning"], "answer_arxiv_id": ["2002.01896v4", "2102.07657"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_19848"} +{"question": "Could you provide me some works that generate future trajectories for already-existing agents in simulation?", "answer": ["TrajGen: Generating Realistic and Diverse Trajectories with Reactive and Feasible Agent Behaviors for Autonomous Driving", "TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors", "Guided Conditional Diffusion for Controllable Traffic Simulation", "BITS: Bi-level Imitation for Traffic Simulation", "Symphony: Learning Realistic and Diverse Agents for Autonomous Driving Simulation"], "answer_arxiv_id": ["2203.16792", "2101.06557", "2210.17366", "2208.12403", "2205.03195"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_19849"} +{"question": "Which studies utilized polar curvature between collections of points to develop a spectral curvature clustering (SCC) algorithm?", "answer": ["Foundations of a Multi-way Spectral Clustering Framework for Hybrid Linear Modeling"], "answer_arxiv_id": ["0810.3724v2"], "source_meta": {"published_time": "20210728"}, "qid": "AutoScholarQuery_train_19850"} +{"question": "Could you tell me what works adopted the approach of optical flow estimation and image warping in flow-based VFI methods?", "answer": ["Spatial Transformer Networks", "Super SloMo: High Quality Estimation of Multiple Intermediate Frames for\n Video Interpolation", "Quadratic video interpolation", "All at Once: Temporally Adaptive Multi-Frame Interpolation with Advanced\n Motion Modeling", "Time Lens++: Event-based Frame Interpolation with Parametric Non-linear\n Flow and Multi-scale Fusion"], "answer_arxiv_id": ["1506.02025", "1712.00080", "1911.00627", "2007.11762", "2203.17191"], "source_meta": {"published_time": "20230831"}, "qid": "AutoScholarQuery_train_19851"} +{"question": "What papers discuss the approach of joint modeling state-action trajectories?", "answer": ["Prompting Decision Transformer for Few-Shot Policy Generalization"], "answer_arxiv_id": ["2206.13499"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19852"} +{"question": "Could you provide me some studies about adapter tuning?", "answer": ["Compacter: Efficient Low-Rank Hypercomplex Adapter Layers", "AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition", "VL-Adapter: Parameter-Efficient Transfer Learning for Vision-and-Language Tasks"], "answer_arxiv_id": ["2106.04647", "2205.13535", "2112.06825"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_19853"} +{"question": "Which papers proposed different prompt learning strategies in natural language processing?", "answer": ["Fantastically Ordered Prompts and Where to Find Them: Overcoming\n Few-Shot Prompt Order Sensitivity", "Language Models are Few-Shot Learners", "How Can We Know What Language Models Know?", "Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2104.08786", "2005.14165", "1911.12543", "2203.02155"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_19854"} +{"question": "What work discussed the limitations of approximating the whole covariance matrix on small models?", "answer": ["Efficient Full-Matrix Adaptive Regularization"], "answer_arxiv_id": ["1806.02958"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_19855"} +{"question": "What papers are about volume rendering?", "answer": ["UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction"], "answer_arxiv_id": ["2104.10078", "2106.10689", "2106.12052", "2206.00665"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_19856"} +{"question": "Who worked on maximizing Chow's excess risk?", "answer": ["Online Active Learning of Reject Option Classifiers", "Exponential Savings in Agnostic Active Learning through Abstention", "Efficient Active Learning with Abstention"], "answer_arxiv_id": ["1906.06166", "2102.00451v3", "2204.00043"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_19857"} +{"question": "What are some studies that focused on improving the architectures of generators and discriminators in GANs?", "answer": ["Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks", "Improved Training of Wasserstein GANs", "Large Scale GAN Training for High Fidelity Natural Image Synthesis", "Progressive Growing of GANs for Improved Quality, Stability, and Variation", "A Style-Based Generator Architecture for Generative Adversarial Networks", "StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets", "StyleGAN-T: Unlocking the Power of GANs for Fast Large-Scale Text-to-Image Synthesis", "Scaling up GANs for Text-to-Image Synthesis"], "answer_arxiv_id": ["1511.06434", "1704.00028", "1809.11096", "1710.10196", "1812.04948", "2202.00273", "2301.09515", "2303.05511"], "source_meta": {"published_time": "20230829"}, "qid": "AutoScholarQuery_train_19858"} +{"question": "What are some studies on automated test generation techniques such as fuzz testing?", "answer": ["On the Unusual Effectiveness of Type-Aware Operator Mutations for Testing SMT Solvers"], "answer_arxiv_id": ["2004.08799"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_19859"} +{"question": "What research has been done on pruning and approximating pre-trained parameter tensors for efficient inference?", "answer": ["The State of Sparsity in Deep Neural Networks", "A Survey of Quantization Methods for Efficient Neural Network Inference"], "answer_arxiv_id": ["1902.09574", "2103.13630"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_19860"} +{"question": "What are the earlier methods used to tackle under-constrained nature in human portrait relighting?", "answer": ["Shape, Illumination, and Reflectance from Shading", "SfSNet: Learning Shape, Reflectance and Illuminance of Faces in the Wild"], "answer_arxiv_id": ["2010.03592", "1712.01261"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19861"} +{"question": "What studies propose techniques to modify the data collection process to increase state diversity?", "answer": ["A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning", "Uncertainty-Aware Data Aggregation for Deep Imitation Learning", "ThriftyDAgger: Budget-Aware Novelty and Risk Gating for Interactive Imitation Learning", "HG-DAgger: Interactive Imitation Learning with Human Experts", "Human-in-the-Loop Imitation Learning using Remote Teleoperation", "Eliciting Compatible Demonstrations for Multi-Human Imitation Learning", "DART: Noise Injection for Robust Imitation Learning"], "answer_arxiv_id": ["1011.0686v3", "1905.02780", "2109.08273", "1810.02890", "2012.06733", "2210.08073v1", "1703.09327"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_19862"} +{"question": "What studies discuss instances when float16 created issues during the training of large-scale models?", "answer": ["OPT: Open Pre-trained Transformer Language Models", "BLOOM: A 176B-Parameter Open-Access Multilingual Language Model"], "answer_arxiv_id": ["2205.01068", "2211.05100"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_19863"} +{"question": "Could you provide some references about the application of diffusion model into text-driven motion generation field?", "answer": ["MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model", "Human Motion Diffusion Model", "FLAME: Free-form Language-based Motion Synthesis & Editing", "Executing your Commands via Motion Diffusion in Latent Space", "ReMoDiffuse: Retrieval-Augmented Motion Diffusion Model"], "answer_arxiv_id": ["2208.15001", "2209.14916", "2209.00349", "2212.04048", "2304.01116"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_19864"} +{"question": "Could you provide me the research that uses KL decomposition to approximate data in spectral space while learning a standard diffusion model?", "answer": ["Spectral Diffusion Processes"], "answer_arxiv_id": ["2209.14125"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_19865"} +{"question": "Which studies focus on traditional methods to learn disentangled representation?", "answer": ["Disentangling Factors of Variation via Generative Entangling", "InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets"], "answer_arxiv_id": ["1210.5474", "1606.03657"], "source_meta": {"published_time": "20230902"}, "qid": "AutoScholarQuery_train_19866"} +{"question": "Which works contributed to the field of video GANs?", "answer": ["Generating Videos with Scene Dynamics", "MoCoGAN: Decomposing Motion and Content for Video Generation", "Adversarial Video Generation on Complex Datasets", "StyleGAN-V: A Continuous Video Generator with the Price, Image Quality and Perks of StyleGAN2"], "answer_arxiv_id": ["1609.02612", "1707.04993", "1907.06571", "2112.14683"], "source_meta": {"published_time": "20221121"}, "qid": "AutoScholarQuery_train_19867"} +{"question": "Which works employed GANs to generate images from keypoints?", "answer": ["LatentKeypointGAN: Controlling Images via Latent Keypoints", "GANSeg: Learning to Segment by Unsupervised Hierarchical Image\n Generation"], "answer_arxiv_id": ["2103.15812v5", "2112.01036"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_19868"} +{"question": "What works proposed program synthesis methods that take input-output examples?", "answer": ["DeepCoder: Learning to Write Programs", "SpreadsheetCoder: Formula Prediction from Semi-structured Context"], "answer_arxiv_id": ["1611.01989", "2106.15339"], "source_meta": {"published_time": "20220412"}, "qid": "AutoScholarQuery_train_19869"} +{"question": "Which papers propose re-sampling and re-weighting strategies for supervising learning with imbalanced datasets?", "answer": ["Relay Backpropagation for Effective Learning of Deep Convolutional Neural Networks", "Deep Active Learning over the Long Tail", "Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss", "Rethinking Class-Balanced Methods for Long-Tailed Visual Recognition from a Domain Adaptation Perspective"], "answer_arxiv_id": ["1512.05830", "1711.00941", "1906.07413", "2003.10780"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_19870"} +{"question": "Which studies adopted the concept of active learning in the context of debiasing?", "answer": ["A Survey of Active Learning for Natural Language Processing"], "answer_arxiv_id": ["2210.10109"], "source_meta": {"published_time": "20240823"}, "qid": "AutoScholarQuery_train_19871"} +{"question": "Which works discuss the application of GFlowNets in biological sequence design?", "answer": ["Biological Sequence Design with GFlowNets"], "answer_arxiv_id": ["2203.04115"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_19872"} +{"question": "Which research introduced the NeRFs for modeling 3D shapes in text-to-3D applications?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_19873"} +{"question": "Which works belong to the category of offline RL that focus on learning from mixed-quality datasets with reward labels?", "answer": ["Addressing Function Approximation Error in Actor-Critic Methods", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Conservative Q-Learning for Offline Reinforcement Learning", "Behavior Regularized Offline Reinforcement Learning", "Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems", "D4RL: Datasets for Deep Data-Driven Reinforcement Learning"], "answer_arxiv_id": ["1802.09477", "1906.00949", "2006.04779", "1911.11361", "2005.01643", "2004.07219"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19874"} +{"question": "Could you provide research papers that contributed to refining the designs and efficiency of deep learning-based matchers?", "answer": ["Learning to Match Features with Seeded Graph Matching Network", "ClusterGNN: Cluster-based Coarse-to-Fine Graph Neural Network for\n Efficient Feature Matching", "LightGlue: Local Feature Matching at Light Speed"], "answer_arxiv_id": ["2108.08771", "2204.11700", "2306.13643"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19875"} +{"question": "Which papers have contributed to the applications of autonomous driving in Optical character Recognition (OCR)?", "answer": ["OCR-RTPS: An OCR-based real-time positioning system for the valet parking"], "answer_arxiv_id": ["2212.04116"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19876"} +{"question": "Are there any works that mitigated complexity in Vision Transformers by utilizing local attention within fixed window size?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Twins: Revisiting the Design of Spatial Attention in Vision Transformers"], "answer_arxiv_id": ["2103.14030", "2104.13840"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_19877"} +{"question": "What are some of the parametric methods used in prior works on 3D HMR?", "answer": ["3D Hand Shape and Pose from Images in the Wild", "End-to-end Hand Mesh Recovery from a Monocular RGB Image", "Model-based 3D Hand Reconstruction via Self-Supervised Learning"], "answer_arxiv_id": ["1902.03451", "1902.09305", "2103.11703"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_19878"} +{"question": "Which papers used reward functions to improve faithful image generation based on compositional prompts?", "answer": ["T2I-CompBench: A Comprehensive Benchmark for Open-world Compositional\n Text-to-image Generation", "ImageReward: Learning and Evaluating Human Preferences for Text-to-Image\n Generation", "Aligning Text-to-Image Models using Human Feedback", "Training Diffusion Models with Reinforcement Learning"], "answer_arxiv_id": ["2307.06350", "2304.05977", "2302.12192", "2305.13301"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_19879"} +{"question": "Can you name the works that have used out-of-domain visual datasets for representation learning in control?", "answer": ["The (Un)Surprising Effectiveness of Pre-Trained Vision Models for Control", "CLIPort: What and Where Pathways for Robotic Manipulation", "Simple but Effective: CLIP Embeddings for Embodied AI", "Learning to See before Learning to Act: Visual Pre-training for Manipulation", "Where are we in the search for an Artificial Visual Cortex for Embodied Intelligence?"], "answer_arxiv_id": ["2203.03580", "2109.12098", "2111.09888", "2107.00646", "2303.18240"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_19880"} +{"question": "Which papers discussed techniques for local image editing such as swapping certain parts between images or modifying style at specific regions?", "answer": ["Editing in Style: Uncovering the Local Semantics of GANs", "Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval", "Spatially Controllable Image Synthesis with Internal Representation Collaging", "Exploiting Spatial Dimensions of Latent in GAN for Real-time Image Editing", "Rewriting a Deep Generative Model", "Attribute-specific Control Units in StyleGAN for Fine-grained Image Manipulation", "StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation", "Low-Rank Subspaces in GANs", "Decorating Your Own Bedroom: Locally Controlling Image Generation with Generative Adversarial Networks", "EditGAN: High-Precision Semantic Image Editing", "StyleFusion: A Generative Model for Disentangling Spatial Segments"], "answer_arxiv_id": ["2004.14367", "2107.06256", "1811.10153", "2104.14754", "2007.15646", "2111.13010", "2011.12799", "2106.04488", "2105.08222", "2111.03186", "2107.07437"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_19881"} +{"question": "Can you list some works where automated characterizations of animal behavior are used?", "answer": ["A framework for studying behavioral evolution by reconstructing ancestral repertoires", "Task Programming: Learning Data Efficient Behavior Representations"], "answer_arxiv_id": ["2007.09689", "2011.13917"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_19882"} +{"question": "Which work is about using GPT-4 to evaluate text through chain-of-thoughts prompting?", "answer": ["G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment"], "answer_arxiv_id": ["2303.16634"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_19883"} +{"question": "Could you provide the work that encourages transformation consistency between predictions with augmentations applied to the image inputs and the segmentation outputs?", "answer": ["Transformation-consistent Self-ensembling Model for Semi-supervised Medical Image Segmentation"], "answer_arxiv_id": ["1903.00348"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_19884"} +{"question": "What studies are related to indiscriminate poisoning attacks on linear regression and support vector machine?", "answer": ["Poisoning Attacks against Support Vector Machines"], "answer_arxiv_id": ["1206.6389"], "source_meta": {"published_time": "20220222"}, "qid": "AutoScholarQuery_train_19885"} +{"question": "What studies adopt image-text contrastive learning methods in medical report analysis?", "answer": ["Contrastive Learning of Medical Visual Representations from Paired\n Images and Text", "Unsupervised Multimodal Representation Learning across Medical Images\n and Reports", "Multi-Granularity Cross-modal Alignment for Generalized Medical Visual\n Representation Learning"], "answer_arxiv_id": ["2010.00747", "1811.08615", "2210.06044"], "source_meta": {"published_time": "20240203"}, "qid": "AutoScholarQuery_train_19886"} +{"question": "What works proposed to enhance models’ generalizationability using data augmentation?", "answer": ["Adversarial Domain Adaptation with Domain Mixup", "Improving Out-of-Distribution Robustness via Selective Augmentation", "Discover and Cure: Concept-aware Mitigation of Spurious Correlation"], "answer_arxiv_id": ["1912.01805", "2201.00299", "2305.00650"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19887"} +{"question": "Any works about primal approaches that exclude the dual variables from the optimization framework for solving safe RL problems?", "answer": ["CRPO: A New Approach for Safe Reinforcement Learning with Convergence Guarantee", "Sauté RL: Almost Surely Safe Reinforcement Learning Using State Augmentation"], "answer_arxiv_id": ["2011.05869", "2202.06558"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_19888"} +{"question": "Which paper described that the DPMs use a bidirectional iterative chain for the generation process?", "answer": ["Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2006.11239"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19889"} +{"question": "Which papers have addressed handling occlusion in object detection?", "answer": ["Robust Object Detection under Occlusion with Context-Aware CompositionalNets", "Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers", "A Tri-Layer Plugin to Improve Occluded Detection"], "answer_arxiv_id": ["2005.11643", "2103.12340", "2210.10046"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_19890"} +{"question": "What studies introduce Convolutional Networks (ConvNets) for visual architecture?", "answer": ["Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning", "Deep Residual Learning for Image Recognition"], "answer_arxiv_id": ["1602.07261", "1512.03385"], "source_meta": {"published_time": "20230201"}, "qid": "AutoScholarQuery_train_19891"} +{"question": "Which works focused on probing structured linguistic knowledge in LMs?", "answer": ["What you can cram into a single vector: Probing sentence embeddings for\n linguistic properties", "BERT Rediscovers the Classical NLP Pipeline"], "answer_arxiv_id": ["1805.01070", "1905.05950"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_19892"} +{"question": "Which works studied Text-conditional auto-regressive models in generating high-quality and diverse images from textual conditions?", "answer": ["Zero-Shot Text-to-Image Generation", "Make-A-Scene: Scene-Based Text-to-Image Generation with Human Priors", "CogView: Mastering Text-to-Image Generation via Transformers", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2102.12092", "2203.13131", "2105.13290", "2206.10789"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19893"} +{"question": "Are there any researches that attempt to cast offline RL as a sequence-to-sequence model?", "answer": ["Decision Transformer: Reinforcement Learning via Sequence Modeling", "Offline Reinforcement Learning as One Big Sequence Modeling Problem"], "answer_arxiv_id": ["2106.01345", "2106.02039"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_19894"} +{"question": "What research work establishes a connection between NPG and a specific form of policy mirror descent (PMD) for the tabular setting?", "answer": ["On the Convergence Rates of Policy Gradient Methods"], "answer_arxiv_id": ["2201.07443"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_19895"} +{"question": "What studies are about tuning DP algorithms in private machine learning?", "answer": ["Differentially Private Learning with Adaptive Clipping"], "answer_arxiv_id": ["1905.03871"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_19896"} +{"question": "Which studies suggested to share all parameters between different factorizations of the sequence in language models?", "answer": ["XLNet: Generalized Autoregressive Pretraining for Language Understanding", "Compact Bidirectional Transformer for Image Captioning"], "answer_arxiv_id": ["1906.08237", "2201.01984"], "source_meta": {"published_time": "20230313"}, "qid": "AutoScholarQuery_train_19897"} +{"question": "Which works have been proposed in the field of automatic graphical design for various scenarios, including indoor scenes?", "answer": ["Fast and Flexible Indoor Scene Synthesis via Deep Convolutional Generative Models"], "answer_arxiv_id": ["1811.12463"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19898"} +{"question": "What are the earlier works on side-channel attacks, specifically on reconstructing the target graph from released graph embeddings?", "answer": ["Quantifying Privacy Leakage in Graph Embedding", "DeepWalking Backwards: From Embeddings Back to Graphs"], "answer_arxiv_id": ["2010.00906", "2102.08532"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_19899"} +{"question": "What studies talk about the convergence guarantee of policy gradient in Cooperative MGs with softmax parameterization?", "answer": ["Convergence and Price of Anarchy Guarantees of the Softmax Policy Gradient in Markov Potential Games"], "answer_arxiv_id": ["2206.07642"], "source_meta": {"published_time": "20230602"}, "qid": "AutoScholarQuery_train_19900"} +{"question": "Which works are about the effective method of Adversarial Training that defends against many attacks?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks"], "answer_arxiv_id": ["1412.6572", "1706.06083"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_19901"} +{"question": "Could you point to a study that proposes a different algorithm for fine-grained, compositional text-to-image generation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20221209"}, "qid": "AutoScholarQuery_train_19902"} +{"question": "Could you provide me with some studies about Parameter-efficient Fine-Tuning (PEFT) methods?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "LoRA: Low-Rank Adaptation of Large Language Models", "Visual Prompt Tuning"], "answer_arxiv_id": ["1902.00751", "2106.09685", "2203.12119"], "source_meta": {"published_time": "20240417"}, "qid": "AutoScholarQuery_train_19903"} +{"question": "Which papers describe methods like SLAC and SOLAR that learn latent representations of the dynamics jointly with optimization of the target policies?", "answer": ["Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Model", "SOLAR: Deep Structured Representations for Model-Based Reinforcement Learning"], "answer_arxiv_id": ["1907.00953", "1808.09105"], "source_meta": {"published_time": "20230128"}, "qid": "AutoScholarQuery_train_19904"} +{"question": "What are the studies using training epochs and dataset subsets commonly as fidelity variables in deep learning?", "answer": ["Freeze-Thaw Bayesian Optimization", "Bayesian Optimization for Iterative Learning"], "answer_arxiv_id": ["1406.3896", "1909.09593"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_19905"} +{"question": "Have there been studies analysing the effect of long-tail tokens on incremental contextual learning?", "answer": ["Understanding In-Context Learning via Supportive Pretraining Data", "Meta-learning via Language Model In-context Tuning"], "answer_arxiv_id": ["2306.15091", "2110.07814"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_19906"} +{"question": "Which works discuss model-free methods that are motivated by Q-learning in RL?", "answer": ["Provably Efficient Q-Learning with Low Switching Cost"], "answer_arxiv_id": ["1905.12849"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19907"} +{"question": "Could you provide me some research regarding pruning periodically during training?", "answer": ["The State of Sparsity in Deep Neural Networks"], "answer_arxiv_id": ["1902.09574"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_19908"} +{"question": "Which work reconstructs rejection-ABC for providing exact inference instead of approximate inference in GBI?", "answer": ["Approximate Bayesian computation (ABC) gives exact results under the assumption of model error"], "answer_arxiv_id": ["0811.3355"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_19909"} +{"question": "Could you provide me some studies that propose methods about removing features?", "answer": ["Visualizing and Understanding Convolutional Networks", "Explaining image classifiers by removing input features using generative models", "Interpretable Explanations of Black Boxes by Meaningful Perturbation"], "answer_arxiv_id": ["1311.2901", "1910.04256", "1704.03296"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_19910"} +{"question": "What architectures of Vision Transformers (ViTs) were designed to improve accuracy-efficiency trade-offs using a pyramid-like structure?", "answer": ["Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions", "CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification", "Rethinking Spatial Dimensions of Vision Transformers", "Multiscale Vision Transformers", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows"], "answer_arxiv_id": ["2102.12122", "2103.14899", "2103.16302", "2104.11227", "2103.14030"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_19911"} +{"question": "Can you provide me with a work that extends model-based offline RL algorithms by incorporating a latent-state dynamics model for high-dimensional visual observation spaces?", "answer": ["Offline Reinforcement Learning from Images with Latent Space Models"], "answer_arxiv_id": ["2012.11547"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_19912"} +{"question": "What works have applied deep learning to accelerate numerical simulations?", "answer": ["Machine Learning for Fluid Mechanics", "NVIDIA SimNet™: an AI-accelerated multi-physics simulation framework"], "answer_arxiv_id": ["1905.11075", "2012.07938"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_19913"} +{"question": "Which papers focus on expanding the context length of LLMs?", "answer": ["Train Short, Test Long: Attention with Linear Biases Enables Input\n Length Extrapolation", "YaRN: Efficient Context Window Extension of Large Language Models", "Unlimiformer: Long-Range Transformers with Unlimited Length Input"], "answer_arxiv_id": ["2108.12409", "2309.00071", "2305.01625"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_19914"} +{"question": "What research works consider the stable matching setting in MPMAB?", "answer": ["Competing Bandits in Matching Markets", "Bandit Learning in Decentralized Matching Markets", "Learning Equilibria in Matching Markets from Bandit Feedback"], "answer_arxiv_id": ["1906.05363", "2012.07348", "2108.08843"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19915"} +{"question": "Are there any examples of studies that tried to select rationales belonging to different aspects at once?", "answer": ["Multi-Dimensional Explanation of Target Variables from Documents", "Rationalization through Concepts"], "answer_arxiv_id": ["1909.11386", "2105.04837"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_19916"} +{"question": "What studies introduced LLMs specifically for target location finding and dynamic task decomposition in EIF?", "answer": ["Prompter: Utilizing Large Language Model Prompting for a Data Efficient\n Embodied Instruction Following", "LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large\n Language Models"], "answer_arxiv_id": ["2211.03267", "2212.04088"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_19917"} +{"question": "In what studies was the log-barrier regularizer used?", "answer": ["Learning in Games: Robustness of Fast Convergence", "More Adaptive Algorithms for Adversarial Bandits", "Efficient Online Portfolio with Logarithmic Regret", "On First-Order Bounds, Variance and Gap-Dependent Bounds for Adversarial Bandits"], "answer_arxiv_id": ["1606.06244", "1801.03265", "1805.07430", "1903.07890"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19918"} +{"question": "Are there any studies on cross-lingual transfer and alignment across programming languages?", "answer": ["On the Transferability of Pre-trained Language Models for Low-Resource\n Programming Languages", "CIRCLE: Continual Repair across Programming Languages", "MetaTPTrans: A Meta Learning Approach for Multilingual Code\n Representation Learning", "On Negative Interference in Multilingual Models: Findings and A\n Meta-Learning Treatment"], "answer_arxiv_id": ["2204.09653", "2205.10956", "2206.06460", "2010.03017"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_19919"} +{"question": "What researches have been conducted regarding pruning for model compression?", "answer": ["PARP: Prune, Adjust and Re-Prune for Self-Supervised Speech Recognition", "Structured Pruning of Self-Supervised Pre-trained Models for Speech\n Recognition and Understanding"], "answer_arxiv_id": ["2106.05933", "2302.14132"], "source_meta": {"published_time": "20240220"}, "qid": "AutoScholarQuery_train_19920"} +{"question": "Any studies about mitigating turbulence in moving object scenarios where the dynamic regions can be isolated?", "answer": ["Atmospheric turbulence mitigation for sequences with moving objects\n using recursive image fusion", "Image Reconstruction of Static and Dynamic Scenes through Anisoplanatic\n Turbulence"], "answer_arxiv_id": ["1808.03550", "2009.00071"], "source_meta": {"published_time": "20240108"}, "qid": "AutoScholarQuery_train_19921"} +{"question": "What is the work that used geometric knowledge distillation to enable efficient inference on smaller graphs?", "answer": ["Geometric Knowledge Distillation: Topology Compression for Graph Neural Networks"], "answer_arxiv_id": ["2210.13014"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_19922"} +{"question": "What are some of the research that generate human motion through unconditioned signals?", "answer": ["Perpetual Motion: Generating Unbounded Human Motion", "Human Motion Diffusion as a Generative Prior"], "answer_arxiv_id": ["2007.13886", "2303.01418"], "source_meta": {"published_time": "20240407"}, "qid": "AutoScholarQuery_train_19923"} +{"question": "Which works are about automated mechanism design for finding optimal auctions with multiple items and bidders?", "answer": ["Complexity of Mechanism Design", "Payment Rules through Discriminant-Based Classifiers"], "answer_arxiv_id": ["1408.1486v1", "1208.1184"], "source_meta": {"published_time": "20230520"}, "qid": "AutoScholarQuery_train_19924"} +{"question": "Which studies explored the use of an additional QA model over extracted reasoning paths?", "answer": ["Multi-hop Question Answering via Reasoning Chains"], "answer_arxiv_id": ["1910.02610"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_19925"} +{"question": "What work trained a full language model for each class in conditional sampling?", "answer": ["GeDi: Generative Discriminator Guided Sequence Generation"], "answer_arxiv_id": ["2009.06367"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_19926"} +{"question": "Which papers focused on discrete-time dynamic graphs (DTDGs) in early approaches?", "answer": ["EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs", "dyngraph2vec: Capturing Network Dynamics using Dynamic Graph Representation Learning", "ROLAND: Graph Learning Framework for Dynamic Graphs"], "answer_arxiv_id": ["1902.10191", "1809.02657", "2208.07239"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_19927"} +{"question": "Could you point to studies discussing the potential bias in language models due to the demographic background of human raters?", "answer": ["On Releasing Annotator-Level Labels and Information in Datasets", "Whose Ground Truth? Accounting for Individual and Collective Identities Underlying Dataset Annotation"], "answer_arxiv_id": ["2110.05699", "2112.04554"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_19928"} +{"question": "Could you provide me some studies related to Quantization-aware training (QAT)?", "answer": ["Bi-Real Net: Binarizing Deep Network Towards Real-Network Performance", "PACT: Parameterized Clipping Activation for Quantized Neural Networks", "Binarizing Sparse Convolutional Networks for Efficient Point Cloud\n Analysis"], "answer_arxiv_id": ["1811.01335", "1805.06085", "2303.15493"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19929"} +{"question": "Could you provide research work on learning with label differential privacy for classification tasks?", "answer": ["Private Learning and Sanitization: Pure vs. Approximate Differential Privacy"], "answer_arxiv_id": ["1407.2674v1"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_19930"} +{"question": "Which methods derive density values from the outputs of a signed distance function for rendering similar to NeRF?", "answer": ["UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction", "NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "Volume Rendering of Neural Implicit Surfaces"], "answer_arxiv_id": ["2104.10078", "2106.10689", "2106.12052"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_19931"} +{"question": "What works explored equivariant representation learning using autoencoders such as transforming autoencoders and Homeomorphic VAEs?", "answer": ["Explorations in Homeomorphic Variational Auto-Encoding", "Unsupervised Learning of Group Invariant and Equivariant Representations"], "answer_arxiv_id": ["1807.04689", "2202.07559"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_19932"} +{"question": "What is a paper that explored the interaction between hierarchical learning and imitation in goal-conditioned settings?", "answer": ["Hierarchical Imitation and Reinforcement Learning"], "answer_arxiv_id": ["1803.00590"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_19933"} +{"question": "Can you tell me about the studies that proposed a two-stage pipeline in the context of image segmentation in open-vocabulary segmentation task?", "answer": ["Decoupling Zero-Shot Semantic Segmentation", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion\n Models"], "answer_arxiv_id": ["2112.07910", "2210.04150", "2303.04803"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19934"} +{"question": "Which papers have shown the challenges associated with training pixel diffusion models for high resolution images on ImageNet without guidance or cascades?", "answer": ["Diffusion Models Beat GANs on Image Synthesis", "Classifier-Free Diffusion Guidance", "Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2105.05233", "2207.12598", "2106.15282"], "source_meta": {"published_time": "20221222"}, "qid": "AutoScholarQuery_train_19935"} +{"question": "Who proposed the SOTA online RL algorithm OAC that inspired the design of CFPI operators?", "answer": ["Better Exploration with Optimistic Actor-Critic"], "answer_arxiv_id": ["1910.12807"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_19936"} +{"question": "What research papers discuss pre-training models in computer vision on large-scale datasets?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Big Transfer (BiT): General Visual Representation Learning", "How to train your ViT? Data, Augmentation, and Regularization in Vision\n Transformers"], "answer_arxiv_id": ["2010.11929", "2103.14030", "1912.11370", "2106.10270"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_19937"} +{"question": "What papers are about multimodal summarization where the input contains both text and video?", "answer": ["Kosmos-G: Generating Images in Context with Multimodal Large Language\n Models", "A Video Is Worth 4096 Tokens: Verbalize Videos To Understand Them In\n Zero Shot", "Multimodal Few-Shot Learning with Frozen Language Models", "Hierarchical3D Adapters for Long Video-to-text Summarization"], "answer_arxiv_id": ["2310.02992", "2305.09758", "2106.13884", "2210.04829"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_19938"} +{"question": "What works propose neural architecture encoding?", "answer": ["Neural Architecture Optimization", "Progressive Neural Architecture Search", "AlphaX: eXploring Neural Architectures with Deep Neural Networks and\n Monte Carlo Tree Search", "Semi-Supervised Neural Architecture Search", "BANANAS: Bayesian Optimization with Neural Architectures for Neural\n Architecture Search"], "answer_arxiv_id": ["1808.07233", "1712.00559", "1903.11059", "2002.10389", "1910.11858"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_19939"} +{"question": "Could you provide me some examples of research that predicted visual emotions by extracting high-level semantics?", "answer": ["Learning Multi-level Deep Representations for Image Emotion\n Classification", "Stimuli-Aware Visual Emotion Analysis", "SOLVER: Scene-Object Interrelated Visual Emotion Reasoning Network"], "answer_arxiv_id": ["1611.07145", "2109.01812", "2110.12334"], "source_meta": {"published_time": "20240109"}, "qid": "AutoScholarQuery_train_19940"} +{"question": "Could you share examples of research that proposed various Mixture of Experts (MoE) architectures?", "answer": ["Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity", "GLaM: Efficient Scaling of Language Models with Mixture-of-Experts"], "answer_arxiv_id": ["2101.03961", "2112.06905"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_19941"} +{"question": "What research constructs a certifiable classifier by adding smooth noise to the original classifier?", "answer": ["Certified Adversarial Robustness via Randomized Smoothing"], "answer_arxiv_id": ["1902.02918"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_19942"} +{"question": "Are there any works that joinly use points and lines during localization?", "answer": ["GlueStick: Robust Image Matching by Sticking Points and Lines Together", "Pose Refinement with Joint Optimization of Visual Points and Lines", "3D Line Mapping Revisited"], "answer_arxiv_id": ["2304.02008", "2110.03940", "2303.17504"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_19943"} +{"question": "What are some previous works that use gradient methods to seek empirical robustness in adversarial behaviour rather than robust MDP solutions?", "answer": ["Robust Adversarial Reinforcement Learning", "Action Robust Reinforcement Learning and Applications in Continuous Control", "Soft-Robust Actor-Critic Policy-Gradient"], "answer_arxiv_id": ["1703.02702", "1901.09184", "1803.04848"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19944"} +{"question": "What studies have introduced frequency-related losses for better controlling the PD trade-off and easing the training of GANs?", "answer": ["Frequency Separation for Real-World Super-Resolution", "Fourier Space Losses for Efficient Perceptual Image Super-Resolution", "SWAGAN: A Style-based Wavelet-driven Generative Model", "On the Frequency Bias of Generative Models", "SSD-GAN: Measuring the Realness in the Spatial and Spectral Domains"], "answer_arxiv_id": ["1911.07850", "2106.00783", "2102.06108", "2111.02447", "2012.05535"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_19945"} +{"question": "Could you list some works that used external feature extractors in SDMs?", "answer": ["Domino: Discovering Systematic Errors with Cross-Modal Embeddings", "Understanding Failures of Deep Networks via Robust Feature Extraction"], "answer_arxiv_id": ["2203.14960", "2012.01750"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_19946"} +{"question": "Could you provide some studies that proposed evaluation metrics that depend only on the real support, which are robust to outlying features?", "answer": ["Reliable Fidelity and Diversity Metrics for Generative Models"], "answer_arxiv_id": ["2002.09797"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_19947"} +{"question": "What are the works studying sublinear constraints like memory or sample complexity under a machine learning enhancing classical algorithmic problems framework?", "answer": ["Learning-Based Low-Rank Approximations", "Composable Sketches for Functions of Frequencies: Beyond the Worst Case", "Learning-based Support Estimation in Sublinear Time", "Triangle and Four Cycle Counting with Predictions in Graph Streams", "Learning the Positions in CountSketch"], "answer_arxiv_id": ["1910.13984", "2004.04772", "2106.08396", "2203.09572", "2007.09890"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_19948"} +{"question": "What were the recent studies that used contrastive losses in self-supervised representations?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Contrastive Multiview Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Momentum Contrast for Unsupervised Visual Representation Learning"], "answer_arxiv_id": ["1807.03748", "1906.05849", "2002.05709", "1911.05722"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_19949"} +{"question": "Which research papers proposed mechanisms for computing the mean of a set of vectors under communication constraint in the context of privacy-aware compression for distributed mean estimation?", "answer": ["Protection Against Reconstruction and Its Applications in Private Federated Learning", "Optimal Algorithms for Mean Estimation under Local Differential Privacy", "The Discrete Gaussian for Differential Privacy", "The Distributed Discrete Gaussian Mechanism for Federated Learning with Secure Aggregation", "The Skellam Mechanism for Differentially Private Federated Learning", "Privacy-Aware Compression for Federated Data Analysis", "Breaking the Communication-Privacy-Accuracy Trilemma", "Optimal Compression of Locally Differentially Private Mechanisms", "Joint Privacy Enhancement and Quantization in Federated Learning"], "answer_arxiv_id": ["1812.00984", "2205.02466", "2004.00010", "2102.06387", "2110.04995", "2203.08134", "2007.11707", "2111.00092", "2208.10888"], "source_meta": {"published_time": "20221108"}, "qid": "AutoScholarQuery_train_19950"} +{"question": "What works revealed that FedAvg may not converge if data from different clients is non-i.i.d?", "answer": ["Federated Learning with Non-IID Data", "On the Convergence of FedAvg on Non-IID Data", "Achieving ​ Linear ​ Speedup ​ with ​​ Partial ​​ Worker ​​ Participation ​​ in ​​ Non-IID ​​ Federated ​​ Learning"], "answer_arxiv_id": ["1806.00582", "1907.02189", "2101.11203"], "source_meta": {"published_time": "20220128"}, "qid": "AutoScholarQuery_train_19951"} +{"question": "What research papers discuss about the efficient exploration methods?", "answer": ["Centralized Cooperative Exploration Policy for Continuous Control Tasks", "Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor", "Reinforcement Learning with Deep Energy-Based Policies", "Exploration by Random Network Distillation", "Curiosity-driven Exploration by Self-supervised Prediction"], "answer_arxiv_id": ["2301.02375", "1801.01290", "1702.08165", "1810.12894", "1705.05363"], "source_meta": {"published_time": "20231017"}, "qid": "AutoScholarQuery_train_19952"} +{"question": "What work is about synthetically generating hallucination datasets by inserting knowledge conflicts into the output?", "answer": ["Entity-Based Knowledge Conflicts in Question Answering"], "answer_arxiv_id": ["2109.05052"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_19953"} +{"question": "What studies have reported effects on performance related to the construction of in-context examples or prompts?", "answer": ["In-context Examples Selection for Machine Translation"], "answer_arxiv_id": ["2212.02437"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_19954"} +{"question": "Which papers conducted research on Action Quality Assessment (AQA) in sports?", "answer": ["Am I a Baller? Basketball Performance Assessment from First-Person\n Videos", "What and How Well You Performed? A Multitask Learning Approach to Action\n Quality Assessment", "Learning to score the figure skating sports videos", "Action Quality Assessment Across Multiple Actions", "Learning To Score Olympic Events", "TSA-Net: Tube Self-Attention Network for Action Quality Assessment"], "answer_arxiv_id": ["1611.05365", "1904.04346", "1802.02774", "1812.06367", "1611.05125", "2201.03746"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_19955"} +{"question": "Which work pointed out that the diverse samples from large language models often include correct programs?", "answer": ["Evaluating Large Language Models Trained on Code"], "answer_arxiv_id": ["2107.03374"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_19956"} +{"question": "Which paper presents the Multi-Level Firing (MLF) unit?", "answer": ["Multi-Level Firing with Spiking DS-ResNet: Enabling Better and Deeper Directly-Trained Spiking Neural Networks"], "answer_arxiv_id": ["2210.06386"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_19957"} +{"question": "Which works studied learning multimodal joint representations for 3D shapes?", "answer": ["Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining", "CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D Recognition"], "answer_arxiv_id": ["2302.02318", "2303.11313"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19958"} +{"question": "Which papers discuss that Multitask Learning can learn commonalities and differences across different tasks?", "answer": ["Multi-Task Learning with Deep Neural Networks: A Survey", "Multi-Task Learning for Dense Prediction Tasks: A Survey", "An Overview of Multi-Task Learning in Deep Neural Networks"], "answer_arxiv_id": ["2009.09796", "2004.13379", "1706.05098"], "source_meta": {"published_time": "20231102"}, "qid": "AutoScholarQuery_train_19959"} +{"question": "Which study challenges the notion that non-robust features are useful, stating that they may actually produce poorer performance when transferred?", "answer": ["Adversarial Examples Are Not Real Features"], "answer_arxiv_id": ["2310.18936"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_19960"} +{"question": "Could you provide me studies concerning prompt learning which aimed to enhance generalization performance in Vision Language Models?", "answer": ["Conditional Prompt Learning for Vision-Language Models", "Task Residual for Tuning Vision-Language Models", "MaPLe: Multi-modal Prompt Learning", "Self-regulating Prompts: Foundational Model Adaptation without Forgetting"], "answer_arxiv_id": ["2203.05557", "2211.10277", "2210.03117", "2307.06948v2"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_19961"} +{"question": "What works have used TDA-based measures to quantify the difference between training data and new data generated by generative models?", "answer": ["Manifold Topology Divergence: a Framework for Comparing Data Manifolds"], "answer_arxiv_id": ["2106.04024"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_19962"} +{"question": "What is the method mentioned which offers non-convex implementations through a first-order optimizer?", "answer": ["CUP: A Conservative Update Policy Algorithm for Safe Reinforcement Learning"], "answer_arxiv_id": ["2202.07565"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_19963"} +{"question": "Could you provide me some works about point-based methods in the analysis of 3D point cloud shape?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "Dynamic Graph CNN for Learning on Point Clouds"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1801.07829"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_19964"} +{"question": "Can you provide the research that employed visual prompt at the pixel level?", "answer": ["ProSFDA: Prompt Learning based Source-free Domain Adaptation for Medical\n Image Segmentation"], "answer_arxiv_id": ["2211.11514"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19965"} +{"question": "Any studies about introducing self-training approach for visual question answering?", "answer": ["Q: How to Specialize Large Vision-Language Models to Data-Scarce VQA\n Tasks? A: Self-Train on Unlabeled Images!"], "answer_arxiv_id": ["2306.03932"], "source_meta": {"published_time": "20240406"}, "qid": "AutoScholarQuery_train_19966"} +{"question": "Are there research addressing non-zero average cosine similarity between word vectors in contextual NLP models?", "answer": ["How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddings"], "answer_arxiv_id": ["1909.00512"], "source_meta": {"published_time": "20221215"}, "qid": "AutoScholarQuery_train_19967"} +{"question": "Which studies focus on expressive human pose and shape (EHPS), adding hands and face to the estimation?", "answer": ["Expressive Body Capture: 3D Hands, Face, and Body from a Single Image", "Monocular Total Capture: Posing Face, Body, and Hands in the Wild", "Monocular Expressive Body Regression through Body-Driven Attention", "FrankMocap: A Monocular 3D Whole-Body Pose Estimation System via Regression and Integration", "Monocular Real-time Full Body Capture with Inter-part Correlations", "Collaborative Regression of Expressive Bodies using Moderation", "PyMAF-X: Towards Well-aligned Full-body Model Regression from Monocular Images"], "answer_arxiv_id": ["1904.05866", "1812.01598", "2008.09062", "2108.06428", "2012.06087", "2105.05301", "2207.06400"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_19968"} +{"question": "Are there works that focused on localizing first-person-view camera in bird’s-eye-view aerial images?", "answer": ["Where am I looking at? Joint Location and Orientation Estimation by\n Cross-View Matching"], "answer_arxiv_id": ["2005.03860"], "source_meta": {"published_time": "20221219"}, "qid": "AutoScholarQuery_train_19969"} +{"question": "What work constructed an additional IoU predictor and an IoU-guided NMS scheme?", "answer": ["Acquisition of Localization Confidence for Accurate Object Detection"], "answer_arxiv_id": ["1807.11590"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_19970"} +{"question": "What research involves saliency methods specifically for time series?", "answer": ["Benchmarking Deep Learning Interpretability in Time Series Predictions", "What went wrong and when? Instance-wise feature importance for time-series black-box models", "Explaining Time Series Predictions with Dynamic Masks"], "answer_arxiv_id": ["2010.13924", "2003.02821", "2106.05303"], "source_meta": {"published_time": "20210729"}, "qid": "AutoScholarQuery_train_19971"} +{"question": "Could you provide me some research articles that focus on developing communication efficient federated learning algorithms?", "answer": ["A Communication-Efficient Collaborative Learning Framework for Distributed Features", "Communication-Efficient Distributed Learning via Lazily Aggregated Quantized Gradients", "FedPAQ: A Communication-Efficient Federated Learning Method with Periodic Averaging and Quantization", "Federated Learning with Compression: Unified Analysis and Sharp Guarantees", "Adaptive Quantization of Model Updates for Communication-Efficient Federated Learning", "DAdaQuant: Doubly-adaptive quantization for communication-efficient Federated Learning"], "answer_arxiv_id": ["1912.11187", "1909.07588", "1909.13014v4", "2007.01154v2", "2102.04487", "2111.00465"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_19972"} +{"question": "What research is the most closely relevant to the present work about fully polynomial-time tester-learners for halfspaces?", "answer": ["An Efficient Tester-Learner for Halfspaces"], "answer_arxiv_id": ["2302.14853"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_19973"} +{"question": "Which studies discuss the issue of foundation models sometimes producing outputs that do not align with human values?", "answer": ["RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models"], "answer_arxiv_id": ["2009.11462v2"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_19974"} +{"question": "Which work incorporated phylogenetic information for the classification of metagenomics data using CNNs?", "answer": ["Phylogenetic Convolutional Neural Networks in Metagenomics"], "answer_arxiv_id": ["1709.02268"], "source_meta": {"published_time": "20230217"}, "qid": "AutoScholarQuery_train_19975"} +{"question": "In which paper has the Barker acceptance test been used in the context of a differentially private Metropolis-Hastings?", "answer": ["Differentially Private Markov Chain Monte Carlo"], "answer_arxiv_id": ["1901.10275"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_19976"} +{"question": "Which research proposes a formulation that treats model outputs as parameters of a Dirichlet distribution in the analysis of decision certainty and computational parsimony?", "answer": ["Evidential Deep Learning to Quantify Classification Uncertainty"], "answer_arxiv_id": ["1806.01768"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_19977"} +{"question": "Are there any studies that focus on precise control in denoising diffusion probabilistic models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "DragonDiffusion: Enabling Drag-style Manipulation on Diffusion Models"], "answer_arxiv_id": ["2302.05543", "2302.08453", "2307.02421"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_19978"} +{"question": "Which works study the capacity of a graph convolution for one-layer networks and its out-of-distribution generalization potential?", "answer": ["Graph Convolution for Semi-Supervised Classification: Improved Linear Separability and Out-of-Distribution Generalization"], "answer_arxiv_id": ["2102.06966"], "source_meta": {"published_time": "20220420"}, "qid": "AutoScholarQuery_train_19979"} +{"question": "In point cloud processing, what studies have attempted to treat wireframe reconstruction as an edge point classification task?", "answer": ["EC-Net: an Edge-aware Point set Consolidation Network", "PCEDNet : A Lightweight Neural Network for Fast and Interactive Edge Detection in 3D Point Clouds"], "answer_arxiv_id": ["1807.06010", "2011.01630v4"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_19980"} +{"question": "Any works introducing the benefits of sparse MoEs such as an enhancement in its adversarial robustness?", "answer": ["On the Adversarial Robustness of Mixture of Experts"], "answer_arxiv_id": ["2210.10253"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_19981"} +{"question": "Which work leverages spatial boxes and RoI-aligned features for input and training in region-specific LMMs?", "answer": ["GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest"], "answer_arxiv_id": ["2307.03601"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_19982"} +{"question": "What previous works used geometry regularization, semantic consistency, depth supervision etc. to optimize a NeRF using sparse-view data for few-shot reconstruction?", "answer": ["RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs", "Depth-supervised NeRF: Fewer Views and Faster Training for Free", "PS-NeRF: Neural Inverse Rendering for Multi-view Photometric Stereo", "SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse\n Views", "Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis", "InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Rendering"], "answer_arxiv_id": ["2112.00724", "2107.02791", "2207.11406", "2206.05737", "2104.00677", "2112.15399"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_19983"} +{"question": "Could you provide me some studies that modify the learning objective to prevent representation drift when encountered with new classes?", "answer": ["Co2L: Contrastive Continual Learning", "Task Agnostic Representation Consolidation: a Self-supervised based Continual Learning Approach", "New Insights on Reducing Abrupt Representation Change in Online Continual Learning"], "answer_arxiv_id": ["2106.14413", "2207.06267", "2104.05025"], "source_meta": {"published_time": "20230214"}, "qid": "AutoScholarQuery_train_19984"} +{"question": "What studies show the potential for retrieval augmentation in image generation?", "answer": ["KNN-Diffusion: Image Generation via Large-Scale Retrieval"], "answer_arxiv_id": ["2204.02849"], "source_meta": {"published_time": "20231122"}, "qid": "AutoScholarQuery_train_19985"} +{"question": "Are there any works about exploiting different forms of weak supervision and human annotations in weakly-supervised 3D segmentation?", "answer": ["Multi-Path Region Mining For Weakly Supervised 3D Semantic Segmentation on Point Clouds", "Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using Bounding Boxes", "One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation", "Scribble-Supervised LiDAR Semantic Segmentation"], "answer_arxiv_id": ["2003.13035", "2206.01203", "2104.02246", "2203.08537"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_19986"} +{"question": "What research introduced the methodology of using support prototypes to enhance discriminative capabilities?", "answer": ["Learning Support and Trivial Prototypes for Interpretable Image\n Classification"], "answer_arxiv_id": ["2301.04011"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_19987"} +{"question": "Which paper discusses on the suodular multiway partition problem (Suodular-MP)?", "answer": ["Local Distribution and the Symmetry Gap: Approximability of Multiway Partitioning Problems"], "answer_arxiv_id": ["1503.03905v1"], "source_meta": {"published_time": "20220812"}, "qid": "AutoScholarQuery_train_19988"} +{"question": "What papers present BOBW algorithm with the data-dependent bound?", "answer": ["More Adaptive Algorithms for Adversarial Bandits", "Adversarially Robust Multi-Armed Bandit Algorithm with Variance-Dependent Regret Bounds"], "answer_arxiv_id": ["1801.03265", "2206.06810"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_19989"} +{"question": "What research transformed image-point cloud registration into point cloud visibility classification for direct 2D-3D registration?", "answer": ["DeepI2P: Image-to-Point Cloud Registration via Deep Classification"], "answer_arxiv_id": ["2104.03501"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19990"} +{"question": "Which papers present conventional representations for static 3D objects and scenes using voxels?", "answer": ["3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction", "Dense 3D Object Reconstruction from a Single Depth View", "Robust Attentional Aggregation of Deep Feature Sets for Multi-view 3D Reconstruction"], "answer_arxiv_id": ["1604.00449", "1802.00411", "1808.00758"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_19991"} +{"question": "Which papers discuss improving the privacy/utility trade-off in ML models training with DP-SGD through the use of large batch sizes?", "answer": ["Large-Scale Differentially Private BERT", "Large Language Models Can Be Strong Differentially Private Learners"], "answer_arxiv_id": ["2108.01624", "2110.05679"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_19992"} +{"question": "Can you name the papers that illustrated how a smaller but officially fine-tuned LLM can achieve similar evaluation results on natural language generation?", "answer": ["LLaMA: Open and Efficient Foundation Language Models", "TIGERScore: Towards Building Explainable Metric for All Text Generation\n Tasks", "Generative Judge for Evaluating Alignment"], "answer_arxiv_id": ["2302.13971", "2310.00752", "2310.05470"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_19993"} +{"question": "What are the works that reduced the decoding cost by making decoders shallow?", "answer": ["Deep Encoder, Shallow Decoder: Reevaluating Non-autoregressive Machine Translation", "Confident Adaptive Language Modeling"], "answer_arxiv_id": ["2006.10369", "2207.07061"], "source_meta": {"published_time": "20230215"}, "qid": "AutoScholarQuery_train_19994"} +{"question": "Are there any research works that propose methods for generating pruning masks?", "answer": ["Fire Together Wire Together: A Dynamic Pruning Approach with Self-Supervised Mask Prediction", "NISP: Pruning Networks using Neuron Importance Score Propagation", "Discrimination-aware Channel Pruning for Deep Neural Networks", "DMCP: Differentiable Markov Channel Pruning for Neural Networks"], "answer_arxiv_id": ["2110.08232", "1711.05908", "1810.11809", "2005.03354"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_19995"} +{"question": "Which research focused on deeply fusing visual and text knowledges to enhance VLM?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "FLAVA: A Foundational Language And Vision Alignment Model"], "answer_arxiv_id": ["2201.12086", "2112.04482"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_19996"} +{"question": "Which papers discussed that choosing a specific ordering does not rigorously correspond to maximum-likelihood estimation?", "answer": ["Efficient Graph Generation with Graph Recurrent Attention Networks", "Order Matters: Probabilistic Modeling of Node Sequence for Graph Generation"], "answer_arxiv_id": ["1910.00760", "2106.06189"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_19997"} +{"question": "What is the original work that combined neural networks and differential equations leading to the emergence of neural ordinary differential equations?", "answer": ["Neural Ordinary Differential Equations"], "answer_arxiv_id": ["1806.07366"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_19998"} +{"question": "Which works have developed efficient algorithms for learning Gaussian Mixture Models (GMMs) robustly?", "answer": ["Robustly Learning any Clusterable Mixture of Gaussians", "Robust Learning of Mixtures of Gaussians", "Settling the Robust Learnability of Mixtures of Gaussians"], "answer_arxiv_id": ["2005.06417", "2007.05912", "2011.03622"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_19999"} +{"question": "What prior works took on finding subnetworks inside a dense random network in a FL setting?", "answer": ["HideNseek: Federated Lottery Ticket via Server-side Pruning and Sign Supermask", "FRL: Federated Rank Learning"], "answer_arxiv_id": ["2206.04385", "2110.04350"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_20000"} +{"question": "Which research papers introduced widespread benchmarks for evaluating LLMs?", "answer": ["A Survey on Evaluation of Large Language Models", "Measuring Massive Multitask Language Understanding", "Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them"], "answer_arxiv_id": ["2307.03109", "2009.03300", "2210.09261v1"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_20001"} +{"question": "Which studies highlight the unreliability of evaluation metrics in image generation?", "answer": ["Effectively Unbiased FID and Inception Score and where to find them", "On Aliased Resizing and Surprising Subtleties in GAN Evaluation"], "answer_arxiv_id": ["1911.07023", "2104.11222"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_20002"} +{"question": "What works have included captioning losses in their study?", "answer": ["MERLOT: Multimodal Neural Script Knowledge Models", "LAVENDER: Unifying Video-Language Understanding as Masked Language\n Modeling"], "answer_arxiv_id": ["2106.02636", "2206.07160"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_20003"} +{"question": "Which works focus on adapting standardization statistics in normalization layers?", "answer": ["Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shift", "The Norm Must Go On: Dynamic Unsupervised Domain Adaptation by Normalization", "MixNorm: Test-Time Adaptation through Online Normalization Estimation", "Improving robustness against common corruptions by covariate shift adaptation", "Test-time Batch Statistics Calibration for Covariate Shift", "SITA: Single Image Test-time Adaptation", "TTN: A Domain-Shift Aware Batch Normalization in Test-Time Adaptation", "Learning Instance-Specific Adaptation for Cross-Domain Segmentation"], "answer_arxiv_id": ["2006.10963", "2112.00463", "2110.11478", "2006.16971v2", "2110.04065", "2112.02355", "2302.05155", "2203.16530"], "source_meta": {"published_time": "20230912"}, "qid": "AutoScholarQuery_train_20004"} +{"question": "What studies utilize the parallel scan algorithm to improve parallel training in traditional recurrent neural networks?", "answer": ["Parallelizing Linear Recurrent Neural Nets Over Sequence Length", "Simplified State Space Layers for Sequence Modeling"], "answer_arxiv_id": ["1709.04057", "2208.04933"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_20005"} +{"question": "What studies have used multi-view stereo from dense input views to create mesh surfaces as a proxy for image warping?", "answer": ["Free View Synthesis", "Stable View Synthesis"], "answer_arxiv_id": ["2008.05511", "2011.07233"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20006"} +{"question": "Which studies are associated with black-box defense approach at the model level?", "answer": ["One-Pixel Signature: Characterizing CNN Models for Backdoor Detection", "Black-box Detection of Backdoor Attacks with Limited Information and Data", "AEVA: Black-box Backdoor Detection Using Adversarial Extreme Value Analysis"], "answer_arxiv_id": ["2008.07711", "2103.13127", "2110.14880"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_20007"} +{"question": "Are there any works that assesses requirements for fitting discrete latent variable models using NAS-X and SIXO?", "answer": ["SIXO: Smoothing Inference with Twisted Objectives"], "answer_arxiv_id": ["2206.05952v2"], "source_meta": {"published_time": "20230828"}, "qid": "AutoScholarQuery_train_20008"} +{"question": "Which papers discussed the application of deep learning to software maintenance?", "answer": ["Improving Automatic Source Code Summarization via Deep Reinforcement\n Learning"], "answer_arxiv_id": ["1811.07234"], "source_meta": {"published_time": "20230716"}, "qid": "AutoScholarQuery_train_20009"} +{"question": "Which works applied curiosity-driven intrinsic reward to tackle the issue of delayed rewards in reinforcement learning?", "answer": ["Curiosity-driven Exploration by Self-supervised Prediction", "Episodic Multi-Agent Reinforcement Learning with Curiosity-Driven Exploration", "Language as a Cognitive Tool to Imagine Goals in Curiosity-Driven Exploration"], "answer_arxiv_id": ["1705.05363", "2111.11032", "2002.09253"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_20010"} +{"question": "What works have proposed methods to enhance LLMs with fine-grained understanding capabilities?", "answer": ["GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest", "Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic", "VisionLLM: Large Language Model is also an Open-Ended Decoder for\n Vision-Centric Tasks", "Kosmos-2: Grounding Multimodal Large Language Models to the World", "Ferret: Refer and Ground Anything Anywhere at Any Granularity"], "answer_arxiv_id": ["2307.03601", "2306.15195", "2305.11175", "2306.14824", "2310.07704"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20011"} +{"question": "What studies focused on composition of scenes from object-“slots” in the domain of object-centric learning?", "answer": ["Object-Centric Learning with Slot Attention", "Multi-Object Representation Learning with Iterative Variational Inference", "Conditional Object-Centric Learning from Video"], "answer_arxiv_id": ["2006.15055", "1903.00450", "2111.12594"], "source_meta": {"published_time": "20230710"}, "qid": "AutoScholarQuery_train_20012"} +{"question": "What studies used the concept of equal representation in fairness, like fair clustering variation and fair data summarization?", "answer": ["Fair Clustering Through Fairlets", "Fair and Diverse DPP-based Data Summarization"], "answer_arxiv_id": ["1802.05733", "1802.04023"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20013"} +{"question": "What works discuss time-dependent interpolation of radiance field in dynamic NeRFs?", "answer": ["Nerfies: Deformable Neural Radiance Fields", "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields", "Neural 3D Video Synthesis from Multi-view Video"], "answer_arxiv_id": ["2011.12948", "2106.13228", "2103.02597"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20014"} +{"question": "Which papers discuss the concept of G-equivariant function in relation to group transformations?", "answer": ["Group Equivariant Convolutional Networks"], "answer_arxiv_id": ["1602.07576"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_20015"} +{"question": "Which papers focus on producing adversarial future trajectories given agents’ initial states?", "answer": ["AdvSim: Generating Safety-Critical Scenarios for Self-Driving Vehicles", "Learning to Collide: An Adaptive Safety-Critical Scenarios Generating Method", "KING: Generating Safety-Critical Driving Scenarios for Robust Imitation via Kinematics Gradients"], "answer_arxiv_id": ["2101.06549", "2003.01197", "2204.13683"], "source_meta": {"published_time": "20231105"}, "qid": "AutoScholarQuery_train_20016"} +{"question": "Which work pointed out that the order of the two texts affects evaluation results when LLM-based evaluators are used?", "answer": ["Large Language Models are not Fair Evaluators"], "answer_arxiv_id": ["2305.17926"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_20017"} +{"question": "Where can I find research about V2I cooperative perception models?", "answer": ["VINet: Lightweight, Scalable, and Heterogeneous Cooperative Perception\n for 3D Object Detection", "PillarGrid: Deep Learning-based Cooperative Perception for 3D Object\n Detection from Onboard-Roadside LiDAR", "V2X-ViT: Vehicle-to-Everything Cooperative Perception with Vision\n Transformer", "Vehicle-Infrastructure Cooperative 3D Object Detection via Feature Flow\n Prediction", "CoBEVFusion: Cooperative Perception with LiDAR-Camera Bird's-Eye View\n Fusion"], "answer_arxiv_id": ["2212.07060", "2203.06319", "2203.10638", "2303.10552", "2310.06008"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_20018"} +{"question": "Which works investigate methods that directly process document images without external OCR tools?", "answer": ["Pix2Struct: Screenshot Parsing as Pretraining for Visual Language\n Understanding", "MatCha: Enhancing Visual Language Pretraining with Math Reasoning and\n Chart Derendering"], "answer_arxiv_id": ["2210.03347", "2212.09662"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20019"} +{"question": "Who proposed utilization of 'reverse' transition kernel to obtain a further lower-bound on the ELBO in MCVI?", "answer": ["Markov Chain Monte Carlo and Variational Inference: Bridging the Gap"], "answer_arxiv_id": ["1410.6460"], "source_meta": {"published_time": "20211125"}, "qid": "AutoScholarQuery_train_20020"} +{"question": "Could you provide me some researches that applied relaxed linear programming for image robustness verification?", "answer": ["Provable Defenses against Adversarial Examples via the Convex Outer Adversarial Polytope", "Towards Fast Computation of Certified Robustness for ReLU Networks"], "answer_arxiv_id": ["1711.00851", "1804.09699"], "source_meta": {"published_time": "20230616"}, "qid": "AutoScholarQuery_train_20021"} +{"question": "Are there any examples of hybrid models combining both hyperbolic and Euclidean geometries?", "answer": ["Graph Geometry Interaction Learning"], "answer_arxiv_id": ["2010.12135"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_20022"} +{"question": "Any works about Wasserstein natural gradient methods that replace Euclidean gradients?", "answer": ["Kernelized Wasserstein Natural Gradient", "Optimal transport natural gradient for statistical manifolds with continuous sample space", "Wasserstein Proximal of GANs"], "answer_arxiv_id": ["1910.09652", "1805.08380", "2102.06862"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_20023"} +{"question": "What studies have been done for single tumor segmentation?", "answer": ["The state of the art in kidney and kidney tumor segmentation in\n contrast-enhanced CT imaging: Results of the KiTS19 Challenge", "Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors\n in MRI Images"], "answer_arxiv_id": ["1912.01054", "2201.01266"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_20024"} +{"question": "Can you name some research papers where the authors developed Vision Transformer-based (Vit-based) models?", "answer": ["Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet", "Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions", "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "Going deeper with Image Transformers", "XCiT: Cross-Covariance Image Transformers", "Focal Self-attention for Local-Global Interactions in Vision Transformers"], "answer_arxiv_id": ["2101.11986", "2102.12122", "2103.14030", "2103.17239", "2106.09681", "2107.00641"], "source_meta": {"published_time": "20230424"}, "qid": "AutoScholarQuery_train_20025"} +{"question": "Could you provide me with studies that added consistency regularization explicitly?", "answer": ["Temporal Ensembling for Semi-Supervised Learning", "Learning with Pseudo-Ensembles", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence"], "answer_arxiv_id": ["1610.02242v3", "1412.4864", "2001.07685v2"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_20026"} +{"question": "What works modeled BEV features as queries at different positions?", "answer": ["BEVSegFormer: Bird's Eye View Semantic Segmentation From Arbitrary\n Camera Rigs", "PETR: Position Embedding Transformation for Multi-View 3D Object\n Detection", "BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera\n Images via Spatiotemporal Transformers"], "answer_arxiv_id": ["2203.04050", "2203.05625", "2203.17270"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_20027"} +{"question": "Can you give me some example works which apply meta-learning in few-shot learning?", "answer": ["Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks", "Meta-Learning with Implicit Gradients", "A contrastive rule for meta-learning"], "answer_arxiv_id": ["1703.03400", "1909.04630", "2104.01677v3"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_20028"} +{"question": "What studies in literature have analyzed deep ReLU networks?", "answer": ["An Improved Analysis of Training Over-parameterized Deep Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Global Convergence of Deep Networks with One Wide Layer Followed by Pyramidal Topology", "On the Proof of Global Convergence of Gradient Descent for Deep ReLU Networks with Linear Widths", "Tight Bounds on the Smallest Eigenvalue of the Neural Tangent Kernel for Deep ReLU Networks"], "answer_arxiv_id": ["1906.04688", "1811.03962", "2002.07867", "2101.09612", "2012.11654"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_20029"} +{"question": "Which work used gradient-based meta-learning but adapted whole weights of the shared Multi Layer Perceptron (MLP) in the context of generalizable INRs?", "answer": ["Learned Initializations for Optimizing Coordinate-Based Neural Representations"], "answer_arxiv_id": ["2012.02189"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_20030"} +{"question": "Which work leveraged unlabeled narrated videos for DVC pre-training?", "answer": ["Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense\n Video Captioning"], "answer_arxiv_id": ["2302.14115"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_20031"} +{"question": "Could you mention some papers that discussed BERT-style pretraining?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "BEiT: BERT Pre-Training of Image Transformers"], "answer_arxiv_id": ["1810.04805", "2106.08254"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_20032"} +{"question": "What are some recent studies that tackle the overestimation issue in reinforcement learning?", "answer": ["Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning", "Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning", "MOReL: Model-Based Offline Reinforcement Learning"], "answer_arxiv_id": ["1910.00177", "2006.04779", "2110.06169", "2005.05951"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_20033"} +{"question": "What studies are about test-time adaptation for robustness against distribution shift?", "answer": ["Adversarial Attacks are Reversible with Natural Supervision", "MEMO: Test Time Robustness via Adaptation and Augmentation"], "answer_arxiv_id": ["2103.14222", "2110.09506"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_20034"} +{"question": "Could you provide me some studies that use condition further generation on the execution states in code generation literature?", "answer": ["Write, Execute, Assess: Program Synthesis with a REPL", "Representing Partial Programs with Blended Abstract Semantics", "Show Your Work: Scratchpads for Intermediate Computation with Language Models"], "answer_arxiv_id": ["1906.04604", "2012.12964", "2112.00114"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_20035"} +{"question": "What was the recent work that also consists of user-annotated chapters, similar to VidChapters-7M?", "answer": ["Multi-modal Video Chapter Generation"], "answer_arxiv_id": ["2209.12694v1"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_20036"} +{"question": "What papers proposed methods to address the inconsistency between teacher and student features in semantic segmentation tasks?", "answer": ["Knowledge Adaptation for Efficient Semantic Segmentation", "Channel-wise Knowledge Distillation for Dense Prediction", "Cross-Image Relational Knowledge Distillation for Semantic Segmentation"], "answer_arxiv_id": ["1903.04688v1", "2011.13256v4", "2204.06986"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_20037"} +{"question": "In what paper was a pre-training method suggested for predicting cognitive task types during an fMRI scan?", "answer": ["Attend and Decode: 4D fMRI Task State Decoding Using Attention Models"], "answer_arxiv_id": ["2004.05234v2"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_20038"} +{"question": "What literature discusses machine learning in the context of program equilibrium?", "answer": ["Learning in two-player games between transparent opponents"], "answer_arxiv_id": ["2012.02671"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_20039"} +{"question": "Could you provide me the paper that suggested a method of using the pull-back metric defined with arbitrary decoders on the latent space?", "answer": ["Pulling back information geometry"], "answer_arxiv_id": ["2106.05367v2"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_20040"} +{"question": "What work utilizes existing high-performance encoders and decoders for multimodal generation across various modalities in LLMs?", "answer": ["NExT-GPT: Any-to-Any Multimodal LLM"], "answer_arxiv_id": ["2309.05519"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_20041"} +{"question": "What research introduced the use of word embedding to a time-space conditioned neural mapper?", "answer": ["A Neural Space-Time Representation for Text-to-Image Personalization"], "answer_arxiv_id": ["2305.15391"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_20042"} +{"question": "Could you provide me a study that complements this research's Riemannian manifold perspective?", "answer": ["Contrastive and Non-Contrastive Self-Supervised Learning Recover Global and Local Spectral Embedding Methods"], "answer_arxiv_id": ["2205.11508"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20043"} +{"question": "Which papers introduced the use of an adversarial attack to text data?", "answer": ["Explaining and Harnessing Adversarial Examples", "TextBugger: Generating Adversarial Text Against Real-world Applications"], "answer_arxiv_id": ["1412.6572", "1812.05271"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_20044"} +{"question": "Which studies proposed methods to evaluate multi-agent simulation?", "answer": ["TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors", "SimNet: Learning Reactive Self-driving Simulations from Real-world Observations", "Symphony: Learning Realistic and Diverse Agents for Autonomous Driving Simulation", "Nocturne: a scalable driving benchmark for bringing multi-agent learning one step closer to the real world", "BITS: Bi-level Imitation for Traffic Simulation", "InterSim: Interactive Traffic Simulation via Explicit Relation Modeling"], "answer_arxiv_id": ["2101.06557", "2105.12332", "2205.03195", "2206.09889", "2208.12403", "2210.14413"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_20045"} +{"question": "Which papers have used differentiable rendering frameworks for coupling with visual observations in 3D reconstruction context?", "answer": ["Differentiable Rendering: A Survey"], "answer_arxiv_id": ["2006.12057"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_20046"} +{"question": "Which studies attempted to capture feature learning and predictor variance from perturbative series around infinite-width dynamics?", "answer": ["Finite Depth and Width Corrections to the Neural Tangent Kernel", "Asymptotics of Wide Networks from Feynman Diagrams", "The Principles of Deep Learning Theory"], "answer_arxiv_id": ["1909.05989", "1909.11304", "2106.10165"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_20047"} +{"question": "Could you provide me some studies that focus on specific areas such as eyes and lips to discern fake facial media?", "answer": ["Lips Don't Lie: A Generalisable and Robust Approach to Face Forgery\n Detection"], "answer_arxiv_id": ["2012.07657"], "source_meta": {"published_time": "20231216"}, "qid": "AutoScholarQuery_train_20048"} +{"question": "What research works contribute to the understanding of DNN models being vulnerable to adversarial attacks?", "answer": ["Intriguing properties of neural networks", "Explaining and Harnessing Adversarial Examples", "Interpretable Deep Learning under Fire"], "answer_arxiv_id": ["1312.6199", "1412.6572", "1812.00891"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_20049"} +{"question": "What papers studied Language-Level Models (LLMs) emphasizing compliance with arbitrary human instructions?", "answer": ["GPT-4 Technical Report"], "answer_arxiv_id": ["2303.08774"], "source_meta": {"published_time": "20231112"}, "qid": "AutoScholarQuery_train_20050"} +{"question": "Could you provide me some studies that incorporated channel attention mechanisms in super-resolution?", "answer": ["Image Super-Resolution Using Very Deep Residual Channel Attention\n Networks"], "answer_arxiv_id": ["1807.02758"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_20051"} +{"question": "What research has been done on measuring the importance of discovered concepts?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with\n Concept Activation Vectors (TCAV)", "Towards Automatic Concept-based Explanations", "Invertible Concept-based Explanations for CNN Models with Non-negative\n Concept Activation Vectors", "CRAFT: Concept Recursive Activation FacTorization for Explainability", "A Holistic Approach to Unifying Automatic Concept Extraction and Concept\n Importance Estimation"], "answer_arxiv_id": ["1711.11279", "1902.03129", "2006.15417", "2211.10154", "2306.07304"], "source_meta": {"published_time": "20240119"}, "qid": "AutoScholarQuery_train_20052"} +{"question": "Which pre-trained language models generate summaries in an end-to-end manner?", "answer": ["BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer", "PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization"], "answer_arxiv_id": ["1910.13461", "1910.10683", "1912.08777"], "source_meta": {"published_time": "20220921"}, "qid": "AutoScholarQuery_train_20053"} +{"question": "What papers introduce the construction of more challenging datasets for image retrieval?", "answer": ["Deep Metric Learning via Lifted Structured Feature Embedding", "The iNaturalist Species Classification and Detection Dataset", "Google Landmarks Dataset v2 A Large-Scale Benchmark for Instance-Level Recognition and Retrieval"], "answer_arxiv_id": ["1511.06452", "1707.06642", "2004.01804"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_20054"} +{"question": "Could you provide me some studies where probabilistic regression for uncertainty prediction has been used?", "answer": ["Uncertainty Estimates and Multi-Hypotheses Networks for Optical Flow", "On the uncertainty of self-supervised monocular depth estimation", "Deep Uncertainty Quantification: A Machine Learning Approach for Weather Forecasting"], "answer_arxiv_id": ["1802.07095", "2005.06209", "1812.09467"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_20055"} +{"question": "Which studies worked on human motion modeling conditioned on past motion?", "answer": ["Convolutional Sequence to Sequence Model for Human Dynamics", "HP-GAN: Probabilistic 3D human motion prediction via GAN", "DLow: Diversifying Latent Flows for Diverse Human Motion Prediction", "Long-term Human Motion Prediction with Scene Context", "Physics-based Human Motion Estimation and Synthesis from Videos", "Contact-aware Human Motion Forecasting"], "answer_arxiv_id": ["1805.00655", "1711.09561", "2003.08386", "2007.03672", "2109.09913", "2210.03954"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_20056"} +{"question": "Which papers explored pseudo-labeling methods in SSL, generating artificial labels of unlabeled data?", "answer": ["Meta Pseudo Labels", "PseudoSeg: Designing Pseudo Labels for Semantic Segmentation"], "answer_arxiv_id": ["2003.10580", "2010.09713"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_20057"} +{"question": "Which papers focused on representation learning in the Neural Tangent Kernel (NTK) setting?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "Scaling Limits of Wide Neural Networks with Weight Sharing: Gaussian Process Behavior, Gradient Independence, and Neural Tangent Kernel Derivation", "Feature Learning in Infinite-Width Neural Networks"], "answer_arxiv_id": ["1806.07572", "1902.04760", "2011.14522v3"], "source_meta": {"published_time": "20210830"}, "qid": "AutoScholarQuery_train_20058"} +{"question": "Which research proposed an efficient robustness strategy using conditional value at risk (CVar) to mitigate spurious correlations?", "answer": ["Large-Scale Methods for Distributionally Robust Optimization"], "answer_arxiv_id": ["2010.05893"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_20059"} +{"question": "Which works originally propose the GLM-tron algorithm for learning the well-specified Generalized Linear Model?", "answer": ["Efficient Learning of Generalized Linear and Single Index Models with Isotonic Regression"], "answer_arxiv_id": ["1104.2018"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_20060"} +{"question": "Which works focus on visual reference segmentation?", "answer": ["Singular Value Fine-tuning: Few-shot Segmentation requires\n Few-parameters Fine-tuning", "Segment Everything Everywhere All at Once", "Prior Guided Feature Enrichment Network for Few-Shot Segmentation"], "answer_arxiv_id": ["2206.06122", "2304.06718", "2008.01449"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_20061"} +{"question": "Does any work address model heterogeneity in Federated Learning using knowledge distillation-based methods?", "answer": ["FedMD: Heterogenous Federated Learning via Model Distillation", "Group Knowledge Transfer: Federated Learning of Large CNNs at the Edge", "P"], "answer_arxiv_id": ["1910.03581", "2007.14513", "0704.0320"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_20062"} +{"question": "What works propose the integration of 2D images and 3D data for visual grounding?", "answer": ["SAT: 2D Semantics Assisted Training for 3D Visual Grounding", "Bottom Up Top Down Detection Transformers for Language Grounding in Images and Point Clouds"], "answer_arxiv_id": ["2105.11450", "2112.08879"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_20063"} +{"question": "What are the works that expand LLMs’ capabilities into the multimodal domain?", "answer": ["Language Is Not All You Need: Aligning Perception with Language Models", "MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large\n Language Models", "Visual Instruction Tuning", "GraphText: Graph Reasoning in Text Space", "NExT-GPT: Any-to-Any Multimodal LLM"], "answer_arxiv_id": ["2302.14045", "2304.10592", "2304.08485", "2310.01089v1", "2309.05519"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_20064"} +{"question": "Who proposed ImageNet-C to benchmark the common corruption robustness of deep learning models?", "answer": ["Benchmarking Neural Network Robustness to Common Corruptions and Perturbations"], "answer_arxiv_id": ["1903.12261"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_20065"} +{"question": "Which works introduced the discrete time-step diffusion models and the parameterized reversal process for generating novel data?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["1503.03585", "2006.11239"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_20066"} +{"question": "Any works about the technique of clustering for local image editing in the context of Generative Adversarial Networks (GANs)?", "answer": ["Editing in Style: Uncovering the Local Semantics of GANs", "Decorating Your Own Bedroom: Locally Controlling Image Generation with Generative Adversarial Networks", "StyleFusion: A Generative Model for Disentangling Spatial Segments"], "answer_arxiv_id": ["2004.14367", "2105.08222", "2107.07437"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_20067"} +{"question": "What works directly apply SDS or DDS to NeRF editing?", "answer": ["ED-NeRF: Efficient Text-Guided Editing of 3D Scene with Latent Space\n NeRF", "FocalDreamer: Text-driven 3D Editing via Focal-fusion Assembly", "DreamEditor: Text-Driven 3D Scene Editing with Neural Fields"], "answer_arxiv_id": ["2310.02712", "2308.10608", "2306.13455"], "source_meta": {"published_time": "20231123"}, "qid": "AutoScholarQuery_train_20068"} +{"question": "What research investigates the role of non-robust features in adversarial attacks according to IIyas et al.?", "answer": ["Adversarial Examples Are Not Bugs, They Are Features"], "answer_arxiv_id": ["1905.02175"], "source_meta": {"published_time": "20230830"}, "qid": "AutoScholarQuery_train_20069"} +{"question": "Which studies conducted an optimization on continuous space by implementing the trust region approach?", "answer": ["Scalable Global Optimization via Local Bayesian Optimization", "Increasing the Scope as You Learn: Adaptive Bayesian Optimization in Nested Subspaces"], "answer_arxiv_id": ["1910.01739", "2304.11468"], "source_meta": {"published_time": "20230702"}, "qid": "AutoScholarQuery_train_20070"} +{"question": "What papers discuss applying knowledge distillation for better optimization in diffusion models?", "answer": ["Progressive Distillation for Fast Sampling of Diffusion Models", "Knowledge Distillation in Iterative Generative Models for Improved\n Sampling Speed", "On Distillation of Guided Diffusion Models", "Consistency Models"], "answer_arxiv_id": ["2202.00512", "2101.02388", "2210.03142v3", "2303.01469"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_20071"} +{"question": "Which research papers discuss the 'dirty-label' data poisoning attack?", "answer": ["Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning"], "answer_arxiv_id": ["1712.05526v1"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_20072"} +{"question": "What papers discussed the use of Mean absolute error (MAE) loss in addressing label noise?", "answer": ["Robust Loss Functions under Label Noise for Deep Neural Networks"], "answer_arxiv_id": ["1712.09482"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_20073"} +{"question": "In what works the 'virtual node' concept was introduced?", "answer": ["Graph Classification via Deep Learning with Virtual Nodes"], "answer_arxiv_id": ["1708.04357"], "source_meta": {"published_time": "20230521"}, "qid": "AutoScholarQuery_train_20074"} +{"question": "Which papers discuss black-box attacks aiming to mislead end-to-end acoustic systems?", "answer": ["SirenAttack: Generating Adversarial Audio for End-to-End Acoustic Systems", "Who is Real Bob? Adversarial Attacks on Speaker Recognition Systems"], "answer_arxiv_id": ["1901.07846v2", "1911.01840"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_20075"} +{"question": "Which papers directly used PowerAgg in their research?", "answer": ["DeeperGCN: All You Need to Train Deeper GCNs", "Generalizing Aggregation Functions in GNNs: High-Capacity GNNs via Nonlinear Neighborhood Aggregators", "Learned-Norm Pooling for Deep Feedforward and Recurrent Neural Networks"], "answer_arxiv_id": ["2006.07739v1", "2202.09145", "1311.1780"], "source_meta": {"published_time": "20230624"}, "qid": "AutoScholarQuery_train_20076"} +{"question": "What papers contain a study regarding the brevity of HTML documents and representation of these documents in web navigation?", "answer": ["A data-driven approach for learning to control computers", "WebShop: Towards Scalable Real-World Web Interaction with Grounded\n Language Agents", "Reinforcement Learning on Web Interfaces Using Workflow-Guided\n Exploration", "DOM-Q-NET: Grounded RL on Structured Language"], "answer_arxiv_id": ["2202.08137", "2207.01206", "1802.08802", "1902.07257"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_20077"} +{"question": "What works propose a contrastive loss for aligning pairs of image and text in multi-modal text guided image models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_20078"} +{"question": "Any research papers on automated architecture search algorithm for building a tree-structured network?", "answer": ["Learning to Branch for Multi-Task Learning", "Automated Search for Resource-Efficient Branched Multi-Task Networks"], "answer_arxiv_id": ["2006.01895", "2008.10292"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_20079"} +{"question": "What work, which utilizes a deterministic model, employs an initialization scheme that requires numerous forward passes through G and f?", "answer": ["FuseDream: Training-Free Text-to-Image Generation with Improved CLIP+GAN Space Optimization"], "answer_arxiv_id": ["2112.01573"], "source_meta": {"published_time": "20230426"}, "qid": "AutoScholarQuery_train_20080"} +{"question": "What papers propose neural network architectures specifically for uncertainty estimation, such as Bayesian Neural Networks and Deep Ensembles?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles"], "answer_arxiv_id": ["1506.02142", "1612.01474"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_20081"} +{"question": "What studies have shown that approximatability by a shallow (depth 3 network) is a necessary condition for learnability via a deeper network?", "answer": ["The Connection Between Approximation, Depth Separation and Learnability in Neural Networks"], "answer_arxiv_id": ["2102.00434"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_20082"} +{"question": "What papers show application of Diffusion Models for conditional Novel View Synthesis in the field of 3D reconstruction?", "answer": ["Novel View Synthesis with Diffusion Models", "Zero-1-to-3: Zero-shot One Image to 3D Object", "One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape\n Optimization"], "answer_arxiv_id": ["2210.04628", "2303.11328", "2306.16928"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_20083"} +{"question": "What research papers implemented diffusion models in image perception tasks such as object detection and image segmentation?", "answer": ["DiffusionDet: Diffusion Model for Object Detection", "GMMSeg: Gaussian Mixture based Generative Semantic Segmentation Models"], "answer_arxiv_id": ["2211.09788", "2210.02025"], "source_meta": {"published_time": "20230706"}, "qid": "AutoScholarQuery_train_20084"} +{"question": "What research paper discusses the challenge of existing methods in machine unlearning needing full dataset access for retraining?", "answer": ["Adaptive Machine Unlearning", "Fast Yet Effective Machine Unlearning", "Machine Unlearning: Linear Filtration for Logit-based Classifiers", "Making AI Forget You: Data Deletion in Machine Learning"], "answer_arxiv_id": ["2106.04378", "2111.08947", "2002.02730", "1907.05012"], "source_meta": {"published_time": "20240516"}, "qid": "AutoScholarQuery_train_20085"} +{"question": "Is there any study that discusses affordance in the context of object to object interaction?", "answer": ["O2O-Afford: Annotation-Free Large-Scale Object-Object Affordance Learning"], "answer_arxiv_id": ["2106.15087"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_20086"} +{"question": "Could you cite some works that regularized discriminators for training stability in GANs?", "answer": ["Improved Training of Wasserstein GANs", "Improving GANs with A Dynamic Discriminator", "Generator Knows What Discriminator Should Learn in Unconditional GANs", "GLeaD: Improving GANs with A Generator-Leading Task"], "answer_arxiv_id": ["1704.00028", "2209.09897", "2207.13320", "2212.03752"], "source_meta": {"published_time": "20230829"}, "qid": "AutoScholarQuery_train_20087"} +{"question": "Are there any studies that directly learn the articulated 3D shape without using traditional template-based or skeleton-based methods?", "answer": ["DOVE: Learning Deformable 3D Objects by Watching Videos", "MagicPony: Learning Articulated 3D Animals in the Wild"], "answer_arxiv_id": ["2107.10844", "2211.12497"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_20088"} +{"question": "What studies are about the augmentation of latent vectors produced by passing data inputs through a neural network?", "answer": ["Manifold Mixup: Better Representations by Interpolating Hidden States", "A Closer Look At Feature Space Data Augmentation For Few-Shot Intent Classification"], "answer_arxiv_id": ["1806.05236", "1910.04176"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_20089"} +{"question": "Are there any works utilizing CLIP to manipulate a shape or NeRF with text?", "answer": ["Text2Mesh: Text-Driven Neural Stylization for Meshes", "ClipMatrix: Text-controlled Creation of 3D Textured Meshes", "CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields"], "answer_arxiv_id": ["2112.03221", "2109.12922", "2112.05139"], "source_meta": {"published_time": "20220909"}, "qid": "AutoScholarQuery_train_20090"} +{"question": "Any works about point-level fusion methods in cross-modal 3D object detection?", "answer": ["Deep Continuous Fusion for Multi-Sensor 3D Object Detection", "PointPainting: Sequential Fusion for 3D Object Detection", "EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection"], "answer_arxiv_id": ["2012.10992", "1911.10150", "2007.08856"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_20091"} +{"question": "Are there any studies that convert the registration problem into a classification and inverse camera projection optimization problem?", "answer": ["DeepI2P: Image-to-Point Cloud Registration via Deep Classification"], "answer_arxiv_id": ["2104.03501"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_20092"} +{"question": "Can you provide me with examples of research papers where the MAZE method is utilized for open-ended learning?", "answer": ["Heterogeneous Multi-agent Zero-Shot Coordination by Coevolution"], "answer_arxiv_id": ["2208.04957"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20093"} +{"question": "Give me some studies that use a pretrained self-supervised model to train a classifier to detect and mitigate spurious correlations.", "answer": ["Group Robust Classification Without Any Group Information"], "answer_arxiv_id": ["2310.18555"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_20094"} +{"question": "Which research works dealt with the approximation of the joint posterior using variational methods?", "answer": ["DiBS: Differentiable Bayesian Structure Learning", "BCD Nets: Scalable Variational Approaches for Bayesian Causal Discovery"], "answer_arxiv_id": ["2105.11839", "2112.02761"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20095"} +{"question": "Are there any studies using adversarial learning in UDA?", "answer": ["Learning Transferable Features with Deep Adaptation Networks", "ADVENT: Adversarial Entropy Minimization for Domain Adaptation in\n Semantic Segmentation", "CyCADA: Cycle-Consistent Adversarial Domain Adaptation"], "answer_arxiv_id": ["1502.02791", "1811.12833", "1711.03213"], "source_meta": {"published_time": "20230316"}, "qid": "AutoScholarQuery_train_20096"} +{"question": "What works developed first method that can enforce general equality constraints onto deep learning models?", "answer": ["DC3: A learning method for optimization with hard constraints"], "answer_arxiv_id": ["2104.12225"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_20097"} +{"question": "Which study used white balance, image contrast enhancement, and gamma transformation to preprocess the original image for end-to-end restoration of hazy images?", "answer": ["Gated Fusion Network for Single Image Dehazing"], "answer_arxiv_id": ["1804.00213"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_20098"} +{"question": "What are the works related to Textual Entailment (TE) and it's application in assessing the semantic alignment between images and text (VE)?", "answer": ["Visual Entailment: A Novel Task for Fine-Grained Image Understanding"], "answer_arxiv_id": ["1901.06706"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_20099"} +{"question": "What papers analyzed the limitations of expressive power in popular GNN designs?", "answer": ["Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning", "Graph Neural Networks Exponentially Lose Expressive Power for Node Classification", "How Neural Networks Extrapolate: From Feedforward to Graph Neural Networks", "Not too little, not too much: a theoretical analysis of graph (over)smoothing"], "answer_arxiv_id": ["1801.07606", "1905.10947", "2009.11848", "2205.12156"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_20100"} +{"question": "Could you provide me some works that utilize the compiler's intermediate representation to ground code generation?", "answer": ["Unleashing the Power of Compiler Intermediate Representation to Enhance\n Neural Program Embeddings", "Code Translation with Compiler Representations"], "answer_arxiv_id": ["2204.09191", "2207.03578"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_20101"} +{"question": "Are there any follow-up works that have continued to use the attention alignment idea?", "answer": ["LAVT: Language-Aware Vision Transformer for Referring Image Segmentation", "GRES: Generalized Referring Expression Segmentation", "PolyFormer: Referring Image Segmentation as Sequential Polygon\n Generation"], "answer_arxiv_id": ["2112.02244", "2306.00968", "2302.07387"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20102"} +{"question": "Which studies have successfully recovered interreflection- and shadow-free SVBRDF materials and manipulate large outdoor urban scene through proper modeling of the indirect illumination and direct illumination visibility, and disentangling complex geometry and materials from lighting effects?", "answer": ["Modeling Indirect Illumination for Inverse Rendering", "Neural Fields meet Explicit Geometric Representation for Inverse Rendering of Urban Scenes"], "answer_arxiv_id": ["2204.06837", "2304.03266v1"], "source_meta": {"published_time": "20240110"}, "qid": "AutoScholarQuery_train_20103"} +{"question": "What research showed that memorization increases with scale in the context of models trained on web-scale corpora?", "answer": ["Quantifying Memorization Across Neural Language Models"], "answer_arxiv_id": ["2202.07646"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_20104"} +{"question": "Could you list the research papers where diffusion models were used for predicting actions?", "answer": ["Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning", "Diffusion Policy: Visuomotor Policy Learning via Action Diffusion"], "answer_arxiv_id": ["2208.06193", "2303.04137v5"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_20105"} +{"question": "Could you name some works that applied NeRF in 3D reconstruction?", "answer": ["Neuralangelo: High-Fidelity Neural Surface Reconstruction", "MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient\n Neural Field Rendering on Mobile Architectures", "Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields", "Baking Neural Radiance Fields for Real-Time View Synthesis"], "answer_arxiv_id": ["2306.03092", "2208.00277", "2111.12077", "2103.14645"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_20106"} +{"question": "Which works proposed the use of surrogate models in black-box optimization to reduce costs?", "answer": ["Conditioning by adaptive sampling for robust design"], "answer_arxiv_id": ["1901.10060"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_20107"} +{"question": "What studies leveraged scene graph generation to improve object detection and visual relationship detection?", "answer": ["Exploiting Scene Graphs for Human-Object Interaction Detection"], "answer_arxiv_id": ["2108.08584"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_20108"} +{"question": "Which papers deal with the 3D avatar generation from text?", "answer": ["AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars", "DreamFace: Progressive Generation of Animatable 3D Faces under Text\n Guidance", "DreamAvatar: Text-and-Shape Guided 3D Human Avatar Generation via\n Diffusion Models", "AvatarCraft: Transforming Text into Neural Human Avatars with\n Parameterized Shape and Pose Control", "DreamHuman: Animatable 3D Avatars from Text", "DreamWaltz: Make a Scene with Complex 3D Animatable Avatars"], "answer_arxiv_id": ["2205.08535", "2304.03117", "2304.00916", "2303.17606", "2306.09329", "2305.12529"], "source_meta": {"published_time": "20231228"}, "qid": "AutoScholarQuery_train_20109"} +{"question": "Which studies propose a defense for traditional data poisoning using deterministic base learners?", "answer": ["Deep Partition Aggregation: Provable Defenses against General Poisoning Attacks", "Improved Certified Defenses against Data Poisoning with (Deterministic) Finite Aggregation", "Lethal Dose Conjecture on Data Poisoning", "Run-Off Election: Improved Provable Defense against Data Poisoning Attacks", "On Practical Aspects of Aggregation Defenses against Data Poisoning Attacks"], "answer_arxiv_id": ["2006.14768", "2202.02628", "2208.03309", "2302.02300", "2306.16415"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_20110"} +{"question": "Could you provide me a study related to Byzantine robustness in the asynchronous communication and unconstrained topologies settings?", "answer": ["Collaborative Learning in the Jungle (Decentralized, Byzantine, Heterogeneous, Asynchronous and Nonconvex Learning)"], "answer_arxiv_id": ["2008.00742"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_20111"} +{"question": "Which papers discussed the translation equivariance property of convolutional neural networks?", "answer": ["On Translation Invariance in CNNs: Convolutional Layers can Exploit Absolute Spatial Location"], "answer_arxiv_id": ["2003.07064"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20112"} +{"question": "Which works offer optimal error for empirical risk minimization in the problem of convex optimization?", "answer": ["Revisiting differentially private linear regression: optimal and adaptive prediction & estimation in unbounded domain"], "answer_arxiv_id": ["1803.02596"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_20113"} +{"question": "What works utilize tree search guided by a self-critic reward model to find optimal problem-solving paths?", "answer": ["Tree of Thoughts: Deliberate Problem Solving with Large Language Models"], "answer_arxiv_id": ["2305.10601"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_20114"} +{"question": "Which works have developed advanced adversarial training techniques building upon the foundation of AT?", "answer": ["Theoretically Principled Trade-off between Robustness and Accuracy", "Adversarial Weight Perturbation Helps Robust Generalization", "Geometry-aware Instance-reweighted Adversarial Training", "LAS-AT: Adversarial Training with Learnable Attack Strategy"], "answer_arxiv_id": ["1901.08573", "2004.05884", "2010.01736", "2203.06616"], "source_meta": {"published_time": "20240315"}, "qid": "AutoScholarQuery_train_20115"} +{"question": "What papers discuss the use of the model-free method in non-stationary RL?", "answer": ["Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments", "Online Meta-Learning", "Optimizing for the Future in Non-Stationary MDPs", "Towards Safe Policy Improvement for Non-Stationary MDPs", "Model-Free Non-Stationary RL: Near-Optimal Regret and Applications in Multi-Agent RL and Inventory Control", "Dynamic Regret of Policy Optimization in Non-stationary Environments"], "answer_arxiv_id": ["1710.03641", "1902.08438", "2005.08158", "2010.12645", "2010.03161v4", "2007.00148"], "source_meta": {"published_time": "20230926"}, "qid": "AutoScholarQuery_train_20116"} +{"question": "Can you list the works that exploit the distances between labels to preserve ordinality for ordinal classification?", "answer": ["Class Distance Weighted Cross-Entropy Loss for Ulcerative Colitis Severity Estimation", "RankSim: Ranking Similarity Regularization for Deep Imbalanced Regression"], "answer_arxiv_id": ["2202.05167", "2205.15236"], "source_meta": {"published_time": "20230121"}, "qid": "AutoScholarQuery_train_20117"} +{"question": "Which works developed methods that utilize Graph Neural Networks (GNNs) for face geometry estimation?", "answer": ["Towards High-Fidelity 3D Face Reconstruction from In-the-Wild Images\n Using Graph Convolutional Networks"], "answer_arxiv_id": ["2003.05653"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_20118"} +{"question": "Could you provide me some studies examining the correlation of next word prediction task with downstream performance?", "answer": ["Same Pre-training Loss, Better Downstream: Implicit Bias Matters for\n Language Models", "Training Trajectories of Language Models Across Scales", "Scaling Laws vs Model Architectures: How does Inductive Bias Influence\n Scaling?"], "answer_arxiv_id": ["2210.14199", "2212.09803", "2207.10551"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_20119"} +{"question": "Which works studied the heavy-hitter problem in the local model?", "answer": ["Local, Private, Efficient Protocols for Succinct Histograms", "Collecting Telemetry Data Privately", "Communication Complexity in Locally Private Distribution Estimation and Heavy Hitters", "Heavy Hitters and the Structure of Local Privacy", "Practical Locally Private Heavy Hitters"], "answer_arxiv_id": ["1504.04686", "1712.01524", "1905.11888", "1711.04740", "1707.04982"], "source_meta": {"published_time": "20230222"}, "qid": "AutoScholarQuery_train_20120"} +{"question": "What works propose a general framework for computing approximate gradients of metrics?", "answer": ["Training Deep Neural Networks via Direct Loss Minimization"], "answer_arxiv_id": ["1511.06411"], "source_meta": {"published_time": "20230216"}, "qid": "AutoScholarQuery_train_20121"} +{"question": "Which papers present both asymptotic and non-asymptotic convergence analysis for the deterministic or stochastic bilevel optimization?", "answer": ["Bilevel Programming for Hyperparameter Optimization and Meta-Learning", "Truncated Back-propagation for Bilevel Optimization", "A Value-Function-based Interior-point Method for Non-convex Bi-level Optimization", "Approximation Methods for Bilevel Programming", "Bilevel Optimization: Convergence Analysis and Enhanced Design", "A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic", "Efficiently Escaping Saddle Points in Bilevel Optimization"], "answer_arxiv_id": ["1806.04910", "1810.10667", "2106.07991", "1802.02246", "2010.07962", "2007.05170v4", "2202.03684v2"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_20122"} +{"question": "Are there works on critic regularization methods modifying the Q-function's objective?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Actionable Models: Unsupervised Offline Reinforcement Learning of Robotic Skills", "COMBO: Conservative Offline Model-Based Policy Optimization", "AlgaeDICE: Policy Gradient from Arbitrary Experience", "Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble", "Pessimistic Bootstrapping for Uncertainty-Driven Offline Reinforcement Learning", "The Importance of Pessimism in Fixed-Dataset Policy Optimization"], "answer_arxiv_id": ["2006.04779", "2104.07749", "2102.08363", "1912.02074", "2110.01548", "2202.11566", "2009.06799"], "source_meta": {"published_time": "20230724"}, "qid": "AutoScholarQuery_train_20123"} +{"question": "What works are based on the InfoNCE in the field of contrastive methods?", "answer": ["Representation Learning with Contrastive Predictive Coding", "Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1807.03748", "1911.05722", "2002.05709"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_20124"} +{"question": "Which paper developed the method known as SD-R used for image variation?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2112.10752"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_20125"} +{"question": "Can you cite studies that introduced implicit structural guidance into MaskGiT?", "answer": ["MaskSketch: Unpaired Structure-guided Masked Image Generation"], "answer_arxiv_id": ["2302.05496"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_20126"} +{"question": "Which works demonstrate the application of DPR in knowledge-intensive tasks?", "answer": ["Reading Wikipedia to Answer Open-Domain Questions", "Leveraging Passage Retrieval with Generative Models for Open Domain\n Question Answering", "REALM: Retrieval-Augmented Language Model Pre-Training", "Latent Retrieval for Weakly Supervised Open Domain Question Answering", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks"], "answer_arxiv_id": ["1704.00051", "2007.01282", "2002.08909", "1906.00300", "2005.11401"], "source_meta": {"published_time": "20240213"}, "qid": "AutoScholarQuery_train_20127"} +{"question": "Who had studied the Correlation Clustering problem under the assumption of random or semi-random errors made by the classifier?", "answer": ["Correlation Clustering with Noisy Partial Information"], "answer_arxiv_id": ["1406.5667"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_20128"} +{"question": "Could you provide me some studies that use function approximation for density ratio?", "answer": ["AlgaeDICE: Policy Gradient from Arbitrary Experience", "Offline Reinforcement Learning with Realizability and Single-policy Concentrability", "Offline Reinforcement Learning Under Value and Density-Ratio Realizability: The Power of Gaps", "Minimax Value Interval for Off-Policy Evaluation and Policy Optimization"], "answer_arxiv_id": ["1912.02074", "2202.04634", "2203.13935", "2002.02081"], "source_meta": {"published_time": "20221228"}, "qid": "AutoScholarQuery_train_20129"} +{"question": "What works proposed G-Openmax for generating unknown samples using generative models?", "answer": ["Generative OpenMax for Multi-Class Open Set Classification"], "answer_arxiv_id": ["1707.07418"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_20130"} +{"question": "Which studies focus on using contrastive learning in LiDAR-based AD scenarios?", "answer": ["ProposalContrast: Unsupervised Pre-training for LiDAR-based 3D Object Detection", "Spatio-temporal Self-Supervised Representation Learning for 3D Point Clouds"], "answer_arxiv_id": ["2207.12654", "2109.00179"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20131"} +{"question": "What papers have focused on generating descriptions of time series in the health domain?", "answer": ["TCube: Domain-Agnostic Neural Time-series Narration"], "answer_arxiv_id": ["2110.05633v1"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_20132"} +{"question": "Which paper offers instance-dependent guarantees for tabular PAC-RL?", "answer": ["Beyond No Regret: Instance-Dependent PAC Reinforcement Learning"], "answer_arxiv_id": ["2108.02717"], "source_meta": {"published_time": "20221009"}, "qid": "AutoScholarQuery_train_20133"} +{"question": "Which papers in the field of offline RL propose to regularize the learned policy to be close to the behavior policy?", "answer": ["Off-Policy Deep Reinforcement Learning without Exploration", "Behavior Regularized Offline Reinforcement Learning", "Keep Doing What Worked: Behavior Modelling Priors for Offline Reinforcement Learning", "A Minimalist Approach to Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning"], "answer_arxiv_id": ["1812.02900", "1911.11361", "2002.08396", "2106.06860", "2110.06169"], "source_meta": {"published_time": "20230721"}, "qid": "AutoScholarQuery_train_20134"} +{"question": "Which research works deal with zero-shot coordination through self-play?", "answer": ["On the Utility of Learning about Humans for Human-AI Coordination"], "answer_arxiv_id": ["1910.05789"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20135"} +{"question": "Which works discuss pre-text tasks in vision-language pre-training?", "answer": ["ViLT: Vision-and-Language Transformer Without Convolution or Region\n Supervision", "Scaling Up Visual and Vision-Language Representation Learning With Noisy\n Text Supervision", "WenLan: Bridging Vision and Language by Large-Scale Multi-Modal\n Pre-Training", "Align before Fuse: Vision and Language Representation Learning with\n Momentum Distillation", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks"], "answer_arxiv_id": ["2102.03334", "2102.05918", "2103.06561", "2107.07651", "1908.02265"], "source_meta": {"published_time": "20230914"}, "qid": "AutoScholarQuery_train_20136"} +{"question": "Any works about the use of diffusion models in video generation?", "answer": ["Structure and Content-Guided Video Synthesis with Diffusion Models", "Video Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data"], "answer_arxiv_id": ["2302.03011", "2204.03458", "2210.02303", "2209.14792"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20137"} +{"question": "What studies have proposed preprocessing defenses against adversarial examples?", "answer": ["Countering Adversarial Images using Input Transformations", "PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examples", "Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples", "On Adaptive Attacks to Adversarial Example Defenses", "Demystifying the Adversarial Robustness of Random Transformation Defenses"], "answer_arxiv_id": ["1711.00117", "1710.10766", "1802.00420", "2002.08347", "2207.03574"], "source_meta": {"published_time": "20221007"}, "qid": "AutoScholarQuery_train_20138"} +{"question": "Which works extended the Rademacher complexity to adversarial settings for linear classifier?", "answer": ["Adversarial risk bounds via function transformation", "Rademacher Complexity for Adversarially Robust Generalization"], "answer_arxiv_id": ["1810.09519", "1810.11914"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_20139"} +{"question": "Which work introduced the foundation of VAE-LSTM-based world models for image-based environments?", "answer": ["World Models"], "answer_arxiv_id": ["1803.10122"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_20140"} +{"question": "What work focused on methods for computing factorizations and their privacy properties?", "answer": ["Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams", "Multi-Epoch Matrix Factorization Mechanisms for Private Machine Learning"], "answer_arxiv_id": ["2202.08312", "2211.06530"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_20141"} +{"question": "Could you tell me about the research which integrates the neural component within the explicit grid in hybrid methods?", "answer": ["Neural Sparse Voxel Fields", "ACORN: Adaptive Coordinate Networks for Neural Scene Representation", "Direct Voxel Grid Optimization: Super-fast Convergence for Radiance\n Fields Reconstruction", "TensoRF: Tensorial Radiance Fields", "F$^{2}$-NeRF: Fast Neural Radiance Field Training with Free Camera\n Trajectories", "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields"], "answer_arxiv_id": ["2007.11571", "2105.02788", "2111.11215", "2203.09517", "2303.15951", "2304.06706"], "source_meta": {"published_time": "20230922"}, "qid": "AutoScholarQuery_train_20142"} +{"question": "What studies developed algorithms with quasi-polynomial sample and computational complexity under observability properties?", "answer": ["Learning in Observable POMDPs, without Computationally Intractable Oracles", "Planning in Observable POMDPs in Quasipolynomial Time"], "answer_arxiv_id": ["2206.03446", "2201.04735"], "source_meta": {"published_time": "20220712"}, "qid": "AutoScholarQuery_train_20143"} +{"question": "Any works about hyperbolic embeddings of the state space of tabular MDPs?", "answer": ["Hyperbolic Embeddings for Learning Options in Hierarchical Reinforcement Learning"], "answer_arxiv_id": ["1812.01487v2"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_20144"} +{"question": "Could you provide me the research that trains an LLM with a hybrid of CoT and PoT rationales?", "answer": ["MAmmoTH: Building Math Generalist Models through Hybrid Instruction\n Tuning"], "answer_arxiv_id": ["2309.05653"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_20145"} +{"question": "What studies have employed individual MLPs to represent a deformation field and a canonical field for capturing complex scene changes over time in the context of Neural Radiance Fields?", "answer": ["D-NeRF: Neural Radiance Fields for Dynamic Scenes", "Nerfies: Deformable Neural Radiance Fields"], "answer_arxiv_id": ["2011.13961", "2011.12948"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_20146"} +{"question": "Could you provide me some papers that have extended the Transformer model to graph tasks?", "answer": ["A Generalization of Transformer Networks to Graphs", "Rethinking Graph Transformers with Spectral Attention", "Your Transformer May Not be as Powerful as You Expect", "Pure Transformers are Powerful Graph Learners", "Recipe for a General, Powerful, Scalable Graph Transformer", "GRPE: Relative Positional Encoding for Graph Transformer", "Global Self-Attention as a Replacement for Graph Convolution", "Rethinking the Expressive Power of GNNs via Graph Biconnectivity"], "answer_arxiv_id": ["2012.09699", "2106.03893", "2205.13401", "2207.02505", "2205.12454", "2201.12787", "2108.03348", "2301.09505"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_20147"} +{"question": "Could you provide me some studies about low-rank approaches for adapting transformers?", "answer": ["LoRA: Low-Rank Adaptation of Large Language Models", "Towards a Unified View of Parameter-Efficient Transfer Learning"], "answer_arxiv_id": ["2106.09685", "2110.04366"], "source_meta": {"published_time": "20240610"}, "qid": "AutoScholarQuery_train_20148"} +{"question": "What studies focus on viewpoint mismatches between domains for domain shift IL with image observation?", "answer": ["Time-Contrastive Networks: Self-Supervised Learning from Video"], "answer_arxiv_id": ["1704.06888"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_20149"} +{"question": "Which research employed a cross-attention structure to attend to visual contexts in the development of multi-modal LLMs?", "answer": ["Flamingo: a Visual Language Model for Few-Shot Learning"], "answer_arxiv_id": ["2204.14198"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_20150"} +{"question": "Which papers were concerned with the method of proxy-guided decoding?", "answer": ["Tuning Language Models by Proxy"], "answer_arxiv_id": ["2401.08565"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_20151"} +{"question": "Are there any researches about allowing for fast communication and gossip rates in decentralized optimization in fixed topology?", "answer": ["An Optimal Algorithm for Decentralized Finite Sum Optimization"], "answer_arxiv_id": ["2005.10675"], "source_meta": {"published_time": "20220726"}, "qid": "AutoScholarQuery_train_20152"} +{"question": "Could you list the studies which examine the optimization of overparameterized neural networks coupling the training dynamics to kernel regression with neural tangent kernel?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Gradient Descent Provably Optimizes Over-parameterized Neural Networks"], "answer_arxiv_id": ["1806.07572", "1811.03962", "1810.02054"], "source_meta": {"published_time": "20230403"}, "qid": "AutoScholarQuery_train_20153"} +{"question": "Could you provide me some works about how discrete and continuous RNNs suffer from vanishing/exploding gradients problem?", "answer": ["Recurrent Neural Networks for Multivariate Time Series with Missing Values", "Phased LSTM: Accelerating Recurrent Network Training for Long or Event-based Sequences", "GRU-ODE-Bayes: Continuous modeling of sporadically-observed time series", "Neural Controlled Differential Equations for Irregular Time Series", "Liquid Time-constant Networks"], "answer_arxiv_id": ["1606.01865", "1610.09513", "1905.12374", "2005.08926", "2006.04439"], "source_meta": {"published_time": "20220926"}, "qid": "AutoScholarQuery_train_20154"} +{"question": "Which works are focused on modeling 3D generation with GAN?", "answer": ["Learning a Probabilistic Latent Space of Object Shapes via 3D\n Generative-Adversarial Modeling", "Learning Representations and Generative Models for 3D Point Clouds", "Point Cloud GAN", "3D Point Cloud Generative Adversarial Network Based on Tree Structured\n Graph Convolutions", "Progressive Point Cloud Deconvolution Generation Network", "Learning Gradient Fields for Shape Generation", "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned\n from Images"], "answer_arxiv_id": ["1610.07584", "1707.02392", "1810.05795", "1905.06292", "2007.05361", "2008.06520", "2209.11163"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_20155"} +{"question": "What papers focus on the one-to-many setting in Referring Expression Segmentation?", "answer": ["Beyond One-to-One: Rethinking the Referring Image Segmentation"], "answer_arxiv_id": ["2308.13853"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_20156"} +{"question": "What studies involved methods which compute future expected disagreement and proactively seek out uncertain areas in the state space?", "answer": ["Planning to Explore via Self-Supervised World Models", "Curious Exploration via Structured World Models Yields Zero-Shot Object Manipulation"], "answer_arxiv_id": ["2005.05960", "2206.11403"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_20157"} +{"question": "What paper showed that shallow neural networks could learn to predict arbitrarily well in the context of near-initialization?", "answer": ["Early-stopped neural networks are consistent"], "answer_arxiv_id": ["2106.05932"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_20158"} +{"question": "What research works have attempted to achieve reductions through sparsifying the attention mechanism in long-range transformers?", "answer": ["Generating Long Sequences with Sparse Transformers", "Reformer: The Efficient Transformer", "Longformer: The Long-Document Transformer", "Efficient Content-Based Sparse Attention with Routing Transformers", "ETC: Encoding Long and Structured Inputs in Transformers", "Big Bird: Transformers for Longer Sequences"], "answer_arxiv_id": ["1904.10509", "2001.04451", "2004.05150", "2003.05997", "2004.08483", "2007.14062"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_20159"} +{"question": "Which works proposed the use of aerial images, LiDAR points and HD panorama in semantic map learning?", "answer": ["HDNET: Exploiting HD Maps for 3D Object Detection", "TorontoCity: Seeing the World with a Million Eyes"], "answer_arxiv_id": ["2012.11704", "1612.00423"], "source_meta": {"published_time": "20220617"}, "qid": "AutoScholarQuery_train_20160"} +{"question": "Could you provide the studies that proposed relaxing invariances by simply averaging data augmentations?", "answer": ["Learning Invariances in Neural Networks", "Last Layer Marginal Likelihood for Invariance Learning", "Learning Invariant Weights in Neural Networks"], "answer_arxiv_id": ["2010.11882", "2106.07512v2", "2202.12439"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_20161"} +{"question": "Which studies utilized Gromov-Wasserstein distance for IL with domain shift?", "answer": ["Cross-Domain Imitation Learning via Optimal Transport"], "answer_arxiv_id": ["2110.03684"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_20162"} +{"question": "What works discuss prior advancements in transfer learning from generic pre-trained models in machine learning?", "answer": ["A Comprehensive Survey on Transfer Learning"], "answer_arxiv_id": ["1911.02685"], "source_meta": {"published_time": "20211216"}, "qid": "AutoScholarQuery_train_20163"} +{"question": "What works propose uncertainty estimation through ensemble learning for Neural Radiance Fields?", "answer": ["Density-aware NeRF Ensembles: Quantifying Predictive Uncertainty in\n Neural Radiance Fields"], "answer_arxiv_id": ["2209.08718"], "source_meta": {"published_time": "20230906"}, "qid": "AutoScholarQuery_train_20164"} +{"question": "Can you point to any studies that have considered the capturing of 3D visual reconstructions of spaces using RGB-D data for multimodal 3D scenes?", "answer": ["ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes", "ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenes"], "answer_arxiv_id": ["1702.04405", "2308.11417"], "source_meta": {"published_time": "20240507"}, "qid": "AutoScholarQuery_train_20165"} +{"question": "Which papers have applied equivariant networks in medical imaging and robotics?", "answer": ["Roto-Translation Equivariant Convolutional Networks: Application to Histopathology Image Analysis", "On-Robot Learning With Equivariant Models", "Sample Efficient Grasp Learning Using Equivariant Models", "Neural Descriptor Fields: \"SE\"⁢(3)-Equivariant Object Representations for Manipulation", "Equivariant Transporter Network", "EDGI: Equivariant Diffusion for Planning with Embodied Agents"], "answer_arxiv_id": ["2002.08725", "2203.04923", "2202.09468", "2112.05124", "2202.09400", "2303.12410"], "source_meta": {"published_time": "20230528"}, "qid": "AutoScholarQuery_train_20166"} +{"question": "What studies proposed the first comprehensive Chinese evaluation package for LLMs?", "answer": ["C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for\n Foundation Models"], "answer_arxiv_id": ["2305.08322"], "source_meta": {"published_time": "20240820"}, "qid": "AutoScholarQuery_train_20167"} +{"question": "Which work incorporates temporal self-attention layers into UNet for video editing?", "answer": ["Tune-A-Video: One-Shot Tuning of Image Diffusion Models for\n Text-to-Video Generation"], "answer_arxiv_id": ["2212.11565"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_20168"} +{"question": "Which papers explore unsupervised object detection in 2D images?", "answer": ["Unsupervised Object Discovery and Localization in the Wild: Part-based\n Matching with Bottom-up Region Proposals", "Localizing Objects with Self-Supervised Transformers and no Labels", "TokenCut: Segmenting Objects in Images and Videos with Self-supervised\n Transformer and Normalized Cut", "Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised\n Semantic Segmentation and Localization", "Unsupervised Salient Object Detection with Spectral Cluster Voting"], "answer_arxiv_id": ["1501.06170", "2109.14279", "2209.00383", "2205.07839", "2203.12614"], "source_meta": {"published_time": "20230511"}, "qid": "AutoScholarQuery_train_20169"} +{"question": "Which papers focus on the research of occupancy prediction and its advantages in 3D scene understanding?", "answer": ["LMSCNet: Lightweight Multiscale 3D Semantic Completion", "MonoScene: Monocular 3D Semantic Scene Completion", "Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction", "Occ3D: A Large-Scale 3D Occupancy Prediction Benchmark for Autonomous\n Driving", "SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Driving", "OccFormer: Dual-path Transformer for Vision-based 3D Semantic Occupancy\n Prediction", "VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene\n Completion", "OccDepth: A Depth-Aware Method for 3D Semantic Scene Completion"], "answer_arxiv_id": ["2008.10559", "2112.00726", "2302.07817", "2304.14365", "2303.09551", "2304.05316", "2302.12251", "2302.13540"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_20170"} +{"question": "What works illustrate how image-to-image tasks are possible by modifying the denoising network?", "answer": ["StyleDrop: Text-to-Image Generation in Any Style", "Palette: Image-to-Image Diffusion Models"], "answer_arxiv_id": ["2306.00983", "2111.05826"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_20171"} +{"question": "Could you provide me some works about hybrid architectures that combine CNNs and ViTs to effectively capture local and global information?", "answer": ["FastViT: A Fast Hybrid Vision Transformer using Structural\n Reparameterization", "EfficientFormer: Vision Transformers at MobileNet Speed", "Rethinking Vision Transformers for MobileNet Size and Speed", "MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer", "Separable Self-attention for Mobile Vision Transformers"], "answer_arxiv_id": ["2303.14189", "2206.01191", "2212.08059", "2110.02178v2", "2206.02680"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_20172"} +{"question": "What works examined certain challenges of diffusion models in image generation, like generating less common concepts or controlling the identity of generated objects?", "answer": ["Re-Imagen: Retrieval-Augmented Text-to-Image Generator", "An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion"], "answer_arxiv_id": ["2209.14491", "2208.01618"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20173"} +{"question": "What work predicted normalising flow distributions over ancestor-conditioned joint rotations?", "answer": ["HuManiFlow: Ancestor-Conditioned Normalising Flows on SO(3) Manifolds\n for Human Pose and Shape Distribution Estimation"], "answer_arxiv_id": ["2305.06968"], "source_meta": {"published_time": "20240330"}, "qid": "AutoScholarQuery_train_20174"} +{"question": "Which works focused on improving the robustness of loss functions to prevent models from overfitting on noisy labels?", "answer": ["Robust Loss Functions under Label Noise for Deep Neural Networks", "Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels", "Symmetric Cross Entropy for Robust Learning with Noisy Labels", "Normalized Loss Functions for Deep Learning with Noisy Labels"], "answer_arxiv_id": ["1712.09482", "1805.07836", "1908.06112", "2006.13554"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_20175"} +{"question": "What papers introduced the coverability coefficient for efficient exploration in online RL?", "answer": ["The Role of Coverage in Online Reinforcement Learning"], "answer_arxiv_id": ["2210.04157"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_20176"} +{"question": "What studies used low-rank tensors to reduce memory usage in Grid-based methods?", "answer": ["TensoRF: Tensorial Radiance Fields", "K-Planes: Explicit Radiance Fields in Space, Time, and Appearance"], "answer_arxiv_id": ["2203.09517", "2301.10241v2"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_20177"} +{"question": "Which articles contributed to the field of behavioral alignment in LLMs?", "answer": ["Scalable agent alignment via reward modeling: a research direction", "Alignment of Language Agents", "Training language models to follow instructions with human feedback", "Enhancing Chat Language Models by Scaling High-quality Instructional Conversations", "Training language models to follow instructions with human feedback", "GPT-4 Technical Report", "GPT-4 Technical Report", "Constitutional AI: Harmlessness from AI Feedback", "UltraFeedback: Boosting Language Models with High-quality Feedback", "Proximal Policy Optimization Algorithms", "Direct Preference Optimization: Your Language Model is Secretly a Reward\n Model", "Let's Verify Step by Step", "Fine-Grained Human Feedback Gives Better Rewards for Language Model\n Training"], "answer_arxiv_id": ["1811.07871", "2103.14659", "2203.02155", "2305.14233v1", "2203.02155", "2303.08774", "2303.08774", "2212.08073", "2310.01377", "1707.06347", "2305.18290", "2305.20050", "2306.01693"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_20178"} +{"question": "What papers collect human judgments on fine-tuned models' decodes to train a reward model that ranks candidates according to human preferences?", "answer": ["Fine-Tuning Language Models from Human Preferences"], "answer_arxiv_id": ["1909.08593"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_20179"} +{"question": "Could you provide me some studies about the Dreamer, a series of latent dynamics models learning via image reconstruction?", "answer": ["Learning Latent Dynamics for Planning from Pixels", "Dream to Control: Learning Behaviors by Latent Imagination", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1811.04551", "1912.01603", "2010.02193"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_20180"} +{"question": "Which work introduced the 'Self-consistency' as a method to suppress the wrong rationales in CoT prompting?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_20181"} +{"question": "Which researches made early approaches with query-support fusion in Few-Shot Segmentation?", "answer": ["CANet: Class-Agnostic Segmentation Networks with Iterative Refinement\n and Attentive Few-Shot Learning", "Prior Guided Feature Enrichment Network for Few-Shot Segmentation"], "answer_arxiv_id": ["1903.02351", "2008.01449"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_20182"} +{"question": "What research has been done on using Vector-Quantized Network for talking face generation?", "answer": ["CodeTalker: Speech-Driven 3D Facial Animation with Discrete Motion Prior"], "answer_arxiv_id": ["2301.02379"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_20183"} +{"question": "Which studies promote pseudo mask labeling using pre-trained VLMs?", "answer": ["Single-Stage Semantic Segmentation from Image Labels", "CLIP is Also an Efficient Segmenter: A Text-Driven Approach for Weakly\n Supervised Semantic Segmentation", "Multi-class Token Transformer for Weakly Supervised Semantic\n Segmentation", "Learning Affinity from Attention: End-to-End Weakly-Supervised Semantic\n Segmentation with Transformers", "Image Segmentation Using Text and Image Prompts", "ReCo: Retrieve and Co-segment for Zero-shot Transfer", "Extract Free Dense Labels from CLIP"], "answer_arxiv_id": ["2005.08104", "2212.09506", "2203.02891", "2203.02664", "2112.10003", "2206.07045", "2112.01071"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_20184"} +{"question": "Which works perform single-task poisoning using arbitrary phrases through data poisoning or directly manipulating the model weights?", "answer": ["Poison Attacks against Text Datasets with Conditional Adversarially Regularized Autoencoder", "You Autocomplete Me: Poisoning Vulnerabilities in Neural Code Completion*", "Be Careful about Poisoned Word Embeddings: Exploring the Vulnerability of the Embedding Layers in NLP Models", "Weight Poisoning Attacks on Pre-trained Models"], "answer_arxiv_id": ["2010.02684", "2007.02220", "2103.15543", "2004.06660"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_20185"} +{"question": "Could you provide some references that focused on the design of effective local modules in the context of 3D representation learning?", "answer": ["PU-GCN: Point Cloud Upsampling using Graph Convolutional Networks", "Dynamic Graph CNN for Learning on Point Clouds", "KPConv: Flexible and Deformable Convolution for Point Clouds", "Tangent Convolutions for Dense Prediction in 3D", "SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional\n Filters", "Relation-Shape Convolutional Neural Network for Point Cloud Analysis", "Point Transformer"], "answer_arxiv_id": ["1912.03264", "1801.07829", "1904.08889", "1807.02443", "1803.11527", "1904.07601", "2012.09164"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_20186"} +{"question": "Could you provide me some studies about the application of video diffusion models in different downstream video synthesis tasks?", "answer": ["Structure and Content-Guided Video Synthesis with Diffusion Models", "Control-A-Video: Controllable Text-to-Video Generation with Diffusion\n Models", "Make-Your-Video: Customized Video Generation Using Textual and\n Structural Guidance", "VideoComposer: Compositional Video Synthesis with Motion Controllability", "Dreamix: Video Diffusion Models are General Video Editors", "MagicEdit: High-Fidelity and Temporally Coherent Video Editing"], "answer_arxiv_id": ["2302.03011", "2305.13840", "2306.00943", "2306.02018", "2302.01329", "2308.14749"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_20187"} +{"question": "Which works focus on out-of-distribution detection based on supervised discriminative models?", "answer": ["Why Normalizing Flows Fail to Detect Out-of-Distribution Data", "Do Deep Generative Models Know What They Don’t Know?", "Likelihood Ratios for Out-of-Distribution Detection", "Input Complexity and Out-of-distribution Detection with Likelihood-based Generative Models", "Likelihood Regret: An Out-of-Distribution Detection Score For Variational Auto-encoder"], "answer_arxiv_id": ["2006.08545", "1810.09136", "1906.02845", "1909.11480", "2003.02977"], "source_meta": {"published_time": "20220308"}, "qid": "AutoScholarQuery_train_20188"} +{"question": "Any works about classifier guidance for generating samples from conditional distributions?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_20189"} +{"question": "Are there any papers that studied the almost-Bayes-optimal solutions and their relation with wide flat landscape regions?", "answer": ["Wide flat minima and optimal generalization in classifying high-dimensional Gaussian mixtures"], "answer_arxiv_id": ["2010.14761"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_20190"} +{"question": "Could you provide the works that introduced the naive estimator and the inverse propensity score estimator?", "answer": ["Unbiased Learning for the Causal Effect of Recommendation"], "answer_arxiv_id": ["2008.04563"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_20191"} +{"question": "What research papers have explored local priors to regularise multi-view geometry estimation problems?", "answer": ["RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from\n Sparse Inputs"], "answer_arxiv_id": ["2112.00724"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_20192"} +{"question": "Which works have been conducted to support Person Re-identification (ReID) task in scenarios with changing environments, perspectives, and poses?", "answer": ["Deep Learning for Person Re-identification: A Survey and Outlook", "Person Re-Identification by Camera Correlation Aware Feature\n Augmentation", "Person Re-identification: Past, Present and Future", "Improving Deep Visual Representation for Person Re-identification by\n Global and Local Image-language Association", "Dual-Path Convolutional Image-Text Embeddings with Instance Loss"], "answer_arxiv_id": ["2001.04193", "1703.08837", "1610.02984", "1808.01571", "1711.05535"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_20193"} +{"question": "Can you list some works which proposed real-world datasets that capture real-world objects under multiple point light sources?", "answer": ["Multi-View Photometric Stereo: A Robust Solution and Benchmark Dataset\n for Spatially Varying Isotropic Materials"], "answer_arxiv_id": ["2001.06659"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_20194"} +{"question": "Which works introduced BERT and its variants in the field of natural language processing?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "ALBERT: A Lite BERT for Self-supervised Learning of Language Representations", "RoBERTa: A Robustly Optimized BERT Pretraining Approach"], "answer_arxiv_id": ["1810.04805", "1909.11942", "1907.11692"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_20195"} +{"question": "What research projects focused on dataset learning for unsupervised optical flow?", "answer": ["AutoFlow: Learning a Better Training Set for Optical Flow", "RealFlow: EM-based Realistic Optical Flow Dataset Generation from Videos", "Self-supervised AutoFlow"], "answer_arxiv_id": ["2104.14544", "2207.11075", "2212.01762"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_20196"} +{"question": "Could you provide me some studies that entail NeRF enhancements for robust training under degraded imaging conditions?", "answer": ["NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw\n Images", "Lighting up NeRF via Unsupervised Decomposition and Enhancement", "Deblur-NeRF: Neural Radiance Fields from Blurry Images"], "answer_arxiv_id": ["2111.13679", "2307.10664", "2111.14292"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_20197"} +{"question": "Which researches proposed methods that leverage geometric insights in bounding box projection to enhance learning?", "answer": ["GPV-Pose: Category-level Object Pose Estimation via Geometry-guided\n Point-wise Voting", "OPA-3D: Occlusion-Aware Pixel-Wise Aggregation for Monocular 3D Object\n Detection"], "answer_arxiv_id": ["2203.07918", "2211.01142"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_20198"} +{"question": "Could you provide me some papers about learning time series representation using non-Transformer-based models?", "answer": ["Unsupervised Scalable Representation Learning for Multivariate Time Series", "Unsupervised Representation Learning for Time Series with Temporal Neighborhood Coding", "Unsupervised Time-Series Representation Learning with Iterative Bilinear Temporal-Spectral Fusion", "TS2Vec: Towards Universal Representation of Time Series"], "answer_arxiv_id": ["1901.10738", "2106.00750", "2202.04770", "2106.10466"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_20199"} +{"question": "What are the traditional methods of Novel View Synthesis (NVS) that rely on a large number of input images?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from\n Multi-View Stereo", "Self-Supervised Visibility Learning for Novel View Synthesis"], "answer_arxiv_id": ["2003.08934", "2103.15595", "2103.15407"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_20200"} +{"question": "What work proposes continual prompt tuning, a parameter-efficient framework that can avoid forgetting and facilitate knowledge transfer between tasks?", "answer": ["Continual Prompt Tuning for Dialog State Tracking"], "answer_arxiv_id": ["2203.06654"], "source_meta": {"published_time": "20240731"}, "qid": "AutoScholarQuery_train_20201"} +{"question": "What studies developed multi-channel neural models for Sign2Text translation?", "answer": ["Multi-channel Transformers for Multi-articulatory Sign Language Translation"], "answer_arxiv_id": ["2009.00299"], "source_meta": {"published_time": "20230502"}, "qid": "AutoScholarQuery_train_20202"} +{"question": "What works consider using subsets of data for hyperparameter tuning of deep learning models?", "answer": ["Grad-Match: Gradient Matching based Data Subset Selection for Efficient Deep Model Training", "Automata : Gradient Based Data Subset Selection for Compute-Efficient Hyper-parameter Tuning"], "answer_arxiv_id": ["2103.00123", "2203.08212"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_20203"} +{"question": "What researches have been conducted on gradient checkpointing to reduce the memory requirement?", "answer": ["Training Deep Nets with Sublinear Memory Cost"], "answer_arxiv_id": ["1604.06174"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_20204"} +{"question": "What study proposed the contact maps by exploiting the consistency between hand contact points and object contact regions?", "answer": ["Hand-Object Contact Consistency Reasoning for Human Grasps Generation"], "answer_arxiv_id": ["2104.03304"], "source_meta": {"published_time": "20220705"}, "qid": "AutoScholarQuery_train_20205"} +{"question": "What research proposed the continual domain adaptation method based on pruning-aided weight modulation to reduce catastrophic forgetting?", "answer": ["Continual Domain Adaptation through Pruning-aided Domain-specific Weight\n Modulation"], "answer_arxiv_id": ["2304.07560"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_20206"} +{"question": "Which research papers discuss extending NeRF to other tasks like generative modelling or dynamic scenes?", "answer": ["pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis", "GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis", "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", "Neural Scene Graphs for Dynamic Scenes"], "answer_arxiv_id": ["2012.00926", "2007.02442", "2011.13084", "2011.10379"], "source_meta": {"published_time": "20230610"}, "qid": "AutoScholarQuery_train_20207"} +{"question": "Which paper considered a version of the interaction screening estimator?", "answer": ["Learning Continuous Exponential Families Beyond Gaussian"], "answer_arxiv_id": ["2102.09198"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_20208"} +{"question": "Which papers regarding using gradient-based bi-level optimization in hyperparameter optimization?", "answer": ["Investigating Bi-Level Optimization for Learning and Vision from a Unified Perspective: A Survey and Beyond", "Gradient-based Bi-level Optimization for Deep Learning: A Survey"], "answer_arxiv_id": ["2101.11517", "2207.11719"], "source_meta": {"published_time": "20230107"}, "qid": "AutoScholarQuery_train_20209"} +{"question": "What works proposed the ODIN score and Mahalanobis distance-based confidence score to improve out-of-distribution uncertainty estimation in the early stages?", "answer": ["Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks"], "answer_arxiv_id": ["1706.02690", "1807.03888"], "source_meta": {"published_time": "20230801"}, "qid": "AutoScholarQuery_train_20210"} +{"question": "Which works focused on utility to the emergent communication approach?", "answer": ["Learning to Communicate with Deep Multi-Agent Reinforcement Learning", "Emergent Multi-Agent Communication in the Deep Learning Era"], "answer_arxiv_id": ["1605.06676", "2006.02419"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_20211"} +{"question": "What papers have a focus on GNN architectures, mainly on Graph Convolutional Networks (GCN) and its variants?", "answer": ["Adversarial Attacks on Neural Networks for Graph Data", "Robustness of Graph Neural Networks at Scale", "Reliable Graph Neural Networks via Robust Aggregation", "Adversarial Attacks on Graph Neural Networks via Meta Learning", "Adversarial Attack on Graph Structured Data", "Spectral Adversarial Training for Robust Graph Neural Network", "Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning"], "answer_arxiv_id": ["1805.07984", "2110.14038v4", "2010.15651", "1902.08412", "1806.02371", "2211.10896", "1801.07606"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_20212"} +{"question": "What works have proposed methods to edit implicit representations for the generation of compositional 3D avatars?", "answer": ["i3DMM: Deep Implicit 3D Morphable Model of Human Heads", "3D-aware Blending with Generative NeRFs", "NeRFFaceEditing: Disentangled Face Editing in Neural Radiance Fields", "SSDNeRF: Semantic Soft Decomposition of Neural Radiance Fields", "Learning Locally Editable Virtual Humans"], "answer_arxiv_id": ["2011.14143", "2302.06608", "2211.07968", "2212.03406", "2305.00121"], "source_meta": {"published_time": "20240216"}, "qid": "AutoScholarQuery_train_20213"} +{"question": "Could you give me an example of a study that extended these algorithms to the federated setting using the concept of pseudo-gradients?", "answer": ["Adaptive Federated Optimization"], "answer_arxiv_id": ["2003.00295"], "source_meta": {"published_time": "20230123"}, "qid": "AutoScholarQuery_train_20214"} +{"question": "What works fine-tuned pre-trained Stable Diffusion model on specific domains to adapt to RCA?", "answer": ["Freestyle Layout-to-Image Synthesis"], "answer_arxiv_id": ["2303.14412"], "source_meta": {"published_time": "20240304"}, "qid": "AutoScholarQuery_train_20215"} +{"question": "Which works discussed data redundancy reduction methods for 3D HPE?", "answer": ["Exploiting Temporal Contexts with Strided Transformer for 3D Human Pose\n Estimation", "P-STMO: Pre-Trained Spatial Temporal Many-to-One Model for 3D Human Pose\n Estimation", "DeciWatch: A Simple Baseline for 10x Efficient 2D and 3D Pose Estimation", "Uplift and Upsample: Efficient 3D Human Pose Estimation with Uplifting\n Transformers"], "answer_arxiv_id": ["2103.14304", "2203.07628", "2203.08713", "2210.06110"], "source_meta": {"published_time": "20231120"}, "qid": "AutoScholarQuery_train_20216"} +{"question": "Could you provide some references developing popular distribution-based models for uncertainty quantification?", "answer": ["Predictive Uncertainty Estimation via Prior Networks", "Evidential Deep Learning to Quantify Classification Uncertainty", "Posterior Network: Uncertainty Estimation without OOD Samples via Density-Based Pseudo-Counts"], "answer_arxiv_id": ["1802.10501", "1806.01768", "2006.09239"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_20217"} +{"question": "What research exists on the topic of neuron-symbolic computation?", "answer": ["Neural-Symbolic Computing: An Effective Methodology for Principled Integration of Machine Learning and Reasoning"], "answer_arxiv_id": ["1905.06088"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_20218"} +{"question": "Can you provide papers that support the development of a direct approach in Speech-to-Speech Translation?", "answer": ["Leveraging Weakly Supervised Data to Improve End-to-End Speech-to-Text\n Translation", "Direct speech-to-speech translation with discrete units", "Enhanced Direct Speech-to-Speech Translation Using Self-supervised\n Pre-training and Data Augmentation", "TranSpeech: Speech-to-Speech Translation With Bilateral Perturbation", "Many-to-Many Spoken Language Translation via Unified Speech and Text\n Representation Learning with Unit-to-Unit Translation"], "answer_arxiv_id": ["1811.02050", "2107.05604", "2204.02967", "2205.12523", "2308.01831"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_20219"} +{"question": "Which works conduct the similar comparisons of FMPE against NPE?", "answer": ["Flexible statistical inference for mechanistic models of neural dynamics", "Automatic Posterior Transformation for Likelihood-free Inference"], "answer_arxiv_id": ["1711.01861", "1905.07488"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_20220"} +{"question": "Which works utilize the SDF to achieve the implicit reconstruction of 3D objects at the instance level?", "answer": ["DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation"], "answer_arxiv_id": ["1901.05103"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_20221"} +{"question": "Are there any studies that pre-train a classifier for each source domain and then adversarially map the target images into each trained feature space?", "answer": ["Multi-source Distilling Domain Adaptation"], "answer_arxiv_id": ["1911.11554"], "source_meta": {"published_time": "20220930"}, "qid": "AutoScholarQuery_train_20222"} +{"question": "What researches explore negative sampling within a large mini-batch in contrastive instance discrimination?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Learning Transferable Visual Models From Natural Language Supervision", "SimCSE: Simple Contrastive Learning of Sentence Embeddings"], "answer_arxiv_id": ["2002.05709", "2103.00020", "2104.08821"], "source_meta": {"published_time": "20220716"}, "qid": "AutoScholarQuery_train_20223"} +{"question": "Which research documents have mentioned about the hardware innovation on N:M sparse tensor core?", "answer": ["An Algorithm-Hardware Co-Optimized Framework for Accelerating N:M Sparse Transformers"], "answer_arxiv_id": ["2208.06118"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_20224"} +{"question": "Can you tell me about the research that proposed the Beta Embedding as a method that supports a comprehensive set of operations in FOL?", "answer": ["Beta Embeddings for Multi-Hop Logical Reasoning in Knowledge Graphs"], "answer_arxiv_id": ["2010.11465"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20225"} +{"question": "What papers proposed the frequently studied attack methods such as FGSM, PGD and C&W Attack?", "answer": ["Towards Deep Learning Models Resistant to Adversarial Attacks", "Towards Evaluating the Robustness of Neural Networks"], "answer_arxiv_id": ["1706.06083", "1608.04644"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_20226"} +{"question": "Which works used large scale text-image datasets to learn how to predict images from text inputs?", "answer": ["Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation"], "answer_arxiv_id": ["2205.11487", "2206.10789"], "source_meta": {"published_time": "20230102"}, "qid": "AutoScholarQuery_train_20227"} +{"question": "Are there any works that propose a differentiable raycasting method to forecast 2D occupancy states?", "answer": ["Point Cloud Forecasting as a Proxy for 4D Occupancy Forecasting"], "answer_arxiv_id": ["2302.13130"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_20228"} +{"question": "Which papers proposed fine-tuning techniques for personalized Text-to-Image synthesis?", "answer": ["DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation"], "answer_arxiv_id": ["2208.12242"], "source_meta": {"published_time": "20230519"}, "qid": "AutoScholarQuery_train_20229"} +{"question": "Which work developed a text encoder compatible with a pre-trained vision encoder in the context of Vision-Language Models?", "answer": ["LiT: Zero-Shot Transfer with Locked-image text Tuning"], "answer_arxiv_id": ["2111.07991"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20230"} +{"question": "What are the related works that propose the generation of the protein backbone and of the protein sequence given the backbone structure separately?", "answer": ["Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem", "Generating Novel, Designable, and Diverse Protein Structures by Equivariantly Diffusing Oriented Residue Clouds", "Protein structure generation via folding diffusion"], "answer_arxiv_id": ["2206.04119", "2301.12485", "2209.15611"], "source_meta": {"published_time": "20230728"}, "qid": "AutoScholarQuery_train_20231"} +{"question": "Are there any works reporting on the application of the prompting technique for fine-tuning pre-trained ViT models to downstream tasks?", "answer": ["Visual Prompt Tuning"], "answer_arxiv_id": ["2203.12119"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_20232"} +{"question": "Which papers proved asymptotic convergence guarantees of popular distributional RL algorithms like C51 and QR-DQN?", "answer": ["An Analysis of Categorical Distributional Reinforcement Learning", "An Analysis of Quantile Temporal-Difference Learning"], "answer_arxiv_id": ["1802.08163v1", "2301.04462"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_20233"} +{"question": "What works have enabled the application of more powerful GNN models in the field of molecular chemistry?", "answer": ["How Powerful are Graph Neural Networks?", "Principal Neighbourhood Aggregation for Graph Nets", "Weisfeiler and Lehman Go Cellular: CW Networks", "DeeperGCN: All You Need to Train Deeper GCNs", "Directional Graph Networks", "Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting"], "answer_arxiv_id": ["1810.00826", "2004.05718", "2106.12575", "2006.07739v1", "2010.02863", "2006.09252"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_20234"} +{"question": "Could you provide me research that directly aggregate 2D features with MLPs or transformers?", "answer": ["IBRNet: Learning Multi-View Image-Based Rendering", "Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction", "Unsupervised Learning of 3D Object Categories from Videos in the Wild", "pixelNeRF: Neural Radiance Fields from One or Few Images", "ContraNeRF: Generalizable Neural Radiance Fields for Synthetic-to-real Novel View Synthesis via Contrastive Learning", "Neural Rays for Occlusion-aware Image-based Rendering", "ViewFormer: NeRF-free Neural Rendering from Few Images Using Transformers", "GRF: Learning a General Radiance Field for 3D Representation and Rendering", "Is Attention All That NeRF Needs?"], "answer_arxiv_id": ["2102.13090", "2109.00512", "2103.16552", "2012.02190", "2303.11052", "2107.13421", "2203.10157", "2010.04595", "2207.13298"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_20235"} +{"question": "Which papers showcased theoretical expressivity by injecting positional encodings in graph neural networks?", "answer": ["Identity-aware Graph Neural Networks"], "answer_arxiv_id": ["2101.10320"], "source_meta": {"published_time": "20231028"}, "qid": "AutoScholarQuery_train_20236"} +{"question": "What studies discussed graph-based methods for clustering, particularly the use of GNNs for feature extraction?", "answer": ["Graph Contrastive Learning with Adaptive Augmentation", "Graph Communal Contrastive Learning", "X-GOAL: Multiplex Heterogeneous Graph Prototypical Contrastive Learning"], "answer_arxiv_id": ["2010.14945", "2110.14863", "2109.03560"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_20237"} +{"question": "What works made use of Adapt-ML-Prod to control the generalization error?", "answer": ["A Second-order Bound with Excess Losses"], "answer_arxiv_id": ["1402.2044"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_20238"} +{"question": "Can you name the papers which propose a graph convolutional network to predict the heat map for optimal solution of a TSP instance?", "answer": ["An Efficient Graph Convolutional Network Technique for the Travelling Salesman Problem"], "answer_arxiv_id": ["1906.01227"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_20239"} +{"question": "What studies have attempted at improving the efficiency of spherical CNNs?", "answer": ["Spin-Weighted Spherical CNNs", "Efficient Generalized Spherical CNNs", "Scattering Networks on the Sphere for Scalable and Rotationally Equivariant Spherical CNNs"], "answer_arxiv_id": ["2006.10731", "2010.11661", "2102.02828"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_20240"} +{"question": "What works have proposed rendering novel views using pseudo geometry?", "answer": ["SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image"], "answer_arxiv_id": ["2204.00928"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_20241"} +{"question": "What studies propose using explainability techniques to find spurious attributes?", "answer": ["Finding and Fixing Spurious Patterns with Explanations", "Identifying Spurious Correlations and Correcting them with an Explanation-based Learning", "Meaningfully Debugging Model Mistakes using Conceptual Counterfactual Explanations"], "answer_arxiv_id": ["2106.02112", "2211.08285", "2106.12723"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_20242"} +{"question": "Which studies utilized learned correction to add a residual term to the output of a numerical step?", "answer": ["Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers"], "answer_arxiv_id": ["2007.00016"], "source_meta": {"published_time": "20211127"}, "qid": "AutoScholarQuery_train_20243"} +{"question": "Are there any works that discuss 'zero-concentrated differential privacy' (zCDP)?", "answer": ["Concentrated Differential Privacy: Simplifications, Extensions, and Lower Bounds", "Concentrated Differential Privacy"], "answer_arxiv_id": ["1605.02065", "1603.01887"], "source_meta": {"published_time": "20220310"}, "qid": "AutoScholarQuery_train_20244"} +{"question": "Which works proposed methods in continuum of learning for data selection?", "answer": ["iCaRL: Incremental Classifier and Representation Learning", "An Empirical Study of Example Forgetting during Deep Neural Network Learning", "End-to-End Incremental Learning", "Gradient based sample selection for online continual learning"], "answer_arxiv_id": ["1611.07725", "1812.05159", "1807.09536", "1903.08671"], "source_meta": {"published_time": "20220519"}, "qid": "AutoScholarQuery_train_20245"} +{"question": "What are the existing works that provide certification methods offering sound guarantees against fooling a model's prediction?", "answer": ["Feature-Guided Black-Box Safety Testing of Deep Neural Networks", "A Game-Based Approximate Verification of Deep Neural Networks with Provable Guarantees"], "answer_arxiv_id": ["1710.07859", "1807.03571"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_20246"} +{"question": "Which works are about class-agnostic motion prediction methods for enhancing system redundancy?", "answer": ["MotionNet: Joint Perception and Motion Prediction for Autonomous Driving\n Based on Bird's Eye View Maps", "Long-Term Occupancy Grid Prediction Using Recurrent Neural Networks", "Convolutional LSTM Network: A Machine Learning Approach for\n Precipitation Nowcasting"], "answer_arxiv_id": ["2003.06754", "1809.03782", "1506.04214"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_20247"} +{"question": "Which works have extended the FL methods for heterogeneous clients to aim for personalization?", "answer": ["Personalized Federated Learning with Gaussian Processes", "Self-Aware Personalized Federated Learning", "Personalized Federated Learning with Moreau Envelopes", "Towards Personalized Federated Learning", "Personalized Federated Learning: A Meta-Learning Approach", "Adaptive Personalized Federated Learning", "Efficient Split-Mix Federated Learning for On-Demand and In-Situ Customization"], "answer_arxiv_id": ["2106.15482", "2204.08069v1", "2006.08848", "2103.00710", "2002.07948", "2003.13461", "2203.09747"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_20248"} +{"question": "Could you mention works where progress in LLMs inspired the development of versatile and generalist agents?", "answer": ["A Generalist Agent"], "answer_arxiv_id": ["2205.06175"], "source_meta": {"published_time": "20221006"}, "qid": "AutoScholarQuery_train_20249"} +{"question": "Are there any studies on corruption robustness in video classification?", "answer": ["Benchmarking the Robustness of Spatial-Temporal Models Against Corruptions", "Robustness Guarantees for Deep Neural Networks on Videos"], "answer_arxiv_id": ["2110.06513", "1907.00098"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_20250"} +{"question": "What papers made innovations in the framework’s training and sampling techniques of diffusion models?", "answer": ["Denoising Diffusion Implicit Models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Classifier-Free Diffusion Guidance", "Cascaded Diffusion Models for High Fidelity Image Generation"], "answer_arxiv_id": ["2010.02502", "2206.00364v2", "2207.12598", "2106.15282"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_20251"} +{"question": "Any related works on the use of domain randomization to improve generalization and sim-to-real transfer?", "answer": ["Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World", "Asymmetric Actor Critic for Image-Based Robot Learning"], "answer_arxiv_id": ["1703.06907", "1710.06542"], "source_meta": {"published_time": "20221212"}, "qid": "AutoScholarQuery_train_20252"} +{"question": "Which paper introduces a denoising training method to speed up DETR training?", "answer": ["DN-DETR: Accelerate DETR Training by Introducing Query DeNoising"], "answer_arxiv_id": ["2203.01305"], "source_meta": {"published_time": "20220307"}, "qid": "AutoScholarQuery_train_20253"} +{"question": "What works developed stochastic algorithms for DAM, inspired by the min-max objective?", "answer": ["Stochastic AUC Maximization with Deep Neural Networks", "Large-scale Robust Deep AUC Maximization: A New Surrogate Loss and Empirical Studies on Medical Image Classification"], "answer_arxiv_id": ["1908.10831", "2012.03173"], "source_meta": {"published_time": "20230514"}, "qid": "AutoScholarQuery_train_20254"} +{"question": "Can you name the study that proposed a two-branch graph transformer network to address the text from syntax and logic?", "answer": ["Logiformer: A Two-Branch Graph Transformer Network for Interpretable\n Logical Reasoning"], "answer_arxiv_id": ["2205.00731"], "source_meta": {"published_time": "20240529"}, "qid": "AutoScholarQuery_train_20255"} +{"question": "What work discusses the asymptotic normality of various kernelized variants of score matching?", "answer": ["Minimum Stein Discrepancy Estimators"], "answer_arxiv_id": ["1906.08283v3"], "source_meta": {"published_time": "20230603"}, "qid": "AutoScholarQuery_train_20256"} +{"question": "Which works learned prototypical shapes, rather than sticking to a predefined family of parametric primitives?", "answer": ["Learning elementary structures for 3D shape generation and matching", "Representing Shape Collections with Alignment-Aware Linear Models"], "answer_arxiv_id": ["1908.04725", "2109.01605v2"], "source_meta": {"published_time": "20230419"}, "qid": "AutoScholarQuery_train_20257"} +{"question": "Any works about proposing communication-efficient algorithms for finite-sum structure and variance reduction in centralized settings and (strongly) monotone s in decentralized settings?", "answer": ["Distributed Methods with Compressed Communication for Solving Variational Inequalities, with Theoretical Guarantees", "Optimal Algorithms for Decentralized Stochastic Variational Inequalities"], "answer_arxiv_id": ["2110.03313v3", "2202.02771"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_20258"} +{"question": "What studies talk about node injections or deletions in GNN?", "answer": ["Robustness of Graph Neural Networks at Scale"], "answer_arxiv_id": ["2110.14038"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_20259"} +{"question": "Can you suggest a paper that showed the stabilizability of 𝑊-games under λ-bounded distributions in an aim to relax the requirements of PAC-stabilizability?", "answer": ["PAC learning and stabilizing Hedonic Games: towards a unifying approach."], "answer_arxiv_id": ["2301.13756"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_20260"} +{"question": "Which study replaces the isotropic Gaussian prior with a distribution of mean and covariance computed on the data?", "answer": ["PriorGrad: Improving Conditional Denoising Diffusion Models with Data-Dependent Adaptive Prior"], "answer_arxiv_id": ["2106.06406"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20261"} +{"question": "What research measures the transferability of adversarial robustness between data distributions using conditional Wasserstein distance?", "answer": ["Robust Learning Meets Generative Models: Can Proxy Distributions Improve Adversarial Robustness?"], "answer_arxiv_id": ["2104.09425"], "source_meta": {"published_time": "20221018"}, "qid": "AutoScholarQuery_train_20262"} +{"question": "Which works discuss the encoding of image-text pairs into a joint feature space for learning the semantic alignment between vision and language?", "answer": ["Behind the Scene: Revealing the Secrets of Pre-trained Vision-and-Language Models", "Learning Relation Alignment for Calibrated Cross-modal Retrieval"], "answer_arxiv_id": ["2005.07310", "2105.13868"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_20263"} +{"question": "Which works have researched multi-camera collaboration and formulated it as a multi-agent learning problem?", "answer": ["Pose-Assisted Multi-Camera Collaboration for Active Object Tracking", "Learning Multi-Agent Coordination for Enhancing Target Coverage in Directional Sensor Networks", "Coordinate-Aligned Multi-Camera Collaboration for Active Multi-Object Tracking"], "answer_arxiv_id": ["2001.05161", "2010.13110", "2202.10881"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_20264"} +{"question": "What works have presented studies on the multiple forms of the BYOL predictor network?", "answer": ["VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction"], "answer_arxiv_id": ["2105.04906", "2103.03230"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20265"} +{"question": "Could you provide me some papers about encoding the geometry of point clouds via neural networks and supervised learning?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric\n Space", "Relation-Shape Convolutional Neural Network for Point Cloud Analysis", "Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs", "Dynamic Graph CNN for Learning on Point Clouds", "Point Transformer", "Fast Point Transformer", "Point Transformer V2: Grouped Vector Attention and Partition-based\n Pooling"], "answer_arxiv_id": ["1612.00593", "1706.02413", "1904.07601", "1711.09869", "1801.07829", "2012.09164", "2112.04702", "2210.05666"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20266"} +{"question": "Which researches proposed the use of group representation theory for rotation-equivariant maps?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural\n networks for 3D point clouds", "SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks"], "answer_arxiv_id": ["1802.08219", "2006.10503"], "source_meta": {"published_time": "20221126"}, "qid": "AutoScholarQuery_train_20267"} +{"question": "Can you tell me about research that has proposed frameworks for conceptualizing CSKBs?", "answer": ["CAT: A Contextualized Conceptualization and Instantiation Framework for\n Commonsense Reasoning"], "answer_arxiv_id": ["2305.04808"], "source_meta": {"published_time": "20240114"}, "qid": "AutoScholarQuery_train_20268"} +{"question": "Which studies used pseudocode to improve the performance of NLP tasks?", "answer": ["Prompting with Pseudo-Code Instructions"], "answer_arxiv_id": ["2305.11790"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_20269"} +{"question": "Which studies have made assumptions about the full state-action space coverage of offline datasets in their theoretical frameworks?", "answer": ["Batch Policy Learning under Constraints", "Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction"], "answer_arxiv_id": ["1903.08738", "1906.00949"], "source_meta": {"published_time": "20220523"}, "qid": "AutoScholarQuery_train_20270"} +{"question": "Which reference suggests that Nash-V can output a nearly-minimax policy through weighted averaging?", "answer": ["V-Learning—A Simple, Efficient, Decentralized Algorithm for Multiagent RL"], "answer_arxiv_id": ["2110.14555"], "source_meta": {"published_time": "20230305"}, "qid": "AutoScholarQuery_train_20271"} +{"question": "Could you provide me some works about applying similar strategies as orthogonal learning in Federated Learning (FL)?", "answer": ["Federated Orthogonal Training: Mitigating Global Catastrophic Forgetting\n in Continual Federated Learning", "GradMA: A Gradient-Memory-based Accelerated Federated Learning with\n Alleviated Catastrophic Forgetting"], "answer_arxiv_id": ["2309.01289", "2302.14307"], "source_meta": {"published_time": "20230824"}, "qid": "AutoScholarQuery_train_20272"} +{"question": "Who showed that the probing outcomes using various prompts can be contradictory or unreliable?", "answer": ["Measuring and Improving Consistency in Pretrained Language Models"], "answer_arxiv_id": ["2102.01017"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_20273"} +{"question": "What are some studies that discuss neural implicit representation in Neural Implicit SLAM?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "DIST: Rendering Deep Implicit Signed Distance Function with\n Differentiable Sphere Tracing", "UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction", "Volume Rendering of Neural Implicit Surfaces", "DeepSDF: Learning Continuous Signed Distance Functions for Shape\n Representation"], "answer_arxiv_id": ["2003.08934", "1911.13225", "2104.10078", "2106.12052", "1901.05103"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_20274"} +{"question": "Can you provide some studies about Diffusion-based inverse model (DIM)?", "answer": ["Solving Inverse Problems in Medical Imaging with Score-Based Generative Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Pretraining is All You Need for Image-to-Image Translation", "Improving Diffusion Models for Inverse Problems using Manifold Constraints", "Denoising Diffusion Restoration Models", "Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model", "JPEG Artifact Correction using Denoising Diffusion Restoration Models", "Diffusion Posterior Sampling for General Noisy Inverse Problems"], "answer_arxiv_id": ["2111.08005", "2204.06125", "2205.12952", "2206.00941", "2201.11793", "2212.00490", "2209.11888", "2209.14687"], "source_meta": {"published_time": "20230212"}, "qid": "AutoScholarQuery_train_20275"} +{"question": "Could you provide me some works about direct speech-to-speech translation approach in machine translation?", "answer": ["Textless Speech-to-Speech Translation on Real Data"], "answer_arxiv_id": ["2112.08352"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_20276"} +{"question": "Could you mention some works that used the Lagrangian-based method to solve the CMDP?", "answer": ["Risk-Constrained Reinforcement Learning with Percentile Risk Criteria", "Constrained Policy Optimization via Bayesian World Models"], "answer_arxiv_id": ["1512.01629", "2201.09802"], "source_meta": {"published_time": "20220529"}, "qid": "AutoScholarQuery_train_20277"} +{"question": "Can you tell me about the works that have used diffusion models for their stability at scale?", "answer": ["Hierarchical Text-Conditional Image Generation with CLIP Latents", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2204.06125", "2112.10741", "2205.11487", "2112.10752"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_20278"} +{"question": "What are the works that have explored ways to extend the capabilities of language models to comprehend visual content?", "answer": ["VL-BERT: Pre-training of Generic Visual-Linguistic Representations", "Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks", "ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for\n Vision-and-Language Tasks"], "answer_arxiv_id": ["1908.08530", "2004.06165", "1908.02265"], "source_meta": {"published_time": "20240112"}, "qid": "AutoScholarQuery_train_20279"} +{"question": "Can you provide the work that introduces structured packing for long-context data?", "answer": ["Structured Packing in LLM Training Improves Long Context Utilization"], "answer_arxiv_id": ["2312.17296"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_20280"} +{"question": "What works have explored the architecture of diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "High-Resolution Image Synthesis with Latent Diffusion Models", "Denoising Diffusion Implicit Models", "Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["2006.11239", "2112.10752", "2010.02502", "1907.05600"], "source_meta": {"published_time": "20230802"}, "qid": "AutoScholarQuery_train_20281"} +{"question": "Which studies focus on improving the synthesis quality of text-to-image generation through denoising diffusion probabilistic models?", "answer": ["Elucidating the Design Space of Diffusion-Based Generative Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Denoising Diffusion Probabilistic Models", "Score-Based Generative Modeling through Stochastic Differential Equations", "High-Resolution Image Synthesis with Latent Diffusion Models"], "answer_arxiv_id": ["2206.00364", "1503.03585", "1907.05600", "2006.11239", "2011.13456", "2112.10752"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20282"} +{"question": "Which paper is about SceneRF, an extension of PixelNeRF, that incorporates a probabilistic ray sampling strategy, Spherical U-Net, and depth penalties?", "answer": ["SceneRF: Self-Supervised Monocular 3D Scene Reconstruction with Radiance\n Fields"], "answer_arxiv_id": ["2212.02501"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_20283"} +{"question": "Could you provide some studies that discuss improvements on multi-step reasoning tasks?", "answer": ["Chain-of-Thought Prompting Elicits Reasoning in Large Language Models", "Large Language Model Guided Tree-of-Thought", "Tree of Thoughts: Deliberate Problem Solving with Large Language Models", "Algorithm of Thoughts: Enhancing Exploration of Ideas in Large Language\n Models"], "answer_arxiv_id": ["2201.11903", "2305.08291", "2305.10601", "2308.10379"], "source_meta": {"published_time": "20240521"}, "qid": "AutoScholarQuery_train_20284"} +{"question": "Which works have introduced novel modules for depth estimation like fully-connected CRFs and an Internal Discretization module?", "answer": ["iDisc: Internal Discretization for Monocular Depth Estimation"], "answer_arxiv_id": ["2304.06334"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_20285"} +{"question": "Which work first introduced the integration of semantics into NeRF?", "answer": ["In-Place Scene Labelling and Understanding with Implicit Scene Representation"], "answer_arxiv_id": ["2103.15875"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_20286"} +{"question": "Which paper proposed a combination of LSTD updates for the linear weights on top of DQN representations?", "answer": ["Shallow Updates for Deep Reinforcement Learning"], "answer_arxiv_id": ["1705.07461"], "source_meta": {"published_time": "20230529"}, "qid": "AutoScholarQuery_train_20287"} +{"question": "Which papers discussed learning with verifications as a strategy to optimize the mathematical proficiencies of LLMs?", "answer": ["Training Verifiers to Solve Math Word Problems", "Making Large Language Models Better Reasoners with Step-Aware Verifier", "Let's Verify Step by Step", "Learning Math Reasoning from Self-Sampled Correct and Partially-Correct\n Solutions"], "answer_arxiv_id": ["2110.14168", "2206.02336", "2305.20050", "2205.14318"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_20288"} +{"question": "Could you provide me works where model agnostic meta-learning has been effectively used to study computer vision tasks?", "answer": ["MetaPix: Few-Shot Video Retargeting", "Self-Supervised Video Representation Learning with Meta-Contrastive\n Network", "Video Deblurring by Fitting to Test Data", "Tracking by Instance Detection: A Meta-Learning Approach", "One to Many: Adaptive Instrument Segmentation via Meta Learning and\n Dynamic Online Adaptation in Robotic Surgical Video"], "answer_arxiv_id": ["1910.04742", "2108.08426", "2012.05228", "2004.00830", "2103.12988"], "source_meta": {"published_time": "20240306"}, "qid": "AutoScholarQuery_train_20289"} +{"question": "What studies are related to node/edge perturbation techniques in the augmentation phase of the Graph Contrastive Learning framework?", "answer": ["Graph Contrastive Learning with Augmentations", "An Empirical Study of Graph Contrastive Learning"], "answer_arxiv_id": ["2010.13902", "2109.01116"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_20290"} +{"question": "Which papers look at filtering out of domain data by training an out of domain classifier?", "answer": ["Multi-Task Curriculum Framework for Open-Set Semi-Supervised Learning", "Trash to Treasure: Harvesting OOD Data with Cross-Modal Matching for Open-Set Semi-Supervised Learning", "OpenMatch: Open-set Consistency Regularization for Semi-supervised Learning with Outliers", "RETRIEVE: Coreset Selection for Efficient and Robust Semi-Supervised Learning", "RealMix: Towards Realistic Semi-Supervised Deep Learning Algorithms", "Out-distribution aware Self-training in an Open World Setting", "OpenCoS: Contrastive Semi-supervised Learning for Handling Open-set Unlabeled Data"], "answer_arxiv_id": ["2007.11330", "2108.05617", "2105.14148", "2106.07760v2", "1912.08766", "2012.12372", "2107.08943"], "source_meta": {"published_time": "20230228"}, "qid": "AutoScholarQuery_train_20291"} +{"question": "Which papers describe neural networks augmenting numerical solvers by learning data-driven discretizations for PDEs?", "answer": ["Learning data driven discretizations for partial differential equations", "Machine learning accelerated computational fluid dynamics"], "answer_arxiv_id": ["1808.04930", "2102.01010"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_20292"} +{"question": "Would you suggest works that introduced techniques for data handling such as class-balanced resampling or loss reweighting?", "answer": ["Federated Visual Classification with Real-World Data Distribution", "Addressing Class Imbalance in Federated Learning"], "answer_arxiv_id": ["2003.08082", "2008.06217"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_20293"} +{"question": "Could you provide some works that investigate the potential of extracting harmful behavior in safety-trained language models?", "answer": ["Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned", "Exploiting Programmatic Behavior of LLMs: Dual-Use Through Standard Security Attacks", "Multi-step Jailbreaking Privacy Attacks on ChatGPT", "Fundamental limitations of alignment in Large Language Models", "On the Impossible Safety of Large AI Models"], "answer_arxiv_id": ["2209.07858", "2302.05733", "2304.05197", "2304.11082", "2209.15259"], "source_meta": {"published_time": "20230705"}, "qid": "AutoScholarQuery_train_20294"} +{"question": "Could you provide me with some studies based on region proposal ranking for referring image detection?", "answer": ["Learning to Compose and Reason with Language Tree Structures for Visual\n Grounding", "Modeling Relationships in Referential Expressions with Compositional\n Modular Networks", "Grounding Referring Expressions in Images by Variational Context", "Parallel Attention: A Unified Framework for Visual Object Discovery\n through Dialogs and Queries"], "answer_arxiv_id": ["1906.01784", "1611.09978", "1712.01892", "1711.06370"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_20295"} +{"question": "What works extended CLIP to pixel-level prediction for open vocabulary semantic segmentation?", "answer": ["CAT-Seg : Cost Aggregation for Open-Vocabulary Semantic Segmentation", "Image Segmentation Using Text and Image Prompts"], "answer_arxiv_id": ["2303.11797", "2112.10003"], "source_meta": {"published_time": "20231008"}, "qid": "AutoScholarQuery_train_20296"} +{"question": "Which papers have researched on online unsupervised reinforcement learning?", "answer": ["URLB: Unsupervised Reinforcement Learning Benchmark"], "answer_arxiv_id": ["2110.15191"], "source_meta": {"published_time": "20231109"}, "qid": "AutoScholarQuery_train_20297"} +{"question": "Can you list some studies that incorporated video inputs for 3D pose estimation?", "answer": ["VoxelTrack: Multi-Person 3D Human Pose Estimation and Tracking in the\n Wild", "4D Association Graph for Realtime Multi-person Motion Capture Using\n Multiple Video Cameras"], "answer_arxiv_id": ["2108.02452", "2002.12625"], "source_meta": {"published_time": "20231118"}, "qid": "AutoScholarQuery_train_20298"} +{"question": "Can you mention some studies related to the ensemble methods in imbalanced learning?", "answer": ["ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot", "Long-tailed Recognition by Routing Diverse Distribution-Aware Experts", "BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed Visual Recognition"], "answer_arxiv_id": ["2108.02385", "2010.01809", "1912.02413"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_20299"} +{"question": "Which research explore decomposing images into a set of composable energy functions representing different concepts?", "answer": ["Unsupervised Compositional Concepts Discovery with Text-to-Image\n Generative Models"], "answer_arxiv_id": ["2306.05357"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_20300"} +{"question": "Are there any studies that proposed test-time adaptation methods to alleviate the domain shift issue by online updating the pre-trained model on the test data?", "answer": ["Test-Time Training with Self-Supervision for Generalization under Distribution Shifts", "Tent: Fully Test-Time Adaptation by Entropy Minimization", "MEMO: Test Time Robustness via Adaptation and Augmentation", "Efficient Test-Time Model Adaptation without Forgetting", "Continual Test-Time Domain Adaptation", "Contrastive Test-Time Adaptation", "Towards Stable Test-time Adaptation in Dynamic Wild World", "Self-Supervised Aggregation of Diverse Experts for Test-Agnostic Long-Tailed Recognition"], "answer_arxiv_id": ["1909.13231", "2006.10726", "2110.09506", "2204.02610", "2203.13591", "2204.10377", "2302.12400", "2107.09249"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_20301"} +{"question": "Could you provide some studies that introduced continuous fitted value iteration?", "answer": ["Value Iteration in Continuous Actions, States and Time"], "answer_arxiv_id": ["2105.04682"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20302"} +{"question": "What study introduced the first weakly-supervised COD framework?", "answer": ["Weakly-Supervised Camouflaged Object Detection with Scribble Annotations"], "answer_arxiv_id": ["2207.14083"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_20303"} +{"question": "What research derives the volume rendering equation under the assumption of piecewise constant opacity and color?", "answer": ["Local Light Field Fusion: Practical View Synthesis with Prescriptive Sampling Guidelines"], "answer_arxiv_id": ["1905.00889"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_20304"} +{"question": "Can you give examples of studies using magnetic markers per hand and optical markers per object in human-human handover research?", "answer": ["H2O: A Benchmark for Visual Human-human Object Handover Analysis"], "answer_arxiv_id": ["2104.11466"], "source_meta": {"published_time": "20231001"}, "qid": "AutoScholarQuery_train_20305"} +{"question": "What research papers touch on the application of Graph Neural Networks in various fields like computer vision and natural language processing?", "answer": ["Few-Shot Learning with Graph Neural Networks", "Bidirectional Graph Reasoning Network for Panoptic Segmentation", "Contrastive Language-Image Pre-Training with Knowledge Graphs", "Encoding Sentences with Graph Convolutional Networks for Semantic Role Labeling"], "answer_arxiv_id": ["1711.04043", "2004.06272", "2210.08901", "1703.04826"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_20306"} +{"question": "Could you give me examples of studies where MPMAB assumes that players do not get any reward when a collision occurs?", "answer": ["Multi-Player Bandits – a Musical Chairs Approach", "Multi-Player Bandits Revisited", "SIC - MMAB: Synchronisation Involves Communication in Multiplayer Multi-Armed Bandits", "Selfish Robustness and Equilibria in Multi-Player Bandits", "Decentralized Multi-player Multi-armed Bandits with No Collision Information", "Heterogeneous Multi-player Multi-armed Bandits: Closing the Gap and Generalization", "Towards Optimal Algorithms for Multi-Player Bandits without Collision Sensing Information", "Multiplayer bandits without observing collision information"], "answer_arxiv_id": ["1512.02866", "1711.02317v3", "1809.08151", "2002.01197", "2003.00162v1", "2110.14622v2", "2103.13059", "1808.08416v2"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20307"} +{"question": "What works proposed the method of using a permutation-based method to approximate SII and STI?", "answer": ["The Shapley Taylor Interaction Index", "Faith-Shap: The Faithful Shapley Interaction Index"], "answer_arxiv_id": ["1902.05622", "2203.00870"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_20308"} +{"question": "Any earlier works on self-attention?", "answer": ["Long Short-Term Memory-Networks for Machine Reading", "A Decomposable Attention Model for Natural Language Inference", "A Deep Reinforced Model for Abstractive Summarization", "A Structured Self-attentive Sentence Embedding"], "answer_arxiv_id": ["1601.06733", "1606.01933", "1705.04304", "1703.03130"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_20309"} +{"question": "Which references discuss autoregressive methods for tackling human motion composition?", "answer": ["Perpetual Motion: Generating Unbounded Human Motion", "TEACH: Temporal Action Composition for 3D Humans", "MultiAct: Long-Term 3D Human Motion Generation from Multiple Action\n Labels", "The Wanderings of Odysseus in 3D Scenes", "Perpetual Humanoid Control for Real-time Simulated Avatars", "Synthesizing Long-Term Human Motions with Diffusion Models via Coherent\n Sampling"], "answer_arxiv_id": ["2007.13886", "2209.04066", "2212.05897", "2112.09251", "2305.06456", "2308.01850"], "source_meta": {"published_time": "20240223"}, "qid": "AutoScholarQuery_train_20310"} +{"question": "Which paper presented the concept of label-only attack in Membership Inference Attack studies?", "answer": ["Label-Only Membership Inference Attacks"], "answer_arxiv_id": ["2007.14321"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_20311"} +{"question": "Can you provide a research paper which conducted theoretical study of convergence properties of 𝚒𝙻𝚀𝚁, 𝚒𝙻𝚀𝙶, and 𝙳𝙿𝙿?", "answer": ["Iterative Linearized Control: Stable Algorithms and Complexity Guarantees"], "answer_arxiv_id": ["1908.07615"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_20312"} +{"question": "Are there any studies about area under different metrics like AUPRC and AUTKC?", "answer": ["Stochastic Optimization of Areas Under Precision-Recall Curves with Provable Convergence", "Exploring the Algorithm-Dependent Generalization of AUPRC Optimization with List Stability", "Optimizing Partial Area Under the Top-k Curve: Theory and Practice", "OpenAUC: Towards AUC-Oriented Open-Set Recognition"], "answer_arxiv_id": ["2104.08736", "2209.13262", "2209.01398", "2210.13458"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_20313"} +{"question": "Which works utilized a hybrid training approach on both image and video datasets?", "answer": ["Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval", "Swap Attention in Spatiotemporal Diffusions for Text-to-Video Generation", "Advancing High-Resolution Video-Language Representation with Large-Scale\n Video Transcriptions"], "answer_arxiv_id": ["2104.00650", "2305.10874", "2111.10337"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_20314"} +{"question": "What studies discuss about Bias, a method where only the bias term is updated?", "answer": ["BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models"], "answer_arxiv_id": ["2106.10199"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_20315"} +{"question": "Could you provide examples of datasets proposed for anomaly detection after KolektorSDD?", "answer": ["VT-ADL: A Vision Transformer Network for Image Anomaly Detection and\n Localization"], "answer_arxiv_id": ["2104.10036"], "source_meta": {"published_time": "20240319"}, "qid": "AutoScholarQuery_train_20316"} +{"question": "Which papers present numerical conjugation using computational methods?", "answer": ["A fast approach to optimal transport: The back-and-forth method"], "answer_arxiv_id": ["1905.12154"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_20317"} +{"question": "Which works explore enhancing instruction-following abilities during the fine-tuning step of language models?", "answer": ["Multitask Prompted Training Enables Zero-Shot Task Generalization", "Finetuned Language Models Are Zero-Shot Learners", "How Far Can Camels Go? Exploring the State of Instruction Tuning on Open\n Resources"], "answer_arxiv_id": ["2110.08207", "2109.01652", "2306.04751"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_20318"} +{"question": "Can you provide studies that introduced simulated multi-agent perception datasets?", "answer": ["V2X-Sim: Multi-Agent Collaborative Perception Dataset and Benchmark for\n Autonomous Driving", "OPV2V: An Open Benchmark Dataset and Fusion Pipeline for Perception with\n Vehicle-to-Vehicle Communication", "V2X-ViT: Vehicle-to-Everything Cooperative Perception with Vision\n Transformer"], "answer_arxiv_id": ["2202.08449", "2109.07644", "2203.10638"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_20319"} +{"question": "What studies have equipped neural models with specialized modules for certain operations?", "answer": ["Training Verifiers to Solve Math Word Problems", "Just Add Functions: A Neural-Symbolic Language Model"], "answer_arxiv_id": ["2110.14168", "1912.05421"], "source_meta": {"published_time": "20221118"}, "qid": "AutoScholarQuery_train_20320"} +{"question": "What studies build 3D shapes conditioned on partial shapes?", "answer": ["ComplementMe: Weakly-Supervised Component Suggestions for 3D Modeling", "PQ-NET: A Generative Part Seq2Seq Network for 3D Shapes"], "answer_arxiv_id": ["1708.01841", "1911.10949"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_20321"} +{"question": "What works discussed non-rigid 3D surface registration approaches?", "answer": ["Quasi-Newton Solver for Robust Non-Rigid Registration", "Robust Non-Rigid Registration with Reweighted Position and\n Transformation Sparsity"], "answer_arxiv_id": ["2004.04322", "1703.04861"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_20322"} +{"question": "Could you provide some works comparing GBDTs to NNs on tabular data?", "answer": ["Well-tuned Simple Nets Excel on Tabular Datasets", "Revisiting Deep Learning Models for Tabular Data", "TabNet: Attentive Interpretable Tabular Learning", "Tabular Data: Deep Learning is Not All You Need", "Deep Neural Networks and Tabular Data: A Survey"], "answer_arxiv_id": ["2106.11189", "2106.11959", "1908.07442", "2106.03253", "2110.01889"], "source_meta": {"published_time": "20230504"}, "qid": "AutoScholarQuery_train_20323"} +{"question": "What research enabled acceleration of Vision Transformer without model modification and used methods like pruning or merging input tokens?", "answer": ["AdaViT: Adaptive Vision Transformers for Efficient Image Recognition", "Learned Token Pruning for Transformers", "Token Merging: Your ViT But Faster", "Token Merging for Fast Stable Diffusion", "Less is More: Pay Less Attention in Vision Transformers"], "answer_arxiv_id": ["2111.15668", "2107.00910", "2210.09461", "2303.17604", "2105.14217"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_20324"} +{"question": "Could you provide me some studies that involve generating text conditioned on other modalities for cross-modal text generation?", "answer": ["BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "HRVDA: High-Resolution Visual Document Assistant"], "answer_arxiv_id": ["2201.12086", "2404.06918v1"], "source_meta": {"published_time": "20240618"}, "qid": "AutoScholarQuery_train_20325"} +{"question": "What research used diffusion models to imitate human behavior and parameterize policy in offline RL?", "answer": ["Imitating Human Behaviour with Diffusion Models", "Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning"], "answer_arxiv_id": ["2301.10677", "2208.06193"], "source_meta": {"published_time": "20221128"}, "qid": "AutoScholarQuery_train_20326"} +{"question": "Which studies present a unified formulation for class and content disentanglement in generative models?", "answer": ["Demystifying Inter-Class Disentanglement"], "answer_arxiv_id": ["1906.11796"], "source_meta": {"published_time": "20220322"}, "qid": "AutoScholarQuery_train_20327"} +{"question": "What application uses SAM’s segmentation results as prior information to assist downstream tasks?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks", "Inpaint Anything: Segment Anything Meets Image Inpainting", "Edit Everything: A Text-Guided Generative System for Images Editing", "Matte Anything: Interactive Natural Image Matting with Segment Anything\n Models", "Personalize Segment Anything Model with One Shot", "Segment Anything Model (SAM) Enhanced Pseudo Labels for Weakly\n Supervised Semantic Segmentation", "Token Contrast for Weakly-Supervised Semantic Segmentation", "SSA: Semantic Structure Aware Inference for Weakly Pixel-Wise Dense\n Predictions without Cost", "Deep learning universal crater detection using Segment Anything Model\n (SAM)"], "answer_arxiv_id": ["2401.14159", "2304.06790", "2304.14006", "2306.04121", "2305.03048", "2305.05803", "2303.01267", "2111.03392", "2304.07764"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_20328"} +{"question": "Could you provide me research that proposed an unsupervised learning framework called EGN for general CO problems?", "answer": ["Erdos Goes Neural: an Unsupervised Learning Framework for Combinatorial Optimization on Graphs"], "answer_arxiv_id": ["2006.10643v4"], "source_meta": {"published_time": "20230108"}, "qid": "AutoScholarQuery_train_20329"} +{"question": "Which works provide large-scale datasets that contributed to advancements in vision foundation models?", "answer": ["LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs", "LAION-5B: An open large-scale dataset for training next generation\n image-text models"], "answer_arxiv_id": ["2111.02114", "2210.08402"], "source_meta": {"published_time": "20240601"}, "qid": "AutoScholarQuery_train_20330"} +{"question": "Which studies demonstrated the solution for 2t0s games can be reduced to solving a 2p0s game through intra-team policy announcements?", "answer": ["A Marriage between Adversarial Team Games and 2-player Games: Enabling Abstractions, No-regret Learning, and Subgame Solving"], "answer_arxiv_id": ["2206.09161"], "source_meta": {"published_time": "20230122"}, "qid": "AutoScholarQuery_train_20331"} +{"question": "Which works first tackled novel category discovery (NCD) as a transfer learning problem?", "answer": ["Learning to Discover Novel Visual Categories via Deep Transfer Clustering", "Learning to cluster in order to transfer across domains and tasks", "Multi-class Classification without Multi-class Labels"], "answer_arxiv_id": ["1908.09884", "1711.10125", "1901.00544"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_20332"} +{"question": "Which papers applied denoising score matching (DSM) to protein structure prediction and protein-ligand docking?", "answer": ["Generative Modeling by Estimating Gradients of the Data Distribution", "EBM-Fold: Fully-Differentiable Protein Folding Powered by Energy-based Models", "DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking"], "answer_arxiv_id": ["1907.05600", "2105.04771", "2210.01776"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_20333"} +{"question": "What research works have been made in open-vocabulary semantic segmentation (OVSS)?", "answer": ["Decoupling Zero-Shot Semantic Segmentation", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels", "A Simple Baseline for Open-Vocabulary Semantic Segmentation with\n Pre-trained Vision-language Model", "FreeSeg: Unified, Universal and Open-Vocabulary Image Segmentation"], "answer_arxiv_id": ["2112.07910", "2210.04150", "2112.12143", "2112.14757", "2303.17225"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_20334"} +{"question": "What research papers proposed the use of image-based reconstruction systems like NeuS?", "answer": ["NeuS: Learning Neural Implicit Surfaces by Volume Rendering for\n Multi-view Reconstruction", "SyncDreamer: Generating Multiview-consistent Images from a Single-view\n Image"], "answer_arxiv_id": ["2106.10689", "2309.03453"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_20335"} +{"question": "In which papers did researchers adopt point discrimination as a pretext task for 3D self-supervised representation learning?", "answer": ["PointContrast: Unsupervised Pre-training for 3D Point Cloud\n Understanding", "Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene\n Contexts", "Masked Scene Contrast: A Scalable Framework for Unsupervised 3D\n Representation Learning"], "answer_arxiv_id": ["2007.10985", "2012.09165", "2303.14191"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_20336"} +{"question": "What are some studies that worked on contextual machine translation?", "answer": ["Does Neural Machine Translation Benefit from Larger Context?", "Modeling Coherence for Discourse Neural Machine Translation", "Fill in the Blanks: Imputing Missing Sentences for Larger-Context Neural Machine Translation"], "answer_arxiv_id": ["1704.05135", "1811.05683", "1910.14075"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_20337"} +{"question": "Do you have any examples of instruction-following models that have been successful to due recent datasets?", "answer": ["LLaMA: Open and Efficient Foundation Language Models"], "answer_arxiv_id": ["2302.13971"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_20338"} +{"question": "What are some relevant works on data summarization?", "answer": ["Randomized Rounding for the Largest Simplex Problem", "Twice-Ramanujan Sparsifiers"], "answer_arxiv_id": ["1412.0036", "0808.0163"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_20339"} +{"question": "Can you mention some works that use neural representations for surface completion?", "answer": ["Occupancy Networks: Learning 3D Reconstruction in Function Space", "DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation"], "answer_arxiv_id": ["1812.03828", "1901.05103"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_20340"} +{"question": "Can you name the studies which introduced an IoU-aware classification score to improve the rank of detection results?", "answer": ["VarifocalNet: An IoU-aware Dense Object Detector", "Detection Transformer with Stable Matching", "Align-DETR: Improving DETR with Simple IoU-aware BCE loss"], "answer_arxiv_id": ["2008.13367", "2304.04742", "2304.07527"], "source_meta": {"published_time": "20231013"}, "qid": "AutoScholarQuery_train_20341"} +{"question": "Any research that utilizes mixup augmentations to enhance contrastive learning or improving masked image modeling?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Hard Negative Mixing for Contrastive Learning", "i-Mix: A Domain-Agnostic Strategy for Contrastive Representation Learning", "Un-Mix: Rethinking Image Mixtures for Unsupervised Visual Representation Learning", "Boosting Discriminative Visual Representation Learning with Scenario-Agnostic Mixup", "Architecture-Agnostic Masked Image Modeling – From ViT back to CNN", "Mixed Autoencoder for Self-supervised Visual Representation Learning"], "answer_arxiv_id": ["2002.05709", "2010.01028v2", "2010.08887", "2003.05438", "2111.15454", "2205.13943", "2303.17152"], "source_meta": {"published_time": "20220321"}, "qid": "AutoScholarQuery_train_20342"} +{"question": "Could you provide me some references about parametric human mesh recovery methods that use statistical models?", "answer": ["End-to-end Recovery of Human Shape and Pose", "Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the\n Loop", "Delving Deep Into Hybrid Annotations for 3D Human Recovery in the Wild", "Synthetic Training for Accurate 3D Human Pose and Shape Estimation in the Wild", "CLIFF: Carrying Location Information in Full Frames into Human Pose and\n Shape Estimation", "Implicit 3D Human Mesh Recovery using Consistency with Pose and Shape\n from Unseen-view"], "answer_arxiv_id": ["1712.06584", "1909.12828", "1908.06442", "2009.10013v2", "2208.00571", "2306.17651"], "source_meta": {"published_time": "20240530"}, "qid": "AutoScholarQuery_train_20343"} +{"question": "What papers focus on learning mappings from the problem parameters to the solution using methods like DeepONet and neural operator?", "answer": ["DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators", "Fourier Neural Operator for Parametric Partial Differential Equations"], "answer_arxiv_id": ["1910.03193", "2010.08895"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_20344"} +{"question": "What are the studies implementing Optimal Transport in domain adaptation?", "answer": ["DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised\n Domain Adaptation", "Unbalanced minibatch Optimal Transport; applications to Domain\n Adaptation", "Unified Optimal Transport Framework for Universal Domain Adaptation", "Unsupervised Cross-Domain Image Retrieval via Prototypical Optimal\n Transport"], "answer_arxiv_id": ["1803.10081", "2103.03606", "2210.17067", "2402.18411"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20345"} +{"question": "Could you provide me some papers about employing adapter modules in NLP for transferring Transformer-based pre-trained models?", "answer": ["Parameter-Efficient Transfer Learning for NLP", "AdapterFusion: Non-Destructive Task Composition for Transfer Learning", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["1902.00751", "2005.00247", "2106.09685"], "source_meta": {"published_time": "20240212"}, "qid": "AutoScholarQuery_train_20346"} +{"question": "Any research papers that performed learning a sequence-level adversarial-ranking reward?", "answer": ["Adversarial Ranking for Language Generation"], "answer_arxiv_id": ["1705.11001"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20347"} +{"question": "What works initiated the study of generalization bounds related to the information between the output and input of the learner?", "answer": ["Information-theoretic analysis of generalization capability of learning algorithms", "Learners that Use Little Information", "How much does your data exploration overfit? Controlling bias via information usage."], "answer_arxiv_id": ["1705.07809", "1710.05233", "1511.05219"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20348"} +{"question": "Which works traditionally depend on extracting content from target images and matching style via second-order statistics in the domain of style transfer?", "answer": ["ArtFlow: Unbiased Image Style Transfer via Reversible Neural Flows", "AdaAttN: Revisit Attention Mechanism in Arbitrary Neural Style Transfer"], "answer_arxiv_id": ["2103.16877", "2108.03647"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20349"} +{"question": "Where has the application of DPMs in image synthesis been discussed?", "answer": ["Diffusion Models Beat GANs on Image Synthesis"], "answer_arxiv_id": ["2105.05233"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20350"} +{"question": "Which papers utilized in-context prompting for LLMs?", "answer": ["Language Models are Few-Shot Learners", "Calibrate Before Use: Improving Few-Shot Performance of Language Models", "Improving In-Context Few-Shot Learning via Self-Supervised Training"], "answer_arxiv_id": ["2005.14165", "2102.09690", "2205.01703"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_20351"} +{"question": "Could you provide me some works using data selection techniques on vision models?", "answer": ["Semantic Redundancies in Image-Classification Datasets: The 10% You Don’t Need", "Training Data Subset Search with Ensemble Active Learning", "Accelerating Deep Learning by Focusing on the Biggest Losers", "Trivial or impossible—dichotomous data difficulty masks model differences (on ImageNet and beyond)", "Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt", "Deep Learning on a Data Diet: Finding Important Examples Early in Training", "An Empirical Study of Example Forgetting during Deep Neural Network Learning"], "answer_arxiv_id": ["1901.11409", "1905.12737", "1910.00762v1", "2110.05922", "2206.07137", "2107.07075", "1812.05159"], "source_meta": {"published_time": "20230823"}, "qid": "AutoScholarQuery_train_20352"} +{"question": "Which studies examined the neural tangent kernel approximation that linearizes the network function with regard to its parameters?", "answer": ["Neural Tangent Kernel: Convergence and Generalization in Neural Networks"], "answer_arxiv_id": ["1806.07572"], "source_meta": {"published_time": "20221130"}, "qid": "AutoScholarQuery_train_20353"} +{"question": "Can you refer to the works that leverage neural radiance field (NeRF) for 3D modeling?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis", "Nerfstudio: A Modular Framework for Neural Radiance Field Development", "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding", "D-NeRF: Neural Radiance Fields for Dynamic Scenes", "PlenOctrees for Real-time Rendering of Neural Radiance Fields", "NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections", "Neural 3D Reconstruction in the Wild"], "answer_arxiv_id": ["2003.08934", "2302.04264", "2201.05989", "2011.13961", "2103.14024", "2008.02268", "2205.12955"], "source_meta": {"published_time": "20240201"}, "qid": "AutoScholarQuery_train_20354"} +{"question": "Which work introduced the CLIP model for embedding images and text in shared representation space?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_20355"} +{"question": "What papers found that applying locality-enhanced attention mechanism can increase the training speed for instance and panoptic segmentation?", "answer": ["Masked-attention Mask Transformer for Universal Image Segmentation", "Deformable DETR: Deformable Transformers for End-to-End Object Detection", "Conditional DETR for Fast Training Convergence"], "answer_arxiv_id": ["2112.01527", "2010.04159", "2108.06152"], "source_meta": {"published_time": "20230629"}, "qid": "AutoScholarQuery_train_20356"} +{"question": "What researches applied representation interpolation methods to NLP tasks?", "answer": ["MixKD: Towards Efficient Distillation of Large-scale Language Models", "MixText: Linguistically-Informed Interpolation of Hidden Space for Semi-Supervised Text Classification"], "answer_arxiv_id": ["2011.00593", "2004.12239"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_20357"} +{"question": "Could you name some research papers where non-adversarial training methods are utilized in Data-Free Knowledge Distillation?", "answer": ["Large-Scale Generative Data-Free Distillation", "Data-Free Knowledge Distillation with Soft Targeted Transfer Set Synthesis"], "answer_arxiv_id": ["2012.05578", "2104.04868"], "source_meta": {"published_time": "20230924"}, "qid": "AutoScholarQuery_train_20358"} +{"question": "Which papers present models that implicitly learn valuable representations from large-scale image-text pairs?", "answer": ["High-Resolution Image Synthesis with Latent Diffusion Models", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"], "answer_arxiv_id": ["2112.10752", "2204.06125", "2205.11487"], "source_meta": {"published_time": "20230811"}, "qid": "AutoScholarQuery_train_20359"} +{"question": "Could you provide me some studies about Neural LP?", "answer": ["Differentiable Learning of Logical Rules for Knowledge Base Reasoning"], "answer_arxiv_id": ["1702.08367"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_20360"} +{"question": "What paper computed the mean and variance of the model’s confidence for the target label throughout training to identify interesting examples in the context of natural language processing?", "answer": ["Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamics"], "answer_arxiv_id": ["2009.10795"], "source_meta": {"published_time": "20220920"}, "qid": "AutoScholarQuery_train_20361"} +{"question": "What are the key works on differentially private empirical risk minimization (DP-ERM)?", "answer": ["Differentially Private Empirical Risk Minimization", "Differential Privacy Dynamics of Langevin Diffusion and Noisy Gradient Descent"], "answer_arxiv_id": ["0912.0071v5", "2102.05855v5"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_20362"} +{"question": "Which studies combine diffusion models with 3D priors for novel view synthesis?", "answer": ["Sparse3D: Distilling Multiview-Consistent Diffusion for Object\n Reconstruction from Sparse Views", "SparseFusion: Distilling View-conditioned Diffusion for 3D\n Reconstruction", "ViewFormer: NeRF-free Neural Rendering from Few Images Using\n Transformers", "Novel View Synthesis with Diffusion Models", "Generative Novel View Synthesis with 3D-Aware Diffusion Models", "NerfDiff: Single-image View Synthesis with NeRF-guided Distillation from\n 3D-aware Diffusion"], "answer_arxiv_id": ["2308.14078", "2212.00792", "2203.10157", "2210.04628", "2304.02602", "2302.10109"], "source_meta": {"published_time": "20231206"}, "qid": "AutoScholarQuery_train_20363"} +{"question": "What are some studies about few-shot learning in relation to NeRF?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis", "LOLNeRF: Learn from One Look", "SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image", "NerfingMVS: Guided Optimization of Neural Radiance Fields for Indoor Multi-view Stereo", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo"], "answer_arxiv_id": ["2012.02190", "2104.00677", "2111.09996", "2204.00928", "2109.01129", "2103.15595"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_20364"} +{"question": "Which works have recently proposed stationarity measures involving value function or its regularized variant?", "answer": ["BOME! Bilevel Optimization Made Easy: A Simple First-Order Approach"], "answer_arxiv_id": ["2209.08709"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_20365"} +{"question": "What research introduced the idea of using diffusion models for pure behavior cloning, without requiring Q-value maximizing?", "answer": ["Know Your Boundaries: The Necessity of Explicit Behavioral Cloning in Offline RL", "Imitating Human Behaviour with Diffusion Models"], "answer_arxiv_id": ["2206.00695", "2301.10677"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_20366"} +{"question": "What work proposed a unified vision transformer compression technique?", "answer": ["Unified Visual Transformer Compression"], "answer_arxiv_id": ["2203.08243"], "source_meta": {"published_time": "20240323"}, "qid": "AutoScholarQuery_train_20367"} +{"question": "What works utilize mesh-based methods for human rendering?", "answer": ["Textured Neural Avatars", "Real-time Deep Dynamic Characters", "Dressing Avatars: Deep Photorealistic Appearance for Physically\n Simulated Clothing", "Driving-Signal Aware Full-Body Avatars"], "answer_arxiv_id": ["1905.08776", "2105.01794", "2206.15470", "2105.10441"], "source_meta": {"published_time": "20231210"}, "qid": "AutoScholarQuery_train_20368"} +{"question": "Which work proposed a Transformer-based method to enhance representation discriminability for both point clouds and textual queries?", "answer": ["Text to Point Cloud Localization with Relation-Enhanced Transformer"], "answer_arxiv_id": ["2301.05372"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_20369"} +{"question": "Could you provide me some works regarding the usage of LLMs to evaluate responses that align with specific personas?", "answer": ["Faithful Persona-based Conversational Dataset Generation with Large\n Language Models"], "answer_arxiv_id": ["2312.10007"], "source_meta": {"published_time": "20240227"}, "qid": "AutoScholarQuery_train_20370"} +{"question": "What work generalizes to ImageNet?", "answer": ["3D generation on ImageNet"], "answer_arxiv_id": ["2303.01416"], "source_meta": {"published_time": "20230707"}, "qid": "AutoScholarQuery_train_20371"} +{"question": "Can you mention studies that proposed learning schedules for masks?", "answer": ["HRank: Filter Pruning using High-Rank Feature Map", "Soft Masking for Cost-Constrained Channel Pruning", "CHEX: CHannel EXploration for CNN Model Compression", "Rigging the Lottery: Making All Tickets Winners"], "answer_arxiv_id": ["2002.10179", "2211.02206", "2203.15794", "1911.11134"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_20372"} +{"question": "Which papers discuss curiosity-driven approaches in online pretraining for RL?", "answer": ["Exploration by Random Network Distillation", "Self-Supervised Exploration via Disagreement", "Planning to Explore via Self-Supervised World Models", "Curiosity-driven Exploration by Self-supervised Prediction"], "answer_arxiv_id": ["1810.12894", "1906.04161", "2005.05960", "1705.05363"], "source_meta": {"published_time": "20231031"}, "qid": "AutoScholarQuery_train_20373"} +{"question": "What are the studies that applied adapters with incremental learning in the computer vision field?", "answer": ["Incremental Learning Through Deep Adaptation"], "answer_arxiv_id": ["1705.04228"], "source_meta": {"published_time": "20240513"}, "qid": "AutoScholarQuery_train_20374"} +{"question": "Which research applied differentiable SDP relaxations to solve Sudoku puzzles and Rubik’s cube?", "answer": ["SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver", "Learning Symmetric Rules with SATNet"], "answer_arxiv_id": ["1905.12149", "2206.13998"], "source_meta": {"published_time": "20231003"}, "qid": "AutoScholarQuery_train_20375"} +{"question": "Which works adopted SPL for weakly supervised object detection?", "answer": ["Self Paced Deep Learning for Weakly Supervised Object Detection", "Bridging Saliency Detection to Weakly Supervised Object Detection Based on Self-paced Curriculum Learning"], "answer_arxiv_id": ["1605.07651", "1703.01290v1"], "source_meta": {"published_time": "20230310"}, "qid": "AutoScholarQuery_train_20376"} +{"question": "Which papers discuss anomaly detection approaches relying on autoencoders?", "answer": ["Improving Unsupervised Defect Segmentation by Applying Structural\n Similarity to Autoencoders"], "answer_arxiv_id": ["1807.02011"], "source_meta": {"published_time": "20231125"}, "qid": "AutoScholarQuery_train_20377"} +{"question": "Which research discussed industry benchmarking system in the machine learning industry?", "answer": ["MLPerf Training Benchmark", "MLPerf Inference Benchmark", "TBD: Benchmarking and Analyzing Deep Neural Network Training", "AIBench Training: Balanced Industry-Standard AI Training Benchmarking"], "answer_arxiv_id": ["1910.01500v3", "1911.02549", "1803.06905", "2004.14690"], "source_meta": {"published_time": "20220720"}, "qid": "AutoScholarQuery_train_20378"} +{"question": "What studies focus on estimating 2D positions of human joints from images or videos?", "answer": ["Cascaded Pyramid Network for Multi-Person Pose Estimation", "Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields", "Deep High-Resolution Representation Learning for Visual Recognition", "Poseur: Direct Human Pose Regression with Transformers"], "answer_arxiv_id": ["1711.07319", "1611.08050", "1908.07919", "2201.07412"], "source_meta": {"published_time": "20240508"}, "qid": "AutoScholarQuery_train_20379"} +{"question": "Which works discuss the shift in volume density representation from occupancy to Signed Distance Fields (SDF)?", "answer": ["Vox-Fusion: Dense Tracking and Mapping with Voxel-based Neural Implicit\n Representation", "ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of\n Signed Distance Fields", "Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural\n Real-Time SLAM", "NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM"], "answer_arxiv_id": ["2210.15858", "2211.11704", "2304.14377", "2302.03594"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_20380"} +{"question": "Which studies work on discrete diffusion models?", "answer": ["Argmax Flows and Multinomial Diffusion: Learning Categorical Distributions", "Structured Denoising Diffusion Models in Discrete State-Spaces", "ImageBART: Bidirectional Context with Multinomial Diffusion for Autoregressive Image Synthesis", "Global Context with Discrete Diffusion in Vector Quantised Modelling for Image Generation", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Improved Vector Quantized Diffusion Models"], "answer_arxiv_id": ["2102.05379", "2107.03006", "2108.08827", "2112.01799", "2111.14822", "2205.16007"], "source_meta": {"published_time": "20221127"}, "qid": "AutoScholarQuery_train_20381"} +{"question": "What studies have continued the adoption of GNNs as a general purpose vision backbone?", "answer": ["PVG: Progressive Vision Graph for Vision Recognition", "MobileViG: Graph-Based Sparse Attention for Mobile Vision Applications"], "answer_arxiv_id": ["2308.00574", "2307.00395"], "source_meta": {"published_time": "20240510"}, "qid": "AutoScholarQuery_train_20382"} +{"question": "What are the studies that focused on clients’ incentives to participate in FL training process?", "answer": ["Incentive Mechanisms for Federated Learning: From Economic and Game Theoretic Perspective", "Coopetition Against an Amazon"], "answer_arxiv_id": ["2111.11850", "2005.10038"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_20383"} +{"question": "Could you provide me some researches proposing partial differential operators to maintain equivariance?", "answer": ["PDO-eConvs: Partial Differential Operator Based Equivariant Convolutions", "Steerable Partial Differential Operators for Equivariant Neural Networks"], "answer_arxiv_id": ["2007.10408", "2106.10163"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_20384"} +{"question": "What papers propose methods that deal with overconfident softmax scores when predicting OOD inputs?", "answer": ["Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images", "Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problem"], "answer_arxiv_id": ["1412.1897", "1812.05720"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_20385"} +{"question": "Which works mark the paradigm shift with the introduction of deep learning methodologies in 2D-3D lifting?", "answer": ["C3DPO: Canonical 3D Pose Networks for Non-Rigid Structure From Motion", "Deep Non-Rigid Structure from Motion", "Unsupervised 3D Pose Estimation with Non-Rigid Structure-from-Motion\n Modeling"], "answer_arxiv_id": ["1909.02533", "1908.00052", "2308.10705"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_20386"} +{"question": "Which references could I use to understand more about controllable diffusion models?", "answer": ["Adding Conditional Control to Text-to-Image Diffusion Models", "Composer: Creative and Controllable Image Synthesis with Composable\n Conditions", "Palette: Image-to-Image Diffusion Models", "An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "SinFusion: Training Diffusion Models on a Single Image or Video"], "answer_arxiv_id": ["2302.05543", "2302.09778", "2111.05826", "2208.01618", "2208.12242", "2211.11743"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_20387"} +{"question": "What works have been done to address the scalability issues for inductive representation learning over graph structured data?", "answer": ["Simplifying Graph Convolutional Networks"], "answer_arxiv_id": ["1902.07153"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_20388"} +{"question": "Which studies highlight the importance of equality, apart from fairness, in Federated Learning?", "answer": ["Tilted Empirical Risk Minimization", "Ditto: Fair and Robust Federated Learning Through Personalization", "Fair Resource Allocation in Federated Learning", "Agnostic Federated Learning"], "answer_arxiv_id": ["2007.01162", "2012.04221", "1905.10497", "1902.00146"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_20389"} +{"question": "What are some one-stage methods for EHPS?", "answer": ["One-Stage 3D Whole-Body Mesh Recovery with Component Aware Transformer", "SMPLer-X: Scaling Up Expressive Human Pose and Shape Estimation"], "answer_arxiv_id": ["2303.16160", "2309.17448"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_20390"} +{"question": "What works deal with the LLM-centric planning with feedback from different sources such as environment, human, and model?", "answer": ["Voyager: An Open-Ended Embodied Agent with Large Language Models", "Reflexion: Language Agents with Verbal Reinforcement Learning", "Describe, Explain, Plan and Select: Interactive Planning with Large\n Language Models Enables Open-World Multi-Task Agents", "Leveraging Pre-trained Large Language Models to Construct and Utilize World Models for Model-based Task Planning", "Language Models can Solve Computer Tasks"], "answer_arxiv_id": ["2305.16291", "2303.11366", "2302.01560", "2305.14909v2", "2303.17491"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_20391"} +{"question": "What are some works on unified models for video segmentation tasks?", "answer": ["Mask2Former for Video Instance Segmentation", "Video K-Net: A Simple, Strong, and Unified Baseline for Video\n Segmentation", "Video-kMaX: A Simple Unified Approach for Online and Near-Online Video\n Panoptic Segmentation", "DVIS: Decoupled Video Instance Segmentation Framework", "Universal Instance Perception as Object Discovery and Retrieval", "TarViS: A Unified Approach for Target-based Video Segmentation", "Tracking Anything with Decoupled Video Segmentation"], "answer_arxiv_id": ["2112.10764", "2204.04656", "2304.04694", "2306.03413", "2303.06674", "2301.02657", "2309.03903"], "source_meta": {"published_time": "20240228"}, "qid": "AutoScholarQuery_train_20392"} +{"question": "Could you provide me some studies about joint modeling of motion and language based on datasets having both motion and language modalities?", "answer": ["BABEL: Bodies, Action and Behavior with English Labels", "AMASS: Archive of Motion Capture as Surface Shapes"], "answer_arxiv_id": ["2106.09696", "1904.03278"], "source_meta": {"published_time": "20230515"}, "qid": "AutoScholarQuery_train_20393"} +{"question": "Which research works are about the development of density and coverage metrics?", "answer": ["Reliable Fidelity and Diversity Metrics for Generative Models"], "answer_arxiv_id": ["2002.09797"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20394"} +{"question": "What works have proposed diffusion models for predicting a grid-based representation in 3D shapes?", "answer": ["DiffRF: Rendering-Guided 3D Radiance Field Diffusion", "Diffusion-SDF: Text-to-Shape via Voxelized Diffusion", "Neural Wavelet-domain Diffusion for 3D Shape Generation"], "answer_arxiv_id": ["2212.01206", "2212.03293", "2209.08725"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_20395"} +{"question": "Which studies have found that adding random noise can improve the robustness against accumulated error at the inference time in Scientific ML?", "answer": ["Graph Networks as Learnable Physics Engines for Inference and Control", "Learning to Simulate Complex Physics with Graph Networks", "Learning Mesh-Based Simulation with Graph Networks", "Learned Coarse Models for Efficient Turbulence Simulation"], "answer_arxiv_id": ["1806.01242", "2002.09405", "2010.03409", "2112.15275"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_20396"} +{"question": "Which studies relate to predictable behavior learning and propose to minimize model errors?", "answer": ["Robust Predictable Control"], "answer_arxiv_id": ["2109.03214"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_20397"} +{"question": "Are there any studies that have leveraged a pre-trained speech processing model for time series classification?", "answer": ["Voice2Series: Reprogramming Acoustic Models for Time Series Classification"], "answer_arxiv_id": ["2106.09296"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_20398"} +{"question": "Could you provide me some studies which deal with the use of different auxiliary losses in layer-wise training?", "answer": ["Layer-wise training of deep networks using kernel similarity", "Putting An End to End-to-End: Gradient-Isolated Learning of Representations", "Layer-wise Learning of Stochastic Neural Networks with Information Bottleneck", "The HSIC Bottleneck: Deep Learning without Back-Propagation", "Revisiting Locally Supervised Learning: an Alternative to End-to-end Training", "Training Neural Networks with Local Error Signals", "A More Biologically Plausible Local Learning Rule for ANNs"], "answer_arxiv_id": ["1703.07115", "1905.11786", "1712.01272", "1908.01580", "2101.10832", "1901.06656", "2011.12012"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_20399"} +{"question": "Which works treat FL as a coalitional game theory problem and discuss between-client heterogeneity?", "answer": ["Model-sharing Games: Analyzing Federated Learning Under Voluntary Participation", "Optimality and Stability in Federated Learning: A Game-theoretic Approach"], "answer_arxiv_id": ["2010.00753", "2106.09580"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_20400"} +{"question": "What works illustrated that single-label approaches in representation learning resulted in sub-optimal models at the patch level?", "answer": ["Multi-modal Masked Autoencoders Learn Compositional Histopathological Representations"], "answer_arxiv_id": ["2209.01534"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_20401"} +{"question": "What research papers have looked into using a coarse-to-fine strategy with spatial propagation networks in depth enhancement?", "answer": ["Learning Affinity via Spatial Propagation Networks"], "answer_arxiv_id": ["1710.01020"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_20402"} +{"question": "What papers study linear bandits with a hidden low-rank structure?", "answer": ["Stochastic Linear Bandits with Hidden Low Rank Structure", "Bilinear Bandits with Low-rank Structure", "Low-Rank Generalized Linear Bandit Problems", "Optimal Gradient-based Algorithms for Non-concave Bandit Optimization"], "answer_arxiv_id": ["1901.09490", "1901.02470v2", "2006.02948", "2107.04518"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20403"} +{"question": "Could you list some studies dealing with subset selection methods by diversity criterion?", "answer": ["On Training Instance Selection for Few-Shot Neural Text Generation", "Semantic Redundancies in Image-Classification Datasets: The 10% You Don’t Need"], "answer_arxiv_id": ["2107.03176", "1901.11409"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_20404"} +{"question": "What works discuss the effectiveness of training image foundation models with large datasets?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale", "Masked Autoencoders Are Scalable Vision Learners", "BEiT: BERT Pre-Training of Image Transformers", "Flamingo: a Visual Language Model for Few-Shot Learning", "CoCa: Contrastive Captioners are Image-Text Foundation Models", "ImageBind: One Embedding Space To Bind Them All"], "answer_arxiv_id": ["2010.11929", "2111.06377", "2106.08254", "2204.14198", "2205.01917", "2305.05665v2"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20405"} +{"question": "Can you point me to some works done on STS tasks in multilingual and cross-lingual settings?", "answer": ["SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluation"], "answer_arxiv_id": ["1708.00055v1"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_20406"} +{"question": "What papers discussed recent linear attention formulations that potentially enable widespread use of attention in reinforcement learning?", "answer": ["Linear Transformers Are Secretly Fast Weight Programmers", "Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention"], "answer_arxiv_id": ["2102.11174", "2006.16236"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_20407"} +{"question": "What research demonstrates the exploitation of causal dependencies between action dimensions and reward terms to facilitate policy learning?", "answer": ["Causal Policy Gradient for Whole-Body Mobile Manipulation"], "answer_arxiv_id": ["2305.04866"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_20408"} +{"question": "Could you provide me some works that extended model-based RL in a number of settings?", "answer": ["DayDreamer: World Models for Physical Robot Learning", "Discovering and Achieving Goals via World Models", "Deep Hierarchical Planning from Pixels", "Transformers are Sample-Efficient World Models", "TransDreamer: Reinforcement Learning with Transformer World Models", "DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations", "Mastering Diverse Domains through World Models"], "answer_arxiv_id": ["2206.14176", "2110.09514", "2206.04114", "2209.00588", "2202.09481", "2110.14565", "2301.04104"], "source_meta": {"published_time": "20230628"}, "qid": "AutoScholarQuery_train_20409"} +{"question": "Could you provide me some works that employ feature-based methods in knowledge distillation algorithms?", "answer": ["Knowledge Distillation with the Reused Teacher Classifier", "Distilling Knowledge via Knowledge Review", "A Comprehensive Overhaul of Feature Distillation", "Knowledge Distillation via the Target-aware Transformer", "FitNets: Hints for Thin Deep Nets", "Contrastive Representation Distillation", "Variational Information Distillation for Knowledge Transfer", "Online Knowledge Distillation for Efficient Pose Estimation", "Class Attention Transfer Based Knowledge Distillation"], "answer_arxiv_id": ["2203.14001", "2104.09044", "1904.01866", "2205.10793", "1412.6550", "1910.10699", "1904.05835", "2108.02092", "2304.12777"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_20410"} +{"question": "What research showcased that transformers can perform in-context learning by simulating gradient descent?", "answer": ["Transformers Learn In-Context by Gradient Descent", "What learning algorithm is in-context learning? Investigations with linear models", "Looped Transformers as Programmable Computers"], "answer_arxiv_id": ["2212.07677", "2211.15661", "2301.13196v1"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_20411"} +{"question": "Which studies have been dedicated towards Compositional Generalization?", "answer": ["Measuring Compositional Generalization: A Comprehensive Method on Realistic Data", "Generalization without Systematicity: On the Compositional Skills of Sequence-to-Sequence Recurrent Networks", "COGS: A Compositional Generalization Challenge Based on Semantic Interpretation", "Systematic Generalization with Edge Transformers", "Compositional Generalization in Semantic Parsing: Pre-training vs. Specialized Architectures", "Span-based Semantic Parsing for Compositional Generalization", "Meta-Learning to Compositionally Generalize", "Learning to Recombine and Resample Data for Compositional Generalization", "Rearranging the Familiar: Testing Compositional Generalization in Recurrent Networks", "Improving Compositional Generalization in Semantic Parsing", "Learning to generalize to new compositions in image understanding", "Neural Baby Talk", "Compositional Generalization in Image Captioning", "C-VQA: A Compositional Split of the Visual Question Answering (VQA) v1.0 Dataset", "Separating Skills and Concepts for Novel Visual Question Answering", "Systematic Generalization: What Is Required and Can It Be Learned?", "CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning"], "answer_arxiv_id": ["1912.09713", "1711.00350", "2010.05465", "2112.00578", "2007.08970", "2009.06040", "2106.04252", "2010.03706", "1807.07545v1", "2010.05647", "1608.07639", "1803.09845", "1909.04402", "1704.08243", "2107.09106", "1811.12889", "1612.06890"], "source_meta": {"published_time": "20231116"}, "qid": "AutoScholarQuery_train_20412"} +{"question": "Could you provide the references of the works that utilize ShapeNet and OmniObject3D datasets for category-specific mesh generation?", "answer": ["ShapeNet: An Information-Rich 3D Model Repository", "OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic\n Perception, Reconstruction and Generation"], "answer_arxiv_id": ["1512.03012", "2301.07525"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_20413"} +{"question": "What works have looked at how fine-tuning can adapt pretrained features to a target distribution?", "answer": ["How transferable are features in deep neural networks?", "CNN Features off-the-shelf: an Astounding Baseline for Recognition"], "answer_arxiv_id": ["1411.1792", "1403.6382"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_20414"} +{"question": "Which papers presented the utilization of octrees in accelerating 3D GANs?", "answer": ["PlenOctrees for Real-time Rendering of Neural Radiance Fields", "Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D\n Shapes", "NerfAcc: Efficient Sampling Accelerates NeRFs"], "answer_arxiv_id": ["2103.14024", "2101.10994", "2305.04966"], "source_meta": {"published_time": "20240104"}, "qid": "AutoScholarQuery_train_20415"} +{"question": "Which work proposed a solution to the problem of non-differentiability in the initial design of selective rationalization frameworks by using differentiable sampling techniques?", "answer": ["Interpretable Neural Predictions with Differentiable Binary Variables", "How Does Selective Mechanism Improve Self-Attention Networks?"], "answer_arxiv_id": ["1905.08160", "2005.00979"], "source_meta": {"published_time": "20230625"}, "qid": "AutoScholarQuery_train_20416"} +{"question": "Which works have investigated the impact of pretraining on feature representations in computer vision models?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "Bootstrap Your Own Latent A New Approach to Self-Supervised Learning", "Representation Learning via Invariant Causal Mechanisms", "Why Do Better Loss Functions Lead to Less Transferable Features?"], "answer_arxiv_id": ["2002.05709", "2006.07733", "2010.07922", "2010.16402"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_20417"} +{"question": "Could you provide me some works that derived NTKs for convolutional neural networks?", "answer": ["On Exact Computation with an Infinitely Wide Neural Net", "Enhanced Convolutional Neural Tangent Kernels"], "answer_arxiv_id": ["1904.11955", "1911.00809v1"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_20418"} +{"question": "What papers introduced neural networks for simulating movement trajectories or estimating physical parameters?", "answer": ["NeuralSim: Augmenting Differentiable Simulators with Neural Networks", "Neural Implicit Representations for Physical Parameter Inference from a\n Single Video"], "answer_arxiv_id": ["2011.04217", "2204.14030"], "source_meta": {"published_time": "20231121"}, "qid": "AutoScholarQuery_train_20419"} +{"question": "What works explore the idea of large kernel paralleling or stacking?", "answer": ["Large Kernel Matters —— Improve Semantic Segmentation by Global Convolutional Network", "Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning"], "answer_arxiv_id": ["1703.02719", "1602.07261"], "source_meta": {"published_time": "20220707"}, "qid": "AutoScholarQuery_train_20420"} +{"question": "What studies have considered the removal of subsets of training data for purposes like model debugging or the evaluation of explanation techniques?", "answer": ["Towards A Rigorous Science of Interpretable Machine Learning", "A Benchmark for Interpretability Methods in Deep Neural Networks", "A Consistent and Efficient Evaluation Strategy for Attribution Methods"], "answer_arxiv_id": ["1702.08608", "1806.10758", "2202.00449"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_20421"} +{"question": "Can you identify the studies that have reformed the task of medical image segmentation as a sequence-to-sequence prediction task by using the vision transformer (ViT) architecture?", "answer": ["TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation", "UNETR: Transformers for 3D Medical Image Segmentation", "Class-Aware Adversarial Transformers for Medical Image Segmentation"], "answer_arxiv_id": ["2102.04306", "2103.10504", "2201.10737"], "source_meta": {"published_time": "20230203"}, "qid": "AutoScholarQuery_train_20422"} +{"question": "Could you provide me some papers that discuss adapter-based methods in PETL?", "answer": ["CLIP-Adapter: Better Vision-Language Models with Feature Adapters", "Parameter-Efficient Transfer Learning for NLP", "Compacter: Efficient Low-Rank Hypercomplex Adapter Layers", "Parameter-efficient Multi-task Fine-tuning for Transformers via Shared Hypernetworks", "VL-Adapter: Parameter-Efficient Transfer Learning for Vision-and-Language Tasks", "Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling"], "answer_arxiv_id": ["2110.04544", "1902.00751", "2106.04647", "2106.04489", "2112.06825", "2111.03930"], "source_meta": {"published_time": "20230904"}, "qid": "AutoScholarQuery_train_20423"} +{"question": "What papers described the method for recognizing a group activity from a whole image without any person features?", "answer": ["Detector-Free Weakly Supervised Group Activity Recognition"], "answer_arxiv_id": ["2204.02139"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_20424"} +{"question": "Which studies focus on the notion of bisimulation abstraction for representation learning?", "answer": ["Metrics for Finite Markov Decision Processes"], "answer_arxiv_id": ["1207.4114"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_20425"} +{"question": "What paper proposed the method of Hypothesis Search for improving the induced hypotheses of LLMs?", "answer": ["Hypothesis Search: Inductive Reasoning with Language Models"], "answer_arxiv_id": ["2309.05660"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_20426"} +{"question": "Could you provide me some papers in which the SignNet approach is used?", "answer": ["Recipe for a General, Powerful, Scalable Graph Transformer", "Transformers Meet Directed Graphs", "Efficiently predicting high resolution mass spectra with graph neural networks"], "answer_arxiv_id": ["2205.12454", "2302.00049", "2301.11419"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20427"} +{"question": "Which works explore the improvement of k-means approximation ratio by adding local search steps after the seeding procedure?", "answer": ["k-means++: few more steps yield constant approximation"], "answer_arxiv_id": ["2002.07784v1"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_20428"} +{"question": "Which works discuss the polynomial growth in the number of neurons and exponential growth in the number of input dimensions of a neural network's regions?", "answer": ["Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problem", "Complexity of Linear Regions in Deep Networks"], "answer_arxiv_id": ["1812.05720", "1901.09021"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_20429"} +{"question": "Which papers studied the emergence of outliers at scale that hamper quantization attempts?", "answer": ["Outlier Suppression: Pushing the Limit of Low-bit Transformer Language Models"], "answer_arxiv_id": ["2209.13325"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20430"} +{"question": "What works studied the generalization ability of machine-generated text detection?", "answer": ["Release Strategies and the Social Impacts of Language Models", "On the Zero-Shot Generalization of Machine-Generated Text Detectors", "M4: Multi-generator, Multi-domain, and Multi-lingual Black-Box\n Machine-Generated Text Detection", "Smaller Language Models are Better Black-box Machine-Generated Text\n Detectors"], "answer_arxiv_id": ["1908.09203v2", "2310.05165", "2305.14902", "2305.09859"], "source_meta": {"published_time": "20240218"}, "qid": "AutoScholarQuery_train_20431"} +{"question": "Which papers discussed self-calibration in continuation of the work by Neural Radiance Fields?", "answer": ["Self-Calibrating Neural Radiance Fields"], "answer_arxiv_id": ["2108.13826"], "source_meta": {"published_time": "20230126"}, "qid": "AutoScholarQuery_train_20432"} +{"question": "Which studies fine-tuned models by introducing learnable prompt tokens in PEFT?", "answer": ["The Power of Scale for Parameter-Efficient Prompt Tuning"], "answer_arxiv_id": ["2104.08691"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_20433"} +{"question": "Are there any works featuring the reconstruction of 3D volumes from 2D microscopy images using implicit neural representations?", "answer": ["Reconstructing continuous distributions of 3D protein structure from\n cryo-EM images", "FPM-INR: Fourier ptychographic microscopy image stack reconstruction\n using implicit neural representations"], "answer_arxiv_id": ["1909.05215", "2310.18529"], "source_meta": {"published_time": "20240316"}, "qid": "AutoScholarQuery_train_20434"} +{"question": "What works use quantile-regression-based conformal prediction methods?", "answer": ["Conformalized Quantile Regression", "Nested conformal prediction and quantile out-of-bag ensemble methods"], "answer_arxiv_id": ["1905.03222", "1910.10562"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_20435"} +{"question": "What studies focus on the importance of disclosing the details of evaluation procedures for text-to-speech research?", "answer": ["Why We Should Report the Details in Subjective Evaluation of TTS More Rigorously"], "answer_arxiv_id": ["2306.02044"], "source_meta": {"published_time": "20230613"}, "qid": "AutoScholarQuery_train_20436"} +{"question": "What studies have attempted to extend conformal prediction to handle time series data?", "answer": ["Exact and Robust Conformal Inference Methods for Predictive Machine Learning With Dependent Data"], "answer_arxiv_id": ["1802.06300"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_20437"} +{"question": "What research papers reported on the shortcomings of feature attribution methods through evaluation metrics?", "answer": ["Sanity Checks for Saliency Maps", "A Benchmark for Interpretability Methods in Deep Neural Networks", "Do Input Gradients Highlight Discriminative Features?"], "answer_arxiv_id": ["1810.03292", "1806.10758", "2102.12781"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_20438"} +{"question": "Are there any papers that used code-like structures for causal reasoning tasks?", "answer": ["Causal Reasoning of Entities and Events in Procedural Texts"], "answer_arxiv_id": ["2301.10896"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_20439"} +{"question": "Who conducted a study on audio-visual event recognition and localization using positive sample propagation?", "answer": ["Positive Sample Propagation along the Audio-Visual Event Line"], "answer_arxiv_id": ["2104.00239"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_20440"} +{"question": "Are there any works that use auto-encoders for image reconstruction?", "answer": ["AutoLink: Self-supervised Learning of Human Skeletons and Object\n Outlines by Linking Keypoints", "Unsupervised Discovery of Object Landmarks as Structural Representations"], "answer_arxiv_id": ["2205.10636", "1804.04412"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_20441"} +{"question": "Could you provide me some studies about using lightweight MLP structures in time series forecasting?", "answer": ["Less Is More: Fast Multivariate Time Series Forecasting with Light Sampling-oriented MLP Structures"], "answer_arxiv_id": ["2207.01186"], "source_meta": {"published_time": "20231110"}, "qid": "AutoScholarQuery_train_20442"} +{"question": "Which works propose methods for physically modeling both the environment and radar systems?", "answer": ["MaxRay: A Raytracing-based Integrated Sensing and Communication\n Framework"], "answer_arxiv_id": ["2112.01751"], "source_meta": {"published_time": "20240428"}, "qid": "AutoScholarQuery_train_20443"} +{"question": "What are some of the studies that require multi-view RGB video for human body reconstruction?", "answer": ["Neural Body: Implicit Neural Representations with Structured Latent\n Codes for Novel View Synthesis of Dynamic Humans", "Surface-Aligned Neural Radiance Fields for Controllable 3D Human\n Synthesis", "Structured Local Radiance Fields for Human Avatar Modeling", "AvatarReX: Real-time Expressive Full-body Avatars", "PoseVocab: Learning Joint-structured Pose Embeddings for Human Avatar\n Modeling", "HumanRF: High-Fidelity Neural Radiance Fields for Humans in Motion", "HandNeRF: Neural Radiance Fields for Animatable Interacting Hands", "RelightableHands: Efficient Neural Relighting of Articulated Hand Models", "LISA: Learning Implicit Shape and Appearance of Hands", "DoubleField: Bridging the Neural Surface and Radiance Fields for\n High-fidelity Human Reconstruction and Rendering", "HumanNeRF: Efficiently Generated Human Radiance Field from Sparse Inputs", "Animatable Neural Radiance Fields for Modeling Dynamic Human Bodies", "ARAH: Animatable Volume Rendering of Articulated Human SDFs", "TAVA: Template-free Animatable Volumetric Actors"], "answer_arxiv_id": ["2012.15838", "2201.01683", "2203.14478", "2305.04789", "2304.13006", "2305.06356", "2303.13825", "2302.04866", "2204.01695", "2106.03798", "2112.02789", "2105.02872", "2210.10036v1", "2206.08929"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_20444"} +{"question": "Are there studies that use a single model to represent the body and the garment?", "answer": ["3D Human Body Reconstruction from a Single Image via Volumetric\n Regression", "DeepHuman: 3D Human Reconstruction from a Single Image", "PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization", "PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution\n 3D Human Digitization", "Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view\n Human Reconstruction", "Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing", "Learning to Reconstruct People in Clothing from a Single RGB Camera", "Tex2Shape: Detailed Full Human Body Geometry From a Single Image", "ARCH: Animatable Reconstruction of Clothed Humans", "ARCH++: Animation-Ready Clothed Human Reconstruction Revisited", "ICON: Implicit Clothed humans Obtained from Normals", "PaMIR: Parametric Model-Conditioned Implicit Representation for\n Image-based Human Reconstruction"], "answer_arxiv_id": ["1809.03770", "1903.06473", "1905.05172", "2004.00452", "2006.08072", "2204.08906", "1903.05885", "1904.08645", "2004.04572", "2108.07845", "2112.09127", "2007.03858"], "source_meta": {"published_time": "20231117"}, "qid": "AutoScholarQuery_train_20445"} +{"question": "What works introduced adversarial examples in computer vision?", "answer": ["Intriguing properties of neural networks", "Explaining and Harnessing Adversarial Examples"], "answer_arxiv_id": ["1312.6199", "1412.6572"], "source_meta": {"published_time": "20240501"}, "qid": "AutoScholarQuery_train_20446"} +{"question": "Are there any papers on the use of specialized expert models with large language models?", "answer": ["HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face"], "answer_arxiv_id": ["2303.17580"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_20447"} +{"question": "What studies use biological entity-based graphs for providing deeper insights?", "answer": ["Towards Explainable Graph Representations in Digital Pathology", "Quantifying Explainers of Graph Neural Networks in Computational\n Pathology"], "answer_arxiv_id": ["2007.00311", "2011.12646"], "source_meta": {"published_time": "20231222"}, "qid": "AutoScholarQuery_train_20448"} +{"question": "What work explored the impact of language-based models on computer vision?", "answer": ["Language in a Bottle: Language Model Guided Concept Bottlenecks for\n Interpretable Image Classification", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "Verbs in Action: Improving verb understanding in video-language models"], "answer_arxiv_id": ["2211.11158", "2211.09800", "2304.06708"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_20449"} +{"question": "Which studies focus on constructing RL agents that learn policies and compress sequences of observations?", "answer": ["Robust Predictable Control"], "answer_arxiv_id": ["2109.03214"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_20450"} +{"question": "Which work first applied the orthogonal constraint to Vision Transformer?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20220527"}, "qid": "AutoScholarQuery_train_20451"} +{"question": "Which works explored sampling and post-processing modules for enhancing point features in point cloud semantic segmentation?", "answer": ["Learning to Sample", "PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Sampling", "Modeling Point Clouds with Self-Attention and Gumbel Subset Sampling", "SASA: Semantics-Augmented Set Abstraction for Point-based 3D Object Detection", "JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Clouds", "Contrastive Boundary Learning for Point Cloud Segmentation"], "answer_arxiv_id": ["1812.01659", "2003.00492", "1904.03375", "2201.01976", "2007.06888", "2203.05272"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_20452"} +{"question": "Which studies employ late fusion mechanisms in multimodal integration for cancer prognosis?", "answer": ["Pathomic Fusion: An Integrated Framework for Fusing Histopathology and\n Genomic Features for Cancer Diagnosis and Prognosis"], "answer_arxiv_id": ["1912.08937"], "source_meta": {"published_time": "20230413"}, "qid": "AutoScholarQuery_train_20453"} +{"question": "Could you provide some studies about context encoder?", "answer": ["Equivariant Graph Neural Networks for 3D Macromolecular Structure", "Vector Neurons: A General Framework for SO(3)-Equivariant Networks"], "answer_arxiv_id": ["2106.03843", "2104.12229"], "source_meta": {"published_time": "20230522"}, "qid": "AutoScholarQuery_train_20454"} +{"question": "Can you list some notable open-source 100K+ LLMs?", "answer": ["YaRN: Efficient Context Window Extension of Large Language Models"], "answer_arxiv_id": ["2309.00071"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_20455"} +{"question": "Which studies introduced regularized versions of WDRO?", "answer": ["Sinkhorn Distributionally Robust Optimization", "Regularization for Wasserstein Distributionally Robust Optimization"], "answer_arxiv_id": ["2109.11926v4", "2205.08826"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_20456"} +{"question": "Which work first proposed Group-wise Correlation Volume in stereo matching?", "answer": ["Group-wise Correlation Stereo Network"], "answer_arxiv_id": ["1903.04025"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_20457"} +{"question": "What is the work that utilizes internet-scale data through learning an inverse dynamics model?", "answer": ["Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos"], "answer_arxiv_id": ["2206.11795"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_20458"} +{"question": "Which studies are related to the improvement of R@1 when learning a graph neural network in conjunction with class centroids?", "answer": ["Learning Intra-Batch Connections for Deep Metric Learning"], "answer_arxiv_id": ["2102.07753"], "source_meta": {"published_time": "20221004"}, "qid": "AutoScholarQuery_train_20459"} +{"question": "Could you name the models trained on several million 3D models to generate point clouds and parameters of implicit function?", "answer": ["Point-E: A System for Generating 3D Point Clouds from Complex Prompts", "Shap-E: Generating Conditional 3D Implicit Functions"], "answer_arxiv_id": ["2212.08751", "2305.02463"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_20460"} +{"question": "Could you provide some papers that explored parameters of potential functions using meta-learning technique?", "answer": ["Reward Shaping via Meta-Learning", "Learning to Utilize Shaping Rewards: A New Approach of Reward Shaping"], "answer_arxiv_id": ["1901.09330", "2011.02669"], "source_meta": {"published_time": "20231029"}, "qid": "AutoScholarQuery_train_20461"} +{"question": "Which papers scaled multimodal approaches to hundreds of millions of text-image pairs using deep learning and contrastive learning?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_20462"} +{"question": "Any work simplifying the operations in the non-local block while maintaining performances?", "answer": ["GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond"], "answer_arxiv_id": ["1904.11492"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_20463"} +{"question": "Could you provide me with research that applies greedy per-neuron quantization procedure in adaptive rounding?", "answer": ["Global Convergence of Stochastic Gradient Descent for Some Non-convex Matrix Problems"], "answer_arxiv_id": ["1411.1134"], "source_meta": {"published_time": "20230725"}, "qid": "AutoScholarQuery_train_20464"} +{"question": "What papers propose strategies such as contrasting negative samples, variance-covariance regularization, or by maximizing the entropy of the representations?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "A Simple Framework for Contrastive Learning of Visual Representations", "VICReg: Variance-Invariance-Covariance Re-gularization for Self-Supervised Learning", "Barlow Twins: Self-Supervised Learning via Redundancy Reduction", "Self-labelling via simultaneous clustering and representation learning", "Unsupervised Learning of Visual Features by Contrasting Cluster Assignments", "Emerging Properties in Self-Supervised Vision Transformers", "Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samples", "Masked Siamese Networks for Label-Efficient Learning"], "answer_arxiv_id": ["1911.05722", "2002.05709", "2105.04906", "2103.03230", "1911.05371", "2006.09882", "2104.14294", "2104.13963", "2204.07141"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_20465"} +{"question": "What research introduced the visual features into the QFormer alongside the instruction?", "answer": ["InstructBLIP: Towards General-purpose Vision-Language Models with\n Instruction Tuning"], "answer_arxiv_id": ["2305.06500"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_20466"} +{"question": "Can you name some methods that utilize the coarse-to-fine optimization approach?", "answer": ["Magic3D: High-Resolution Text-to-3D Content Creation", "Progressive3D: Progressively Local Editing for Text-to-3D Content\n Creation with Complex Semantic Prompts"], "answer_arxiv_id": ["2211.10440", "2310.11784"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_20467"} +{"question": "Are there any studies that extracted relations from the feature map and transferred the relation for knowledge distillation?", "answer": ["Relational Knowledge Distillation"], "answer_arxiv_id": ["1904.05068"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20468"} +{"question": "What papers propose structure-based methods for camera pose prediction?", "answer": ["From Coarse to Fine: Robust Hierarchical Localization at Large Scale", "SuperGlue: Learning Feature Matching with Graph Neural Networks", "SuperPoint: Self-Supervised Interest Point Detection and Description", "LoFTR: Detector-Free Local Feature Matching with Transformers", "DISK: Learning local features with policy gradient", "DSAC - Differentiable RANSAC for Camera Localization", "Visual Camera Re-Localization from RGB and RGB-D Images Using DSAC", "Learning Less is More - 6D Camera Localization via 3D Surface Regression"], "answer_arxiv_id": ["1812.03506", "1911.11763", "1712.07629", "2104.00680", "2006.13566", "1611.05705", "2002.12324", "1711.10228"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_20469"} +{"question": "Whose research work first established downstream guarantees for contrastive learning?", "answer": ["A Theoretical Analysis of Contrastive Unsupervised Representation Learning"], "answer_arxiv_id": ["1902.09229v1"], "source_meta": {"published_time": "20230304"}, "qid": "AutoScholarQuery_train_20470"} +{"question": "What papers focus on learning generalizable representations to improve the generalization ability in GCRL?", "answer": ["Universal Successor Representations for Transfer Reinforcement Learning", "Universal Successor Features Approximators"], "answer_arxiv_id": ["1804.03758", "1812.07626"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20471"} +{"question": "Could you name the paper that demonstrates similar in-context behaviors in transformer model as in the current research?", "answer": ["In-context Reinforcement Learning with Algorithm Distillation"], "answer_arxiv_id": ["2210.14215"], "source_meta": {"published_time": "20230526"}, "qid": "AutoScholarQuery_train_20472"} +{"question": "Can you tell me about any papers that have investigated the utilization of conditional generative models in AVS methods?", "answer": ["Contrastive Conditional Latent Diffusion for Audio-visual Segmentation", "Multimodal Variational Auto-encoder based Audio-Visual Segmentation", "Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2307.16579", "2310.08303", "2006.11239"], "source_meta": {"published_time": "20230406"}, "qid": "AutoScholarQuery_train_20473"} +{"question": "Can you point me to works focusing on the readiness aspect of data, including its completeness and accessibility?", "answer": ["Data Readiness Levels", "Data Readiness Report"], "answer_arxiv_id": ["1705.02245", "2010.07213"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_20474"} +{"question": "Could you provide me some studies about the manipulation of deformable objects?", "answer": ["VisuoSpatial Foresight for Physical Sequential Fabric Manipulation", "Learning Visible Connectivity Dynamics for Cloth Smoothing", "FabricFlowNet: Bimanual Cloth Manipulation with a Flow-based Policy", "Learning to Manipulate Deformable Objects without Demonstrations", "Learning Rope Manipulation Policies Using Dense Object Descriptors Trained on Synthetic Depth Data", "Self-Supervised Learning of State Estimation for Manipulating Deformable Linear Objects", "Learning to Manipulate Deformable Objects without Demonstrations", "Learning to Control PDEs with Differentiable Physics", "3D Neural Scene Representations for Visuomotor Control", "Visual Closed-Loop Control for Pouring Liquids", "PlasticineLab: A Soft-Body Manipulation Benchmark with Differentiable Physics", "Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids", "RoboCraft: Learning to See, Simulate, and Shape Elasto-Plastic Objects with Graph Networks", "Deformable Elasto-Plastic Object Shaping using an Elastic Hand and Model-Based Reinforcement Learning", "DiSECt: A Differentiable Simulation Engine for Autonomous Robotic Cutting"], "answer_arxiv_id": ["2102.09754", "2105.10389", "2111.05623v3", "1910.13439", "2003.01835", "1911.06283v3", "1910.13439", "2001.07457", "2107.04004", "1610.02610", "2104.03311", "1810.01566", "2205.02909", "2107.06924", "2105.12244"], "source_meta": {"published_time": "20230327"}, "qid": "AutoScholarQuery_train_20475"} +{"question": "In which publication can you find a comprehensive comparison of Griddly, NetHack, and MineRL with the initial Neural MMO?", "answer": ["The Neural MMO Platform for Massively Multiagent Research"], "answer_arxiv_id": ["2110.07594"], "source_meta": {"published_time": "20231107"}, "qid": "AutoScholarQuery_train_20476"} +{"question": "What works extended the setting of stochastic rank-1 matrix bandits?", "answer": ["Stochastic Low-Rank Bandits", "Bilinear Bandits with Low-rank Structure", "Optimal Gradient-based Algorithms for Non-concave Bandit Optimization", "Low-Rank Generalized Linear Bandit Problems"], "answer_arxiv_id": ["1712.04644", "1901.02470v2", "2107.04518", "2006.02948"], "source_meta": {"published_time": "20220908"}, "qid": "AutoScholarQuery_train_20477"} +{"question": "Could you provide me the studies that focus on counterfactual explanations?", "answer": ["Counterfactual Visual Explanations", "Explainable Image Classification with Evidence Counterfactual", "Counterfactual States for Atari Agents via Generative Deep Learning"], "answer_arxiv_id": ["1904.07451", "2004.07511", "1909.12969"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_20478"} +{"question": "What are the experimental works that highlight the need for new explanations to understand generalization in deep learning?", "answer": ["Understanding deep learning requires rethinking generalization", "Exploring Generalization in Deep Learning"], "answer_arxiv_id": ["1611.03530v2", "1706.08947"], "source_meta": {"published_time": "20220616"}, "qid": "AutoScholarQuery_train_20479"} +{"question": "What work propounded modeling functions with diffusion by defining a Gaussian measure on Hilbert spaces?", "answer": ["Diffusion Generative Models in Infinite Dimensions"], "answer_arxiv_id": ["2212.00886v2"], "source_meta": {"published_time": "20221104"}, "qid": "AutoScholarQuery_train_20480"} +{"question": "Which works contributed to the progress in text-to-image diffusion models?", "answer": ["Zero-Shot Text-to-Image Generation", "Hierarchical Text-Conditional Image Generation with CLIP Latents", "High-Resolution Image Synthesis with Latent Diffusion Models", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "SDXL: Improving Latent Diffusion Models for High-Resolution Image\n Synthesis", "RAPHAEL: Text-to-Image Generation via Large Mixture of Diffusion Paths", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models", "IPDreamer: Appearance-Controllable 3D Object Generation with Complex\n Image Prompts"], "answer_arxiv_id": ["2102.12092", "2204.06125", "2112.10752", "2205.11487", "2205.11487", "2307.01952", "2305.18295", "2211.01324", "2112.10741", "2310.05375"], "source_meta": {"published_time": "20231226"}, "qid": "AutoScholarQuery_train_20481"} +{"question": "Which work proposed the Segment Anything Model (SAM)?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20240504"}, "qid": "AutoScholarQuery_train_20482"} +{"question": "Which works employ supervised training for image super-resolution?", "answer": ["GLEAN: Generative Latent Bank for Large-Factor Image Super-Resolution", "Panini-Net: GAN Prior Based Degradation-Aware Feature Interpolation for\n Face Restoration", "Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure\n Synthetic Data"], "answer_arxiv_id": ["2012.00739", "2203.08444", "2107.10833"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_20483"} +{"question": "What attempts have been made to mitigate the computational complexity of the attention mechanism used in image segmentation?", "answer": ["Rethinking Vision Transformers for MobileNet Size and Speed", "Separable Self-attention for Mobile Vision Transformers", "SwiftFormer: Efficient Additive Attention for Transformer-based\n Real-time Mobile Vision Applications"], "answer_arxiv_id": ["2212.08059", "2206.02680", "2303.15446"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20484"} +{"question": "Are there any studies that present semi-automatic construction of knowledge bases?", "answer": ["UnCommonSense: Informative Negative Knowledge about Everyday Concepts"], "answer_arxiv_id": ["2208.09292"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_20485"} +{"question": "What works are about training a model for real-world noise in unsupervised learning-based image denoising?", "answer": ["C2N: Practical Generative Noise Modeling for Real-World Denoising", "Modeling sRGB Camera Noise with Normalizing Flows"], "answer_arxiv_id": ["2202.09533", "2206.00812"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_20486"} +{"question": "What papers discuss the usage of evolutionary algorithms in performing text-to-image adaptation?", "answer": ["Gradient-Free Textual Inversion"], "answer_arxiv_id": ["2304.05818"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20487"} +{"question": "Could you provide me examples of studies that used edge maps for image super-resolution tasks?", "answer": ["Edge-Informed Single Image Super-Resolution"], "answer_arxiv_id": ["1909.05305"], "source_meta": {"published_time": "20200331"}, "qid": "AutoScholarQuery_train_20488"} +{"question": "What papers have addressed the word frequency bias in foundational models?", "answer": ["A Simple Zero-shot Prompt Weighting Technique to Improve Prompt Ensembling in Text-Image Models"], "answer_arxiv_id": ["2302.06235"], "source_meta": {"published_time": "20231012"}, "qid": "AutoScholarQuery_train_20489"} +{"question": "Which paper proposed an Epipolar Attention Module to fuse single-view and multi-view geometric information?", "answer": ["MVS2D: Efficient Multi-view Stereo via Attention-Driven 2D Convolutions"], "answer_arxiv_id": ["2104.13325"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_20490"} +{"question": "What works have aimed to adjust the edge weights in the coarsened graph through graph neural networks (GNNs)?", "answer": ["Graph Coarsening with Neural Networks"], "answer_arxiv_id": ["2102.01350"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_20491"} +{"question": "What studies used a fusion of text, audio and upper body gestures to learn an upper body gesture avatar?", "answer": ["Speech Gesture Generation from the Trimodal Context of Text, Audio, and\n Speaker Identity"], "answer_arxiv_id": ["2009.02119"], "source_meta": {"published_time": "20231207"}, "qid": "AutoScholarQuery_train_20492"} +{"question": "Could you provide me some studies about using additive attention on tabular tasks?", "answer": ["Fastformer: Additive Attention Can Be All You Need"], "answer_arxiv_id": ["2108.09084"], "source_meta": {"published_time": "20230510"}, "qid": "AutoScholarQuery_train_20493"} +{"question": "What studies focus on providing recent offline-to-online solutions in RL?", "answer": ["Conservative Q-Learning for Offline Reinforcement Learning", "Offline Reinforcement Learning with Implicit Q-Learning", "Supported Policy Optimization for Offline Reinforcement Learning", "Cal-QL: Calibrated Offline RL Pre-Training for Efficient Online Fine-Tuning", "Revisiting the Minimalist Approach to Offline Reinforcement Learning"], "answer_arxiv_id": ["2006.04779", "2110.06169", "2202.06239", "2303.05479", "2305.09836"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_20494"} +{"question": "In what research does the author report using a logarithmic quantization scheme for 4-bits Transformer?", "answer": ["Neural Machine Translation with 4-Bit Precision and Beyond"], "answer_arxiv_id": ["1909.06091"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20495"} +{"question": "Can you mention a research that explores drag-based editing with diffusion models?", "answer": ["DragonDiffusion: Enabling Drag-style Manipulation on Diffusion Models"], "answer_arxiv_id": ["2307.02421"], "source_meta": {"published_time": "20230626"}, "qid": "AutoScholarQuery_train_20496"} +{"question": "What studies use pretrained encoders for extracting data representations in developing multimodal LLMs?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "ImageBind: One Embedding Space To Bind Them All"], "answer_arxiv_id": ["2103.00020", "2201.12086", "2305.05665v2"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_20497"} +{"question": "What research inherited the center-based framework and proposed the omission of the estimation of the 2D bounding box in 3D object detection?", "answer": ["SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint\n Estimation"], "answer_arxiv_id": ["2002.10111"], "source_meta": {"published_time": "20240404"}, "qid": "AutoScholarQuery_train_20498"} +{"question": "What works focused on generating group dances?", "answer": ["Music-Driven Group Choreography", "Dance with You: The Diversity Controllable Dancer Generation via\n Diffusion Models"], "answer_arxiv_id": ["2303.12337", "2308.13551"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_20499"} +{"question": "Could you provide me some perturbation-based works in XAI methodology?", "answer": ["“Why Should I Trust You?” Explaining the Predictions of Any Classifier", "A Unified Approach to Interpreting Model Predictions"], "answer_arxiv_id": ["1602.04938", "1705.07874"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_20500"} +{"question": "What researches have been conducted about learning the causal structure in offline reinforcement learning (RL)?", "answer": ["Offline Reinforcement Learning with Causal Structured World Models"], "answer_arxiv_id": ["2206.01474"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_20501"} +{"question": "What research has been conducted exploring this sensitivity in the context of meta-learning?", "answer": ["Understanding and correcting pathologies in the training of learned optimizers"], "answer_arxiv_id": ["1810.10180"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_20502"} +{"question": "What papers studied relevant techniques on control of a multi-camera team?", "answer": ["FlyCap: Markerless Motion Capture Using Multiple Autonomous Flying Cameras", "Active Perception based Formation Control for Multiple Aerial Vehicles", "3D Human Reconstruction in the Wild with Collaborative Aerial Cameras"], "answer_arxiv_id": ["1610.09534v3", "1901.07813", "2108.03936"], "source_meta": {"published_time": "20230307"}, "qid": "AutoScholarQuery_train_20503"} +{"question": "Any works that discusses other concepts of interactions using marginal contributions?", "answer": ["How does this interaction affect me? Interpretable attribution for feature interactions"], "answer_arxiv_id": ["2006.10965"], "source_meta": {"published_time": "20230302"}, "qid": "AutoScholarQuery_train_20504"} +{"question": "What work extended the concept of neural ODEs to stochastic processes by using stochastic latent variables?", "answer": ["Neural ODE Processes"], "answer_arxiv_id": ["2103.12413"], "source_meta": {"published_time": "20230226"}, "qid": "AutoScholarQuery_train_20505"} +{"question": "What papers analyze the effect of clipping in DP-(S)GD?", "answer": ["Understanding Gradient Clipping in Private SGD: A Geometric Perspective"], "answer_arxiv_id": ["2006.15429"], "source_meta": {"published_time": "20220621"}, "qid": "AutoScholarQuery_train_20506"} +{"question": "Which research revisited the use of large kernel-sized convolutions for ConvNets?", "answer": ["Swin Transformer: Hierarchical Vision Transformer using Shifted Windows", "A ConvNet for the 2020s"], "answer_arxiv_id": ["2103.14030", "2201.03545"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20507"} +{"question": "What is the reference for the model known as Perceiver IO, which adapts the original Perceiver to be versatile and linearly scalable?", "answer": ["Perceiver IO: A General Architecture for Structured Inputs & Outputs"], "answer_arxiv_id": ["2107.14795"], "source_meta": {"published_time": "20240718"}, "qid": "AutoScholarQuery_train_20508"} +{"question": "What papers have used the mixup-style interpolation technique to generate augmented images for domain generalization?", "answer": ["Improving Out-of-Distribution Robustness via Selective Augmentation"], "answer_arxiv_id": ["2201.00299"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_20509"} +{"question": "What works have studied the expressive power of graph neural networks under the assumption of unique node identifiers?", "answer": ["What Graph Neural Networks Cannot Learn: Depth vs Width"], "answer_arxiv_id": ["1907.03199"], "source_meta": {"published_time": "20230130"}, "qid": "AutoScholarQuery_train_20510"} +{"question": "Which research work introduced the Segment Anything Model (SAM)?", "answer": ["Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks"], "answer_arxiv_id": ["2401.14159"], "source_meta": {"published_time": "20231217"}, "qid": "AutoScholarQuery_train_20511"} +{"question": "Can you cite research papers that implemented neural rendering approaches focusing on surface modeling of masked objects from monocular RGB and RGB-D?", "answer": ["Unbiased 4D: Monocular 4D Reconstruction with a Neural Deformation Model", "Neural Surface Reconstruction of Dynamic Scenes with Monocular RGB-D\n Camera"], "answer_arxiv_id": ["2206.08368", "2206.15258"], "source_meta": {"published_time": "20231202"}, "qid": "AutoScholarQuery_train_20512"} +{"question": "Which studies focused on models performing in-context learning on VL tasks?", "answer": ["An Empirical Study of GPT-3 for Few-Shot Knowledge-Based VQA", "Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering", "Multimodal Chain-of-Thought Reasoning in Language Models", "Unified-IO: A unified model for vision, language, and multi-modal tasks", "Prompt-aligned Gradient for Prompt Tuning"], "answer_arxiv_id": ["2109.05014", "2209.09513", "2302.00923", "2206.08916", "2205.14865"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_20513"} +{"question": "Could you list works that explored the idea of distilling the outcomes of planner into the goal-conditioned policy?", "answer": ["Combining Q-Learning and Search with Amortized Value Estimates"], "answer_arxiv_id": ["1912.02807"], "source_meta": {"published_time": "20230320"}, "qid": "AutoScholarQuery_train_20514"} +{"question": "Are there any papers that explored aggregating multiple responses in multi-pass ICL through self-consistency techniques?", "answer": ["Self-Consistency Improves Chain of Thought Reasoning in Language Models"], "answer_arxiv_id": ["2203.11171"], "source_meta": {"published_time": "20230211"}, "qid": "AutoScholarQuery_train_20515"} +{"question": "What works evaluated the expressivity of GNNs by measuring correspondence to Weisfeiler-Leman (WL) graph isomorphism tests?", "answer": ["Weisfeiler and Leman go Machine Learning: The Story so far"], "answer_arxiv_id": ["2112.09992"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_20516"} +{"question": "What research sources carried out an overview of prior datasets and studies in patent subject classification and summarization?", "answer": ["S2ORC: The Semantic Scholar Open Research Corpus", "Neural Text Generation from Structured Data with Application to the Biography Domain"], "answer_arxiv_id": ["1911.02782", "1603.07771"], "source_meta": {"published_time": "20220708"}, "qid": "AutoScholarQuery_train_20517"} +{"question": "Which studies discussed augmentation methods in terms of both input and feature spaces for long-tailed recognition?", "answer": ["M2m: Imbalanced Classification via Major-to-minor Translation", "Feature Space Augmentation for Long-Tailed Data", "MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition"], "answer_arxiv_id": ["2004.00431", "2008.03673", "2103.12579"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_20518"} +{"question": "Which paper first introduced the Composed Image Retrieval (CIR) task?", "answer": ["Composing Text and Image for Image Retrieval - An Empirical Odyssey"], "answer_arxiv_id": ["1812.07119"], "source_meta": {"published_time": "20240324"}, "qid": "AutoScholarQuery_train_20519"} +{"question": "What papers are about the better generalization abilities of Neural Radiance Fields?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "GRF: Learning a General Radiance Field for 3D Representation and Rendering"], "answer_arxiv_id": ["2012.02190", "2010.04595"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_20520"} +{"question": "Could you mention studies that provide infinite-horizon sample-efficient algorithms under the random access generative model setting?", "answer": ["Learning with Good Feature Representations in Bandits and in RL with a Generative Model", "Best Policy Identification in Linear MDPs"], "answer_arxiv_id": ["1911.07676", "2208.05633"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_20521"} +{"question": "Which papers have designed instance-optimal algorithms for specific interactive decision making problems based on the UCB foundations?", "answer": ["Multi-Armed Bandits with Correlated Arms", "Structure Adaptive Algorithms for Stochastic Bandits"], "answer_arxiv_id": ["1911.03959", "2007.00969v1"], "source_meta": {"published_time": "20220606"}, "qid": "AutoScholarQuery_train_20522"} +{"question": "What works used norm-based generalization analysis as a method for theoretical studies?", "answer": ["Norm-Based Capacity Control in Neural Networks"], "answer_arxiv_id": ["1503.00036"], "source_meta": {"published_time": "20230301"}, "qid": "AutoScholarQuery_train_20523"} +{"question": "What researches explain the role of differential privacy in deep neural networks?", "answer": ["Deep Learning with Differential Privacy", "Differential Privacy Has Disparate Impact on Model Accuracy"], "answer_arxiv_id": ["1607.00133", "1905.12101"], "source_meta": {"published_time": "20221020"}, "qid": "AutoScholarQuery_train_20524"} +{"question": "What papers describe the use of Fourier transform in developing network architectures?", "answer": ["Clebsch–Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network"], "answer_arxiv_id": ["1806.09231"], "source_meta": {"published_time": "20231106"}, "qid": "AutoScholarQuery_train_20525"} +{"question": "What works stick with using standard activation functions in neural architecture search?", "answer": ["A Survey on Neural Architecture Search", "Neural Architecture Search: A Survey", "Neural Architecture Search with Reinforcement Learning"], "answer_arxiv_id": ["1905.01392", "1808.05377", "1611.01578"], "source_meta": {"published_time": "20230113"}, "qid": "AutoScholarQuery_train_20526"} +{"question": "Could you provide me some works dealing with mechanistic interpretability of transformers?", "answer": ["Dissecting Recall of Factual Associations in Auto-Regressive Language Models", "Locating and Editing Factual Associations in GPT", "Progress measures for grokking via mechanistic interpretability", "In-context Learning and Induction Heads", "Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small"], "answer_arxiv_id": ["2304.14767", "2202.05262", "2301.05217", "2209.11895v1", "2211.00593"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20527"} +{"question": "What studies develop autoregressive models to generate atoms and bonds step-by-step?", "answer": ["A 3D Generative Model for Structure-Based Drug Design", "Generating 3D Molecules for Target Protein Binding", "Pocket2Mol: Efficient Molecular Sampling Based on 3D Protein Pockets"], "answer_arxiv_id": ["2203.10446", "2204.09410", "2205.07249"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_20528"} +{"question": "Could you list some studies that explored MAE on point clouds by masked modeling of 3D point cloud coordinates?", "answer": ["Masked Autoencoders for Point Cloud Self-supervised Learning"], "answer_arxiv_id": ["2203.06604"], "source_meta": {"published_time": "20221216"}, "qid": "AutoScholarQuery_train_20529"} +{"question": "Which works proposed the fill-in-the-middle autoregressive infilling approach?", "answer": ["Efficient Training of Language Models to Fill in the Middle"], "answer_arxiv_id": ["2207.14255"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_20530"} +{"question": "What studies have proposed methods for taxonomy completion?", "answer": ["Taxonomy Completion via Triplet Matching Network", "TaxoEnrich: Self-Supervised Taxonomy Completion via Structure-Semantic\n Representations", "Enhancing Taxonomy Completion with Concept Generation via Fusing\n Relational Representations"], "answer_arxiv_id": ["2101.01896", "2202.04887", "2106.02974"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_20531"} +{"question": "What papers provides examples of Kernel-Based Efficient Transformers (KETs)?", "answer": ["Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention"], "answer_arxiv_id": ["2006.16236"], "source_meta": {"published_time": "20231018"}, "qid": "AutoScholarQuery_train_20532"} +{"question": "Could you provide me some information about the studies on model-free methods for robust discounted MDPs?", "answer": ["Reinforcement Learning under Model Mismatch", "Robust Reinforcement Learning using Least Squares Policy Iteration with Provable Performance Guarantees", "Online Robust Reinforcement Learning with Model Uncertainty", "Policy Gradient Method For Robust Reinforcement Learning", "Action Robust Reinforcement Learning and Applications in Continuous Control", "Towards Theoretical Understandings of Robust Markov Decision Processes: Sample Complexity and Asymptotics", "Sample Complexity of Robust Reinforcement Learning with a Generative Model", "Robust Markov Decision Process: Beyond Rectangularity", "Partial Policy Iteration for L1-Robust Markov Decision Processes"], "answer_arxiv_id": ["1706.04711", "2006.11608", "2109.14523", "2205.07344", "1901.09184", "2105.03863", "2112.01506v3", "1811.00215", "2006.09484v1"], "source_meta": {"published_time": "20230517"}, "qid": "AutoScholarQuery_train_20533"} +{"question": "What works aim to select effective distillation regions for better feature imitation in the context of Knowledge Distillation?", "answer": ["Distilling Object Detectors with Fine-grained Feature Imitation", "General Instance Distillation for Object Detection"], "answer_arxiv_id": ["1906.03609", "2103.02340"], "source_meta": {"published_time": "20230620"}, "qid": "AutoScholarQuery_train_20534"} +{"question": "Could you provide the works where learning the prompts by backpropagating the task-specific loss while freezing the VLM encoders is presented?", "answer": ["Learning to Prompt for Vision-Language Models", "Conditional Prompt Learning for Vision-Language Models", "Compositional Prompt Tuning with Motion Cues for Open-vocabulary Video Relation Detection"], "answer_arxiv_id": ["2109.01134", "2203.05557", "2302.00268"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_20535"} +{"question": "Could you provide some studies that focused on prompting generative LMs with enriched task requirements and examples?", "answer": ["Evaluating ChatGPT's Information Extraction Capabilities: An Assessment\n of Performance, Explainability, Calibration, and Faithfulness", "Exploring the Feasibility of ChatGPT for Event Extraction"], "answer_arxiv_id": ["2304.11633", "2303.03836"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_20536"} +{"question": "Could you provide me some papers discussing the use of natural language to refer target objects to robots for goal-conditioned manipulation tasks?", "answer": ["Language-Conditioned Imitation Learning for Robot Manipulation Tasks"], "answer_arxiv_id": ["2010.12083"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_20537"} +{"question": "What studies proposed label selection strategies for generating high-quality supervision?", "answer": ["FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning", "FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling", "DMT: Dynamic Mutual Training for Semi-Supervised Learning", "In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning", "Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic Segmentation", "Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation", "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence", "Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation"], "answer_arxiv_id": ["2205.07246", "2110.08263", "2004.08514", "2101.06329", "2003.03773", "2111.12903", "2001.07685v2", "2208.09910"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_20538"} +{"question": "Who investigated a generalized form of the term 1∧‖φ​(𝒙t)‖𝑿t−1−121 in pursuit of variance reduction in non-gaussian linear regression models?", "answer": ["The Elliptical Potential Lemma for General Distributions with an Application to Linear Thompson Sampling"], "answer_arxiv_id": ["2102.07987v3"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_20539"} +{"question": "Could you provide me studies showing the performance of instance discrimination task in self-supervised learning?", "answer": ["A Simple Framework for Contrastive Learning of Visual Representations", "An Empirical Study of Training Self-Supervised Vision Transformers", "SimCSE: Simple Contrastive Learning of Sentence Embeddings", "With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["2002.05709", "2104.02057", "2104.08821", "2104.14548"], "source_meta": {"published_time": "20220716"}, "qid": "AutoScholarQuery_train_20540"} +{"question": "Can you tell me about the research that demonstrated that LLMs can be easily distracted by Irrelevant context and proposed several approaches for filtering out irrelevant information?", "answer": ["Large Language Models Can Be Easily Distracted by Irrelevant Context"], "answer_arxiv_id": ["2302.00093"], "source_meta": {"published_time": "20230323"}, "qid": "AutoScholarQuery_train_20541"} +{"question": "What papers contributed to the field of image manipulation by providing techniques for editing through fine-tuning or changing cross-attention values at the inference stage?", "answer": ["An Image is Worth One Word: Personalizing Text-to-Image Generation using\n Textual Inversion", "Multi-Concept Customization of Text-to-Image Diffusion", "DreamBooth: Fine Tuning Text-to-Image Diffusion Models for\n Subject-Driven Generation", "TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Composition", "InstructPix2Pix: Learning to Follow Image Editing Instructions", "Prompt-to-Prompt Image Editing with Cross Attention Control", "Subject-Diffusion:Open Domain Personalized Text-to-Image Generation\n without Test-time Fine-tuning", "Compositional Text-to-Image Synthesis with Attention Map Control of\n Diffusion Models"], "answer_arxiv_id": ["2208.01618", "2212.04488", "2208.12242", "2307.12493", "2211.09800", "2208.01626", "2307.11410", "2305.13921"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_20542"} +{"question": "What researches use learning-based methods for lens blur rendering, particularly leveraging light field rendering and neural rendering?", "answer": ["Learning-Based View Synthesis for Light Field Cameras", "Aperture Supervision for Monocular Depth Estimation", "Aperture Supervision for Monocular Depth Estimation", "SteReFo: Efficient Image Refocusing with Stereo Vision", "Rendering Natural Camera Bokeh Effect with Deep Learning", "Deep Shading: Convolutional Neural Networks for Screen-Space Shading", "DeepLens: Shallow Depth Of Field From A Single Image", "MPIB: An MPI-Based Bokeh Rendering Framework for Realistic Partial\n Occlusion Effects", "BokehMe: When Neural Rendering Meets Classical Rendering"], "answer_arxiv_id": ["1609.02974", "1711.07933", "1711.07933", "1909.13395", "2006.05698", "1603.06078", "1810.08100", "2207.08403", "2206.12614"], "source_meta": {"published_time": "20230817"}, "qid": "AutoScholarQuery_train_20543"} +{"question": "Which paper introduces a proposal-level fusion framework that employs a Soft-Polar-Association and Spatio-Contextual Fusion Transformer?", "answer": ["CRAFT: Camera-Radar 3D Object Detection with Spatio-Contextual Fusion\n Transformer"], "answer_arxiv_id": ["2209.06535"], "source_meta": {"published_time": "20240325"}, "qid": "AutoScholarQuery_train_20544"} +{"question": "What studies employ sparse regularization on the scaling factors of batch normalization to facilitate channel pruning?", "answer": ["Learning Efficient Convolutional Networks through Network Slimming"], "answer_arxiv_id": ["1708.06519"], "source_meta": {"published_time": "20240321"}, "qid": "AutoScholarQuery_train_20545"} +{"question": "What work introduces sub-classes in the prompts based on the main class in images?", "answer": ["DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic\n Segmentation Using Diffusion Models"], "answer_arxiv_id": ["2303.11681"], "source_meta": {"published_time": "20231219"}, "qid": "AutoScholarQuery_train_20546"} +{"question": "What research describes UCGs being used in applications such as training recurrent neural networks, meta-training learned optimizers, and learning reinforcement learning policies?", "answer": ["Unbiased Gradient Estimation in Unrolled Computation Graphs with Persistent Evolution Strategies"], "answer_arxiv_id": ["2112.13835"], "source_meta": {"published_time": "20230421"}, "qid": "AutoScholarQuery_train_20547"} +{"question": "Which datasets require causal reasoning for answering the questions with regard to images?", "answer": ["From Recognition to Cognition: Visual Commonsense Reasoning", "VisualCOMET: Reasoning about the Dynamic Context of a Still Image"], "answer_arxiv_id": ["1811.10830", "2004.10796v3"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_20548"} +{"question": "What research has been done on regression-based methods for clothed 3D human recovery from a single image?", "answer": ["PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human\n Digitization", "PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution\n 3D Human Digitization", "PaMIR: Parametric Model-Conditioned Implicit Representation for\n Image-based Human Reconstruction", "ICON: Implicit Clothed humans Obtained from Normals", "ECON: Explicit Clothed humans Optimized via Normal integration", "ARCH: Animatable Reconstruction of Clothed Humans", "ARCH++: Animation-Ready Clothed Human Reconstruction Revisited", "Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view\n Human Reconstruction"], "answer_arxiv_id": ["1905.05172", "2004.00452", "2007.03858", "2112.09127", "2212.07422", "2004.04572", "2108.07845", "2006.08072"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20549"} +{"question": "What papers discuss the applications of Stable Diffusion models?", "answer": ["DiffEdit: Diffusion-based semantic image editing with mask guidance", "Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "UniControl: A Unified Diffusion Model for Controllable Visual Generation\n In the Wild", "T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Inversion-Based Style Transfer with Diffusion Models"], "answer_arxiv_id": ["2210.11427", "2211.12572", "2305.11147", "2302.08453", "2211.13203"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_20550"} +{"question": "What studies use value factorization approaches for decentralized execution of MARL policies?", "answer": ["Rethinking the Implementation Tricks and Monotonicity Constraint in Cooperative Multi-Agent Reinforcement Learning", "QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement learning", "Weighted QMIX: Expanding Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning"], "answer_arxiv_id": ["2102.03479", "1905.05408", "2006.10800"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_20551"} +{"question": "Which paper argues that specifying the target object via vectors and one-hot vectors is not flexible enough to support continued learning in deployed robots?", "answer": ["Language-Conditioned Imitation Learning for Robot Manipulation Tasks"], "answer_arxiv_id": ["2010.12083"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_20552"} +{"question": "What are the papers that have used Markov and Conditional Random Fields in computer vision applications?", "answer": ["Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials", "Conditional Random Fields as Recurrent Neural Networks", "Semantic Image Segmentation with Deep Convolutional Nets and Fully\n Connected CRFs"], "answer_arxiv_id": ["1210.5644v1", "1502.03240", "1412.7062"], "source_meta": {"published_time": "20230814"}, "qid": "AutoScholarQuery_train_20553"} +{"question": "What are some studies in the field of retrieval-augmented language models?", "answer": ["RePlug: Retrieval-Augmented Black-Box Language Models", "Improving language models by retrieving from trillions of tokens", "Few-shot Learning with Retrieval Augmented Language Models", "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks", "Unsupervised Cross-Task Generalization via Retrieval Augmentation", "ReSel: N-ary Relation Extraction from Scientific Text and Tables by Learning to Retrieve and Select", "ReACC: A Retrieval-Augmented Code Completion Framework", "Weakly-Supervised Scientific Document Classification via Retrieval-Augmented Multi-Stage Training"], "answer_arxiv_id": ["2301.12652", "2112.04426", "2208.03299", "2005.11401", "2204.07937", "2210.14427v1", "2203.07722", "2306.07193"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_20554"} +{"question": "Which work studies the use of standard knowledge distillation in Adversarial Distillation?", "answer": ["Adversarially Robust Distillation"], "answer_arxiv_id": ["1905.09747"], "source_meta": {"published_time": "20240311"}, "qid": "AutoScholarQuery_train_20555"} +{"question": "Can you list some works related to data augmentation in time series modeling?", "answer": ["Time Series Data Augmentation for Deep Learning: A Survey", "An Empirical Survey of Data Augmentation for Time Series Classification with Neural Networks"], "answer_arxiv_id": ["2002.12478", "2007.15951"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_20556"} +{"question": "Which works have explored language agents in simulation environments that mimic real-world scenarios?", "answer": ["SmartPlay: A Benchmark for LLMs as Intelligent Agents", "AgentBench: Evaluating LLMs as Agents", "MindAgent: Emergent Gaming Interaction", "Playing repeated games with Large Language Models"], "answer_arxiv_id": ["2310.01557", "2308.03688", "2309.09971", "2305.16867"], "source_meta": {"published_time": "20240208"}, "qid": "AutoScholarQuery_train_20557"} +{"question": "What work has been done to theoretically study the bottleneck structure used in MLPs?", "answer": ["Vector-Valued Variation Spaces and Width Bounds for DNNs: Insights on Weight Decay Regularization"], "answer_arxiv_id": ["2305.16534"], "source_meta": {"published_time": "20230623"}, "qid": "AutoScholarQuery_train_20558"} +{"question": "Any works about variational autoencoder models that allow to disentangle multiple factors?", "answer": ["Learning Hierarchical Features from Generative Models"], "answer_arxiv_id": ["1702.08396"], "source_meta": {"published_time": "20230330"}, "qid": "AutoScholarQuery_train_20559"} +{"question": "Please indicate papers where joint training of the generator and ranker were proposed within a multi-task framework.", "answer": ["Generate & Rank: A Multi-task Framework for Math Word Problems"], "answer_arxiv_id": ["2109.03034"], "source_meta": {"published_time": "20220721"}, "qid": "AutoScholarQuery_train_20560"} +{"question": "Which are the papers that presented methods for enhancing agent’s ability through multimodal capability?", "answer": ["Visual Programming: Compositional visual reasoning without training", "HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging\n Face", "Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation\n Models"], "answer_arxiv_id": ["2211.11559", "2303.17580", "2303.04671"], "source_meta": {"published_time": "20240214"}, "qid": "AutoScholarQuery_train_20561"} +{"question": "In what paper the researcher adopted implicit differentiation on the equilibrium state to train SNN?", "answer": ["Training Feedback Spiking Neural Networks by Implicit Differentiation on the Equilibrium State"], "answer_arxiv_id": ["2109.14247"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_20562"} +{"question": "Could you provide me some works related to one-shot models for graph generation?", "answer": ["Variational Graph Auto-Encoders", "Constrained Generation of Semantically Valid Graphs via Regularizing Variational Autoencoders", "Graphite: Iterative Generative Modeling of Graphs", "Permutation Invariant Graph Generation via Score-Based Generative Modeling"], "answer_arxiv_id": ["1611.07308", "1809.02630", "1803.10459", "2003.00638"], "source_meta": {"published_time": "20230125"}, "qid": "AutoScholarQuery_train_20563"} +{"question": "Which studies used the SSL techniques in areas such as weather and climate prediction?", "answer": ["ClimaX: A foundation model for weather and climate"], "answer_arxiv_id": ["2301.10343"], "source_meta": {"published_time": "20230711"}, "qid": "AutoScholarQuery_train_20564"} +{"question": "What studies addressed the straggling worker problem in deep neural network training using redundancy methods?", "answer": ["Revisiting Distributed Synchronous SGD", "Stochastic Gradient Coding for Straggler Mitigation in Distributed Learning"], "answer_arxiv_id": ["1604.00981", "1905.05383v1"], "source_meta": {"published_time": "20230618"}, "qid": "AutoScholarQuery_train_20565"} +{"question": "What work introduced large kernel design into 3D networks?", "answer": ["LargeKernel3D: Scaling up Kernels in 3D Sparse CNNs"], "answer_arxiv_id": ["2206.10555"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_20566"} +{"question": "Could you provide me research about state-of-the-art methods can efficiently calculate the determinant of the Jacobian dot product?", "answer": ["Rectangular Flows for Manifold Learning"], "answer_arxiv_id": ["2106.01413"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_20567"} +{"question": "Could you provide me a study where variational method was applied for complex scenarios?", "answer": ["The Deep Ritz method: A deep learning-based numerical algorithm for solving variational problems"], "answer_arxiv_id": ["1710.00211"], "source_meta": {"published_time": "20230223"}, "qid": "AutoScholarQuery_train_20568"} +{"question": "Which study proposed the RUST method that can be trained for novel view synthesis without access to camera poses?", "answer": ["RUST: Latent Neural Scene Representations from Unposed Imagery"], "answer_arxiv_id": ["2211.14306"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_20569"} +{"question": "What work applies regional guidance during the sampling process using merged weights to resolve the issue of concept blending?", "answer": ["Mix-of-Show: Decentralized Low-Rank Adaptation for Multi-Concept\n Customization of Diffusion Models"], "answer_arxiv_id": ["2305.18292"], "source_meta": {"published_time": "20240405"}, "qid": "AutoScholarQuery_train_20570"} +{"question": "What papers explored differential neural architecture search methods to learn the bit-width?", "answer": ["Single Path One-Shot Neural Architecture Search with Uniform Sampling", "Rethinking Differentiable Search for Mixed-Precision Neural Networks", "Search What You Want: Barrier Panelty NAS for Mixed Precision\n Quantization", "Generalizable Mixed-Precision Quantization via Attribution Rank\n Preservation", "SEAM: Searching Transferable Mixed-Precision Quantization Policy through\n Large Margin Regularization"], "answer_arxiv_id": ["1904.00420", "2004.05795", "2007.10026", "2108.02720", "2302.06845"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_20571"} +{"question": "Which works improve task performance of a deep learning model by training on multiple datasets, especially in depth estimation?", "answer": ["Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer", "Omnivore: A Single Model for Many Visual Modalities"], "answer_arxiv_id": ["1907.01341", "2201.08377"], "source_meta": {"published_time": "20231108"}, "qid": "AutoScholarQuery_train_20572"} +{"question": "What works introduce directed hypergraph ORC?", "answer": ["Ollivier Ricci Curvature of Directed Hypergraphs"], "answer_arxiv_id": ["1907.04727"], "source_meta": {"published_time": "20221021"}, "qid": "AutoScholarQuery_train_20573"} +{"question": "Which papers have attempted to keep the learned policy within the support of the behavior policy by Maximum Mean Discrepancy (MMD) or explicit density estimation in Offline RL?", "answer": ["Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction", "Supported Policy Optimization for Offline Reinforcement Learning"], "answer_arxiv_id": ["1906.00949", "2202.06239"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_20574"} +{"question": "What datasets include object attributes but are relatively small and focus on crowdsourcing the labels?", "answer": ["Semantic Jitter: Dense Supervision for Visual Comparisons via Synthetic Images"], "answer_arxiv_id": ["1612.06341"], "source_meta": {"published_time": "20230429"}, "qid": "AutoScholarQuery_train_20575"} +{"question": "Could you tell me the research papers that discuss the use of re-sampling in class-imbalanced learning?", "answer": ["SMOTE: Synthetic Minority Over-sampling Technique"], "answer_arxiv_id": ["1106.1813"], "source_meta": {"published_time": "20231027"}, "qid": "AutoScholarQuery_train_20576"} +{"question": "Can you name a paper where transformer-based pre-trained language models were used to improve Answer Sentence Selection?", "answer": ["TANDA: Transfer and Adapt Pre-Trained Transformer Models for Answer\n Sentence Selection"], "answer_arxiv_id": ["1911.04118"], "source_meta": {"published_time": "20240812"}, "qid": "AutoScholarQuery_train_20577"} +{"question": "Could you provide me some studies on the sub-problem of tensor network-rank selection (TN-RS)?", "answer": ["Bayesian CP Factorization of Incomplete Tensors with Automatic Rank Determination", "Smooth PARAFAC Decomposition for Tensor Completion", "On Algorithms for and Computing with the Tensor Ring Decomposition", "Bayesian Tensorized Neural Networks with Automatic Rank Selection", "Approximately Optimal Core Shapes for Tensor Decompositions"], "answer_arxiv_id": ["1401.6497", "1505.06611", "1807.02513", "1905.10478", "2302.03886"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_20578"} +{"question": "What works have been utilized in the application of Vision-Language Models, such as text-driven image or video generation and captioning?", "answer": ["Zero-Shot Text-to-Image Generation", "Photorealistic Text-to-Image Diffusion Models with Deep Language\n Understanding", "High-Resolution Image Synthesis with Latent Diffusion Models", "eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert\n Denoisers", "Scaling Autoregressive Models for Content-Rich Text-to-Image Generation", "MagicVideo: Efficient Video Generation With Latent Diffusion Models", "Imagen Video: High Definition Video Generation with Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "BLIP: Bootstrapping Language-Image Pre-training for Unified\n Vision-Language Understanding and Generation", "BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image\n Encoders and Large Language Models", "Flamingo: a Visual Language Model for Few-Shot Learning", "Visual Instruction Tuning", "Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense\n Video Captioning", "End-to-end Generative Pretraining for Multimodal Video Captioning"], "answer_arxiv_id": ["2102.12092", "2205.11487", "2112.10752", "2211.01324", "2206.10789", "2211.11018", "2210.02303", "2209.14792", "2304.08818", "2201.12086", "2301.12597", "2204.14198", "2304.08485", "2302.14115", "2201.08264"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20579"} +{"question": "Can you provide me papers that focus on tabular or linear models in offline RL?", "answer": ["Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism", "Pessimistic Q-Learning for Offline Reinforcement Learning: Towards Optimal Sample Complexity", "Towards Instance-Optimal Offline Reinforcement Learning with Pessimism", "Is Pessimism Provably Efficient for Offline RL?", "Corruption-Robust Offline Reinforcement Learning", "On Instance-Dependent Bounds for Offline Reinforcement Learning with Linear Function Approximation", "Pessimistic Bootstrapping for Uncertainty-Driven Offline Reinforcement Learning"], "answer_arxiv_id": ["2103.12021v2", "2202.13890", "2110.08695v1", "2012.15085", "2106.06630", "2211.13208", "2202.11566"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_20580"} +{"question": "Which researches discuss the selection-based methods of PEFT that involve only fine-tuning bias?", "answer": ["BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based\n Masked Language-models"], "answer_arxiv_id": ["2106.10199"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_20581"} +{"question": "What studies have focused on fairness within each attribute to prevent bias from affecting the final decision?", "answer": ["Equality of Opportunity in Supervised Learning", "On Fairness and Calibration", "A Reductions Approach to Fair Classification"], "answer_arxiv_id": ["1610.02413", "1709.02012", "1803.02453"], "source_meta": {"published_time": "20220531"}, "qid": "AutoScholarQuery_train_20582"} +{"question": "What research discusses the concept of phases in the forward process of diffusion models?", "answer": ["Stable target field for reduced variance score estimation in diffusion models", "Elucidating the Design Space of Diffusion-Based Generative Models", "Perception Prioritized Training of Diffusion Models", "Tackling the Generative Learning Trilemma with Denoising Diffusion GANs"], "answer_arxiv_id": ["2302.00670", "2206.00364", "2204.00227", "2112.07804"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_20583"} +{"question": "Which study validated a more robust lottery ticket hypothesis, essentially asserting an exceptionally high expressive power?", "answer": ["Proving the Lottery Ticket Hypothesis: Pruning is All You Need"], "answer_arxiv_id": ["2002.00585"], "source_meta": {"published_time": "20200819"}, "qid": "AutoScholarQuery_train_20584"} +{"question": "Could you provide me some studies that employed prefix tuning as part of their PETL methods?", "answer": ["Prefix-Tuning: Optimizing Continuous Prompts for Generation"], "answer_arxiv_id": ["2101.00190"], "source_meta": {"published_time": "20240303"}, "qid": "AutoScholarQuery_train_20585"} +{"question": "Which paper cites the issue with policy gradient methods where they cannot be extended to the offline RL settings due to the distribution shift problem?", "answer": ["Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems"], "answer_arxiv_id": ["2005.01643"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_20586"} +{"question": "What papers reported on the invention of a novel conditional method to direct DDPM generation towards reference images?", "answer": ["ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models"], "answer_arxiv_id": ["2108.02938"], "source_meta": {"published_time": "20220316"}, "qid": "AutoScholarQuery_train_20587"} +{"question": "Which studies explored optimization theory for deep learning, specifically how first order methods identify the optimum of neural networks?", "answer": ["Gradient Descent Finds Global Minima of Deep Neural Networks", "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks", "Gradient Descent Provably Optimizes Over-parameterized Neural Networks", "A Convergence Theory for Deep Learning via Over-Parameterization", "Learning and Generalization in Overparameterized Neural Networks, Going Beyond Two Layers", "Convergence Analysis of Two-layer Neural Networks with ReLU Activation"], "answer_arxiv_id": ["1811.03804", "1901.08584", "1810.02054", "1811.03962", "1811.04918", "1705.09886"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_20588"} +{"question": "What works have introduced benchmarks for web navigation tasks?", "answer": ["A data-driven approach for learning to control computers", "WebShop: Towards Scalable Real-World Web Interaction with Grounded\n Language Agents", "Mapping Natural Language Instructions to Mobile UI Action Sequences", "META-GUI: Towards Multi-modal Conversational Agents on Mobile GUI", "A Dataset for Interactive Vision-Language Navigation with Unknown\n Command Feasibility"], "answer_arxiv_id": ["2202.08137", "2207.01206", "2005.03776", "2205.11029", "2202.02312"], "source_meta": {"published_time": "20240206"}, "qid": "AutoScholarQuery_train_20589"} +{"question": "Which papers contributed to the task of code generation in software development?", "answer": ["Evaluating Large Language Models Trained on Code", "Program Synthesis with Large Language Models", "RepoCoder: Repository-Level Code Completion Through Iterative Retrieval\n and Generation", "McEval: Massively Multilingual Code Evaluation"], "answer_arxiv_id": ["2107.03374", "2108.07732", "2303.12570", "2406.07436v1"], "source_meta": {"published_time": "20240624"}, "qid": "AutoScholarQuery_train_20590"} +{"question": "Any studies on contrastive learning during pre-training of VLMs?", "answer": ["Just Ask: Learning to Answer Questions from Millions of Narrated Videos", "Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval"], "answer_arxiv_id": ["2012.00451", "2104.00650"], "source_meta": {"published_time": "20240103"}, "qid": "AutoScholarQuery_train_20591"} +{"question": "What papers suggest that unstable convergence occurs when the loss landscape of neural networks forms a local forward-invariant set near the minima?", "answer": ["Understanding the Unstable Convergence of Gradient Descent"], "answer_arxiv_id": ["2204.01050"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_20592"} +{"question": "Bearing multicompartment-learning in mind, which works were based on it and hence, demonstrated the significance of external contexts?", "answer": ["MIMIC-IT: Multi-Modal In-Context Instruction Tuning", "MMICL: Empowering Vision-language Model with Multi-Modal In-Context\n Learning", "Generative Multimodal Models are In-Context Learners", "Fine-tuning Multimodal LLMs to Follow Zero-shot Demonstrative\n Instructions"], "answer_arxiv_id": ["2306.05425", "2309.07915", "2312.13286", "2308.04152"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_20593"} +{"question": "Are there any research papers on improving controllability by focusing on motion generation conditioned on different types of goals?", "answer": ["ActFormer: A GAN-based Transformer towards General Action-Conditioned 3D\n Human Motion Generation", "Action-Conditioned 3D Human Motion Synthesis with Transformer VAE"], "answer_arxiv_id": ["2203.07706", "2104.05670"], "source_meta": {"published_time": "20240423"}, "qid": "AutoScholarQuery_train_20594"} +{"question": "Could you provide any synthetic datasets that are widely used for ground truth benchmarking?", "answer": ["Learning Non-Lambertian Object Intrinsics across ShapeNet Categories", "NeRD: Neural Reflectance Decomposition from Image Collections", "De-rendering the World’s Revolutionary Artefacts", "De-rendering 3D Objects in the Wild", "ABO: Dataset and Benchmarks for Real-World 3D Object Understanding"], "answer_arxiv_id": ["1612.08510v1", "2012.03918", "2104.03954", "2201.02279", "2110.06199"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_20595"} +{"question": "Can you provide any research that focused on the adversarial regime with a self-bounding constraint in bandit problems?", "answer": ["Tsallis-INF: An Optimal Algorithm for Stochastic and Adversarial Bandits"], "answer_arxiv_id": ["1807.07623"], "source_meta": {"published_time": "20230209"}, "qid": "AutoScholarQuery_train_20596"} +{"question": "Which papers highlight that a naive approach in Predict+Optimize is prone to produce highly suboptimal models?", "answer": ["The Offset Tree for Learning with Partial Labels", "Melding the Data-Decisions Pipeline: Decision-Focused Learning for Combinatorial Optimization"], "answer_arxiv_id": ["0812.4044", "1809.05504"], "source_meta": {"published_time": "20230718"}, "qid": "AutoScholarQuery_train_20597"} +{"question": "What approach does ATVHunter, a representative approach for identifying vulnerable Android library versions, adopt?", "answer": ["ATVHunter: Reliable Version Detection of Third-Party Libraries for\n Vulnerability Identification in Android Applications"], "answer_arxiv_id": ["2102.08172"], "source_meta": {"published_time": "20230809"}, "qid": "AutoScholarQuery_train_20598"} +{"question": "What papers mentioned that diffusion models excel in controllable generation?", "answer": ["T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for\n Text-to-Image Diffusion Models", "Adding Conditional Control to Text-to-Image Diffusion Models", "ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models", "Diffusion Models Beat GANs on Image Synthesis", "LoRA: Low-Rank Adaptation of Large Language Models"], "answer_arxiv_id": ["2302.08453", "2302.05543", "2108.02938", "2105.05233", "2106.09685"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_20599"} +{"question": "What works have used planning through the learned world model for model-based RL?", "answer": ["Recurrent World Models Facilitate Policy Evolution", "Mastering Atari with Discrete World Models"], "answer_arxiv_id": ["1809.01999", "2010.02193"], "source_meta": {"published_time": "20231214"}, "qid": "AutoScholarQuery_train_20600"} +{"question": "Which papers discussed scalable graph transformer architectures?", "answer": ["Exphormer: Sparse Transformers for Graphs"], "answer_arxiv_id": ["2303.06147"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20601"} +{"question": "What research articles focus on single-turn evaluation for LLM-based chat models?", "answer": ["Measuring Massive Multitask Language Understanding", "AGIEval: A Human-Centric Benchmark for Evaluating Foundation Models", "Beyond the Imitation Game: Quantifying and extrapolating the\n capabilities of language models"], "answer_arxiv_id": ["2009.03300", "2304.06364", "2206.04615"], "source_meta": {"published_time": "20231011"}, "qid": "AutoScholarQuery_train_20602"} +{"question": "Could you provide me some studies on achieving group robustness without group labels?", "answer": ["No Subclass Left Behind: Fine-Grained Robustness in Coarse-Grained Classification Problems", "Environment Inference for Invariant Learning", "Learning to Split for Automatic Bias Detection"], "answer_arxiv_id": ["2011.12945", "2010.07249", "2204.13749"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_20603"} +{"question": "Which paper mentions the effects of privacy-preserving techniques on the convergence of model weights?", "answer": ["Do not Let Privacy Overbill Utility: Gradient Embedding Perturbation for Private Learning"], "answer_arxiv_id": ["2102.12677"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_20604"} +{"question": "What are some works about the use of white-box optimization to solve SAE problems?", "answer": ["OptNet: Differentiable Optimization as a Layer in Neural Networks", "Momentum-Net: Fast and convergent iterative neural network for inverse problems"], "answer_arxiv_id": ["1703.00443", "1907.11818"], "source_meta": {"published_time": "20230925"}, "qid": "AutoScholarQuery_train_20605"} +{"question": "Could you provide me some studies that require more direct supervision in causal representation learning?", "answer": ["Weakly Supervised Disentangled Generative Causal Representation Learning"], "answer_arxiv_id": ["2010.02637"], "source_meta": {"published_time": "20230601"}, "qid": "AutoScholarQuery_train_20606"} +{"question": "Which researchers employed learnable queries for personalized segmentation in order to represent expert annotations?", "answer": ["Transformer-based Annotation Bias-aware Medical Image Segmentation"], "answer_arxiv_id": ["2306.01340"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_20607"} +{"question": "Any works delving into Open-set Semi-supervised Learning, considering that unseen classes may exist in training data?", "answer": ["Trash to Treasure: Harvesting OOD Data with Cross-Modal Matching for Open-Set Semi-Supervised Learning"], "answer_arxiv_id": ["2108.05617"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_20608"} +{"question": "What works incorporate foundational language models into downstream tasks?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language\n Understanding", "Language Models are Few-Shot Learners"], "answer_arxiv_id": ["1810.04805", "2005.14165"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_20609"} +{"question": "What research work proposed parallel or 'batched' models like temporal CNNs and transformers?", "answer": ["An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling", "Attention Is All You Need"], "answer_arxiv_id": ["1803.01271", "1706.03762"], "source_meta": {"published_time": "20231006"}, "qid": "AutoScholarQuery_train_20610"} +{"question": "What studies have provided a detailed examination of various forms of invariance-based methods derived from underlying causal graphs?", "answer": ["Counterfactual Invariance to Spurious Correlations: Why and How to Pass Stress Tests"], "answer_arxiv_id": ["2106.00545"], "source_meta": {"published_time": "20231221"}, "qid": "AutoScholarQuery_train_20611"} +{"question": "What work is a variant of the plane sweep algorithm that constructs a plane sweep volumes from deep features?", "answer": ["MVSNet: Depth Inference for Unstructured Multi-view Stereo"], "answer_arxiv_id": ["1804.02505"], "source_meta": {"published_time": "20231213"}, "qid": "AutoScholarQuery_train_20612"} +{"question": "Which study presents a method that jointly forecasts both human and object motion sequences given an initial sequence observation?", "answer": ["InterDiff: Generating 3D Human-Object Interactions with Physics-Informed\n Diffusion"], "answer_arxiv_id": ["2308.16905"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_20613"} +{"question": "Which study proposed the Max Sliced Wasserstein (MSW) distance?", "answer": ["Max-Sliced Wasserstein Distance and its use for GANs"], "answer_arxiv_id": ["1904.05877"], "source_meta": {"published_time": "20201028"}, "qid": "AutoScholarQuery_train_20614"} +{"question": "What studies use cross-attention for semantic segmentation while grouping tokens?", "answer": ["GroupViT: Semantic Segmentation Emerges from Text Supervision"], "answer_arxiv_id": ["2202.11094"], "source_meta": {"published_time": "20221017"}, "qid": "AutoScholarQuery_train_20615"} +{"question": "Which studies have discussed the concept of spectral bias in multi-layer perceptron (MLP)?", "answer": ["On the Spectral Bias of Neural Networks", "Understanding training and generalization in deep learning by Fourier\n analysis", "A Closer Look at Memorization in Deep Networks", "The Convergence Rate of Neural Networks for Learned Functions of\n Different Frequencies", "On the Inductive Bias of Neural Tangent Kernels", "Fourier Features Let Networks Learn High Frequency Functions in Low\n Dimensional Domains"], "answer_arxiv_id": ["1806.08734", "1808.04295", "1706.05394", "1906.00425", "1905.12173", "2006.10739"], "source_meta": {"published_time": "20240115"}, "qid": "AutoScholarQuery_train_20616"} +{"question": "What research papers proposed methods that involve the model rating its own uncertainty?", "answer": ["Teaching Models to Express Their Uncertainty in Words", "Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence\n Scores from Language Models Fine-Tuned with Human Feedback"], "answer_arxiv_id": ["2205.14334", "2305.14975"], "source_meta": {"published_time": "20240309"}, "qid": "AutoScholarQuery_train_20617"} +{"question": "Can you list the papers that introduced the reinforcement-learning algorithm for interpretable decision making?", "answer": ["Recurrent Models of Visual Attention", "A Probabilistic Hard Attention Model For Sequentially Observed Scenes", "EDDI: Efficient Dynamic Discovery of High-Value Information with Partial VAE"], "answer_arxiv_id": ["1406.6247", "2111.07534", "1809.11142"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_20618"} +{"question": "What was the first work to apply an encoder-only Transformer architecture to non-overlapping image patches in an image classification task?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at\n Scale"], "answer_arxiv_id": ["2010.11929"], "source_meta": {"published_time": "20230527"}, "qid": "AutoScholarQuery_train_20619"} +{"question": "Is there any research on the concept of learnable interpolators in the context of collocation methods?", "answer": ["MAgNet: Mesh Agnostic Neural PDE Solver"], "answer_arxiv_id": ["2210.05495"], "source_meta": {"published_time": "20231025"}, "qid": "AutoScholarQuery_train_20620"} +{"question": "Which work provided enhancements to WebSplit dataset by augmenting the LSTM with a copy mechanism?", "answer": ["Incorporating Copying Mechanism in Sequence-to-Sequence Learning"], "answer_arxiv_id": ["1603.06393"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_20621"} +{"question": "Which research works illustrate that the accuracy of image classification models improves with larger data sizes?", "answer": ["Revisiting Unreasonable Effectiveness of Data in Deep Learning Era"], "answer_arxiv_id": ["1707.02968"], "source_meta": {"published_time": "20230915"}, "qid": "AutoScholarQuery_train_20622"} +{"question": "What papers have applied SAM to challenging situations such as medical images, camouflaged and transparent objects?", "answer": ["Segment Anything Model for Medical Image Analysis: an Experimental Study", "Segment Anything in Medical Images", "Can SAM Segment Anything? When SAM Meets Camouflaged Object Detection", "SAM Struggles in Concealed Scenes -- Empirical Study on \"Segment\n Anything\"", "Segment Anything Model (SAM) Meets Glass: Mirror and Transparent Objects\n Cannot Be Easily Detected", "Segment Anything Is Not Always Perfect: An Investigation of SAM on\n Different Real-world Applications"], "answer_arxiv_id": ["2304.10517", "2304.12306", "2304.04709", "2304.06022", "2305.00278", "2304.05750"], "source_meta": {"published_time": "20231127"}, "qid": "AutoScholarQuery_train_20623"} +{"question": "Can you provide examples of research that developed GNN models where message passing operations are used to augment node features?", "answer": ["Simplifying Graph Convolutional Networks", "Are Powerful Graph Neural Nets Necessary? A Dissection on Graph Classification", "SIGN: Scalable Inception Graph Neural Networks", "On Graph Neural Networks versus Graph-Augmented MLPs"], "answer_arxiv_id": ["1902.07153", "1905.04579", "2004.11198", "2010.15116"], "source_meta": {"published_time": "20221218"}, "qid": "AutoScholarQuery_train_20624"} +{"question": "Which research proposed Res-FFT-ReLU-Block for frequency selection in image deblurring?", "answer": ["Intriguing Findings of Frequency Selection for Image Deblurring"], "answer_arxiv_id": ["2111.11745"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_20625"} +{"question": "What works propose a unified optimal transport-based framework for encouraging both global cluster discrimination and local consistency of samples?", "answer": ["Unified Optimal Transport Framework for Universal Domain Adaptation"], "answer_arxiv_id": ["2210.17067"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_20626"} +{"question": "Any works analyze label noise dynamics in the central phase of the training?", "answer": ["Label noise (stochastic) gradient descent implicitly solves the Lasso for quadratic parametrisation"], "answer_arxiv_id": ["2206.09841"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_20627"} +{"question": "Which papers proposed methods to denoise a single image in test-time training approaches?", "answer": ["Deep Image Prior", "Score Priors Guided Deep Variational Inference for Unsupervised\n Real-World Single Image Denoising"], "answer_arxiv_id": ["1711.10925", "2308.04682"], "source_meta": {"published_time": "20240415"}, "qid": "AutoScholarQuery_train_20628"} +{"question": "Which papers discuss the perspective of group fairness metrics in terms of confusion matrices for each subgroup?", "answer": ["FACT: A Diagnostic for Group Fairness Trade-offs", "Fairness in Criminal Justice Risk Assessments: The State of the Art"], "answer_arxiv_id": ["2004.03424", "1703.09207"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_20629"} +{"question": "Which papers deal with Concept Activation Vectors, Concept Bottleneck Models, and Concept Embedding Models for concept-based explanations?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)", "Concept Bottleneck Models"], "answer_arxiv_id": ["1711.11279", "2007.04612"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_20630"} +{"question": "What studies have adapted transformer-based models, which have shown remarkable success in various vision tasks, to temporal action detection?", "answer": ["ActionFormer: Localizing Moments of Actions with Transformers"], "answer_arxiv_id": ["2202.07925"], "source_meta": {"published_time": "20240329"}, "qid": "AutoScholarQuery_train_20631"} +{"question": "Which work introduced the transformer model originally used for NLP tasks?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20230329"}, "qid": "AutoScholarQuery_train_20632"} +{"question": "Which research proposed Token Merging (ToMe) that helps in reducing token count?", "answer": ["Token Merging: Your ViT But Faster"], "answer_arxiv_id": ["2210.09461"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_20633"} +{"question": "Which studies discuss about measures to robustify deep neural networks?", "answer": ["Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Parseval Networks: Improving Robustness to Adversarial Examples", "Theoretically Principled Trade-off between Robustness and Accuracy", "DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensembles", "Building Robust Ensembles via Margin Boosting", "Evaluating Robustness of Neural Networks with Mixed Integer Programming", "Training for Faster Adversarial Robustness Verification via Inducing ReLU Stability", "Certified Defenses against Adversarial Examples", "Randomized Smoothing of All Shapes and Sizes"], "answer_arxiv_id": ["1702.01135", "1706.06083", "1704.08847", "1901.08573", "2009.14720", "2206.03362", "1711.07356", "1809.03008", "1801.09344", "2002.08118"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_20634"} +{"question": "Could you provide me some examples of specific-domain CSC datasets?", "answer": ["MCSCSet: A Specialist-annotated Dataset for Medical-domain Chinese\n Spelling Correction", "General and Domain Adaptive Chinese Spelling Check with Error Consistent\n Pretraining", "Rethinking Masked Language Modeling for Chinese Spelling Correction"], "answer_arxiv_id": ["2210.11720", "2203.10929", "2305.17721"], "source_meta": {"published_time": "20231119"}, "qid": "AutoScholarQuery_train_20635"} +{"question": "What studies propose models addressing catastrophic forgetting using regularization-based methods?", "answer": ["Overcoming catastrophic forgetting in neural networks", "Continual Learning Through Synaptic Intelligence"], "answer_arxiv_id": ["1612.00796", "1703.04200"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20636"} +{"question": "Can you tell me papers that shows the application of Graph Neural Networks in the field of biochemistry?", "answer": ["Neural Message Passing for Quantum Chemistry", "Conditional Graph Information Bottleneck for Molecular Relational Learning", "Shift-Robust Molecular Relational Learning with Causal Substructure"], "answer_arxiv_id": ["1704.01212", "2305.01520", "2305.18451"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_20637"} +{"question": "Could you point to a research paper that introduces a generative model capable of producing fake news?", "answer": ["Defending Against Neural Fake News"], "answer_arxiv_id": ["1905.12616"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_20638"} +{"question": "What studies focus on the continuous-time diffusion models that unify the denoising diffusion model and the denoising score matching models?", "answer": ["Score-Based Generative Modeling through Stochastic Differential Equations", "Generative Modeling by Estimating Gradients of the Data Distribution"], "answer_arxiv_id": ["2011.13456", "1907.05600"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_20639"} +{"question": "What datasets have been created for building functional use classification?", "answer": ["Building Instance Classification Using Street View Images", "Bounding Boxes Are All We Need: Street View Image Classification via\n Context Encoding of Detected Buildings"], "answer_arxiv_id": ["1802.09026", "2010.01305"], "source_meta": {"published_time": "20231009"}, "qid": "AutoScholarQuery_train_20640"} +{"question": "Whose work is about constructing an equivariant update function based on spherical harmonics in tensor field networks?", "answer": ["Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds"], "answer_arxiv_id": ["1802.08219"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_20641"} +{"question": "Which works focus on strategic learning and classification in online learning settings?", "answer": ["Strategic Classification from Revealed Preferences", "Learning Strategy-Aware Linear Classifiers"], "answer_arxiv_id": ["1710.07887", "1911.04004"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_20642"} +{"question": "What research has been done in constrained counting and sampling in the field of neuro-symbolic learning?", "answer": ["DeepProbLog: Neural Probabilistic Logic Programming", "A Semantic Loss Function for Deep Learning with Symbolic Knowledge", "A Knowledge Compilation Map", "A Scalable Approximate Model Counter", "Constrained Sampling and Counting: Universal Hashing Meets SAT Solving", "Uniform Solution Sampling Using a Constraint Solver As an Oracle", "Approximate Counting, the Lovász Local Lemma and Inference in Graphical Models", "Fast sampling and counting k-SAT solutions in the local lemma regime"], "answer_arxiv_id": ["1805.10872", "1711.11157", "1106.1819v1", "1306.5726v3", "1512.06633v1", "1210.4861v1", "1610.04317", "1911.01319v1"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_20643"} +{"question": "Which dataset came with a 3D scan?", "answer": ["Ego4D: Around the World in 3,000 Hours of Egocentric Video"], "answer_arxiv_id": ["2110.07058"], "source_meta": {"published_time": "20230614"}, "qid": "AutoScholarQuery_train_20644"} +{"question": "What research used the SimCSE and GPT models for evaluating the C-STS dataset?", "answer": ["SimCSE: Simple Contrastive Learning of Sentence Embeddings", "Language Models are Few-Shot Learners", "GPT-4 Technical Report"], "answer_arxiv_id": ["2104.08821", "2005.14165", "2303.08774"], "source_meta": {"published_time": "20240606"}, "qid": "AutoScholarQuery_train_20645"} +{"question": "Which publications discuss the improvement of the generalization ability of neural networks through flat minimizers?", "answer": ["Fantastic Generalization Measures and Where to Find Them", "Relative Flatness and Generalization", "Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data"], "answer_arxiv_id": ["1912.02178", "2001.00939", "1703.11008"], "source_meta": {"published_time": "20230607"}, "qid": "AutoScholarQuery_train_20646"} +{"question": "What work proposed the ProtoPNet method?", "answer": ["This Looks Like That: Deep Learning for Interpretable Image Recognition"], "answer_arxiv_id": ["1806.10574"], "source_meta": {"published_time": "20240413"}, "qid": "AutoScholarQuery_train_20647"} +{"question": "Which research train models specifically to solve tasks such as updating information, fixing grammar errors, or improving citations?", "answer": ["FRUIT : Faithfully Reflecting Updated Information in Text", "JFLEG: A Fluency Corpus and Benchmark for Grammatical Error Correction", "Parallel Iterative Edit Models for Local Sequence Transduction", "Improving Wikipedia Verifiability with AI"], "answer_arxiv_id": ["2112.08634", "1702.04066", "1910.02893", "2207.06220"], "source_meta": {"published_time": "20220824"}, "qid": "AutoScholarQuery_train_20648"} +{"question": "Which studies proposed the planning in robust Reinforcement Learning (RL)?", "answer": ["Policy Gradient Method For Robust Reinforcement Learning"], "answer_arxiv_id": ["2205.07344"], "source_meta": {"published_time": "20230516"}, "qid": "AutoScholarQuery_train_20649"} +{"question": "What works have been done on transformers in the context of graph learning?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_20650"} +{"question": "Which research papers proposed loss functions for top-k accuracy?", "answer": ["Top-k Multiclass SVM", "Loss Functions for Top-k Error: Analysis and Insights", "Smooth Loss Functions for Deep Top-k Classification", "Differentiable Top-k Classification Learning"], "answer_arxiv_id": ["1511.06683", "1512.00486", "1802.07595", "2206.07290v1"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_20651"} +{"question": "What papers are about applying recurrent nets to tackle POMDPs?", "answer": ["Memory-based control with recurrent neural networks", "Learning to reinforcement learn", "Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPs", "Memory-based Deep Reinforcement Learning for POMDPs", "Recurrent Off-policy Baselines for Memory-based Continuous Control", "Open-Ended Learning Leads to Generally Capable Agents", "Legged Locomotion in Challenging Terrains using Egocentric Vision"], "answer_arxiv_id": ["1512.04455", "1611.05763", "2110.05038", "2102.12344", "2110.12628", "2107.12808", "2211.07638"], "source_meta": {"published_time": "20230712"}, "qid": "AutoScholarQuery_train_20652"} +{"question": "What studies focus on neural network robustness against out-of-domain samples?", "answer": ["Masked Autoencoders Are Scalable Vision Learners", "Discrete Representations Strengthen Vision Transformer Robustness", "The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution\n Generalization", "AugMix: A Simple Data Processing Method to Improve Robustness and\n Uncertainty"], "answer_arxiv_id": ["2111.06377", "2111.10493", "2006.16241", "1912.02781"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_20653"} +{"question": "Which studies connected volume-based geometric regularization to ASC- and SSC-like conditions to enhance NMF identifiability?", "answer": ["Robust Volume Minimization-Based Matrix Factorization for Remote Sensing and Document Clustering", "Identifiability of the Simplex Volume Minimization Criterion for Blind Hyperspectral Unmixing: The No Pure-Pixel Case"], "answer_arxiv_id": ["1608.04290v1", "1406.5273"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20654"} +{"question": "Which studies proposed the gradient-based explanation methods meant for modifying the input in deep learning models?", "answer": ["SmoothGrad: removing noise by adding noise", "Axiomatic Attribution for Deep Networks", "Attribution in Scale and Space"], "answer_arxiv_id": ["1706.03825", "1703.01365", "2004.03383"], "source_meta": {"published_time": "20220610"}, "qid": "AutoScholarQuery_train_20655"} +{"question": "What papers discuss different methods using the 'frozen CLIP' paradigm for zero-shot segmentation?", "answer": ["Decoupling Zero-Shot Semantic Segmentation", "A Simple Baseline for Open-Vocabulary Semantic Segmentation with Pre-trained Vision-language Model", "FreeSeg: Unified, Universal and Open-Vocabulary Image Segmentation", "Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP", "Scaling Open-Vocabulary Image Segmentation with Image-Level Labels"], "answer_arxiv_id": ["2112.07910", "2112.14757", "2303.17225", "2210.04150", "2112.12143"], "source_meta": {"published_time": "20230930"}, "qid": "AutoScholarQuery_train_20656"} +{"question": "Which papers propose methods for detecting Out-of-distribution (OOD) data?", "answer": ["A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks", "Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks", "A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks", "Energy-based Out-of-distribution Detection", "Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samples", "Deep Anomaly Detection with Outlier Exposure", "Unsupervised Out-of-Distribution Detection by Maximum Classifier Discrepancy", "POEM: Out-of-Distribution Detection with Posterior Sampling"], "answer_arxiv_id": ["1610.02136", "1706.02690", "1807.03888", "2010.03759", "1711.09325", "1812.04606", "1908.04951", "2206.13687"], "source_meta": {"published_time": "20231016"}, "qid": "AutoScholarQuery_train_20657"} +{"question": "Could you provide me some studies about low-rank parameter factorization?", "answer": ["Exploiting Linear Structure Within Convolutional Networks for Efficient\n Evaluation", "Speeding up Convolutional Neural Networks with Low Rank Expansions", "Convolutional neural networks with low-rank regularization"], "answer_arxiv_id": ["1404.0736", "1405.3866", "1511.06067"], "source_meta": {"published_time": "20230411"}, "qid": "AutoScholarQuery_train_20658"} +{"question": "Which works are about foundation models trained on web-scale datasets using self-supervised learning?", "answer": ["BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", "Language Models are Few-Shot Learners", "High-Resolution Image Synthesis with Latent Diffusion Models", "Zero-Shot Text-to-Image Generation", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1810.04805", "2005.14165", "2112.10752", "2102.12092", "2103.00020"], "source_meta": {"published_time": "20231130"}, "qid": "AutoScholarQuery_train_20659"} +{"question": "Could you tell me the researches that studied and instantiated the concept of conditional GFlowNet for MOO?", "answer": ["Multi-Objective GFlowNets"], "answer_arxiv_id": ["2210.12765"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_20660"} +{"question": "Could you provide me some works about VAE-based models that predict multiple futures by sampling multiple latent vectors?", "answer": ["The Trajectron: Probabilistic Multi-Agent Trajectory Modeling With Dynamic Spatiotemporal Graphs", "Trajectron++: Dynamically-Feasible Trajectory Forecasting With Heterogeneous Data", "Multiple Futures Prediction"], "answer_arxiv_id": ["1810.05993", "2001.03093", "1911.00997"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_20661"} +{"question": "What researches extracted object masks and representations using unlabeled object-centric datasets, similar to the DINOSAUR method?", "answer": ["Bridging the Gap to Real-World Object-Centric Learning"], "answer_arxiv_id": ["2209.14860"], "source_meta": {"published_time": "20220711"}, "qid": "AutoScholarQuery_train_20662"} +{"question": "Could you provide me some studies that explore strategies from semi-supervised learning and self-supervised learning in SSDA?", "answer": ["A Survey on Deep Semi-supervised Learning", "Online Meta-Learning for Multi-Source and Semi-Supervised Domain\n Adaptation"], "answer_arxiv_id": ["2103.00550", "2004.04398"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_20663"} +{"question": "Are there any studies focused on finetuning language models?", "answer": ["Large Language Models Can Self-Improve", "STaR: Self-Taught Reasoner Bootstrapping Reasoning With Reasoning"], "answer_arxiv_id": ["2210.11610", "2203.14465"], "source_meta": {"published_time": "20230210"}, "qid": "AutoScholarQuery_train_20664"} +{"question": "What are some research works that discuss instance discrimination?", "answer": ["Big Self-Supervised Models are Strong Semi-Supervised Learners", "Momentum Contrast for Unsupervised Visual Representation Learning", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2006.10029", "1911.05722", "2103.00020"], "source_meta": {"published_time": "20230412"}, "qid": "AutoScholarQuery_train_20665"} +{"question": "Which papers provide theoretical analyses of the benefits of group invariance in learning settings?", "answer": ["Group Invariance, Stability to Deformations, and Complexity of Deep Convolutional Representations"], "answer_arxiv_id": ["1706.03078"], "source_meta": {"published_time": "20220929"}, "qid": "AutoScholarQuery_train_20666"} +{"question": "What is the work that models various complex relational patterns by representing temporal evolutions as rotations in quaternion vector space?", "answer": ["RotateQVS: Representing Temporal Information as Rotations in Quaternion\n Vector Space for Temporal Knowledge Graph Completion"], "answer_arxiv_id": ["2203.07993"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_20667"} +{"question": "What works describe the use of SSMs for video classification?", "answer": ["Long Movie Clip Classification with State-Space Video Models", "Selective Structured State-Spaces for Long-Form Video Understanding"], "answer_arxiv_id": ["2204.01692v3", "2303.14526"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20668"} +{"question": "What CLIP-based models have been used to transfer knowledge from pre-trained CLIP models?", "answer": ["Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2103.00020"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_20669"} +{"question": "What is an example of research aimed at improving the inference speed of pre-trained ViTs?", "answer": ["Token Merging: Your ViT But Faster"], "answer_arxiv_id": ["2210.09461"], "source_meta": {"published_time": "20231005"}, "qid": "AutoScholarQuery_train_20670"} +{"question": "What works contributed to the field of identifiable representation learning by introducing structure on the data generating process?", "answer": ["Properties from Mechanisms: An Equivariance Perspective on Identifiable Representation Learning", "Contrastive Learning Inverts the Data Generating Process"], "answer_arxiv_id": ["2110.15796", "2102.08850v4"], "source_meta": {"published_time": "20221129"}, "qid": "AutoScholarQuery_train_20671"} +{"question": "Which studies generated RIRs from reverberant audio clips?", "answer": ["Filtered noise shaping for time domain room impulse response estimation from reverberant speech", "IR-GAN: Room impulse response generator for far-field speech recognition", "TS-RIR: Translated synthetic room impulse responses for speech augmentation"], "answer_arxiv_id": ["2107.07503", "2010.13219", "2103.16804"], "source_meta": {"published_time": "20230727"}, "qid": "AutoScholarQuery_train_20672"} +{"question": "Which study deployed Proximal Policy Optimization (PPO) in RLHF for fine-tuning Large Language Models?", "answer": ["Proximal Policy Optimization Algorithms"], "answer_arxiv_id": ["1707.06347"], "source_meta": {"published_time": "20230410"}, "qid": "AutoScholarQuery_train_20673"} +{"question": "Which studies introduced constraint-based bilevel methods?", "answer": ["On Differentiating Parameterized Argmin and Argmax Problems with Application to Bi-level Optimization", "A Review on Bilevel Optimization: From Classical to Evolutionary Approaches and Applications"], "answer_arxiv_id": ["1607.05447", "1705.06270"], "source_meta": {"published_time": "20230530"}, "qid": "AutoScholarQuery_train_20674"} +{"question": "Can you provide me works on deep learning-based methods for point cloud analysis?", "answer": ["PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation", "Dynamic Graph CNN for Learning on Point Clouds", "PointFlowNet: Learning Representations for Rigid Motion Estimation from\n Point Clouds", "Spectrum AUC Difference (SAUCD): Human-aligned 3D Shape Evaluation", "PC-HMR: Pose Calibration for 3D Human Mesh Recovery from 2D\n Images/Videos", "High Fidelity 3D Hand Shape Reconstruction via Scalable Graph Frequency\n Decomposition"], "answer_arxiv_id": ["1612.00593", "1801.07829", "1806.02170", "2403.01619", "2103.09009", "2307.05541"], "source_meta": {"published_time": "20240312"}, "qid": "AutoScholarQuery_train_20675"} +{"question": "Which works discuss about federated extensions of DP-SGD?", "answer": ["Differentially Private Federated Learning: A Client Level Perspective", "Learning Differentially Private Recurrent Language Models", "Differentially Private Federated Learning on Heterogeneous Data"], "answer_arxiv_id": ["1712.07557", "1710.06963", "2111.09278v3"], "source_meta": {"published_time": "20230224"}, "qid": "AutoScholarQuery_train_20676"} +{"question": "Which works propose custom approaches to adapt state space models to higher-dimensional data or long sequences?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context", "Memorizing Transformers"], "answer_arxiv_id": ["2010.11929", "1901.02860", "2203.08913"], "source_meta": {"published_time": "20230213"}, "qid": "AutoScholarQuery_train_20677"} +{"question": "What papers discussed XGNN, the state-of-the-art model-level explanation method for GNNs?", "answer": ["XGNN: Towards Model-Level Explanations of Graph Neural Networks"], "answer_arxiv_id": ["2006.02587"], "source_meta": {"published_time": "20220915"}, "qid": "AutoScholarQuery_train_20678"} +{"question": "Which papers implemented Mixture-of-Experts transformer models?", "answer": ["Outrageously Large Neural Networks: The Sparsely-Gated\n Mixture-of-Experts Layer", "Switch Transformers: Scaling to Trillion Parameter Models with Simple\n and Efficient Sparsity", "GShard: Scaling Giant Models with Conditional Computation and Automatic\n Sharding", "Taming Sparsely Activated Transformer with Stochastic Experts"], "answer_arxiv_id": ["1701.06538", "2101.03961", "2006.16668", "2110.04260"], "source_meta": {"published_time": "20231201"}, "qid": "AutoScholarQuery_train_20679"} +{"question": "Which works have discussed and proposed solutions to overconfident scores produced by neural networks?", "answer": ["Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problem", "A Survey of Uncertainty in Deep Neural Networks", "Revisiting the Calibration of Modern Neural Networks", "Improving model calibration with accuracy versus uncertainty optimization", "Improving Uncertainty Calibration of Deep Neural Networks via Truth Discovery and Geometric Optimization", "Towards calibrated and scalable uncertainty representations for neural networks"], "answer_arxiv_id": ["1812.05720", "2107.03342", "2106.07998", "2012.07923", "2106.14662", "1911.00104"], "source_meta": {"published_time": "20221122"}, "qid": "AutoScholarQuery_train_20680"} +{"question": "What works have considered stochastic robustness certificates?", "answer": ["A Framework of Randomized Selection Based Certified Defenses Against Data Poisoning Attacks", "Intrinsic Certified Robustness of Bagging against Data Poisoning Attacks"], "answer_arxiv_id": ["2009.08739", "2008.04495"], "source_meta": {"published_time": "20230205"}, "qid": "AutoScholarQuery_train_20681"} +{"question": "Which research papers treat pose estimation as a regression task in direct methods?", "answer": ["PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in\n Cluttered Scenes", "Explaining the Ambiguity of Object Detection and 6D Pose From Visual\n Data"], "answer_arxiv_id": ["1711.00199", "1812.00287"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_20682"} +{"question": "Which works use 3D voxel grids for reconstruction?", "answer": ["Convolutional Occupancy Networks", "Atlas: End-to-End 3D Scene Reconstruction from Posed Images", "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video", "TransformerFusion: Monocular RGB Scene Reconstruction using Transformers", "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM"], "answer_arxiv_id": ["2003.04618", "2003.10432", "2104.00681", "2107.02191", "2112.12130"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_20683"} +{"question": "Could you provide me research papers about importance sampling-based estimator in the context of off-policy evaluation?", "answer": ["Consistent On-Line Off-Policy Evaluation", "Bootstrapping with Models: Confidence Intervals for Off-Policy Evaluation", "Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation", "Importance Resampling for Off-policy Prediction", "Towards Optimal Off-Policy Evaluation for Reinforcement Learning with Marginalized Importance Sampling", "CoinDICE: Off-Policy Confidence Interval Estimation", "GenDICE: Generalized Offline Estimation of Stationary Values"], "answer_arxiv_id": ["1702.07121", "1606.06126", "1810.12429", "1906.04328", "1906.03393", "2010.11652", "2002.09072"], "source_meta": {"published_time": "20221229"}, "qid": "AutoScholarQuery_train_20684"} +{"question": "What are some studies that discussed weight-oriented setup used in Trojan Attacks?", "answer": ["TBT: Targeted Neural Network Attack with Bit Trojan", "TrojViT: Trojan Insertion in Vision Transformers"], "answer_arxiv_id": ["1909.05193", "2208.13049"], "source_meta": {"published_time": "20230303"}, "qid": "AutoScholarQuery_train_20685"} +{"question": "Could you provide me some works about non-rendering based methods used for shadow generation?", "answer": ["Shadow Generation for Composite Image in Real-world Scenes"], "answer_arxiv_id": ["2104.10338"], "source_meta": {"published_time": "20240322"}, "qid": "AutoScholarQuery_train_20686"} +{"question": "Any research which used threshold according to loss on labeled data and adjusted it based on fixed mechanism?", "answer": ["Dash: Semi-Supervised Learning with Dynamic Thresholding"], "answer_arxiv_id": ["2109.00650"], "source_meta": {"published_time": "20220515"}, "qid": "AutoScholarQuery_train_20687"} +{"question": "Which research studies implement the coupling layer as a scale-and-shift operation?", "answer": ["P", "Density estimation using Real NVP"], "answer_arxiv_id": ["0704.0320", "1605.08803"], "source_meta": {"published_time": "20220614"}, "qid": "AutoScholarQuery_train_20688"} +{"question": "Could you provide me some works that shifted towards investing in 3D-aware image synthesis?", "answer": ["Efficient Geometry-aware 3D Generative Adversarial Networks", "pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware\n Image Synthesis", "StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image\n Synthesis", "GANcraft: Unsupervised 3D Neural Rendering of Minecraft Worlds", "StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation", "VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids"], "answer_arxiv_id": ["2112.07945", "2012.00926", "2110.08985", "2104.07659", "2112.11427", "2206.07695"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_20689"} +{"question": "Could you provide me some works about map-centric SLAM using voxel grids or points as the underlying 3D representation?", "answer": ["BundleFusion: Real-time Globally Consistent 3D Reconstruction using\n On-the-fly Surface Re-integration", "A Framework for the Volumetric Integration of Depth Images"], "answer_arxiv_id": ["1604.01093", "1410.0925"], "source_meta": {"published_time": "20231211"}, "qid": "AutoScholarQuery_train_20690"} +{"question": "Could you provide me some studies about surface rendering?", "answer": ["Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision", "Multiview Neural Surface Reconstruction by Disentangling Geometry and Appearance", "Learning Signed Distance Field for Multi-view Surface Reconstruction"], "answer_arxiv_id": ["1912.07372", "2003.09852", "2108.09964"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_20691"} +{"question": "Can you name research that extended the use of an end-to-end video-language model incorporating multiple video datasets for various downstream tasks?", "answer": ["All in One: Exploring Unified Video-Language Pre-training"], "answer_arxiv_id": ["2203.07303"], "source_meta": {"published_time": "20240403"}, "qid": "AutoScholarQuery_train_20692"} +{"question": "Can you tell me about some works that extended and improved NeRF for multiple instances and without camera ground-truths?", "answer": ["pixelNeRF: Neural Radiance Fields from One or Few Images", "GRF: Learning a General Radiance Field for 3D Representation and Rendering", "MINE: Towards Continuous Depth MPI with NeRF for Novel View Synthesis", "MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo", "IBRNet: Learning Multi-View Image-Based Rendering", "Geometry-biased Transformers for Novel View Synthesis", "NeRF-⁣-: Neural Radiance Fields Without Known Camera Parameters", "GNeRF: GAN-based Neural Radiance Field without Posed Camera"], "answer_arxiv_id": ["2012.02190", "2010.04595", "2103.14910", "2103.15595", "2102.13090", "2301.04650", "2102.07064", "2103.15606"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_20693"} +{"question": "What papers discuss shifting focus to the specific forms of symbols?", "answer": ["Lemur: Harmonizing Natural Language and Code for Language Agents", "Harnessing the Power of Large Language Models for Natural Language to\n First-Order Logic Translation"], "answer_arxiv_id": ["2310.06830", "2305.15541"], "source_meta": {"published_time": "20231115"}, "qid": "AutoScholarQuery_train_20694"} +{"question": "Which works propose the concept of Low-shot Object Detection employing image-conditioned detection?", "answer": ["Few-shot Object Detection via Feature Reweighting", "One-Shot Object Detection with Co-Attention and Co-Excitation", "OS2D: One-Stage One-Shot Object Detection by Matching Anchor Features"], "answer_arxiv_id": ["1812.01866", "1911.12529", "2003.06800"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_20695"} +{"question": "Could you provide me some studies on uncertainty estimation in the field of pixel classification and image segmentation?", "answer": ["Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning", "Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles", "Confidence-Aware Learning for Deep Neural Networks", "Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understanding", "What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?", "Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypotheses", "Uncertainty Estimates and Multi-Hypotheses Networks for Optical Flow", "A Probabilistic U-Net for Segmentation of Ambiguous Images"], "answer_arxiv_id": ["1506.02142", "1612.01474", "2007.01458", "1511.02680", "1703.04977", "1612.00197", "1802.07095", "1806.05034"], "source_meta": {"published_time": "20220603"}, "qid": "AutoScholarQuery_train_20696"} +{"question": "Which studies discuss adversarial training (AT) for enhancing model’s adversarial robustness?", "answer": ["Explaining and Harnessing Adversarial Examples", "Towards Deep Learning Models Resistant to Adversarial Attacks", "Theoretically Principled Trade-off between Robustness and Accuracy"], "answer_arxiv_id": ["1412.6572", "1706.06083", "1901.08573"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_20697"} +{"question": "Any works associated with multi-rater medical image segmentation that also consider label diversity caused by different expert preferences and inherent data uncertainty?", "answer": ["Transformer-based Annotation Bias-aware Medical Image Segmentation", "Probabilistic Modeling of Inter- and Intra-observer Variability in\n Medical Image Segmentation"], "answer_arxiv_id": ["2306.01340", "2307.11397"], "source_meta": {"published_time": "20240320"}, "qid": "AutoScholarQuery_train_20698"} +{"question": "What researches used goals or trajectory-level aggregates as inputs to a conditional policy in RCSL?", "answer": ["End-to-end Driving via Conditional Imitation Learning", "Learning to Reach Goals via Iterated Supervised Learning", "Generalized Decision Transformer for Offline Hindsight Information Matching"], "answer_arxiv_id": ["1710.02410", "1912.06088", "2111.10364"], "source_meta": {"published_time": "20221024"}, "qid": "AutoScholarQuery_train_20699"} +{"question": "Are there papers that deal with the Labels-at-Client scenario where some clients are fully labeled while others only contain unlabeled samples?", "answer": ["SemiFed: Semi-supervised Federated Learning with Consistency and Pseudo-Labeling", "RSCFed: Random Sampling Consensus Federated Semi-supervised Learning", "Federated Semi-supervised Medical Image Classification via Inter-client Relation Matching"], "answer_arxiv_id": ["2108.09412", "2203.13993", "2106.08600v1"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_20700"} +{"question": "Could you mention the works that have extended the data augmentation method like Mixup?", "answer": ["A Survey of Mix-based Data Augmentation: Taxonomy, Methods, Applications, and Explainability", "An overview of mixing augmentation methods and augmentation strategies", "CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features", "SaliencyMix: A Saliency Guided Data Augmentation Strategy for Better Regularization", "AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty", "Manifold Mixup: Better Representations by Interpolating Hidden States", "Improving Out-of-Distribution Robustness via Selective Augmentation"], "answer_arxiv_id": ["2212.10888", "2107.09887", "1905.04899", "2006.01791", "1912.02781", "1806.05236", "2201.00299"], "source_meta": {"published_time": "20230204"}, "qid": "AutoScholarQuery_train_20701"} +{"question": "Which papers have documented issues with optimizing Transformer-based architectures with SGD instead of Adam?", "answer": ["Understanding the Difficulty of Training Transformers", "Robust Training of Neural Networks using Scale Invariant Architectures"], "answer_arxiv_id": ["2004.08249", "2202.00980"], "source_meta": {"published_time": "20221011"}, "qid": "AutoScholarQuery_train_20702"} +{"question": "What works proposed defenses during training to produce a backdoor-free classifier from a possibly poisoned training set?", "answer": ["Spectral Signatures in Backdoor Attacks", "Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering", "Robust Anomaly Detection and Backdoor Attack Detection Via Differential Privacy", "Backdoor Defense via Decoupling the Training Process", "Learning with Bad Training Data via Iterative Trimmed Loss Minimization", "Reverse Engineering Imperceptible Backdoor Attacks on Deep Neural Networks for Detection and Training Set Cleansing", "On the Effectiveness of Mitigating Data Poisoning Attacks with Gradient Shaping", "DP-InstaHide: Provably Defusing Poisoning and Backdoor Attacks with Differentially Private Data Augmentations"], "answer_arxiv_id": ["1811.00636", "1811.03728", "1911.07116", "2202.03423", "1810.11874", "2010.07489", "2002.11497", "2103.02079"], "source_meta": {"published_time": "20231026"}, "qid": "AutoScholarQuery_train_20703"} +{"question": "Which works demonstrated an exponential gap between end-to-end-based verification sample complexity and the decomposition-based verification sample complexity?", "answer": ["On the Sample Complexity of End-to-end Training vs. Semantic Abstraction Training"], "answer_arxiv_id": ["1604.06915"], "source_meta": {"published_time": "20220406"}, "qid": "AutoScholarQuery_train_20704"} +{"question": "Which works discuss Nonlinear Independent Component Analysis (ICA) regarding images?", "answer": ["Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere", "Contrastive Learning Inverts the Data Generating Process", "Representation Learning with Contrastive Predictive Coding"], "answer_arxiv_id": ["2005.10242", "2102.08850v4", "1807.03748"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_20705"} +{"question": "What are some works related to image colorization using generative models?", "answer": ["UniColor: A Unified Framework for Multi-Modal Colorization with\n Transformer", "L-CAD: Language-based Colorization with Any-level Descriptions using\n Diffusion Priors"], "answer_arxiv_id": ["2209.11223", "2305.15217"], "source_meta": {"published_time": "20240401"}, "qid": "AutoScholarQuery_train_20706"} +{"question": "What papers mentioned the use of models with large image diffusion for image correspondence?", "answer": ["A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot\n Semantic Correspondence", "Diffusion Hyperfeatures: Searching Through Time and Space for Semantic\n Correspondence", "Unsupervised Semantic Correspondence Using Stable Diffusion", "Emergent Correspondence from Image Diffusion"], "answer_arxiv_id": ["2305.15347", "2305.14334", "2305.15581", "2306.03881"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_20707"} +{"question": "Have there been any studies focusing on initialization in FL?", "answer": ["FedNLP: Benchmarking Federated Learning Methods for Natural Language Processing Tasks", "Pretraining Federated Text Models for Next Word Prediction", "Pretrained Models for Multilingual Federated Learning"], "answer_arxiv_id": ["2104.08815v3", "2005.04828", "2206.02291"], "source_meta": {"published_time": "20220623"}, "qid": "AutoScholarQuery_train_20708"} +{"question": "Are there any studies that use a joint embedding model between agent states and associated natural language commands for agent reward shaping?", "answer": ["A Narration-based Reward Shaping Approach using Grounded Natural Language Commands"], "answer_arxiv_id": ["1911.00497"], "source_meta": {"published_time": "20230921"}, "qid": "AutoScholarQuery_train_20709"} +{"question": "What works study design of recommender systems which optimize for both user and producer utility?", "answer": ["Towards Content Provider Aware Recommender Systems: A Simulation Study on the Interplay between User and Provider Utilities"], "answer_arxiv_id": ["2105.02377v1"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_20710"} +{"question": "Can you name some studies that have applied diffusion models in image-to-image translation?", "answer": ["Plug-and-Play Diffusion Features for Text-Driven Image-to-Image\n Translation", "BBDM: Image-to-image Translation with Brownian Bridge Diffusion Models", "General Image-to-Image Translation with One-Shot Image Guidance", "Adding Conditional Control to Text-to-Image Diffusion Models"], "answer_arxiv_id": ["2211.12572", "2205.07680", "2307.14352", "2302.05543"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20711"} +{"question": "Which references discussed LiDAR-based BEV methods?", "answer": ["PointPillars: Fast Encoders for Object Detection from Point Clouds", "PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection", "Embracing Single Stride 3D Object Detector with Sparse Transformer", "AFDetV2: Rethinking the Necessity of the Second Stage for Object\n Detection from Point Clouds", "DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets"], "answer_arxiv_id": ["1812.05784", "1912.13192", "2112.06375", "2112.09205", "2301.06051"], "source_meta": {"published_time": "20240613"}, "qid": "AutoScholarQuery_train_20712"} +{"question": "Could you provide me some studies that utilize core-set techniques or clustering methods in the latent feature space in diversity-based Active Learning?", "answer": ["Active Learning for Convolutional Neural Networks: A Core-Set Approach", "Sequential Graph Convolutional Network for Active Learning"], "answer_arxiv_id": ["1708.00489", "2006.10219"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_20713"} +{"question": "What studies use neuro2vec for representation learning by predicting the Fourier transform of the masked part of the input signal?", "answer": ["neuro2vec: Masked Fourier Spectrum Prediction for Neurophysiological Representation Learning"], "answer_arxiv_id": ["2204.12440"], "source_meta": {"published_time": "20231024"}, "qid": "AutoScholarQuery_train_20714"} +{"question": "Which works have designed unified segmentation decoders for different task prompts?", "answer": ["Universal Instance Perception as Object Discovery and Retrieval", "Segment Everything Everywhere All at Once"], "answer_arxiv_id": ["2303.06674", "2304.06718"], "source_meta": {"published_time": "20231204"}, "qid": "AutoScholarQuery_train_20715"} +{"question": "Which paper introduced the model DeepID2 that uses a contrastive loss?", "answer": ["Deep Learning Face Representation by Joint Identification-Verification"], "answer_arxiv_id": ["1406.4773"], "source_meta": {"published_time": "20231104"}, "qid": "AutoScholarQuery_train_20716"} +{"question": "Can you cite researches that introduced Relative approach for encoding positional information in Transformers?", "answer": ["Self-Attention with Relative Position Representations"], "answer_arxiv_id": ["1803.02155"], "source_meta": {"published_time": "20230531"}, "qid": "AutoScholarQuery_train_20717"} +{"question": "Which studies have used a local mapping network to improve similarity in the encoder-based stream?", "answer": ["ELITE: Encoding Visual Concepts into Textual Embeddings for Customized\n Text-to-Image Generation"], "answer_arxiv_id": ["2302.13848"], "source_meta": {"published_time": "20240301"}, "qid": "AutoScholarQuery_train_20718"} +{"question": "Which studies argue that existing synthetic datasets do not have enough categorical attributes representing 'semantic' factors?", "answer": ["Disentangling by Factorising", "A Fine-Grained Analysis on Distribution Shift"], "answer_arxiv_id": ["1802.05983", "2110.11328"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_20719"} +{"question": "Which work introduced the concept of catastrophic forgetting, a phenomenon in continual learning?", "answer": ["Continual Lifelong Learning with Neural Networks: A Review"], "answer_arxiv_id": ["1802.07569"], "source_meta": {"published_time": "20230719"}, "qid": "AutoScholarQuery_train_20720"} +{"question": "Which works explore the generation of object states?", "answer": ["Learning Universal Policies via Text-Guided Video Generation", "Learning Procedure-aware Video Representation from Instructional Videos\n and Their Narrations", "P3IV: Probabilistic Procedure Planning from Instructional Videos with\n Weak Supervision", "Manipulate by Seeing: Creating Manipulation Controllers from Pre-Trained\n Representations"], "answer_arxiv_id": ["2302.00111", "2303.17839", "2205.02300", "2303.08135"], "source_meta": {"published_time": "20231212"}, "qid": "AutoScholarQuery_train_20721"} +{"question": "Could you provide me some studies that show robust overfitting is caused by the falsely memorized non-robust features?", "answer": ["Nash Equilibria and Pitfalls of Adversarial Training in Adversarial Robustness Games"], "answer_arxiv_id": ["2210.12606v3"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20722"} +{"question": "Which studies highlighted the issue of representation effort spent on recovering the 'noise' part of time series data?", "answer": ["Time Series Data Augmentation for Deep Learning: A Survey", "An Empirical Survey of Data Augmentation for Time Series Classification with Neural Networks", "ARDA: Automatic Relational Data Augmentation for Machine Learning"], "answer_arxiv_id": ["2002.12478", "2007.15951", "2003.09758"], "source_meta": {"published_time": "20230609"}, "qid": "AutoScholarQuery_train_20723"} +{"question": "Are there any studies which expanded on the theoretical foundations of GFlowNets with respect to variational methods?", "answer": ["Unifying Generative Models with GFlowNets and Beyond"], "answer_arxiv_id": ["2209.02606"], "source_meta": {"published_time": "20221002"}, "qid": "AutoScholarQuery_train_20724"} +{"question": "Could you provide me some works that describe methods to address the unreliability of LLMs with complex input-output mappings?", "answer": ["Natural Language to Code Translation with Execution", "CodeT: Code Generation with Generated Tests", "Coder Reviewer Reranking for Code Generation", "LEVER: Learning to Verify Language-to-Code Generation with Execution", "Self-collaboration Code Generation via ChatGPT", "Teaching Large Language Models to Self-Debug"], "answer_arxiv_id": ["2204.11454", "2207.10397", "2211.16490", "2302.08468", "2304.07590", "2304.05128v2"], "source_meta": {"published_time": "20230929"}, "qid": "AutoScholarQuery_train_20725"} +{"question": "Could you provide me some researches about Sample Reweighting which used to address the issue of label noise?", "answer": ["Robust Probabilistic Modeling with Bayesian Data Reweighting"], "answer_arxiv_id": ["1606.03860"], "source_meta": {"published_time": "20230920"}, "qid": "AutoScholarQuery_train_20726"} +{"question": "What papers introduced InstructGPT?", "answer": ["Training language models to follow instructions with human feedback"], "answer_arxiv_id": ["2203.02155"], "source_meta": {"published_time": "20231007"}, "qid": "AutoScholarQuery_train_20727"} +{"question": "Could you provide me some studies that have built graphs with entity nodes for factoid question answering?", "answer": ["Knowledge Guided Text Retrieval and Reading for Open Domain Question\n Answering", "Cognitive Graph for Multi-Hop Reading Comprehension at Scale"], "answer_arxiv_id": ["1911.03868", "1905.05460"], "source_meta": {"published_time": "20240221"}, "qid": "AutoScholarQuery_train_20728"} +{"question": "Could you provide me some works showing that Transformer architecture can discover efficient learning algorithms?", "answer": ["What Can Transformers Learn In-Context? A Case Study of Simple Function Classes", "What learning algorithm is in-context learning? Investigations with linear models", "In-context Reinforcement Learning with Algorithm Distillation"], "answer_arxiv_id": ["2208.01066", "2211.15661", "2210.14215"], "source_meta": {"published_time": "20230314"}, "qid": "AutoScholarQuery_train_20729"} +{"question": "What studies use a masking pretext task in self-supervised methods?", "answer": ["OmniMAE: Single Model Masked Pretraining on Images and Videos"], "answer_arxiv_id": ["2206.08356"], "source_meta": {"published_time": "20231019"}, "qid": "AutoScholarQuery_train_20730"} +{"question": "Which paper used featurized spheres to represent the scene in the context of view synthesis with point clouds?", "answer": ["Pulsar: Efficient Sphere-based Neural Rendering"], "answer_arxiv_id": ["2004.07484"], "source_meta": {"published_time": "20220512"}, "qid": "AutoScholarQuery_train_20731"} +{"question": "Which works perturb the speech variations during the analysis stage to encourage the synthesis stage to use the more stable representations?", "answer": ["Unsupervised Speech Decomposition via Triple Information Bottleneck", "Neural Analysis and Synthesis: Reconstructing Speech from Self-Supervised Representations"], "answer_arxiv_id": ["2004.11284", "2110.14513"], "source_meta": {"published_time": "20220525"}, "qid": "AutoScholarQuery_train_20732"} +{"question": "What work used a differentiable PatchMatch-based pruner to prune the disparity search space?", "answer": ["DeepPruner: Learning Efficient Stereo Matching via Differentiable\n PatchMatch"], "answer_arxiv_id": ["1909.05845"], "source_meta": {"published_time": "20240317"}, "qid": "AutoScholarQuery_train_20733"} +{"question": "What research highlights the risk of catastrophic forgetting in fine-tuning the VLM encoder for specific tasks?", "answer": ["Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution"], "answer_arxiv_id": ["2202.10054v1"], "source_meta": {"published_time": "20230822"}, "qid": "AutoScholarQuery_train_20734"} +{"question": "Which studies have achieved great results using InfoNCE loss in instance discrimination?", "answer": ["Momentum Contrast for Unsupervised Visual Representation Learning", "Improved Baselines with Momentum Contrastive Learning", "A Simple Framework for Contrastive Learning of Visual Representations"], "answer_arxiv_id": ["1911.05722", "2003.04297", "2002.05709"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_20735"} +{"question": "Which studies utilize motion information or scene flow to detect moving objects?", "answer": ["Motion Inspired Unsupervised Perception and Prediction in Autonomous\n Driving", "4D Unsupervised Object Discovery"], "answer_arxiv_id": ["2210.08061", "2210.04801"], "source_meta": {"published_time": "20240305"}, "qid": "AutoScholarQuery_train_20736"} +{"question": "Could you provide me any works which used off-the-shelf search engines to enhance the retrieval-augmented ICL process?", "answer": ["Making Pre-trained Language Models Better Few-shot Learners", "REPLUG: Retrieval-Augmented Black-Box Language Models", "Semantic-Oriented Unlabeled Priming for Large-Scale Language Models"], "answer_arxiv_id": ["2012.15723", "2301.12652", "2202.06133"], "source_meta": {"published_time": "20231220"}, "qid": "AutoScholarQuery_train_20737"} +{"question": "Which studies focus on multi-view photometric stereo (MVPS) that aims at 3D reconstruction under varying light conditions?", "answer": ["Multi-View Photometric Stereo: A Robust Solution and Benchmark Dataset\n for Spatially Varying Isotropic Materials", "ReLight My NeRF: A Dataset for Novel View Synthesis and Relighting of\n Real World Objects"], "answer_arxiv_id": ["2001.06659", "2304.10448"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_20738"} +{"question": "Which papers mentioned about the split-conformal prediction framework?", "answer": ["Distribution-Free Predictive Inference For Regression"], "answer_arxiv_id": ["1604.04173"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_20739"} +{"question": "Which studies extend SR-STE towards using adaptive N:M ratios across layers and steps?", "answer": ["Training Recipe for N:M Structured Sparsity with Decaying Pruning Mask"], "answer_arxiv_id": ["2209.07617"], "source_meta": {"published_time": "20230202"}, "qid": "AutoScholarQuery_train_20740"} +{"question": "Could you provide me some studies that used quantization to reduce the memory and compute requirements?", "answer": ["LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale", "ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers", "SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models", "GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers"], "answer_arxiv_id": ["2208.07339", "2206.01861", "2211.10438", "2210.17323"], "source_meta": {"published_time": "20230525"}, "qid": "AutoScholarQuery_train_20741"} +{"question": "What papers would give me a deeper understanding of the properties and methods for the fast computation of the Berk-Jones statistics?", "answer": ["Fast calculation of p-values for one-sided Kolmogorov-Smirnov type statistics"], "answer_arxiv_id": ["2009.04954"], "source_meta": {"published_time": "20221227"}, "qid": "AutoScholarQuery_train_20742"} +{"question": "Who proposed Masked image modeling?", "answer": ["An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale", "SiT: Self-supervised vIsion Transformer", "Are Large-scale Datasets Necessary for Self-Supervised Pre-training?", "Masked Autoencoders Are Scalable Vision Learners", "SimMIM: a Simple Framework for Masked Image Modeling", "BEiT: BERT Pre-Training of Image Transformers", "data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language", "Masked Feature Prediction for Self-Supervised Visual Pre-Training", "MultiMAE: Multi-modal Multi-task Masked Autoencoders"], "answer_arxiv_id": ["2010.11929", "2104.03602", "2112.10740", "2111.06377", "2111.09886", "2106.08254", "2202.03555", "2112.09133", "2204.01678"], "source_meta": {"published_time": "20230621"}, "qid": "AutoScholarQuery_train_20743"} +{"question": "Which works originally developed the diffusion models?", "answer": ["Denoising Diffusion Probabilistic Models", "Deep Unsupervised Learning using Nonequilibrium Thermodynamics", "Generative Modeling by Estimating Gradients of the Data Distribution", "Score-Based Generative Modeling through Stochastic Differential Equations"], "answer_arxiv_id": ["2006.11239", "1503.03585", "1907.05600", "2011.13456"], "source_meta": {"published_time": "20231205"}, "qid": "AutoScholarQuery_train_20744"} +{"question": "Which work proposed Analytical Marching, an alternative method to extract the 0-isosurface of a CPWA neural implicit shape?", "answer": ["Learning and Meshing from Deep Implicit Surface Networks Using an Efficient Implementation of Analytic Marching"], "answer_arxiv_id": ["2106.10031"], "source_meta": {"published_time": "20230612"}, "qid": "AutoScholarQuery_train_20745"} +{"question": "Could you provide me some studies about image transformations used for data augmentations?", "answer": ["Mitigating Adversarial Effects Through Randomization", "Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks"], "answer_arxiv_id": ["1711.01991", "1904.02884"], "source_meta": {"published_time": "20230131"}, "qid": "AutoScholarQuery_train_20746"} +{"question": "Which works propose weakly supervised image segmentation methods?", "answer": ["Fully Convolutional Multi-Class Multiple Instance Learning", "Revisiting Dilated Convolution: A Simple Approach for Weakly- and Semi-\n Supervised Semantic Segmentation", "GroupViT: Semantic Segmentation Emerges from Text Supervision", "SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary\n Semantic Segmentation", "MaskCLIP: Masked Self-Distillation Advances Contrastive Language-Image\n Pretraining", "Decoupling Zero-Shot Semantic Segmentation", "Image Segmentation Using Text and Image Prompts", "ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation"], "answer_arxiv_id": ["1412.7144", "1805.04574", "2202.11094", "2211.14813", "2208.12262", "2112.07910", "2112.10003", "2212.03588"], "source_meta": {"published_time": "20231218"}, "qid": "AutoScholarQuery_train_20747"} +{"question": "What are some works that discuss ways to tackle slow convergence on complex datasets of symmetric losses?", "answer": ["Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels", "Normalized Loss Functions for Deep Learning with Noisy Labels"], "answer_arxiv_id": ["1805.07836", "2006.13554"], "source_meta": {"published_time": "20221208"}, "qid": "AutoScholarQuery_train_20748"} +{"question": "Which studies proposed the first solutions to the data-free model extraction problem?", "answer": ["Data-Free Model Extraction", "MAZE: Data-Free Model Stealing Attack Using Zeroth-Order Gradient Estimation"], "answer_arxiv_id": ["2011.14779v2", "2005.03161"], "source_meta": {"published_time": "20230918"}, "qid": "AutoScholarQuery_train_20749"} +{"question": "What papers discussed the combination of models with different architectures to utilize the growing PTM hubs?", "answer": ["Domain Generalization using Pretrained Models without Fine-tuning", "ZooD: Exploiting Model Zoo for Out-of-Distribution Generalization"], "answer_arxiv_id": ["2203.04600", "2210.09236"], "source_meta": {"published_time": "20230605"}, "qid": "AutoScholarQuery_train_20750"} +{"question": "What research papers focused on the transformer model?", "answer": ["Attention Is All You Need"], "answer_arxiv_id": ["1706.03762"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_20751"} +{"question": "Could you tell me about research works that proposed the use of a fully accessible LLM for detecting hallucinations when direct access to output uncertainty is not possible?", "answer": ["SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for\n Generative Large Language Models"], "answer_arxiv_id": ["2303.08896"], "source_meta": {"published_time": "20231231"}, "qid": "AutoScholarQuery_train_20752"} +{"question": "What Diffusion Model approaches integrate a two-stage method for processing functional data?", "answer": ["SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation", "3DShape2VecSet: A 3D Shape Representation for Neural Fields and\n Generative Diffusion Models", "DiffRF: Rendering-Guided 3D Radiance Field Diffusion"], "answer_arxiv_id": ["2212.04493", "2301.11445", "2212.01206"], "source_meta": {"published_time": "20231126"}, "qid": "AutoScholarQuery_train_20753"} +{"question": "Which works showed that large language models acquire surprising in-context reasoning capabilities?", "answer": ["Language Models are Few-Shot Learners", "Holistic Evaluation of Language Models", "Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models"], "answer_arxiv_id": ["2005.14165", "2211.09110", "2206.04615"], "source_meta": {"published_time": "20230524"}, "qid": "AutoScholarQuery_train_20754"} +{"question": "Which research publications have used linear-polarization images in the application of shape reconstruction?", "answer": ["Shape from Polarization for Complex Scenes in the Wild", "3D Human Shape Reconstruction from a Polarization Image", "Deep Shape from Polarization"], "answer_arxiv_id": ["2112.11377", "2007.09268", "1903.10210"], "source_meta": {"published_time": "20231129"}, "qid": "AutoScholarQuery_train_20755"} +{"question": "Could you please cite some works that show the incorporation of multi-task data in inverse reinforcement learning?", "answer": ["Learning a Prior over Intent via Meta-Inverse Reinforcement Learning", "Meta-Inverse Reinforcement Learning with Probabilistic Context Variables", "Multi-task Maximum Entropy Inverse Reinforcement Learning", "Repeated Inverse Reinforcement Learning"], "answer_arxiv_id": ["1805.12573", "1909.09314", "1805.08882v2", "1705.05427"], "source_meta": {"published_time": "20230901"}, "qid": "AutoScholarQuery_train_20756"} +{"question": "Which studies proposed methods for interacting two-hand reconstruction from monocular RGB?", "answer": ["Interacting Attention Graph for Single Image Two-Hand Reconstruction", "InterHand2.6M: A Dataset and Baseline for 3D Interacting Hand Pose\n Estimation from a Single RGB Image", "Im2Hands: Learning Attentive Implicit Representation of Interacting\n Two-Hand Shapes", "FourierHandFlow: Neural 4D Hand Representation Using Fourier Query Flow", "Reconstructing Interacting Hands with Interaction Prior from Monocular\n Images", "A2J-Transformer: Anchor-to-Joint Transformer Network for 3D Interacting\n Hand Pose Estimation from a Single RGB Image", "Bringing Inputs to Shared Domains for 3D Interacting Hands Recovery in\n the Wild", "Decoupled Iterative Refinement Framework for Interacting Hands\n Reconstruction from a Single RGB Image", "Monocular 3D Reconstruction of Interacting Hands via Collision-Aware\n Factorized Refinements"], "answer_arxiv_id": ["2203.09364", "2008.09309", "2302.14348", "2307.08100", "2308.14082", "2304.03635", "2303.13652", "2302.02410", "2111.00763"], "source_meta": {"published_time": "20240326"}, "qid": "AutoScholarQuery_train_20757"} +{"question": "Any works that have applied KD to handle task of semantic segmentation?", "answer": ["Structured Knowledge Distillation for Dense Prediction"], "answer_arxiv_id": ["1903.04197"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_20758"} +{"question": "Can you share some studies that adopted reward learning from preferences in NLP tasks?", "answer": ["Bayesian Attention Modules", "Training language models to follow instructions with human feedback", "Fine-Tuning Language Models from Human Preferences", "Learning with Different Amounts of Annotation: From Zero to Many Labels", "ALLSH: Active Learning Guided by Local Sensitivity and Hardness", "WebGPT: Browser-assisted question-answering with human feedback", "Knowing More About Questions Can Help: Improving Calibration in Question Answering", "Teaching language models to support answers with verified quotes", "Passage-Mask: A Learnable Regularization Strategy for Retriever-Reader Models", "Causal-aware Safe Policy Improvement for Task-oriented dialogue"], "answer_arxiv_id": ["2010.10604", "2203.02155", "1909.08593", "2109.04408", "2205.04980", "2112.09332", "2106.01494", "2203.11147", "2211.00915", "2103.06370"], "source_meta": {"published_time": "20230220"}, "qid": "AutoScholarQuery_train_20759"} +{"question": "What research papers are about improving adversarial robustness in the context of knowledge distillation?", "answer": ["The Enemy of My Enemy is My Friend: Exploring Inverse Adversaries for Improving Adversarial Training", "Exploring Memorization in Adversarial Training", "Theoretically Principled Trade-off between Robustness and Accuracy", "Adversarially Robust Distillation", "Distillation as a Defense to Adversarial Perturbations against Deep Neural Networks", "Reliable Adversarial Distillation with Unreliable Teachers", "Revisiting Adversarial Robustness Distillation: Robust Soft Labels Make Student Better"], "answer_arxiv_id": ["2211.00525", "2106.01606", "1901.08573", "1905.09747", "1511.04508", "2106.04928", "2108.07969"], "source_meta": {"published_time": "20231101"}, "qid": "AutoScholarQuery_train_20760"} +{"question": "Which research works have focused on integrating images with entity features in knowledge graphs?", "answer": ["Image-embodied Knowledge Representation Learning"], "answer_arxiv_id": ["1609.07028"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_20761"} +{"question": "Which papers proposed a self-supervised framework that predicts angles and inter-residue distances?", "answer": ["Structure-aware Protein Self-supervised Learning"], "answer_arxiv_id": ["2204.04213"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20762"} +{"question": "Any works about program synthesis being a novel direction for financial NLP?", "answer": ["PIXIU: A Large Language Model, Instruction Data and Evaluation Benchmark\n for Finance", "WHEN FLUE MEETS FLANG: Benchmarks and Large Pre-trained Language Model\n for Financial Domain", "BBT-Fin: Comprehensive Construction of Chinese Financial Domain\n Pre-trained Language Model, Corpus and Benchmark", "Select and Trade: Towards Unified Pair Trading with Hierarchical\n Reinforcement Learning"], "answer_arxiv_id": ["2306.05443", "2211.00083", "2302.09432", "2301.10724"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_20763"} +{"question": "Can you provide works that mention the theoretical foundation of NeRF and its application in neural volume rendering?", "answer": ["NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis"], "answer_arxiv_id": ["2003.08934"], "source_meta": {"published_time": "20240118"}, "qid": "AutoScholarQuery_train_20764"} +{"question": "Could you provide me some studies that discuss the use of marginal likelihood to select regularization strength, invariances, architectures, and representations in the context of deep learning?", "answer": ["Global inducing point variational posteriors for Bayesian neural networks and deep Gaussian processes", "Scalable Marginal Likelihood Estimation for Model Selection in Deep Learning"], "answer_arxiv_id": ["2005.08140", "2104.04975"], "source_meta": {"published_time": "20230606"}, "qid": "AutoScholarQuery_train_20765"} +{"question": "Which studies found that deeper layers are responsible for forgetting in transfer learning settings using CKA?", "answer": ["Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics"], "answer_arxiv_id": ["2007.07400"], "source_meta": {"published_time": "20221028"}, "qid": "AutoScholarQuery_train_20766"} +{"question": "Could you provide me examples of works that reported on the problems with using scaling methods?", "answer": ["Verified Uncertainty Calibration"], "answer_arxiv_id": ["1909.10155"], "source_meta": {"published_time": "20230518"}, "qid": "AutoScholarQuery_train_20767"} +{"question": "Could you provide me works that embedded analytic equations into neural solvers to preserve physical laws?", "answer": ["ContactNets: Learning Discontinuous Contact Dynamics with Smooth, Implicit Representations", "Data-Augmented Contact Model for Rigid Body Simulation"], "answer_arxiv_id": ["2009.11193", "1803.04019"], "source_meta": {"published_time": "20221207"}, "qid": "AutoScholarQuery_train_20768"} +{"question": "Which research papers discussed the use of human annotations to build stereotype resources in the field of NLP?", "answer": ["Re-contextualizing Fairness in NLP: The Case of India", "StereoSet: Measuring stereotypical bias in pretrained language models"], "answer_arxiv_id": ["2209.12226", "2004.09456"], "source_meta": {"published_time": "20230720"}, "qid": "AutoScholarQuery_train_20769"} +{"question": "Are there any references exploring the need for user-defined concepts in concept-based models?", "answer": ["Language in a Bottle: Language Model Guided Concept Bottlenecks for\n Interpretable Image Classification"], "answer_arxiv_id": ["2211.11158"], "source_meta": {"published_time": "20240124"}, "qid": "AutoScholarQuery_train_20770"} +{"question": "What methods can be used to properly limit exploration within a high probability safety region in a safe exploration RL?", "answer": ["Learning-based Model Predictive Control for Safe Exploration", "Safe Exploration for Interactive Machine Learning", "Safe Exploration in Continuous Action Spaces", "Robust Regression for Safe Exploration in Control", "Safe Reinforcement Learning via Shielding", "Neurosymbolic Reinforcement Learning with Formally Verified Exploration", "Safe Multi-Agent Reinforcement Learning via Shielding"], "answer_arxiv_id": ["1803.08287", "1910.13726", "1801.08757", "1906.05819v2", "1708.08611", "2009.12612", "2101.11196"], "source_meta": {"published_time": "20231203"}, "qid": "AutoScholarQuery_train_20771"} +{"question": "What research provides upper bounds for minimax optimization in unconstrained settings?", "answer": ["Last Iterate is Slower than Averaged Iterate in Smooth Convex-Concave Saddle Point Problems"], "answer_arxiv_id": ["2002.00057"], "source_meta": {"published_time": "20230425"}, "qid": "AutoScholarQuery_train_20772"} +{"question": "Which work introduced the concept of Diffusion Models?", "answer": ["Deep Unsupervised Learning using Nonequilibrium Thermodynamics"], "answer_arxiv_id": ["1503.03585"], "source_meta": {"published_time": "20230604"}, "qid": "AutoScholarQuery_train_20773"} +{"question": "Any works about the extension of text-to-image models for video generation?", "answer": ["GODIVA: Generating Open-DomaIn Videos from nAtural Descriptions", "CogVideo: Large-scale Pretraining for Text-to-Video Generation via\n Transformers", "N\\\"UWA: Visual Synthesis Pre-training for Neural visUal World creAtion", "Video Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "Imagen Video: High Definition Video Generation with Diffusion Models", "VideoFusion: Decomposed Diffusion Models for High-Quality Video\n Generation", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "MagicVideo: Efficient Video Generation With Latent Diffusion Models"], "answer_arxiv_id": ["2104.14806", "2205.15868", "2111.12417", "2204.03458", "2209.14792", "2210.02303", "2303.08320", "2304.08818", "2211.11018"], "source_meta": {"published_time": "20240314"}, "qid": "AutoScholarQuery_train_20774"} +{"question": "What papers are about consistency regularization in Test-time adaptation?", "answer": ["Continual Test-Time Domain Adaptation", "Robust Mean Teacher for Continual and Gradual Test-Time Adaptation", "Contrastive Learning with Stronger Augmentations"], "answer_arxiv_id": ["2203.13591", "2211.13081", "2104.07713"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_20775"} +{"question": "Which works follow the process of sampling programs and maximizing the probability of the correct ones in weakly-supervised semantic parsing?", "answer": ["From Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood", "A Discrete Hard EM Approach for Weakly Supervised Question Answering", "Merging Weak and Active Supervision for Semantic Parsing"], "answer_arxiv_id": ["1704.07926", "1909.04849", "1911.12986"], "source_meta": {"published_time": "20220528"}, "qid": "AutoScholarQuery_train_20776"} +{"question": "Could you name some papers that proposed reconstruction methods from a federated-learning setup?", "answer": ["Deep Leakage from Gradients", "Deep Models Under the GAN: Information Leakage from Collaborative Deep Learning", "Inverting Gradients - How easy is it to break privacy in federated learning?", "See through Gradients: Image Batch Recovery via GradInversion", "Evaluating Gradient Inversion Attacks and Defenses in Federated Learning", "Fishing for User Data in Large-Batch Federated Learning via Gradient Magnification"], "answer_arxiv_id": ["1906.08935", "1702.07464", "2003.14053", "2104.07586", "2112.00059", "2202.00580"], "source_meta": {"published_time": "20230704"}, "qid": "AutoScholarQuery_train_20777"} +{"question": "Could you provide some studies about applying flows in causal discovery and various causal applications?", "answer": ["Differentiable Causal Discovery from Interventional Data", "Causal Autoregressive Flows", "CausalAF: Causal Autoregressive Flow for Safety-Critical Driving Scenario Generation", "Harmonization with Flow-based Causal Inference"], "answer_arxiv_id": ["2007.01754", "2011.02268v2", "2110.13939", "2106.06845"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_20778"} +{"question": "What works are about training-free samplers that balance the number of function evaluations and generation quality?", "answer": ["Denoising Diffusion Implicit Models", "Pseudo Numerical Methods for Diffusion Models on Manifolds", "Fast Sampling of Diffusion Models with Exponential Integrator", "DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps"], "answer_arxiv_id": ["2010.02502", "2202.09778", "2204.13902", "2206.00927"], "source_meta": {"published_time": "20230615"}, "qid": "AutoScholarQuery_train_20779"} +{"question": "What studies use LiDAR point clouds as extra data to improve the performance of monocular 3D object detection?", "answer": ["Monocular 3D Object Detection Leveraging Accurate Proposals and Shape\n Reconstruction", "MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty\n Propagation", "MonoDistill: Learning Spatial Features for Monocular 3D Object Detection"], "answer_arxiv_id": ["1904.01690", "2103.12605", "2201.10830"], "source_meta": {"published_time": "20240229"}, "qid": "AutoScholarQuery_train_20780"} +{"question": "Could you provide me some studies about video generation using Diffusion models?", "answer": ["Imagen Video: High Definition Video Generation with Diffusion Models", "Align your Latents: High-Resolution Video Synthesis with Latent\n Diffusion Models", "Make-A-Video: Text-to-Video Generation without Text-Video Data", "GLIDE: Towards Photorealistic Image Generation and Editing with\n Text-Guided Diffusion Models"], "answer_arxiv_id": ["2210.02303", "2304.08818", "2209.14792", "2112.10741"], "source_meta": {"published_time": "20230928"}, "qid": "AutoScholarQuery_train_20781"} +{"question": "Which papers established domain alignments through various strategies for point cloud domain adaptation?", "answer": ["PointDAN: A Multi-Scale 3D Domain Adaption Network for Point Cloud\n Representation", "Unsupervised Domain Adaptive 3D Detection with Multi-Level Consistency"], "answer_arxiv_id": ["1911.02744", "2107.11355"], "source_meta": {"published_time": "20240327"}, "qid": "AutoScholarQuery_train_20782"} +{"question": "What were the previous regularization techniques used for continual learning?", "answer": ["Overcoming catastrophic forgetting in neural networks", "CPR: Classifier-Projection Regularization for Continual Learning", "Continual Learning for Text Classification with Information Disentanglement Based Regularization"], "answer_arxiv_id": ["1612.00796", "2006.07326", "2104.05489v2"], "source_meta": {"published_time": "20240302"}, "qid": "AutoScholarQuery_train_20783"} +{"question": "Which works proposed using a replay buffer in order to reduce computational complexity in Energy-based models (EBMs)?", "answer": ["Implicit Generation and Generalization in Energy-Based Models"], "answer_arxiv_id": ["1903.08689"], "source_meta": {"published_time": "20230306"}, "qid": "AutoScholarQuery_train_20784"} +{"question": "What research describes the concept of 'Child-tuning' where they iteratively update a subset of parameters?", "answer": ["Raise a Child in Large Language Model: Towards Effective and\n Generalizable Fine-tuning"], "answer_arxiv_id": ["2109.05687"], "source_meta": {"published_time": "20231215"}, "qid": "AutoScholarQuery_train_20785"} +{"question": "Which papers used knowledge models to generate tail inferences from narrative statements?", "answer": ["Dynamic Neuro-Symbolic Knowledge Graph Construction for Zero-shot\n Commonsense Question Answering", "MISC: A MIxed Strategy-Aware Model Integrating COMET for Emotional\n Support Conversation"], "answer_arxiv_id": ["1911.03876", "2203.13560"], "source_meta": {"published_time": "20240226"}, "qid": "AutoScholarQuery_train_20786"} +{"question": "Which work artificially imposes colors on the MNIST handwritten digits dataset?", "answer": ["Invariant Risk Minimization"], "answer_arxiv_id": ["1907.02893"], "source_meta": {"published_time": "20230923"}, "qid": "AutoScholarQuery_train_20787"} +{"question": "What papers are about identifying optimal methods for selecting examples in In-context learning?", "answer": ["Demystifying Prompts in Language Models via Perplexity Estimation", "Complexity-Based Prompting for Multi-Step Reasoning", "An Information-theoretic Approach to Prompt Engineering Without Ground\n Truth Labels", "Automatic Chain of Thought Prompting in Large Language Models", "Automatic Prompt Augmentation and Selection with Chain-of-Thought from\n Labeled Data"], "answer_arxiv_id": ["2212.04037", "2210.00720", "2203.11364", "2210.03493", "2302.12822"], "source_meta": {"published_time": "20231111"}, "qid": "AutoScholarQuery_train_20788"} +{"question": "Could you provide some research about Concept-based Explainability?", "answer": ["Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)", "Towards Automatic Concept-based Explanations", "On Completeness-aware Concept-Based Explanations in Deep Neural Networks"], "answer_arxiv_id": ["1711.11279", "1902.03129", "1910.07969"], "source_meta": {"published_time": "20221013"}, "qid": "AutoScholarQuery_train_20789"} +{"question": "What works have been done in the zero-shot forecasting setting?", "answer": ["Meta-learning framework with applications to zero-shot time-series forecasting", "N-BEATS: Neural basis expansion analysis for interpretable time series forecasting", "Zero-shot and few-shot time series forecasting with ordinal regression recurrent neural networks"], "answer_arxiv_id": ["2002.02887", "1905.10437", "2003.12162"], "source_meta": {"published_time": "20231103"}, "qid": "AutoScholarQuery_train_20790"} +{"question": "Could you provide me an example of a research that adopted temporary logical rules extracted via temporary random walks for better interpretability?", "answer": ["TLogic: Temporal Logical Rules for Explainable Link Forecasting on\n Temporal Knowledge Graphs"], "answer_arxiv_id": ["2112.08025"], "source_meta": {"published_time": "20240528"}, "qid": "AutoScholarQuery_train_20791"} +{"question": "Which works added attention scores in the previous layer to the current one for layer interaction?", "answer": ["RealFormer: Transformer Likes Residual Attention", "Evolving Attention with Residual Convolutions"], "answer_arxiv_id": ["2012.11747", "2102.12895"], "source_meta": {"published_time": "20230208"}, "qid": "AutoScholarQuery_train_20792"} +{"question": "What studies apply DEQs in various tasks such as language modeling, image generation, inverse problems in imaging and others?", "answer": ["Deep Equilibrium Models", "Implicit Graph Neural Networks", "Multiscale Deep Equilibrium Models", "Deep Equilibrium Approaches to Diffusion Models", "Deep Equilibrium Architectures for Inverse Problems in Imaging", "Connections between Deep Equilibrium and Sparse Representation Models with Application to Hyperspectral Image Denoising", "Deep Equilibrium Optical Flow Estimation"], "answer_arxiv_id": ["1909.01377", "2009.06211", "2006.08656", "2210.12867", "2102.07944", "2203.15901", "2204.08442"], "source_meta": {"published_time": "20231014"}, "qid": "AutoScholarQuery_train_20793"} +{"question": "In what papers were the impacts of increasing corpus diversity on incremental contextual learning studied?", "answer": ["On the Effect of Pretraining Corpora on In-context Learning by a\n Large-scale Language Model"], "answer_arxiv_id": ["2204.13509"], "source_meta": {"published_time": "20240219"}, "qid": "AutoScholarQuery_train_20794"} +{"question": "Which research papers attempted to learn key points for robotic manipulation tasks?", "answer": ["kPAM: KeyPoint Affordances for Category-Level Robotic Manipulation", "Keto: Learning Keypoint Representations for Tool Manipulation", "Unsupervised Learning of Intrinsic Structural Representation Points", "KeypointDeformer: Unsupervised 3D Keypoint Discovery for Shape Control", "Unsupervised Learning of Visual 3D Keypoints for Control"], "answer_arxiv_id": ["1903.06684", "1910.11977", "2003.01661", "2104.11224", "2106.07643"], "source_meta": {"published_time": "20231225"}, "qid": "AutoScholarQuery_train_20795"} +{"question": "Which papers originally proposed and studied the saturation effect in linear inverse problems?", "answer": ["Global Saturation of Regularization Methods for Inverse Ill-Posed Problems"], "answer_arxiv_id": ["1008.0108v1"], "source_meta": {"published_time": null}, "qid": "AutoScholarQuery_train_20796"} +{"question": "Which studies have developed methodologies for enforcing certain fairness metrics?", "answer": ["Fairness Reprogramming"], "answer_arxiv_id": ["2209.10222"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20797"} +{"question": "What works have been developed towards using the distributionally robust optimization (DRO) framework to improve robustness with the help of group information?", "answer": ["Does Distributionally Robust Supervised Learning Give Robust Classifiers?", "Coping with label shift via distributionally robust optimisation"], "answer_arxiv_id": ["1611.02041", "2010.12230"], "source_meta": {"published_time": "20230608"}, "qid": "AutoScholarQuery_train_20798"} +{"question": "Can you specify studies that investigated the applicability of diffusion to discrete and categorical variables?", "answer": ["Structured Denoising Diffusion Models in Discrete State-Spaces", "Vector Quantized Diffusion Model for Text-to-Image Synthesis", "Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning"], "answer_arxiv_id": ["2107.03006", "2111.14822", "2208.04202"], "source_meta": {"published_time": "20230207"}, "qid": "AutoScholarQuery_train_20799"} +{"question": "What paper proposed a lightweight image captioning model incorporating retrieved concepts?", "answer": ["EVCap: Retrieval-Augmented Image Captioning with External Visual-Name\n Memory for Open-World Comprehension"], "answer_arxiv_id": ["2311.15879"], "source_meta": {"published_time": "20240604"}, "qid": "AutoScholarQuery_train_20800"} +{"question": "What works have extensively studied the problem of repeated Stackelberg games?", "answer": ["No-Regret Learning in Dynamic Stackelberg Games", "Decentralized Cooperative Reinforcement Learning with Hierarchical Information Structure"], "answer_arxiv_id": ["2202.04786", "2111.00781"], "source_meta": {"published_time": "20230127"}, "qid": "AutoScholarQuery_train_20801"} +{"question": "What are the papers that discuss 3D geometric representation learning for 3D molecular conformation?", "answer": ["SchNet – a deep learning architecture for molecules and materials", "Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium Molecules", "Geometric and Physical Quantities improve E(3) Equivariant Message Passing", "Spherical Message Passing for 3D Graph Networks", "Equivariant message passing for the prediction of tensorial properties and molecular spectra", "Rotation Invariant Graph Neural Networks using Spin Convolutions", "GemNet: Universal Directional Graph Neural Networks for Molecules"], "answer_arxiv_id": ["1712.06113", "2011.14115", "2110.02905", "2102.05013", "2102.03150", "2106.09575", "2106.08903v10"], "source_meta": {"published_time": "20220627"}, "qid": "AutoScholarQuery_train_20802"} +{"question": "What works proposed to distill a 2D text-to-image generation model to generate 3D shapes from texts?", "answer": ["DreamFusion: Text-to-3D using 2D Diffusion", "Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation"], "answer_arxiv_id": ["2209.14988", "2212.00774v1"], "source_meta": {"published_time": "20231023"}, "qid": "AutoScholarQuery_train_20803"} +{"question": "Could you provide me some works that extended VAE for sequential data generation?", "answer": ["Disentangled Sequential Autoencoder", "S3VAE: Self-Supervised Sequential VAE for Representation Disentanglement and Data Generation", "Contrastively Disentangled Sequential Variational Autoencoder"], "answer_arxiv_id": ["1803.02991", "2005.11437", "2110.12091"], "source_meta": {"published_time": "20220815"}, "qid": "AutoScholarQuery_train_20804"} +{"question": "Which papers study the convergence of gradient descent for two-layer ReLU networks?", "answer": ["Fine-Grained Analysis of Optimization and Generalization for\n Overparameterized Two-Layer Neural Networks", "Gradient Descent Finds Global Minima of Deep Neural Networks"], "answer_arxiv_id": ["1901.08584", "1811.03804"], "source_meta": {"published_time": "20240328"}, "qid": "AutoScholarQuery_train_20805"} +{"question": "Who studied a notion of overstability in natural language processing?", "answer": ["Adversarial Examples for Evaluating Reading Comprehension Systems"], "answer_arxiv_id": ["1707.07328"], "source_meta": {"published_time": "20230501"}, "qid": "AutoScholarQuery_train_20806"} +{"question": "Which papers presented the application of Sparse MoEs in vision models?", "answer": ["Scaling Vision with Sparse Mixture of Experts"], "answer_arxiv_id": ["2106.05974"], "source_meta": {"published_time": "20220608"}, "qid": "AutoScholarQuery_train_20807"} +{"question": "Which works mentioned methods of reconstructing merged tokens to their original positions for semantic segmentation?", "answer": ["Making Vision Transformers Efficient from A Token Sparsification View", "Token Merging: Your ViT But Faster", "Not All Patches are What You Need: Expediting Vision Transformers via\n Token Reorganizations", "A Fast Training-Free Compression Framework for Vision Transformers", "Beyond Attentive Tokens: Incorporating Token Importance and Diversity\n for Efficient Vision Transformers"], "answer_arxiv_id": ["2303.08685", "2210.09461", "2202.07800", "2303.02331", "2211.11315"], "source_meta": {"published_time": "20240614"}, "qid": "AutoScholarQuery_train_20808"} +{"question": "Can you name some of the notable work on deep clustering methods?", "answer": ["Unsupervised Deep Embedding for Clustering Analysis", "SpectralNet: Spectral Clustering using Deep Neural Networks", "Contrastive Clustering"], "answer_arxiv_id": ["1511.06335", "1801.01587", "2009.09687"], "source_meta": {"published_time": "20230427"}, "qid": "AutoScholarQuery_train_20809"} +{"question": "Which works propose an alternative to classical provers by automating the interaction with proof assistants?", "answer": ["Deep Network Guided Proof Search"], "answer_arxiv_id": ["1701.06972"], "source_meta": {"published_time": "20230627"}, "qid": "AutoScholarQuery_train_20810"} +{"question": "Which works explored online optimization for non-convex functions?", "answer": ["Learning in Non-convex Games with an Optimization Oracle", "Online Non-Convex Learning: Following the Perturbed Leader is Optimal", "Second-order Online Nonconvex Optimization", "Non-convex online learning via algorithmic equivalence"], "answer_arxiv_id": ["1810.07362", "1903.08110", "2001.10114", "2205.15235"], "source_meta": {"published_time": "20230807"}, "qid": "AutoScholarQuery_train_20811"} +{"question": "Could you provide me some studies about self-supervised learning for tracking-by-matching methods?", "answer": ["Tracking Emerges by Colorizing Videos", "Self-supervised Learning for Video Correspondence Flow", "Joint-task Self-supervised Learning for Temporal Correspondence", "Learning Correspondence from the Cycle-consistency of Time", "Space-Time Correspondence as a Contrastive Random Walk", "Rethinking Self-supervised Correspondence Learning: A Video Frame-level Similarity Perspective"], "answer_arxiv_id": ["1806.09594", "1905.00875", "1909.11895", "1903.07593", "2006.14613", "2103.17263"], "source_meta": {"published_time": "20230523"}, "qid": "AutoScholarQuery_train_20812"} +{"question": "Which papers conducted research on weakly supervised semantic segmentation?", "answer": ["Object Detectors Emerge in Deep Scene CNNs", "Self-taught Object Localization with Deep Networks", "Weakly Supervised Object Localization with Multi-fold Multiple Instance\n Learning", "Token Contrast for Weakly-Supervised Semantic Segmentation", "Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image\n Segmentation", "On Regularized Losses for Weakly-supervised CNN Segmentation", "Reliability Does Matter: An End-to-End Weakly Supervised Semantic\n Segmentation Approach", "Universal Weakly Supervised Segmentation by Pixel-to-Segment Contrastive\n Learning", "Object Region Mining with Adversarial Erasing: A Simple Classification\n to Semantic Segmentation Approach", "Self-Erasing Network for Integral Object Attention", "Hide-and-Seek: Forcing a Network to be Meticulous for Weakly-supervised\n Object and Action Localization", "Adversarial Complementary Learning for Weakly Supervised Object\n Localization"], "answer_arxiv_id": ["1412.6856", "1409.3964", "1503.00949", "2303.01267", "1603.06098", "1803.09569", "1911.08039", "2105.00957", "1703.08448", "1810.09821", "1704.04232", "1804.06962"], "source_meta": {"published_time": "20231128"}, "qid": "AutoScholarQuery_train_20813"} +{"question": "Are there any works explaining empirically observed faster convergence of over-parameterized QNNs?", "answer": ["Theory of overparametrization in quantum neural networks", "A Convergence Theory for Over-parameterized Variational Quantum Eigensolvers", "Quantum Kerr Learning", "Laziness, Barren Plateau, and Noise in Machine Learning"], "answer_arxiv_id": ["2109.11676", "2205.12481v1", "2205.12004", "2206.09313"], "source_meta": {"published_time": "20220830"}, "qid": "AutoScholarQuery_train_20814"} +{"question": "What research papers developed an asynchronous event-based architecture for graph neural networks?", "answer": ["AEGNN: Asynchronous Event-based Graph Neural Networks"], "answer_arxiv_id": ["2203.17149"], "source_meta": {"published_time": "20220613"}, "qid": "AutoScholarQuery_train_20815"} +{"question": "Which works deal with maximizing the similarity between related data and minimizing the similarity of unrelated data in representation learning?", "answer": ["Representation Learning with Contrastive Predictive Coding", "A Simple Framework for Contrastive Learning of Visual Representations", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["1807.03748", "2002.05709", "2103.00020"], "source_meta": {"published_time": "20220609"}, "qid": "AutoScholarQuery_train_20816"} +{"question": "Which works represent synthetic pretraining efforts?", "answer": ["LIME: Learning Inductive Bias for Primitives of Mathematical Reasoning", "Does Pretraining for Summarization Require Knowledge Transfer?", "Task2Sim: Towards Effective Pre-training and Transfer from Synthetic Data"], "answer_arxiv_id": ["2101.06223", "2109.04953", "2112.00054"], "source_meta": {"published_time": "20231030"}, "qid": "AutoScholarQuery_train_20817"} +{"question": "What studies are based on entropy-based approaches for TTA?", "answer": ["Tent: Fully Test-time Adaptation by Entropy Minimization", "Efficient Test-Time Model Adaptation without Forgetting", "Towards Stable Test-Time Adaptation in Dynamic Wild World"], "answer_arxiv_id": ["2006.10726", "2204.02610", "2302.12400"], "source_meta": {"published_time": "20231124"}, "qid": "AutoScholarQuery_train_20818"} +{"question": "Which research works have greatly advanced text-to-image generation with the availability of web-scale datasets of text-image pairs, large language models and large vision-language models?", "answer": ["LAION-5B: An open large-scale dataset for training next generation\n image-text models", "Exploring the Limits of Transfer Learning with a Unified Text-to-Text\n Transformer", "Learning Transferable Visual Models From Natural Language Supervision"], "answer_arxiv_id": ["2210.08402", "1910.10683", "2103.00020"], "source_meta": {"published_time": "20231208"}, "qid": "AutoScholarQuery_train_20819"} +{"question": "Which works studied the Message Passing Neural Network (𝖬𝖯𝖭𝖭)?", "answer": ["Neural Message Passing for Quantum Chemistry"], "answer_arxiv_id": ["1704.01212"], "source_meta": {"published_time": "20230206"}, "qid": "AutoScholarQuery_train_20820"} +{"question": "What research has been done on tallying settings that are special cases of WTB where m